Andres Velasco Economia

Brookings Institution Press
Washington, D.C.
www.brookings.edu
Volume 5 Number 2
2005
Spring
Cover design by Rogue Element
Journal of the Latin American and Caribbean Economic Association
economi
www.lacea.org
economa
Latin American and Caribbean Economic Association
Volume 5 Number 2
2005
Spring
Lora and Olivera on Electoral

Consequences of the Washington
Consensus
Suominen and Estevadeordal

on Rules of Origin in Trade Agreements
Majnoni and Powell on Basel II

and Bank Capital Requirements
Echeverry, Ibez, Moya,

and Hilln on Reforming Public
Transport in Bogot
Brookings / LACEA
Kaplan, Martnez Gonzlez,

and Robertson on Wages after
Displacement
Volume 5 Number 2
economa
2005
Spring
EDITOR
Andrs Velasco
LATIN AMERICAN AND CARIBBEAN

ECONOMIC ASSOCIATION
BROOKINGS INSTITUTION PRESS
Washington, D.C.
Articles in this publication were developed by the authors for the biannual Economa meetings.
In all cases the papers are the product of the authors thinking alone and do not imply endorsement by the staff members, officers, or trustees of the Brookings Institution or of LACEA, or of
those institutions with which the authors are affiliated.
Copyright 2005
LATIN AMERICAN AND CARIBBEAN ECONOMIC ASSOCIATION
www.lacea.org
Published by
BROOKINGS INSTITUTION PRESS
1775 Massachusetts Avenue, N.W., Washington, DC 20036
www.brookings.edu
ISSN 1529-7470
ISBN 0-8157-2079-3
For information on subscriptions, standing orders, and individual copies, contact Brookings Institution Press, P.O. Box 465, Hanover, PA 17331-0465. Or call 866-698-0010. E-mail brookings
@tsp.sheridan.com. Visit Brookings online at http://bookstore.brookings.edu.
Brookings periodicals are available online through Online Computer Library Center (contact the
OCLC subscriptions department at 800/848-5878, ext. 6251) and Project Muse (http://muse.jhu.edu).
Authorization to photocopy items for internal or personal use or the internal or personal use of specific clients
is granted by the Brookings Institution for libraries and other users registered with the Copyright Clearance
Center Transactional Reporting Service, provided that the basic fee is paid to the Copyright Clearance Center,
222 Rosewood Drive, Danvers, MA 01923. For more information, please contact CCC at 508/750-8400. This
authorization does not extend to other kinds of copying, such as copying for general distribution, or for creating
new collective works, or for sale. Specific written permission for such copying must be obtained from the Permissions Department, Brookings Institution Press, 1775 Massachusetts Avenue, N.W., Washington, DC 20036,
fax 202/797-6195; e-mail permissions@brookings.edu.
economa
Volume 5 Number 2
2005
Spring
Editors Summary
vii
EDUARDO LORA and MAURICIO OLIVERA
The Electoral Consequences of the Washington Consensus

Comments by Sebastin Galiani and Ernesto Dal B
ANTONI ESTEVADEORDAL
and KATI SUOMINEN
Rules of Origin in Preferential Trading Arrangements:

Is All Well with the Spaghetti Bowl in the Americas?
Comments by Pablo Sanguinetti and Alberto Trejos
GIOVANNI MAJNONI
46
63
93
and ANDREW POWELL
Reforming Bank Capital Requirements:

Implications of Basel II for Latin American Countries
Comments by Patricia Correa and Philip Brock
105
141
JUAN CARLOS ECHEVERRY, ANA MARA IBEZ,

ANDRS MOYA,
and LUIS CARLOS HILLN
The Economics of TransMilenio,

a Mass Transit System for Bogot
151
Comments by Mauricio Crdenas and Andrs Gmez-Lobo
189
DAVID S. KAPLAN, GABRIEL MARTNEZ GONZLEZ,
and RAYMOND ROBERTSON
What Happens to Wages after Displacement?

Comments by Narcio Menezes-Filho and Omar Arias
235
197
LATIN AMERICAN AND CARIBBEAN ECONOMIC ASSOCIATION
The Latin American and Caribbean Economic Association (LACEA), or

Asociacin de Economa de Amrica Latina y el Caribe, is an international
association of economists with common research interests in Latin
America. It was formed in 1992 to facilitate the exchange of ideas among
economists and policymakers.
Membership in LACEA is open to all individuals or institutions
professionally concerned with the study of Latin American and Caribbean
economies. For membership information, please visit the LACEA website
at www.lacea.org and click on join LACEA.
OFFICERS
President
Mariano Tommasi, Universidad de San Andrs, Argentina
Vice President
Andrs Velasco, Harvard University and Universidad de Chile
Past Presidents
Sebastin Edwards, University of CaliforniaLos Angeles
Guillermo Calvo, Inter-American Development Bank and University of Maryland
Nora Lustig, Universidad de las Amricas, Puebla
Albert Fishlow, Columbia University
Secretary
Ariel Fiszbein, World Bank
Treasurer
Sergio Schmukler, World Bank
EXECUTIVE COMMITTEE
Nancy Birdsall, Center for Global Development
Franois Bourguignon, World Bank
Raquel Fernndez, New York University
Francisco H. G. Ferreira, Pontifcia Universidade Catlica, Rio de Janeiro, and
World Bank
Nora Lustig, Universidad de las Amricas, Puebla

Jos A. Ocampo, United Nations
Guillermo Perry, World Bank
Carola Pessino, Universidad Torcuato Di Tella, Argentina
Carmen Reinhart, University of Maryland

Andrs Rodrquez-Clare, Inter-American Development Bank
Jaime Saavedra, World Bank
Cristina T. Terra, Fundao Getlio Vargas
ECONOMIA
Editor
Andrs Velasco, Harvard University and Universidad de Chile
Editorial Associate
Jennifer Hoover
Managing Editor
Magdalena Balcells
Editorial Board
Rafael Di Tella, Harvard University
Eduardo Engel, Yale University
Francisco H. G. Ferreira, Pontifcia Universidade Catlica, Rio de Janeiro, and
World Bank
Carmen Pags, World Bank

Roberto Rigobn, Massachusetts Institute of Technology
Andrs Rodrguez-Clare, Inter-American Development Bank
Roberto Steiner, International Monetary Fund
Miguel Urquiola, Columbia University
Sebastin Edwards (ex officio), University of CaliforniaLos Angeles
Mariano Tommasi (ex officio), Universidad de San Andrs, Argentina
ECONOMIA PANEL FOR VOLUME 5
Mauricio Crdenas, Fedesarrollo-Colombia

Gerardo Esquivel Hernndez, Colegio de Mxico
Ronald Fischer, Universidad de Chile
Ilan Goldfajn, Pontifcia Universidade Catlica, Rio de Janeiro
Eduardo Levy Yeyati, Universidad Torcuato Di Tella, Argentina
Mara Soledad Martnez Pera, World Bank
Francisco Rodrguez, Instituto de Estudios Superiores de Administracin,
Venezuela
Nouriel Roubini, New York University

Jaime Saavedra, World Bank
Ernesto Schagrodsky, Universidad Torcuato Di Tella, Argentina
AUTHORS AND DISCUSSANTS
Omar Arias, World Bank

Philip Brock, University of Washington
Mauricio Crdenas, Fedesarrollo-Colombia
Patricia Correa, Super Intendencia de Bancos de Colombia
Ernesto Dal B, University of CaliforniaBerkeley
Juan Carlos Echeverry, Universidad de los Andes
Antoni Estevadeordal, Inter-American Development Bank
Sebastin Galiani, Universidad de San Andrs
Andrs Gmez-Lobo, Universidad de Chile
Luis Carlos Hilln, Universidad de los Andes
Ana Mara Ibez, Universidad de los Andes
David S. Kaplan, Instituto Tecnolgico Autnomo de Mxico
Eduardo Lora, Inter-American Development Bank
Giovanni Majnoni, World Bank
Gabriel Martnez Gonzlez, Inter-American Conference on Social Security
Narcio Menezes-Filho, University of So Paulo
Andrs Moya, Universidad de los Andes
Mauricio Olivera, Inter-American Development Bank
Andrew Powell, Universidad Torcuato Di Tella
Raymond Robertson, Macalester College
Pablo Sanguinetti, Universidad Torcuato Di Tella
Kati Suominen, Inter-American Development Bank
Alberto Trejos, Instituto Centroamericano de Administracion de Empresas INCAE
ANDRS VELASCO
Editors Summary
early two decades after a wave of policy changes swept through Latin
America, economic reforms continue to be the focus of much discussion. Critics claim that the promarket reforms have failed to
deliver economic growth, and that the time has come to try something else.
Advocates claim that the reforms were never given a fair chancetoo little was done, often too late. Complete the reform process, they claim, and
growth will come.
Both sides do agree on one point: Latin America seems to be suffering
from reform fatigue, and another wave of reforms is unlikely to happen any
time soon. Certainly not in countries led by left-leaning populists, such as
Argentinas Nstor Kirchner or Venezuelas Hugo Chvez. The reform
momentum has even stalled in countries led by promarket conservatives
Mexico under Vicente Fox and Colombia under Alvaro Uribe are two
examples. If such reforms are now unpopular in many quarters, did the
politicians who initially adopted them bear a political cost? Was the Washington Consensus electorally bad for friends of Washington? That is the
question studied by Eduardo Lora and Mauricio Olivera in the lead article
of this, the tenth issue of Economa.
Lora and Olivera analyze the outcome of sixty-six presidential elections and eighty-one parliamentary elections in seventeen Latin American
countries from 1985 to 2002. Their general conclusion is striking: reforming parties and politicians were rewarded electorally only when reforms
involved macroeconomic stabilization and a sharp reduction in inflation;
otherwise, their reforming zeal cost them dearly at the polls. Economic outcomes do matter for electoral outcomes. Lora and Olivera find that the
incumbents party is rewarded in presidential elections for reductions in the
inflation rate and in legislative elections for increases in the growth rate.
Changes in unemployment and income distribution, however, do not appear
to influence voters behavior.
vii
viii
E C O N O M I A , Spring 2005
What is even more surprising is that, at the polls, policies matter irrespective of their resultsthat is, their effects on growth or inflation. Electorates seem not to like reform policies of the kind applied in Latin
America in the 1990s. In a regression with electoral outcomes on the right
hand side, reform indexes have a negative and significant effect, even when
the authors control for changes in inflation and growth. The point estimate of the effect of policies on electoral results implies that the incumbents party typically lost 15 percent of its vote in presidential elections on
account of the average amount of promarket reforms introduced during
its term. More aggressive reformers (say, those reforming one standard
deviation above the mean) sacrificed 27 percent of their vote on account of
promarket reforms. Statistically, this seems to be a very robust result for
presidential elections.1
Moreover, lying about ones true intentions does not seem to be a good
electoral strategy. Several Latin American politiciansincluding Fujimori
in Peru, Menem in Argentina, and more recently Gutirrez in Ecuador
first ran as opponents of the Washington Consensus, then followed orthodox policies. The paper shows that a candidate that said one thing on tax
policy and then did another was, on average, punished more severely at
the polls. Campaign promises do not seem to matter for the effect of other
policies on voting behavior.
These results raise two kinds of questions. For academics, the issue is
why inputs (policies) matter and not just outputs. Is it ideology, pure and
simple? Or is it that because outcomes represent an extremely noisy signal
of politicians competence, the choice of policies conveys some information that voters find useful? For policymakers, the question is political:
what has to change in Latin America before ambitious reforms become feasible again? Are all large-scale reforms out of the question, or only those
that carry the Washington Consensus label? Both sets of questions remain
very much open.
The unpopularity of the reforms does not mean, however, that policy is
frozen everywhere. Trade is one area in which reform has not stopped dead
1. The total effect of reforms on electoral outcomes is the sum of two effects: a direct
effect that runs from policies to votes and an indirect effect that runs from policies to economic outcomes to votes. The first is typically large and negative, while the second is positive insofar as the reforms lowered inflation and stimulated growth. The figures given
correspond to the total effectthat is, after the positive indirect effects have been taken into
account. The direct negative effects are much larger. See the paper for details.
Andrs Velasco
ix
on its tracks. Liberalization agreements, of both the bilateral and regional

kind, continue to be signed, though at a less frenzied pace than a decade
ago. Nearly fifty deals have been forged in the Americas since 1990s. But
this veritable spaghetti bowl of overlapping and sometimes contradictory
agreements has costs as well as benefits. One cost, studied by Antoni Estevadeordal and Kati Suominen in the second paper of this issue, results from
the rules of origin applied.
At heart, the matter is simple: if Brazil gives Paraguay preferential
access to its market, Brazilian policymakers want to make sure that the new
imports entering Brazil are, in fact, made in Paraguay and not in a third
country attempting to benefit from Paraguays preferential status. But what
sounds simple in theory becomes devilishly complicated in practice. What,
precisely, is a Paraguayan good? Goods often have imported inputs, and in
few or no items is 100 percent of value added likely to originate in
Paraguay. Where, then, should a country draw the line?
Rules of origin attempt to settle the issue, but in doing so they face many
pitfalls. If the set of rules is too stringent, then the bulk of Paraguayan
goods may be left out of the Brazilian market. Indeed, such rules can be
used as protectionist devices that effectively undercut trade preferences and
contradict the avowed liberalizing intent of free trade agreements. Another
problem is that rules of origin are almost inevitably complex (the paper
identifies several kinds, each with its own subcategories). Applying them
can be very costly, especially for the poorer economies in the region.
Estevadeordal and Suominen offer three main conclusions. First, putting
stringent rules of origin into an agreement makes it more politically feasible, since the rules can be used as a tool to pay off protectionist interests.
Second, there is evidence that restrictive rules of origin undercut the liberalizing potential of free trade agreements. NAFTA is a particularly egregious example of this, with many Mexican goods subjected to rules that
verge on ludicrous. It is unfortunate, thereforeand this is the third conclusion of the paperthat the NAFTA model of rules of origin is increasingly being used in other agreements in the region. This does not bode
well for free trade in the Americas.
Not all is lost, however. NAFTA-type rules are at least precise, and they
leave less room for arbitrary application than do other types of rules of
origin used in earlier agreements. Moreover, the growing homogeneity of
rules that follow the NAFTA model simplifies the life of customs officials
and lowers transaction costs. Last, and most important, the NAFTA rules of
origin have what trade experts call lenient facilitation devices. In English,
this means that the rules themselves include ways to reduce their restrictiveness. A key aspect is diagonal cumulation, which allows countries tied
by the same set of origin rules to use products that originate in any part of
the common rules-of-origin zone as if they originated in the exporting
country. Therefore, argue the authors, the rules of origin in a future Free
Trade Area of the Americasif one ever materializesshould not be all
that restrictive. One can only hope they are right.
Financial regulation is another area in which policy is changing, as a
result of both internal needs and international changes in standards. The
1988 Basel Accord on bank capitalthe so-called Basel I agreementis
now in place throughout the region, and discussion has shifted to whether
and how Latin American countries should apply Basel II. It is widely
accepted that bank capital ought to be regulated, but how to do so remains
open to debate. The simple approach of Basel I divides assets into very
broad risk categories and establishes an 8 percent minimum capital requirement for risky assets. The potential for arbitraging ones way around this
simple rule has grown, however, as risk management becomes more
sophisticated. In response, Basel II goes well beyond simple quantitative
requirements, proposing two basic approaches: the standardized approach,
which uses external credit rating agencies together with a table that maps
those ratings directly into capital requirements; and the internal ratingsbased (IRB) approach, in which the banks themselves estimate their customers default probability (without relying on external rating agencies) and
then use a particular formula to determine capital requirements as a function of the estimated default probability.
The third paper in this issue, by Giovanni Majnoni and Andrew Powell,
focuses on a key aspect of Basel II application. Many emerging markets
do not have many (or any) external rating agencies, so the standardized
approach may not be applicable. The internal ratings-based approach, in
turn, is complex, and its application and supervision may stretch limited
supervisory resources. Majnoni and Powell suggest a simplification of the
IRB approach that could be used as a transition arrangement. In their
centralized ratings-based (CRB) approach, banks would rate their clients,
but the regulator would determine the rating scale and the way in which
the banks ratings map into default probabilities. Using a centralized scale
would facilitate comparison across banks and greatly ease the monitoring of banks ratings. Those requirements would also be easier to monitor,
Andrs Velasco
xi
since the regulator would determine how banks ratings feed into capital
requirements.
The hard part of the approach is deciding what kinds of standards the
regulator should apply, since what works in rich countries may not work
in emerging economies. Basel IIs IRB approach suggests a formula for calculating a banks capital requirement as a function of three basic variables:
default probability, exposure at default, and loss given default. A regulator
might then ask a bank to hold provisions and capital to cover a specified
percentage of the distribution of losses to ensure the continued solvency
of the bank except in highly extraordinary circumstances. The calibration of
the Basel II IRB formula uses a value at risk of 99.9 percent with a horizon of one yearthat is, a bank is only expected to use up its capital in
one year with a probability of 0.1 percent, or once every 1,000 years.
Majnoni and Powell employ a bootstrapping technique to calculate loss
distribution functions for Argentina, Brazil, and Mexico, using data on loan
performance from public credit registries. They then use these functions to
estimate the size of expected and unexpected losses of an average-sized
bank with a loan portfolio randomly drawn from the universe of loans
within the financial system. Their results show that these three countries
have significantly higher default probabilities than Group of Ten (G10)
countries. As a result, both current practice under Basel I and the suggested standards under Basel II may be inadequate. To achieve a 99 percent
level of protection, capital requirements would need to be close to 15 percent, which is significantly higher than the 8 percent level recommended
in Basel I. Even higher levels would be required to achieve 99.9 percent
protection, as intended in Basel II. They also find that Basel IIs IRB
approach would result in levels of 9095 percent protection rather than the
99.9 percent goal. This is not surprising, since the IRB was calibrated for
the safer economies of G10 countries.
If bank regulation needs modernizing in Latin America, public transport does too. The spectacle of streets packed with old buses spewing black
smoke is all too common in many cities of the region, from Mexico City
to Quito and from So Paulo to Santiago. Poor public transport induces
more private cars to enter the streets, worsening congestion and pollution.
If you think that this is a textbook case of the state not doing the job of
providing public services, think again. Bus systems are private in many
cities in Latin America, and that does not seem to solve the problem. As
Juan Carlos Echeverry, Ana Mara Ibez, Andrs Moya, and Luis Carlos
xii
Hilln document in their paper in this volume, the market for urban buses
is ripe with market failures: unclear definition of property rights on the
curbside and on the road; cartelization that results in fares set above the
competitive equilibrium levels; misalignment between the incentives of bus
drivers and owners, in a typical principal-agent conflict; and congestion and
pollution externalities. In many developing countries, these market failures are exacerbated by weak regulation and enforcement. The result often
is too many buses each carrying too few passengers in unsafe conditions,
clogging the streets and soiling the air as they move (or fail to move) along.
One city in Latin America to have tackled the problem head-on is
Bogot, Colombia. Its so-called TransMilenio system is now being imitated
in Quito and Santiago, among others, as well as several cities in Colombia. Echeverry, Ibez, Moya, and Hilln explain the logic behind the new
system and analyze is effects. The key elements of the new system are as
follows: (i) a hybrid public-private system, with concession contracts for
private service providers; (ii) competition for the road (rather than on
the road) in a tendering process, with fare-setting based on long-term
investment recovery; (iii) remuneration based on kilometers traveled rather
than passengers transported, so as to prevent drivers from fighting over passengers on the street; (iv) separation between the transportation service and
the fare collection process; and (v) exclusive road and curb-side service in
metro-like stations.
Congestion, pollution, traffic accidents, travel times, and waiting times
all fell dramatically along the corridors where TransMilenio was first put to
work. The system was initially hailed as the solution of Bogots serious
transport problems. Not all results were unambiguously positive, however,
as the paper makes clear. Increased ridership resulted in jammed buses
and rising waiting times. Moreover, the full system covering the entire
city is not expected to be operating until 2015. This gradual transition did
not help: older buses were displaced to secondary streets, where traffic
and pollution increased.
A cost-benefit analysis of the system as is, with approximately 25 percent of the routes in operation, reveals welfare gains for users of the new
routes, but an overall negative effect stemming primarily from increases
in travel time for passengers using the traditional transport system. Since
congestion costs are highly nonlinear, the welfare losses from heightened
congestion in unserved corridors more than offset the benefits from TransMilenio, even though those benefits are sizeable. The authors conclude by
Andrs Velasco
xiii
arguing that the adoption of a new public transport system must be coupled with improved regulation of all other public transport providers, so as
to avoid the problem that arose in Bogot.
What happens to workers wages and employment prospects once they
are displaced from their current jobsfor instance, by trade reform? If they
are likely to be re-hired quickly at comparable wages, then no policy
response is called for; but if some wage losses are large and lasting, then
targeted help for displaced workers may be called for. David Kaplan,
Gabriel Martnez, and Raymond Robertson study the issue for the case of
Mexico, using an administrative data set that allows them to follow individual workers over a period of thirty-two quarters in four regions that vary
significantly in labor market conditions. They focus on the differences in
institutions, inequality, and labor market conditions that may explain differences in wage behavior after displacement.
One striking result is the heterogeneity of worker experiences, which
range from large wage losses to many instances of gains after displacement.
This is consistent with earlier results for other countries, but it cannot be
attributed to differences in institutions (rates of unionization) or inequality, which are quite similar across Mexico. Rather, Kaplan, Martnez, and
Robertson argue that labor market conditions, which vary quite a bit across
time and regions within Mexico, explain the heterogeneity of experiences.
In good times and in the most economically active regions, postdisplacement wages are generally higher than they were in the previous jobs. However, workers who are fired during times of high unemployment and in
less economically active regions experience lasting effects on wages. If any
public assistance is to be disbursed, Kaplan, Martnez, and Robertson
argue, it should go to these workers.
All papers but one included in this issue were presented at the panel
meeting held in San Jos, Costa Rica, in October 2004. The local hosts, and
particularly Juan Rafael Vargas, provided much help. As usual, associate
editors of Economa, members of the 2004 panel, and outside discussants
and referees have done an outstanding job. Thanks is due to them all.
EDUARDO LORA
MAURICIO OLIVERA
The Electoral Consequences of the

Washington Consensus
o country in Latin America escaped the dictums of the Washington
Consensus. From Brazil under left-leaning Fernando Henrique Cardoso to Mexico under ultra-orthodox economist Ernesto Zedillo
and Peru under Alberto Fujimoris yoke, macroeconomic imbalances were
brought under control, barriers to international trade were lifted, and stateowned enterprises were privatized. Whether this one-size-fits-all prescription was imposed from outside or adopted at will by the governments elected
on the promise of improving the lot of their peoples may be a matter of
debate. But all sides seem to agree on one point: the results did not meet
the expectations created both by outsiders and by those in power.
Up to the mid-1980s only two countries in Latin America had adopted
a package of policies similar to what became to be known as the Washington Consensus at the turn of the decade. Those two were undemocratic
Chile and impoverished Bolivia, by then among the most politically and
economically unstable countries, if not in the world, then certainly in Latin
America. Extreme cases, extreme policies: that was a common interpretation of the two experiences. Less common was the expectation that those
policies were about to be adopted by virtually every Latin American country in the next few years, both those in which democracy had been the rule
Lora and Olivera are with the Inter-American Development Bank.

We are grateful for the valuable research assistance of Carlos Andrs Gmez. We also
thank Benito Arrunada, Mauricio Crdenas, Stephen Kay, Ugo Panizza, Andres RodrguezClare, Mariano Tommasi, Jessica Wallack, and seminar participants at ISNIE-University
Pompeu Fabra, LACEA-PEG, Econnet-IADB, and the Economia panel meeting for comments and suggestions. We are especially grateful to Rafael Di Tella, Sebastian Galiani, and
Ernesto Dal B for their detailed and very useful comments and suggestions. We would like
also to thank Sebastian Saiegh for allowing us to use his data on political coalitions.
2 E C O N O M I A , Spring 2005
for decades, like Colombia, Costa Rica, and Venezuela, and those where
the third wave of democratization was just arriving, such as Argentina,
Brazil, and Uruguay.
The years of high expectations, both about democratization and about
Washington Consensustype policies, are over. Latin Americans are still
convinced democrats, but enthusiasm has waned. Three out of every four
Latin Americans see democracy as the best form of governmentor rather,
as the least bad, since 68 percent think that democracy is not functioning well in their countries. Latin Americans are even more sceptical about
the benefits of promarket economic policies. Only one out of four Latin
Americans considers privatization to have been beneficial for his or her country and barely 16 percent think that the market economy is doing a good job.1
Malaise is getting the upper hand in a number of places. Electricity and
water privatizations were blocked in Arequipa (Peru) and Cochabamba
(Bolivia), following violent clashes between vociferous opponents and the
police. An ambitious project to attract foreign direct investment to
Bolivias gas sector was derailed by the Indian communities. While these
events may be dismissed as isolated expressions of popular feeling, a new
crop of presidents from Nstor Kirchner in Argentina to Lucio Gutirrez
in Ecuador and Tabar Vsquez in Uruguay has won clear majorities in
popular elections after campaigning against the excesses of marketoriented policies.
In an attempt to establish whether this malaise is justified or not, economists have devoted substantial effort to assessing the economic and social
consequences of the Washington Consensus policies. The dominant view
seems to be that they have had positive effects on economic growth and
income levels, though there is intense debate over the size of those effects,
over whether they are transient or permanent, and over the importance of
each of the components of the Washington Consensus. The dominant view
also holds that the effects have been muted by lack of regulatory and institutional support for the liberalization efforts, though the specific forms of
regulation and institutions necessary for that purpose are far from clear.
Even more intense is the debate over the social and distributional effects of
fiscal stabilization and promarket reforms, which are the two main pillars of
the Washington Consensus.2
1. Opinion data come from the 2003 issue of Latinobarmetro, a public opinion survey
conducted by the Corporacin Latinobarmetro, Santiago, Chile.
2. These debates are surveyed in Lora and Panizza (2002); Kuczynski and Williamson
(2003); and Lora, Panizza and Quispe-Agnoli (2004).
Eduardo Lora and Mauricio Olivera
However, the future of these policies will depend not so much on their
efficacy but on whether they receive the support of the electorate. On this,
the state of knowledge is much more scant and fragmentary, as will be seen
below. This paper attempts to help fill that vacuum by evaluating through
econometric methods the electoral consequences of the Washington Consensus. Although our approach is backward looking, it sheds considerable
light on the future. Our study shows that the electorate cares not only about
the outcomes of the policies (maybe about only some outcomes and not
others), but also about the policies themselves, irrespective of whether they
produce good or bad (observable) outcomes. In addition, the electorate
seems to care about whether the policies adopted by a government are
in line with the ideology of the incumbents party and with preelectoral
promises. Furthermore, in presidential regimes voters cast separate votes
for the executive and the legislature, and outcomes and policies affect each
vote differently. The presidential vote is more volatile and more susceptible to economic outcomes and policies, but votes for the legislature are not
completely immune: policies in which the legislature clearly plays a role,
such as privatizations, tend to have electoral consequences. These results
provide a nuanced landscape for the future of Washington Consensus policies, where neither bold backslashes nor aggressive promarket reforms
should be expected in the future. Not only is the time of high expectations
over; perhaps the time for deep reforms is also past.
In the next section of the paper we present a short survey of the literature
assessing the electoral consequences of the Washington Consensus policies
and derive our empirical hypotheses. On that basis, we then discuss the theoretical and econometric approaches that support the empirical analysis.
In subsequent sections we describe the data, present the econometric findings, and discuss our conclusions.
A note on terminology is in order before proceeding. Neo-liberal,
market-oriented, orthodox, and a variety of other labels have been
attached to the set of economic policies in vogue since the early nineties in
Latin America and elsewhere. We use these terms interchangeably, but not
loosely: for the sake of clarity and brevity, this paper deals with the ten policies summarized in the classic article by Williamson that made the term
Washington Consensus famous.3 We assume that all those labels refer to
that same set of policies (as detailed below in the section titled Data).
3. Williamson (1990a).
Review of the Empirical Literature and Some Testable Hypotheses

The most straightforward view of the response of the electorate to economic policies is based on the economic voting argument: people base
their electoral decisions on cost-benefit calculations. If the policies bring
net benefits to them, they cast their votes to support the government, or the
party, administering those policies; if the policies bring losses to them,
they lend their support to the candidate, or the party, opposing them. Economic voting is usually assumed to be retrospective: voters observe past
performance and assume that past trends will persist into the future if the
government or the party remains in power. If those trends are deemed
acceptable, given a set of standards or expectations voters decide to reelect
the incumbent, or his party if the option of reelection does not exist. Therefore, in retrospective economic voting policies play no direct role, since
voters decide entirely on the base of past outcomes.4
Considerable evidence from advanced industrial democracies supports the
view that past economic performance influences peoples voting decisions
and their support for governments.5 An important empirical finding from this
literature is that voters base their decisions on aggregate (or sociotropic)
economic outcomes such as growth, inflation, and unemployment, rather
than on individual (or pocketbook) outcomes. Most of the empirical literature on developed countries comes from single-country analyses, based
either on time-series electoral outcomes or public opinion polls. The economic voting hypothesis is more robust for public opinion polls than for
actual electoral outcomes.6 Empirical studies of electoral behavior in the
United States using state-level data lend support to the simple economic
voting hypothesis, in the sense that voters are able to evaluate their states
economic performance relative to that of the national economy. Furthermore, they (irrationally) reward state governors for economic fluctuations
that are unrelated to gubernatorial actions, which implies that they have limited ability to filter aggregate economic information.7 The ability of voters
4. Stokes (2001b, pp. 118) provides a concise review and discussion of the theoretical
underpinnings of retrospective economic voting.
5. Based on the seminal work by Downs (1957); among the initial papers on economic
voting in the United States are Kramer (1971); Meltzer and Vellrath (1975); and Arcelus and
Meltzer (1975).
6. Lewis-Beck (1988) is a salient example of the early empirical literature based on opinion polls in European countries. For a review of this literature, see Stokes (2001b, pp. 28).
7. Wolfers (2002).
to gather and update information is a central issue in the theoretical and

empirical literature on economic voting.8 Although some evidence points
out to the presence of prospective as well as retrospective behavior, uncertainty about the workings of the economy and the relatively high cost of
gathering and processing the information necessary to forecast outcomes
is consistent with the importance of retrospective voting in the empirical
findings.9
Empirical support for the economic voting hypothesis in Latin America
has been uncovered by Karen Remmer, Michael Coppedge, Kenneth Roberts
and Erik Wibbels, and Susan Stokes.10 A concise summary of these findings
is presented in table 1. Based on data for twenty-one competitive elections
between 1982 and 1990, Remmer has found that conditions of economic
crisis undermine support for incumbents and provoke high levels of electoral volatility.11 The magnitude of the electoral change is found to be associated with the depth of the crisis during the campaign period, with variations
in exchange rates, GDP, and inflation highly correlated with various
indicators of electoral outcomes. Her results also suggest that the effect
of economic conditions on electoral instability are mediated by the structure of the party system (insulating two-party systems from the volatility
experienced by more fragmented systems). However, as Stokes points
out, these results are anomalous given the predictions of normal economic
voting, as she finds that incumbent parties suffered larger losses at the
polls when inflation went down (significant) and when GDP rose (not
significant).12
In a subsequent paper, Remmer presents new estimates on the influence
of inflation and growth on the incumbent vote in presidential elections.13
Her new database covers forty-nine elections for seven countries between
1983 and 1999. Her results indicate that after controlling for the advantage
of incumbency as well as major differences in the structure of party systems, electoral outcomes are strongly influenced, in the direction expected,
by macroeconomic performance in the year before the election. That is, inflation is found to be negatively correlated with electoral support, whereas
8. For a review of this debate, see Duch and Stevenson (2004); and Keech (1995).
9. On prospective behavior, see, for instance, Lewis-Beck (1988).
10. Remmer (1991, 2003); Coppedge (2001); Roberts and Wibbels (1999); Stokes (2001b).
11. Remmer (1991).
12. Stokes (2001b, p. 27).
13. Remmer (2003).
T A B L E 1 . Summary of Empirical Findings on Economic Voting in Latin America
Study
Dependent variable
Remmer (1991)
Electoral volatility
Remmer (2003)
Vote shares
Electoral volatility
Roberts and
Wibbels (1999)
Coppedge (2001) Vote shares
Probability of a
Stokes (2001b)
security-oriented
candidate being
elected
Election type
(number of
countries)
Period
Estimation
method
Presidential (12)
198290
Pooled OLS
Presidential (8)
198399
Pooled OLS
Legislative and
presidential (16)
Legislative (11)
Presidential (15)
198097
Pooled OLS
197895
198295
Pooled OLS
Probit
Main results
Inflation ; GDP
growth +a
Inflation ; GDP
growth +
Inflation a; GDP
growth +
Inflation
Inflation ; GDP
growth +
Source: Authors calculations.

a. Not significant.
growth is positively correlated with it. Furthermore, inflation is significant

in all the regressions presented, while growth is more significant for the
elections held in the 1990s than for those in the 1980s, indicating that the
sensitivity of the electorate to economic performance has increased rather
than waned over time.
Coppedges empirical work focuses on the impact of changes in inflation
on legislative vote shares. His dependent variable consists of 132 changes in
legislative vote shares for major parties in eleven countries from 1978 to
1995. His only indicator of economic performance is the change in (the
log of) inflation from the last year of the previous government to the last
year of the current government. By interacting this variable with appropriate dummies, Coppedge finds that changes (whether increases or decreases)
in inflation affect electoral support for the incumbents parties in the
expected way, while only increases in inflation improve the vote share of
the opposition parties. However, these results apply only to parties with a
fluid base, that is, parties that do not count on a strong party identification.
When there is such identification, voters are reluctant to question their
party identification on the basis of macroeconomic outcomes.
Roberts and Wibbels consider economic voting as a possible explanation of electoral volatility in Latin America. Their database includes fiftyeight congressional elections and forty-three presidential elections in sixteen
Latin American countries during the 1980s and 1990s. Their results show
that economic performance has an effect on electoral stability. Economic
growth stabilizes partisan support in legislative elections, whereas sharp
changes in the rate of inflation from one administration to the next, whether
positive or negative, produce the opposite effect. Short-term inflation influences support for incumbent presidents, but growth changes have only a
weak effect on the vote for incumbents, which suggests that voters are
more inclined to hold them directly accountable for monetary stability
than economic growth. Although electoral volatility is influenced by economic performance, it is also related to the institutional characteristics of
political regimes and party systems, and to the structure and organization
of class cleavages.14
In her study of neoliberalism by surprise, Stokes uses data from
twenty-three elections in the 1980s and 1990s in order to assess how the
electorate judges incumbents who, having campaigned for stability-oriented
or protectionist policies, once in office switch to market-oriented ones. She
finds that for both, switchers and non-switchers, economic growth
and inflation affect their vote share in the expected ways. Furthermore,
voters are more sensitive to economic outcomes in the case of switchers,
although this result is not statistically significant (more on these results
below).15
These empirical studies taken together lend support to the retrospective
economic voting argument in both presidential and legislative elections.
They make clear that voting decisions are also influenced by political, institutional, and structural factors and that some of these factors may influence
the severity with which voters judge economic outcomes. Therefore, based
on these studies, two testable propositions are derived:
1. Electoral support for the incumbents party is higher, the better the
aggregate economic outcomes during his or her administration.
2. The sensitivity of electoral support to economic outcomes depends on
the institutional characteristics of the political regime and the party system.
As mentioned, in normal economic voting only past outcomes influence
peoples views. However, as in all six of the Stokes case studies on market reforms in new democracies, people sometimes react to economic
deterioration by supporting the government more strongly; and conversely,
they sometimes respond to economic improvements with pessimism and
opposition.16 Normal economic voting is not the only pattern, especially in
the process of deep economic reform. If there are good reasons to believe
14. Roberts and Wibbels (1999), quote from p. 584.
15. Stokes (2001a).
16. Stokes (2001a).
that past circumstances are not good indicators of the future, information
other than past economic outcomes may influence peoples electoral decisions. For instance, voters may recognize that past circumstances were
affected by factors beyond the governments control and exonerate the
incumbent from the responsibility for past declines in their welfare. Voters may then forecast their future welfare as a function of government policy, rather than as an extrapolation of the past. This sounds simpler than
it is, of course, because future government policies are unknown and
because the relationship between policies and outcomes is diffuse. Peoples expectations of future policies may be formed on the basis of the
policies adopted or announced by the incumbent or on the basis of his
partys ideology. These policy expectations may then be translated into
expected outcomes through a set of beliefs and hypotheses about their possible consequences.
It is often implicitly assumed that peoples (average) beliefs conform
to the actual functioning of the real world. If that is so, assessing the
effects of economic policies would help explain voters electoral decisions.
Economists have devoted considerable effort to evaluating the impact of
Washington Consensus policies on economic growth, income distribution, employment levels, and a host of other variables.17 However, there
has been no comparable effort to examine whether these results are consistent with how the electorate responds to those policies. The only study
on the subject, by Carlos Gervasoni, has found positive correlations
between several indicators of heterodox (that is, anti-neoliberal) policies
and losses in the vote shares of the parties of the incumbents who adopted
those policies.18 The variable with the largest and most significant effect is
money supply growth. Import protection indicators are also significant,
whereas fiscal deficit and the share of the state in GDP are not significant.
These results suggest that Washington Consensus policies do not entail
electoral costs and may even produce electoral benefits, probably because
they bring positive economic effects. It is suggestive that the most significant policy variable is the money supply, because it is well known that
inflation is, ultimately, a monetary phenomenon, and as mentioned, empirical evidence suggests that inflation is a key economic outcome influencing
electoral decisions.
17. For surveys of the literature, see Inter-American Development Bank (2003, chap. 5);
Kuczynski and Williamson (2003); and Lora and Panizza (2002).
18. See Gervasoni (1997), citing a 1995 study.
However, it is a great leap of faith to assume that peoples beliefs conform to the actual consequences of policies. In mapping policies on outcomes, ideology and leaders opinions may be more important for most
people than their limited understanding of how policies work their influence
through the social and economic structures to affect production, employment, or income distribution. Evidence on how those factors influence
electoral responses to economic policies is very scant. However, in-depth
case studies on Argentina and Venezuela by Javier Corrales clearly show
that the reaction of the electorate to the adoption of neoliberal economic
policies in the 1990s was mediated by the party structure and other institutional factors.19 The cohesion and tactics of the Partido Peronista help
explain the electorates support of the neoliberal reforms in Argentina in
the early 1990s, as well as their demise a decade later. Venezuelas Accin
Democrtica lacked that cohesion, and its reforms were soon rejected by
the electorate.
If voters care about policies and not only about past outcomes, the policy announcements of presidential candidates will be a key source of information. However, campaign promises are often poor predictors of actual
policy: according to Stokes, of the thirty-three Latin American governments that adopted promarket reforms between 1982 and 1995, only about
half (seventeen) hinted during their campaigns that such reforms were
going to be implemented.20 This raises several empirical issues. First, do
policy announcements in fact influence electoral decisions? Empirical evidence from the United States and other advanced industrialized economies
shows that they do: people seem to base their opinions in part on campaign
announcements, and voters punish ambiguous campaigns.21 Of course,
some promises may resonate more than others, depending on, among other
things, economic circumstances. For thirty-eight Latin American elections
in the 1980s and 1990s, Stokes finds that stability-oriented candidates (as
opposed to market-oriented ones) stand a better chance of being elected,
the lower the rates of GDP growth and inflation.22
A second empirical issue is whether deviating from campaign promises
carries electoral costs for the incumbent. Although deviations may in
principle be costly, they may produce a positive payoff if they signal the
19.
20.
21.
22.
Corrales (2002).
Stokes (2001a).
For a brief review of this topic, see Stokes (2001a, pp. 45).
Stokes (2001a, pp. 9397).
incumbents commitment to achieving highly desirable economic outcomes at the expense of more immediate partisan support.23 According to
Stokes, deviating from campaign promises does carry electoral costs,
although only weakly.24 However, since her estimates control for economic outcomes, this result implies that policy switches may still have a
positive electoral payoff if the new policies bring substantial economic
improvement. Neoliberalism by surprise may still be a good political
strategy.25
A common theme in the literature on economic voting is the conditional
nature of voters responses to economic outcomes and policies. As mentioned, the severity of their judgment depends on their attachment to the
party in power, the structure of the party system, and other institutional
considerations. It also depends, although weakly, on whether the policies
adopted by the incumbent are in line with his campaign pronouncements.
An additional variation on this theme holds that the electorate is better prepared to support untested policies, even if they may cause short-term duress
or if they run counter to established beliefs, when economic conditions have
deteriorated.26 However, once conditions improve or simply stabilize, tolerance subsides and support for further reforms wanes. Therefore, while
uncertainty is welcome at the outset of the reform process, certainty is the
key factor for its consolidation. Based on case studies of Peru and Argentina,
Kurt Weyland offers persuasive evidence that the public was supportive to
the reform process while there was a perception of acute economic crisis.27
Even though the reformers were reelected, support for their economic programs was already diminishing. Corrales endorses this view in his analysis
of the reform process in Argentina and Venezuela, although he acknowledges that in the latter case support for reform was never very strong.28
Therefore, the literature on economic voting suggests that policies, not
only outcomes, may influence electoral decisions. As with outcomes, voters
position with respect to policies may be mediated by a host of factors,
including ideological considerations, policy pronouncements during the
23. For a theoretical approach, see Cukierman and Tommasi (1998).
24. Stokes (2001a, p. 95).
25. Cukierman and Tommasi (1998); Navia and Velasco (2003).
26. This behavioral hypothesis is based on seminal work by Thaler and others (1997),
Kahneman and Tversky (1979), and Tversky and Kahneman (1991), who find that people are
more prone, even eager, to assume risks after experiencing losses.
27. Weyland (2002).
28. Corrales (2002).
11
electoral campaign, and the state of the economy at the time of elections.
This gives rise to the following additional testable propositions:
3. Electoral support for the incumbents party depends on the economic
policies adopted. Policies may carry electoral costs even when they deliver
good economic outcomes.
4. The electorates tolerance of unpopular policies depends on the ideology of the incumbents party, his or her campaign statements, and the
initial state of the economy.
Empirical Approach
None of the empirical literature just reviewed offers a full-fledged theoretical model of electoral behavior, and we have no intention of providing
one. However, the series of hypotheses arising from that literature can be
organized in a simple framework such that the persistence of the vote for
the incumbents party is a function of a vector of economic outcomes and
a vector of policies (both relative to their past values):
X
P
Vt
= A t t ut ,
Xt 1 Pt 1
Vt 1
where Vt and Vt 1 are the share of the vote for the incumbents party at the
end and the beginning, respectively, of its term in office; Xt and Xt 1
are the economic outcomes at the time of each election; and Pt and Pt 1
are the policies at those two moments. A is the set of other parameters that
may influence the stability of the vote for the party in office, and ut is an
error term. and are our parameters of interest. In this simple framework, hypothesis 1 states that is positive for economic outcomes that are
desirable, such as growth, or negative for undesirable ones, such as inflation or unemployment (and assumes that is zero, since it ignores the
influence of policies). Hypothesis 2 postulates that is a function of some
features of the political system, such as party fragmentation or the ideological polarization of the party system. The stronger these features, the
higher the electorates response to the economic outcomes. Hypothesis 3,
which postulates that the electorate cares about the choice of policies,
implies that is not zero but probably negative if the policies are market
oriented. Finally, hypothesis 4 states that some aspects of the political
and economic context when the incumbents party was initially elected
may affect the way the electorate judges the adoption of policies. This
hypothesis can be incorporated in our framework by assuming that is a

function of those factors. More specifically, will be smaller (in absolute
value) when the policies adopted were those announced by the incumbent
during his election campaign, when they are in line with his partys ideology, or when the economy started from a situation of crisis. Although our
framework is general enough to test further hypotheses, due to sample size
limitations and for the sake of parsimony and tractability, we restrict its
application to the hypotheses identified in the literature review.
Our economic voting framework is relevant both for presidential and
for legislative elections. An important feature of presidential systems is the
separation of powers between the legislative and the executive, aimed at
imposing checks and balances in order to discipline parties and make them
accountable.29 Since checks and balances force the two powers to agree on
policies, voters should be expected to pass judgment on the performance
of the incumbents party in both branches on the basis of economic outcomes and policy decisions. Of course, we should expect that the influence
of each policy on presidential vis--vis legislative elections will depend on
whether such policy is controlled exclusively by the executive or not. While
legislatures have very little influence on monetary, exchange rate, and tariff
policies in most Latin American countries, they do have a strong (even overriding) influence on tax policies, privatization decisions, and the regulation
of financial, capital, and labor markets. As Brian Crisp and Gregg Johnson
show, contrary to widespread belief, Latin American legislatures make use
of their powers to influence the timing and depth of promarket reforms.30
And according to Roberts and Wibbels, the electorate holds each branch of
power more accountable for some outcomes than for others.31 When assessing the role of the legislature in policy decisions in Latin America, it is important to keep in mind that the incumbents party (or the coalition of parties
backing the incumbent) usually holds the majority in that body (see below).
To estimate the relevant parameters, the previous expression can be
written in logs as
d log(Vt ) = + log( F ) + d log( Xt ) + d log( pt ) + t ,
where d log(Vt) corresponds to the change in (the log of) the share of votes
for the incumbent party between t, the time when its performance is eval29. Persson, Roland, and Tabellini (1997).
30. Crisp and Johnson (2003).
31. Roberts and Wibbels (1999).
13
uated, and t 1, when it was elected for office; d log(Xt) and d log(Pt) are
the changes in (log measures of the) outcomes and policies, respectively;
t is equivalent to log(ut); and + log(F ) is equal to log(A), with as a
constant parameter and F as a set of political control variables.
We estimate separate models for presidential and legislative elections
with panel data for seventeen countries starting from the mid-1980s
described below. Potential problems of heteroscedasticity and endogeneity need to be addressed in this type of specification. The former may arise
from country or party heterogeneity and is dealt with by the use of White
robust standard errors. The endogeneity problem stems from potential omitted
variables, since differentiating countries solely by the economic and policyrelated variables included in sets X and P may not capture all the sources of
heterogeneity.32 This is partly dealt with by the inclusion as controls of a set
of political variables (represented by F). However, other country-related
factors might bias the estimations if they are correlated with the explanatory
variables. To take care of this problem, we run all the regressions with country fixed effects (although, admittedly, our sample size is too small to get
precise estimation of these effects).33 The fixed effects estimator is
d log(Vt ) = + log( Ft ) + d log( Xt ) + d log( pt ) + C + t ,
where C is the set of country dummies.
Data and Sources

Table 2 presents the structure of our database, and table 3 shows correlations
between the more relevant variables. The database includes a total of sixtysix presidential elections and eighty-one legislative elections in seventeen
32. We assume that the two other sources of endogeneityreverse causality and measurement errorare not latent in our model. Reverse causality is not a concern, since voters
evaluate the incumbents behavior after policies and outcomes are known. Measurement error
problems may be present, depending on the actual process of expectations formation. However,
ample empirical evidence provides support for the hypotheses of retrospective voting, which
for our framework implies that expectations are formed on the basis of past outcomes only.
33. All the regressions were also run without fixed effects: while virtually all the conclusions are the same, in these regressions, some of the explanatory variables (especially those
measuring promarket policies) show higher levels of significance. We have also run the regressions including a common time trend, or including five-year period fixed effects, without any important divergence from the results presented below. Results are available upon
request from the authors.
3 (198999)
4 (198597)
3 (198998)
3 (198999)
5 (198602)
5 (198602)
5 (19862000)
4 (198898)
4 (198499)
4 (198599)
5 (19852001)
3 (19882000)
3 (19902001)
4 (19852000)
4 (19892003)
3 (198499)
4 (19882000)
66
Presidential
8 (198599)
4 (198597)
4 (198698)
4 (19892001)
5 (198698)
5 (19862002)
4 (19862000)
7 (198698)
6 (19852000)
5 (198597)
5 (19852001)
6 (19852000)
3 (19902001)
4 (19852000)
4 (19892003)
3 (198499)
4 (19882000)
81
Legislative
2
3
2
2
2
2
3
4
2
4
2
2
1
3
1
2
4
2.4
Presidency
3
3
2
2
1
2
3
2
2
4
2
1
2
4
1
2
1
2.2
Largest share in the legislature
Number of parties that held . . .
2.77
4.06
6.60
4.90
2.66
2.31
2.48
6.05
2.68
3.31
2.18
2.38
2.05
3.80
2.21
3.19
3.92
3.39
Mean
2.30
3.42
2.76
4.84
2.21
2.21
2.18
4.29
2.41
2.35
2.00
1.85
2.05
2.50
1.88
2.92
2.34
2.62
Minimum
Fragmentationa
3.06
5.08
8.27
4.99
3.09
2.56
2.88
7.56
3.06
4.44
2.58
2.82
2.05
5.83
2.54
3.32
5.79
4.11
Maximum
0.23
0.52
0.25
0.16
0.16
0.42
0.55
0.36
0.39
0.24
0.42
0.32
0.58
0.51
0.40
0.42
0.30
0.37
Polarization index
Source: Payne and others (2002), complemented with the Political Database of the Americas (Organization of American States and Georgetown University).
a. Effective number of parties in the legislature.
b. In Chile, the effective number of parties differs from the number of coalitions (Concertacin and Alianza por Chile), which are close to 2 in effective terms and of which only Concertacin has held the presidency.
Argentina
Bolivia
Brazil
Chileb
Colombia
Costa Rica
Dominican Republic
Ecuador
El Salvador
Guatemala
Honduras
Mexico
Nicaragua
Peru
Paraguay
Uruguay
Venezuela
Total or average
Country
Number of elections (period)
T A B L E 2 . Structure of the Data Set
Presidential elections
Votes (share)
Fragmentation
Polarization
Promises
Ideology
Growth (log, change)
Inflation (loss of
purchasing
power, change)
Unemployment (change)
Gini index (change)
Macro index
(log, change)
Structural index
(log, change)
Institutional Index
(log, change)
0.24
0.55
0.13
0.16
0.25
0.05
0.18
0.24
0.02
0.22
0.14
0.24
0.07
0.09
0.27
0.33
0.46
0.19
1.00
0.09
0.05
0.03
0.02
1.00
0.32
0.20
0.16
0.16
0.21
1.00
0.32
0.45
0.06
0.15
0.08
0.27
0.02
0.16
0.21
1.00
0.37
0.25
0.00
0.05
0.11
0.16
0.28
0.09
1.00
0.14
0.01
0.24
0.25
0.41
0.27
0.23
1.00
0.10
0.26
0.12
1.00
0.48
0.13
0.24
0.33
0.27
1.00
0.09
0.15
0.37
0.00
1.00
0.03
0.02
1.00
0.44
1.00
1.00
(continued )
Inflation (loss
Institutional
Votes
Growth
of purchasing Unemployment Gini index Macro index Structural index
index
(share) Fragmentation Polarization Promises Ideology (log, change) power, change)
(change)
(change) (log, change) (log, change) (log, change)
T A B L E 3 . Correlations
0.07
0.12
0.61
0.26
0.29
0.10
0.20
0.15
0.26
0.31
0.08
0.13
0.17
0.29
0.04
0.14
0.47
0.18
0.09
0.32
1.00
0.36
0.23
0.24
0.34
0.38
0.19
1.00
0.40
0.38
0.03
1.00
0.16
0.32
0.20
0.33
1.00
0.11
0.43
0.19
0.09
0.19
Legislative elections
Votes (share)
Fragmentation
Polarization
Promises
Ideology
Inflation (loss of
purchasing
power, change)
Unemployment
(change)
Gini index
(change)
Macro index
(log, change)
Structural index
(log, change)
Institutional Index
(log, change)
0.21
0.34
0.08
0.07
0.43
0.24
1.00
0.05
0.15
0.08
0.83
0.01
0.38
0.30
1.00
0.23
0.20
0.63
0.17
0.34
1.00
0.47
0.01
0.10
0.18
1.00
0.20
0.51
0.08
1.00
0.29
0.15
1.00
0.32
1.00
1.00
Inflation (loss
Institutional
Votes
Growth
of purchasing Unemployment Gini index Macro index Structural index
index
(share) Fragmentation Polarization Promises Ideology (log, change) power, change)
(change)
(change) (log, change) (log, change) (log, change)
T A B L E 3 . Correlations (continued )
17
Latin American countries over the period 19852002. Party alternation

was moderate over this period: the average number of parties that held
power or had a majority was 2.4 for presidential elections and 2.2 for legislative elections (with a maximum of 4 and a minimum of 1). However, the
effective number of parties (also known as political fragmentation) was
higher: 3.4 on average, with a maximum of 8.3 in Brazil and 7.6 in
Ecuador.34 Except for Mexico during the 1980s and Paraguay at the end on
that decade, none of the seventeen countries showed party fragmentation
below 2, implying a generally healthy level of political competition. The
ideological polarization of the political system was low during the period,
as measured by a polarization index that computes the (weighted average)
distance between the ideological positions of the parties on a scale from
0 to 1; parties are classified as extreme left, center left, center right, or
extreme right. When all the parties have the same ideological position, the
index takes the value 0, when half of them (measured by the number of
votes) are extreme left and the other half are extreme right, the index takes
the value 1.35 The average value of the index in our data set was 0.37, with
a maximum of 0.58 for Nicaragua and a minimum of 0.16 for Chile and
Colombia.
Dependent Variable
Our dependent variable is the change in the share (in logs) of votes36 for
the incumbents party in presidential elections, and for the majority party
34. The effective number of parties is calculated using the Laako-Taagepera index, defined as the inverse of the sum of the squares of the shares (measured by the number of seats)
of all the parties in the legislature; Payne and others (2002).
35. More precisely, the index is calculated in two steps. First, the average position of
the electorate on a left-right scale (APLR) is calculated as a weighted average of the party
positions on a scale from 1 to +1, where the weights are the shares of the votes:
APLR = 1*(% votes obtained by parties on the extreme left) 0.5*(% votes for parties
on the center left) + 0.5*(% votes for parties on the center right) + 1*(% votes for parties on
the extreme right).
In the second step, the polarization index (IP) is calculated as a weighed deviation from
the APLR:
IP = 1 APLR * (% votes left) + 0.5 APLR * (% votes center left) + 0.5
APLR * (% votes center right) + 1 APLR * (% votes right).
A minimum of 0 is reached when all the votes are in one ideological bloc; and a maximum of 1, when half of the votes are at each extreme. Ideological orientations are taken
from Coppedge (1997) and the World Banks Database of Political Institutions, 2002
(www.worldbank.org/research/bios/pkeefer.htm).
36. The share of votes comes from Payne and others (2002).
in the legislature in legislative elections. Since we use logs for both the
dependent and (when possible) the independent variables, the estimated
coefficients can be interpreted as elasticities.
Some calculations were necessary in order to compute the share of votes,
especially for presidential elections, when party coalitions or party dissolutions had taken place before and after the elections, as well as to be able to
account for new independent parties. These calculations treat coalitions as
regular parties. The vote for the coalition party in the election previous to its
creation is simply computed as the sum of the votes of the joining parties.
When parties break up, the same procedure is used for the following elections. Table 4, which presents summary statistics for the most important
variables, shows that the share of votes varies from 0 to 64 percent for
presidential elections and 62 percent for legislative elections, with means
of 35 percent and 36 percent, respectively.
Political Variables
The political variables used as independent variables attempt to measure
key dimensions of the party system and the political environment. Following the literature review, they are to be included in the regressions both as
independent controls and/or interacted with the variables measuring economic outcomes. Fragmentation (or the effective number of parties) and
polarization, already described, are the two basic dimensions of the party
system. In addition, we use a dummy for divided governments (when the
presidents party is not the largest party in the legislature).37
We also use several variables intended to measure the electorates expectations about the future orientation of economic policies. The first, named
promarket promises, measures to what extent the positions adopted by
incumbents during their election campaigns were promarket; it is a rescaled version of a variable computed by Stokes.38 The second, named
right-oriented ideology, a measure taken from the World Banks Database
37. Taken from Payne and others (2002). Divided government is not frequent in Latin
America, in contrast to the United States; Alesina, Londregan, and Rosenthal (1993);
Alesina and Rosenthal (1995, 1996); Fiorina (1992). The only cases in our data set are
mainly concentrated in Brazil and Ecuador (six), with one in the Dominican Republic. More
recently, the PRI lost its monopoly power in Mexico.
38. Based on an ordinal variable computed by Stokes (2001a, p. 3) that classifies forty presidential pre-electoral campaigns according to the importance assigned by the candidates to
issues of economic security vis--vis economic efficiency, the promises variable takes values
on a scale from 0 to 1, where higher values indicate more efficiency-oriented messages.
Dependent
Votes (shares)
Political
Fragmentation
Polarization (01)
Divided government (dummy)
Rule of law
Promises (01)
Ideology (01)
Switch index with ideology (1, +1)
Switch index with promises (1, +1)
Outcome
Inflation (log, change)
Crisis
Unemployment (log, change)
Gini index (log, change)
Macro index (log, change)
Structural fiscal balance (ratio to GDP, change)
Real exchange rate (detrended in logs, change)
Social expenditures (share of GDP, change)
Structural reforms index
Structural reform index (log, change)
Trade index (log, change)
Financial index (log, change)
Tax index (log, change)
Privatizations index (change)
Institutional index
Variable
T A B L E 4 . Summary Statistics
0.03
0.01
0.01
0.00
0.01
0.09
0.01
0.02
0.00
0.01
0.12
0.00
0.00
0.01
0.23
0.18
0.30
0.10
0.09
0.27
49
52
52
44
49
51
44
51
50
49
49
51
52
52
51
0.19
0.06
0.03
0.06
0.00
0.19
0.06
0.00
0.04
0.00
2.70
0.33
0.00
0.44
0.50
0.50
0.00
0.01
3.33
0.38
0.13
0.23
0.55
0.60
0.00
0.05
52
51
47
16
33
42
39
32
0.36
Median
0.35
Mean
48
Number of
observations
0.21
0.30
0.44
0.23
0.17
0.35
0.40
0.04
0.26
0.02
0.24
0.08
0.04
0.03
0.03
1.62
0.22
0.34
0.50
0.26
0.33
0.08
0.12
0.17
Standard
deviation
Presidential elections
0.32
0.14
0.15
0.09
0.10
1.70
0.12
0.67
0.05
0.78
1.15
1.38
0.81
0.97
1.26
1.00
0.07
0.61
0.02
0.09
0.32
0.15
0.29
0.03
0.22
8.27
0.92
1.00
1.19
1.00
1.00
0.48
0.24
0.64
Maximum
0.74
0.25
0.00
0.08
0.10
1.85
0.03
0.00
0.81
0.25
0.00
0.11
0.06
0.00
Minimum
68
68
70
72
72
71
70
63
70
69
71
72
72
63
69
72
67
72
16
25
68
39
72
Number of
observations
0.17
0.15
0.23
0.10
0.06
0.19
0.08
0.00
0.01
0.01
0.07
0.02
0.02
0.00
0.00
3.37
0.34
0.01
0.23
0.56
0.55
0.00
0.36
Mean
0.12
0.04
0.03
0.03
0.00
0.10
0.04
0.00
0.01
0.00
0.04
0.02
0.00
0.00
0.01
2.80
0.30
0.00
0.44
0.50
0.33
0.00
0.38
Median
0.19
0.29
0.35
0.23
0.15
0.28
0.34
0.03
0.22
0.02
0.20
0.06
0.03
0.03
0.03
1.54
0.19
0.12
0.50
0.27
0.32
0.08
0.14
Standard
deviation
Legislative elections
0.09
0.24
0.15
0.29
0.01
0.51
0.99
0.07
0.61
0.03
0.71
0.19
0.14
0.06
0.10
1.85
0.03
0.00
0.81
0.25
0.00
0.11
0.00
Minimum
0.78
1.15
1.38
0.71
0.97
0.93
1.70
0.11
0.67
0.06
0.40
0.15
0.00
0.07
0.10
8.27
0.91
1.00
1.19
1.00
1.00
0.48
0.62
Maximum
of Political Institutions and a study by Coppedge, classifies parties on a left

to right scale according to their economic ideology.39 In order to test the
Stokes hypothesis on the electoral effects of switching, we have created
two types of switch indexes, one measuring the deviation between the amount
of promarket reforms implemented by the administration (see below for the
description of this variable) and the promarket promises during the campaign, and the other measuring the deviation between the reforms and the
measure of right-oriented ideology of the party.40 Note that only the latter
is applicable to legislative elections.
Economic Outcomes
Following the empirical literature on economic voting, we focus on inflation
and growth as the two main economic outcomes of interest, but we also test
other variables, such as unemployment and income concentration. We measure inflation as the average annual loss of purchasing power of a currency
unit, rather than as the increase in the price index, since this reduces the
extreme observation problem that arises with the cases of high or hyperinflation. We apply the formula 1 (1/(1 + )), where is the price increase
during the last year of the administration. Economic growth is measured
as the rate of annual change (in logs) in GDP. In addition to inflation and
growth, we test for the influence of two other outcomes: the unemployment
rate and the Gini coefficient of distribution of per capita household income.41
Policy Variables
As mentioned in the introduction, we define the Washington Consensus in
accordance with the list of policies included in Williamson.42 Since those
policies cover a variety of areas, from fiscal to institutional, we use the
39. World Bank, Database of Political Institutions, 2002 (www.worldbank.org/research/
bios/pkeefer.htm); Coppedge (1997).
40. The switch indexes range from 1 to +1. In the first type, 1 indicates that having
adopted the most pro-efficiency stance during the campaign, the candidate does not implement any promarket reform once in office; +1 indicates that having adopted the most prosecurity position in the campaign, the candidate once in office becomes the most aggressive
promarket reformer. The formula is then SI = [change in reforms (PROMISES median
PROMISES)]. In the other type of switch index, the variable PROMISES is replaced by our
measure of party ideology.
41. Prices and GDP are taken from International Monetary Fund, World Economic Outlook (online). Unemployment is from ECLAC (various years). Gini coefficients for incomes
are from Deininger and Squire (1998).
42. Williamson (1990a).
21
followingadmittedly somewhat arbitraryclassification (numbers in

parentheses refer to Williamsons list):
Macroeconomic policies: fiscal discipline (1), public expenditure on
social services and infrastructure (2), and competitive exchange rates (5).
Structural reforms: tax reform (flat, low, and effective tax rates) (3),
interest rate liberalization (4), trade liberalization (6), liberalization of
foreign direct investment inflows (7), and privatization (8).
Institutional reforms: deregulation of entry and exit (9) and protection
of property rights (10).
The most important distinction is that between macroeconomic policies and structural reforms, the latter referring to sectoral or microeconomic
policies that affect the functioning of specific markets (imports, credit,
infrastructure services, and so forth). The inclusion of public expenditure
on social services and infrastructure as a macroeconomic policy is arbitrary but justifiable for the sake of simplicity. Institutional reforms include
protection of property rights, a policy that is usually seen not as a core element of the Washington Consensus (as a matter of fact, it was added by
Williamson as an afterthought) but rather as a key element of what analysts starting with Moiss Naim have referred to as second generation
reforms.43 However, these also include regulatory institutions, modernization of the state apparatus (especially for the provision of social services),
and reform of the judiciary sector, none of which are considered here.
We use quantitative indicators to measure eight of the ten policies that
constitute the Washington Consensus, as well as composite indexes for
macroeconomic policies and structural reforms. We do not have quantitative indicators for foreign direct investment policies or deregulation of
entry and exit. Therefore, these policies are not included in our reform
indexes. A brief description of the policy indicators follows (further
details are in the footnotes):
Fiscal discipline is measured by the fiscal balance of the central
government, adjusted by the endogenous influence of the economic cycle
and changes in the terms of trade on fiscal revenues. The purpose of
these adjustments is to isolate the exogenous or policy component of the
fiscal balance, which is a better measure of fiscal discipline than the
observed fiscal balance.44 Fiscal balance, fiscal revenue, and GDP data
43. Nam (1994).
44. Specifically, we subtract from the fiscal balance of the central government the revenue that is associated with either the economic cycle or the terms of trade cycle (obtained
applying standard Hodrick-Prescott filters).
used in this calculation are from the World Bank and terms of trade data
are from the Economic Commission for Latin America and the Caribbean
(ECLAC).45
Public expenditure in social services includes only education and
health expenditures, based on data from ECLAC and complemented with
data from World Bank.46
The measure of competitive exchange rates is the log distance between
the observed real exchange rate and its trend, computed with a standard
Hodrick-Prescott filter.47
Tax reform is taken from previous work by Lora, who constructs a
composite index of the levels and effectiveness of corporate, personal, and
value added taxes.48
Interest rate liberalization is measured by Loras index of financial
liberalization, which includes information on interest rate freedom, reserve
requirements, and quality of regulation and supervision of the financial
sector.
Trade liberalization is also taken from Lora, who uses an index that
combines import tariff averages and dispersion.
Privatization is measured by Loras index of the cumulated value of
the sales of state-owned firms to the private sector, as a share of the GDP.
Protection of property rights is a combined measure of the risk of
expropriation and the risk of repudiation of government contracts, on a
scale from 0 to 1 (the higher the index, the lower the risk).49
The composite index of macroeconomic policies is a simple average
of the indicators of its three components scaled from 0 to 1, where 0 corresponds to the lowest observation and 1 to the highest observation for the
whole period and set of countries in the sample.
The composite index for structural reforms is calculated as the simple average of the indexes for tax reform, financial liberalization, trade liberalization, and privatization (each of which is also calculated on a scale
from 0 to 1).50
45. World Bank, World Development Indicators (online); ECLAC (various years).
46. World Bank, World Development Indicators (online); ECLAC (various years).
47. Real exchange rate data are from IMF, World Economic Outlook (online).
48. Tax reform, interest rate liberalization, and privatization are all from Lora (2001).
49. Taken from the International Country Risk Guide, 2004 (www.icrgonline.com).
50. Note that this composite index is not identical to the total reform index computed
by Lora (2001), since the latter includes labor reform, which is not among the Washington
Consensus policies.
23
All variables are measured as changes between the previous election

year and the current election year. Since taking the current election year is
somewhat arbitrary, we checked the robustness of our main results by also
using the year prior to the election year.51
Econometric Results
Before discussing the hypotheses in detail, it is helpful to convey the thrust
of our findings. The regression summarized in table 5 indicates that the
electorate is highly sensitive to one economic outcomeinflationand
strongly rejects the adoption of promarket policies. Our estimates imply
that the typical reduction in the rate of inflation, from say 20 percent to
8 percent during a presidents tenure, boosts the vote for his party by
21 percent.52 However, if that same incumbent also introduces the average
amount of promarket reform, the resultant party losses account for 23 percent of the vote. Put a different way, the adoption of the standard Washington Consensus package brings positive electoral payoffs only when
implemented in a period of high inflation. Thus, if the same dose of promarket reform is adopted as part of a package that reduces inflation from
100 percent to 8 percent, the net electoral effect is a handsome 82 percent
increase in vote share.
Admittedly, our basic regression overstates the negative effect of the
promarket policies because those policies may help to reduce the rate of
inflation and increase the rate of growth.53 Taken to the extreme, this argument would imply that the total effect of the adoption of promarket policies would be the sum of the direct effect captured in the coefficient of the
regression in table 5 and the indirect effects of the changes in the rates of
inflation and growth. Based on this calculation (see tables 6A and 6B), the
total effect does appear to be substantially milder: 0.97 instead of 1.57.
Nevertheless, it would still be substantial, as it would imply that the typi-
51. These results, which are not included in this version of the paper, are available upon
request from the authors.
52. This reduction corresponds to the average value of our measure of the change of
inflation.
53. The regression includes several other control variables that may also affect the vote
(see notes to table 5).
T A B L E 5 . Impact of Economic Outcomes and Washington Consensus Policies
in Presidential Elections, 19852002: Country Fixed Effects Results
Independent variablesa
Economic outcomes
Inflation (change in loss of purchasing power)
Growth (change in growth rate, log)
Washington Consensus reforms
Structural reforms index (log, change)
Constant
Summary statistic
Number of observations
Number of countries
R2
Country fixed effects
Dependent variable: Change in vote shareb

2.030 (2.09)*
1.016 (0.74)
1.569 (2.98)***
0.627 (0.85)
37
17
0.80
Yes

*Significant at 10 percent; ***significant at 1 percent.
a. The regression also includes as control variables measures of divided government, polarization, and fragmentation (see text for
definitions and method of calculation).
b. The dependent variable is the change in the log of the vote share of the incumbent presidents party. Robust t statistics are in
parentheses.
cal reformist government must still sacrifice 15 percent of the vote for the
sake of the reforms.54 However, this calculation most likely overestimates
the effects of the reforms on growth and inflation, as we have not isolated
the influence of other factors on these two variables. Therefore, the central
conclusion is that even if we grant that promarket reforms have strong beneficial effects on growth and inflation, their electoral cost is far from negligible.
Apart from promarket reforms, the other Washington Consensus policies do not affect the electorates behavior. Likewise, we find no robust
evidence that any economic outcomes other than inflation affect the vote
in presidential elections. We do find that these results are affected by some
features of the political system. In legislative elections the results are less
straightforward, as they are strongly mediated by several contextual and
political variables.
Do Outcomes Matter?
We start our empirical analysis by testing the simplest version of the economic voting model, in which voters update their opinion on the incumbents party based entirely on the changes observed since the last election
54. Note that the total effect would be reduced only slightly (to 0.84) if the indirect
effect through growth, which has the wrong sign, is not included.
25
T A B L E 6 A . Rough Estimate of Total Effect of Promarket Reforms

on the Presidential Vote (Elasticities)
Effect
a
Effect of reforms on inflation or growth

Effect of inflation or growth on the vote
Indirect effect of reforms on the vote (via inflation or growth)
Direct effectb
Total effect (indirect + direct)
Inflation
Growth
0.361
2.030
0.733
0.133
1.016
0.135
Total
0.598
1.569
0.971

a. See table 6B for regressions.
b. As estimated in table 5.
T A B L E 6 B . Estimates for Inflation and Growth Used in Table 6A

Constant
Summary statistic
Number of countries
R2
Inflationa (1)
Growthb (2)
0.361 (1.88)*
0.103 (0.48)
0.133 (2.10)**
0.063 (0.78)
49
17
0.37
Yes
49
17
0.26
Yes

*Significant at 10 percent; **significant at 5 percent.
a. Dependent variable is the independent variable from table 5; that is, the change in the inflation rate, where inflation is measured
as the annual loss of purchasing power of the currency. Robust t statistics are in parentheses.
b. Dependent variable is the independent variable from table 5; that is, the change in the growth rate, in logs. Robust t statistics are
in parentheses.
in the key economic variables, Xt. As mentioned, we include as additional

controls a set of political variables (represented by F below) that may affect
the stability of the vote share, namely, our measures of political fragmentation, polarization, and divided government (lagged to reduce endogeneity
and better capture the political environment prevailing during the administration).55 Since other country-specific factors may also have an influence on
the persistence of the vote for the incumbents party, we attempt to isolate
them by using fixed effects. We start with
d log(Vt ) = + log( Ft 1 ) + d log( Xt ) + C + t .
55. Divided government is a dummy equal to 1 when the party with the greatest representation in the legislature is not the incumbents party (this is unusual in Latin America:
in our database, it occurs in only seven instances). In regressions not shown, a dummy for
midterm elections was also included in legislative elections. It was never significant and it
did not affect any of the results.
43
17
0.47
Yes
0.804
(1.73)*
0.345
(1.05)
0.588
(1.31)
(1)
43
17
0.48
Yes
0.453
(1.41)
0.938
(0.80)
0.442
(0.82)
(2)
43
17
0.48
Yes
1.028
(2.13)**
0.588
(1.19)
0.939
(0.79)
0.539
(1.11)
(3)
43
17
0.6
Yes
0.743
(1.30)
1.127
(1.80)*
0.121
(0.18)
0.700
(0.73)
0.451
(0.86)
(4)

a. See text for descriptions and method of construction of independent variables. Robust t statistics are in parentheses.
Summary statistic
Number of countries
R2
Constant
Unemployment rate (change)
Gini index (change)
Economic outcomes
Divided government (dummy, lagged)
Polarization (lagged)
Political
Fragmentation (lagged)
Independent variable
43
17
0.48
Yes
1.030
(1.99)*
0.034
(0.03)
0.589
(1.14)
0.938
(0.78)
0.537
(1.09)
(5)
41
17
0.48
Yes
1.157
(1.82)*
0.624
(0.23)
0.669
(1.19)
1.192
(0.70)
0.584
(1.05)
(6)
37
15
0.56
Yes
2.510
(1.06)
0.824
(1.01)
0.364
(0.45)
3.636
(1.11)
0.629
(1.98)*
(7)
Dependent variable: Change in log of vote share of incumbent presidents party
T A B L E 7 . Impact of Economic Outcomes in Presidential Elections, 19852002: Country Fixed Effects Resultsa
37
15
0.74
Yes
1.924
(2.08)*
2.683
(1.45)
2.256
(0.70)
4.307
(1.24)
0.806
(0.99)
0.546
(0.59)
3.150
(1.44)
0.171
(0.48)
(8)
43
17
0.66
Yes
0.786
(1.56)
1.674
(1.91)*
2.828
(1.70)
0.023
(0.04)
0.509
(0.64)
0.269
(0.76)
(9)
27
Table 7 lends some support to this simple version of the economic voting
hypothesis: in all the regressions, changes in inflation have the expected sign
and have a significant impact on the presidential vote. However, changes in
growth rates are seldom significant and when included in a regression with
inflation show the wrong sign. Results for unemployment and inequality
are similarly weak. When all four economic variables are included in the
same regression, inflation remains the only significant variable. In legislative elections (see table 8) growth is the only one that is sometimes
significantbut it is not when all economic variables are included in the
same regression. Therefore, inflation and growth seem to matter for the leading party or parties, but through different channels. The size of the coefficients suggests that the incumbent loses 12 percent of his vote for
each (additional) 1 percent of (annual) loss in the purchasing power of the
currency in the last year of his administration (with respect to the loss in
the year prior to his administration). Likewise, the largest party in the legislature increases its share of seats by about 1 percent for each (additional)
1 percent of economic growth in the year before the election (with respect
to the year immediately before the previous election). Neither changes in
the unemployment rate nor income distribution changes appear to have a
clear effect on electoral behavior.
These conclusions must now be qualified in accordance with our second hypothesis, namely, that the electorates response to the economic
outcomes, , depends on several features of the political system, F (some
of which, as tables 7 and 8 show, also have a direct influence on voters
behavior):
= + log( F0 ).
Replacing in the previous equations gives (with fixed effects)
d log (Vt = + log ( Ft 1 + d log ( X t
+ log ( F0 d log ( X t + C + t .
Note that in the interaction terms we use the values of F at the earliest period of our sample, F0, in order to reduce endogeneity. However,
we use the values of F at the beginning of each electoral cycle, Ft 1, to
directly control for these variables, since the inclusion of country fixed effects
74
17
0.53
Yes
0.911
(3.33)***
0.007
(0.05)
0.868
(3.86)***
(1)
74
17
0.30
Yes
0.204
(1.19)
0.068
(0.21)
0.279
(1.52)
(2)
74
17
0.53
Yes
0.998
(3.31)***
0.894
(4.01)***
0.240
(0.70)
0.014
(0.09)
(3)
71
17
0.40
Yes
0.649
(2.77)***
0.086
(0.38)
0.572
(2.99)***
0.157
(0.53)
0.011
(0.08)
(4)

*Significant at 10 percent; **significant at 5 percent; ***significant at 1 percent.
Summary statistic
Number of countries
R2
Constant
Unemployment (change)
Gini index (change)
Economic outcomes
Political
71
17
0.43
Yes
0.526
(2.26)**
0.861
(1.77)*
0.507
(3.26)***
0.125
(0.49)
0.052
(0.41)
(5)
72
17
0.55
Yes
1.027
(3.39)***
1.322
(0.79)
0.915
(4.23)***
0.304
(0.78)
0.017
(0.12)
(6)
65
15
0.60
Yes
1.529
(0.86)
1.356
(3.69)***
1.056
(4.42)***
1.116
(2.26)**
0.029
(0.19)
(7)
(8)
62
15
0.47
Yes
0.07
(0.26)
0.064
(0.10)
0.043
(0.03)
1.070
(0.59)
0.937
(2.75)***
0.705
(2.87)***
0.853
(2.08)**
0.001
(0.00)
Dependent variable: Change in log of vote share of party with most seats in legislature
T A B L E 8 . Impact of Economic Outcomes in Legislative Elections, 19852002: Country Fixed Effects Resultsa
71
17
0.43
Yes
0.542
(2.33)**
0.053
(0.23)
0.913
(1.64)
0.529
(2.85)***
0.120
(0.46)
0.051
(0.40)
(9)
29
precludes the use of time-invariant F0. None of the results reported below is
sensitive to whether we include the set of F variables as direct controls.
Tables 9 and 10 suggest that the electorates response to the economic
outcomes is indeed affected by the structure of the political system, and in
the expected manner. In presidential elections (table 9), the more fragmented the party system, the more harshly the electorate punishes the
incumbents party for an increase in the inflation rate.56 The intuition
behind this result is that in more fragmented party systems there is more
competition for votes, and probably also more information available to
the voters and a wider choice of policy proposals, all of which enhance
the response of the electorate to changes in the economic situation. One
should expect this response to be stronger in presidential than in legislative elections, given the winner-take-all nature of the former. A divided
government affects the response of the electorate to inflation in a similar
way. However, due to the small number of cases of divided government,
we do not attach much relevance to this result.57 In contrast to party fragmentation, the degree of polarization does not seem to have any significant
influence on the electorates response to the economic outcomes in presidential elections.
In legislative elections (table 10), the opposite is the case: while the interaction terms between economic outcomes and fragmentation are not significant, the interaction with ideological polarization is significant for inflation
and for growth. This implies that the more distanced the economic policy
platforms of the parties, the stronger the swings of the electorate in response
to changes in the macroeconomic outcomes. From regression 5, when the
degree of polarization is high (0.53), each percentage point of extra growth
brings an increase of about 1 percent in the vote for the largest party in the
legislature, while this elasticity becomes negative (0.4) when the degree of
polarization is low (0.15). Our results indicate that the legislative vote is also
sensitive to inflation outcomes, depending on the degree of ideological
polarization of the party system, with implied elasticities of 0.3 when
polarization is high and 0.38 when polarization is low (regression 2).
To summarize, our results suggest that economic outcomes do matter in
presidential as well as in legislative elections, though in different ways.
The executive is held more accountable for increases in inflation, and
56. However, this result does not hold in a similar regression without fixed effects (results
available from the authors upon request).
57. Furthermore, similar regressions for growth show implausibly high coefficients for
the interaction term (GROWTH*DIVIDED GOVERNMENT).
Growth*initial polarization
Growth*initial fragmentation
Inflation*divided government
Inflation*initial polarization
Inflation*initial fragmentation
Economic outcomes and interactions

Political
3.153
(2.66)**
3.351
(3.50)***
1.316
(2.27)**
0.426
(0.77)
0.5
(1.46)
(1)
5.039
(1.62)
3.288
(1.86)*
0.187
(0.24)
0.054
(0.08)
0.19
(0.40)
(2)
4.868
(19.97)***
0.402
(1.89)*
0.692
(1.53)
0.205
(0.40)
0.373
(0.92)
(3)
1.717
(0.48)
1.751
(0.40)
0.732
(1.15)
1.002
(0.83)
0.492
(0.91)
(4)
2.849
(0.40)
1.198
(0.49)
0.667
(1.35)
1.04
(0.78)
0.664
(1.00)
(5)

(6)
1.032
(0.95)
0.141
(0.19)
1.518
(1.11)
1.028
(1.59)
T A B L E 9 . Impact of Economic Outcomes Interacted with Political Features in Presidential Elections, 19852002: Country Fixed Effects Resultsa
43
17
0.80
Yes
1.545
(2.45)**
43
17
0.70
Yes
0.474
(0.52)

Summary statistic
Number of countries
R2
Constant
Growth*divided government
43
17
0.87
Yes
0.938
(2.10)**
43
17
0.49
Yes
1.191
(1.95)*
43
17
0.49
Yes
1.142
(2.03)*
43
17
0.59
Yes
36.076
(1.80)*
0.427
(0.67)
Inflation*divided government
Inflation*initial polarization
Inflation*initial fragmentation
Economic outcomes and interactions

Political
0.560
(0.61)
0.680
(0.66)
0.606
(3.25)***
0.135
(0.45)
0.007
(0.05)
(1)
1.781
(3.03)***
0.646
(2.16)**
0.566
(2.77)***
0.205
(0.76)
0.069
(0.50)
(2)
0.495
(1.37)
0.322
(1.14)
0.565
(2.96)***
0.132
(0.52)
0.039
(0.31)
(3)
0.035
(0.03)
0.523
(3.32)***
0.124
(0.48)
0.038
(0.29)
(4)
1.024
(1.54)
0.566
(3.75)***
0.078
(0.35)
0.002
(0.02)
(5)
2.678
(1.54)
0.468
(2.56)**
0.066
(0.25)
0.158
(1.05)
(6)
0.429
(0.57)
1.263
(1.77)*
0.543
(1.68)*
0.557
(2.82)***
0.127
(0.57)
0.044
(0.31)
(7)
T A B L E 1 0 . Impact of Economic Outcomes Interacted with Political Features in Legislative Elections, 19852002: Country Fixed Effects Resultsa
71
17
0.41
Yes
0.687
(3.00)***
71
17
0.46
Yes
0.699
(2.89)***
71
17
0.45
Yes
0.630
(2.53)**

Summary statistic
Number of countries
R2
Constant
Growth*divided government
Growth*initial polarization
Growth*initial fragmentation
71
17
0.44
Yes
0.553
(2.32)**
0.981
(0.69)
71
17
0.48
Yes
0.631
(2.84)***
3.781
(3.40)***
71
17
0.45
Yes
2.072
(1.15)
0.340
(1.26)
71
17
0.51
Yes
0.647
(2.66)**
2.548
(1.80)*
more so in highly fragmented party environments. The largest party in the

legislature (which usually is the incumbents) is rewarded when economic
growth improves, and this reaction seems to increase with the degree of
ideological polarization.58 Party polarization even makes the legislative
vote sensitive to changes in the inflation rate.59 Our results thus lend support to hypotheses 1 and 2 above.
Do Policies Matter?
The next step is to establish whether the electorate cares about policies,
and not only about outcomes. For parsimony, and given our limited sample sizes, we ignore the influence that the features of the political system
may have on voters sensitivity to the economic outcomes. We also ignore
other factors that may affect how the electorate feels about the adoption of
certain policies and focus on the direct electoral effects of the policies
themselves, as captured in :
d log(Vt ) = + log( Ft 1 ) + d log( Xt ) + d log( Pt ) + C + t .
The first four regressions in table 11 assess the influence on presidential elections of the set of macroeconomic policy indicators defined earlier.
The only indicator that shows some significance is the structural fiscal balance, which appears with a negative sign in regression 2, implying that the
electorate reacts against fiscal restraint (however, the coefficient implies
that this effect is very small). Note that inflation always keeps the right
sign and remains significant in this particular regression, although it loses
its significance in some others. Therefore, although the electorate seems to
want price stability, it does not rewardand may even punishan incumbent for some of the macroeconomic policies that may be needed to achieve
those outcomes, such as stronger fiscal balance.
The electorate is more emphatically opposed to some of the promarket
reforms, according to regressions 5 to 9. The coefficients for the total index
of reforms and for trade liberalization policies are highly significant, with
elasticities of 1.57 and 0.84, respectively. Regression 5 is the basis for
the analysis in tables 3 and 4, where we show that the total electoral pay58. None of our main conclusions, in either this or the following sections, is altered
when the regressions are run for the share of votes of the incumbents party (results available from the authors upon request).
59. All these results persist when the set of political control variables is excluded from
the regressions.
35
off of the reforms remains strongly negative, even if we take into account
the full indirect effects implied in the correlations between the changes in
the reform index and the changes in inflation and growth. As mentioned, the
point estimate of the direct effect implies that the incumbents party typically lost 23 percent of its vote in presidential elections on account of
the average amount of promarket reforms introduced during its term (or
15 percent if we take into account our rough estimate of indirect effects).
More aggressive reformerssay, those reforming 1 standard deviation
above the meanwould sacrifice 40 percent of their vote on account of all
the promarket reforms (or 27 percent with the indirect effects). As the
remainder of the paper shows, the negative electoral payoff of the adoption of promarket reforms is a remarkably robust result.
Regression 10 evaluates the effect of the protection of property rights
and finds that it does not influence the behavior of the electorate. Regression 11 is an attempt to summarize the influence of all the Washington
Consensus policies, using the composite indexes for the macroeconomic
and structural policies, along with the index of property rights. This regression indicates that while the electorate does not hold strong views on
macroeconomic or property rights policies, it does on promarket policies.
Finally, the last two regressions in table 11 test the robustness of the policy variables that were found to be significant in previous regressions,
namely, the fiscal balance, the total reform index, and the trade liberalization index. Only the total reform index is robust to the inclusion of the
other variables.
In summary, this evidence lends support to the hypothesis that the electorate rewards the incumbents party for good macroeconomic results
inflation, in particularbut punishes it for the adoption of the promarket
policies endorsed by the Washington Consensus.
Table 12 presents a similar set of regressions for legislative elections.
Those that test the significance of the macroeconomic policy indicators are
consistent with the conclusion that the electorate does not care about these
policies. However, in regression 3 the real exchange rate is significant at
10 percent with a positive sign, suggesting that the electorate favors more
depreciated exchange rates.60 The set of regressions dealing with the various indicators of promarket reforms suggests that they do not carry electoral
costs in legislative elections. Since some of these policies fall under the
control of the executive, this result is not surprising. However, as we show
below, privatizations, which are strongly influenced by the legislature, do
60. However, this result does not hold in a similar regression without fixed effects.
Trade reform index (log, change)
Macroeconomic reforms index (log, change)
Economic outcomes
Political
1.087
(1.21)
0.934
(0.95)
0.695
(1.99)*
(2)
0.463
(1.72)
0.066
(2.20)**
1.544 1.82
(1.61)
(2.15)*
2.861 2.788
(1.75)* (1.59)
0.122
(0.16)
0.341
(0.42)
0.19
(0.51)
(1)
(4)
(5)
0.403
(0.88)
6.79
(1.31)
1.569
(2.98)***
1.793 1.558 2.03

(1.54) (1.35) (2.09)*
2.14 2.998 1.016
(1.47) (1.61) (0.74)
0.263 0.276 1.436

(0.36) (0.36) (1.55)
0.365 0.391
2.298
(0.48) (0.48) (1.72)
0.396 0.211
0.103
(1.07) (0.61) (0.51)
(3)
0.844
(2.24)**
1.736
(1.81)*
1.256
(0.99)
0.945
(0.97)
1.834
(1.21)
0.199
(0.57)
(6)
(9)
0.099 0.118
(0.12) (0.15)
0.35
0.608
(0.42) (0.61)
0.258 0.401
(0.69) (0.87)
(8)
(11)
0.052 1.722
(0.06) (2.11)*
0.159 2.114
(0.21) (2.17)**
0.209 0.107
(0.61) (0.46)
(10)
1.347
(1.93)*
3.357
(2.48)**
0.194
(0.69)
(12)
0.007
(0.41)
1.938
1.825
(3.95)*** (4.23)***
0.353
(1.43)
2.381 1.674 1.899 2.424 3.195

2.735
(2.00)* (1.52) (1.86)* (2.10)* (3.16)*** (3.82)***
3.447 2.652 2.54 3.05 1.793
2.211
(1.78)* (1.55) (1.63)
(1.82)* (1.32)
(1.22)
0.043
(0.05)
0.165
(0.23)
0.177
(0.53)
(7)
T A B L E 1 1 . Impact of Economic Outcomes and Washington Consensus Policies in Presidential Elections, 19852002: Country Fixed Effects Resultsa
0.845
(1.82)
0.012
(0.38)
2.159
(3.11)**
2.003
(1.17)
1.026
(0.84)
4.456
(2.60)**
0.404
(0.65)
(13)
40
15
0.67
Yes
0.792
(1.18)
33
15
0.79
Yes
1.837
(1.83)*
40
17
0.64
Yes
39
16
0.64
Yes
0.933 0.85
(1.56) (1.47)
37
17
0.80
Yes
0.627
(0.85)

Summary statistic
Number of countries
R2
Constant
Institutional reforms index (log, change)
Financial reform index (log, change)
37
17
0.76
Yes
0.061
(0.08)
39
17
0.70
Yes
0.759
(0.92)
0.073
(0.37)
0.776
(1.50)
40
17
0.63
Yes
40
17
0.69
Yes
0.826 0.647
(1.31) (1.03)
0.064
(0.09)
39
16
0.71
Yes
36
15
0.89
Yes
0.232 0.049
(0.73) (0.20)
0.654 0.805
(0.79) (1.07)
31
15
0.91
Yes
0.197
(0.29)
31
15
0.91
Yes
0.62
(0.62)
0.877
(1.43)
Macroeconomic reforms index (log, change)
Economic outcomes
Political
0.08
(0.74)
0.077
(0.33)
0.948
(1.60)
0.588
(3.24)***
0.138
(0.54)
0.038
(0.29)
(1)
0.002
(0.20)
0.086
(0.33)
1.037
(1.50)
0.618
(3.01)***
0.174
(0.58)
0.02
(0.12)
(2)
0.281
(1.78)*
0.048
(0.21)
0.916
(1.69)*
0.59
(3.65)***
0.11
(0.48)
0.034
(0.27)
(3)
0.089
(0.36)
1.034
(1.78)*
0.593
(3.19)***
0.142
(0.54)
0.024
(0.17)
(4)
0.007
(0.03)
0.786
(1.26)
0.524
(2.60)**
0.26
(1.10)
0.043
(0.32)
(5)
0.025
(0.10)
0.712
(1.16)
0.533
(2.43)**
0.292
(1.30)
0.038
(0.29)
(6)
(8)
0.023
(0.09)
0.599
(1.05)
0.051
(0.22)
1.086
(1.84)*
0.595
0.615
(3.03)*** (3.15)***
0.314
0.173
(1.27)
(0.66)
0.021 0.024
(0.16)
(0.18)
(7)
0.609
(3.19)***
0.279
(1.03)
0.035
(0.26)
(10)
0.018
0.025
(0.07) (0.10)
1.084
0.703
(1.89)* 1.16
0.507
(2.51)**
0.07
(0.26)
0.061
(0.48)
(9)
0.167
(1.12)
0.071
(0.28)
0.133
(0.23)
0.524
(2.64)**
0.438
(2.17)**
0.033
(0.23)
(11)
0.327
(1.84)*
0.006
(0.03)
0.693
(1.21)
0.551
(3.00)***
0.224
(1.04)
0.034
(0.25)
(12)
T A B L E 1 2 . Impact of Economic Outcomes and Washington Consensus Policies in Legislative Elections, 19852002: Country Fixed Effects Resultsa
0.305
(1.78)*
0.004
(0.02)
0.653
(1.13)
0.572
(2.65)**
0.276
(1.35)
0.018
(0.13)
(13)
0.134
(0.71)
0.083
(0.73)
0.057
(0.66)
0.318
(1.03)
0.131
(0.83)
0.296
(1.44)
0.103
(0.58)
0.23
(0.78)
0.087
(0.71)
68
15
0.45
Yes
61
15
0.41
Yes
68
17
0.49
Yes
67
16
0.41
Yes
65
17
0.48
Yes
65
17
0.48
Yes
67
17
0.46
Yes
69
17
0.47
Yes
69
17
0.46
Yes
68
16
0.47
Yes
63
15
0.53
Yes
64
15
0.52
Yes
64
15
0.53
Yes
0.191
0.298
(1.12)
1.6
0.634 0.683 0.615 0.65
0.583 0.615 0.718 0.663
0.518 0.741 0.7
0.592 0.638
(2.69)*** (2.24)** (2.86)*** (2.60)** (2.27)** (2.23)** (3.06)*** (2.69)*** (2.01)** (2.90)*** (2.45)** (2.23)** (2.26)**
0.283
(0.13)

Summary statistic
Number of countries
R2
Constant
Institutional reforms index (log, change)
Financial reform index (log, change)
Trade reform index (log, change)
have electoral implications in some political contexts. As in the previous

set of regressions, policies concerning property rights do not have significant effects on the behavior of voters. The regressions that include the three
summary indexes confirm that none of them is significant. The two final
regressions indicate that the real exchange rate index remains weakly significant when other policy variables are included. Therefore, evidence on the consequences of the Washington Consensus policies in legislative elections is not
robust. Somewhat surprisingly, the policy indicator that turns out to be
more robust is outside the direct influence of the legislature.
The main conclusion that emerges from the empirical evidence presented so far is that the electorate is in favor of some economic outcomes
as well as some economic policies. Inflation and the advancement of some
promarket reforms are key reasons for withdrawing support from the incumbents party in presidential elections. (For legislative elections, the evidence so far is very scant, regarding both outcomes and policies). It is very
unlikely that the negative payoffs of promarket reforms in presidential
elections would be countered by their positive effects on inflation, growth,
or other economic or social outcomes, because the electorate does not
seem to be very sensitive to these variables. It is only fair to conclude that
the electorate dislikes promarket policies, irrespective of their results.
However, these conclusions require some additional testing, because the
response of the electorate may depend on political, institutional, and economic circumstances, as stated in hypothesis 4.
Does Context Matter?

The sensitivity of the electorate to Washington Consensus policies may be
influenced by a host of contextual variables, such as the ideology of the
incumbents party, the incumbents promises during the election campaign, and whether the economy was in crisis at the time of the previous
elections.61 As mentioned above, to treat this hypothesis we endogenize
the coefficient as follows:
= + PROMISES IDEOLOGY + CRISIS.
Replacing in the previous equations gives (with fixed effects)
61. Crisis is measured as the (log) distance between GDP and its trend when GDP is
below its trend, and 0 otherwise.
41
d log (Vt = + log ( Ft 1 + d log ( X t + d log ( Pt + PROMISES
d log ( Pt + CRISIS d log ( Pt + C + t .

Tables 13 and 14 present the relevant results from this specification.62
The context in which reforms take place does not seem to affect the electorates sensitivity to those reforms. As shown in table 13, the only exception occurs when the tax reform index is interacted with our switch index,
measured with respect to promises.63 The negative coefficient in regression
3 implies that the adoption of measures that make the tax system more neutral and effective leads to vote gains when the incumbent has campaigned
on the adoption of promarket policies, but brings losses when the incumbent has argued against them on campaign but switched once in power. In
legislative elections, contextual factors seem to play an important role for
privatizations. In regression 1 of table 14 the coefficient for the privatizations variable is negative and significant and the coefficient for the interaction term (IDEOLOGY*PRIVATIZATIONS) is positive and significant.
The values of the coefficients suggest that while privatizations do carry
electoral costs, these are reduced by about a third when the largest party in
the legislature is market oriented. Regression 4 includes two interaction
terms found significant in previous regressions, namely, (IDEOLOGY*
PRIVATIZATIONS) and (GROWTH*POLARIZATION). It finds that both
remain strongly significant. These results confirm the importance of ideology in legislative elections. It is revealing that the influence of ideology is
detected in connection with privatizations, because this is the area of reform
in which the legislature plays the most important role and on which public
opinion is strongest.
Conclusion
This paper has assessed the electoral consequences of Washington Consensus policies in Latin America on the basis of testable hypotheses derived
from econometric and case studies on the subject. The results lend qualified support for our main four hypotheses, as follows.
62. A more complete set of results is available from the authors upon request.
63. The switch indexes are defined above. We also tested a switch index measured with
respect to the ideology of the party, and those same indexes in absolute values (which measure whether the incumbent has lied or not, regardless of the direction of the switch). None
of these alternative measures was found to be significant.
T A B L E 1 3 . Impact of Washington Consensus Policies Interacted with Contextual
Features in Presidential Elections, 19852002: Country Fixed Effects Resultsa
Dependent variable: Change in log of vote
share of incumbent presidents party
Political
Economic outcomes
Promises*tax reforms index
(1)
(2)
(3)
0.377
(0.65)
1.124
(1.33)
0.000
(0.00)
0.762
(1.70)
0.156
(0.30)
0.317
(0.91)
0.192
(0.38)
1.143
(1.34)
0.000
(0.00)
0.808
(1.75)*
0.063
(0.13)
0.382
(1.19)
0.696
(1.57)
0.524
(0.49)
0.517
(1.34)
0.897
(1.17)
0.701
(1.85)*
0.319
(0.37)
0.622
(1.50)
0.909
(1.18)
1.564
(1.92)*
1.424
(1.62)
0.283
(0.36)
0.394
(1.65)
0.061
(0.11)
0.154
(0.65)
Ideology*tax reforms index
1.658
(2.72)**
Promises switch index*tax reforms index
0.190
(0.32)
1.057
(2.27)**
0.371
(0.67)
6.489
(0.48)
1.115
(2.31)**
27
14
0.67
Yes
38
17
0.64
Yes
26
14
0.72
Yes
38
17
0.64
Yes
Crisis*tax reforms index

Constant
Summary statistic
Number of countries
R2
(4)

1. Electoral support for the incumbents party is higher, the better the
aggregate economic outcomes during his or her administration. The incumbents party is rewarded in presidential elections for reductions in the rate
of inflation; and in legislative elections, for increases in the rate of growth
(although the latter result is not robust in this first hypothesis). Neither
changes in unemployment nor changes in income distribution appear to
influence voters behavior.
43
T A B L E 1 4 . Impact of Washington Consensus Policies Interacted with Contextual

Features in Legislative Elections, 19852002: Country Fixed Effects Resultsa
Dependent variable: Change in log of vote
share of party with most seats in legislature
Political
Economic outcomes
Ideology*privatization reforms index
(1)
(2)
(3)
(4)
0.568
(2.80)***
0.045
(0.16)
0.104
(0.79)
0.657
(3.36)***
0.185
0.7
0.085
0.62
0.658
(3.25)***
0.098
(0.35)
0.042
(0.28)
0.59
(2.90)***
0.003
0.01
0.018
0.12
0.278
(1.19)
1.168
(1.99)*
0.089
0.4
1.198
(2.05)**
0.251
(0.94)
0.638
(0.94)
0.291
1.32
0.539
0.64
2.076
(2.97)***
0.680
(2.45)**
0.333
0.79
0.583
(1.61)
2.068
(3.01)***
0.67
(2.48)**
0.046
0.19
Promises switch index*privatizations index

Crisis*privatization reforms index
15.173
(1.61)
Growth*polarization
Constant
Summary statistic
Number of countries
R2
0.525
(2.04)**
0.684
(2.34)**
0.760
(2.90)***
3.546
(3.12)***
0.626
(2.43)**
67
14
0.54
Yes
67
17
0.5
Yes
69
14
0.5
Yes
67
17
0.58
Yes

2. The sensitivity of electoral support to economic outcomes depends

on the institutional characteristics of the political regime and the party system. We find that in presidential elections, the more fragmented the party
system, the higher the payoff of inflation rate decreases. There is also some
evidence that in presidential elections a divided government increases the
payoff of the inflation rate decreases or increases in the rate of economic
growth (however, this finding is based on a very small sample). In legislative
elections, there is strong evidence that party polarization enhances the

electoral payoff of higher rates of growth.
3. Electoral support for the incumbents party depends on the economic policies adopted. Policies may carry electoral costs even when they
deliver good economic outcomes. We find strong evidence that voters care
not only about economic outcomes but also, in some cases, about policies.
While the electorate seems to be blind to macroeconomic policies, it is
antagonistic to promarket policies beyond their effects on growth or inflation. Promarket reforms in general carry very large electoral costs for the
incumbents party in presidential elections. If the context of these reforms
is not taken into consideration, the evidence of adverse payoffs in legislative elections is weak.
4. The electorates tolerance of unpopular policies depends on the ideology of the incumbents party, his or her campaign statements, and the
initial state of the economy. Ideology does influence the reaction of the
electorate in legislative elections, according to our results. While the electorate dislikes privatization measures, it is more tolerant of them when the
largest party in the legislature has a promarket ideology. In presidential
elections, some evidence suggests that the electorate punishes the incumbent for the adoption of tax reforms when they run counter to his or her
campaign statements.
In synthesis, adopting the Washington Consensus was costly to the
reformers, although these costs were mitigated in some circumstances.
The parties in power were able to harvest juicy electoral dividends only
when the government pursued ambitious stabilization policies in highinflation economies. These findings seem to fit well the salient facts of the
last two decades, where a few incumbents were favored by the electorate
for their success in taming inflation, but little electoral recognition was
given to those who advanced the other macroeconomic and structural policies deemed necessary to accelerate growth and ensure stability. It might
be tempting to conclude that the days of economic orthodoxy are numbered.
However, it is unclear that reversing the reforms will produce electoral benefits. To date, the experience of reversals is limited to a few countries, and
it is too soon to assess their political payoffs.
The strongest conclusion of this paperthat promarket reforms carry
large electoral costs, irrespective of their macroeconomic effectsmay
not surprise political scientists, but it certainly will surprise many economists: why should the electorate reject policies that improve aggregate
economic outcomes and welfare? Although this paper does not address
45
this question, some results (not reported) suggest that many of the simpler
hypotheses proposed to answer it are at best incomplete. It has been widely
argued that such rejection is due to the social and distributional effects of
the reforms, but we have not found any evidence either that voting decisions are directly affected by social or distributional outcomes or that the
electorates response to the reforms is influenced by them.64 It has also
been argued that frustration with the reforms is due to their weak economic
impact in countries that lack the institutional support needed to harvest the
benefits of market liberalization.65 Again, we find no evidence in support
of this view. Relatedly, several authors have suggested that opposition to
promarket policies is stronger when those who make the liberalization
decisions or benefit from them are perceived to be corrupt.66 However, we
do not find that any measure of perceived corruption helps explain the
electorates response to the adoption of promarket reforms.67
Many other hypotheses beyond those that we have been able to test
are possible. Based on psychological theory and experimentation, Sergio
Pernice and Federico Sturzenegger have argued that public opposition to
a successful reform process can be explained by universal cognitive
biasesconfirmatory bias and self-serving biasif the principles of the
reform are at odds with their beliefs and self-serving view of the world.68
And Sanjay Jain and Sharun Mukand have developed a theoretical model
to explain why successful reforms may run aground: if the reform process
tilts the political balance in a way that makes the redistribution of the benefits less likely, public opinion may turn against the continuation of the
reform process.69 Why Latin Americans reject promarket reforms at the
polls remains an open question.
64. For a summary of such arguments, see Lora and Panizza (2002); and Lora, Panizza
and Quispe-Agnoli (2004).
65. See Lora and Panizza (2002), on the basis of public opinion data.
66. Di Tella and MacCulloch (2004) have uncovered empirical evidence consistent
with this hypothesis.
67. For instance, when we include the interaction between our measure of reform and
a measure of control of corruption (taken from International Country Risk Guide, 2004;
www.icrgonline.com) in the basic regression from table 5, the coefficient is positive but not
significant.
68. Pernice and Sturzenegger (2003).
69. Jain and Mukand (2003).
Comments
Sebastin Galiani: This paper by Lora and Olivera is very interesting.
It investigates whether the Latin American governments that adopted
market-oriented reforms during the late 1980s and early 1990s were rewarded with votes by the electorate. Certainly, this is an interesting and
much-debated question. It is also a very hard one. There are many
sources of complexity.
A virtue of the paper is that the authors keep the analysis at an explorative level. Identifying the impact of the reforms implemented during the
1990s on the voting behavior of the electorate is an intrinsically delicate
matter, since there is no good way to assess how the same electorate would
have voted had the reforms not been implemented (in precisely the way
they were). Another serious difficulty is the lack of a more detailed theory
to help disentangle the effects under consideration from the data.
The authors rely on cross-country panel data to attempt to control for factors that could potentially impact the effects they are interested in. Clearly,
this is the best strategy for the data sets they use. They analyze both presidential and legislative elections, with separate empirical models.
A first question, which is very important for their analysis, is whether
the electorate cares about policies per se, or only in relation to the effects
they have on outcomes. This would determine whether the effect of economic reforms on voting behavior is a structural parameter or not. The
authors assume that voters do care about policies per se, and that they vote
on the basis of policies in the recent past.
Of course, the reforms of the 1990s were not implemented in a vacuum.
They could be expected to affect variables such as growth and inflation
and were also likely to hit unemployment, although transitorily. All of
these variables are likely to influence voters, and thus should be included
as controls in the regression models. The authors do so.1 What is more,
1. They also include other time-varying political control variables. The authors report
that the results are robust to the inclusion of year effects, but it is not clear why they do not
just report those results instead, since they encompass the ones presented here.
46
47
they also rightly compute the total effect (in addition to the direct effect)
of the reforms on voting, using ancillary models that estimate the effects
of the promarket reforms on these macroeconomic outcomes.
In presidential elections, the authors find that the incumbents party is
rewarded for reducing inflation, but once this is controlled for, the electorate seems to oppose market-oriented reforms.2 Even when the total
effect of these reforms is computed, it appears that reforming parties paid
a price for adopting them. Given the relative importance of inflation among
the explanatory variables, it would have been useful to check extensively
the robustness of this result. Unfortunately, there are episodes of very high
inflation that might be driving the results.
In any event, one question naturally arises: why did so many governments adopt promarket reforms? A common argument today is that structural reforms were supposed to deliver more growth than they did. However,
even if this were the case, growth is found to have little effect on presidential elections. Thus, if one sticks to the estimated model, one needs to look
elsewhere for the answer. Perhaps these reforms were also an essential component of the stabilization programs adopted by the countries that reduced
inflation during this period, in a way not captured by the ancillary models
reported in the paper? This seems to be the case for Argentina.3
Another important issue is how the reforms affect voting, once the main
economic outcomes are controlled for. It is true that voting may well reflect
the taste of the electorate for these policies. But it could also capture other
things. It is likely to depend on the way these policies were implemented.
It might perhaps capture distributive effects. For example, voters might not
have fixed ideas about trade liberalization in general. Maybe, in deciding
how to vote, an individual only considers his own economic situation,
which could have been affected by this particular policy reform and is also
affected by the macroeconomic performance of his country.
Consider privatization. Firms improved substantially after privatization.
And consumers, in general, also benefited.4 But not everyone gained: displaced workers lost earnings and employment security, even in the long run.5
2. They find that the results for legislative elections are different. This is troublesome,
because there is no theory to help interpret it. Why should one expect inflation to be more
influential in presidential elections, for example, and growth to be more influential in legislative elections?
3. Galiani, Heymann, and Tommasi (2003).
4. See, for example, Galiani and others (2005); Galiani, Gertler, and Schargrodsky (2005).
5. See, for example, Galiani and Sturzenegger (2005).
Although the average benefits outweigh the costs, privatization might have
a small positive effect on the welfare of each consumer but a large cost
among the small group of displaced workers. The former might not decide
how to vote on the basis of the outcome of privatization, but the latter would
certainly do so. This is just a speculative counterexample. These issues need
to be explored further. A promising research strategy would be to exploit
panel data on voting at the smallest electoral district level, where the proportions of potential winners and losers from each reform can be identified
using census data.
Ernesto Dal B: The 1990s saw several governments in Latin America
implement various combinations of the reforms commonly associated with
the Washington Consensus. One important question concerns the effects of
these reforms. A vast body of work in several different applied literatures
has dealt with, for example, the impact of trade liberalization, deregulation, and privatization. As the reform process seems to be losing momentum across the continent, a second important question has emerged: what
are the constraints on the reforms sustainability?
This paper by Lora and Olivera relates to the second, which is a crucial
question if one believes that market-friendly reforms are, under some guise,
beneficial. In a democratic government, reforms must appeal to the electorate, or a large enough portion of it. This, in turn, is a precondition for
politicians to have a stake in implementing such reforms. Lora and Oliveras
central claim is that apart from macroeconomic stabilization programs, the
effect of the Washington Consensus policies tends to be to drive votes away:
privatization, tax reform, trade and financial liberalization generally cost
votes.1 If this is true and more reforms are desirable, the challenge appears
substantial: why expect political entrepreneurs to provide policies that lose
votes? This question is crucial because we would like both to see policies
that can bring Latin America prosperity and to preserve democracy.
The authors study variations in the vote shares of the party associated
with the incumbent president or the dominant party in the legislature.
1. Williamson (1990b) includes ten priorities in his synthesis on what Washington
means by policy reform: fiscal discipline; tax reform; the liberalization of interest rates,
foreign investment, and trade; a competitive exchange rate; privatization; deregulation; the
redirection of public expenditure toward social areas such as health, education, and infrastructure; and the protection of property rights. In practice, some reforms were emphasized
over others. As I argue below, there are reasons to believe this bias may have had important
consequences.
49
The explanatory factors are economic outcomes (such as inflation and

growth), political controls (such as political polarization or party fragmentation), and policies in place (such as privatizations, trade opening,
macroeconomic stabilization, or a composite index of structural reforms).
Coefficients on inflation tend to be negative and sometimes significant;
those on growth tend to be quite unstable, but appear positive and significant in specifications related to the party dominating the legislature;
and those on policies associated with the Washington Consensus tend to
be negative. For example, the indexes of both structural reforms and trade
liberalization appear to lose votes to presidents (see table 11), while privatizations tend to cost dominant legislative parties votes (see table 14).
The authors interpret these results as evidence that macroeconomic stabilization programs that reduce inflation pay off quite handsomely in terms
of votes and growth tends to pay off as well, while policies associated
with the Washington Consensus are, in themselves, vote losers. The fact
that these policy variables have negative coefficients in the presence of
economic outcome variables is interpreted as evidence that the public
has an intrinsic negative attitude toward such policies. In other words,
voters disapprove of market-friendly reforms per se, regardless of their
impact on economic performance. This is the second important claim in
the paper. Calculations are provided to show that although reforms may
yield votes by virtue of their beneficial effects on economic performance,
they lose votes overall because Latin Americans dislike market-friendly
reforms per se.
Two Observations
The problems of estimating the impact of a policy (and hence how it will
affect voters opinions) are well known. Here, I argue that the potential
benefits of policy reform might be underestimated, in terms of both their
economic and their electoral impacts. For instance, consider a few countries set on various growth trajectories. Assume now that, foreseeing collision, policy reform is implemented more aggressively when countries
get to very bad stages of their growth trajectoriesthat is, reforms are
endogenous. It may be possible for the reforms to improve growth outcomes relative to what the countries would have experienced without the
reforms, while estimates indicate that reforms are associated with less,
rather than more, growth. The key problem is that we lack the counterfactual trajectory for each country, corresponding to less or no reform. When
estimates of the reforms impact on economic outcomes are used in com-
bination with estimates of the outcomes impact on votes, the electoral

payoff from the reforms will tend to be underestimated.
A second possible problem could arise when reformist governments are
elected right after an acute crisis. Suppose, now, that the old administration, which is seen to be responsible for the crisis, is heavily punished at
the polls. This is likely to get the reformer-to-be government elected with
a very large share of votes. Once reforms have been put in place and the
crisis resolved, vote shares may well regress to the mean corresponding to
normal times. Successful reformers might then be reelected because voters approve of their work, but their reelection vote shares may be lower
than when they were first elected. In this case, estimates will show reforms
as destroying votes even when they really do not; in the absence of the
reforms, the incumbent reformers would have been hit even harder at the
polls. True, inflation reduction appears to continue to earn votes, according to the authors results. Thus if the bias I mention is affecting both the
reform and the outcome coefficients, it is not enough to alter the sign of
the inflation coefficient. However, it is still possible that the electoral
impacts of both reforms and inflation are biased downward.
The Specification
The basic specification driving the analysis in this paper is reported in column 1 of table 5 and in table 11. This specification includes both economic
outcomes and policy indexes as explanatory variables of variation in vote
shares. One problem with this specification is that it could be unfair to the
policies. After all, if the policies are improving economic outcomes and
these outcomes are included as regressors, the coefficient on the policy variable will never capture some of the policys electoral rewards. These will be
attributed to the improved economic outcomes. Also, to the extent that the
policies affect the economic outcomes, the regressors are not independent.
A better approach might be to estimate the impact of policies on all relevant outcomes, and then estimate the impact of those outcomes on votes.
A crucial question, of course, is how to identify all the relevant outcomes
that the policies might affect. Moreover, this strategy does not attempt to
determine whether voters disapprove of reforms for reasons other than
their economic consequences. The authors are highly concerned about this
second issue, and the inclusion of policy variables next to economic outcome variables is meant to capture any intrinsic value that the public may
place on policies.
51
To get around the problem of evaluating the electoral impact of policies

in a specification that includes outcomes, the authors attempt to separate
the effect of policies due to the intrinsic value voters may perceive from
the effect due to their impact on economic outcomes. In order to do this,
the authors regress both growth and inflation on the index of structural
reforms, which yields significant estimates of how much reforms reduced
inflation and fostered growth (reported in table 6); call this effect A. Given
that improvements in inflation and growth are estimated to yield votes
(column 1 of table 5)call this effect Bthe authors compute the electoral gains from reform through improved performance by multiplying
effects A and B. Then they add the resulting (positive) estimate to the
negative coefficient on policies themselves in the basic regression from
column 1 of table 5. They conclude that the net effect that Washington
Consensus policies have on votes is negative: the direct intrinsic value
effect is larger than the outcome improvement effect.
On Interpretation
The authors emphasize that the negative coefficient on the structural reforms
index represents an intrinsic distaste on the part of voters for market-oriented
policies. One could argue against this interpretation. The reason relates
to the question of defining the complete set of relevant outcomes. To be
concrete, suppose that reforms reduce inflation and improve growth but
also severely damage the environment, causing air quality to drop to
unhealthy levels. Now, suppose one runs the authors regression with
the change in the share of votes as the dependent variable and inflation,
growth, and a reform policy index as independent ones, without including an air quality variable. If people punish the government for the
decline in air quality, using the authors interpretation for this regression
one would conclude that the public dislikes market-friendly reforms. In
this case, that conclusion is wrong. People may not care in the least about
policy labels but care strongly about air quality, which has been affected
by the policies.
Therefore, I believe that the conclusion that Latin American people dislike
market-friendly reforms per se is premature. A much larger set of controls for
various relevant outcomes should be explored in the future. I discuss below
some possible reasons why voters may be unhappy with reforms irrespective
of their ideological profile. One potential candidate is the deterioration in the
social atmosphere that some countries have experienced. As reforms went
deeper in Argentina, for example, there were alarming increases in crimes
against property.2 A relatively safe society just a few decades ago, its social
air has now become a lot less easy to breathe; even middle-class people now
discuss defensive tactics against express kidnapping.
Why Are Latin Americans Unhappy with Reforms?

For the reasons explained above, I view the evidence that reforms cost votes
as indicative rather than definitive. In particular, I am skeptical that reforms
may cost votes because of their labels. Even if they do, I would still argue that
what is now being interpreted as ideological motivation may be capturing
other concerns. However, when these findings are considered next to the available survey evidence indicating that Latin Americans are dissatisfied with
some structural reformsespecially privatizationa rather negative picture
emerges.3 This is puzzling. The reforms are backed by reasonable theory, and
moreover, several of them have had good effects. Most economists would
have expected the electorate to endorse the reform process strongly and visibly. Why do we find ourselves analyzing evidence that seems to point the
other way?
As the profession expected, a number of reforms have had measurable
benefits. I will not attempt a systematic overview of the impact of structural reforms here. An arbitrary selection of the literature suggests that
even privatizations, which are strongly opposed in opinion polls, have had
a number of benefits. Some such studies have appeared in this journal.
David McKenzie and Dilip Mookherjee, for example, review evidence on
the effects of privatizations from studies applying to four Latin American
countries: Argentina, Bolivia, Mexico, and Nicaragua.4 The emerging picture shows improvements in labor efficiency and access to services, without generalized negative effects on the income distribution or poverty.
There is evidence that privatizations have increased firm efficiency.5
Sebastian Galiani, Paul Gertler and Ernesto Schargrodsky show that in
Argentina the privatization of water led to improvements in service and
2. Between 1991 and 1998, the years during which most reforms were introduced
and per capita GDP increased by about 40 percent, crimes against property rose by around
71 percent.
3. See the polls conducted by Latinobarmetro. See also Lora, Panizza and QuispeAgnoli (2004) on reform fatigue.
4. McKenzie and Mookherjee (2003).
5. See, for instance, Boubakri and Cosset (1998) for a study of developing countries that
includes some Latin American cases.
53
thus to reductions in child mortality.6 The privatization of power generation in Argentina led to a large increase in generating capacity, which in
turn contributed to lower energy prices. The privatization of energy distribution, in particular, when coupled with regulation through price caps, had
beneficial effects on productivity.7
Given the need for electoral support in order for the reform agenda to
progress, it seems important to ask why, if reforms have a number of benefits, one does not see stronger support for them. There are several possible reasons.
One answer is given by Lora and Olivera: voters are strongly ideological. There is little reformers can do about this. If ideology is a fixed taste
that goes against economic convenience, then maximizing social happiness may require abstaining from further reforms.
It is possible that reforms constitute an investment, and as such entail costs
during the first few years, whereas benefits will tend to accrue later on. If
voters are not aware of this temporal pattern, they may be quick to punish
reformers when experiencing the costs. Voters may not realize that judged in
terms of net present value contribution, the reformers deserve to be rewarded.
Moreover, reforms may create very diffuse benefits and highly concentrated costs, and voters may react more strongly in the face of large variations in payoffs. For example, privatizations tend to increase consumer
surplus but generate layoffs. Suppose that one such reform creates a small
benefit for 80 percent of the population, so that this fraction of the population is now likely to reelect the reformer with a 60 percent chance when
he faces a challenger (assume no abstentions nor a third candidate). This
gives the reformer a 48 percent vote share from the winners. But the
remaining 20 percent of the population is badly hurt by the reform, so only
5 percent of this group will reelect the reformer. Thus the overall vote
share for the reformer is only 49 percent of the vote, which is insufficient
to obtain reelection even though a large majority gained from the reform.
Economists can get biased estimates when they lack the appropriate
counterfactual, and so can voters. Therefore, voters may not associate
reforms with improved economic outcomes, even when they have been
beneficial. This outcome is even more likely if voters compare their governments performance to that of neighboring nations, and it is the countries on the worst trajectories that implement more reforms.
6. Galiani, Gertler, and Schargrodsky (2005).
7. Estache and Rossi (2005).
It may be the case that reforms are made possible by one of two things
(or both): a large fiscal crisis or a substantial amount of corruption.8 Either
of these two factors could facilitate reform by softening the resistance of
vested interests. Traditional economics emphasizes the role of compensating transfers to facilitate Pareto-improving moves. These transfers can
be hard to implement openly. Corruption could then facilitate reform,
because it allows reformers to pass compensating transfers to losers or
players with veto power over the reform process. But if crisis and corruption facilitate reform, voters may associate reforms with these two traits,
which in themselves have a number of negative consequences that may
cost votes. In this case, it is not reform that costs votes. The electoral consequences of the facilitating factors are confused with those of reforms.9
Voters may also be motivated by spite and considerations of fairness.
Even when reforms create a substantial consumer surplus, individuals may
be negative about them as a result of the fortunes that some key players
have made in the process.
Another interpretation is that voters are cruel pragmatists. They elect
leaders who they think will be pragmatic enough to effect the transfers that would make reform happen. Examples of successful reformers
who were later accused of corruption are Fujimori in Peru and Menem in
Argentina. Once these leaders have delivered the reforms, why would
rational voters keep them around? The social benefits from their (alleged)
corruption have been realized, and there would be only costs to be reaped
further down the line.
Voters assessments of reform may be affected by the economic cycle.
The debate on the negative opinions of reform is relatively new. As such, it
may be marked by the fact that since about 1998 a few Latin American
countries have encountered new macroeconomic problems and less abundant foreign capital. It would be interesting to see if opinions remain negative during a future wave of capital inflows and macroeconomic recovery.
It is possible that voters may associate reforms (rightly or spuriously)
with a number of outcomes that economists have not paid enough attention
to when discussing policy reform. Here, I discuss only one such possibility:
reforms may be correlated with a deterioration in the social atmosphere.
8. On fiscal crisis as a catalyst, see, for instance, Drazen and Grilli (1993).
9. In their conclusion, Lora and Olivera note in passing that they did not find any connection between perceptions of corruption and electoral results. It would be worthwhile to
explore this connection further, under alternative specifications.
55
By this I mean that while reforms were being implemented, perceptions of

security, social cohesion, and trust in others may have suffered.
The Social Constraint to Policy Analysis

I begin with a simple conjecture: policy reform indexes may be positively
correlated with data on crimes against property and kidnapping with extortion. (This may not be the fault of the reforms; the issue also requires
investigation.) These phenomena, like air quality in my earlier example,
are essentially missing in the specification used by Lora and Olivera. If my
conjecture is true, it may imply that voters do not care about reforms from
an ideological standpoint, but associate them with decreased security and
enjoyment of public spaces and social life.
Note that unemployment figures, inequality, or simple poverty measures may not capture all the relevant variation when it comes to explaining how social issues affect electoral results. Indicators of marginality, the
importance of the shadow economy, the average spell of unemployment,
and the concentration of longer spells at certain skill levels may be more
telling. In this connection, a pertinent question is to what extent reforms
change not the levels, but the profiles of employment.10 For instance, if
reform pushes some workers to a different sector, it may impose on them
a large loss of industry-specific human capital. Some of these workers may
decide to drop out of the labor force and not even look for a job. Measured
unemployment will remain the same, while marginality increases.
Although the original body of Washington Consensus recommendations did include elements such as public expenditures in social areas and
the protection of property rights, these do not seem to have been the main
focus of the reform process. There are various possible reasons for this.
The most energetic reforms may have been those fueled by business interests. Also, it may have been easier to pursue reforms that entailed reducing
the scope of public intervention than those that would make intervention
more sophisticated. Further, in the absence of clear theoretical predictions
of how reforms may impact welfare when property rights are not perfectly
enforceable, the protection of property may not have been seen as a vital
component of the reform agenda.
10. Galiani and Hopenhayn (2003) show that after a decade of reforms in Argentina,
taking into account reincidence, the risk of unemployment rose significantly and was distributed unequally in the labor force.
A reassessment of the implementation constraints to policy reform could

be productive. Political economics has done a lot to clarify how political
constraints may shape the process of endogenous policy formation. An
analogous analysis of how social constraints may limit the set of desirable
and implementable policies is still lacking.
My study with Pedro Dal B takes a step in this direction, arguing that
social conflict may impose a social constraint on policy analysis.11 The
reason is that in a world where property rights cannot be perfectly protected, certain policies could unleash a social backlash of forced redistribution. We examine a simple, small open economy model in the presence
of one critical distortion: individuals may engage in appropriation activities; for example, they may become members of a guerrilla organization
that extorts payments from nonmembers, or simply become criminals. The
size of this appropriation sector is governed by a balance between the
opportunity costs and the returns to appropriation activities.
The key assumption in the model is that appropriation activities are
more labor intensive than the overall economy. (The model uses the labels
capital and labor, but these could be interpreted to mean, respectively,
skilled labor and unskilled labor; we would then say that appropriation is
relatively intensive in unskilled labor.) Shocks or policies that hurt laborintensive activities (or that favor capital-intensive ones) drive wages down
relative to the mass or disputable wealth in the economy. This lowers the
main opportunity cost of appropriation activities relative to their returns,
so appropriation expands. An example is trade liberalization that increases
the perceived price of capital-intensive goods relative to labor-intensive
goods.12 Similar effects could follow from deregulation and foreign investment that increase productivity in capital-intensive industries more than in
labor-intensive ones. These forces clearly have the potential to generate
efficiency gains and make society wealthier. However, by hurting the relative value of labor in a world where appropriation is relatively labor inten-
11. Dal B and Dal B (2004).

12. Under the skilled versus unskilled labor interpretation, this increases the wage gap
across skills, increasing inequality, and it drives appropriation activities up. Thus the model
can account for simultaneous increases in GDP per capita, inequality, skill premiums, and
crime, as experienced by Argentina in the 1990s. One prediction of the model is that not all
forms of inequality will affect crime, so simple measures of inequality may not always be
neatly associated with crime. Appropriation activities are only affected by forms of inequality that affect the structure of remuneration across factors (or skill levels). Otherwise, the
balance between the costs and returns to appropriation may remain unaltered.
57
sive, such forces may unleash an appropriation backlash that leaves all
agents in the economy worse off.
We tend to consider a number of reforms convenient because we expect
their efficiency gains will make societies more prosperous. However,
when the protection of property rights is imperfect, we are in a secondbest world. The welfare implications of reforms may then depend on new
considerations, and the set of reforms that are desirable and sustainable
will change. It may pay to reassess the Washington Consensus from this
perspective.
Conclusion
Lora and Olivera have produced an interesting, thought-provoking piece.
The problem they tackle belongs to a set of complicated issues on which
it is hard to derive definitive conclusions. More work is needed. Their evidence suggests that market-friendly reforms may not have produced handsome electoral payoffs beyond those secured in the macroeconomic
stabilization phase. This is consistent with survey evidence on public attitudes toward reforms and with the sentiment one perceives in casual interactions on the continent.
One challenge that lies ahead is to reconcile the evidence of important
benefits from some reformsnotably privatizationswith the fact that
many Latin Americans seem to reject the idea that they have benefited
from the reforms. I have here proposed a few potential approaches,
including (1) examining how opinions on reforms are affected by factors
ranging from the economic cycle to the perceived fairness of their implementation, (2) analyzing how corruptiona much vilified traitmay
have been instrumental in pushing reforms forward, and (3) investigating
the role of a social constraint on the implementation of reforms. In particular, a reassessment of the reform process in terms of its impact on social
peace and criminal activities would be profitable. This should entail more
reflection on what constitutes desirable and sustainable policy in a world
where the protection of life and property can only be imperfect.
References
Alesina, Alberto, and Howard Rosenthal. 1995. Partisan Politics, Divided Government, and the Economy. Cambridge University Press.
. 1996. A Theory of Divided Government. Econometrica 64(6): 131141.
Alesina, Alberto, John Londregan, and Howard Rosenthal. 1993. A Model of
the Political Economy of the United States. American Political Science
Review 87(1): 1233.
Arcelus, Francisco, and Allan H. Meltzer. 1975. The Effect of Aggregate Economic Variables on Congressional Elections. American Political Science
Review 69(4): 123239.
Boubakri, Narjess, and Jean Claude Cosset. 1998. The Financial and Operating
Performance of Newly Privatized Firms: Evidence from Developing Countries. Journal of Finance 53(3): 1081110.
Coppedge, Michael. 1997. A Classification of Latin American Political Parties.
Working Paper 244. University of Notre Dame (November).
. 2001. Political Darwinism in Latin Americas Lost Decade. In Political Parties and Democracy, edited by Larry Diamond and Richard Gunther.
John Hopkins University Press.
Corrales, Javier. 2002. Presidents without Parties: The Politics of Economic
Reform in Argentina and Venezuela in the 1990s. Pennsylvania State University Press.
Crisp, Brian F., and Gregg B. Johnson. 2003. Mandates, Powers and Policies.
American Journal of Political Science 47(1): 12842.
Cukierman, Alex, and Mariano Tommasi. 1998. When Does It Take a Nixon to
Go to China? American Economic Review 88(1): 18097.
Dal B, Ernesto, and Pedro Dal B. 2004. Workers, Warriors, and Criminals:
Social Conflict in General Equilibrium. University of California, Berkeley,
mimeo.
Deininger Klaus, and Lyn Squire. 1998. New Ways of Looking at Old Issues:
Inequality and Growth. Journal of Development Economics 57(2): 25987.
Di Tella, Rafael, and Robert MacCulloch. 2004. Why Doesnt Capitalism Flow
to Poor Countries? Harvard Business School and Princeton University,
mimeo.
Downs, Anthony. 1957. An Economic Theory of Democracy. New York: Harper.
Drazen, Allan, and Vittorio Grilli. 1993. The Benefit of Crises for Economic
Reforms. American Economic Review 83(3): 598607.
Duch, Raymond M., and Randy Stevenson. 2004. The Economy: Do They Get It
Right? University of Houston and Rice University, mimeo (April).
Economic Commission for Latin America and the Caribbean (ECLAC). Various
years. Panorama preliminar de las economas de Amrica Latina y El Caribe.
Santiago, Chile: United Nations.
59
Estache, Antonio, and Martn Rossi. 2005. Do Regulation and Ownership Drive
the Efficiency of Electricity Distribution? Evidence from Latin America. Economics Letters 86(2): 25357.
Fiorina, Morris P. 1992. An Era of Divided Government. Political Science
Quarterly 107(3): 387410.
Galiani, Sebastin, Paul Gertler, and Ernesto Schargrodsky. 2005. Water for
Life: The Impact of the Privatization of Water Supply on Child Mortality.
Journal of Political Economy 113(February): 83120.
Galiani, Sebastin, Daniel Heymann, and Mariano Tommasi. 2003. Great Expectations and Hard Times: The Argentine Convertibility Plan. Economia 3(2):
10960.
Galiani, Sebastin, and Hugo Hopenhayn. 2003. Duration and Risk of Unemployment in Argentina. Journal of Development Economics 71(1): 199212.
Galiani, Sebastin, and Federico Sturzenegger. 2005. Structural Reform and Displaced Workers Earnings. Universidad Torcuato Di Tella, mimeo.
Galiani, Sebastin, and others. 2005. The Benefits and Costs of Privatization in
Argentina: A Microeconomic Analysis. In Privatization in Latin America:
Myth and Realities, edited by A. Chong and F. Lopez-De-Silanes. Stanford
University Press (forthcoming).
Gervasoni, Carlos. 1997. La sostenibilidad electoral de los programas de estabilizacin y reforma estructural: los casos de Argentina y Peru. Universidad
Catlica Argentina, Universidad Torcuato Di Tella, mimeo.
Inter-American Development Bank. 2003. Good Jobs Wanted. Economic and
Social Progress in Latin America series. Washington.
Jain, Sanjay and Mukand, Sharun. 2003. Redistributive Promises and the Adoption of Economic Reform. American Economic Review 93(1): 25664.
Kahneman, Daniel, and Amos Tversky. 1979. Prospect Theory: An Analysis of
Decision under Risk. Econometrica 47(2): 26391.
Keech, W. 1995. Economic Politics: The Costs of Democracy. Cambridge University Press.
Kramer, G. H. 1971. Short-term Fluctuations in U.S. Voting Behavior, 18961964.
American Political Science Review 65(March): 13143.
Kuczynski, Pedro-Pablo, and John Williamson, eds. 2003. After the Washington
Consensus. Restarting Growth and Reform in Latin America. Washington:
Institute for International Economics (March).
Lewis-Beck, Michael S. 1988. Economics and Elections. University of Michigan
Press.
Lora, Eduardo. 2001. Structural Reforms in Latin America: What Has Been
Reformed and How to Measure It. Research Department Working Paper 462.
Washington: Inter-American Development Bank.
Lora, Eduardo, and Ugo Panizza. 2002. Structural Reforms under Scrutiny.
Research Department Working Paper 470. Washington: Inter-American Development Bank.
Lora, Eduardo, Ugo Panizza, and Myriam Quispe-Agnoli. 2004. Reform Fatigue:
Symptoms, Reasons, Implications. Federal Reserve Bank of Atlanta Economic Review (2nd quarter): 128.
McKenzie, David, and Dilip Mookherjee. 2003. The Distributive Impact of Privatization in Latin America: An Overview of Evidence from Four Countries.
Economa 3(2): 161233.
Meltzer, Allan H., and Marc Vellrath. 1975. The Effect of Economic Policies on
Votes for the Presidency: Some Evidence for Recent Elections. Journal of
Law and Economics 18(3): 78189.
Nam, Moiss. 1994. Latin America: The Second Stage Reform. Journal of
Democracy 5(4): 3248.
Navia, Patricio, and Andrs Velasco. 2003. The Politics of Second Generation
Reforms. In After the Washington Consensus: Restarting Growth and Reform
in Latin America, edited by Pedro-Pablo Kuczynski and John Williamson.
Washington: Institute for International Economics (March).
Payne, J. Mark, and others. 2002. Democracies in Development: Politics and
Reform in Latin America. Washington: Inter-American Development Bank.
Pernice, Sergio, and Federico Sturzenegger. 2003. Culture and Social Resistence
to Reforms: A Theory about the Endogeneity of Public Beliefs with an Application to the Case of Argentina. CEMA and Universidad Torcuato Di Tella,
unpublished.
Persson, Torsten, Gerard Roland, and Guido Tabellini. 1997. Separation of Powers and Political Accountability. Quarterly Journal of Economics 112(4):
1163202.
Remmer, Karen L. 1991. The Political Impact of Economic Crisis in Latin America in the 1980s. American Political Science Review 85(3): 777800.
. 2003 Elections and Economics in Contemporary Latin America. In
Post-Stabilization Politics in Latin America, edited by Carol Wise and Riordan
Roett. Brookings.
Roberts, Kenneth M., and Erik Wibbels. 1999. Party System and Electoral
Volatility in Latin America: A Test of Economic, Institutional, and Structural
Explanations. American Political Science Review 99(3): 57590.
Stokes, Susan. 2001a. Mandates and Democracy. Neoliberalism by Surprise in
Latin America. Cambridge University Press.
. 2001b. Public Support for Market Reforms in New Democracies. Cambridge University Press.
Thaler, Richard H., and others. 1997. The Effect of Myopia and Loss Aversion
on Risk Taking: An Experimental Test. Quarterly Journal of Economics
112(2): 64761.
Tversky, Amos, and Daniel Kahneman. 1991. Loss Aversion in Riskless Choice:
A Reference-Dependent Model. Quarterly Journal of Economics 106(4):
103961.
61
Weyland, Kurt. 2002. The Politics of Market Reform in Fragile Democracies.

Argentina, Brazil, Peru and Venezuela. Princeton University Press.
Williamson, John. 1990a. Latin American Adjustment: How Much Has Happened.
Washington: Institute of International Economics.
. 1990b. What Washington Means by Policy Reform. In Latin American
Adjustment: How Much Has Happened? edited by J. Williamson. Washington:
Institute for International Economics.
Wolfers, Justin. 2002. Are Voters Rational? Evidence from Gubernatorial Elections. Harvard University, Department of Economics, mimeo (April).
ANTONI ESTEVADEORDAL
KATI SUOMINEN
Rules of Origin in Preferential

Trading Arrangements: Is All Well with
the Spaghetti Bowl in the Americas?
referential trading arrangements (PTAs) have proliferated spectacularly over the past decade around the world.1 The number of
PTAs in force soared from fifty in 1990 to some 230 by the end of
2004, and it is expected to rise to 300 in the course of 2005. Governing
more than a third of global trade, PTAs have sparked intense policy interest at the multilateral level. They are among the top priorities of the ongoing Doha Round of trade negotiations of the World Trade Organization
(WTO).2
The Western Hemisphere has been a major source for the expansion of
the worlds PTAs. The regions countries have signed some forty free
trade agreements with each other or with extrahemispheric parties since
1990. Mexico and Chile have been particularly prolific integrators: Mexico
has signed twelve PTAs and Chile seven.3 While the bulk of their agreements are with partners in the Americas, both countries have also integrated with European and Asian economies. For its part, the United States
has concluded four free trade agreements in the Americas and six with
Estevadeordal and Suominen are with the Inter-American Development Banks Integration
and Regional Programs Department, Integration, Trade, and Hemispheric Issues Division.
We are grateful to Andrs Rodrguez, Pablo Sanguinetti, Alberto Trejos, and Andrs
Velasco for outstanding comments.
1. PTAs include free trade agreements, customs unions, common markets, and single
markets.
2. The Doha Declaration states, We also agree to negotiations aimed at clarifying and
improving disciplines and procedures under the existing WTO provisions applying to
regional trade agreements. The negotiations shall take into account the developmental
aspects of regional trade agreements.
3. The figures refer to formal free trade agreements and exclude the economic complementation agreements.
63
nonhemispheric partners, and it is proceeding toward another six. In total,

the countries of the Americas are negotiating or opening negotiations for
more than two dozen new PTAs. Prominent ongoing initiatives include
the Free Trade Area of the Americas (FTAA) talks encompassing thirtyfour countries and the negations between the Southern Common Market
(Mercosur) and the European Union aimed at connecting the worlds two
largest customs unions.
The hemispheric PTA spree has forged a veritable spaghetti bowl of
multiple and often overlapping agreements (figure 1). The various rules
included in each PTAsuch as standards, safeguards, government procurement, and investmententangle the bowl further. While PTAs can
generate important economic benefits, the PTA spaghetti bowl carries two
risks. First, the manifold trade rules of PTAs can introduce policy frictions
that increase the costs of trading. Each new rule in each PTA represents a
new policy for firms to consider in their export, outsourcing, and investment decisions. Each also has legal, administrative, and economic implications for the PTA partner countries. Not all PTA rules necessarily work to
expand trade from its pre-PTA levels. Second, differences in rules across
PTAs can translate into transaction costs to countries dealing on two or more
F I G U R E 1 . The Americas and Trans-Pacific PTA Spaghetti Bowl
Hong Kong
EU, EFTA
FTAA
APEC
PR China
Brunei
Taiwan
Cambodia
Canada
Costa CACM
Rica El Salvador
Guatemala
Honduras
Nicaragua
Korea
Thailand
Malaysia
Philippines
Japan
USA
Panama
Mercosur
Paraguay
Argentina Brazil
Uruguay
Mexico
Myanmar
Bolivia
ASEAN
Singapore
New Zealand
Laos
Indonesia
Vietnam
Chile
Ecuador
Andean
Peru Community
Australia
Papua New Guinea

FTAA
APEC
Intra-Americas in force
Intra-Asia-Pacific in force
Intra-Asia-Pacific signed
Trans-Pacific signed
Colombia
Venezuela
Russia
Source:Devlin and Estevadeordal (2004).
Bahamas
Dominican
Republic
CARICOM
Dominica Trinidad & Tobago
Suriname Grenada Barbados
Jamaica St. Vincent & Grenadines
Guyana Antigua & Barbuda
St. Kitts & Nevis Belize
Haiti St. Lucia
Antoni Estevadeordal and Kati Suominen
65
PTA fronts simultaneously. This is a particular consideration in the Americas, where each country belongs to an average of four PTAs.4
Rules of origin are a key market access rule (or discipline, in the jargon
of trade negotiators) in PTAs. Rules of origin are the crucial gatekeepers
of commerce: a product shipped from an exporting PTA member must
meet the corresponding rule of origin to receive preferential treatment
from the importing member. Rules of origin epitomize the hemispheres
policy problem: a growing number of the regions PTAs carry complex
and restrictive rules of origin, and the many rules-of-origin regimes differ from each other. Consequently, the rules-of-origin spaghetti could
hold back the trade-creating potential of the hard-earned PTAs.
This paper presents an in-depth diagnosis of rules-of-origin regimes in
the Americas and offers policy recommendations for the region to
counter the potential negative effects of rules of origin. We hope to make
two contributions: to deepen understanding of the types and effects
of rules of origin used in Western Hemisphere PTAs, and to add rigor to
the policy debate on the implications of PTAs to the multilateral trading
system.5
The paper is organized in four parts. The first part surveys the state and
latest trends in the rules-of-origin regimes in the Americas. The next section summarizes the recent research on the political economy reasons
behind the choice of rules-of-origin instrument in PTAs. The third section
does the same for the economic effects of rules of origin, and discusses the
implications of the research findings to the hemispheres PTA spaghetti
bowl. The fourth part contains our policy recommendations, and a final
section concludes.
The Current Status of Rules-of-Origin Regimes in the Americas

Rules of origin can be divided into nonpreferential and preferential rules of
origin. Individual countries use nonpreferential rules of origin to distinguish
foreign from domestic products when applying other trade policy instruments,
4. The calculation includes continental Latin America, Canada, and the United States,
but not the countries of the Caribbean. The figures exclude partial scope agreements and
economic complementation agreements.
5. See Estevadeordal and Suominen (2005a), Suominen (2004), and WTO (2002b) for
further discussions on rules-of-origin regimes around the world.
such as antidumping and countervailing duties, safeguard measures, origin marking requirements, discriminatory quantitative restrictions or tariff quotas, and rules on government procurement. The WTO is in the final
stages of the decade-long process of harmonizing nonpreferential rules of
origin at the multilateral level.6 Preferential rules of origin, the focus of
this paper, are employed in PTAs and in the context of a generalized system
of preferences. They define the processes to be performed and inputs to be
incorporated in a product in the territory of an exporting PTA member
in order for the product to qualify for preferential access to an importing PTA member. The justification for preferential rules of origin is to
curb trade deflectionto prevent products originating from non-PTA
members from being transshipped through a low-tariff PTA partner to a
high-tariff one under the PTA-provided preferential treatment. Rules of
origin, in short, are tools for keeping non-PTA parties from free riding on
the PTA preferences. They are an inherent feature of free trade agreements in which the members external tariffs diverge or in which the
members wish to retain their individual tariff policies vis--vis the rest of
the world. Rules of origin are also used in aspiring customs unions to govern sectors for which the members have yet to establish a common external tariff.7
Rules of origin have become increasingly important over the past decade.
This is due both to the globalization of productionthe growth of international trade in goods manufactured in multiple countriesand to the
fact that todays PTAs quickly reduce the preferential tariff, the more traditional tool regulating preferential market access, to zero across most
product categories. Indeed, rules of origin are currently a key arbitrator of
the effectiveness of multilateral trade rules requiring PTAs to cover substantially all trade between the partner countries and not to raise barriers
vis--vis third parties.
Preferential rules of origin have thus far eluded multilateral regulation.
As a result, a wide repertoire of rules-of-origin types and combinations
has developed around the world. This section surveys the rules of origin
employed in the Western Hemispheres PTAs.
6. See Estevadeordal and Suominen (2005a) and Suominen (2004) for details on the
harmonization process.
7. Rules of origin are thus employed in the vast majority of PTAs. The Asia-Pacific
Economic Cooperation (APEC) forum is a prominent exception in that it operates on the
concept of open regionalism, with the preferential tariffs essentially being extended also to
nonmembers.
67
Product-Specific Rules of Origin:Toward Product-Specific Tailoring

Preferential rules of origin were a simple affair in most of the world until a
few years ago.8 In the Americas, trade agreements formed before the 1990s
generally put in place one, often vaguely defined, rule of origin applicable
to all products. Over time, the lack of precision in origin requirements
became much criticized for allowingand indeed requiringsubjective
case-by-case origin determinations.9 The growth in international trade and
the globalization of production expanded the constituency that paid attention to rules of origin, while paradoxically complicating correct judgments
on origin.
Perhaps the single most important event that raised the profile of rules of
origin in the Americas was the U.S. Customs finding in 1991 that the domestic content in Honda Civics imported from Canada fell below 50 percent
the threshold required for claiming origin under the U.S.-Canada free trade
agreement of 1989 and the preceding U.S.-Canada Auto Pact of 1965. The
finding fuelled the U.S. automotive industrys concerns about the intensifying Japanese competition and the moves by Japanese companies to use
Canada as a production base. For its part, Canada, which was concerned
about the loss of foreign investment, claimed that U.S. origin determinations were arbitrary at best. The Honda case had important repercussions
for the 1994 North American Free Trade Agreement (NAFTA) negotiations. Pressured by vehicle lobbies, NAFTA negotiators put in place a
highly precise regional value content of 62.5 percenta level that Japanese
firms were not expected to meet. The rules of origin lobbying spread
through other economic sectors. Fearing that Asian and European firms
would establish simple touch-up assembly operations in Mexico in order
to gain duty-free access to the North American markets, U.S. industries
called for tailor-made rules of origin that would be stringent enough to
keep extraregional parties out, yet lenient enough to allow U.S. multinationals to retain their extraregional outsourcing linkages. Mexico, in
turn, generally pushed for rules of origin that would not deter foreign
investment. This bargaining resulted in the 150-page NAFTA rules-oforigin protocol, which carries individualized rules of origin for some
5,000 different products.
8. The exception to the global pattern of general and vaguely defined rules of origin was
the European Community, which as early as the 1970s had preferential rules-of-origin
regimes with different rules of origin governing the various product categories.
9. Reyna (1995, p. 7).
Product-by-product rules-of-origin tailoring became the norm in PTA

talks across the hemisphere and around the world. The drive toward precise rules of origin was reflected by the 1999 multilateral Kyoto Convention, which established five main criteria for determining origin.10 The first
is the wholly obtained or produced criterion, which asks whether the commodities and related products have been entirely grown, harvested, or
extracted from the soil in the territory of the exporting PTA member or
manufactured in that member from any of these products. This rule is
met through not using any second-country components or materials. The
remaining four criteria are more complex and are packaged together under
the substantial transformation criterion. Of the four individual criteria, the
first involves a change in tariff classification in the territory of a PTA member between a product imported from an extra-PTA party and the product
that is being exported within the PTA. The change may be required to occur
at the level of chapter (two digits under the Harmonized Commodity
Description and Coding System), heading (four digits), subheading (six
digits), or item (eight to ten digits). The second criterion is an exception
attached to a change in tariff classification rule, which generally prohibits
the use of nonoriginating materials from a specific subheading, heading,
or chapter. The third defines the value content and prescribes either a minimum percentage of the products value that must originate in the territory
of the exporting PTA member (domestic or regional value content) or a
maximum percentage of the products value that can originate outside the
PTA members territory (import content). Finally, technical requirements
prescribe or prohibit the use of certain inputs or the realization of certain
processes in the production. Technical requirements are particularly prominent in rules of origin governing apparel products.
Rules-of-origin regimes use these four substantial transformation rules
both alone and in combination with each other. The staple of regimes is the
change in tariff classification. This rule is used at different levels: more than
half of NAFTA rules of origin are based on a change in chapter, while many
other regimes use mainly a change in heading. Table 1 displays the percentage shares of various combinations of rules-of-origin components in
selected PTAs in the Americas and elsewhere. The table reveals the high
10. The Revised Kyoto Convention is an international instrument adopted by the World
Customs Organization (WCO) to standardize and harmonize customs policies and procedures around the world. The WCO adopted the original Convention in 1974. The revised
version was adopted in June 1999.
69
T A B L E 1 . Distribution of Rules-of-Origin Combinations, Selected PTAs in the Americas

Requirementa
NC
NC + ECTC
NC + TECH
NC + ECTC + TECH
NC + VC
NC + ECTC + VC
NC + VC + TECH
NC + Wholly obtained
chapter
NC + Wholly obtained
heading
Subtotal
CI
CI + ECTC
CI + TECH
CI + ECTC + TECH
CI + VC
CI + ECTC + VC
CI + VC + TECH
Subtotal
CS
CS + ECTC
CS + TECH
CS + ECTC + TECH
CS + VC
CS + ECTC + VC
CS + VC + TECH
CS + ECTC + VC + TECH
Subtotal
CH
CH + ECTC
CH + TECH
CH + ECTC + TECH
CH + VC
CH + ECTC + VC
CH + VC + TECH
CH + ECTC + VC + TECH
Subtotal
NAFTA
U.S.Chile
G3
0.54
0.51
4.05
MercosurChile
Andean
Community
ChileKorea
U.S.Jordan
E.U.Mexico
E.U.Chile
0.39
2.04
1.39
0.39
2.39
1.39
83.94
10.91
1.57
0.20
11.90
1.57
0.20
16.06
7.62
7.62
0.70
0.70
0.51
0.02
100.00
0.54
0.78
0.53
4.05
0.00
100.00
1.29
100.00
24.82
26.16
0.04
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
1.29
2.52
0.04
0.40
16.56
5.57
0.14
0.04
0.42
0.04
1.54
0.73
0.10
0.04
4.60
0.20
0.20
1.90
1.78
0.27
0.27
2.37
2.25
32.99
5.13
32.86
4.56
6.66
12.68
0.86
6.66
12.78
0.37
0.02
0.02
58.34
57.25
0.02
0.02
0.10
1.68
0.47
2.11
0.16
0.04
0.83
4.35
22.77
7.88
0.00
17.09
19.18
0.02
0.14
3.54
0.58
0.10
23.70
11.19
0.34
0.44
3.25
0.48
16.45
13.45
0.97
0.26
2.01
46.00
40.65
39.40
0.00
9.99
23.97
46.02
100.00
0.00
46.87
9.12
0.14
20.04
8.06
4.82
4.42
2.95
0.49
0.00
59.57
0.00
(continued)
T A B L E 1 . Distribution of Rules-of-Origin Combinations, Selected PTAs
in the Americas (continued )
Requirementa
NAFTA
CC
CC + ECTC
CC + TECH
CC + ECTC + TECH
CC + VC
CC + ECTC + VC
CC + VC + TECH
CC + ECTC + VC + TECH
30.95
17.71
0.02
5.76
U.S.Chile
G3
23.18
5.83
0.06
8.08
0.06
21.09
5.90
5.43
6.65
0.14
MercosurChile
Andean
Community
ChileKorea
U.S.Jordan
22.49
4.71
0.08
5.67
1.80
E.U.Mexico
E.U.Chile
2.16
1.02
0.04
11.25
2.16
1.02
0.04
11.02
2.67
0.20
Subtotal
54.44
37.21
42.08
0.00
0.00
34.75
0.00
14.47
14.24
Total
100
100
100
100
100
100
100
100
100
Source: Adapted from Estevadeordal and Suominen (2005a); Suominen (2004).

a. The notation on requirements is as follows: NC: No change in tariff classification required; CI: Change in tariff item; CS: Change in
tariff subheading; CH: Change in tariff heading; CC: Change in tariff chapter; ECTC: Exception to change in tariff classification; VC: Value
content; and TECH: Technical requirement. Calculations are made at six-digit level of the Harmonized System.
degree of diversity in rules-of-origin regimes in the Americas. Nevertheless, four main hemispheric rules-of-origin families can be identified.11 One
extreme is populated by the older trade agreements such as the Latin American Integration Agreement (LAIA), which uses one general rule applicable
to all products (either a change of heading level or a 50 percent regional
value content). The LAIA model has been the point of reference for the
Andean Community and Caribbean Community rules-of-origin regimes.
At the other extreme lie the so-called new generation PTAs such as
NAFTA. The NAFTA model served as the reference point for numerous
recent bilateral agreements, including the U.S.Central America free trade
agreement (CAFTA) and the U.S.-Chile, Chile-Canada, Mexico-Bolivia,
Mexico-Chile, MexicoCosta Rica, Mexico-Nicaragua, MexicoNorthern
Triangle (El Salvador, Guatemala, and Honduras), and the Group of Three
(or G3, encompassing Mexico, Colombia, and Venezuela) free trade agreements. The model, particularly the versions employed in the U.S.-Chile
free trade agreement and CAFTA, is also widely viewed as the likeliest
blueprint for the FTAA rules of origin. The NAFTA-based rules-of-origin
regimes are complex: depending on the product, the rules of origin may
require a change of chapter, heading, subheading, or item, and the change
11. Garay and Cornejo (2002).
71
of tariff classification is often combined with an exception, regional value

content (generally ranging from 35 to 60 percent), or technical requirements. Like many other rules-of-origin regimes in the world, the NAFTAmodel regimes contain an alternative list of product-specific rules of origin
for selected products, which enables an exporter to choose between two
types of rules of origin. The list is relatively extensive in NAFTA, covering nearly 40 percent of the products, and its rules of origin are as complex
as those on the main list.
Mercosur rules of origin and the rules of origin in the Mercosur-Bolivia
and Mercosur-Chile free trade agreements are based on the change-ofheading criterion and different combinations of regional value content and
technical requirements. They fall between the LAIA and NAFTA extremes in
their degree of complexity. The Central American Common Markets rulesof-origin regime can be placed between those of Mercosur and NAFTA.12
U.S. bilateral free trade agreements with some extrahemispheric
partnerssuch as Jordan and Israeldiverge markedly from the NAFTA
model, incorporating value content rules of origin alone. The rules of origin
of the U.S.-Singapore and U.S.-Australia free trade agreements, however,
resemble NAFTA in their complexity. The recently forged ChileSouth
Korea and Mexico-Japan free trade agreements also feature sectoral selectivity la NAFTA. The future Canada-Singapore, Mexico-Singapore, and
Mexico-Korea free trade agreements, among others, will likely compound
the spread of the NAFTA model in Asia and the Pacific. Meanwhile, the
European Unions free trade agreements with Mexico and Chile carry the
European Unions standard, harmonized pan-European rules of origin.13
How Restrictive Are Rules of Origin?

Making meaningful cross-product comparisons across the many types of
rules of origin requires a parsimonious tool. Estevadeordals restrictiveness index provides such a tool.14 The indexs observation rule is based on
12. The Central American Common Market chiefly uses a change in tariff classification
only. The regime is more precise and diverse than Mercosur, however, because it requires
the change to take place at the chapter, heading, or subheading level, depending on the product in question.
13. See Estevadeordal and Suominen (2003).
14. Estevadeordal (2000). The index was subsequently made more generalizable in
Estevadeordal and Suominen (2005a) and Suominen (2004). Carrer and de Melo (2004)
compare Estevadeordals index with an ordering emerging from cost estimates of different
types of rules of origin; they find the index to be consistent with the cost ranking.
the length of the jump over the Harmonized Systems tariff lines required
by rules of origin: a change of chapter is more restrictive than a change of
heading, a change of heading more restrictive than a change of subheading, and so on. Value content and technical requirements add to the rules
restrictiveness.
Figure 2 reports the restrictiveness values of rules of origin in some of the
main PTAs. Since it is based on coding at the six-digit level, it also reveals
the degree of interproduct dispersion of restrictiveness values, which serves
as a measure of the selectivity of regimes. The final bar represents the likeF I G U R E 2 . Restrictiveness of Rules of Origin in Selected PTAsa
8
7
6
5
4
3
2
1
0
al
nti
ere
ref
np
No
TA
AF ile
h
.-C
E.U exico
.-M
E.U rope
u
n-E
Pa el
ra
-Is
US dan
r m.
-Jo
US n Com ia
v
a
de oli
An sur-B
rco hile
Me ur-C
s
rco
Me rea
-Ko
ile ivia
Ch -Bol ica
o
xic ta R
Me -Cos
o
xic
Me
G3
FTA
CA ile
h
.-C
U.S
FTA
NA
Source:Adapted from Estevadeordal and Suominen (2005a); Suominen (2004).

a.The box plots represent interquartile ranges (IQR), with the box extending from the twenty-fifth percentile to the seventy-fifth
percentile. The line in the middle of the box represents the median fiftieth percentile of the data. The whiskers emerging from the
boxes extend to the lower and upper adjacent values. The upper adjacent value is defined as the largest data point less than or equal to
x(75) + 1.5 IQR. The lower adjacent value is defined as the smallest data point greater than or equal to x(25) + 1.5 IQR. Observed points
more extreme than the adjacent values are individually plotted (outliers and extreme values are marked usingandsymbols,
respectively).
73
liest outcome of the harmonization process of nonpreferential rules of origin.

Two issues stand out. The first is the presence of rules-of-origin families.
Regimes drawing on the NAFTA model are highly similar in terms of overall restrictiveness and selectivity, as are regimes drawing on the Mercosur
and European Union models. Second, the NAFTA-type rules-of-origin
regimes are by far the most restrictive and selective in the hemisphere and,
indeed, the world. This finding is particularly important in light of the spread
of the NAFTA-model rules of origin across the hemisphere.
Research shows that food, textiles, and apparel products tend to have
the highest restrictiveness values across regimes.15 This provides precursory evidence that rules of origin may be arbitrated by the same political
economy variables that drive tariffs, particularly in the industrialized countries. Nonpreferential rules of origin feature some selectivity, which suggests the operation of political economy dynamics also at the multilateral
leveland the endogeneity of the nonpreferential rules of origin to the
existing preferential rules-of-origin regimes.
Comparing Regimewide Rules of Origin

Rules-of-origin regimes also vary in their use of general, regimewide rules
of originthat is, rules of origin that apply similarly to all products in a
regime. Some of the most commonly used regimewide rules of origin include
the following:
De minimis levels, which allow a specified maximum percentage of
nonoriginating materials to be used without affecting origin.
Cumulation provisions, which enable producers of one PTA member
to use materials from other members without losing the preferential status
of the final product. The three types of cumulation are bilateral cumulation,
which operates between two PTA partners (that is, firms operating in one
partner country can use products that originate in the other and still qualify
for preferential treatment when exporting the product), diagonal cumulation, which allows countries tied by the same set of preferential origin rules
to use products that originate in any part of the common rules-of-origin
zone as if they originated in the exporting country, and full cumulation,
which extends diagonal cumulation to allow the use of goods processed in
any part of the common rules-of-origin zone even if these do not qualify as
originating products.
15. See, for example, Estevadeordal (2000); Estevadeordal and Suominen (2005a);
Suominen (2004); and Sanguinetti and Bianchi (2005).
Prohibition of duty drawback, which precludes the refunding of tariffs on nonoriginating inputs that are subsequently included in a final product that is exported to a PTA partner. Drawback in the context of a PTA is
viewed as providing a cost advantage to producers who gear their final
goods to export over producers who sell their final goods in the domestic
market. However, ending drawback increases the cost of nonoriginating
components to producers who have thus far benefited from it.16
Certification method, which defines the instance authorized to certify
an origin claim. The main methods are self-certification by exporters; certification by the exporting countrys government or a certifying agency; and a
two-step combination of the private self-certification and the public governmental certification. High bureaucratic hurdles for obtaining a certificate of
origin lower the incentives for exporters to seek preferential treatment.
Whereas de minimis and cumulation clauses insert leniency in the application of product-specific rules of origin, drawback prohibition and complex certification methods may have the opposite effect, namely, increasing
the difficulty of complying with the rules-of-origin regime.17
Table 2 compares the regimewide rules of origin in the various rulesof-origin regimes. Bilateral cumulation is applied in virtually all regimes,
but use of other regimewide components varies considerably. Again, the different rules-of-origin families stand out. The NAFTA-model regimes set de
minimis levels at 710 percent, preclude diagonal and full cumulation, do
not permit drawback (often after a certain transition period), and are based
on self-certification.18 There are exceptions; for example, CAFTA, the latest
of the NAFTA-model regimes, allows cumulation within Central America,
16. Many PTAs in the Americas include duty drawback provisions in the market access
chapter rather than in the rules-of-origin protocol. However, the implications of ending
drawback are very similar to the implications of stringent rules of origin, namely, increasing production costs for exporters. Cadot, de Melo, and Olarreaga (2001) show that duty
drawback may have a protectionist bias due to reducing producers interest in lobbying against
protection of intermediate products.
17. Nonmembers of a cumulation area may view the cumulation system as introducing
another layer of discrimination in that it provides incentives for member countries to outsource from within the cumulation zone at the expense of extrazone suppliers.
18. Two qualifications are in order. First, the de minimis principle has numerous exceptions in most regimes. For example, in NAFTA, it does not extend to dairy products, edible
products of animal origin, citrus fruit and juice, instant coffee, cocoa products, and some
machinery and mechanical appliances. Many regimes also calculate de minimis levels in
textile products as the percentage of weight rather than the value of the final product. Second,
although NAFTA prohibits drawback, it has launched a refund system, whereby the producer will be refunded the lesser of the amount of duties paid on imported goods and on the
exports of the good to another NAFTA member.
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Not mentioned
Not mentioned
Yes
Yes
Yes
Yes
Yes
Not mentioned
Yes
Yes
Yes
10
Not mentioned
2
10
2c
Yes (except automotive)

Yes
Yes
7
7
7
8
9
Not mentioned
Not mentioned
10
Not mentioned
Not mentioned
10
10
10
7
10
10
Roll-up
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Bilateral
Yes
Not mentioned
Not after 10 years
No
Yes
Yes (full in EEA)b

No
Yes (full)
No
No
Not mentioned
No after 7 years
No after 8 years
Not mentioned
Not mentioned
No after 5 years
No after 5 years
Yes
Not mentioned
Yes
Not mentioned
No after 2 years
No after 4 years
No after 7 years
No after 12 years
Not mentioned
Drawback
allowed?
No
No
Yes in chap. 62 with Mexico
and Canada
No
No
No
No
No
No
No
No
No
No
No (OP and ISI allowed)a
No
No
Diagonal
Cumulation
Source: Adapted from Estevadeordal and Suominen (2005a); Suominen (2004).

a. Both originating products (OP) and the integrated sourcing initiative (ISI) operate in the U.S.-Singapore Free Trade Agreement. ISI applies to nonsensitive, globalized sectors, such as information technologies. Under
the scheme, certain information technology components and medical devices are not subject to rules of origin when shipped from either of the parties to the FTA. The scheme is designed to reflect the economic
realities of globally distributed production linkages and to encourage U.S. multinationals to take advantage of ASEAN countries respective comparative advantages.
b. European Economic Area (EEA).
c. The section on trade remedies mentions that one of the criteria for imposing a countervailing duty within the block is that the targeted subsidy is not less than the 2 percent de minimis.
G3
MexicoCosta Rica
Mexico-Bolivia
Chile-Korea
Canada-Chile
Mercosur-Chile
Mercosur-Bolivia
Central American Common Market (CACM)
U.S.-Jordan
U.S.-Israel
U.S.-Singapore
E.U.-Mexico
E.U.-Chile
Other regions
Pan-European
ASEAN Free Trade Agreement (AFTA),
AustraliaNew Zealand Closer Economic Relations Trade
Agreement (ANZCERTA)
Southern African Development Community (SADC)
Common Market for Eastern and Southern Africa (COMESA)
Americas
NAFTA
U.S.-Chile
CAFTA
Region and preferential trade agreement
De minimis
(percent)
T A B L E 2 . Regimewide Rules of Origin in Selected Preferential Trade Arrangements
Mexico, and Canada of materials that Central America may use for producing U.S.-bound goods.19 The clause covers only a limited quota, however, and it enters into force only after Canada and Mexico agree on it. The
Central American Common Markets regimewide rules of origin resemble
those of the NAFTA model, but they do not prohibit drawback. Mercosurs free trade agreements do not have de minimis levels or cumulation
provisions, they phase drawback out in five years, and they are based on
public certification.20 The European Unions rules-of-origin regimes stand
out for employing diagonal cumulation extensively across Europe.21
As in the case of the restrictiveness of product-specific rules of origin,
the facilitation provided by regimewide rules of origin to the application
of the rules-of-origin regime can be systematically assessed through an
index. The facilitation index developed by Estevadeordal and Suominen
incorporates de minimis levels, diagonal cumulation, full cumulation, and
drawback (all of which can be expected to cut producers production costs
by amplifying their pool of low-cost inputs), as well as self-certification
(which can keep producers administrative costs lower than the other
methods).22 Figure 3 shows the behavior of the index. Regimes styled
after NAFTA and the European Union feature the highest levels of facilitation, while the Mercosur- and LAIA-based rules of origin score relatively low. The result suggests some correlation between the restrictiveness and facilitation indexes: regimes with the highest restrictiveness
of product-specific rules of origin tend to also have the highest facilitation values.
Many rules-of-origin regimes have devised further, more idiosyncratic
ad hoc mechanisms to help the members adjust to the rigors of rules of
origin.23 Some such mechanisms include phase-in periods for stringent
value content rules of origin; a number of different options for calculating value content rules of origin; and tariff preference levels, which allow
19. See chapter 62 of CAFTA. (The full text of CAFTA is available at www.sice.
oas.org/Trade/CAFTA/CAFTADR_e/CAFTADRin_e.asp.)
20. The Mercosur rules-of-origin regime is similar, but allows for drawback. Drawback
is, however, prohibited for Argentine and Brazilian imports of intermediate automotive
products when the final product is exported within Mercosur.
21. The absorption by the European Union customs union of the ten new member countries implied that thirty-four of the pan-European free trade agreements vanished overnight.
Prior to the accession, the diagonal cumulation incorporated sixteen partners and covered
no fewer than fifty free trade agreements.
22. Estevadeordal and Suominen (2005a); Suominen (2004).
23. For a more thorough treatment, see Estevadeordal and Suominen (2005a).
77
F I G U R E 3 . Facilitation of Regimewide Rules of Origin in Selected PTAs

4
NAFTA
U.S.Chile
CAFTA
G3
Mexico- Mexico- Chile- Merco- Merco- Andean U.S.- U.S.- Pan- E.U.- E.U.Costa Bolivia Korea sursur- Comm. Jordan Israel Europe Mexico Chile
Rica
Chile Bolivia
AFTA
Source:Adapted from Estevadeordal and Suominen (2005a); Suominen (2004).
goods that would not otherwise satisfy the rules-of-origin protocol to

qualify for preferential treatment up to a yearly quota. While most regimes
that employ these mechanisms make them available to all members, some
regimes provide them to one or some of the PTA partners only (for instance
to accommodate country-specific endowments, production structures, and
development levels).
Trends in Rules of Origin in the Hemisphere

The main finding of the above analysis is that rules-of-origin regimes
based on the NAFTA model are among the most restrictive and selective
in the world. The analysis also reveals two key temporal trends in the
Western Hemisphere. First, the so-called new generation regimes of the
1990s score the highest for restrictiveness, selectivity, and facilitation values. Second, the restrictiveness of NAFTA-style agreements has fallen
somewhat over time. Some of the most marked declines are in the mineral,
leather, plastic, apparel, and footwear sectors. In very general terms, this
means, for instance, that some producers based in Costa Rica have greater
leeway to procure inputs or perform operations outside the PTA zone
under the recently signed CAFTA than they do under the 1995 Costa
RicaMexico free trade agreement. What is more, NAFTA itself is liberalizing some of its rules of origin.24 The Working Group in charge of the
rules-of-origin review process is designing new rules of origin on the basis
of consultations with consumers and producers and a review of the rulesof-origin protocols that each NAFTA member country has negotiated in
their post-NAFTA free trade agreements, such as the United States with
Singapore or Mexico with the European Union. If this latter process results
in interregime rules-of-origin borrowing, it could enhance convergence
between the NAFTA rules of origin and the rules of origin of other regimes
around the world.
The Political Economy of Rules of Origin

This section examines why rules of origin are chosen as policy instruments
in preferential trade. After all, integrating governments could simply exclude
the potentially trade-deflecting sectors from the PTAs coverage, or put in
place a common external tariff covering all products. We also consider
why restrictive and selective rules-of-origin regimes have gained ground
over the past few years.
The Choice of Rules of Origin as a Policy Instrument

Studies on political economy widely concur that using rules of origin as a
key policy instrument serves to pay off import-competing lobbies jeopar-
24. The initial set of revised NAFTA rules of origin took effect on 1 January 2003; they
involve alcoholic beverages, petroleum/topped crude, esters of glycerol, pearl jewelry,
headphones with microphones, chassis fitted with engines, and photocopiers. See Regulations Amending the NAFTA Rule of Origin Regulations, Canada Gazette, 1 January 2003
(available at canadagazette.gc.ca/partII/2003/20030115/html/sor24-e.html). In July 2004,
the trade ministers of the NAFTA countries instructed the trilateral Working Group on
Rules of Origin to extend the liberalization drive to chemicals, pharmaceuticals, plastics
and rubber, motor vehicles and their parts, footwear, copper, and all items with a zero mostfavored-nation tariff for all of the NAFTA members. See A Decade of Achievement,
NAFTA Free Trade Commission, 16 July 2004 (available at www.freetradealliance.org/
pdf/2004%20Advocacy/JointStatement.pdf).
79
dized by PTA formation.25 Rules of origin can be employed to favor intraPTA industry linkages over linkages between the PTA and the rest of the
world and thus to indirectly protect PTA-based input producers vis--vis
their rivals outside the PTA.26 If rules of origin provide captive markets
downstream, they may even be superior for the import-competing intermediate producer lobbies than exclusions of their products from the PTA.27
Furthermore, stringent rules of origin can also extend protection to uncompetitive intra-PTA final-good producers. This happens when their extraPTA competitors are too hard-pressed to switch to the components
prescribed by the rules of origin. Even if an extra-PTA firm were to move
operations to the PTA market, the edge of producers with existing intraPTA supply links would continue until the new entrants regional sourcing met the rules of origin.28
Rules of origin, in short, enable governments to balance the competing
claims of export lobbies, which seek a liberalizing PTA in which all products are subjected to tariff phase-outs, and import-competing lobbies,
which are intent on halting all liberalization. Rules of origin compensate
and can even benefit import-competing lobbies, while export interests
accept stringent product-specific rules of origin as a preferable and politically attainable alternative to a PTA rife with exclusions.29 Indeed, regimes
with the most stringent rules of origin also tend to feature the highest facilitation values, which may be a sign of counter-lobbying by exporters threatened by the restrictive rules of origin.
Empirical work supports the hypotheses about the protectionist impulses
behind rules of origin. Both Estevadeordal and Suominen find that restrictive rules of origin tend to be put in place in sectors that are also marked by
high most-favored-nation tariffs and long preferential tariff liberalization
25. Rules of origin are a particularly useful trade policy instrument for two reasons.
First, like tariffs, rules of origin are a highly targetable instrument because they are often
negotiated at the product level. Second, unlike the tariff, rules of origin can be defined in
technical and diverse terms, so they can be tailored differently for each individual good, while
their presumed protection can be hidden since rules of origin are not as immediately quantifiable as a tariff.
26. Krueger (1993); Krishna and Krueger (1995).
27. Suominen (2003).
28. Graham and Wilkie (1998). Given that rules of origin hold the potential for increasing local sourcing, governments can also use them to encourage investment in sectors that
provide high value-added or jobs (Jensen-Moran, 1996; Hirsch, 2002).
phase-out schedules.30 Sanguinetti and Bianchi encounter similar evidence

in Mercosurs rules-of-origin regime.31
Explaining Rules of Origin Trends

Because they use rules of origin as a trade policy instrument, governments
expend considerable time and resources on the tedious, technical, and
often highly contentious crafting of the rules-of-origin protocols. But why
have rules of origin become more restrictive over time? One possibility is
that the liberalization of most-favored-nation treatment and the growth of
global trade have strengthened export lobbies, while antagonizing importcompeting interests. Governments find themselves under growing pressures from export interests to produce deeper trade liberalization, so they
have had to develop targetable and effective tools, such as product-specific
rules of origin, to compensate the potential losers from liberalization. In
the case of NAFTA, for example, neither the deep preferences nor the sustained political support for the agreement would have been possible without a stringent rules-of-origin regime. Earlier integration schemes, such as
LAIA, were less liberalizing than NAFTA; they managed the potential
losers concerns in the tariff schedules, which obviated the need to create
a sturdy set of new compensation tools within the PTA. It is thus no accident that the ambitious liberalization of todays PTAs is accompanied by
restrictive rules of origin.
Another, complementary explanation is that the growing propensity
to fragment global production presents a threat to import-competing
intermediate-good providers, who, in turn, see stringent rules of origin as
an opportunity to discourage final-good producers from outsourcing or
shifting production abroad.32 This notion implies that PTA formation
could be driven by protectionist interests.
As noted above, however, the restrictiveness of rules of origin appears
to have declined somewhat in the microcosm of NAFTA-model regimes
over the past decade. This trend has three potential explanations beyond
the potential strengthening of export lobbies in the Americas since the
mid-1990s. First, NAFTA partners have had time to learn about the implications of the different types of rules of origin. NAFTA is one of the first
regimes in the world to establish rules of origin tailored individually for
30. Estevadeordal (2000); Suominen (2003, 2004).
31. Sanguinetti and Bianchi (2005).
81
each product; both governments and business lobbies thus lacked

information on the effects of rules of origin when the NAFTA rules were
first negotiated. NAFTA-based exporters and producers are widely perceived as having grown to find the rules-of-origin regime excessively
restrictive.33
Second, the newer regimes may be endogenous to the prior ones. Countries integrating with the United States after Mexico did sonamely,
Chile and the Central American countriesmay have sought terms that
are more favorable than those attained by Mexico in order to rapidly bring
themselves on a par with the existing U.S. partners in the U.S. preferential
market.34
The third explanation negates the other two: the reduced restrictiveness
may have little to do with temporal dynamics, but rather may simply be
caused by other variables, such as bilateral trade volumes and the types of
goods produced by the different partners. One hypothesis is that newer
regimes may have achieved the same level of protection provided by
NAFTA through using less stringent rules of origin. Detailed time-series
data on the utilization rates of tariff preferences in the different NAFTAmodel PTAs would help illuminate whether this is the case.35
The Effects of Rules of Origin

We now turn to the potential economic effects of rules of origin. Recent
research indicates that rules of origin can increase firms administrative
and production costs, and both costs can introduce protectionist biases that
hamper the free flow of trade and investment. The differences across rulesof-origin regimes may generate transaction costs, but these have yet to receive
empirical scrutiny. We consider the three costs in turn.
33. In theory at least, stringent NAFTA rules of origin may have caused competitive
extraregional producers to move their production facilities to the NAFTA region. In
response, affected intraregional producers who initially favored restrictive rules of origin
may have grown disposed to loosening the rules-of-origin regime.
34. Perhaps less plausibly, the fact that all recent free trade agreements have been negotiated in the shadow of the FTAA process may have provided the NAFTA-model adherents
incentives to define a rules-of-origin model that is more acceptable to all countries of the
hemisphere than the FTAA rules-of-origin regime. This assumes that the adherents to the
NAFTA model favor the adoption of the FTAA.
35. The pattern would not be universal, however, given that the European Union has
implemented the identical rules-of-origin regime across all its partners.
Administrative and Production Costs

The administrative costs of rules of origin stem from the procedures
required for ascertaining compliance with the regime. They are essentially
bookkeeping costs for the exporterthe paperwork and costs associated
with certifying originand the costs incurred by the partner countrys customs in verifying origin. The administrative costs can be considerable
even in regimes using self-certification.36 Cadot, Estevadeordal, and SuwaEisenmann disentangle NAFTAs administrative costs into those associated
with rules of origin and those that are not; they find the former to approximate
two percent of the value of Mexican exports to the U.S. market.37
The production costs of rules of origin arise from the various technical
criteria imposed by the rules-of-origin regime. If rules of origin encourage final-good producers to use intra-PTA sources even when cheaper
supplies are available in the rest of the world, then the rules of origin raise
production costs and thus likely dampen the PTAs trade-creating potential. Rules of origin can also create trade diversion in intermediates if
they give an unnatural boost to intra-PTA purchases of intermediate
goods. However, if the costs of complying with the rules-of-origin regime
exceed the benefits of the PTA-conferred preferences, then final-good
producers might cease to use the preferential channel, obtaining intermediates from the rest of the world and exporting final goods under the mostfavored-nation regime instead. Status quo would ensue, with the PTA
having no effect on trade. Meanwhile, the various facilitating regimewide
rules of origin should have the opposite effect, helping the PTA channel
to flourish.
36. Many regimes call for self-certification, including NAFTA, CAFTA, and the
U.S.-Chile, MexicoCosta Rica, Canada-Chile, Central American Common Market
(CACM), CACM-Chile, Chile-Korea, U.S.-Singapore, and U.S.-Jordan agreements. The
Mexico-Bolivia agreement implements self-certification after an initial four-year period of
two-step private and public certification. The pan-European, European UnionMexico,
and European UnionChile agreements are mostly based on two-step private and public
certification, with limited self-certification. The G3 agreement, LAIA, the Common Market for Eastern and Southern Africa (COMESA), and the Southern African Development
Community (SADC) specify two-step private and public certification, whereas Mercosur, Mercosur-Chile, Mercosur-Bolivia, Andean Community, Caribbean Community
(CARICOM), and most rules-of-origin regimes in Asia and the Pacific rely on public
certification or delegate certification to a private entity. See Estevadeordal and Suominen
(2005a); Suominen (2004).
37. Cadot, Estevadeordal, and Suwa-Eisenmann (2005).
83
Estevadeordal and Suominen provide the most comprehensive analysis to

date on the implications of rules of origin for trade, based on a 155-country
gravity model spanning twenty-one years.38 They reach four main conclusions. First, restrictive and selective product-specific rules of originthat is,
rules of origin that can be expected to increase production and administrative costsundermine bilateral trade flows. This indicates that stringent rules
of origin do undermine PTAs trade-creating potential. Second, de minimis
levels, diagonal and full cumulation, drawback, and self-certificationwhich
can be expected to reduce a rules-of-origin regimes production and administrative costsfoster bilateral trade. This suggests that lenient regimewide
rules of origin may counteract the negative effects of stringent productspecific rules of origin. Third, restrictive rules of origin in final goods
encourage bilateral trade in intermediate goods. As such, restrictive rules of
origin may result in trade diversion to the PTA area. Fourth, the trade effects
of rules of origin change over time: the negative effects of stringent rules
of origin gradually decrease, while the positive effects of permissive
regimewide rules of origin increase. This suggests that exporters learn to
comply with product-specific rules of origin and to take advantage of
regimewide rules of origin.
Other, single-regime studies on the trade effects of rules of origin
reach similar results, as do the closely related studies on usage rates of
PTA preferences.39 Estevadeordal and Miller document missed preferences (or utilization rates below 100 percent) between the United States
and Canada, which they attribute to the tightening of the rules of origin
under NAFTA in 1994.40 Cadot, Estevadeordal, and Suwa-Eisenmann
38. Estevadeordal and Suominen (2005b); Suominen (2004).

39. Cadot, Estevadeordal, and Suwa-Eisenmann (2005), focusing on NAFTA, show
that stringent rules of origin have undermined Mexicos aggregate exports to the United
States. The United States played a key role in establishing NAFTAs Uniform Regulations
and rules-of-origin enforcement mechanisms. In January 1995, the United States found a
high compliance rate among Mexican and Canadian exporters and producers on rules of origin, at 90 percent and 80 percent, respectively (Reyna, 1995, pp. 3738). Appiah (1999) also
examines NAFTA, but using a three-country, multisector computable general equilibrium
(CGE) model; he finds that rules of origin distort trade flows, diverting resources from their
most efficient uses and undercutting global welfare. James (2004) posits that NAFTA preferences and restrictive rules of origin have undercut Asian textile and apparel exports to the
United States. Flatters and Kirk (2005) find that restrictive South African Development
Community (SADC) rules of origin work against efficiency gains that the members could
reach through outsourcing outside the PTA area.
40. Estevadeordal and Miller (2002).
link the mere 64 percent utilization rate of NAFTA preferences to stringent rules of origin.41
In addition to their short-run trade effects, stringent rules of origin may
cause investment diversion in the long run. This occurs when extra-PTA
final-good producers move production to the PTA area with the sole purpose of meeting the rules of origin, even if the PTA is not the most efficient location for production. Rules of origin can also divert investment
within the PTA. Final-good producers may want to get around rules of origin by moving production to the territory of the PTA partner that has the
largest domestic demand or the lowest external tariff on third-country
inputs (or both)such as the United States in NAFTA.42 From a theoretical perspective, requirements for a high regional value content can paradoxically encourage investment to the PTA country that has the highest
production costs (that is, is the most inefficient producer), because goods
made in member countries with low production costs may be hard-pressed
to meet the rules of origin. Rodriguez theorizes that stringent rules of origin
can lead to distortions in production structures within the PTA, while Estevadeordal, Lpez-Crdova, and Suominen encounter preliminary empirical
evidence that flexible rules of origin are conducive to foreign direct investment (FDI) inflows.43
Transaction Costs
Analysts have yet to understand whether differences among rules-oforigin regimes generate transaction costs and impart economic effects.44
41. Cadot, Estevadeordal, and Suwa-Eisenmann (2005). Krueger (1993) reports that
under NAFTAs predecessor (the U.S.-Canada free trade agreement), Canadian producers
opted to pay the tariff rather than go through the administrative hurdles to meet the rules of
origin. Brenton (2003) and Inama (2004) show that rules of origin shape developing countries odds of qualifying for treatment under the generalized system of preference.
42. For example, a Mexican and a U.S. firm selling on the U.S. market and purchasing
their inputs from outside the NAFTA region would be treated unequally under NAFTA: the
Mexican firm would be disadvantaged vis--vis the U.S. firm because it fails to meet the
rules of origin required to export to the U.S. market (Graham and Wilkie, 1998, p. 110).
43. Rodriguez (2001); Estevadeordal, Lpez-Crdova, and Suominen (2004).
44. Garay and Cornejo (2002) provide the only rigorous examination of the diversity in
rules of origin across regimes. They evaluate the correlations between types of rules of origin in NAFTA, CACM, Mercosurs free trade agreements, and Mexicos free trade agreements. The study finds that only 10 percent of the product-specific rules of origin are exactly
identical or highly similar between the regimes, although up to 75 percent of the rules of origin in most chapters within both the Mercosur and Mexican regimes are highly similar.
85
Any adverse effects would clearly be heaviest for countries that are party
to several relatively different rules-of-origin models, such as Chile and
Costa Rica. These so-called spoke countries require customs that are wellequipped to verify and implement the different rules-of-origin regimes, and
they may eventually have to tailor their production structures differently
for each PTA market.45 This generates transaction costs that would be nil
in a world with one rules-of-origin model. The costs will be highest for
small producers in spokes with a narrow domestic sourcing base. In contrast, producers and customs alike in rules-of-origin hubssuch as the
European Union, Mercosur, Mexico, and the United Statesescape most
of these costs. If the transaction costs of operating on many PTA fronts
become excessive, then producers in spoke countries may be compelled to
specialize for one preferential channel over the others.46 At the global
level, the market specialization induced by rules of origin could give rise
to policy-driven, trade-diverting PTA hubs.
Other factors, however, could mitigate the costs associated with crossregime differences. First, a small producer generally produces only a few
items and would thus need to apply only a couple of different rules of
origin when exporting to the various preferential markets. Multinational
companies selling a variety of goods in different markets may face greater
complexity, but they are also better equipped to economize any transaction costs given their superior human, technical, and financial capacities.
Second, even when rules of origin differ across PTA markets, a single
production process may qualify for preferential treatment in each market.
45. Consider a Chilean producer of typewriters (heading 8469): the firm will have to
comply with rules of origin that stipulate a ceiling of 50 percent import content to enter the
European Union; a change of subheading (except from subheading 8469.12) to enter the
United States; a change of heading to enter Korea (except from heading 84.13 or, alternatively, a change from heading 84.13, provided the regional value content is not less than
45 percent using the build-down method or not less than 30 percent using the build-up
method); and a 60 percent regional value content (that is, a ceiling of 40 percent import content) to enter Mercosur. Meanwhile, a European Union producer in the same heading can
use the same rules of origin50 percent import contentto enter Mexico, Chile, South
Africa, and the whole pan-European system. This example also illustrates the comparative
complexities faced by customs: if each rules-of-origin regime stipulates rules of origin for
5,000 products, the Chilean customs would basically have to verify 20,000 different rules
of origin, whereas customs in the European Union countries would only need to verify
5,000 rules of origin.
46. Inter-PTA divergences also allow countries wishing to join these preferential
arrangements to engage in PTA shopping, choosing to join the agreements that best accommodate their existing domestic standards and interests, rather than joining PTAs that are liberalizing, neutral vis--vis third parties, and welfare-enhancing.
Much depends on the idiosyncrasies of the product and production process

in question. Finally, regimewide facilitation mechanisms can go a long
way toward reducing the effects of the cross-regime incompatibilities in
product-specific rules of origin.
Avenues for Future Research

Our understanding of the effects of rules of origin is far from complete. The
costs of differences across rules-of-origin regimes await analysis. Three
avenues for future empirical research would be particularly fruitful. The
first involves the long-term effects of rules of origin, particularly in light
of the interplay of intermediate and final goods markets. While restrictive
rules of origin may initially dampen intra-PTA trade in final goods by
increasing the cost of intermediate goods, the subsequent decline in the
demand of intermediates should lower their price and thus revive both the
demand for them and the intra-PTA trade in final goods.47
The second avenue for research is the economic impact of the certification and verification costs of rules of origin, together with the potential
trade-off between the different certification methods, on the one hand, and
verification costs, on the other. For example, do regimes using the selfcertification method increase the costs of verifying origin, such that the
low transaction costs of certifying origin translate into high transaction
costs of verifying origin?
Finally, the welfare effects of rules of origin remain uncharted.48 Capturing welfare effects will undoubtedly prove challenging, given that rulesof-origin regimes carry frictionsincluding restrictiveness, selectivity, and
various regimewide componentsthat work in different directions.
Policy Recommendations on Rules-of-Origin Systems for the Americas

This study has analyzed the structure and evolution of rules-of-origin
regimes in the Americas and reviewed the latest research results on the
effects of rules of origin. The main findings are three-fold: stringent rules
of origin can be used as a tool to pay off protectionist sectors in a PTA and
thus to foster the political prospects of PTAs; the NAFTA rules-of-origin
47. See Ju and Krishna (1998) and Duttagupta and Panagariya (2001).
48. Appiah (1999) finds that rules of origin undermine welfare in the case of NAFTA,
although his operationalization of rules of origin in a CGE framework is rather crude.
87
model gaining force in the Americas carries relatively restrictive rules of

origin; and restrictive rules of origin can undercut the liberalizing potential of PTAs. Taken together, these findings raise concerns about the ultimate economic effects of the Americas expanding PTA bowl. They also
raise legal concerns: stringent rules of origin may breach Article XXIV of
the General Agreements on Tariffs and Trade (GATT), which in paragraph
8(b) defines a free trade area as a group of two or more customs territories
in which the duties and other restrictive regulations of commerce . . . are
eliminated on substantially all the trade between the constituent territories
in products originating in such territories. Indeed, the WTO has recently
recognized rules of origin as constituting part of other regulations of
commerce.49 Since rules of origin have implications for extra-PTA parties access to the PTA market, they also risk violating paragraph 5 of Article XXIV, which prohibits PTAs that raise barriers toward the rest of the
world from the pre-PTA levels.50
The evolution of the hemispheres rules-of-origin regimes also provides reasons for optimism, however, and the regions countries have a
number of policy options for reducing the potential negative effects of
rules of origin. The rest of this section addresses these two issues.
Encouraging Patterns in Rules of Origin in the Americas

The countries of the Americas have five reasons to be optimistic about the
evolution of the regional rules-of-origin regimes. Each also augurs well
for the design and implementation of the FTAA rules of origin. First, the
most recent rules-of-origin regimes based on the NAFTA modelnamely,
the U.S.-Chile free trade agreement and CAFTAincorporate simpler,
more practical, and less restrictive product-specific rules of origin than
NAFTA. This evinces a trend toward market-friendly rules of origin in the
hemisphere. The NAFTA review process will provide a further boost to
the NAFTA systems liberalization of its rules of origin.
Second, the various regimes designed after NAFTA are fairly similar
vis--vis each other, in both the types of rules of origin specified and their
level of restrictiveness. This can help reduce any potential transaction
49. See, for instance, WTO (2002a). Ambiguities remain as to the meaning of substantially all the trade.
50. The WTO Negotiation Group on Rules is advocating a case-by-case analysis of the
potentially restrictive effects of preferential rules of origin on extra-PTA parties (WTO,
2002a).
costs for NAFTA-model adherents that export under preferential terms to

two or more NAFTA-model PTAs. NAFTAs review of its rules of origin
may engender further interregime compatibilities, thereby paving the way
for diagonal cumulation linking the NAFTA-model free trade agreements.
Third, the NAFTA-style regimes apply relatively lenient facilitation terms.
This helps alleviate the compliance costs of the product-specific rules of origin. Even more encouraging is the movement toward somewhat higher de
minimis levels and the willingness to experiment with diagonal cumulation,
as evidenced in CAFTA. Cumulation is crucial even in the presence of identical product-specific rules of origin across PTAs. Augier, Gasiorek, and
Lai-Tong find that bilateral trade is up to 52 percent lower than expected
between two spoke countries that have identical rules-of-origin protocols
with the same hub, but that are not linked by diagonal cumulation.51
Fourth, the NAFTA model has now been adopted in numerous free trade
agreements. The current adherents will thus find it fairly easy to negotiate,
adopt, and implement future free trade agreements. Should the FTAA come
to carry NAFTA-type rules of origin, the costs of adjusting to its rules-oforigin regime would be low for a good part of the hemisphere.
Finally, negotiators on rules of origin throughout the Americas, and
particularly in free trade agreements based on the NAFTA model, have
proved their willingness to revise existing rules-of-origin regimes to make
them more flexible. NAFTAs review of its rules of origin is the clearest
example, demonstrating commitment to keeping North Americas rules of
origin apace with changes in technology and the globalization of production, and potentially marking a growing role of export interests in setting
trade policy.
More generally, the precision of the NAFTA-model rules of origin is
superior to the vaguely defined and subjective rules of origin of the past.
Precision provides clarity and certainty to traders and customs alike. Because
the NAFTA regime is based on the change in tariff classification, it provides a fairer, more transparent, and more easily verifiable rules-of-origin
model than regimes based on value content, which paradoxically can be
hard to meet in countries with low production costs and are difficult to
implement in the face of fluctuations in exchange rates and changes in production costs. Precise rules of origin do not need to be restrictive rules of
origin; the NAFTA review process may well yield rules of origin that are
both precise and flexible.
51. Augier, Gasiorek, and Lai-Tong (2004).
89
Tackling the Negative Effects of Rules of Origin: Flexible Rules of Origin

plus Hemispheric Cumulation
The positive trends in the Americas notwithstanding, potential sources of
friction remain: stringent and selective rules of origin still govern many
sectors, and the various regimes differ markedly, even across the subset of
regimes based on the NAFTA model. How can entrepreneurs obtain inputs
from the cheapest sources, firms exploit cross-border economies of scale, and
multinational companies make sweeping investment decisions based on economic efficiency? How can producers in spoke countries qualify for all the
preferential markets simultaneously without undue transactions costs? What
are the best ways to counter the rise of trade- and investment-diverting hubs?
The simplest way to preempt the negative effects of rules of origin
would be to bring most-favored-nation tariffs to zero globally, although
this is not likely to become politically palatable in the near future. A further
option would be to move from free trade agreements to customs unions
with low common external tariffs, thereby eliminating rules of origin altogether, or, alternatively, to harmonize preferential rules of origin at the
multilateral level, which would ensure compatible requirements across
spoke producers export markets. However, the founding of customs
unions with an across-the-board common external tariff has proved difficult outside the European Union; rules of origin will thus remain an issue
as long as a common external tariff does not cover all product categories.52
Meanwhile, the prospect of global harmonization of preferential rules of
origin is still relatively distant.
Two shorter-term policy options are more realistic. First, the existing
regional rules-of-origin spaghetti bowl can be revised. PTA members
should strive to design and revise their rules-of-origin regimes to establish
transparent, simple, precise, nonrestrictive product-specific rules of origin,
such as a change in heading or subheading, and they should put in place
lenient regimewide rules of origin, in particular a high de minimis level.
Such rules of origin alone would reduce the frictions within and between
PTAs. The hemispheres PTAs should be interconnected through diagonal
cumulationa task that would be relatively uncomplicated to implement in
the presence of readily harmonized origin regimes and would pave the way
52. NAFTA contains a small sectoral customs union, with a common external tariff
governing certain automatic data-processing goods and their parts. The tariff ranges from
zero to 3.9 percent. See NAFTA Annex 308. (The full text of NAFTA is available at
www.sice.oas.org/trade/nafta/naftatce.asp.)
to a regionwide trade and production base. The countries of the Americas

should also improve training for exporters and customs about the technical
requirements and implementation of rules of origin. These measures would
help shorten the learning lags associated with rules of origin, reduce the
administrative hurdles facing both exporters and customs, provide small
countries access to larger pools of intermediate goods, and allow spoke
economies to trade on several different fronts by applying the same rules of
origin. This, in turn, would ensure that the hemisphere continues to enjoy
the benefits of open regionalism. The movement from the complex and
restrictive NAFTA rules-of-origin regime to the simpler and less restrictive
U.S.-Chile free trade agreement and CAFTA is an encouraging step, and it
should be furthered in future regimes.
Second, the FTAA would automatically sort out the rules-of-origin
spaghetti bowl and put in place a hemispherewide cumulation zoneno
small feat given that the countries of the Americas contain a sizable subsample of the worlds PTAs.53 An optimal FTAA rules-of-origin outcome
would establish simple, nonrestrictive product-specific rules of origin and,
again, lenient regimewide rules of origin. The overall framework could be
buttressed with ad hoc innovative measures designed to accommodate the
partners idiosyncratic production patterns and capabilities. Thus construed, the FTAA would also prove that the hemispheres existing PTAs
represent genuine building blocks for regionwide trade liberalization.54
To be sure, all hopes should not be pinned on the FTAA. Much work
remains to be done to reconcile the various partner countries rules-oforigin preferences, and the FTAA project per se has been troubled over
the past several months. Nonetheless, the FTAA might prove to be the
only way to integrate the NAFTA- and Mercosur-model rules-of-origin
53. Countries (and regions such as Mercosur) have thus far submitted rules-of-origin
proposals for chapters 140. Each product tends to feature five to ten different proposals.
54. The hemispheres trade ministers proposed in November 2003 a two-tiered FTAA,
with the first tier of keen integrators adopting deep commitments and wide tariff liberalization across the tariff universe and the second tier opting for shallower commitments and a
narrower list of liberalized products. Two cumulation zones would likely resultone with
all member countries and a narrower range of goods, and another with the wider liberalizers in the additional set of goods. Cumulation in both tiers could be complemented with
some ad hoc tools, such as phasing in the rules-of-origin regime, particularly for the smaller
countries. Should this structure result, countries in the two tiers would be able to cumulate
in the products they have liberalized with partners that have liberalized the same goods.
Wider liberalizers would thus cumulate among each other in a broad range of goods, while
all countries would cumulate in the narrower range to which the second-tier countries have
acceded. See Blanco, Zabludovsky, and Gmez Lora (2004).
91
regimes.55 It could also facilitate the prospects of multilateral harmonization of preferential rules of origin.
Whether accomplished through interlinking PTAs or through the FTAA,
a hemispherewide cumulation zone appears to be the most promising
optionparticularly when its rules of origin are flexible enough to prevent
trade diversion. A sustained fluid operation of a hemispherewide cumulation zone will require solid verification tools. Poor verification is a major
problem in most of Latin America, and it has been accentuated by the
growing inflows of goods, particularly from Asia. This situation could
provoke a backlash against regional trade liberalization. The strong verification regime that CAFTA introduces in the textile and apparel sector
could serve as a starting point, along with technical assistance to countries
with the most feeble verification systems. Information technology should
be fully harnessed to facilitate verification.56
The countries of the Americas cannot afford to pursue new policies
only within the hemisphere, but should push their WTO partners on two
fronts. First, they should call for launching the harmonization of the worlds
preferential rules-of-origin regimes. This option is increasingly timely
given the proliferation of free trade agreements with different rules of origin around the world and, in particular, the establishment of free trade
agreements between the hemispheres countries and extraregional partners. Harmonizing multilateral rules of origin is hardly a novel idea, but
rather is a long-standing international commitment: the Uruguay Round
Agreement on Rules of Origin stipulates that once the signatories conclude the harmonization of nonpreferential rules of origin, they will move
to harmonize preferential rules of origin, using the relatively flexible and
simple harmonized nonpreferential rules of origin as a blueprint.
The second multilateral policy that the countries of the Americas should
pursue is the lowering of tariffs and nontariff barriers. The higher the PTA
partners most-favored-nation barriers, the wider the preferential margins
and the greater the willingness of firms in the partner countries to comply
55. Even if the different hemispheric rules-of-origin regimes were left to coexist with
the FTAA rules of origin (as occurred with the Central American Common Market and
CAFTA rules of origin), exporters would be better off for two reasons: first, firms could
choose between two alternative rules of origin when trading with their pre-FTAA PTA partners, and second, the FTAA rules of origin could be less restrictive, in practice, than the
prior PTA rules of origineven if they are more restrictive on paperbecause the FTAA
cumulation zone is vastly expanding the pool of inputs available to any member country.
We thank Jeremy Harris for pointing this out.
56. See Cornejo (2004).
even with costly and distortionary rules of origin. The expansion of the
PTA spaghetti bowl must be accompanied by open regionalism, in which
most-favored-nation liberalization proceeds hand-in-hand with preferential
opening.57
Conclusion
This paper has analyzed the various rules-of-origin regimes in the Americas, reviewed the latest research findings on the effects of rules of origin,
and provided policy recommendations for the regions countries to reduce
the adverse economic impact of rules of origin. We have found that the
NAFTA rules-of-origin model, which is expanding in the hemisphere, carries restrictive and complex rules of origin, and such rules of origin can
counteract PTA-inspired trade liberalization. These findings raise concerns about the hemispheres increasingly complex rules-of-origin bowl.
The worrisome features can be tamed, however, through regional cooperation, in particular the adoption of simple and transparent product-specific
rules of origin, the incorporation of mechanisms to promote regimewide
flexibility, and the implementation of cross-PTA diagonal cumulation.
Given the globalization of regional integrationthat is, the movement of
regional partners to negotiate interregional agreementsthe countries of the
Americas should also live up to the Uruguay Round commitment of harmonizing preferential rules of origin at the global level.
Preferential rules of origin matter only as long as there are multilateral
barriers to trade. If there is a silver bullet for reducing the negative effects
of rules of origin, it is the multilateral liberalization of tariffs and nontariff
barriers. If the Doha Round negotiators succeed in producing deep cuts in
most-favored-nation tariffs and nontariff barriers, and if the proliferation of
PTAs engenders a dynamic of competitive liberalization worldwide, the
importance of preferential rules of origin as gatekeepers of commerce will
progressively dissolve.
57. See Bergsten (1997) and Wonnacott (1996). Wonnacott suggests that free trade
agreements should be replaced by customs unions or a hybrid arrangements of customs
unions and free trade agreements, lest the benefits of preferential opening be lost.
Comments
Pablo Sanguinetti: This is a very interesting paper that deals with an
important and often neglected aspect of preferential trading agreements
(PTAs), namely, the determination of rules of origin. Rules of origin are
the regulations that determine under what circumstances a good is considered to be produced in the region and thus able to enjoy the preferential
tariff treatment. The definition of these regimes, which is mainly the concern of lawyers and policy practitioners, could have important economic
impacts on trade and investment flows. Rules of origin have therefore
become an alternative trade policy instrument targeted by governments
and especially by the private sector in the integrating countries.
The paper does four things. First, it offers a very complete and detailed
survey of the various rules-of-origin regimes that have been put in place in
the context of the huge increase in PTA initiatives for the world economy
and the Americas in particular over the last fifteen years. Second, the paper
draws on the political economy literature to examine why the use of rules
of origin has become such an important policy for government and private
sector lobbies and why the level of restriction implied by rules of origin
has increased over time. Third, given what the (positive) theory predicts
regarding why rules of origin are established, the paper summarizes the
evidence about the effects of these regulations on trade and investment
flows. Finally, the paper ends with policy recommendations. I concentrate
my comments on the first two of these issues: the features of the various
rules-of-origin regimes in the Americas and the political economy aspects
of these rules.
On the Extension of the Restrictive NAFTA Model in the Americas

The paper concludes that the NAFTA model of rules of origin, which has
been widely applied in the Americas since the second half of the 1990s,
is much more restrictive and selective than the rules included in previous
93
agreements like the old LAIA system and also those applied in Mercosur
and in the free trade agreements between Mercosur and Chile and Bolivia.
Is this bad news for free trade in the region? To a certain extent, it is not
surprising that the NAFTA model has been extended to various free trade
agreements in the Americas. Many of these new free trade agreements
were signed by NAFTA member countries (including all the bilateral free
trade agreement signed by the United States), which presumably would
establish similar rules in their new agreements in the interest of internal
consistency and for the same political reasons that originated the NAFTA
rules-of-origin system. It is also not surprising that the NAFTA-type rulesof-origin regime is more restrictive than those established in previous preferential trade initiatives. As the authors mention, initiatives such as LAIA
were much less ambitious than NAFTA, and many sectors and goods were
exempted from free trade. Import-competing producers did not have to ask
for an alternative mechanism to receive some sort of import relief because
they were already excluded from the agreements. The free trade agreements signed since the beginning of the 1990s, however, are more in
accordance with Article 24 of GATT in that they cover a significant part
of trade and go much deeper in terms of eliminating trade barriers (even
compared to unilateral or multilateral liberalization schemes). Governments and import-competing sectors naturally try to target additional
measures like rules of origin to ease the cost of adjustment for sensitive
sectors. This reasoning implies that this development is not necessarily
bad news for free trade in the Americas, since the extension of rules of origin is precisely a reaction to further trade integration. On the other hand,
the impact of these added restrictions may partially undo the gains from
liberalization resulting from decreasing tariffs.
Mercosur differs from NAFTA in that it is an incomplete customs
union, which has certain advantages. Since the main normative argument
for adopting rules of origin is to avoid trade deflection (that is, imports
entering the member country with the lowest tariff and being reshipped to
the other partners with no additional tariffs), rules of origin are not relevant for items that have already converged to the common external tariff
of the trade union. One would therefore expect a more lenient regime in
Mercosur than in NAFTA. By the same argument, Mercosur rules of origin should also be less restrictive than those included in the free trade
agreements signed by Mercosur with other countries, like Chile and Bolivia.
Table 3 presents an index that measures the degree of restrictiveness of the
Mercosur, Mercosur-Bolivia, and Mercosur-Chile regimes. The index
95
T A B L E 3 . Mercosur: Rules-of-Origin Index, by Manufacturing Sectora

Sector
Mercosur
Mercosur-Bolivia
Mercosur-Chile
Food, beverages, and tobacco

Textiles, apparel, and leather
Wood products
Paper and printing
Chemicals
Nonmetallic products
Basic metal products
Metal products, machinery, and equipment
Other manufacturing products
1.3
1.8
1.0
1.2
2.5
1.1
1.7
1.6
1.0
1.7
2.9
1.7
1.4
2.7
1.2
2.6
2.0
1.3
1.4
2.9
1.4
1.3
2.6
1.1
2.6
1.9
1.2
Total
1.7
2.3
2.2
Source: Sanguinetti and Bianchi (2005).

a. The index ranges from one to four, with one being the most lenient regime and four the most restrictive.
ranges from one to four, with one being the most lenient regime and four
the most restrictive.1 The overall level of restriction implied by rules-oforigin rules is 1.7 for intra-Mercosur trade, 2.2 for Mercosur-Chile trade,
and 2.3 for Mercosur-Bolivia. The table also shows that sectors like textiles, chemicals and basic metal products (steel) are among those most
affected by these regulations. Estevadeordal and Suominen find similar
results for NAFTA.
Despite the fact that Mercosur is an incomplete customs union (so that
rules of origin should only matter for items that are exempted from the common external tariff, as mentioned above), in practice, the rules-of-origin
regime is applied to all items independently of whether they are included in
the common external tariff. This evidence confirms that these rules are used
not only for the normative prescription of avoiding trade deflection, but also
as a policy tool that could potentially offer some type of import protection.
On the Political Economy of Rules of Origin

Given that rules of origin can function as a protectionist device within the
context of free trade agreements, how does a political economy approach
change the normative prescription about the emergence of free trade agreements and the role of these regulations? What are their determinants, and
1. This index is developed in Sanguinetti and Bianchi (2005) and closely follows the
methodology presented in Estevadeordal (2000).
how do they relate to other key trade policy variables like tariff preferences?
The paper addresses some of these concerns, but I wish to offer some comments to complement the authors discussion.2
Grossman and Helpman provide a political economy model of the emergence of free trade agreements.3 According to their approach, the decision
of whether to form a free trade agreement is subject to political pressures
from the potential losers and winners of trade creation and trade diversion.
Grossman and Helpman use the term enhanced protection to describe trade
diversion and reduced protection for trade creation (relative to the tariffridden situation prevalent before the free trade agreement). This approach
suggests that exporters that stand to gain the most from trade diversion in
the partner country will be most in favor of establishing the trade agreement, while import-competing sectors that will suffer from trade creation
originating in imports from the other members will most vividly oppose
the free trade agreement. Thus producers will support a free trade agreement when the probability of generating trade diversion is maximized
and trade creation is minimized. This is the case when, from a normative
point of view, a free trade agreement is not fully justifiable. In practice,
the final result will depend on how efficient these different groups are in
influencing government policy through lobby activity and how the government objective function weights consumer welfare vis--vis that of producer groups.
The original Grossman and Helpman model does not address the issue
of intermediate inputs, so it cannot be easily applied to study the endogenous determination of rules of origin. This extension is provided by Cadot,
Estevadeordal, and Suwa-Eisenmann, who present a simple partial equilibrium model in which two countries (North and South) engage in a free
trade agreement and both tariff preferences and rules of origin are jointly
determined.4 They focus on a case in which intermediate-good interests in
the North wish to use the free trade agreement to create a captive market
for their product. These interests lobby their government (though political
contribution, as in Grossman and Helpman) to establish strong rules of origin to obligate Southern final-good producers to source in the North in
order to qualify for preferential access. This clearly reduces the effective
protection that the Southern producers receive for entering into the finalgood market in the North. The authors assume that the South is always on
2.
3.
4.
My comments are based on Sanguinetti and Bianchi (2005).

Grossman and Helpman (1995).
Cadot, Estevadeordal, and Suwa-Eisenmann (2003).
97
its participation constraint (that is, effective protection is zero).5 In this

context, deeper tariff preferences for the final goods can sustain stricter
rules of origin. This, in turn, favors the Northern producers because
it raises both the demand for their product and, more important, the
intermediate-good price. The model thus delivers the interesting prediction that this price is not tariff ridden, but depends on demand and supply
(as if the market for this product were closed). This is not surprising; rules
of origin function as a type of quantitative restriction. This framework
leads to the testable implication that the restrictiveness of rules of origin
and tariff preferences are positively associated. This positive association
is documented in Estevadeordal for NAFTA and in Sanguinetti and
Bianchi for Mercosur.6
Summary Remarks
As I indicated at the beginning, this paper by Estevadeordal and Suominen
is a very interesting piece of work that carefully analyzes the political, economic, and policy implications of rules-of-origin regimes in the Americas.
I hope this survey-type of work encourages further research on the topic.
Alberto Trejos: I quite like this paper, which thoroughly addresses the
topic of rules of origin in current and future free trade agreements in the
Americas. Motivations for this kind of work include concerns that the growing complexity of the administration of rules-of-origin regimes will be
compounded as very disparate rules are implemented across different
agreements; the problem that many free trade agreements may use stringent rules of origin as an alternative (and less visible) mechanism for
maintaining high rates of protection; and the possibility that such disparate
rules of origin will turn free trade agreements into a stumbling block,
rather than a building block, in the process of world trade liberalization.
Understanding this topic is necessary if governments are to design the
correct policies, including better free trade agreements, in the future. The
5. In this case, exports of the final good will not increase significantly as a consequence
of the free trade agreement initiative. Thus the lobby for stronger tariff preferences by the
intermediate-good industry in the North will not face strong opposition from the final-good
industry in the same country. There will be very low trade creation in final goods and a
strong trade diversion in intermediates.
6. Estevardeordal (2000); Sanguinetti and Bianchi (2005).
majority of world trade (especially within the Americas) happens today in

the context of free trade agreements or other preferential arrangements in
which rules of origin are applied. Previous work by the same authors illustrates that in the Western Hemisphere the prevalent rules-of-origin
regimes are indeed more restrictive and heterogeneous than in the rest of
the world.
When rules of origin are binding, they can have some of the same
effects as tariffs and other barriers to trade. They discourage trade, require
learning, reduce the rate of utilization of free trade agreements, and redirect investment and trade. Furthermore, the costs of compliance can be
very high, reaching 2 percent of the total value of trade in some cases.
While not as effective as tariffs when used as trade barriers (especially in
comprehensive free trade agreements in which tariff phase-out takes place
across almost all goods), they provide protectionist measures that the general public does not always see and that policymakers have a hard time
quantifying.
The authors measure and assess rules-of-origin regimes according to
the stringency of the rules, the cost of implementing them, their nature,
and their heterogeneity within and across agreements. They find a very
high diversity of rules of origin in the existing free trade agreements and
preference regimes in the Americas, both across agreements and across
goods within a given agreement. The rules of origin can also be very stringent, especially in older free trade agreements.
At the same time, the authors demonstrate that there are some sources
of optimism on this topic. First, newer agreements are less restrictive. Second, as economies become more open and the results of enhancing trade
are appreciated, it becomes easier for governments to negotiate agreements
that boldly go beyond their predecessors. Third, countries that are now
negotiating new free trade agreements show signs of significant learning
from a decade or so of implementing their older agreements. Fourth, to
remain competitive in an environment where others are doing the same,
negotiators of new agreements are producing further liberalization than in
previous agreements. Finally, the most recent free trade agreements have
been negotiated in the context of an imminent Free Trade Area of the
Americas (FTAA), which would much reduce the effectiveness of rules of
origin as trade barriers. (This factor will probably be less meaningful in
free trade agreements negotiated after the modest results of the Miami
ministerial of 2003, which much delayed the expected time of completion
of a comprehensive FTAA.)
99
I would add to these causes for optimism the fact that recent agreements
include a variety of new flexibilities to make rules of origin less stringent.
De minimis clauses, phase-ins, tariff preference levels, and, most important, accumulation of origin are the most important of such flexibilities.
The authors similarly mention the possibility of building on the progress
at the WTO on multilateral harmonization of rules of origin in a mostfavored-nation basis; I am not optimistic about achieving relevant progress
there at this time.
While criticisms of the restrictiveness of rules-of-origin regimes are
largely valid, the political economy of trade negotiations is such that restrictive rules of origin are often the only way to maintain a particular product
in the tariff phase-out commitments of a free trade agreement. Not only do
rules of origin give the local producer of the good more protection (in
which case the rules of origin undo some of the progress attained in the
phase-out), but restrictive rules of origin create other winners (the regional
producers of the key inputs to that good), often tilting the balance. Trade
diversion toward the parties involved in a free trade agreement is always
politically more feasible than trade diversion away from them, and this is
used in negotiations to generate political backup for further liberalization.
A flexible rule of origin (which is always preferable, of course) may
reduce the feasibility of achieving a quick tariff reduction in the first place
by shifting the sourcing of materials to third countries. Under that light,
one may see restrictive rules of origin as a necessary, and transitory, evil
in some cases.
The authors neglect to look carefully at the growing web of subregional
agreements in the hemisphere. Mercosur, the Andean Pact, CARICOM, and
the Central American Common Market involve plans of economic integration that go much further than current free trade agreements. These efforts
will probably converge to a situation in which nations that belong to the
same subregional group, in their efforts to construct customs unions, will
homogenize their existing bilateral agreements with third parties, committing to the same rules of origin and allowing for origin accumulation among
the subregional partners. This will probably take a long time to come to
fruition, but when it does it will significantly simplify the spaghetti bowl
problem and reduce the distortionary impact of rules of origins.
The authors should also address the question of how rules-of-origin
regimes differ across free trade agreements in another way: while rulesof-origin procedures may be very heterogeneous across different goods
within a given free trade agreement, specific goods might be treated similarly
across different free trade agreements. My impression is that this is the

case for some of the problematic goods, so the effects of current rules-oforigin regimes on FTAA and on future integration are less daunting than a
first read of the paper may suggest.
In general, accumulation of origin that is not limited to subregional
partners is a significant source of optimism that the hemispheres rules of
origin will become less onerous, both as trade barriers and as administrative costs. For example, four distinct (but quite similar) agreements existing today bind together, in all directions, a group of four nations (namely,
Canada, Chile, Mexico, and the United States). Costa Rica will join this
group with the enactment of CAFTA, as will the other Central American
Common Market partners once their agreement with Canada is in place. It
should be feasible and desirable for nations in this list to allow, in their
bilateral agreements, origin accumulation with other nations in the list, as
the direct market access to those other parties has already been granted.
That bottom-up mechanism may result in a better way to construct hemispheric integration and solve the problems of the complexity and stringency of rules of origin.
In conclusion, this is a very good and important paper. It is not easy
to figure out how to address this question systematically, and the technical work required for that purpose is certainly daunting. The authors
clearly do a good job there. They ask the right questions and raise many key
points. Perhaps some topics (origin accumulation, in particular) deserve
more attention than was given to them, but the effort clearly achieves
progress.
References
Appiah, Alex Jameson. 1999. Applied General Equilibrium Model of North
American Integration with Rules of Origin. Ph.D. Dissertation, Simon Fraser
University (Canada).
Augier, Patricia, Michael Gasiorek, and Charles Lai-Tong. 2004. The Impact of
Rules of Origin on Trade Flows. Paper prepared for the conference Rules of
Origin in Regional Trade Agreements: Conceptual and Empirical Approaches. Inter-American Development Bank (Integration and Regional Programs Department), Centre for Economic Policy Research, and DELTA/INRA,
Washington, 2021 February.
Bergsten, C. Fred. 1997. Open Regionalism. In Whither APEC: The Progress to
Date and Agenda for the Future, edited by C. Fred Bergsten. Washington:
Institute of International Economics.
101
Blanco M., Herminio, Jaime Zabludovsky K., and Sergio Gmez Lora. 2004.
Una llave para la integracin hemisfrica. Documento de divulgacin
IECI-03. Buenos Aires: Instituto para la Integracin de Amrica Latina y el
Caribe (INTAL) and Inter-American Development Bank, Integration and
Trade Division.
Brenton, Paul. 2003. Integrating the Least Developed Countries into the World
Trading System: The Current Impact of EU Preferences under Everything but
Arms. Policy research working paper 3018. Washington: World Bank.
Cadot, Oliver, Jaime de Melo, and Marcelo Olarreaga. 2001. Can Duty Drawbacks Have a Protectionist Bias? Evidence from Mercosur. Working paper
2523. Washington: World Bank.
Cadot, Olivier, Antoni Estevadeordal, and Akiko Suwa-Eisenmann. 2003. Rules
of Origin as Export Subsidies. Washington: Inter-American Development
Bank. Mimeographed.
. 2005 (forthcoming). An Assessment of Rules of Origin: The Case of
NAFTA. In The Origin of Goods: A Conceptual and Empirical Assessment of
Rules of Origin in PTAs, edited by Olivier Cadot and others. Oxford University Press and Centre for Economic Policy Research.
Carrer, Cline, and Jaime de Melo. 2004. Are Different Rules of Origin Equally
Costly? Estimates from NAFTA. Discussion paper 4437. London: Centre for
Economic Policy Research.
Cornejo, Rafael. 2004. Recientes innovaciones en los regmenes de origen y su
incidencia en el proceso de verificacin: el caso del CAFTA. Mimeographed.
Devlin, Robert, and Antoni Estevadeordal. 2004. Trade and Cooperation: A
Regional Public Goods Approach. In Regional Public Goods: From Theory to
Practice, edited by Antoni Estevadeordal, Brian Frantz, and Tam R. Nguyen.
Washington: Inter-American Development Bank and Asian Development Bank.
Duttagupta, Rupa, and Arvind Panagariya. 2001. Free Trade Areas and Rules of
Origin: Economics and Politics. Working paper 03/229. Washington: International Monetary Fund.
Estevadeordal, Antoni. 2000. Negotiating Preferential Market Access: The Case
of the North American Free Trade Agreement. Journal of World Trade 34(1):
14166.
Estevadeordal, Antoni, Jos Ernesto Lpez-Crdova, and Kati Suominen. 2004.
Impact of NAFTA on the Location of Foreign Direct Investment in Mexico.
Washington: Inter-American Development Bank. Mimeographed.
Estevadeordal, Antoni, and Eric Miller. 2002. Rules of Origin and the Pattern of
Trade between the U.S. and Canada. Washington: Inter-American Development Bank, Integration, Trade and Hemispheric Issues Division. Mimeographed.
Estevadeordal, Antoni, and Kati Suominen. 2003. Rules of Origin in FTAs in
Europe and in the Americas: Issues and Implications for the EU-MERCOSUR
Inter-Regional Association Agreement. In Market Access for Goods and Services in the EU-Mercosur Negotiations, edited by Alfredo G. A. Vallado and
Roberto Bouzas. Paris: Chaire Mercosur de Sciences Po. Also published as

Regras de origem em reas de livre comrcio: implicaes para negociaes,
Revista Brasileira de Comercio Exterior (July/September 2004).
. 2005a (forthcoming). Rules of Origin: A World Map. In The Origin of
Goods: A Conceptual and Empirical Assessment of Rules of Origin in PTAs,
edited by Olivier Cadot and others. Oxford University Press and Centre for
Economic Policy Research.
. 2005b. What Are the Effects of Rules of Origin on Trade? Washington:
Inter-American Development Bank. Mimeographed.
Flatters, Frank, and Robert Kirk. 2005 (forthcoming). Rules of Origin as Tools of
Development? Some Lessons from SADC. In The Origin of Goods: A Conceptual and Empirical Assessment of Rules of Origin in PTAs, edited by Olivier Cadot
and others. Oxford University Press and Centre for Economic Policy Research.
Garay, Luis, and Rafael Cornejo. 2002. Metodologa para el anlisis de regimenes de origen: aplicacin en el caso de las Amricas. INTAL-ITD-STA
working paper 8. Washington: Inter-American Development Bank.
Graham, Edward M., and Christopher Wilkie. 1998. Regional Economic Agreements and Multinational Firms: The Investment Provisions of the NAFTA. In
Global Competitive Strategies in the New World Economy, edited by Hafiz
Mirza. Cheltenham: Edward Elgar.
Grossman, Gene M., and Elhanan Helpman. 1995. The Politics of Free Trade
Agreements. American Economic Review 85(4): 66790.
Hirsch, Moshe. 2002. International Trade Law, Political Economy and Rules of
Origin: A Plea for a Reform of the WTO Regime on Rules of Origin. Journal
of World Trade 36(2): 17189.
Inama, Stefano. 2004. Rules of Origin in GSP and ACP. United Nations Conference on Trade and Development. Mimeographed.
James, William. 2004. Rules of Origin, Tariff Discrimination and Trade Diversion: A Case Study of Asian Textiles and Apparel Exports. Paper prepared for
the conference Rules of Origin in Regional Trade Agreements: Conceptual
and Empirical Approaches. Inter-American Development Bank (Integration
and Regional Programs Department), Centre for Economic Policy Research,
and DELTA/INRA, Washington, 2021 February.
Jensen-Moran, Jeri. 1996. Trade Battles as Investment Wars: The Coming Rules
of Origin Debate. Washington Quarterly 19(1): 23953.
Ju, Jiandong, and Kala Krishna. 1998. Firm Behavior and Market Access in a Free
Trade Area with Rules of Origin. Working paper 6857. Cambridge, Mass.:
National Bureau of Economic Research.
Krishna, Kala, and Anne O. Kruger. 1995. Implementing Free Trade Areas:
Rules of Origin and Hidden Protection. In New Directions in Trade Theory,
edited by Alan Deardorff, James Levinsohn, and Robert Stern. University of
Michigan Press.
Krueger, Anne O. 1993. Free Trade Agreements as Protectionist Devices: Rules
of Origin. Working paper 4352. Cambridge, Mass.: National Bureau of Economic Research.
103
Reyna, Jimmie V. 1995. Passport to North American Trade: Rules of Origin and
Customs Procedures under NAFTA. Colorado Springs: Shepards/McGraw-Hill.
Rodriguez, Peter. 2001. Rules of Origin with Multistage Production. World
Economy 24(1): 21020.
Sanguinetti, Pablo, and Eduardo Bianchi. 2005 (forthcoming). Implementing
Preferential Trade Agreements in the Southern Cone Region of Latin America:
Rules of Origin. In The Origin of Goods: A Conceptual and Empirical Assessment of Rules of Origin in PTAs, edited by Olivier Cadot and others. Oxford
University Press and Centre for Economic Policy Research.
Suominen, Kati. 2003. Selective Liberalization in Response to Globalization: Rules
of Origin as Determinants of Market Access Provisions in PTAs. Integration and
Trade 19(7): 15385.
. 2004. Rules of Origin in Global Commerce. Ph.D. Dissertation, University of California at San Diego.
Wonnacott, Paul. 1996. Beyond NAFTAThe Design of a Free Trade Agreement of the Americas. In The Economics of Preferential Trading Agreements,
edited by J. Bhagwati and A. Panagariya, pp. 79107. Washington: AEI Press.
WTO (World Trade Organization). 2002a. Coverage, Liberalization Process and
Transitional Provisions in Regional Trade Agreements: Background Survey by
the Secretariat. Working document WT/REG/W/46. Geneva: World Trade
Organization, Committee on Regional Trade Agreements.
. 2002b. Rules-of-Origin Regimes in Regional Trade Agreements: Background Survey by the Secretariat. Working document WT/REG/W/45. Geneva:
World Trade Organization, Committee on Regional Trade Agreements.
GIOVANNI MAJNONI
ANDREW POWELL
Reforming Bank Capital Requirements:

Implications of Basel II for
Latin American Countries
he appropriate regulation of banks is a hotly contested topic in both
industrialized and developing countries. This year the Basel Committee on Banking Supervision put forward a controversial proposal to
overhaul the 1988 Basel Accord that has long guided the regulation of bank
capital across over a hundred countries.1 The thirteen member countries of
the Basel Committee on Banking Supervision are due to apply Basel II, as
the new accord is called, by 2007. If and how other countries should apply
Basel IIand therefore whether the new Accord will be successful as a
standardremains an open question. If Basel II is applied across the globe,
then its details will be extremely relevant; if not, it will be important to understand why many emerging countries decided to retain Basel I despite its
well-known drawbacks.2
Majnoni is with the World Bank; Powell is with the Universidad Torcuato Di Tella. This
paper is part of a project on Credit Information, Credit Risk Measurement, and Solvency
Ratios in Emerging Countries supported by a grant from the World Bank Research Committee. We thank Margaret Miller and Nataliya Mylenko for their collaboration on this project. Andrew Powell gratefully acknowledges the World Bank, which he visited between
August and December 2003, for financial support.
We are grateful to Matas Gutierrez from the Central Bank of Argentina, Jos Luis
Negrn, Javier Mrquez, and Alberto Romero from the Bank of Mexico, and Ricardo
Schectman from the Central Bank of Brazil for invaluable help in the quantitative analysis.
We thank Jerry Caprio, Mark Carey, Patricia Correa, Charles Goodhart, Michael Gordy,
Patrick Honohan, Patricia Jackson, Rafael Repullo, Roberto Steiner, Kostas Tsatsaronis,
and Andrs Velasco for very helpful comments.
1. See the Financial Stability Forum website on financial standards (www.fsforum.org).
2. Below we discuss the possibility that too many countries will adopt Basel II, in
which case its role as a standard in creating peer group pressure would have been too great!
105
It is widely accepted that bank capital should be regulated, but how to do

so remains open to debate.3 The simple approach of Basel I divides assets
into very broad risk categories and establishes an 8 percent minimum capital requirement for risky assets.4 However, as bank risk management has
become more sophisticated and as the possibilities for transforming asset
risk have grown, the potential distortions created by these simple rules and
the opportunities for arbitraging across them have multiplied.5 By contrast,
Basel II goes well beyond simply recasting quantitative requirements.
Making capital requirements more risk sensitive and reducing regulatory
arbitrage are main objectives of the new accord.6 Basel II proposes two
basic approaches: the standardized approach, which uses external credit
rating agencies together with a table that maps those ratings directly into
capital requirements; and the internal ratings-based (IRB) approach, in
which the banks themselves estimate their customers default probability
without relying on external rating agenciesand then use a particular formula specified in Basel II to determine capital requirements as a function of
the default probability and other parameters.7
This paper focuses on one specific but critical issue and on a set of more
general questions. We analyze whether the IRB approach as calibrated is
appropriate for the Latin American context. We believe that this is the first
paper to estimate credit risk across a set of emerging economies using a simple and homogeneous methodology. We find significant differences between
our estimates from the region and those from the Group of Ten (G10) coun3. Supporters see bank capital regulation as a response to the moral hazard of an
inevitable public safety net for banks (see, for example, Mishkin, 2001; Goodhart and
others, 1999) or as part of a scheme to emulate the incentives for owners and managers in
firms where debt holders are more sophisticated than bank depositors (see Dewatripont and
Tirole, 1994, for their representation hypothesis). Members of the free-banking school disagree with bank capital regulation (for excellent reviews, see Freixas and Rochet, 1999,
pp. 26065; Berger, Herring, and Szego, 1995).
4. Lower risk categories include mortgages, contingent facilities, short-term loans to
other banks, and lending to members of the Organization for Economic Cooperation and
Development (OECD). See the original Basel Committee on Banking Supervision documents and the literally hundreds of comments on Basel II at www.bis.org.
5. The standard criticism is that banks have incentives to sell or securitize assets for
which capital requirements do not bind and buy assets when requirements would bind. In
this way, banks would transform the risk on their balance sheets to ensure that capital
requirements were always binding.
6. The new Accord has three pillars: quantitative requirements, supervisory review,
and market discipline. While we focus on the first pillar, we briefly discuss the other pillars
in the next section.
7. See Basel Committee on Banking Supervision (2003).
Giovanni Majnoni and Andrew Powell
107
tries. These differences have strong implications for the application of the
IRB approach.
We also discuss Basel II implementation for Latin America and more
generally across all emerging countries. Typically there are few external
rating agencies in these countries, so the standardized approach would have
little effect in linking regulatory capital to risk.8 But the IRB approach may
not be calibrated appropriately for emerging markets, and its implementation and supervision may stretch limited supervisory resources. Given the
data on compliance with the Basel Core Principles for Effective Banking
Supervision across Latin America, it may not be advisable for many countries to implement the IRB approach for a long time.
Given this situation, we suggest an innovative simplification of the IRB
approach that builds on current policies regarding provisioning in some
emerging countries and that may be used as a transition arrangement
toward the IRB approach. We call this the centralized ratings-based (CRB)
approach. Under this approach, banks would rate their clients, but the regulator would determine the rating scale and the way in which the banks
ratings map into default probabilities. The use of a centralized scale would
facilitate comparison across banks and greatly ease the monitoring of
banks ratings.9 Those requirements would also be easier to monitor, since
the regulator would determine how banks ratings would feed into capital
requirements.
Countries must choose whether to stay on Basel I or, if not, which Basel
II alternative to apply (here we include our proposed CRB approach). To
date there is little guidance on this important decision. We therefore develop
a Basel II decision tree to assist countries deciding whether to adopt Basel
II and, if so, how. Our broad advice is that many countries should stay on
Basel I or only adopt Basel II for a subset of banks at least for several years
beyond 2007. Regulators should not move to complex rules too quickly sim-
8. This may also be the case for smaller and regional banks in G10 countries. These
banks are unlikely to be systemic, however, whereas systemic banks in emerging countries
will typically have mostly nonrated assets. See Ferri, Liu, and Majnoni (2001) for a discussion on the global pattern of ratings.
9. Bank ratings could be compared directly in the case of banks lending to the same
corporate client. Bank ratings for similar types of loans (to companies in the same economic
sector, business line, or region) could also be compared and outliers investigated. Some G10
regulators informally acknowledge that even where the IRB approach is likely to be employed,
supervisors will compare banks internal ratings of important corporate clients (as they do
today) and for that purpose will no doubt map ratings into a centralized scale.
ply because of peer group pressure or pressure from the large international
banks. We also argue that countries in the region should seriously consider
the CRB approach, and we suggest ways in which this may be made compatible with Basel II for the purpose of assessing standards.10
The paper proceeds as follows. In the next section we provide a highly
synthetic account of the new Accord. The paper then introduces our methodology for testing the calibration of the proposed requirements for the IRB
approach and also presents and discusses these results. We go on to consider broader questions regarding Basel II implementation. The closing
section concludes with a discussion of policy.
Basel II: A Synthetic Account

At first sight, the Basel II documentation is daunting. While the Accord
itself is less than 300 pages, fully understanding those 300 pages requires
studying several hundred pages of supporting documents. The Accord
specifies a set of new alternative approaches for minimum capital requirements (Pillar 1: Quantitative Requirements), states how those requirements
should be supervised (Pillar 2: Supervisory Review), and finally defines what
banks should reveal to the market regarding the risk of their assets and how
(including whether) they satisfy regulatory requirements (Pillar 3: Market
Discipline). The idea is that the three pillars are complementary and mutually reinforcing.11
The Basel Core Principles for Effective Banking Supervision already
encapsulate Basel I and much of the second pillar of Basel II.12 Moreover,
for the simpler Basel II approaches, the third pillar (on market discipline)
10. Countries may be concerned about how banking regulations and supervision will be
assessed by the International Monetary Fund and the World Bank, in the context of the
Financial Sector Assessment Programs (FSAPs).
11. The definition of capital has not changed from Basel I, but the new Accord includes
important changes in the level of consolidation that banking supervisors should apply
(scope of application) and for the first time introduces rules on lending to affiliated companies (related lending).
12. The Core Principles refer to supervision. They cover what banks should report to the
supervisor, but not what banks should disclose to the market (Basel II, pillar 3). Strictly
speaking, following Basel I is neither necessary nor sufficient for a country to be compliant
with the sixth Core Principle (on capital adequacy). In practice, however, Basel I is normally considered a necessary condition, and the Financial Stability Forum deems it one of
the critical financial standards that countries should implement (see www.fsforum.org).
109
is reduced to the banks obligation to publish its capital requirement and

its actual level of capital. Thus what is really new in Basel II are the various Pillar 1 alternatives regarding actual capital requirements. We therefore focus on the first pillar in this paper.13
Pillar 1 contains three main approaches: the simplified standardized
approach, the standardized approach, and the internal ratings-based (IRB)
approach, which breaks down into two options (the foundation IRB and
the advanced IRB approaches). Pillar 1 also covers alternatives for the measurement of basic credit risk, credit risk mitigation techniques, securitization risk, and operational risk.14 With regard to the first three alternatives,
Basel II attempts to improve on the treatment in Basel I; in the case of the
fourth alternative, this is the first time that an actual quantitative requirement for operational risk has been included in the Basel recommendations.
The alternatives are illustrated in table 1. In practice, a relatively simple
approach for underlying credit risk assessment would normally be combined with simple approaches for the other topics. The simplified standardized approach, for example, explicitly combines the simplest approaches
for credit risk evaluation, credit risk mitigation, securitization risk, and
operational risk. At the other extreme, the advanced internal ratings-based
approach would normally be accompanied by advanced approaches elsewhere, in particular the advanced measurement approach for operational risk.
One potentially important issue for emerging economies is that the new
capital requirements are calibrated so that, on average, the capital requirement for a standard G10 bank would remain around 8 percent under the
standardized approach, and capital requirements for an average G10 bank
would fall under the IRB approach. This implies that under the standardized approach, the increase in the new requirement for operational risk
would be roughly offset by the reduction in requirements for credit risk,
given the ratings distribution in a typical G10 country. In the case of a
developing country with low ratings penetration, the proportion of unrated
claims on bank balance sheets is likely to be much higher than for G10
13. This does not imply that the second pillar is unimportant. Compliance with the Core
Principles is weak in developing countries, and the second pillars tighter definitions on
aspects of the supervisory process highlight the importance of making progress in these areas.
14. Credit risk mitigation techniques mainly refer to contracts that use securities as
guarantees, such as repurchase agreements (repos) and credit derivatives; they do not refer
to real guarantees such as mortgages, for which there are rules under basic credit risk evaluation. Securitization risk covers both investment in a securitized instrument and the
retained risk of originating a securitization of assets on a bank portfolio.

T A B L E 1 . Different Options Proposed in the First Pillar of Basel II
Approach
Basic credit risk

measurement technique
Credit risk mitigation
Securitization risks
Operational risk
SSA banks can only

invest (cannot offer
enhancements or
liquidity facilities).
Risk weight = 100
percent.
Standardized: uses
export credit agency
ratings (only
investing banks can
use below BB+)
Basic indicator:
Capital = 15% gross
income
Simplified
standardized
Export credit agencies

Simple: risk weight of
(www.oecd.org,
collateral substitutes
trade directorate, ECA
that of claim
page)
Standardized
Export credit agencies or

credit rating agencies
(such as S&P,
Moodys, Fitch)
Simple: same as simplified

standardized approach.
Comprehensive: exposure
amount reduced
subject to claim and
collateral haircuts.
Internal
ratings-based
Foundation
Banks internal ratings

for default
probability and Basel
II formula sets capital
requirement (loss
given default 45% for
senior and 75% for
subord).
Comprehensive, loss given

default adjusted given
reduction in exposure
and capital
requirement given by
Basel formula.
IRB approach: Investing

banks may use bank
ratings according to a
standard scale.
Originators may use
supervisory formula.
Advanced
Banks set internal rating

(default probability),
Loss given default
exposure at default
and maturity. Capital
requirement still
given by Basel
formula.
Own model determines

Loss given default and
exposure at default;
capital requirement
given by formula.
As for Foundation IRB

approach.
Basic indicator, or
standardized
approach where
bank capital =
weighted sum of
gross income
across activities.
More sophisticated
banks will be
expected to
graduate to the
advanced
measurement
approach where
capital requirement is given by
own risk measurement system.
As for Foundation IRB
approach.
banks, and the distribution of rated claims will probably also be different.15
This implies that the operational risk requirement may approximate a simple add-on, increasing overall capital requirements. While this may not be
undesirable, it may be an impediment to the implementation of Basel II in
some countries.
15. Local supervisors may also employ local ratings (that is, ratings conducted
according to a national or local scale rather than an international scale). However, the three
major rating agencies (Fitch, Moodys, and Standard and Poors) all warn customers that
local ratings are not necessarily comparable across countries. This raises an important issue
for the use of the standardized approach as a financial standard.
111
This paper focuses on Pillar 1namely, the measurement of underlying

credit risk. Given the fundamental problem of asymmetric information
between the regulator and the regulated institution, banks will generally
have superior measures of clients risks than will the regulator. However,
if the underlying motive for regulation is moral hazard, then it is clearly
problematic to allow the bank itself to use its own estimates of client risk
or its own assessment of a portfolio of such risks.16 The solution to this
conundrum in the new Accord is to use either external credit rating assessments or bank ratings subject to supervision of the rating methodology and
a specific formula that maps those ratings into capital requirements.
Since rating penetration is typically low in developing countries, the
standardized approach will buy very little in terms of linking regulatory
capital to risk. This argument implies that many regulators may well be
interested in considering the IRB approach or our proposed CRB approach.
Countries may also adopt a mixed approach whereby some banks remain
on Basel I or adopt the standardized approach while more sophisticated or
larger banks adopt the IRB approach. The United States may provide
something of a model in which very few banks will be forced to adopt the
IRB approach, some others may be permitted to adopt the IRB model, and
the vast majority of banks will remain on Basel I. When we translate the
U.S. model to an emerging market, we find that if the regulators aim is for
regulatory capital in the financial system as a whole to reflect risk more
closely, then the proportion of assets that will be covered by the IRB
approach will be relatively large (that is, the banks that adopt the IRB model
are likely to be few in number but large in size). These arguments suggest
that the calibration of the IRB approach is an important issue for emerging
economies considering whether and how to adopt Basel II. This is the topic
of the next section.
On Basel II Calibration: Methodology

The new capital Accords internal ratings-based approach suggests a formula for calculating a banks capital requirement as a function of three
basic variables: default probability, exposure at default, and loss given
16. Assessing credit portfolio risk implies assumptions not only on individual default
probabilities, but also on the multivariate distribution of those default probabilities. For simpler assumptions on distributions, this implies assumptions on the mean and variancecovariance matrix of default probabilities.
default.17 The expected loss for a single claim is the multiplication of the
three variables if expressed in appropriate units. However, the provisions
a bank holds are typically identified with expected losses, and capital is
identified with a value-at-risk (VaR) concept. A regulator might then ask
a bank to hold provisions and capital to cover a specified percentage of the
distribution of losses to ensure the continued solvency of the bank except
in highly extraordinary circumstances.
This is illustrated in figure 1, where the segment OB represents the total
value at risk associated with a specific probability of occurrence, that is,
the level of losses that corresponds to a certain percentile of the distribution function of expected losses (99 percent in the figure). A regulator
might then ask a bank to hold general provisions to cover the expected loss
represented by the segment OA, where A is the mean of the loss distribution, and to hold capital to cover the unexpected loss represented by the
segment AB (that is, the difference between the total value at risk defined
with respect to a particular percentile of the distribution and the mean or
expected loss). The calibration of the Basel II IRB formula employs a value
at risk of 99.9 percent with a horizon of one year. Hence a bank is only
expected to use up its capital in one year with a probability of 0.1 percent,
or once every 1,000 years.18
The inputs for the Basel II IRB formula are the parameters for a single
loan or claim, and the output is the capital requirement for that instrument. Subject to an underlying assumption regarding the correlation of
asset risks, this single-instrument approach should approximate the result
of a portfolio credit risk model. An econometric methodology is typically
employed to estimate individual instrument default probabilities, and these
estimates are then fed into a model with other parameter estimates to obtain
the loss distribution curve for the portfolio.19 Simplifying assumptions are
employed in both estimating the parameters and developing the model.
Commonly used models include Moodys KMV option-based model,
the McKinsey macroeconomic simulation model, the CreditMetrics model
17. See Altman and Saunders (1997) for a useful discussion.
18. This assumes that draws from the distribution are independent over time.
19. This may be for a particular bank or a specific business line of a bank. As we directly
mimic the portfolio of a bank, we do not discuss further the important issue of aggregation.
Suffice to say that the Basel II IRB formula is by business line (sovereign, commercial, and
retail) and is calibrated with particular assumptions regarding asset correlations in each sector. The results are simply added, implying an assumed perfect positive correlation between
business lines.
113
Probability
Figure 1. Loan Loss Probability Distribution Stemming from Credit Risk
General
provisions
Capital
99th percentile
Expected
losses
O
Unexpected
losses
A
Loss
from J. P. Morgans RiskMetrics division, and the Credit Suisse First

Bostons CreditRisk+ model. The latter of these models is arguably the
simplest to implement, but even here implementation relies on key additional assumptions such as the number of risk factors, the estimation of
factor volatilities and loadings, and the correlation of default probabilities.20 Moreover, the estimation of each model relies on a set of quite specific data requirements and assumptions that make cross-country and even
cross-institution comparisons problematic. These model-based methodologies of estimating the credit loss distributions of a loan portfolio are thus
subject to both estimation risk of the parameters of a single instrument and
model risk (in that the assumptions of the portfolio model may be incorrect). Furthermore, the Basel II IRB formula introduces approximation
risk. Capital requirements are calculated on each single instrument model
and then they are simply added across all instruments. The aggregated single instrument formula yields the appropriate loss distribution for the portfolio only for a particular correlation of risks between instruments. If
actual correlations differ substantially from this assumption, then this
approximation to the risk of a portfolio may cease to be valid.
20. See Balzarotti, Falkenheim, and Powell (2002) and Balzarotti, Castro, and Powell
(2004) on the implementation of CreditRisk+ in Argentina. See Mrquez and others (2003)
for the case of Mexico and Foglia (2003) for a discussion and model-based estimates of
credit risk using Italian credit registry data.
In this paper we take a quite different approach: we adopt a bootstrapping technique that essentially enables us to mimic the shape of the loss
distribution function of any specified loan portfolio.21 This approach minimizes the impact of estimation errors and maximizes the degree of comparability across countries. Even without an underlying model, bootstrapping
techniques can be used to simulate the frequency distribution of credit
losses. The resampling approach is very flexible and lends itself to many
alternative simulation exercises aimed, for example, at measuring the exposure to credit losses of portfolios characterized by different loan sizes,
maturities, ratings, geographic locations, or economic sectors.
The empirical exercises performed in this paper are for Argentina, Brazil,
and Mexico. In each of the three countries, the central bank maintains a
public credit registry that contains information on a very large number of
loans in the financial system.22 Each financial system requires a clear
amount of capital plus provisions every year to confront total credit losses.
However, that observation tells us little about the required capital and provisions for an average bank in that year. Conditional on the overall macroeconomic conditions, the losses suffered by an average bank depend on
the sensitivity of the banks loan portfolio performance to the prevailing
economic conditions and the idiosyncratic risk of the portfolio. The Basel
II IRB approach assumes that the correlation structure of a bank portfolio
is known and summarizes credit loss correlations as sensitivities to a single factor, but credit risk correlations are not known with certainty and a
single factor model can at best be thought of as an approximation to a more
complex reality.
The technique we employ generates conditional loss distribution functions based on overall economic performance, the correlation of credit losses,
and any residual idiosyncratic risks in a large number of sample portfolios.
We then use these distributions to measure the expected and unexpected
losses. In other words, conditional on the overall performance of the financial system over the period of analysis, our results provide a measure of
the level of expected and unexpected losses of a bank of average size with
21. Here we are following Carey (2002). Also see Carey (1998) for further analysis of
credit risk in G10 portfolios.
22. See Miller (2003) for details on public credit registries around the world; see Powell
and others (2004) for an empirical analysis of the value of public credit registries in
Argentina, Brazil, and Mexico and a discussion of their use for predicting credit losses.
115
a loan portfolio randomly drawn from the universe of loans within the
financial system.23
We limit our empirical analysis to only one specific year owing to changes
in definitions, the scope of coverage, and data quality across the credit registers from the three countries. Our findings should therefore be regarded
as illustrative of a methodology that needs to be repeated over several
years to achieve its full empirical relevance. This point is shown in greater
detail in figure 2, which shows a sequence of conditional distributions estimated at different points in time (namely, t1, t2, and t3, which represent a
sequence of good-bad-good years over a hypothetical economic cycle) and
the unconditional distribution resulting from pooling the data from all the
conditional distributions. Our estimates reflect the events of the chosen
year and thus cannot be taken as representative of the unconditional distribution. However, we chose a year close to the cyclical trough for each
country (the period t2 in the characterization of figure 2), so our estimates
may properly reflect those observations that carry a greater weight in shaping the right tail of the unconditional distribution of credit losses. For
instance, the Argentine data are for 2001. In that year, a recession led to a
fall of GDP equal to 4.4 percent in real terms and a deepening crisis. Bank
deposits were frozen in December 2001, and there was considerable economic and political uncertainty that resulted in the removal of the president amid riots. Over the same period, Mexico experienced a stagnation of
economic activity with zero GDP growth and a reduction of the ratio of
bank credit to GDP to 11.9 percent, the lowest value of the last decade. In
Brazil, a slowdown in economic activity brought GDP growth down to
only 1.3 percent and led to a contraction of bank credit in real terms. To
summarize, while the results naturally reflect a period in time, the snapshot
captures economic stress in all three countries.
23. Our one-point-in-time distributions might be thought of as distributions across idiosyncratic risk or, alternatively, of distributions of correlations of asset risk within our sample portfolios. For a one-factor model, the systemic risk of a portfolio might be approximated
by average correlations as portfolio size increases. In practice, however, asset correlations
may differ substantially across bank portfolios if asset correlations depend on many factors,
including sector, loan, and borrower characteristics, and if portfolios are lumpy in terms of
their exposures across these factors. Indeed, one common explanation of why one bank may
fail during a recession whereas another does not is based on differences in exposure to systemic factors rather than pure idiosyncratic risk. The Basel II IRB formula assumes that
there is a single systemic factor, that bank portfolios have zero idiosyncratic risk, and that
asset correlations are identical for companies of the same size (and decrease with company
size), but correlations are always assumed to be known and stable.

Figure 2. Conditional and Unconditional Credit Loss Distributions
Conditional
Unconditional
Losses
(%)
UL2
UL
UL3
UL1
EL2
EL
EL3
EL1
t1
t2
t3
Time
Frequency
The first step of the procedure consists of extracting from the public
credit registry a large pool of performing loans to the nonfinancial corporate sector at a particular date. This pool reflects the overall risks of lending to the corporate sector in each particular country. Second, we define
default as the event of payments that are over ninety days past due.24 Third,
we classify loans into two categories according to whether they maintain
their initial status or default over the following twelve months. Fourth,
from this pool of loans, we randomly sample a predefined number of loans
(in our case 500), intended to mimic the loan portfolio of a medium-sized
bank.25 Given a predefined recovery ratio, we compute the value of the
losses of the sampled portfolio, expressing this as a fraction of the face
value.26 Fifth, we replicate the last step a large number of times (20,000 in
24. This follows the typical definition of a nonperforming loan according to international best practices and to one of the criteria set out by the Basel Committee on Banking
Supervision.
25. Modifying the sampling procedures would enable the selection of predefined risk
profiles and the analysis of different risks embedded in bank portfolios (see Carey, 2002).
26. We assume a predefined recovery ratio of 50 percent of the face value of a defaulted
loan. The Basel Committee on Banking Supervision employs a loss given default of 45 percent for the foundation IRB approach. Our conversations with the central banks indicated
that this might be somewhat low for emerging markets, so we selected the figure of 50 percent
for the purposes of our calibration exercises. The bootstrapping methodology would be considerably more precise if data on loss given default were available at the individual loan level.
117
this case) to generate a frequency distribution of credit losses. This frequency distribution simulates the actual distribution of credit losses faced
by banks in that financial system at that time. Finally, we use the frequency
distribution of credit losses to calculate a number of statistics in a reasonably homogeneous way across countries. In particular, we calculate the
expected and unexpected losses up to different statistical tolerance values
across the different portfolios. We compare these estimates in each country against current regulations (inspired by Basel I), the actual provisions
and capital of banks, and a simulated capital requirement using estimated
default probabilities and following the Basel II IRB formula.
Having described the methodology, we reiterate the caveat that the
results reflect a snapshot of a particular country in a particular year. The
results cannot and should not be interpreted as average values representing credit risk exposure over different time horizons or over the full business cycle. Moreover, the results are dependent on the universe of loans
collected by the public credit registry. For Argentina and Mexico, we are
confident that this universe of commercial loans is representative of the
financial system as a whole, but in the case of Brazil, we could only access
the universe of larger corporate loans, for reasons explained below.27 Our
results should thus be interpreted as tentative and conditional on the time
and loan universes obtained. Nevertheless, they are highly suggestive, tend to
back up other evidence, and could provide the catalyst for similar studies
in other countries or in the same countries over longer time periods.
Calibrating Basel II for Emerging Countries: The Results

This section first discusses the data that we use for the analysis and then
details the results of the bootstrapping sampling methodology described
above. Finally, we draw the main implications of our results for emerging
economies.
Description of the Data

Figures 3 to 5 illustrate, for the three countries, the frequency distribution
of three variables involved in the experiment. Panel A reports the frequency
distribution of the size of individual loans, which are extracted from the pub27. The Basel II formula was calibrated on commercial loans, so we feel that this choice
is appropriate.

Figure 3. Argentina: Relevant Frequency Distributions from the Resampling Exercise
A. Frequency distribution of bank loan size
Frequency (percent)
60
50
40
30
20
10
21
71
2,4
21
2,2
71
2,1
21
1,9
71
1,8
21
1,6
71
1,5
21
1,3
71
1,2
1,0
92
77
62
47
32
17
21
Exposure size (thousands of U.S. dollars)
73
11
6
15
9
20
2
24
5
28
9
33
2
37
5
41
8
46
2
50
5
54
8
59
1
63
4
67
8
72
1
76
4
80
7
85
1
89
4
93
7
Frequency (percent)
B. Frequency distribution of 20,000 sampled portfolios

4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
Portfolio size (millions of U.S. dollars)
9
0.1
0
0.1
2
0.1
3
0.1
5
0.1
6
0.1
8
0.1
9
0.2
0
0.2
2
0.2
3
0.2
5
0.2
6
0.2
8
0.2
9
0.0
0.0
6
0.0
5
0.0
0.0
0
0.0
0.0
Frequency (percent)
C. Frequency distribution of credit losses

4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
Portfolio losses (percent)
Source: Central Bank of the Argentine Republic.

Figure 4. Brazil: Relevant Frequency Distributions from the Resampling Exercise
A. Frequency distribution of bank loan size
Frequency (percent)
30
25
20
15
10
5
1
59
2
70
3
81
4
92
5
1,0
36
1,1
47
1,2
58
1,3
69
1,4
80
1,5
91
1,7
02
1,8
13
48
37
25
14
37

B. Frequency distribution of 20,000 sampled portfolios
Frequency (percent)
3. 0
2. 5
2. 0
1. 5
1. 0
0. 5
27
5
33
4
39
2
45
1
50
9
56
8
62
6
68
5
74
3
80
2
86
0
91
9
97
7
1,0
36
1,0
94
1,1
53
1,2
11
1,2
70
0. 0
Portfolio size (milions of U.S. dollars)
Source: Central Bank of Brazil.
.39
.48
16
.56
15
.64
14
.72
13
.80
12
11
6
.88
10
9.9
9.0
8.1
7.2
6.2
5.3
4.4
3.5
2.6
0.7
3.0
2.5
2.0
1.5
1.0
0.5
0.0
1.7
Frequency (percent)
C. Frequency distribution of credit losses
119

Figure 5. Mexico: Relevant Frequency Distributions from the Resampling Exercise
0
89
8
99
6
1,0
94
1,1
92
1,2
90
1,3
88
1,4
86
1,5
84
80
70
60
50
40
4
21
2
31
0
16
11
Frequency (percent)
A. Frequency distribution of bank loan size: Large banks

45
40
35
30
25
20
15
10
5
0
29
1,1
41
60
1,0
.5
38
99
.9
92
85
78
71
64
57
50
9
43
1
37
30
23
16
95
26
Frequency (percent)
B. Frequency distribution of 20,000 sampled portfolios: Large banks

40
35
30
25
20
15
10
5
0
Portfolio size (millions of U.S. dollars)

Source: Bank of Mexico.
.0
35
.4
33
.9
30
.3
.8
28
25
.3
23
.7
20
.6
.2
18
15
.1
13
.6
10
8.0
5.5
3.0
0.4
Frequency (percent)
C. Frequency distribution of credit losses: Large banks

40
35
30
25
20
15
10
5
0
121
lic credit registries and which represent the universe from which our samples of 20,000 loans are drawn; panel B shows the frequency distribution
of the dollar value of the 20,000 randomly selected portfolios; and panel
C shows the distribution of credit losses of the 20,000 randomly selected
portfolios as a fraction of the face value of the respective portfolios.
A visual inspection of the charts shows clear differences among the
three samples. Mexico has the lowest concentration in terms of loan size,
and about 80 per cent of all the loans extracted from the credit register are
smaller than U.S.$100,000. The same figure for Argentina is about 60 percent. We adopted a different sampling procedure for Brazil because of the
huge size of the credit market and following advice from the Central Bank.
Specifically, we included only those companies whose gross exposures
with the financial system were above U.S.$300,000, and hence the only
smaller loans included are those to companies that had other larger loans
outstanding (about 40 percent of loans were less than U.S.$100,000). We
were advised that this sampling methodology would capture the major
credit risks in the Brazilian financial system. The charts for Argentina and
Mexico thus show considerably more skewed distributions for both the
value and credit losses of the portfolios than the comparable distribution
computed for Brazil. As a supplement to the visual information provided
by the charts, table 2 summarizes a set of descriptive statistics of the distribution of the 20,000 randomly sampled portfolios.
T A B L E 2 . Simulated Loan Portfolios: Descriptive Statisticsa

Millions of U.S. dollars
Country and period
Argentina (Dec 2000 to Dec 2001)
Brazil (Oct 2001 to Oct 2002)
Mexico (Dec 2000 to Dec 2001)
No.
observations
Mean
Median
Mode
Standard
deviation
Minimum
Maximum
70,017
41,784
188,165
242
551
85
215
538
62
182
510
33
104
110
89
72
275
16
943
1,306
1,477

a. Data refer to the twelve-month period indicated for each country. The number of observations are the number of bank loans to
nonfinancial entities above a minimum amount, which were extracted from the national credit registers of each country at the beginning of the twelve-month period. The criteria underlying the selection of loans from the credit registers differ slightly across countries.
For Argentina and Mexico, a loan refers to the overall position of a single borrower with the banking system as a whole. For Argentina,
the positions selected are those equal to or larger than U.S.$21,000. For Mexico, we include both loans equal to or larger than
U.S.$20,000 (which are reported on a compulsory basis) and smaller loans that are reported on a voluntary basis. For Brazil, the minimum size is U.S.$300,000, but positions with different banks that concur to define the total exposure are treated as distinct individual
loans. The descriptive statistics (mean, median, mode, standard deviation, minimum, and maximum) refer to the distribution of the
value of the 20,000 portfolio of 500 loans each, randomly sampled from the pool of loans described above. Exact sources and definitions
of each variable can be found in the main text.

Table 3. Capital and Provisions: Unexpected and Expected Losses Estimations Based on
Simulation Resultsa
Percent
Country and period

United States (198991)b
United States (1929)b
Unexpected loss
Default
probability
Expected
loss
95%
99%
99.9%
Basel (Jan 2004)
9.60
8.32
2.70
3.00
6.24
4.80
4.16
1.35
1.50
3.12
7.30
3.51
4.44
1.62
2.54
14.80
6.07
16.58
2.55
3.80
21.80
10.46
31.64
3.91
5.36
14.93
14.15
9.68
10.07
12.70
Source: Authors calculations and Carey (2002).

a. Expected loss is given by the mean value of the simulated distribution of credit losses. Simulations are based on the random
extraction without replacement of 500 loans from the pool of loans registered in the credit register of each country, to simulate a standard bank loan portfolio. The extraction is repeated 20,000 times (this time with replacement) to obtain 20,000 portfolios. The distribution of credit losses for each portfolio provides the 20,000 observations used to simulate the distribution of credit losses. Unexpected
losses at different levels of probability represent the value of credit losses (as a percentage of the face value of the portfolio) corresponding to the percentile on the right tail of the distribution minus the expected loss given by the mean value of the distribution. Basel
unexpected losses indicates the value of unexpected losses computed according to the algorithm proposed by the Basel Committee on
Banking Supervision (2004b). The value of default probabilities used in the algorithm is given by twice the value of the expected loss,
assuming the same 50 percent loss given default used in the simulation exercise.
b. Results for the United States are from Carey (2002), who simulates credit losses based on a loan portfolio that mimics the risk
exposure of a medium-sized bank in the United States and the default probability values observed at times of systemic distress, such as
the moderate recession of 198991 and the severe 1929 recession.
The Main Results

Table 3 presents the main results of the bootstrapping exercise for each
country, including the value of expected losses and the value of unexpected losses associated with different percentile levels of the right tail of
the simulated distribution of credit losses. Here we refer to the total losses
(equal to the sum of the expected and unexpected components) as the value
at risk. For the three countries considered, expected losses proved to be in
an interval between 1 percent and 5 percent. If we assume provisions to cover
this amount, the amount of capital necessary to provide protection for unexpected losses up to 99 percent of the distribution is about 15 percent for
both Argentina and Mexico. The capital that would be required to cover
99.9 percent of the distribution is significantly higher, at 21 percent for
Argentina and 31 percent for Mexico. In the case of Brazil, the results suggest that capital of just over 6 percent is required for the 99 percent confidence limit and 10.5 percent for the 99.9 percent confidence value.28
28. These lower values may reflect the fact that the loan size for Brazil is significantly
greater than that for Argentina and Mexico.
123
Table 4. Capital and Provisions in Argentina: Unexpected and Expected Losses Estimations
Based on Loan Universes with Different Minimum Exposure Thresholds
Percent
Minimum size of individual
borrowers loan exposure
U.S.$20,000
U.S.$300,000
Unexpected loss
Default
probability
Expected
loss
95%
99%
99.9%
9.6
8.2
4.8
4.1
7.3
3.6
14.8
5.6
21.8
8.0

a. Expected loss is given by the mean value of the simulated distribution of credit losses. Simulations are based on the random
extraction without replacement of 500 loans from the pool of loans registered in the credit register of each country, to simulate a standard bank loan portfolio. The extraction is repeated 20,000 times (this time with replacement) to obtain 20,000 portfolios. The distribution of credit losses for each portfolio provides the 20,000 observations used to simulate the distribution of credit losses. Unexpected
losses at different levels of probability represent the value of credit losses (as a percentage of the face value of the portfolio) corresponding to the percentile on the right tail of the distribution minus the expected loss given by the mean value of the distribution. Basel
unexpected losses indicates the value of unexpected losses computed according to the algorithm proposed by the Basel Committee on
Banking Supervision (2004b). The value of default probabilities used in the algorithm is given by twice the value of the expected loss,
assuming the same 50 percent loss given default used in the simulation exercise.
The results for Brazil reflect the fact that the category of commercial
loans is restricted to borrowers with consolidated borrowing of more than
U.S.$300,000. To test the effect of this different definition of the loan category, we report a summary of the results for the Argentine bootstrapping
with the same restriction as Brazil (see table 4). In this case, the capital that
would be required for Argentine banks is reduced to 5.6 percent (for 99 percent unexpected losses) or 8 percent (for 99.9 percent unexpected losses),
which is actually somewhat lower than the estimated requirements for
Brazil. These results suggest that Brazils risks are roughly in line with
those of Argentina when we take into account the different definition of
the loan universe. They also call into question the adjustment made for
lending to small and medium-sized enterprises in Basel II, which reduces
capital requirements for this sector. In other words, the reduction in required
capital, which reflects the additional diversification of risks, appears to be
more than outweighed by increased default probabilities in our sample of
emerging economies.
We compared our results with the level of capital that would be generated using the estimated probability of default and the formula proposed
by the Basel Committee for the foundation IRB approach.29 The Basel for29. We used the formula for assessing the capital requirement for the corporate portfolio as described in the Basel Committees third consultative paper (Basel Committee on
Banking Supervision, 2003) and revised in January 2004 (Basel Committee on Banking
Supervision, 2004b).

Table 5. Capital and Provisions: Unexpected and Expected Losses Estimations Based on
Standard and Poors Observed Average Default Frequency
Percent
Country and period
Rating
Default probability
Expected loss
Basel unexpected
loss (Jan 2004)

CCC
BB+
BBB+
27.87
1.38
0.37
13.94
0.69
0.19
19.57
8.07
4.75
Source: Authors calculations, based on data from Standard and Poors.

a. Rating represents the lowest value of domestic currency sovereign rating expressed by Standard and Poors over the time period
considered. This is only partially true in the case of Argentina, where we have conventionally selected a rating of CCC, although formal
rating was suspended in November on the eve of the government default. The default probability is computed on the basis of the
historical average one-year default frequency on Standard and Poorsrated corporate bonds; loss given default is assumed equal to
50 percent. Basel unexpected loss refers to the value of unexpected loss computed according to the algorithm proposed by the Basel
Committee on Banking Supervision (2004b) and using the default probability value corresponding to the value of expected losses over
loss given default.
mula applied to simulated default probabilities generates capital requirements of 14.9 percent, 14.1 percent, and 9.7 percent for Argentina, Brazil,
and Mexico, respectively (see the last column of table 3). These are considerably lower than our simulations at the 99.9 percent level of confidence
for Argentina and Mexico and higher than the 11.5 percent requirement
that we computed for Brazil.
We also compared our results to those of a similar exercise conducted
by Carey, which was intended to mimic the risk exposure of a representative U.S. bank in the period 198991 and also in a period of high stress for
the financial system (namely, 1929).30 These results are included in table 3.
The expected loss for a U.S. bank in 1929 was about 3.1 percent, which
falls between our estimates of 1.4 percent for Mexico and 4.2 percent for
Argentina and Brazil. However, the estimates of unexpected loss for the
United States in 1929 are significantly below our estimates for the three
emerging countries at each statistical confidence level. Finally, our results
can also be compared to the capital requirements generated by the Basel
formula, using a one-year default probability appropriate to the Standard
and Poors sovereign rating in domestic currency (see table 5).31 The results
for expected and unexpected losses are also given. Table 6 provides details
on the domestic currency sovereign ratings across the whole region, as
well as the simulated Basel II IRB capital requirements.
30. See Carey (2002).
31. The default probabilities are estimated by Standard & Poors based on their historical data of defaults by rating category, including corporate claims.
125
Table 6. Capital and Provisions in Latin American Countries Based on Standard and
Poors Ratings
Percent
Standard and
Poors classification
AAA
AA
A
BBB
BB
B
CCC
Country
Chile
Barbados, Mexico, and Trinidad
and Tobago
Colombia
Belize, Brazil, Costa Rica, El Salvador,
Guatemala, Panama, and Peru
Bolivia, Jamaica, Suriname, Uruguay,
and Venezuela
Ecuador, Paraguay, and Dominican
Republic
Expected
loss
Basel unexpected
loss (Jan 2004)
Sum of expected and

unexpected losses
0.00
0.00
0.34
0.61
0.34
0.62
0.02
0.17
1.56
4.75
1.59
4.91
0.62
8.07
8.69
2.71
12.54
15.25
12.54
19.57
32.11

a. Country classification refers to Standard and Poors domestic currency sovereign rating as of December 2003. The expected loss is
the product of the default probability and the loss given default. The default probability is computed on the basis of the historical average one-year default frequency on Standard and Poorsrated corporate bonds; the loss given default is assumed equal to 50 percent.
Basel unexpected loss refers to the value of unexpected losses computed according to the algorithm proposed by the Basel Committee
on Banking Supervision (2004b) and using the default probability value corresponding to the value of expected losses over loss given
default.
Three conclusions can be drawn from the results. The first is that the
Basel formula, applied to Argentina and Mexico, generates levels of protection inferior to the advertised 99.9 percent.32 In the case of Brazil, our
results indicate the opposite, but they are affected by the different universe
of loans used. As the IRB approach is currently calibrated, the degree of
protection would be in the range of 9599 percent of the credit loss distribution for Argentina and Mexico. Another way to state this result is that
achieving the advocated 99.9 percent level of protection would require
substantially higher capital requirements than those advocated in Basel II.33
Moreover, the fact that different levels of capital are required to achieve
the level of protection theoretically granted by the Basel IRB curve calls
into question the curve calibration. In other words, given the types of default
32. This conclusion is conditional on events in the year chosen for the analysis.
33. The Basel II IRB formula in fact calculates assets at risk, and capital requirements
are defined as 8 percent of those assets at risk. Assets at risk are then 12.5 times required
capital. Our results may be interpreted as saying that assets at risk must be larger (maintaining the 8 percent capital requirement) or that the 8 percent should be increased to
achieve 99.9 percent protection. This discussion also assumes that provisions cover expected
losses.
probabilities in emerging countries, it would not necessarily be appropriate

to apply the IRB curve, as written, to individual instruments to ensure a
good approximation to portfolio risk.
The second conclusion is that emerging countries face a difficult choice
if they wish to apply the IRB approach. They may either implement the
IRB curve as it is written and hence very likely opt for a lower degree of
protection than that envisaged by the Basel Committee on Banking Supervision, or they must attempt to recalibrate the curve to obtain a degree of
protection closer to the 99.9 percent as suggested by the Basel Committee.
This should not come as a surprise, since the same issue is present with
Basel I. Many countries adopted the Basel I methodology, but they applied
a higher minimum than the recommended 8 percent.34
A third conclusion of our results is that for emerging countries, the
foundation IRB level of capital requirement is (notwithstanding its benevolent risk calibration) likely to give higher capital requirements than the
existing 8 percent minimum of the Accord. In the case of Argentina, the
foundation IRB approach gives a requirement of about 15 percent, but in
fact it is close to actual capital requirements in Argentina.35 In Mexico and
Brazil, the foundation IRB approach would, according to our simulations,
yield requirements of around 10 percent and 14 percenthigher than
Basels 8 percent and higher than current levels in both Mexico (8 percent)
and Brazil (11 percent).36
In the discussion above we compared the foundation IRB approach
with the actual default experience to measure the default probabilities. We
obtain much lower figures, however, when we use the Standard and Poors
historical mappings of ratings to default probabilities and the relevant sovereign rating. Rating agencies typically interpret such a rating as the floor
to nongovernmental ratings, and therefore the associated default probabil34. A country could simply state that capital requirements are, for example, 10 percent
of Basel II IRB-calculated assets at risk, but there seems little point in adopting a relatively
sophisticated formula and then applying an ad hoc adjustment without considering what
the effect of that adjustment would be on the level of protection within, say, a value-at-risk
framework.
35. Argentina had a baseline 11.5 percent capital requirement, but various add-ons
implied that the overall requirement was close to 15 percent of assets at risk as calculated
under Basel I.
36. Moreover, this does not take into account the additional operational risk capital
requirement (set equal to 15 percent of gross income for the basic indicator approach; see
figure 1). On the other hand, we have not computed whether the enhanced rules on credit
risk mitigation techniques or securitization risk would significantly change this conclusion.
127
ity, expected loss, and unexpected loss can be considered as a floor for the
corporate sectors default probability, expected loss, and unexpected loss.37
Table 4 shows that, notwithstanding the lower level of capital requirements derived from Standard and Poors ratings and default probabilities,
increased capital charges are likely to emerge for most Latin American
and Caribbean countrieswhere the sum of expected and unexpected
losses already exceeds the value of the all-encompassing 8 percent capital
requirement of the current Accord.
In short, our results suggest that while the foundation IRB approach
implies a rather generalized increase in capital requirements, it may not
afford the 99.9 percent protection advocated by the Basel Committee on
Banking Supervision given the default probabilities encountered in
emerging countries. This result also calls into question the calibration of
the published curve for use in these environments. If we assume that provisions cover expected losses, then achieving 99 percent protection
would require capital levels significantly higher than the Basel I recommendation of 8 percent and around a 15 percent overall requirement.
Achieving 99.9 percent protection would require even higher levels of
capital. While these levels were close to Argentinas overall capital
requirement in 2001, they represent a steep increase in capital requirements for many countries.
Policy Implications for Latin America and Emerging Countries

The previous section focused on the appropriate calibration of the Basel II
IRB approach. The results are highly relevant for the wider discussion of
appropriate Basel II application to Latin America and emerging countries.38 That is the focus of this section. The IRB approach reflects recent
developments in the internal risk management of larger G10 banks. Many
large banks have developed their own rating methodologies and have tested
how their own ratings map into default probabilities and value at risk
both on an individual claim and on a portfolio basisusing their own
37. Having said that, each of the three major rating agencies now allows private institutions to break through the sovereign floor, although each is subject to slightly different rules.
38. Kupiec (2001) also discusses Basel II calibration for developing countries. However, he considers the original Basel II curve and examines specific assumptions on particular types of loans in a model-based approach.
credit risk portfolio models.39 Supervisors across G10 countries have

largely been playing catch-up in their understanding of these models.40
The starting point is very different, however, in a typical country in Latin
America.
Is Latin America Ready for the IRB Approach?

Banks in emerging countries are generally less advanced in terms of developing and using internal rating methodologies, mappings those ratings
into default probabilities, and establishing portfolio models of credit risk.
In many emerging countries, the supervisory agencys main motivation for
moving towards the Basel II IRB approach may be to improve banks own
internal risk management, rather than to catch up with what banks are already
doing.41 Moreover, supervisors tend to have significantly less resources in
emerging countries, and they lack supervisory human capital, information
systems, and both legal and real power.42
The statistics on compliance with the Basel Core Principles for Effective Banking Supervision convey a picture of inadequate banking supervision across many emerging countries worldwide and across Latin America
in particular. The average emerging country is compliant with just seven
of the thirty Basel Core Principles.43 Figure 6 illustrates emerging economies compliance with a set of critical Basel Core Principles. Emerging
countries fare poorly, to say the least, and Latin America performs worse
than the average of this group. This suggests that it may be many years
before supervisors in these countries would be advised to adopt the IRB
approach, given the heavy burden on scarce supervisory resources implied.
39. In fact, the concern of many larger G10 banks is that Basel II does not give them
sufficient freedom to use their own portfolio models of credit risk and that they must use the
IRB formula to approximate the risk of a loan portfolio. See the comments by several large
banks on the proposals at www.bis.org.
40. The Basel Committee on Banking Supervision has decided to maintain the formula
rather than allow banks to use internal models for multiple reasons, including the issues of
parameter and model risk and perhaps the fundamental moral hazard reasons discussed in
the introduction.
41. The more sophisticated emerging markets will present exceptions to this, and local
banks that are branches or subsidiaries of large G10 banks are likely to have benefited from
the risk management methodologies implemented across the globe.
42. See Pagano (2001) for a set of papers on issues related to the legal system and credit
risk in Latin America.
43. There are actually twenty Core Principles; here we count the subprinciples of principle 1 as principles in their own right to obtain thirty.
Compliant
Independence & resources (1.2)
Capital adequacy (6)
Connected lending (10)
Country risk (11)
Market risks (12)
Other risks (13)
Cons. supervision (20)
Remedial measures (22)
Globally Cons. Supervision (23)
Largely compliant
10
20
30
50
Percent
Materially noncompliant
40
60
Noncompliant
70
Figure 6. Emerging Countries Compliance with Specific Basel Core Principles for Effective Banking Supervision
80
Not applicable
90
100
As discussed earlier, however, the standardized approach may yield little in relating regulatory capital to risk because of the low penetration of
rating agencies in emerging countries. Therefore an intermediate approach
is warranted to serve as a transition measure to the IRB model. We refer
to this as the centralized ratings-based approach.
The Centralized Ratings-Based Approach

Our proposed centralized ratings-based (CRB) approach is similar in spirit
to the IRB framework in that banks would place their clients into a set of
rating buckets based on their estimated default probabilities, and then each
rating bucket would translate into an average default probability that is then
mapped to a capital requirement using a formula along the lines of the
Basel II IRB model. We simplify the methodology significantly, however,
because the rating methodology and the mapping to capital requirements
are determined by the regulator. The Basel II IRB approach stipulates that
there would be a minimum of seven rating buckets. Under the CRB approach,
the regulator would define the default probabilities (mean, minimum, and
maximum) that would correspond to each of a minimum of seven buckets.
This scale might conform to one used by a leading rating agency, but bank
supervisors may wish to define the scale to reflect the risk characteristics
of their own country and any objectives they wish to achieve.44 Banks
would then simply slot their clients into the buckets suggested by the regulator based on their estimation of each borrowers probability of default.
This approach suffers from one disadvantagenamely, that each bank
would be forced to use the same rating scale (though not necessarily the
same rating for each client, since banks opinions might differ). This means
that a bank specializing in one type of business or region of a country would
have to use the same rating scale as a bank in another line of business or
region. Put another way, because the buckets would essentially be defined
by the minimum and maximum default probabilities, the default probability range of each bucket may not be ideal for every bank. Some banks may
have a large number of clients in one or two buckets of a CRB approach,
whereas if they used an internal scale, they could break those buckets down
44. For a supervisory-based application using cluster analysis, see Foglia, Iannotti, and
Marullo Reedtz (2001). Rating agency scales are typically through the cycle, whereas internal bank rating scales tend to specify a twelve-month or other horizon. Supervisors may
wish to adopt a through-the-cycle scale to reduce concerns of procyclicality.
131
into finer ones with a smaller range of default probabilities, thereby achieving a more precise measure of required bank capital. The supervisor would
most likely define the buckets to be appropriate for the largest, systemic
institutions in the banking sector.45 Consequently, these problems would
be limited to relatively small institutions. One solution would be to adopt
the U.S. model, which calls for such institutions to remain on Basel I or
adopt the standardized approach of Basel II.
Countries may be concerned that the CRB approach would not be seen
as compatible with Basel II. Some emerging countries, especially in Latin
America, already use a type of CRB approach for calculating provisions,
and the level of provisions tends to be high in the region.46 In our simulations we defined default as more than ninety days past due and a loss given
default of 50 percent, whereas many countries in Latin America ask for
100 percent provisions for noncollateralized loans in this category.
This discussion underlines the need for a highly coordinated system for
loan loss reserves and capital requirements. It is the sum of provisions and
capital that should be compared against the value at risk (the sum of expected
and unexpected losses)and not necessarily provisions against expected
loss and capital against the unexpected component.47 If for some reason
(legal or otherwise) there are impediments to increasing capital to cover
unexpected losses relative to the desired level of protection, then provisions might be increased over and above the level of expected loss.
The methodology should gauge the overall value at risk of loans rather
than their expected or unexpected loss components. This calls into question the common system combining a general loan loss reserve, a specific
loan loss classification and provision depending on past performance (say,
according to the traditional five-category classification), and a specified level
of capital.
45. In a more complex proposal, the regulator could allow the use of more than one centralized scale to reflect different banking specializations. The Hong Kong Monetary Authority planned the introduction of a loan classification regulation similar to that described in
this paper (with loan grades characterized by an upper and lower default probability for each
grade; see Hong Kong Monetary Authority (2002)).
46. Colombia, for example, is developing a system labeled SARC (Sistema de Administracin de Riesgos de Crdito) to quantify loan loss reserves based on individual banks
assessments of expected losses (internal models). Argentina and Brazil also have databases
that include a rating scale determined by the regulator, which is used to monitor provisioning.
47. We abstract here from a discussion of the potential dynamic aspects of banks provisioning policies or counter cyclical regulations regarding loan loss reserves; see Cavallo
and Majnoni (2002) and Laeven and Majnoni (2003) for discussions.
How would capital requirements be defined under a CRB approach?

One possibility would be to rely on observed default probabilities, use the
Basel II IRB curve to calculate unexpected losses, and set provisions and
capital to cover expected and unexpected losses, respectively.48 In this case,
provisions could be derived as a residual from the following expression,
based on the estimated value at risk (VaRe):
General Provisions = VaRe Capital
If a regulator could not alter provisioning rules for some reason, a second
approach could set capital equal to value at risk minus the allowable provisions along the lines suggested by the Basel Committee in the revisions
to the third consultative paper.49 This approach is intended to ensure a more
rational integration of bank capital requirements and loan loss reserves.
In this case, capital would be computed as a residual:
Capital = VaRe General Provisions
A third approach, which is appropriate for regulators who have the freedom to alter provisions but who prefer compatibility with Basel II and a
simple rule for capital, would be to adopt the Basel II standardized approach
but then establish forward-looking provisions determined by the value at
risk minus capital requirement specified by the Basel II standardized
approach. This method thus uses the CRB approach to enhance forwardlooking provisioning rules in a fashion that is totally consistent with Basel
II.50 Moreover, banks internal rating methodologies should develop over
time, allowing such banks to move over to the full IRB approach when they
are ready.
The CRB approach has many advantages as a transition tool. First, it lowers the monitoring costs for bank supervisors. Supervisors would have to
verify the quality of banks methodologies for slotting clients into the rele48. Whether a regulator is free to do this would depend on the particular constraints,
legal or otherwise. However, if regulators do not have the freedom to determine provisioning or capital rules, then they would not comply fully with the first Basel Core Principle on
independence and autonomy.
49. Basel Committee on Banking Supervision (2004a, 2004b).
50. In practice, regulators in Latin America tend to have more freedom to determine
provisions than capital. It thus is not surprising that provisions are typically higher than a
standard expected loss calculation.
133
vant buckets; to do so, they could very easily establish a homogeneous scale
for comparing different banks ratings of particular corporate clients, corporate clients of similar characteristics, corporate clients in particular economic sectors or regions, and so forth. The homogeneous scale would also
allow the supervisor to easily verify loan classifications based on the default
probabilities. Second, the CRB approach provides a consistent treatment of
capital and loan loss reserves, which is a vital component of Pillar 2 and
Basel Core Principle 8. Third, a risk-based capacity is developed within the
banking system, independently of each countrys decision to officially remain
in Basel I or to adopt the different options of Basel II. Fourth, the homogeneity of bank classification schemes implies that bank data could easily be
aggregated at a country level, thereby generating an important (and thus far
largely missing) source of data for prudential monitoring at the macroeconomic level. These data would be useful for analyzing changes in the asset
structure of the financial system, developing tools to consider aggregate
financial sector risk, and predicting where problems might occur and their
potential depth. Finally, if countries (in a region or more widely) could coordinate the number and definitions of their rating buckets, then this would
enhance aggregation and comparability across countries. Under the standardized approach, local regulators will likely use incomparable local ratings,
and comparability will undoubtedly be lost under the IRB approach, in which
individual banks will use their own rating methodologies. From this perspective, the CRB appears particularly suited for ensuring the dissemination
of a new risk-based regulatory standard.
Developing a Basel II Decision Tree

Many emerging countries face a difficult decision of whether to stay with
Basel I or move to Basel II.51 If they choose the latter, they must consider
which of the many alternatives to adopt, or whether to implement a mixed
approach leaving some banks on a simpler approach and allowing or forcing a selection of banks to incorporate a more advanced alternative. Five
country characteristics that may aid to guide these important choices: the
degree of compliance with the Basel Core Principles and hence with Basel
II Pillar 2; the penetration of rating agencies and the operation of the ratings market in general; the current level of bank capital and the feasibility
of increases in bank capital ratios in the short term; the size of, or the
51. This section draws on Powell (2002, 2004).
desire to develop, domestic capital markets; and the availability of information and the degree of sophistication of both banks and the supervisor
in assessing and monitoring loan loss provisioning.
As discussed above, evidence from Financial Sector Assessment Programs completed by the International Monetary Fund and the World Bank
illustrates that many countries are far from being fully compliant with the
Basel Core Principles for Effective Banking Supervision and, on average,
emerging countries lag behind their G10 counterparts.52 Of particular concern is the lack of (i) effective consolidated supervision, (ii) supervisory
independence, resources, and authority, and (iii) effective prompt corrective action. If supervisors lack resources and the basics of effective bank
supervision, then correcting this deficiency should be the first priority, and
the implementation of complex rules on capital requirements may well be
counter productive. Basel II also introduces a significant change in the
level of consolidation required for banking supervisionfrom the bank
itself to its holding company. Since many countries do not comply with more
modest versions of consolidated supervision, these countries remain far
from the spirit of the Basel II proposals.
However, full compliance with the Basel Core Principles for Effective
Banking Supervision is too strict a precondition for moving to Basel II.
After all, many G10 countries are not compliant with all the Core Principles. A country should be compliant with the Core Principles to the degree
required to implement the appropriate alternative chosen within the Basel
II framework. For example, if a supervisor does not have the resources
(including data, information, technical competence, staffing, and management) to consider whether the calibration of the Basel II IRB approach is
appropriate to that country, or to monitor effectively how banks would
apply the IRB methodology, then a simpler alternative should be adopted.
Many emerging countries will probably opt for the simpler Basel II
approaches, including the simplified standardized approach and the standardized approach. An important difference between the two is that the
latter allows for the use of credit ratings from private agencies, whereas
the former only uses the ratings of official export credit guarantee agencies
for sovereign risk assessment. The problem for many emerging countries,
however, is that markets for credit ratings are shallow, so the standardized
approach would not improve much on either the simplified standardized
approach or Basel I in terms of aligning capital requirements with risk.
52. See the joint IMFWorld Bank project (IMF and World Bank, 2002)
135
Adopting the standardized approach may create incentives for such ratings markets to develop, but this brings its own dangers in terms of companies buying a good rating and provoking a race to the bottom in ratings
quality. The second key characteristic, then, is the state of the ratings market. The standardized approach makes little sense for a country with no
ratings market to speak of. Such a country should stick with Basel I or
adopt the simplified standardized approach. Alternatively, if the countrys
compliance with the Core Principles is reasonably high, the authorities
could consider the CRB approach as a potential precursor to the Basel II
IRB approach. For a country with an active ratings market, the standardized approach makes more sense.
If a country adopts either the simplified standardized approach or the
standardized approach, Basel II will likely increase bank capital requirements.53 The source of the extra capital charge is operational risk. An
increase in capital requirements may not be a bad thing, but an emerging
country deciding whether to adopt Basel II should carefully consider the
current level of bank capital and the feasibility of increasing required bank
capital. This is the third characteristic listed above.
Basel II also includes enhancements for the credit risk implications of
securitization risk and for credit risk mitigation techniques. A country
with a fairly inactive ratings market may still benefit from the use of ratings in these areas. For example, if a country has an active market for securitized claims (a market that is currently growing in importance in some
countries), then those claims will most likely be rated and the Basel II
standardized approach regarding securitization risk might be gainfully
adopted. This may not seem to be a critical feature, but if a country wishes
to develop capital markets, then it needs to ensure that banks have the right
incentives to securitize claims. Basel II does a better job here than Basel I.
A similar argument can be made for credit risk mitigation techniques.
Basel II makes useful improvements in this area, so it may be appropriate
if markets using securities as collateral are important or if a country wishes
to develop them. The fourth characteristic listed is thus the importance of
local capital markets and the desire to develop them.
The final characteristic suggested above is the sophistication of the supervisor and banks in terms of provisioning rules, monitoring, and control. Basel
53. This may also depend on the long list of items under national discretion. Two such
issues are the risk weights for mortgages and for retail exposures. Capital requirements are
more likely to rise if the more generous treatments are not applied. We do not go further into
the specific items left to national discretion.
II replaces a set of ad hoc rules regarding capital requirements with a more

robust estimate of credit risk reflecting value at risk. Value at risk may be
decomposed into expected and unexpected losses subject to a statistical
tolerance value. As discussed, current theory holds that provisions should
reflect expected loss, whereas capital should reflect unexpected loss. For
an economist, the appropriate level of provisioning and capital for credit
risk then both come from the same probability distribution; they simply
reflect different statistics of that same distribution.
Under this more general approach, a supervisor that has advanced in terms
of more forward-looking provisioning rules has also advanced in terms of
considering finer risk-based capital rules. In several countries in the region,
supervisors have set up centralized databases to monitor the large debtors
of the financial system and ensure that each lender knows the total debt
outstanding of large borrowers. In some cases these databases have been
expanded to cover most loans of the financial system and are used to monitor and control provisioning requirements. Miller presents a review of the
design and uses of these databases.54 While in most countries such requirements are not forward looking but reflect arrears, the move to a forwardlooking system for provisioning and capital is certainly made more feasible
if such a database is in place. For example, some countries have now incorporated a bank rating into these databases that includes not only backwardlooking variables, but also cash-flow-type analyses.
In sum, the key characteristic is the sophistication of the supervisor and
banks in terms of information on provisioning and loan losses. A supervisor that has regularly tracked loan losses across banks and has developed
monitoring tools such as transition probability matrixes and simple credit
scoring techniques to monitor provisioning rules is in a much better position to implement the Basel II IRB approach or our simpler centralized
ratings-based approach than a supervisor that has no experience in these
areas. Still, the IRB or CRB approach will probably only be appropriate
for large, relatively sophisticated banks. A country with a highly concentrated banking sector in which a few large, sophisticated banks control a
large percentage of the sector will encounter added benefits in moving to
the CRB or IRB formula, at least for those banks.
The decision tree in figure 7 illustrates how the above five characteristics may affect the Basel decision and provides a simple navigational aid
for countries regarding the Basel standards. Countries that do not comply
54. Miller (2003).
Basel I
Very weak compliance with

Core Principles
Basel I
Not feasible/desirable to
increase capital
requirements
F I G U R E 7 . A Basel II Decision Tree
Basel II A
Pillar I: Simplified
standardized approach
Feasible/desirable to
increase capital
requirements
Relatively high compliance

with Core Principles, but
not high enough for
internal ratings-based
approach
Centralized ratingsbased approach as
transition to internal
ratings-based approach
Basel II B
Pillar I: Standardized
approach
Internal ratingsbased approach
More sophisticated
provisioning rules
Very good compliance

with Core Principles
Rating penetration
may be low
Compliance with Core

Principles is adequate for
monitoring external
ratings
More liquid capital

markets or reasonable
rating penetration
Minimum compliance with Core

Principles to comply with Pillars II
and III and monitor operational risk
The Basel II Decision
with many of the basic Basel Core Principles are probably advised to stay
with Basel I. However, a country that wishes to increase bank capital requirements should consider the Basel II simplified standardized approach if the
extra burden of supervising operational risk is feasible. Countries that
have only a shallow market for ratings will get limited benefits from the
standardized approach and should be advised that this will also lead to an
increase in capital requirements. They should stick with the simplified standardized approach if supervisory resources are limited. Countries that have
deeper capital markets or a strong desire to develop them should reconsider
the standardized approach for its enhancements to securitization risk and
credit risk mitigation techniques. Finally, countries that have made advances
in terms of forward-looking provisioning rules and that have the information and systems to control banks provisioning practices are better placed
to consider the CRB or even the IRB approach.55
Conclusions
In this paper we have discussed the implementation of the Basel II Accord
in emerging countries, with an emphasis on Latin America. The discussion
suggests three broad concerns with the new accord. First, given the low
penetration of rating agencies, the Basel II standardized approach (which
uses external ratings to gauge credit risk) will do little to link regulatory
capital to risk in Latin America. For countries adopting the standardized
approach, moving to Basel II will imply only a marginal correction of other
problems in Basel I, and it will not address the fundamental problems of
Basel I that motivated the new accord.56 Second, the more advanced Basel
II internal ratings-based approach may require recalibration given our estimates of credit risk. It also appears complex and will stretch scarce supervisory resources in many countries. Finally, the essence of a standard may
be lost if many countries adopt the standardized approach (using incomparable local ratings) or the IRB approach (using many different private
banks ratings and default-probability estimation methodologies).
We propose an intermediate approach between the standardized and
IRB approaches, which we call the centralized ratings-based approach.
55. We perceive a loss of comparability across countries as a cost of the IRB approach.
56. Implementing the standardized approach has advantages stemming from improvements in the treatment of securitization risk and credit risk mitigation techniques.
139
Our approach might be used as a transition measure to the IRB methodology; it might be employed to more fully integrate capital and provisioning
regulations; and it might allow increased coordination on a standardized
risk-based reserving policy across countries in the region or beyond. A
main difference with the IRB approach is that although banks would rate
their clients (and estimate default probabilities), the regulator would define
the rating scale and the way in which the rating buckets would map to
default probabilities. This approach could be used to set forward-looking
provisioning requirements only. A country could then adopt the Basel II
standardized approach, set provisions using the CRB methodology to cover
the value at risk minus the standardized approachs capital, and thereby
ensure that banks total reserves (provisions plus capital) covered the entire
value at risk up to the desired level of protection.
We employed a homogeneous bootstrapping methodology to analyze
credit risk in three emerging markets in Latin America. The bootstrapping
methodology implies that our estimates are free from the usual problems
of parameter estimation error and model error that plague standard attempts
to measure portfolio credit risk. At the same time, our results should be
taken as indicative only, and we hope that future research will attain further precision by extending the empirical methodology over time. The results
indicate that to achieve a 99 percent level of protection (in other words,
such that capital covers the unexpected loss to 99 percent of the distribution), capital requirements would need to be significantly higher than the
8 percent level recommended in Basel I and closer to 15 percent. Even higher
levels would be required to achieve 99.9 percent protection, as intended in
Basel II. We also find that the Basel II foundation IRB approach, while
resulting in increases in capital requirements above Basel I, would result
in levels closer to the 9095 percent protection rather than the 99.9 percent
level stated as used in its calibration. We believe that further research is
required in this area to consider if and how the Basel II IRB methodology
might be recalibrated for countries that have default probabilities significantly higher than G10 countries.
We also discuss Basel II implementation in Latin America more generally and develop a simple Basel II decision tree. Countries should consider
five characteristics when deciding whether to stay on Basel I or implement
Basel IIand if the latter, how. An important characteristic is how countries comply with the Basel Core Principles for Effective Banking Supervision and, hence, with the second pillar of Basel II. If compliance with the
Core Principles is weak, then countries should consider staying on Basel I.
If operational risk can be monitored and capital requirements increased, then

the simplified standardized approach may be in order. If the ratings market is reasonably deep, if the country has a strong desire to deepen capital
markets, and if supervising external ratings is feasible, then the standardized
approach may be appropriate. As the degree of supervisory sophistication
improves, especially in regard to the regulation and supervision of forwardlooking provisions, then the CRB and, eventually, the IRB approaches
may be considered.
Basel II may imply the end of a standard rather than the establishment
of a new one, given the multiple Basel II alternatives, the reliance on incomparable local ratings, and the use of individual bank-generated ratings and
default probability estimations. Put simply, two countries with 14 percent
assets at risk under Basel II may actually be quite different. The CRB
approach is an attempt to develop a more homogeneous system that is compatible with Basel II, suitable to the context of emerging country supervisors, and consistent with the notion of maintaining a standard.57
57. We have not discussed a number of issues pertinent to Latin America that receive
less attention in the new Accord: namely, bank lending to its own government, lending in
foreign currency (domestic dollarization), and related lending. The Basel Accord addresses
all three cases, but the treatment should be tightened in each case.
Comments
Patricia Correa: Endless pages have been written on the potential flaws
that could prevent Basel II from achieving the ultimate goal of increasing
financial stability worldwide by improving bank risk management and
making capital requirements more sensitive to risk. Criticisms, which are
sometimes contradictory, can be grouped into five categories. First, the new
accord offers alternative approaches for measuring capital requirements
(two versions of the standardized approach and two of the internal ratingsbased, or IRB, approach), but it does not create proper incentives to use the
most risk-sensitive approach, thereby opening the door for arbitrage. Second, in countries with little capital market and financial development, reliable external ratings are not available for most of the assets in the banks
credit portfolio; in such cases, the standardized approach will do little to
link better capital to risk and would be, at best, a poor substitute to Basel I.
Third, Basel II relies heavily on methodologies that lead to capital requirements that tend to accentuate the cycle, which can eventually increase,
rather than diminish, financial instability. Whether this flaw can be attributed to the methodologies themselves or to the length of the databases
employed to make the respective estimates is a controversial issue, but this
topic is certainly an important one in the discussion. Fourth, the new framework will most probably increase the amount of minimum regulatory capital regardless of the measurement methodology chosen, and many banks,
particularly in emerging economies, are simply not ready to meet the additional capital requirement. Finally, regulatory and supervisory bodies in
most emerging economies are not prepared to meet the challenges posed
under the second pillar of the accord, owing to lack of infrastructure, inadequate human capital, and so forth.
Majnoni and Powells paper represents an important contribution to the
understanding of the potential impact of Basel II on emerging economies
and sheds light on the validity or relevance of the aforementioned criticisms. The authors contribution is particularly valuable in two aspects: it
141
is the first attempt, to my knowledge, to estimate capital requirements in

Latin American economies using the same methodology, enabling comparison across countries; and the paper goes beyond simply criticizing the
new accord and, within the spirit of Basel II, constructively proposes a new
approach to capital requirements (namely, the centralized ratings-based
approach). However, issues related to the estimation of recovery ratios,
which represent a key component of capital requirement estimation, are
left out of both the quantitative and qualitative analyses in the paper. This
makes the analysis of the challenges ahead incomplete, and it leads to
some erroneous conclusions about the relative advantages and disadvantages of Basel II over Basel I. The authors recommendations are also somewhat unclear and self-contradictory regarding how far and how fast these
countries should move toward implementing better risk management systems and more risk-sensitive capital requirements. Below I present more
specific comments on each of the papers three main components: a quantitative study estimating capital requirements that accomplish the objectives of Basel II in three emerging economies; a qualitative analysis of the
practical difficulties of implementing Basel II, with an assessment of a
countrys readiness to adopt Basel II; and policy recommendations.
With regard to the empirical analysis, the authors use a bootstrap
methodologywhich they argue is free of the usual problems that plague
traditional econometric techniques (namely, parameter estimation and model
errors and risks)to estimate default probabilities for the loan portfolio of
Argentina, Brazil, and Mexico. They describe their specific assumptions
about the provisioning system and their allegedly representative sample of
credit data.1 They then calculate the capital requirements that would be
necessary to cover credit risk under two circumstances: first, using the
foundation IRB approach and the default probability risk weight mapping
curve proposed by Basel II; and second, using the bootstrapping methodology to calculate the capital requirements needed to cover the value at
risk of those portfolios at a 99.9 percent confidence rate.
When comparing the two results, the authors conclude that, except for
Brazil, the Basel II formula generates levels of protection inferior to the
advertised 99.9 percent. The authors therefore call into question the cali-
1. Their assumptions include a recovery ratio of 50 percent, a level of provisions that

covers expected losses; and a definition of default as the event of more than ninety days
past due payments. The sample periods are 200001 for Argentina and Mexico and 200102
for Brazil.
143
bration of the Basel II IRB curve, stating that it is not clear that it would
be appropriate for emerging economies to apply the IRB curve to individual instruments to ensure a good approximation of risk. They suggest that
as with Basel I, countries should or may choose to recalibrate the curve or
make the requirement more stringent (for example, establishing a 10 percent capital adequacy requirement instead of 8 percent). They also conclude that the foundation IRB approach is, notwithstanding its benevolent
risk calibration, likely to set higher capital requirements than the existing 8
percent under Basel I.2
While the methodology employed by the authors has its advantages, the
data limitations and the short period chosen for the estimations (issues that
are acknowledged in the paper) make it premature to categorically conclude that the Basel II models need to be recalibrated. Only after many
years of experience and collection of quality data would it be possible to
quantify those default rates and value-at-risk levels appropriately and,
incidentally, reduce the procyclicality inherent to all approaches based on
empirical estimations (even those using bootstrapping). Several related
questions come to mind. Has the bootstrapping methodology been contrasted with that of Basel II using G10 data? If so, for what periods? How
sensitive are the comparisons of bootstrapping and Basel II to the period
chosen? Are Brazils small credit loans necessarily more risky than large
ones, as assumed by the authors?
In any event, these limitations should not serve as an excuse for regulators and banks to not move forward in refining internal risk measurements.
Neither should the fact that Basel II implies more capital requirements. To
advance Basel IIs general goal, the problem of raising additional capital
could be solved by phasing in the meeting of the new requirements, rather
than halting progress in the implementation of IRB systems.
I turn now to the papers qualitative analysis. In assessing the difficulties that emerging economies may have in applying the new capital accord,
Majnoni and Powell outline the minimum conditions that countries have
to meet before they begin implementation, and they summarize documentation that proves that many emerging economies do not yet meet these
requirements. I disagree with the authors approach to tackling these problems, which is basically to strengthen institutions before starting to implement Basel II. It is true that the ideal conditions are not present in many
countries (I would dare say any country), and many institutional weaknesses
2. Specifically, 15 percent for Argentina and 1014 percent in Mexico and Brazil.
prevail in the supervisory agencies and the banking industry. My experience as a banking supervisor in Colombia, however, made me a firm
believer that the only way to create such conditions, particularly regarding
the second pillar, is precisely by moving forward and setting clear goals
and deadlines. In this ever-evolving field, which is more art than science,
learning by doing is the only way to succeed, as has long been the process
in the developed world. A sure way to delay preparedness is to postpone
the definition of policy goals such as the development of good risk assessment within banks and matching regulatory capital. Again, gradually phasing in the objective is preferable to not starting the run.
Finally, the papers policy recommendations aim to facilitate the transition toward the Basel II IRB regime in emerging economies. The authors
propose a centralized ratings-based (CRB) approach to cover credit risks,
which is compatible with the IRB model and which has the following characteristics: banks estimate default probabilities according to their own
internal models, as in Basel II; the regulator defines the rating scale to be
used and the mapping of each rating bucket to a range of default probabilities (in Basel II this is done independently by each bank, not uniformly by
the regulator); and loan loss provisions are defined as the expected loss
given default for each category of loans, and regulatory capital is defined
as the total value at risk minus the expected loss (or, in the case of legal or
other problems with changing the capital regime during the transition
period, provisions could be defined as the difference between the desired
total level of protection and the current capital requirement, and provisions
could thus be over the expected loss).
This transition regime has a number of advantages. It coordinates the
system for loan loss provisioning and capital requirements, and it would
facilitate comparisons across banks and the handling and interpretation of
data on credit risk. This proposal is in many ways similar to the system
being implemented in Colombia. It is extremely appealing and should
receive more attention and backing by international regulators. While it
certainly simplifies matters for both regulators and banks, it is perfectly
compatible with the spirit and ultimate goal of Basel II.
Philip Brock: In 1988 the Basel Committee on Banking Supervision formulated the first Basel accord for bank capital requirements (Basel I). Its
purpose was to raise overall levels of capital adequacy in the thirteen
member countries while simultaneously homogenizing standards. Basel I
was a great success, with over a hundred countries adopting the framework.
145
Concerns arose, however, over the side effects of the accord. Among other
issues, the capital standards of Basel I are relatively insensitive to the riskiness of bank portfolios, and the accord creates incentives to engage in regulatory arbitrage, whereby banks increase their risk within the parameters
of Basel I without raising levels of capital.1
These concerns with the somewhat blunt nature of the Basel I capital
standards led to the forging of a second accord (Basel II) in June 2004,
which seeks to make bank capital more responsive to credit risk. Basel II
offers four approaches to calculating bank capital. The first two, the standardized and simplified standardized approaches, map the ratings of credit
rating agencies into capital requirements. The second two rely on banks
own internal ratings-based (IRB) models to generate levels of capital adequacy. Although this menu of approaches to capital requirements addresses
the concern that Basel I is not sensitive enough to bank risk, Basel II has
its own flaws. The two standardized approaches have highlighted concerns about the ability of risk-rating agencies to provide meaningful assessments of bank risk.2 The two IRB approaches rely heavily on value-at-risk
(VaR) models that only provide point estimates of the loss distribution,
leaving substantial room for so-called spike-the-firm events involving
high losses with low probability.3 Regulatory arbitrage could also occur
across banks adopting different approaches (for example, the standardized
versus IRB approaches).4
Majnoni and Powells paper centers on the adaptation of Basel II to
Latin American financial systems. Their first concern is the lack of penetration of credit-rating agencies in Latin America, which makes the implementation of Basel IIs standardized approach difficult. The second is the
accuracy of the VaR approach for calculating capital adequacy levels. A
centerpiece of the paper is the use of a bootstrapping methodology to calculate levels of capital adequacy that cover losses in 99 and 99.9 percent
of potential outcomes in any given year. The authors apply this bootstrapping methodology to loans from Argentina, Brazil, and Mexico. When they
compare the bootstrapping methodology with a VaR model calibrated using
Basel II values, they find that that the VaR model underestimates the amount
of capital that banks should be holding in each of the three countries. The
1.
2.
3.
4.
Dewatripont and Tirole (1994); Saidenberg and Schuermann (2003).

Danielsson and others (2001); Goodhart (2004).
Danielsson and others (2001).
Repullo and Surez (2004).
authors find, among other factors, that the Basel II risk weights for small
and medium-sized business loans are too low for Latin America, in
comparison with the empirically derived results from the bootstrapping
exercises.
The authors concerns with the application of the standardized and IRB
approaches of Basel II leads them to propose a hybrid approach, which
they call the centralized ratings-based (CRB) approach. This approach
relies on bank supervisors to develop risk ratings for banks based on the
information that banks provide to the supervising authorities. In contrast
with Basel II, the CRB imposes uniform (rather than bank-specific) risk
weights across categories of loans for all banks, but the risk weights are
determined with the active interaction of the banks and the bank supervisor. The CRB approach is similar to approaches currently in use in several Latin American countries.
A primary purpose of the Basel capital accords is to promote the stability
of financial systems. As with Basel I and II, there are some worries associated with the CRB approach. Like Basel II, the CRB approach may lead to
procyclical capital requirements, since lower measured credit risk will lead
to lower capital-asset ratios during extended periods of good banking performance. This is less apt to be the case with Basel I, in which capital
requirements respond less to changes in risk. Regulatory capture is another
concern with the CRB approach. In particular, the regulator may come under
pressure at times to be lenient in the classification of bank loans.
Any capital adequacy framework that a country adopts may destabilize,
as well as stabilize, the financial system. Basel I, Basel II, and the CRB
approach all strengthen bank supervision, but they may result in unwanted
risk taking. Much risk faced by banks is macroeconomic, and this type of
risk is underemphasized in Basel II.5 Other financial sector policies can
partially address this macroeconomic risk. For example, evidence indicates that policies geared toward reducing dollarization in Latin America
would stabilize financial systems.6 Policy measures to cushion the impact
of sudden stops of foreign capital would also increase the stability of the
banking systems.7 Ultimately, the success of Basel II or the CRB approach
in Latin America will depend on the accompanying policy measures taken
to stabilize the economies against macroeconomic shocks.
5.
6.
7.
Blaschke and others (2001); Carling and others (2002); Sorge (2004).
Herrera and Valdes (2004); Levy-Yeyati (2004).
Calvo, Izquierdo, and Meja (2004).
147
References
Altman, Edward I., and Anthony Saunders. 1997. Credit Risk Measurement:
Developments over the Last Twenty Years. Journal of Banking and Finance
21(1112): 172142.
Balzarotti, Veronica, Christian Castro, and Andrew Powell. 2004. Reforming
Capital Requirements in Emerging Countries: Calibrating Basel II Using Historical Argentine Credit Bureau Data and CreditRisk+. Business School
working paper. Universidad Torcuato Di Tella.
Balzarotti, Veronica, Michael Falkenheim, and Andrew Powell. 2002. On the
Use of Portfolio Risk Models and Capital Requirements in Emerging Markets:
The Case of Argentina. World Bank Economic Review 16(2): 197212.
Basel Committee on Banking Supervision. 2003. Overview of the New Basel
Capital Accord. Consultative document 3 (April). Basel: Bank for International Settlements.
. 2004a. An Explanatory Note on the Basel II IRB Risk Weight Functions. Technical paper (October). Basel: Bank for International Settlements.
. 2004b. Modifications to the Capital Treatment for Expected and Unexpected Credit Losses in the New Basel Accord. Technical paper (January).
Basel: Bank for International Settlements.
Berger, Allen N., Richard R. Herring, and Giorgio P. Szego. 1995. The Role of
Capital in Financial Institutions. Journal of Banking and Finance 19(34):
393430.
Blaschke, Winfrid J., and others. 2001. Stress Testing of Financial Systems: An
Overview of Issues, Methodologies, and FSAP Experiences. Working paper
01/88. Washington: International Monetary Fund.
Calvo, Guillermo, Alejandro Izquierdo, and Luis-Fernando Meja. 2004. On the
Empirics of Sudden Stops: The Relevance of Balance-Sheet Effects. Working
paper 10520. Cambridge, Mass.: National Bureau of Economic Research.
Carey, Mark S. 1998. Credit Risk in Private Debt Portfolios. Journal of Finance
53(4): 136387.
. 2002. A Guide to Choosing Absolute Bank Capital Requirements.
Journal of Banking and Finance 26(5): 92951.
Carling, Kenneth, and others. 2002. Capital Charges Under Basel II: Corporate
Credit Risk Modelling and the Macroeconomy. Working paper series 142.
Stockholm: Sveriges Riksbank.
Cavallo, Michelle, and Giovanni Majnoni. 2002. Do Banks Provision for Bad
Loans in Good Times? Empirical Evidence and Policy Implications. In Ratings, Rating Agencies and the Global Financial System, edited by Richard
Levich, Giovanni Majnoni, and Carmen Reinhart. New York: Kluwer Academic Publishers.
Danielsson, Jon, and others. 2001. An Academic Response to Basel II. Special
paper 130. London School of Economics, Financial Markets Group.

Dewatripont, Mathias, and Jean Tirole. 1994. The Prudential Regulation of Banks.
MIT Press.
Ferri, Giovanni, Li-Gang Liu, and Giovanni Majnoni. 2001. The Role of Rating
Agency Assessments in Less Developed Countries: Impact of the Proposed
Basel Guidelines. Journal of Banking and Finance 25(l): 11548.
Foglia, Antonella. 2003. Using Credit Register and Company Account Data for
Measuring Credit Risk: A Supervisory Approach. Rome: Bank of Italy.
Mimeographed.
Foglia, Antonella, Simona Iannotti, and Paolo Marullo Reedtz. 2001. The Definition of the Grading Scales in Banks Internal Rating Systems. Rome: Bank of
Italy. Mimeographed.
Freixas, Xavier, and Jean-Charles Rochet. 1999. Microeconomics of Banking,
4th ed. MIT Press.
Goodhart, Charles. 2004. Some New Directions for Financial Stability? Per
Jacobsson lecture. London School of Economics, Financial Markets Group.
Goodhart, Charles, and others. 1999. Financial Regulation. London: Routledge.
Herrera, Luis Oscar, and Rodrigo Valdes. 2004. Dedollarization, Indexation and
Nominalization: The Chilean Experience. Working paper 261. Santiago: Central Bank of Chile.
Hong Kong Monetary Authority. 2002. Revision of Loan Classification System.
Mimeographed.
IMF (International Monetary Fund) and World Bank. 2002. Implementation of
the Basel Core Principles for Effective Banking Supervision: Experiences,
Influences, and Perspectives. Washington: International Monetary Fund. Available at www.imf.org/external/np/mae/bcore/2002/092302.htm.
Kupiec, Paul. 2001. Is the New Basel Accord Incentive Compatible? Washington: International Monetary Fund. Mimeographed.
Laeven, Luc, and Giovanni Majnoni. 2003. Loan Loss Provisioning and Economic Slowdowns: Too Much, Too Late? Journal of Financial Intermediation 12(2): 17897.
Levy Yeyati, Eduardo. 2004. Financial Dollarization: Evaluating the Consequences. Universidad Torcuato Di Tella. Mimeographed.
Mrquez, Javier, and others. 2003. Credit Information, Credit Risk Measurement, and the Regulation of Bank Capital and Provisions in Mexico. Mexico
City: Bank of Mexico. Mimeographed.
Miller, Margaret. 2003. Credit Reporting Systems and the International Economy.
MIT Press.
Mishkin, Frederic S., ed. 2001. Prudential Supervision: What Works and What
Doesnt. University of Chicago Press.
Pagano, Marco, ed. 2001. Defusing Default: Incentives and Institutions. Johns
Hopkins University Press.
Powell, Andrew. 2002. A Capital Accord for Emerging Economies? Policy
research working paper 2808. Washington: World Bank.
149
. 2004. Basel II and Developing Countries: Sailing through the Sea of

Standards. Universidad Torcuato Di Tella. Mimeographed.
Powell, Andrew. and others. 2004. Improving Credit Information, Bank Regulation, and Supervision: On the Role and Design of Public Credit Registries.
Policy research working paper 3013. Washington: World Bank.
Repullo, Rafael, and Javier Surez. 2004. Loan Pricing under Basel Capital
Requirements. Madrid: Center for Monetary and Financial Studies.
Saidenberg, Marc R., and Til Schuermann. 2003. The New Basel Capital Accord
and Questions for Research. Working paper 03-14. University of Pennsylvania, Wharton School, Financial Institutions Center.
Sorge, Marco. 2004. Stress-Testing Financial Systems: An Overview of Current
Methodologies. Working paper 165. Basel: Bank for International Settlements.
JUAN CARLOS ECHEVERRY

ANA MARA IBEZ
ANDRS MOYA
LUIS CARLOS HILLN
The Economics of TransMilenio,

a Mass Transit System for Bogot
he provision of urban public mass transportation in Latin America has
exhibited marked swings during the last century. Various cities have
experimented with public and private provision, as well as different
types of market regulation ranging from almost total liberalization to intervention in fares, route allocations, bus size, service quality, and exclusive
road lanes. Despite the importance of this sector for the quality of social
life and urban economic development, the literature contains little theoretical research on the market failures characteristic of this industry. There
are also few studies on the consequences of alternative provision and regulation arrangements. This paper contributes to this second area of research.
We describe the experience of Bogot, Colombia, where a new hybrid system of urban public transport was put in place at the beginning of this
centuryhence its name, TransMilenio. The new system was designed after
decades of learning about the failures of both publicly and privately owned
systems. It currently supplies more than 20 percent of daily trips. By 2015,
the complete TransMilenio system is expected to transport 80 percent of
the citys population at an average speed of 25 kilometers per hour with a
service quality similar to an underground metro system.
Several market failures affect the provision of urban mass transit: unclear definition of property rights on the curbside and on the road; collusion,
Echeverry, Ibez, and Moya are with the Universidad de los Andes, Department of Economics, in Bogot, Colombia. Hilln is with the Colombian Ministry of the Environment.
We express our thanks to Marcela Melndez and Edgar Sandoval, who worked on a first
draft of this paper. We are particularly grateful to Andrs Gmez-Lobo for on-going cooperation with successive drafts, as well as to Mauricio Crdenas, Kenneth McConnell, Norman
Offstein, and Andrs Velasco. Excellent research assistance was provided by ngela Fonseca
and Mnica Hernndez.
151
which results in fares set above competitive equilibrium levels; misalignment of the incentives of bus drivers and owners (a typical principal-agent
problem); and congestion and pollution.1 In many developing countries,
these market failures are exacerbated by weak regulation and enforcement. In Bogot over the last three decades, city transportation was in the
hands of private entrepreneurs, and local authorities were in charge of regulating the system and maintaining the road infrastructure. This arrangement suffered to some degree from all of these market failures. At the end
of the 1990s, the system had an excess supply of usually empty and slow
buses, low-quality service, and widespread inefficiency. Average travel
time to work was one hour and ten minutes; obsolete vehicles were used;
average speed was only ten kilometers per hour during peak hours; 70 percent of air pollution in the central corridors was generated by traffic; and
accidents were frequent.
TransMilenio, a hybrid public-private scheme, was designed to overcome these market failures and improve urban transport quality. The first
market failure, the unclear definition of property rights on the curbside
and on the road, was solved via exclusive lanes and the construction of
restricted-access elevated stations. Passengers pay for access to a system
of buses and stations, as in an underground metro system, and not for access
to the vehicles, as was previously the case. TransMilenio uses the left lane
of the streets, and there is no staircase for accessing the buses, which facilitates transfers in different directions and improves the speed of passengers movements.
The second failure TransMilenio addressed involved fares. These were
traditionally set above competitive equilibrium levels, largely as a result
of capture of the regulator. The theoretical literature identifies reasons for
market power by the bus owners that can also support overpricing. Overinvestment in busses was profitable, leading to excess supply, low use, overcrowding of streets, low speeds, and pollution. In contrast, TransMilenio
fares are set at the level at which they finance the long-term cost of provision, defined through a route-tendering process in which potential providers compete for the exclusive use of the roads based on the lowest cost
bidding. Fares evolve based on the change in input prices and the number
of passengers transported.
TransMilenio also faced a third market failure, which resulted from the
private solution to the agency problem between affiliating firms and bus
1. Estache and Gmez-Lobo (2005).
Juan Carlos Echeverry, Ana Mara Ibez, Andrs Moya, and Luis Carlos Hilln
153
driver-owners. Under the traditional scheme, firms extracted their revenues

from affiliated buses, while the driver-owners revenues depended on the
number of passengers. This arrangement caused an excess supply of buses
and infighting among them for each extra passenger, with undesirable effects
on safety, pollution, and congestion. TransMilenio introduced a new arrangement in which bus owners are private firms that charge per kilometer
traveled, not per passenger, and drivers are employees whose salary is determined by a labor contract that is unaffected by the number of passengers
traveling. The public sector participates in the network configuration and
system regulation and supervision, and the private sector operates the buses.
In addition, a different private company is in charge of fare collection.
After its launch in January 2001, TransMilenio significantly improved
traveling conditions for its users and reduced traffic on its corridors. The
system reduced travel times for TransMilenio users by 32 percent, particulate matter pollution fell by 9 percent in some areas of the city, and accident rates dropped by 90 percent in the TransMilenio corridors.
But there have been problems, as well. Perhaps the most significant
shortcoming of TransMilenio is related to the transition between the traditional transportation system and the new one on currently unserved corridors. To secure support for TransMilenio, the local government provided
politically powerful traditional transportation firms with more routes in the
remaining corridors, where they relocated most of the vehicles displaced by
TransMilenio. Congestion, pollution, and travel times have worsened for
riders of the traditional system.
We carry out a cost-benefit analysis to assess the impact of the first
phase of TransMilenio on travel times, the environment, and traffic accidents in the whole city. Despite the sizable benefit that TransMilenio has
bestowed on its users, the net benefits for the whole city of implementing
the first phase appear negative. This is the result of the spillovers caused
by the slow scrapping rates and the maintenance of weak regulation of the
traditional system.2
This paper provides policy lessons for cities in developing countries planning to undertake similar reforms in this industry. The cost-benefit analysis
identifies policy mistakes during the implementation of TransMilenio that
2. Contracts for TransMilenio private operators required them to buy traditional system
buses and scrap them, to prevent those buses from being relocated to unserved TransMilenio
corridors. The authorities did not effectively enforce this process, however, so traditional
system buses that were displaced by TransMilenios operation were, in fact, relocated to
other corridors.
led to negative spillovers and reduced the benefits of TransMilenio. Recognizing these risks up front is useful for avoiding similar mistakes in
cities that are planning to reform their mass transit system. The paper is
organized as follows. The next section reviews the market failures prevailing in the provision of public transport, describes Bogots mass transit system before TransMilenio, illustrates the main characteristics of the
new system, and discusses some political economy issues of its implementation. Both positive and negative changes in the quality of life resulting from the new mass transit system, as well as its particular adoption and
a cost-benefit analysis, are included in the subsequent section. A final section presents our concluding remarks.
Mass Transit Market Failures in Bogot and the TransMilenio Reform

Many cities in developing countries have experienced a pendulum in the
operation of their public transit system. Bogot, for instance, evolved from
privately owned trolley buses at the end of the nineteenth century to a fully
public bus system in the first half of the twentieth. In the second half of the
last century, developing countries witnessed a slow transition back to private ownership owing to the lack of flexibility and productive inefficiency
that characterized state ownership and operation. The mixed success of the
last decades regarding private provision, liberalization, and market competition, especially in cities in developing countries, led to a reconsideration
of this model and the adoption of so-called hybrid models that combine
public and private features. The main reason for this paradigm shift is the
evidence of market failures that hinder a fully private provision from reaching socially optimal outcomes.
Market failures in the provision of urban mass transit include unclear
definition of property rights on the curbside and on the road; the fact that
fares are set above competitive equilibrium levels; a principal-agent problem stemming from a possible misalignment of the incentives of bus drivers
and owners; and externalities of street congestion and air and noise pollution. In addition to these market failures, developing countries frequently
suffer policy failures in the form of weak regulation and enforcement.
Property rights usually are not clearly defined on the road or on the
curbside.3 The absence of properly defined rights to waiting passengers
3. Klein, Moore, and Reja (1997).
155
implies that the fare received from a pedestrian at the curb does not belong
exclusively to the firm authorized to operate on that route. In the absence
of regulation and control, other means of transportation can interlope and
offer pedestrians a ride, which they may accept because waiting time is
costly.4 The implication of this lack of property rights differs in developed
and developing countries. Demand for mass transit in developed countries
can be low (that is, markets are thin), in which case it needs to be induced
by a regular, high-quality service. Once this demand is created, incentives
arise for illegal interlopers to start operating. Hence investments by legal
operators may not be recuperated. This induces underinvestment. In contrast, demand in developing countries tends to be high (that is, markets are
thick).5 Excessive bus entry is the norm, resulting in strong competition for
passengers on the curb. This strong competition spurs distortions in
investment, such as deficient service quality and the use of small vehicles,
which are more maneuverable but produce more pollution, congestion,
larger investment per seat, and safety problems.6
In developing countries, fares are often set above competitive equilibrium levels. This promotes excessive entry of buses. Because buses are not
perfect substitutes, price competition is not an effective mechanism for
regulating the optimal quantity of buses in the market. To minimize waiting time, riders prefer to use the first bus that arrives even though a cheaper
bus may come along in a few minutes. Time, not fares, might be the most
important decision variable for the rider, so the bus can exercise its market power by raising fares. As a result, prices are set above efficient levels
and returns on investment are high, creating incentives for entry and an
excessive number of buses on the road.7 This feature can also result from,
or be exacerbated by, the capture of the regulator. The excess supply of
buses, paired with deficient service, leads to congestion, pollution, and traffic accidents.
Another source of market failures results from the private solution to the
principal-agent problem. This solution, which tries to cope with the misalignment between the interests of the bus driver and the bus owner, introduces further inefficiencies in the provision of mass transit. The profits of
bus owners depend on the number of passengers carried per bus. To align
the interests of bus owners and drivers, the owners typically pay the drivers
4.
5.
6.
7.
Evidence shows that passengers consider waiting time more costly than travel time.
Most of the population in these countries lacks transportation alternatives.
Estache and Gmez-Lobo (2005).
Evans (1987).
based on the number of passengers carried. This contract between the owner
and the driver, however, introduces further complications for the provision
of mass transit. On the one hand, bus drivers compete for passengers to
maximize their payments, causing negative externalities in terms of safety
problems and congestion. On the other, this compensation scheme promotes the use of smaller buses, despite the congestion and pollution problems, because they are better equipped to compete for passengers.
The problems described above are deepened and new inefficiencies
arise when drivers are responsible for fare collection. High monitoring
costs promote the sale of vehicles to the drivers and cause the atomization
of the bus industry in terms of bus ownership. Although this atomization
is effective for aligning the interests of the owner and drivers, it leads to
additional costs. The large number of owners makes it difficult to exploit
economies of scale on coordination and economies of density. Deficient
regulation, weak enforcement, and capture of institutions also worsen market failures.
The Mass Transit System in Bogot before TransMilenio

Essentially all of the market failures mentioned above characterized the
public transportation system of Bogot in the second half of the twentieth
century. This industry went through three periods in terms of ownership
and regulation: (i) simultaneous provision and ownership by the state and
the private sector (through the mid-1980s); (ii) private firms specialized in
intermediating routes between the regulator and the bus owners and drivers
(mid-1980s to the early 1990s); and (iii) private provision and the emergence of different types of buses, with fares dependent on the quality of
service, such as carrying seated passengers only, providing clear bus stops,
and opening new routes (early 1990s to the present).
In an initial phase, mass transit was exclusively provided by a public
company, resulting in economic inefficiencies, an excess of drivers per
vehicle, and inflexibility in supplying sufficient transport services and new
routes for a growing city.8 Increasingly, the regulatory agencies found that
private firms could supply new routes and either complement or replace the
existing ones at cost-efficient levels. The resulting private scheme had
three types of actors: the local government, bus owners, and firms acting as
intermediaries between them. Each bus owner operated a specific route
after purchasing one of the slots approved by the authority and then paying
8. Urrutia (1981).
157
a monthly affiliation fee to the intermediary.9 Originally, the intermediary

firms owned the buses, and the drivers were in charge of fare collection.
The drivers income depended on the number of passengers carried, which
served to align the interests of drivers and bus owners. However, this solution led to strong competition among drivers of different companies. Moreover, most routes sought to cover the central part of the city, where the
highest number of passengers is concentrated. Overlapping routes exacerbated traffic congestion, street competition, and traffic accidents.
In the early 1980s, fare setting depended on recurrent negotiations with
the regulator and local authorities and on a subsidy system implemented
to avoid transferring high fuel costs to passenger fares. The high return on
the bus investment promoted oversupply, as fares and subsidies surpassed
marginal costs. The gradual elimination of subsidies in the 1980s led to
two developments. First, higher fares were granted to unsubsidized vehicles, which were obliged to comply with higher quality standards than
their subsidized counterparts. The average fare increased from U.S.$0.10
in 1975 to U.S.$0.40 in 2004, in constant 1999 dollars (see figure 1). Since
Bogot did not experiment with fare liberalization, these fare increments
were mostly the result of the strong bargaining power that private transport firms attained in the 1980s.10 Second, the absence of price competition made the market a poor regulating mechanism for the optimal quantity
of buses. Returns to investment were high, which created incentives for
the entry of an excessive number of buses. Figure 2 illustrates the evolution of the public bus fleet in Bogot. The increasing trend accelerated in
the first half of the 1990s, peaking at approximately 22,000 vehicles. This
was accompanied by a sharp decline in average occupancy after 1985. The
daily number of passengers per vehicle stabilized briefly in the early 1990s
and then fell dramatically from 538 in 1992 to 294 in 2003. Frequent fare
hikes offset the subsequent income losses and served as an incentive for
the entry of more vehicles, resulting in further reductions in the number of
passengers per vehicle.
The industry further separated the affiliating firms from the bus owners
and drivers. The atomization of the ownership structure was caused by the
high costs of monitoring the drivers revenue collection. The solution of
9. In 1995 the cost per slot was between U.S.$2,300 and U.S.$4,600, the affiliation fee
ranged from U.S.$100 to U.S.$600, and the monthly fee varied between U.S.$5 and U.S.$35
(Lleras, 2003).
10. Private transporters started to participate in politics either directly via their own candidates in the city council or indirectly via funding the campaigns of traditional politicians.

F I G U R E 1 . Bus Fare and Average Occupancy, 19732003
Average occupancy
650
Fare (1999 U.S.$ cents)

45
600
40
550
35
500
30
450
25
400
20
350
15
Passengers per vehicle
Fare
300
1970
1975
1980
1985 Year 1990
1995
10
2000
2005
Source: Authors calculations, based on data from the National Administrative Department of Statistics (DANE).
this agency problem led the affiliating firms to specialize in obtaining

routes from the authorities and affiliating as many buses as possible. Since
the aging of the existing fleet increased operating costs, the firms sold old
vehicles, usually to the drivers themselves, and owned only new buses.11 At
this point, the number of vehicles on the streets was completely determined
by the affiliating firms, because the public company had been liquidated.
The firms obtained as many routes as possible and earned their income
basically from affiliation rather than from the number of passengers transported. The firms also established a cartel to consistently push for bus fare
hikes and the authorization to extend the service life of buses.12 This struc11. The increase of driver-owners as a solution to the limited monitoring capacity of the
affiliating firms is also documented for the case of Santiago de Chile (Estache and GmezLobo, 2005).
12. Bus fares were historically set through negotiation processes with the local authorities. Only in 1997 was an effort put forth to tie them to a transport costs basket (Decree
3109 of 1997). Estache and Gmez-Lobo (2005) propose a model for explaining the overprovision of services in a private equilibrium when fares are too high. Their key argument
is the fact that buses are not perfect substitutes, but differentiated products. Hence the bus
stopping in front of a passenger enjoys a market power over its competitors due to the cost
159
F I G U R E 2 . Public Transit Vehicle Fleet in Bogot, 19702003a

Busetas
10,000
9,000
8,000
7,000
Vehicles
25,000
BUSETA
EJECUTIVA
SUPEREJECUTIVA
TOTAL BUSETAS
VEHICLES
20,000
6,000
15,000
5,000
4,000
10,000
3,000
2,000
5,000
1,000
1970
1975
1980
1985
Year
1990
1995
2000
2005
Source: Authors calculations, based on data from DANE.

a. While busetas carry an average of 25 passengers and operate with low-quality standards, busetas ejecutivas and superejecutivas,
introduced in the early 1990s, offer an average capacity of 50 seated passengers and better service and, consequently, require higher
fares than traditional busetas.
ture intensified the struggle for economic rents, increased the number of
buses on the streets, and contributed to the aging of the vehicle fleet. Market and policy failures thus accentuated the problem of oversupply.
We estimated the economic rents for Bogots traditional mass transit
system by simulating the contracts between affiliating firms and ownerdrivers (see appendix A for a description of the methodology). This industry operated at efficiency levels (nearly zero profits) in only two periods:
the end of the 1970s and the beginning of the 1990s (see figure 3). Outside
of these two periods, economic rents were positive, substantial, and highly
of waiting, which is highly valued by eventual passengers; this permits the charging of a
higher fare. Consequently, rising tariffs create excessive returns to investment and promote
the entry of new buses beyond the socially efficient level (see pp. 1215). The capture of (or
political influence on) the regulator is yet another reason for high fares. As a result of excessive entry, buses in Bogot and Santiago, Chile, have witnessed a sharp decline in occupancy, as illustrated below.

F I G U R E 3 . Daily Simulated Economic Rent per Vehicle, 19732003
Bus drivers (1999 U.S.$)
20.00
Fleet administrator (1999 U.S.$)

40.00
18.00
Driver-owners
36.00
16.00
Fleet administrator
32.00
28.00
12.00
24.00
10.00
20.00
8.00
16.00
6.00
12.00
4.00
8.00
2.00
4.00
19
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
2099
2000
2001
2002
03
14.00
Year
Source: Authors calculations, based on DANE (2003).
variable. These rents rose in the 1980s, when fares grew from U.S.$.10 to
U.S.$.30, which was associated with the introduction of smaller capacity
vehicles. Rents accrued to both the fleet administrating firms and the bus
owner-drivers.
As figure 3 illustrates, however, the composition of rents changed in the
1990s, favoring the affiliating firms vis--vis bus owner-drivers, and total
daily rents never returned to the peak observed at the end of the 1980s. A
new development was observed at the beginning of the 1990s, when the
affiliating firms lobbied for further fare increases based on different levels
of service quality. The so-called executive buses carried only seated passengers, used relatively new vehicles, and offered faster service at higher
fares than traditional buses. This behavior reveals either the capture of the
regulator or the authorities ignorance regarding the true costs of bus transportation, or both.
In sum, the market failure stemming from the absence of price competition in urban bus transport was exacerbated by a policy failure that granted
161
free affiliating privileges and periodic fare negotiation with a cartel, leading to systematic economic rents. In addition, this incentive scheme forced
bus drivers to compete for passengers along the central corridors. Drivers
therefore did not respect assigned bus stops or delineated areas for bus
transit in the pursuit of passengers.13
At the end of the 1990s, the excess bus supply, coupled with a substantial increase of private vehicles, exceeded the traffic capacity of Bogot.
From 1991 to 1995, the total number of cars registered in Bogot increased
by 75 percent, and 40 percent of the countrys vehicles were circulating in
the city.14 By 1998, private cars occupied 64 percent of the road space and
mobilized only 19 percent of the population.15 As a result, gridlock was
commonplace, accidents abounded, and travel times were unbelievably
high. Estimates show that in 1995 the average number of daily trips per
household was 11.9, the average number of daily trips per person was 1.7,
and the total number of daily trips was about 10 million.16 The average
speed for public transportation during peak hours was ten kilometers per
hour, which could drop to five kilometers per hour in the center of the
city.17 Inequality of travel times between public transport users and car
owners was sizeable. Mean travel time was 66.8 minutes for public transport users, whereas car owners faced a mean travel time of 42.6 minutes.18
These problems finally led to a rethinking of the entire bus transport system in Bogot, which culminated in the design and implementation of the
first phase of the TransMilenio system in 2001.
The Features of TransMilenio

The authorities designed and adopted TransMilenio to tackle the consequences of the market and policy failures described above. The key elements of the new system are concession contracts for service providers
(namely, TransMilenio transporters and feeder buses); vertical separation
of the transportation service and fare collection; bus remuneration based
on kilometers traveled rather than passengers transported; fare setting based
13. The road infighting among drivers for every extra passenger is locally known as the
penny war (la guerra del centavo).
14. Lozano (2003).
15. Chaparro (2002).
16. In 2000, Hong Kong, which has a population similar to Bogot (7,394,170 inhabitants), reported 12 million total daily trips.
17. Chaparro (2002).
18. Lleras (2003).
on long-term investment recovery following a tendering process (that is,

competition for the road); exclusive road lanes (the left lane of each road,
to facilitate turning at intersections by private automobiles); and exclusive
curbside service in metro-like stations, which constitute an organized system of express and slow routes and facilitate transfers.
As Lleras states, TransMilenio is a flexible, rubber-tired rapid transit
mode that combines stations, vehicles, services, and driving lanes . . . into
an integrated system.19 The first phase of TransMilenio was designed for
35,000 passengers per hour per direction; it covers 42.4 kilometers of
exclusive bus lanes along three of the main transit corridors. In these corridors, central lanes are dedicated exclusively to TransMilenio operations,
and passengers at the stations are their exclusive customers. The stations
are in the middle of the road like a metro system, which facilitates bus
transfers. The bus fleet consists of 470 buses for the first phase of the system,
with better mechanical conditions and environmental performance than
those of the traditional system. The bus stops include fifty-seven stations,
located every 700 meters, equipped with pay booths, registering machines,
surveillance cameras, and infrastructure such as bridges, pedestrian crossings, and traffic lights designed to ease the entrance of passengers into the
system. At the end of the corridors, three principal access stations serve as
a meeting point for feeder buses and buses from the traditional system that
work in neighboring municipalities. Feeder bus passengers have an integrated tariff, so riders do not have to pay twice for using the feeder system
and TransMilenio. Feeder buses, which share corridors with the traditional
system, have bus stops every 300 meters in the lower socioeconomic areas
of the city and are synchronized by the operators with a satellite system to
minimize travel and wait times for passengers.20
TransMilenio solved the agency problem present between owners and
drivers by establishing a prepayment scheme, in which users buy tickets in
booths located in the stations, as in underground systems around the world.
Therefore, bus drivers do not have to deal with collecting payment fares
like they do in the traditional system. Previously, drivers incomes depended
on the number of passengers traveling, whereas TransMilenio drivers are
salaried employees with no direct relation to fare collection.
The market failure deriving from an opaque property right definition for
roads and curbsides, which is present in different degrees in thick and thin
19. Lleras (2003).
20. Each TransMilenio bus has a capacity of transporting 160 passengers, while feeder
buses can transport seventy passengers.
163
mass transit markets, was solved via exclusive bus transit lanes and exclusive stations.
Finally, the recurrent problem of setting fares above market equilibrium
levels was solved with concession contracts awarded through public competition. The concessionaires income is no longer associated with the number
of passenger-trips serviced. Transport providers, both local and foreign, are
called to associate and participate under a new set of rules. The new transport
firms own a number of buses that meet certain specifications and whose operation is subject to the leadership of a central authority, TransMilenio S.A.
These buses have a specified service life stipulated at the outset of the contracts. The concessions expire when the vehicle fleet reaches an average
mileage of 850,000 kilometers, with no individual bus reaching more than
1,000,000 kilometers. If the average mileage threshold is reached in less than
ten years, then the concession extends to the tenth year. If the average
mileage threshold is not reached by the tenth year, the concession lasts until
this happens or until the fifteenth year, whichever comes first.21 The new contracts also establish a payment per kilometer traveled. The number of kilometers traveled depends on the manner in which the central authority
dispatches service and, ultimately, on demand. The reward per kilometer
traveled is the central variable by which the firms compete for these contracts.
TransMilenio S.A. carries out thorough calculations of the costs involved in the provision of the transportation service, as well as demand
forecasts, which enables the authority to set the range of acceptable fees
per kilometer traveled. This range guarantees a fair return to the participating investors and is used as a parameter in the selection process of the
concessionaires.22 TransMilenio thus represents a hybrid private-public
model that replaces competition on the road by competition for the road.23
These measures turned the entire incentive scheme of the previous system upside down. They eliminated the affiliating firm and the license business. They also removed any gains from cutting maintenance costs, since
the service is to be provided over a specified period under safety and efficiency standards: cutting down on costs today implies incurring greater
costs tomorrow to comply with the norms. These measures gave rise to a
21. There is a restriction by which the concessionaire is forbidden to add new vehicles
to the fleet to deliberately bring its average mileage down when a certain portion of the concession period has expired. This contract resembles those analyzed by Engel, Fischer, and
Galetovic (2001).
22. The range is given by a real return on investment between 14 percent and 16 percent.
true transportation firm able to earn a return on its capital comparable to

what it would receive in any other business of comparable risk.
Feeder buses are also organized through concession contracts awarded
through competition. These contracts are slightly different from the main
service contracts: they are awarded for a ten-year period subject to the condition that no individual vehicle should exceed 950,000 kilometers during
the time of the concession, and income to the feeder buses is defined as a
combination of revenues paid per kilometer traveled and per passenger
served. The reward system is designed to prevent opportunistic behavior
from the transport firms. Feeder buses bring passengers from outskirt locations into the TransMilenio system, but most of the time they do not transit on exclusive public transport lanes, making it difficult to control the
kilometers they travel. The regulating authority, TransMilenio S.A., has
calculated the maximum fixed operation costs that a firm will incur per
bus, and it agrees to remunerate the firms based on the number of passengers carried and kilometers traveled. Under this rule, traveling without
passengers becomes unprofitable for the feeder bus.
Taking the responsibility of collecting fares away from the bus drivers
and centralizing the management of fare revenues under an independent
fiduciary entity further contribute to the improved operation of the new system.24 Booths at the TransMilenio stations collect the systems revenues
and are managed through a fiduciary contract. This not only relieves the
system of information asymmetries and aids in the collection of taxes, but
also has a considerable impact on road safety.25
The last important element of the new scheme is the fare-setting procedure. Under TransMilenio, travel fares cease to be negotiable. The written
contracts subject the concessionaires to operate under the fares set by the
public regulating authority, TransMilenio S.A. The fare-setting procedure
and fare adjustment over time are part of the contracts.26 TransMilenio
travel fares are set to cover the long-run average costs of operation, includ24. In the previous system, passengers paid the drivers on entering the buses, which was
a distraction to the drivers and contributed to congestion and accidents.
25. At the moment, the regulator lacks a direct monitoring instrument for the fare collection contract.
26. The procedure for calculating the technical fare for TransMilenio comprises four
steps. First, the basis of the calculation is the pesos per kilometer offered by the main transport and feeder transport concessionaires, and the pesos per ticket offered by the revenue
collection concessionaire. The calculation also contemplates the cost of the fiduciary revenue management and TransMilenios management (both of which enter the formula as percentage shares). Second, the weight of each input in total operation costs is calculated using
165
ing the administration costs of the regulating authority and costs of the
fiduciary contract through which fare revenues are managed.
Authorities decided during the design stage that fares would not be set
to recover the infrastructure investments, since that would render the system unfeasible. TransMilenio fares set to recover the systems operating
costs were already significantly higher than the current bus fares. Imposing
on TransMilenio the obligation of recovering the infrastructure investments
or its maintenance costs would have resulted in negative profitability,
thereby eliminating the participation of the private sector. In addition, the
intervention on TransMilenio corridors includes a full renovation of two
lanes for private cars and the curbside; benefits from this investment accrue
not only to TransMilenio providers and users, but also to other population
groups. Thus investment in the infrastructure required to operate the system
was said to be social investment.
After the opening of the first TransMilenio lines, traditional buses continue to mobilize nearly 80 percent of the population. The TransMilenio
concession contracts obligated the operators to purchase and scrap buses
that previously operated in TransMilenio corridors. Only 1,410 of these
buses were effectively scrapped, however, and the remaining 4,670 buses
were relocated into unserved TransMilenio corridors. Although the
demand for the old system dropped, causing revenues per vehicle to fall, a
loose regulation of the system permitted the excessive entry and overprovision of services in these corridors. Moreover, the fares for the traditional
system were still allowed to rise after TransMilenio came into operation,
as illustrated in figure 4. The introduction of TransMilenio thus did not
spur a substantial reduction in the transport fleet of buses from the traditional system.
Some Political Economy Considerations of TransMilenios Adoption

Before the adoption of TransMilenio, the local authorities biggest concern
was the opposition the system could face from traditional transporters.
These companies had derived substantial rents over the decades, and they
the numbers contained in the contracts, and costs are adjusted assuming a monthly average
of 6,400 km per vehicle. Third, to conform to available coin values, the passengers fare is
rounded to the nearest multiple of fifty. Finally, the difference per ticket goes into a fund
that compensates for cost increases not reflected in the fare. Costs are revised monthly, and
the fare is revised (up or down) to the next closest multiple of fifty when the technical fare
changes by more than twenty-five pesos.

F I G U R E 4 . Nominal Fares: Buses versus TransMilenio, 19952003
Fare
1200
1000
800
Buses
TransMilenio
600
400
200
De
c-9
Ap 5
rAu 96
g-9
De 6
c-9
Ap 6
rAu 97
g-9
De 7
c-9
Ap 7
rAu 98
g-9
De 8
c-9
Ap 8
rAu 99
g-9
De 9
c-9
Ap 9
rAu 00
g-0
De 0
c-0
Ap 0
rAu 01
g-0
De 1
c-0
Ap 1
rAu 02
g-0
De 2
c-0
Ap 2
rAu 03
g-0
De 3
c-0
3
0
Month
Source: TransMilenio S. A.
might feel threatened by the TransMilenio intervention. However, the firms

were also aware that problems related to public transportation, such as congestion, pollution, and accidents, had reached worrying levels. A growing
consensus thus emerged in the late 1990s in favor of a regime change, in
the form of either a subway or a transformation of the bus system.
Bogot witnessed several episodes of protest against the new system.27
The strikes and protests lost momentum quite rapidly, however, and
demonstrations on other issues consistently outnumbered TransMileniorelated incidents. Between 1999 and 2000, five events were organized against
TransMilenio, and TransMilenio-related protests represented only one-fifth
of total political demonstrations in the sector even at their peak. Activity
against this initiative had already ceased by 2001.
An important factor facilitating the transition was the national governments decision to support a new transportation system for Bogot. The
central governments willingness to partly finance the associated infrastructure investments became a critical argument to convince transporters
that this was the right time for reform.28
27. Estache and Gmez-Lobo (2005) document similar reactions in the case of Santiago,
Chile.
28. In the 1990s, Colombian presidents supported either a subway or a bus solution.
Elected Bogot mayors followed suit because any solution would require substantial financial
resources. Consequently, local authorities choices regarding the new public transport system
were partly determined from the outside. The city authorities revealed the TransMilenio blue
print in 1999 after a new national government rejected the subway alternative for Bogot.
167
Finally, the municipal government decided that any solution would be

pursued in association with the traditional transportation firms of Bogot,
and they used the allocation of new routes in the non-TransMilenio corridors to reward affiliating firms for their willingness to participate in TransMilenio. The traditional transporters relocated their buses to new routes
granted by TransMilenio authorities themselves, and not by the traditional
regulator. The municipal representatives thus chose not to abandon the old
businesses in the proportion in which TransMilenio replaced them, but
simply to relocate them. Unfortunately, this produced damaging effects on
the other corridors in terms of speed, congestion, pollution, and riding time,
as illustrated below.
Cost-Benefit Analysis of the First Phase of TransMilenio

The ultimate goal of TransMilenio is to improve the quality of life in
Bogot. Although only 25 percent of the system has been put in place so
far, its impact is indicative of what can be achieved. Travel times for TransMilenio passengers have dropped, and traffic congestion, air pollution, noise
levels, and frequency of traffic accidents have decreased significantly in
TransMilenio corridors. However, some unexpected negative spillovers
have emerged. Traffic congestion and pollution have heightened along
corridors not served by TransMilenio owing to the slow scrapping rate of
buses in the traditional public transit system. This section presents a costbenefit analysis of the first wave of TransMilenio contracts.
The Impact of TransMilenio on the Quality of Life in Bogot

Soon after TransMilenio came into operation, many public transportation users switched to this new public transit mode. The year before the
introduction of TransMilenio, 69 percent of individuals relied on public
transportation. After TransMilenio became available in 2001, 6 percent
of individuals switched from traditional public services to the new alternative. This share of the population using TransMilenio has expanded continuously, and the new system served 13 percent of the population in
2003.29 TransMilenio has not increased the overall attractiveness of public
transit, however, since the share of the population using cars has remained
29. The first phase of TransMilenio, which represents 25 percent of the total expected
network, covers 13 percent of the public transportation demand in Bogot.
constant since its adoption. Thus the demand for public transport as whole
did not expand, but rather the appearance of a new transportation mode led
to a distribution of this demand between TransMilenio and the traditional
system.
Demand for public transportation stems mainly from the lower-income
households, yet TransMilenio users are concentrated in the fourth and
fifth income quintiles, as seen in figure 5. In contrast, demand for traditional buses is primarily from the first, second, and third quintiles. The
highest income quintiles thus use TransMilenio more than the lower in-
F I G U R E 5 . Use of Public Transportation, by Income Quintile

Percent of
population
75
A. Transport Mode by Income Quintile
Bus
50
25
Private vehicle
0
Percent of
population
10
3
Income quintile
B. TransMilenio Usage by Income Quintile
8
5
3
0
Source: Authors calculations, based on DANE (2003).
3
Income quintile
169
T A B L E 1 . Time Spent Traveling from Home to Work

Percent of individuals
Time spent
Less than one hour
Between one and two hours
More than two hours
2001
2002
2003
74
23
4
68
29
2
74
17
5
Source: Authors calculations, based on Napoleon Franco (www.napoleon.com.co)
come quintiles.30 The top two quintiles are also the main users of private
vehicles.
TransMilenios impact on travel times and average speed varies widely
across users of TransMilenio, private vehicles, and the traditional public
transport system. Two years after TransMilenio began operating, the average trip time in Bogot decreased to 35 minutes, from 44 minutes in 2001.
The average speed of cars increased to 20 miles per hour, from 16.8 miles
per hour in 2000.31 Table 1 shows that the percentage of individuals who
spent more than an hour traveling from home to work fell from 23 percent
to 17 percent, while the frequency of trips under one hour rose during 2003.
The benefits of travel time reductions accrued mainly to TransMilenio
users. The average speed of other forms of public transportation actually
dropped, which caused travel times to increase by 10 percent.32 Slow scrapping of buses from the traditional system may be causing the uneven distribution of benefits. An important share of these buses were relocated to
corridors not served by TransMilenio, worsening congestion. Attributing
worsening congestion solely to the relocation of buses is not accurate,
however; slow adoption of other measures to control traffic congestion
and the increase in the number of taxis, as a result of simultaneous interventions, are partially responsible, as well.
Lleras studies TransMilenios impact on public transportation users,
mainly through travel times.33 In 2002 a revealed preference survey was
30. Of all TransMilenio users, 62 percent (325,925 users) correspond to the fourth and
fifth income quintile, while 38 percent (203,830 users) correspond to the first three income
quintiles (DANE, 2003).
31. See www.transitobogota.gov.co.
32. To date, the impact on travel times for riders using unserved corridors has not been
officially measured. The union of small providers of public transportation argues that travel
times in unserved corridors increased by 10 percent as a consequence of bus relocation
(Pico y Placa seguir en discussion, El Tiempo, 11 March 2001).
33. Lleras (2003).
applied to 2,095 public transport riders who could choose between TransMilenio and the traditional system for their transportation. Respondents
were interviewed in the street near areas where routes from the two systems run. The survey elicited information about the transportation mode
selected, the expected attributes of the trip, and various socioeconomic
characteristics. The data were used to estimate random utility models,
which provide the coefficients to calculate the value of time for users of
both systems in the different stages of the process (for example, walk-in
and walk-out times). The study shows that declines in travel times were
not uniform across TransMilenio users. TransMilenio passengers starting
the trip in the vicinity of the main corridors travel 12 minutes less per trip
than passengers of the traditional system. In contrast, passengers requiring
one or more transfers did not experience drops in travel times. In fact, total
travel time is two minutes shorter in the traditional system because of the
waiting time required for the TransMilenio feeder routes.34
Waiting-in and waiting-out time increased for all TransMilenio users,
however, because these times are extremely low for the traditional bus
system.35 TransMilenio users have to buy tickets, wait in line, exit the station, and walk to the final destination, whereas users of the traditional system enter the bus at any location, pay the bus driver directly, and stop the
bus at the point nearest to their destination, since there are no official bus
stops in Bogot. Improvements in TransMilenio travel times arise, therefore, from in-vehicle travel time.
Estimations of the value of time for TransMilenio passengers, vis--vis
users of the traditional system, show unambiguous improvements in traveling conditions for the former. Overall, people are willing to pay less for
savings in travel time than people using the traditional system, indicating
that TransMilenio is a less painful experience. For example, the value
of waiting time for the traditional system is U.S.$3.08 per hour, while in
TransMilenio it is U.S.$1.14 per hour.36
These figures indicate that the first phase of TransMilenio enhanced
travel conditions for a certain population group, although improvements
34. Lleras (2003).
35. Waiting-in time refers to the time spent waiting for the bus; waiting-out time is calculated from the time the passengers step off the bus until they arrive at their destination.
Lleras (2003) estimates that waiting-in and waiting-out times increased 2.95 and 5.16 minutes, respectively, for TransMilenio users vis--vis users of the traditional public transport
system.
36. Lleras (2003).
171
F I G U R E 6 . Traffic Accidents in TransMilenio Corridors before and after the Intervention

Number of incidents
1,600
Before TransMilenio
2002
2003
1,400
1,200
1,000
800
600
400
200
0
Traffic collisions
Pedestrian accidents
Injuries
Fatalities
Source: TransMilenio S. A.
should not be entirely attributed to the new mass transit system.37 Nevertheless, benefits from the new system are not distributed widely across the
entire population; gains were perceived largely by TransMilenio users, in
particular downtown dwellers.
The quality of services also improved, and TransMilenio corridors
experienced a decline in the external costs caused by public transport
namely, accidents and pollution. Traffic accidents decreased dramatically
in TransMilenio corridors. Figure 6 depicts the incidence of traffic collisions, pedestrian accidents, injuries, and fatalities in TransMilenio corridors
before and after the system came into operation. After two years, traffic
collisions and pedestrian accidents decreased by 94 percent, injuries to
passengers by 76 percent, and fatalities by 94 percent. This is, by all accounts,
an impressive performance.
Finally, air pollution, which is a major concern in Bogot, exhibited a
declining trend in TransMilenio corridors. Half of the districts in the city
currently exceed the particulate matter (PM-10) and ozone pollution limits. Studies indicate that automobiles are the most significant emission
37. Bogots authorities implemented several programs to transform traffic conditions,
including mobilization restrictions during traffic peaks, investment in road infrastructure,
and increased traffic fines.

F I G U R E 7 . Average Daily Readings of Particulate Matter, Citywide and in Monitoring
Stations near TransMilenioa
Micrograms per
cubic meter
80
1999
2000
2001
60
40
20
0
Daily average -Monitoring
station 1
Daily average - Monitoring

station 2
Daily average - Bogot
Source: Authors calculations based on data from the Bogot Air Quality Monitoring Network (Red de Monitoreo de la Calidad del Aire
de Bogot, or RMCAB).
a. The RMCAB maintains thirteen monitoring stations, located throughout Bogot, that measure hourly emissions for a group of
pollutants.
source in Bogot, contributing 70 percent of air pollution.38 Reductions in

traffic congestion increased the speed of other vehicles and thus curbed
emissions in TransMilenio corridors. TransMilenio also appears to have
improved air quality by transporting more passengers in less time with better vehicles.39 In 200002 the citywide average of PM-10 emissions grew
by 23 percent, but it fell by 8 and 11 percent in the two TransMilenio corridors shown in figure 7. Although the environmental authorities of Bogot
adopted a group of measures to control air pollution over the last decade,
evidence indicates TransMilenio contributed substantially in this respect.40
38. Cavallazi (1996).
39. This superior performance of TransMilenio vehicles contributes to the control of
mobile source emissions. TransMilenio buses transport 1,596 passengers daily, which is five
times the number of passengers for traditional public buses, and the average speed is 17 to
44 percent faster. The average age of public buses is fifteen years. TransMilenio buses have
catalytic converters, emissions are below the limits required by the Euro II norm, some vehicles use natural gas, and noise levels are less than ninety decibels (Ibez and Uribe, 2003).
40. Measures include the use of catalytic converters for new cars, mandatory inspection
and maintenance, and mobilization restrictions.
173
Mndez analyzes the impact of several pollution programs implemented in Bogot.41 Her study evaluates the evolution of particulate matter and ozone in monitoring stations located near TransMilenio corridors
and far from industrial areas from 1997 to 2002. Mndez defines categorical variables to denote the implementation of three policy interventions
(namely, the introduction of TransMilenio, mobilization restrictions,
and mandatory inspection and maintenance) and then estimates time
series regressions to predict the trend of particulate matter and ozone
levels, controlling for these policy interventions. Her results reveal that
TransMilenio is the most effective program for curbing pollution, with a
much stronger impact than programs specifically designed to control
emissions. Mandatory inspection and maintenance reduced ozone levels
by 13.6 percent, while restrictions on car mobilization lowered ozone
levels by 21 percent. The first phase of TransMilenio, however, produced a 28.8 percent decline in ozone levels. Similar estimations for particulate matter indicate that TransMilenio abated PM-10 pollution levels
by 9.2 percent.
Air pollution levels rose in other areas of the city, because the slow
scrapping rate of buses from the traditional system triggered the relocation of buses to non-TransMilenio corridors. To evaluate the negative
pollution spillovers of TransMilenio, we used a difference-in-differences
approach to compare readings from a monitoring station located near a
TransMilenio corridor to readings at a baseline monitoring station. The
baseline monitoring station should have two characteristics: it should
have similar particulate matter readings to the TransMilenio monitoring
station before entry into operation of the system; and it should be located
near a non-TransMilenio corridor without negative spillovers from bus
relocation. Other interventions in the city can also affect the results,
namely, the entrance of new taxis and the construction of new TransMilenio corridors. The evolution of particulate matter in both monitoring
stations is illustrated in figure 8.
Readings of PM-10 soared in the monitoring stations with negative spillovers after TransMilenio fully entered into operation in June 2001.42 It
remained constant in the monitoring station without the negative spillovers. The difference-in-differences calculations indicate that bus reloca41. Mndez (2004).
42. TransMilenio started partial operations in January 2001, and the full system with the
three trunk corridors was in place in June 2001.

F I G U R E 8 . Particulate Matter Readings from Monitoring Stations with and without
Negative TransMilenio Spillovers, 19982002
Monthly average PM-10
70
60
With
Without negative spillovers
With negative spillovers
50
40
30
20
Mar-98
May-98
Jul-98
Sep-98
Nov-98
Jan-99
Mar-99
May-99
Jul-99
Sep-99
Nov-99
Jan-00
Mar-00
May-00
Jul-00
Sep-00
Nov-00
Jan-01
Mar-01
May-01
Jul-01
Sep-01
Nov-01
Jan-02
Mar-02
May-02
Jul-02
10
Month
Source: Authors calculations, based on data from RMCAB.
tion caused particulate matter emissions to increase by 10.5 percent, more

than offsetting the reductions in PM-10 readings in TransMilenio corridors. Although the change cannot be attributed solely to TransMilenio,
this was the main intervention in Bogots traffic system in the period.
Air pollution can cause serious damages to health, and particulate matter and ozone levels appear to increase the incidence of acute respiratory
illnesses in Bogot. Lozano estimates a concentration response function
that links daily respiratory hospital admissions in 1998; the study finds a
strong link between incidence of respiratory health admissions and particulate matter levels.43 In fact, a 25 percent increase in particulate matter
causes a 21.8 percent rise in respiratory health admissions, while decreasing particulate matter emissions by 25 percent results in a 17 percent reduction in respiratory health admissions.
The impact of the first phase of TransMilenio reveals positive results
for the people attended by the system, but ambiguous results for the rest of
the city. Travel times for TransMilenio users fell, safety improved, congestion dropped, and air quality increased, yet these benefits are restricted
43. Lozano (2003).
175
to TransMilenio corridors. Meanwhile, the relocation of buses increased

congestion in non-TransMilenio corridors, causing a deterioration in traveling conditions for passengers of at least some corridors of the traditional
system and a rise in pollution levels. Once the full system comes into operation, the negative spillovers should decrease significantly or disappear.
Cost-Benefit Analysis
Understanding the welfare gains and losses from the introduction of TransMilenio is crucial for evaluating whether investments in this new scheme
produced net benefits for the population and identifying future adjustments to the system. Welfare changes can be approximated by measuring
variations in the consumer and producer surplus caused by the adoption of
TransMilenio. We used the above information and data on investments,
costs, and revenues of the systems operation to conduct a cost-benefit
analysis. Only the first phase of TransMilenio (25 percent of the system)
was evaluated. Ideally, we should estimate the consumer and producer
surplus before and after TransMilenio and then calculate the changes in
both surpluses; however, data restrictions limited our analysis to calculating changes in both surpluses as a result of the adoption of the system. A
detailed description of the assumptions and methodology used to perform
the analysis is included in appendix B. Table 2 presents a summary of
results.44
Who wins and who loses with TransMilenio? The previous section
suggests how benefits and costs are distributed among different groups of
the population. We now undertake a detailed cost-benefit analysis to clarify those findings. On the production side, benefits arise for the private
TransMilenio operators but are lost for the owners of the scrapped buses.
The producer surplus of private TransMilenio operators represents welfare gains. In contrast, the former producer surplus that the owners of the
scrapped buses no longer collect represents welfare losses. The shaded
areas depicted in figure 9 denote those welfare gains and losses.
On the demand side, the evaluation of TransMilenio must account for
changes in travel times faced by TransMilenio users and users of the traditional system. An appropriate welfare indicator for these changes is the
willingness to pay for improvements in travel time. The value of time is
defined as the marginal rate of substitution between travel time and a mon44. Detailed calculations of the costs and benefits are available on request.

F I G U R E 9 . Welfare Gains for Private TransMilenio Operators and Welfare Losses
for Scrapped Busesa
A. Gains for private TransMilenio operators
Fare
B. Losses for owners of scrapped buses

Fare
STS(F)
STM(F)
FATM
FBTM
TATM
Number of
trips
TBTM
Number of
trips
a. The shaded area denotes welfare gains (panel A) and losses (panel B). The subscript BTM indicates the fare and number of
passengers in TransMilenio corridors before TransMilenio, and the subscript ATM indicates the fare and number of passengers after
TransMilenio. Supply curves for private TransMilenio operators and for owners of scrapped buses (STM and STS, respectively) are
assumed to depend on fares (F).
etary cost. Figure 10 presents welfare changes resulting from the adoption
of TransMilenio. When TransMilenio started operations, TransMilenio
users experienced two welfare enhancing effects: travel time decreased,
and the ride was more pleasant than on the traditional system.45 The movement from point A to point B in the indifference curve map for TransMilenio users shows the combined effect on utility; X 20 X 12 measures
willingness to pay for these improvements in traveling conditions. Riders
of the traditional system faced costs from increases in travel time, while their
traveling conditions were practically unchanged. As a result, travel times
rose, but the marginal rate of substitution between savings in travel time
and the numeraire good remained constant. By moving from the status quo
(point C) to the new condition (point D), riders of the traditional system
experienced welfare losses, represented by X 22 X 23.
The adoption of TransMilenio also produced both positive and negative
impacts on pollution readings, leading to gains and losses for the popula45. The improved quality of the ride reduces willingness to pay for improvements in
travel time.
177
F I G U R E 1 0 . Welfare Effects from Impacts on Traveling Conditions and Travel Timesa

A. TransMilenio users
B. Traditional system passengers
X2
X 20
X2
Losses
A
X 22
X 23
Gain
U0
U1
X 21
U0
X1
U1
X1
a. The horizontal axis represents savings in travel time (X1), and the vertical axis represents a numeraire good measured in monetary
units (X2).
tion. Residents located near TransMilenio corridors experienced welfare

gains from drops in pollution, while bus relocation heightened congestion
and deteriorated air quality in some of the unserved TransMilenio corridors. To establish the welfare impact of these opposite effects, we use the
marginal damage function (which corresponds to drops in utility as a result
of pollution, measured in monetary terms) to calculate welfare changes
from the adoption of TransMilenio. Figure 11 shows how to value these
welfare impacts for both groups of the population.
Finally, society as a whole benefits from the decrease in fatalities in
TransMilenio corridors stemming from enhanced traffic and security conditions. We use the value of a statistical life, which shows aggregated individuals willingness to pay for risk reductions, to measure reductions in
the risk of death in TransMilenio corridors.
This series of figures offers an insight into who wins and who loses after
the adoption of TransMilenio. We now perform a cost-benefit analysis to
establish whether the benefits offset the costs or whether the costs from
negative spillovers offset the benefits provided to TransMilenio users. The
cost-benefit analysis incorporates the revenues and costs from TransMilenio operations, as well as initial investment. Operational information
comprises revenues from fare collections, operational costs for concession-

F I G U R E 1 1 . Welfare Gains for Residents near TransMilenio Corridors and Welfare Losses
for Residents near Unserved Corridors
A. Gains to residents near
TransMilenio corridors
Marginal
damage
B. Losses to residents near unserved corridors

with negative spillovers
Marginal
damage
EATM
EBTM
PM-10
EBTM
EATM
PM-10
aires, fare collection expenses, administrative expenditures of TransMilenio

S.A., and management costs for the fiduciary. Infrastructure investments
and maintenance costs of corridors are also incorporated.
We also calculated the forgone revenues of firms operating previously in
TransMilenio corridors. Before TransMilenio came into operation, nearly
6,080 buses operated in those corridors; of those, approximately 1,410 were
scrapped, and the remaining 4,670 were relocated to unserved TransMilenio
corridors. Forgone revenues were estimated for the scrapped buses.
Our estimation of welfare losses and gains from variations in travel
times for users of both TransMilenio and the traditional system considers
all stages of traveling from one destination to another: waiting time, invehicle travel time, and walking-in and walking-out time. Lleras calculates
the value of time for each stage.46 Public transportation demand encompasses two groups: users that require one or more transfers and users that
travel directly to their destination. The estimation includes the values calculated for both segments.
We also consider the benefits and costs associated with changes in particulate matter levels. Mndez provides figures from emissions reductions
46. Lleras (2003).
179
in TransMilenio corridors.47 Emissions increases from negative spillovers

were calculated using a difference-in-differences approach. We calculated
the impact of particulate matter pollution on the incidence of acute respiratory illnesses based on a concentration-response function estimated by
Lozano for Bogot.48 We then valued the changes in the incidence of
acute respiratory illness as a consequence of TransMilenio using willingness to pay for reductions in respiratory health admissions, also reported
by Lozano.49
Finally, we calculated the gains from reductions in traffic fatalities in
TransMilenio corridors. We estimated the risk of death before and after
TransMilenio based on the number of passengers and the number of deaths.
Bowland and Beghin report the value of a statistical life (VSL) for Chile;
we adjusted these values for Colombia using the GNP per capita for both
countries.50
The analysis does not incorporate two main welfare benefits accruing
from TransMilenio as a result of data restrictions, namely, gains from
reduced traffic collisions and injuries (reported in figure 6) and gains from
the reduction in petty thefts and general insecurity in TransMilenio corridors. Data simply are not available for calculating the welfare benefits
from these variables.
Table 2 presents the net present value of the cost-benefit analysis for
different discount rates.51 The calculation of the net present value occurs
in two stages: we first calculate the net benefits for TransMilenio corridors, without accounting for negative spillovers on unserved corridors and
then incorporate the effects on unserved TransMilenio corridors. Results
reveal that the benefits in TransMilenio corridors significantly offset costs;
net benefits are U.S.$3.7 million for a 9 percent discount rate. Once negative spillovers are incorporated in the analysis, net benefits are negative, at
47. Mndez (2004).
48. Lozano (2003).
49. Lozano (2003).
50. Bowland and Beghin (1998).
51. We also calculated net benefits, considering infrastructure investments (a net loss of
U.S.$64 million for a discount rate of 9 percent). However, there are several arguments
against including infrastructure investments in the cost-benefit analysis. First, the bulk of
infrastructure investment is street pavement, most of which would have been done anyway
and is the responsibility of city authorities. Second, a full renovation of two lanes for private cars and the curbside was required for TransMilenios operation, such that benefits
from this investment not only accrue to TransMilenio providers and users, but also to other
population groups. Third, this infrastructure lasts for decades, allowing city dwellers to reap
the benefits over a very long period.

T A B L E 2 . Cost-Benefit Analysis of TransMilenios First Phase
Millions of 2002 dollars
Discount rate
Item
Forgone revenues, traditional system
Forgone operational costs, traditional system
Revenues, TransMilenio
Operational costs, TransMilenio
Costs from increased waiting time for TransMilenio, no transfer
Costs from increased waiting time, transfer
Costs from increased walking in and out time, no transfer
Costs from increased walking in and out time, transfer
Benefits from reductions in travel time, no transfer
Benefits from reductions in travel time, transfer
Benefits from a 9.1 percent reduction in PM-10 in TransMilenio corridors
Benefits from reductions in mortality
Net present value, TransMilenio
Travel time costs for traditional system users, no transfer
Travel time costs for traditional system users, transfer
Costs from a 10.5 percent increase in PM-10 in traditional corridors
Net present value, traditional system
Total net present value
7 percent
9 percent
12 percent
3.628
2.702
6.858
6.005
0.608
1.686
2.766
3.026
6.217
5.852
0.283
0.832
5.03
2.720
2.025
5.046
4.410
0.447
1.240
2.034
2.225
4.572
4.303
0.212
0.625
3.71
2.114
1.574
3.870
3.378
0.343
0.950
1.559
1.705
3.504
3.298
0.165
0.486
2.85
5.014
10.239
0.357
15.61
10.59
3.762
7.682
0.268
11.71
8.00
2.925
5.973
0.208
9.11
6.26
U.S.$8 million. The increased travel times for users of the traditional system, resulting from bus relocation, drive these results. Pollution in nonTransMilenio corridors also reduces benefits, but the contributions are not
significant and are partially offset by drops in pollution in TransMilenio
corridors. Benefits stemming from less in-vehicle travel time for TransMilenio users, operational revenues, and reductions in mortality, although
significant, are not sufficient to cover these costs.
Because congestion costs are highly nonlinear, the congestion relief
brought about by TransMilenio has been more than offset by increases in
congestion in others areas. Currently, TransMilenio covers 13 percent of
the public transportation demand, whereas the public transportation system covers the remaining 87 percent. TransMilenio has indeed improved
the traveling conditions of its users, but it has worsened the conditions of
a high proportion of the populationnamely, users of the traditional public system.
Negative spillovers have resulted from three features of the current
mass transit system of Bogot. First, only 25 percent of the TransMilenio
181
network has been constructed so far. Because the TransMilenio network is

only partially operating, the benefits arising from the economics of coordination and density are not being fully exploited. Second, the scrapping
rates required in the first phase of TransMilenio contracts are insufficient.
Buses that were previously operating in TransMilenio corridors and that
should have been scrapped were relocated to unserved corridors, which
heightened congestion. Finally, high fares and imperfect controls on the
entry of new buses into the traditional public system promoted the entry of
new buses and slowed the exit of old ones. In fact, the second wave of contracts defines more stringent scrapping rates than did the first phase. Fares
are still excessively high, creating an incentive for the entry of new buses.52
In addition, controls on the entry of new buses are difficult to enforce,
allowing so-called pirate buses to operate citywide.
There are two important caveats to these results. First, our analysis does
not incorporate some important benefits, like fewer traffic accidents and
better security in TransMilenio corridors, because of data restrictions. These
benefits would certainly change the results. Second, given that TransMilenio
will eventually serve the vast majority of the road network, negative spillovers from bus relocation should disappear once the network is complete.
Nevertheless, the analysis sheds some light on drawbacks of the TransMilenio system that should be adjusted for the next phases. The design of
the system should take into account negative spillovers to the traditional
public transportation system in order to minimize them. Moreover, TransMilenio and the traditional system should not operate separately, but should
be integrated into a single mass transit system. Finally, the prevailing incentives (such as high fares) that spur the entry of new buses, despite the
fact that many of them travel empty, should be eliminated.53
Concluding Remarks: The Road Ahead

TransMilenio has become a cornerstone in the history of urban public
transportation in Colombia. Today at least five Colombian cities with over
half a million people each are applying for central government resources
to replicate the Bogot system. Other cities in Latin America are planning
to follow suit. The design of TransMilenio mimicked the interventions in
52. Castro and others (2001).
mass transit systems of Curitiba, Brazil, and Quito, Ecuador, but it introduced some novelties that have become its trademark. The new transit
system is a hybrid model that combines public planning of the network
structure, route tendering conditions, regulation, and supervision, with private operation of the separate functions of revenue collection and transport
service. Among the most salient features are the financial fiduciary management; the division of services among separate private providers that
simultaneously administer the firm TransMilenio S.A.; the flexible contracts for bus operation; the separation of concessions for feeder buses and
regular buses; the payment per kilometer instead of per passenger; the definition of property rights for the road and the curbside; and the use of the
faster left lanes.
This paper is the first to provide a full economic account of the origins,
design, political economy, and costs and benefits of TransMilenio. The
system had a sizeable impact on users by improving travel conditions significantly. Congestion, pollution, and traffic accidents fell sharply in TransMilenio corridors. But the type of transition adopted for corridors not
covered by TransMilenio caused unforeseen negative spillovers, as a consequence of slow scrapping rates and bus and route relocation. Although
the cost-benefit analysis for the first phase of the corridors covered by
TransMilenio is positive, the citywide net effect is negative mainly as a
result of increases in travel time for passengers using the traditional transport system. To minimize the negative spillovers during the full implementation of TransMilenio, integration of the traditional and new systems
should continue, and strict regulation of the traditional public transportation system should be crafted.
There remain potential vulnerabilities. As Estache and Gmez-Lobo
indicate, the institutional capacity of the national and municipal planning
authorities for defining the network configuration, quality requirements,
and service levels are crucial.54 One of the advantages of the private system
is its flexibility to modify routes, cover new developing areas of the city,
and bring bus routes closer to consumers. These challenges now need to
be answered satisfactorily by good planning and regulation. The tendering
system requires special care to ensure competition and avoid collusion
among potential bidding concessionaires, as has been argued in the case of
Pereira, the second Colombian city to implement a TransMilenio-type system. The application of contracts also needs to be monitored. This is par54. Estache and Gmez-Lobo (2005).
183
ticularly true for the revenue collection contract, whose performance is

central to the systems profitability.
Recent events demonstrate other deficiencies in Bogot. The number of
daily passengers has increased from 800,000 to 900,000, and the number
of vehicles on the road remains the same. Passenger jamming has become
the rule, and the problem is an especially severe problem during peak
hours. Long lines to purchase access cards are coupled with crowds waiting at the curb to enter the buses. People find it difficult to board the buses.
Waiting time, a critical variable in any public transportation system and one
of TransMilenios big achievements early on, has started to rise. Security
has deteriorated, and theft is common. As a result, people have begun to
complain about the TransMilenio monopoly.55
Additional vulnerabilities are related to financing and political support.
TransMilenio depends on fiscal resources from the city and the nation.
Both political will and financial difficulties could prove obstacles in the
future. The last two municipal administrations strengthened the systems
institutional framework. The new national and municipal administrations
that entered office in August 2002 and 2004 had earlier criticized aspects
of TransMilenio, but since taking office these new leaders have embraced it
and promised to improve it.
Finally, the fact that a democratically elected mayor appoints the head
of the public company, TransMilenio S.A., poses a potential problem. One
or two individuals can decide the future of the entire system. The system
needs to strengthen an impersonal and institutionalized regulatory arrangement. One such decision might involve the scrapping process of the first
lines of TransMilenio.
The implementation of TransMilenio holds several policy lessons for
cities in developing countries planning to reform their public transportation system. The design of TransMilenio successfully reduced many market failures plaguing the provision of public transportation in developing
countries. This hybrid system, in which the public sector participates in the
regulation of the system and network configuration while the private sector operates the buses, avoids the problems inherent in both public monopoly or unregulated private provision. Breaking the link between passengertrips serviced and profits eliminates the incentives to atomize the bus
industry, removes the perverse signals to use smaller buses, and enhances
55. Transmi-Lleno (or Transmi-full) became slang for TransMilenio. On 10 March
2004, passengers protested by sitting on the road after an accident delayed a long queue of
TransMilenio buses for hours.
traffic conditions. The allocation of exclusive property rights over a route

abolishes the excess supply of buses, and the competitive tendering of the
right to operate on a route pushes fares towards efficient levels.
But the gradual implementation of the new system along with the parallel
operation of a poorly regulated traditional system brought unexpected negative results. Since congestion costs are highly nonlinear, the welfare losses
from heightened congestion in unserved TransMilenio corridors more than
offset the benefits from TransMilenio, even though those benefits are sizable.
This suggests the adoption of a new public transportation system must run
parallel to an appropriate regulation of all other public transit providers.
Appendix A: Econometric Estimates for Rents

To calculate the economic rents for affiliating firms and bus drivers, we
estimated the demand per vehicle and the income per passenger using time
series regressions, using two approaches. The first approach estimates an
ordinary least squares (OLS) regression that corrects for autocorrelation;
the second approach estimates an autoregressive moving average exogenous variables (ARMAX) model.
The observed demand per vehicle is estimated with a distributed lag
model in which a lagged variable for the number of passengers is included
as an explanatory variable:
qt = 0 + 1qt 1 + 2 vt2 + t + t 1
where qt is the daily number of passengers per vehicle, qt 1 is the lagged
daily number of passengers, and vt represents the number of buses. We
estimated several models. The model that best approximates the functional
form is presented table A1.
The ARMAX model provides the best estimation, and autocorrelation
is eliminated. The signs of the coefficients are as expected. 0 is equivalent to autonomous demand, 1 incorporates the influence of prior periods,
and 2 includes the influence of size of the total fleet, which is negative in
all the estimates.
To estimate income per passenger, the production function curve for a
congestion good is estimated using a polynomial regression model.
Yt = 0 + 1V + 2 v 2 + 3 v 3 + t + t 1 ,
185
T A B L E A 1 . Econometric Estimates of Demanda

Coefficient
0
1
2
OLS model
ARMAX model
605.2
0.15 106
123.34
0.8064
0.167 10
40648.9
0.24*
0.3363**
6003.83
=0
Sum of the square of the errors

E(t)b
*Durbin Watson; ** statistically significant at the 2 percent level.

a. All figures are statistically significant at the 1 percent level unless otherwise specified.
b. E(t) = 0 implies that autocorrelation does not persist.
T A B L E A 2 . Econometric Estimates of Income per Passenger

Coefficient
0
1
2
3
OLS model
AR(1) model
ARMAX model
52.8123
0.1212 101
0.1062 105
0.2484 1010
21.4539
0.5918 103
0.1053 106
0.2776 1011
33.5797
0.7134 102
0.6694 106
0.1538 10
284.49
0.7466*
0.9674
139.77
0
0.8316
137.78
=0
10
Sum of the square of the errors

E(t)a
*Durbin Watson.
a. E(t) = 0 implies that autocorrelation does not persist.
where Y represents the income per passenger, which is a variable that

approximates the fare. To correct for autocorrelation, ARMAX and firstorder autoregressive, or AR(1), models were estimated. The results are
presented in table A2.
Calculating rents involves calculating the gains and costs to each vehicle. Gains are determined by the number of vehicles in circulation, for
which a production function of the public transportation sector is defined.
Costs per vehicle include fuel expenses, parts replacement, maintenance,
depreciation, insurance costs, taxes, and purchase value of the vehicle.1 We
calculated the average cost per vehicle for each year. Rents for the affiliating firms were estimated by adding up the benefits for the buses they own,
revenues from routes allocations, and operational costs.
Contracts between the affiliating firm and bus owners are simulated by
maximizing benefits from affiliating firms subject to positive benefits for the
1. The prevailing interest rate for each year was used.
bus owners. We thus obtained the optimal number of vehicles per affiliating
firm, the affiliation fee, and rents for the affiliating firm and bus owners.
Appendix B: Methodology of the Cost-Benefit Analysis

We performed the cost-benefit analysis for the length of the first wave of
the contract: 200015. All values are in millions of 2002 U.S. dollars. Population growth rates for 2004 and 2005 are projections from the National
Administrative Department of Statistics (DANE); for 2006 onwards, we
used the average population growth for 200105.
We calculated TransMilenio revenues by multiplying the number of
TransMilenio users by the technical fare. For 2001, 2002, and 2003, we
used observed values for the number of passengers, as reported by TransMilenio S.A. For 2004 onwards, the revenues were calculated assuming
that the number of passengers increased with the population growth rate.
We estimated the revenues for the traditional bus system as follows.
First, the number of passengers on the Caracas and Avenida 80 corridors
before the implementation of TransMilenio was obtained from DANE for
1996 and adjusted for 2001 using population growth rates for Bogot.
Castro and others provide the revenues and costs per passenger for the different types of buses that operated in these trunk corridors before TransMilenio.2 We estimated the number of buses that operated in these trunk
corridors (approximately 6,080 buses) based on the number of routes in
these corridors and the average number of buses per route; this information was provided by Express del Futuro, one of the TransMilenio concessionaires. With regard to scrapping, TransMilenio S.A. reports that for
every TransMilenio bus that entered the system, three traditional system
buses were scrapped. Nearly 470 TransMilenio buses came into operation, so 1,410 traditional buses (23.2 percent of the total) were effectively
scrapped; we assume that the remaining 4,670 buses (76.8 percent) were relocated to other corridors. We calculated the forgone revenues of the traditional buses as the number of passengers multiplied by the revenue per
passenger for buses in the Caracas and Avenida 80 corridors. We estimated
the gains from forgone operational costs of traditional buses previously
operating in TransMilenio corridors as the number of passengers multiplied by the operational cost per passenger for buses in the Caracas and
187
Avenida 80 corridors. Operational revenues (costs) for traditional buses that

relocated to unserved TransMilenio corridors were calculated as the revenues (costs) of traditional buses multiplied by the percentage of buses
that relocated to other corridors (76.8 percent).
We calculated five types of operational costs: (a) feeder line costs: the
number of passengers in the feeder lines multiplied by the feeder rate per
passenger; (b) trunk corridor cost: the number of kilometers multiplied by
the bus fare per kilometer; (c) fare collection costs: a rate per passenger
multiplied by the number of passengers; (d) operational costs for TransMilenio S. A: 3 percent over the technical fare multiplied by the number of
passengers; and (e) fiduciary contract cost: 0.0387 percent of total revenues. For the first four operational costs, we used observed values reported
by TransMilenio S.A. for 2001, 2002, and 2003. For 2004 onwards, the
number of passengers increases by the population growth rate.
Infrastructure costs for the first phase of TransMilenio are based on values reported by Lleras.3 Costs include trunk corridors, bus stations, bus terminals, access to stations, bus depots, operations control center, and buses.
Maintenance costs are the total maintenance costs reported by the Instituto de Desarrollo Urbano (IDU) and TransMilenio.
Time costs break down into several components. The value of time is
from Lleras.4 All times were assumed to remain constant over the period
of analysis. (a) Waiting time: we calculated increases in waiting time as
the difference in waiting time of TransMilenio and traditional system
users and then multiplied this figure by the value of waiting time for TransMilenio users and the number of TransMilenio users. (b) Walking-in and
walking-out time: we calculated increases in walking-in and walking-out
times as the difference between TransMilenio and traditional system users
and then multiplied this figure by the value of walking time for TransMilenio users times the number of TransMilenio users. (c) In-vehicle travel time: we calculated reductions in in-vehicle travel time as the difference
between TransMilenio and traditional system users and then multiplied
this figure by the value of travel time for traditional system users and the
number of traditional system users. (d) Travel time for users of the traditional system: increases in travel time for users of the traditional system
amount to ten percent according to Lleras.5 We then multiplied this increase
3. Lleras (2003).
4. Lleras (2003).
5. Lleras (2003).
by the value of travel time for traditional system users and the number of
traditional system users that transit throughout corridors to which traditional buses were relocated. (e) Negative congestion spillovers in traditional system corridors: five city corridors received the bulk of relocated
buses when TransMilenio began its operation (namely, Avenida Calle 68,
Carrera 7a, Carrera 13, Carrera 15, and Avenida 19). Information regarding the number of daily passengers that transit these selected corridors was
provided by the Districts Transit Secretariat. (f ) TransMilenio times: we
calculated all TransMilenio times for two segments of the population. Segment one includes users who do not use feeder lines, as their trips originate in the vicinity of the Trunk Corridors (55 percent of the population,
according to Lleras).6 Segment two encompasses the remaining 45 percent,
who start their trips in areas served by feeder lines and thus have to engage
in one or more transfers. (g) Traditional system times: we calculated all
traditional system times for two segments of the population. Segment one
includes those who only take one bus to their destination (55 percent of
the population), while segment two (the remaining 45 percent) includes
those who take two or more buses to their destinations.
We estimated the effects on air pollution as follows. Reductions in PM10 emissions associated with TransMilenio operations in trunk corridors
were based on Mndez.7 Emission reductions or increases in TransMilenio
corridors are assumed to affect 20 percent of the population. Finally, we
conducted a difference-in-differences analysis to determine the increases in
PM-10 emissions associated with traffic spillovers to non-TransMilenio
corridors.
To determine reductions in mortality, VSL estimates for Chile were
adjusted for Colombia using the per capita GNP of both countries.8 We
then calculated the risk of death for public transportation users before and
after TransMilenio as the number of deaths divided by the number of public transit users.
To calculate TransMilenios net present value, we used social discount
rates of 7.00 percent, 9.33 percent, and 12.00 percent.
6. Lleras (2003).
7. Mndez (2004).
8. VSL estimates for Chile are from Bowland and Beghin (1998).
Comments
Mauricio Crdenas: This is a very interesting paper on a subject that is
highly relevant and relatively ignored by the profession. The success story
of TransMilenio has received ample press coverage in Colombia and elsewhere, but little academic attention. This public policy initiative was able
to change an entrenched status quo and deliver positive results in a short
period. Since other large cities are following the example of Bogot, it is
valuable to understand the main features of the new scheme, including its
limitations and main challenges.
The authors discuss many dimensions of the problem of organizing
transportation in large cities. The paper provides useful analysis and data
on the optimal amount of intervention in urban transportation and the type
of intervention that works best in developing countries. The paper discusses the cost effectiveness of different kinds of intervention and their
impact on living standards, particularly the health impact of air quality.
The experience of Bogot is characteristic of many large cities in the
developing world. Bogot tried for decades to deal with market failures
associated with transportation systems, experimenting with solutions such
as trolley buses and publicly owned and operated buses. Based on the
information shown in the paper, these policies tended to exacerbate, rather
than resolve, market failures. This is what the authors call the pervasiveness of huge service inefficiencies, and it is reflected in variables such as
the excessive number of buses, the low average speed of circulation, the
high frequency of accidents, and the low air quality. TransMilenio has not
solved all the problems, however, and many inefficiencies are still present
today. For example, the authorities of Bogot have been unable to regulate
the flow of buses originating in conjoining municipalities that circulate in the
city. Also, TransMilenio has increased multileg trips and, hence, the total
fare costs for the average user.
The paper emphasizes the excessive number of private cars in the city.
However, contrary to the conventional wisdom that guides policies in the
sector, Bogot has only 130 cars per 1,000 inhabitants, far fewer than
189
other cities of similar size and per capita income. For example, Curitiba,
Brazil, which was one of the early and influential adopters of a bus rapid
transit system, has 300 cars per 1,000 inhabitants. Other international
comparisons suggest that rather than having too many cars, Bogot has
too few kilometers of paved roads. Since per capita income is rising, vehicle demand will undoubtedly continue to increase rapidly, more so after
trade liberalization has lowered the cost of many durable goods, including automobiles. Consequently, while TransMilenio can solve some traffic problems and reduce commuters travel time, the contribution will be
short-lived unless important investments are made in additional infrastructure or policies are adopted to discourage the use of cars through
market mechanisms.
The institutional aspects of TransMilenio are also of interest. The creation of an entity relatively insulated from political competition was critical for attracting individuals with better technical qualifications, who are
paid competitive salaries. The paper sometimes favors the role of individuals, but the fact that the system has been maintained and improved with
the passage of time speaks well of institutional features. The adoption and
expansion of the system has been the responsibility of three successive
municipal administrations that have shown important ideological differences on other issues. This suggests that the institutions were designed in
a way that prevented policy volatility.
The public policies behind the TransMilenio system have not only been
stable, in the sense of remaining in place beyond the tenure of political
officeholders, but they have also been flexible. The system is not based on
rigid rules, so the authorities have been able to fine-tune important aspects,
most notably in relation to the bidding criteria and the contractual design.
Recent contracts have thus transferred more risks and fewer revenues to
the private sector concessionaires. The various agents involved in the process have exhibited a good amount of coherence and coordination, including the different agencies in charge of building the infrastructure and the
company that operates the system. The policymaking process has generated incentives for increasing efficiency in delivery. Most important,
policies have generated broad (as opposed to concentrated) benefits, resembling true public goods.
This brings me to the political economy of the implementation, which
is one the most interesting and relevant aspects of the paper. I agree with
much of what is said and with the overall spirit of the argument, but four
aspects merit greater emphasis. First, the institutional reform in the city of
191
Bogot (implemented in 1993) was a consequence of the 1991 Constitution. This enhanced the powers of the mayor and reduced the influence of
the city council. Prior to 1993, the city council coadministered the city,
especially in relation to public utilities and transportation policies. Corruption was rampant. Privatizing public utilities and changing transportation policies would have been very difficult without that reform. Second,
privatization and fiscal reform allowed the city to improve its finances,
providing the necessary resources for a large investment project such as
TransMilenio. Third, given the electoral importance of the capital city, the
presidential candidates in the 1998 election supported the idea of committing national funds to the public transportation system for Bogot. The
question was whether to build a metro or a bus-based system like TransMilenio. The economic technocracy of the national government ruled out
the metro and opted for TransMilenio because it was the only fiscally feasible alternative. In fact, the national government actually chose the TransMilenio solution while the city administration was still considering the
metro as an option. This decision carried political costs because it was
seen by the public as an inferior solution to the transportation problems of
the city, but in truth it is a much better solution. TransMilenios area of
influence (500 meters on each side of the busway) covers 85 percent of the
urban area, whereas the metro would have covered only 8 percent of the
city. In addition, the metros infrastructure would have cost twice as much
as TransMilenios and its operation twenty times more. Fourth, multilateral
banks did not have an important role in this process. Conversations on a
credit facility for the construction of infrastructure for a bus rapid transit
system started in 1986. These loans never materialized and had endless
obstacles. Cities in need of overhauling their transportation systems cannot
count exclusively on the advice of multilateral banks.
As the authors rightly point out, congestion, pollution, and traffic accidents fell significantly in TransMilenio corridors, explaining why the
partial cost-benefit analysis is positive. When they look at the general equilibrium effects, they find that the opposite is actually true, mainly because
congestion increased in the unserved corridors, where more buses continue
to compete for passengers. The authors use difference-in-differences calculations to show that emissions have increased in unserved corridors,
more than offsetting the reductions in TransMilenio corridors. These calculations are sensitive to the number of data points (before and after TransMilenio) and the number of monitoring stations, among other factors. The
value of time, the value of life, and the conversion of emissions into monetary
amounts also make the results sensitive to specific assumptions. Therefore, the results should not be taken as unambiguous proof that TransMilenio has had a short-run negative impact. However, I fully share the
authors view in favor of speeding up the process of implementation of
TransMilenio and improving the regulatory framework that applies to traditional public transportation in order to minimize the potential downsides of the scheme. These are important lessons for other cities building
on the experience of Bogot.
Andrs Gmez-Lobo: The paper analyzes the origins, justification, and
results of TransMilenio, an urban transport plan introduced in Bogot,
Colombia, in 2000. Urban transport policy is not a prominent issue in the
discussions of economists who study Latin America, but it nonetheless
merits more attention from specialists. The macroeconomic costs of having
a large portion of the population wasting several productive hours a day
traveling to work or the extra costs for companies that need to distribute
products throughout a city are rarely measured. Urban transport efficiency
may be just as important for economic development as the sophistication
and efficiency of other network industries such as telecommunications.
This paper, insofar as it presents a novel experience in urban transport policy, is relevant for countries throughout the region.
Gwilliam points out that in developing countries urban bus services are
usually characterized by competition in the market.1 Competition among
urban transport providers has been suspect as a welfare maximizing policy, however, since Chadwicks famous article advocating competition for
the market rather than in the market for certain industries, including urban
transport.2 Interestingly, the market failures of competitive urban bus transport markets are still not well understood from a theoretical perspective.
Congestion in the absence of road pricing is an obvious and well-understood
problem, but what is more puzzling is why fares seem to increase rather
than decrease when competition is introduced.3 The ensuing high tariffs
generate excessive entry into the industry, capacity utilization of each bus
diminishes, and the final result is economically inefficient and socially
harmful (for example, reduced safety and increased pollution).
This paper briefly reviews the arguments behind the observed market
failures in urban bus markets. This is not the best part of the paper since
1. Gwilliam (2001).
2. Chadwick (1859).
193
some of the arguments are presented very succinctly and are thus out of
focus. For example, the authors mention the principal-agent relationship
between bus owners and drivers as a market failure. When owners are interested in the number of passengers carried, they will naturally try to align
drivers incentives with their own by offering drivers a share of each passengers fare. There is no market failure here. The problem is that the
owner does not take into account the externalities that these incentives
generate in the form of more aggressive driving and frequency distortions.
The main aim of the paper, however, is not to review the theoretical literature, but rather to describe the transport reform in Bogot and undertake a cost-benefit analysis of the first stages of this reform. There is a
growing consensus among policymakers that state-of-the-art regulatory
policy in the transport sector, at least for middle-income countries, involves what Estache and Gmez-Lobo call the hybrid system.4 In this
scheme, the authorities design the network and impose quality standards,
frequencies, and tariff integration, while the private sector operates the services under a concession system. Revenues are centralized, and operators
are not paid according to the number of passengers carried but receive
either a fixed fee or rate per kilometer traveled. This hybrid scheme avoids
the most important failures associated with state-provided monopoly transport services, as well as the market failures of competition in the streets.
This policy has been adopted in London and Curitiba, Brazil, and it will
also be implemented in Santiago, Chile, in 2005. Bogotas TransMilenio
experience provides another interesting example of the application of this
hybrid model, and its documentation and evaluation provides interesting
lessons for policymakers worldwide.
What can one learn, then, from the experience of Bogot? The current
paper raises two sets of issues that are particularly interesting: first, the
design of the concessions and the way risks are allocated among agents;
and second, the evaluation of the benefits and costs of the reform, which
produces two unexpected results. With regard to the design issue, a key
point is the way contracts were designed for the bus operators in the main
corridors. Since the authorities determine dispatch frequencies, the number
of kilometers served is not under the operators control. Operators are paid
according to the kilometers actually served, however, to avoid competition
on the road. It seems rational that operators should not face a demand risk
that is not under their control, whereas the state is better suited to adminis4. Estache and Gmez-Lobo (2005).
ter this risk. This was achieved in Bogot through a novel variable-length
concession that is reminiscent of Engel, Fisher, and Galetovics leastpresent-value-of-revenue concessions.5 If at the end of the tenth year the
concessionaire has not served the number of kilometers stipulated in the
contract, the concession is automatically extended. This design has several
attractive properties, and a similar system has been adopted for the future
reform in Santiago, Chile.
At the same time, this scheme has the disadvantage that since operators
are shielded from demand risk, they will make no extra effort to take-on
extra passengers or change route designs when demand patterns change.
This will probably be a greater problem in routes outside the main corridors and in the periphery of the city where new neighborhoods may generate new transport demands. The TransMilenio planners wisely relaxed
the payment method for feeder services by combining a payment per kilometer traveled with a payment per passenger carried, thereby providing
incentives to cater to passengers needs and demands in the feeder zones.
On the evaluation of the first stage of TransMilenio, the paper presents
two surprising results that should be given serious attention by policymakers in other countries embarking on similar reforms. First, the reform
was regressive. Those that benefited (that is, TransMilenio users) include
a higher proportion of wealthy individuals than those that lost (namely,
users of the traditional system). Given that infrastructure investments,
which represent a large fraction of the costs of the reform, will be paid for
not by users but by taxpayers in general, the distributional impact of the
first stage of the reform is worrying. Second, the overall cost-benefit analysis shows that the extra pollution and congestion generated on unserved
corridors more than offset the benefits on TransMilenio corridors.
The above two results suggest that it may be socially preferable to undertake a more comprehensive reform rather than implementing the system in
a piecemeal fashion as in Bogot. In Santiago, Chile, the reform of the bus
transport system scheduled to start in August 2005 is all-encompassing,
simultaneously changing the whole transport system of the metropolitan
area. A few years will have to pass before it is evident whether the added
costs and complexities of a citywide reform, as in Santiago, more than offset the benefits of avoiding some of the negative effects of the piecemeal
approach used in Bogot.
5. Engel, Fisher, and Galetovic (2001).
195
References
Bowland, Brad, and John Beghin. 1998. Robust Estimates of Value of a Statistical Life for Developing Economies: An Application to Pollution and Mortality
in Santiago. Staff general research paper 4046. Iowa State University, Department of Economics.
Cavallazi, Marcelo. 1996. Contaminacin atmosfrica en Bogot. Revista Cmara
de Comercio 97. Bogot.
Castro, Ral, and others. 2001. Clculo de la tarifa ptima para el sistema de
transporte masivoTransMilenio. Final report presented to TransMilenio
S.A. Universidad de los Andes, Department of Economics.
Chadwick, Edwin. 1859. Results of Different Principles of Legislation and Administration in Europe; of Competition for the Field, as Compared with Competition within the Field, of Service. Journal of the Statistical Society of
London 22(3).
Chaparro, Irma. 2002. Evaluacin de impacto socioeconmico del transporte
urbano en la ciudad de Bogot: el caso del sistema de transporte masivo, TransMilenio. Serie recursos naturales e infraestructura 48. Santiago: United Nations
Economic Commission for Latin America and the Caribbean.
DANE (Departamento Administrativo Nacional de Estadstica). 2003. Encuesta
de Calidad de Vida. Bogot.
Engel, Eduardo M. R., Ronald D. Fischer, and Alexander Galetovic. 2001. LeastPresent-Value-of-Revenue Auctions and Highway Franchising. Journal of
Political Economy 109(5): 9931020.
Estache, Antonio, and Andrs Gmez-Lobo. 2005. Limits to Competition in Urban
Bus Services in Developing Countries. Transport Reviews 25(2):13958.
Evans, Andrew. 1987. A Theoretical Comparison of Competition with Other
Economic Regimes for Bus Services. Journal of Transport Economics and
Policy 21(1): 736.
Gwilliam, Kenneth M. 2001. Competition in Urban Passenger Transport in the
Developing World. Journal of Transport Economics and Policy 35(1): 99118.
Ibez, Ana Mara, and Eduardo Uribe. 2003. Medio ambiente y desarrollo econmico: priorizacin de la inversin ambiental con criterios econmicos. CEDE
Document 2003-33. Universidad de los Andes. Bogot.
Klein, Daniel B., Adrian Moore, and Binyam Reja. 1997. Curb Rights: A Foundation for Free Enterprise in Urban Transit. Brookings.
Lleras, Germn. 2003. Bus Rapid Transit: Impacts on Travel Behavior in Bogot.
Masters thesis, Massachusetts Institute of Technology, Department of Urban
Studies and Planning.
Lozano, Nancy. 2003. Air Pollution in Bogot, Colombia: A ConcentrationResponse Approach. Masters thesis, University of Maryland at College Park,
Department of Agriculture and Resource Economics.

Mndez, Mildred. 2004. Anlisis de intervencin: efectividad de las polticas
para reduccin de la contaminacin por fuentes mviles en Bogot. Masters
thesis, Universidad de los Andes, Department of Economics. Bogot.
Urrutia, Miguel. 1981. Buses y busetas: una evaluacin del transporte urbano en
Bogot. Bogot: Fedesarrollo.
DAVID S. KAPLAN
GABRIEL MARTNEZ GONZLEZ
RAYMOND ROBERTSON
What Happens to Wages after Displacement?

conomic shocks and policy reforms can induce large changes in
establishment-level employment. Since wage losses from displacement can be large and long-lasting, policymakers often express a
desire to support displaced workers. When resources are limited, policymakers need to target support to the workers who need it most.
But the academic literature offers little guidance on how to do this.
There is little agreement on how wages change after displacement. The
influential works of Jacobson, LaLonde, and Sullivan document large adverse effects of displacement on workers in the United States.1 Subsequent
studies also find that displacement has significant long-term adverse
effects.2 More recent international comparisons, however, find zero or
positive wage changes following displacement. Abbring and others find
no change in wages in the United States, and Bender and others find positive wage changes following displacement in France and Germany.3
Kaplan is with the Instituto Tecnolgico Autnomo de Mxico (ITAM); Martnez

Gonzlez is with the Inter-American Conference on Social Security; and Robertson is with
Macalester College.
This paper is part of a project with the Inter-American Development Bank called Market Institutions, Labor Market Dynamics, Growth and Productivity: An Analysis of Latin
America and the Caribbean. We gratefully acknowledge the assistance of Hector Macas
of the Mexican Social Security Institute (IMSS), financial support from the Asociacin
Mexicana de Cultura, and research assistance from Diago Dieye and Allison Hicks. We also
thank Omar Arias, Emek Basker, David Drukker, Francisco H. G. Ferreira, Tricia Gladden,
John Haltiwanger, Daniel Hamermesh, Adrianna Kugler, Narcio Menezes-Filho, Carmen
Pags, Ken Troske, Andrs Velasco, and Sarah West for extremely helpful comments.
1. Jacobson, LaLonde, and Sullivan (1993a, 1993b).
2. Most studies focus on the United States, including Caballero, Engle, and Haltiwanger
(1997), Davis and Haltiwanger (1999), Stevens (1995, 1997), Revenga, Riboud, and Tan
(1994), and Marcal (2001). Others focus on other developing countries, such as MenezesFilho (2004), Burda and Mertens (2001), Couch (2001), Fallick (1996), Kletzer (1998), and
Ruhm (1991a, 1991b).
3. Abbring and others (2002); Bender and others (2002).
197
The literature also offers conflicting explanations of why these estimates vary. Kuhn suggests that differences in inequality and institutions
in France, Germany, and the United States can explain the different estimates for these countries.4 Alternatively, Howland and Peterson, Carrington, Jacobson, LaLonde, and Sullivan, and Farber suggest that labor
market conditions can affect postdisplacement wages.5 Since a wide variation in displacement costs makes targeting aid difficult, the efficiency gains
from identifying determinants of postdisplacement wages are potentially
significant.
This paper studies the Mexican labor market to contrast various explanations for differences in postdisplacement wage changes. We hope to
identify patterns that may help policymakers target aid to displaced workers. An environment with varying temporal and regional economic conditions and with economic conditions and institutions substantially different
from those in the countries previously studied is ideally suited to identify
such patterns. If institutions vary little across regions, then the institutional
hypothesis would be an unlikely explanation of differences across regions
in postdisplacement wages.
Mexico meets these conditions. Differences between Mexico and other
countries, as well as differences within Mexico over time and space, can
help us identify these patterns in postdisplacement wage changes. First,
wage dispersion is higher in Mexico than in France, Germany, or the United
States.6 If inequality drives differences in postdisplacement wages, then
Mexican workers should have much more negative postdisplacement experiences than observed in these countries.
Second, institutions such as workers separation costs, the legislated
costs of displacement (to the firm), and unions are very different in Mexico
than in other countries. Mexican workers are much less likely to leave firms
voluntarily than workers in other countries, which suggests that they have
4. Kuhn (2002).
5. Howland and Peterson (1988); Carrington (1993); Jacobson, LaLonde, and Sullivan
(1993b, chap. 6); Farber (2003). When examining local labor market conditions, Jacobson,
LaLonde, and Sullivan (1993b) compare two Pennsylvania regions over the same time
period. Carrington (1993) and Howland and Peterson (1988) provide much wider geographic coverage, but these studies are not directly comparable to Jacobson, LaLonde, and
Sullivan because they use cross-section data that are subject to recall error rather than tracking the actual wages of workers over time.
6. The Deininger and Squire data set (available at www.worldbank.org/research/
growth/dddeisqu.htm) shows that Mexicos historically averaged Gini coefficient (52.92) is
higher than that of the United States (35.79), France (37.71), and Germany (32.91).
David S. Kaplan, Gabriel Martnez Gonzlez, and Raymond Robertson
199
higher separation costs. Mexicos mandatory severance pay is higher than

what is legislated in France, Germany, and the United States (the latter two
have no legislated severance pay). Unions also have much less wage-setting
power in Mexico than in the comparison countries, which can lead to negative union-wage differentials.7 If unions explain the difference in results
across countries, then Mexican workers should have much more negative
postdisplacement experiences than is the case elsewhere.
Third, Mexicos geographic regions exhibit little variation in unionization and inequality, but large differences in economic conditions.8 Mexican labor laws regarding severance payments, strikes, prohibitions against
nominal wage reductions, the legal recourses of workers in case of unfair
treatment, and guarantees of profit sharing are determined at the federal
level and therefore do not vary across regions. Moreover, no important
labor market reforms occurred in the period we study.
In short, we use this variation in economic conditions (but not institutions) to compare local labor market conditions and postdisplacement wages
over time and space. Since inequality and institutions vary less within
Mexico than across the countries previously studied in the literature,
heterogeneity in postdisplacement wages within Mexico probably cannot
be explained by inequality and institutions. This points instead toward an
important role for labor market conditions. At the same time, our results
help explain the variation found in the literature. It thus seems very likely
that they can be applied beyond Mexico to target aid to displaced workers
when and where it is most needed.
Our approach differs from previous studies in two key ways. First, we
use a very simple, but formal, theoretical framework that illustrates how
differences in institutions, such as separation costs for both the firm and
the worker, play a key role in the postdisplacement experience. The model
also shows how a displaced worker might earn higher wages after being
displaced and yet not have wanted to leave the original job in search of
higher wages prior to displacement. More important, the model shows
how unemployment rates (through time) and differences in economic activity (through space) can lead to negative, zero, or positive postseparation
wage changes.9 Second, we employ a near-census-sized administrative data
set that allows us to directly compare displacement experiences across time
7. Panagides and Patrinos (1994).
8. On unionization, see Fairris and Levine (2004).
9. The model also shows how the comparison group plays a key role, as Kuhn (2002)
suggests.
and regions. Since we want our results to be as directly comparable with

previous studies as possible, we use the methodological gold standard
established by Jacobson, LaLonde, and Sullivan.10 We use the same standards as in previous studies to identify displaced workers, match workers
to firms, and track workers as they move between firms. We are therefore
able to avoid the so-called recall bias from displaced-worker surveys and
directly compare our results with earlier studies.
In line with the model, we find that different external conditions can
cause wages to go up, go down, or stay constant after displacement. Workers who are displaced during good times experience higher wages than nondisplaced workers (including both nondisplaced workers who remained
employed in firms that underwent large employment contractions and nondisplaced workers who never worked in these firms), while workers displaced in bad times can experience very large losses.11 Furthermore, the
effects of separating in bad times linger: workers who separate when unemployment is high never seem to catch up to workers who separate when
unemployment is low. This effect is most pronounced in relatively less
economically dynamic geographic regions.
Our basic results are robust to the effects of age, attrition, tenure, and
switching sectors. Like previous studies, we find that displaced workers
with longer tenure experience larger losses than workers with shorter tenure
in some periods but not others. Variation in economic conditions through
time may therefore explain why Kreichel and Pfann argue that tenure does
not account for observed wage differences, while other studies, such as
Carrington, support the tenure explanation.12 We also find effects of changing sectors that are similar to previous studies. Our main conclusion is that
changing local labor market conditions produce a wide range of displacement effects and therefore might be the key to understanding when displacement hurts workers.
10. For example, Jacobson, LaLonde, and Sullivan (1993a) analyze the effects of displacement on workers using matched firm-worker data from the United States. Their results
suggest that workers begin to experience falling wages before they are displaced and that earnings recovery may take more than five years. Hamermesh (1989) and Davis and Haltiwanger
(1992) show that adjustment costs at the firm level are generally nonlinear and significantly
affect employment decisions. Other studies examine earnings losses before displacement
(de la Rica, 1995) and how changing labor market conditions affect displacement (Stevens,
2001; Clark, Herzog, and Schlottmann, 1998; Helwig, 2001).
11. We find loss levels that are very similar to those documented by Jacobson, LaLonde,
and Sullivan (1993a, 1993b).
12. Kreichel and Pfann (2003); Carrington (1993).
201
We present our analysis in five sections. We start by presenting our simple theoretical framework. The subsequent section describes the source,
collection, and limitations of our data, discusses the Mexican economic
environment, defines the term displaced worker, and finally describes the
various comparison groups. We then explain our empirical approach and
present our results. A final section concludes.
Theory
This section illustrates how differences in economic conditions can result
in either an increase or a decrease in wages following displacement. The
model modifies McLaughlins theory of quits and layoffs by incorporating
a separation cost that the worker bears in the case of a quit but that the firm
pays to the worker in the case of a layoff.13 We present the model in its
simplest form to illustrate the concepts that guide our empirical work.
Workers receive a wage, w, and produce value to the firm, W. We
assume that the value to the firm is a function denoted W(X, G) in which X
represents worker-specific characteristics and G represents firm-specific
characteristics, including the firms output price and productivity shocks.
Workers have the ability to search on the job. Denote as E(r) the expected
value of an outside wage offer from a firm that values the worker at R. We
assume the outside offer is costlessly verifiable once it is made, and that
the expected value of the offer is a function of external characteristics,
including the number of firms that value the workers skills (following
Stevens), the unemployment rate (which reduces the expected value), and
the economic activity in the region (which increases the expected value).14
We employ McLaughlins important distinction between layoffs and
quits. McLaughlin defines a quit as the result of a firm-refused, workerinitiated attempt to increase wages and a layoff as the result of a firminitiated, worker-refused attempt to lower wages. Firms (workers) have
the option to accept proposals for changes in the wage, and they will do so
as long as the value of the wage is not larger (smaller) than the value of the
output to the firm, or the reservation wage. We modify McLaughlins model
by adding a separation cost. The separation cost may stem from the loss of
shared surplus from firm-specific training, an institutional arrangement that
13. McLaughlin (1991).
14. Stevens (1994).
encourages long-term employment, or other reasons.15 The relevant characteristic is that this cost is paid to the worker in the event of a layoff. This
is particularly relevant for the Mexican case, because Mexico, like some
European and many Latin American countries, follows a more preventative stance.16 Article 50 of the Mexican Federal Labor Law mandates that
workers hired for an indefinite period who are laid off (without cause) are
entitled to twenty days pay for each year of service.17 This mandate may
significantly increase the cost of separation in Mexico.18
The effect of this type of legislation on employment flows is still under
debate.19 One particularly relevant study argues that this kind of legislation
in Brazil creates the incentive for workers to negotiate with firms to make
quits look like layoffs, in order to receive this payment.20 This behavior
creates a procyclical turnover pattern, because workers may be particularly interested in getting their separation payment in good times to start
new businesses or invest in areas with higher returns. Kaplan, Martnez
Gonzlez, and Robertson examine job creation and job destruction in
Mexico.21 They find that the pattern of job destructionespecially the
component that is due to contraction (layoffs)is weakly counter cyclical. Over the 19862001 period, the component of job destruction stemming from firm contraction moved negatively with the net growth rate of
employment. We are therefore confident that the kind of adverse incentives and false layoffs documented in Brazil do not affect our results.22
15. Hashimoto (1979, 1981).

16. Kuhn (2002).
17. This provision applies to contracts of indefinite length. The United States and
Germany have no legislated severance pay, although in the United States the industry standard is one to two weeks per year of service and in Germany severance pay is generally
included in the social plan. In France, workers with more than two years of service receive
0.1 months of salary per year of service (Kuhn, 2002).
18. McLaughlins (1991) analysis of data from the Panel Study of Income Dynamics
(PSID) suggests that separations in the United States are evenly divided between quits and
layoffs. Mexicans, in contrast, are much less likely to leave voluntarily, which may imply a
larger separation cost. See figure 3.
19. See Heckman and Pags (2000) and Robertson and Dutkowsky (2002) for examples
of estimates of labor market adjustment costs in Latin America and a discussion of their link
to labor market legislation.
20. Gonzaga (2003).
21. Kaplan, Martnez Gonzlez, and Robertson (2004).
22. We do not claim that this behavior does not occur in Mexico or that these concerns
are not relevant for Mexico. The aggregate statistics, however, seem to suggest that such
behavior does not have a significant effect on our results.
203
F I G U R E 1 . Workers Decision Rule
E(r)
E(r)
II
III
IV
E(r) C
E(r)
The workers decision rule is most clearly illustrated graphically, as

seen in figure 1. Given an initial wage of w, the worker will voluntarily
leave the firm as long as the expected wage, E(r), minus the separation
cost, C, is greater than the current wage. In figure 1, all workers with
wages in region IV will quit and seek employment elsewhere. Workers
with wages in regions I, II, and III will remain in the firm. The separation
cost will lead some workers to remain in the firm even if they have higher
expected wages elsewhere (as shown in region III).
Now consider the effect of an adverse shock that lowers the workers
value to the firm from w to w. Such an adverse shock could stem from an
adverse productivity or price shock or from some other factor. In this case,
the firm will offer the worker a wage reduction, which the worker can
either accept or reject. If the worker accepts the lower wage, the worker
remains with the firm; if the worker refuses the wage cut, the worker is laid
off. According to Mexican law, the worker must receive a separation payment, which is presumably intended to compensate the worker for the
separation. The effect of this payment means that workers with expected
wages above the new wage, w, will be laid off. Some of those workers will
receive lower wages than in their prior employment (workers with
expected wages in region II), but other workers will now take jobs with
higher wages (workers with expected wages in region III). Workers with
wages in regions II and III will both have higher postdisplacement wages
than workers in the same firm (region I).
The effect of displacement on wages is therefore ambiguous: wages
may either rise or fall after displacement. Wages may go up because the
separation cost keeps workers from voluntarily moving to take advantage
of higher potential wages in other firms. This result illustrates why a
worker who involuntarily separates from a firm may have higher wages
after separating, while lacking the incentive to take a higher-paying job
(before incurring the separation cost) prior to being displaced.
For a given level of separation costs, the value of wages in other jobs
(that is, the expected value of the outside offers) depends on several factors.
If labor markets are not perfect and worker experience is valued outside a
single firm, then increasing the number of firms that value the workers
experience or that would compete for workers will drive up the outside
wage offers into region III.23 Alternatively, a higher unemployment rate
reduces the expected value of wage offers into region II, which implies
that displaced workers would tend to enter firms with lower wages. The
model thus illustrates that the heterogeneity of results found in the literature (negative, zero, and positive) is consistent with a simple theory and
that this heterogeneity can be linked to institutions and labor market conditions in ways that can be empirically compared.
When assessing a theoretical framework such as this one, it is useful to
consider alternative explanations that may generate similar results. Most
theory in this literature focuses on explanations for lower postdisplacement
wages. Suggestions include loss of firm-specific capital and seniority.
These concepts can be easily incorporated into the model above, but they
offer little guidance for the case of higher postdisplacement wages. Higher
postdisplacement wages are somewhat more difficult to reconcile if workers can move to higher paying jobs. This argument, however, assumes that
moving is basically costless. As explained above, a positive moving cost
directly addresses this concern. Furthermore, the model described above
differs from previous approaches that focus on either positive or negative
23. See Stevens (1994).
205
wage changes because it shows how differing labor market conditions

could generate either positive or negative postdisplacement wages.
The Data
Mexican labor laws require all private sector firms to report wage and
employment information on all employees to the Mexican Social Security
Institute (Instituto Mexicano del Seguro Social, or IMSS). In practice, however, firms report information on roughly half the private sector employees.
Firms may choose not to formally register in order to evade taxes and social
security contributions. Academic studies of Mexicos informal sector use
the act of reporting to the IMSS as a criterion for formal sector participation.
The IMSS records thus represent a census of private firms in the formal sector of the Mexican economy.24 Our data come directly from these records.
The IMSS data are collected at the firm level rather than at the establishment level. Each formal sector firm in Mexico has a firm identifier
called its registro patronal. The registro patronal is similar to the employer
identification number (EIN) that is commonly used as a firm identifier in
U.S. data sets. Just as several subsidiary EINs in the United States might be
owned by one parent firm, several registros patronales might be owned by
the same parent company in Mexican data. The registro patronal may
incorporate more than one establishment in a single firm (again like EIN in
U.S. data), but in almost all cases, we identify individual establishments (or
plants in the case of manufacturing industries). We use the registro patronal to link observations over time, to follow workers as they move among
firms, and to track workers wages within their given firm over time.25
As an initial check of data quality, we compare our sample with official
IMSS employment statistics. The motivation behind this comparison is
that the IMSS reports formal employment statistics based on their data,
which are used as an indicator of Mexican employment, but their method
for calculating these statistics is not known to us. A favorable comparison
24. Public sector workers and members of the military have social security accounts
with other agencies.
25. Firms could potentially change their registros patronales from time to time for
administrative reasons, and this would generate false births and deaths stemming from
changes in the registro patronal for continuing firms. In practice, however, very few firms
(fewer than four) closed entirely and opened again in the next quarter with the same employees. These firms were dropped from the sample.
with official statistics would indicate that we have a reliable data set, and
in fact the figures match up quite well in a comparison with data from official IMSS statistics.26
Our data represent all sectors of the Mexican economy.27 To verify coverage, we compared the 1993 average employment in manufacturing in our
data (2,958,715.5) with the 1993 average total employment in the 1993 Mexican Industrial Census (3,246,039.0). Our data thus cover about 91.1 percent of
total manufacturing employment. This leads us to conclude that the distinction between formal and informal labor markets, which is so important in
developing economies, is mainly an issue outside the manufacturing sector.
Since our data are effectively a census of formal sector employment, we
are particularly concerned about the rate of attrition in our sample. Workers may leave our sample for three reasons: they may leave the labor force,
become unemployed, or enter the informal sector. To get a sense of the
rate of attrition in our sample, we focus on workers who worked at least
one quarter in 1993. Of these workers, 78 percent worked at least once in
1994, and 57 percent worked at least once in 2000. About 87 percent of the
workers who appear in our sample for at least one quarter in 1993 appear
in our sample for at least one quarter between 1994 and 2000.
Analytically, treating workers who leave the labor force or become unemployed is straightforward. The complication in our data arises because a
potentially significant number of workers who leave our sample may enter
the informal sector. The informal sector has traditionally been thought of as
an employer of last resort, in which workers earn lower wages and experience
inferior working conditions. Maloney challenges this view for Latin America generally and for Mexico in particular.28 He shows that workers who
become self-employed in the informal sector often earn 25 percent higher
wages, on average, than they did as salaried workers in the formal sector.
Salaried workers in the informal sector, however, always earn less than their
formal sector counterparts. This result suggests that there is no clear presumption of bias, or, more specifically, the direction of a bias from not being
able to account for informal sector employment is unclear. In the empirical
section below, we address this potential bias by comparing results across
samples that include and exclude workers who drop out of our sample.
26. The official data are from www.imss.gob.mx/ventunica/memoria_2001/2/024000.htm.
A table showing this comparison is available on request.
27. Our data cover all economic sectors and are classified using a four-digit industry
code that is similar, but not identical, to the U.S. 1987 SIC code.
28. Maloney (2004).
207
We analyze employee-level records for the period 1993 to 2000, measuring wages on 31 March, 30 June, 30 September, and 31 December of
each year. This yields thirty-two quarters of data. While the period is determined by data availability, it is a particularly interesting time in which to
study displacement because it encompasses several reforms and macroeconomic events, including the implementation of the North American
Free Trade Agreement (NAFTA) in 1994 and the December 1994 collapse
of the peso, which induced a serious recession that lasted until 1996.
The Mexican Economic Environment

In addition to the firm-worker identifier, the data also include details such
as geographic region, sector, and the gender and birth date (month and year)
of each worker. Regional heterogeneity in Mexico emerges as a result of
historical differences in production (centered on Mexico City), concentration of foreign investment in the U.S.-Mexican border region, and the persistent poverty and lack of growth in southern Mexico. In particular,
Mexican regions have had very different experiences with adjustment.29
We therefore focus on four Mexican regions: the border, the north, central
Mexico, and the south.30
The simple model presented earlier suggests that differences in the concentration of economic activity can affect postdisplacement wages. Mexico
exhibits significant regional heterogeneity. Manufacturing is predominantly located in the border, north, and central regions. The highest rates of
employment growth and investment have been concentrated in the border
region, possibly as a result of maquiladora investment.31 The south, in contrast, concentrates on tourism (most notably in the state of Quintana Roo,
home of Cancn) and oil. While aggregate statistics suggest that manufacturing is similar in the border, north, and center regions, the regions are
actually quite different. Figure 2 (panel A) shows the evolution of per
capita gross domestic product (GDP) for each region over time. The central
29. Conroy and West (2000).
30. We define the four regions as encompassing the following Mexican states: the border region: Baja California, Coahuila de Zaragoza, Chihuahua, Nuevo Len, Sonora, and
Tamaulipas; the north: Aguascalientes, Baja California Sur, Durango, Guanajuato, Hidalgo,
Jalisco, Nayarit, Quertaro de Arteaga, San Luis Potos, Sinaloa, Veracruz-llave, and
Zacatecas; the central region: Distrito Federal, Mxico, Morelos, Puebla, and Tlaxcala; and
the south: Campeche, Colima, Chiapas, Guerrero, Michoacn de Ocampo, Oaxaca, Quintana
Roo, Tabasco, and Yucatn.
31. Feenstra and Hanson (1997); Robertson (2000).

F I G U R E 2 . Regional Differences in per Capita GDP and Unemployment Rates
a
A. Per capita GDP
Thousands 1993 pesos

Border
Central
34
North
South
29
24
19
14
1993
1995
1997
Year
1999
2001
B. Unemployment rates
Percent
7.5
Border
Central
North
South
4.5
1.5
1992q3 1993q3 1994q3 1995q3 1996q3 1997q3 1998q3 1999q3 2000q3 2001q3 2002q3
Time Period
a. The four lines represent GDP per capita for the different regions. The GDP of each region is calculated as the sum of the real state GDP of
all states in each region. The population of each region in each year was calculated using a linear growth trend from data between the 1990
and 2000 population censuses. The real state GDP is in thousands of 1993 pesos. The distribution of states into regions is identified in the text.
b. Regional unemployment rates are the simple averages of city-level unemployment rates in each region, as identified in the text.
Tic marks are at third quarter of given year.
209
and border regions are much more economically dynamic than the north.
The central region has the largest amount of measured economic activity,
although it has the fewest states. The border and the north regions have similar levels of total GDP, but GDP per capita is much higher in the border
region, suggesting that economic activity is more concentrated in the border
region than in the north. The border region also had a higher GDP growth
rate than the north for most of the sample period. All regions experienced a
sharp recession in 1995; the years following 1995 are recovery years. We
expect that, to be consistent with the model, postdisplacement wages would
generally be lower in the north than in the central and border regions.
The model also suggests that unemployment rates can affect postdisplacement wages. Panel B of the figure presents regional unemployment
rates calculated as a simple average of the official urban unemployment for
the main cities in each state. The recession is especially evident here.
Unemployment rates are highest in the third quarter of 1995, and they fall
steadily in the third quarters of subsequent years in our sample. Unemployment rates track each other quite closely across regions, but the south tends
to have lower unemployment rates than the rest of the country. The north
has the highest peak unemployment rate. Prior to the collapse, the border
region had higher unemployment rates than other regions, but rates in the
border region fell faster than the rest of the country during the recovery
period. We expect that, to be consistent with the model, postdisplacement
wages would generally be lower for workers who separated during the
height of the recession.
In contrast, institutions (such as unions) and inequality differ very little
across regions in Mexico. Fairris and Levine find unionization rates for
1998 of 0.21 both in states that share a border with the United States and
in states that do not.32 In general, they find little heterogeneity in unionization rates across regions in Mexico. Inequality varies little across regions, as well. As a measure of income inequality, we calculated the Gini
coefficient of the natural logarithm of the real daily wage (the wage measure used in the empirical work below) for each region in our data. In the
first quarter of 1993, the Gini coefficients for the border, north, central,
and south regions were 0.424, 0.422, 0.422, and 0.439, respectively.33
32. Fairris and Levine (2004).
33. For the first quarter of 1991, the Gini coefficients were, in the same order, 0.406,
0.409, 0.419, and 0.405. The regional Gini coefficients generally track each other closely
over time, rising after Mexicos entrance into GATT and then leveling off when NAFTA
went into effect. See Robertson (2004) for further discussion of Mexican wage inequality.
These rates are much more similar to each other than they are to the Gini
coefficients of France, Germany, and the United States.
Description of Comparison Groups and Definition of Displaced Workers

We divide our sample of approximately 80 million observations in several
ways. Given the overwhelming size of the data set, we focus on workers
displaced between the third and fourth quarters of 1995, 1996, and 1997.
These periods match three possibly distinct conditions: high unemployment (1995), sharply falling unemployment (1996), and relatively low and
stabilizing unemployment (1997). These three displacement periods also
maximize the time necessary to identify wage movements before and after
displacement.
The issues related to defining displaced workers are widely discussed
in the literature. Administrative data, such as those used by Jacobson,
LaLonde, and Sullivan, generally do not include direct information on the
cause of separation. The cause of separation is important because workers
who leave voluntarily are more likely to have more positive economic prospects beyond their current firm. Including voluntary separations would
therefore bias the estimated effects of displacement upwards.
To get a sense of the magnitude of voluntary and involuntary displacement, we draw from the National Urban Employment Survey (Encuesta
Nacional de Empleo Urbano, or ENEU). This household survey is like the
U.S. Current Population Survey in that it is used to calculate measures of
unemployment. The survey contains a question that can be used to determine whether a worker separated voluntarily or involuntarily from the firm.
The average responses over time are shown in figure 3. Two facts are
immediately apparent. First, workers in Mexico are more likely to leave
their firm involuntarily than voluntarily, which might suggest a high voluntary separation cost for workers. Second, the rate of involuntary (voluntary) separation is highest (lowest) during the three years on which we
focus (1995, 1996, and 1997). These rates follow the business cycle (they
are consistent with figure 2, panel B) and suggest that our focus years are
the least susceptible to selection bias.34
34. Gonzaga (2003) suggests that workers may negotiate with firms to create the
impression that they were fired when they actually quit, in order to receive severance payments mandated by Brazilian labor law. Our Mexican data do not seem to exhibit the same
kinds of patterns (for example, the cyclicality of separations) that this phenomenon apparently causes in Brazil.
211
F I G U R E 3 . Separation Rates by Motivea

Involuntary Separation Rate
Voluntary Separation Rate
80
72
64
56
48
40
32
24
16
8
87.1
89.1
91.1
93.1
95.1
Time Period
97.1
99.1
01.1
a. Separation rates are calculated based on the Mexican quarterly Encuesta Nacional de Empleo Urbano. The two rates do not add up
to 100 because we excluded separation resulting from injury and other exogeneous factors.
We follow previous studies in our attempt to minimize this bias by

focusing on workers who left firms with significant contractions. We created two samples to identify displaced workers. We first identified firms
that contracted more than 60 percent between the third and fourth quarters
of the reference year, from an initial employment of 50 or more workers.
Our second sample uses a 30 percent contraction threshold. We labeled
these displacing firms.35 The logic behind this decision, which is well established in the literature, is that workers leaving these distressed firms are
most likely to be immune from the selection bias that arises from voluntary separation.
We broke down each of the two samples into four subsamples of displaced workers to examine the possible effects of tenure, sample attrition,
35. Jacobson, LaLonde, and Sullivan (1993a) label all firms that contract more than
30 percent from an early sample average as contracting firms.
age, and sector changes. To focus on the effects of tenure, we restricted our
sample to workers who were in the displacing firm for the entire period up
until the displacement event, left the firm at the time of displacement,
found work in another firm either immediately or after spending some time
out of our sample, and then worked in one firm until the end of the sample period. We refer to this group as C1. Workers who are out of the sample may be either unemployed or working in the informal sector.36 Since
we cannot distinguish between these two conditions, we created another
subsample of displaced workers who found jobs immediately after the displacement event and then remained employed for the rest of the sample.
We label this group C2. Our third group comprises workers who worked
at the same (displacing) firm for less than two years prior to displacement
and then worked at one firm following displacement (C3); these workers
may not be in the sample for the entire period. This short-tenure criterion
contrasts directly with the long-tenure criterion for workers in the first
group. The final group consists of workers who worked at one firm prior
to displacement, are in the sample in all periods, but may have worked at
several firms following displacement (C4).
We contrast the wage patterns of these workers against two comparison groups. Employees in the first group (A) worked in every quarter at
firms that did not experience large contractions in any quarter during our
period of study. Given the large size of the data set, we selected these
workers from a 1 percent sample of all workers in nondisplacing firms.
For 1995, our sample of this group begins with 3.87 million observations,
or about 121,000 per quarter. The second group (B) consists of workers
who worked at a displacing firm in every period of the sample (that is,
workers in displacing firms who remain with the firm after the displacement event).37
About 18 percent of the observations represent multiple firms per worker
in each quarter. This could be due to the fact that workers could hold several formal sector jobs, change jobs frequently within the quarter, or are
not coded correctly. The problem of multiple jobs becomes slightly more
serious when considering displacement because being displaced from your
second job may not have the same implications as being displaced from
36. See appendix A for a discussion of the age differences of those who are displaced
and those who leave the sample.
37. Since we want to compare the wages of workers who remain in displacing plants,
we omit plants that shut down completely.
213
your primary job. Since we do not have data on hours, it is difficult to

determine which is the primary job, which entries are errors, and which
entries represent changing jobs. We therefore drop all workers with multiple observations per quarter from the sample.
We impose additional sample restrictions. We restrict the sample to
workers between fifteen (in the first year) and seventy-one years of age (in
the last year), those with positive earnings, and those who are not missing
sectoral information. We also restrict the sample to those who are not
missing any geographic data.
Summary Statistics
Tables 1 and 2 contain summary statistics for the 1995 ABC1 and ABC2
samples (that is, samples covering groups A, B, and either C1 or C2) by
sector and region. The 1995 sample refers to the sample for the 1995 displacement event and contains observations for every quarter between 1993
and 2000. The summary statistics in tables 1 and 2 summarize data for
1996 from these samples. Our measure of wages is the natural log of the
real daily wage.38 Several interesting results emerge. Workers are generally youngest in the central and border regions, and wages are lowest in the
border. The border and the north have higher employment shares in manufacturing than in services.
Table 3 disaggregates the A, B, and C components of the sample and
compares the sample summary statistics before and after the 1995 displacement event by summarizing the data for 1994 and 1996. The table
shows that the wages of all workers fell between 1994 and 1996. Interestingly, the average wages of workers who remained in displacing firms fell
by more than workers who were displaced.
One potential concern is that the ages of displaced workers and workers who left the sample affect our results. Appendix A formally compares
the ages of workers in each subsample and those who leave our data. This
comparison reveals two main results. First, for workers who remained in
the sample, there is no statistical difference between workers who were
displaced and workers who remained in displacing firms in the 1995 and
1996 sample (workers who remained in displacing firms were, on average,
38. We converted nominal wages to real wages using the national-level consumer price
index available at www.banxico.org, based on the index values corresponding to the month
of observation.

T A B L E 1 . Summary Statistics for the ABC1 Displacement Sample, 1995:3 to 1995:4a
Sector and summary
statistic
Agriculture
Wage
Age
Percent female
No. observations
Mining
Wage
Age
Percent female
No. observations
Manufacturing
Wage
Age
Percent female
No. observations
Transport equipment
Wage
Age
Percent female
No. observations
Construction
Wage
Age
Percent female
No. observations
Utilities
Wage
Age
Percent female
No. observations
Services
Wage
Age
Percent female
No. observations
Total
Wage
Age
Percent female
No. observations
Geographic region
Border
North
Central
South
Total
2.94
37.51
8.79
2,674
2.87
37.52
13.11
3,096
2.80
39.51
10.59
236
2.46
44.2
18.78
905
2.84
38.46
12.1
6,911
3.34
33.62
1.94
1,029
3.36
35.06
1.8
778
3.04
36.92
11.11
72
3.63
35.79
2.96
540
3.40
34.66
2.4
2,419
3.25
30.4
33.04
22,455
3.52
33.18
29.89
45,223
3.39
33.63
29.61
36,741
3.47
32.69
17.67
9,909
3.42
32.74
29.36
114,328
3.50
29.4
28.03
2,472
3.61
32.72
14.91
1,254
3.86
33.64
10.2
2,107
...
...
...
0
3.65
31.65
18.77
5,834
3.21
35.81
3.05
15,042
3.23
33.89
2.78
12,888
3.20
33.82
11.62
5,997
3.46
35.05
9.01
8,033
3.26
34.79
5.33
41,960
4.03
38
14.44
561
4.32
39.51
16
2,025
4.15
39.01
12.6
1,619
3.96
38.36
12.59
588
4.18
39.02
14.25
4,793
3.20
35.38
35.75
19,492
3.35
36.09
42.42
35,706
3.49
34.81
39.76
42,777
3.32
36.04
42.19
17,636
3.37
35.49
40.28
115,611
3.23
33.58
24.91
63,725
3.42
34.57
29.66
100,970
3.45
34.33
32.43
89,549
3.38
35.17
27.06
37,612
3.38
34.36
29.14
291,856
. . . No observations in this category.

a. Data are for 1996.
215
T A B L E 2 . Summary Statistics for the ABC2 Displacement Sample, 1995:3 to 1995:4a

Sector and summary
statistic
Agriculture
Wage
Age
Percent female
No. observations
Mining
Wage
Age
Percent female
No. observations
Manufacturing
Wage
Age
Percent female
No. observations
Transport equipment
Wage
Age
Percent female
No. observations
Construction
Wage
Age
Percent female
No. observations
Utilities
Wage
Age
Percent female
No. observations
Services
Wage
Age
Percent female
No. observations
Total
Wage
Age
Percent female
No. observations
Geographic region
Border
North
Central
South
Total
2.89
40.21
14.49
552
2.71
41.33
13.31
1,052
2.79
44
7.32
164
2.39
45.72
18.33
720
2.66
42.53
14.63
2,488
3.70
37.59
3.41
352
3.49
36.75
1.01
396
3.28
38.71
14.29
28
3.92
34.81
1.47
272
3.67
36.58
2.29
1,048
3.63
33.94
32.79
8,344
3.79
35.76
25.82
12,020
3.70
37.73
24.8
13,500
3.70
36.28
17.31
2,888
3.71
36.11
26.36
36,752
3.77
30.93
35.73
1,500
3.84
33.5
14.68
436
4.18
35.94
7.21
1,276
...
...
...
0
3.94
33.27
21.54
3,212
3.45
39.22
14.56
632
3.38
35.32
7.78
1,028
3.74
34.19
7.69
780
4.02
36.54
47.77
628
3.62
36.09
17.34
3,068
4.16
39.79
17.27
440
4.04
39.83
20.97
744
4.27
40.37
6.95
1,208
4.01
40.4
14.12
340
4.16
40.13
13.32
2,732
3.55
38.33
44.8
7,884
3.38
38
45.72
13,936
3.81
38.15
39.7
15,112
3.48
37.84
48.66
5,540
3.58
38.09
43.79
42,472
3.59
36.01
35.85
19,704
3.55
37.08
33.5
29,612
3.79
37.91
29.94
32,068
3.53
37.88
35.43
10,388
3.64
37.23
32.98
91,772
. . . No observations in this category.

a. Data are for 1996.

T A B L E 3 . Descriptive Statistics by Displacement Status before and after Displacement Eventa
Year before displacement (1994)
Statistic
Wage
Age
Percent female
Agriculture
Mining
Manufacturing
Transport equipment
Construction
Utilities
Services
Border
North
Central
South
No. observations
Year after displacement (1996)
C1
C2
C1
C2
3.820
(0.829)
35.962
(9.902)
0.349
(0.477)
0.035
(0.184)
0.012
(0.107)
0.330
(0.470)
0.049
(0.216)
0.012
(0.109)
0.038
(0.192)
0.524
(0.499)
0.263
(0.441)
0.271
(0.445)
0.369
(0.482)
0.097
(0.295)
62,260
4.189
(0.806)
33.519
(9.311)
0.299
(0.458)
0.002
(0.047)
0.014
(0.117)
0.640
(0.480)
0.007
(0.081)
0.049
(0.216)
0.014
(0.119)
0.274
(0.446)
0.103
(0.304)
0.468
(0.499)
0.255
(0.436)
0.174
(0.379)
23,480
3.541
(0.785)
33.042
(10.641)
0.342
(0.474)
0.018
(0.133)
0.008
(0.088)
0.319
(0.466)
0.010
(0.101)
0.179
(0.383)
0.008
(0.087)
0.459
(0.498)
0.181
(0.385)
0.374
(0.484)
0.305
(0.460)
0.140
(0.347)
251,313
3.907
(0.774)
34.301
(9.455)
0.248
(0.432)
0.019
(0.135)
0.000
(0.000)
0.207
(0.405)
0.000
(0.000)
0.074
(0.261)
0.001
(0.036)
0.699
(0.459)
0.145
(0.352)
0.284
(0.451)
0.523
(0.500)
0.048
(0.213)
6,032
3.569
(0.854)
37.962
(9.902)
0.349
(0.477)
0.035
(0.184)
0.012
(0.107)
0.330
(0.470)
0.049
(0.216)
0.012
(0.109)
0.038
(0.192)
0.524
(0.499)
0.263
(0.441)
0.271
(0.445)
0.369
(0.482)
0.097
(0.295)
62,260
3.854
(0.848)
35.519
(9.311)
0.299
(0.458)
0.002
(0.047)
0.014
(0.117)
0.640
(0.480)
0.007
(0.081)
0.049
(0.216)
0.014
(0.119)
0.274
(0.446)
0.103
(0.304)
0.468
(0.499)
0.255
(0.436)
0.174
(0.379)
23,480
3.278
(0.737)
33.134
(10.474)
0.273
(0.446)
0.023
(0.149)
0.007
(0.081)
0.382
(0.486)
0.013
(0.112)
0.194
(0.396)
0.010
(0.100)
0.372
(0.483)
0.218
(0.413)
0.355
(0.478)
0.294
(0.456)
0.133
(0.340)
206,116
3.607
(0.804)
36.301
(9.455)
0.248
(0.432)
0.042
(0.202)
0.000
(0.000)
0.192
(0.394)
0.000
(0.000)
0.193
(0.395)
0.002
(0.045)
0.570
(0.495)
0.145
(0.352)
0.288
(0.453)
0.519
(0.500)
0.048
(0.213)
6,032
a. Displacement occurred between 1995:3 and 1995:4. The groups are defined as follows: group A: workers who are not in displacing
firms and remain in the same firm; group B: workers who are in displacing firms but do not separate from displacing firms; group C1:
workers who are in displacing firms, separate from those firms, and are not necessarily employed in every period in the sample; and
group C2: workers who are in displacing firms, separate from those firms, and are employed in every period in the sample. Standard
errors are in parentheses.
1.32 years older than displaced workers in 1997). The point estimates
suggest that workers displaced in 1995 and 1996 were slightly older than
workers who remained in displacing firms. Second, workers who left the
sample were younger than those who remained in the sample. Other studies suggest that workers in Mexico often leave the formal sector to become
entrepreneurs in the informal sector.39 Since risk is often associated with
39. Maloney (1998, 2004); Maloney and Krebs (1999).
217
youth, our results seem to be consistent with the idea that when displaced,
young workers may find the informal sector attractive and therefore remain out of our sample.
Table 3 also shows that sample ABC2 is balanced, but ABC1 is not,
since workers in C1 are allowed to exit and return to the sample after the
displacement event. When we consider only a balanced panel of workers,
the average age of all workers is two years higher in the later period. The
table also shows that displaced workers who found a job right away were
most likely to be male, while the difference in the percent of females in C1
and C2 suggests that workers who were displaced and exited the sample
were more likely to be female.
The table also includes information about the sectoral distribution of
each group in each period (the sum over the sectors in each column
equals one hundred percent). Since workers who did not change firms
did not change sectors, the sectoral distribution of the first two groups
remains constant. One might expect that the sectoral distribution of the
two groups of workers who were in displacing firms would be identical,
but we restrict the sample to workers who were employed in every period.
Therefore, differences in the sectoral distribution between the second two
groups reflect the differences in future employment patterns. In our sample, no workers who were displaced from transportation equipment or
mining remained in those sectors when they were displaced. They could
be excluded from the sample as a result of extended search times or they
could have moved to other sectors, such as construction or agriculture.
The percent of displaced workers in construction and agriculture more
than doubles following displacement (for workers who immediately
found employment).
Regional differences in displacement patterns are also evident in table 3.
The majority of employment is in the central region, but so are most of the
displaced workers who immediately found jobs. The north has the highest
share of workers who either were in displacing firms and did not leave or
were displaced and exited the sample at some point. This may be consistent with the shift in production from the central region to the north, as
described by Hanson.40 On the other hand, the overall regional pattern of
employment in table 3 displays a large degree of stability, suggesting that
few workers who were displaced in a particular region moved to other
regions in the very short run to begin other jobs.
40. Hanson (1998).
Empirical Approach
To maximize comparability with studies in developed countries, we employ
the methodological gold standard established by Jacobson, LaLonde, and
Sullivan.41 We first define displacement indicators as Djit, which equals
one if the worker separated from a displacing firm (and zero otherwise) to
compare the wages of displaced workers with all other workers. After presenting these initial results below, we redefine the displacement indicator
to identify workers in each of three groups ( j = 1, 2, 3). The first variable
takes on the value of one for workers who were not in displacing firms, and
zero otherwise (group A). The second takes on a value of one for workers
in displacing firms who remained with the same firm, and zero otherwise
(group B). The third variable takes on a value of one if the workers left
firms that contracted more than 60 percent in the quarter in which they separated (that is, they are in one of the C samples). We estimate each aggregated sample separately (ABC1, ABC2, and so forth). We begin with the
following specification.
(1) wit = ai + t + x it + j j Dji + j t Dji it jt + it .
The dependent variable is the natural log of the real wage, which is calculated by adjusting the nominal wages variable by the Mexican national
consumer price index using 1994 as the base year. The ai term captures
individual-specific fixed effects that take on a value of one for each individual in the sample. The parameter t represents time-specific effects.
Each estimated equation includes a dummy variable for each quarter-year
(for thirty-one of thirty-two periods, omitting the first quarter in the sample). The vector xit represents other time-varying characteristics of workers, including age. We also include the indicator for the individuals
displacement group status, excluding the workers not in displacing firms
as a control group. We then interact the time effects with the displacement
group indicators to compare wages in each group before and after the displacement event. We estimate this equation separately for each of the four
geographic regions in our sample.42
By fully interacting displacement status with the time effects (dummy
variables for each quarter), we allow the time trend for displaced workers
42. We estimate separate equations for each region because the sample sizes are so
large.
219
to differ from the time trend for nondisplaced workers. These differential
time trends are identified off differences over time in wage changes between
displaced workers and nondisplaced workers. We would expect, for example, that wage changes over time would be fairly similar for displaced
workers and nondisplaced workers before the displacement event, but that
wage changes would begin to differ sharply after the displacement event.
This is, in fact, what we normally observe.43
Results
We begin by estimating equation 1 by ordinary least squares (OLS) for
each region. All but thirteen of the 124 estimated marginal effects of the
displacement x time variables (thirty-one coefficients for each of four
regions) for sample ABC1 are statistically significant at the 5 percent
level. The R2 are all 0.90 or higher. A nearly identical pattern of significance emerges for sample ABC2.44 The standard errors are generally very
small. Since we are particularly concerned about the pattern of the relative
wages of displaced workers wages over time, however, a graphical presentation may more effectively facilitate comparisons across years and
sectors.45 Figures 4 and 5 graph the estimated coefficients for the 1995,
1996, and 1997 displacement samples. The patterns of standard errors and
diagnostic statistics are similar for the other years.46
As in Jacobson, LaLonde, and Sullivan, wages in all periods and all regions fell prior to displacement; in contrast with Jacobson, LaLonde, and
Sullivan, no region exhibits a sharp drop in wages at the time of displacement.47 Figures 4 and 5 do show significant effects of displacement,
but these effects vary by region and time of displacement. Figure 4, for
example, reveals that workers who were displaced in 1995, the trough of
the recession, did worse than other workers. Workers displaced in later
years, however, recovered. The heterogeneity across time is especially
evident in the central and border regions. Workers in the relatively poor
43. We are required to use some normalization for both time trends. For both displaced
workers and nondisplaced workers, we set the coefficient equal to zero for the dummy variable corresponding to nine quarters before the (potential) displacement event.
44. The tables are available on request.
45. Appendix B provides a more extensive evaluation of statistical significance and
standard errors.
46. These results are also available on request.

F I G U R E 4 . Effects of Displacement by Region: Sample ABC1a
1995
1997
1996
Border
North
Central
South
.6
.45
.3
.15
0
-.15
-.3
.6
.45
.3
.15
0
-.15
-.3
-16 -12 -8 -4 0
-16 -12 -8 -4 0
8 12 16 20
Quarters from Displacement
8 12 16 20
a. The sample is defined as workers who were in the displacing firm for the entire period up until the displacement event, left the
firm at the time of displacement, found work in another firm either immediately or after spending some time out of our sample, and
then worked in one firm until the end of the sample period (sample ABC1 in the text). Estimated coefficients of the time x displacement
status effects for displaced workers are from equation 1. Groups A and B were both omitted, so these coefficient estimates represent the
difference between the wages of displaced workers and all other workers in the sample. The reference time period is nine quarters prior
to the displacement event for each sample.
north show few effects of displacement and fewer of recovery. While the
wage trends in the border and central regions become positive at the time
of displacement, wage trends remain flat in the north and (to a lesser
degree) the south.
The second important message emerging from figures 4 and 5 is that
the effects of being displaced in 1995 seem to be permanent, or at least
long-lasting. That is, there is little evidence of recovery over the sample
period. This is especially true in the border and central regions. The results of being displaced at times of peak unemployment are therefore similar to the findings of Jacobson, LaLonde, and Sullivan.48 Being displaced
at different times generates different results that include positive postdisplacement wages.
221
F I G U R E 5 . Effects of Displacement by Region: Sample ABC2a

1995
1997
1996
Border
North
Central
South
.6
.45
.3
.15
0
-.15
-.3
.6
.45
.3
.15
0
-.15
-.3
-16 -12 -8 -4 0
-16 -12 -8 -4 0
8 12 16 20
8 12 16 20
firm at the time of displacement, found work in another firm immediately (that is, they spent no time out of our sample), and then
worked in one firm until the end of the sample period (sample ABC2 in the text). Estimated coefficients of the time x displacement
status effects for displaced workers are from equation 1. Groups A and B were both omitted, so these coefficient estimates represent the
to the displacement event for each sample. The omitted category is workers who were not in displacing firms and remained in the same
firm for the entire sample.
The differences between figures 4 and 5 suggest that workers who may
not be in the sample in all periods may suffer more serious repercussions
from displacement than other workers. In general, workers who were employed immediately did better than workers who were out of the sample
for any length of time. The difference between 1995 and 1996 becomes
more pronounced in the north and somewhat less pronounced in other regions when we focus on those in the sample in all periods. Even workers
who were employed immediately in the north continued to experience
falling wages, whereas workers in the dynamic central and border regions
did much better when employed immediately.
Figures 4 and 5 compare the wages of displaced workers with all other
workers, which allows us to compare our results with other studies in the

F I G U R E 6 . Effects of Displacement by Region Relative to Workers Who Remained
in the Firm
a
A. Sample BC1
1995
1997
1996
Border
North
Central
South
.6
.45
.3
.15
0
-.15
-.3
.6
.45
.3
.15
0
-.15
-.3
-16 -12 -8 -4 0
8 12 16 20
-16 -12 -8 -4 0
8 12 16 20

firm at the time of displacement, found work in another firm either immediately or after spending some time out of our sample, and
then worked in one firm until the end of the sample period (sample ABC1 in the text). Estimated coefficients of the time x displacement
status effects for displaced workers are from equation 1. Group B was omitted, so these coefficient estimates represent the difference
between the wages of displaced workers who remained in the displacing firm. The reference time period is nine quarters prior to the
displacement event for each sample.
(continued )
literature. We can also divide the comparison groups into nonseparating

workers who are in distressed firms and those who are not in distressed
firms. Figure 6 contains the results from the comparison of displaced workers with workers who remain in contracting firms for the entire sample.49
The main result of this comparison is that workers who were displaced
experienced large wage gains relative to workers who remained in distressed firms. In nearly every region and every time period, workers who
separated from distressed firms experienced higher wages relative to
49. The results from comparisons with workers in nondisplacing firms only are very
similar to those described above, so we do not discuss them in detail.
223
F I G U R E 6 . Effects of Displacement by Region Relative to Workers Who Remained

in the Firm (Continued )
b
B. Sample BC2
1995
1997
1996
Border
North
Central
South
.6
.45
.3
.15
0
-.15
-.3
.6
.45
.3
.15
0
-.15
-.3
-16 -12 -8 -4 0
8 12 16 20
-16 -12 -8 -4 0
8 12 16 20

b. The sample is defined as workers who were in the displacing firm for the entire period up until the displacement event, left the
firm at the time of displacement, found work in another firm immediately (that is, they spent no time out of our sample), and then
worked in one firm until the end of the sample period (sample BC2 in the text). Estimated coefficients of the time x displacement status
effects for displaced workers are from equation 1. Group B was omitted, so these coefficient estimates represent the difference between
the wages of displaced workers who remained in the displacing firm. The reference time period is nine quarters prior to the
workers who stayed behind. This result is consistent with the model presented earlier (specifically, regions II and III of figure 1).
The second result that emerges from figure 6 is that many of the patterns described in the analysis of figures 4 and 5 remain: wages fall prior
to displacement and the long-run effects of being displaced in 1995 are
less positive than the effects of being displaced in recovery years. Displacement in the border and central regions is followed by higher wages
than in the north and south. Workers displaced in 1995 in the north and
south took much longer to recover than their counterparts in the border
and central regions. As the economy recovered, so did the prospects of
displaced workers.

F I G U R E 7 . Effects of Displacement by Region: Short-Tenure Workers
A. Sample ABC3a
1995
1997
1996
Border
North
Central
South
.6
.45
.3
.15
0
-.15
-.3
.6
.45
.3
.15
0
-.15
-.3
-16 -12 -8 -4 0
8 12 16 20
-16 -12 -8 -4 0
8 12 16 20

a. The sample is defined as workers who may not have been in the sample for the entire period, who worked at the same
(displacing) firm for less than two years prior to displacement, and then worked at one firm following displacement (sample ABC3 in the
text). This short tenure contrasts directly with the long tenure of workers in the first group. Estimated coefficients of the time x
displacement status effects for displaced workers are from equation 1. Groups A and B were both omitted, so these coefficient estimates
represent the difference between the wages of displaced workers and all other workers in the sample. The reference time period is nine
quarters prior to the displacement event for each sample. The omitted category is workers who were not in displacing firms and
remained in the same firm for the entire sample.
(continued )
One of the findings in the current literature is that tenure increases the
adverse effects of displacement. To investigate the effects of tenure, we
created a sample similar to the first except that we dropped all workers
with more than two years tenure in the displacing firm. We performed the
same empirical exercise using this sample and present the results in figure 7. The results in panels A and B can be directly compared with the
results in figures 4 and 6 (panel A). Panel A of figure 7 suggests that
short-tenure workers did better than workers with longer tenure. This is
consistent with worker training and other hypotheses in the literature.
Short-tenure workers displaced in 1995, however, did worse than workers displaced in 1996 and 1997 in the border and central regions, since the
225
F I G U R E 7 . Effects of Displacement by Region: Short-Tenure Workers (Continued )

b
B. Sample BC3
1995
1997
1996
Border
North
Central
South
.6
.45
.3
.15
0
-.15
-.3
.6
.45
.3
.15
0
-.15
-.3
-16 -12 -8 -4 0
8 12 16 20
-16 -12 -8 -4 0
8 12 16 20

b. The sample is defined as workers who may not have been in the sample for the entire period, who worked at the same
(displacing) firm for less than two years prior to displacement, and then worked at one firm following displacement (sample BC3 in the
text). This short tenure contrasts directly with the long tenure of workers in the first group. Estimated coefficients of the time x
displacement status effects for displaced workers are from equation 1. Group B was omitted, so these coefficient estimates represent the
difference between the wages of displaced workers who remained in the displacing firm. The reference time period is nine quarters
prior to the displacement event for each sample.
latter immediately earned higher wages and the former experienced a downward trend in wages. Time of displacement also induces more heterogeneity in the northern region: short-tenure workers in the north who separated
in 1995 did much worse than long-tenure workers who displaced at the
same time.
Tenure does not seem to matter when comparing displaced and nondisplaced workers from displacing firms, in the sense that the overall results in figures 6 (panel A) and 7 (panel B) are very similar. Workers in all
periods and regions eventually did better than workers who stayed behind.
Workers displaced during recovery periods did the best in all regions, and
workers displaced in the border and central regions did better than workers in the north and south regardless of tenure. Therefore, differences in

F I G U R E 8 . Effects of Displacement by Region: Postdisplacement Movers
a
A. Sample ABC4
1995
1997
1996
Border
North
Central
South
.6
.45
.3
.15
0
-.15
-.3
.6
.45
.3
.15
0
-.15
-.3
-16 -12 -8 -4 0
8 12 16 20
-16 -12 -8 -4 0
8 12 16 20

a. The sample is defined as workers who worked in one firm prior to displacement, were in the sample in all periods, and may have
worked at several firms following displacement (sample ABC4 in the text). Estimated coefficients of the time x displacement status
effects for displaced workers are from equation 1. Groups A and B were both omitted, so these coefficient estimates represent the
to the displacement event for each sample. The omitted category is workers who were not in displacing firms and remained in the same
firm for the entire sample.
(continued )
the effects of displacement do not seem to be driven by tenure so much as

by local labor market conditions.
Figure 8 presents our results for the sample in which workers are allowed
to change firms several times following displacement but remain in the
sample. We focus on this sample because workers who switch more often
may have either lower search costs, which would suggest that their wages
would be higher, or less potential to accumulate firm-specific capital, which
would suggest that their wages would be lower. The results in panels A
and B of the figure suggest that, generally, workers who switch more often
may be no less susceptible to the effects of the time of displacement than
workers who switch less often. In fact, the results are almost identical to
figures 5 and 6 (panel B), respectively, which implies that the results are
227
F I G U R E 8 . Effects of Displacement by Region: Postdisplacement Movers (Continued )

b
B. Sample BC4
1995
1997
1996
Border
North
Central
South
.6
.45
.3
.15
0
-.15
-.3
.6
.45
.3
.15
0
-.15
-.3
-16 -12 -8 -4 0
8 12 16 20
-16 -12 -8 -4 0
8 12 16 20

b. The sample is defined as workers who worked in one firm prior to displacement, were in the sample in all periods, and may have
worked at several firms following displacement (sample BC4 in the text). Estimated coefficients of the time x displacement status
effects for displaced workers are from equation 1. Group B was omitted, so these coefficient estimates represent the difference between
the wages of displaced workers who remained in the displacing firm. The reference time period is nine quarters prior to the
not driven by the restriction that workers stay in the same firm for the
remainder of the sample. Overall, the regional and temporal heterogeneity
seem to matter more than the workers switching cost.
One of our concerns about displacement is that workers may lose specific human capital. This problem may increase with the degree of change
a worker experiences from the original position. To consider some of the
potential effects of the loss of such capital, we differentiated the effects
of displacement for workers who remained in the same two-digit sector
and those who changed sectors. To isolate the comparison, we focused on
sample ABC4, which is the sample in which workers are observed for all
periods but may change firms more than once following displacement. We
then created an indicator variable based on whether the worker changed
sectors at the time of displacement. The results are presented in figure 9.

F I G U R E 9 . Sectoral Effects of Displacement
A. Workers who changed two-digit sectors
1995
1997
1996
.6
.45
.3
.15
0
-.15
-.3
-16
-12
-8
-4
0
4
8
12
16
20
12
16
20
B. Workers who did not change two-digit sectors

1995
1997
1996
.6
.45
.3
.15
0
-.15
-.3
-16
-12
-8
-4
0
4
8
229
Panel A compares workers who change sectors and all other workers, while
panel B compares workers who remain in the same sector and all other
workers. For these regressions, we pooled all of the regions and controlled
for region-specific effects using regional dummy variables.
As in previous studies, our results suggest that Mexican workers experience some loss to capital from changing sectors or, perhaps more precisely,
a gain from remaining in the same sector. Workers who changed sectors at
the time of displacement do not seem that much different from other workers for the majority of the sample. Displaced workers who remained in the
same sector, however, follow a pattern similar to those workers in previous
samples. Wages fell prior to displacement, and rose following displacement if the workers were displaced after 1995.
As an additional robustness check, we considered all of the samples and
results described above using a contraction of 30 percent, rather than 60 percent, as our criterion for identifying displacing firms. If selection bias
severely affects our sample, then the bias would be larger with firms under
the 30 percent contraction criterion, because workers leaving firms that
contract by 30 percent would probably include a higher proportion of voluntary separations. These results are nearly identical numerically and qualitatively to the results presented above.50 We find no evidence of a rising
problem of selection bias when we expand the sample. This may be due to
the fact that we include individual-specific fixed effects in all of the regressions, and these effects may effectively be capturing unobserved characteristics that are correlated with ability and other features that could drive
selection bias. This result, along with the fact that we follow established
approaches designed to minimize selection bias, leads us to believe that
selection bias does not significantly drive our results.
Conclusions
Given limited resources and a desire to support displaced workers, policymakers could increase the efficiency of support programs if they targeted
aid when and where it is needed most. Studies on the effects of displacement on wages have generated a wide range of results, but they do little
theoretically or empirically to formally explain the underlying sources of
this heterogeneity. Previous studies suggest that differences in institutions,
50. The results are available on request.
inequality, or labor market conditions might explain the results, but no study
that we are aware of compares these possibilities using matched workerfirm data over time.
In this paper we examine the costs of displacement to workers using an
administrative data set that allows us to follow workers over thirty-two
quarters and four regions that vary significantly in labor market conditions.
By following an established empirical methodology to estimate postdisplacement wages, we focus on the differences in institutions, inequality, and
labor market conditions in a single study in an attempt to understand the
difference in results.
Several findings emerge. First, our results exhibit the same heterogeneity
found in the current literature. We find a range of postdisplacement experiences from negative (such as those documented by Jacobson, LaLonde, and
Sullivan) to positive (such as those documented in Kuhn).51 Since inequality and institutions (unions) are similar throughout Mexico but the empirical results vary through time and space, we therefore conclude that our
analysis provides little support, if any, for the institutional explanation.
This conclusion is further backed by international comparisons. If national
institutions alone explained the differences in results between Germany
and the United States, we would expect to see little heterogeneity within
Mexico rather than the very wide range of results we find.52 Furthermore,
if France and Germany have positive postdisplacement because wages are
more compressed than in the United States, then we would expect the displacement effects in Mexico to be mainly negative because Mexicos
inequality is greater than that of the United States. Instead, we find much
heterogeneity in the results, with many instances of positive postdisplacement wages.
We conclude that differences in local labor market conditions (over
both space and time) are most consistent with our results. We do find large,
negative, and lasting effects of displacement on wages for workers who
are displaced during times of high unemployment and in less economically active regions.53 Postdisplacement wage changes are typically zero
or positive in good times and in the most economically active regions. This
51. Jacobson, LaLonde, and Sullivan (1993a, 1993b); Kuhn (2002).

52. The difference in the results of Jacobson, LaLonde, and Sullivan (1993b) and
Abbring and others (2002), which both focus on the United States, also weighs against the
institutional explanation.
53. Jacobson, LaLonde, and Sullivan (1993b) find similar results.
231
is similar to recently documented patterns in France, Germany, and the

United States.
Our results are robust to changes in the definition of displaced worker.
For example, we consider the implications of displacement for workers
who may not be employed immediately following displacement, workers
who are employed after possibly being outside the labor force, workers with
different levels of tenure, and workers who change sectors. Our results are
generally consistent with other studies that focus on tenure and sectoral
changes. We also explore different reference groups and find strong and
consistent evidence that displaced workers earn significantly higher wages
than their coworkers that were left behind, which seems consistent with a
very simple theoretical model and Kuhns reminder that comparison group
matters.54
We also find that of our four regions (the border, the north, central
Mexico, and the south), the border region has the displacement wage pattern most like that of the United States. Other studies have shown that
labor markets in this region are the most integrated with the United States,
so this similarity may not be surprising. The main example of this is the
decline in wages prior to displacement documented by Jacobson, LaLonde,
and Sullivan.55 This feature is not present in all regions or at all times of
displacement in Mexico, but it emerges most frequently in Mexicos border
region.
The main policy recommendation that emerges from our results is that
targeting aid to displaced workers during recessions and in less economically active areas has potentially significant efficiency gains. These workers tend to suffer larger and more lasting adverse effects from displacement
than other workers, which suggests that targeted aid may be especially valuable. This recommendation clearly assumes that behavior does not change
with policy. But clearly, potential changes in behavior must be taken into
account when considering changes to policy.
Appendix A: Age, Displacement, and Attrition

We formally compared the ages of displaced workers and workers who
left our sample. We are concerned about any difference in age because it
54. Kuhn (2002).
could affect our results through a sort of selectivity bias. If workers who
left the sample are systematically younger than the workers who stayed,
for example, then the remaining workers wages may appear higher simply because these were older workers with more experience.
The results illustrate several points. First, workers in displacing firms
are generally younger than workers in nondisplacing firms. The age of displaced workers is not statistically different from that of workers who remained in displacing firms for 1995 and 1996, although the point estimates
suggest that displaced workers were slightly older. In 1997, displaced
workers were younger than workers who remained in displacing firms, and
the difference (about 1.32 years) is statistically significant.
Second, we find that workers who left the sample were significantly
younger than workers who remained in the sample. As discussed in the
text, this seems to be consistent with other papers that find relatively high
rates of entrepreneurship in the informal sector. Our results may be biased
downwards (upwards) if these workers earn higher (lower) wages, on
average, than workers in the formal sector. These results are shown in
table A1.
T A B L E A 1 . Age Comparisons across Samplesa

Age
Explanatory variable
Age and attrition
Always in sample
Constant
No. observations
R2
Age relative to displaced workers
All other workers
In displaced firms
Constant
No. observations
R2
1995
1996
1997
2.308
(0.215)**
32.503
(0.032)**
112,032
0.00
5.814
(0.129)**
30.937
(0.031)**
116,437
0.02
6.174
(0.146)**
30.313
(0.029)**
136,321
0.01
2.222
(0.207)**
0.108
(0.230)
34.811
(0.191)**
23,711
0.01
1.034
(0.140)**
0.171
(0.183)
36.751
(0.116)**
26,537
0.00
2.186
(0.152)**
1.324
(0.193)**
36.487
(0.130)**
24,579
0.01
*Statistically significant at the 5 percent level; ** statistically significant at the 1 percent level.
a. Standard errors are in parentheses.
233
We used two main approaches to address this issue. First, we included

age and individual-specific fixed effects in our wage regressions to control
for differences in age. Second, we explored the robustness of our results to
different samples. The basic patterns emerge in all samples, regardless of
how we control for experience or attrition. Changing samples affects the
absolute, but not the relative, magnitude of our results.
Appendix B: The Statistical Significance of the Differences

We undertook a simple analysis to determine whether the differences between regions and periods are statistically significant. Given our large sample sizes, the standard errors are generally quite small, suggesting that the
differences between regions and periods are often statistically significant.
For example, figure B1 graphs the 95 percent confidence intervals for the
estimates of the three periods (1995, 1996, and 1997) for the border region
graph in figure 4. The graph suggests that the differences across periods
are probably not statistically significant prior to displacement, but clear
F I G U R E B 1 . Ninety-Five Percent Confidence Intervals for the Border Region in Figure 4
.15
1997
1996
0
1995
-.15
-.3
-18
-12
-8
-4
4
time
12
16
20

F I G U R E B 2 . Ninety-Five Percent Confidence Intervals for Three Regions in the 1995
Sample in Figure 4
.15
North
0
Central
-.15
Border
-.3
-12
-8
-4
4
time
12
16
20
differences emerge after displacement. With regard to differences across

regions, figure B2 graphs the effects of displacement in 1995 for the border, north, and central regions from figure 4 with the 95 percent confidence
intervals. The differences between the lines again suggest that the differences between regions are statistically significant, especially in the postdisplacement period.
Comments
Narcio Menezes-Filho: This interesting paper on the costs of job displacement in Mexico is commendable for several reasons. First, it deals
with an important issue, namely, the earnings trajectory of individuals who
change their jobs. While there is an established literature on this subject
for developed countries, studies on developing countries are scant, despite
the fact that most of the recent reforms that provoked labor reallocation
occurred in the latter, where the problems of poverty and inequality are
severe. Second, the paper addresses this issue using very good data
essentially a census of private firms in the formal sector of the Mexican
economyover a long period. Finally, the analysis is thorough, as the
authors submit their results to various robustness tests using different subsamples. The main drawback of the paper lies in the interpretation of the
results, as I detail below.
The paper does not fully distinguish among the different explanations
for the costs of displacement. It presents interesting graphical analyses of
the wage changes for different periods and regions of displacement, but it
offers very little formal statistical analysis as to whether these differences
are statistically significant (apart from two figures in appendix B). The
authors could have pooled the data and explicitly tested for differences in
the displacement effect across regions and periods or included indicators
of labor conditions at the time of displacement, such as regional unemployment, directly in the regression. Moreover, institutions and inequality
may differ across regions in Mexico. Their explanatory power should be
tested as well, if the aim is to provide a formal test of the different explanations for the displacement effect. It does not suffice to state that
inequality and institutions vary less within Mexico than across countries,
so they are not likely to be the main explanations for the different costs of
displacement.
The authors could also have spent more time interpreting the results,
since reconciling them with the theory is not straightforward. The model
235
predicts, for example, that postdisplacement wages should be lower in the

north than in the central and border regions, but figures 4 and B2 show
exactly the opposite, as wages fall by more in the border and central areas.
The results change completely in figure 5, where the panel sample is used,
(and again in figure 6, panel A), yet the authors do not present a fully convincing explanation for these changes.
The main problem with the data is that, as in other studies, the authors
cannot distinguish between displacements and voluntary separations. As
figure 3 shows, the rate of involuntary displacements varies over time,
reaching its peak in 1995, when displacement seems to be most damaging
to the worker. Does this rate vary across regions, as well? This selection
problem could explain some of the variation in displacement effects documented in the paper. To minimize this selection problem, the authors focus
on workers from displacing firms, that is, firms that contracted more than
60 percent between the third and fourth quarters of a given year (30 percent in another subsample). According to the authors, these workers are
less likely to have voluntarily separated from their firm than workers in
firms that did not implement such massive layoffs. But why were these
workers displaced instead of the workers who continued working at the
displacing firm? According to the model, the displaced workers were those
with expected wages above the new proposed wage. If this is the case, why
were their wages falling prior to displacement, as the various figures show?
Another question that deserves a more careful explanation is why product and labor market conditions vary so much across regions and over time.
The paper does not investigate the reasons for such differences in any
detail. Do good firms and workers, for example, endogenously locate in the
border regions so as to enjoy its good prospects? Is this choice driven by
unobservables? In other words, an endogeneity problem may underlie the
differences in the displacement effect across regions.
In sum, this paper represents an important step toward better understanding the displacement problem in developing countries. The results as
a whole are very interesting, but they deserved a more careful explanation,
especially in view of the selection problems mentioned above.
Omar Arias: The paper discusses the impact of displacement (resulting
from layoffs or voluntary separation) on future earnings performance using
Mexican data. The topic is certainly of utmost relevance for Latin America and the Caribbean in light of the limited reform of overly protective job
regulations and the need for well-designed support for displaced workers.
237
The paper provides an extensive and concise review of the literature in this
area and presents novel results that highlight the heterogeneous potential
earnings impacts of displacement depending on labor market conditions.
The paper will be useful for both researchers and policymakers to better
understand the role of factors mediating the impact of displacement on
earnings and factor these into policy design.
The paper uses a unique panel data set for a large sample of Mexican
workers registered in social security over a reasonably long period. The
authors discuss the methodological difficulties of isolating the impact of
displacement on future earnings. Three sets of issues merit special attention: the problems caused by omitted variables and self-selection (sample
composition biases); attrition or incomplete employment spells (censoring
biases); and the existence of heterogeneous impacts. The first two refer to
the inability to appropriately control for worker and firm characteristics
that may be correlated with both displacement probabilities and postdisplacement earnings, to the restriction to workers with social security
registration (that is, formal sector), and to the possibility that workers who
drop out of the sample may have different characteristics and earnings performance than those who stay. The paper proposes several ways to address
these issues and discusses the implications for the robustness of the results. The third point relates to the fact that average postdisplacement earnings may vary widely across workers depending on context-specific factors
and workers skills. The paper argues that the empirical results favor an
important role for varying labor market conditions over that of local institutions and inequality. I focus my comments on some questions for future
research with regard to the methodological approach and the robustness of
the empirical evidence to discern competing explanations of impacts.
Although not framed in this way, the paper deals with an impact evaluation problem, in which the treatment effect corresponds to the change in
displaced workers earnings. The counterfactual is given by the change in
earnings that would have occurred had these workers not been displaced, and
it is approximated by the change in earnings of comparable nondisplaced
workers. This raises issues familiar from the impact evaluation literature:
identifying the parameter(s) of interest, whether the control (comparison)
groups are good proxies of the counterfactual, and validity of the identifying assumptions. The recent evaluation literature highlights that alternative treatment (impact) parameters could be of interest, although they are
not always identifiable. For example, one may want to measure the average impact of displacement (the effect on any randomly selected worker),
the average effect on the treated (the impact on formal workers actually
displaced), or a local average effect on the treated (the impact for workers
close to displacement thresholds, such as those fired first in a recession).1
These parameters have different interpretations and, more important, lead
to different implications regarding the impact of displacements. For example, the latter parameter tends to capture impacts on marginal workers
(that is, those displaced at the margin during layoffs). These impacts may
depend on both observed skills (for example, human capital measures like
years of education or tenure) and unobserved skills (such as individual
ability or labor market connections). The parameter thus fails to fully capture the impact of large-scale layoffs such as those that would occur in
major recessions or economic restructuring. The paper analyzes multiple
treatment groups that seem to resemble local average impacts, and it is not
entirely clear that the estimated effects readily generalize to the impacts of
displacements of any size or to well-defined groups of workers. Consequently, the results may have limited application for interventions targeted
to massive numbers of workers.
The question of the validity of comparison groups is fundamental to the
results. Lacking other identifying restrictions, this boils down to whether
earnings trends prior to displacement were the same in the displaced
(treated) and nondisplaced (control) samples. The paper does not explicitly discuss this identifying condition. It does not seem to hold for all of
the displaced samples considered since wages decline prior to displacement in some regions or periods. If these trends were not matched in the
corresponding comparison groups of nondisplaced workers, it would raise
questions of possible biases arising from dissimilar composition of the samples (that is, differences in worker characteristics across groups) or mean
reversion (in which earnings eventually move back towards their mean). It
would be useful for future work to discuss these issues in detail.
The paper maintains that the results support a greater role for labor market conditions in mediating the impacts of displacements vis--vis other
factors such as labor institutions and inequality. While well argued and
suggestive, this claim deserves further exploration in future studies. First,
the reported similarity of inequality levels within Mexico does not conform to results from other studies that find significant differences in inequality levels across Mexican regions.2 The reported Gini coefficients are
1.
2.
For a clear exposition of this, see Moffitt (1999).

See, for example, Andaln-Lpez and Lpez-Calva (2002).
239
obtained from the sample of formal sector workers under study (who are
likely to have equally dispersed earnings across regions), while the relevant statistics should cover the entire local labor markets. Second, varying
regional capacities to enforce labor legislation may lead to de facto regional differences in relevant regulations. Finally, it is ultimately difficult
to separate labor market conditions (outcomes), such as unemployment,
from the characteristics of labor institutions. For example, differences in
the enforcement of regulations that prescribe high severance payments or
nonwage benefits correlate with differences in the rates of unemployment
or informal employment.
Thus the reported variation in displacement effects across regions and
time does not support definitively disregarding the potential role of
inequality and institutions in mediating the impacts of displacement.
Future empirical research should delve further into the questions raised by
the new results of the paper and their implications for informing the design
of policies to better balance protection against job loss and more flexible
labor regulations in the region.
References
Abbring, Jaap H., and others. 2002. Displaced Workers in the United States and
the Netherlands. In Losing Work, Moving on: International Perspectives on
Worker Displacement, edited by Peter J. Kuhn, pp. 10594. Kalamazoo: W. E.
Upjohn Institute for Employment Research.
Andaln-Lpez, Mabel, and Luis Lpez-Calva. 2002. The Many Mexicos:
Income Inequality and Polarization in Urban Mexico during the 1990s. Paper
prepared for the Cornell-LSE-WIDER Conference on Spatial Inequality. Cornell University, London School of Economics, and the World Institute for
Development Economics Research, London, 2830 June.
Bender, Stefan, and others. 2002. Worker Displacement in France and Germany. In Losing Work, Moving on: International Perspectives on Worker Displacement, edited by Peter J. Kuhn, pp. 375470. Kalamazoo: W. E. Upjohn
Institute for Employment Research.
Burda, Michael C., and Antje Mertens. 2001. Estimating Wage Losses of Displaced Workers in Germany. Labour Economics 8(1): 1541.
Caballero, Ricardo J., Eduardo M. R. A. Engle, and John C. Haltiwanger. 1997.
Aggregate Employment Dynamics: Building from Microeconomic Evidence.
American Economic Review 87(1): 11537.
Carrington, William J. 1993. Wage Losses for Displaced Workers: Is It Really
the Firm that Matters? Journal of Human Resources 28(3): 43562.
Clark, Don P., Henry W. Herzog Jr., and Alan M. Schlottmann. 1998. Import Competition, Employment Risk, and the Job-Search Outcomes of Trade-Displaced
Manufacturing Workers. Industrial Relations 37(2): 182206.
Conroy, Michael E., and Sarah E. West. 2000. The Impact of NAFTA and the
WTO on Chiapas and Southern Mexico. In Poverty or Development: Global
Restructuring and Regional Transformations in the U.S. South and the Mexican South, edited by Richard Tardanico and Mark Rosenberg, pp. 4155. New
York: Routledge.
Couch, Kenneth A. 2001. Earnings Losses and Unemployment of Displaced
Workers in Germany. Industrial and Labor Relations Review 54(3): 55972.
Davis, Steven J., and John C. Haltiwanger. 1992. Gross Job Creation, Gross Job
Destruction, and Employment Reallocation. Quarterly Journal of Economics
107(3): 81963.
. 1999. Gross Job Flows. In Handbook of Labor Economics, vol. 3B,
edited by Orley Ashenfelter and David Card. Amsterdam: Elsevier.
De la Rica, Sara. 1995. Evidence of Preseparation Earnings Losses in the Displaced Worker Survey. Journal of Human Resources 30(3): 61021.
Fairris, David, and Edward Levine. 2004. Declining Union Density in Mexico,
19842000. Monthly Labor Review 127(9): 1017.
Fallick, Bruce C. 1996. A Review of the Recent Empirical Literature on Displaced Workers. Industrial and Labor Relations Review 50(1): 516.
241
Farber, Henry S. 2003. Job Loss in the United States, 19812001. Working
paper 471. Princeton University, Industrial Relations Section.
Feenstra, Robert C., and Gordon H. Hanson. 1997. Foreign Direct Investment and
Relative Wages: Evidence from Mexicos Maquiladoras. Journal of International Economics 42(3): 37193.
Gonzaga, Gustavo. 2003. Labor Turnover and Labor Legislation in Brazil.
Economa 4(1): 165222.
Hamermesh, Daniel S. 1989. Labor Demand and the Structure of Adjustment
Costs. American Economic Review 79(4): 67489.
Hanson, Gordon H. 1998. North American Economic Integration and Industry
Location. Oxford Review of Economic Policy 14(2): 3044.
Hashimoto, Masanori. 1979. Bonus Payments, On-the-Job Training, and Lifetime Employment in Japan. Journal of Political Economy 87(5, part 1):
1086104.
. 1981. Firm-Specific Human Capital as a Shared Investment. American
Economic Review 71(3): 47582.
Heckman, James, and Carmen Pags. 2000. The Cost of Job Security Regulation:
Evidence from Latin American Labor Markets. Working paper 7773. Cambridge, Mass.: National Bureau of Economic Research.
Helwig, Ryan T. 2001. Worker Displacement in a Strong Labor Market. Monthly
Labor Review 124(6): 1328.
Howland, Marie, and George E. Peterson. 1988. Labor Market Conditions and
the Reemployment of Displaced Workers. Industrial and Labor Relations
Review 42(1): 10922.
Jacobson, Louis S., Robert J. LaLonde, and Daniel G. Sullivan. 1993a. Earnings
Losses of Displaced Workers. American Economic Review 83(4): 685709.
. 1993b. The Costs of Worker Dislocation. Kalamazoo: W. E. Upjohn Institute for Employment Research.
Kaplan, David, Gabriel Martnez Gonzlez, and Raymond Robertson. 2004.
Worker and Job Flows in Mexico. Macalester College.
Kletzer, Lori G. 1998. Job Displacement. Journal of Economic Perspectives
12(1): 11536.
Kreichel, Ben, and Gerard A. Pfann. 2003. Heterogeneity among Displaced
Workers. Paper prepared for the conference on Analyses of Labor Reallocation. W. E. Upjohn Institute and the University of Michigan Business School,
William Davidson Institute, Augusta, Mich., 2123 August.
Kuhn, Peter J. 2002. Summary and Synthesis. In Losing Work, Moving on:
International Perspectives on Worker Displacement, edited by Peter J. Kuhn,
pp. 1104. Kalamazoo: W. E. Upjohn Institute for Employment Research.
Maloney, William F. 1998. Are Labor Markets in Developing Countries Dualistic? Policy research working paper 1941. Washington: World Bank.
. 2004. Informality Revisited. World Development 32(7): 115978.

Maloney, William F., and Tom Krebs. 1999. Quitting and Labor Turnover: Microeconomic Evidence and Macroeconomic Consequences. Policy research working paper 2068. Washington: World Bank.
Marcal, Leah E. 2001. Does Trade Adjustment Assistance Help Trade-Displaced
Workers? Contemporary Economic Policy 19(1): 5972.
McLaughlin, Kenneth J. 1991. A Theory of Quits and Layoffs with Efficient
Turnover. Journal of Political Economy 99(1): 129.
Menezes-Filho, Narcio. 2004. Costs of Displacement in Brazil. University of
So Paulo.
Moffitt, Robert A. 1999. Models of Treatment Effects When Responses Are
Heterogeneous. Proceedings of the National Academy of Sciences 96(12):
657576.
Panagides, Alexis, and Harry A. Patrinos. 1994. Union-Nonunion Wage Differentials in the Developing World: A Case Study of Mexico. Policy research
working paper 1269. Washington: World Bank.
Revenga, Ana, Michelle Riboud, and Hong Tan. 1994. The Impact of Mexicos
Retraining Program on Employment and Wages. World Bank Economic Review 8(2): 24777.
Robertson, Raymond. 2000. Wage Shocks and North American Labor Market
Integration. American Economic Review 90(4): 74264.
. 2004. Relative Prices and Relative Wages: Evidence from Mexico.
Journal of International Economics 64(2): 387409.
Robertson, Raymond, and Donald H. Dutkowsky. 2002. Labor Adjustment Costs
in a Destination Country: The Case of Mexico. Journal of Development Economics 67(1): 2954.
Ruhm, Christopher J. 1991a. Are Workers Permanently Scarred by Job Displacements? American Economic Review 81(1): 31924.
. 1991b. Displacement Induced Joblessness. Review of Economics and
Statistics 73(3): 51722.
Stevens, Ann Huff. 1995. Long-Term Effects of Job Displacement: Evidence
from the Panel Study of Income Dynamics. Working paper 5343. Cambridge,
Mass.: National Bureau of Economic Research.
. 1997. Persistent Effects of Job Displacement: The Importance of Multiple Job Losses. Journal of Labor Economics 15(1, part 1): 16588.
. 2001. Changes in Earnings Instability and Job Loss. Industrial and
Labor Relations Review 55(1): 6078.
Stevens, Margaret. 1994. A Theoretical Model of On-the-Job Training with
Imperfect Competition. Oxford Economic Papers 46(4): 53762.

Andres Velasco Economia

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Andres Velasco Economia

Uploaded by

Copyright:

Available Formats

Brookings Institution Press

Cover design by Rogue Element

Journal of the Latin American and Caribbean Economic Association

Latin American and Caribbean Economic Association

Lora and Olivera on Electoral

Suominen and Estevadeordal

Majnoni and Powell on Basel II

Echeverry, Ibez, Moya,

Kaplan, Martnez Gonzlez,

Journal of the Latin American and Caribbean Economic Association

LATIN AMERICAN AND CARIBBEAN

Journal of the Latin American and Caribbean Economic Association

EDUARDO LORA and MAURICIO OLIVERA

The Electoral Consequences of the Washington Consensus

and KATI SUOMINEN

Rules of Origin in Preferential Trading Arrangements:

and ANDREW POWELL

Reforming Bank Capital Requirements:

JUAN CARLOS ECHEVERRY, ANA MARA IBEZ,

and LUIS CARLOS HILLN

The Economics of TransMilenio,

Comments by Mauricio Crdenas and Andrs Gmez-Lobo

DAVID S. KAPLAN, GABRIEL MARTNEZ GONZLEZ,

and RAYMOND ROBERTSON

What Happens to Wages after Displacement?

LATIN AMERICAN AND CARIBBEAN ECONOMIC ASSOCIATION

The Latin American and Caribbean Economic Association (LACEA), or

Nora Lustig, Universidad de las Amricas, Puebla

Carmen Reinhart, University of Maryland

Carmen Pags, World Bank

Mauricio Crdenas, Fedesarrollo-Colombia

Nouriel Roubini, New York University

AUTHORS AND DISCUSSANTS

Omar Arias, World Bank

on its tracks. Liberalization agreements, of both the bilateral and regional

The Electoral Consequences of the

Lora and Olivera are with the Inter-American Development Bank.

Eduardo Lora and Mauricio Olivera

Review of the Empirical Literature and Some Testable Hypotheses

Eduardo Lora and Mauricio Olivera

to gather and update information is a central issue in the theoretical and

Source: Authors calculations.

growth is positively correlated with it. Furthermore, inflation is significant

Eduardo Lora and Mauricio Olivera

Eduardo Lora and Mauricio Olivera

Eduardo Lora and Mauricio Olivera

hypothesis can be incorporated in our framework by assuming that is a

Eduardo Lora and Mauricio Olivera

Data and Sources

Largest share in the legislature

Number of parties that held . . .

Number of elections (period)

T A B L E 2 . Structure of the Data Set

Source: Authors calculations.

Eduardo Lora and Mauricio Olivera

Latin American countries over the period 19852002. Party alternation

of Political Institutions and a study by Coppedge, classifies parties on a left

Eduardo Lora and Mauricio Olivera

followingadmittedly somewhat arbitraryclassification (numbers in

Eduardo Lora and Mauricio Olivera

All variables are measured as changes between the previous election

Dependent variable: Change in vote shareb

Source: Authors calculations.

Eduardo Lora and Mauricio Olivera