CityVitals

Series: Economic Integration

Why It Matters and How Cities Can Get More of It
Paul Jargowsky, University of Texas at Dallas
Todd Swanstrom, University of Missouri St. Louis
Why It Matters and How Cities Can Get More of It
Project Sponsor


Tabl e of Contents

Introduction 1
The Reali ty of Economi c Segregation 2
Why Do We Care about Economi c Integration? 5
Measuri ng Economi c Integration 9
Segregation of the Poor from the Non-Poor 13
Concentration of Poverty 15
The Metropoli tan Context of Economi c Integration 16
Weak Market Metropolitan Areas: Excess Supply 17
Strong Market Metropolitan Areas: Rising Demand, Limited Supply 21
Poli cy Approaches to Economi c Integration 25
Inclusionary Zoning 28
Using Public Transit to Promote Economic Integration 32
Concl usi on: Where Do We Go from Here? 35
Endnotes 37

Int roducti on
One of the most striking characteristics of U.S. cities is residential segregation.
Historically, researchers and policymakers have focused on segregation by race,
especially the segregation of African Americans into troubled black ghettos in
central cities. Segregation of blacks from whites in most U.S. metropolitan areas
rose steadily from the turn of the century until about 1970.

While racial
segregation is still high by historical and international standards, there has been
a slow and steady decline in segregation by race since 1970, driven by the Fair
Housing Act of 1968 and steady progress in race relations.

Progress against racial segregation is certainly good news, and it helps blacks
and other minority group members enjoy better access to jobs and educational
opportunities. Unfortunately, there is a countertrend that may undercut the
benefits of reduced racial segregation. Closely tied to rapid suburbanization in
many cities and regions, economic segregation, the degree of residential
separation of families based on their incomes, has been rising in U.S.
metropolitan areas since at least 1970.

The result is that low-income persons

are spatially isolated in neighborhoods with few resources, high crime and little
access to economic opportunities. A growing body of research shows
segregation by income has harmful effects on low-income people, including
worse economic outcomes for adults, higher school dropout and teenage
pregnancy rates, and worse academic achievement for schoolchildren. Poor
people who live in mixed-income areas do better than poor people who live in
areas of concentrated poverty. As policymakers think about growth and
development in their metropolitan regions, they need to implement policies that
lead to economic integration to improve the spatial access to opportunity for low-
income citizens.

Now may be a particularly favorable time for policymakers to consider policies
that promote economic integration. For one thing, in many cities, rapid
suburbanization has begun to reach limits imposed by commuting time and
increased fuel costs. The housing finance crisis also calls into question the
approach to urban and suburban development that has dominated the last
several decades. Appealing to the free market, policies have promoted the flight
of higher-income households to the suburban fringe, leaving behind pockets of
concentrated poverty. In many neighborhoods, the large number of foreclosures
provides the opportunity for and points to the necessity of creative approaches to
redevelopment. In this paper, we seek to give policymakers a guide to the
concept of economic integration and to highlight particular cities that are more
integrated. We then examine a series of policies that have been tried or
advocated to promote economic integration.

The Reali ty of Economi c Segregati on

Every metropolitan area has a wide variety of neighborhoods, including some very
rich and some very poor neighborhoods. Often the poorest neighborhoods are
clustered together in the central city, not far from the central business district,
while many of the wealthiest neighborhoods are clustered in the outer suburbs.
There are exceptions, of course. East St. Louis in the St. Louis metropolitan area
and Camden, NJ, in the Philadelphia metropolitan area are high-poverty
neighborhoods not even located in the same state as the central business district
of those metropolitan areas. And there are often pockets of great wealth within
central city boundaries, such as the Turtle Creek neighborhood of Dallas, home
to some of the nations wealthiest families.

Even though all metropolitan areas have richer and poorer neighborhoods, they
are not all segregated by income to the same degree. Some cities have a far
more economic integration, the mixing of households of different income levels
within neighborhoods. If economic segregation is harmful to low-income
persons and creates barriers to economic success, then economic integration
has the potential provide them with greater access to economic opportunities.
Map 1:
Memphis
Metro
Area
Neighborhoods of
different incomes
cluster together,
with the highest
incomes in the
eastern suburbs.
43592
53902
0
78974
75351
34545
62692
82847
45511
70437
44550
72468
64511
25395
61982
60877
58565
53339
61250
66810
55417
61819
94900
41940
0
26250
55172
76839
25080
40046
45909
100393
63125
54023
30291
141729
33492
80395
40431
51786
56250
61142
30869
33333
21522
27058
89377
46794
38542
62032
39988
65953
101643
67730
27656
42460
40885
23054
52380
36014
34655
40974
45661
23179
37011
79292
23738
74375
20630
63481
30458
52066
17623
46071
57448
68089
24401
40042
37485
33047
41828
52561
100858
22865
125206
43764
45015
75582
40736
52298
16703
21086
61188
27382
58948 46341
29195
27202
79768
32383
60250
26793
30993
99159
24891
71696
43098
78543
33696
46737
31000
48517
31193
28444
40754
42656
26377
41084
34273
36025
24000
22554
19229
41832 50199
17346
31011
46445
46494
42333
74280
50146
67241
50968
56638
47598
25221
34911
77778
46397
30220
61000
29861
17697
20296
37153
62997
39830
40431
51858
32841
32622
35981
12015
40583
29923
82414
35000
60169
33274
26071
13185
34667
25000
17080
14417
50202
40662
22088
15295
29438
39924
54028
30980
31719
21875
38581
14778
28464
28405
17457
26886
54196
25170
24220
32390
37885
51786
17898
25096
25949
23981
15387
19943
6950
27829
40345
14126
46719
7446
30703
29196
18571
25571
7885
20365
10291
30291
10452
25860
7906
13444
51429 14107
14863
64412
18984
10417
10144
8689
47069
Memphis
OliveBranch
Bartlett
Collierville
Southaven
Lakeland
Arlington
Germantown
_
I240
_
I55
_
I40

Several examples will help to illustrate the patterns of economic segregation
that frequently plague our metropolitan areas. Map 1 shows the median
household income for neighborhoods in the Memphis metropolitan area, based
on the 2000 U.S. Census. The pattern is striking. The central business district is
an island, to the left side of the map, in a sea of neighborhoods with median
household incomes in the lowest category (below $25,000). Around the poor area
is a region of middle-income neighborhoods in the $25,000 to $50,000 range, with
a few neighborhoods in the $50,000 to $75,000 concentrated on the east side of
the city. Outside the boundary of the central city, the suburbs consist primarily
of neighborhoods in the highest income categories: $75,000 to $100,000 and those
above $100,000.

Downtown Dallas is featured in Map 2. Neighborhoods are sharply differentiated
by income level. For example, the independent political jurisdictions of Highland
Park and University Park, the home of Southern Methodist University, have
extremely high median household incomes, as do the neighborhoods to the north
of SMU. One neighborhood has a median household income above $200,000 and
several exceed $150,000. Just across sharp boundaries formed by highways and
major streets, there a solid blocks of middle class neighborhoods in the $25,000
to $50,000 range. A short distance away, just on the other side of downtown and
Dallas shows
extreme variations,
with household
incomes over
100,000 in one area
(near SMU), but
consistently below
25,000 in another
area south of I -30.
Map 2:
Downtown
Dallas
Median Household Income
29063
31961
31961
87834
56064
29908
51681
35375
27262
66629
20877
65901
70136
49071
8553
32659
10800
7094
30275
32375 41344
49312
43363
47389
47979
88101
19500
33386
42536
15625
76326
32277
76525
75797
149428
139177
120247
12685
111457
105352
172732
142199
60593
154606
200001
117433
29602
119653
49655
45152
97684
40465 86398
50160
28481
33365
61615
70929
76200
104601
184920
36250
61994
69591
19914
30341
51037
67337
50673
37667
24022
22308
20625
54069
71739
62025
52165
48594
31384
31875
33579
41809
31034
9824
32465
30929
84418
200001
25690
27500
26382
42781
26519
45625
52895
31065
26660
58929
26648
17768
25262
100807
40175
151032
31712
14811
41973
46234
37672
6250
51838
38653
25422
50020
102658
27843
52188
30266
35721
49094
14341
30164
40519
56912
31892
39946
25489
59444
39922
18176
42669
51442
25843
24576
38750
21403
32121
26525
62308
25182
37718
24880
31250
15817
26919
38946
40145 38521
48945
31336
Dallas
_
I30
_
I35E
_
I635
_
I45

south of Interstate 30, there is an extremely poor group of neighborhoods within

median incomes below $25,000, including a neighborhood with a median income
of $9,624. These neighborhoods, despite their close proximity, are virtually
different worlds. They look and feel quite different, they contain different
societies of people who distrust each other and rarely interact, and their children
attend very different schools.

St. Louis, shown in Map 3, is another prime example of economic segregation.
The city is divided into several distinct zones by income level, with very low-
income areas near the river both north and south of the Central Business District,
and across the Mississippi River in East St. Louis. Moving west in the Central
Corridor, neighborhood income levels rise and exhibit a high degree of
clustering. This pattern is typical of many large metropolitan areas, especially
those in the Northeast and Midwest.

Map 3:
St. Louis
Median Household Income
44352
13625
56310
30958
72500
52472
12660
27665
62523
63958
78522
0 89370
153366 89027
73472
90324
117381
28395
43531
47428
134028
48523
0
47383
43373
71996
22938
33243
23929
41024
35565
108090
39036
39509
65310
74079
66981
63493
45881
50653
41738
29886
57221
44339
41822
37500
50156
30777
75177
44782
36569
36322
53633
30861
19853
45436
31693
38219
32233
79167
64443
49026
103798
66290
34300
75471
39235
37006
52655
36098
43608
40290
74774
17191
34107
41879
42545
32083
44672
31168
60511
36583
32823
15412
26127
39762
79461
35824
68343
29087
49938
41054
36917
42479
75162
44550
30605
26833
37584
33318
34513
61806
31642
39602
61235
41013
37420
34543
66454
30245
32764
87969
70905
42490
43433
18346
51143
32167
31125
34826
40238
34912
46774
40880
58984
28687
50025
66602
28738
40058
29135
32124
40502 37993 39394
51091
32349
9290
37483
43616
16833
44449
31304
33307
60272
22040
25035
24911
64619
20292
19684
23750
37440
64419
21356
79467
49574
26167
48152
51635
52692
36498
64167 61678
50427
63750
23791
32480
28987
29033
23741
65568
39063
32135
73125
34167
42958
12872
58523
23038
34844
30703
33239
26964
27268
17477
35000
25925
33623
24207
37623
40455
22047
13580
30861
32800
32519
24375
29821
24035
17002
41392
30529
23836
17102
21794
25195
6875
8064
32932
18162
26020
21052
20041
21230
20701
28180 14827
17927
15403
24124
26392
21802
18434
25709
28227
20875
16571
18451
36716
17013
38311
20000
10145
34884
28401
38958
17219
27309
65169
28000
22446
21005
30353
17852
25850
27475
21575
16886
22337
31484
27656
32917
49201
28651
15509 24750
21952
22143
11543
12094
33883
22857
35536
65428
51150 49250
70542
78908
Ladue
Bridgeton
Maryland Heights
Kirkwood
Cahokia
Affton
Fenton
_
I270
_
I70
_
I44
_
I170
_
I255

Why Do We Care about Economi c Integrati on?

Economic segregation, and the harm that it does, has lived to some extent in the
shadow of racial segregation. Given the history of segregation by race in the
United States, a great deal of public policy concern focused on highly segregated
concentrations of African Americans in the nations central cities. This attention
was clearly needed, given the extreme deprivation, high crime and social
disorder in those neighborhoods, which contributed to highly destructive race
riots in the 1960s. In the aftermath of these riots, the Kerner Commission
appointed by President Johnson delivered a stunning indictment of the American
institutions, policies and practices that created and sustained the ghetto.

While
race was the primary organizing principal of the ghetto, the vast difference in
economic status between blacks and whites meant that racial concentration
created and exacerbated economic segregation. Race prejudice created the
ghettos but the concentration of poverty within them caused much of the harm.

In the years since, there has been slow but steady progress in reducing racial
segregation, and the levels in the fast-growing cities in the West and Southwest
are substantially lower than the older industrial cities of the Northeast and
Midwest. At the same time, levels of economic segregation have mostly been
increasing. As we strive for a post-ghetto world, we are unfortunately replacing
one form of segregation with another. There are a variety of reasons why policy
makers should not squander the gains of declining racial segregation by
allowing this substitution to take place.

When metropolitan areas are economically segregated, every problem in every
policy area becomes harder to address.

Low-income adults in highly

economically segregated metropolitan areas are less likely to come into contact
with employed persons who can act as mentors and provide information about
employment opportunities. Lower-income neighborhoods are usually located far
from the areas with the fastest growth of low-skill jobs. When low-income
neighborhoods are created by a process of out-migration of the middle class, as
they often are, the result is low-density neighborhoods with few role models.
Such neighborhoods are harder to police and often experience high levels of
crime, violence, gang activity and drug abuse.

Public health is worse in high-poverty neighborhoods for a variety of reasons.
First, health service providers are reluctant to locate in poor neighborhoods.
Second, the stress of dealing with high crime rates and the deteriorated
condition of the housing stock in such communities contribute to poor health
outcomes.

Poor communities often have to shoulder the fiscal burden of providing the
social and public health needs of low-income residents, increasing the tax
burden on businesses and the middle class. In turn, this discourages private
sector investment and encourages further flight to wealthier suburban
communities, increasing economic segregation. Declining or stagnant property
values in and near concentrations of poverty inhibit capital formation, shutting off
one of the key routes to social mobility.

As Figure 1 shows, family structure is very different in high-poverty
neighborhoods. Married-couple families are the dominant family structure in
neighborhoods with poverty rates of less than 40 percent. For neighborhoods
with poverty in excess of 40 percent, female-headed families are the dominant
family structure.

Similarly, as indicated in Figure 2, more than half of all males age 16 or older are
employed in lower-poverty neighborhoods, but in high-poverty neighborhoods,
the opposite is true.

Adolescents are particularly vulnerable to the effects of living in separate
communities. When adolescents live in communities where the modal family
type is a single mother with children, where the majority of men are not in the
labor force, and where college graduation is rare, they are more at risk of falling
into those behaviors themselves.

Perhaps most importantly, economic segregation has implications for education
because schools are creatures of neighborhoods. Often the relationship is
legally encoded in school attendance zone boundaries. Even for schools that
draw on larger and less precise areas, such as magnet and charter schools,
commuting time and transportation costs often restrict attendance to those who
live nearby. Schools usually closely reflect the racial and economic composition
of the surrounding community. When they do differ, public schools will tend to
have a greater proportion of minority and low-income children due to life cycle
differences and differential selection into private schools and home schooling.
Thus, schools are often even more segregated by race and income than the
surrounding community.

As a result, when poor families reside in different neighborhoods than middle-
and upper-income families, their children will likely attend different schools

than more affluent children. Over time, the schools themselves become
different. Schools in poorer neighborhoods have greater needs than schools with
more advantaged children. Teachers and school administrators may develop
lower academic expectations when they deal predominantly with poor children,
many of whom do not have resources and support in the home. There is also a
feedback effect, because school performance plays a role in parents decisions
about where to live, and the effect is greatest for the parents who have the most
resources to spend on housing and who place the greatest value on education.
The neighborhood drags down the school, and the school drags down the
neighborhood, ad infinitum.

Anyone who has been a teacher understands that learning is, in part, a group
process. Students bring their own skills and abilities, but they learn from and
with their fellow students. Children, especially adolescents, want to fit in, so
they may adopt the prevailing attitude towards the value of academic
achievement. In some inner-city schools, working hard and getting good grades
is derided as acting white. Even students who resist caving in to peer pressure
may still be impeded in learning if enough classmates are disruptive and slow
the pace of instruction. These so-called peer effects on students have been
documented in a number of carefully controlled studies.

Over and above peer effects, neighborhood conditions may have spill over effects
on academic achievement. Neighborhoods develop cultures that in turn
influence their members. Students may lack positive role models if few
members of a community graduate from college or have steady employment in
high-skilled occupations that require education. The noted sociologist William
Julius Wilson wrote that children in such communities may come to question the
value of education if they seldom interact on a sustained basis with people who
are employed.

Parents are not immune to neighborhood influences either. It

may be harder for parents to keep pressure on their children to achieve
academically if other parents in the neighborhood do not demand the same effort
from their childrens friends and classmates.

Economic integration ameliorates all of these effects. When there is less
difference between neighborhoods in economic status, the quality of public
goods and economic opportunities available to any given individual is less
dependent on that persons own economic status. The conflict between local
jurisdictions within a metropolitan area is also reduced if there is less of a
disconnect between social needs and fiscal capacity.

Measuri ng Economi c Integrati on

Economic integration can take many forms. It is highly unlikely that economic
integration will be so complete that every neighborhood has exactly the same
mean income. However, every neighborhood can contain a mixture of housing
types of varying proportions. In that case there would still be some economic
differentiation between neighborhoods, but not the kind of extreme variation
shown in Maps 1-3. Economic integration could also operate at a higher scale, so
that there was still great variation among census tracts, but rich and poor
census tracts were not as clustered as those shown in those maps. In that case,
you might still have rich and poor neighborhoods, but they would not also be
located in separate political jurisdictions, dependent on different tax bases, and
physically remote from one another.

Economic integration, stated simply, means that households of all different
income levels have realistic housing options in all parts of a metropolitan area,
so that they can pursue opportunities throughout the region.
Household I ncome I ntegrati on
A previous CEOs for Cities report, City Vitals, presented a measure of economic
integration based work by the Lewis Mumford Center at the State University of
New York at Albany.

The measure uses Census data from the 2000 Census.

There is no national definition of what constitutes a neighborhood, but the
Census Bureau has divided the nation up into parcels known as census tracts,
which average 4,000 in population and can serve as a proxy for neighborhoods.
Within each neighborhood, they report the number of households with household
income in 1999 in categories, such as less than $10,000, $10,000 to $14,999, and so
on up to $200,000 and higher.

Given that there are hundreds of neighborhoods in many large cities and sixteen
different income brackets, some way must be found to reduce all the information
to a single number representing the degree of economic integration in a
metropolitan area. The Mumford Institute created three groups: low income
(less than $30,000 in household income), middle income ($30,000 to $59,999),
and high income ($60,000 or more). They then computed an Index of
Dissimilarity between the low-income and high-income groups. (The Index of
Dissimilarity measures economic segregation by reporting the share of persons
in one group who would have to move to achieve an even distribution of the
groups across neighborhoods.) City Vitals inverts the scores so that they
measure economic integration, i.e., the proportion of persons who would not
have to move to achieve an even distribution of the three groups.

We recomputed this measure from the raw Census data for all metropolitan
areas in the United States.

Economic integration for the top 50 metropolitan

areas is shown in Figure 3. In the least segregated areas such as Nassau-
Suffolk, San Francisco and San Jose, fewer than three out of 10 households
would have to move to create a perfectly even distribution of households of
different incomes across neighborhoods. In the most economically segregated
areas, such as Austin, San Antonio, Denver and Houston, nearly half of all
households would have to move to obtain perfect economic integration. The
national average economic integration score for all 331 metropolitan areas was 66
percent, meaning that one third of all households would have to move to achieve
an even distribution of high- and low-income households.

There is not an immediately obvious pattern. It is true that Portland, which
actively works to contain sprawl and manage growth, is one of the most
economically integrated cities. And cities like Dallas, Houston and Phoenix, with
unbridled suburban sprawl and a fondness for gated communities, are among
the least economically integrated. Other economically integrated cities have
geographic constraints on growth such as San Francisco and Boston, which
force infill development and redevelopment of older areas. But the high-ranking
cities also include San Jose and Nassau/Suffolk, which seems counterintuitive.

This economic integration measure, patterned after the City Vitals economic
integration score, is based on the application of fixed income levels, regardless
of differences in relative wealth and differences in cost of living among
metropolitan areas. The cut off values used by the Mumford Center $30,000
and $60,000 have very different meanings in New York than in Memphis or
Buffalo. The proportion of the households with incomes below $30,000 ranges
from less than 20 percent in San Jose to more than 40 percent in New Orleans.
Similarly, the proportion with household incomes of $60,000 or more is over 60
percent in San Jose and about 25 percent in New Orleans.

With such wide variations in the relative position of the two income groups, we
decided to develop a new measure of relative economic integration that takes into
account the differences among metropolitan areas in their income distributions.
The procedure is as follows:

First, we calculated the dollar values that divided the income distribution into
three equal sized groups within each metropolitan area. These are shown in
the Appendix. Note that the boundary between the middle and upper thirds
in New Orleans is less than the boundary between the lower and middle
thirds in San Jose.
Second, we applied these thresholds to each census tract within a
metropolitan area to get the count of relatively poor and relatively rich
households in each neighborhood. Because the threshold often was in the
middle of one of the household income brackets given in the data, we use
linear interpolation to allocate the households to the appropriate groups.
Third, we calculated the Relative Economic Integration score as 1 minus the
Index of Dissimilarity between the upper third and the lower third of
households within each metropolitan area. It is the proportion of households
that would not have to move to achieve a perfectly even distribution of
household income across neighborhoods.

Table 1 shows the result of the analysis. The economic integration scores and
relative rankings do change, although for the most part the differences are small

Rank Metropolitan Areas

Relative
Economic
Integration Diffrence
Rank
Change
1 Nassau/Suffolk, NY 72.9% -2.4% 0
2 Greensboro/Winston-Salem/High Point, NC 69.1% 1.0% -2
3 Portland/Vancouver, OR/WA 67.4% 0.1% -2
4 San Jose, CA 67.1% -4.6% 2
5 San Francisco, CA 66.9% -2.1% 2
6 Pittsburgh, PA 66.3% 0.9% -3
7 Seattle/Bellevue/Everett, WA 65.0% -0.7% 0
8 Tampa/St. Petersburg/Clearwater, FL 65.0% 2.1% -7
9 Orlando, FL 65.0% 1.1% -1
10 Boston, MA/NH 64.4% -2.7% 4
11 Nashville, TN 63.3% 0.3% -3
12 Providence/Fall River/Warwick, RI/MA 63.1% -0.7% 1
13 Charlotte/Gastonia/Rock Hill, NC/SC 63.0% -0.1% 0
14 Raleigh/Durham/Chapel Hill, NC 62.3% -0.6% -2
15 Bergen/Passaic, NJ 62.3% -3.3% 7
16 Fort Lauderdale, FL 62.0% -0.2% -1
17 St. Louis, MO/IL 61.3% 0.0% -2
18 Minneapolis/St. Paul, MN/WI 61.2% -0.2% 0
19 New Orleans, LA 61.0% -0.3% -1
20 Salt Lake City/Ogden, UT 60.8% 1.2% -10
21 Sacramento, CA 60.5% -0.2% 0
22 Riverside/San Bernardino, CA 60.4% -0.2% 0
23 Miami, FL 60.4% 0.6% -5
24 Cincinnati, OH/KY/IN 60.3% -0.2% -2
25 Norfolk/Virginia Beach/Newport News, VA/NC 59.8% 0.9% -8
26 Indianapolis, IN 59.8% 0.0% -1
27 Cleveland/Lorain/Elyria, OH 59.6% -0.1% -2
28 Atlanta, GA 59.6% -1.0% 4
29 Detroit, MI 59.5% -1.0% 6
30 Chicago, IL 59.3% -1.2% 5
31 Orange County, CA 59.2% -4.0% 19
32 Fort Worth/Arlington, TX 58.4% 0.3% -4
33 San Diego, CA 58.4% -1.0% 2
34 Kansas City, MO/KS 58.3% 0.3% -4
35 Los Angeles/Long Beach, CA 58.1% -1.0% 3
36 Las Vegas, NV/AZ 57.7% 0.6% -4
37 Milwaukee/Waukesha, WI 56.9% -0.1% -4
38 Washington, DC/MD/VA/WV 56.6% -1.8% 4
39 New York, NY 56.4% -1.6% 2
40 Philadelphia, PA/NJ 56.4% -1.2% 1
41 Columbus, OH 56.1% 0.1% -2
42 Phoenix/Mesa, AZ 56.1% 0.1% -2
43 San Antonio, TX 55.9% 1.2% -6
44 Oakland, CA 55.4% -2.9% 9
45 Baltimore, MD 55.0% -0.6% -1
46 Dallas, TX 54.5% -1.1% 1
47 Houston, TX 54.4% -0.9% 0
48 Denver, CO 54.2% -0.5% 0
49 Austin/San Marcos, TX 53.5% -0.7% -1
50 Newark, NJ 53.1% -3.1% 8
Table 1: Relative Economic Integration of the 50 Largest Metropolitan Areas, 2000
relative to the gap between the most and least economically integrated
metropolitan areas. The changes in the economic integration score range from -
4.6 percentage points to +2.1 percentage points. The Orange County MSA drops 19
places, from 12
th
to 31
st
place. Boston drops from 6
th
to 10
th
. Salt Lake Citys rank
rises 10 places.

The contrast between the absolute and relative Economic Integration scores
demonstrates that the metric used to measure economic integration clearly
matters. For that reason, we provide several additional measures of economic
segregation that provide alternative perspectives.
Segregati on of the Poor f rom the Non-Poor
From a public policy perspective, we are particularly concerned with the extent
to which the spatial organization of neighborhoods affects the poorest members
of the community. The federal poverty definition classifies a person as poor if
they live in a household with income below a specified threshold level that is
adjusted for family size and is updated every year for inflation. For example, in
2008 the poverty level for a family of four is $21,200. Unfortunately, the federal
poverty measure is not adjusted for local differences in the cost of living, and it
does not take into account in-kind benefits such as Medicaid and Food Stamps.
With all its faults, it is still the most widely available measure of poverty and is
familiar to researchers and policy makers.

One consequence of economic segregation is that the poor mainly come into
contact with other poor persons. For that reason, we have computed several
segregation measures indicating the segregation of the poor from the non-poor.
They are the Index of Dissimilarity (D) described above and three additional
measures: the Variance Ratio Index (V), the Gini Index of Segregation (G) and the
Theil Index (T). In contrast with the Economic Integration measure presented
above, these segregation measures emphasize the spatial distribution of those at
the lower end of the income distribution.

These figures, and the ranks of the 50 largest metropolitan areas under each
measure, are presented in Table 2. With a few small exceptions, the four
different measures rank the metropolitan areas very similarly. However, the
focus on the segregation of the poor from everyone else, compared to low-
income from high-income, has a large effect on the outcomes. Portland has
little segregation of the poor from the non-poor regardless of the measure used.
Philadelphia and Milwaukee consistently have the worst segregation of the poor.
Yet Milwaukee was only ranked 37
th
on the Relative Economic Integration
measure. Milwaukee falls from 18
th
place to 42
nd
(on D). Other metropolitan
areas that look substantially worse when the focus is on the poor include
Cleveland, Detroit and Chicago. Some cities look more integrated (or,
equivalently, less segregated) by the poor/non-poor measures, such as Los
Angeles, Los Vegas and Miami.

Table 2: Segregation of the Poor from Nonpoor, 50 Largest Metropolitan
Areas, 2000: Dissimilarity (D), Gini (G), Variance (V), and Thiel (T)
Metropolitan Area D G V T D G V T
Portland, OR/WA 0.27 0.38 0.05 0.07 1 1 3 1
Orlando, FL 0.29 0.41 0.06 0.08 2 4 7 6
Nassau, NY 0.29 0.40 0.03 0.07 3 2 1 2
Seattle, WA 0.29 0.41 0.05 0.08 4 5 4 4
Charlotte, NC/SC 0.30 0.42 0.06 0.08 5 10 6 7
San Jose, CA 0.30 0.41 0.04 0.07 6 3 2 3
Miami, FL 0.30 0.41 0.09 0.09 7 6 19 11
San Francisco, CA 0.30 0.42 0.05 0.08 8 7 5 5
Tampa, FL 0.30 0.42 0.07 0.09 9 8 11 8
Greensboro, NC 0.30 0.42 0.07 0.09 10 9 10 10
Fort Lauderdale, FL 0.31 0.42 0.07 0.09 11 11 12 9
Riverside, CA 0.31 0.43 0.08 0.09 12 12 14 12
Las Vegas, NV/AZ 0.33 0.44 0.07 0.09 13 13 8 13
Pittsburgh, PA 0.33 0.46 0.09 0.11 14 15 22 17
Los Angeles, CA 0.34 0.45 0.10 0.11 15 14 34 16
Salt Lake City, UT 0.34 0.47 0.07 0.10 16 18 9 14
Nashville, TN 0.34 0.47 0.09 0.11 17 17 27 20
Raleigh, NC 0.34 0.47 0.09 0.11 18 19 26 21
San Diego, CA 0.35 0.47 0.09 0.11 19 19 23 19
San Antonio, TX 0.35 0.47 0.09 0.11 20 16 24 18
Orange County, CA 0.36 0.47 0.07 0.10 21 21 13 15
Houston, TX 0.36 0.48 0.09 0.11 22 22 29 22
Atlanta, GA 0.36 0.49 0.10 0.12 23 25 32 28
Fort Worth, TX 0.36 0.49 0.08 0.11 24 23 17 23
New York, NY 0.36 0.49 0.12 0.13 25 24 40 32
Sacramento, CA 0.37 0.50 0.10 0.12 26 26 31 25
New Orleans, LA 0.37 0.50 0.14 0.13 27 27 47 37
Dallas, TX 0.37 0.50 0.09 0.12 28 28 25 26
Wash., DC/MD/VA/WV 0.37 0.50 0.09 0.12 29 30 18 27
Norfolk, VA/NC 0.37 0.51 0.11 0.13 30 33 36 35
Indianapolis, IN 0.37 0.50 0.08 0.11 31 29 15 24
Oakland, CA 0.38 0.51 0.09 0.12 32 31 28 30
Boston, MA/NH 0.38 0.51 0.09 0.12 33 32 20 28
Bergen, NJ 0.39 0.52 0.09 0.13 34 35 21 33
Denver, CO 0.39 0.53 0.08 0.12 35 36 16 31
Kansas City, MO/KS 0.39 0.53 0.09 0.13 36 38 30 36
Cincinnati, OH/KY/IN 0.39 0.54 0.13 0.15 37 39 43 42
Phoenix, AZ 0.40 0.53 0.11 0.14 38 37 37 38
St. Louis, MO/IL 0.40 0.54 0.12 0.14 39 40 38 39
Providence, RI/MA 0.40 0.52 0.11 0.13 40 34 35 34
Austin, TX 0.41 0.54 0.12 0.15 41 41 39 41
Minneapolis, MN/WI 0.41 0.55 0.10 0.15 42 43 33 40
Columbus, OH 0.41 0.55 0.12 0.15 43 42 41 43
Baltimore, MD 0.43 0.57 0.13 0.16 44 44 44 44
Chicago, IL 0.44 0.58 0.14 0.17 45 46 46 46
Detroit, MI 0.45 0.58 0.13 0.16 46 45 45 45
Cleveland, OH 0.45 0.59 0.15 0.17 47 47 48 48
Newark, NJ 0.46 0.59 0.13 0.17 48 48 42 47
Philadelphia, PA/NJ 0.46 0.59 0.15 0.18 49 49 49 49
Milwaukee, WI 0.51 0.64 0.17 0.21 50 50 50 50
Rank Score

0 10 20 30 40 50
Chicago, IL
Norfolk, VA/NC
Miami, FL
Los Angeles, CA
Cleveland, OH
Cinci nnati, OH/KY/IN
Philadelphia, PA/NJ
Milwaukee, WI
New Or leans, LA
New York, NY
All Poor Black Poor
Hispanic Poor Non-Hi spanic White Poor
Concentrati on of Poverty
Policymakers are particular concerned with the most extreme form of
segregation of the poor: high-poverty ghettos and barrios. Such neighborhoods
are dominated by housing projects and vacant housing units and often suffer
high levels of crime and violence. The formation of such neighborhoods is driven
by high overall poverty levels and the intersection of racial and economic
segregation. One way to measure the concentration of poverty is the percentage
of the poor residents of a given metropolitan area who live in extremely poor
neighborhoods, defined as census tracts with poverty rates of 40 percent or
more.

Figure 4 shows the 10 metropolitan areas of the top 50 with the highest
concentrations of poverty. In New York, for example, one fourth of the poor live in
high-poverty ghettos and barrios. In low economic segregation metro areas such
as Portland and San Francisco, fewer than 2 percent of the poor reside in such
neighborhoods. Also shown in the figure are the concentration of poverty levels
for the Black poor, the Hispanic poor, and the Non-Hispanic White poor. The
differences among the groups are striking. The Black Poor are substantially
more concentrated than the Hispanic poor, except in Philadelphia and New York.
The Hispanics in those areas are predominantly Puerto Ricans, compared to
Mexicans in the other metro areas depicted. The White poor almost never live in
high-poverty neighborhoods, except in New York, which is affected by a higher
than normal presence of white immigrants from central Europe.

Metropoli tan Areas wi t h t he Hi ghest Concentrations of Poverty

The Met ropoli tan Context of Economi c Integrati on

Metropolitan development patterns influence the ability of communities to
achieve economic integration. Mixed-income communities are easier to develop
in metropolitan areas with strong housing markets where developers have
incentives to develop at higher levels of density to accommodate affordable
workforce housing. In metropolitan areas with weak housing markets it is more
difficult to entice developers to build affordable housing and more affluent
households choose to segregate themselves in new single-family developments
on the suburban fringe.

The strength of metropolitan housing markets depends on the balance of the
supply and demand for housing. The supply of housing is determined by public
policies that either subsidize or constrain new development and availability of
developable land. The demand for housing is driven by job growth and
population growth, often driven by immigration.
Weak
Market
Strong
Market
Las Vegas
Atlanta
Chicago
St. Louis
Portland, OR
San Francisco
Pittsburgh
Cleveland
New Orleans
D
e
m
a
n
d

f
o
r

H
o
u
s
i
n
g

Supply of Housing
Uncontained
Suburban Growth
Contained
Suburban Growth

Declining
Population
Increasing
Population
(Immigration)
Stable

As Figure 5 illustrates, American metropolitan areas vary in the ratio of supply

and demand, with weak market metros having a surplus of housing and strong
market regions having a shortage of housing. (Metropolitan areas are placed in
the graph for illustrative purposes only, not based on any quantitative
determination.)

The strength of housing markets also varies within metropolitan areas. A
metropolitan area can have an overall strong housing market but the market can
be very weak in some central city neighborhoods with high poverty rates.
Conversely, even weak metropolitan housing markets have islands of renewal in
inner urban areas that policymakers can build upon to promote economic
integration. In any case, local leaders need to understand both the general
metropolitan context and the specific neighborhood contexts in order to choose
the best approach for promoting economic integration.
Weak Market Metropoli tan Areas: Excess Supply
Weak housing markets are created by an excess of supply and/or a shortage of
demand for housing. The United States has supported a range of policies that
have subsidized an oversupply of single-family housing for middle and affluent
households in the suburbs. This has led inevitably to concentrated poverty and
weak housing markets in central cities and inner suburbs. Metropolitan areas
with suburban sprawl and weak housing demand due to job loss and population
out-migration face a double whammy that makes building mixed-income
communities all that more difficult.

In most countries, the rich live in historic housing near the urban center while
the poor are relegated to newer housing on the urban fringe. The historic
American pattern has been the opposite: the poor generally live in older housing
near the center and higher- income households occupy new housing in the
suburbs. The main reason for this is that the United States has spent
considerably more on homeowner tax deductions, highways and other subsidies
of suburban development than it has spent on new housing for the poor. Less
than a third of poor households receive any form of housing subsidy and those
subsidies have at best not been effective in de-concentrating poverty, and at
worst have exacerbated the problem. In effect, the de facto low-income housing
policy in the U.S. has been to build new housing for the affluent in the suburbs
and let their old housing filter down to the poor. The result of the housing
filtering chain, or what some call the thrift shop approach to affordable
housing, has been economic segregation.

The housing filtering chain did help low-income Americans improve the physical
attributes of their housing. Overcrowding is down and most poor households

now have running water and other basic amenities. The problem with the
American pattern of subsidizing the construction of new single-family homes in
the suburbs is that it has been a mighty engine of economic segregation.
Indeed, over time, suburban development has become even more oriented
toward the secession of the successful, to use Robert Reichs felicitous phrase.
New housing construction has always been geared to the affluent but this
tendency grew more pronounced in recent decades. After World War II
developers built relatively inexpensive housing in Levittown-like developments
for returning veterans. They continued this pattern of building relatively modest
suburban homes until the 1960s and 1970s. After that, new home construction
shifted from the middle of the income distribution into the top of the income
distribution. Between 1950 and 2000 the average size of a new home increased by
more than 50 percent (from 1,470 sf to 2,265 sf).

In 2000, the average new house

was almost two-thirds more expensive than in 1960 (in constant dollars), and the
share of new housing purchased by the top 20 percent of the income range
increased dramatically. McMansions, large (3,000 to 5,000 square feet) mass-
produced homes in suburban subdivisions, have become more and more
common. The correlation between new housing and economic segregation is
strong: the newer the housing in a neighborhood, the higher the average income
in that neighborhood.

The demand for housing in the core parts of the region is determined not only by
immigration and population growth but by the attractiveness of the central city.
One of the main causes of weak housing markets in central cities is job
decentralization. In weak market areas, jobs have decentralized and
suburbanization is driven in part by a desire to be closer to jobs. In a study of the
100 largest metropolitan areas, extremely decentralized metro areas with less
than 10 percent of jobs located within 10 three miles of the Central Business
District included Detroit, St. Louis, Riverside and Gary, Indiana.

Source for 85 metropolitan areas: Larry C. Ledebur and William R. Barnes, "City Distress, Metropolitan Disparities, and
Metropolitan Growth," comb. rev. ed., Research Report of the National League of Cities (Washington, D.C.: National
League of Cities, September 1992), p. 2, and authors' calculation for 2000.
The decanting of the middle and upper classes from the cities to the suburbs is
the main reason that the income of central cities fell behind the income of
suburbs. This also created inner suburban decline. A study of the thirty-five
largest metropolitan areas found that 37 percent of all suburbs actually lost
population in the 1990s.

Another study of over 5,000 suburbs found that the

proportion of suburban residents living in poor suburbs more than doubled
from 1980 to 2000 (from 8.4 to 18.1 percent).

Suburban sprawl becomes especially damaging in metropolitan areas with weak
job and population growth. Weak market cities tend to be located in the Midwest
and Northeast, and include metropolitan areas like Buffalo, Scranton, Cleveland,
Akron, Youngstown, Detroit, Flint and Milwaukee. In these metropolitan areas,
the housing market becomes especially weak because the supply of housing far
outstrips demand. Metropolitan areas in the bottom third of job and population
growth from 1960 to 2000 actually built housing at six times the rate of population
growth.

Table 3 lists the top ten metropolitan areas in the ratio of building
permits to new households in the 1990s. During this period, Buffalo actually
issued 3.89 building permits for every one new household.

This massive
oversupply of housing led inevitably to abandonment and concentrated poverty in

the urban core. The correlation between excess home construction and central
city decline is strong.

Table 3. Top Ten Metro Areas, Ratio of Bui l di ng Permi ts to Househol d
Growt h, 1990-2000

Rank Name Permits Household Change Ratio

1. Buffalo-Niagara Falls, NY MSA 26,881 6,916 3.89
2. Pittsburgh, PA MSA 55,936 19,252 2.91
3. Scranton-Wilkes-Barre 13,462 5,331 2.53
Hazelton, PA MSA
4. Youngstown-Warren, OH MSA 15,505 6,613 2.34
5. Dayton-Springfield, OH MSA 33,888 15,326 2,21
6. Hartford, CT MSA 67,227 37,660 1.79
7. St. Louis, MO-IL MSA 109,944 64,650 1.70
8. Syracuse, NY MSA 16,222 9,627 1.69
9. Cleveland-Lorain-Elyria, OH 70,718 47,376 1.49
PMSA
10. Gary, IN PMSA 30,304 20,375 1.49

Source: Thomas Bier & Charlie Post, Vacating the City: An Analysis of New Home Construction and Household Growth, in
Redefining Urban & Suburban America: Evidence from Census 2000, edited by Alan Berube, Bruce Katz & Robert E. Lang
(Washington, D.C.: Brookings Institution Press, 2006), p. 185.

In weak market metropolitan areas the outward movement of middle and
affluent households and rising concentrated poverty creates a reinforcing cycle
of decline. The desire of affluent households to separate themselves from the
poor artificially drives new home construction. In other words, in areas with wide
income inequalities and uncontained suburban sprawl, the desire of the middle
class to separate from problematic areas of concentrated poverty artificially
inflates the demand for new housing on the urban fringe. Excess housing
construction causes abandonment, undermining the ability of inner urban areas
to jump-start their housing markets. As the poor become more concentrated in
areas with vacant and abandoned homes, their social problems worsen and
upper-income households have even more motivation to separate themselves. In
weak market sprawling metropolitan areas, a vicious cycle of suburban flight
and inner urban decay is set in motion.

Efforts to increase economic integration must be adapted to conditions in weak
market cities. The policy tools must be powerful enough to counter the negative
trends that are undermining housing markets in poor areas. Strategies must
face the realities of consumer choice and build the market for working and
middle class households. The recent rash of foreclosures only makes this task

all the more daunting. In areas with weak markets, foreclosures often lead to
vacant and abandoned properties, which can cripple the local housing market.
Cities need to prevent foreclosures or break the link between foreclosures and
vacancy and abandonment by helping CDCs to purchase, rehabilitate, and market
the properties or land bank key properties as part of a strategy for mixed-income
revitalization. Above all, cities need to recognize, given the regional context of
population stagnation and surplus housing construction, that they need to right
size themselves, strategically investing in selected neighborhoods that can
attract a socio-economic mix.

Even weak cities have areas with strong

amenities such as parks, pedestrian spaces or public transit that can be
leveraged to market housing to middle- and working-class households.
Increasingly weak market cities are recognizing that a bricks and mortar
approach to urban revitalization is not enough and, in fact, can backfire. In
metropolitan areas where new home construction greatly exceeds household
formation, CDCs are put in the position of running up the down escalator. An
evaluation of CDCs in Cleveland, for example, concluded that every new unit of
housing added to the supply in one neighborhood may only result in
abandonment somewhere else in the city. As a result, Neighborhood Progress
Inc. (NPI), an intermediary which provides technical and planning support for
CDCs, has embarked on an ambitious Strategic Investment Initiative that
concentrates funding in six neighborhoods viewed as having a reasonable chance
of becoming stable, mixed-income neighborhoods.

Housing investments are

linked to investments in schools, open space, transportation and other anchors in
order to make these areas regional neighborhoods of choice.

This initiative
recognizes that in a shrinking city like Cleveland not all neighborhoods can be
revitalized and policymakers must make tough choices to concentrate resources
in neighborhoods that have inherent strengths that can be built upon.
Strong Market Metropoli tan Areas: Ri si ng Demand,
Li mi ted Supply
One of the best things the federal and state governments could do to encourage
greater economic integration would be to eliminate policies that subsidize
suburban sprawl and encourage affluent households to meet their housing
demand by building new homes on the suburban fringe. Smart growth and fix-
it-first policies that encourage infill development turn housing demand inward
toward the urban core, increasing the possibilities for stable integrated
neighborhoods. Smart growth aids economic integration efforts and protects the
natural environment.

In strong housing markets, prices increase rapidly, outpacing incomes. Table 4
shows the top ten metropolitan areas in the nation in terms of housing
affordability, defined as the ratio between an affordable house (costing no more

than three times the 2006 median income). High housing prices occur for two
basic reasons either escalating demand outstrips supply or the supply of new
housing is constrained and cannot keep up with demand. In the Los Angeles-
Long Beach-Santa Ana metro area, the average sales price is more than three
times the affordable home price.

TABLE 4. Ten strongest metropoli tan housi ng markets, Ratio of
Affordable Home Pri ce to Actual Sales Pri ce, 2006
R Name Affordable Home Actual Sales Ratio Price

1. Los Angeles- $183,900 $584,800 .31
Long Beach-Santa Ana, CA
2. San Diego-Carlsbad- $194,700 $601,800 .32
San Marcos, CA
3. Honolulu, HI $213,900 $630,000 .34
4. San Francisco-Oakland- $258,900 $736,800 .35
Fremont, CA
5. San Jose-Sunnyvale- $291,300 $775,000 .38
Santa Clara, CA
6. New York-Wayne- $212,700 $539,400 .39
White Plains, NY-NJ
7. Riverside-San Bernadino- $172,500 $400,100 .43
Ontari, CA
8. Miami-Fort Lauderdale- $167,700 $371,200 .45
Miami Beach, FL
9. NY: Neward-Union, NJ-PA $212,700 $433,000 .49
10. Sacramento-Arden- $196,200 $374,800 .52
Arcade-Roseville, CA
Source: Based on calculations be David Rusk using National Association of Realtors data on sales prices and HUD Area
Median Income (AMI)
The main reason for excess housing demand is a strong economy that fuels
population growth. Attracted by jobs, immigrants fuel population growth. Many
cities in the South and the West fit this pattern, including Miami, Tampa-St.
Petersburg, Dallas, San Antonio, Las Vegas and San Diego. In every one of these
metropolitan areas, the central city gained population from 2000-2006.

Housing demand is also turning inward, boosting markets in inner urban areas.
A major reason for this is the clustering of jobs there. Many jobs, especially the
higher paid, skilled information processing jobs, need to locate in dense clusters
that encourage innovation and the sharing of cutting-edge information. In dense

employment metro areas, such as New York, San Francisco and Portland, Ore.,
more than 25 percent of the jobs in the region are located within three miles of
the Central Business District.

As people move out to the suburban frontier,

they are confronted with longer commutes, and urban living looks more
attractive. In addition, the rising price of energy is making long suburban
commutes and large, detached single-family homes more expensive. During
the recent run-up in energy prices, homes in the urban center did better than
homes on the suburban fringe.

This will happen again when energy prices

rise. People are being pulled to the center by the attractions of energy-saving,
mixed-use, pedestrian friendly, more compact forms of development.

The other reason for strong housing markets is that the supply of new housing
on the urban fringe is contained by natural barriers, market dynamics or
intentional policies. When suburban sprawl is contained, whether by natural
barriers or deliberate policies, upper-income households have no choice but to
search for housing near the center, often competing with the poor in gentrifying
neighborhoods. Natural barriers that limit suburban sprawl include oceans,
mountains and protected wetlands. Caught between the Atlantic and the
Everglades, some Florida metropolitan areas have been forced to grow at higher
rates of density. Hemmed in by the Pacific Ocean and the desert and attracting
huge numbers of immigrants, Los Angeles is now one of the densest
metropolitan areas in the country.

Public policies also restrict new suburban construction, including restrictive
zoning laws, such as minimum lot requirements and open space preservation.
Federally owned land is another obstacle to expansion. At a certain point,
developable suburban land within a reasonable commuting distance of jobs is
exhausted. The New York, Boston, Seattle and San Francisco metropolitan areas
all qualify as regions where the growth of the housing stock on the urban fringe is
constrained. Instead of the rich occupying new housing on the urban fringe,
increasingly, they compete for housing in gentrifying areas of the urban core. In
these metropolitan areas, the segregation machine of suburban sprawl has
ground to a halt.

In most cases, urban containment is the result of a series of uncoordinated
policies, market forces and natural limits. But a few metropolitan areas have
implemented intentional policies to contain sprawl and force development back to
the center. Urban containment policies include requiring that infrastructure be
in place before development occurs and various growth management policies,
such as public land purchases. The most powerful form of urban containment is
a greenbelt. A greenbelt is a ring around a metropolitan area that restricts
development to the area inside the urban growth boundary. European cities are
famous for their greenbelts. When you leave London, for example, you will see a

clear boundary between the city and the countryside. In most American cities the
dividing line between city and country is blurred. Many Americans have adopted
open country living, building homes on lots of 5, 10, or more acres on the exurban
fringe. A driver approaching most metropolitan areas in the U. S. gradually sees
homes sprinkled across the countryside before noticing the first suburban
development. The income segregation is often very explicit with signs declaring
Homes starting in the $300s.

Portland, Ore., is an exception. As you drive into Portland you go through what
looks like a nature preserve until you cross the urban growth boundary and
relatively dense urban development suddenly appears. When first implemented
in 1979, Portlands UGB was quite loose. It took 20 years for development to reach
it. But it is clear now that the UGB has had a major impact on development.
Before the UGB Portland was suburbanizing at a rapid rate and consuming land
at a rate way beyond its population growth. Since 1979, Portland has become one
of the rare American cities where the consumption of land has not exceeded
population growth.

The evidence is strong that urban containment policies reduce segregation. One
study found that 10 years of strong urban containment reduces the level of racial
segregation by an additional 1.4 percent than would have occurred without
containment policies.

According to our measure, Portland, which has the

strongest urban containment policies in the nation, is one of the least
economically segregated metropolitan areas in the nation. In addition, the central
city of Portland has done well relative to the suburbs, with the per capita income
of the central city at 96 percent that of the suburbs in 2000.

However, the UGB has not been without controversy. One critique of the UGB is
that it has increased housing costs. In fact, the UGB has been loosened over the
years. According to law, the UGB must allow for 20 years worth of growth.
Scholars disagree about the effect of the UGB on housing prices. The consensus
seems to be that it has had a moderate impact on housing prices, perhaps $10,000
per house. Greenbelts increase land costs but more compact development also
saves money on infrastructure and city services.

One thing is clear: Portlands

UGB has disrupted the pattern of meeting demand for high-end housing by
building new single-family homes on large lots on the urban fringe. In Portland,
this demand is met in all sorts of ways, including luxury condos in the central city.

Urban containment policies are not the sole answer to rising economic
segregation for two reasons. First, urban containment policies face entrenched
political opposition. With a small minority presence, Portland did not have to
overcome racial fears. Assembling a unique coalition of urban and agricultural
interests, Portland drew its growth boundary early and far out, not threatening

existing landowners. Once the snowball of suburban sprawl gathers

momentum, it is difficult to stop with local policies. Powerful land interests and
homebuilders on the suburban fringe are able to veto strong urban containment
policies.

Second, even if a metropolitan area is able to enact strong urban containment
policies, there is no guarantee that it will increase economic integration. Polices
that disrupt the pattern of building high-income housing on the urban fringe
provide an opportunity to reduce economic segregation but do not guarantee it. In
Portlands case, a series of policies, including accelerated approval of housing
permits, massive investments in public transit and a requirement that 20 percent
of all new housing be affordable for low- and moderate-income households,
ensured that new housing would include a mix of incomes.

Even though strong urban containment policies are politically difficult, urban
containment will occur spontaneously over time through spontaneous market
processes and physical limits. Rising traffic congestion, longer commuting
times and higher gasoline prices will limit how far commuters will be willing to
drive for new homes in the suburbs. Suburban zoning laws will restrict the
supply of housing in suburbia and affluent households will be forced to turn to
the urban core to meet their demand for high-end housing. The gradual closing
of the suburban frontier will open up opportunities to promote economic
integration. But urban containment and strong urban housing markets will not
automatically reduce economic segregation. Without the political will to enact
policies that support economic integration, we could end up copying the worst of
the European pattern, with the middle and upper classes dominating in the
historic urban cores and the poor relegated to depressing suburban ghettos. We
need innovative and flexible policies to ensure that when metropolitan areas turn
inward to accommodate growth they do not sort the affluent into gated enclaves
and the poor into ghettos, barrios and slums. At the same time, we need policies
appropriate for weak market metropolitan areas still characterized by suburban
sprawl and central city decline.
Poli cy Approaches to Economi c Integrati on
There are many different ways to promote economic integration. One way would
be to lift up the incomes of people living in poor areas through job training or
economic development initiatives. The movement of jobs, especially entry-level
jobs, to remote suburban areas is an important cause of urban poverty
concentrations and promoting jobs and economic development in poor areas is a
crucial way to promote economic integration. In this paper we focus on housing
policies but it is important to understand that housing policies must be

coordinated with jobs and economic development policies if we are to promote

sustainable economic integration.

Table 5. Policy Approaches for Promoting Economic Integration
LAND USE AND
PLANNING
REGIONALISM SUBSIDIES FOR
AFFORDABLE
HOUSING IN
HIGHER INCOME
AREAS
DISCOURAGING
FLIGHT FROM
DECLINING OR
GENTRIFYING
NEIGHBORHOODS
Removing
exclusionary zoning
barriers (e.g.,
minimum lot size &
parking requirements;
zoning out multi-
family housing, etc.)
Regional fair-share
housing (Mt. Laurel
in New Jersey)
Targeting federal
housing subsidies to
mixed-income areas
(HOPE VI, Low-
Income Housing Tax
Credits, New Market
Tax Credits, Section 8
Vouchers, etc.)
Home equity assurance
programs (Oak Park,
IL)
Zoning for affordable
and manufactured
housing in rural areas
Regional tax-base
sharing by
municipalities
(Minneapolis-St.
Paul metro area) or
by school districts
(Texas)
Targeting local
subsidies to mixed-
income communities
(tax increment
financing (TIF), tax
abatements, etc.)
Community land trusts
(nonprofits own land
and keep some housing
perpetually affordable);
Limited-Equity
Cooperatives (tenants
share in overall
ownership and price
increases are limited)
Inclusionary zoning
and Transit-Oriented
Development (TOD)
Consolidated school
districts (Charlotte-
Mecklenburg)
Private financial aid
for mixed-income
communities
(resource-efficient
and location-
efficient mortgages)
Nonprofit owned rental
housing (community-
development
corporations)

Table 5 summarizes the main housing policy approaches to economic integration.
Some policies are designed to prevent middle-income families from fleeing
declining neighborhoods by protecting their home equity. Other policies aim to
prevent families from being displaced from gentrifying neighborhoods using
community land trusts or limited-equity co-ops. Policies that consolidate school
districts or reduce funding inequities among school districts reduce incentives for
higher-income households to avoid living close to lower-income households. One
of the most powerful ways to promote economic integration would be to remove
exclusionary zoning laws that prevent affordable housing from being built near
suburban jobs.

Successful economic integration requires working with the market, not against
it. Policies must be place-based, sustaining stable market demand in
neighborhoods for both low- and middle-income households. Middle and
affluent households always have the alternative to move out. And if policies
concentrate on improving the incomes of poor families, without a strong place-
based strategy, many households will simply move out once their incomes move
up. In order to sustain stable integrated neighborhoods, economic integration
policies must be adapted to each metropolitan context.

In weak market regions experiencing suburban sprawl, powerful policies will be
needed to overcome centrifugal tendencies. Given a surplus of housing in many
central cities and inner suburbs, housing subsidies will need to be carefully
targeted to create neighborhoods of choice for middle- and working-class
families. Strong neighborhoods with self-sustaining markets as well as weak
neighborhoods suffering rapid abandonment should not be the targets of
housing subsidies. Target neighborhoods should be chosen because they
possess anchors, such as public transit, parks, or pedestrian access to job
clusters that increase the chances of making them true neighborhoods of choice.
Housing policies should be coordinated with other supportive policies, such as
economic development, job training, schools, parks and public transit. The most
difficult challenge for local governments will be resisting the law of political
dispersion the tendency to spread the money around to benefit the most
political interests. Spreading the money around, the so-called peanut butter
approach, ensures that the neighborhoods will continue to be overwhelmed by
weak market housing dynamics.

Strong market regions have a completely different challenge in promoting
economic integration. In strong markets the problem is too much demand for
housing, not too little. The main challenge in strong market regions is that
gentrification will lead to displacement of the poor and areas will re-segregate.
Research suggests that gentrification may have more benefits for low-income

residents than is generally thought. Many long-time residents find ways to stay
in gentrifying neighborhoods and the influx of higher income households
reduces crime and creates job and retail opportunities.
26
Nevertheless,
displacement of low-income households can be a problem in gentrifying
neighborhoods. Policy tools to limit displacement , including community land
trusts, limited-equity co-ops and nonprofit ownership of rental housing. (See
Table 5.)

Over the years, the most common method for promoting mixed-income
neighborhoods has been to subsidize affordable housing in middle-income areas
where it would not otherwise be built or give people vouchers so that they can
move to middle-class areas. These approaches have one major drawback: they
cost money. Indeed, because housing is so expensive they cost a considerable
amount of money per household. Especially in strong market metropolitan areas
with rising housing prices subsidizing economic integration with taxpayer-funded
incentives is prohibitively expensive. With the looming recession placing fiscal
stress on federal, state and local governments, housing subsidy funds will be
limited. It is for this reason that we have chosen to concentrate on two place-
based strategies that local governments can use to promote economic integration
by leveraging market dynamics without taxpayer funding.
Incl usi onary Zoni ng
Inclusionary zoning (IZ) is one way to promote economic integration without
taxpayer subsidies. IZ relies upon leveraging developer commitments to
affordable housing by allowing them to build at higher densities. The success of
IZ depends upon a strong real estate market. Ironically, IZ is often made possible
by exclusionary zoning. To the extent that zoning requires single-family homes on
large lots, the supply of housing will not be able to keep pace with demand. IZ
laws enable developers to build at higher densities if they agree to set aside a
certain number of units for low- and moderate-income families. Developers not
only have an incentive to build units with a mix of incomes but they have a strong
motivation to ensure that the affordable units do not harm the value of the market-
rate units.

All IZ laws operate under the same principles, but they can vary significantly in
their application.
27
Some IZ laws are voluntary: developers can accept or reject
the density bonus in exchange for providing affordable units. More municipalities
are now switching to mandatory IZ programs, which are more effective at
producing affordable units. Usually the laws only apply to larger developments of
50 units or more. The set aside for affordable units varies from 5 to 35 percent and
the targeted incomes vary from 30 to 120 percent of area median income. The two
factors interact so that the lower the income group targeted, the lower the set

aside. IZ laws need to be carefully calibrated to protect the profitability of

developers. In strong housing markets IZ laws can generate more units for lower-
income families. In weak markets, IZ may need to be accompanied by subsidies
for the developer.

One of the nations first and probably most successful IZ ordinances was adopted
in 1973 by Montgomery County, Maryland. According to Maryland law, with the
exception of a few municipalities, Montgomery County exercises all land use
planning and zoning powers. The law requires that in all housing developments
of 50 units or more 15 percent of the housing must be set aside for the lowest one-
third of county residents. Today, the cut off is 65 percent of the median household
income in the county. Montgomerys law has an innovative dimension that enables
it to address the needs of low-income renters. The local Housing Authority can
purchase up to one-third of the affordable units and sell or rent them to low-
income families. In return, developers can receive a density bonus of up to 22
percent. To date the IZ ordinance has produced over 11,000 affordable units that
are dispersed throughout middle-class housing developments in Montgomery
County at little or no cost to the taxpayers.

IZ represents a dramatic shift away from Euclidian zoning laws. Named after
the 1926 Supreme Court ruling (Village of Euclid v. Ambler Realty Co., 1926) that
upheld zoning power in the United States, Euclidean zoning is based on a map
that sorts different types of land uses into separate sections of the city. Each
classification (single-family, rental, retail, etc.) is strictly separated from the
others. IZ is a non-Euclidean approach to zoning that sets up incentives to achieve
public goals and then leaves it largely up to developers on how to achieve those
goals. Under this more flexible approach, developers have invented ways to build
affordable housing units that blend into the neighborhood. If the market-rate
housing has a brick faade, they put brick on the affordable units. Instead of
building two duplexes side-by-side, they will build them back-to-back so that
from the street it looks like only one duplex. As a result of developer creativity, it is
almost impossible to spot the affordable units and there is no stigma attached to
living in one.

Affordable Dwelling Units in Fairfax Count, Virginia, Required by Inclusionary Zoning
Can you tell which townhouses are affordable and which ones are market rate (selling for
$750,000)?

The key to the political success of IZ ordinances is that they frame the issue not in
terms of welfare or public housing but as a matter of correcting a market failure.
Instead of using the terms public housing or subsidized housing, IZ advocates
use of the term workforce housing. Their mantra is: If someone is good enough
to work here, they ought to be good enough to live here. The lack of affordable
housing is not caused by people making insufficient incomes, but by the market
failing to meet housing demand. Exclusionary zoning laws artificially restrict the
supply of affordable housing. Montgomery County experienced a huge boom in
jobs in recent decades but without the IZ law many of those workers would have
been forced to live outside the county, clogging the highways. IZ is framed as
meeting multiple objectives: by increasing economic integration it reduces the
crime and social decay that follows concentrated poverty; it reduces roadway
congestion by enabling workers to live closer to where they work; by economically
balancing communities, it reduces fiscal stress.

Above all, political success requires convincing citizens that the affordable units
will not harm property values. Research has shown this to be true.
28
If the
demand for affordable housing is satisfied over a large geography, no single
neighborhood is forced to accept a large proportion of affordable units (which

could negatively affect neighboring property values). If all developers are required
to set aside units for affordable housing, no developer feels unfairly burdened.
This is another argument for mandatory IZ laws.

Finally, IZ does not cost the taxpayers anything. How, then, does affordable
housing get built if no public money is involved? Essentially, developers are able
to meet the demand for low-income housing the way they have always done by
building smaller units in more dense configurations. In the 19
th
and early 20
th

centuries developers made profits by building multi-story tenements in central
cities. In the case of IZ the units can be affordable because the cost of the land is
effectively zero. The affordable units are added on to an existing site plan. The
profit rate on the affordable units can be zero or even negative. But because the
density bonus extends beyond the affordable units, the developer is able to make
additional profits by building more market-rate units on the land. IZ creates value
by letting the market operate to satisfy demand. Developers would not participate
if they did not make a profit.

In strong market regions with rising land values and strong demand for housing,
IZ provides a way to promote economic integration at no cost to the taxpayers. By
2008, about 500 cities and counties encompassing 38 million residents had
enacted mandatory IZ laws. One out of eight Americans now resides in
communities requiring mixed-income housing.
29
Responding to Long Islands
severe shortage of workforce housing, in July 2008, the New York Legislature
mandated that all 109 villages and towns in Nassau and Suffolk counties enact IZ
laws. Inclusionary zoning expert David Rusk has estimated what would happen
if IZ were enacted in the 100 largest American metropolitan areas. Assuming a 15
percent set aside that would cover 80 percent of all new construction between
1980 and 2000, the result would be the creation of 2.6 million additional
affordable units and a 37 percent reduction in economic segregation.
30
This
would result in dramatic reductions in crime and other social problems without
major taxing and spending programs.

IZ has been limited almost entirely to strong market regions. But even in weak
market regions, there are housing submarkets that are strong enough to
leverage investment in affordable workforce housing. Strong housing
submarkets are created by access to amenities, such as walking distance to
employment centers, parks and open space, pedestrian-friendly public spaces
and retail developments, and access to safe and convenient public
transportation.

Usi ng Publi c Transi t to Promote Economi c I ntegrati on

Housing policy and transportation policy tend to be developed in separate policy
silos. HUDs dont talk to DOTs. This is a mistake. Transportation investments
shape housing demand in powerful ways. Suburbanization would not have been
possible without massive investments in suburban road capacity. Years ago in
cities, multi-story apartment buildings, often with shops on the first floor,
clustered along trolley lines, and streetcar suburbs developed at trolley hubs.
Transportation policy can be used to shape housing development. Linking
transportation to housing development is a powerful way to build mixed-income
communities.

In the last 20 years, a new generation of fixed-rail transit systems has evolved.
Starting with San Diego in the early 1980s metropolitan areas have invested in new
light rail systems that are faster than trolleys, but stop more frequently than the
heavy rail suburban commuter trains. Bus rapid transit (BRT) lines, in which
buses are given dedicated lanes and priority at traffic lights, are being developed
in many cities and can provide many of the same benefits as light rail. The
demand for housing near light rail station lines is soaring. Today, about 6 million
households live within a half mile of a transit station. The demand for housing
adjacent to transit is projected to reach 16 million by 2030.
31
To meet this demand,
10 million housing units will need to be built within a 10-minute walk of transit
stations. This movement toward denser, mixed-use forms of development
presents a golden opportunity to create economically integrated communities.

Transportation has taken up a larger and larger proportion of household budgets,
especially for low-income families. Since the 1930s the amount the average
American family spends on transportation has increased by about 500 percent (in
real terms). As a proportion of family income, transportation costs have more than
doubled.
32
A major reason for this is the sprawled out pattern of suburban
development. As women have entered the workforce, owning two automobiles has
become a necessity. In 2007 the cost of owning an automobile averaged $9,498.
33

Today, the average American family spends 19 percent of its budget on
transportation with the burden increasing as family incomes decline. High-
income families spend only about 9 percent of their incomes on transportation
compared to 55 percent for very low-income families. The rising cost of gasoline
is essentially a regressive tax.
34
The average American household now spends
about 4 percent of its income on gas, but low-income households, typically driving
older cars that get lower gas mileage, spend twice that percentage or more.

For most Americans, transportation costs are fixed because they must use an
automobile for nearly all their trips. But people who live in transit-rich,
pedestrian-friendly neighborhoods have a choice. Accordingly, the amount of

money spent on transportation varies depending on where you live. In reality,

transportation costs are tightly tied to housing. Under the slogan, drive til you
qualify, households trade longer commutes in exchange for more affordable
suburban housing on cheaper land. Housing near the center of urban areas that
is well serviced by transit is more expensive per square foot but enables
households to cut their transportation costs. Recognizing this, Fannie Mae
created a Location Efficient Mortgage that enables borrowers to qualify for a
larger loan if they are buying in areas that have lower average transportation costs.

The proportion of household budgets spent on transportation varies across
metropolitan areas, from 21 percent in Houston, Cleveland and Detroit, to 15
percent in Baltimore, Portland and New York.
35
Higher levels of density and
public transit enable households to cut automobile use and decrease
expenditures on transportation. Transportation costs also vary within
metropolitan areas. The Urban Markets Initiative has developed a new tool called
the Affordability Index that shows how much a household can save by living in a
transit-rich environment. In Minneapolis-St. Paul, monthly costs of transportation
varied from $446 to $941. Moving from a transit-poor to a transit-rich
neighborhood would save the average household $5,940 a year.
36
For a low-
income family, this would be huge.

Substantial new investments are being made in new light rail systems. The
Federal New Starts program has been funded at less than $2 billion a year and
only a handful of metropolitan areas get assistance in any year. But many
metropolitan areas are passing local taxes to expand transit. In 2004 Denver
voters passed a half-cent sales tax that will fund a $4.7 billion expansion of the
light rail system. Charlotte voters also approved a half-cent sales tax that will
finance a $9 billion light rail system scheduled for completion in 2030. Light rail
systems are sold to the voters for a wide range of benefits, including cutting traffic
congestion, reducing gasoline consumption, improving air quality and attracting
new investment to the region. In addition, light rail systems are touted for
stimulating dense, mixed-use development and helping low-income people who
cannot afford to own a car.

Light rail systems are valuable not just because they respond to present demand
for transit but because they shape future demand for transit. Transit-oriented
development (TOD) is development within a half mile of a transit station (about a
10-minute walk) that is high-density, pedestrian friendly, mixed-use and includes
station-focused public spaces. TOD is built primarily by private developers but it
has extensive public benefits that justify government support: TOD can increase
property values around stations and therefore enhance tax revenues; well-
designed TODs can reduce crime by creating eyes on the street and 24-hour
activity; TOD increases transit ridership by giving residents access to more

destinations by transit and on foot; TOD promotes active, healthy lifestyles; and
TOD saves infrastructure costs by reducing the need for parking.

TOD can also improve social equity and reduce economic segregation. Low-
income households are good to have in TOD developments because they tend to
use transit more than higher-income households. In 2001, those making less than
$20,000 a year accounted for 38 percent of all transit riders, far more than their 14
percent share of the urban population.
37
Mixed-income TOD also provides low-
income people better access to jobs and therefore increases their chances of
upward mobility.

Mixed-income TODs can be promoted by governments in a number of different
ways. TOD overlay zoning districts have been implemented that reduce the
requirements for parking, develop design standards, and encourage denser
developments around stations. In strong markets, TOD overlay zoning can permit
higher levels of density in exchange for a certain percentage of affordable units.
Low-income households are less likely to own a car and therefore the government
can reduce the parking requirement by 75 percent (from one parking space per
middle-income unit to one-quarter of a space per low-income unit).
38
At $20,000
to $30,000 per underground parking space, this can be a powerful incentive for
developers to include affordable housing. In weaker housing markets,
governments can further incentivize mixed-income TODs using Tax Increment
Financing (TIF) districts or Low-Income Housing Tax Credits (LIHTC).

The potential savings to low-income households of living in a TOD are significant,
but only rarely is it possible for households to entirely give up access to a car. In
most American metropolitan areas families need a car periodically to make major
purchases or to visit friends or relatives in another part of the region. Normally,
there is no way to own part of a car, but now with the spread of carsharing you
can gain access to an automobile on a part-time basis. A nonprofit in the Bay Area,
City CarShare, began operating in 2001. Subsequently private companies, FlexCar
and ZipCar, have entered the business. Flex cars are parked on city streets or in
parking garages, and after joining and doing a background check, members can
use the cars on a per hour basis, usually for less than $10 an hour. A study of
CarShare members found that nearly 30 percent had gotten rid of one or more
cars and nearly two-thirds said they had decided not to purchase another car.
39

Imagine what it would mean to a family of three earning the federal poverty cut-off
($17,600 in 2008) if they could dispense with the cost of owning a car (annual cost
$9,498) and instead use public transit and carsharing at half that amount or less.

In short, the development of new light rail systems across American cities offers
an opportunity to build economically integrated transit villages. Even in weak
market metropolitan areas land values are rising around transit stations, justifying

denser development. Local governments can induce mixed-income developments

by allowing denser development in exchange for setting aside some units for low-
and moderate-income housing. Transportation policy can be used to leverage
housing dollars and opportunities for low-income families. If land values are not
high enough to leverage developer commitments for workforce housing,
governments can use housing subsidies, such as Low-Income Housing Tax
Credits, to make it profitable. Targeting subsidies to mixed-income TOD makes
sense because of the added benefits of reducing traffic congestion and increasing
economic integration. With political will, it can be done in any city where transit
increases land values around stations.
Conclusi on: Where Do We Go f rom Here?
Given the enormity of the recent financial bailout, the new administration in
Washington will undoubtedly be forced to cut social spending. State and local
governments also face cutbacks. How can the United States address serious
social problems failing schools, crime, chronic joblessness, teen pregnancy,
drug abuse in a time of fiscal retrenchment? One way is to promote economic
integration. The research is clear that concentrated poverty has negative effects
on school performance and a range of social outcomes independent of individual
characteristics, including poverty. Even if we cannot reduce poverty we can reduce
its effects by creating more mixed-income communities. Though racial
segregation has been declining slowly, economic segregation has been increasing
steadily, and we need to do something about it.

Economic integration varies significantly across metropolitan areas. As we show,
there are many different ways to measure economic integration. For this study we
devised a Relative Economic Integration Index that takes into account the huge
differences in income ranges across metropolitan areas. We also examined the
segregation of the poor from the rest of society and the concentration of poverty.
The different ways of measuring economic integration are related but measure
different aspects of the same phenomena. Notwithstanding the complexities of
measuring economic integration, the federal government should make tracking
and setting goals for increasing economic integration, part of its policy.

Policies for promoting economic integration should not tell people where to live.
As we put it: Economic integration, stated simply, means that households of all
different income levels have realistic housing options in all parts of a
metropolitan area, so that they can pursue opportunities throughout the region.
A certain amount of economic segregation is normal. But present levels of
economic segregation are excessive. They are driven not just by individual
choices but by public policies, including exclusionary zoning laws, excess
highway capacity and fiscal inequities across municipalities and school districts.

Federal and state governments should level the playing field, eliminating the
market distortions created by public policies.

Policies to foster mixed-income communities must be based on the realities of
the market. Suburban sprawl has been an engine of economic segregation.
Smart growth and urban containment policies would help to level the playing
field, fostering economic integration. Urban containment policies are politically
difficult but containment is happening spontaneously as regions exhaust
developable land, traffic congestion increases, commutes lengthen and gas
prices rise. As the suburban frontier gradually closes, new housing demand will
turn inward, presenting opportunities to leverage market forces to promote
economic integration.

Intelligent economic integration policies require us to transcend the policy silos
that too often reinforce the status quo. Transportation policy is housing policy.
Transportation policy should not just aim at mobility, which often results in a
predominant focus on highways, reinforcing sprawl. Transportation policy
should aim at accessibility, which requires a multimodal approach that
integrates transportation with land use. Transportation policy should be used to
build sustainable mixed-income communities. Housing policy is education
policy. Housing policies that promote mixed-income communities can improve
school performance. Flexible policies that work with the market can give the
taxpayers more bang for their buck. Indeed, as we show, inclusionary zoning and
transit-oriented development can leverage market values to increase economic
integration, with all its collateral benefits, at little or no cost to taxpayers.

37

37
Endnotes

1
David M. Cutler, Edward L. Glaeser, and Jacob L. Vigdor, "The Rise and Decline of the American Ghetto,
Journal of Political Economy 107 (1999), pp. 455-506.
2
See Paul A. Jargowsky, "Take the money and run: economic segregation in U.S. metropolitan areas,"
American Sociological Review, 61 (1996), pp. 984-998; Tara Watson, Metropolitan Growth, Inequality, and
Neighborhood Segregation by Income, Brookings-Wharton Papers on Urban Affairs 2006, pp. 1-45; Alan
Berube and Elizabeth Kneebone, Reversal of Fortune: A New Look at Concentrated Poverty in the
2000s, Metropolitan Policy Program at Brookings (Washington, DC: Brookings Institution, 2008).
3
Report of the National Advisory Commission on Civil Disorders (New York: E.P. Dutton & Co., Inc., 1968).
4
For syntheses of the research on the policy challenges posed by economic segregation, see Paul A.
Jargowsky, Poverty and Place: Ghettos, Barrios, and the American City (New York: Russell Sage
Foundation, 1997); Alan Altshuler, et al, Governance and Opportunity in Metropolitan America
(Washington, D.C.: Natioanl Academy Press, 1999); Peter Dreier, John Mollenkopf, and Todd Swanstrom,
Place Matters: Metropolitics for the Twenty-First Century, rev. ed. (Lawrence, KS: University Press of
Kansas, 2005).
3
For example, see E. A. Hanushek, J. F. Kain, J. M. Markman, and S. G. Rivkin, "Does peer ability affect
student achievement?" Journal of Applied Econometrics 18 (2003), pp. 527-544; A. A. Summers and B. L.
Wolfe, "Do Schools Make a Difference," American Economic Review 67 (1977), pp. 639-652; and R. W.
Zimmer and E. F. Toma, "Peer effects in private and public schools across countries," Journal of Policy
Analysis and Management 19 (2000), pp. 75-92.
6
William Julius Wilson, The Truly Disadvantaged: The Inner City, the Underclass, and Public Policy
(Chicago: The University of Chicago Press, 1987), p. 57.
7
Joseph Cortright, City Vitals (Chicago: CEOs for Cities, undated), pp. 36-37.
8
The complete list of income brackets is:
Less than $9,999
$10,000 to $14,999
$15,000 to $19,999
$20,000 to $24,999
$25,000 to $29,999
$30,000 to $34,999
$35,000 to $39,999
$40,000 to $44,999
$45,000 to $49,999
$50,000 to $59,999
$60,000 to $74,999
$75,000 to $99,999
$100,000 to $124,999
$125,000 to $149,999
$150,000 to $199,999
$200,000 and over

9
We combine Metropolitan Statistical Areas (MSAs) with Primary Metropolitan Statistical Areas (PMSAs).
We dont report figures for Consolidated Metropolitan Statistical Areas (CMSAs), which are composed of
PMSAs and are probably too large to be considered a cohesive social and economic unit.

10
Paul A. Jargowsky, Poverty and Place: Ghettos, Barrios, and the American City (New York: Russell Sage
Foundation, 1997).
11
Rachel Dwyer, Expanding Homes and Increasing Inequalities: U.S. Housing Development and the
Residential Segregation of the Affluent, Social Problems 54 (1), 2007, pp. 23-46.
12
Paul Jargowsky, Sprawl, Concentration of Poverty, and Urban Inequality, in Urban Sprawl: Causes,
Consequences, and Policy Responses, edited by Gregory D. Squires (Washington, D.C.: Urban Institute
Press, 2002), p. 57.
38

38

13
Edward Glaeser, Matthew Kahn & Chenghuan Chu, Job Sprawl: Employment Location in U.S.
Metropolitan Areas, The Brookings Institution, Center on Urban & Metropolitan Policy, May 2001.
14
Lucy and Phillips, 2003.
13
Todd Swanstrom, Colleen Casey, Robert Flack & Peter Dreier, Pulling Apart: Economic Segregation
Among Suburbs and Central Cities in Major Metropolitan Areas, Brookings Institution Living Cities
Census Series, October 2004. Poor suburbs are defined as suburbs with per capita incomes less than
75 percent of the per capita income for the metropolitan area.
16
Tara Watson, New Housing, Income Inequality, and Distressed Metropolitan Areas, Brookings
Institution, Metro Economy Series, September 2007.
17
Tom Bier & Charlie Post, Vacating the City: An Analysis of New Home Construction and Household
Growth, in Redefining Urban & Suburban America, edited by Alan Berube, Bruce Katz, and Robert E.
Lang (Washington, D.C.: Brookings Institution Press, 2006), pp. 167-189.
18
PowerPoint presentation by Allan Mallach at Strengthening Neighborhoods in Weak Markets, Federal
Reserve Bank of St. Louis, September 24-25, 2008. For a discussion of the right-sizing movement see
Miriam Axel-Lute, Small is Beautiful Again, Shelterforce (Summer 2007).
19
Neighborhood Progress, Inc., Reinvigorating the Urban Marketplace: Clevelands Strategic Investment
Initiative (www.neighborhoodprogress.org; accessed December 17, 2008).
20
Bruce Katz, Neighborhoods of Choice and Connection: The Evolution of American Neighborhood Policy
and What It Means for the United Kingdom, Brookings Institution, Research Brief, July 2004.
21
Danielle Arigoni, Affordable Housing and Smart Growth: Making the Connection (Washington, D.C.:
National Neighborhood Coalition, 2001).
22
Glaeser, Kahn & Chu, op. cit.
23
Peter S. Goodman, Fuel Costs Shift Equation for Life in Far Suburbs, New York Times (June 25, 2008).
24
Arthur C. Nelson, Thomas W. Sanchez & Casey J. Dawkins, The Effect of Urban Containment and
Mandatory Housing Elements on Racial Segregation in US Metropolitan Areas, 19990-2000, Journal of
Urban Affairs 26 (3), 2004, p. 348.
23
Robert W. Burchell, et al, Sprawl Costs: Economic Impacts of Unchecked Development (Washington,
D.C.: Island Press, 2005).
26
Jacob L. Vigdor, Does Gentrification Harm the Poor? Brookings-Wharton Papers on Urban Affairs
(Washington, D.C.: Brookings Institution Press, 2002), pp. 133-182; Lance Freeman, There Goes the
Hood: Views of Gentrification from the Ground Up (Philadelphia, PA: Temple University Press, 2006.
27
This section is based on a PowerPoint presentation by David Rusk, Inclusionary Zoning 101: A
Presentation for ISAAC, Kalamazoo, Michigan, May 21, 2008.
28
See Innovative Housing Institute and Greater Washington Research Council, The House Next Door,
(www.inhousing.org/housenexhtm); and MITs Technology Center for Real Estate, Effects of Mixed-
Income, Multi-Family Rental Housing Developments on Single-Family Home Values,
(http:??mit.edu/cre/research/hai/pdf/40B_report_HAI_0405.pdf).
29
Calculation by David Rusk (personal communication).
30
Based on calculations by David Rusk using 2000 Census data and the Urban Institutes economic
segregation index.
31
Shelley Poticha, Building Housing Near Transit: A Long-Lasting Affordability Strategy, Congressional
Testimony Before the Appropriations Subcommittee on Transportation, Housing and Urban Development,
and Related Agencies, U. S. House of Representatives, March 8, 2007. This estimate was made before the
recent increase in gas prices.
32
As reported in Jerome M. Segal, Graceful Simplicity: Toward a Philosophy and Politics of Simple Living
(New York: Henry Holt, 1999), p. 55.
33
American Automobile Association, Your Driving Costs (Heathrow, Florida: AAA, 2007). The estimate is
based on gasoline costing $2.256 a gallon.
34
Poticha, op cit.

Center for Neighborhood Technology and Surface Transportation Policy Project, Driven to Spend:
Pumping Dollars out of Our Households and Communities, June 2005, p. 6.

Center for Transit-Oriented Development and Center for Neighborhood Technology, The Affordability
Index: A New Tool for Measuring the True Affordability of a Housing Choice, Brookings Institutioen,
Urban Markets Initiative, Market Innovation Brief, January 2006, p. 10.

John Pucher, Public Transportation, in The Geography of Urban Transportation, 3rd ed., edited by
Susan Hanson and Genevieve Giuliano (New York: Guilford Press, 2004), p. 212.

Dena Belzer, et al, The Case for Mixed-Income Transit-Oriented Development in the Denver Region
(Center for Transit-Oriented Development), p. 42.

Robert Cervero and Yu-Hsin Tsai, "San Francisco City CarShare: Travel-Demand Trends and Second-
Year Impacts" (August 1, 2003). Institute of Urban & Regional Development. IURD Working Paper Series.
Paper WP-2003-05.
http://repositories.cdlib.org/iurd/wps/WP-2003-05.

Appendi x:


Househol d I ncome Threshol ds f or Membershi p i n Top
or Bot tom Thi rd, by Metropol i tan Area, 2000
Metropolitan Area Thresholds ($)
Botton Third Top Third
New Orleans, LA 23,022 51,680
Miami, FL 23,334 52,864
Pittsburgh, PA 25,014 53,798
Buffalo, NY 25,481 55,139
New York, NY 25,572 62,199
Tampa, FL 26,145 52,707
San Antonio, TX 27,011 54,976
Providence, RI/MA 27,300 59,473
Los Angeles, CA 27,953 62,542
Greensboro, NC 28,374 56,474
Fort Lauderdale, FL 28,468 59,671
Cleveland, OH 28,785 59,312
Riverside, CA 28,946 60,171
Orlando, FL 29,906 57,688
Norfolk, VA/NC 30,252 58,253
Cincinnati, OH/KY/IN 30,287 61,915
Las Vegas, NV/AZ 30,313 58,861
Houston, TX 30,554 64,596
St. Louis, MO/IL 30,703 61,902
West Palm Beach, FL 30,821 64,406
Nashville, TN 30,927 60,665
Columbus, OH 31,203 62,096
Phoenix, AZ 31,515 62,093
Milwaukee, WI 31,664 63,738
Indianapolis, IN 31,759 63,179

Philadelphia, PA/NJ 31,805 67,861

San Diego, CA 32,318 66,736
Sacramento, CA 32,376 65,225
Fort Worth, TX 32,388 63,937
Kansas City, MO/KS 32,471 63,638
Charlotte, NC/SC 32,522 63,375
Detroit, MI 33,004 69,214
Portland, OR/WA 33,308 64,458
Dallas, TX 33,595 68,460
Austin, TX 33,799 68,216
Baltimore, MD 33,823 69,159
Chicago, IL 35,662 72,195
Denver, CO 36,269 70,371
Boston, MA/NH 36,856 78,435
Atlanta, GA 36,887 71,574
Seattle, WA 37,448 72,499
Newark, NJ 37,695 81,878
Minneapolis, MN/WI 38,621 72,853
Bergen, NJ 40,274 84,930
Oakland, CA 40,338 83,431
Orange County, CA 40,982 82,342
San Francisco, CA 42,659 91,862
Wash., DC/MD/VA/WV 43,962 86,299
Nassau, NY 48,481 93,030
San Jose, CA 52,342 102,814

Source: 2000 Census, Summary File 3, Table P-52, Calculations by the author.