Essay
gaoxiaone
3.pdf
Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=rtpg20
The Professional Geographer
ISSN: 0033-0124 (Print) 1467-9272 (Online) Journal homepage: https://www.tandfonline.com/loi/rtpg20
Home Ownership, Minorities, and Urban Areas: The American Dream Writ Local
Lawrence A. Brown & Michael D. Webb
To cite this article: Lawrence A. Brown & Michael D. Webb (2012) Home Ownership, Minorities, and Urban Areas: The American�Dream Writ Local, The Professional Geographer, 64:3, 332-357, DOI: 10.1080/00330124.2011.601183
To link to this article: https://doi.org/10.1080/00330124.2011.601183
Published online: 25 Aug 2011.
Submit your article to this journal
Article views: 445
View related articles
Citing articles: 8 View citing articles
Home Ownership, Minorities, and Urban Areas: The
American Dream Writ Local∗
Lawrence A. Brown and Michael D. Webb The Ohio State University
This article focuses on home ownership changes since 1990, particularly by minority or low-income populations—known colloquially as the American dream. This longtime centerpiece of U.S. policy has been primarily viewed in terms of national outcomes. Here, we address differences in a spatial context, focusing on the forty-nine Metropolitan Statistical Areas (MSA/CMSAs) with greater than 1 million population in 2000. The central question is how ownership change differed among urban areas. Following the annual State of the Nation’s Housing report, we spotlight the percentage point change (PPC) of home ownership by all, white, black, Hispanic, Asian, and minority (as-a-group) households—as these vary among urban areas for 1990–2000, 2000–2007, and 1990–2007. Statistically, we find that PPCs for each racial and ethnic group tend to move in tandem; that there are considerable differences among MSAs in PPC performance; that these differences tend to cluster spatially in a manner that reflects regional dynamics but that, overall, the goal of reducing the minority–white gap in home ownership has not been realized. Regarding specific vari- ables, metropolitan growth and, to a lesser extent, MSA size, best account for change in home ownership; subprime lending is not significant. In addition, consideration of unexplained variance leads us to conclude that, as a complement to the approach taken here, a more qualitative strategy would significantly increase understanding of this important issue—a strategy that focuses on institutional structures, supply-side actors, advocacy groups, financial organizational practices, community procedures, and the like—with key informant interviews as a central component. Key Words: American dream policies, home ownership, housing, metropolitan areas, race/ethnicity.
∗Earlier versions of this article were presented in 2009 at the meeting of the North American Regional Science Council and in 2010 at meetings of the Western Regional Science Association, Association of American Geographers, International Geographical Union, and European Regional Science Association. Comments from persons attending those meetings are appreciated, as are those of anonymous reviewers that significantly improved the final product. The article emanated from an idea by Jennifer Evans-Cowley of Ohio State’s City and Regional Planning Program, while serving as a member of an MA thesis committee. Vital suggestions regarding relevant literature were provided by Mat Coleman and Kevin Cox of The Ohio State University, Geography, and an anonymous reviewer. Also of central importance were the efforts of Wenqin Chen and Ohio State’s Center for Urban and Regional Analysis in assembling data and related materials.
The Professional Geographer, 64(3) 2012, pages 332–357 C© Copyright 2012 by Association of American Geographers. Initial submission, May 2010; revised submission, November 2010; final acceptance, December 2010.
Published by Taylor & Francis Group, LLC.
Home Ownership, Minorities, and Urban Areas 333
Este artı́culo se centra en los cambios en el acceso a la casa propia desde 1990, particularmente por las minorı́as o poblaciones de bajos ingresos—coloquialmente conocido como el sueño americano. Esta eterna pieza central de la polı́tica de EE.UU. ha sido vista principalmente en función a los resultados nacionales. En este artı́culo tocamos las diferencias en un contexto espacial, centrándonos en las cuarenta y nueve Áreas Estadı́sticas Metropolitanas (MSA/CMSAs) con más de 1 millón de habitantes en el año 2000. La principal interrogante es cómo el cambio de propiedades difiere entre las zonas urbanas. A raı́z del informe anual del Estado de la Vivienda en la Nación, destacamos la variación en puntos porcentuales (PPC) en el acceso a la casa propia para todos los hogares, de blancos, negros, hispanos, asiáticos, y de las minorı́as (como-grupo)—ya que estas varı́an entre las áreas urbanas en los periodos 1990–2000, 2000–2007 y 1990–2007. Estadı́sticamente, encontramos que las PPCs para cada grupo racial y étnico tienden a moverse en tándem, que hay considerables diferencias entre las MSAs en el funcionamiento de la PPC, que estas diferencias tienden a agruparse espacialmente de una manera que refleja la dinámica regional, pero que, en general, el objetivo de reducir la brecha minorı́a-blanco en el acceso a la casa propia no se ha concretado. En cuanto a las variables especı́ficas, el crecimiento metropolitano y, en menor medida, el tamaño de MSA, representan mejor este cambio en el acceso a la casa propia, el préstamo hipotecario no es significativo. Además, al considerar las varianzas inexplicadas nos llevan a concluir que, como complemento al enfoque adoptado aquı́, una estrategia más cualitativa aumentarı́a significativamente la comprensión de este importante tema—una estrategia que se centra en estructuras institucionales, agentes de oferta, grupos de apoyo, prácticas financieras organizacionales, procedimientos de la comunidad, y similares—con entrevistas a informantes clave como componente central. Palabras claves: polı́ticas del sueño americano, acceso a la casa propia, vivienda, áreas metropolitanas, raza/etnia.
T he American dream—owning one’shome—has long been a fundamental goal of U.S. domestic policy. Franklin D. Roo- sevelt’s New Deal reforms included creation of the Federal Housing Administration (FHA) and predecessors of government-sponsored entities (GSEs) such as Fannie Mae and Freddie Mac. More recent initiatives are a response to ownership rate decreases during the 1980s and early 1990s, and the lag of minority home ownership rates behind those of whites. These include strengthening prohi- bitions on discriminatory lending through the Community Reinvestment Act and altering GSE practices to loosen downpayment re- quirements. Government regulation practices also were modified in ways that encouraged innovation in mortgage products, ultimately leading to subprime, no-documentation, and adjustable-rate loans.
Calibrating the effects of such policies has largely focused on national outcomes (Table 1). For example, Harvard’s Joint Center for Hous- ing Studies (JCHS), in its annual State of the Nation’s Housing report (2009, 37, Table A-4), shows that between 1994 and 2008 ownership rates for whites grew from 70.0 to 75.0 per- cent, blacks from 42.5 to 47.9, Hispanics from 41.2 to 49.1, Asians from 50.8 to 59.8, and mi- norities overall from 43.2 to 50.6. Nevertheless, white and minority rates moved more or less in tandem, hovering in at ∼25 points and lead- ing the 2003 report (JCHS 2003, 16) to note poignantly that “Strong gains notwithstanding,
the gap between white and minority home own- ership rates has improved little in 40 years.”1
Here, with the national perspective as a backdrop, we consider how the American dream has played out among urban areas. In particular, knowing that government actions invariably lead to outcomes that are spatially differentiated, even when promoted as place- neutral, JCHS reports provide only a partial (and potentially misleading) portrait of mi- nority accessibility to housing and disregard a number of relevant factors. Said another way, despite public exhortations and financial in- centives, supply and demand levels reflect lo- cal characteristics such as demographic profiles and strength of the economy, and accessibility varies accordingly. This article seeks to iden- tify such characteristics for metropolitan areas and delineate differences in their levels of home ownership.2
Our sample consists of the forty-nine continental Metropolitan Statistical Areas (MSAs/CMSAs) that were larger than 1 mil- lion in population in 2000.3 Focus is on owner- ship change among whites, blacks, Hispanics, Asians, and minorities overall for the periods from 1990 to 2000 and 2000 to 2007 drawing on data from, respectively, the decennial Cen- sus and American Community Survey (ACS). Differences among metropolitan areas are highlighted and linked to appropriate char- acteristics. Our primary objective is to better understand how and why national policies and related exhortations have housing outcomes
T a b
le 1
H o
m e
ow n
e rs
h ip
ra te
s, p
e rc
e n
ta g
e p
o in
t ch
an g
e by
ra ce
an d
e th
n ic
it y,
an d
w h
it e
–m in
o ri ty
d iff
e re
n ti al
, 1
9 9
4 –2
0 0
8
W h
it e –
% p
o in
t %
p o
in t
% p
o in
t %
p o
in t
A s ia
n -
% p
o in
t %
p o
in t
m in
o ri
ty Y e a r
A ll %
c h
a n
g e
W h
it e %
c h
a n
g e
B la
c k %
c h
a n
g e
H is
p a n
ic %
c h
a n
g e
o th
e r%
c h
a n
g e
M in
o ri
ty %
c h
a n
g e
g a p
1 9 9 4
6 4 .0
0 .0
7 0 .0
0 .0
4 2 .5
0 .0
4 1 .2
0 .0
5 0 .8
0 .0
4 3 .2
0 .0
2 6 .8
1 9 9 5
6 4 .7
0 .7
7 0 .9
0 .9
4 2 .9
0 .4
4 2 .0
0 .8
5 1 .5
0 .7
4 3 .7
0 .5
2 7 .2
1 9 9 6
6 5 .4
1 .4
7 1 .7
1 .7
4 4 .5
2 .0
4 2 .8
1 .6
5 1 .5
0 .7
4 4 .9
1 .7
2 6 .8
1 9 9 7
6 5 .7
1 .7
7 2 .0
2 .0
4 5 .4
2 .9
4 3 .3
2 .1
5 3 .3
2 .5
4 5 .8
2 .6
2 6 .2
1 9 9 8
6 6 .3
2 .3
7 2 .6
2 .6
4 6 .1
3 .6
4 4 .7
3 .5
5 3 .7
2 .9
4 6 .8
3 .6
2 5 .8
1 9 9 9
6 6 .8
2 .8
7 3 .2
3 .2
4 6 .7
4 .2
4 5 .5
4 .3
5 4 .1
3 .3
4 7 .4
4 .2
2 5 .8
2 0
0 0
6 7 .4
3 .4
7 3 .8
3 .8
4 7 .6
5 .1
4 6 .3
5 .1
5 3 .9
3 .1
4 8 .1
4 .9
2 5 .7
C e n
s u
s 2 0 0 0
6 5 .3
2 .5
7 1 .7
4 .2
4 2 .7
2 .7
4 2 .1
-1 .9
5 2 .4
2 .1
4 5 .0
2 .5
2 6 .7
2 0 0 1
6 7 .8
3 .8
7 4 .3
4 .3
4 8 .4
5 .9
4 7 .3
6 .1
5 4 .7
3 .9
4 9 .0
5 .8
2 5 .3
2 0 0 2
6 7 .9
3 .9
7 4 .7
4 .7
4 8 .2
5 .7
4 7 .0
5 .8
5 5 .0
4 .2
4 8 .9
5 .7
2 5 .8
2 0 0 3
6 8 .3
4 .3
7 5 .4
5 .4
4 8 .8
6 .3
4 6 .7
5 .5
5 6 .7
5 .9
4 9 .5
6 .3
2 5 .9
2 0 0 4
6 9 .0
5 .0
7 6 .0
6 .0
4 9 .7
7 .2
4 8 .1
6 .9
5 9 .6
8 .8
5 1 .0
7 .8
2 5 .0
2 0 0 5
6 8 .9
4 .9
7 5 .8
5 .8
4 8 .8
6 .3
4 9 .5
8 .3
6 0 .4
9 .6
5 1 .3
8 .1
2 4 .5
2 0 0 6
6 8 .8
4 .8
7 5 .8
5 .8
4 8 .4
5 .9
4 9 .7
8 .5
6 1 .1
1 0 .3
5 1 .3
8 .1
2 4 .5
2 0 0 7
6 8 .1
4 .1
7 5 .2
5 .2
4 7 .8
5 .3
4 9 .7
8 .5
6 0 .3
9 .5
5 0 .9
7 .7
2 4 .3
2 0 0 8
6 7 .8
3 .8
7 5 .0
5 .0
4 7 .9
5 .4
4 9 .1
7 .9
5 9 .8
9 .0
5 0 .6
7 .4
2 4 .4
C e n
s u
s 2 0 0 8
6 7 .0
4 .3
7 4 .0
6 .5
4 3 .0
3 .0
4 6 .5
2 .6
6 2 .0
1 1 .6
4 7 .6
5 .1
2 6
.4
N o
te :
Jo in
t C
e n
te r
fo r
H o
u si
n g
S tu
d ie
s p
e rc
e n
t ow
n e
rs h
ip d
at a
w e
re d
ra w
n fr
o m
th e
U .S
. C
e n
su s
H o
u si
n g
V ac
an cy
S u
rv ey
; ye
ar -t
o -y
e ar
p e
rc e
n ta
g e
p o
in t
ch an
g e
(P P
C )
w as
co m
p u
te d
by au
th o
rs , as
w as
th e
w h
it e
–m in
o ri ty
d iff
e re
n ti al
. P
P C
w as
co m
p u
te d
as ow
n e
rs h
ip ra
te (t
) –
ow n
e rs
h ip
ra te
(1 9
9 4
). C
e n
su s
2 0
0 0
n u
m b
e rs
w e
re co
m p
u te
d by
au th
o rs
u si
n g
ta b
u la
ti o
n s
o f
th e
2 0
0 0
U .S
. C
e n
su s
o f
P o
p u
la ti o
n ;
re la
te d
P P
C s
w e
re co
m p
u te
d by
au th
o rs
u si
n g
th e
1 9
9 0
C e
n su
s as
b as
e .
C e
n su
s 2
0 0
8 p
e rc
e n
ta g
e s
w e
re co
m p
u te
d by
au th
o rs
u si
n g
ta b
u la
ti o
n s
o f
th e
A m
e ri ca
n C
o m
m u
n it
y S
u rv
ey ;
re la
te d
P P
C s
w e
re co
m p
u te
d by
au th
o rs
u si
n g
th e
1 9
9 0
C e
n su
s o
f P o
p u
la ti o
n as
b as
e .
S o
u rc
e :
Jo in
t C
e n
te r
fo r
H o
u si
n g
S tu
d ie
s, H
ar va
rd U
n iv
e rs
it y
(2 0
0 9
), S
ta te
o f
th e
N at
io n’
s H
o u
si n
g 2
0 0
9 ,
A p
p e
n d
ix Ta
b le
A -4
: H
o m
e ow
n e
rs h
ip R
at e
s by
A g
e ,
R ac
e /E
th n
ic it
y, an
d R
e g
io n
: 1
9 9
4 –2
0 0
8 .
334
Home Ownership, Minorities, and Urban Areas 335
that vary appreciably from place to place. In this regard, that metropolitan areas group rather nicely in terms of home ownership out- comes suggests an important role for regional assemblages of socioeconomic characteristics and trends therein.
The article begins with a synopsis of earlier research that underpins and provides context for our inquiry. We then shift to empirical anal- yses of home ownership gains for 1990 to 2000, 2000 to 2007, and 1990 to 2007.
Background
A large number, if not the majority, of broad national policies are place blind, but their imple- mentation is conditioned by, and reflects, an area’s economic base, institutional flexibility, political constituencies, socioeconomic compo- sition, bureaucratic culture, economic health, and the like. Accordingly, policies that are su- perficially aspatial, and often promoted as such, in fact lead to highly differentiated outcomes from one place to another.4 Those related to the American dream are no exception.
One source of such variation is local condi- tions that impact the supply and demand for housing and, thus, the level of home ownership and changes therein. These include, as instru- mental factors, life cycle, user costs, and liquid- ity constraint effects on the decision to rent or own.
Life cycle adherents emphasize the demo- graphic profile of a population (W. A. V. Clark, Deurloo, and Dieleman 1984, 2003; Kendig 1984; G. Painter and Lee 2009) whereby a per- son’s housing career reflects factors such as age, marital status, children, wealth, and the like. In simplified terms, the scenario would be renting after moving away from one’s family, followed by ownership and, later in life, by again rent- ing. The prime owning years are, roughly, ages thirty-five to sixty-five but punctuated by peri- ods, perhaps lengthy, of renting necessitated by relocation, divorce, economic misfortune, and the like. In terms of spatial variation, life cycle effects would be seen in that, for example, some cities have an older demographic profile (e.g., Florida retirement communities, Jacksonville) and others a younger profile (e.g., Austin, Los Angeles).5
The user cost framework focuses on differ- entials between owning and renting (Hender- shott and Slemrod 1983; Smith, Rosen, and
Fallis 1988; Gallin 2008; Verbrugge 2008; Gar- ner and Verbrugge 2009). Homeowner costs and benefits not experienced by renters include mortgage, maintenance, opportunities forgone due to the investment, depreciation and appre- ciation, tax savings, and so on. As the ben- efits of ownership increase relative to costs (e.g., from rising home values or falling interest rates), the ownership rate should expand, and vice versa. In terms of spatial variation, user cost effects would be seen in that, for exam- ple, some cities have a robust housing mar- ket as indicated by the 1990–2000 change in housing value (e.g., Denver, San Francisco), whereas others were less so (e.g., Buffalo, Rochester).
Liquidity constraints reference barriers to home ownership such as downpayment or other up-front costs (Zeldes 1989; Engelhardt 1996; Guiso and Jappelli 2002; Ortalo-Magne and Rady 2006). To meet these financial obliga- tions, renter households (or owners seeking to move on) set aside portions of their income, which might be constrained (or assisted) by costliness of the rental or owner markets, in- come level, personal habits such as a house- hold’s propensity to either save or consume, and so on. Under this approach, a decline in the liquidity constraint (e.g., by mortgage instru- ments that lessen the downpayment require- ment or a robust job market) would relate to an increase in ownership. In terms of spatial vari- ation, liquidity constraint effects would be seen in that, for example, some cities had a higher ratio of median house value to median house- hold income in 2000 (e.g., Los Angeles, New York, San Francisco), and others had a lower ratio (e.g., Buffalo, Pittsburgh, Rochester).
Local conditions, such as those just dis- cussed, are embedded in, and influenced by, the broader regional (space-) economy. Par- ticularly relevant for MSAs is the detritus of economic restructuring under the Fordist/ post-Fordist transition. In this regard, Brown and Sharma (2010, 16) called attention to sunk costs in differentiating American manufacturing belt (AMB, or Rust Belt) from Sun Belt locales. Specifically, these are
investments in material or human capital at a particular place . . . and the aggregate of this affects . . . regional and MSA profiles. Hence, AMB urban areas whose economic base was strongly Fordist entered the post–World War II period with an overload of sunk costs
336 Volume 64, Number 3, August 2012
relative to MSAs with less Fordist involvement, generally found in the Sun Belt. The ensuing post-Fordist transition left Rust Belt MSAs reel- ing with unemployment, a deflated economic base, an under-used (if not unusable) built en- vironment. . . . Rust Belt MSAs were “structure limited” (G. J. Clark, 1994, 12) and for some time, if not yet today, remained caught in an inertial vortex—i.e., “struggling to make a suc- cessful transition from an economy based on routine manufacturing to one based on more knowledge-oriented activities” (Vey 2007, 4) or “struggling to find their competitive niche . . . one foot planted in a waning industrial era, the other in the emerging global economy” (Austin and Affolter-Caine 2006, 4).
By 2000, however,
the sunk cost phenomenon that held back AMB/Rust Belt MSAs prior to 1990 [appears to have] abated. These economies have been reinvigorated over the past 20 years, albeit not to their former level; new economic enter- prise has formed and often prospered; and sunk costs have largely been absorbed or written off (Austin and Affolter-Caine 2006; Vey 2007). An element of this . . . is dramatic change in demo- graphic profiles, labor force characteristics, ed- ucational levels, human capital capabilities, and the like. (Brown and Sharma 2010, 23)
Elsewhere, however, there was a regional econ- omy that differs dramatically from the AMB, a new industrial complex that Markusen (1991) called the Gun Belt, a distinct creation of gov- ernment policy. She argued that World Wars I and II were
equipped from mass-production industrial plants . . . [while] the cold war was distinctively postmodern . . . World Wars were . . . con- ducted with endless battalions of men, armed with huge quantities of guns and ammunition, and transported on thousands of land-based ve- hicles, ships, and fighter planes . . . made pos- sible by the extraordinary productivity of the modern industrial economy . . . [whereas] the cold war era was characterized by the replace- ment of manpower with highly sophisticated, electronics-intensive, deadly precision machin- ery [such that] machines, not men, formed the core of a nation’s fighting ability. (394)
As this shift occurred, Cold War pressures to build military capability would seem to greatly favor traditional manufacturing and the AMB heartland, especially because it was already aligned with the military through joint war ef-
forts. In fact, however, success with the new warfare required a different approach and cul- ture of production. Accordingly, the beneficia- ries were not AMB locales so much as Los Angeles, Southern California, Silicon Valley, Boston’s Route 128, Austin–Houston–Dallas, Texas, Colorado Springs–Denver, Huntsville, Alabama agglomerations, and the like—in Markusen’s words, a new Gun Belt.
In sum, MSA housing markets reflect lo- cal characteristics such as those captured by life cycle, user costs, and liquidity constraint effects—and a myriad of others. The dis- position of these mirrors local conditions, and those, in turn, reflect the broader space- economy within which the MSA is situated. In general, then, characteristics of a city located in the AMB should differ visibly from those of a city in the Sun Belt or Gun Belt. Hence, we might expect to find different profiles of home ownership gains, overall and by race or ethnicity, and to find that these occur in spa- tially distinct clusters that correspond with re- gional differences such as those associated with the AMB/Sun–Gun Belt.
An Urban System Perspective on the
American Dream
Urban agglomerations are examined as discrete units, focusing on the forty-nine MSA/CMSAs with a population of 1 million or greater in 2000. These range from Louisville at 1.0 million to New York at 21.2 million, with a distinct break at Philadelphia (6.2 million); only San Francisco (7.0 million), Washington–Baltimore (7.6 million), Chicago (9.2 million), Los Angeles (16.4 million), and New York were larger. In our opinion, this sample provides a better rendering of current- day urban dynamics than the entire range of ∼300 MSA/CMSAs (Frey et al. 2004). Fol- lowing the State of the Nation’s Housing pro- tocol, we employ the percentage point change (PPC) in home ownership for all households, whites, blacks, Hispanics, Asians, and minority- as-a-group (PPC-All, PPC-White, PPC-Black, PPC-Hispanic, PPC-Asian, PPC-Minority).
To highlight that American dream policy outcomes differ dramatically from place to place, consider Figure 1, which graphs the 1990–2000 PPC in ownership for whites alone
Home Ownership, Minorities, and Urban Areas 337
Figure 1 Homeownership percentage point change (PPC) for metropolitan areas, PPC-White versus PPC-Minority, 1990–2000. (Color figure available online.)
(PPC-White) versus change for minorities as a group (PPC-Minority). The 1990s range is from a low of 1.69 PPC-White in Rochester to 8.79 in Las Vegas and for minority own- ership from –2.73 in Louisville to 7.55 PPC in Las Vegas. Also relevant is the linearity of this cross-tabulation, wherein MSAs tend to be low or high for both total/white and minority change (r = 0.649, 0.592 respectively). Further, the low end is marked by Cleveland, Grand Rapids, Louisville, Oklahoma City, Philadel- phia, Pittsburgh, Providence, and Rochester and the high end by Austin, Las Vegas, Den- ver, Houston, and Atlanta—somewhat of an AMB/Rust Belt versus Sun Belt dichotomy—a distinction borne out by other members of the low-low and high-high quadrants.
In addition to finding at least a degree of re- gional differentiation in the outcomes of Amer- ican dream policies, our national picture is
less buoyant than that presented by State of the Nation’s Housing 2009. Hence, even though our calculations are based on 1990 rather than 1994, they are lower for 2000—PPC- All, 2.5 versus 3.4; PPC-Black, 2.7 versus 5.1; PPC-Hispanic, –1.9 versus 5.1; PPC-Asian, 2.1 versus 3.1; and PPC-Minority, 2.5 versus 4.9—only PPC-White, 4.2 versus 3.8, is higher (Table 1).
Clusters of Change To examine MSA differences in greater detail, we again consider the six categories of change but, to better facilitate comparison, expressed as standard deviations from the mean of each; that is, PPC-All-Std, PPC-White-Std, PPC-Black- Std, PPC-Hispanic-Std, PPC-Asian-Std, and PPC-Minority-Std. Then, MSA/CMSAs are grouped using the K -means cluster algorithm of SPSS, Version 17 (2009).6 Six clusters
338 Volume 64, Number 3, August 2012
Figure 2 Metropolitan area groupings based on racial/ethnic similarity in homeownership percentage point changes (PPCs), 1990–2000. (Color figure available online.)
emerge, each represented by an average value for each PPC variable (Figure 2). Because these values reference standard deviations, ∼+0.75 or greater is considered to mark a high range of MSA/CMSA change, ∼–0.75 or less marks a low range, and between ∼+0.75 and ∼–0.75 marks a band of average change.
Cluster I includes twenty MSA/CMSAs that represent average change in home ownership but on the low side of that band. This ap- plies to all racial and ethnic categories except Asian, which falls slightly into the low-range band. Urban agglomerations include several in the AMB (e.g., Buffalo, Cleveland, Detroit, Philadelphia, Pittsburgh), all but one West Coast MSAs, and three in northern Florida (Tampa, Orlando, Jacksonville).
Cluster II includes four MSAs: Atlanta, Austin, Denver, and Las Vegas. These expe- rienced extremely high change by all racial and ethnic categories except Hispanics. By contrast, Cluster III—Grand Rapids, Portland Oregon, and Oklahoma City—falls distinctly in the low-change band for whites, blacks, all, and minority-as-a-group but in the average band for Hispanics and Asians. Cluster V (Charlotte, Greensboro, Indianapolis, Nashville) shows av- erage levels of change, except for Hispan-
ics, whose growth in home ownership lagged considerably.
Greater differentiation between racial and ethnic groups is found in Cluster IV, which represents moderately high change for whites, Asians, and the population overall and Cluster VI, which represents high change for blacks, Hispanics, and minorities as a group. Cluster IV includes eleven MSAs that are largely in the South Central, South- west, and Southeast United States—Dallas, Houston, Memphis, New Orleans, Norfolk, Phoenix, Raleigh–Durham, Salt Lake City, and San Antonio. Cluster VI includes seven MSAs—Boston, Hartford, New York, and Washington–Baltimore, which represent the U.S. East Coast megalopolis; and Chicago, Mi- ami, and West Palm Beach, which also are megalopolis components.
A striking aspect of these groupings is their geography. Cluster I distinguishes constella- tions of the AMB, West Coast, and northern Florida; Cluster VI highlights megalopolis, es- pecially on the East Coast; Clusters IV and V draw attention to Sun Belt segments, espe- cially the South Central but also Southeast and Southwest; and Cluster II represents emerg- ing centers of the New South and Southwest
Home Ownership, Minorities, and Urban Areas 339
(Birdsall et al. 2005; Berube et al. 2010). Given that metros within each regional constellation experienced (more or less) similar American dream policy outcomes, might this connec- tion provide a segue way toward understand- ing why policy impacts differ from place to place?
Another noteworthy aspect is the magni- tude of change.7 Overall ownership increased ∼2.5 PPC between 1990 and 2000 (Table 1), but this is exceeded by Cluster II (∼+5.7 PPC)—emerging centers of the New South and Southwest—where all racial and ethnic categories substantially increased their own- ership stake—and patently trailed by Cluster III (∼+1.2 PPC), where the ownership stake of blacks and Hispanics actually decreased. Regarding particular categories, for the entire set of MSA/CMSAs whites, blacks, Asians, and minority experienced ∼+4.2, 2.7, 2.1, and 2.5 PPC, respectively, but Hispanics decrease ∼–1.9 PPC (Table 1). Considering each cluster, however, emerging centers of the New South and Southwest (Cluster II) show ∼+5.9 PPC for blacks, +4.4 for Hispanics, +8.2 for Asians, and ∼+6.4 PPC for minorities overall. Hispanics also gained significantly in megalopolis MSAs (Cluster VI), ∼+4.1 PPC, but fell behind in Cluster V (Charlotte, Greens- boro, Indianapolis, and Nashville), with ∼–17.1 PPC. Asians include a loss in Cluster I (∼–0.89 PPC), representing the AMB, West Coast, and northern Florida, but gains elsewhere such as ∼+8.2 PPC in Cluster II, emerging centers of the New South and Southwest, and ∼5.8 in Cluster III (Grand Rapids, Portland, and Ok- lahoma City).
Clearly then, American dream policy im- pacts are highly divergent in their spatial manifestation—not only from one urban ag- glomeration to another but also for ethnic groups and clusters of urban areas. It also appears that whites and blacks benefited sig- nificantly more than Hispanics or Asians. Re- garding the latter, however, in-migration can be a distorting factor in that new arrivals might be more likely to rent or cohabit, a practice that especially pertains to the foreign born and that, under the accounting scheme used here, could (and probably does) reduce the relative share of home ownership. Correlation and re- lated analyses, reported later, permit a better understanding of such dynamics.
Correlations of Change The preceding section established that Amer- ican dream policy impacts vary from one urban area to another and exhibit a spa- tial order in terms of regional differences. To explore further, we now consider the re- lationship of PPCs (PPC-All, PPC-White, PPC-Black, PPC-Hispanic, PPC-Asian, PPC- Minority) with variables that might be consid- ered to have explanatory power. Our protocol is to set out a rationale (R) for each set of re- lationships and evaluate it through the relevant zero-order correlations (Table 2).
R-1: One conjecture is that the milieu of change applies to all racial and ethnic groups, such that the PPCs of each correlate sig- nificantly with one another. Table 2 indi- cates that this is generally true in that all correlations are positive, and most are statis- tically significant. Noteworthy, however, is that Hispanic PPCs are significant only with PPC-Minority. This might indicate that the distribution of Hispanics among MSAs dif- fers from other population groups or that forces working toward Hispanic ownership differ from or operate independently of those for other minorities.8
R-2: Earlier research suggests that instrumen- tal factors in housing ownership include life-cycle, liquidity constraint, and user cost effects, as reviewed earlier.9 The results, however, are disappointing. Whereas life cy- cle effects would be hypothesized to increase PPC, instead we find an inverse relation- ship, indicating that MSAs with a higher percentage of the population between thirty- five and sixty years old experienced less change in ownership than those with a lower percentage. The liquidity constraint variable produces a similarly contradictory finding. More generally, a consistent pattern does not emerge and, for the most part, life cycle, liq- uidity constraint, and user cost effects are not statistically significant.
R-3: Another premise is that 1990 ownership levels relate to subsequent change—either by signifying the degree to which demand has been met (indicated by a negative re- lationship) or that the prevailing mood is consistent with further ownership (positive relationship). The saturation hypothesis is supported by white and total ownership in
340 Volume 64, Number 3, August 2012
Table 2 Correlations: Percentage point change and selected variables
Rationale Variables PPC-All PPC-White PPC-Black PPC-
Hispanic PPC-Asian PPC-
Minority
1 PPC-All 1.000 .941∗∗ .521∗∗ .255 .538∗∗ .649∗∗ 1 PPC-White .941∗∗ 1.000 .454∗∗ .260 .574∗∗ .593∗∗ 1 PPC-Black .521∗∗ .454∗∗ 1.000 .272 .272 .867∗∗ 1 PPC-Hispanic .255 .260 .272 1.000 .000 .550∗∗ 1 PPC-Asian .538∗∗ .574∗∗ .272 .000 1.000 .396∗∗ 1 PPC-Minority .649∗∗ .593∗∗ .867∗∗ .550∗∗ .396∗∗ 1.000 2 Life Cycle 90,%35–60 –.074 –.090 .187 –.183 –.034 .090 2 Life Cycle 00,%35–60 –.256 –.306∗ .045 –.271 –.287∗ –.118 2 Life Cycle PPC 90–00 –.281 –.333∗ –.252 –.115 –.399∗∗ –.346∗ 2 Liquidity Constraint 90
(Value/Income) –.238 –.172 .278 .344∗ –.172 .195
2 Liquidity Constraint 00 (Value/Income)
–.169 –.165 .181 .213 –.011 .131
2 Liquidity Constraint PPC 90–00
.226 .107 –.285∗ –.367∗∗ .308∗ –.193
2 User Cost, Value 90–00
.258 .152 –.128 –.233 .399∗∗ –.073
3 Own-%90-All –.329∗ –.440∗∗ –.340∗ –.383∗∗ –.220 –.393∗∗ 3 Own-%90-White –.334∗ –.440∗∗ –.140 –.312∗ –.258 –.191 3 Own-%90-Black .092 .082 –.233 –.169 –.092 –.197 3 Own-%90-Hispanic .180 .093 –.278 –.232 .003 –.203 3 Own-%90-Asian –.117 –.116 –.108 .219 –.440∗∗ –.009 3 Own-%90-Minority .165 .147 –.230 –.044 .003 –.151 4 Growth-%90–00-All .680∗∗ .743∗∗ .313∗ .056 .634∗∗ .390∗∗ 4 Growth-%90–00-
White .635∗∗ .647∗∗ .187 –.135 .597∗∗ .194
4 Growth-%90–00-Black .373∗∗ .445∗∗ .159 .011 .536∗∗ .254 4 Growth-%90–00-
Hispanic .061 .125 .130 –.685∗∗ .430∗∗ –.068
4 Growth-%90–00-Asian .348∗ .396∗∗ .223 –.386∗∗ .408∗∗ .134 4 Growth-%90–00-
Minority .396∗∗ .480∗∗ .136 .100 .593∗∗ .267
5 SameHs-%95–00 –.586∗∗ –.671∗∗ –.205 –.102 –.579∗∗ –.303∗ 5 ForgnBorn-%00 .058 .130 .340∗ .565∗∗ .124 .486∗∗ 5 ForgnBorn-%80–00 .120 .202 .356∗ .534∗∗ .210 .510∗∗ 5 ForgnBorn-%90–00 .195 .273 .394∗∗ .510∗∗ .308∗ .547∗∗ 5 ForgnBorn-%80–90 .024 .106 .295∗ .542∗∗ .080 .444∗∗ 5 ForgnBorn-%lt80 –.077 –.031 .282∗ .588∗∗ –.067 .400∗∗ 6 Pop-#00-All –.181 –.104 .149 .340∗ –.141 .245 6 Pop-#00-White –.182 –.118 .141 .340∗ –.143 .235 6 Pop-#00-Black –.096 –.055 .286∗ .282∗ –.154 .338∗ 6 Pop-#00-Hispanic –.127 –.012 .068 .348∗ –.049 .217 6 Pop-#00-Asian –.236 –.150 .083 .271 –.139 .147 6 Pop-#00-ForgnBorn –.188 –.095 .135 .331∗ –.085 .242 7 HousesBlt-%90–00 .615∗∗ .677∗∗ .248 –.145 .590∗∗ .278 7 HousesBlt-%80–90 .455∗∗ .577∗∗ .253 .171 .476∗∗ .314∗ 7 HousesBlt-%70–80 .333∗ .372∗∗ .046 .197 .391∗∗ .205 7 HousesBlt-%lt70 –.587∗∗ –.678∗∗ –.242 –.042 –.598∗∗ –.323∗ 7 HousesBlt-
MedianYear .571∗∗ .658∗∗ .218 –.032 .598∗∗ .289∗
8 SubPrNum-%04 .170 .161 –.022 –.086 .004 .075 8 SubPrVol-%04 .107 .106 –.026 –.046 –.030 .069 8 SubPrNum-%05 .266 .258 .121 .080 .076 .277 8 SubPrVol-%05 .191 .183 .131 .162 –.004 .284∗ 8 SubPrNum-%06 .221 .212 .164 .212 –.006 .330∗ 8 SubPrVol-%06 .162 .149 .199 .284∗ –.053 .350∗
Note: N = 49. Rationale: These numbers refer to a rationale as to why (or how) each set of variables might relate to the percentage point change (PPC) in ownership for a particular racial or ethnic category. These rationales are elaborated in the text and labeled as R-1, R-2, R-3, and so on. Bolded values are any that are significant at either the .05 or .01 level. ∗Significant at .05 level. ∗ ∗Significant at .01 level.
Home Ownership, Minorities, and Urban Areas 341
1990, which relate negatively to all PPCs. Neither hypothesis, however, is supported (in terms of both coefficient sign and sig- nificance) by black, Hispanic, Asian, or mi- nority ownership in 1990. We thus con- clude that 1990 ownership is not a significant factor.
R-4: Growth of an urban area or its racial and ethnic components might heighten demand or, alternatively, by increasing renting or co- habiting, artificially lower ownership PPCs. In support of the demand hypothesis, per- centage growth between 1990 and 2000 for the entire population relates directly to all PPCs and, except for PPC-Hispanic, signif- icantly so. White and minority percentage growth renders a similar picture.
Confounding, however, is the picture pre- sented by Hispanics and Asians. Hispanic percentage growth is negatively related to PPC-Hispanic, consistent with the hypothe- sis that in-migration artificially lowers our PPC statistic, but this does not hold for Asians (where growth from 1990 to 2000 relates directly and significantly to PPC- Asian). Further, Asian percentage growth relates inversely to PPC-Hispanic, whereas Hispanic percentage growth relates directly to PPC-Asian. As in R-1, this assemblage of outcomes might indicate that the spatial distributions of Hispanics and Asians dif- fer from one another; that forces working toward Hispanic or Asian ownership are dis- tinct from each other; or that that the ob- servation(s) is an aberration stemming from high levels of Hispanic in-migration that sig- nificantly exceeds Asians or artificially dis- torts its home ownership levels.
We conclude that urban area growth, over- all and by its racial or ethnic segments, is strongly related to upward shifts in home ownership. In the case of Hispanics and Asians, however, the process seems to be more complex than simply referencing growth.
R-5: Implicit in R-4 is the degree to which home ownership change reflects actions of 1990 residents or, alternatively, of in- migrants post-1990. One indicator is per- centage of the population in the same house in 2000 as in 1995 (SameHs-%95–00) relative to PPC. We find an inverse rela- tionship for all racial and ethnic categories,
indicating that longer term, rather than re- cent, residents are becoming homeowners. This relationship is not, however, signifi- cant for blacks and Hispanics, suggesting that new arrivals in these groups might become homeowners more quickly than others. It ap- pears, then, that a noteworthy share of new white and Asian owners derive from longer term residents, whereas in-migrants play a more significant role in black and Hispanic PPC.
Supporting this conclusion are correlations with percentage of the population that is for- eign born (ForgnBorn-%00) overall and by decade of entry (e.g., ForgnBorn-%90–00). These relate directly to PPC for all groups and significantly so for blacks, Hispanics, and minorities overall. It seems, then, that the American dream remains a significant moti- vator, especially for new Americans.
R-6: Size of the population overall and of its racial and ethnic segments might impact an area’s attraction or provide a stimulus to real estate developers, financial service firms, lo- cal government, racial and ethnic advocacy groups, community organizations, growth coalitions, and the like. This relationship holds for blacks (Pop-#00-Black), which cor- relates directly (and significantly) with PPC- Black, PPC-Hispanic, and PPC-Minority; similarly, Hispanic population (Pop-#00- Hispanic) and foreign-born as a group (Pop- #00-ForgnBorn) also correlate directly with PPC-Hispanic. Other relationships between population size and home ownership change are not significant. We conclude, then, that larger urban agglomerations are not more likely than smaller ones to receive the ben- efits of American dream policies (contrary, perhaps, to common wisdom) but also that size of a particular racial or ethnic compo- nent might be relevant.
R-7: The era in which an urban area grew up also might impact minority home owner- ship. Some would relate this to community culture wherein older cities are more conservative and bureaucratized, and newer cities are more open to change and flexibility (Mollenkopf 1983; Davis 1991; Goetz 1994). Growth era(s) also reflect the establishment of new neighborhoods, which coincides with increased ownership or better housing stock becoming available to lower income
342 Volume 64, Number 3, August 2012
households, a reverse-filtering process (We- icher and Thibodeau 1988; Somerville and Holmes 2001; Skaburskis 2006; Brown and Chung 2008; Leinbarger 2008). Under this reasoning, an MSA that grew significantly in more recent decades would experience increased home ownership in all racial and ethnic groups, and that likelihood would be greater in the 1990s than in the 1980s, and so on (HousesBlt-%90–00, HousesBlt- %80–90, HousesBlt-%70–80; HousesBlt- %lt70, HousesBlt-MedianYear). This relationship applies to whites, Asians, and the MSA overall but not to blacks and Hispanics. This is, however, consistent with the likelihood of a reverse-filtering process wherein blacks and Hispanics are moving to higher quality housing but not the most contemporary of such.
R-8: Subprime lending is commonly linked to American dream policies, both as a stim- ulus to ownership and, subsequently, as a deleterious offshoot. Nevertheless, we find virtually no relationship between the per- centage of either the number or dollar vol- ume of subprime loans in 2004, 2005, 2006, (SubPrNum-%04, SubPrVol-%04, etc.) and PPC.10 The exceptions are Hispanics with 2006 and minorities overall with 2005 and 2006, but these relationships are only marginally significant.
Multiple Regression Having reviewed the relationship between PPC and single variables, we now turn to multiple re- gression. Structural dimensions represented by our variables are identified through principal components analysis, and the resulting princi- pal component scores are regressed against the PPC for each racial and ethnic group (PPC- All, PPC-White, PPC-Black, PPC-Hispanic, PPC-Asian, PPC-Minority).11
Five dimensions emerge, which account for 83.1 percent of total variance (Table 3). PC- I (+) mirrors MSAs that experienced growth between 1990 and 2000 for all racial and eth- nic groups and whose housing stock tends to be more recent (1980s, 1990s)—in con- trast (–) to MSAs with older housing stock (pre-1970) and little growth. The latter also covaries with the proportion of population between thirty-five and sixty years old (Life
Cycle PPC 90–00), which suggests out- migration by younger age cohorts. PC-II (+) reflects the gross size or population of an MSA, with all racial and ethnic cohorts being rep- resented; that this covaries with liquidity con- straints (median housing value/median family income) suggests an attenuating, or dampen- ing, effect on home ownership. PC-III (+) per- tains to subprime lending, which appears more prevalent in MSAs with a lesser liquidity con- straint. PC-IV (+) indicates growth from 1990 to 2000 in the Hispanic and Asian population segments; covariance with the life cycle variable suggests that this growth tends to be in the age group from thirty-five to sixty years old. PC-V (+) emphasizes change from 1990 to 2000 in the liquidity constraint (+), use cost (+), and life cycle (–) indexes, such as to increase the difficulty of home ownership.
Considering each regression (Table 4), PPC for the entire population (PPC-All) relates di- rectly to growth in all racial and ethnic segments and newer housing stock (PC-I).
A similar relationship also holds for PPC- White. Here, however, there is some indica- tion that white gains are slowed by Hispanic growth, Asian growth, or both (PC-IV, nega- tive relationship, statistically relevant at the 10 percent level).
PPC-Black also relates to overall growth (PC-I), but the size of an MSA (PC-II) also is relevant. Hence, blacks tend to gain in urban settings that are not only faster growing but also larger.
PPC-Hispanic relates to size of an MSA (PC-II). Its inverse relationship with Hispanic and Asian growth (PC-IV) is attributed to our method of deriving PPCs wherein, as noted earlier, the level of in-migration can compro- mise that variable, especially for Hispanics.
PPC-Asian relates to growth in all racial and ethnic segments and newer housing stock (PC-I). More unique is its positive relationship with ownership constraints (PC-V). We sug- gest that this reflects the bifurcation wherein many better-off Asians are located in larger ur- ban areas and are able to purchase in the upscale locales of those places, locales that also would be highly constrained in terms of indexes used here (Alba et al. 1999; Logan, Alba, and Zhang 2002).
Minorities as a group (PPC-Minority) relates to both overall growth (PC-I) and the size of
Home Ownership, Minorities, and Urban Areas 343
Table 3 Principal components, selected variables
Component
Variables I II III IV V
Life Cycle 90,%35–60 .145 .268 –.121 .739 .225 Life Cycle 00,%35–60 –.159 –.018 –.177 .735 –.019 Life Cycle PPC 90–00 –.507 –.495 –.072 –.108 –.421 Liquidity Constraint 90 (Value/Income) –.096 .714 –.455 –.154 –.327 Liquidity Constraint 00 (Value/Income) –.062 .677 –.524 –.141 .170 Liquidity Constraint PPC 90–00 .099 –.456 .169 .105 .835 User Cost, Value 90–00 .140 –.109 –.197 –.021 .918 Growth-%90–00-All .954 –.027 .033 –.003 .113 Growth-%90–00-White .851 –.271 –.064 .109 .201 Growth-%90–00-Black .751 –.134 .112 .012 –.098 Growth-%90–00-Hispanic .561 –.222 –.033 .413 .124 Growth-%90–00-Asian .732 –.275 .220 .407 –.004 Growth-%90–00-Minority .828 .000 –.029 –.016 .060 SameHs-%95–00 –.869 –.059 .050 .307 –.136 ForgnBorn-%00 .176 .797 .021 –.341 –.145 Pop-#00-All –.228 .908 .066 .183 –.058 Pop-#00-White –.261 .853 .037 .258 –.019 Pop-#00-Black –.215 .742 .258 .331 –.065 Pop-#00-Hispanic –.093 .878 .093 –.138 –.149 Pop-#00-Asian –.165 .908 –.192 .019 –.059 HousesBlt-%90–00 .948 –.150 .073 .032 .154 HousesBlt-%80–90 .763 –.039 .039 –.407 –.288 HousesBlt-%70–80 .430 .064 .104 –.703 .141 HousesBlt-%lt70 –.909 .078 –.080 .322 –.001 HousesBlt-MedianYear .905 –.082 .070 –.270 .045 SubPrNum-%04 –.010 –.250 .929 –.021 .039 SubPrVol-%04 –.086 –.253 .915 .041 –.067 SubPrNum-%05 .106 .018 .968 –.110 .100 SubPrVol-%05 .019 .078 .955 –.070 –.019 SubPrNum-%06 .152 .135 .923 –.163 –.011 SubPrVol-%06 .089 .188 .881 –.151 –.113
Explained variance 27.2 20.3 19.3 9.2 7.1 Cumulative 27.2 47.5 66.8 76.0 83.1 Varimax rotation Factor loading > 0.40 or < 0.40, deemed significant for interpretation xxx
Note: PPC = percentage point change.
an MSA (PC-II), thus realizing home owner- ship gains in urban settings that are not only faster growing but also larger. Of further inter- est is the subprime exposure component (PC- III, significant at 10 percent level), which is di- rectly related to gains in home ownership—and only so for minorities as a group, not for other racial and ethnic segments.
In summary, home ownership increases co- vary with growth in an urban area overall and, for minority groups, urban area size. This combination is likely to unleash a dynamic that opens up both newly developed and well- established areas for habitation, attracts en- trepreneurial and community interest in pro- moting home ownership, and provides fertile ground for advocacy groups.
A noteworthy counterpoint, however, is the variance not accounted for by factors consid- ered statistically—from 50.1/51.0 percent for Asians/whites to 67.4 percent for minorities overall and 77.7 percent for blacks. Even if sta- tistically significant, the correlations leave am- ple room for further inquiry.
Following on this observation, consider metropolitan areas that were not well pre- dicted by regressions. Using a standardized residual of ∼+0.75 to indicate an underpre- diction of PPC and ∼–0.75 for overpredic- tion, we ordered PPC-All, PPC-White, and PPC-Minority (each separately). Ideally, and under American dream expectations, an in- crease in ownership for one racial and eth- nic group would be echoed by others. In fact,
344 Volume 64, Number 3, August 2012
Table 4 Regression: Percentage point change (PPC) vs. principal component scores
Dependent variable Statistics
PC-I Growth newer
housing
PC-II Population
size
PC-III Subprime exposure
PC-IV Hispanic,
Asian growth
PC-V Ownership constraints r r 2
PPC-All Beta .551 –.065 .170 –.179 .191 .637 .405 t value 4.687 –.556 1.448 –1.518 1.622 Significance .000 .581 .155 .136 .112
PPC-White Beta .650 .011 .163 –.187 .075 .700 .490 t value 5.971 .103 1.497 –1.719 .687 Significance .000 .918 .142 .093 .496
PPC-Black Beta .310 .276 .078 .110 –.180 .473 .223 t value 2.310 2.056 .579 .816 –1.342 Significance .026 .046 .566 .419 .187
PPC- Hispanic
Beta –.037 .427 .088 –.405 –.178 .622 .387 t value –.307 3.574 .736 –3.390 –1.490 Significance .760 .001 .465 .002 .144
PPC-Asian Beta .631 –.013 –.023 –.134 .287 .707 .499 t value 5.852 –.120 –.217 –1.240 2.658 Significance .000 .905 .829 .222 .011
PPC- Minority
Beta .345 .381 .228 –.069 –.071 .571 .326 t value 2.759 3.046 1.822 –.555 –.571 Significance .008 .004 .075 .582 .571
Note: PPC = percentage point change. Bold values represent statistical significance.
ten urban areas were underpredicted for PPC- White, but only six of these were also under- predicted for PPC-Minority (Austin, Boston, Denver, Chicago, Las Vegas, New Orleans); only seven of the twelve areas overpredicted for white also were overpredicted for mi- nority (Charlotte, Greensboro, Los Angeles, Louisville, Oklahoma City, Portland, Tampa); and the relationship between PPC-White and PPC-Minority (for the entire sample) is r = 0.593 or 35 percent explained variance.
It is a concern, of course, that the levels of explained variance leave much to be desired in terms of better understanding home ownership in the United States. Of more immediate rele- vance, however, are the majority–minority dis- parities. Ten urban areas outperformed expec- tations with regard to PPC-White, but only six of these did similarly for PPC-Minority. Our unease is exacerbated upon considering that the group performing better than expected by our regression (i.e., were underpredicted) had a median PPC of 5.93 for white and 5.21 for minority, whereas the underperforming group had 3.03 for white and 0.02 for minority. Aside from the considerable difference between the under- and overpredicted, this also buttresses contentions that the system works more effec- tively for whites than for minorities, both in
terms of the level of and spatial disparity in ownership. In this comparison, however, note that minorities in more responsive MSAs are only slightly behind whites (5.21 PPC vs. 5.93 PPC median), whereas the disparity in less re- sponsive MSAs is considerable (0.02 PPC vs. 3.03 PPC median).
Why are there such considerable differences among leader and laggard urban areas, which presumably pursue the American dream of homeownership with equal vigor? Similarly, within these cities, why do we consistently find white–minority differences, sometimes starkly so? One hypothesis might focus on local mechanisms that impact the supply, or avail- ability, of houses. Other factors include differ- ences in financial institutional practices regard- ing lending, the degree to which Community Reinvestment Act guidelines are followed, community relations, community activist orga- nizations and their effectiveness, differences in local economies and the economic standing of minorities, community traditions, and the like. We return to this concern at the article’s end.
Post-2000 ACS data enable extension of analysis through 2007.12 Differences from the 2000 Census in- clude alterations to MSA/CMSA boundaries
Home Ownership, Minorities, and Urban Areas 345
Figure 3 Home ownership percentage point change (PPC) for metropolitan areas, PPC-White versus PPC-Minority, 2000–2007. (Color figure available online.)
and joining Miami–Fort Lauderdale with West Palm Beach into a single urban unit. Hence, our sample is forty-eight MSA/CMSAs for the period from 2000 to 2007.13
First consider the graphs of PPC-Minority versus PPC-White (Figure 3). PPC-White varies from –1.13 for Las Vegas to 4.0 for Dal- las and PPC-Minority from –2.93 for Salt Lake City to 8.38 for Providence.14 Like 1990 to 2000, metropolitan areas tend to be either low or high for both all/white and minority owner- ship change, but the 2000 to 2007 relationships are somewhat weaker (r = 0.687, 0.492 respec- tively). Nevertheless, there is a dramatic shift in which MSAs either jump ahead or lag—for example, among the ten highest PPC-All in 1990 to 2000, only two are highest in 2000 to 2007 and four are among the lowest PPC-All in 2000 to 2007—and regarding the ten low-
est for 1990 to 2000 PPC-All, only one is in the same category in 2000 to 2007 and two are in the highest category. Similarly, divergence is found for PPC-White and PPC-Minority. Hence, it appears that population subgroups move somewhat in tandem within a given MSA, but that there is considerable difference be- tween an MSA’s performance from one time period to another.
Regarding how home ownership changes cluster in 2000 to 2007, we again use PPC- All, PPC-White, PPC-Black, PPC-Hispanic, PPC-Asian, and PPC-Minority; each variable is expressed as standard deviations from its mean; and SPSS’s K-means cluster algorithm is used. Because Las Vegas and Salt Lake City are ex- treme outliers (obvious from Figures 3A and 3B), they were not included. Three clusters emerge.
346 Volume 64, Number 3, August 2012
Figure 4 Metropolitan area groupings based on racial/ethnic similarity in homeownership percentage point changes (PPCs), 2000–2007. (Color figure available online.)
Cluster I (Figure 4) includes nine ur- ban agglomerations distinguished by high in- creases in PPC-Hispanic and average out- comes for other segments. Largely located in the Southeast, or New South, Cluster I includes Charlotte, Greensboro, Indianapo- lis, Jacksonville, Louisville, Memphis, Norfolk, Raleigh–Durham, and Rochester. These urban areas are, in general, smaller than others (me- dian population Cluster I = 1,511,231, Cluster II = 2,908,612, Cluster III = 2,263,536) and ex- perienced a higher percentage increase in His- panics between 2000 and 2007 (Cluster I = 87.3 percent; Cluster II = 39.0 percent, Cluster III = 49.8 percent).
Cluster III (Figure 4) includes sixteen urban agglomerations distinguished by high PPC in- creases for the entire population, minorities in general, whites, and blacks. This cluster in- cludes Atlanta, Boston, Chicago, Cincinnati, Columbus, Dallas, Hartford, Houston, Mil- waukee, Nashville, New Orleans, New York,
Providence, Sacramento, San Antonio, and Washington–Baltimore. Regionally, emerging centers of the Southeast are again typical, but we also find South Central representation, ur- ban areas making up the East Coast megalopo- lis, and Chicago as the central point of another megalopolis.
Cluster II (Figure 4) consists of the remain- ing twenty-one MSA/CMSAs, which experi- enced average PPC changes among all racial and ethnic categories. The PPC differential be- tween Clusters II and III is mirrored in the growth of their racial and ethnic groups—that is, comparing Cluster III versus II in terms of median growth, all households = 12.1 per- cent versus 7.7 percent; white households = 6.8 percent versus 3.5 percent; black = 13.2 percent versus 8.5 percent; Hispanic = 49.8 percent versus 39.0 percent; Asian = 45.2 percent versus 40.7 percent; and minority households overall = 23.8 percent versus 20.9 percent.
Home Ownership, Minorities, and Urban Areas 347
Regarding correlations, PPC for racial and ethnic groups are directly related to one an- other, indicating that increased home owner- ship is experienced by all, as in 1990 to 2000. Again, however, Hispanics and Asians appear less connected, in that most of their links are not statistically significant.
1990–2007 Ownership Changes Among Metropolitan Areas To close analyses, we consider the entire period from 1990 to 2007 in terms of all, white, and minority PPCs; for 1990 to 2000, 2000 to 2007, and 1990 to 2007; with urban areas listed in the order of their 1990 to 2007 minority PPC (Table 5). Also shown is the difference between minority and white PPCs for each time period and each MSA/CMSA.
Regarding correlations between PPCs them- selves, within each time era they are positive and significant in all instances, and comparing 1990 to 2000 or 2000 to 2007 separately with 1990 to 2007 yields correlations that are posi- tive and significant in all instances but one.15 Accordingly, there is evidence that minority gains occur more or less in tandem with oth- ers; for example, PPC-Minority-90–07 relates directly to PPC-All and PPC-White for 1990 to 2000, 2000 to 2007, and 1990 to 2007.
But the degree to which minorities also gain ground, or even remain in place (relative to whites), is yet another question. Although the minority–white differential correlates well with PPC-Minority (both positive and significant), it is never significant with PPC-All and PPC- White and often signed in the wrong direction (i.e., negative).
Moving away from the entire set of MSAs, then, we narrow our focus on the PPC- Minority/White link to single urban agglom- erations; that is, to what degree minori- ties and whites move in tandem within a given metropolitan area. Hence, each urban area, for each time period, is designated in one of three categories for PPC-All, PPC- White, PPC-Minority, and the minority–white difference—greater than 0.75 standard devia- tion from the variable’s mean, less than 0.75, and between –0.75 and +0.75 (Table 5).
Considering 1990 to 2007, eleven MSA/CMSAs are in the High PPC-Minority bracket, but only five of these show similarly
high performances for PPC-White (Atlanta, Chicago, Dallas, Houston, Austin). On the other hand, minority gains exceed white in eight of the eleven (Atlanta, Chicago, Washington–Baltimore, Boston, Las Vegas, Hartford, Miami, New York). At the other extreme, twelve MSA/CMSAs are in the Low PPC-Minority bracket; these are matched with seven that had similarly low performance for PPC-White (Oklahoma City, Buffalo, Greensboro, Pittsburgh, Tampa, Cleveland, Philadelphia), and in no instance does a low-performing area show PPC-Minority exceeding PPC-White. Regarding the average performance group for PPC-Minority, which included twenty-five places, seven had high PPC-White (Minneapolis, Denver, San An- tonio, Phoenix, Norfolk, Raleigh–Durham, Columbus), five had low PPC-White (Los Angeles, San Francisco, Portland, Grand Rapids, Rochester), three had minority gains exceeding white (San Diego, Los Angeles, San Francisco), and six had white exceeding minority (San Antonio, Phoenix, Norfolk, Raleigh–Durham, Nashville, Columbus).
Similar mixtures are found for 1990 to 2000 and 2000 to 2007, and if considered simulta- neously, we see that MSA/CMSAs in the 1990 to 2007 high PPC-Minority group are simi- larly high in either 1990 to 2000, 2000 to 2007, or both and likewise for 1990 to 2007 low PPC-Minority. A similar finding holds for PPC-White, PPC-All, and minority–white dif- ference. Further, where PPC classifications dif- fer for an MSA/CMSA, the gap is marginal in the sense that, say, PPC-White is in the high category, PPC-Minority is in the medium cate- gory, but there are virtually no instances where the comparison is high versus low.16
At a local perspective, then, one racial and ethnic group’s experience might differ from that of another, but this is likely to be a dif- ference of degree rather than kind (i.e., some- what lower/higher, better/worse rather than attention/neglect). Nevertheless, the range of numerical differences between places is note- worthy. Hence, for the entire period from 1990 to 2007, minority PPC ranges from 12.08 in Atlanta to –1.77 in Louisville, a 13.85 gap; White PPC ranges from 11.52 in Hous- ton to 3.09 in Pittsburgh, an 8.43 gap; and minority–white differential ranges from 5.07 in Washington–Baltimore to –7.05 in Louisville,
T a b
le 5
H o m
e ow
n e rs
h ip
p e rc
e n ta
g e
p o in
t ch
an g e
(P P
C ) ov
e r
th re
e e ra
s
1 9
9 0
– 2
0 0
0 P
P C
2 0
0 0
– 2
0 0
7 P
P C
1 9
9 0
– 2
0 0
7 P
P C
A m
e ri
c a n
d re
a m
A m
e ri
c a n
d re
a m
A m
e ri
c a n
d re
a m
d is
p a ri
ty d
is p
a ri
ty d
is p
a ri
ty
M e tr
o p
o li ta
n A
re a
A ll
h o
u s e -
h o
ld s
W h
it e
M in
o ri
ty M
in o
ri ty
v s .
W h
it e
A ll
h o
u s e -
h o
ld s
W h
it e
M in
o ri
ty M
in o
ri ty
v s ,
W h
it e
A ll
h o
u s e h
o ld
s W
h it
e M
in o
ri ty
M in
o ri
ty v s .
W h
it e
A tl an
ta 3
.7 2
6 .1
3 6 .9
0 0 .7
7 2 .6
9 3 .1
1 5
.1 9
2 .0
7 6
.4 2
9 .2
5 1
2 .0
8 2 .8
4
C h ic
ag o -G
ar y
3 .4
1 4 .7
9 5
.9 7
1 .1
7 3
.9 3
3 .8
3 6
.0 0
2 .1
7 7 .3
4 8 .6
3 1
1 .9
5 3 .3
2
W as
h in
g to
n -B
al ti m
o re
2 .8
2 3 .4
1 6 .6
2 3 .2
1 3
.0 2
3 .0
0 4 .8
9 1 .8
9 5
.8 4
6 .4
2 1
1 .4
9 5 .0
7
B o
st o n -W
o rc
e st
e r
2 .3
2 3 .8
5 4 .4
0 0 .5
5 3 .5
8 3
.7 1
7 .2
6 3 .5
5 5 .9
1 7 .5
9 1
0 .6
7 3 .0
7
L as
Ve g
as 6
.5 2
8 .7
9 7 .5
5 –1
.2 4
– 1 .6
2 –1
.1 3
2 .6
8 3 .8
1 4 .9
0 7 .7
1 1
0 .2
2 2 .5
1
H ar
tf o rd
1 .5
4 3 .2
7 4
.4 2
1 .1
5 3
.3 9
3 .8
8 5 .7
7 1 .8
9 4 .9
2 7 .1
6 9 .8
6 2 .6
9 M
ia m
i-F o
rt L au
d e rd
al e -P
al m
B e
ac h
3 .0
9 4 .3
6 6 .9
3 2 .5
7 1 .3
5 1 .9
5 2 .7
4 0 .7
8 4 .4
3 6 .3
6 9 .6
5 3 .2
9
D al
la s-
Fo rt
W o rt
h 2 .8
3 5
.6 2
3 .3
0 –2
.3 1
3 .5
4 4
.0 0
6 .0
0 2 .0
0 6
.3 8
9 .6
1 9 .2
8 –0
.3 3
N ew
Yo rk
0 .9
7 3 .3
7 4 .0
4 0
.6 8
3 .0
7 3 .0
0 5 .5
3 2 .5
3 4 .0
4 6 .3
7 9
.0 3
2 .6
6
H o
u st
o n
4 .4
7 7 .5
9 4 .4
2 –3
.1 7
3 .0
6 3 .8
8 4 .2
8 0 .4
0 7
.5 5
1 1 .5
2 8 .6
8 –2
.8 4
A u
st in
6 .9
9 8 .6
1 5 .8
8 –2
.7 2
1 .8
6 2 .7
6 1 .7
5 -0
.0 1
8 .8
9 1
1 .3
5 7 .7
7 –3
.5 7
S an
D ie
g o
1 .6
2 3 .7
0 2 .2
2 –1
.4 7
1 .8
4 1
.5 1
4 .9
0 3 .4
0 3 .4
7 5 .2
6 7 .1
5 1
.9 0
M in
n e ap
o lis
-S t.
P au
l 3 .5
8 5 .3
2 3 .2
1 –2
.1 0
2 .2
6 2 .9
6 3 .8
0 0 .8
5 5
.8 5
8 .2
7 7 .1
3 –1
.1 4
S ac
ra m
e n to
2 .3
4 4 .0
7 1 .7
7 –2
.2 9
2 .8
3 2 .4
4 5
.3 3
2 .8
9 5 .1
7 6 .5
6 7 .0
7 0 .5
1 P
ro vi
d e
n ce
0 .6
7 2 .7
6 –0
.7 4
–3 .5
0 3 .7
4 3
.7 3
8 .3
8 4 .6
5 4 .4
0 6 .4
9 7 .0
6 0 .5
7 M
ilw au
ke e
1 .7
2 3 .3
1 3 .2
1 – 0 .1
0 2 .9
7 3 .7
3 3 .8
7 0 .1
5 4 .7
1 7 .0
0 6 .8
3 –0
.1 8
N ew
O rl e an
s 3 .1
8 4 .4
5 3 .7
7 –0
.6 7
3 .9
5 3 .3
9 2 .9
0 –0
.4 9
7 .1
2 7 .8
0 6 .5
8 –1
.2 2
Lo s
A n
g e
le s
0 .7
9 3 .0
6 2 .6
9 – 0 .3
7 1 .6
5 1
.1 1
3 .5
7 2
.4 6
2 .4
3 4 .1
7 6 .2
7 2
.1 0
D e
nv e
r-B o
u ld
e r
4 .9
3 6 .2
1 4 .5
8 –1
.6 3
1 .8
3 2 .7
9 1
.0 9
–1 .7
0 6
.7 8
9 .0
2 5 .7
4 –3
.2 8
S an
A n
to n
io 3
.7 9
6 .4
0 2 .3
1 4
.0 9
3 .0
4 3 .4
3 3 .2
2 –0
.2 1
6 .8
3 9 .8
4 5 .6
7 – 4 .1
7
S an
Fr an
ci sc
o -O
ak la
n d
1 .2
8 2 .7
3 2 .1
2 –0
.6 2
1 .9
1 2 .0
5 3 .1
9 1 .1
4 3 .1
8 4
.8 0
5 .5
7 0
.7 7
S e at
tl e -T
ac o m
a 1 .9
2 3 .4
5 1 .3
1 –2
.1 4
1 .7
7 2 .0
0 3 .6
5 1 .6
5 3 .6
9 5 .4
3 5 .0
7 –0
.3 6
P o
rt la
n d
1 .1
8 2 .6
8 0 .3
0 –2
.3 8
1 .8
2 1 .8
7 5
.2 8
3 .4
1 2
.9 9
4 .5
4 5 .0
6 0 .5
3 G
ra n d
R ap
id s
1 .3
0 2 .5
8 1 .0
1 –1
.5 7
2 .5
0 2 .3
2 3 .6
0 1 .2
8 3 .8
2 4
.8 9
4 .8
4 –0
.0 5
P h
o e
n ix
4 .2
9 6 .6
5 2 .3
6 – 4 .2
9 0
.8 0
1 .8
9 2 .3
3 0 .4
4 5 .0
9 8
.4 9
4 .6
6 – 3 .8
3
M e
m p
h is
3 .6
7 5 .2
4 4
.5 8
–0 .6
6 0
.4 3
2 .6
6 – 0 .1
2 –2
.7 9
4 .1
3 7 .8
1 4 .6
1 –3
.2 0
N o
rf o
lk 3 .3
4 5
.6 4
2 .2
3 – 3 .4
1 2 .0
1 2 .6
7 2 .2
6 –0
.4 1
5 .3
4 8
.2 4
4 .4
9 – 3 .7
5 D
e tr
o it -A
n n
A rb
o r
3 .0
0 4 .0
6 2 .6
6 –1
.4 0
1 .7
3 2 .3
3 1 .5
0 –0
.8 3
4 .7
5 6 .3
9 4 .1
2 –2
.2 7
348
T a b
le 5
H o m
e ow
n e rs
h ip
p e rc
e n ta
g e
p o in
t ch
an g e
(P P
C )
ov e r
th re
e e ra
s (C
o n ti n u e d )
1 9
9 0
– 2
0 0
0 P
P C
2 0
0 0
– 2
0 0
7 P
P C
1 9
9 0
– 2
0 0
7 P
P C
A m
e ri
c a n
d re
a m
A m
e ri
c a n
d re
a m
A m
e ri
c a n
d re
a m
d is
p a ri
ty d
is p
a ri
ty d
is p
a ri
ty
M e tr
o p
o li ta
n A
re a
A ll
h o
u s e -
h o
ld s
W h
it e
M in
o ri
ty M
in o
ri ty
v s .
W h
it e
A ll
h o
u s e -
h o
ld s
W h
it e
M in
o ri
ty M
in o
ri ty
v s ,
W h
it e
A ll
h o
u s e h
o ld
s W
h it
e M
in o
ri ty
M in
o ri
ty v s .
W h
it e
O rl an
d o
2 .1
1 4 .5
7 2 .9
5 –1
.6 2
0 .5
2 2 .2
9 1
.1 9
–1 .1
0 2
.6 3
6 .8
8 4 .0
8 –2
.8 0
R al
e ig
h -D
u rh
am 3
.7 4
5 .5
2 2 .9
5 –2
.5 7
1 .5
4 2 .6
7 1 .1
4 –1
.5 3
5 .2
8 8
.2 4
4 .0
5 – 4 .1
8
N as
h vi
lle 2 .7
6 4 .1
7 1 .1
6 –3
.0 1
3 .2
9 3 .7
4 2 .7
8 –0
.9 6
6 .0
5 7 .9
3 4 .0
1 –3
.9 2
C in
ci n
n at
i 2 .5
4 3 .5
1 1 .6
0 –1
.9 1
2 .5
3 2 .6
8 2 .7
3 0 .0
5 5 .0
6 6 .2
2 3 .9
8 –2
.2 3
S t.
Lo u
is 2 .6
7 3 .7
0 2 .6
7 –1
.0 4
1 .8
8 2 .3
0 1
.1 0
–1 .1
9 4 .5
6 5 .9
7 3 .7
2 –2
.2 5
C o
lu m
b u
s 2 .5
3 4 .2
8 0
.1 3
–4 .1
4 3
.4 2
3 .9
3 2 .9
8 –0
.9 5
5 .9
4 8 .2
1 3 .2
4 – 4 .9
6
R o
ch e
st e
r 0
.2 7
1 .6
9 –0
.1 7
–1 .8
7 1 .4
8 1 .6
9 3 .6
9 2
.0 0
1 .7
5 3 .3
9 3 .2
4 –0
.1 5
C h
ar lo
tt e
1 .5
5 3 .7
2 1 .0
4 –2
.6 7
– 0 .2
8 1 .5
1 2 .1
5 0 .6
3 1
.2 7
5 .2
2 3 .1
8 –2
.0 4
O kl
ah o
m a
C it
y 0
.4 1
2 .1
7 –1
.3 9
–3 .5
6 1 .4
4 1 .9
5 3 .6
9 1 .7
5 1
.8 5
4 .1
5 2 .4
8 –1
.6 8
B u ff
al o -N
ia g ar
a Fa
lls 1 .6
6 2
.6 9
2 .0
6 –0
.6 4
1 .0
3 1 .8
0 0
.5 8
–1 .2
3 2 .6
9 4 .5
0 2 .1
5 –2
.3 5
G re
e n
sb o
ro -W
in st
o n
-S al
e m
0 .9
9 3 .0
0 0 .8
1 –2
.2 0
-0 .0
4 0 .5
4 1 .3
7 0 .8
3 0 .9
6 3 .5
5 2 .1
3 –1
.4 2
Ja ck
so nv
ill e
2 .5
2 4 .5
1 – 0 .3
6 –4
.8 6
1 .4
2 1 .9
9 2 .5
9 0 .6
1 3 .9
4 6 .4
8 2
.0 7
–4 .4
1
In d
ia n
ap o
lis 3 .1
4 4 .6
6 0 .3
8 –4
.2 8
1 .7
1 2 .6
6 1 .5
2 – 1 .1
4 4 .8
3 7 .3
1 1 .7
6 –5
.5 5
P itt
sb u
rg h
1 .2
8 1 .9
5 –0
.0 5
–2 .0
0 0
.8 1
1 .1
6 1 .3
9 0 .2
3 2
.0 8
3 .0
9 1 .0
8 –2
.0 1
Ta m
p a-
S t.
P e
te rs
b u
rg 1 .4
8 3 .1
1 0
.0 2
–3 .0
8 0
.0 7
1 .2
2 0 .7
8 –0
.4 5
1 .5
6 4 .3
4 0 .7
9 –3
.5 6
S al
t L
ak e
C it
y 3
.9 3
5 .0
7 3 .3
5 –1
.7 2
-1 .1
4 0 .0
2 –2
.9 3
–2 .9
4 2
.8 2
5 .1
6 0
.3 4
–4 .8
2
C le
ve la
n d
-A kr
o n
1 .8
1 2
.9 4
1 .1
8 –1
.7 6
0 .7
9 1 .8
9 -1
.4 7
–3 .3
7 2
.6 0
4 .8
1 –0
.2 4
–5 .0
6
K an
sa s
C it
y 2 .3
7 3 .8
0 -0
.0 7
–3 .8
6 1 .7
8 2 .6
0 –0
.2 9
–2 .8
9 4 .1
6 6 .3
7 -0
.3 3
–6 .7
0
P h
ila d
e lp
h ia
0 .4
4 2 .3
3 –1
.0 0
–3 .3
3 1 .0
5 1 .8
7 0 .3
1 –1
.5 6
1 .4
9 4 .1
8 –0
.7 4
–4 .9
2
Lo u
is vi
lle 0
.9 6
2 .6
9 –2
.7 3
–5 .4
2 2 .2
2 2 .6
2 1 .1
6 –1
.4 6
3 .1
6 5 .2
9 – 1 .7
7 –7
.0 5
P P
C >
+0 .7
5 S
D in
d ic
at e
d by
x x
x
P P
C <
–0 .7
5 S
D in
d ic
at e
d by
x x
x M
e d
ia n
2 .4
5 3 .8
2 2 .2
7 –1
.9 6
1 .8
4 2 .5
2 2 .7
6 0 .4
2 4 .5
0 6 .4
5 4 .7
5 –2
.0 2
A v e
ra g
e 2 .5
1 4 .2
1 2 .3
9 –1
.8 2
1 .8
8 2 .4
2 2 .8
6 0 .4
4 4 .4
0 6 .6
3 5 .1
6 –1
.4 7
S D
1 .4
7 1 .6
2 2 .3
5 1 .8
5 1 .2
8 1 .0
7 2 .2
2 1 .9
3 1 .8
5 2 .0
0 3 .4
9 2 .9
6 0 .7
5 D
e v ia
ti o
n A
b o
v e
M e a n
3 .6
1 5 .4
3 4 .1
5 –0
.4 4
2 .8
4 3 .2
2 4 .5
2 1 .8
9 5
.7 9
8 .1
3 7 .7
8 0 .7
5
0 .7
5 D
e v ia
ti o
n B
e lo
w M
e a n
1 .4
1 2 .9
9 0 .6
2 –3
.2 1
0 .9
3 1 .6
2 1 .2
0 –1
.0 0
3 .0
1 5 .1
3 2 .5
4 –3
.6 9
349
350 Volume 64, Number 3, August 2012
Figure 5 Metropolitan area groupings based on minority and white percentage point changes (PPCs) and the minority–white differential in homeownership, 1990–2007. (Color figure available online.)
Home Ownership, Minorities, and Urban Areas 351
a 12.12 gap. Such differences indicate a signif- icant degree of variation between metropoli- tan areas, which translates to substantial spatial variation within the United States, especially for minorities and the effectiveness of minority catch-up efforts.
Turning then to map representation, the spatial manifestation of PPC-Minority for 1990 to 2007 (Figure 5A) is relatively dis- tinct. High values are found in megalopolis complexes of the East Coast (Washington– New York–Boston corridor), Miami–Fort Lauderdale–Palm Beach, Chicago, and Texas (Dallas–Houston–Austin) and also in Atlanta and Las Vegas. Low PPC-Minority cuts a swath through the AMB. Average change also forms an AMB presence, but its distinct band along the West Coast (including Phoenix and Denver) stands out.
Regarding PPC-White (Figure 5B), high values occur in the megalopolis complexes of Texas (Dallas, Houston, Austin, San An- tonio) and Chicago and scattered across mid-sized MSA/CMSAs such as Atlanta, Raleigh-Durham, and Norfolk in the South; Columbus and Minneapolis in the Midwest; and Denver and Phoenix in the West. Low lev- els of PPC-White occur throughout the AMB (Grand Rapids, Cleveland, Pittsburgh, Buffalo, Rochester, Philadelphia), on the West Coast (Los Angeles, San Francisco, Portland), and scattered locales in the South (Greensboro, Ok- lahoma City, Tampa).
Regarding minority gains (Figure 5C), of particular prominence are megalopolis com- plexes of the East Coast (Washington–New York–Boston corridor), Chicago, Miami, West Coast (Los Angeles, San Diego, San Fran- cisco, Las Vegas), and Atlanta. Places that suffered in the minority–white balance, the low category, include AMB locales such as Cleveland, Columbus, Indianapolis, Louisville, Kansas City, and Philadelphia; places spread across the South such as Jacksonville, San Anto- nio, Nashville, Raleigh–Durham, and Norfolk; and in the West, Salt Lake City and Phoenix.
Concluding Observations
This article applies a spatially refracted lens to efforts that purport to advance the American dream of owning one’s home—long viewed as
a fundamental goal of U.S. domestic policy. Our focus is on urban areas, represented by the forty-nine MSA/CMSAs that housed at least 1 million people in 2000, for the periods from 1990 to 2000, 2000 to 2007, and 1990 to 2007. Following standard procedure in this arena, as represented by the yearly State of the Nation’s Housing report (e.g., JCHS 2009), the variable of interest is PPC; for example, the percent- age of blacks who owned their home in 2000 minus that percentage for 1990. Also in line with standard procedures, our racial and ethnic categories are all, white, black, Hispanic, Asian, and minority as a group. However, whereas JCHS assessments are for the United States as a unit, our concern shifts the scale downward to urban areas or, as our title states, the American dream writ local.
In this context, that home ownership gains (and losses) vary substantially from place to place is a significant finding. Considering the entire housing stock, for example, we find that PPC-All varies from 6.99 (Austin) to 0.27 (Rochester) for 1990 to 2000; from 3.95/3.93 (New Orleans/Chicago) to –1.62 (Las Vegas) for 2000 to 2007; and from 8.89 PPC (Austin) to 0.96 (Greensboro) for 1990 to 2007. Among racial and ethnic groups, the high-to-low range is narrower for whites (e.g., for 1990–2007, 11.52 [Houston] to 3.09 [Pittsburgh]) and greater for minorities (e.g., for 1990–2007, 12.08 [Atlanta] to –1.77 [Louisville]). This is especially noteworthy in that a major goal of recent American dream policies is to increase home ownership among minorities and suc- cess in this is emphasized, if not loudly touted. Nevertheless, our measure of minority minus white PPCs for 1990 to 2007 ranged from 5.07 (Washington–Baltimore) to –7.05 (Louisville), and only fourteen MSA/CMSAs returned a positive score.
Our study thus indicates that, contrary to widely recognized assessments, (1) there is sig- nificant spatial variation in the impact of Amer- ican dream policies; (2) outcomes for particular racial and ethnic groups also vary spatially; and (3) at least among urban areas, whites gain no- ticeably more from these policies than minori- ties do. Is this success?17
PPCs generally move in tandem with one an- other, especially for the total population (all), whites, blacks, and minority as a group, which display correlation coefficients that are both
352 Volume 64, Number 3, August 2012
high and significant for each time frame. This is less so for Asians and Hispanics. Indeed, group- ing MSA/CMSAs on the basis of their PPCs generally yields an all–white–black–minority cluster and, separately, clusters where Hispan- ics, Asians, or both are dominant.
Regarding factors related to the expansion of home ownership, one set of variables repre- sents widely accepted models concerning own- ing one’s domicile, rather than renting—the life cycle, use cost, and liquidity constraint frame- works. For the most part, these did not perform as one would hypothesize.
Instead, we find that expansion largely re- lates to urban growth and, less pervasively, to size of an urban area or its various popu- lation components. Specific variables include the 1990 to 2000 and 2000 to 2007 percent- age growth of each racial ethnic group, pop- ulation size in 2000 and 2007 for each racial and ethnic group, and percentage of houses built in selected decades (e.g., earlier than 1970, 1970–1980, 1980–1990, and 1990–2000). Growth of an urban area reflects the likeli- hood of increased demand, and era in which a city grew reflects past demand. Although less important statistically, population size remains significant as a stimulus to real estate develop- ers, financial service firms, local government, racial and ethnic advocacy groups, community organizations, growth coalitions, and the like. Further, the combination of population growth and size contributes to unleashing a dynamic that opens up both newly developed and well- established areas for habitation; enhances home ownership opportunities, especially for minori- ties; and thereby alters a city’s racial and ethnic landscape.18
Growth and size are, however, only a part of the story, a point brought forward by the vari- ance not accounted for through factors consid- ered statistically—from 50.1/51.0 percent for Asians/whites to 67.4 percent for minorities overall and 77.7 percent for blacks. One en- try point to carrying analysis further is urban areas that are either over- or underpredicted by our statistical models of PPC. With re- gard to white and minority, for example, of the ten MSAs performing better than expected with regard to whites in 1990 to 2000 (stan- dardized residual > 0.75), only six perform similarly for minorities as a group—Austin, Boston, Chicago, Denver, Las Vegas, and New
Orleans. Similarly, of the eleven MSAs that performed worse than expected with regard to whites, (only) Miami–Fort Lauderdale per- forms better than expected for minorities as a group. What is going on? Do these places have a common dimension or differ from others in some fundamental aspect(s)? Another entry point is simply PPC differences. For ex- ample, over the entire period from 1990 to 2007, there is a 13.85 PPC-Minority gap be- tween Atlanta (12.08) and Louisville (–1.77) and a 12.12 gap in the minority–white differen- tial between Washington–Baltimore (5.07) and Louisville (–7.05).
The persistence of place performance also is highlighted by our maps, wherein broad regional effects and differences become ap- parent. Considering Figure 5 especially, sub- regions of positive change include the East Coast megalopolis and Texas conurbations as well as Chicago, Atlanta, and Miami. Positive change also was found for the West Coast megalopolis of Los Angeles–Las Vegas–San Diego–San Francisco but in a dif- ferent way. There, MSA/CMSAs recorded average or low PPC levels but, on bal- ance, experienced net gains in minority home ownership. Performance at less impressive levels, low or medium PPCs and net losses for minorities, characterizes much of the AMB. Nevertheless, despite its AMB connection, Philadelphia sparks interest insofar as it is a prominent element of the East Coast mega- lopolis, seemingly similar to Boston in urban heritage and current-day dynamics, yet consis- tently lags in home ownership growth. This is an enticing comparison—similar in many re- spects but opposite in terms of over- and under- performance relative to the American dream. Why?
On balance, however, the emergence of distinct regional groupings is significant—not simply random patterns or ones that reflect an urban area’s size, racial and ethnic com- position, or other obvious characteristics. Fur- ther, these resonate with areas highlighted by Markusen’s (1991) Gun Belt and Berube et al.’s (2010) seven-group classification of urban areas, not to mention the traditional AMB–Sun Belt dichotomy.19 This suggests that a broad spatial process might be at work, one that deserves further attention as research progresses.
Home Ownership, Minorities, and Urban Areas 353
Obvious elements of this process include those traditionally highlighted by discourses on the post-Fordist transition. Arguably, however, many (most?) of these elements are subsumed by the growth and size of an MSA/CMSA’s population. Although these emerged as major factors in our regression analyses, at best these accounted for 50 percent of the variance in home ownership growth, as noted earlier.
Accordingly, although further analyses might continue on a statistical path, a qualitative approach could be considerably more fruitful—one focused on institutional structures, advocacy groups, financial orga- nizations, and the like. The fundamental question is why there are considerable dif- ferences among leader and laggard urban ar- eas, which presumably pursue the American dream of owning one’s home with equal vigor. Are there local mechanisms that impact the supply or availability of housing or differ- ences in financial institutional practices regard- ing lending, the degree to which Community Reinvestment Act guidelines are followed, community relations, and so on? Are there ma- terial differences in community activist organi- zations and their effectiveness, differences in local economies and the economic standing of minorities, differences in community tra- ditions, and so on? Central to this approach would be key informant interviews, especially of persons who have experience in, or know about, more than one urban area. Further, given the noted differences between Boston and Philadelphia, that comparison might prove rewarding. Alternatively, consideration might be given to MSA/CMSAs that were underpre- dicted for both whites and minorities (Austin, Boston, Denver, Chicago, Las Vegas, New Orleans) or overpredicted (Charlotte, Greens- boro, Los Angeles, Louisville, Oklahoma City, Portland, Tampa)—MSA/CMSAs that repre- sent notably different regional groupings.
Particular attention also might be given to the subprime lending system that has been justified as a segue to affordable housing and an important component of American dream policies. Here, however, variables reflecting subprime activity for 2004, 2005, and 2006 were neither significant nor noteworthy in any analyses—a surprising outcome, even knowing that the subprime system has been diverted to a plethora of other uses. Chambers, Garriga, and
Schlagenhauf (2009) found otherwise, focus- ing on both financial market innovations and demographic structure overall (i.e., not place- specific). In this regard, inquiry at a local level would be appropriate to better understand use of the subprime system, the degree to which home ownership is a motivator for borrowing, and whether subprime lending has led to per- manent gains in minority ownership or taken away gains once made. Regarding the latter, observations on the subprime crisis often high- light high foreclosure rates among minorities, arguing that their gains, especially relative to whites, have been seriously eroded (Hagerty and Gepfert 2007; Leland 2008; Wyly et al. 2008; Kaplan and Sommers 2009). This is cur- rently under study using data on foreclosures and mortgage originations in Columbus and Lima, Ohio, for 2000 to 2008.
In addition to understanding American dream policy impacts on urban agglomerations or nationally, its effects within urban areas also are relevant. Consider, for example, shifts in the social landscape when neighborhoods ex- perience a marked change in minority home ownership and socioeconomic status. Are there deleterious effects as the marginally better off relocate to the urban fringe, leaving behind higher concentrations of the very poor, low- ering the tax base, increasing free-ridership for urban amenities, and so on? Alternatively, is there an increased demand for social services as the result of in-migration or proliferation of minorities and persons of lower socioeconomic status? Are American dream impacts beneficial as the result of increased neighborhood stabil- ity related to a higher level of home ownership or detrimental as the result of lessened neigh- borhood safety, destabilized house values, or an increased need for community services such as public safety, school classrooms, health in- terventions, or English-language instruction? A simple example of such impacts is a demo- graphic shift toward younger people, thus in- creasing the need for school services, or toward older people, thus increasing the need for com- munity services.
This article provides a significant amount of descriptive findings concerning home own- ership differentials among metropolitan areas and some understanding of these occurrences. At the end, however, it points to the need for further research to better understand why
354 Volume 64, Number 3, August 2012
American dream policy impacts have varied significantly—and raises further questions con- cerning the differential effects of these policies on areas within the urban area itself. Greater attention to such matters is warranted, and it is hoped that this thrust will be invigorated by this article. �
Notes
1 Attention also is given to region (Northeast, Mid- west, South, West) and age of householder (un- der thirty-five, thirty-five to forty-four, forty-five to fifty-four, fifty-five to sixty-four, and sixty-five and over).
2 Our concern with urban areas might readily extend to other scales. For example, to what degree are neighborhoods stabilized by higher rates of home ownership or alternatively, due to the demograph- ics of new homeowners, made more dependent on community services?
3 In the 2000 U.S. Census, Consolidated Metropoli- tan Statistical Areas (CMSAs) consist of more than one Primary Metropolitan Statistical Area (PMSA), each of which is more or less equivalent to an MSA; see Frey et al. (2004). Conurbations used here include eighteen CMSAs and thirty-one MSAs.
4 Examples of broad policy impacts with a distinct spatial dimension include Jackson’s (1985) account of suburbanization in the United States and Ko- dras and Jones’s (1990) Geographic Dimensions of United States Social Policy. The latter considers Aid to Families with Dependent Children, food stamps, health, homelessness, neighborhood re- habilitation, and public education. Also, whereas this article emphasizes differences in local con- ditions, an interesting counterpoint is J. Painter (2006), who talks of the prosaic dimension of pol- icy, calling attention to the “mundane practices through which something we label ‘the state’ be- comes present in everyday life” (753); that is, the outcome of policies
depends on and proceeds through mundane practices undertaken by thousands of individ- ual state officials and citizens . . . [providing] considerable scope for . . . qualitative and quan- titative social and spatial variation . . . [imple- mentation] necessarily proceeds unevenly . . . so [that] geographical variations in the provision of health care, policing, education and so on are not “aberrations” but integral to the opera- tion of modern state institutions. . . . The com- plex geographies of central-local relations con- tribute to the production of unintended state effects. (764)
5 MSAs mentioned as examples reflect actual com- putations of representative variables; that is, for life cycle, proportion of the population aged thirty to sixty; for user cost, 1990 to 2000 absolute gain in housing value; and for liquidity constraint, ra- tio of median housing value to median income in 2000.
6 K-means cluster is a method for identifying ho- mogeneous groups of MSA/CMSAs (observa- tions or sample points) based on their values for each of the six ownership change variables. This employs an iterative procedure that minimizes within-group variance and maximizes between- group variance. The authors most commonly use K-means clustering in connection with principal component scores (representing latent dimensions among variables) to group or classify observations (e.g., Brown, Mott, and Malecki 2007). This ap- proach was bypassed here because the six PPC variables are meaningful in themselves, distinct, and comparable to one another when standard- ized.
7 Numbers reported here are actual PPCs, not the standardized.
8 Alternatively, this observation might simply be an aberration stemming from high levels of Hispanic in-migration that, as noted earlier, might artifi- cially diminish (or distort) ownership levels.
9 Life cycle variables (Life Cycle 90,%35–60; Life Cy- cle 00,%35–60) are measured as the proportion of an MSA’s population between ages thirty-five and sixty (i.e., Population 35–60/Total Population); change between 1990 and 2000 is in PPC terms (i.e., Life Cycle 00—Life Cycle 90). Liquidity con- straint variables (Liquidity Constraint 90; Liquidity Constraint 00) are measured as the ratio of median value of owner-occupied housing to median family income; change between 1990 and 2000 is, again, in PPC terms. User cost is only calibrated in terms of its change from 1990 to 2000, calibrated as the difference between the median value of owner- occupied housing for each year (following Smith, Rosen, and Fallis 1988).
10 An important caution here is that our data are ag- gregated and not subdivided according to whether subprime use is for house purchase rather than maintenance, bill payment, lifestyle maintenance, and so on. Accordingly, our results do not con- tradict studies such as Chambers, Garriga, and Schlagenhauf (2009), which find that home own- ership increases are primarily due to mortgage innovations. Also relevant is Brooks and Ford (2007) on “The United States of Subprime” (which also provided the subprime data used in our analyses).
11 The principal components analysis employs only a selection of the variables in Table 2. This was done on the basis of their significance in the
Home Ownership, Minorities, and Urban Areas 355
zero-order correlation analyses, related discus- sion, or conceptual relevance.
12 The ACS replaces the long-form element of the decennial census with monthly data that, through aggregation, will ultimately provide information equivalent to decennial censuses. ACS data are currently available on a yearly basis for geograph- ical units that are 65,000 or larger in popula- tion and three-year aggregations (beginning with 2005–2007, 2006–2008, etc.) for areas with more than 20,000 population. In 2010, the ACS plans to aggregate five years of data, which will provide information for all Census geographies, including block groups and Census tracts. For further de- tails, see Mather, Rivers, and Jacobsen (2005).
13 We will continue to use the term MSA/CMSA. Under ACS terminology, however, MSA could signify either a Metropolitan or Micropolitan Sta- tistical Area (MetroSA, MicroSA). There also is a CSA (Combined Statistical Area) designation, a new unit comprised of “adjacent metropolitan and micropolitan statistical areas” (ProximityOne 2010). Regarding these distinctions and other ACS particulars, see Frey et al. (2004), who, relevant to this article, noted, “For those interested in comparing metropolitan areas across the country, there is now really only one choice: the Metropoli- tan Statistical Area” (5).
14 New Orleans had PPC-All of 3.95, but this is likely distorted by the effects of Hurricane Katrina (Au- gust 2005).
15 The one instance is PPC-White-90–07 and PPC- Minority-00–07, which have a positive but non- significant correlation.
16 Three exceptions are New York, All-PPC, 1990–2000 and Las Vegas, All-/White-PPC, 2000–2007, low in high category; and Salt Lake City, All-PPC, 1990–2000, high in low category.
17 State of the Nation’s Housing reports draw on a na- tional sample, the U.S. Census National Housing Survey, which also would represent rural areas and urban places smaller than 1 million population. Further, tabulations of home ownership from de- cennial censuses and ACS, our data source, can be inaccurate as the result of in- or out-migration and immigration, which particularly pertain to minor- ity groups. Given these drawbacks, our findings are taken with caution but are nevertheless rele- vant in charting directions for future inquiry and raising a flag of skepticism concerning progress elicited by American dream policies.
18 Flippen (2010, 864) concluded somewhat differ- ently that “the key contributor to minority home ownership is . . . a sizable coethnic base, and not how fast the base is expanding” but stated fur- ther that the role of growth is actually a function of “lower housing values, higher share[s] of new housing, and lower segregation in those areas.”
She also found that “residential segregation has a clearly negative impact on minority homeown- ership . . . [that] lower homeownership propensi- ties are not a general characteristic of highly seg- regated cities but rather that segregation affects minorities in particular” (863–64). Related to the latter, correlations between entropy in- dexes of intermixing for each of our forty-nine MSA/CMSAs, drawn from Brown and Sharma (2010), and home ownership PPCs led us to con- clude that a segregation index should not be in- cluded in further analyses.
19 The Berube et al. (2010) groups are Bor- der Growth, Diverse Giant, Industrial Core, Mid-Sized Magnet, New Heartland, Next Fron- tier, and Skilled Anchor.
Literature Cited
Alba, R. D., J. R. Logan, B. J. Stults, G. Marzan, and W. Zhang. 1999. Immigrant groups in the suburbs: A reexamination of suburbanization and spatial assimilation. American Sociological Review 64:446– 60.
Austin, J., and B. Affolter-Caine. 2006. The vital center: A federal-state compact to renew the Great Lakes region. Washington, DC: Brookings Insti- tution Metropolitan Policy Program.
Berube, A., W. H. Frey, A. Singer, and J. H. Wil- son. 2010. State of metropolitan America: On the front lines of demographic transformation. Washing- ton, DC: Brookings Institution Metropolitan Pol- icy Program.
Birdsall, S. S., E. J. Palka, J. C. Malinowski, and M. L. Price. 2005. Regional landscapes of the United States and Canada. 6th ed. New York: Wiley.
Brooks, R., and C. M. Ford. 2007. The United States of subprime: Data show bad loans permeate the nation: Pain could last years. Wall Street Journal 11 October:A1.
Brown, L. A., and S.-Y. Chung. 2008. Market- led pluralism: Rethinking our understanding of racial/ethnic spatial patterning in U.S. cities. An- nals of the Association of American Geographers 98:180–212.
Brown, L. A., T. E. Mott, and E. J. Malecki. 2007. Immigrant profiles of U.S. urban areas and agents of resettlement. The Professional Geographer 59:56– 73.
Brown, L. A., and M. Sharma. 2010. Metropoli- tan context and racial/ethnic intermixing in resi- dential space: U.S. metropolitan statistical areas, 1990–2000. Urban Geography 31:1–28.
Chambers, M., C. Garriga, and D. E. Schla- genhauf. 2009. Accounting for changes in the homeownership rate. International Economic Review 50:677–726.
356 Volume 64, Number 3, August 2012
Clark, G. J. 1994. Strategy and structure: Corporate restructuring and the scope and characteristics of sunk costs. Environment and Planning A 26:9–36.
Clark, W. A. V., M. Deurloo, and F. M. Dieleman. 1984. Housing consumption and residential mobil- ity. Annals of the Association of American Geographers 74:29–43.
———. 2003. Housing careers in the United States, 1968–93: Modelling the sequencing of housing states. Urban Studies 40:143–60.
Davis, J. E. 1991. Contested ground: Collective action and the urban neighborhood. Ithaca, NY: Cornell Univer- sity Press.
Engelhardt, G. V. 1996. Consumption, down pay- ments, and liquidity constraints. Journal of Money, Credit, and Banking 28:255–71.
Flippen, C. A. 2010. The spatial dynamics of strat- ification: Metropolitan context, population redis- tribution, and black and Hispanic homeownership. Demography 47:845–68.
Frey, W. H., J. H. Wilson, A. Berube, and A. Singer. 2004. Tracking metropolitan America into the 21st century: A field guide to the new metropolitan and mi- cropolitan definitions. Washington, DC: Brookings Institution Metropolitan Policy Program, Living Cities Census Series.
Gallin, J. 2008. The long-run relationship between house prices and rents. Real Estate Economics 36:635–58.
Garner, T. I., and R. Verbrugge. 2009. Reconciling user costs and rental equivalence: Evidence from the US Consumer Expenditure Survey. Journal of Housing Economics 18:172–92.
Goetz, E. G. 1994. Expanding possibilities in local development policy: An examination of U.S. cities. Political Research Quarterly 47:85–109.
Guiso, L., and T. Jappelli. 2002. Private transfers, borrowing constraints and the timing of home- ownership. Journal of Money, Credit, and Banking 34:315–39.
Hagerty, J. R., and K. Gepfert. 2007. One family’s journey into a subprime trap. Wall Street Journal 16 August:A1.
Hendershott, P. H., and J. Slemrod. 1983. Taxes and the user cost of capital for owner-occupied hous- ing. American Real Estate and Urban Economics Asso- ciation Journal 10:375–93.
Jackson, K. T. 1985. Crabgrass frontier: The subur- banization of the United States. New York: Oxford University Press.
Joint Center for Housing Studies at Harvard Univer- sity. 2003. State of the nation’s housing 2003. Cam- bridge, MA: Joint Center for Housing Studies at Harvard University.
———. 2009. State of the nation’s housing 2009. Cam- bridge, MA: Joint Center for Housing Studies at Harvard University.
Kaplan, D. H., and G. G. Sommers. 2009. An analysis of the relationship between housing foreclosures, lending practices, and neighborhood ecology: Ev- idence from a distressed county. The Professional Geographer 61:101–20.
Kendig, H. L. 1984. Housing careers, life cycle and residential mobility: Implications for the housing market. Urban Studies 21:271–83.
Kodras, J. E., and J. P. Jones III, eds. 1990. Geographic dimensions of United States social policy. London: Ed- ward Arnold.
Leinbarger, C. 2008. The next slum? Atlantic Monthly 301 (2): 70–75.
Leland, J. 2008. Baltimore finds subprime crisis snags women. New York Times 15 January:A1.
Logan, J. R., R. D. Alba, and W. Zhang. 2002. Im- migrant enclaves and ethnic communities in New York and Los Angeles. American Sociological Review 67:299–322.
Markusen, A. 1991. The military–industrial divide. Environment and Planning D: Society and Space 9:391–416.
Mather, M., K. L. Rivers, and L. A. Jacobsen. 2005. The American Community Survey. Population Bul- letin 60–3. Washington, DC: Population Refer- ence Bureau.
Mollenkopf, J. 1983. The contested city. Princeton, NJ: Princeton University Press.
Ortalo-Magne, F., and S. Rady. 2006. Housing mar- ket dynamics: On the contribution of income shocks and credit constraints. Review of Economic Studies 73:459–85.
Painter, G., and K.-O. Lee. 2009. Housing tenure transitions of older households: Life cycle, demo- graphic, and familial factors. Regional Science and Urban Economics 39:749–60.
Painter, J. 2006. Prosaic geographies of stateness. Po- litical Geography 25:752–74.
ProximityOne. 2010. Metropolitan areas: Assessing competitive position and change. What is a Com- bined Statistical Area?http://proximityone.com/ metros.htm#csa (last accessed 30 December 2010).
Skaburskis, A. 2006. Filtering, city change and the supply of low-priced housing in Canada. Urban Studies 43:533–58.
Smith, L. B., K. T. Rosen, and G. Fallis. 1988. Re- cent developments in economic models of housing markets. Journal of Economic Literature 26:29–64.
Somerville, C. T., and C. Holmes. 2001. Dynamics of the affordable housing stock: Microdata analysis of filtering. Journal of Housing Research 12:115–40.
SPSS. 2009. SPSS, Version 17. Chicago: SPSS, Inc. Verbrugge, R. 2008. The puzzling divergence of
rents and user costs. The Review of Income and Wealth 54:671–99.
Vey, J. S. 2007. Restoring prosperity: The state role in revitalizing America’s older industrial cities.
Home Ownership, Minorities, and Urban Areas 357
Washington, DC: Brookings Institution Metropolitan Policy Program.
Weicher, J. C., and T. G. Thibodeau. 1988. Filtering and housing markets: An empirical analysis. Journal of Urban Economics 12:21–40.
Wyly, E. K., M. Moos, H. Foxcroft, and E. Kabahizi. 2008. Subprime mortgage segmentation in the American urban system. Tijdschrift voor Economische en Sociale Geografie 99:3–23.
Zeldes, S. P. 1989. Consumption and liquidity con- straints: An empirical investigation. The Journal of Political Economy 97:305–46.
LAWRENCE A. BROWN is a Distinguished University Professor in the Department of Ge-
ography at The Ohio State University, 154 N. Oval Mall, Columbus, OH 43210–1361. E-mail: [email protected]. His interests include migration, urban geography, and most recently, race-ethnicity and housing issues, including foreclosure and policy outcomes.
MICHAEL D. WEBB is a PhD student in the De- partment of Geography at The Ohio State Univer- sity, 154 N. Oval Mall, Columbus, OH 43210–1361. E-mail: [email protected]. His research interests include housing markets and gentrification, with particular attention to the political arrangements that support or inhibit infill development and revitalization.
