Financing Community Development in the US: A Comparison of “War on Poverty” and 1990s-Era Policy Approaches

Abstract

This paper first analyzes the role of finance in community development, and then contrasts 1990s development-financing policies with those undertaken decades earlier in the context of the federal “War on Poverty.” It is shown that the Clinton-Bush era programs in support of community development were less extensive than War on Poverty programs. Further, they were weighted toward large banks and large commercial real-estate projects, and did not take into account the implications of the changing strategies and activities of private-sector banking firms. Consequently, programs implemented in the 1990s and afterward by the Clinton and Bush Administrations had relatively little impact. This paper then suggests some lessons for the new Administration.

Introduction

At the end of April 1992, Los Angeles erupted in violence after the televised brutal beating of Rodney King by Los Angeles police officers. The episode paralleled the 1965 Watts riots. It seemed momentarily that if that one long historical cycle—urban violence, policy intervention, benign neglect, festering social inequalities—had been completed, another had begun.

But history never performs the same drama twice. The 1965 riots were centered in Watts and bounded by a police/military blockade around a predominantly African American area known thenceforth as South Central Los Angeles.¹ The 1992 events also began in South Central, but soon stretched to Koreatown, Culver City, even trendy Melrose Avenue. And in 1992, many Latinos, having recently moved in large numbers to South Central, were involved. The targets of violence were often stores owned by Koreans, who had become the principal “middleman minority” operating stores in inner-core Los Angeles. It was widely noted that the 1992 uprising constituted the first multi-cultural riot in contemporary U.S. history.²

The 1992 civil unrest in Los Angeles helped launch a new round of federal policies, aimed at spurring asset-building and business activity in lower-income and non-white communities. This marked the first sustained federal urban-policy effort since the 1960s. Here too, differences between 1965 and 1992 policy responses are striking. This paper analyzes these post-1992 federal policies, notably the Community Development Financial Institution (CDFI) program and the New Markets initiative, which aim at building assets and strengthening businesses in lower-income and non-white communities. This analysis first considers the role of finance in community development in general terms; then, it contrasts 1990s development-financing policies with those undertaken decades earlier in the context of the federal “War on Poverty.”

We proceed as follows. “Conceptualizing banking/financial and asset-building practices: a framework” investigates the basic elements of asset-building processes in urban areas, especially in lower-income urban areas. This discussion points out that financing structures are at the heart of asset-building processes; and these structures pose challenges for any development effort, whether in upper-income suburban or lower-income inner-core areas. “A short history of urban asset-building initiatives before 1990” then summarizes federal urban programs aimed at asset-building and business development, from the 1960s through the Nixon era. “The transformation of banking and financial practices” then examines another crucial context of the newer federal urban initiatives—the transformation of banking and financial-industry practices. Then “Contemporary asset-building in urban areas: from CDFIs to New Market tax credits” returns us to Los Angeles in 1992. It describes the federal urban asset-building policies of the Clinton and Bush Administrations.³ “Evaluating the post-1990 federal financing policies” then analyzes these newer federal asset-building urban policies—CDFI and New Markets—in light of past federal policies and of the changing structure of urban financial markets. “The new urban policy and evolving financial markets: some conclusions” draws out some policy implications.

Conceptualizing banking/financial and asset-building practices: a framework

Long years after Watts burned the first time, there remains an appalling lack of clarity about basic concepts in urban development—beginning with the notion of “community-development financial institution” itself. This term became centrally important in policy discussion and advocacy in the 1990s. In the past decade, a global movement centered on the problem of financial exclusion and advocating the creation of CDFIs has emerged. Consider this passage from Inspiration for Abroad, written for an Australian credit union (Rees 2002):

The Community Development Financial Institution (CDFI) movement began in the United States in the 1990s and has spread to the UK. CDFIs range from small non-profit micro-enterprise lenders with a few hundred borrowers, to substantial banking institutions with thousands of borrowers and loans worth hundreds of millions of dollars. Whatever their size, CDFIs share a commitment to investing in local, disadvantaged communities to build wealth, generate jobs, improve housing and stabilise communities.

This report goes on to list five types of CDFIs: community development credit unions; community development banks (CDBs); community development loan funds (CLFs); community development corporations (CDCs); and micro finance funds. The second and third definitions are of special interest here. CDBs are described in Inspiration from Abroad as for-profit regulated entities that “target disadvantaged communities to provide banking services, loans, and community revitalisation programs. There are only a few such banks in the United States.” By contrast, CLFs are “non-profit, unregulated and uninsured entities. They administer loan funds for community development ... Some offer limited micro-enterprise programs. [They] can make loans that banks and credit unions would hesitate to make ... Many accept private investment. Offerings tend to be below market rate.”

The statement in the above quote is accurate, as is the list of five institutional types. But the evident enthusiasm shown herein for creating new institutions and approaches—repeated in innumerable other programs, including New America’s asset-building program—should be tempered by some lessons from past experience, and by a deeper appreciation of practices in banking markets in the contemporary economy.

For one thing, virtually every commercial bank the world over would affirm its “commitment to investing in local, disadvantaged communities to build wealth ..” This core feature of the definition of CDFIs does not exclude non-CDFI banks. Second, the notion that a CDFI “movement” arose in the 1990s represents a misreading of history. CDCs were created some 30 years before the 1990s, based on the developmental needs and circumstances of largely segregated, lower-income urban areas in the U.S. The notion of socially responsive banking had been an object of community-based struggle in US cities and rural areas at least two decades before the 1990s. Micro-finance came into being informally long before that; but as a formalized contemporary movement, its roots lie in the Grameen Bank model launched in Bangladesh two decades ago. This mischaracterization of the historical trajectories that underlie these five institutional forms leads to a view that the Clinton Administration’s CDFI program was the apotheosis of a unique new movement to recruit finance to the cause of economic growth and social justice. The possibility that the modest CDFI initiative of the 1990s might lack the capacity to make good on its promises emerges only if a longer historical frame is adopted.

Another problematic element of the above characterization is the implicit claim that the CDFI movement is somehow independent of banking and financial market dynamics. CDBs are characterized as for-profit entities targeting lower-income communities; but the description of community-development loan funds acknowledges that these funds make loans that “banks and credit unions would hesitate to make.” So which is it—is the CDFI movement able to show banks the world over how to “do good by doing well” in market competition; or is this movement an attack on the market?

The observation that there are only a few CDBs in the U.S. suggests that African-American-owned and other minority-owned banks are excluded in this characterization. For example, recent experience suggests that minority-owned banks, as for-profit entities, have had to be cautious in loan making—to the point that scholarly researchers have examined whether they discriminate by race (Black et al. 1997). Does the fact that there are loans they “hesitate to make” mean that minority-owned banks should not be considered CDBs or potential members of a global CDFI “movement”? If minority-owned banks were considered not to be members of the CDFI movement, then what does that suggest about the relationship between non-white communities and banks and this movement? The tension between seeking adequate returns and meeting social needs is, in any case, at the heart of the dilemma of financing community development (Brimmer 1972; Jegen 1998). Only by considering explicitly the dynamics of financial markets can the role of non-market and/or community-asset-building institutions be understood.

More broadly, the very term “community development financial institution” permits other hard choices and persistent challenges to be overlooked. It suggests that the provision of finance has something intrinsically to do with community development, and that community-based economic development will somehow follow as a logical accompaniment of a well-implemented financing scheme—whether that scheme emphasizes micro-loans or mezzanine financing.

These ideas are problematic because, simply put, financial practices are present in any community, whether or not any community-development finance effort has been undertaken therein. Such financing efforts, when undertaken, often fail. It is tempting to generalize from individual case-studies, as proponents of the Grameen-Bank microfinance model sometimes do. We must, however, leave open the possibility of a more sober assessment, as Brimmer (1992) has advised. Analytical sobriety, in this case, starts with an appropriate analytical framework.

A Framework

To begin, we can identify three core financial services: the payments mechanism—that is, the receipt of income and the payment of bills owed; financial savings instruments; and credit or capital processes, wherein money is provided in advance of income. Income/payment and credit services sometimes are interlocked: for example, pay-day loans permit payments to be made in advance of receipt of income via paycheck-secured debt. In the U.S., most economic units obtain these services by maintaining accounts with firms operating within the formal financial sector—that is, firms that meet strict regulatory standards for safety and soundness of customers’ funds, and which receive all available governmental and quasi-governmental protections (such as deposit insurance). However, these services may be provided by the informal financial sector, whose members lack access to governmental protections and are less heavily regulated. Payments services are sometimes bundled with savings or income services, as when checking account costs are waived for customers who directly deposit their paychecks into those accounts. Payments services can be disproportionately costly for lower-income units, especially those without access to the formal financial sector.

Of special importance for our discussion are capital and credit processes provided for firm or household development. Development in this limited context means any set of actions undertaken with the aim of increasing net worth via the creation or expansion of assets. Asset-based development processes, in turn, can be either real-estate-based (commercial, residential, or mixed-use); or they can involve the creation and expansion of assets unrelated to real-estate development. Often the two types of development overlap. Dymski and Veitch (1996) have suggested that real-estate-based development poses special problems because of its spatial fixity; this feature implies that such assets are linked irretrievably to the idiosyncratic and unique history of its surrounding area—they are irreversible, and their value depends on spillover effects from the development (or lack thereof) of nearby assets. This implies project risk, in the case of real-estate development, and enterprise risk, in the case of businesses. These risks are amplified by problems of asymmetric information, rivalry, and coordination among lenders, builders, and occupants. Standardization of housing units and the spread of franchise retail outlets can be interpreted, in part, as efforts to reduce project and enterprise risk. Some risk is, however, irreducible; this risk must be borne by whomever puts up the money. Small business start-ups are especially risky, accounting for a large (even majority) share of job creation and job destruction.

Development—whether of enterprises, of real estate-based assets, or both—requires large outlays of resources: outlays for assets per se, and those required to use those assets productively. Individuals or firms undertaking development must either possess sufficient wealth to purchase and use assets autarchically, or they must have access to financing. Asset acquisition may be financed externally either by seeking partners in start-up financing (as with venture capital) or by refinancing assets (especially homes) whose net worth can be leveraged. Assets used in firm operations are sometimes supported by working-capital financing for payroll, materials, and other running expenses.

External financing generates additional risks beyond the risks already inherent in the project itself: especially default risk—the possibility that the borrower will not repay a loan in a timely fashion and interest-rate risk—the possibility that the rate embedded in a loan contract will be less than the cost of funds. These two financial risks must be apportioned between borrower and lender, in the context of both asymmetric information and Keynesian uncertainty about any project’s outcome. Financial institutions considering loans for asset acquisition or use do what they can to minimize their exposure to risks. They typically require collateral that can be seized or attached in the case of borrower non-performance; they charge fees up-front; they join together in lending consortia; and they investigate borrowers’ track records. For small-business loans, third-party guarantees are sometimes available—this is the case for many categories of Small Business Administration (SBA) loans. Once loans have been made, they are sometimes sold off to secondary markets.

Development finance in economic theory

Development finance has been under-analyzed in the economics literature, for several reasons. First, the development/finance literature is focused narrowly on macreconometric relationships (see Demirgüç-Kunt and Levine 2001). Second, the economics literature on banking and credit markets (see Freixas and Rochet 1997) suggests that banks routinely make loans in support of borrowers’ “investment projects.” This characterization is especially prevalent in the literature on development and credit (see Stiglitz and Uy 1996). This suggestion has been gladly accepted by innumerable economic theorists and policy-makers, as it justifies the application of many principal-agent models to banking ‘puzzles’. The problem is that this suggestion is wrong: banks almost never make loans to support any borrower’s acquisition of productive assets per se. Such financing must come from elsewhere; bank loans instead finance borrowers’ use of assets they already have. Because of this oversight, a key problem in development financing—the financing of asset acquisition—has received virtually no attention in the formal economics literature.

Third, economists have largely ignored the implications of the differences between the formal and informal financial sectors, in terms of both spatial coverage and behavioral practices. Many households and businesses in the U.S. obtain financial services from the informal financial sector: about 10–15 percent, according to the Federal Reserve’s Surveys of Consumer Finances (SCF), are “unbanked”; other experts have asserted that this percentage is substantially higher. Those using informal financial services are disproportionately lower-income; they tend to be concentrated spatially and to pay more for the financial services they receive. According to the 2001 SCF, approximately one-third of those in the lowest income quintile in the U.S. lack a bank account—that is, are completely outside the formal banking system.

Economists have tended to view informal financial markets as providing the same core financial services that the formal-sector does: see, for example, Caskey’s (1994) argument that pawnshops constitute the short-term loan market for the poor. While true as far as it goes, this characterization overlooks the fact that the terms and conditions for services provided through the informal sector are generally far worse than those provided via the formal sector. Indeed, lower-income and majority non-white neighborhoods often are sites of financial practices custom-made to make money in the context of asset-poor units with low and unstable income streams. High fees, interest rates, and penalties often lead to systematic asset decumulation, as the subprime crisis has dramatically shown.

This leads to perverse spatial financial dynamics in many lower-income areas, wherein day-to-day financial practices reduce asset balances, for economic units that systematically have low or even negative net worth. By contrast, in upper-income areas, financial dynamics often generate increasing asset balances, even for units with large debt loads. In general, spatially-segmented financial dynamics vastly complicate the problem of economic development in lower-income and majority non-white communities.

In sum, while development always requires financing, securing this financing can be especially problematic in lower-income and majority non-white communities for several reasons: first, all development projects are subject to problems of Keynesian uncertainty and asymmetric information, and these problems may loom larger in these communities than elsewhere; second, wealth levels in these communities are typically far lower than elsewhere; third, financial dynamics in these communities often encourage wealth decumulation. Over time, the spatially-fixed assets that do exist in areas with all three development problems often decline due to inadequate maintenance and to low levels of local asset turnover and investment. Of course, this underinvestment in physical and business assets is often paralleled by underinvestment in social assets.

The special challenges associated with renewing development in such areas, then, force some choices. One key choice is how to achieve development: to focus on transforming the assets and income flows of the businesses and residents already in a lower-income area, or on meeting the needs of this population on the assumption that its core income/wealth characteristics will not change over time? Another choice is whether to focus on building up the economic characteristics of a lower-income area for its current residents, or on transforming the economic characteristics of the area so that it becomes attractive to a new set of residents? That is, is one attempting to transform the situation of lower-income people, to open up an area, or both? This need not be an either/or choice; but it often is. Is the task to create dedicated assets for a community with fixed needs, locked into non-mainstream markets, or is it to facilitate the inclusion of marginal populations and areas in mainstream asset-building processes?

A short history of urban asset-building initiatives before 1990

Anti-poverty initiatives in the U.S. have sometimes emphasized asset development; but these programs have often had other goals, some conflicting with wealth creation. As Katz (1989) pointed out, welfare programs have often attempted to target the “deserving poor”—those categorically unable to work—while denying benefits to the “undeserving poor.” The latter category consists of those who have assets and/or who have the capacity to work, but who are not working. The pre-1996 federal AFDC (Aid to Families with Dependent Children) program, for example, prohibited benefits payments to families whose assets exceeded certain thresholds.

Federal anti-poverty policies have sometimes incorporated asset-building program elements. One early experiment in this direction was the 1966 CAAP (Community Action Against Poverty) Program, which provided community residents with the resources needed to organize for independent change and to apply for local public-good assets such as Head Start Programs and community-health centers. Employment development was a central concern of CAAP programs as well. This last concern led to the provision of funds for the establishment of community-development corporations (CDCs). CDCs were designed like miniature versions of the development banks that were then engaged in nation-building throughout Latin America: they encouraged business start-ups, provided technical assistance, and sometimes had access to concessionary finance for community-based businesses. In 1964, the Small Business Administration (created in 1953) launched the Equal Opportunity Loan Program (EOLP), which relaxed credit and collateral requirements for poverty-level applicants seeking financial backing for the creation of new small businesses. The EOLP targeted African Americans—the first time federal policy had done so.

Supplementing these efforts were a variety of housing programs, especially the §235/236 and §221(d)3 programs, which provided subsidies for low-income housing units, and public housing per se, which maintained and expanded the nation’s stock of publicly-owned housing (a program initially passed in 1949). Urban renewal was also funded federally. Facetiously dubbed “Negro removal” by community activists, this program provided a legal mechanism wherein governmentally-appointed redevelopment authorities obtained rights of eminent domain over areas that were declared “blighted.” Land classified in this way was typically bundled and then used to launch civic projects or new commercial construction projects. Similar in spirit to urban renewal were historical-preservation programs, many of which aimed at rescuing low-income areas of cities for upper-income residents by restoring the class character of residential areas that had been occupied by lower-income households. In sum, two distinctly different kinds of asset-building occurred on the urban space occupied by the poor: first, public and publicly-subsidied housing was put into place on urban space not considered desirable for other uses; second, redevelopment, historical-preservation, zoning, and other policies were used to free urban space occupied by lower-income and minority communities but desired by the politically connected and economically powerful.

The five-year Nixon Administration decisively reshaped urban anti-poverty policy. President Nixon issued Executive Order 11458 in 1969, establishing the Office of Minority Business Enterprise; in 1971, he signed Executive Order 11625 requiring all federal agencies to develop plans for Minority Business Enterprise contracting. Complementing these efforts was a 1969 amendment to the Small Business Administration act, establishing the SBA §8(a) program, which specified that a percentage of federal contracts would be awarded to minority businesses. In 1971, the §7(a) loan guarantee program, for loans to minority-owned businesses, was passed, as was the §504 Certified Development Company Loan Program, for the purchase of fixed assets by minority-owned businesses. This cluster of mandates and programs embodied the idea of “Black capitalism,” and implemented a new approach to asset-building in minority communities.⁴

Despite this affirmation of Black economic power, the Nixon Administration viewed urban anti-poverty programs as diffuse and uncoordinated, and too politically independent. In 1969, it replaced many individual program elements (including CAAP) with the Model Cities Program (MCP). MCP was originally designed as a program that would permit a number of well-funded “model cities” to demonstrate different approaches to poverty reduction and urban renewal. When implemented, the idea of a few cities winning out in a competition for the available funds—so that best practices could be generated and, later, imitated—was replaced by an allocation process that doled out available federal monies to all cities that met eligibility criteria and filed applications for the money.

The Nixon Administration was responding to local elected officials’ joint demands: first, for more resources to respond to social- and physical-infrastructure needs; second, for more control over the level and location of the federal government’s urban grant-in-aid programs. In terms of the second demand, the MCP was a half-way measure: it retained some elected community-based representatives on Model Cities area councils. The Nixon Administration then finished the job of streamlining federal grant-in-aid programs and centralizing local political controls over those programs. It obtained passage of several block-grant programs in 1974. One of these was the Community Development Block Grant program (CDBG), which pooled the various asset-building and housing-related programs aimed at reducing poverty and achieving redevelopment in stagnant urban areas. The Law Enforcement Assistance Act (LEAA), another block grant program, provided support for local police and law-enforcement expenditures; a third block grant, General Revenue Sharing (GRS), transferred discretionary funds to local political jurisdictions at all levels. Localities wanting CDBG or LEAA funds had to submit applications illustrating how particular federal goals were being satisfied (for example, regarding the percentage of CDBG funds being spent in low- and moderate-income census tracts). The application-grant aspect of the MCP was carried forward in the Urban Development Action Grant (UDAG) program—a special pool of funds for high-priority local development projects (with no poverty-reduction proviso).

Subsidized-housing policy was drastically reshaped. The alphabet soup of federal programs was replaced by the comprehensive §8 program. This act shifted the balance of federal housing expenditures from the supply to the demand side. Low-income households would not be assigned federally-financed or federally-subsidized units, but instead provided with vouchers that could used to seek out housing on the open market. This program has never worked out as planned: in most cities, waiting lists for §8 vouchers are scandalously long, and these vouchers are normally used in apartment complexes in low-income areas.

So as early as 1974, asset-building directed to low-income people, whether for economic development or housing, was no longer a component of anti-poverty programs. Instead, cities were given more tools to use federal funds for asset-building projects that might have, at best, indirect benefits for lower-income people (Philadelphia Plan jobs, for example). The Reagan Administration began a pattern of declining support for federal anti-poverty and urban policy that has continued until today: LEAA, UDAG, and GRS were eliminated; CDBG was provided an ever-declining portion of the federal budget. Faced with declining federal support for urban asset-building and poverty-reduction, states and localities began creating enterprise zones in the late 1970s. These zones offer tax incentives to firms that locate production sites and/or create jobs in designated areas.⁵

In sum, four principal asset-building strategies were used prior to 1990 to engender asset-based development in lower-income and minority areas in U.S. cities: (1) activities aimed at creating more real-estate development and at removing urban ‘blight’—that is, the expansion of space for mainstream real-estate-based asset-building; (2) efforts at encouraging community-based financial and business enterprises; (3) direct and indirect government efforts to provide housing for lower-income households in areas set aside as less desirable; (4) bank initiatives under the CRA to provide credit for asset-building in marginalized urban areas. Different groups and interests continue to pursue all four approaches in different ways until the present day.

The transformation of banking and financial practices

We now undertake a rapid sketch of urban financial structures and banking practices since the War on Poverty was initiated in 1964.⁶ Banking in the early 1960s largely retained the shape it had been given in the 1930s: banking markets were segmented by geography and product line; banking loan and deposit customers had virtually no alternatives to banks’ instruments; banks’ maximum demand and time deposit rate maxima were set under Regulation Q; and demand deposits were protected by deposit insurance. A robust home-mortgage market had arisen, with half of all mortgages made under the Federal Housing Administration (FHA) or Veterans Administration (VA) programs. Most mortgages originated by savings and loan associations or savings banks (and the remainder by commercial banks); and these lenders held them to maturity, up to 30 years after they were originated. While bank branches permeated white areas of cities, few or none operated in minority areas. Due to this financial exclusion, minorities either operated their own ethnic lending circles or opened their own ethnic banks. African American banks, plentiful before the financial crashes of the pre-War period, started to emerge again in African American urban population centers in the post-War period. The capacity of these banks was limited; consequently, minority areas had more than their share of pawnshops, check-cashing stores, and finance companies.

By the mid-1960s, banks were beginning to emerge from a long era of competitive lethargy. The impetus came from money-center banks, who were losing high-balance customers to broker-dealers at that time. The money-center banks created a number of liability-side innovations to retain these customers; so doing endogenized their liabilities and permitted the emergence of liability management (wherein banks identified asset growth targets and then funded them through purchases of funds in money markets). In the mid-1970s, banks of all sizes faced a further customer-loss problem: the newly-established money-market mutual funds created liquid short-term savings vehicles paying more than Regulation Q permitted. This time, banks had no effective response; instead, they lost funds via “disintermediation” throughout the 1970s. The bleeding of deposits from the banking system forced restrictions on bank loan growth. The banks’ more creditworthy loan customers turned increasingly to direct-credit sources (commercial paper and bonds).

One might suppose that, faced with declining loan demand, banks would reach out more heavily to potential loan customers in their market areas. But by the end of the 1960s, banks’ market areas were themselves in the midst of change. “White flight”—that is, the systematic relocation of white middle-class and working-class families from urban neighborhoods bordering minority areas, to primarily white suburban areas—was affecting virtually all U.S. cities. This destabilized established loan-market relationships for many banks. Many banks reacted badly—making loans to overseas borrowers or financing ‘block-busting’ behavior. Banks sometimes would neither nurture the potential loan demand of the new (heavily-minority) residents of the areas they served nor the demand of existing (heavily white) residents. These problems especially affected housing finance. For one thing, a large proportion of housing finance had been provided through the FHA program, which used explicitly racist criteria in loan approval; for another, even whites could not get home loans in their neighborhoods, as the unstable racial character of these areas suggested that pledged collateral might decline markedly in value.

These patterns of financial exclusion coincided with the emergence in many cities of independent, turf-based community organizations inspired by the Civil Rights movement and activated by the successes of the War on Poverty years. These organizations embraced the problem of access to banking credit; the term ‘redlining’ was coined to denote banks’ refusal to make loans—especially home loans—in inner-city areas. Redlined areas in this era contained both majority non-white and majority white neighborhoods. Consequently, multi-racial alliances for “reinvestment” by banks emerged.⁷ These alliances mobilized widespread political support, and succeeded in passing two crucial pieces of federal legislation: the Home Mortgage Disclosure Act of 1975 and the Community Reinvestment Act of 1977. These alliances also pressured the FHA to eliminate its racist loan-approval criteria.

At the end of the 1970s, the U.S. banking industry created in the cauldron of the Great Depression began to come apart. Vietnam War spending and two energy-price shocks in the 1970s, combined with the breakdown of the Bretton Woods system, led to unsustainable “stagflation” conditions. The Federal Reserve reacted aggressively, spiking interest rates to unprecedented nominal levels in 1979 and keeping them there through 1982. This led to two sharp recessions in 1980 and 1982 and a steep decline in global oil prices. Consequently, commercial banks in Texas and Oklahoma underwent a systemic collapse and interest-rate margins in the U.S. savings and loan industry turned upside down. Meanwhile, epochal bank and savings-and-loan reform legislation was passed by Congress in 1980, 1982, and 1989 (as well as in later years). And the Latin American loan markets to which most large banks had turned as of the late 1970s experienced a systemic crisis in 1982, triggered by Mexico’s debt default.

These huge dislocations led to a dramatic reshaping of the banking industry. The instrument of this transformation was a bank merger wave, which began in 1981 and continues to the present (Dymski 1999). The first phase coincided with distress mergers with failed or troubled banks and savings and loans in the 1982–88 period; these mergers were orchestrated or permitted by federal regulatory authorities. The Reagan Administration applied anti-trust laws selectively in the area of banking. The Federal Reserve, for its part, increasingly took the public stance that U.S. banking law was outmoded; it regarded the U.S. as overbanked, and regarded geographic and product-line restrictions on bank expansion to be outmoded. With money-center banks largely sidelined due to their large bad-debt burdens from the Latin debt crisis, super-regional banks such as North Carolina National Bank (later Nationsbank) and Bank America emerged as winners in contests for banks and savings and loans elsewhere in the US. By 1990, the five largest banks in Texas were headquartered in other states.

This expansion of superregional banks across state lines, at a time of technological change, permitted the emergence of new consumer banking strategies. The banks expanding into new state and regional markets pursued an “upscale retail banking” strategy. This strategy involves, first of all, identifying and winning the loyalty of a middle-to-upper-income customer base, the members of which are candidates for multiple financial products—consumer loans (for autos, higher education, housing rehabilitation), credit cards, insurance, retirement and college accounts, mutual funds, and so on. Second, this strategy involves brand standardization and aggressive marketing campaigns, and the heavy use of computer-based evaluation methods; emphasis shifts from the branch loan officer’s personal knowledge of clients to the bank’s back-office capacity to tailor products for a vast customer base. Mergers emerge naturally in this strategy: the bigger the customer base, the more revenue potential for any new product. The new frontier of consumer banking was constituted by asset management, mutual funds, and real-estate-based services.

This new strategy left gaps and holes in banking markets, which smaller banks could exploit. The huge reduction in the commercial-bank population through mergers has been partially offset by a burst of new bank charters. Smaller banks have expanded in areas in which megabanks have weak market presence: in particular, local business banks and minority-owned banks have moved into niches in urban areas.

Trends in Banking since 1990

While super-regional banks initially outmaneuvered money-center banks in competing for the markets opened up by deregulation and the merger wave, from the mid-1990s on, a small cluster of US megabanks—Chase, Bank America, Citibank, and Wells Fargo—gradually regained the upper hand, and now dominate all other banks in terms of both their market reach and their balance-sheet size.

These megabanks are somewhat different in character from the money-center banks that dominated U.S. banking in the 1960s. Bank America has had severe problems in trying to make the transition to being a major player on Wall Street (McGeehan and Rivas 2003). Even Citibank has continued to generate reliable profits only in consumer banking; broker/dealer activity, trading, and investment-banking have either created losses and generated disappointing earnings. Diversification itself proved no panacea for these contemporary megabanks: synergies among lending-based mass-market strategies, investment-banking and broker-dealer activities, asset management, and venture-capital strategies have proven less reliably profitable than was initially imagined. In particular, high-end financial services have continued to evolve rapidly, not stabilizing in a form that delivered reliable net cash-flows.

Opportunities for enhancing megabank cash-flows soon emerged anew, however, due to another aspect of the evolution of consumer banking. Upscale retail banking changed the terms on which lower-income households and small businesses could access the financial system. In the New Deal banking system, cross-subsidies implicitly existed between customers with high wealth balances and those with little wealth—and also between ‘blue chip’ businesses and ‘mom and pop’ businesses. In post-1980s banking, megabanks no longer systematically subsidized low-balance customers’ services with returns from high-balance customers; for the latter were the subject of intense competition by other banks and non-banks. So with some exceptions—some states’ requirement that seniors receive ‘lifeline’ bank accounts—lower-balance customers and small businesses were left to fend for themselves. They were increasingly charged high fees to maintain bank accounts. Consequently, there was an upward shift in ‘unbanked’ households and small businesses.

In earlier days, such individuals were served by check-cashers, pawnbrokers, and loan stores—all locally funded and operated. This changed for several reasons. First, increasing inequality in the distribution of income and wealth increased the revenues available from providing financial services to lower-income customers. Further, the growing numbers of immigrants, especially migrant workers, created new market opportunities for remittances and payment services.⁸ Megabanks began offering an increased number of options such as collateralized credit cards, debit-cards and ATMS, and money-cards. They also targeted migrant Latino and Asian customers with special ATM accounts and other inducements, directly or through subsidiaries.

Due to advances in securitization technology and in procedures for pricing and off-loading risk, increasing portions of informal finance were incorporated into Wall Street. The leading example is termed predatory lending: that is, housing-based loans targeted at borrowers with enhanced risk. Predatory lending initially encompassed second mortgages and refinancing loans for existing—largely minority—homeowners, home-purchase loans for new homeowners in minority areas, payday loans, among other loan types. The interest rates, fees, and penalties on these loans were considerably higher, and trigger clauses for non-payment considerably tighter, than for conventional loans. These loans grew explosively in minority areas, especially for African-American and Latino borrowers, from the early 1990s until the early 2000s. So a minority loan applicant who in previous years would be denied a mortgage loan was instead provided one—albeit a loan that entailed substantially higher carrying costs, more financial fragility, and more risk of loss. It hardly bears repeating that predatory—subprime—lending in the home-mortgage market was increasingly used in the period 2004–06 to support home purchases in the broader US housing market—in some cases due to a housing bubble (Florida, California, Nevada), in other cases due to the collapse of household income flows (Michigan). Given the focus of this paper on the pre-subprime boom/bust period (see footnote 3), the unfolding and collapse of subprime lending is not discussed further here.⁹

This discussion of US banking transformation has said little or nothing about lending to small and medium businesses. As noted, banks have not historically made loans to create small businesses or to permit them to start new lines of business. This did not change in the last 20 years, with the exception of some microfinance and revolving-loan funds in a number of US cities. While households have become a central focus of banks’ liability and asset strategies, small businesses have largely not. Lending to these businesses proved difficult to incorporate into the loan-selling technologies that U.S. banks were pioneering. For one thing, small-business loans are more heterogeneous than other forms of credit—they are used for idiosyncratic purposes, have non-uniform risks, and have mismatched maturation dates. For another, they are fewer in number and in dollar volume than are housing-based loans. Consequently, they were both difficult and costly to securitize. Only in the height of the securitization epidemic that accompanied the subprime boom, 2004–06, was a significant amount of small-business lending securitized. But this came to an abrupt end with the collapse of the asset-backed commercial paper market, and in any event securitization had little impact on the overall level of small-business lending.

Contemporary asset-building in urban areas: from CDFIs to new market tax credits

We return now to the weeks and months after the Los Angeles civil unrest of 1992, to ask what urban policy actions were taken in the following decade, and what the surrounding financial context was. Reporters investigating “causes of the Riots” quickly focused in on access to capital and credit, along with access to jobs, as key explanatory factors. They found widespread problems in all of these areas, especially for minority entrepreneurs.¹⁰ In late Spring, Presidential candidates Clinton and Dole walked ravaged Los Angeles streets and promised federal aid. Bankers and business leaders were loaded onto buses at the First AME Church and given tours of South Central LA. It seemed that other cities might burn, and that a rebirth of federal policy initiatives in economically-deprived inner-core communities might be at hand.

But a long, hot summer of 1992 did not ensue. And the federal policy response was tepid—a drop in the fiscal bucket, compared to the scale and ambition of the “War on Poverty.” The spotlight cast by Los Angeles on the economic challenges of minority and lower-income urban communities did yield two federal program responses once the Clinton administration took office: first, a federal “enterprise community” initiative based on the “enterprise zone” policies developed by numerous states and localities; second, equity support for “community-development financial institutions” that meet social-banking needs in underserved market areas.

Both policies were familiar to the President from his days as governor of Arkansas; Chicago’s SouthShore Bank had established the Good Hope microfinance fund in Arkansas during Clinton’s term there, and Clinton had met with Mohammed Yunus, the charismatic founder of the Grameen Bank. Eventually other policies were added to this roster: the ‘brownfields’ program for recycling old industrial sites in cities; the establishment of quantitative targets and tightened accountability for financial intermediaries subject to the Community Reinvestment Act; and in 2000 the New Markets Initiative.

What factors explain this choice of urban policies, and not others? One factor is the presence of a conservative Congress through most of the Clinton years; so Congressional approval of new initiatives required buy-in from Republicans. For example, the press release announcing the New Markets Tax Credit program characterized it as a partnership between President Clinton and Republican Speaker of the House Dennis Hastert.¹¹ Another factor is the Clinton Administration’s “third-way” approach to social policy. Clinton wanted to be free of old thinking, and to embrace the new. This approach was summarized by Clinton domestic-policy advisor Gene Sperling in a 1999 speech:

“When the President and the Vice President ran for office, there was an increasingly counter-productive tug-of-war between government and laissez-faire approaches to our Nation’s struggling urban and rural communities. The President and Vice President believe that there is a third way—an activist effort by government to bring private sector capital, free enterprise and entrepreneurial activity to our nation’s underserved areas.

“In order to accomplish this goal and breathe life into their third way vision, President Clinton and Vice President Gore set out in 1993 to put in place a comprehensive community empowerment agenda. Among other things, this agenda has included the creation of Empowerment Zones and Enterprise Communities, the establishment of the Community Development Financial Institution Fund, ..., reform of the Community Reinvestment Act regulations, a greater commitment to affordable housing, and the New Markets Initiative. ...

“This approach has proven successful.” (Sperling 1999)

Sperling’s speech goes on to document program success by mentioning the dollars spent and the number of programs authorized. As of 1999, some 24 urban and 8 rural “empowerment zones” had been designated, along with 115 “enterprise communities.” Another proof of success was these programs’ leverage—for every federal dollar spent, he argues, some $17 in non-federal investments were achieved via the CDFI program.¹²

The CDFI Program

Awards were made under the CDFI program for the first time in 1996. As of 2003, some eight rounds of funding were awarded under this program’s administration, which is housed within the Treasury Department. All entities seeking funds under the CDFI program must first meet the test for certification as CDFIs. This involves six requirements, per the CDFI website¹³: “(1) The organization individually must have a primary mission of promoting community development; (2) The organization must be a Financing Entity; (3) The organization must principally serve a Target Market; (4) The organization must provide Development Services in conjunction with its financing activities; (5) The organization must maintain Accountability to its defined Target Market; (6) The organization must be a Non-Governmental entity, and must not be controlled by one or more governmental entities.” The term “target market” here refers either to areas designated as EC/EZ’s or to areas that are considered considerably distressed in terms of income and/or housing conditions.¹⁴ As of November 1, 2004, 736 entities were officially certified as CDFIs.

The CDFI umbrella has encompassed several different programs: separate financial-assistance and technical-assistance components for CDFIs; a Bank Enterprise Assistance program for conventional banks providing special facilities for officially-designated CDFIs; and as of 2002, a program to encourage CDFI development on Native American reservations. Figure 1 depicts the funding levels authorized for these sub-components through 2003; overall, $708 million was authorized and $650 million awarded, to a total of 1582 recipients. The financial-assistance program for CDFIs accounted for $390 million, 60% of funds awarded, with 419 awardees (and 1294 applicants). The grants under the financial-assistance program were awarded for purposes such as providing support for low-income housing construction, funding micro-finance lending programs, assisting small businesses in Target Areas, and so on. The technical-assistance program accounted for just $45 million, with 394 awardees (112 of which were start-up CDFIs).

Fig. 1

Allocations under Community Development Financial Institution Program, fiscal years 1996–2003 ($ in million)

The Bank Enterprise program, in turn, awarded $187 million to some 784 awardees (of 1094 applicants) through 2004. In the BEA program’s first several years, megabanks—informally, bank holding companies whose asset size ranks them in the top 15 in the nation—took the lion’s share of the money. In 1996, megabanks received 86% of BEA grant money; in 1997, 59%; and in 1998, 57%. Megabanks’ share has declined steadily thereafter: from 39% in 1999, to 20% in 2000. Under the Bush Administration, megabanks’ share equalled 16% in the 2002 and 2003 awards rounds; in the 2004 awards round, it dropped to just 0.4%. Note that minority-owned banks have consistently been among the awardees under the BEA program. BEA monies are often received for banks that invest in low-income housing projects, that support small-scale businesses, and that provide grants to community-based groups.

Figure 2 shows that average awards were substantially larger in the financial-assistance category ($930,000) than in the remainder of the CDFI program. Average funds awarded declined steadily after 2000.

Fig. 2

Allocations under Community Development Financial Institution Program, fiscal years 1996–2003 ($ in million)

The New Markets Initiative

During the 1990s and 2000s, while the CDFI program was selectively awarding less than a billion dollars to selected borrowers, a broad trend toward greater inequality was at work. Wealth gaps widened, and asset poverty increased, especially for non-white households (Caner and Wolff 2004). As noted in “The transformation of banking and financial practices”, consumer-banking markets were being transformed during these years: low-income customers had to pay higher marginal prices for banking services, while being targeted by sophisticated new loan instruments backed by Wall Street. Indebtedness levels among low-income households rose at a tremendous pace. The inner-core areas of U.S. cities were widely recognized (Wilson 1987) as areas in severe need—cut off from jobs, decent educational opportunities, and health care. One key deficit was access to credit and capital by entrepreneurs and proto-entrepreneurs in lower-income and minority areas. The venture capital that was becoming such a force in industrial expansion elsewhere was sorely needed in inner-core areas, as were working-capital loans.

The CDFI program was a start, but its scale was miniscule compared to overall urban dollar flows. There were two triggers suggesting the need for another round of policy innovation. One impetus came from a network of activists and investors, loosely associated with the emerging alternative mutual and investment funds (featuring “green” investing, “anti-sweatshop” investing, and so on). Some members of this network were interested in building bridges between the worlds of venture capital and community-based investing. The Community Development Venture Capital Alliance, an association of community-based venture-capital and investment firms, was formed in 1994 (as of 1998 it had 45 members; by 2003 its membership was approximately 100, including 80 venture-capital funds).¹⁵ Given the high rate of failure in business and housing start-up activity among community-based firms, expertise-pooling, technical support, and additional resources were clearly needed to establish this venue on a more secure basis.

The second impetus for further asset-building activity came from a very different direction. Inner-core areas might have immense problems of access to capital, but nonetheless cash-flows therein were substantial; the problem was not so much cash-flows per se as insuring that cash-flows were wealth-enhancing and savings-generating. Ironically, the very size of inner-city cash flows attracted the interest of business strategist Michael Porter (1995). His 1995 article in the Harvard Business Review applied his ideas about competitive policy to the U.S. inner city. He made a number of arguments, several of interest here: first, inner-city consumer markets offer a very attractive target for large-scale retailers and service providers; second, land parcels were often too small in inner-city areas to permit efficient businesses to locate there; third, inner-city community groups must welcome business development, not be hostile to it; fourth, government should provide tax incentives facilitating the opening of inner-city markets to mainstream businesses. In this way, the inner city would prosper.

Porter started an organization, the Initiative for a Competitive Inner-City (www.icic.org), which quickly attracted business partners and loaned executives in many cities. There were many limitations in Porter’s conception of what was wrong with the inner-city economy; these were addressed in a special issue of the Review of Black Political Economy, later published as Boston and Ross (1997). ICIC did have positive impacts on many minority entrepreneurs operating in impacted inner-core markets. But its focus was primarily on removing government barriers to better market equilibria, not about the ways in which financial spatial segmentation affected the choices that different participants in different financial locations were free to make. As such, the ICIC program was linked to Porter’s call for government assistance in opening new markets for large firms’ investment.

This is the dual context in which the New Markets initiative was launched by the Clinton Administration. In a July 24, 1999 speech before a conference organized by the Congressional Black Caucus, Gary Gentsler, Treasury undersecretary for domestic finance, described the New Markets concept in terms that were completely consistent with Porter’s vision: “There is enormous economic potential in America's low- and moderate-income communities—potential that can benefit the residents of these communities and the businesses that successfully serve these markets. ... [Among the reasons for this are these:] In New Markets, retail spending power is significant, and in urban areas, customer demand is quite concentrated. ... Many of these markets are not as well served with retail outlets. ... The buying power of minorities is expanding. ... Finally, many urban New Markets have important location advantages. For example, they are close to downtown commercial districts for business service firms, are close to highways and airports for distribution businesses, or have real estate with the appropriate infrastructure for manufacturing businesses.” Gentsler went on to note that the CDFI program and the CRA played an important part in these markets. He then observed,

“But this isn't enough to fully meet the need. In early July, the President made an historic trip to highlight the considerable need and the market opportunities of low and moderate income communitites. His New Markets Initiative, a program of tax incentives and loan guarantees, will further the Administration goals of bringing more private sector capital and expertise to these communities.”

President Clinton’s New Markets initiative was passed into law on December 14, 2000, just before he left office. The official White House statement observed, “This initiative will help encourage private sector equity investment in underserved communities throughout the country to ensure that all Americans share in our nation’s economic prosperity.” There were two principal components of the New Markets initiative (again per the White House press release):

• “The New Markets Tax Credit. The credit will spur $15 billion in equity investment for business growth in low- and moderate-income rural and urban communities throughout the United States and Puerto Rico. The credit, worth over 30 percent of the amount invested (in present value terms), will be available to taxpayers who invest in a wide range of privately managed community development investment funds, such as community development banks and other CDFIs, venture funds, and new investment companies, that finance businesses in low- and moderate-income communities.

• New Markets Venture Capital (NMVC) Firms. NMVC firms will provide incentives to increase the availability of venture capital in low and moderate-income communities for small businesses. Expert guidance will also be made available to small business entrepreneurs in inner city and rural areas. Ten to twenty NMVC firms are planned. The agreement authorizes the SBA to guarantee up to $150 million in loans that will match $100 million in private equity for a total of $250 million and provides $30 million in technical assistance for small businesses.”¹⁶

This legislation, co-sponsored in the Senate by John Kerry, also encompassed a new round of EZ/EC funding, an expansion of the low-income housing tax credit, and some other items aimed at stimulating lower-income areas. So the New Markets initiatives incorporated both the Porter/ICIC vision of incentives for firms investing in lower-income areas, and also the CDVCA-documented need for venture capital inflows to areas without substantial wealth.

The New Markets Tax Credit

The NMTC Program received an authorization to allocate $15 billion in tax credits over its authorized lifespan. This amount was to be allocated as follows: fiscal year (FY) 2002, $2.5 billion; FY 2003, $3.5 billion; FY 2004, $2 billion; FY 2005 and FY 2006, $3.5 billion. Just as CDFIs are the entities officially authorized to receive grants under that program, the NMTC Program is open for taxpayers investing in designated “Community Development Entities” (CDEs). According to the program website (http://cdfifund.gov/programs/programs.asp?programID=5),

“Substantially all of the qualified equity investment must in turn be used by the CDE to provide investments in low-income communities. The credit provided to the investor totals 39% of the cost of the investment and is claimed over a seven-year credit allowance period. In each of the first three years, the investor receives a credit equal to five percent of the total amount paid for the stock or capital interest at the time of purchase. For the final four years, the value of the credit is six percent annually. Investors may not redeem their investments in CDEs prior to the conclusion of the seven-year period.”

So what is a CDE? The Treasury website specified (in late 2004) that a CDE must be a legally established entity, meeting two specific requirements: “(1) The organization must have a primary mission of serving, or providing investment capital for, Low-Income Communities (LICs) or Low-Income Persons; and (2) The organization must be accountable to residents of the LICs that it serves.”

The CDE criteria were (and remain) relatively permissive. LICs are defined as census tracts with more than 20% poverty populations or less than 80% of the median area income. As Swibel (2004) notes, this definition incorporates 39% of all census tracts in the U.S., containing 36% of the nation’s population. Even so, projects outside of LICs can be designated a “target area” if the applicant CDE can show that inadequate access to investment capital exists in that area using “some type of lending or other analysis”¹⁷ The accountability guideline is similarly loose. The CDE must have a representative on its board of directors who represents the interests of the residents of the affected LIC: either by virtue of being a resident, owning a business there, or representing a church or civic organization that operates there. Further, an entity designated as a CDE can be a member of a larger corporate unit: e.g., Bank of America—whose primary mission is not to provide investment capital for low-income communities–can create a CDE. CDEs, in turn, can be “parents” of designated “subsidiaries.” By December 2004, 1,778 CDEs were authorized; the entities included ranged from subsidiary units of megabanks and megacorps to local startups.

Which of the many entities authorized as CDEs to apply under NMTC received funding?¹⁸ The first round of funding under the NMTC program was allocated to 66 CDEs in March 2003. The scale of these credits, on average, dwarfed that of grants under the CDFI program, as Fig. 2 shows; and these credits’ average value increased between the first and second rounds, while those in the CDFI program were falling.

Table 1 shows the allocation of these funds by project type in 2002 and 2004.¹⁹ In 2002, half the tax credits were targeted for commercial real estate projects—in some cases with the aim of spurring business development. Business development itself received just under a quarter of the available allocation; the remainder was designated for mixed-use real estate credits and miscellaneous purposes, including financing CDE activities. Regarding NMTC allocations for the May 2004 round, as before, fully half the available credits were used to support real-estate developments increasing retail, office, and industrial space in target communities. Business development per se again received just over 25% of available credits; remaining funds were divided among other uses, with real estate development figuring prominently once again.

Table 1 Distribution of 2002 and 2004 new market tax credits by project type

Projects linked to both of the competing rationales for the New Market initiative received some funding: Porter’s emphasis on using incentives to bring market competition to lower-income areas is evident, as is some emphasis on business development. But here we must go deeper. The language of the NMTC act specifies ‘equity investment’. How does this translate into support for business development? Figures 3 and 4 analyze the type of investable funds that will be made available for business-development projects (this includes business development per se and also commercial real-estate/business development). Consider the 2002 allocation; Figure 3 differentiates credits, first of all, based on whether they are allocated to banks, corporations, and real-estate developers; to government and redevelopment agencies; or to other grantees. Then within each grantee type, Fig. 3 specifies whether loans alone will be provided, or whether equity or subordinated debt (or both) will be made available.²⁰ In Fig. 3, most of the business-development-related funds received by banks and corporations include equity of some kind (counting subordinated debt as equity). By contrast, the largest category of business-development credits received by other grantees takes the form of loans.

Fig. 3

Investment type for business-related New Markets projects, 2002, by category of grantee ($M)

Fig. 4

Investment type for business-related New Markets projects, 2004, by category of grantee ($M)

Figure 4 shows the same scenario for the larger 2004 allocation. Here again, the ‘other grantee’ category received mostly loans. ‘Banks, corporations, and real-estate developers’ again got a high proportion of equity. Compared with 2002, government/redevelopment grantees were more prominent; again, more of these grantees’ credits involved equity than loans. In sum, in the first two rounds of allocation, equity investments for business development were channeled primarily to established firms and to governments; grantees working directly with startup businesses receive a disproportionate amount of loans. A detailed grantee-by-grantee review of the 2002 and 2004 allocations reveals that a large number emphasize the creation of large national stores, franchises, and commercial space—conforming closely to Porter’s ideas about opening inner-city markets to “mainstream” businesses.

The New Market Venture Capital Program

The idea of the NMVC program is to increase the availability of venture capital and entrepreneurial expertise in underserved communities. Funds for this program were allocated in the 2001 budget. Round I of funding for the NMVC program provided $150 million in guarantees for unsecured corporate bonds (debentures) and $30 million in technical assistance grants. NMVC funds were matched dollar-for-dollar by private sector investments. Six companies were designated as recipients of these funds: two operating in Appalachia, one in New England, and three in the Mid-Atlantic area. Even while five of these first six Round-I grantees were awaiting approval, however, Congress cut the funding for the planned Round-II allocation in 2003.

Evaluating the post-1990 federal financing policies

The two primary urban programs passed into law after 1992—the CDFI program and the New Markets initiative—are paradigmatic “third way” programs: they have open structures allowing for maximum local input, they are cost-efficient, they force government officials to cooperate with private-sector players, and they are also selectively implemented. If we take 1960s’ urban programs as a frame of reference, however, questions quickly emerge about these programs’ effectiveness and focus. The question of effectiveness is especially pertinent for the Clinton EC/EZ programs. As noted above, there is virtually no empirical evidence that zones of this type stimulate employment in lower-income or minority areas. The question of focus is posed for the CDFI program. Clinton always referred to the Grameen Bank of Bangladesh, whose charismatic founder Mohammed Yunus he had met, as an example of how a few well-placed dollars of financing can bring hard-working poor people out of poverty. Most “assessments” of microfinance have in fact come from observers who are struck by the transformative power these programs have on their participants. Still, the path from inspiring examples to systematic evaluation remains to be established.²¹ And very few community development banks are exclusively microfinance funds.²²

But the problem was more extensive than a lack of clear evaluative criteria. More to the point, as Hyman Minsky pointed out, “neither the Administration nor Congress has really formulated the concept of community development banking” (1993: 33), despite the rush to implement the CDFI program. Minsky went on to argue that the concept of community-development banking could be effective in addressing the problems of low-income urban communities only if implemented on a substantial scale (see Figs. 1 and 2), if provided with a mechanism for restoring CDBs’ liquidity, and especially if the problem of start-up equity was addressed. Minsky argued that something like a modern-day Reconstruction Finance Corporation would be needed to provide the missing equity capital that proto-entrepreneurs and small/medium businesses in lower-income communities currently lacked.

For one thing, both programs operate on a competitive-grant basis, not an entitlement or comprehensive-funding basis.²³ In this respect, the continuity of Clinton’s “third way” with Nixon’s and Reagan’s urban approaches is clear: just as federal housing programs for the poor were transformed from a build-to-need basis (using tax-based incentives) to an allocate-spaces-as-available basis, both programs were available only to those municipalities and entities that submitted superior proposals. The local scope-of-action that figured so prominently in Nixon’s block-grant approach is present in these programs; but only a few get to play. So the fundamental premise of the War-on-Poverty programs and also of Nixon’s block-grant approach—that the problems being addressed are deep-seated and structural, reflect historical power and resource imbalances, and therefore must be addressed through the systematic application of federal resources—is overturned. Policy is not a matter of using redistributive and programmatic power in a concentrated structural assault; it is a matter of identifying latent superior ideas. It is as if the federal government is only funding ‘best-practice’ experiments, on the presumption that local governments and lenders will imitate the best ones.

On these terms, the CDFI program has been successful in several respects. First, it provided a programmatic vehicle for both microfinance advocates, for groups and professionals working with megabanks, for smaller community banks, and for the small number of community-development banks. The small amount of available funds (just $81 million per year, on average) was spread across a large number of participants. There was considerable competition for the more sizable financial-assistance awards; but in other portions of the program, most applicants were funded. Second, inasmuch as it was designed as a demonstration program, it could be evaluated not on the basis of whether the scale of lending in lower-income areas of large cities had significantly increased, along with wealth creation and savings levels, but rather on the basis that participating institutions were able to do more than they otherwise could have.²⁴ This modest claim was indisputable; and the modest amount of funds allocated to the program, sufficiently small to avoid the axe in a time of budgetary stringency (for programs such as this).

One helpful component of the new urban policy in the 1990s was the Clinton Administration’s clear commitment to strengthening enforcement and accountability under the Community Reinvestment Act. After 2000, the Republican-dominated Bush Administration worked to weaken the CRA. As this occurred, the list of bank financing activities considered ‘normal business practice’ in inner-core areas, versus those considered doable only given CDFI or NMTC funding, shifted toward the latter. This indirectly reduced the likelihood of any follow-up to the demonstration effects of new-age asset-building policies for low-income spaces. Of course, the explosive growth of predatory loans in these same spaces reduced this likelihood further.

The newer New Markets initiative programs were evaluated formally for the first time in a study released only in October 2008. Perhaps this was less necessary. As Swibel (2004) points out, tax credit programs are always popular—especially those with loose criteria, permitting real-estate developers (among others) to exploit the difference between lower recorded income levels (per the Census) and rising incomes in surrounding real-estate developments..

The New Urban Policy and the Urban Development Crisis

Our review of urban policy development in the 1970s showed that urban governments were given more control over local development projects at the same time that they were coming under pressure from emerging fiscal crises. Initially, federal policies provided for funds that could be used to undertake desired projects locally (notably UDAG and some portions of CDBG funding). However, as federal funds dried up, so did governments’ capacity to oversee development. To some extent, booming real-estate markets and innovative devices such as incremental tax-revenue bonds have made up the gap. From this perspective, the NMTC and, to a lesser extent, the CDFI program, offered new sources of funds for the kind of developments that City Halls tend to prefer—large projects with high sales-tax revenue potential. This may well explain the Round-I to Round-II shift toward the use of NMTCs to support government, real-estate developer, and bank investment efforts (see Fig. 5).

Fig. 5

Entities receiving New Market Tax Credits, 2002 and 2004 allocations (% of total for each year)

The new urban policy and evolving financial markets: some conclusions

This comparative discussion of “War on Poverty” and post-1990 policies to support the financing of community development in US cities is hardly an historical exercise. These programs, having withered in the later years of the Bush Administration, have been renewed with the Obama Administration. For example, in September 2008, CDFI director Donna Gambrell announced $20.1 million in awards for the CDFI program. Less than a year later, on June 29, 2009, Treasury Secretary Timothy Geithner announced $90 million in CDFI awards, along with $1.5 billion in New Markets Tax Credits. These actions were taken by funding made available under the Obama Administration’s American Recovery and Reinvestment Act (ARRA). According to the press release accompanying this action:

“The Recovery Act is playing a critical role in restoring economic growth and strengthening our nation's financial stability by developing and investing in local communities," said Secretary Geithner. "The Recovery Act awards announced today build on the Administration's efforts to get lenders lending again—these awards will help generate capital for small businesses, mortgage loans for homebuyers, and funding for affordable housing projects and other facilities in communities across the country.”

While shorn of “third age” rhetoric, these actions and words indicate the Obama Administration’s concurrence in this approach to urban community-development financing. Despite the large overall expenditures made under the ARRA, these CDFI and NMTC funding levels are modest. Indeed, the New Market commitment of $1.5 billion is less than the $2.65 billion of commitments in 2002 or the $3.5 billion in 2004 (Table 1).

So thus far, at least, the Obama Administration has carried forward the new urban policy with very few changes. Will it have an effect? It is useful to return to the challenge posed in “Conceptualizing banking/financial and asset-building practices: a framework”—that is, should asset-building development projects attempt to transform the people living in lower-income spaces, or should they attempt to transform the space itself, so that residents with higher incomes—and businesses with tonier clients—move in? The 1960s round of urban policy made both attempts at once; but eventually, political considerations and fiscal limitations squeezed out the first approach in favor of the second.

In response to urban social stress in the early 1990s, the new urban policy’s answer was, definitively, the second path. Again, some preliminary moves to follow the first path were erased after a short time. In 2003, Congress killed the one program explicitly aimed at providing venture capital for businesses and potential entrepreneurs in lower-income areas; and we have seen that loans and not equity, in the main, are provided for grantees in the NMTC program that focus on business development and are not controlled by governments, banks, or large businesses. The modestly-successful and very small CDFI program will not be able to affect the overall dynamics of impacted urban communities.

In the past decade, this weak federal response to urban stress has been partially offset by the growth of a new wave of service-sector and low-wage manufacturing jobs (after two decades of inner-core economic hollowing-out and deindustrialization). This new job market, low-wage and low-benefits as it is, has nonetheless supported large inflows of immigrants, documented and undocumented alike, into inner-core urban areas throughout the U.S. But these income flows have benefited pre-existing residents of inner-core urban areas only marginally. And they have been accompanied, as we have seen, by new financial practices. These practices have reinforced the spatial financial segregation of lower-income areas; indeed, the subprime crisis that has resulted from these practices is only worsening it.

In closing, we take one final look at Los Angeles. There, we find a growing Latino presence, notably in South Central, and a resurgence of low-wage manufacturing and service-sector activity. This population inflow has transformed the post-riot dynamics of Los Angeles yet again, even as many African Americans have been participating in return migrations from California to the American South and other places. Now there is an emerging competition for the market in remittances; indeed, the sheer weight of the LA Latino market is leading banks to rethink their urban strategies. The subprime crisis has wrenched Los Angeles, with negative wealth effects and displacements that will take years to sort out.

There is, in the meantime, a lot of money to be made in informal-market banking, as in upscale retail banking. The premise of these two faces of consumer banking—in effect, one for the rich and upper-income, the other for the poor and working class—was, from the beginning, that the banking market can be split into two or more relatively distinct market segments; and once split, the distribution of risks and costs can be fine-tuned based on the characteristics of each segment. The subprime crisis has not changed this premise. Subtle distinctions among the services offered by any of the megabanks—for example, collateralized debt cards for the poor, credit cards for the upper income—can maintain these distinctions. So can services offered separately, such as payday loans.

Banks and other financial firms are well aware of the character and implications of the spatial financial segregation in U.S. cities. The new urban policies have ignored this structural divide, and the way in which market forces react to and build on this divide until now; these policies instead view market forces simply as an opportunity set for selective social mobility.

A shift in perspective in this policy area is sorely needed. Making the people in inner-core urban areas the clients of policy initiative, rather than developers and megabanks, is critical, and will lead to bolder approaches that dare to transform the people and businesses already in lower-income and non-white areas. The fact that approaches of this type have been aborted twice, in short order, in the 1960s and 1990s rounds of federal urban policy does not mean they are too costly, or unworkable. Quite simply, small-scale programs can only very rarely overcome large-scale barriers and dynamics. If the politics of the possible are defined to permit only programs that are non-threatening and non-redistributional, then no urban transformation for the economically excluded can occur.

New ideas, as ever, have been percolating and evolving into yet new proposed policies. To cite two, Stegman et al. (2003) suggest coordinating the use of earned-income tax credits with affordable housing policies; and Williams (2004) argues that credit unions are effective vehicles for pooling savings and creating financial alternatives in lower-income communities. These ideas deserve consideration and governmental program support. There may or may not be another fire next time; but there certainly will be unrelieved asset-poor lives for millions of Americans in the nation’s cities, unless asset-building policies for lower-income households are again made a focal point for national action.