Trade Liberalisation, Economic Crises and Growth: Epublications@Bond

Bond University
ePublications@bond
Globalisation and Development Centre Bond Business School
11-2010
Trade liberalisation, economic crises and growth

Rodney Falvey
Bond University, rodney_falvey@bond.edu.au
Neil Foster
David Greenaway
Follow this and additional works at: http://epublications.bond.edu.au/gdc

Part of the Business Administration, Management, and Operations Commons, Finance and
Financial Management Commons, and the International Business Commons
Recommended Citation
Falvey, Rodney; Foster, Neil; and Greenaway, David, "Trade liberalisation, economic crises and growth" (2010). Globalisation and
Development Centre. Paper 44.
http://epublications.bond.edu.au/gdc/44
This Working Paper is brought to you by the Bond Business School at ePublications@bond. It has been accepted for inclusion in Globalisation and
Development Centre by an authorized administrator of ePublications@bond. For more information, please contact Bond University's Repository
Coordinator.
GDC Working Papers
Rod Falvey
GDC, Bond University
Neil Foster
Foster
University of Vienna
David Greenaway
Greenaway
University of Nottingham
Trade Liberalisation, Economic Crises and Growth
No. 44, November 2010
1
Trade Liberalisation, Economic Crises and
Growth
Rod Falvey
Bond University
Neil Foster
University of Vienna
David Greenaway
University of Nottingham
Abstract
Many economic reforms are undertaken at a time of economic crisis. But is this a good time
for a country to undertake trade reform? In this paper we investigate whether an economic
crisis at the time of trade liberalisation affects a country’s subsequent growth performance.
We employ threshold regression techniques on five crisis indicators commonly used in the
literature, to identify the relevant “crisis values” and to estimate the differential post-
liberalisation growth effects in the crisis and non-crisis regimes. We find that the magnitude
of the acceleration in post-liberalisation growth depends on the characteristics of the crisis.
Although trade liberalisation in both crisis and non-crisis periods raises subsequent growth,
an internal crisis implies a lower acceleration and an external crisis a higher acceleration
relative to the non-crisis regime.
JEL Classification: F14, O40

Keywords: Trade liberalisation, Growth, Crises
Falvey and Greenaway acknowledge financial support from The Leverhulme Trust under
Programme Grant F114/BF. We thank two referees and an Associate Editor for very helpful
comments and advice on an earlier draft. Comments from participants at the Conference on
Economic, Social and Environmental Consequences of the Liberalization of Trade in North
Africa and the Middle East in Rabat, Morocco, 2007 are also acknowledged.

Corresponding author: Rod Falvey, Bond Business School, Bond University, Gold Coast QLD 4229 Australia.
Ph 61755951258; Fax 61755951160. Email: rfalvey@bond.edu.au

Neil Foster, Vienna Institute for International Economics, Rahlgasse 3, Vienna, A-1060. Email:
foster@wiiw.ac.at

David Greenaway, Vice-Chancellor, University of Nottingham, University Park, Nottingham, NG7 2RD.
Email: david.greenaway@nottingham.ac.uk
1. Introduction
Is an economic crisis a good or a bad time for a country to undertake trade liberalisation?
This is a question to which policymakers need an answer, since an economic crisis is often a
politically convenient time to undertake economic reforms. The policy status quo is clearly
unsustainable. But while immediate policy reforms in some areas are clearly called for, it is
not obvious that the reform package should include significant trade liberalisation, though it
often does. Here we present evidence that an economic crisis at the time of trade
liberalisation does affect a country’s post-liberalisation growth performance. Furthermore, its
effects depend on the characteristics of the crisis.
Trade liberalisations have been widespread in the last three decades, particularly among
developing and transition countries. The reasons for this include the perceived limitations of
import substitution as a development strategy1; the weight of empirical evidence suggesting a
positive relationship between openness and growth2; and, not least, the influence of the
International Financial Institutions (IFIs - the World Bank and IMF) which often required
that trade liberalisation be included as part of a package of reforms when agreeing to loans3.
Despite their early promise, recent experience and evidence suggests not all trade reforms
have been as successful as anticipated (Singh 2010). This is partly attributable to weaknesses
in reform packages themselves, including inappropriate timing and sequencing of reforms,
their lack of credibility to private agents and doubts over commitment shown by some
political actors. In many cases it seems a crisis was necessary to trigger the reforms. Could it
be, therefore, that an economic crisis is an unfortunate time to undertake trade reforms?
In this paper we examine whether the extent and type of economic crisis at the time of
liberalisation affects post-liberalisation growth. We consider five crisis indicators commonly
used in the literature (output falls, inflation increases, exchange rate depreciations, increased
external debt to export ratios and increased current account deficits), which we are also able
1
This view is not uncontroversial. Rodrik (1999) argues that IS policies worked quite well at least until the mid-
1970s and that the poor performance of such countries after 1973 was the result of an inability to respond to
macro-shocks and not because of IS policies. Moreover defenders of IS policies argue that it has often been
misinterpreted and that it is not a rationale for indiscriminate protection. They also cite evidence of successful
selective intervention in some of the so-called liberal trading countries of East Asia (Rodrik, 1995; Baldwin,
2003; Cline, 2004).
2
Again, this statement is not uncontroversial. Rodriguez and Rodrik (2000) criticise much of the existing
literature on growth and openness. While not arguing that there is a systematic relationship between inward
orientation and growth, they argue that the evidence linking outward orientation and growth is overstated.
3
For the period 1980-89, 79% of all loans had conditions in the trade policy area, in excess of those which
attached to any other policy (Greenaway, 1998).
2
to combine into two factors roughly representing the internal and external dimensions of a
crisis. We employ threshold regression techniques on our crisis indicators to identify the
relevant “crisis values” and the differential post-liberalisation growth effects in the crisis and
non-crisis regimes. Our results indicate that an economic crisis at the time of liberalisation
does affect post-liberalisation growth, with the direction of the effect depending on the nature
of the crisis. An internal crisis implies lower growth and an external crisis higher growth
relative to the non-crisis regime.
The remainder of the paper is organised as follows. Section 2 reviews the theoretical and
empirical literature linking crises, liberalisation and growth. Section 3 discusses data,
methodology and long-run results, while Section 4 adds in short-run effects. Section 5
extends our analysis and examines robustness of our main results, and Section 6 concludes.
2. Background: Trade Liberalisation and Growth

The potential growth effects of trade liberalisation are well known4. While the immediate
impact is likely to be negative as resources become redundant in areas of comparative
disadvantage, their eventual reallocation into areas of comparative advantage will see a rise in
the growth rate in the medium run as income moves to a higher steady state level 5. Longer
run gains in the growth rate must come through improvements in factor productivity and
these can emerge through a variety of channels. Increased imports of capital and intermediate
goods not available domestically may directly raise the productivity of manufacturing
production (Lee, 1995) and increased trade (exports and imports) with advanced economies
could indirectly raise growth by facilitating knowledge and technology spillovers. Learning
by doing may be more rapid in export industries6. A liberal trading regime may attract export-
platform FDI. The magnitude of these long-run growth effects will vary across countries,
depending on their sectors of comparative advantage in particular.
While the empirical literature on openness and growth is voluminous (Dollar, 1992; Sachs
and Warner, 1995; and Frankel and Romer, 1999 are prominent examples) that on trade
4
Dornbusch (1992) and Krueger (1998) provide useful surveys of the gains from trade liberalisation.
5
The static gains from trade liberalisation need not be limited to such resource allocation gains. Further gains
can arise from reductions in rent seeking, corruption and smuggling. Other gains include those resulting from
economies of scale in exporting industries, reduced market power in protected markets, and increased variety
and quality of imported goods available to domestic producers and consumers.
6
Indirect evidence suggestive of the importance of learning by doing in export industries is provided by the
recent literature on exporting and productivity (for a review see Greenaway and Kneller 2007).
3
liberalisation and growth is more limited. Some comparative cross-country studies have been
undertaken, including Little, Scitovsky and Scott (1970), Krueger (1978), Bhagwati (1978)
and Papageorgiou, Michaely and Choksi (1991) (PMC). The latter is the most sanguine,
concluding that trade liberalisation results in a more rapid growth of exports and GDP,
without significant transitional costs of unemployment7. Other studies find liberalisation
leads to growth in exports and improvement in the current account (although some of this is
because of import compression), and that while some countries have increased investment
following liberalisation, others suffer an investment slump. So the impact on growth may be
positive or negative, although there seem to be more cases of a positive than negative growth
effect (Greenaway, 1998).
Econometric studies are relatively more plentiful. Greenaway, Leybourne and Sapsford
(1997) use a smooth transition model to test for a transition in the level and trend of real GDP
per capita for 13 countries in the PMC sample and relate these to liberalisation. While all
displayed a transition in level or trend, in the majority it was negative8, and where it was
positive it generally could not be related to liberalisation episodes9. Greenaway, Morgan and
Wright (1998, 2002) (GMW) use a dynamic panel model to examine both the short- and
long-run impact of liberalisation on growth in a large sample of countries. Results using three
measures of liberalisation suggest a J-curve effect, growth at first falls but then increases after
liberalisation. Wacziarg and Welch (2003) update the Sachs and Warner (1995) indicator of
liberalisation, and regress per capita output growth on country (and time) fixed effects and
their indicator of liberalisation. They find the difference in growth between a liberalised and
non-liberalised country is 1.53 percentage points. Salinas and Aksoy (2006) use an alternative
indicator10 and find trade liberalisation increases growth by between 1 and 4 percent.
Although the later empirical evidence provides broad support for the hypothesis that trade
liberalisation improves economic growth, this support is far from universal and it is clear
some liberalisations have been more successful than others. Given the variety of
circumstances under which trade liberalisations have occurred this is hardly surprising.
Where liberalisations have been the outcome of a specific policy review process, have had
7
Critiques of these results are provided by Collier (1993) and Greenaway (1993).
8
Maurer (1998) finds in the majority of cases neither a positive nor a negative impact on growth of the
liberalisation episodes defined by PMC.
9
Greenaway and Sapsford (1994) model liberalisation as a discrete break rather than a smooth transition, and
again find little evidence of liberalisation increasing a country’s growth rate.
10
Though their liberalisation dates are generally consistent with those of Wacziarg and Welch.
4
broad political support and been undertaken in a stable economic and political environment
they are likely to be sustained and successful. But in many cases liberalisations have been
undertaken as part of a “package” of reforms emerging from an economic or political crisis.
Crises appear to facilitate some reforms11. Drazen and Grilli (1993) model a “war-of-
attrition” in an economy that has settled into a Pareto-inferior equilibrium, and where reforms
are resisted because of uncertainty over who is more willing to bear the costs. An economic
crisis may then help to move the economy to a welfare-superior path, as reforms that would
be resisted under normal circumstances, may be accepted if the losses from a continuing
crisis are large. Such an approach seems particularly promising for explaining
macroeconomic stabilisations, where the distribution costs are low and there is likely to be
consensus on the policies required, and this is confirmed by the empirical evidence (see for
example, Bruno and Easterly, 1996; Bruno, 1996; Drazen and Easterly, 2001; and Alesina et
al, 2006). But with structural reforms (e.g. trade and labour market reforms) the distributional
costs are higher and there is a lower likelihood of consensus on the appropriate policies
(Rodrik, 1996). The empirical evidence on whether crises facilitate structural reforms is
correspondingly less decisive. Lora (1998) finds empirical support (in Latin America) for the
hypothesis that a crisis involving a decline in real income is likely to facilitate trade reforms,
although he notes that the effect is quantitatively small. Tornell (1998) presents empirical
evidence on the relationships among drastic political change, a major economic crisis
(measured by inflation and a decline in output) and trade liberalisation. Using Probit models
explaining the start of liberalisation he finds that the unconditional probability of reform is
2.7%, increasing to 27% with an economic crisis and 60% with both an economic and
political crisis. Campos et al (2006), however, find that, unlike political crises, economic
crises have no significant impact on the implementation of reforms.
Even if an economic crisis facilitates structural reforms in general, it need not be a good time
to undertake trade liberalisation; for two reasons. First, trade reform works by correcting
distortions in relative prices, but high and variable inflation can confound price signals,
making it difficult to disentangle relative price changes from changes in the general price
level, thereby blunting incentives to reallocate resources (Rodrik, 1992). Moreover, the
11
This is not surprising according to Rodrik (1996), who states that “There is a strong element of tautology in
the association of reform with crisis. Reform naturally becomes an issue only when current policies are not
perceived to be working. A crisis is just an extreme instance of policy failure. That policy reform should follow
crisis, then, is no more surprising than smoke following fire.” (pp. 26-27).
5
slowdown in domestic activity associated with crises can exacerbate transitional
unemployment as resources shift between sectors, increasing opposition to reforms and
increasing the likelihood they will be reversed. Second, if trade liberalisation is to be
successful (and sustained), the private sector must respond to changed incentives, and if
private agents are sceptical of policymakers’ commitment, they will be slow to incur the
(sunk) costs associated with shifting resources between import competing and export sectors.
Short-run adjustment will be prolonged and efficiency gains delayed. In such a situation there
will be few gainers from liberalisation, while some will lose due to increased foreign
competition. Such an outcome is likely to make it politically difficult to sustain reforms as
well as limiting their impact. Thus scepticism on behalf of the private sector may be more
likely for liberalisations undertaken in times of crisis. This may be compounded if
liberalisation is undertaken as part of a package of reforms that countries were obliged to
negotiate to secure financial support from the IFIs (Rodrik, 1989b). In the absence of a crisis
and conditions requiring trade reform laid down by IFIs, it would be clear to the private
sector that a government that undertook liberalisation would be committed to the reforms. In
the presence of intervention from IFIs however, there is an incentive for uncommitted
governments to undertake reform temporarily to receive funds. In this situation it is difficult
for the private sector to distinguish between a government committed to reform and one that
is undertaking reform for financial gain12.
These considerations combine to suggest that a trade liberalisation undertaken at a time of

crisis may reflect weaker commitment from policymakers and higher scepticism from private
agents. If so it will be less likely to be sustained and successful, and less likely to have a
significant growth promoting impact. The nature of the crisis itself may also be important. A
severe “internal” crisis (falling output and high and variable inflation) will distort price
signals and delay growth enhancing benefits. A severe “external” crisis (currency
depreciation, growing current account deficit and high debt to export ratio) will also constrain
growth and is more likely to lead a “not otherwise reform minded government” to undertake
reforms to obtain support from IFIs. Of course the trade liberalisation itself will eventually
free up these constraints, particularly if the external crisis occurs in a highly inward-looking
12
Support from IFIs cannot be taken as a signal of a lack of local commitment, however, since such support can
act as an external anchor strengthening the credibility of reforms and providing short-term finance that can
alleviate the short-term costs for governments committed to reform (Morrissey, 1995).
6
policy regime. In practice an economic crisis will exhibit both internal and external
symptoms, which is why we include indicators of both in our analysis.
One limitation of our analysis should be made clear at the outset. Ideally we would extend it
by employing a measure of the strength or depth of liberalisation to investigate empirically
how it is related to the presence or absence of an economic crisis at the time it was
undertaken. We could then shed further light on why liberalisations might have generated
different growth outcomes. Unfortunately the multidimensional nature of trade policy, in
terms of the numbers of both products traded and types of policies, precludes us from doing
this in any systematic fashion. Aggregate measures of even direct trade policy instruments
(tariffs and quotas) are difficult to formulate (Anderson and Neary, 2005), and even if we
could, the time series data on trade policy measures required for quantification is not
available covering liberalisations for an adequate sample of countries. Given these
constraints, our analysis is largely restricted to whether crises at the time of liberalisation
have implications for subsequent growth, and leaves the investigation of why they might have
such implications to future research once data becomes available.
3. Data, Methodology and Initial Results

The starting point for our empirical analysis is an equation similar to the initial regression
estimated by GMW (2002)13:
𝐼𝑁𝑉
∆ln𝑦𝑖,𝑡 = 𝛽1 l𝑛𝑦𝑖,60 + 𝛽2 𝑆𝑌𝑅𝑖,60 + 𝛽3 ∆ln𝑃𝑂𝑃𝑖,𝑡 + 𝛽4 Δln𝑇𝑇𝐼𝑖,𝑡 + 𝛽5 + 𝛿𝐿𝐼𝐵𝑖,𝑡 +
𝐺𝐷𝑃 𝑖,𝑡
𝜂𝑡 + 𝜀𝑖,𝑡 (1)
where 𝑖 denotes country and 𝑡 time; and
𝑦𝑖,𝑡 = GDP per capita
𝑦𝑖,60 = GDP per capita in 1960 (the base period)
𝑆𝑌𝑅𝑖,60 = Average years of secondary schooling in the population over 15 in 1960
𝑇𝑇𝐼 = Terms of trade index
𝑃𝑂𝑃 = Population
𝐼𝑁𝑉/𝐺𝐷𝑃 = Ratio of gross domestic investment to GDP
13
There are two major differences between (1) and the static model estimated by GMW. First, we replace the
initial level of secondary school enrolment with the average years of secondary schooling in the population over
15 in 1960 as a measure of initial human capital, since as Pritchett (2001) argues, enrolment ratios are not an
ideal measure of the stock of human capital. Second we include a full set of time dummies, 𝜂𝑡 , to account for
time-specific heterogeneity in growth rates across countries.
7
𝐿𝐼𝐵 = Dummy variable taking the value one for all years after and including the
year of liberalisation and zero otherwise.
We estimate this using annual data for a panel of (up to) 75 countries within the period 1960-
2003. Much of the data is from the World Bank’s World Development Indicators (2005)
database; including GDP, population, investment and the terms of trade. Data on schooling is
from Barro and Lee (2001). The indicator of trade liberalisation is from Wacziarg and Welch
(2003)14, and is a broad measure. That the average growth experience is likely to be higher
post-liberalisation is indicated by Figure 1 which plots the average annual growth rates of all
countries in our sample for the 10 years pre- and post-liberalisation. Given the different
timings of liberalisations, it is clear that these averages are based on a different number of
observations for each period. Also reported (the dotted lines) are the average growth rates
over the 10 years pre- and post liberalisation.15 Figure 1 shows that average growth following
liberalisation (around 2 percent) tends to be higher than that prior to liberalisation (around -
0.7 percent), consistent with the econometric results we report below.
The results of estimating equation (1) are reported in Table 1.16 The first regression is our
base specification, excluding the liberalisation dummy. The outcomes for the control
variables are largely in line with existing results, particularly those reported by GMW (2002).
We find negative and significant coefficients on initial GDP per capita and population
growth, and positive and significant coefficients on initial schooling, investment and the
terms of trade index. In regression 2 we add the liberalisation dummy. This leaves the control
variables largely unchanged, with the liberalisation dummy itself positive and significant.
The estimated coefficient indicates that liberalisation has a favourable impact on growth of
14
This is an update of the Sachs and Warner (1995). This is a dummy variable of openness, with a country being
classified as closed if it displayed at least one of five criteria, namely; (i) Average tariff rates of 40% or more,
(ii) Non-tariff barriers covering 40% or more of trade, (iii) A Black Market exchange rate (BMP) that is
depreciated by 20% or more relative to the official exchange rate, on average, (iv) A state monopoly on major
exports, (v) A socialist economic system. The date of liberalisation is then defined as the year in which none of
these criteria are met. The measure was heavily criticised by Rodriguez and Rodrik (2000), who argued that the
information on the BMP and the state monopoly on major exports played the major role in its classification of.
They went on to argue that a high BMP is likely to reflect factors other than trade policy, including
macroeconomic mismanagement, weak enforcement of the rule of law and high levels of corruption, while the
information on the state monopoly of exports works like an Africa dummy. In updating this indicator, Wacziarg
and Welch (2003) note that the liberalisation date is less subject to criticism, and are careful to cross-check their
dates against case studies of reforms in developing countries.
15
The figures for the five- and twenty-year averages are similar.
16
The regressions estimated in Columns 5 and 6 employ data for 58 countries for the 1960s and 1970s, 70
countries for the 1980s and all 75 countries for the 1990s and 2000s.
8
around 2 percent in the years following it, in line with estimates reported by GMW (1998 and
2002), Wacziarg and Welch (2003) and Salinas and Aksoy (2006).
Figure 1: Average Growth Before and After Liberalisation
Table 1: Initial Results

∆ln𝑦 1 2 3 4 5 6
-0.005 -0.005
l𝑛𝑦60
(-3.18)*** (-3.12)***
0.26 0.23 0.27 0.26 0.21 0.20
𝐼𝑁𝑉/𝐺𝐷𝑃
(10.42)*** (9.20)*** (7.73)*** (7.45)*** (7.35)*** (6.95)***
-0.54 -0.55 -0.67 -0.74 -0.36 -0.28
∆ln𝑃𝑂𝑃
(-3.39)*** (-3.43)*** (-1.87)* (-2.03)** (-1.36) (-1.06)
0.005 0.008
𝑆𝑌𝑅60
(2.07)** (0.31)
0.02 0.02 0.02 0.019
Δln𝑇𝑇𝐼
(1.92)* (1.83)* (2.31)** (2.22)**
0.02 0.018 0.028
𝐿𝐼𝐵
(5.47)*** (4.78)*** (7.24)***
Time
Yes Yes Yes Yes Yes Yes
Dummies
Country
No No Yes Yes Yes Yes
dummies
Observations 952 952 1327 1327 2619 2619

F-Statistic 26.23*** 310.95*** 15.25*** 206.43*** 13.02*** 13.57***
R2 0.34 0.36 0.34 0.36 0.27 0.29
Notes: t-statistics in brackets. All models estimated using White Heteroscedasticity-Consistent standard errors.
*, **, *** indicate significance at the 10, 5 and 1 percent level respectively.
One limitation of estimating (1) is that data constraints mean that only 39 out of the 75
countries (and only 952 observations out of a potential 2767) are included. Three variables
9
are responsible for this: initial output per capita, initial schooling and the terms of trade
index. We therefore drop initial output per capita and schooling in regressions 3 and 4,
replacing them with a full set of country dummies. Including country fixed effects allows us
to drop time invariant variables, with the country dummies capturing the impact of country-
specific factors on growth, including initial levels of output per capita and schooling. The
estimated coefficients on the remaining control variables are largely unaffected, as is the
liberalisation dummy which remains highly significant. Finally, in regressions 5 and 6 we
drop the terms of trade variable, which increases our sample to 2619, and allows inclusion of
all 75 countries17. This lowers the coefficient on population, which also becomes
insignificant, but has little impact on investment. That on liberalisation increases in size but is
still within the range of estimates in the literature, and is again highly significant.
The regressions in Table 1 give an estimate of the average impact of trade liberalisation on
growth across all liberalising countries. Using the final regression (6) as a base, we now
explore whether these growth effects differ depending on: (a) whether the country faced an
economic crisis at the time of liberalisation; and (b) if it did, the nature of the crisis. Several
variables have become standard indicators of aspects of an economic crisis (Alesina et al,
2006; Campos et al, 2006): the proportional decline in per capita GDP (𝑂𝑈𝑇), the inflation
rate18 (𝐼𝑁𝐹), the nominal exchange rate (𝑋𝑅), the ratio of debt to exports (𝐷𝐸𝐵𝑇), and the
current account deficit (𝐶𝐴𝐷). Data are again taken from the World Development Indicators
(2005) database.19 Each represents a specific aspect of a crisis. Individually they are
informative, but will be even more so if they can be combined in some way. In particular it is
of interest whether the internal or external dimensions of an economic crisis at the time of
liberalisation have different implications for a country’s subsequent growth performance.
Factor analysis is a method of condensing a number of random variables into a smaller

number of uncorrelated variables20. The first factor accounts for as much of the variability in
the data as possible, and each succeeding factor accounts for as much of the remaining
17
The results for the restricted sample where we retain the terms of trade variable are very similar to those in the
remainder of this section and are provided in Section 5 below for comparison.
18
The results reported are based on the GDP deflator rather than the CPI index, since the GDP deflator is
available for more countries and more years. Our results are robust to the use of either the CPI or GDP deflator.
19
The ratio of debt to exports is calculated as the total value of debt in current US dollars divided by the value
of exports in current US dollars.
20
For an introduction to principal components and factor analysis see Kline (1994). Campos et. al. (2006)
employ principal components to construct an index of social and political instability. Dreher et.al. (2008) use
principal components to construct indices of globalisation.
10
variability as possible. We implement the factor analysis procedure using the original data
on our five crisis variables and employing the maximum likelihood factor method. The
results yield two retained factors, with the rotated factor loadings as reported in Table 2.
While there cannot be said to be a definitive separation of variables, the first factor (which
explains over 80% of the variance in the variables) has its largest positive weightings on
OUT, INF and, to a lesser extent DEBT, while the second (which explains the remainder of
the variance) has its largest positive weightings on CAD and XR. In what follows we
therefore label the first factor INT and interpret it as an indicator of the internal dimension of
the crisis, and the second EXT and interpret it as an indicator of its external dimension..
Table 2: Rotated Factor Loadings

Factor 1 [𝐼𝑁𝑇] Factor 2 [𝐸𝑋𝑇]
𝑂𝑈𝑇 0.497 -0.112
𝐼𝑁𝐹 0.466 0.077
𝑋𝑅 0.023 0.150
𝐷𝐸𝐵𝑇 0.216 -0.033
𝐶𝐴𝐷 -0.094 0.282
Combined the two factors account for all of the variance in the crisis
variables, with 𝐼𝑁𝑇 accounting for 82 percent of the variance of the
crisis variables, and 𝐸𝑋𝑇 18 percent.
For each crisis indicator we calculate a standardised score as:

𝑋𝑗𝑖𝑡 − 𝑋𝑗𝑖𝑡
𝐶𝑅𝐼𝑆𝐼𝑆𝑗𝑖𝑡 =
𝑠𝑗𝑖𝑡
where 𝑋𝑗𝑖𝑡 is the value of indicator 𝑗 in country 𝑖 in period 𝑡, 𝑋𝑗𝑖𝑡 is the average of this
indicator over the five years up to and including 𝑡, and 𝑠𝑗𝑖𝑡 is the standard deviation of the
indicator over this five year period. The interpretation of the standardised score is
straightforward, and standardised scores can be compared since converting our data to scores
results in a distribution with mean 0 and standard deviation 1. A standardised score of 0.5, for
example, indicates that, at the time of liberalisation, the value of this indicator was half a
standard deviation above its recent average. Given the way the indicators are defined, higher
values indicate a deeper crisis. Preliminary evidence that a crisis at the time of liberalisation
may have implications for subsequent growth performance is given in Figure 2. Here we have
divided the liberalising countries into crisis and non-crisis samples according to an arbitrary
threshold of 0.5 for each indicator. As with Figure 1, this reports the average growth rate
across all countries for each year pre and post liberalisation (i.e. 10 years before and 10 years
11
Figure 2: Growth comparison with exogenously chosen crises
OUT INF
XR DEBT
CAD INT
EXT
Non-crisis countries
Crisis countries
12
after), along with the average growth rates over the 10 years pre and post liberalisation. The
solid lines represent non-crisis countries and the dashed lines crisis countries.21 Overall, the
econometric results we report below are nicely reflected in this alternative framework. For
OUT, INF, XR and INT the growth rate in the post-liberalisation period is higher for non-
crisis countries, while the reverse is true for the CAD, DEBT and EXT. The figures also
indicate that growth following liberalisation is higher than that prior to liberalisation for both
regimes.
Of course these indicators only signal a “crisis” if their value exceeds some positive
threshold, which is unknown, a priori. But our interest is not simply in what threshold might
be said to indicate a crisis. Rather we are concerned with what threshold indicates a crisis of
sufficient magnitude that it has implications for the liberaliser’s subsequent growth22. To
determine this we employ the panel threshold regression model of Hansen (1999), and
estimate thresholds for our crisis indicators that allow the coefficient on the liberalisation
dummy to vary discretely depending upon the value of the crisis indicator at the time of
liberalisation. The regression for a single threshold (i.e. two-regime) equation is given by:
𝐼𝑁𝑉
∆ln𝑦𝑖,𝑡 = 𝛽3 ∆ ln 𝑃𝑂𝑃𝑖,𝑡 + 𝛽5 + 𝛿1 𝐿𝐼𝐵𝑖,𝑡 𝐼 𝐶𝑅𝐼𝑆𝑗𝑖𝐿𝐼𝐵 ≤ 𝜆𝑗
+𝛿2 𝐿𝐼𝐵𝑖,𝑡 𝐼 𝐶𝑅𝐼𝑆𝑗𝑖𝐿𝐼𝐵 > 𝜆𝑗 + 𝜐𝑖 + 𝜂𝑡 + 𝜀𝑖,𝑡 (2)

Here the observations are divided into two regimes depending upon whether the value of the
crisis indicator at the time of liberalisation (𝐶𝑅𝐼𝑆𝑗𝑖𝐿𝐼𝐵 ) is smaller or larger than the estimated
threshold for that indicator (𝜆𝑗 ). The impact of liberalisation on growth will be given by 𝛿1
for observations in the low (“non-crisis”) regime (𝐶𝑅𝐼𝑆𝑗𝑖𝐿𝐼𝐵 ≤ 𝜆𝑗 ) and by 𝛿2 for observations
in the high (“crisis”) regime (𝐶𝑅𝐼𝑆𝑗𝑖𝐿𝐼𝐵 > 𝜆𝑗 ). To estimate (2) we firstly have to estimate the
threshold parameter which is taken as the value that minimises the concentrated sum of
squared errors from the least squares regression. To allow us to concentrate on crises we
impose the restriction that the threshold must be positive23. After obtaining an estimate for
the threshold, 𝜆𝑗 , we need to identify whether it is significant or not. In essence, this boils
down to testing whether the threshold regression model is preferred to the linear model, by
testing the null hypothesis that 𝛿1 = 𝛿2 . If we cannot reject this null then the regression
21
Once again, the figures are similar if we consider the five- and twenty-year averages instead.
22
Many studies use these standardised scores to create “crisis” dummy variables. Whether a country is viewed
to be in crisis is determined by imposing some threshold value (usually 1.5 or 2) on the standardised score.
23
To ensure a reasonable number of observations in each regime we generally impose the restriction that at least
ten percent of observations must lie in each.
13
model in (2) collapses to the linear model and the threshold is considered not significant.24 A
complication with this test is that the threshold is not identified under the null hypothesis,
implying that classical tests do not have standard distributions. We follow Hansen (1999) and
bootstrap to obtain the p-value for the test.25
Table 3A: Endogenous Threshold Results

Crisis Indicator
∆ln𝑦
𝑂𝑈𝑇 𝐼𝑁𝐹 𝑋𝑅 𝐷𝐸𝐵𝑇 𝐶𝐴𝐷 𝐼𝑁𝑇 𝐸𝑋𝑇
0.19 0.20 0.21 0.24 0.22 0.25 0.26
(6.25)*** (6.45)*** (7.15)*** (7.43)*** (6.38)*** (7.34)*** (7.40)***
-0.28 -0.23 -0.39 -0.37 -0.54 -0.52 -0.46
∆ln𝑃𝑂𝑃
(-0.99) (-0.82) (-1.46) (-1.05) (-1.78)* (-1.56) (-1.34)
0.038 0.043 0.034 0.023 0.030 0.027 0.020
𝐿𝐼𝐵1
(7.87)*** (7.85)*** (5.73)*** (4.97)*** (5.62)*** (5.50)*** (3.98)***
0.025 0.022 0.024 0.50 0.051 0.012 0.033
𝐿𝐼𝐵2
(4.60)*** (4.85)*** (5.80)*** (4.43)*** (4.33)*** (2.00)** (6.33)***
0.05 0.09 0.9 1.34 1.09 0.51 0.78
𝜆1
(66th) (54th) (34th) (88th) (90th) (75th) (72nd)
p-value 0.009*** 0.00*** 0.045** 0.00*** 0.047** 0.005*** 0.018**
Observations 2494 2458 2384 1890 1961 1774 1774
F-Statistic 12.18*** 12.34*** 13.50*** 9.30*** 8.72*** 10.25*** 10.62***
R2 0.28 0.29 0.29 0.24 0.25 0.28 0.28
Notes: All models include a full set of unreported country and time dummies. t-statistics in brackets based on
White Heteroscedasticity-Consistent standard errors. *, **, *** indicate significance at the 10, 5 and 1 percent
level respectively. The p-value of the significance of the estimated threshold is calculated using the bootstrap
procedure of Hansen (1999).
The results for a single threshold for each indicator are presented in Table 3A, where 𝐿𝐼𝐵1
and 𝐿𝐼𝐵2 refer to the coefficients on liberalisation in the non-crisis and crisis regimes
respectively. In this table 𝜆1 refers to the estimated threshold on each of the crisis variables,
with the figure in brackets reporting the percentile of the distribution at which the threshold
lies. The row entitled p-value reports the p-value from the bootstrap procedure used to test
whether 𝐿𝐼𝐵1 = 𝐿𝐼𝐵2 (i.e. whether 𝛿1 = 𝛿2 ). Despite the variety of indicators used, definite
patterns can be discerned. First, there is at least one significant crisis threshold for all
indicators, and in the majority of cases these thresholds are less than unity and all are less
than the values (1.5 or 2) commonly imposed in the literature. 26 This suggests that less severe
crises may be more important than normally thought. Second, trade liberalisation raises
growth in both crisis and non-crisis regimes, consistent with the results in Figure 2. Third, the
24
Similar test procedures are used to test for multiple thresholds. In the two threshold model for example the
test is defined as a test of two thresholds versus a single threshold.
25
The bootstrap distribution of the test statistic was computed using 1000 replications of the procedure proposed
in Hansen (1999).
26
Table A1 in the Appendix reports whether each particular liberalisation episode would be classed as having
occurred at a time of crisis according to the results in Table 3A.
14
individual indicators fall into two groups in terms of their predictions of the sign of the effect
of a crisis on subsequent growth. Liberalising during a time of crisis involving above
threshold falls in output, increases in inflation or depreciations of the exchange rate is
associated with lower subsequent growth, while liberalising during a crisis involving above
threshold increases in the debt to export ratio or the current account deficit is associated with
enhanced subsequent growth. These results lend support to the arguments, noted above, that
liberalisation at a time of high inflation or unemployment will reduce subsequent growth
benefits by masking relative price signals and delaying resource reallocations. They also
support a view that trade liberalisation is more effective when countries are subject to
external constraints as indicated by DEBT and CAD crises. These issues are investigated
further in the short-run analysis of the next section.
Evidence that different dimensions of an economic crisis may have differing implications for
subsequent growth rates reinforces our interest in exploring their combined effects through
our two estimated factors (𝐼𝑁𝑇 and 𝐸𝑋𝑇). The single threshold results for these indicate that
liberalisation during an internal crisis (𝐼𝑁𝑇 above its threshold) is associated with dampened
growth, while liberalisation during an external crisis (𝐸𝑋𝑇 above its threshold) is associated
with amplified growth. We next use the two independently estimated thresholds to construct
four separate liberalisation dummy variables, each reflecting one of the four possible
situations at the time of liberalisation: 𝐿𝐼𝐵(𝑁, 𝑁) no crisis, 𝐿𝐼𝐵(𝐸, 𝑁) an external but no
internal crisis, 𝐿𝐼𝐵(𝑁, 𝐼) an internal but no external crisis, and 𝐿𝐼𝐵(𝐸, 𝐼) a crisis in both
dimensions. The results are shown as regression 1 in Table 3B. The strongest growth effects
arise when the 𝐸𝑋𝑇 indicator is above its threshold (the coefficient on 𝐿𝐼𝐵(𝐸, 𝑁) is
significantly different from the coefficients on 𝐿𝐼𝐵(𝑁, 𝑁) and 𝐿𝐼𝐵(𝑁, 𝐼), but not that on
𝐿𝐼𝐵(𝐸, 𝐼)). Liberalisation in the absence of a crisis is also associated with significant growth
effects, but liberalisation when there is an internal but no external crisis, has no significant
implications for subsequent growth. While these results are interesting and suggestive, they
are based on dummy variables that are defined by two thresholds each estimated ignoring the
other. Our final step therefore is joint estimation of these thresholds. In view of the apparent
importance of the 𝐸𝑋𝑇 indicator, we use the estimated threshold on 𝐸𝑋𝑇 to divide the sample
into two regimes (𝐸𝑋𝑇 above and below the threshold at the time of liberalisation) and
sequentially search for independent thresholds on 𝐼𝑁𝑇 in each regime. The outcomes are
shown in the final two columns in Table 3B. There is one significant second threshold –
15
indicated by the significant p-value in Column 2, that on 𝐼𝑁𝑇 in the low (non-crisis) regime
for 𝐸𝑋𝑇. Its value is the same as the separately estimated threshold for 𝐼𝑁𝑇, and the results
are virtually identical to those in the second column as a consequence.
In combination, these results support the view that an economic crisis can have a significant
impact on post-liberalisation growth. In particular, liberalisation at a time of internal
economic crisis does not appear to yield subsequent growth benefits of the same magnitude
as those found with liberalisation in the absence of a crisis or where an external crisis is also
present. This is consistent with the discussion in Section 2 which suggested that an internal
crisis would very likely hamper and obscure the potential benefits of a trade liberalisation.
Table 3B: Endogenous Threshold Results

1 2 3
0.26 0.26 0.26
(7.40)*** (7.42)*** (7.40)***
-0.51 -0.51 -0.45
∆ln𝑃𝑂𝑃
(-1.53) (-1.52) (-1.34)
0.025 0.025
𝐿𝐼𝐵(𝑁, 𝑁)
(4.52)*** (4.52)*** 0.020
0.009 0.009 (3.97)***
𝐿𝐼𝐵(𝑁, 𝐼)
(1.49) (1.49)
0.033 0.032
𝐿𝐼𝐵(𝐸, 𝑁)
(6.23)*** 0.033 (6.13)***
0.029 (6.34)*** 0.039
𝐿𝐼𝐵(𝐸, 𝐼)
(2.93)*** (3.35)***
𝜆1 (𝐸𝑋𝑇) 0.78 0.78 0.78
𝜆1 (𝐼𝑁𝑇) 0.51 0.51 0.00
p-value N/A 0.013** 0.584
Observations 1744 1744 1744
F 10.42*** 10.52*** 10.68***
R2 0.28 0.28 0.28
level respectively. The p-value of the significance of the estimated threshold is calculated using the bootstrap
procedure of Hansen (1999).
4. Short-Run Impacts of Liberalisation on Growth

Given our limited sample sizes, our results are likely to reflect a combination of short and
longer run influences. They could therefore be viewed as suggesting that the detrimental
effects of an internal crisis at the time of liberalisation go beyond the short-run. As mentioned
above, GMW (1998, 2002) found evidence of a J-curve effect, whereby growth initially
declines or remains stable following liberalisation, and then increases after a period. We now
modify their approach to consider three issues: first, whether a similar short-run relationship
holds for our sample; second, whether inclusion of short-run effects disturbs our threshold
16
estimates for the long-run growth relationship; third, whether any short-run growth effects of
trade liberalisation are also crisis dependent.
Table 4A: Endogenous Threshold Results (Long-run threshold only)

Crisis Indicator
Linear
0.20 0.19 0.19 0.20 0.23 0.22 0.25 0.25
(6.75)*** (5.95)*** (6.12)*** (6.90)*** (7.23)*** (6.16)*** (7.19)*** (7.25)***
-0.30 -0.28 -0.24 -0.40 -0.36 -0.54 -0.53 -0.47
Δln𝑃𝑂𝑃
(-1.12) (-1.01) (-0.84) (-1.48) (-1.02) (-1.81)* (-1.57) (-1.39)
0.029 0.042 0.048 0.036 0.025 0.034 0.029 0.022
𝐿𝑅1
(7.13)*** (8.03)*** (8.29)*** (5.83)*** (4.42)*** (5.54)*** (5.02)*** (3.83)***
0.029 0.026 0.026 0.051 0.056 0.013 0.035
𝐿𝑅2
(5.14)*** (5.29)*** (5.37)*** (4.66)*** (4.70)*** (2.01)** (5.82)***
-0.022 -0.027 -0.029 -0.016 -0.009 -0.019 -0.010 -0.01

𝑆𝑅(0)1
(-2.73)*** (-3.38)*** (-3.47)*** (-2.53)** (-1.41) (-2.67)*** (-1.51) (-1.60)
0.001 -0.002 -0.002 -0.002 -0.002 -0.005 0.0002 -0.0005
𝑆𝑅(1)1
(0.17) (-0.26) (-0.34) (-0.29) (-0.22) (-0.65) (0.03) (-0.07)
0.003 -0.001 -0.001 0.001 -0.004 -0.0003 -0.002 -0.002
𝑆𝑅(2)1
(0.62) (-0.13) (-0.27) (0.09) (-0.77) (-0.04) (-0.32) (-0.44)
0.021 0.017 0.015 0.015 0.009 0.01 0.009 0.009
𝑆𝑅(3)1
(4.32)*** (3.16)*** (2.92)*** (2.79)*** (1.64) (1.68)* (1.63) (1.53)
𝜆1 0.05 0.09 0.9 1.34 1.09 0.51 0.78

p-value 0.00*** 0.00*** 0.047** 0.00*** 0.004*** 0.007*** 0.019**
Observation 1961 1774 1774
2619 2494 2458 2384 1890
s
F 13.31*** 11.95*** 12.16*** 13.20*** 9.03*** 8.53*** 9.92*** 10.25***
R2 0.30 0.29 0.30 0.29 0.24 0.25 0.28 0.28
Notes: See Table 3B. LRI and SR(J)I refer to the long-run and short-run liberalisation dummies in regime I = 1-
2.
As a first step to capturing both the short-run and long-run effects we estimate:
𝐼𝑁𝑉 3
∆ln𝑦𝑖,𝑡 = 𝛽3 ∆ ln 𝑃𝑂𝑃𝑖,𝑡 + 𝛽5 + 𝛿𝐿𝑅𝑖,𝑡 + 𝜙𝑗 𝑗 =0 𝑆𝑅(𝑗)𝑖,𝑡 + 𝜐𝑖 + 𝜂𝑡 + 𝜀𝑖,𝑡 (3)
Alongside the long-run (post-) liberalisation dummy described above (now relabelled 𝐿𝑅),
this equation includes four additional liberalisation dummies, each corresponding to a single
year – the year of liberalisation (𝑆𝑅(0)) and each of the subsequent three years (𝑆𝑅(1),
𝑆𝑅(2) and 𝑆𝑅(3)). The impact on growth in the year of liberalisation and in each of the
subsequent three years is therefore given by 𝛿 + 𝜙𝑗 : 𝑗 = 0, … ,3. The results are shown in the
second column of Table 4A. Estimated coefficients on 𝐼𝑁𝑉/𝐺𝐷𝑃, Δln𝑃𝑂𝑃 and 𝐿𝑅 are very
similar to those in the corresponding regression in Table 1. The estimates for the short-run
post-liberalisation dummies indicate that growth is significantly lower than the post-
liberalisation average in the year of liberalisation, is no different from this average in the
following two years and is sufficiently higher in the third year to recover what had been lost
17
in the year of liberalisation. Our sample thus replicates the type of J-curve effects found
previously.
To begin the process of examining how these results are affected by a crisis, we initially used
a modified version of equation (3) which estimated common crisis thresholds for all five
post-liberalisation dummies. The broad pattern of outcomes remained as before, but for three
of the indicators we now have a significant second threshold. In the light of this evidence that
different crisis levels may be applicable to the short and long-run growth effects27, we
proceeded in two steps. First, we estimated crisis thresholds for the long-run dummies in (3)
only, applying no thresholds on the short-run dummies. The results are shown in the
remaining columns of Table 4A. The estimated thresholds for the crisis indicators are
identical to those of the preceding section, and coefficients on the long-run post-liberalisation
dummies are the same or slightly higher in both regimes. For all the single indicators (except
DEBT), the estimated coefficients on the short-run dummies show the same J-curve pattern as
the linear case. However, there is enough variation in the effects of these individual indicators
that when they are aggregated (along with DEBT) into the combined indicators no significant
short-run effects are evident. Our second step involves estimating crisis-indicator-based
thresholds for the short-run post-liberalisation dummies, taking as given the estimated
thresholds for the long-run dummies. The equation estimated is:
∆ln𝑦𝑖,𝑡 =
𝐼𝑁𝑉 3
𝛽3 ∆ ln 𝑃𝑂𝑃𝑖,𝑡 + 𝛽5 + 𝛿1 𝐿𝑅𝑖,𝑡 𝐼 𝐶𝑅𝐼𝑆𝑖𝑗𝐿𝐼𝐵 ≤ 𝜆𝐿𝑗 + 𝑗 =0 𝜙1,𝑗 𝑆𝑅(𝑗)𝑖,𝑡 𝐼 𝐶𝑅𝐼𝑆𝑖𝑗𝐿𝐼𝐵 ≤ 𝜆𝑆𝑗 +
3
𝛿2 𝐿𝑅𝑖,𝑡 𝐼 𝐶𝑅𝐼𝑆𝑖𝑗𝐿𝐼𝐵 > 𝜆𝐿𝑗 + 𝑗 =0 𝜙2,𝑗 𝑆𝑅(𝑗)𝑖,𝑡 𝐼 𝐶𝑅𝐼𝑆𝑖𝑗𝐿𝐼𝐵 > 𝜆𝑆𝑗 + 𝜐𝑖 + 𝜂𝑡 + 𝜀𝑖,𝑡 (4)
where 𝜆𝐿𝑗 is the long-run threshold for crisis indicator 𝑗 as reported in Table 4A. The results
are shown in Table 4B.
Only two of the individual crisis indicators (𝑂𝑈𝑇 and 𝑋𝑅) have significant short-run
thresholds – as signified by p-value (𝜆𝑆 ) – and both are higher than their long-run values.
The estimated coefficients on the long-run liberalisation dummies are largely unaffected.
27
The single thresholds and the lower values of the double thresholds were similar to the thresholds reported in
Table 3A (except for XR). The estimated long-run coefficients in the non-crisis regimes were the same or
slightly higher than the corresponding coefficients in Table 3A. Of the indicators with two thresholds, only INF
has coefficients in its crisis regimes that significantly differ from each other, indicating that the second
thresholds arise to accommodate the short-run effects for the other indicators at least.
18
Table 4B: Endogenous Threshold Results (Short-run Thresholds)
Crisis Variables
0.19 0.20 0.21 0.23 0.22 0.25 0.25
(5.97)*** (6.15)*** (6.93)*** (7.24)*** (6.17)*** (7.19)*** (7.21)***
-0.31 -0.23 -0.38 -0.37 -0.53 -0.59 -0.47
Δln𝑃𝑂𝑃
(-1.11) (-0.80) (-1.40) (-1.04) (-1.76)* (-1.73)* (-1.39)
0.041 0.048 0.037 0.025 0.034 0.028 0.022
𝐿𝑅1
(7.80)*** (8.35)*** (6.00)*** (4.34)*** (5.50)*** (4.84)*** (3.86)***
0.033 0.026 0.025 0.055 0.056 0.017 0.034
𝐿𝑅2
(5.80)*** (5.23)*** (5.24)*** (5.31)*** (4.73)*** (2.56)** (5.68)***
-0.023
-0.013 -0.026 -0.01 -0.007 0.007 -0.019
𝑆𝑅 0 1 (-
(-1.85)* (-2.92)*** (-1.56) (-1.21) (0.96) (-2.18)**
2.89)***
-0.114 -0.053 -0.021 -0.021 -0.005 -0.040 -0.0003
𝑆𝑅 0 2
(-4.44)*** (-2.47)** (-2.07)** (-0.83) (-0.45) (-5.34)*** (-0.03)
0.002 -0.005 -0.007 0.002 -0.006 0.005 0.0007

𝑆𝑅 1 1
(0.28) (-0.72) (-0.93) (0.33) (-0.78) (0.69) (0.07)
-0.021 0.017 0.004 -0.022 0.001 -0.010 -0.002
𝑆𝑅 1 2
(-1.19) (0.91) (0.44) (-0.62) (0.11) (-0.79) (-0.29)
0.001 -0.005 -0.011 -0.003 0.004 -0.004 0.0003

𝑆𝑅 2 1
(0.24) (-1.00) (-1.85)* (-0.54) (0.55) (-0.61) (0.04)
-0.011 0.024 0.012 -0.010 -0.014 0.002 -0.006
𝑆𝑅 2 2
(-0.79) (1.27) (1.72)* (-0.81) (-1.75)* (0.16) (-0.96)
0.013 0.012 0.0007 0.012 0.014 0.010 0.012

𝑆𝑅 3 1
(2.44)** (2.31)** (0.12) (1.88)* (2.25)** (1.64) (1.42)
0.029 0.035 0.029 -0.005 -0.005 0.007 0.005
𝑆𝑅 3 2
(2.40)** (2.27)*** (3.68)*** (-0.47) (-0.43) (0.65) (0.73)
𝜆𝐿 𝑒𝑥𝑜𝑔 0.05 0.09 0.9 1.34 1.09 0.51 0.78

𝜆𝑆 0.79 1.6 1.4 1.3 0.59 0.17 0.45
p-value (𝜆𝑆 ) 0.00*** 0.131 0.044** 0.658 0.337 0.011** 0.621
Observation
2494 2458 2384 1890 1961 1774 1774
s
F-Statistic 11.91*** 11.84*** 12.84*** 8.95*** 8.26*** 9.92*** 9.89***
R2 0.30 0.30 0.30 0.24 0.25 0.29 0.28
Notes: 𝑆𝑒𝑒 𝑇𝑎𝑏𝑙𝑒 4𝐴
The estimated coefficients on the short-run dummies again exhibit a similar J-curve pattern.
Compared to the post-liberalisation long-run, there is lower growth in the year of
liberalisation and higher growth three years later. Negative growth in the liberalisation year
is predicted for countries in the high crisis regimes by the OUT and INF indicators, again
confirming concerns that high inflation or unemployment may mask relative price signals
and delay resource reallocation. For the combined indicators, we find a significant short-run
19
threshold for INT, at a value below its long-run threshold. But the only J-curve effect evident
is lower growth in the short-run crisis regime in the year of liberalisation28.
We can now address the three issues noted at the start of this section. First, our results
confirm the presence of the short-run J-curve effects found in the earlier literature. Second,
the long-run results are essentially unaffected by the allowance for short-run effects. The
estimated long-run crisis-thresholds are unchanged. The estimated coefficients tend to be
slightly higher, but the pattern is unchanged. Third, there is evidence that the short-run
growth J-curve is also crisis sensitive. Output, inflation or exchange rate crises at the time of
liberalisation imply lower growth in the liberalisation year but a stronger recovery three years
later. A current account crisis exhibits the opposite pattern. The only significant effects for
the combined indicators are in the liberalisation year, where there is lower growth with an
internal crisis or the absence of an external crisis.
5. Robustness and Extensions

For brevity we generally report the robustness results for the long-run model only.
5.1 Longer-lived crises: If crises are long-lived, the values of our indicators at the time of
liberalisation may not be too different from their average values over the last five years. A
country in such a long-lived crisis may therefore be classed as a non-crisis liberaliser. To test
the sensitivity of our results to the crisis length, we recalculate the crisis scores using a ten-
year rather than five-year average. The results for the five crisis indicators are reported in
Table 5. It is immediately clear that the pattern of coefficients on LIB1 and LIB2 are as
above, with the impact of liberalisation on growth being higher in non-crisis countries for
OUT, INF and XR, and higher in crisis countries for DEBT and CAD. Moreover, the
coefficients in the two regimes are in all cases statistically significant and the p-values
indicate that the coefficients in the two regimes are significantly different.
28
Given the mixed bag of short-run results for the individual indicators, it is perhaps unsurprising that
consideration of separate short and long-run thresholds tends to generate few significant short-run results for our
composite indicators. Given this outcome we see little point in pursuing joint thresholds in the short-run.
20
Table 5: Crisis Results Based on Ten-Year Averages
∆ln𝑦 𝑂𝑈𝑇 𝐼𝑁𝐹 𝑋𝑅 𝐷𝐸𝐵𝑇 𝐶𝐴𝐷
0.17 0.20 0.23 0.24 0.23
(5.42)*** (6.07)*** (8.54)*** (7.86)*** (7.41)***
-0.26 -0.39 -0.40 -0.44 -0.49
∆ln𝑃𝑂𝑃
(-0.82) (-1.21) (-1.39) (-1.32) (-1.63)
0.040 0.042 0.026 0.017 0.016
𝐿𝐼𝐵1
(6.99)*** (6.73)*** (5.81)*** (3.80)*** (3.55)***
0.022 0.024 0.018 0.40 0.031
𝐿𝐼𝐵2
(3.23)*** (3.99)*** (4.46)*** (4.92)*** (5.67)***
𝜆1 0.051 0.73 1.45 1.64 0
p-value 0.002*** 0.00*** 0.057* 0.002*** 0.00***
Observations 2135 2024 2033 1758 1877
F-Statistic 9.61*** 8.87*** 11.10*** 9.82*** 9.35***
R2 0.25 0.23 0.28 0.27 0.26
Notes: See Table 3A
5.2 Excluding transition economies: Transition countries undertook trade liberalisation in

the same years as they dismantled their socialist economies, making it hard to identify the
effects of trade reform. Results when excluding transition countries are reported in Table 6.
They are remarkably similar to those in Section 3. We again find the impact of liberalisation
tends to be higher for non-crisis countries when crises are measured using OUT, INF and XR,
though the difference in coefficients is no longer significant for XR. We find for DEBT and
CAD that liberalisation again tends to have a stronger impact in crisis countries as opposed to
non-crisis countries, though only in the case of DEBT is the difference significant. Finally,
the results on the combined crisis indicators, INT and EXT, are consistent with the results
above, with liberalisation being associated with a stronger impact on growth in non-crisis
countries when INT is our crisis indicator, and crisis countries when EXT is our crisis
indicator.
5.3. Retaining the terms of trade index: We excluded TTI to increase our sample size. We
examine the robustness of our results to this exclusion in two steps. Firstly, we re-estimate
the threshold results for the sub-sample of observations for which we have data on TTI, while
continuing to exclude TTI from the equation (Table 7A). This allows us to examine whether
any differences in results are due to the changed sample size rather than the inclusion of the
TTI variable. Secondly, we re-estimate our thresholds including the TTI variable (Table 7B).
21
Table 6: Endogenous Threshold Results (Excluding Transition Countries)29
Crisis Indicator
∆ln𝑦
0.19 0.20 0.20 0.23 0.21 0.24 0.24
(7.32)*** (7.55)*** (17.31)*** (7.40)*** (6.27)*** (7.88)*** (7.89)***
-0.50 -0.49 -0.54 -0.49 -0.51 -0.49 -0.51
∆ln𝑃𝑂𝑃
(-1.92)* (-1.87)* (-2.06)** (-1.46) (-1.61) (-1.46) (-1.51)
0.023 0.026 0.022 0.019 0.025 0.021 0.015
𝐿𝐼𝐵1
(6.15)*** (6.58)*** (5.92)*** (4.52)*** (5.19)*** (4.93)*** (3.32)***
0.012 0.007 0.026 0.029 0.034 0.011 0.025
𝐿𝐼𝐵2
(2.69)*** (1.78)* (3.03)*** (4.78)*** (5.16)*** (1.60) (5.31)***
𝜆1 0.09 0.41 1.68 0.74 1.08 0.79 0.49
p-value 0.01*** 0.00*** 0.19 0.07* 0.28 0.08* 0.02*
Observations 2259 2223 2240 1824 1831 1703 1703
F-Statistic 14.72*** 14.72*** 14.64*** 9.49*** 9.71*** 10.45*** 10.59***
R2 0.29 0.30 0.29 0.24 0.24 0.28 0.28
Notes: See Table 3A
Table 7A: Thresholds on the TTI sub-sample

0.27 0.28 0.27 0.27 0.25
(7.47)*** (7.72)*** (7.43)*** (6.12)*** (6.06)***
-0.72 -0.81 -0.85 -1.48 -1,05
∆ln𝑃𝑂𝑃
(-1.92)* (-2.21)** (-2.30)** (-2.99)*** (-2.43)***
0.024 0.023 0.023 0.017 0.014
𝐿𝐼𝐵1
(5.69)*** (5.37)*** (4.88)*** (3.60)*** (2.83)***
0.002 0.003 0.012 0.037 0.032
𝐿𝐼𝐵2
(0.35) (0.67) (2.76)*** (4.92)*** (5.31)***
𝜆1 0.04 0.53 1.21 0.74 0
p-value 0.00*** 0.00*** 0.06* 0.004*** 0.001***
Observations 1290 1268 1303 993 1063
F-Statistic 19.03*** 29.34*** 45.49*** 4.18*** 15.05***
R2 0.35 0.36 0.35 0.27 0.28
Notes: See Table 3A
Table 7B: Inclusion of TTI Index

0.27 0.28 0.27 0.27 0.26
(7.56)*** (7.82)*** (7.52)*** (6.18)*** (6.14)***
-0.71 -0.81 -0.84 -1.45 -1.04
∆ln𝑃𝑂𝑃
(-1.87)* (-2.16)** (-2.25)** (-2.89)*** (-2.35)**
0.019 0.019 0.018 0.016 0.019
∆ln𝑇𝑇𝐼
(2.21)** (2.25)** (2.21)** (1.91)* (2.11)**
0.024 0.023 0.023 0.017 0.014
𝐿𝐼𝐵1
(5.68)*** (5.36)*** (4.87)*** (3.58)*** (2.79)***
0.001 0.003 0.012 0.037 0.031
𝐿𝐼𝐵2
(0.25) (0.61) (2.69)*** (4.90)*** (5.32)***
𝜆1 0.02 0.53 1.22 0.74 0
p-value 0.00*** 0.00*** 0.058* 0.003*** 0.001***
Observations 1290 1268 1303 993 1063
F-Statistic 19.23*** 29.06*** 49.59*** 4.19*** 16.20***
R2 0.36 0.37 0.35 0.27 0.29
Notes: See Table 3A
29
The initial results when excluding transition countries are very similar to those reported in Table 1.
22
The results in each case are consistent with those above. In particular, we continue to find
that the relationship between liberalisation and growth is stronger in the non-crisis regime for
OUT, INF and XR, and stronger in the crisis regime for DEBT and CAD.
5.4 Allowing for the effects of crises on growth: We now examine the possibility that
estimated non-linearities in the liberalisation variable simply reflect omitted effects related to
the crisis variables rather than being attributable to liberalisation. Bruno and Easterly (1995)
examined countries that had high-inflation crises (inflation above 40 percent annually for two
or more years) and found that growth fell sharply during the high inflation crisis. They further
showed that growth after the crisis was higher than before, even though inflation had returned
to pre-crisis levels. Easterly (1996) found that GDP growth was already positive in the year
that high inflation declined from its pre-stabilisation peak, with growth becoming stronger in
the periods following. Bruno and Easterly (1996) report that broad reforms were the usual
outcome of high-inflation crises and show a strong association between their measure of
stabilisation from high-inflation and the openness measure of Sachs and Warner (1995).
Table 8A reports the results of adding contemporaneous values of our crisis indicators.30 The
coefficients on the liberalisation dummy are robust to their inclusion, ranging from 0.018 to
0.022, and remain significant. The coefficients on the crisis indicators are as expected,
negative and significant, except for CAD where a significantly positive coefficient is found.
Since a large current account deficit is likely to reflect relatively high domestic expenditure
which is also likely to generate relatively high output, the positive relationship for CAD is not
surprising. Table 8B reports results from estimating thresholds on liberalisation. The
coefficients on the crisis indicators are consistent with those reported in the previous table.
The pattern of threshold results is similar to that in Section 3 (though the positioning of the
threshold often differs). For OUT, INF and XR we find the impact of liberalisation to be
larger in non-crisis countries, and for DEBT we find the opposite. The major difference is that
for a CAD crisis we now find the benefits of liberalisation to be greater in non-crisis
countries.
30
We report, but do not emphasise the results using per capita income growth as our crisis indicator, because of
its definitional relationship with the annual growth rate of per capita income. The correlation between the two is
-0.57. This is not an issue in the main analysis, however, where we only consider the value of the crisis indicator
at the date of liberalisation, which is generally not correlated with the annual growth rate of per capita income.
23
Table 8A: Inclusion of Crisis Indicator Linearly
0.19 0.21 0.21 0.24 0.21
(8.20)*** (7.91)*** (8.69)*** (7.11)*** (5.67)***
0.12 -0.42 -0.49 -0.80 -0.68
∆ln𝑃𝑂𝑃
(0.55) (-1.53) (-1.98)** (-2.11)** (-2.11)**
0.019 0.022 0.018 0.019 0.022
𝐿𝐼𝐵
(5.92)*** (5.98)*** (5.57)*** (5.20)*** (5.65)**
-0.035 -0.012 -0.004 -0.006 0.003
𝐶𝑅𝐼𝑆𝐼𝑆
(-34.48)*** (-9.61)*** (-3.03)*** (-5.41)*** (2.89)***
Observations 2454 2405 2372 1599 1675

F-Statistic 37.99*** 15.38*** 16.81*** 27.34*** 14.18***
R2 0.60 0.33 0.34 0.34 0.34
Notes: See Table 3A.
Table 8B: Threshold Results when including the Crisis variables

0.19 0.20 0.21 0.25 0.22
(7.47)*** (7.35)*** (8.01*** (6.80)*** (5.02)***
0.07 -0.42 -0.53 -0.92 -0.93
∆ln𝑃𝑂𝑃
(0.31) (-1.46) (-2.07)** (-2.17)** (-2.35)**
0.023 0.032 0.027 0.016 0.029
𝐿𝐼𝐵1
(5.93)*** (6.43)*** (4.68)*** (3.84)*** (5.35)***
0.012 0.015 0.019 0.032 0.016
𝐿𝐼𝐵2
(2.73)*** (3.56)*** (5.10)*** (4.56)*** (2.41)**
-0.04 -0.012 -0.004 -0.006 0.004
𝐶𝑅𝐼𝑆𝐼𝑆
(-33.74)*** (-9.39)*** (-2.88)*** (-4.93)*** (3.05)**
𝜆1 0.17 0.03 0.7 1.03 0.69
p-value 0.005*** 0.00*** 0.186 0.014** 0.04**
Observations 2355 2310 2225 1387 1328
F-Statistic 36.24*** 13.80*** 15.06*** 10.34*** 8.72***
R2 0.59 0.32 0.31 0.28 0.27
Notes: See Table 3A
In Table 8C we add to the threshold on liberalisation a second threshold on the crisis variable,
examining whether the relationship between the crisis and growth depends on the level of the
crisis indicator. For OUT, INF and DEBT the threshold results indicate that the link between
crises and growth is negative in both the low and high crisis regime, but the relationship is
stronger in the high crisis regime. This result is therefore similar to that found by Bruno and
Easterly, albeit using a different methodology and different variable definition. Rather than
the rate of inflation, our indicator measures differences in the inflation rate relative to its
average over the recent past. The results for XR are similar, though the coefficient in the low
regime is positive (albeit insignificant). For CAD we obtain positive coefficients in both
regimes, though only significant in the low regime and we cannot reject the hypothesis that
the coefficients are the same. The results for CAD may indicate that our liberalisation results
are to some degree capturing a relationship between CAD and growth.
24
Table 8C: Thresholds on both Liberalisation and Crisis
0.19 0.20 0.21 0.25 0.22
(7.75)*** (7.38)*** (7.99)*** (6.91)*** (5.03)***
0.02 -0.46 -0.60 -0.95 -0.93
∆ln𝑃𝑂𝑃
(0.07) (-1.63) (-2.30)** (-2.26)** (-2.63)***
0.022 0.031 0.027 0.017 0.029
𝐿𝐼𝐵1
(5.93)*** (6.42)*** (4.88)*** (3.90)*** (5.35)***
0.012 0.016 0.018 0.032 0.015
𝐿𝐼𝐵2
(2.69)*** (3.68)** (4.68)*** (4.61)*** (2.36)**
-0.032 -0.008 0.001 -0.005 0.005
𝐶𝑅𝐼𝑆𝐼𝑆𝐿
(-31.27)*** (-7.11)*** (0.69) (-3.42)*** (3.43)***
-0.044 -0.02 -0.01 -0.011 0.0016
𝐶𝑅𝐼𝑆𝐼𝑆𝐻
(-16.68)*** (-6.44)*** (-5.88)*** (-4.15)*** (0.60)
𝜆𝐿𝐼𝐵 0.17 0.03 0.7 1.03 0.69
𝜆𝐶𝑅𝐼𝑆𝐼𝑆 1.25 1.42 1.57 1.45 1.35
p-value 0.00*** 0.00*** 0.00*** 0.012*** 0.26
Observations 2355 2310 2225 1387 1328
F-Statistic 37.54*** 13.64*** 16.58*** 10.26*** 8.66***
R2 0.59 0.33 0.33 0.29 0.27
level respectively. p-value refers to the significance of the threshold on the crisis variable and is calculated
using the bootstrap procedure of Hansen (1999).
While allowing our crisis indicators to have a direct effect on growth appears to have a
limited impact on our liberalisation results (except for CAD), we should still examine
whether they are being driven by the response of the economy to a crisis around the time of
liberalisation. It could be argued that we already account for this by the inclusion of the
short-run liberalisation variables in Table 4A. This is only true, however, if liberalisation is
indeed a normal response to a crisis. Our data indicates this is not the case. For OUT 49 of the
75 observations were negative at the time of liberalisation, with the number being 40 out of
74 for INF. While the number of liberalisations that would be considered to have taken place
in a non-crisis period are smaller for the other indicators (5 out of 68 for XR, 25 out of 53 for
DEBT and 33 out of 57 for CAD), they still suggest that many, if not a majority, of
liberalisations would have taken place in non-crisis conditions. To examine whether our
results are driven by a natural recovery from the crisis rather than a response to liberalisation,
we construct a variable that is equal to one at the peak of any crisis (for each of the five crisis
indicators)31 then calculate lags of it for three periods and include it alongside our
liberalisation variable.
31
A crisis is considered to occur if we obtain positive values of the crisis indicator for three consecutive periods,
with the largest value taken as the peak of the crisis. Note particularly that this variable is defined for any crisis,
independently of whether a trade liberalisation also occurs.
25
Table 9A reports the linear results, where CRISPEAK is the dummy representing the peak of
the crisis. The results are largely consistent with those reported in Table 1, with the
liberalisation dummy taking a value between 0.023 and 0.028. The coefficients on the crisis
indicators tend to be negative and significant for the first couple of periods for OUT, INF and
XR, steadily decreasing in absolute size and becoming insignificant (or becoming positive
and significant) in subsequent years. For DEBT and CAD no significant effects are found.
Table 9A: Linear Results with Crisis Variables Included

0.17 0.18 0.18 0.18 0.18
(5.63)*** (6.18)*** (6.24)*** (6.11)*** (6.01)***
-0.24 -0.32 -0.30 -0.30 -0.29
∆ln𝑃𝑂𝑃
(-0.36) (-1.18) (-1.08) (-1.08) (-1.05)
0.023 0.027 0.028 0.028 0.028
𝐿𝐼𝐵1
(6.03)*** (6.89)*** (6.89)*** (6.97)*** (6.97)***
-0.05 -0.021 -0.012 -0.002 0.005
𝐶𝑅𝐼𝑆𝑃𝐸𝐴𝐾
(-11.53)*** (-3.85)*** (-2.41)** (-0.49) (1.26)
𝐶𝑅𝐼𝑆𝑃𝐸𝐴𝐾 − 1 -0.02 -0.012 0.002 0.001 -0.005
(-4.94)*** (-2.43)** (0.38) (0.26) (-1.01)
𝐶𝑅𝐼𝑆𝑃𝐸𝐴𝐾 − 2 -0.016 -0.0006 0.008 -0.005 -0.01
(-3.74)*** (-0.14) (2.13)** (-1.14) (-2.48)
𝐶𝑅𝐼𝑆𝑃𝐸𝐴𝐾 − 3 -0.007 0.001 0.007 0.002 -0.006
(-1.55) (0.29) (1.78)* (0.45) (-1.40)
Observations 2436 2436 2436 2436 2436

F-Statistic 22.86*** 17.53*** 18.00*** 17.57*** 18.36***
R2 0.35 0.31 0.30 0.30 0.24
Table 9B reports the endogenous threshold results when including these crisis variables. The
coefficients are similar to those in Table 9A with negative coefficients in the year of the peak
of the crisis for OUT, INF and XR, with the coefficients declining and becoming insignificant
in subsequent years. For DEBT and CAD the coefficients tend to be insignificant, with the
exception of the second year after the crisis when a significant negative coefficient is found.
The liberalisation variables are qualitatively similar to those reported in Table 3A, with a
coefficient that is larger in size in non-crisis countries (LIB1) when OUT, INF and XR are
used as indicators, and larger in crisis countries (LIB2) when DEBT and CAD are used. In the
case of OUT and XR however, the differences in the coefficients are no longer found to be
statistically significant (despite being economically important).
26
Table 9B: Long-Run Threshold Results with Crisis Variables Included
0.16 0.18 0.19 0.24 0.21
(5.08)*** (5.78)*** (6.74)*** (7.31)*** (6.22)***
-0.29 -0.33 -0.45 -0.42 -0.64
∆ln𝑃𝑂𝑃
(-1.02) (-1.16) (-1.69)* (-1.08) (-2.12)**
0.030 0.038 0.035 0.024 0.028
𝐿𝐼𝐵1
(6.51)*** (7.23)*** (5.15)*** (5.01)*** (5.52)***
0.016 0.019 0.022 0.053 0.036
𝐿𝐼𝐵2
(3.05)*** (4.25)*** (5.62)*** (4.69)*** (5.52)***
-0.051 -0.02 -0.014 -0.008 0.002
(-11.29)*** (-3.63)*** (-2.74)*** (-1.53) (0.56)
𝐶𝑅𝐼𝑆𝑃𝐸𝐴𝐾 − 1 -0.020 -0.011 -0.003 -0.004 -0.004
(-4.79)*** (-2.20)** (-0.67) (-0.76) (-0.82)
𝐶𝑅𝐼𝑆𝑃𝐸𝐴𝐾 − 2 -0.016 -0.0002 0.004 -0.011 -0.012
(-3.64)*** (-0.05) (1.06) (2.33)*** (-2.92)***
𝐶𝑅𝐼𝑆𝑃𝐸𝐴𝐾 − 3 -0.007 0.001 0.003 -0.003 -0.006
(-1.56) (0.34) (0.82) (-0.82) (-1.49)
𝜆1 0.17 0.15 0.70 1.34 1.06

p-value 0.30 0.00*** 0.34 0.00*** 0.018**
Observations 2337 2301 2239 1783 1840
F-Statistic 18.81*** 14.65*** 17.31*** 10.23*** 8.72***
R2 0.34 0.30 0.31 0.25 0.26
Notes: See Table 3A
Finally, Table 9C reports the results when including short-run liberalisation indicators, which
allow us to examine whether the results on the short-run effects are simply reflecting
adjustment to a crisis, or whether they really capture the short-run effects of liberalisation.
The results on the (long-run) liberalisation variable are consistent with those reported in
Table 9B, as are the coefficients on the crisis peak and its lags. The results show that the
impact of liberalisation on growth is negative in the year of liberalisation (and usually
significant) and in the following two years (though not significant). In the third year the
impact of liberalisation is positive, and often significant. Overall therefore, the J-curve is
confirmed even when a measure capturing any crisis is included.
27
Table 9C: Endogenous Threshold Results (Long-run threshold only)
0.15 0.17 0.18 0.23 0.20
(4.76)*** (5.45)*** (6.51)*** (6.90)*** (6.00)***
-0.29 -0.32 -0.45 -0.43 -0.65
Δln𝑃𝑂𝑃
(-1.02) (-1.14) (-1.68)* (-1.11) (-2.14)**
0.036 0.043 0.038 0.025 0.034
𝐿𝑅1
(7.19)*** (7.76)*** (5.30)*** (4.23)*** (5.53)***
0.023 0.024 0.026 0.038 0.042
𝐿𝑅2
(4.07)*** (4.86)*** (5.51)*** (5.05)*** (5.68)***
-0.028 -0.026 -0.018 -0.010 -0.021

𝑆𝑅(0)1
(-3.74)*** (-3.30)*** (-2.96)*** (-1.43) (-3.11)***
-0.007 -0.005 -0.004 -0.002 -0.006
𝑆𝑅(1)1
(-1.10) (-0.83) (-0.63) (-0.25) (-0.80)
-0.004 -0.004 -0.003 -0.003 -0.003
𝑆𝑅(2)1
(-0.84) (-0.80) (-0.66) (-0.58) (-0.49)
0.011 0.013 0.012 0.010 0.007
𝑆𝑅(3)1
(2.32)** (2.75)*** (2.36)** (1.69)* (1.30)
-0.051 -0.019 -0.013 -0.008 0.002
(-11.28)*** (-3.45)*** (-2.60)*** (-1.36) (0.48)
𝐶𝑅𝐼𝑆𝑃𝐸𝐴𝐾 − 1 -0.020 -0.009 -0.002 -0.003 -0.004
(-4.77)*** (-1.80) (-0.52) (-0.71) (-0.85)
𝐶𝑅𝐼𝑆𝑃𝐸𝐴𝐾 − 2 -0.015 0.001 0.005 -0.011 -0.012
(-3.55)*** (0.29) (1.17) (-2.30)** (-2.96)***
𝐶𝑅𝐼𝑆𝑃𝐸𝐴𝐾 − 3 -0.005 0.002 0.003 -0.004 -0.007
(-1.20) (0.52) (0.76) (-0.92) (-1.54)
𝜆1 0.17 0.16 0.69 1.34 1.11

p-value 0.41 0.00*** 0.29 0.00*** 0.017**
Observations 2337 2301 2239 1783 1840
F-Statistic 9.47*** 7.83*** 8.09*** 5.30*** 5.91***
R2 0.35 0.31 0.31 0.25 0.27
5.5 Liberalisation and investment: As Wacziarg and Welch (2003) observe, the literature
suggests that investment is an important channel through which trade liberalisation may
affect economic growth. This has been ignored in our analysis above. Table 10 reports the
results from estimating the relationship between trade liberalisation and the investment share,
and the impact of our various crises indicators on this relationship. Consistent with the
previous literature we find that liberalisation has a positive and significant effect on
investment. We also find evidence of significant differences in the coefficients in the crisis
regime for each of the indicators, except INF and XR. Moreover, for all indicators (except
DEBT) differences in the coefficients in the two regimes, whether significant or not, work in
the same direction as those on growth. For DEBT the results are in contrast to those for
growth, with the effect of liberalisation on investment being higher in the non-crisis regime,
and indeed negative in the crisis regime. A 𝐷𝐸𝐵𝑇 crisis at the time of liberalisation reduces
the subsequent investment share, perhaps reflecting tightened credit conditions.
28
Table 10: Liberalisation and Investment
Linear 𝑂𝑈𝑇 𝐼𝑁𝐹 𝑋𝑅 𝐷𝐸𝐵𝑇 𝐶𝐴𝐷 𝐼𝑁𝑇 𝐸𝑋𝑇
0.017 0.022 0.016 0.02 0.021 0.005 0.023 0.014
𝐿𝐼𝐵1
(3.79)*** (4.48)*** (2.84)*** (3.20)*** (4.02)*** (0.96) (4.00)*** (2.54)***
-0.0005 0.015 0.018 -0.029 0.024 0.001 0.026
𝐿𝐼𝐵2
(-0.07) (2.77)*** (3.68)*** (-3.49)*** (2.49)** (0.19) (3.54)***
𝜆1 0.05 0.09 0.9 1.34 1.09 0.51 0.78

𝐿𝐼𝐵1 = 𝐿𝐼𝐵2 14.51*** 0.81 0.21 55.36*** 4.53** 9.12*** 3.31*
Observations 2620 2495 2459 2385 1891 1962 1775 1775
F-Statistic 318.1*** 320.1*** 313.7*** 330.9*** 298.7*** 296.1*** 278.9*** 277.7***
R2 0.94 0.94 0.94 0.94 0.94 0.94 0.94 0.94
Notes: All regressions include a full set of country and time fixed effects. t-statistics in brackets. All models
estimated using White Heteroscedasticity-Consistent standard errors. *, **, *** indicate significance at the 10,
5 and 1 percent level respectively.
Since investment appears to respond to trade liberalisation, and the response depends on the
presence or absence of a crisis, some account of its endogeneity should be allowed for (see
Temple, 1999)32. As a step in this direction we adapt the approach of Caner and Hansen
(2004) that allows one to estimate the threshold model in the presence of endogenous
explanatory variables. This proceeds in three stages: in the first stage one uses Two-Stage-
Least-Squares (2SLS) or General Method of Moments (GMM) to obtain predicted values of
the endogenous variables (i.e. investment); in the second stage one replaces the endogenous
variables with their predicted values and estimates the threshold model using OLS in the
same way as in Hansen (1999); finally, given the estimated value of the threshold, one
estimates the coefficients in the two sub-samples using 2SLS (or GMM).33 To implement this
method we require excluded instruments for investment. The instrument we adopt is taken
from Cook (2002) who argues that age structure is an important determinant of both human
and physical capital. We follow his approach and construct a measure of the average age of
the workforce as the sum of the midpoint age of each five-year age group between 15 and 64
multiplied by the share of the population aged between 15 and 64 in that age group. 34 The
32
One possible extension to our approach would be to estimate a simultaneous equation model identifying the
channels through which liberalisation affects growth (as in Wacziarg, 2001), allowing for thresholds related to
crises. But such an extension would require imposing thresholds, since as far as we know methods have not
been developed to estimate thresholds in a simultaneous equation model.
33
We have to adapt this approach since we only consider a threshold on the liberalisation variable and not all
explanatory variables. The Caner and Hansen method involves estimating the coefficients in stage 3 on the two
sub-samples separately. This would imply that the coefficients on investment and population growth would also
be allowed to vary across sub-samples. To maintain consistency therefore, in stage 3 we simply report the
results of 2SLS on a single regression on the full sample. In unreported results we adopt the exact approach of
Caner and Hansen – allowing the coefficients on investment and population growth to differ across sub-samples
– finding that the results on liberalisation are consistent with those reported in Table 11.
34
The F-statistic from the first stage regression is 329.13 (significant at the 1 percent level), with the coefficient
on the average age variable being above 6.0 in absolute value.
29
data used to construct this is taken from IIASA,35 and is available at five-year intervals. Table
11 reports both the linear and threshold results. Interestingly we tend to find that the
coefficient on investment becomes larger when we account for endogeneity. In the linear
model we find a coefficient on liberalisation that is positive and significant, with a value
similar to that reported in Table 1. In the threshold model we tend to find a pattern of
coefficients similar to earlier results, for 𝑂𝑈𝑇 and 𝐼𝑁𝐹 we find that the coefficient on
liberalisation is larger in the non-crisis regime, while for 𝐷𝐸𝐵𝑇 and 𝐶𝐴𝐷 the coefficients are
larger in the crisis regime. The exception is 𝑋𝑅 in which the coefficient on liberalisation is
now larger in the crisis regime, though the coefficients in the two regimes are very similar.
Table 11: Controlling for the Endogeneity of Investment

∆ln𝑦 Linear 𝑂𝑈𝑇 𝐼𝑁𝐹 𝑋𝑅 𝐷𝐸𝐵𝑇 𝐶𝐴𝐷
0.46 0.43 0.39 0.39 0.32 0.37
(3.17)*** (3.18)*** (2.62)*** (2.84)*** (1.49) (1.66)*
-0.48 -0.44 -0.37 -0.36 -0.26 -0.39
∆ln𝑃𝑂𝑃
(-1.61) (-1.56) (-1.32) (-1.30) (-0.77) (-0.99)
0.017 0.024 0.028 0.023 0.015 0.026
𝐿𝐼𝐵1
(3.93)*** (4.63)*** (5.36)*** (4.27)*** (2.80)*** (4.89)***
0.015 0.009 0.025 0.017 0.050
𝐿𝐼𝐵2
(2.22)*** (1.71)* (3.66)*** (3.01)*** (4.38)***
𝜆1 0.55 0.41 1.67 0.03 1.12
Observations 1961 1893 1857 1874 1497 1526
RMSE 0.0466 0.0462 0.0455 0.0453 0.0407 0.0451
Notes: See Table 3A. RMSE is the root mean squared error.
5.6 Liberalisation and openness: As noted at the outset, any attempt to delve into why
liberalisation at a time of crisis might have particular effects on an economy’s subsequent
growth, comes up against the twin barriers of the absence of an appropriate aggregate
measure of the strength of liberalisation and a lack of time series data from which to construct
one. But one variable for which data is readily available is “openness” (the ratio of imports
plus exports to GDP). Since its weaknesses as a measure of a country’s trade policy stance
are well known, we make no pretence to use it for this purpose here. But rightly or wrongly it
is commonly employed in this role, and so it is of interest to examine, as others have done,
whether trade liberalisations are reflected in increased openness measured in this way, and
whether any such relationship is affected by the presence of a crisis. To this end we estimate
𝑂𝑃𝐸𝑁𝑗𝑡 = 𝛼 + 𝛽𝐿𝐼𝐵𝑗 + 𝜐𝑗 +𝜂𝑡 + 𝜀𝑗𝑡 (5)
35
http://www.iiasa.ac.at/Research/POP/edu07/index.html
30
where 𝑂𝑃𝐸𝑁𝑗𝑡 is the ratio of imports plus exports to GDP in country 𝑗 at time 𝑡, and 𝜈𝑗 and 𝜂𝑡
are country and year dummies respectively36. We then impose the thresholds on 𝐿𝐼𝐵 based on
the crisis indicators as estimated above. The results are shown in Table 12A.
Table 12A: Liberalisation and Openness

0.046 0.023 0.053 0.006 0.05 0.053 0.025 0.072
𝐿𝐼𝐵1
(4.20)*** (1.76)* (3.29)*** (0.40) (3.96)*** (4.00)*** (1.89)* (4.85)***
0.087 0.037 0.055 -0.073 -0.041 0.129 -0.019
𝐿𝐼𝐵2
(4.51)*** (2.75)*** (4.20)*** (-2.63)*** (-1.68)* (4.47)*** (-1.07)
𝜆1 0.05 0.09 0.9 1.34 1.09 0.51 0.78

𝐿𝐼𝐵1 = 𝐿𝐼𝐵2 12.89*** 1.02 9.06*** 22.15*** 16.42*** 14.58*** 33.26***
Observations 2575 2450 2414 2352 1894 1964 1783 1783
F-Statistic 443.5*** 430.8*** 425.7*** 444.4*** 391.0*** 403.6*** 388.8*** 389.2***
R2 0.96 0.96 0.96 0.96 0.96 0.96 0.96 0.96
Notes: see Table 10
Several observations follow. A simple linear regression suggests that trade liberalisation
increases openness by 4.6 percentage points on average. But this masks a diversity of
outcomes once our crisis thresholds are taken into account37. A liberalisation during a 𝐷𝐸𝐵𝑇
or 𝐶𝐴𝐷 crisis results in reduced openness. Except for 𝐼𝑁𝐹, where the presence of a crisis has
no significant implications for the effects of liberalisation on openness, for all the other
indicators the relative strengths of the effects of liberalisation on openness in the crisis and
non-crisis regimes is exactly opposite to their relative effects on growth. The individual ratios
of imports and exports to GDP are investigated in Tables 12B and 12C, respectively. They
tell the same story. The linear regression indicates increases in both ratios on average, with
the larger increase in the import ratio. The negative effect on openness of liberalisation
during a 𝐷𝐸𝐵𝑇 or 𝐶𝐴𝐷 crisis seems to apply through imports and exports, as does the
stronger effects in 𝑂𝑈𝑇 and 𝑋𝑅 crises. On the basis of this evidence, one can but conclude
that this measure of openness is not the channel through which a crisis at the time of
liberalisation impacts on a countries subsequent growth performance.
36
Our sample is reduced to 73 countries, since two had no data on 𝑂𝑃𝐸𝑁.
37
This masks a quite diverse range of outcomes in any event. If we re-estimate equation (5) allowing the
coefficient β to differ across countries, we find (depending on the exact specification) positive coefficients in 36
cases (25 significant) and negative coefficients in 37 cases (13 significant). Wacziarg and Welch (2003) obtain a
similar balance of increases and decreases in openness following liberalisation in their sample.
31
Table 12B: Liberalisation and Imports
0.028 0.021 0.036 0.002 0.026 0.031 0.01 0.037
𝐿𝐼𝐵1
(4.56)*** (3.00)*** (4.14)*** (0.22) (3.39)*** (3.78)** (1.18) (4.13)***
0.05 0.027 0.038 -0.031 -0.011 0.063 -0.021
𝐿𝐼𝐵2
(4.82)*** (3.51)*** (5.31)*** (-1.91)* (-0.8) (3.95)*** (-1.99)**
𝜆1 0.05 0.09 0.9 1.34 1.09 0.51 0.78

𝐿𝐼𝐵1 = 𝐿𝐼𝐵2 8.53*** 1.12 17.01*** 15.22*** 11.62*** 12.79*** 34.76***
Observations 2575 2450 2414 2352 1894 1964 1783 1783
F-Statistic 383.18*** 384.7*** 381.0*** 404.5*** 341.9*** 337.8*** 316.9*** 319.8***
R2 0.95 0.95 0.95 0.95 0.95 0.95 0.95 0.95
Notes: see Table 15
Table 12C: Liberalisation and Exports

0.018 0.001 0.017 0.005 0.025 0.023 0.015 0.035
𝐿𝐼𝐵1
(3.11)*** (0.16) (1.99)** (0.50) (4.02)*** (3.57)*** (2.25)** (4.81)***
0.037 0.01 0.018 -0.043 -0.03 0.066 0.002
𝐿𝐼𝐵2
(3.76)*** (1.46) (2.50)** (-3.15)*** (-2.28)** (4.65)*** (0.24)
𝜆1 0.05 0.09 0.9 1.34 1.09 0.51 0.78

𝐿𝐼𝐵1 = 𝐿𝐼𝐵2 15.64*** 0.68 2.19 25.22*** 17.82*** 14.55*** 19.24***
Observations 2575 2450 2414 2352 1894 1964 1783 1783
F-Statistic 364.5*** 351.8*** 347.4*** 358.7*** 337.4*** 364.4*** 345.8*** 343.03***
R2 0.94 0.95 0.95 0.95 0.95 0.95 0.95 0.95
Notes: see Table 15
6. Conclusions
Our evidence supports earlier results that trade liberalisation increases economic growth in
the long-run. We also find evidence of significant crisis thresholds, at levels below those
normally assumed in the literature, for all our crisis indicators. While liberalisation leads to
higher long-run growth whether there is a crisis or not, the characteristics of the crisis appear
to influence the level of post-liberalisation growth. Liberalisation when output is declining,
inflation is increasing or the exchange rate is depreciating at above threshold levels relative to
its recent performance leads to lower subsequent growth. But if the debt to export ratio or the
current account deficit is increasing at above crisis levels at the time of liberalisation, then
growth will be higher than otherwise. Our composite indicators provided some, albeit
tentative support for the notion that an internal crisis tends to dampen the growth effects of
trade liberalisation, while an external crisis tends to amplify them.
The explicit allowance for trade liberalisations to have both short and long-run growth effects
did not materially affect our long-run conclusions. The same pattern of coefficients remained,
with post-liberalisation growth rates estimated to be a little higher if anything. The estimated
short-run coefficients generally supported the conclusion of a J-curve effect found in the
32
earlier literature. Our robustness checks showed that this was adjustment to the liberalisation
and not just recovery from a crisis. Compared to the post-liberalisation average, growth is
lower in the year of liberalisation, and higher three years later. These short-run effects were
also found to be crisis sensitive to some degree, exhibiting a similar pattern to the long-run
effects with respect to the individual crisis indicators. Output, inflation or exchange rate
crises at the time of liberalisation imply lower growth in the liberalisation year, but a stronger
recovery three years later. A current account crisis shows the opposite pattern.
So, is an economic crisis a good or a bad time for a country to undertake trade liberalisation?
Our results suggest that the answer depends on the nature of the crisis. Liberalisations at a
time of external (but not internal) crisis can bring additional growth benefits by alleviating
the constraints imposed by the crisis. But liberalisations at a time of internal crisis may
exacerbate adjustment problems and discourage the resource reallocations which are
necessary for trade liberalisation to be successful. Interestingly, our results suggest that these
crisis-related effects may extend beyond the short-run. Investigation of the precise channels
through which these effects take place awaits availability of the requisite data. But at this
stage it seems unlikely that an increased trade share is playing a role.
33
References
Alesina, A., Ardagna, S. and F. Trebbi, 2006. Who adjusts and when? On the political
economy of reforms. NBER Working Paper no. 12049, NBER.
Anderson, J. E. and J. Peter Neary, 2005 Measuring the Restrictiveness of International
Trade Policy. Cambridge, MA, MIT Press.
Baldwin, R. E., 2003. Openness and growth: What’s the empirical relationship? In
Challenges to Globalization: Analyzing the Economics, Robert E. Baldwin and L. Alan
Winters (eds) (Chicago, University of Chicago Press).
Barro, R. J. and J-W. Lee, 2001. International Data on Educational Attainment: Updates and
Implications. Oxford Economic Papers, 53, 541-563.
Ben-David, D., 1993. Equalizing exchange: Trade liberalisation and income convergence.
Quarterly Journal of Economics, 108, 653-679.
Bhagwati, Jagdish, 1978. Foreign Trade Regimes and Economic Development. Cambridge,
MA, Ballinger Press.
Bruno, M., 1996. Deep Crisis and Reform: What Have We Learned? Washington: DC, World
Bank.
Bruno, M. and W. Easterly, 1995. Inflation crises and long-run growth. NBER Working
Paper 5209, National Bureau of Economic Research.
Bruno, M. and W. Easterly, 1996. Inflation’s children: Tales of crises that beget reforms.
American Economic Review Papers and Proceedings, 86, 213-217.
Caner, M. and B.E. Hansen, 2004. Instrumental variable estimation of a threshold model.
Econometric Theory, 20, 813-843.
Campos, N., F. H. Cheng and J. B. Nugent. Crises, What crises? IZA Discussion Paper no.
2217, Institute for the Study of Labour, Bonn.
Chang, R., L. Kaltani and N. Loayza, 2009. Openness can be good for growth: The role of
policy complementarities. Journal of Development Economics, 90(1), 33-49.
Collier, P., 1993. Higgledy-piggledy liberalisation. The World Economy, 16, 503-12.
Cook, D., 2002. Education and growth: Instrumental variables estimates. Mimeo
Dollar, D., 1992. Outward-oriented developing economies really do grow more rapidly.
Economic Development and Cultural Change, 40, 523-544.
Dornbusch, R., 1992. The case for trade liberalisation in developing countries. Journal of
Economic Literature, 31, 1358-1393.
34
Drazen, A. and W. Easterly, 2001. Do crises induce reform? Simple empirical tests of
conventional wisdom. Economics and Politics, 13, 129-158.
Drazen, A. and V. Grilli, 1993. The benefit of crises for economic reforms. American
Economic Review, 83, 598-607.
Dreher, A., N. Gaston and P. Martens, 2008, Measuring Globalisation: Gauging its
Consequences. New York, Springer.
Easterly, W., 1996. When is stabilization expansionary? Evidence from high inflation.
Economic Policy, 11, 65-107.
Frankel, J. and D. Romer, 1999. Does trade cause growth? American Economic Review, 89,
379-399.
Greenaway, D., 1993. Liberalising foreign trade through rose tinted glasses. Economic
Journal, 103, 208-223.
Greenaway, D., 1998. Does trade liberalisation promote economic development? Scottish
Journal of Political Economy, 45, 491-511.
Greenaway, D. and R. Kneller, 2007. Firm Heterogeneity, Exporting and FDI. Economic
Journal, 117, F134-161.
Greenaway, D., S. Leybourne and D. Sapsford, 1997. Modelling growth (and liberalisation)
using smooth transitions analysis. Economic Inquiry, 35, 798-814.
Greenaway, D., C.W. Morgan and P. Wright, 1998. Trade reform, adjustment and growth:
What does the evidence tell us? Economic Journal, 108, 1547-1561.
Greenaway, D., C.W. Morgan and P. Wright, 2002. Trade liberalisation and growth in
developing countries. Journal of Development Economics, 67, 229-244.
Greenaway, D. and D. Sapsford, 1994. What does liberalisation do for exports and growth?
Weltwirtschaftliches Archiv, 130, 152-174.
Hansen, B. E., 1999. Threshold Effects in Non-Dynamic Panels: Estimation, Testing and
Inference. Journal of Econometrics, 93, 345-368.
Kline, P., 1994. An Easy Guide to Factor Analysis. London, Routledge.
Krueger, A. O., 1978. Foreign Trade Regimes and Economic Liberalisation. Lexington, MA,
Ballinger.
Krueger, A. O., 1997. Trade policy and economic development: How we learn. American
Economic Review, 87, 2-21.
Lee, J.W., 1995. Capital goods imports and long-run growth. Journal of Development
Economics, 48, 19-110.
35
Little, I, T. Scitovsky and M. Scott, 1970. Industry and Trade in Some Developing Countries:
A Comparative Study. Cambridge, Oxford University Press.
Lora, E., 1998. What makes reforms likely? Timing and sequencing of structural reforms in
Latin America. Working Paper no. 424, Inter-American Development Bank.
Maurer, R., 1998. Economic Growth and International Trade with Capital Goods: Theories
and Empirical Evidence. Tubingen, Kiel University.
Papageorgiou, D., M. Michaely and A. Choksi, 1991. Liberalising Foreign Trade. Oxford,
Basil Blackwell.
Pritchett, L., 2001. Where Has All the Education Gone? World Bank Economic Review, 15,
367-391.
Rodríguez, F. and D. Rodrik, 2000. Trade Policy and Economic Growth: A Skeptic's Guide
to the Cross-National Evidence. In Bernanke, B. and K. S. Rogoff (eds.) NBER
Macroeconomics Annual 2000, Camdridge: MA, MIT Press.
Rodrik, D., 1989. Credibility of trade reform. The World Economy, 1, 1-16.
Rodrik, D., 1989b. Promises, promises: Credible policy reform via signalling. Economic
Journal, 99, 756-772.
Rodrik, D., 1995. Getting interventions right: How South Korea and Taiwan grew rich.
Economic Policy, 20, 55-107.
Rodrik, D., 1996. Understanding economic policy reform. Journal of Economic Literature,
34, 9-41.
Rodrik, D., 1999. The New Global Economy and Developing Countries: Making Openness
Work. Washington, DC, Overseas Development Council.
Sachs, J. and A. Warner, 1995. Economic reform and the process of global integration.
Brookings Papers on Economic Activity, 1, 1-118.
Salinas, G. and A. Aksoy, 2006. Growth before and after trade liberalisation. World Bank
Policy Research Working Paper no. 4062, Washington, DC.
Singh, T., 2010. Does international trade cause economic growth?: A Survey. The World
Economy, 33, 1517-1564.
Temple, J., 1999. The new growth evidence. Journal of Economic Literature, 37, 112-156.
Tornell, A., 1998. Reform from within. Mimeo, UCLA.
Wacziarg, R, 2001. Measuring the dynamic gains from trade. World Bank Economic Review,
15, 393-429.
Wacziarg, Romain and K. Welch, 2008. Trade liberalisation and growth: New evidence.
World Bank Economic Review, 22, 187-231.
36
Appendix
Summary Statistics
Variable Observations Mean Standard Minimum Maximum
deviation
∆ln𝑦 2766 0.014 0.055 -0.593 0.221
𝑦60 2194 6.79 1.19 4.52 9.28
𝐼𝑁𝑉/𝐺𝐷𝑃 2620 0.21 0.08 -0.06 0.60
∆ln𝑃𝑂𝑃 2766 0.020 0.011 -0.028 0.060
∆ln𝑇𝑇𝐼 1364 -0.010 0.147 -1.844 1.986
𝑆𝑌𝑅60 2148 0.540 0.626 0.003 2.69
𝐶𝑅𝐼𝑆𝐿𝐼𝐵
-- 𝑂𝑈𝑇 2633 -0.210 0.829 -1.764 1.739
-- 𝐼𝑁𝐹 2597 0.160 0.872 -1.475 1.760
-- 𝑋𝑅 2500 1.04 0.744 -1.271 1.789
-- 𝐷𝐸𝐵𝑇 1967 -0.063 1.054 -1.773 1.735
-- 𝐶𝐴𝐷 2042 -0.2518 0.954 -1.735 1.586
Notes: While the mean and standard deviations of the crisis variables are zero and one respectively, there is no
reason to suppose that the mean of the variables at the time of liberalisation should be zero. Interestingly, for
three of the five crisis variables (per capita output growth, the ratio of debt to exports and the current account
balance) the mean of the crisis variable at liberalisation is negative, indicating that performance according to
these measures was better than average.
37
Table A1: Liberalising Countries and Episodes Undertaken in Crisis According to Table 3A
Country Lib ∆ln𝑦 at Lib 𝑂𝑈𝑇 𝐼𝑁𝐹 𝑋𝑅 𝐷𝐸𝐵𝑇 𝐶𝐴𝐷 𝐼𝑁𝑇 𝐸𝑋𝑇
Year (0.05) (0.09) (0.9) (1.34) (1.09) (0.51) (0.78)
Albania 1992 -0.07492 √ N/A N/A N/A N/A
Argentina 1991 0.10856 √ √
Armenia 1995 0.08343 √
Australia 1964 0.049161 √ N/A N/A N/A N/A
Azerbaijan 1995 -0.13708 √ N/A N/A N/A N/A
Bangladesh 1996 0.02772 √ √ √ √ √
Barbados 1966 0.035858 √ √ N/A N/A N/A N/A
Benin 1990 -0.00014 √
Bolivia 1995 -0.03852 √ √ √ √ √
Botswana 1979 0.079099 N/A N/A N/A N/A N/A √
Brazil 1991 -0.00325 √
Bulgaria 1991 -0.07832 √ √ √ N/A N/A N/A
Burkina Faso 1998 -0.01427 √ √ N/A N/A N/A
Burundi 1999 -0.02942 √
Cameroon 1993 -0.06053 √ √ √
Cape Verde 1991 -0.00544 √ √
Chile 1976 0.018003 √ N/A N/A N/A
Colombia 1986 0.036367 √ √
Costa Rica 1986 0.024531 √
Cote d’Ivoire 1994 -0.02364 √ √ √
Dominican Republic 1992 0.060575 √ √ √
Ecuador 1991 0.028394 √ √
Egypt 1995 0.02622 √ √
El Salvador 1991 -0.00613 √ √ √ √
Ethiopia 1996 0.073819
Gambia 1985 -0.0449 √ √ √ √
Georgia 1996 0.109529 N/A √ N/A N/A N/A
Ghana 1985 0.011366 √
Guatemala 1988 0.013508 √ √
Guinea-Bissau 1987 -0.00348 √ √
Guyana 1988 -0.03144 √ √ √
Honduras 1991 0.001985 √ √ √ √
Hungary 1990 -0.03241 √ √ √ √ √
Indonesia 1970 0.054622 N/A N/A N/A N/A
Ireland 1966 0.006114 √ N/A N/A N/A N/A
Israel 1985 0.016261 √ √ N/A N/A N/A
Jamaica 1962 -0.00324 N/A N/A N/A N/A N/A N/A N/A
Jamaica 1989 0.060218 √ √
Japan 1964 0.100021 N/A N/A N/A N/A
Kenya 1963 0.052643 N/A N/A N/A N/A N/A N/A N/A
Kenya 1993 -0.02314 √ √
Republic of Korea 1968 0.08828 N/A N/A N/A N/A
Kyrgyz Republic 1994 -0.22355 √ N/A N/A N/A N/A
Latvia 1993 -0.03348 N/A N/A N/A N/A
Lithuania 1993 -0.17245 √ N/A N/A N/A N/A N/A
Macedonia 1994 -0.02495 N/A N/A N/A N/A N/A
Madagascar 1996 -0.00965 √
Mali 1988 -0.01237
Mauritania 1995 0.023755 √
Mexico 1986 -0.05861 √ √ √ √
Moldova 1994 -0.36962 √ N/A N/A N/A N/A N/A
Morocco 1984 0.020437 √
Mozambique 1995 0.016492 √ √
Nepal 1991 0.040674 √ √ √
New Zealand 1986 0.019049 √ √ N/A N/A N/A
Nicaragua 1991 -0.03131 √ √
Niger 1994 0.00416 √ √ √
Pakistan 2001 -0.00569 √ √ √
Panama 1996 0.011507 √ √ √
Paraguay 1989 0.028816 √ √
Peru 1991 0.001776 N/A √ N/A N/A
Philippines 1988 0.041637
Poland 1992 N/A N/A √ N/A N/A N/A
Romania 1992 -0.07533 √ √ √
Sierra Leone 2001 0.024827
Singapore 1965 0.086483 N/A N/A N/A N/A
Slovak Republic 1991 -0.15752 √ √ N/A N/A N/A N/A N/A
South Africa 1991 -0.03085 √ √ √ N/A N/A N/A
38
Sri Lanka 1977 0.033611 √ √
Sri Lanka 1991 0.033908 √ √ √ √
Tajikistan 1996 -0.19837 √ N/A N/A N/A
Tanzania 1995 0.005887 √
Trinidad and Tobago 1992 -0.02491 √
Tunisia 1989 0.004536 √ √
Turkey 1989 -0.01919 √ √ √ √
Uganda 1988 0.044017 √ √ √ √
Venezuela 1989 -0.11348 √ √ √ √
Venezuela 1996 -0.02259 √ √ √ √
Zambia 1993 0.039416 √ N/A N/A N/A
Notes: A √ indicates that the country would have been classed as in a crisis according to the relevant crisis
indicator and the results of Table 3A, N/A indicates that the country was not included in the regression, usually
due to a lack of data on the crisis variable.
39

Trade Liberalisation, Economic Crises and Growth: Epublications@Bond

Uploaded by

Document Information

Original Title

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

Trade Liberalisation, Economic Crises and Growth: Epublications@Bond

Uploaded by

Copyright:

Available Formats

Bond University

Trade liberalisation, economic crises and growth

Follow this and additional works at: http://epublications.bond.edu.au/gdc

Trade Liberalisation, Economic Crises and Growth

No. 44, November 2010

JEL Classification: F14, O40

2. Background: Trade Liberalisation and Growth

These considerations combine to suggest that a trade liberalisation undertaken at a time of

3. Data, Methodology and Initial Results

Figure 1: Average Growth Before and After Liberalisation

Table 1: Initial Results

Observations 952 952 1327 1327 2619 2619

Factor analysis is a method of condensing a number of random variables into a smaller

Table 2: Rotated Factor Loadings

For each crisis indicator we calculate a standardised score as:

+𝛿2 𝐿𝐼𝐵𝑖,𝑡 𝐼 𝐶𝑅𝐼𝑆𝑗𝑖𝐿𝐼𝐵 > 𝜆𝑗 + 𝜐𝑖 + 𝜂𝑡 + 𝜀𝑖,𝑡 (2)

Table 3A: Endogenous Threshold Results

Table 3B: Endogenous Threshold Results

4. Short-Run Impacts of Liberalisation on Growth

Table 4A: Endogenous Threshold Results (Long-run threshold only)

-0.022 -0.027 -0.029 -0.016 -0.009 -0.019 -0.010 -0.01

𝜆1 0.05 0.09 0.9 1.34 1.09 0.51 0.78

0.002 -0.005 -0.007 0.002 -0.006 0.005 0.0007

0.001 -0.005 -0.011 -0.003 0.004 -0.004 0.0003

0.013 0.012 0.0007 0.012 0.014 0.010 0.012

𝜆𝐿 𝑒𝑥𝑜𝑔 0.05 0.09 0.9 1.34 1.09 0.51 0.78

5. Robustness and Extensions

5.2 Excluding transition economies: Transition countries undertook trade liberalisation in

Table 7A: Thresholds on the TTI sub-sample

Table 7B: Inclusion of TTI Index

Observations 2454 2405 2372 1599 1675

Table 8B: Threshold Results when including the Crisis variables

Table 9A: Linear Results with Crisis Variables Included

Observations 2436 2436 2436 2436 2436

𝜆1 0.17 0.15 0.70 1.34 1.06

-0.028 -0.026 -0.018 -0.010 -0.021

𝜆1 0.17 0.16 0.69 1.34 1.11

𝜆1 0.05 0.09 0.9 1.34 1.09 0.51 0.78

Table 11: Controlling for the Endogeneity of Investment

Table 12A: Liberalisation and Openness

𝜆1 0.05 0.09 0.9 1.34 1.09 0.51 0.78

𝜆1 0.05 0.09 0.9 1.34 1.09 0.51 0.78

Table 12C: Liberalisation and Exports

𝜆1 0.05 0.09 0.9 1.34 1.09 0.51 0.78

You might also like