Many thanks to Gretchen Minneman and the Spring 2006 Rutgers University Community Development Studio for gathering mortgage foreclosure filings. A very special thanks to the staff at the foreclosure division of the New Jersey Administrative Office of the Courts for facilitating access to the data, to Ken Zimmerman and Ellen Brown at the New Jersey Institute for Social Justice and to Michael Farley, Doris Lewis, Moises Serrano, members of UVSO's predatory lending advisory panel and to the many homeowners who graciously spent their evenings with us. Robert Lake, Elvin K. Wyly, Norman Glickman, Phil Ashton, Dan Hammel, Tom Slater and four IJURR referees provided helpful comments on the manuscript as did Manuel Aalbers, who organized the symposium. All errors or misinterpretations are the responsibility of the author.
Kathe Newman (firstname.lastname@example.org), Program in Urban Planning and Policy Development, Rutgers University, 33 Livingston Avenue, New Brunswick, New Jersey 08901, USA.
For decades community activists fought to increase access to capital for disinvested communities. Now community activists question whether communities have access to capital or capital has access to them. Their concerns are related to the expansion of subprime and predatory lending. The subprime crisis has often been conceptualized as the action of a minor group of predatory lenders, a problem that is individualized, segmented and inherently local. I suggest instead that mortgage capital is the post-industrial widget, the emblematic product of the post-industrial economy. Capital accumulated by owners is extracted from urban communities and flows through brokers and lenders to investors and the subprime lending industry, linking global (and local) capital to place. Place is the node that facilitates capital accumulation, completing the flow from the ethereal world of securities and investment in the secondary circuit of capital to the real-world place of extraction in urban communities. Subprime lending in this context can be seen as a critical component of the financialization of the economy. National housing, macroeconomic and tax policies have expanded the importance of banking and finance within the global and national economy by increasing demand for, and the supply of, mortgage capital. I illustrate the impacts of national financial policy on urban places by examining mortgage foreclosures in Essex County, New Jersey and by talking with residents of one urban community in Newark, New Jersey.
Depuis des décennies, les militants sociaux luttent pour que les communautés en manque d'investissements accèdent plus facilement aux capitaux. Désormais, ils cherchent à savoir si les communautés ont accès au capital ou si le capital a accès aux communautés, inquiets de l'essor des pratiques abusives et des prêts hypothécaires à risque. On a souvent considéré la crise des subprimes comme les agissements d'une minorité de prêteurs au comportement prédateur, situation individualisée, segmentée et forcément locale. À mon avis, les capitaux hypothécaires sont un pur gadget, produit emblématique de l'économie post-industrielle. Les capitaux accumulés par les propriétaires sont prélevés dans les communautés urbaines; ils passent par des courtiers et prêteurs jusqu'aux investisseurs et au secteur des prêts de type subprime, associant des capitaux internationaux (et locaux) à un lieu. Le lieu est la plateforme qui facilite l'accumulation de capital en faisant aboutir le flux qui va du monde intangible des valeurs mobilières et des placements du circuit secondaire des capitaux jusqu'au lieu réel de prélèvement au sein des communautés urbaines. Dans ce contexte, on peut voir les prêts à haut risque comme une composante cruciale de la financiarisation de l'économie. Les politiques nationales fiscales, macro-économiques et du logement ont accentué l'importance de la banque et de la finance dans l'économie mondiale et nationale en accroissant la demande et l'offre de capitaux hypothécaires. Les conséquences de la politique financière nationale sur les lieux urbains sont illustrées par les saisies hypothécaires effectuées dans le comté d'Essex (New Jersey) et par des conversations avec des membres de la communauté urbaine de Newark (New Jersey).
Seventeen ZIP Codes in Newark, NJ, pulled in about $1.5 billion. In all of those ZIP Codes, subprime mortgages comprised more than half of all home loans made (Whitehouse, 2007).
The British-based ‘luxury shirt’ outfitter Thomas Pink has opened a store on Wall Street, a block from the Cipriani Club and across from the Deutsche Bank building . . . Like the flag outside, the store conveys a colorful economic optimism that seems to say, ‘Making money is a rainbow of fun’ (Albo, 2007).
Bear Stearns averted a meltdown this time, but if delinquencies and defaults on subprime loans surge, Wall Street firms, hedge funds and pension funds could be left holding billions of dollars in bonds and securities backed by loans that are quickly losing their value (Creswell and Bajaj, 2007).
The American Dream
Homeownership has been celebrated in the US as a tool to build wealth, revive failing cities and engage the civically disengaged. The US homeownership rate reached 69% in 2006 and many celebrated the vast scores of Americans, including many people of color and lower-income households, who achieved the often elusive American Dream (Wyly et al., 2009, this issue). But, in the US, homeownership is more than the American Dream. It is a central part of the post-industrial economy. Housing-related expenditures accounted for 23% of US gross domestic product in 2005 (Joint Center for Housing Studies, 2006). And home mortgages could be thought of as post-industrial widgets, the raw products necessary for the production of securities, derivatives and the related products of a financialized economy.
The availability of capital fueled redevelopment in cities across the country. Gentrification re-emerged as a central concern in some places, while in others reinvestment was celebrated by cash-strapped cities caught in a competitive neoliberal quest for growth. Many community organizations that had fought for decades to increase access for their neighborhoods to capital began to question whether their neighborhoods have increased access to capital or capital has increased access to them. Initially, they were concerned about predatory lending within the subprime market. Subprime loans include higher interest rates and fees to compensate for the added risk of lending to borrowers who do not qualify for prime loans. Predatory lenders charged excessive fees and interest rates and originated loans that were not beneficial for borrowers (Engel and McCoy, 2002). The subprime market emerged in the early 1990s and expanded rapidly as technological advances like automated underwriting and credit scoring allowed lenders to more finely assess risk and quickly process loans. The increasingly sophisticated use of the secondary market to securitize even high-risk exotic pools of loans by layering credit enhancements provided access to national and global capital. Subprime market share expanded from 8.6% of originations in 2002 to 20% in 2006 (Joint Economic Committee, 2007). In 2005 the total dollar amount of subprime originations was $625 billion dollars (ibid.).
By making capital available to those who could not access it in the past, the expansion of the subprime market had the potential to transform disinvested communities and the borrowers in them. But concerns about the quality of some subprime loans has led some to wonder whether greenlining, a process by which predatory lenders strip capital from neighborhoods and consumers, has replaced redlining. Until recently, most researchers, advocates and lending associations argued that abuses within the subprime market were the result of actions by a minor group of lenders rather than a reflection of industry-wide practices. This argument became hard, if not impossible, to sustain by 2007 as liquidity dried up, foreclosures skyrocketed and the threat of recession loomed over the entire US — if not the world — economy. By October of 2008 the Mortgage Lender's aptly-named ‘Implode-O-Meter’, an online site that tracks the fate of financial institutions,1 had identified nearly 300 financial institutions that had closed, suggesting that the problems in the subprime market, which had been attributed to the actions of a handful of predatory actors, were widespread.
If the subprime crisis is conceptualized as the action of a minor group of predatory lenders, the problem is individualized, segmented and inherently local. But if it is instead seen as embedded in the financial institutional relations that support the economic health of the country, that evolve over time and are globally networked, the problem looks quite different. As finance has become a more significant component of the US economy, the state and the ‘Street’ have shared a common interest in financial services sector growth (Harvey, 1989; Gotham, 2006). Legislative, regulatory and institutional changes facilitated financialization and the growth of the relatively unregulated subprime mortgage market (Newman and Wyly, 2004; NHS, 2004; Chomsisengphet and Pennington-Cross, 2006; Gotham, 2006; Howell, 2006; Apgar et al., 2007).
The financialization of the economy, and in particular the growth of subprime lending, has had direct and real implications for urban communities. The flow of capital from neighborhoods through brokers and lenders to investors and the subprime lending industry links global (and local) capital to urban places. The mortgages provide the connection — the node — to facilitate capital accumulation, completing the flow from the ethereal world of securities and investment in the secondary circuit of capital to the real-world place of extraction in communities (Harvey, 1989).
To understand these connections, I situate the current wave of foreclosures in the broader political economy. I explore the phenomenon of financialization within the transformation of the post-industrial economy and its relationship to place. I review the emerging literature on the role of the state in the expansion of the post-industrial economy, with particular emphasis on changes in the prime and subprime mortgage markets. To illustrate the relationship between financialization and place, I look at mortgage pre-foreclosures in Essex County, New Jersey. Until recently, foreclosures have been explained largely as the result of job loss during economic downturns, housing price decline, or personal events such as divorce or illness. As the volume of subprime originations increased from the 1990s, some researchers and advocates grew concerned about a possible relationship between foreclosures and predatory lending. Poor and/or abusive lending practices, in their view, were playing an instrumental role in foreclosures. The events of 2007–8 suggest that many lenders have been engaging in risky lending practices, especially since 2004. The events magnify the trends in predatory lending and racial/geographic targeting and the effects of these long-term practices on disadvantaged communities and populations.
To illustrate these trends, I turn to a case study of foreclosures in Essex County, New Jersey using data on foreclosures from 2004 to capture foreclosure processes before the recent major expansion of the subprime market. I complement this data with data collected from focus groups with neighborhood residents that suggest a more nuanced picture of the processes through which people access capital. I consider the relationship between the foreclosures in communities and the role of the state in facilitating the financialization of the economy.
Mortgages, the new economy and the city
The increased centrality of credit, finance and banking has been loosely termed ‘financialization’ (Martin and Turner, 2000; French and Leyshon, 2004; Krippner, 2005; Growth and Change, 2007). While the term is used to describe different aspects of finance and the economy, here we are particularly interested in the role that Krippner (2005) explores, a shift from commodity production to finance production. Harvey (1989; 1999) anticipated that the state would play an increasingly important role as fictitious commodities, especially land and money, assumed greater importance in the economy — but he saw finance as distinct from production. ‘Financial markets separate out from commodity and labor markets and acquire a certain autonomy vis-à-vis production. Urban centers can then become centers of coordination, decision-making, and control, usually within a hierarchically organized geographical structure’ (Harvey, 1989: 22). But urban centers have also played a critical role in post-industrial production as places for the production of mortgages, the raw products that fuel the financialized economy.
Reflecting the growing importance of financialization, and the influence of financial institutions in state decision-making, the US federal government (largely viewed as pursuing deregulation policies and a reduction in social policy-making since the Reagan revolution of the 1980s) has been intimately involved in creating policies to increase demand for mortgage products. It constructed an institutional framework that enabled the mortgage-industrial complex, its institutions and actors, to flourish. The UK has adopted similar policies (Martin and Turner, 2000). In much the same way that Keynesiansim guided state policy to address the underconsumption problems of the 1930s, in the post-Keynesian neoliberal age the state has played an aggressive role in ensuring the expansion of the new economy by increasing demand for mortgage products, facilitating investment by de-linking property from place, creating and supporting the expansion of the secondary mortgage market, and changing tax laws to use housing to further commodity consumption (NHS, 2004; Chomsisengphet and Pennington-Cross, 2006; Gotham, 2006; Howell, 2006; Apgar et al., 2007).
In the UK Beaverstock et al. (1992) suggest that there is a relationship between policies that ‘enthusiastically promote home ownership’ and the desire to increase the production and expanded use of credit, or, in other words, to help produce growth by expanding the new economy. In the US, federal policy has shifted since the 1970s towards private market housing solutions. Nixon's moratorium on public housing construction was complemented by a shift to housing vouchers, which are used in the private market, support for federally assisted private multi-family housing, often financed, and later an aggressive push to increase homeownership. One of the central urban programs, the HOPE VI program has facilitated public housing demolition and the resurgence of inner-city markets, re-linking capital to communities (Wyly and Hammel, 1999; Crump, 2002).
The expansion of mortgage production and homeownership was dependent in part on resolving two issues — fixity and liquidity (Gotham, 2006; 2009, this issue). Fixity poses two interrelated problems. First, the market value of a property is dependent, in part, on its location. Second, the investment in real estate traditionally required some knowledge of the property and the place in which it was situated. A series of policy changes initiated in the 1980s helped to de-link property from place by increasing the ease with which investors — even those located on the other side of the globe — could transcend the fixity of property investment (French and Leyshon, 2004; Gotham, 2006; 2009, this issue).
Once de-linked from place, liquidity was necessary to turn properties into commodities. Until the 1980s, most mortgages in the US were originated by depository institutions, whose capacity to make loans was limited by their dependence on deposits and by the need to ensure timely repayment on the mortgages they originated (Stiglitz and Weiss, 1981). The emergence and rapid growth of the US-backed government sponsored enterprises (GSEs) in the 1980s, changed the calculus for mortgage lending. The government's backing of the secondary market provided the support necessary to attract secondary market investors (Follain and Zorn, 1990; Lea, 1990). The ability to package and sell loans to outside buyers meant that originators could increase loan origination volume (dependent on purchases in the secondary market), reduce the cost of lending and loosen underwriting standards.
Since the 1980s, banks, financial investment firms and the US government have enhanced the expansion of the secondary market by increasing liquidity through sophisticated investment tools that, until recently, were thought to all but eliminate investment risk. Collatoralized Mortgage Obligations (CMOs) ‘tranched’ or divided loans into categories of risk allowing investors to invest based on their desired risk and investment return (Kendall and Fishman, 2000; Wyly et al., 2006; 2009, this issue). Changes to federal legislation in the 1980s, such as the 1984 Secondary Mortgage Market Enhancement Act (SMMEA), expanded Wall Street's ability to use these new securities. The 1989 Financial Institutions Reform, Recovery and Enforcement Act (FIRREA) further encouraged secondary market expansion by mandating higher capital requirements for thrifts, pushing them to sell loans they originated into the secondary market (Follain and Zorn, 1990; NHS, 2004; Gotham, 2006). Securitization, tranching and credit enhancement helped to facilitate investment by de-linking it from place and commodifying real estate (Beauregard, 1994; Gotham, 2006). As these processes evolved, investment in housing became ever more intricately woven into the US economy. The Tax Reform Act of 1986 changed the rules on tax deductions, creating an incentive for homeowners to use their homes as banks to fund college educations, purchase new cars, and finance home renovation, further installing housing and finance as a critical segment in the broader economy (NHS, 2004; Chomsisengphet and Pennington-Cross, 2006; Howell, 2006; Apgar et al., 2007).
These regulatory and legislative transformations, along with changes within the banking industry, increased the volume of home mortgage lending and expanded opportunities to many borrowers. But even with these changes, borrowers with less than perfect credit had few mortgage options until the 1990s. Increased liquidity and risk management created through the secondary market, combined with a suite of technological innovations including automated underwriting and credit scoring, facilitated the expansion of subprime lending. Such developments, not incidentally, also fueled the growth of the prime mortgage market (Crews Cutts and Van Order, 2004). Subprime lending can be viewed as an evolutionary innovation that increased the avenues for capital accumulation in the land and housing markets. The later introduction of exotic or nontraditional mortgages and expanded use of ARMs can be thought of in much the same way (Beaverstock et al., 1992). For the financial industry to expand, it needed to originate more loans. Given finite demand for prime loans, lenders increased production by tapping new markets such as subprime borrowers.
The state played a key role in the expansion of the secondary market and the evolution of subprime lending. By creating the secondary market and regulating securities to expand housing production, the federal government facilitated the rise of non-bank mortgage companies that became major actors in the subprime market. At the same time, by creating an alternative source of capital which fueled the growth of the non-bank lenders, securitization reduced the regulatory impact of the Community Reinvestment Act (CRA), which does not regulate non-bank lenders (Follain and Zorn, 1990). In the 1980s, then President Reagan introduced New Federalism, a new relationship between the federal and state governments that would devolve authority and responsibility to the states. During the same time period, the federal government overrode state rules on lending. The 1980 federal Depository Institutions Deregulation and Monetary Control Act (DIDMCA) allowed lenders to exceed state government interest rate caps, permitting high-cost subprime lending. Two years later, Congress passed the Alternative Mortgage Transaction Parity Act (AMTPA) allowing negative amortization, adjustable interest rates and balloon payments, making subprime loans look affordable, while allowing hidden, often predatory, terms (Mansfield, 2000; Chomsisengphet and Pennington-Cross, 2006).
Engel and McCoy (2002: 1259) argue that residents of inner-city communities are more likely to receive subprime credit because they have become ‘disconnected from the credit market and hence are vulnerable to predatory lending hard sell tactics’. Predatory lenders use information asymmetries to exploit such market disconnection (ibid.). Throughout the lending process, lenders and mortgage brokers have more information than borrowers, producing significant power differentials. Borrowers often lack a complete understanding of the borrowing process and their legal rights and rarely obtain their own legal representation (Engel and McCoy, 2002; TRF, 2005; Wyly et al., 2009, this issue). In surveys and interviews with borrowers, the Community Reinvestment Committee (CRC) (Stein and Libby, 2001) and The Reinvestment Fund (TRF) (2005) found that many predatory lending victims do not understand the lending process or shop for their loan, use subprime credit even when they qualify for prime credit, do not understand the terms of their loans and do not read documents at closing. In the CRC study, 36% of respondents did not read their loan documents. The rationales for not reading are telling: ‘61.9% reported that the documents were too lengthy; 38.1% felt pressured; 23.8% reported that the documents were too complex; and 19% reported that they trusted their loan representative’ (Stein and Libby, 2001: 27). The information asymmetries are so great that even the savviest consumer has little chance of identifying the best-priced loan.
There is mounting evidence that subprime and predatory lenders use sophisticated marketing techniques to reach people with little lending experience, education, mobility or access to alternatives (Carr and Kolluri, 2001; Quercia et al., 2004; Stone, 2008). In its survey of borrowers, CRC found that 80% of African American borrowers with incomes above $45,000, who thought they had good credit, did not approach a bank before approaching a subprime lender for a loan (Stein and Libby, 2001). Internet technology and mortgage brokers enable subprime lenders to reduce costs by eliminating branch offices. Lenders reach consumers through direct marketing and mortgage brokers who act as loan officers on the ground as well as home improvement contractors. As subprime lending boomed, the number of mortgage brokerage firms also increased — from 7,000 in 1987 to 44,000 in 2002 — and brokers originated 45% of subprime originations in 2002 (Apgar and Calder, 2005). Broker-originated lending is considerably different from direct institution lending and broker-originated loans are more likely to have poor terms. ‘Borrowers with broker-originated loans were more likely to pay points (25 versus 15%) and more likely to have a loan with a prepayment penalty (26 versus 12%)’ (ibid.: 33). For those who are most in need, with little time or the ability to shop, these offers can be appealing. In one study, 38% of respondents reported that the idea to take out a home loan came from marketing, and the proportions reached 40% for African American respondents and 44% for Latino respondents (Stein and Libby, 2001).
Despite the potential for abuses, brokers, home improvement contractors who link borrowers to financing, and many subprime lenders are not closely regulated (Gale, 2001). Most prime lenders are regulated by one of the four federal regulatory agencies and are subject to periodic reviews and CRA requirements (Immergluck, 2004; Immergluck and Smith, 2005). But subprime capital flows through different conduits and a different regulatory structure (MacDonald, 1996; Apgar et al., 2007). Less than 2% of subprime loans are sold to GSEs, producing what Apgar et al. (2007) call ‘channel specialization’. White borrowers access credit that is originated and travels through channels that are more closely regulated than the credit used by black borrowers: ‘some 44.2% of all blacks (versus 30.1% of whites) obtain a loan from less heavily regulated independent mortgage companies’ (Apgar et al., 2007: iv).
Even though some state governments have passed legislation to limit lending practices associated with predatory lending, such as repeated refinancing without benefit to the consumer and balloon loans, states face significant limitations regulating capital (Quercia et al., 2004). Federal preemption, capital's flexibility and assignee liability, along with media campaigns and pressure from lenders, rating agencies and investment groups that threaten to withdraw from lending activities, all limit state regulatory effectiveness. States that seek to regulate predatory lending face aggressive campaigns by lenders and lending institutions (Newman and Wyly, 2004).
In summary, in the US mortgages were the raw products necessary for the production of financial products that fueled the financialization of the economy. The federal government facilitated the financialization through macro-economic policy and regulatory strategies. Institutional changes within the financial services sector facilitated a new style of lending. The volume of subprime non-bank lenders and subprime mortgage originations soared as lenders sought new markets to produce the raw products for these financial products.
Since subprime lending emerged in the 1990s, researchers and advocacy groups have tried to determine whether there was a connection between foreclosures and subprime lending. In order to do so, they measured the incidence of foreclosure, examined whether the problem was growing worse over time and where it was concentrated, and determined the extent to which subprime lenders originated loans that ended in foreclosure. Studying foreclosure is complicated by a few intersecting problems. First, foreclosure is a process rather than a single event, which creates data collection and measurement challenges. Foreclosures can be measured at the pre-foreclosure stage, when mortgage owners announce their intent to foreclose, or at any point thereafter such as when properties are sold at foreclosure auction. Measuring the process at the pre-foreclosure stage may overestimate the problem as some borrowers may resolve the delinquency or refinance. Measuring it at the auction sale stage misses borrowers who sell properties early, a problem in both hot and declining markets where buyers sell to avoid the costs of foreclosure and to preserve their credit rating. Second, measuring the extent to which foreclosure is a problem is complicated by the challenges involved in calculating a foreclosure rate that meaningfully captures the foreclosure problem and is comparable across places. The numerator could be pre-foreclosures or foreclosure auction sales and the denominator could be the total number of housing units or occupied housing units or the number of loans originated. Because of the different iterations, few foreclosure studies are comparable. A 5% foreclosure rate might be extraordinarily high in one context and negligible in the next, depending on how foreclosures were measured. Third, and not inconsequentially, foreclosure data in most places are not digitally available from public sources and purchasing the data privately is costly, introduces data quality problems and omits much of the data that are contained within the foreclosure documents. The Home Mortgage Disclosure Act (HMDA) provides for an annual national dataset on mortgage applications and originations, but no comparable dataset exists for foreclosures. Fourth, the role of subprime lenders in foreclosure is a primary concern but gathering data that characterize loans as subprime is often difficult if not impossible. Instead, proxies are often used, such as identifying subprime loans based on the characteristics of the lender rather than the loan (Newman, forthcoming).
The National Training and Information Center (NTIC) (1999), a historic participant in the US community reinvestment movement, conducted one of the earliest foreclosure studies in Chicago and a three-county surrounding area based on mortgage foreclosure filings for foreclosures that went to auction between 1993 and 1998. Foreclosures doubled during the study period, subprime lenders accounted for 36% of foreclosures in 1998, and many loans foreclosed in less than 4 years. Subsequent studies replicated this initial study and produced similar findings, including short times to foreclosure, increasing foreclosure rates over time and active subprime participation. Gruenstein and Herbert (2000) studied foreclosures in Atlanta between June 1996 and December 1999 and found a decrease in total foreclosures, a 232% increase in subprime foreclosures, and that subprime loans foreclosed in half the time (2 years) of non-subprime foreclosures. Foreclosures in Cuyahoga County, Ohio increased from 2,582 in 1995 to more than 12,000 in 2005 (Dimora et al., 2005). Foreclosure starts in Chicago and a five-county area skyrocketed from 7,433 in 1995 to 25,145 in 2002 (Immergluck and Smith, 2005; 2006). Many were in neighborhoods with high minority populations and extensive subprime penetration. Newman and Wyly (2004) also identified a connection between foreclosures and subprime lending that was not explained by borrower characteristics.
The Reinvestment Fund (TRF) has conducted some of the most comprehensive and far-reaching foreclosure studies in Philadelphia, Pennsylvania, Delaware and New Jersey. TRF matches foreclosure starts and auctions with loan-level data, including originating lender and loan histories for individual properties. Their understanding of local markets and lenders allowed them to more accurately identify subprime lenders, including those not included in a list of subprime lenders prepared annually by the US Department of Housing and Urban Development, presenting a more complete picture of subprime involvement in foreclosures. When combined with interviews with borrowers, industry representatives and other actors involved in mortgage lending, TRF's studies present a comprehensive account of mortgage foreclosure. In a study of foreclosure starts between 2000 and 2003 for the Pennsylvania Department of Banking, they found that subprime lenders originated 60–75% of foreclosures. Many of these subprime loans were concentrated in moderate-income communities, many of which had sizable minority populations. Even if there is nothing predatory about the loans, they point out that the number of foreclosures and the spatial concentration of subprime lending mean that foreclosure too is spatially concentrated (TRF, 2005).
As researchers struggled to gather data and craft methodologies that would explore whether there was a relationship between subprime lending and foreclosures, the link was visibly made in 2007 when the number of foreclosures surged, producing an international fiscal crisis. Before the economic downturn of late 2007 and 2008, researchers and policy makers anticipated that foreclosure rates on subprime loans originated between 1998 and 2006 would reach 18–20% (Schloemer et al., 2006; JEC, 2007). The US Congress Joint Economic Committee (2007) estimates that there are 179,873 outstanding subprime loans in New Jersey and expects that 35,117 of those will go into foreclosure between the third quarter of 2007 and the fourth quarter of 2009. As the health of the economy declines, foreclosures will likely increase.
The incidence and characteristics of foreclosure can be illustrated through a case study of Essex County, New Jersey. These loans were originated before the most dramatic problems in the subprime market appeared and before the considerable loosening of underwriting practices. They suggest the processes that were at work throughout the 2000s and grew into the substantial problems of recent years.
Foreclosures in Essex County
Essex County lies just west of New York City in northern New Jersey. It is home to Newark, the state's largest city and some of the state's wealthiest suburban communities. In between is a ring of once-thriving inner suburban communities that have suffered disinvestment for more than two decades. The county and most of its constituent communities, including Newark, experienced a rapid upswing in real estate prices from the end of the 1990s. Community organizations that were previously able to access land for free from the city found themselves in competition with for-profit private developers as early as the late 1990s. Essex County was selected for a case study to build on prior work and to build capacity to analyze these processes over time (Community Finance Research Initiative, 2001; Zimmerman et al., 2002; Newman and Wyly, 2004; Community Development Studio, 2006).
The data for this study are 2004 pre-foreclosure filing records gathered by hand from paper legal records at the New Jersey State Administrative Office of the Courts Foreclosure Division in Trenton, New Jersey between Spring 2006 and Winter 2007 and entered into a spreadsheet. Removing tax foreclosures, commercial foreclosures, and multi-family buildings produced a dataset of 2,191 residential mortgage foreclosures.2 The database was enhanced by coding loans to female borrowers based on borrower first name and by identifying originators as subprime using the list prepared by the US Department of Housing and Urban Development (HUD) (FFIEC, 1993–2004; Scheessele, 2005).3 To find the percentage of subprime loans in the 2004 pre-foreclosure dataset, HMDA identification codes were identified for 77% of the loans originated between 1992 and 2004.4 The data were cleaned and geocoded producing 2,083 records that matched 80-100% and 51 that matched at less than 80%.5
Two additional data sources provide context to understand the pre-foreclosures in Essex County. Properties sold at foreclosure auction from 1991 to 2002 suggest the concentration and spatial distribution of foreclosures during the period of subprime expansion (Community Development Studio, 2006). Four focus group interviews in 2003 with 23 homeowners in the Vailsburg section of Newark shed light on the processes through which borrowers got loans, the role of home improvement contractors, mortgage brokers and other local actors, and borrower experience with home loans and lending processes in their community.
Geography of foreclosure
The geography of the 2004 pre-foreclosures matches spatial patterns identified in earlier work (Newman and Wyly, 2004). Pre-foreclosures in Essex County are concentrated in relatively stable, predominantly African American and Latino neighborhoods in Newark and in the adjacent inner-ring cities/suburbs of Irvington, East Orange and Orange. They are sparsely distributed in the County's wealthy western suburbs. Within the city of Newark, foreclosures are absent from areas that were targets of mid-century urban renewal projects that involved major land clearance for the construction of public and federally assisted housing. Foreclosures are also notably absent in the predominantly white communities and in the Ironbound, a predominantly immigrant Portuguese and Brazilian community on Newark's south east corner. Foreclosures are concentrated in the majority-minority low-rise residential neighborhoods, many of which have been the target of focused community development efforts, receiving governmental, private foundation and bank resources. Maps and planning documents from Urban Renewal and Model Cities programs of the 1960s and 1970s show these neighborhoods as places that did not experience the extent of disinvestment and arson that plagued other parts of Newark during that period. They were not torn up by highway construction, nor were they the sites of massive scale high-rise public housing construction.
To better understand where foreclosures were concentrated, I calculated the percentage of loans that were originated between 2000 and 2003 that went into foreclosure in 2004. The Home Mortgage Disclosure Act provides a count of the loans originated each year by census tract. Using loan originations as the denominator provides a numerical indication of areas with severe foreclosure concentrations (see Figure 1) (FFIEC, 1993–2004).6 Seven of the 213 Census tracts in Essex County had 10% or more of the loans originated between 2000 and 2003 enter foreclosure in 2004 (three tracts in Newark, three in Irvington and one in East Orange).7 The foreclosure rate exceeded 5% in 54 Census tracts. In some neighborhoods and regions, the foreclosure rate was much higher. The Upper Clinton Hill neighborhood in Newark's South Ward and the north section of Irvington saw 7% of the loans originated between 2000 and 2003 enter foreclosure in 2004.8
Recent newspaper headlines suggest that loose underwriting in the last few years largely explains the subprime crisis. But Sheriff sale data for 1991–2002 suggest a concentration that looks strikingly like that exhibited by the 2004 pre-foreclosures and a problem that was underway before 2005. Of the 8,763 Sheriff sales in Essex County during the decade, most were concentrated in Newark (3,255) and the surrounding inner-ring cities of Irvington (1,573) and East Orange (1,268) (Community Development Studio, 2006). If we know that these same neighborhoods experienced many foreclosures in the past, is the problem merely a case of poor underwriting decisions in specific cases or are the foreclosures related to a broader set of institutional actors and processes?
The foreclosure filings include information that sheds some light on the processes, institutions and actors at work. The top ten originators of loans in pre-foreclosure in 2004 account for one-quarter of foreclosure starts. Half of these are subprime lenders. And half originated loans that foreclosed in less than 2 years. The average interest rate for loans originated in 2003 that went into foreclosure in 2004 is only slightly above the average prime rate. If interest rate reflects the risk of lending, then these loans to not-so-risky borrowers went into foreclosure very quickly. Another group of lenders originated loans with much higher average interest rates that exceed 11%. These loans took longer to foreclose and women accounted for 44% to 63% of the borrowers.
While interest rates taken alone do not indicate a definitive loan cost, they do provide a picture of the lending that produced the foreclosures. A handful of foreclosures have astronomically high interest rates that look like high credit card rates rather than home mortgage rates. These include a 23.89% interest rate on a $21,680 20-year loan originated in 2002 in Bloomfield and a 21.9% interest rate on a 30-year $30,000 loan in Newark. The interest rates on nearly a fifth of the foreclosure starts were 10% or more.9 To compare the foreclosure filing interest rate with the prime rate at the time of origination, I calculated an average interest rate for the preforeclosures each year for 30-year loans and compared it with the average annual interest rate on prime 30-year loans (Freddie Mac, 2007).10 The two interest rate curves look similar until 1995 when they diverge and the foreclosure-starts interest rate exceeds the prime rate. Between 2000 and 2004, the foreclosure filings interest rate is 1.41 to 1.7 points above the prime rate. The average interest rate on subprime loans is higher than the average foreclosure-start interest rate but it does not fully make up the difference, suggesting that at least some of the other loans are subprime.
The length of time between loan origination and foreclosure provides yet another piece of the puzzle regarding loan quality and/or borrower stability. The 2004 Essex County pre-foreclosures went into foreclosure on average in 3.9 years; subprime loans entered foreclosure in an average 2.6 years. Sixty-three percent of all 2004 pre-foreclosures were originated between 2001 and 2004 and a quarter were originated in 2003. Eighteen percent went into foreclosure less than a year after origination.
Pre-foreclosure filings include the name of the originating lender. This information was used in combination with HMDA to provide the name of the regulating institution at the time of loan origination. Since lenders sometimes change regulators, for each pre-foreclosure, the originating lender was matched to the corresponding HMDA record for each year between 1993 and 2004. For loans originated in 2003, lenders not regulated by one of the four federal regulating agencies accounted for at least 60% of the pre-foreclosures in Essex County. Between 1993 and 2004, 45% of the foreclosure filings were originated by lenders that were not regulated by one of the four federal regulatory agencies; 15% were regulated by the Office of the Comptroller of the Currency, 7% by the Federal Reserve System, 4% by the Federal Deposit Insurance Company, 7% by the Office of Thrift Supervision, and 21% were unidentifiable (See Figure 2). The data appear to suggest that lenders that are not regulated by one of the four federal regulatory agencies played a significant role in originating loans that went into foreclosure and may indicate an increased role for these lenders in 2003 compared to 2000.
The concentration of pre-foreclosures in particular communities, relatively short times to foreclosure, high interest rates, and a possibly increasing role for subprime lenders that are not regulated by one of the four federal regulatory agencies provide some insight into the processes at work, but offer little to explain processes within neighborhoods on the ground. Foreclosures are usually explained by life events and recently they have been described as loans to people who were overly ambitious or naive. But how did these borrowers get the loans and why? To better understand the processes that produced the pre-foreclosures and the context within a neighborhood, we turn now to the Newark neighborhood of Vailsburg.
The view from Vailsburg
Vailsburg is a moderate-income, predominantly African American neighborhood on the west side of Newark between the city of East Orange to its north and the city of Irvington to its south. Vailsburg began to shift from white to black in the 1970s. New residents were middle class but lacked access to mortgage loans because of redlining. Focus groups with Vailsburg homeowners presented a radically different picture of access to capital after 2000. Residents describe a neighborhood bombarded by capital's intermediaries — home improvement contractors, lenders and mortgage brokers — who sought to increase their business through mailings, phone calls and in-house visits. One resident explained: ‘I get a lot of mail, 2 or 3 letters, pieces a day, approved refinancing, sometimes 5 or 6’. Another lamented: ‘My shredder is full of junk mail. I get stacks of information about home improvement loans’. Residents pointed out that the advertisements often sell products unaffordable to the borrower. One resident explained: ‘My husband is retired but they are always offering him loans for $50,000 or $88,000. He can hardly pay the mortgage’.
Even though the advertisements are a nuisance, some residents bought the services offered because they found it difficult to identify alternatives. But those services often came at a high cost. Many of the interviewees noted that contractors often ask about financing before they look at the job that needs to be done and they offer to link homeowners up to financing if they don't already have it. One homeowner explained his experience of getting quotes for his roof: ‘I wanted to put a new roof on my house. The guy went directly to my house. He didn't look at the roof and he started talking about financing to me directly. He didn't look at the roof! He walked directly from the car to the house and you want to talk to me about how much it would cost to put a roof on?’ Residents indicated that they were encouraged to take out loans that exceeded their immediate needs. Home improvement contractors suggested that residents could take out home improvement loans that exceeded their home improvement needs and they could use the extra money to pay for vacations, cars, education, or simply the bills. One resident explained: ‘Home improvement contractors send you a letter and it says the balance of your home is ‘X’ and we can give you ‘X’ amount of dollars to do home repair or more if you want to buy a new car’.
A frequently heard criticism suggests that borrowers would not fall into these lending traps if they were more financially literate and/or had adequate disclosure of loan terms. But Engel and McCoy (2002) have challenged this notion by pointing out the extreme information asymmetries between borrowers and lenders. Even the most savvy borrower could be compromised in a situation where the actors making the loans are preying upon potential borrowers. The Vailsburg residents were savvy and experienced. Nearly all were banked, more than half knew what a credit score was, and two monitored their credit scores. But having access to a savings account and knowing a credit score are not much protection considering the complexity of the mortgage process. Few, in Vailsburg or anywhere else, fully understand points, interest rates, loan durations, fees, or adjustable rate or exotic mortgages.
Residents identified closings as a particularly confusing and stressful time. The volume of legal papers and the complexity of the disclosure rules are overwhelming. More importantly, some homeowners, who discovered problems at closing or at some earlier point during the lending process, found that they had little choice but to proceed. In some cases, residents had to close to purchase new homes. In other cases, homeowners had invested time and money; each fee that residents pay puts them further into the process and the further in, the more compromised they feel. Walking away from a lender midway has practical implications — lost time from work, the time to find new lenders, lost fees, and possibly the loss of their home. The only resident who seemed entirely confident that he had been treated fairly at his mortgage closing used an attorney.
Residents of Vailsburg felt that their neighborhood was a target for unscrupulous actors. One resident explained: ‘Minority neighborhoods are more susceptible to being victimized. We're seen as being less knowledgeable about obtaining contractors for repairs. We're charged a higher rate and higher fee and usually the quality of the work is substandard. And the financing is suspect . . . they will query you about how you can repay the loan, personal savings, have a line of credit or they are willing to establish a line of credit with a lending institution’. Homeowners in Vailsburg, like homeowners in other neighborhoods that have lacked access to financial institutions, are vulnerable to high cost and abusive lending. The residents were clear that they are subjected to aggressive home loan and home improvement marketing and lack accessible alternatives. Vailsburg residents offer some insight into how loans reach community residents. They offer a portrait of a neighborhood that is subject to aggressive home mortgage marketing and residents who lack alternatives.
Once harmed by disinvestment related to redlining, the urban areas of Essex County are now being harmed by an influx of capital related to greenlining. Aggressive marketing pushed loans into communities. The pattern of foreclosures shows concentrations in majority black and Latino neighborhoods with larger homes and equity to strip. For many, homeownership in these communities was not an avenue to wealth, neighborhood stability or urban revitalization.
Mortgage foreclosures can be understood as the plight of individual borrowers who overextended themselves or as the product of unscrupulous lenders seeking to make a profit. Another explanation could be that mortgage lending became intricately woven into the economy through the process of financialization. While the state helped to facilitate economic growth related to financialization, it may not have sufficiently regulated new economy processes and practices. Mortgages acted as post-industrial widgets which helped to link the urban to global financial markets. The effect on the urban has been just as, if not more dramatic than the effect of the re-emergence of the city as the command and control center of the new economy, which has been seen as a driving force producing jobs and demand for housing. The effect on disadvantaged populations and urban places is only beginning to be seen and the full effects are, as yet, unknown. Homeownership has been heralded as the American Dream, a tool to build intergenerational wealth, revitalize disinvested cities and stabilize neighborhoods. And financialization seemed, until recently, to bring little but growth and wealth to many. But with the global economy in turmoil and increasing numbers of borrowers, communities and governments desperately seeking strategies and funding to rescue people, homes, communities, investments and institutions, it is critical to fully explore the processes and institutional relationships that produced the financial crisis.
The initial list totaled 3,068 mortgage foreclosure filings. There were 22 private sales, 54 home equity lines of credit, 75 non-residential and 665 tax foreclosures, reducing our list to 2,191 bank foreclosures. In retrospect, removing the tax sales may result in underestimating the problem. Some borrowers may not realize that their taxes are not in escrow and, while they pay the loan servicer, they may not pay their taxes and thus lose the home to tax foreclosure.
The HUD list identifies lenders as subprime if more than half of their loans are subprime in any given year. This procedure is limited because subprime is defined by the characteristics of the lender rather than the loan. Additionally, small lenders are not required to file HMDA data and are missed using this method.
The unidentified lenders include lenders that do not file HMDA reports and lender names that were unclear in the foreclosure documents.
Addresses were checked using Google Maps to correct misspelling, zip code changes, and other errors. Missing addresses were identified using block and lot numbers through the New Jersey Association of County Tax Boards online parcel identification system (http://njactb.org/).
Both efforts underestimate the extent of the foreclosure problem because they only capture loans that foreclosed in 2004, missing those that foreclosed before and after. They also miss the cumulative effect of foreclosures over time. The foreclosure rate calculated as the number of foreclosure starts by owner-occupied housing units produces a rate of 5% in Irvington and East Orange, 3.7% in Newark, 3.5% in Orange and 1.3% in Belleville and South Orange, similar to what was produced using loan originations.
This does not mean that only 10% of loans originated during that period went into foreclosure; it means that 10% or so foreclosed in 2004, so it likely underestimates the total foreclosure rate of loans originated during that period.
Recent work that collected foreclosure filings for 2005–8 shows considerably higher foreclosure rates in these same places that exceed 20% of loans originated in 2005 and 2006 in some census tracts (Newman, 2008).
Loan position was not identifiable from the foreclosure filings.
Many of the loans with excessively high interest rates were for terms shorter than 30 years.