Multilevel Governance and Metropolitan Regionalism in The USA - URBAN STUDIES

Urban Studies, Vol. 37, No.
56, 851876, 2000
Multilevel Governance and Metropolitan

Regionalism in the USA
Clyde Mitchell-Weaver, David Miller and Ronald Deal Jr
[Paper received in nal form, January 2000]
1. Introduction
Metropolitan regionalism may once again be
on the American political agenda, after a
hiatus of a quarter of a century. Since the
mid 1990s, a burgeoning ow of popular and
academic books and articles, as well as reports from leading liberal US think-tanks, 1
have focused public attention on the problems of big cities and their surrounding regions (Rusk, 1993, 1999; Pierce, 1993;
Downs, 1994; Wallis, 1994a, 1994b, 1994c;
Cisneros, 1995; Walker, 1995; Foster, 1997;
Or eld, 1997; Barnes and Ledebur, 1998;
Lindstrom, 1998). Notably, at the turn of the
new millennium, the Brookings Institution in
Washington, DCfounded during Franklin
Roosevelts New Dealthe Ford Foundation and the Democratic Clinton administration have sounded a clarion call to alert
civic leaders to the growing crisis of the
metropolis (Brookings Institution, 1998;
Ford Foundation, 1999; US HUD, 1999).
The immediate causes of such high-level
concern are threefold. First, socioeconomic
and scal disparities between metropolitan
centres and their outlying settlement clusters
have reached a critical level in the US, and
current domestic demographic trends portend
an ever-worsening gulf in terms of economic
resources. Secondly, sharp competition
within the global economy increasingly
threatens the economic base of US core cities

and their inner-ring suburbs. New productive
investments and industrial growth are predominantly in the outer suburbs and edgecities. And thirdly, urban sprawl
uncontrolled land development and leapfroggingis visibly threatening the sustainability of the physical environment of large
urban communities (Weitz, 1999; Berke and
Conroy, 2000). Twenty- ve years of benign
neglect, in terms of both urban and social
policy, have exacted a very real historical
cost on metropolitan America. The plight of
the cities could be a prime concern in the US
general elections in November 2000.
In major urban areas across the US, an
emerging Regional Coalition is forming
around city-centred and environmental interest-groups (Rothblatt and Sancton, 1998;
Phares, 1999). As we discuss in this paper,
the Coalition argues that metropolitan
governmental fragmentation is the primary
cause of US urban problems, and that some
form of regional governance is the necessary
rst step towards a solution. Furthermore, the
hyper-complex nature of US federalism requires multilevel intervention, using state
and federal powers to reinforce local moves
in the direction of regional co-operation and
consolidation. Intergovernmental strategies
Clyde Mitchell-Weaver and David Miller are in the Graduate School of Public and International Affairs, University of Pittsburgh,
3N28 Forbes Quadrangle, Pittsburgh, PA, USA. Fax: 001 412 648 2605. E-mail: mithweav 1 @pitt.edu (Clyde Mitchell-Weaver);
redsox 1 @pitt.edu (David Miller). Ronald Deal Jr is with the law rm of Kirksey and McNamee, PLC, Brentwood, TN, USA. E-mail:
RDeal@Kirkmac.com.
Downloaded from usj.sagepub.com at UNIV FEDERAL DO PARA on March 16, 2015
0042-0980 Print/1360-063X On-line/00/0560851-26
2000 The Editors of Urban Studies
852
CLYDE MITCHELL-WEAVER ET AL.
are essential to promote metropolitan revenue-sharing, smart growth, the new urbanism and the targeting of skills training,
housing and transport opportunities to match
the changing intraurban location of employment expansion and job needs (Katz, 1994;
Or eld, 1998; Brennan and Hill, 1999; Immergluck, 1999; Katz and Allen, 1999;
American Planning Association, 2000).
In the sections which follow, we present
an overview of the Regional Coalitions
analysis of US metropolitan problems and
their agenda for public action. It is important,
in order to set the context, to begin with an
historical sketch of the role played by metropolitan regionalism in the evolution of US
urban policy.
2. A Note on the History of US Metropolitan Regionalism
Metropolitan regionalism was the rst approach to urban problems in the US, beginning in the early 19th century. Even before
the growth of second industrial revolution
cities, consolidation of city and county governments was undertaken in commercial centres such as New Orleans (1805), Boston
(1821), Nantucket, MA (1821), Baltimore
(1851), Philadelphia (1854), San Francisco
(1856) and St Louis (1876). By the turn of
the 20th century, the modern City of New
York had been created by the merger of New
York, Brooklyn, Queens and Richmond
counties in 1898, and Denver (1904) and
Honolulu (1907) had been added to the list.
Other 19th-century cities became the sole
unit of government and the general metropolitan service provider through annexation
(for example, Boston, Chicago, Detroit and
Pittsburgh) (Stephens and Wikstrom, 2000,
pp. 2934; Herson and Bolland, 1998,
p. 250; Ross and Levine, 1996, p. 324). 2
Private civic organisationswhat today
would be called publicprivate partnershipslike the Chicago Commercial Club
and the National Municipal League, supported metropolitanisation through their regional leadership activities and publication of
the National Municipal Review (later the
National Civic Review). The whole August

1922 issue of the Review was devoted to
Chester Maxeys early synthesis, The political integration of metropolitan communities, which identi ed
the fundamental problem of the metropolis [as] the decentralized or fractionated nature of local government (Stephens
and Wikstrom, 2000, p. 35).
This was followed by another League publication, Paul Studenskis comprehensive
book, The Government of Metropolitan
Areas in the US (1930).
With the onset of the Great Depression,
the US federal government rst became involved with metropolitan problems, inspired
by a Chicago School of local government
reform. Major statements of the Chicago
Schools agenda appeared over the next decade: Roderick D. McKenzie (1933) The
Metropolitan Community; Charles E. Merriam, Spencer D. Parratt and Albert Lepawsky (1933) The Government of the
Metropolitan Region of Chicago; William
Anderson (1934) The Units of Government in
the US; the National Resources Committee
(1937) Our Cities: Their Role in the National
Economy; and Victor Jones great classic
Metropolitan Government (1942), written as
a doctoral dissertation under the direction of
Charles E. Merriam during the late 1930s.
The most important contribution of Chicago
writers, in our view, was to link metropolitan
regional governance with the changing economic and social structure of cities and their
role in the national economy. 3
At the turn of the 20th century, Chicago
was also the rst home of metropolitan planning in the US, which grew up alongside the
city planning movement. Many of the earliest
urban planners were forceful proponents of a
regional perspective, and they soon came to
include entire metropolitan areas within the
scope of their plans. US metropolitan planning was a reaction to the second industrial
revolution, and was composed of four separate but overlapping elements.
Housing reform was the earliest strand.
Investigative commissions (1856, 1884,
METROPOLITAN REGIONALISM IN THE USA
1894) and tenement house laws (1867, 1879,

1895, 1901) in New York led up to the First
National Conference on City Planning, held
in Washington, DC, in 1909, where housing
reform was still a very controversial issue.
Park and boulevard planning and the City
Beautiful movementthe next two elements of metropolitan planningdeveloped
together. Frederick Law Olmsted began the
US tradition of metropolitan park-building
with Central Park in New York City. The
purpose of the Chicago Worlds Fair (1893)
was to show that Americas cities, too, could
aspire to be beautiful. With a layout designed
by Olmsted, it started Daniel H. Burnham on
the road to preparing one of the earliest
metropolitan regional plans, the Plan of
Chicago (Burnham and Bennett, 1909).
In 1904, Burnham was approached by the
Commercial Club of Chicago and asked to
prepare a plan for the city, expanding on his
original scheme for lakefront redevelopment
with a system of regional ring-roads, radial
highways and parkways. After some delay,
work began on the plan in 1907, and two
years later the grand opus emerged as a
limited edition of magni cently illustrated
volumes, selling for $25 each. The plan was
genuinely metropolitan in scope, and by
1925 Chicago had spent some $300 million
to implement it.
The high-water mark of pre-World War II
metropolitan planning in the US was reached
with publication of the Regional Plan for
New York and Its Environs (Committee on
Plan of New York, 192931). The New York
project was started in 1920 by Charles Dyer
Norton, the same insurance executive who
had fathered the Chicago venture. The staffs
director was Thomas Adams. Of the 10 volumes eventually published for the Committee, the rst 2, The Graphic Regional
Plan (1929) and The Building of the City
(1931), contained the bulk of their recommendations. There is little doubt that this
series of impressive volumes represented the
most imposing metropolitan planning effort
ever attempted. Its real departures from the
earlier Chicago Plan, however, lay mainly in
its extensive use of social statistics and its
853
emphasis on newly legitimised land-use controls. (The US Supreme Court upheld the
constitutionality of zoning in the case of
Village of Euclid, Ohio, v. Ambler Realty Co.
(1926).) The principal objective of the plan
remained much the same: to promote the
continued expansion of the metropolis by
developing an ever-greater land area, laced
together by a network of highways.
Government reform was the fourth and
last theme of metropolitan planning. It had
two main goals: professionalising local
government and, as we have already seen,
expanding the geographical boundaries of the
city in order to re ect the new realities of
metropolitan growth. It is here, with the subject of metropolitan budgeting and the territorial expansion of municipal jurisdiction,
that the work of Charles E. Merriam and his
Chicago colleagues discussed above comes
into the picture. Their views on metropolitan
government reform were summarised in
Merriams preface to The Government of the
Metropolitan Region of Chicago, cited earlier:
[This book focuses on]: (1) consideration
of the governmental possibilities of the
Region as a whole; (2) emphasis on the
actual functioning of public agencies
within the Area (3) emphasis on the
principle of interlocking directorates as a
means of obtaining consolidation; (4) attention to the importance of interstate
agreements as a basis of regional organisation; (5) discussion of the possibilities of
independent statehood as a means of
metropolitan development; [and] (6) development of a system of central scal
control over local governments without establishment of a new unit of government
(Merriam et al., 1933, preface).
These themes were to be rediscovered during the 1950s1970s surge of urbanindustrial growth in the US, and again recently in the second half of the 1990s. During this 50-year period, the emphasis of
metropolitan planning changed from guiding
new growth to limiting the geographical expansion of the metropolis.
854
From the late 1940s through the early

1970s, metropolitan regionalism remained a
favoured remedy in the US for dealing with
urban development and the increasingly evident social disparities re ected in the geography of the metropolis. The new Advisory
Commission on Intergovenmental Relations
(ACIR), founded under federal law in 1959,
focused once again on restructuring metropolitan government (Advisory Commission
on Intergovernmental Relations, 1961, 1962,
1966). In his majestic tour dhorizon, Luther
Halsey Gulick (1962), The Metropolitan
Problem and American Ideas, re ected the
spirit of John Kennedys New Frontier.
Gulicks main points, almost paraphrasing
Merriam 30 years earlier (quoted just above),
were that:
(1) our focus on the core city of metropolitan areas must be replaced by a broader
view of socioeconomic and governmental development across the metropolis;
(2) all levels of US governmentespecially
the statesmust be brought to bear on
emerging urban problems; and
(3) some form of metropolitan federalism
was probably the appropriate intergovernmental approach.
The second half of the 1960s brought the
renewed social activism of the civil rights
movement and urban riots in major cities, as
well as New Deal-like expansion of federal
government programmes, as re ected in Bollens and Schmadt (1965) The Metropolis,
and H. Wentworth Eldredges two-volume
1967 collection, Taming Megalopolis. But
with the exception of metropolitan-level coordination agencies, such as Councils of
Governments (COGs) (1965), Metropolitan
Planning Organisations (1966) and A-95
Clearing Houses (1969), the metropolitan
agenda remained very similar to that outlined
by Jones in 1942 (see Committee on Economic Development, 1970; Advisory Commission on Intergovernmental Relations,
197374). Citycounty consolidationsas in
the 19th centuryand limited regional coordination and special service districts were
typical of the period. By the time ACIRs

series on Substate Regionalism appeared in
the early 1970s, the Nixon administration
had taken steps to defund Kennedy/
Johnson-era federal initiatives and, with
them, further moves towards metropolitan
governance. The next 25 years, until the late
1990s, with few exceptions, marked a retreat
from intergovernmental approaches to local
problems. The Reagan Revolution of the
1980s was especially important for this.
Executive Order 12372 ended the A-95 review process and, between 1980 and 1990,
COGs lost federal funding and fell in number
from 670 to 435, or by 35 per cent (Ross and
Levine, 1996, p. 357). ACIR was closed in
1996, and only in the second term of the
Clinton administration has the call for metropolitan change truly become focused.
3. The Major Problem: Governmental
Fragmentation
The doctrine of metropolitan regionalism has
insisted for 70 years that governmental fragmentation is the major source of US urban
problems. Ross and Levine (1996, pp. 310
313), arguing more from conviction than evidence, present the strong program against
fragmentation:
Local autonomy has produced a system of
metropolitan fragmentation whereby the
metropolitan area is divided into many
smaller jurisdictions with no government
possessing the power to look out for the
good of the entire region. No local jurisdiction is required to look at the effects of its
actions on other jurisdictions. Few suburbs
are willing to alter land use, housing, and
school arrangements when such alterations
impose new costs on existing residents.
The consequences of suburban autonomy
and metropolitan fragmentation are numerous. They can brie y be summarized
as follows:
1. Racial imbalance in the metropolis
2. Income and resource imbalance in the
metropolis
855
3. The protection of privilege

4. Increased business power
5. The impact of suburbs on central cities [the] exploitation hypothesis
6. Problems of housing affordability and
homelessness
7. The lack of rational land use planning
and commitment to environmental
values
8. Problems in service provision
citizens denied the advantages of economies of scale (emphasis in original)
Governmental fragmentation is certainly pronounced in urban America, and settlement
patterns are becoming ever-more decentralised. Table 1 was constructed by Stephens
and Wikstrom (2000, p. 8) from Census of
Governments information (taken every ve
years by the US Bureau of the Census, beginning in 1932). The idea of the government
census itself was part of the regionalist
movement.
For the latest reporting year, 1997, there
were more than 87 000 units of local government in the US; all of them devolved legal
creations of the individual states (Judge Dil-
lons Rule; see, City of Clinton, Missouri, v.

Cedar Rapids and Missouri River Railroad
(1868)). In terms of the most basic form of
general-purpose government, counties have
remained steady at a count of just over 3000
for the last seven decades.4 The count of
municipalities (urban places) has gone up
from 16 400 to 19 400, or by 18 per cent; not
considering, of course, the geographical or
population size of the places concerned.
Non-urban towns and townships have fallen
by about the same proportion. This decadeslong pattern of rural decline and urban
growth is re ected more clearly by the abandonment of rural school districts at an amazing overall rate of minus 94 per cent; as well
as the growth of urban-service-type special
districts by 240 per cent.
This last statistic is indicative of the mis t
between the increasing number (and implicitly, the size) of municipalities and the still
larger geographical area absorbed by urban
growth. Looked at from another angle, local
government complexity fell by 57 per cent
with ruralurban change from 1932 to 1972,
and then increased once again by 12 per cent
with continued urbanisation through 1997.
Table 1. Local governments in the US, 193297

Year
1932
(1937)
1942
(1947)
1952
1957
1962
1967
1972
1977
1982
1987
1992
1997
Counties
Municipalities
Towns and
townships
School
districts
Special
districts
Totals
3 062
3 053
3 050
3 049
3 050
3 050
3 043
3 049
3 044
3 042
3 041
3 042
3 043
3 043
16 442
16 332
16 220
16 360
16 778
17 215
17 997
18 048
18 517
18 862
19 076
19 200
19 279
19 372
19 978
19 183
18 919
18 051
17 202
17 198
17 144
17 105
16 991
19 822
16 734
16 691
16 656
16 629
128 548
113 571
108 579
95 521
56 346
50 454
34 678
21 742
15 781
15 174
14 851
14 721
14 422
13 726
14 572
9 867
8 299
9 302
12 319
14 424
18 823
21 264
23 885
25 962
28 078
29 532
31 555
34 683
182 602
162 006
115 067
142 283
105 684
102 341
91 685
81 248
78 218
79 862
81 780
83 186
84 955
87 453
Notes: Various census reports for later years sometimes differ slightly from the date reported in
earlier census reports as to the exact number of local governments for a given Census of Governments
year. The years shown in parentheses (1937 and 1947) are estimates.
Source: Stephens and Wikstrom (2000, p. 8).
Data Sources: 1957 through 1992 Census of Governments, Vol.1 and 1997 fax from Governments
Division of the Census Bureau, May 1998; Graves, (1964, p. 699).
856
This grand total understates settlementpattern change though, because it overlooks

the broad nationwide distribution of declining rural school districts and the comparatively tightly knit pattern of metropoliscentred special district growth.
To calculate changing levels of US metropolitan fragmentation during the most recent
reporting periods for which detailed statistics
are available, we developed our own metropolitan fragmentation index (MFI) for the
census years 1972 and 1992 (see the Appendix for details). Over 300 Metropolitan
Statistical Areas (MSAs) were broken down
into: population-size groupings; and, geographical regions (Northeast, Midwest,
South and West). These categories were
chosen because it is frequently assumed
that large, older MSAs in the Northeast
and Midwest will be more governmentally
fragmented than smaller, newer metropolitan
areas from the South and West (see, for
example, Stephens and Wikstrom, 2000,
pp. 328).
In 1972, 311 metropolitan regions of the

US had an average MFI of 3.8 (from a
theoretical range of 1 to in nity) (see Table
2). The distribution was skewed, in that there
were a few regions with very high scores, as
demonstrated by the median MFI of 3.4.
Heading the list was Philadelphia with an
index of 14.3. Rounding out the top seven
most-fragmented regions were St Louis
(12.3), Boston (11.2), Pittsburgh (10.7),
ScrantonWilkes Barre, PA (9.3), MinneapolisSt Paul (8.5) and Chicago (8.3). At
the other end of the scale was Midland, TX,
with a score of 1.3. The top seven most-centralised MSAs also included Owensboro, KY
(1.4), San Angelo, TX (1.4), Jackson, TN
(1.5), Odessa, TX (1.5), Las Cruces, NM
(1.5) and Tucson, AZ (1.6).
In terms of both size and geographical
location, MFI scores tended to follow the
expected pattern. The scale was statistically
sensitive to population size. The upper part
of Table 2 presents the 1972 MFIs by MSA
population-size groupings. As population in-
Table 2. Metropolitan fragmentation index (MFI), 1972

Count
Median
Mean
25th
percentile
75th
percentile
By MSA size
Large (over 2 million)
Medium-large (12 millions)
Medium (500 000 to 1 million)
Medium-small (250 000 to 500 000)
Small (under 250 000)
24
32
41
73
141
5.65
5.03
4.09
3.46
2.79
6.64
4.99
4.23
3.69
3.04
4.63
3.43
3.36
2.92
2.17
8.23
6.54
4.89
4.45
3.62
Total
311
3.40
3.83
2.58
4.63
By region
Northeast
Midwest
South
West
46
87
122
56
5.38
4.14
2.83
3.10
5.85
4.25
2.98
3.37
4.22
2.89
2.13
2.66
7.29
4.89
3.55
4.17
Total
311
3.40
3.83
2.58
4.63
Population size
F 5 87.52
sig. 000
F 5 43.61
sig. 000
Source: Calculated by the authors from data in US Bureau of the Census (1972).
857
Table 3. Metropolitan fragmentation index (MFI), 1972 (by both MSA size and region)
Population size
Northeast
Midwest
South
West
Total

Medium-small (250 000 to 500 000)
8.12
6.68
6.40
4.96
4.59
8.66
6.30
4.48
4.30
3.53
4.87
3.36
3.69
3.14
2.38
5.00
5.19
3.01
3.10
2.74
6.67
5.38
4.40
3.88
3.31
Total
5.85
4.25
2.98
3.37
3.83
creased, so did the index. Density and higher

populations were related to complexity, as
evidenced by the higher MFIs. A statistically
signi cant relationship was also found between MFI scores and geographical regions,
shown in the lower portion of Table 2. The
Northeast, with an average MFI of 5.9, was
indeed more fragmented than the other regions. Conversely, the South, as generally
assumed, was more centralised than the other
regions.
Combining population size and geographical region by average MFIs for 1972 produces the data presented in Table 3. Several
anomalies should be noted. First, although
the Northeast generally had higher scores
than other regions, large metropolitan areas
in the Midwest were more fragmented than
their Northeast counterparts. Secondly, although the South was generally more centralised than other regions, medium and
medium-small sized MSAs in the West
tended to be more centralised than their
southern counterparts. Finally, in all geographical regions except the West, large
MSAs were substantially more fragmented
than smaller metropolitan areas within the
same region.
Twenty years later, in 1992, the mean MFI
for the same 311 MSAs had increased from
3.8 to 4.2an 8.6 per cent rise in the index
(see Table 4). More importantly, 248 MSAs
or 80 per cent had an increase in their MFIs,
signifying greater fragmentation. An immediate explanatory hypothesis would be to
assume that this additional fragmentation
should follow population change: as popu-
lation increases, so should the MFI. However, the correlation between population
change and changes in the fragmentation index is statistically insigni cant. Two other
factors contributed more to a higher score
than did population. The rst was an increase
in the absolute number of governments per
MSA, and the second was the fact that suburban governmentswhich experienced the
bulk of population growth in metropolitan
areasplayed a greater nancial role in the
delivery of public services than they had 20
years earlier.
Philadelphia continued to have the highest
score on the metropolitan fragmentation index at 15.4. St Louis and Boston retained the
second and third positions, respectively.
Chicago jumped from seventh to fourth with
a 46.1 per cent increase in the index, from
8.3 to 12.1. Pittsburgh, ScrantonWilkes
Barre, PA, and MinneapolisSt Paul completed the top seven most-fragmented MSAs
in 1992. The greatest absolute change in the
MFI occurred in Chicago, 3.8 points. Houston and St Louis were next with a 2.1 increase in their scores. They were followed by
Lake County, IL (1.9) and Joliet, IL (1.9).
Chicagos 46 per cent increase in the MFI
made it the most fragmenting MSA during
the 197292 period. Five other metropolitan
areas also had an increase in their scores
greater than 40 per cent: Houston (44 per
cent); Galveston, TX (44 per cent);
Tuscaloosa, AL (42 per cent); Greeley, CO
(42 per cent); and Midland, TX (41 per cent).
This all suggests that MSAs in the Midwest,
South and West might have been decentralis-
858
Table 4. Metropolitan fragmentation index (MFI), 1992

Count
Median
Mean
25th
percentile
75th
percentile
By MSA size
Medium-small (250 000 to 500 000)
24
32
41
73
141
6.73
5.41
4.34
3.68
3.04
7.59
5.39
4.55
3.97
3.29
4.87
3.56
3.37
3.03
2.30
9.30
7.49
5.14
4.52
4.01
Total
311
3.67
4.16
2.75
4.97
By region
Northeast
Midwest
South
West
46
87
122
56
5.78
4.32
3.04
3.36
6.39
4.62
3.22
3.69
4.63
2.95
2.30
2.76
7.84
5.13
3.90
4.13
Total
311
3.67
4.16
2.75
4.97
Population size
F 5 87.52
sig. 000
F 5 43.61
sig. 000
Source: Calculated by the authors from data in US Bureau of the Census (1992)
ing more rapidly than the nation as a whole,

or than the older, more fragmented northeastern region (Table 5).
Changes in the MFI between 1972 and
1992 by population-size group and geographical region are presented in Tables 6 and 7.
Two notable trends emerged. First, overall
population size was not a statistically
signi cant factor. Indeed, all population
groups were fragmenting at approximately
the same rate. A higher rate of increase for
large metropolitan areas, 14.3 per cent, was,
however, suggestive that areas which already
had a higher MFI were apt to fragment faster

than MSAs that were not already as fragmented as others. This observation was supported by index changes in the 75th
percentile MSAs in medium-large and small
metropolitan areas. These MSAs, relative to
their population groups, had comparably
higher MFIs. They also experienced more
accelerated growth in their scores: 14.5 per
cent and 10.8 per cent, respectively.
Secondly, geographical region was also
not signi cantly associated with changes in
the fragmentation index. However, although
Table 5. Metropolitan fragmentation index (MFI), 1992 (by both MSA size and region)
Population size
Northeast
Midwest
South
West
Total

Medium-small (250 000 to 500 000)
8.93
7.04
7.22
5.40
4.99
10.36
6.96
4.77
4.67
3.75
5.61
3.56
3.92
3.31
2.61
5.71
5.85
3.11
3.35
2.99
7.65
5.85
4.76
4.19
3.59
Total
6.39
4.62
3.22
3.69
4.16
859
Table 6. Metropolitan fragmentation index, 197292 (percentage change)

Population size
Count
Mean
25th percentile
75th percentile
By MSA size
Medium-small (250 000 to 500 000)
24
32
41
73
141
14.2
8.2
7.6
8.1
8.6
5.2
3.8
0.3
3.8
6.0
13.0
14.5
5.1
1.6
10.8
Total
311
8.7
6.6
7.3
By region
Northeast
Midwest
South
West
46
87
122
56
9.0
7.4
8.9
10.3
9.7
2.1
8.0
7.8
7.5
4.9
9.9
2 1.0
Total
311
8.7
6.6
7.3
F 5 0.15
sig. ns
F 5 1.299
sig. ns
Source: Calculated by the authors from the data available in Tables 25.
Table 7. Metropolitan fragmentation index, 197292 (percentage change)
Population size
Medium-small (250 000 to 500 000)
Total
Northeast
Midwest
South
West
Total
10.4
5.5
11.1
9.1
9.2
19.6
10.4
7.0
8.3
5.6
14.8
7.4
6.8
6.9
10.2
13.8
11.6
6.8
9.8
10.5
14.2
8.2
7.6
8.1
8.6
9.0
7.4
8.9
10.3
8.7
F 5 0.549
sig. ns
Source: Calculated by the authors from the data available in Tables 25.
location was statistically insigni cant, the

possible implications of this fact were far
from insigni cant: metropolitan America
was becoming more governmentally fragmented regardless of geographical location.
As can be seen in the lower section of Table
6, all regions experienced fragmentation.
To interpret further these changes, we assigned MSAs into the six groupings for 1972
shown in Table 8. Twelve per cent were
classi ed as centralised and 4.8 per cent as
super decentralised. Of the centralised
MSAs, 36 of 37 were located in the South or

West. Conversely, all 15 of the super decentralised metropolitan areas were in the
Northeast and Midwest.
Another grouping was made to capture the
trend in fragmentation between 1972 and
1992. Five groups were created based on the
rate of change in MFIs, as seen in Table 9.5
As mentioned earlier, 80 per cent of metropolitan areas fell into a decentralising category. Conversely, only 63 MSAs (out of 311)
were centralising. In the Northeast and West
860
Table 8. Fragmentation group membership by region, 1972

Group 1972
Highly centralised
Moderately centralised
Slightly decentralised
Moderately decentralised
Highly decentralised
Super decentralised
Score
Northeast
Midwest
South
West
Total
1.002.00
2.003.00
3.004.00
4.005.00
5.007.50
7.50 1
2
8
10
15
11
1
21
19
26
16
4
26
42
33
17
4
10
15
17
5
9
37
80
77
58
44
15
46
87
122
56
311
Total
Source: Calculated by the authors.
Table 9. Fragmentation trend group membership by region, 197292

197292 trend group
Percentage change
Northeast
Midwest
South
West
Total
Centralising
Slow centralising
Slow decentralising
Rapid decentralising
Hyper decentralising
2 5 or more
05
010
1020
20 or more
3
24
15
4
5
18
34
20
10
16
13
41
33
19
7
1
16
22
10
28
35
115
90
43
46
87
122
56
311
Total
this represented only 6.5 per cent and 14.3

per cent, respectively. Over 26 per cent of
Midwest MSAs were centralising, as were 24
per cent of Southern areas. At the high end of
the scale, 42.8 per cent of all US metropolitan areas fell into the rapid or hyper decentralising categories.
The percentage of southern and western
MSAs that were included in this latter grouping exceeded those in the Northeast (41.3 per
cent) and Midwest (34.5 per cent). In fact,
57.1 per cent of western MSAs and 42.6 per
cent of southern areas increased their fragmentation over the 20 years by greater than 10
per cent. So while the Northeast and Midwest
were the most fragmented regions, the South
and West were the most fragmenting. The
highly centralised category, see Table 10,
had 37 entries in 1972, but only 20 in 1992
a signi cant decline. While at the other extreme, the number of super decentralised
MSAs across the country increased from 15 to
25, for a 67 per cent gain in only 20 years.
The trend through the early 1990s in metro-
politan America was, thus, toward increasingly uniform governmental fragmentation.

The questions then are: what has happened
during the rest of the decade; and, how are
we to evaluate these continuing changes, in
relation to the Regional Coalitions policy
agenda? We address these issues in the next
two sections.
4. Hollowing-out of the US Metropolis
Tentative answers to such questions require a
broad perspective. First, we will consider
demographic changethe root cause of
governmental adaptationduring the second
half of the 1990s. A hundred years ago,
metropolitan growth in the US was rst documented by Adna F. Weber (1899) in The
Growth of Cities in the Nineteenth Century.6
Three-score years later, Jean Gottman (1959)
wrote his classic study of Megalopolis: The
Urbanized Northeastern Seaboard of the US,
analyzing the vertical arm of the
Boswash/Great Lakes urban system com-
861
Table 10. Fragmentation group membership by region, 1992

Group 1992
Highly centralised
Moderately centralised
Slightly centralised
Moderately decentralised
Highly decentralised
Super decentralised
Score
Northeast
Midwest
South
West
Total
1.002.00
2.003.00
3.004.00
4.005.00
5.007.50
7.50 1
1
5
10
16
14
22
13
24
19
9
17
42
34
22
6
1
3
14
24
5
9
1
20
79
76
61
50
25
46
87
122
56
311
Total
plex. John Friedmann (Friedmann and

Miller, 1965) expanded the metropolitan
concept to include The Urban Fields, like
Southern California: a veritable island on the
land. C. F. Whebell (1969) proposed his
Corridors: a theory of urban systems to
explain Canadian metropolitan clusters along
the St Lawrence River. And Brian Berry
(1973a) completed this celebration of metropolitan growth with his massive, nationwide
empirical study of Growth Centers in the
American Urban System. This was the end of
an era, however. In the same year Berry
(1973b) also published The Human Consequences of Urbanisation, and by 1980 wrote
of Urbanisation and counter urbanisation in
the United States. From this intellectual
questioning of large-scale urban growth during the second half of the 1970s and throughout the 1980s, the 1990s proved a time of
actual metropolitan decline in America.
It is impossible to present an accurate
analysis of changing US urban population
patterns until after the decennial census of
2000. But the US Bureau of the Census has
released population estimates for the decade
198797. These show that average 10-year
population growth for MSAs was 9.9 per
cent, compared with a national average of
10.4 per cent (Mitchell-Weaver et al., 2000,
p. 4)i.e. the growth rate for MSAs was 0.5
per cent lower than the national average.
Declines in the central cities have been
common as Americans headed for suburbs
in recent decades, but now the metropoli-
tan areascity and suburbare losing to

the countryside (Pittsburgh TribuneReview, 1997, p. A8).
In the same vein,
Growth areas are able to offer country
living with proximity to the city For the
rst time, more Americans are migrating
from major metropolitan areas to rural areas, often referred to as exurbs (Pittsburgh
Post-Gazette, 1997, p. S8).
Such marginal net rates change annually.
While the 199596 report concluded that
More people left than moved into metro
areas in 199596, the 1997 County Population Estimates argued that the fastestgrowing counties [were] predominantly
Southern, Western and Metropolitan (US
Bureau of the Census, 1997, p. 1; US Bureau
of the Census, 1998, p. 1).
But this could be misleading, because the
averages are skewed by changes in the
largest US county, Los Angeles, with more
than 9 million residentslarger than about
one-third of all US states and most members
of the United Nations. California reversed its
population decline during the mid 1990s: in
1994, at the bottom of its long recession, the
state lost 430 000 residents, but in 1997 its
population grew by 410 000 (Baker, 1998).
Thus, Southern Californias turnaround,
alone, from post-Cold War aerospace layoffs, signi cantly changed the averages for
national metropolitan net growth in the latest
reporting year, 1997. The reality is that, from
862
at least 199097, many US metropolitan

counties, even by generous geographical
de nitions of MSAs, lost population.
Los Angeles County accounts for approximately 200 per cent of all net gain/loss estimates during the latest two reporting years.
National averages for metropolitan growth
are even more equivocal, however, given the
national-regional nature of both rates of
growth and absolute population growth by
MSA. According to the 1997 County Population Estimates, the top 10 counties in terms
of growth rate were all located in the US
West and South. The 10 biggest numerical
gainers in population were, in order: Maricopa, AZ; Los Angeles; Clark, NV; Orange,
CA; San Diego; Harris, TX; Riverside, CA;
Broward, FL; Dallas and Collin, TX (US
Bureau of the Census, 1998, p. 2). LA, Orange, San Diego and Riverside, California,
are all part of the Southern California conurbation. Maricopa, AZ, and Clark, NV, are
both just next door. With the exception of
Broward, FL, all other major growth counties
lie in metropolitan central Texas. The point
is that these 10 counties experienced an increase of nearly one-half a million inhabitants, which means that real growth in the
majority of other MSAs must have been
negative. This also suggests that in 6 or 7 of
the 10 US federal government regions,
metropolitan hollowing-out may be the
norm.
Hollowing-outthe doughnut effect
occurs when regional population shifts move
both people and jobs away from cities towards surrounding outlying suburbs. Suburbs grew twice as fast in 199697 as
central-city counties. More strikingly, in this
same reporting period, metropolitan counties
followed a similar pattern, with 1.3 per cent
growth rates, while outlying counties in the
same general population cluster increased by
2.6 per cent. This was most marked in metropolitan areas like MinneapolisSt Paul, Atlanta, Nashville, TN, DallasFort Worth and
San Antonio, TX (Baker, 1998). In economic
terms,
could it be that its not the suburbs that
depend on the hub, but that it is instead the

exact opposite? (Pittsburgh TribuneReview, 1998).
Concerning the so-called dependency hypothesis, quoted earlier from Ross and
Levine (1996) (see p. 312), if anything, it
seems that the core city is probably
dependent on its edge cities and suburbs for
its economic vitality (Pittsburgh TribuneReview, 1998). A recent study of the Pittsburgh MSA found similar results, with the
highest wage bill in the region paid by outlying low-skill manufacturing industries
(Mitchell-Weaver et al., 2000, p. 16).
So what does this suggest about core
periphery relations in US metropolitan regions? It contradicts the classical hypothesis
summarised by John Friedmann (1972) in his
A general theory of polarized development. Starting with Gunnar Myrdal (1957),
polarised development theory was always
painted with very broad brush strokes, and its
geographical scale of referenceto say the
leastwas ambiguous. In terms of US
metropolitan regions, its stark spatial dualism
would appear to be unfounded. Rather than
endless cumulative causation or revolutionary decentralisation, core-area rot and
dribbling peripheral growth would seem to
be the unspectacular outcome. Perhaps this is
one possible manifestation of Friedmanns
urban elds.
However this may be, there is something
signi cantly more complex unfolding here
than can be portrayed by simple physical
analogue models (such as doughnuts, or
water droplets spreading across a table top)
or virtual-reality representations on a computer. It also appears to be something quite
different from unending urban-like sprawl
textured by Chancy Harris and Edward Ullmans (1954) multiple-nuclei pattern of urban growth (see Friedmann and Weaver,
1979). As quoted above, Americans would
appear to be actually moving out of the
metropolis into rural areas and smaller towns
and cities, some of them apparently freestanding; escaping the metropolis and resettling the countryside. This observation
cannot be proved or documented for two or

three more years, until the rst statistical
summaries are produced comparing the 1990
and 2000 US Censuses of Population. Even
then, the Census Bureaus propensity to
gather and analyse minimal information
about non-metropolitan areas may lead to
some rather unrepresentative conclusions.
It could be argued that politically and socially conservative groups and upper-income
classes within the population have frequently
tried to escape the hassle, expense and danger of living in the city centre, throughout the
long history of ruralurban relations (see
Mitchell-Weaver, 1995). This though would
miss the crucial structural economicgeographical changes taking place in the US
human landscape. A simple scenario summarising US experience might run something
like the following.
For over a hundred years after the American Civil War (186165), the US space economy was integrated on a nearly continental
basis by the processes and social forces of
the second industrial revolution. Manifest
destiny, the God-given right to control
North America, was not limited by preindustrial economic organisation and social
formations, like continentalist imperial Russia or China. The US national economy,
ordered into a hierarchical system of urban
places, marched across the continent. Large
urban corridors along the eastern seaboard
and Great Lakes, and then in California and
Texas, made the US an urban-industrial behemoth, with a highly concentrated though
decentralised metropolitan core. During the
rst quarter-century following World War II,
rapid metropolitan growth and continuing international and ruralurban migration fuelled
a dramatic process of suburbanisation, based
on cheap energy, the automobile, suburban
roads and highways and middle-income
ight from the increasingly heterogeneous,
polluted, worn-out and crime-ridden metropolitan core cities rst to inner-ring suburbs, then to outer-ring communities and
nally exurbs. The centre city became
poverty-stricken as it lost higher-income
groups, jobs, tax base: economic-geographi-
863
cal power in general. And by the late 1960s

and early 1970s, it became the focus of
concentrated government attention and redistribution payments, because of widespread
social unrest, civil disobedience and rioting.
Urban policy and new planning measures
were a response to domestic political crisis
(Friedmann, 1973).
By the end of the 1970s, the same middleand higher-income people who were moving
to the urban fringe voted by an overwhelming majority for passage of Proposition 13 in
California, which cut property taxes in the
state by 60 per cent. (Property taxes or rates
are the basis of urban public nance in the
US.) Proposition 13 proponents argued that
government spending could be cut by as
much as 25 per cent without any deterioration in public services (to middle-class
home-owners). A wide spectrum of voters
resented the growth of governmental subsidies for the poor, with two-thirds of those
citizens who voted yea believing that welfare
payments should be cut. This marked the
beginnings of the conservative Reagan Revolution of the 1980s, when all manner of
federal support for urban-focused redistribution programmes was drastically cut back
(Herson and Bolland, 1998, pp. 352355).
During the 1990s, these same people apparently continued voting with their feet, moving out of megalopolis back to the
countryside.
Since the 1890 US Census, rural areas had
been losing population. Farmers left the land.
Central-place market centres dried up and
became ghost towns, and the settlement system was dominated by higher-level service
and manufacturing centres in the urban hierarchy. One hundred years laterstarting perhaps only a decade agothe politically and
economically dominant white middle class
began, selectively, moving back to small and
mid-sized cities, and rural areas on the extreme periphery of urban clusters along the
US Defense and Interstate Highways System.
This population shift has been accompanied
by a similar locational change in jobs and
service activities. The outlying metropolitan
periphery and freestanding towns of 5000
864
15 000 (and larger) are now the sites of

major green eld (automated, low-skill) manufacturing plants, vast shopping malls
offering everything that can be bought in
the big cityand tracts of pleasant suburblike, upper-middle-class homes, surrounded
by large lawns and protected by geographical
isolation and locally controlled police forces.
If this is indeed a real demographic shift, a
trend which continues and accelerates into
the 21st century, then metropolitan America
is likely to become even more fragmented.
The large city will experience a changed
ecology and increasing economic, social and
political dif culties (Waste, 1998, p. 6). So
what is to be done in public-policy terms? In
the next section, we examine this second
question, how does the agenda of the new
Regionalist Coalition propose to cope with
the hyper-fragmentation of urban America?
5. The Regional Coalition Agenda
David B. Walker (1987) published a widely
used breakdown of types of intergovernmental co-ordination in the venerable and ad-
mirably persistent National Civic Review.

We present his classi cation scheme here in
Table 11. Walker grouped modes of intergovernmental co-operation according to the
perceived degree of political dif culty in initiating and managing them. The rst two
groups primarily refer to regional approaches
to urban-type service delivery; limited regional governance, if you will. The easiest
are informal, non-binding and contractual arrangements, as well as federal or state encouraged or mandated co-ordination.
Moderately dif cult co-ordination generally
includes special districts, annexation and reformed urban counties (i.e. counties with
modern governmental structures that provide
urban-type services).
The Very dif cult category (numbers
1517), refers to actual units of metropolitan
government, and what Stephens and Wikstrom (2000, p. 102, T5.1) call limited regional structures. This group requires a
fuller explanationbecause it is here that the
Regional Coalition Agenda tends to lead, and
it is here that it will be the most dif cult to
achieve its objectives.
Table 11. Types of intergovernmental co-ordination, governance and

government
Relatively easy
1 Informal co-operation
2 Interlocal service agreements
3 Joint powers agreements
4 Extraterritorial powers
5 Regional councils of government (COGs)
6 Federally encouraged single-purpose districts
7 State planning and development districts (SPDDs)
8 Contracting from private vendors
Moderately dif cult
9 Local special districts
10 Transfer of functions
11 Annexation
12 Regional special districts and authorities
13 Metropolitan multipurpose districts
14 Reformed urban county
Very dif cult
15 One-tier consolidation: city-county and area-wide consolidation
16 Two-tier restructuring: federal structures
17 Three-tier reform: metropolitan-wide structures
Source: Adapted from Walker (1987, p. 16).
One-tier consolidation means a totally

new level of regional government. Historically, it has been established by city-county
and area-wide consolidation. There were several such mergers in the 19th century, as
mentioned above in section 2; 710, depending on how one counts. These were typically
established by state legislative authority.
Since World War II, there have been in the
neighbourhood of 20, including well-known
examples like Baton Rouge, LS (1949),
Nashville, TN (1962), Jacksonville, FL
(1968) and Indianapolis, IN (1970). Most
were in the South and required voter approval. The last notable one, in 1970,
Unigov, which involved a northern industrial city, Indianapolis, could only be
achieved by legislative at. Ten smaller
growth anticipation consolidations took
place between 1971 and 1992 in the South
and West (speci cally, Alaska).
Consolidations are the epitome of metropolitan regionalism, bringing everything under one governmental unit. They are
extremely hard to get past the voters, however. In 1997, the Province of Ontario proposed an area-wide consolidation of six
existing cities that since 1953 had made up
Toronto Metro, a two-tier structure. Seventy ve per cent of voters, perhaps following the
lead of the six municipal mayors, all of
whom opposed amalgamation, voted against
the measure. Nevertheless, provincial
of cials enacted Bill 103 in May 1997, and
in January 1998 the six cities and Metro were
replaced by a unitary City of Toronto
Government. It is hard to imagine such highhanded manipulation across the border in the
US at this point in time.
Two-tier restructuring involves metropolitan-level federalism. Area-wide functions are separated from local ones. The
best example is MiamiDade County,
Florida (1957), a two-tier urban county.
Localities control zoning, lot sizes, education
and other lifestyle choices (amenities),
while the county arranges for system-serving
functions such as highways, transport, sewers, water treatment, solid waste disposal,
etc. This allows the region to bene t from
865
rather obvious scale economies, without local citizens losing control of their neighborhoods. There are distinct advantages and
limitations, in terms of metropolitan
ef ciency and equity. But it is frequently
advocated as the best possible compromise
(Stephens and Wikstrom, 2000, pp. 167
174).
There are merely two examples of threetier reform in the US today: MinneapolisSt
Paul and Portland, OR. Both are celebrated
by the new Regional Coalition (see Leo,
1998, and Rusk, 1999). They represent a
halfway house between federated structures
and consolidated metropolitan government.
Each has unique characteristics. The older of
the two, MinneapolisSt Paul, was established in 1967. It was done by the state
legislature of Minnesota, and has a 16-member regional council, appointed by the governor, which exercises policy-review powers
and provides regionwide services. The region
also bene ts from revenue sharing, evening
out the scal resources of local governmental
units.
Portland Metropolitan Service District
came into existence in 1979. It was based on
a long history of regional co-operation, involving special service districts and the local
COG, the Columbia Regional Association of
Governments. Metro was part of a package
proposed by Oregon governor Tom McCall
as a radical no growth policy, to protect
northern Cascadia from the fate of sprawlinfested California. It was passed by the state
legislature and approved by the voters. Metro
has a seven-member elected council with
policy-review and service-delivery functions.
As part of its land-use planning mandate, it
operates a regional greenbelt, called the
metropolitan urban growth boundary, UGB.
The UGB was approved by the statewide
Oregon Land Conservation and Development
Commission, and speci cally operates to
stop urban sprawl beyond the boundary and
promote more intensive, new urbanismtype development in the City of Portland and
its inner-ring suburbs (see Weitz, 1999).
Portland and MinneapolisSt Paul, without
creating another level of general government,
866
represent the ideals of the new Regional

Coalition Agenda. Stateregionallocal federalism provides co-ordinated planning and
infrastructure development, regionwide provision of municipal services, metropolitan
revenue sharing, and smart growth, limiting
sprawl and redeveloping the core. From this
perspective, all that remains to be added is a
reactivated national government, creating a
more powerful nancial and legal carrot-andstick system. This will probably depend, in
the rst instance, on the outcome of the
November 2000 elections.
Publicists for the new Regional Coalition
Agenda present a now familiar argument,
almost ex nihilo. Bruce Katz, director of the
Brookings Institutions Center on Urban and
Metropolitan Policy, gives an overview of
the Coalitions argument in the Brookings
Review (Katz and Bernstein, 1998):
Americans have been content, for the most
part, with a public sector that consists of a
fragmented maze of local governments
and special districts and a private sector
that builds mostly unrelated subdivisions
rather than integrated communities (p. 4).
While hundreds of independent jurisdictions still partition most of our metropolitan areas, their economic activities are
borderless (p. 4).
In the past few years, metropolitanism has
reemerged as a notable force in dozens of
major metropolitan regionsand it is even
beginning to alter market practices (emphasis added; p. 4).
They are appalled by explosive sprawl into
peripheral farmlands and open space, rising suburban traf c congestion, and
slower growth or absolute decline in many
central cities and older suburbs (pp. 45).
In many American metropolitan areas, especially in the Northeast and Midwest,
central cities and older inner-ring suburbs
have been left behind Consequently,
these once-proud places now harbor higher
and higher concentrations of the poor, particularly the minority poor, without the
scal capacity to grapple with the consequences: joblessness, family fragmen-
tation, failing schools, and deteriorating

commercial districts (emphasis added;
p. 5).
In metropolitan areas across the country,
these changes are creating an impetus for
the formation of new, powerful, sometimes-majority coalitions at local and regional levels. Elected of cials from cities
and inner suburbs; downtown corporate,
philanthropic, and civic interests; minority
and low-income community representatives; environmentalists; no-growth advocates in the new suburbs; farmers and rural
activists; and religious leaders all are realizing that they lose as sprawl accelerates
(p. 5).
Some states are joining the action
Maryland enacted smart growth New
Jersey is considering preserv[ing] hundreds of thousands of acres Minnesota
has upgraded metropolitan government Oregon continues its landmark
land use law Missouri, Ohio and Pennsylvania debate similar reforms (p. 6).7
Even the federal government has gotten
into the act [when] Congress preserved
[the] metropolitan focus when it reauthorised the [1991 Intermodal Surface]
[T]ransportation [A]ct earlier this year
(p. 6).
In the same issue of the Brookings Review,
Anthony Downs (1998, p. 11) summarises
the recommendations in his 1994 book:
The rst [speci c tactic to stop sprawl] is
some type of urban growth boundary to
limit the outward draining of resources
from the core areas.
The second tactic is regional coordination
and rationalization of local land use planning, done by some regional planning
body, such as the Metropolitan Council in
the Twin Cities.
The third tactic is some form of regional
tax-base sharing, with all additions to
commercial and industrial tax bases shared
among all communities in the region, not
just captured by the places where those
developments are built.
The fourth tactic is regionwide develop-
ment of housing for low-income households, either by regional vouchers or regional new subsidies or by requiring
developers to build a share of affordable
housing in each new project.
A fth tactic is regional operation of public transit systems and highways, including
new facility construction.
A nal tactic is vigorous regional enforcement of laws against racial discrimination. 8
Effectively adopting any of these tactics,
or certainly most of them together, would
likely require a strong regionwide implementing body.
Myron Or eld, head of the Metropolitan
Area Research Corporation in Minneapolis,
gives an in-depth statement of the Regional
Coalitions analysis in Metropolitics: A Regional Agenda for Community and Stability
(1997), and provides a summary for action in
hyper-fragmented Chicago by the US Congress (1998): 9
In order to stabilize the central cities and
older suburbs and prevent metropolitan
polarisation, there are six substantive reforms that must be accomplished on a
metropolitan scale. The reforms are interrelated and reinforce each other substantively and politically. They are: (1)
property tax-base equity; (2) reinvestment;
and (3) fair housing. Together these reforms provide resource equity, support the
physical rebuilding necessary to bring
back the middle class and private economy, and gradually relieve the concentrated
social
need
that
exists
disproportionately in older suburban communities. The second three(4) land planning/growth management coordinated with
infrastructure; (5) welfare reform/public
works; and (6) transport/transit reform
reinforce the rst three and allow them to
operate ef ciently and sustainably. In addition, these reforms provide for growth
that is balanced socioeconomically, accessible by transit, economical with
governmental resources, and environmentally conscious (Or eld, 1998, p. 35).
867
In June 1999, the US Department of Housing

and Urban Development published its third
annual The State of the Cities report. It contains a succinct statement of the waning Clinton administrations urban policy which, in
effect, is adoption of the Regional Coalitions essential analysis and agenda (US
HUD, 1999):
Three Major Findings
Finding #1. Thanks to a booming national
economy, most cities are experiencing a
strong scal and economic recovery. However, too many central cities are still left
behind and continue to face the challenges
of population decline, loss of middle-class
families, slow job growth, income inequality, and poverty.
Finding #2. Some older suburbs are experiencing problems once associated with
urban areasjob loss, population decline,
crime, and disinvestment. Simultaneously,
many suburbs, including newer ones, are
straining under sprawling growth that creates traf c congestion, overcrowded
schools, loss of open spaces, and other
sprawl-related problems, and a lack of affordable housing.
Finding #3. There is a strong consensus
on the need for joint city/suburb strategies
to address sprawl and the structural decline of cities and older suburbs. We now
have an historic opportunity for cooperation between cities and counties, urban as well as suburban, to address the
challenges facing our metropolitan areas
(emphasis in the original, p. vi).
The 21st Century Agenda for Cities and Suburbs is composed of four parts:
(1) Opening Doors to New Markets. The
Administrations New Markets Initiative
is designed to ensure communities can
access the risk capital and technical expertise they need to take advantage of
their untapped markets for labor, retail,
and land.
(2) Investing in Americas Working Men and
Women. The Agenda provides tools to
ensure that central city residents have the
868
skills needed for todays job market and

the means to learn about and access jobs
that may be distant from their neighborhoods.
(3) Expanding Homeownership and Affordable Rental Housing. Homeowners can
build strong neighborhoods both in cities
that are beginning to do better and in
those that have been left behind. Providing more assistance for rental housing is
critical: for alleviating the distress of
worst case housing needs and homelessness; for overcoming the housing/jobs
mismatch created by metropolitan development patterns; and for providing
families with the support and stability
they need to become part of the new labor
markets
(4) Promoting Smarter Growth and Livable
Communities. To realize the billions in
savings that could be generated by
strengthening existing developed communities, the Agenda includes a major
initiative to promote livable communities. The agenda also includes measures
to ensure public safety, strengthen our
schools, and preserve natural resources
and historic amenities. By providing
communities with strong tools to tackle
these challenges, we can help enhance
their attractiveness for residents, businesses, and investors (US HUD, 1999,
pp. 4647).
6. Conclusions
The Clinton administration has adopted the
language of the new Regional Coalition
Agenda. Its urban and socioeconomic programmes are presented as the needed federal
boost to get metropolitan regions cooperating in terms of governance: taxation,
service delivery and growth management. It is
therefore appropriate to make a number of
critical observations. First, the Administration
is now a lame duck, and will be even more
hard pressed than before to move ambitious
legislation through an opposition-controlled
Congress. These issues should be part of the
2000 electoral campaign and brought before
the voters, if they have merit. Otherwise, they

will remain an arcane subject for elite thinktanks and Beltway experts.
Secondly, the Regional Coalition and the
Clinton administration both bring a heavy
load of ideological baggage to the metropolitan growth debate. Rather than focusing on
shared governance in system-serving functions that has demonstrated some appeal in
MSAs across the US (Wallis, 1993), they
both emphasise government intervention in
so-called lifestyle choices, immediately raising the suspicions of the middle-class majority. Urban policy is being used as a
substitute for national social programmes. If
urban problems are important public issues, in
and of themselves, as we believe, then every
effort should be made to broaden the spectrum of supporters for workable solutions. As
Bruce Katz enumerated in the last section, the
regional coalition can contain a wide crosssection of mainstream interests. Interests that
must be mobilised if metropolitan regionalism
is to be more successful now than in the past
in North America.
Kathryn Foster (2000, p. 91) observed, in a
review of Donald Rothblatts Metropolitan
Governance Revisited (1998):
Top-down directives, though out of favor,
are necessary for managing metropolitan
development and ensuring scal equalization. These are increasingly unlikely in
Canada and a long shot in the US. Nonpublic groups, a potential regional force, lack
unity and coherence. Voluntary consensus
building is nice but not enough to shape
regional patterns.
State and federal government must actif
there is to be actionand in the US this
means that the majority must approve of
governments intentions.
There are even more fundamental questions to be asked, though, than worries about
political timing and strategies. The new Regional Coalition has gone some way to gain
public attention in the press (see, for example,
Pierce, 1998, 2000; Firestone, 1999), but
seems largely innocent of an historical imagination, and even less concerned with an
empirical understanding of changing regional

settlement patterns and economic structures.
Their objectivescentralised control of urban development and redistribution of income
(Niebanek, 2000, p. 92)are unclouded by
analyses of American political history and the
US space economy. 10
In this paper, we have begun to ll in the
lacunae. Metropolitan regionalism and planning have played a unique role in the US
urban reform movement. From the early years
of the republic, local government consolidation has been seen as a tool to overcome
municipal fragmentation, and thus solve a
long list of urban problems. Before the days
of activist state and federal government,
metropolitanism was one of the few approaches open to reformers to cope with what
Sinclair Lewis called the shame of the cities.
There is a large collection of public documents and a wealth of scholarly publications
that paint in vivid colours the hopes of the
reformers and detail their successes and failures. This is now the third generation of
metropolitan reform in the US (see Wallis,
1994a, 1994b), and we would be profoundly
negligent not to learn from earlier experience.
This includes the limited acceptability of centralised metropolitan government within US
political culture, with its celebration of the
local community and individual liberties.
The distinction drawn so often between
place prosperity and people prosperity by
the great American planner, Harvey S.
Perloff, seems crucial here. Metropolitan governance may be the appropriate means of
improving the quality of life and competitiveness of particular places, but it probably is not
the way to ensure the well-being of people. If
people suffer from class-based and race-based
disadvantages in the US, it is class- and
race-based problems that should be addressed
in the national political discourse and by
national policy (Powell, 1998)not the form
of metropolitan government, which Kathryn
Foster (2000, p. 90) dares to suggest
simply does not matter that much for regional development. Regardless of form,
metropolitan systems prove too power-
869
less, purposeless, or discouraged to shape

urban outcomes.
John Kains (1992) famous spatial mismatch
hypothesis, and the related central-city
suburb dependency hypothesis, need to be
thoroughly analysedand perhaps rejected.
Decades ago, when these ideas were rst
formulated, the upstream and downstream
linkages within metropolitan economies were
well documented. Regional industrial complexes within the urban economy were responsible for much of an MSAs growth and
job creation. Suburbs and residential growth
corridors depended on the central city and
centralised industrial zones for their jobs and
economic well-being. None of these relationships necessarily holds true today, with the
hollowing-out of metropolitan cores, and investment and job creation in the exurbs and
rural-green eld locations purposely targeted
by international corporations. We need a new
theoretical basis for our understanding of
urban and regional economics in the US.
Generalisations, based primarily upon the circumstances prevailing in the 1960s and 1970s
(or the 1920s and 1930s), cannot provide an
adequate basis for informed public policy.
Pop economics supplied by journalists cannot
ll the gap.
Peter Gordon and Harry Richardson (1998)
of the University of Southern California point
this out from another perspective, carrying on
Richardsons well-known crusade against
optimum city size concepts at the turn of the
1970s. They argue that we simply do not
know enough about the costs and bene ts of
urban sprawl to make informed policy decisions. We need to do our homework before
writing slogans and entering the national political fray. Janet Rothenberg Pack (1998) of
the Wharton School at the University of
Pennsylvania, writing in the Brookings
Review, argues that we also do not know
enough about citysuburb relations, and that
her preliminary work suggests that citysuburban links change signi cantly from one
national region to another, as well as within
the regions themselves. We need much more
detailed analyses of various groups of MSAs
870
before appropriate public policy decisions

can be made. Arthur Nelson and Kathryn
Fosters (1999) careful study of Metropolitan governance structure and income
growth is a step forward in creating this
new knowledge-base about urban political
boundaries and socioeconomic processes.
Much more work is needed, however.
Our own analysis here of governmental
fragmentation in US MSAs points to another
necessary research thrust. For the rst time,
we are able to make precise statistical generalisations about metropolitan fragmentation
which are sensitive to most of the important
aspects of the fragmentation debate. As new
data become available from the 1997 Census
of Governments, the model needs to be updated, and analyses of the relationship between the metropolitan fragmentation index
(MFI) and a number of critical socioeconomic variables can be undertaken. Analyses
of the MFI and particular forms of metropolitan governance should also be high on the
21st century research agenda. We can then
begin to make speci c public-policy recommendations about when different types of
metropolitan governance might be a practical
alternative, and when it is unlikely to improve a particular urban setting.
Metropolitan regionalism has a contribution to make in solving US urban problems at
the turn of the 21st century. This idea cannot
be approached in a doctrinaire manner
though. Regional governance must stand or
fall on its merits. It should not be expected to
provide a conduit for other liberal or conservative social policy agendas. In the last
weeks of 1999, the Brookings Institution
shelved a book project entitled The Interdependence of Central Cities and Suburbs, a
project meant apparently to demonstrate the
ravages of suburban-dominated central-city
dependency: the cities left behind.11 This
seems to us to be a healthy retreat from
ideology. A reasonable debate of the US
metropolitan crisis, informed by an understanding of changing ruralurban settlement
patterns and increasing nationwide metropolitan fragmentation, may draw in responsible elements from both the major political
parties. This could then be incorporated in

election-year political platforms, and the issue would be put before the electorate. This
is the only way the Regional Coalition
Agenda can nd a place in the new US
administration that will take of ce in 2001.
Notes
1. The dual use of the term liberal in US
political discourse must be kept in mind.
Confusingly, liberal can refer either to free
market economy solutions to public problems or to progressive, activist government.
In this paper, we mean to invoke the latter
set of ideas and concepts.
2. This section is drawn primarily from
Stephens and Wikstrom (2000), Kweit and
Kweit (1999), Herson and Bolland (1998),
Ross and Levine (1996) and Friedmann and
Weaver (1979). Stephens and Wikstroms
200-page treatment of Metropolitan Government and Governance is invaluable. Chapter
11, Beyond the Central City: Cities, their
Suburbs and their States (pp. 230265), of
the recent second edition of Herson and Bollands well-known urban politics text adds a
number of important insights. And Ross and
Levines three solid chapters on metropolitan
regionalism offer a concise but systematic
source of reference. All three books adhere
very closely to the political science and sociology literature, however, and must be complemented with material from the
traditionally more technical areas of urban
studies: urban-regional economics, urbaneconomic geography and urban and regional
planning.
3. The crucial linkage of metropolitan political
institutions and political geography with the
empirical analysis of other socioeconomic
factors is still uncommon over 50 years later.
4. Counties are the fundamental unit of local
government in the US, originally charged
with the maintenance of roads and some
form of local police protection. Municipalities are urban localities, which means they
must deliver a range of urban servicesfor
example, water, sewer, re, police, parks,
solid waste disposal. Towns and townships
are typically basic sub-divisions of counties,
but in some states they may also have urbantype service responsibilities. School districts
are special-purpose local governments which
only provide elementary, secondary and perhaps community-college level education to
residents. Special districts are urban-type
service units outside the realm of education,
5.
6.
7.
8.
9.
10.
such as water, sewer or re protection.

Names, legal descriptions and responsibilities vary for all of these different designations from state to state and over time.
MSAs with negative change of 5 per cent
were classi ed as centralising; those with a
negative score of less than 5 per cent were
classi ed as slow centralising. Negative
change means a decrease in the MFI. The
lower the index, the more the MSA is governmentally centralised. Therefore, a decrease in the index represents a centralising
trend, and the resulting percentage would be
a minus or negative number.
Originally a PhD thesis, submitted to Columbia University in 1898.
The Commonwealth of Pennsylvania passed
Senate Bill 300 at the end of December
1999. This bill amends the states 1968 Planning Code, and allows local government jurisdictions to band together and designate
urban growth boundaries (URBs). Unlike
the Oregon law, however, there is no state
requirement to create URBs around existing
cities.
The dean of American land-use lawyers,
Charles M. Haar (1996a, 1996b) of Harvard
University, argues that court-mandated provision of affordable housing to overcome
segregation must be focused on the metropolitan level.
Special thanks to Heather Tureen and unknown members of the Brookings Institution
staff for helping us track down stray publications.
Paul Niebanek (2000, p. 92), the conscience
of US urban planning, categorised the substance of the new Regional Coalition Agenda
in a review of David Rusks Inside Game/
Outside Game: Winning Strategies for Saving Urban America (1999):
require regional land use planning;
ensure that all suburbs have their fair
share of low- and moderate-income
housing;
implement regional revenue sharing.
11. Telephone conversation between the senior

author and an anonymous employee of the
Brookings Center on Urban and Metropolitan Policy, 29 December 1999.
12. The use of the standard deviation as a measure of fragmentation has limited utility. Because it is a measure of dispersal around the
mean, one government with all the expenditures and 100 governments each with 1 per
cent of the expenditures would both have a
standard deviation of 0. Yet, the rst case
would represent a highly centralised system
and the second a highly decentralised system.
871
13. Unfortunately, after 1992, austerity measures

at the Census Bureau have drastically reduced the information available to the public.
Detailed data for the 1997 Census of Governments are much abbreviated and still unavailable at the time of writing.
References
ADVISORY COMMISSION ON INTERGOVERNMENTAL
RELATIONS (1961) Governmental Structure,
Organization and Planning in Metropolitan
Areas. Washington, DC: US Government
Printing Of ce.
RELATIONS (1962) Alternative Approaches to
Governmental Reorganization in Metropolitan
Areas. Washington, DC: US Governmental
Printing Of ce.
RELATIONS (1966) Metropolitan America:
Challenge to Federalism. Washington, DC: US
Governmental Printing Of ce.
RELATIONS (197374) Regional Decision Making. Vols. IVI. Washington, DC: US Governmental Printing Of ce.
AMERICAN PLANNING ASSOCIATION (2000) Special
issue on smart growth, Planning, 66(1), pp. 4
31.
ANDERSON, W. (1934) The Units of Government
in the United States. Chicago, IL: Public Service Administration Service.
BAKER, G. (1998) West is still best for a nation to
move, Financial Times, 19 March, p. 7.
BARNES, W. R. and LEDEBUR, L. C. (1998) The
New Regional Economies: The US Common
Market and the Global Economy. Thousand
Oaks, CA: Sage Publications.
BERKE, P. R. and CONWAY, M. M. (2000) Are we
planning for sustainable development? An
evaluation of 30 comprehensive plans, Journal
of the American Planning Association, 66(1),
pp. 2133.
BERRY, B. J. L. (1973a) Growth Centers in the
American Urban System, 2 vols. Cambridge,
MA: Ballinger.
BERRY, B. J. L. (1973b) The Human Consequences of Urbanization. New York: St
Martins Press.
BERRY, B. J. L. (1980) Urbanization and counter
urbanization in the United States, Annals of the
American Academy of Political and Social
Science, 451, pp. 1320.
BOLLENS, J. C. and SCHMADT, H. J. (1965) The
Metropolis: Its People, Politics, and Economic
Life. New York: Harper and Row.
BOLLENS, S. A. (1986) A political-ecological
analysis of income inequality in the metro-
872
politan area, Urban Affairs Quarterly, 22(2),

pp. 221241.
BRENNAN , J. and HILL, E. W. (1999) Where are
the jobs?: cities, suburbs, and the competition
for employment. Series 99112, Center on
Urban and Metropolitan Policy, The Brookings
Institution, Washington, DC.
BROOKINGS INSTITUTION (1998) The new metropolitan agenda, Brookings Review, 16(4),
pp. 238.
BURNHAM, D. H. and BENNETT , E. H. (1909) Plan
of Chicago, C. Moore (Ed. ) Chicago: The
Commercial Club.
CISNEROS , H. G. (1995) Regionalism: The New
Geography of Opportunity. Washington, DC:
US Department of Housing and Urban Development.
COMMITTEE ON ECONOMIC DEVELOPMENT (1970)
Reshaping Government in Metropolitan Areas.
New York: CED.
COMMITTEE ON PLAN OF NEW YORK (192931) The
Regional Plan for New York and Its Environs,
10 vols. New York: Russell Sage Foundation.
DOLAN . D. A. (1990) Local government fragmentation: does it drive up the cost of government?
Urban Affairs Quarterly, 26(2), pp. 2845.
DOWNS, A. (1994) New Visions for Metropolitan
America. Washington, DC: The Brookings Institution.
DOWNS, A. (1998) How Americas cities are
growing: the big picture, Brookings Review,
17(4), pp. 811.
ELDREDGE, H. W. (Ed.) (1967) Taming Megapolis, 2 vols. Garden City, N.Y: Anchor Books.
FIRESTONE, D. (1999) Runaway growth around
Atlanta, Pittsburgh Post-Gazette, 21 November, p. A16.
FORD FOUNDATION (1999) Risking the future of
our cities, Ford Foundation Report, 30(3),
pp. 37.
FOSTER , K. A. (1997) Regional impulses, Journal
of Urban Studies, 19, pp. 375403.
FOSTER , K. A. (2000) Review of Metropolitan
Governance Revisited: American/Canandian
Intergovernmental Perspectives, edited by D.
N. ROTHBATT and A. SANCTON, Journal of the
American Planning Association, 66(1), pp. 90
91.
FRIEDMANN, J. (1972) A general theory of polarized development, in: N. HANSON (Ed.) Growth
Centres in Regional Economic Development,
pp. 82107. New York: Free Press.
FRIEDMANN, J. (1973) Retracking America: A
Theory of Transactive Planning. Garden City,
NY: Anchor Books.
FRIEDMANN J. and MILLER , J. (1965) The Urban
Fields, Journal of the American Institute of
Planners, 31, pp. 312320.
FRIEDMANN, J. and W EAVER, C. (1979) Territory
and Function: The Evolution of Regional Plan-
ning. Berkely, CA: University of California

Press.
GOODMAN, J. S. (1980) The Dynamics of Urban
Growth and Politics. New York: Macmillan.
GORDON, P. and RICHARDSON, H. W. (1998) Prove
it: the costs and bene ts of sprawl, Brookings
Review, 16(4), pp. 2325.
GOTTMAN , J. (1959) Megalopolis: The Urbanized
Northeastern Seaboard of the United States.
Cambridge, MA: MIT Press.
GRAVES, W. B. (1964) American Intergovernmental Relations. New York: Charles Scribners
Sons.
GULICK , L. H. (1962) The Metropolitan Problem
and American Ideas. New York: Alfred Knopf.
HAAR, C. M. (1996a) Suburbs under Siege: Race,
Space, and Audacious Judges. Princeton, NJ:
Princeton University Press.
HAAR, C. M. (1996b) Can the courts break the
affordable housing deadlock in metropolitan
areas of the USA.?, Ekistics, pp. 376378
(JanuaryJune), pp. 152157.
HARRIS, C. D. and ULLMAN, E. L. (1954) The
nature of cities, Annals of the Academy of
Political and Social Science, 242, pp. 717.
HAWKINS, B. (1971) The environmental base of
urban government reforms, in: B. HAWKINS
(Ed.) Politics and Urban Policies, pp. 4447.
Indianapolis, IN: Bobbs-Merrill.
HERSON, L. and BOLLARD, J. (1998) The Urban
Web: Politics, Policy and Theory, 2nd edn.
Chicago, IL: Nelson-Hall Publications.
HILL. R. C. (1974) Separate and unequal: governmental inequality, American Political Science
Review, 68, pp. 15571568.
IMMERGLUCK, D. (1999) Cities and Finance Jobs:
The Effects of Financial Services Restructuring
on the Location of Employment. Discussion
Paper and Survey Services 99:111, Center on
Urban and Metropolitan Policy, The Brookings
Institution, Washington, DC.
JONES, V. (1942) Metropolitan Government.
Chicago, IL: University of Chicago Press.
KAIN , J. F. (1992) The spatial mismatch hypothesis: three decades later, Housing Policy
Debate, 3(1), pp. 1726.
KATZ, B. and ALLEN , K. (1999) Help wanted:
connecting inner-city job seekers with suburban
jobs, Brookings Review, 17(4), pp. 3135.
KATZ, B. and BERNSTEIN , S. (1998) The new
metropolitan agenda, Brookings Review, 16(4),
pp. 47.
KATZ, P. (1994) The New Urbanism: Toward
an Architecture of Community. New York:
McGraw-Hill.
KWEIT, R. W. and KWEIT, M. G. (1999) People
and Politics in Urban America, 2nd edn. New
York: Garland Publishing.
LEO, C. (1998) Regional growth management
regime: the case of Portland, Oregon, Journal
of Urban Affairs, 20(4), pp. 363394.
LEWIS, P. (1996) Shaping Suburbia: How Political Institutions Organize Urban Development.
Pittsburgh, PA: University of Pittsburgh Press.
LINDSTROM , B. (1998) Regional cooperation and
sustainable growth: nine Councils of Government in Northeastern Illinois, Journal of Urban
Affairs, 20(3), pp. 327342.
MAXEY, C. (1922) The political integration of
metropolitan communities, National Municipal
Review, 11, pp. 229253.
MCKENZIE , R. D. (1993) The Metropolitan Community. New York: McGraw-Hill.
MERRIAM, C. E., PARRATT, S. D. and LEPAWSKY,
A. (1933) The Government of the Metropolitan
Region of Chicago. Chicago, IL: University of
Chicago Press.
MITCHELL -WEAVER, C. (1995) The urban and the
rural: an historical geographic overview, in: G.
BENKO and U. STROHMAYER (Eds) Geography,
History, and Social Sciences, pp. 97132.
Boston: Kluwer Academic Publishers.
MITCHELL -WEAVER C., DEITRICK, S. and
RIGOPOULOU, A. (2000) Linkages between reconversion of old industrial milieux, intraregional dynamics and continentalization: the
case of Pittsburgh, in: J. M. FONTAN (Ed.) Entre
la Metropolisation et le Village Global: Les
Scenes Territoriales de la Reconversion. Montreal: Presses de lUniversite du Quebec (forthcoming).
MYRDAL, G. (1957) Economic Theory and Underdeveloped Regions. London: Duckworth.
NATIONAL RESOURCES COMMITTEE (1937) Our Cities: Their Role in the National Economy.
Washington, DC: NRC.
NELSON, A. C. and FOSTER , K. A. (1999) Metropolitan governance structure and income
growth, Journal of Urban Affairs, 21(3),
pp. 309324.
NIEBANEK, P. (2000) Review of Inside Game/
Outside Game: Winning Strategies for Saving
Urban America by D. Rusk, Journal of the
American Planning Association, 66(1), pp. 92
93.
ORFIELD, M. (1997) Metropolitics: A Regional
Agenda for Community and Stability, rev. edn.
Washington, DC: Brookings Institution.
ORFIELD, M. (1998) Chicago metropolitics: a
regional agenda for members of the US
Congress. Discussion Paper and Survey
Series, 98:1, Center on Urban and Metropolitan
Policy, The Brookings Institution, Washington,
DC
PACK, J. R. (1998) Metropolitan areas: regional
differences, Brookings Review, 16(4), pp. 26
30.
PARKS, R. B. and OAKERSON, R. J. (1992) Metropolitan Organization: The Allegheny County
Case. Prepared for ACIR. Washington, DC: US
Government Printing Of ce.
873
PHARES, D. (Ed.) (1999) Metropolitan Governance without Metropolitan Government?

Brook eld, VT: Ashgate Publishing Co.
PIERCE, N. (1993) Citistates: How Urban America
Can Prosper in a Competitive World. Washington, DC: Seven Locks Press.
PIERCE, N. (1998) Atlanta pays price for growth,
Pittsburgh Tribune-Review, 1 December, p. A7.
PIERCE, N. (2000) Smart regions: smart politics?,
Pittsburgh Tribune-Review, 17 January, p. A7.
Pittsburgh Post-Gazette (1997) City folks take the
highway out of towns, 26 November, pp. 51
and 58.
Pittsburgh Tribune-Review (1997) Americans
leaving town, 4 December, p. A8.
Pittsburgh Tribune Review (1998) Who depends
on whom? 10 May, p. A6.
POWELL, J. A. (1998) Race space: what really
drives metropolitan growth, Brookings Review,
16(4), pp. 2022.
ROSS, B. H. and LEVINE, M. A. (1996) Urban
Politics: Power in Metropolitan America, 5th
edn. N. Itasca, IL: F. E. Peacock.
ROTHBLATT, D. N. and SANCTON, A. (Eds) (1998)
Metropolitan Governance Revisited: American/
Canadian Intergovernmental Perspectives.
Berkeley, CA: Institute of Governmental Studies, University of California.
RUSK, D. (1993) Cities Without Suburbs. Washington, DC: Woodrow Wilson Press.
RUSK, D. (1999) Inside Game/Outside Game:
Winning Strategies for Saving Urban America.
Washington, DC: Brookings Institution Press.
SCHERER, F. M. and ROSS, D. (1990) Industrial
Market Structure and Economic Performance.
Boston: Houghton Mif in.
SHEPHERD, W. G. (1985) The Economics of
Industrial Organization. Engelwood Cliffs, NJ:
Prentice-Hall.
STEPHENS, G. R. and WIKSTROM, N. (2000)
Metropolitan Government and Governance:
Theoretical Perspectives, Empirical Analysis,
and the Future. New York: Oxford University
Press.
STUDENSKI, P. (1930) The Government of Metropolitan Areas in the United States. New York:
National Municipal League.
US BUREAU OF THE CENSUS (1972) Census of
Governments 1972. Washington, DC: US
US BUREAU OF THE CENSUS (1992) Census of
Governments 1992. Washington, DC: US
US BUREAU OF THE CENSUS (1997) Geographical
Mobility: March 19951996. Census Bureau,
Department of Commerce, Washington, DC.
US BUREAU OF THE CENSUS (1998) 1997 County
Population Estimates. Washington, DC: US
874
US HUD (DEPARTMENT OF HOUSING AND URBAN

DEVELOPMENT ) (1999) The State of the Cities,
1999. third annual report. Washington, DC: US
HUD. June.
WALKER , D. B. (1987) Snow White and the 17
dwarfs: from metro cooperation to governance,
National Civic Review, 76, pp. 1428.
WALKER , D. B. (1995) The Rebirth of Federalism:
Slouching Toward Washington. Chatham, NJ:
Chatham House.
WALLIS, A. D. (1993) Governance and the civic
infrastructure of metropolitan regions, National
Civic Review, 82, pp. 131137.
WALLIS A. D. (1994a) Inventing regionalism: the
rst two waves, National Civic Review, 83,
pp. 170171.
WALLIS A. D. (1994b) The third wave: current
trends in regional governance, National Civic
Review, pp. 290310.
WALLIS A. D. (1994c) Inventing regionalism: a
two-phase approach, National Civic Review,
83, pp. 447468.
WASTE, R. J. (1998) Independent Cities: Rethinking US Urban Policy. New York: Oxford
University Press.
WEBER, A. F. (1899) The Growth of Cities in the
Nineteenth Century. New York: Macmillan.
WEITZ , J. (1999) Sprawl Busting: State Programs
to Guide Growth. Chicago, IL: American Planning Association Press.
WHEBELL, C. F. (1969) Corridors: a theory of
urban systems, Annals of the Association of
American Geographers, 59, pp. 126.
ZEIGLER, D. J. and BRUNN, S. D. (1980) Geopolitical fragmentation and the pattern of growth
and need: de ning the cleavage between Sunbelt and Frostbelt metropolises, in: S. D.
BRUNN and J. O. WHEELER (Eds) The American
Metropolitan System, pp. 7792. London: V. H.
Winston.
Appendix. Methodology for Creation of a

Metropolitan Fragmentation Index (MFI)
Comparative analyses of governmental fragmentation in metropolitan areas across the US fall
broadly into two methodological groups. The rst
approach is a simple process of counting the
governments, either in absolute terms, as in the
text in Table 1, or on some per capita basis. Dolan
(1990, p. 28) de ned local government fragmentation as the proliferation of government units that
may exist within a given metropolitan region.
This measure was built on the earlier work of
Goodman (1980), who identi ed four types of
fragmentation, two of which were counts of (1)
incorporated municipalities, and (2) special districts, public authorities and school districts. Hill
(1974), in an effort to assess inequality among

residents of metropolitan areas, used the number
of municipalities and the number of municipalities per capita as measures of government fragmentation. Bollens (1986) was also interested in
inequality within metropolitan areas, and used the
number of non-centre city municipalities with
over 10 000 population per 100 000 of non-centre
population as a measure of fragmentation. Zeigler
and Brunn (1980) used the number of local governments per 100 000 inhabitants to distinguish
geopolitical patterns in the US Frostbelt from the
Sunbelt. Hawkins (1971) developed a measure of
fragmenation as total governments per 100 000
population in an attempt to determine the impact
of fragmentation on the cost of government. Parks
and Oakerson (1992) used governments per
10 000 inhabitants as a fragmentation score.
The idea has merit that the more governments
there areeither in absolute or in per capita
termsthe more fragmented is the metropolitan
region. Creating a government puts into play another actor with political power and rights of
entry into the public decision-making process.
One signi cant problem is that it fails to provide
a measure of the role each government plays in
what it contributes to the regions. As such, having
a signi cant number of governments that exist on
paper can overin ate that statistic as a meaningful indicator.
Indeed, several of the authors cited above addressed this weakness. Dolan tried to compensate
by introducing the concept of scal dispersion
fragmentation, de ned as the standard deviation
of the per capita expenditures of the governments
in the region under study.12 Bollens added the
percentage on non-central-city population that
lives in incorporated municipalities with over
10 000 population as a measure. Zeigler and
Brunn attempted to reduce several dimensions
into a single index by using the number of governments as a direct proportion and the percentage
of the population living in the centre city as an
inverse proportion.
Regardless of the work of these writers to add
a geopolitical dimension to their fragmentation
index, none of the studies has added a time dimension. This generally can be understood because the authors were using their measure of
fragmentation to explain some other condition in
metropolitan areas of the US. As such, they failed
to assess how fragmentation changed over time.
The second broad approach to measuring fragmentation applies a methodology from the private
business sector. It relates the market share of
several rms in a competitive arena, and is often
referred to as the HirshHer ndal Index (HHI).
Relative power is measured by market share. If
one rm has 90 per cent of the market, whether 50
players or 5 players share the remaining 10 per
cent is of marginal interest. These small players

have little political power. Indeed, Scherer and
Ross (1990, p. 72) observed that the HHI
weights more heavily the value for large rms
than for small (also see Sheperd, 1985). The
technique employed is the use of the squared
percentage of each players market share. As applied to local governments in a metropolitan area,
a measure of expenditures on some array of public services usually substitutes for sales by the
private-sector rm.
Lewis (1996) employed a variation of this approach in his political fragmentation index. Using
the sum of the squared percentages of total expenditures in relation to the degree of expenditures,
this index created a single number that is more
sensitive to the total level of expenditures than to
the distribution of those expenditures within the
metropolitan area.
Although both methodologies capture important principlesthe rst a measure of political
power and the second a measure of economic
powerthey need to be combined so that both
may make a contribution to the resulting scale.
The problem now boils down to a mathematical
one. How does one mathematically represent
these two perspectives on a single scale? A colleague suggested to us that the square root of the
squared contributions could be substituted for the
square of the contributions. Whereas, the square
of the percentage contributions has the impact of
exaggerating the contribution of the larger players, the square root of the percentage contribution
gives greater mathematical value to the smaller
units. Basing the scale on the percentage contribution of each player serves to re ect the economic
dimension, while using the square root of the
contribution re ects the political dimension of
power derived from the semi-sovereignty of political jurisdictions in a metropolitan environment.
In the process of using the squared-percentage
approach, the resulting scale ranges from 0 to 1.
As the scale approaches 1, the greater is the
concentration of market power. Hence, a low
score represents a more fragmented system. By
switching to the square root, the scale starts at 1
and goes, theoretically, to in nity. Like the rst
scale, 1 represents pure concentration or one
player with 100 per cent of the market. Fragmentation, however, is represented by higher numbers.
Developing fragmentation scores for each region in the US using the revised formula requires
a data source that has information on each
governmental jurisdiction within each metropolitan region. The Census of Governments, taken
every ve years, develops a summary of expenditures, revenues and intergovernmental transfers
for virtually every local government in the US.13
Included in the analysis are all general-purpose
875
governments such as counties, boroughs, towns

and townships. Also included are all single-purpose governments such as school districts, utility
authorities and special districts. It is possible to
group these census data into the appropriate
Metropolitan Statistical Areas (MSA), at least
back to 1972. As a result, 1972 is used here as the
beginning year for our studies, and subsequent
analyses were undertaken for 1977, 1982, 1987
and 1992.
The US Bureau of the Census has categorised
metropolitan regions into metropolitan statistical
areas (MSAs). However, the Census of Governments, although undertaken by the Bureau of the
Census, has never structured its reports on MSAs.
Hence, categorising local governments into their
respective MSA is not an easy task. In order to
have those MSAs approximate their current structure, 1992 was selected as the base year. Counties
and municipalities within counties were coded,
based on the 1992 de nition of each MSA. That
coding was also used to group the 1972 data. As
a result, the representation of the 1972 data is as
the MSA was de ned in 1992. In our case, of 336
MSAs currently identi ed, only 311 could actually be used in the analysis. Those 311 metropolitan regions contain slightly less than 33 000
individual governments from which data were
used.
Selecting a measure for the fragmentation index was accomplished in the following manner.
Table A1. Expenditures used in the metropolitan
fragmentation index (MFI)
Fire
Judicial
Central staff
Buildings
Health
Hospitals
Welfare
Highways
Community
Development
Libraries
Parking
Parks and recreation
Police
Natural resources
Sewerage
Solid waste
Water
Electricity
Gas
Transit
General expenses
876
Expenditures serve as a reasonable surrogate for

political power in a metropolitan region. The act
of making expenditures is a representation of
choice in that they re ect not only the expenditures being made, but also the universe of expenditures that could have been made somewhere
else. At a broader level, they also represent a
choice of who (or what government unit) will
make those expenditures.
To capture the range of services traditionally
considered in the US to be the domain of local
government, selected types of operational expense
were used (listed in Table A1).
Computationally, the total operational expenditures in any or all of the categories shown in
Table A1 were added together for each government unit to generate the total spending by that
government. Obviously, some governments, like
special purpose districts, might spend in only one
of the categories, while others, like many city
governments, might spend in almost all categories. Each governments percentage of the total
spending was computed and the square root was
taken and added together to generate the fragmentation index for each MSA.
For a more detailed but still preliminary analysis, MSAs were grouped into two categories
based on our current understanding of what factors seem to affect the structure of metropolitan
organisational designs: population and geography.

Multilevel Governance and Metropolitan Regionalism in The USA - URBAN STUDIES

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

Multilevel Governance and Metropolitan Regionalism in The USA - URBAN STUDIES

Uploaded by

Copyright:

Available Formats

Urban Studies, Vol. 37, No.

56, 851876, 2000

Multilevel Governance and Metropolitan

threatens the economic base of US core cities

CLYDE MITCHELL-WEAVER ET AL.

National Civic Review). The whole August

Downloaded from usj.sagepub.com at UNIV FEDERAL DO PARA on March 16, 2015

METROPOLITAN REGIONALISM IN THE USA

1894) and tenement house laws (1867, 1879,

Downloaded from usj.sagepub.com at UNIV FEDERAL DO PARA on March 16, 2015

CLYDE MITCHELL-WEAVER ET AL.

From the late 1940s through the early

typical of the period. By the time ACIRs

Downloaded from usj.sagepub.com at UNIV FEDERAL DO PARA on March 16, 2015

METROPOLITAN REGIONALISM IN THE USA

3. The protection of privilege

lons Rule; see, City of Clinton, Missouri, v.

Table 1. Local governments in the US, 193297

CLYDE MITCHELL-WEAVER ET AL.

This grand total understates settlementpattern change though, because it overlooks

In 1972, 311 metropolitan regions of the

Table 2. Metropolitan fragmentation index (MFI), 1972

METROPOLITAN REGIONALISM IN THE USA

Large (over 2 million)

creased, so did the index. Density and higher

Downloaded from usj.sagepub.com at UNIV FEDERAL DO PARA on March 16, 2015

CLYDE MITCHELL-WEAVER ET AL.

Table 4. Metropolitan fragmentation index (MFI), 1992

ing more rapidly than the nation as a whole,

had a higher MFI were apt to fragment faster

Large (over 2 million)

METROPOLITAN REGIONALISM IN THE USA

Table 6. Metropolitan fragmentation index, 197292 (percentage change)

location was statistically insigni cant, the

MSAs, 36 of 37 were located in the South or

Downloaded from usj.sagepub.com at UNIV FEDERAL DO PARA on March 16, 2015

CLYDE MITCHELL-WEAVER ET AL.

Table 8. Fragmentation group membership by region, 1972

Table 9. Fragmentation trend group membership by region, 197292

this represented only 6.5 per cent and 14.3

politan America was, thus, toward increasingly uniform governmental fragmentation.

Downloaded from usj.sagepub.com at UNIV FEDERAL DO PARA on March 16, 2015

METROPOLITAN REGIONALISM IN THE USA

Table 10. Fragmentation group membership by region, 1992

plex. John Friedmann (Friedmann and

tan areascity and suburbare losing to

Downloaded from usj.sagepub.com at UNIV FEDERAL DO PARA on March 16, 2015

CLYDE MITCHELL-WEAVER ET AL.

at least 199097, many US metropolitan

depend on the hub, but that it is instead the

Downloaded from usj.sagepub.com at UNIV FEDERAL DO PARA on March 16, 2015

METROPOLITAN REGIONALISM IN THE USA

cannot be proved or documented for two or

cal power in general. And by the late 1960s

Downloaded from usj.sagepub.com at UNIV FEDERAL DO PARA on March 16, 2015

CLYDE MITCHELL-WEAVER ET AL.

15 000 (and larger) are now the sites of

mirably persistent National Civic Review.

Table 11. Types of intergovernmental co-ordination, governance and

METROPOLITAN REGIONALISM IN THE USA

One-tier consolidation means a totally

Downloaded from usj.sagepub.com at UNIV FEDERAL DO PARA on March 16, 2015

CLYDE MITCHELL-WEAVER ET AL.