Policy Matters A Quarterly Publication of the University of California, Riverside

Transcription

1 Policy Matters A Quarterly Publication of the University of California, Riverside EDITORS Mindy Marks Karthick Ramakrishnan EDITORIAL BOARD John Cioffi Exequiel Ezcurra David Fairris Kevin Esterling Linda Fernandez Martin Johnson Robert Ream Ellen Reese Sharon Walker ADVISORY BOARD Terry Amsler Mark Baldassare Steve Cullenberg Anil Deolalikar Gary Dymski Andrés Jiménez Ron Loveridge Mark Pisano Lars Walton Funding provided by the College of Humanities, Arts, and Social Sciences (CHASS) and the Public Policy Initiative at UCR. For more information and archives, visit policymatters.ucr.edu Vanesa Estrada Correa The Housing Downturn and Racial Inequality Executive Summary Racial inequality in homeownership is a key dimension of wealth inequality in America and has been a major concern of housing policy reforms since the 1990s. Since the housing market downturn began in 2006, however, little attention has been given to its implications for racial inequality. Questions of racial inequality are particularly important in diverse states like California, where no racial group can claim majority status. This article addresses three crucial questions which policy makers should consider when evaluating solutions for the current housing crisis. First, did the most recent housing boom alleviate racial inequalities in homeownership? Next, how has the downturn affected disparities across racial and ethnic groups? Finally, how has the downturn affected the economic outlook and housing-related burdens in rapidly changing communities? This article reviews the policy and market decisions which brought about the housing boom and bust, and explores the implications of these decisions for people seeking to become homeowners and those already owning. In terms of the downturn and its effects on racial inequality, this report is a preliminary look at the effects of a still-unfolding economic recession. There are indicators of financial distress and foreclosures which allow us to evaluate how neighborhoods are changing and families are affected. The evidence so far indicates that Latinos and African Americans are more likely to have loans and live in areas that are experiencing the highest rates of foreclosure. Also, mortgage counselors report that minority clients receive worse outcomes than white clients in the foreclosure prevention and loan modification process. Finally, survey evidence from Riverside County, formerly one of the fastest growing regions in the country, indicates that that Latino, African American, and Asian American homeowners are experiencing a far greater financial strain from housing (or housing strain ) than whites in the region. Taken together, these findings suggest that more comprehensive and far-reaching solutions to the housing crisis are necessary to remedy deep racial inequalities in wealth in California and, by extension, other parts of the United States. Vanesa Estrada Correa is Assistant Professor of Sociology at UC Riverside. Her research interests include social stratification, race/ethnicity, migration, demography, urban sociology and public policy. Her current work focuses on housing and neighborhoods. Vanesa can be reached by vanesa.estrada@ucr.edu or by phone at (951)

2 Introduction Inequality in homeownership is a key dimension of wealth inequality in America: not only does home equity represent the majority of most peoples wealth in the United States, but disparities in homeownership are often transfered from one generation to another. Wealth inequality, in turn, is an important component of social inequality in the United States, particularly along racial and ethnic lines. This report focuses on the extent to which the housing crisis of the past several years has affected racial and ethnic disparities in homeownership. Why are these inequalities a matter of public policy? First, research has shown that lower levels of wealth inequality are generally good for sustainable economic growth. Furthermore, higher rates of homeownership in neighborhoods have traditionally been associated with access to better schools, jobs, and public services. Historically, public policies have also played an important role in the prevalence of racially restrictive policies such as redlining and restrictive housing covenants which prohibited residential desegregation and economic mobility via homeownership. (Massey and Denton 1998). At the federal level, homeownership inequalities have received substantial attention in the domestic policy agendas of both Democrat and Republican administrations. Two main goals of housing policies have been to increase aggregate homeownership rates and decrease the racial gap in homeownership. In 1994, Bill Clinton issued an Executive Order requiring federal agencies to affirmatively further fair housing and increase homeownership among young, low-income and minority families. When George W. Bush unveiled his Blueprint for the American Dream in 2002, his goal was to continue work begun in the Clinton administration to make homeownership a reality for at least 5.5 million additional minority families by 2010 and narrow the large gap in homeownership between whites and minorities with public-private partnerships. Along with these national domestic efforts during the 1990s and 2000s, many new mortgage lending products were introduced which were specifically targeted to address the historic barriers for first-time homebuyers. This new generation of affordable mortgages were characterized by liberal financing terms and underwriting standards, including: reductions in downpayments, raising the allowable ratio of mortgage to debt, and establishing credit through alternate means such as rent or utility payment histories for applicants with no prior credit history. Lenders were also allowed to consider borrowers with blemished credit by paying attention to extenuating circumstances such as job loss or illness. Finally, property and neighborhood appraisal guidelines became more flexible to allow for the financing of a broader array of housing in different areas. This new generation of affordable mortgages were characterized by liberal financing terms and underwriting standards, including: reductions in downpayments, raising the allowable ratio of mortgage to debt, and establishing credit through alternate means such as rent or utility payment histories for applicants with no prior credit history. During this time, subprime mortgage lending market grew at a rapid rate. Subprime loans carried a higher rate of interest than prime loans to compensate for the increased credit risk. They were designed for borrowers with blemished or limited credit histories who were unable to qualify for conventional mortgages. Between 1993 and 1999 subprime lending industry grew by 900%. This growth had important implications for racial inequality because minorities were much more likely than whites to get subprime loans, regardless of their income level. Even in upper-income African American neighborhoods, borrowers were one-and-a-half times as likely to have a subprime loan than persons in low-income white neighborhoods (US Department of Housing and Urban Development 2000). How did the housing boom affect inequality? What were the effects of these public policy efforts on racial inequality? Despite many policy changes over the 2 a Quarterly Publication of the University of California, Riverside

3 past decade and a half, the trend in homeownership inequality remained relatively flat. Figure 1 shows the homeownership rates since 1970 for household heads by race/ethnicity. While there were modest gains in aggregate homeownership levels among all groups, the racial gap in homeownership remained unchanged. Figure 1. Percent of household heads who own their home, by race and ethnicity different generations, in order to illustrate differences between the traditional practice of 20% down payments and fixed interest rates, and the more recent policy regime of low down payments and subprime lending. The cohort born in the 1940s was age during the 1970s and early 1980s, and hence their homeownership experiences were under the old policy of conventional mortgages. By contrast, the generation born in the 1960s were young adults in the 1990s and early 2000s, with policies that explicitly sought to make homeownership affordable to young, low-moderate income, and minority families. Figure 2. Proportion of racial inequalities explained by homeownership transitions Source: U.S. Census, calculations by the author. What are the dynamics underlying these aggregate trends? The rates graphed in Figure 1 represent opposing flows of households both entering and exiting from homeownership. Thus, the unchanged racial gap in homeownership over the past four decades could be due to either 1) unchanging rates of entering and exiting homeownership, or 2) offsetting rates of entering and exiting homeownership. My past research (Estrada, 2007) shows that it is the latter: Since the 1990s, the probability of leaving homeownership increased just as the probably of becoming a homeowner also increased. Furthermore, these changes occurred unevenly across racial groups; blacks experienced a much larger increase in the probability of leaving ownership than whites. Thus, despite policy efforts to expand homeownership and increase minority homeownership in particular, there was virtually no change in racial inequality in homeownership. Figure 2 illustrates this finding through a comparison of the homeownership transitions of young adults age for two Source: Panel Study of Income Dynamics (Estrada 2007) What Figure 2 illustrates is that despite the stability in the aggregate level of racial inequality over time, the dynamics underlying this trend have shifted dramatically. Each pie chart describes the relative contribution of transitions into homeownership and out of homeownership to overall racial inequality in the number of years spent owning between ages Transitions out of ownership were only a small contributor to racial inequality in homeownership, accounting for only 18% of the white-black gap for the cohort born in the 1940s. By contrast, transitions out of homeownership grew to account for nearly half of racial inequality for the cohort born in the 1960s. 46% of the A Quarterly Publication of the University of California, Riverside 3

4 3 year gap in years spent owning for the 1960s cohort was due to blacks being more likely than whites to leave homeownership and return to renting. Historically, homeownership loss has been a relatively rare and temporary event for households. Current trends indicate however that ownership loss has emerged as a more common event which now exacerbates racial inequalities in homeownership and offsets the public policy gains made to expand homeownership for racial minorities in the 1990s and 2000s. mitigation, mortgage delinquency and default resolution, predatory lending prevention and post-purchase counseling. The CRC surveys ask to counselors evaluate the challenges facing their clients who are at risk for foreclosure as well as the outcomes of these cases. This provides a unique source of data from the perspective of those working to help homeowners avoid foreclosure and see firsthand what is happening in the relatively new policy efforts to modify and re-negotiate the terms of mortgage loans at risk for foreclosure. The Housing Downturn and Racial Inequality While some analysts have declared that this crisis has affected all homeowners equally, there is substantial evidence indicating that foreclosures have disproportionately affected low income and minority communities. These inequalities are an important consideration to policy makers as they may exacerbate racial inequalities in wealth and class inequalities more generally. Exploring racial differences in housing foreclosure is not straightforward since foreclosure data are not reported by race. There are, however, several sources of data which do include demographic characteristics such as race. I explore two such sources of data here. One source is collected on the ground through a survey of mortgage counselors. These counselors interact with distressed homeowners first-hand and are in a unique position to summarize their experiences. The other source is a view from the air at the geographical distribution of loans both in foreclosure and at risk of foreclosure in the future. Through examining the extent to which these areas overlap with racially segregated communities, we can evaluate whether the impact of this crisis is truly being experienced evenly across residents of different racial and ethnic backgrounds. Problems facing distressed homeowners Since 2007, the California Reinvestment Coalition (CRC) conducted a surveys of California mortgage counseling agencies which provide services to homeowners such as loss Table 1. Mortgage counselors reporting reasons for racial disparities in foreclosure prevention Reason for racial disparity Percent reporting Language barriers with staff 56% Minority clients have lower income 50% Minority clients seriously underwater 41% Minority clients unable to document income 39% Discrimination 18% Don t know 24% Source: California Reinvestment Coalition Evidence from the surveys suggests that there are numerous factors which can account for racial disparities in the foreclosure prevention process. Despite California laws which require translation of mortgage documents, 64% of mortgage counseling agencies in the CRC September 2008 survey reported that it was very common that their clients are not English speakers, yet have English loan documents which they could not read (CRC, 2008). Further, 67% of counselors reported in the March 2009 survey that minorities receive worse outcomes in foreclosure prevention (such as loan modification) relative to white clients (CRC, 2009). Table 1 reports the reasons most commonly given by mortgage counselors in the March 2009 survey explaining why their minority clients received worse outcomes than white clients in the foreclosure prevention process. The most common reason, reported by 56% of counselors, was language barriers with the loss mitigation staff. Also important to note is that 4 a Quarterly Publication of the University of California, Riverside

5 other reasons given pointed to largely structural barriers such as racial disparities in income level and income documentation, or the loan-to-value ratio as a result of disparities in home value depreciation. The least common reason cited was explicit racial discrimination on the part of the lender. the costs of foreclosures to communities have become apparent. When homes foreclose, homeowners and banks are not the only ones who lose money; the losses spread even further. The Federal Reserve Bank of San Francisco estimates that each foreclosure is associated with a 0.9% decrease in value for all properties within 1/8 of a mile. Foreclosed properties which remain vacant without Geography of the housing crisis California has been among the states with the highest levels of foreclosure in the country. The Mortgage Bankers Association has found California to be above the national average in foreclosure starts in its quarterly National Delinquency Survey since This corresponds to the same period in which California s housing price index has fallen (Federal Housing Finance Agency, 2008). Within California, the steepest declines have affected some of the state s most rapidly changing areas. These have been regions of significant population growth and in-migration, areas with newly established communities and burgeoning industries. Table 2 lists the top 20 metropolitan areas nationwide with the highest levels of foreclosure density in August 2008 ( REO means real estate owned, or property owned by a bank). Twelve of these top 20 metropolitan areas are in California, four of which have the highest levels of foreclosure in the country. While these percentages may seem small, REOs are a final stage of foreclosure and considered the most conservative estimate for the extent of the crisis. The table also shows that inland areas of California have been particularly hard hit with foreclosures. These areas correspond to the places where housing prices have declined the most. Since the housing prices peaked in 2006, home prices have fallen more in neighborhoods located further away from urban centers (Carreras, 2008). Analysis of foreclosure starts reveal that the crisis is not distributed evenly across cities. For example, maps of foreclosures by zip code compiled by the Federal Reserve Bank of San Francisco show that foreclosures are concentrated in low and modest income neighborhoods. What are the implications of this trend? As we have seen foreclosures unfold across the state, attention or maintenance can attract crime, pose safety hazards and further reduce neighborhood property values. The average municipal cost for each foreclosure is about $7000 (Carreras, 2008). Table 2. Top 20 MSAs with Highest Estimated Real Estate Owned (REO) Levels, August 2008 A B C Merced, CA Stockton, CA Modesto, CA Riverside/SB/Ontario, CA Las Vegas-Paradise, NV Vallejo-Fairfield, CA Detroit, MI Madera, CA Bakersfield, CA Yuba City, CA Salinas, CA Sacramento area, CA Flint, MI El Centro, CA Jackson, MI Phoenix, AZ Minneapolis-St. Paul, MN Cape Coral-Fort Myers, FL Lansing-East Lansing, MI Fresno, CA Note: A= Unadjusted Prime/Near-Prime REO, B=Unadjusted Subprime REO, C= Total Estimated REO. Source: Federal Reserve Bank of Atlanta. Another way to forecast how the housing crisis may continue to unfold is to consider mortgages which are most at risk for foreclosures in the future. These are high cost mortgages, adjustable rate mortgages which are scheduled to adjust or balloon to higher payments in the future, and subprime loans, A Quarterly Publication of the University of California, Riverside 5

6 which together represent a large share of current foreclosures. While the subprime market grew exponentially during the housing boom and reached 1 in 4 borrowers whom bought a home during that time (FFEIC, 2005), this lending activity was not distributed equally by race and ethnicity. Data collected under the Home Mortgage Disclosure Act (HMDA) requires lenders to report mortgage lending activity, including characteristics of loan applicants and real estate properties. Here I focus on subprime loans specifically reporting findings from my own research and others using HMDA data. This data shows a consistent trend of minorities representing a larger share of loans than white applicants among subprime loans. For example in 1998 subprime loans represented 14% of all loans nationally, but 30% of loans to blacks and 20% of loans to Latinos. In 2007, over 30% of loans to Latino and African American homebuyers in Riverside and San Bernardino counties were subprime loans while fewer than 15% of White homebuyers had subprime loans (Stein, 2009). In addition to the trend of subprime loans made disproportionately to minority applicants, studies also show that they are more likely to occur in minority neighborhoods than white neighborhoods. The Center for Community Capital generated maps of US Congressional districts which illustrate the overlap between low income neighborhoods, minority neighborhoods, and subprime lending. Figure 3 maps Congressional District 45 in Riverside County, by census tract. The panel on top shows low-income neighborhoods and neighborhoods with high concentrations of black and Latino residents. The large area of crosshatching shows there is substantial overlap between these measures. The panel below shows neighborhoods by the percent of subprime loans made in the period. Subprime lending in neighborhoods above 20% is considered high by national standards. The map illustrates that nearly all of the district had a high level of subprime lending activity where 20-40% of loans made between were subprime and some areas had even higher levels of 40-60% of loans subprime. Comparing the maps in both panels show that neighborhoods with high levels of subprime loans also were low income and minority neighborhoods. Figure 3. Subprime lending in Congressional District 45 (Riverside County) by neighborhood income and percent minority Analysis of mortgage application data from HMDA reveals that within Riverside County in 2007, the subprime loans were not only disproportionately given to minority applicants, but also concentrated in minority neighborhoods (Stein, 2009). In other analysis of HMDA data, Estrada and Cort (2007) found that in Los Angeles County African American and Latino loan applicants were actually more likely to have their loans approved when they were seeking to buy a home in a minority neighborhood than a white neighborhood, even after controlling for characteristics such as applicant income and selling price of the home. At the time these loans were made, this pattern served 6 a Quarterly Publication of the University of California, Riverside

7 to reinforce residential segregation among minority homeowners. In the context of the current housing crisis with these loans accounting for the majority of foreclosed properties, it also means that Latino and African American neighborhoods are experiencing higher levels of instability from housing turnover and vacancies. The Housing Crisis and Rapidly Changing Communities While thus far the data presented has illustrated that foreclosure impacts are distributed unevenly, a remaining question is how are communities coping with these impacts? Understanding the consequences of these inequalities may better help us to understand where to concentrate future policy efforts. I take as a case study Riverside County, an area experiencing rapid change. This county has historically been one of rapid growth both in terms of population and industry. In the recent downturn, it has among the highest rates of housing price decline, foreclosure, and unemployment in the nation. In 2008, the UCR Survey Research Center conducted the Riverside County Survey. This survey was a phone survey of a random sample of Riverside County residents asking their opinions on a number of local policy matters, including the housing and the economy. In May 2009, I re-fielded a subset of questions on housing and the economy from this survey to provide an updated picture of Riverside a year further into economic recession. The 2008 survey was fielded over a period of 5 weeks and yielded 1000 respondents; the 2009 survey was fielded over a period of 1 week and yielded 194 respondents. Here I analyze data from both surveys to show how residents from one of the nation s metropolitan areas most impacted by foreclosure and unemployment view their neighborhoods and their futures. What follows is a discussion of descriptive statistics for the variables presented in Table 3. I then present multivariate analyses which examine the effect of individual-level characteristics such as race and income on respondents experiences of the downturn and their future outlook. Table 3. Responses to housing-related items in the Riverside County Survey Housing Strain Avg % of income spent on mortgage Neighborhood Satisfaction Very Satisfied Somewhat Satisfied Somewhat Dissatisfied Very Dissatisfied Don t Know/Refused Do you think buying a home in Riverside County today is Excellent Good Fair Poor Don t Know/Refused Do you think that a year from now you and your family will be Better off Worse off Same Don t Know/Refused Total number of respondents Source: Riverside County Survey (Burnham, Johnson, and Allison 2008, Estrada 2009) Housing Strain Similar to the foreclosure rate, the unemployment in Riverside County is among the highest in the nation. With its high proportion of adjustable-rate and subprime mortgages as well as severe drop in home values, an increasing proportion of its pre-2007 homeowners are face rising housing costs. Housing strain thus serves as one indicator of how citizens may be faring in this economic downturn. Table 3 shows the average percent of income spent on mortgage costs. In the 2008 Riverside County Survey, the average homeowner reported paying about 29% of A Quarterly Publication of the University of California, Riverside 7

8 monthly family income on the mortgage. In 2009, homeowners reported paying 32% of monthly family income on mortgage. These costs are well above the national average, which is 20% for owners (US Census Bureau, 2007). For the purposes of this analysis, I define housing strain as spending more than 30% of income on mortgage costs. Neighborhood Satisfaction Neighborhood satisfaction is an important indicator of how residents are reacting to the ways in which their neighborhoods may be changing. The past couple of years have brought tremendous change not only for residents whom have lost their homes to foreclosure, but also for neighbors whom have seen For Sale signs come up and down in their own blocks. Table 3 shows the response to the question: Overall, how satisfied are you with the neighborhood you live in? In 2008, 8% reported being Very Dissatisfied with their neighborhood and 16% reported being Very Dissatisfied with their neighborhood in This increase in dissatisfaction with the neighborhood could be attributed to multiple causes: it could be the case that long-time neighborhood residents are becoming dissatisfied with changes in their neighborhood, it could also be that movers have left their previous neighborhoods and feel their current neighborhood is worse in comparison. Here I define neighborhood satisfaction as feeling very or somewhat satisfied with the neighborhood. Future Outlook A number of questions in the Riverside County Survey probed respondents to consider the future outlook for Riverside. One such question related to the tremendous decrease in average home values in Riverside Do you think that buying a home in Riverside County today is an excellent, good, fair or poor investment? This question may capture both sentiments about whether homeowners in Riverside are currently reaping benefits as well as whether prospective buyers should look to Riverside as an opportunity for future return. One may interpret, for example, the slight shift towards feeling it is an excellent investment shown in Table 3, increasing from 16.6% in the 2008 survey to 18% in the 2009 survey, to optimism that the housing market has reached its bottom in terms of home prices and that those who buy today are likely to see a growth in their investment. This analysis defines a positive outlook on this item as feeling that buying a home in Riverside is an Excellent or Good investment. Another survey item asked respondents, Do you think that a year from now you and your family will be better off, worse off, or just about the same as now? The distribution of responses to this item in table 3 reveals some uncertainty about whether or not the markets have truly hit bottom and are headed to recovery. While in 2008 most residents seemed to anticipate that their situation would worsen in the upcoming year with 46% responding worse off, in 2009 many still do not see things getting better with 41% responding that they will be the same a year from now and an increased proportion are unsure with the don t know responses increasing from just 1% in 2008 to 8% in In this analysis a positive outlook on this item is responding that you will be better off a year from now. Group Disparities in Riverside While the results so far provide an overall picture of the housing burden in Riverside County, and people s outlooks regarding the housing market and the economy more generally, it is important to also examine if there are differences across groups in the region. It is also important to look at the data when controlling for various types of factors such as income, education, and race to see which factor may be more important than others. Combining the Riverside Community Surveys from 2008 and 2009, I ran a multivariate regression that included the following factors: education level, marital status, whether children under age 18 lived in the household, race/ethnicity, income, and whether the respondent rented or owned. Table 4 presents the results from relationships that were significant after controlling for these factors. The first set of results present differences in housing strain (defined as spending more than 30% of income on mortgage 8 a Quarterly Publication of the University of California, Riverside

9 payments). Having children makes a significant difference, as families with children are 11% more likely to report housing strain compared with families without children, after controlling for other factors such as income. Also families with lower income are more likely to experience housing strain (results available on request). Race also has a substantial effect on whether the respondent experienced housing strain. After controlling for income and other factors, Latinos were the ones most likely to esperience housing strain (63%), followed by Asian Americans (54%), African Americans (53%), and whites (41%). These findings of racial differences in mortgage costs by race are consistent with racial disparities in receiving subprime loans reported earlier in analyses of HMDA data. Table 4. Analysis of housing strain and family economic outlook among Riverside County residents Under housing strain Percent No children in household 44 Children in household 55 White 41 Black 53 Latino 63 Asian 54 Family better off a year from now No children in household 32 Children in household 49 Not married 31 Married 43 Renter 51 Owner 35 Note: Housing strain is defined as the probability of spending more than 30% of income on mortgage costs. The results here are illustrations of relationships that are significant after simultaneously controlling for various factors such as income, education, and race. When running such models, there are no significant racial and ethnic differences in economic outlook. The next set of results present differences in the outlook that I will be better off a year from now. This time, those with children in the household expect to fare better than those without children in the household, and those who are married expect to do better than those who are not married. Finally, we see that homeowners are less likely than renters to expect their economic outlook to improve. Interestingly, however, these results seem to run contrary to the findings in another question on the survey on whether or not buying a home in Riverside is an excellent or good investment. Homeowners are 70% more likely to agree with this outlook compared to renters. These seemingly contradicting findings could reflect the sentiment among homeowners that while real estate may be better for those currently looking to enter the market, their own economic outlooks remain discouraging. The third model predicting neighborhood satisfaction produced no statistically significant results. This is most likely due to the low variation in this variable, as 90% of respondents reported feeling very or somewhat satisfied with their neighborhoods. Nevertheless this is an important indicator to continue monitoring, especially given the extent of rapid neighborhood change in this region. Policy Implications Historically, homeownership loss has been a relatively rare and temporary event for households. In recent years, however, we have seen a dramatic change in national housing markets which have brought record high levels of housing distress and foreclosure. This report has explored the implications for racial inequality of these events through multiple research questions. Here I summarize key findings and explore the public policy implications. First, I explored the impact of the housing boom on racial inequality from a national perspective. The major housing policy efforts made during our housing boom from both Clinton and George W. Bush administrations made substantial efforts to diminish racial inequality through expanding homeownership with targeted programs and products for first A Quarterly Publication of the University of California, Riverside 9

10 time homebuyers and those young, low-moderate income families whom traditionally have not been able to finance conventional mortgages. These policy efforts were also aptly timed with a national climate of economic prosperity and record low interest rates. Through a longitudinal analysis of the homeownership transitions of young households during this period, I found that compared to earlier cohorts whom experienced the previous public policy regime of conventional mortgages, this more recent cohort did experience some gains in attaining homeownership, however these gains were offset by increases in their likelihood to leave homeownership and return to renting. Current trends indicate that ownership loss has become more likely to occur and when it does occur it is less likely that families will become homeowners again in the future. The increase in homeownership loss during this period accounts for nearly half of racial inequality in time spent owning, whereas for the earlier cohort it accounted for less than a fifth of the racial gap. Thus, homeownership loss has not only become a prevalent reality for households, it is emerging as a key factor for racial inequality for homeownership and ultimately a new stratifying factor for wealth inequality more broadly. This finding poses some challenges for housing policy makers. A key lesson is that when barriers to homeownership entry lowered, so did the barriers to exit. The families which were able to take advantage of the new generation of mortgage products we now see have become the most vulnerable to ownership loss due to falling equity in their homes and upwardly adjusting mortgage costs. An emphasis on the next generation of housing policy should concern be the stability of homeownership and sustainability for families over time. Another question addressed here was how the economic downturn has affected racial inequality. This analysis provides some preliminary insights on a still-unfolding crisis through exploring both where foreclosures have occurred and are likely to continue to occur as well as data from a survey of mortgage lending counselors who work to help families avoid foreclosure. Key findings from the CRC survey indicate that many mortgage counselors find that their minority clients receive worse outcomes than their white clients in the foreclosure prevention/loan modification process. They point to structural barriers which should be the focus of future policy reforms in this area, in particular language barriers and the translation of mortgage documents were top cited problems. More generally, the process of loan modification needs attention to explore racial disparities in outcomes explicitly. Analysis of HMDA data showed that the subprime lending market, which grew vigorously during the 1990s, was disproportionately concentrated among minority homebuyers and in minority neighborhoods. While this finding is not necessarily new and has been the subject of many lawsuits to subprime lenders, it raises an important reminder to policy makers to remain vigilant about monitoring predatory lending practices. Finally, analysis of data from the 2008 and 2009 Riverside County Surveys provided some insight into how communities impacted by the housing crisis are coping generally and showed some interesting racial disparities. In general, homeowners in Riverside have a mixed outlook on their economic future, although marriage and income had protective effects which increased positive outlook. With respect to racial inequality, race emerged as a significant factor predicting housing strain with black, Latino and Asian homeowners being significantly more likely to spend more than 30% of their income on mortgage costs than white homeowners. This provides another view into the consequences of racial disparities in subprime lending which occurred in Riverside and illuminates how minority families may be especially vulnerable to homeownership loss as home values fall and unemployment rises in the region. I have presented a picture of our recent housing boom and bust with an emphasis on the consequences for families and communities, specifically with respect to racial inequality. While many economists have forecasted housing trends into the future and estimated market impacts and financial losses, this represents a very narrow conception of the crisis we face. The racial dimensions of the crisis have been understudied for 10 a Quarterly Publication of the University of California, Riverside

11 two main reasons. First, homeownership loss has historically not been prevalent and contributed little to racial inequality in homeownership overall. Second, administrative data on foreclosure are not reported by race. The racial disparities discussed here raise the importance of reforms requiring the reporting of these data by race so that future analyses can explore these issues further. Homeowners in Riverside have a mixed outlook on their economic future, although marriage and income had protective effects which increased positive outlook. With respect to racial inequality, race emerged as a significant factor predicting housing strain with black, Latino and Asian homeowners being significantly more likely to spend more than 30% of their income on mortgage costs than white homeowners. Our society has long considered homeownership to be a public good with its benefits extending beyond individual homeowners to neighborhoods more broadly. And these benefits are measured not only in terms of home value and property tax base, but also a number of intangible aspects of neighborhood quality associated with having a community of stakeholders invested in making a neighborhood the cleanest, safest, and most beautiful it can be. As neighborhood stability erodes with a high volume of housing turnover, to what extent will these intangible benefits also diminish? Future research needs to explore how growing inequalities in homeownership relate to overall inequalities in wealth, and how housing turnover in neighborhoods will affect segregations patterns. The findings presented here provide early indications that the housing market downturn is affecting families and cities unevenly and exacerbating racial and ethnic inequalities. Will recovery happen evenly or will some groups remain stuck behind? What will be the reverberations in the next generation for these financial losses? These inequalities are crucial considerations for policy makers seeking solutions for recovery. The author is grateful to Martin Johnson for providing the 2008 Riverside County Survey data and to Kevin Stein for providing the California Reinvestment Coalition mortgage counselor survey data. References Burnham, Rick, Martin Johnson, and Juliann Allison Riverside County Survey. Edward J. Blakely Center for Sustainable Suburban Development, UC Riverside. Carreras, Joseph Foreclosure Outlook for the Gateway Cities Subregion. Southern California Association of Governments. Estrada, Vanesa Racial Inequality and Homeownership Transitions in the United States, UCLA: Doctoral dissertation. Estrada, Vanesa Survey of Riverside County Residents. [Dataset]. Department of Sociology and Survey Research Center, UC Riverside. Estrada, Vanesa and David Cort Race, Place and Homeownership: A Multi-level Analysis of Los Angeles Mortgage Lending Markets. Unpublished manuscript. Federal Housing Finance Agency Housing Price Index data. Federal Financial Instutions Examination Council Home Mortgage Disclosure Act data. Massey, Douglass, and Nancy Denton American Apartheid: Segregation and the Making of the Underclass. Boston, MA: Harvard University Press. Stein, Kevin Saving Our Homes: Strategies and Solutions to the Inland Empire s Foreclosure Crisis California Reinvestment Coalition. U.S. Department of Housing and Urban Development Unequal Burden: Income and Racial Disparities in Subprime Lending in America. U.S. Census Bureau American Housing Survey, Table Housing and Household Economic Statistics Division. A Quarterly Publication of the University of California, Riverside 11

12 UPCOMING TOPICS No Child Left Behind Others, based on our ongoing CALL FOR SUBMISSIONS Policy Matters VOLUME 3, ISSUE 2 Vanesa Estrada Correa The Housing Downturn and Racial Inequality For interviews or more information, contact: Vanesa Estrada Correa (951) vanesa.estrada@ucr.edu For more information, and to access out archives, visit policymatters.ucr.edu To contact editors via Mindy Marks, mindy.marks@ucr.edu Karthick Ramakrishnan,* karthick@ucr.edu *Action editor for Volume 3, Issue 2 Policy Matters c/o Department of Political Science 900 University Avenue University of California Riverside, CA 92521