EVALUATING URBAN ENTERPRISE ZONES AS A POLICY FOR INNER-CITY MINORITY COMMUNITIES

Can enterprise zones have any long-term effects on the economic status of minority populations or on the concentration of poverty in urban neighborhoods? This question was tackled from two perspectives. First, do enterprise zone interventions make sense in the context of existing models of racial inequality and ghetto development. Second, do the existing literature and evaluation studies provide evidence that enterprise zones are benefiting the urban poor in general and poor minorities in particular? This study examines five selected official evaluations of state enterprise zone programs looking at how incentives are targeted toward specific populations and places and whether benefits to targeted communities were measured. The goals of state enterprise zone programs definitely include economic advancement for the disadvantaged and reducing poverty, but these goals are not directly reflected by the structure of enterprise zone programs or program evaluations. Existing measures of benefits to targeted populations or areas, generally yield negative results. The principal finding from this review of enterprise zone evaluations is the weak link between the structure of the programs, what they intend to accomplish, and what accomplishments evaluators measure. Existing evaluation research on enterprise zones does not provide evidence that existing enterprise zone programs are decreasing racial income disparity or improving the conditions of the urban poor. Enterprise zones are the wrong tool to fix the lack of economic opportunity in blighted urban neighborhoods because there is a poor fit between the structure of enterprise zones and the goal of providing opportunities for economic advancement to the urban poor.


INTRODUCTION
The phenomena of concentrated urban poverty in the U.S. has a complex relationship with race. Urban minority populations have suffered disproportionately from de-industrialization and economic instability (Wilson 1987). Yet, the most talked about urban economic development program designed to create opportunity in poverty areas does not explicitly addresses questions of race or even disadvantage Qones 1987, Harris 1992, Glover 1993. Many advocates of enterprise zones hope they will help alleviate racial economic disparity that leaves concentrations of poor minority populations in depressed urban areas. This research project looks at how urban enterprise zones address the issue of racial inequity through their goals, program structure, and measured outcomes.
Variations on the enterprise zone theme have been the centerpiece of the White House urban agenda under Presidents Reagan, Bush, and Clinton (Bendick and Rasmussen 1986, Boyle 1995, Fulton and Newman 1994, Mier and Pelzer 1982. Thirty-eight states have their own enterprise zone programs (Rubin and College 1994;Wong 1996). Yet despite this popularity, nobody has demonstrated the connection between enterprise zones and the problems of concentrated urban poverty and inequality that they seek to address. Poor inner-city neighborhoods have high concentrations of minorities, enterprise zones target these "depressed" sections of inner cities, but enterprise zone programs make few explicit claims about who receives the benefits. Given 1 the presence of pronounced racial inequality, the unanswered question is how enterprise zones address race. To be consistent with the goals of community and economic advancement, enterprise zones should provide tangible benefits for minority members of the community. Proponents of enterprise zones claim they will provide new economic opportunities in an expanding economic pie (Green 1990). But the rhetoric of zones is not simply about growth, it is targeted growth that will touch the most distressed communities. In contrast, most reports of the benefits of enterprise zones give figures of jobs created, dollars invested, and numbers of new business starts without reference to the distribution of those benefits. The simple fact that businesses and jobs locate within designated enterprise zones does not necessarily translate into increased opportunities for racial minority residents of the community. A full understanding of the benefits of urban enterprise zones must take a critical approach and ask who receives those benefits.

Purpose and Significance
This project looks at how state urban enterprise zone programs and evaluations studies address race. The U.S. Department of Housing and Urban Development (HUD) is about to commission an evaluation of the federal enterprise zone program, and many states have yet to complete program evaluations. This research could contribute to improving program evaluation designs and to a better understanding of the merits of enterprise zone programs. A better conceptual understanding of enterprise zones combined with more sensitive evaluation criteria will increase the likelihood that these programs contribute to increased economic opportunity for minority residents of inner cities and to urban revitalization in general. Most importantly, if there is no evidence to suggest that enterprise zones can contribute to economic opportunity in urban poverty areas, then it is time to reconsider the whole concept.

Organization of the Study
Chapters 2, 3, and 4 review relevant literature on enterprise zones, economic theories of poverty areas, and research on the benefits of enterprise zones for minority populations.I I outline several causes of racially segregated urban poverty areas and discuss the ways enterprise zones may address those underlying causes.
Chapter 5 involves close textual analysis of five state enterprise zone evaluation studies focusing on program goals, targeting mechanisms, and incentives. I selected Colorado, Florida, New Jersey, Texas, and Wisconsin for their geographic diversity, program variety, inclusion of urban enterprise zones, and availability of an official evaluation study report. While these states and reports are diverse, they may not be entirely representative of all zone programs and evaluations. This study identifies the extent to which benefits to minority populations are anticipated and measured through written program goals, incentives, and evaluations. I examine stated goals and program explanations for references to (1) race, (2) disadvantage (as an attribute of people), and (3) distress (as an attribute of a place). This study identifies programmatic links to the objective of increasing economic opportunity for inner-city minorities present in (1) zone selection criteria, (2) 1 The scant existing literature deals primarily with blacks to the exclusion of other minority groups. While the evidence discussed in this paper is primarily looking at black-white income disparity, many of the ideas apply to other racial/ ethnic groups that are affected by concentrated urban poverty and income inequality. I suggest that future evaluation research on enterprise zones be cast more broadly to look at the program's effects on inner-city minority residents including Hispanics, Asians, and Native Americans. 3 requirements to qualify for incentives, and (3) the types of incentives offered.
Finally, I examine the indicators used in program evaluation studies to determine if they reveal any information about program participation by minority members of the population or benefits to them. The concluding chapter sums up the relationships between the idea of revitalizing poverty areas and enterprise zone goals, programs, and achievements.

Government Policy and Targeted Redevelopment
"There has been a long-standing national interest in distributing the nation's economic growth in a manner that reduced the number of places and groups that are left outside the economic mainstream" (Hanson 1983: 61).
Policies promoting westward expansion, port development, grants to canal and railroad companies, regulation of interstate transportation rates, and the national highway system served to connect the national economy and disperse economic growth. The Full Employment Act, the Tennessee Valley Authority, the Area Redevelopment Act, the Appalachian Regional Commission, the antipoverty program, and the Model Cities program worked towards achieving inter-group and inter-regional equity through economic opportunity.
Enterprise zones are only the most recent example in a long history of targeted economic development programs. Federal programs focusing on minority opportunity were developed in response to the urban riots in the sixties. The Urban Employment Opportunities Development Act of 1967 was a federal program to subsidize the location of corporate branch plants in poor urban areas. Community development corporations were created by part of the Special Impact Program of the U.S. Office of Economic Opportunity (Harrison 1974). Until the late seventies, loans were the chief component of minority business assistance. During the 1980s targeted procurement programs became the backbone of minority business programs, with total dollar amounts increasing substantially.
Despite a notable federal commitment to lessening inequality, other federal programs including home finance, public housing, and highway construction, have contributed to the spatial concentration of poverty. Since World War II, policies favoring decentralized suburban growth assisted the exodus of upper and middle-income households from central cities, leaving many urban neighborhoods disproportionately poor and disproportionately minority (Florida and Feldman 1988). Despite the relatively open borders, segregation persists. As workers and consumers moved out of the city, so did many shops, services, and jobs. Many urban areas have experienced longterm decline or stagnation in economic activity. Ironically, civil rights legislation helped undermine black-owned businesses in black communities by opening up new markets and allowing residents to purchase goods and services outside the area Qones 1987). Increased residential choice for blacks led to selective out-migration which has left ghettos very poor albeit still disproportionately minority (Bates 1995).

The Origin and Adaptation of Enterprise Zones
The origin of enterprise zones is usually traced back to British Labour Party activist, Peter Hall, who proposed "freeports" in 1977 as a last-ditch effort to revitalize severely blighted areas. Hall hoped to allow private business to re-create the success of Hong Kong's economic transformation within the most severely blighted and abandoned areas of Liverpool and Glasgow (Hall 1982). These enterprise zones would have import duty and tax exemption, no barriers to international trade or investment, deregulation, and a drastic reduction of social services.
The Heritage Foundation brought the enterprise zone idea across the Atlantic in 1979 with a pamphlet calling for deregulation, free trade, and the elimination of the minimum wage in order to allow small start-up companies to thrive and create wealth (Green 1990). By lowering costs, enterprise zones would help the creation of new firms. Like Hall's "freeports," the Heritage Foundation invoked enterprise zones as an approach to encourage economic reuse of depressed urban areas by making private investment more profitable through tax and regulatory relief (Green 1990, Hall 1981, Hall 1982).
On U.S. soil, enterprise zones quickly evolved away from the international free trade zone concept toward a focus on fostering new entrepreneurial activity. As enterprise zones became more widely discussed in the U.S., the discourse began to include benefits for the urban poor who would participate in the economic revitalization. American adaptations of enterprise zones focus specifically on creating new private-sector jobs instead of focusing broadly on the economic prosperity of the firm. A goal of many enterprise zones is to increase economic opportunities for residents within the zone. Strong supporters of enterprise zones, like Representative Jack Kemp, predicted deregulation and tax breaks for distressed areas would be a way to create indigenous economic growth with minimal losses in public revenue. Kemp saw enterprise zones as "a way to uncork the entrepreneurial spirit that lies dormant in every downtrodden urban neighborhood" (Guskind 1990). Proponents claimed the zones would create new businesses 7 and new jobs without diverting people and economic activity from elsewhere (Green 1990). Support and funding for federal enterprise zones finally coalesced in the wake of the 1992 Los Angeles riots (Gunn 1993, Boyle 1995. Civil unrest created a strong national mandate to take action and became the touchstone in many urban policy debates. The connection drawn between urban riots in Los Angeles and Miami and enterprise zones reinforced the expectation that enterprise zones would ease the strain in poor, predominately black and Hispanic neighborhoods.

Components of Enterprise Zone Programs
It is difficult to make categorical statements about all enterprise zones since state, local, and federal programs are diverse and they have changed significantly in the last fifteen years. The original enterprise zone proposals in the U.S. were "federal, supply-side, anti-regulatory, conservative Republican" programs designed to "attract new, small businesses to the inner city," while by the time enterprise zones were implemented they were "state or local, private-public, reregulatory partnerships" with the additional goal of retaining existing businesses in urban, rural, and suburban areas (Wolf 1990).
The common elements that define enterprise zones are tax and regulatory changes targeted to an economically distressed geographical area. As legislatures wrestled with program details, enterprise zones became an collection of favorite pieces of past programs melded with supply-side incentives.
Enterprise zone eligibility criteria descend directly from Urban Development Action Grants (UDAG) program criteria (DED 1987, Green 8 1990. States adopted these criteria trying to second-guess federal officials even before a federal enterprise program materialized (Green 1990). State and federal criteria include measures of both physical blight and residents' economic well-being: high unemployment, persistent poverty, low household income, declining population, UDAG eligibility,s high proportion of population on public assistance, the minority composition of the population, housing vacancies, and physical deterioration (DED 1987, Glover 1993 (Goldsmith 1982).
At the other end of the spectrum, President Clinton's Enterprise Communities and Empowerment Zones Act is not structured as a simple supply-side incentive package; it is a more comprehensive reinvestment package targeted at designated depressed areas. It integrates the supply-side enterprise zone approach with a community planning and empowerment approach descended from Model Cities (Rubin and College 1994). This expanded scope draws on the lesson from early enterprise zones that tax su.s. Code 42-69-5318. UDAG eligibility and selection criteria include the existence of areas of concentrated poverty within the city in contiguous census tracts housing 10% of the population in which at least 70 percent of the residents have incomes below 80 percent of the median income of the city and at least 30 percent of residents have incomes below the national poverty level. Other criteria used to score applications include demonstrated results in providing low and moderate income housing, age of housing, extent of population lag, growth of per capita income, unemployment, and surplus labor. breaks and regulatory relief are weak policies when not combined with other programs. The act includes wage credits for hiring zone residents and direct benefits like day care, drug-use prevention, crime control, and job training (Szabo and DeMott 1993).
Enterprise zone programs at the state and local level have expanded from preferential tax treatment and more flexible application of regulations to encompass labor-related incentives, special government services, infrastructure investments, and preferential access to government grants (Gunn 1993, Rubin andCollege 1994). The most common labor-related incentive is tax credits for hiring disadvantaged persons. Twenty-nine out of thirty-seven states with enterprise zone legislation have tax incentives for selective hiring (Rubin and College 1994). Examples of special government services and infrastructure include crime control, public child care, job training, community development programs, low interest loans, zone marketing, supplementary social services, landscaping, and road improvements. Most state and local enterprise zone initiatives still include some combination of property tax abatements, tax credits for hiring new employees, sales tax reductions, and deductions for capital improvements (Guskind 1990 andJones 1987).

Summary
Like most economic development strategies, the early enterprise zone idea presumed a simple model that ignores many complexities of economic and social reality. Increasing investment and employment cannot solve every facet of the urban crisis (Munt 1991). An assortment of neo-liberal programs have been incorporated into enterprise zones to remedy community needs unaddressed by supply-side incentives. However, while enterprise zones are touted as an antidote to urban blight, the discourse about enterprise zones largely ignores factors like racial inequality, segregation, and discrimination.
Enterprise zones are not all derived from a single economic theory of the causes of urban distress and the diversity of programs reflects this lack of unified theoretical foundation. What enterprise zones have in common is the goal of increasing local economic prosperity and employment through geographically targeted incentives. A targeted economic development program like enterprise zones is consistent with long-held public policy goals of extending economic opportunity to all places and groups. The following chapters explore the question of whether enterprise zones further these goals in theory or in practice.

Enterprise Zones and Racial Inequality
To understand how enterprise zones work we need to ask what causes the concentration of poverty in ghetto areas and how enterprise zones interact with those forces. This chapter is divided into three sections. The first section highlights evidence that race is a significant variable in urban poverty both in terms of individual earnings and ghetto residence1. The second section outlines the major models of economic inequality and urban poverty areas. The final section identifies the broad categories of economic development strategies for combating inequality and urban poverty and draws the connections between those strategies and the models that underlie them. The close of the chapter explains how enterprise zones fit into the overall framework of models and strategies.
The theory of enterprise zones is only racially neutral on the surface because it does not address issues of racism, access, and participation.
Strategies that ignore issues of race are not truly race-neutral in practice because of preexisting inequalities and barriers to success (Goldsmith and Blakely 1992). Because of the inequality of economic fortunes associated with race we cannot assume that enterprise zones will benefit minorities equally with whites simply because urban enterprise zones target distressed 1"Ghetto" is a frequently used term without a fixed operational definition. Although race and poverty are correlated, ghettos are most frequently defined by the latter. Lynn (1990) and Jargowsky (1994) define ghetto as central city census tracts with overall poverty rates of 40 percent or more.
neighborhoods. Almost all enterprise zone programs ignore minority populations' special circumstances. Enterprise zone programs do not explicitly emphasize minority employment or ownership (Glover 1993, Jones 1987, Munt 1991. This leads to the questions of whether enterprise zones are well adapted to the needs of inner-city minority populations and whether enterprise zones can have any long-term effects on the socioeconomic status of minority populations.

Existing Racial Inequality and Concentrated Urban Poverty
Their are many dimensions to urban poverty, but from most angles race and ethnicity remain important characteristics. Race matters in unemployment rates, skill levels, occupation, wages, and geography. Poverty is becoming increasingly concentrated and racial inequity is growing.
Over the last several decades poverty has become increasingly concentrated in inner-city neighborhoods. As ghettos expand, the incidence of ghetto poverty continues to vary sharply by race. In 1980, 2 percent of all U.S. non-Hispanic white poor people, 21 percent of all U.S. black poor people, and 16 percent of all U.S. Hispanic poor people lived in ghettos. During the eighties "ghetto poverty among blacks increased both in terms of the number of blacks living in ghettos and as a percentage of the black population" Oargowsky 1994: 288). Almost two-thirds of the ghetto poor are black, and most of the rest are Hispanic (Lynn 1990). Poverty is increasing most rapidly among African-American and American Hispanics (Mueller 1990).
Only a small percentage of the poor remain poor for significant parts of their lives, but blacks are disproportionately represented among the persistently poor. Between 1974 and 1983, 5 percent of the population was 13 below the poverty line for at least 80% of that period. African Americans made up 66% of this group in long-term poverty (Mueller 1990).
High unemployment and low skill levels are defining characteristics of the ghetto (Bates 1995). Unemployment rates vary significantly by race both within and outside the ghetto, with black unemployment rates remaining roughly twice as high as those of the national economy. Unskilled inner-city workers are much less mobile occupationally and geographically than the population as a whole (Hall 1982).
Blacks are disproportionately represented among employees in declining industries (Wilson 1987). A national study of the effects of deindustrialization found that industries most affected by plant closings had relatively high levels of black workers. Industries that suffered most from foreign competition from 1964 to 1975 had an average minority work force of 11.5 percent compared to 7.4 percent in growth industries. Blacks have also been less well positioned to benefit from new employment in growth industries because the black population is disproportionately concentrated in slow-growth metropolitan areas (Bartik 1993).
Among employed workers, racial economic disparities are evident. Maume (1996) found that the wage gap between white and black workers has widened from $2.48 per hour in 1976 to $2.66 in 1985 in constant dollars.
After controlling for education, training, and experience Maume estimates that 26% of the wage gap in 1985 was racial discrimination, up from 16% in 1976. Harrison and Gorham (1992) reached similar conclusions about the deterioration of black earning power relative to whites at all levels of educational achievement.
Minorities are extremely underrepresented in managerial employment and entrepreneurship even within ghetto areas (Glover 1993, Goldsmith andBlakely 1992). The majority of ghetto firms are owned by white people living outside the ghetto (Bates 1993). Black-owned businesses employ fewer people and make smaller profits than white-owned firms within the ghetto.

Economic Theories of Poverty Areas
Poli ti cal and academic interest in ghetto economic development swelled in the wake of urban riots in the sixties. Policy specialists began to recognize the importance of economic opportunity within poor minority neighborhoods in addition to the more traditional goal of racial integration (Harrison 1974). Promoting economic opportunity became part of the discussion on promoting social stability (Hanson 1983).
The relationship between race and income has been carefully documented in inner-city neighborhoods. Researchers looked for explanations for the growing poverty of inner-city neighborhoods and for the persistent inequity along racial lines. Racial economic inequality is undeniably an enduring characteristic of the economic structure of the United States, although academicians and politicians disagree about the forces that perpetuate inequality. The following is a summary of the basic explanations for the economics of poor inner-city minority neighborhoods.

Neo-Classical Economic Theory
There are two conventional approaches to racial discrimination and economic disparities along racial lines: utility theory and human capital theory. The utility theory approach stems from work done in the 1950s by Gary S. Becker (1971). Those who discriminate find non-monetary satisfaction in not employing or working with minority workers.
"Individuals are assumed to act as if they have 'tastes for discrimination,' and these tastes are the most important immediate cause of actual discrimination," (Becker 1971: 122). Becker used census data from  and a straight forward neoclassical free market model to calculate the "discrimination coefficient'' which facilitated empirical comparisons of discrimination between places, occupations, and through time.
The human capital approach looks to education, skills, and health to explain productivity differences reflected in wage differences. Minorities are less productive and consequently lower paid because of differences in the education they have been able to attain. The "flawed character" model is a version of the human capital approach that focuses on behavioral inadequacies of the poor often associated with attitudes or culture. The culture of poverty and racial inferiority theories are the two most common examples of the flawed-character perspective (Schiller 1988). The human capital model denies the role of discrimination and instead attributes inequality to impersonal economic forces (Fusfeld and Bates 1984).
Several advocates of the human capital explanation argue that intergroup inequality in the post civil rights era does not result from discrimination. Shulman and Darity (1989) call this the declining discrimination hypothesis. Studies supporting the declining discrimination hypothesis compare employed blacks and whites of the same age, education, or other characteristics to show declining earnings differences. The theoretical arguments accompanying this sort of analysis assert that efficient labor markets work impartially with regard to race. The implication is that the persistence of inequality results from cultural and human capital deficiencies in the minority population. Shulman and Darity present convincing arguments that the declining discrimination hypothesis is inconsistent with economic data if labor force participation and underemployment are considered. Examination of labor market experiences during the eighties revealed increasing real wage differences between blacks and whites (Maume 1996).
Utility and human capital approaches explain the persistent poverty of predominately minority ghetto areas without reference to market failure or structural processes that create inequality. The market will work to increase racial equality because discrimination is inefficient. Both neoclassical approaches to racial differences in income suggest that the solution requires changing people, either through changing the attitudes of those who discriminate or through increasing the human capital of poor individuals as a way of increasing productivity. Spatial Mismatch Hypothesis Kain (1968) argued that a "spatial mismatch" between new suburban employment locations and exclusionary housing practices reduced job opportunities for blacks living in the inner city. Kain explained that employers located in the suburbs were reluctant to risk offending white residents by bringing black workers into the area, resulting in little demand for black workers. This resembles Becker's argument except the white suburban consumers, rather than the employers, discriminate. On the supply side, distance from suburban jobs meant blacks were less likely to hear about job opportunities and would have to spend more time and money to commute when they did find employment. New jobs may be so poorly paid that they do not justify relocating, or they may be in areas without affordable housing options. More recent supporters of the mismatch hypothesis include Leonard (1987), Wilson (1987), Kasarda (1989), and Holzer and Ihlanfeldt (1996).
Empirical support for the spatial mismatch hypothesis comes from studies on residential segregation and journeys to work. In a study of residential and employment patterns in Chicago and Los Angeles, Leonard (1987: 323) concludes that "residential segregation strongly influences black employment patterns and limits the efficacy of efforts to integrate the workplace." The magnitude of race-specific earnings differentials resulting from the spatial mismatch has grown over time (Holzer and Ihlanfeldt 1996). Holzer and Ihlanfeldt have found through their research on 3000 employers in four metropolitan areas that employers located closer to black residences and public transportation are more likely to hire black employees, and that black earnings increase with the distance of their employer from black population centers. Their research also validates the strong effect of the skill mismatch which suggests that human capital investment is a policy response of equal importance to residential desegregation and transportation. Bartik' s work (1993) on economic development and black economic success supports the spatial mismatch hypothesis with a new twist. Bartik found that black earnings are highly responsive to demand in the metropolitan labor market, but that blacks tend to live in slow growth metropolitan areas. Residential location of blacks is an important determinate of earnings both within and between metropolitan areas. The implication of much of the work relating to the spatial mismatch hypothesis is that local labor market conditions have strong effects on the earnings of black central city residents. Increases in the demand for labor have positive effects on black household earnings despite confounding factors emphasized by the human capital model. While Kain focused on employment discrimination as the predominate demand side factor preventing labor from finding employment, more recent advocates of the spatial mismatch hypothesis have shifted attention to job skill requirements as the predominate demand-side factor aggravating the effects of residential segregation. The spatial restructuring of metropolitan America has resulted in a skill mismatch as well as a spatial mismatch between residential patterns and employment (Galster and Keeney 1988). The numbers of blue-collar jobs in central cities have declined since the sixties, partly due to a shift of production away from the central city and partly due to increased efficiency and expanded foreign production. A significant number of jobs remain in central business districts, but many require specific sets of qualifications and skills which make them less accessible to poor inner city residents.
The research on the decline of entry-level jobs in the inner city provides more direct evidence that these demographic and employment trends have produced a serious mismatch between the skills of inner-city blacks and the opportunities available to them. Substantial job losses have occurred in the very industries in which urban minorities have the greatest access, and substantial employment gains have occurred in the highereducation-requisite industries that are beyond the reach of most minority workers. (Wilson 1987: 109) The interaction between the spatial mismatch and the growing skill mismatch is a bridge between the human capital and spatial models for uneven development. The recognition of this interrelationship is a prelude to the circular causation model discussed below.

Structural Explanations
In contrast to neoclassical theory in which the invisible hand of the market works to discourage racial discrimination and lessen inequalities, structural explanations assert that ghetto poverty is an artifact of the larger economic system. The ghetto is viewed as a separate economy peripheral to the economic mainstream. Researchers explain the creation and perpetuation of ghetto poverty through social and economic processes that create social and market dualism. This type of explanation looks at historical context and inter-group power relationships (Mueller 1990). Individual acts of discrimination may be harmful, but it is the economic system that ensures that large numbers of minorities will work for low wages and live in deteriorating areas. The most extreme voices for this school argue that the creation of the underclass and areas of physical blight are an inevitable outcome of capitalist economic development.
In the basic segmented labor market framework, workers find it socially, psychologically, and technically difficult to move from one sector of the economy to another. Race, sex, and class discrimination help maintain this dualism. Jobs in the primary labor market pay relatively good wages and employers value worker skill and longevity (Gordon 1972). Jobs in the secondary labor market are characterized as low wage, menial jobs, with poor conditions and no path for career advancement. Employers in the secondary labor market do not place a high value on longevity or skill, and they do not invest significantly in worker training. Ghetto residents are mostly confined to the secondary labor market and often move between low-wage work, informal economic activities, public assistance, and government job training programs (Harrison 1974). This framework is articulated through the internal colonialism model, the less-developed-nations model, and the circular causation model which differ from each other in details and in emphasis.

The Colony Analogy
The relationship between ghettos and the national economy bares some similarity to internal colonialism: the ghetto is described as a 'less developed country' with a severe 'balance of payments' deficit and with 'foreign' control of the most important local political and economic institutions" (Harrison 1974). The legal relationship of the ghetto to the larger economy does not resemble colonialism, but Blauner (1969) argues that the process of interaction is similar. The four common elements of the process are (1) involuntary entry, (2) policies imposing mainstream values at the expense of indigenous ways of life, (3) institutionalization of colonial rule through local government, (4) the oppressed group is "seen as inferior or different in terms of alleged biological characteristics" and "exploited, controlled, and oppressed socially and psychically by a superiorordinate group" (Blauner 1969quoted by Harrison 1974. Outsiders control many important ghetto institutions, like schools and police, contributing to the sense that the ghetto is an occupied territory.

The Less-Developed Nation Analogy
Concepts from the study of less-developed nations in the world economy have been applied to understanding American urban ghettos.
Structural dualism between the core and periphery can apply equally to both.
The ghetto serves as a vast reservoir of unemployed and under-employed labor that is constantly replenished by migration and population growth. The pool of unemployed workers keeps wage levels low which contributes to the 21 economy by keeping costs low. Economic growth in the advanced sector does not place strong upward pressure on wages in the periphery because the core has a tendency to use relatively capital-intensive technology in spite of the abundant supply of cheap labor. Actions in the public sector also contribute to economic dualism by failing to provide adequate transportation, education, and health facilities in the peripheral sector. The ghetto and mainstream markets are not integrated sufficiently to allocate investment by relative rates of return within and between the sectors (Harrison 1974).
The primary source of income in the ghetto is from low-wage employment. Unskilled labor is the community's major export. Consumer and capital goods are imported. Purchases are financed out of labor earnings and transfer payments. Most income is earned outside the ghetto (Harrison 1974). Money flows out of the ghetto go to absentee landlords and purchases of goods and services. Several studies document the "balance or trade" deficit in which minority ghetto neighborhoods pay substantially more in taxes than is returned in public expenditures in the area (Schaffer 1973, Harrison 1974. Outward flows of income, capital, and human resources to the rest of the economy serve to keep the ghetto in a permanently underdeveloped state and feed the economic interests outside the ghetto that have developed around those income flows, ... The entire economy outside of the ghetto benefits from the income, capital, and manpower resources that are drawn out, just as it benefits from a pool of low-wage labor that provides relatively low-cost services to those outside (Fusfeld and Bates 1984:145) The less developed nation analogy suggests that like less developed nations, ghetto economic development cannot be achieved through monetary transfers: the connections between the depressed economy and the mainstream economy need to be strengthened. The net outward drain of 22 resources must stop to allow sufficient investment in physical and human resources.

Circular Causation
The circular causation explanation is the most complex and complete attempt to understand the processes of ghetto formation. It combines human capital, spatial mismatch, discrimination, structural, and resource drain explanations. More significantly, circular causation recognizes the self-reenforcing nature of underinvestment, disadvantage, and underachievement, that complicates the path to advancement. Gunnar Myrdal (1964) first introduced the term "cumulative causation" in The American Dilemma to point out the negative feedback loops that make it impossible to disentangle cause from effect in understanding the poverty of the underclass. Fusfeld and Bates (1984) look at institutionalized economic processes to explain the perpetuation of ghettos. Businesses in inner-city minority communities lack of one or more of the essential elements of a healthy business: (1) talented entrepreneurs, (2) access to information, (3) access to financial capital, and (4) access to product markets. Neither black-owned businesses nor other businesses have flourished in inner-city minority communities. Successful ghetto firms tend to be smaller, use less labor and are more likely to be headed by drop-outs than by college educated entrepreneurs.
There is movement into and out of the ghetto. Fusfeld and Bates identify three barriers to upward mobility that restrain movement out of the ghetto: (1) race or ethnicity, (2) work and cultural patterns fostered in the ghetto that do not allow many people to fit easily into mainstream society, and (3) attitudes of people outside the ghetto which result in poor services in the ghetto and minimal acceptance of ghetto residents. These barriers are played out in housing and job discrimination.
The ghetto economy perpetuates its own poverty because low incomes mean a low standard of living in terms of food, housing, health, sanitation, and education. High levels of crime are both a result and a contributing cause to poor economic opportunities in poverty areas. These conditions reproduce low labor productivity and perpetuate low incomes. "The drain of resources out of urban poverty areas--manpower, capital, income--serves to reinforce the poverty" (Fusfeld and Bates 1984: 151). Cumulatively these resource flows create a self-sustaining system.
Economic growth is particularly difficult for the ghetto economy. Its weak infrastructure, lack of local initiative and entrepreneurship, and the shortage of capital make it difficult to generate a growth process. They create instead a self-generating poverty cycle. More important, the tendency for resources and income to drain out of the ghetto economy means that even if the forces of development were to appear, much of their strength would be dissipated before they had a significant impact on the ghetto itself. Any program or programs that seek to improve the economy of urban poverty areas must reverse the drain of skilled manpower, capital, and income if a cumulative process of growth is to be established. (Fusfeld and Bates 1984: 152) The process of circular causation with cumulative effects tends to preserve the ghetto in a relatively stable position from one generation to the next, even though population size and income levels may change. Programs that provide improved services to individuals, like education, training, housing, and health services, address the symptoms, not the causes of urban ghettos. If they are large enough and sustained long enough, such programs can make a contribution, but helping individuals will not increase prosperity in the ghetto as long as the social and economic processes that create the ghetto remain in place (Fusfeld and Bates 1984).
The fundamental processes that create and perpetuate the ghetto include a constant drain in human and capital resources combined with informal barriers that inhibit movement out of the ghetto for many people. Fusfeld and Bates (1984) state that the ghetto becomes home to people excluded from other social subsystems for a combination of the following five reasons: (1) race, (2) recent arrival, (3) cultural differences, (4) low earning power from lack of skills, poor health, etc., and (5) low income. The "underclass" can be distinguished from low income individuals in general by its isolation from mainstream values, behaviors, and the labor market (Ricketts 1992).
There are no physical barriers between the ghetto and the rest of society, and no formal methods by which individuals are 'committed' to life in the ghetto. The barriers are economic and social rather than physical, and the selection process is informal. (Fusfeld and Bates 1984: 149) Professional people who provide personal and business services leave, as do other talented people who can earn more money outside the ghetto.
The lack of political power in federal, state, and sometimes local public finance decisions lead to inadequate local control and public investment.
Capital resources leave through the allocation of public resources, the investment practices of banks, and through deferred maintenance on real property.
The circular causation explanation includes the notion of structural discrimination. In contrast to the conventional economic view that the market will weed out discrimination in the absence of what Becker (1971) 25 calls a "taste for discrimination," discrimination is part of the structure of the economy. By gaining convenience and low wages, employers may maximize profits while contributing to racial inequality. Prejudice and institutional employment practices work together to make race a barrier to economic advancement. Much discrimination in the market is unintentional and hidden beneath traditional modes of operation (Schiller 1988).
Discrimination and inequality in the educational system compounds discrimination in the labor market to the disadvantage of minority workers.
Some employers willfully exclude racial minorities, others "rely on recruitment procedures that have the same effect on minority racial groups" (Schiller 1988: 162). Relying heavily on existing employees for recruits works against would-be minority applicants who do not have a network of contacts already in better employment positions. Low-income minorities are seldom aware of unadvertised employment opportunities and are rarely brought to the attention of recruitment personnel (Schiller 1988: 154) Traditional recruitment agencies and advertising forums are more accessible to current members of the economic mainstream. Minority employees hesitate to apply for jobs where they risk embarrassment or harassment.
Breaking the cyclical processes of ghetto formation requires intervention in job market discrimination and reversing the drain of skilled personnel, capital, and income. Policy interventions to relocate residents or job opportunities to a closer geographic proximity would have minimal impact since they do not deal with the human factors contributing to pervasive poverty (Galster 1992). The ghetto experience becomes internalized in the values and behaviors of members of the underclass, complicating the transition to the mainstream economy.

Economic Development Strategies for Poverty Areas
The multidimensional problems of poverty areas call for a comprehensive strategy. Economic development programs to counter ghetto poverty and racial inequality typically involve one or more of the following.

Minority Entrepreneurial Strategy
The minority entrepreneurial model focuses on conventional capitalist institutions like private ownership and profit maximization. Some public sector programs focus on improving the profitability and viability of blackowned businesses. Such support is justified by past inequities. The Small Business Administration, (SBA) and Office of Economic Opportunity provided low-interest loans for small minority-owned businesses. Minority business set aside programs for public procurement and technical assistance also hold potential in business development (Harrison 1974).

Community Economic Development
Community economic development focuses on collective action, ownership of property by community residents, and institution building (Harrison 1974). Leaders urge residents to spend money within the community to stem the outward flow of wealth. Community development corporations work to improve the entire community by channeling capital, technical skills, and various kinds of assistance into target inner-city neighborhoods (Fusfeld and Bates 1984). These efforts focus on consciousness raising and reducing the outward flow of resources from the area.

Equal Opportunity
The equal opportunity approach focuses on breaking down the barriers that hold talented poor people from climbing the economic ladder.
Affirmative action, minority scholarships, anti-discrimination laws, and minority set-asides in government procurement are the most common examples. Encouraging and assisting individuals in their career growth sets an example that people can work their way out of the ghetto. Monitoring of discriminatory employment practices also keeps pressure on firms to provide equal opportunities. Multicultural education efforts attempt to change attitudes and behaviors that lie at the heart of intergroup alienation and discrimination.

Human Capital Investment
Investment in human capital seeks to improve the earning power of the poor through increased productivity. Individuals need education, health, and marketable skills to succeed in the economic mainstream (Hanson 1983).
Federal and state funding of human capital development in low revenue jurisdictions is necessary to provide quality services in poor areas.

Locational Intervention
In the spatial mismatch hypothesis the economic troubles of the ghetto are primarily spatial. Getting outside firms to locate in the ghetto bringing jobs with them or relocating families to affordable housing in the suburbs can counter the concentration of the urban jobless in poverty areas. Policies include enhanced public transit access to suburban work sites and inclusionary zoning to promote dispersed low income housing around the metropolitan area.

Job Creation
The government can work toward full employment by creating public service employment for the urban jobless or through macroeconomic policies promoting job growth (Wilson 1987). The key is increasing the availability of employment at living wages.

Connections Between Theory and Practice
The following chapter documents that state enterprise zones combine community economic development, human capital investment, locational intervention, and job creation economic development strategies (See Table 3  there is a weak connection between the policy and some element of the theory ++ the policy deals directly with at least one of the elements of the theory +++ the policy addresses the primary element of the theory *this policy was not discussed in this chapter, but is included in the table because of its primary importance in enterprise zone programs. If zones were effective at creating jobs within zones they would be combating the spatial mismatch by bringing jobs closer to poor populations. This review of enterprise zones and of theories of ghetto economic development suggest that a successful enterprise zone must (1) attenuate the drain of human and financial capital, (2) counter labor market discrimination, (3) provide adequate levels of public investment in human capital, and (4) facilitate the integration of ghetto and mainstream economies.
In addition, if small business owners are to benefit from the program as entrepreneurs, their needs for financial capital, access to product markets, and technical assistance must be overcome. Increased economic growth at the 30 state level may trickle down into areas of concentrated poverty, but without intervention, the forces of economic dualism will maintain pockets of concentrated poverty even amid prosperity.

Benefits of Enterprise Zones for Inner City Minorities
As a prelude to Chapter Five, which looks at official evaluation studies of state enterprise zone programs, this chapter reviews the existing literature on the benefits of enterprise zones for members of minority communities as workers and entrepreneurs. There is no substantial body of research on this topic, but a handful of writers have published their preliminary findings and raised some important questions, criticisms and hypotheses.
The most patent oversight in existing evaluations of enterprise zones is the general inattention to the demographic characteristics of the individuals and areas that benefit. All but one or two studies ignore the race issue entirely. For this reason the empirical evidence presented in this chapter is patchy and inconclusive. The second half of this chapter outlines a number of criticisms of enterprise zones relating to their potential to benefit minorities: failure to effectively target needy minority populations, lack of support for small start-ups with limited capital and expertise, failure to assure adequate levels of investment in infrastructure and human capital development, and inattention to diverse needs. There is growing evidence that tax incentives by themselves are not the right tool to reverse the cycle of decay in distressed urban areas or to combat the racial segregation and inequality that characterize such neighborhoods. However, even enterprise zone critics maintain optimism that more comprehensive enterprise zone programs may be able to show tangible results.

Minority participation in zone employment
By 1993, thirty-seven states had enterprise zone legislation and 25 reported job creation resulting from their programs (Rubin and College 1994). Wilder and Rubin (1996)  Beyond the question of who benefits from zones, there is doubt about their true job creation impact. Some enterprise zones have not created job 33 growth at all. Most studies measure jobs created or retained on an absolute basis, instead of measuring net job growth. Of the studies that also measured job loss, many have found it to be significant (Wilder and Rubin 1996). In fact some experienced net job losses. Both the success stories and the low performers need to be interpreted with caution because they are not controlled to separate job change attributable to the zone from the underlying economic trends. Dowall, Beyler, and Wong (1994) were forced to conclude that zone initiatives in California "were not conclusively associated with any of the estimated positive outcomes (Wilder and Rubin 1996). The U.S.
General Accounting Office (1989) study of Maryland zones found no clear causal link between job growth and zone designation or incentives. On a more positive side, Rubin and Wilder (1989) used shift-share analysis and found positive job growth attributable to the Evansville Zone in Indiana.
Many evaluations use methods that ignore the issue entirely.
Despite HUD' s consistently positive evaluations of enterprise zones, the Congressional Black Caucus disputes the zones' effectiveness (Harris 1992). The Board of Economists assembled by Black Enterprise magazine for a symposium on enterprise zones gave them a less than favorable review. The Board concluded that enterprise zones' ten-year history demonstrates an emphasis on traditional economic development and public relations without any serious effort to target minority workers. No comprehensive research shows enterprise zones to be ineffective at improving economic conditions for minorities, yet evidence from case studies illustrates some of the ways programs can fail. Irons (1994) reports that enterprise zones frequently fail to serve the targeted community. A 175 acre enterprise zone was created contiguous to a low-income housing project with predominantly black residents. The zone's stated objective was to provide opportunities for the housing project's residents. Irons laments: Our objective was to get the businesses in the enterprise zone to hire the people in the housing project. But there were all kinds of excuses: 'We couldn't train them. We couldn't find them. We couldn't keep them, just one excuse after another. It was hard to refute some of this ... but we felt we were being circumvented. (cited in McCoy 1993) The enterprise zone administrators did not have tools to identify and avoid racially discriminatory hiring practices by zone firms or to train targeted residents to meet employer needs.  (Levitan and Miller 1992). Firms cited the need for better trained workers as a key reason they did not hire more zone residents. These negative examples suggest some of the barriers that may undermine the potential benefits of enterprise zones for minority communities.
Enterprise zones do not have magical qualities that can overcome all physical, social, and economic barriers to revitalization. In this respect, zone critics are correct in arguing that the myriad social and physical problems plaguing many urban neighborhoods (e.g. decaying infrastructure, high crime rates, inadequate school systems) are not responsive to targeted development incentives. (Wilder and Rubin 1996) The one unexpected finding about enterprise zones and minority communities comes from regression analysis performed by Erickson, Friedman, and McCluskey (1989) in a cross sectional-study aimed at understanding the conditions associated with successful enterprise zones (Wilder and Rubin 1996). The percentage of minorities in the zone population was one of only a few variables positively correlated with investment and employment growth. There is no obvious causal relation, but at the very least this indicates that many of the more successful zones are located in areas with higher minority population.

Type of enterprise zone employment for minorities
Few studies of enterprise zones have documented the skill levels, wages, or duration of the jobs created (Wilder and Rubin 1996). One valid concern is that a policy that provides minimum wage jobs to minority workers will not significantly improve the long-term opportunities and economic prosperity of minority families. Any meaningful job program for disadvantaged minority populations must have a long-term goal of helping minorities into career-path jobs that pay more than minimum wage.
Within enterprise zones, minority-owned firms and minority employees tend to be most highly concentrated in low-wage services (Glover 1993). Twelve percent of new jobs in the 101 enterprise zones surveyed by Glover were minimum wage jobs, compared to 7.9 percent of jobs in the national economy. Seventy-eight percent of zone administrators reported that at least some of the new jobs in their zones paid minimum wage. These figures are unsurprising since enterprise zones are targeted toward firms that are particularly sensitive to factor costs and because much of the available labor in enterprise zones is low-skilled.

Minority entrepreneurs participation in zone programs
Only two researchers have looked at participation in zone programs by minority entrepreneurs. In Glover's survey only 5.3% of firms within zones were minority-owned, making minority firm ownership levels within enterprise zones similar to comparable economic areas without enterprise zone designation. Many minority owners in the survey did not participate in the enterprise zone program (Glover 1993). While the sample of minority firms is quite small, it appears that the pattern of minority firm participation in the zones differs from non-minority participation. Jones (1987) also found low levels of participation by minority-owned businesses in his case study. Of the sixty black-owned businesses in Decatur, Illinois, ten are in the zone and only one participates in any way in the zone program. While the numbers are too small to draw firm conclusions, these studies raise strong doubts about whether enterprise zones assist minority-owned firms. Levels of participation of minority-owned firms should be included in future evaluation studies to provide more useful data.
Glover's survey provides one clue to understanding low minority owner participation. Glover (1993) found that minority-owned firms in enterprise zones are concentrated in personal and business services. Overall, personal and business services tend to be underrepresented in the enterprise zones Glover surveyed. Older zones, with older firms have larger manufacturing firms and a more diversified mix of smaller firms supporting greater numbers of small service firms, like food services, that serve the employee and business markets. These service firms are "likely to be relatively small, have a higher probability of minority ownership, and a higher percentage of minority employment" (Glover 1993:87). Because minorities own more service firms and start-up barriers are lower, encouraging economic diversification within the zones may lead to higher minority business ownership and employment.

The Participation Gap: Barriers and Flaws
Targeting disadvantaged and minority populations Although job creation and economic recovery are explicit goals of enterprise zone programs, zones do not explicitly address the underlying problem of low rates of minority business ownership and employment (Glover 1993). For the most part, enterprise zones fail to acknowledge the specificity of race; they marginalize minority concerns instead of strategically incorporating them into program goals. Jones (1987) surveyed local black business people as part of his case study to learn about some of the reasons for non-participation. Local black business people gave three reasons for low program participation Oones 1987: 4): 1. Unfamiliarity with the program operation and types of incentives offered.
2. The program's complexity, possibly requiring the advice of tax consultants.
3. The program seemed to benefit large firms, not small blackowned firms that needing access to venture capital and technical assistance in starting small businesses.
Black entrepreneurs have historically been excluded from economic development initiatives, and in the absences of efforts to specifically reach out to them and to help them overcome barriers to participation, minority business people may exclude themselves through non-participation. The failure to assess the needs of minority entrepreneurs and to market the program to minority entrepreneurs are two of the program's shortcomings.

Start-up barriers
Enterprise zones are supposedly the key to generating new small businesses in the inner-city. Peter Hall argued that the abundance of regulations surrounding opening and running a business inhibits inner-city residents from exercising their entrepreneurial spirit (Hall 1982). The pure supply-side enterprise zone concept does not address the barriers inner-city entrepreneurs face in securing start-up capital. Without capital, minorityowned firms will not be able to take advantage of the other incentives offered in enterprise zones.
Some enterprise zone incentives benefit certain types of firms. Tax relief provides minimal incentive for start-up businesses since such businesses rarely have much income to pay taxes on in the first years, but tax relief can be a monumental benefit to larger firms with professional accountants (Hall 1982, Harrison 1982. Enterprise zones could even hurt minority enterprises by bringing in outside firms that compete with existing firms for the local market. Thus the type of incentives included in prototypical enterprise zones are prejudicial in practice against minority entrepreneurs and start-up firms in general.

Public goods and services
Many critics of the enterprise zone approach claim that supply-side incentives cannot jump start depressed geographic areas without direct public investment in social and physical infrastructure (Gunn 1993, Levitan andMiller 1992). Successful enterprise zone programs therefore must include direct public expenditures to attract and complement private investment.
However, in some places the withdrawal of public investment is so severe that it outweighs incentives provided by enterprise zone programs. A 1988 study of the Anacostia Community enterprise zone in southwestern Washington indicated that problems of high crime, high insurance, and banks' reluctance to lend inhibited any new businesses from locating in the zone communities (Glover 1993). As Harrison (1982) notes: The rapid deterioration of these public goods in the older cities -and their unfinished status in the new cities of the sunbelt --has become a serious bottleneck to further private investment in these areas. In short the fiscal crisis of the state directly contradicts the developmental objectives of the whole enterprise zone concept. (Harrison 1982: 425).
Large-scale changes are necessary to overcome the barriers in severely depressed urban areas. Enterprise zones do not change the fiscal inequalities of the current tax system that virtually ensure that some areas will be unable to provide adequate services and infrastructure. Some enterprise zones give preference to zones for state and federal grant money to provide a temporary infusion of outside money. As long as important public services and infrastructure are funded from local taxes, there is little hope to overcome the geographic inequalities of income and quality of life.
Lack of jobs is only one factor in the complex cycle of poverty. Investment in affordable housing, education, job training, and crime prevention benefit low-income minority populations (Glover 1993). Even relatively unskilled jobs require some basic skills and basic stability. To address these deficits, some state zone programs combine more comprehensive social programs with other incentives. Investment in both people and places are important foundations for economic growth. Elling and Sheldon (1991) found that state economic development policies were a key variable in zone success. They found the most successful zones complemented tax incentives with other kinds of support (Wilder and Rubin 1996).
Community involvement Gunn (1993) concluded that state enterprise zone programs, like the ones in New Jersey and Indiana, that emphasize neighborhood or community involvement in solving long-range problems are more successful than those, like Connecticut's, that only emphasize tax incentives.
"The economic, material, social, and political linkages that make a small area into a neighborhood or community are not addressed by enterprise zones.
Yet these linkages are the are critical determinants of the vitality of the community investment process" (Clarke 1982). Jones discovered some deficiencies in enterprise zones that limit the effectiveness of the program for the black community: For black communities in areas such as Decatur, enterprise zone programs may have only a minimal effect on neighborhood revitalization and business growth and development. However, with provisions for public participation of all sectors of the community in program design, implementation and evaluation; management assistance for potential entrepreneurs, marketing the program to black businesses and residents, and investment pools for development financing, enterprise zones may become more product than promise. Gones 1987: 6)

41
In sum, the incomplete information about the benefits of enterprise zones for minority populations tells a mixed story suggesting substantial variation between zones. There is little evidence to show that "pure" enterprise zones that rely on tax and regulatory breaks are effective, but researchers like Jones and Gunn are still optimistic about diversified programs that involve community members in planning for improved services, job training, economic development, and education (Jones 1987, Gunn 1993. The next step is to design enterprise zone evaluations with these questions in mind so that more firm conclusions can be drawn.

ISSUES OF RACE AND MINORITY EQUITY IN STATE SPONSORED ENTERPRISE ZONE EVALUATIONS
This chapter examines how existing evaluations of enterprise zones address the topic of minority welfare. As explained in Chapter 1, the evaluations reviewed here come from Colorado (1995), Florida (1993), New Jersey (Rubin and Armstrong 1989), Texas (Alwin 1994), andWisconsin (1993). Each one is an official study conducted to inform state decisionmakers about the efficacy of their enterprise zone program. The chapter is divided into three sections focusing on the program goals articulated in the reports, evidence of targeting benefits toward people or places, and finally a look at the conclusions evaluators drew about program success.
The goals, targeting mechanisms, and evaluation criteria of enterprise zone programs help explain why so few claims have been published about enterprise zone benefits for urban minorities. There is general inattention to racial inequity or its particular relationship to areas of concentrated urban poverty. Zone programs target poor persons by various means based on criteria such as residence in an area of concentrated poverty and individual economic disadvantage, but goals or evaluation criteria do not generally emphasize the same characteristics as the zone incentives. The agenda is divided between assisting disadvantaged people and improving the zone as a place. Zone evaluators do not fully appreciate the distinction between targeting benefits toward a specific population of people vs. a geographic area.
Programs can be successful at one without succeeding at the other (e.g., gentrification). The principal finding from this review of enterprise zone evaluations is the weak link between the structure of the programs, what they intend to accomplish, and what accomplishments they measure. The lack of intellectual clarity identified in Chapter 4, between the structure of zone programs and our understanding of concentrated poverty areas, is reflected in the mismatch observed in the state evaluation reports.
Two operational definitions need to be clarified before jumping into the discussion in this chapter: benefiting or targeting people vs. places. I make the distinction based on the unit of analysis implied by the wording in the evaluation report. When a goal or incentive uses individuals as the unit of analysis then I identify it as targeting people, whereas if the geographic area of the zone was the unit of analysis, it targets a place. The distinction might not always be intentional on the part of the writer. For example, employing the unemployed is a goal targeting people. Decreasing the unemployment rate in the zone is a goal targeting a characteristic of a place. Creating jobs for zone residents targets people. Creating jobs in the zone targets a place.
As an illustration of the distinction, creating jobs or reducing unemployment in a zone could be done without any positive economic impact on the zone residents: if new jobs go to outsiders, people commute into the zone from outside without employing more zone residents. The zone unemployment rate could decline as the result of increased employment of residents or because unemployed residents moved out or stopped actively seeking work. The differences can be important.

The Relevance of Racial Inequity and Urban Poverty by State
Racial inequity and urban poverty are not of equal importance in every state in terms of the number of persons below the poverty line or the 44 proportion concentrated in the inner city, but they are relevant to discussions on poverty in all five of the study states (Table 5-1). In all cases, race and Hispanic origin and location of residence are important predictors of poverty.
All five states have large differences between the poverty rates for white, non Hispanics compared to the rest of the population (Figures 5-1 and 5-2). Also unsurprisingly, poverty rates in metropolitan areas are higher within the    The goal of the American strain of urban enterprise zones is to bring about a change in the conditions of our devastated inner cities--to tum around their pattern of decay. To achieve such a reversal, the areas selected as enterprise zones would, ideally, be endowed with an image of economic buoyancy which, in turn, would help the zone attract business, employ the unemployed, and improve the lives of zone residents and the areas in which they live and work.

47
Targeting on Three Levels Enterprise zones target groups of people and geographic areas in three ways: zone selection criteria, qualifications for receiving incentives, and the nature of the incentives themselves. The first two are obvious targeting mechanisms, but the nature of the incentives influences who benefits in discernible ways. Among the five states, the most popular way to target people was to tie benefits like corporate income tax credits to hiring zone residents or previous recipients of public assistance. Incentives targeted toward places seek to create jobs, induce businesses to fix up property, and increase business investment within the zones by tying tax breaks to certain activities that take place within the zones. Only the New Jersey report (Rubin and Armstrong 1989) mentions minority status at all. The following state-bystate analysis gives more detail about how each program is targeted on each of the three levels.
Taken together, the targeting mechanisms demonstrate an intent to target economically disadvantaged people and places. The dual agenda of targeting people and targeting places, makes it hard to determine what outcome is desired and how it should be measured. Evaluators had difficulty measuring success with either approach because of the dearth of appropriate data and the use of inadequate evaluation methods. There is conflict in zone selection criteria between giving zone designation to the most distressed areas and giving the incentives to less poor areas that show greater development potential. Criteria in all five states seek to leverage tax breaks and investment in zones though a competitive zone selection process. The zone selection criteria and observations about their application in each state tell a very interesting story about how variable zone selection is between states.
Colorado's selection criteria do not ensure that the areas selected will be extreme poverty areas. Unemployment at least 125% the state average, slow population growth, and per capita income less than 75% the state average are Texas' criteria focus on identifying contiguous areas that are both poor and deteriorating. Areas must be at least a square mile in area with over 150% of the state unemployment rate sustained over a 12 month period or have a 9% population loss over 6 years. If that test is met, areas must meet one of several distress scenarios: poverty, UDAG eligibility, chronic abandonment, tax arrearages, substantial loss of business or be declared part of a state or federal disaster area. City or county government may designate up to three zones as long as they meet minimum standards. The real competition in Texas is for Enterprise Project designation. Similar to Florida, Texas divides the scoring process for Enterprise Projects between distress criteria and local effort using a 60-40 formula.
New Jersey's selection criteria focus on potential for success and local commitment, with relatively weak criteria dealing with economic distress.
There is a long list of preference policies for zone selection, including "the degree of commitment by public and private entities to utilize minority contractors and assure equal opportunities for employment in construction or reconstruction in the area" (Rubin and Armstrong 1989: 16) This is the only reference to benefiting minorities in any of the selection criteria.
Wisconsin selects zones based on the quality of the local development plan and the zone meeting two out of six economic distress criteria. These criteria include high unemployment, AFDC dependence, decline in assessed property values, low income, UDAG eligibility, and at least 5% of the work force in the jurisdiction being permanently laid-off in the preceding 18 months. These criteria would allow long-term poverty areas to be designated as zones but also would allow working class areas that have experienced recent economic downturns to qualify.

Requirements to Receive Incentives
Unsurprisingly, many surveys indicate businesses prefer zone incentives with the fewest strings attached. Many states have strict requirements for businesses certification to receive tax credits in conjunction with the program. Administrators have found a trade-off between administrative simplicity and sending the desired economic message to firms through specific requirements. Where requirements are weak, like giving job credits for all new hires in the zone, documentation of who receives jobs, employment duration, and wages is generally lacking.
Conditions imposed to qualify for incentives in different states vary widely, falling evenly between incentives tied to benefiting people and those targeted toward places. The following state by state overview focuses only on conditions imposed to qualify for specific benefits, not on the benefits themselves. Table 5-3 summarizes the requirements tied to incentives. The states intend these conditions to target benefits and motivate specific actions.   While race has no bearing on qualifying for any of the incentives, the definitions of "targeted populations" used in Florida, New Jersey, Texas, and Wisconsin can do exactly that in a non-discriminatory manner. The only way to judge whether these definitions target minorities would be to document the percentage of minorities that fall within the targeted groups and the percentage of minorities receiving jobs tied to zone incentives.
The requirements in Colorado are the least likely to provide benefits to minorities. The tax credits for new hires make no restrictions on whom the 54 firm hires. Colorado qualifications place a high emphasis on fixing up physical property, but even those improvements may not benefit poor neighborhoods if they are not located within areas of concentrated poverty.
Alternatively, physical rehabilitation without programs to benefit residents can lead to gentrification and displacement. By going beyond zone boundaries, the Florida and New Jersey incentives are the most focused on employment and income growth for low income individuals.

Nature of Incentives
Each of the five evaluation reports listed available incentives. These lists tend to emphasize tax credits and may not be a comprehensive list of aid available from the local zone administration agency (fable 5-4). Job training and financial programs that are generally available to businesses in the state would not be listed as incentives. Of the five states, only the Texas evaluation mentions business assistance and financial programs as integral parts of the overall zone program.  Small firms find the least benefit in enterprise zone incentives. Rubin and Armstrong (1989) found a strong negative correlation between firm size and the likelihood that the survey respondent reported the enterprise zone tax benefits had no impact on "their decisions to locate in, or expand operations in, the enterprise zone." Firms moving into the zone from out of state are more likely to report that zone benefits have an impact on their location decision. The small, marginal firms that enterprise zones were originally designed to assist, make up the majority of businesses within the zones, but program incentives apparently provide more important benefits to larger firms.  Rubin and Armstrong (1989): 26 In the survey firms were asked to rate the importance of zone benefits from 1, for the most important, to 6. Firm in all size classes rated exemption from sales tax on personal property and on materials as the most important zone benefits. The popularity of sales taxes stems not from their monetary value, but from their universal availability. Sales tax benefits are available to all kinds of firms within the zone and are not tied to hiring requirements.
Retail establishments, which account for 25% of all zone businesses in New Jersey, are not eligible for corporate tax credits even if they hire targeted workers.  Rubin and Armstrong 1989: 29 Corporate Business Tax and Unemployment Insurance tax benefits require firms to meet specific hiring criteria and I or require extensive record keeping. Many firms noted in their responses that they cannot hire the type of employees required for Urban Enterprise Zone tax credit eligibility and still conduct a profitable business. In fact, many respondent firms suggested that a better trained labor force would improve business conditions in the UEZs. Other firms have stated that the paperwork associated with the benefits is too time consuming relative to the value of the benefit offered. When asked to specify why they do not use the UEZ benefits, the . . primary reasons given were: -Employees not qualified 41 % -Unaware of benefits 13% -Bureaucratic requirements 8% -Not a corporation 6% (Rubin and Armstrong 1989: 29-30) This evidence from New Jersey's experience leads to a few general observations. Ironically while, New Jersey got the most marks for targeting incentives through requirements in Table 5-2, the evaluation suggests that on net, the incentives offered are not targeted in a way to maximize opportunities for minorities or to maximize job growth. The connection between zone incentives and hiring "disadvantaged" zone workers is tenuous. Existing urban education and job training is inadequate to bridge the gap between employer expectations and the apparent skills of disadvantaged applicants. From a minority equity standpoint, the most important incentives are those that help minority-owned businesses or create better jobs for minority workers. Minority workers may benefit from enterprise zones indirectly through the improved vitality of zone businesses, but many firms do not hire "qualified" zone workers for new jobs. The kind of enterprise zone incentives most typically offered do not assist start-up firms as much as larger establishments. Incentives like sales tax relief are geographically targeted through zone selection but are otherwise not strategically aimed at aiding specific workers or firms with high potential for creating jobs.

Summary of Conclusions Made by Evaluators
Two evaluations, Colorado, and Florida found a mismatch between goals and programs as implemented. Colorado identified problems with selection of territory for inclusion in zones. Florida found the mismatch between incentives offered and the needs of zone firms. The lack of critical data was a significant barrier to drawing conclusions about zone success. The conclusions evaluators were able to make with available data limited the discussion to things that were easiest to measure. Only New Jersey surveyed firms to collect data beyond administratively collected data, government statistics, and observations by zone administrators.
The five studies did draw some conclusions with relevance to helping the poor. Colorado looked at evidence for job creation, information about whether jobs were filled by zone residents or resulted in an increase in per capita income, and changes in the statutory distress indicators: per capita income, unemployment, population change. Colorado could find no evidence of progress in these areas. The major contribution of the 59 evaluation, in this regard, was suggestions of how to improve the program so impacts can be assessed.

Implementation Process
Program Impact The manner in which the program has been implemented is inconsistent with statutory goals. The huge extent of current zones, the amendment of boundaries to include new areas that do not meet selection criteria, and the disproportionate use of investment tax credits are the primary evidence for this conclusion.
Low participation by businesses reveals a poor fit between goals and some of the incentives. The program does not address some of the more critical needs of small businesses.
Business participation is less than expected. Factors contributing to the low-level of business participation include "limited business awareness of the program, limited availability of suitable development sites, and the reluctance of some private corporations disclose business plans as part of the certification process" (Wisconsin 1993: 5).
Without a verification procedure for claimed tax credits there is no way to assure that businesses complied with their initial plans.

60
Existing reporting mechanisms are inadequate to assess program impact. The irregular and discontinuous boundaries of the zones make it impossible to use census data to evaluate change in socio-economic indicators.
No definitive judgment of program success. Areas continue to be econorrucally distressed. Some zones have shown limited improvement, and most have not deteriorated substantially.
Program had negligible impact on the hiring of AFDC recipients or JTP A participants. The program may have had an rm pact on the hiring of zone residents, but credits were claimed for less than 2% of the eligible work force.
The program's overalJ impact on unemployment and other socioeconomic conditions have been minimal. The Annual Report overstates the benefits of the program. Companies locating in enterprise zones are not necessarily newly created or relocating from outside Texas. The program is not having a large impact on the most distressed areas m the state.
The program's impact on the tax base of zones has l:leen minimal.
The enterprise zone program had substantial positive impact on New jersey's economy. The benefits outweigh the costs.
No impact. It is unlikely that the program is meeting its statutory gOals at the existing low levels of participation.
It is too early to assess the long-term effectiveness of improving the economic health of the zones. The Texas State Auditor (Alwin 1994) looked to see if the zones were targeted toward the poorest areas and found they were not. The report concluded that the Texas program did not target areas with the highest unemployment as evidenced by the fact that jobs were certified in only 2 of the 10 counties with the highest levels of unemployment. Certified actual job creation by designated projects occurred in 21 counties. ln 7 of those counties, the number of jobs created amounts to less than 0.1 percent of the labor force.  of clarity about the goals of the program. There is a real mix between an emphasis toward benefiting people and improving places. This study focuses on a very narrow goal of enterprise zones, which has not been the primary concern of other evaluators and authors. But, even when we go beyond the goal of addressing minority inequity, existing evaluations come up short in demonstrating benefits on a variety of levels. In the end, we have weak and non-comprehensive evidence that any of the goals have been achieved. New Jersey is the only one of the five studies in where evaluators felt satisfied that they had found a positive impact of the program, and that was the only study in which evaluators did not even raise questions about how well benefits were targeted or who benefited. They even assumed values for key indicators, like wages, that they should have measured. The most common conclusion by evaluators is that they had insufficient evidence to draw sound conclusions. The disappointing evidence of program benefits does not prove that zones are an entirely ineffective urban policy tool, but it does highlights some of the weaknesses in program data collection and evaluation design. A good evaluation design must spell out the expected outcomes of the program and assemble evidence to determine whether or not the program resulted in the desired outcome.
The data problems faced by state program evaluators highlight the importance of designing programs to allow evaluation. Documenting use of zone incentives and associated investments needs to be part of program administration. In addition, zones should be contiguous and correspond to census geography enough that available distress statistics will allow evaluators to measure change.
In the case of enterprise zones, we are trying to evaluate change in the zone over time as a result of the program. Time series studies need measurements at multiple points in time to show the overall trend before and after the program began. With time series studies there is a very real possibility that something other than the program may be causing observed changes. A convincing evaluation design would measure change over time that is different from change in comparable areas.
A quasi-experimental design, must substitute for a controlled experiment because zones are not selected randomly. We can select comparison areas that share many characteristics with the zone to provide supporting evidence as to whether the trends observed in the zone are a result of the zone program or external circumstances. The multiple time series design combining time series measurements and comparison groups is ideal for enterprise zones because it provides some measure of control for likely economic trends in absence of the program.
Overall, the plethora of targeting mechanisms demonstrates an intent to target zones to economically disadvantaged populations and locations.
When state evaluators asked whether the program produced positive economic benefits and for whom, they generally were unable to produce positive evidence. Despite the rhetoric about enterprise zones, there is a lack of thorough evaluation research and no compelling theoretical argument to suggest that enterprise zones will accomplish equity goals. The divided focus between targeting places and targeting people, makes less clear what outcome is desired and how it should be measured. There is a poor fit between enterprise zone programs and the goal of improving the economic status of disadvantaged minorities.

64
CHAPTER6 CONCLUSION Enterprise zones, in general, are the wrong tool to fight poverty in blighted urban neighborhoods. There is a poor fit between the structure of enterprise zones and the goal of providing opportunities for economic advancement to the urban poor. Nevertheless, no well-designed evaluation research has been conducted to allow conclusions of much certainty on either side of the issue. We are left with doubt about whether enterprise zones improve the economic conditions in zones, little reason to believe they help put the urban poor on the track of economic advancement, and no reason to believe that zones are a step toward reducing racial inequality.

Recommendations for Future Evaluations
The evidence from evaluation studies reviewed in this report is patchy and plagued with methodological problems. From the standpoint of this study, existing evaluation research has not addressed all the essential questions to understand who benefits from zones, but more glaringly, most evaluations do not measure program impact. Enterprise zones do not, by their nature, defy evaluation. Based on the strengths and oversights of earlier studies I recommend future evaluations use a multiple time series comparison evaluation design.
A multiple time series design measuring changes in particular indicators over time in the zones and comparison areas is the most promising quasi-experimental design to distinguish between program impacts and coincidental changes. When practical, measurement should be made at multiple points in time to distinguish real trends from random or cyclical fluctuation. Comparison areas are useful in program design to distinguish the regional economic conditions facing the zone from the impact of the zone program.
While no study will be able to encompass all important questions, examination of minority participation in zone programs as new employees or as entrepreneurs would add greatly to the understanding of the impact of enterprise zones. We need more information about new jobs, including wages, duration of employees in jobs, and whether the job offers opportunities for advancement. Studies should also measure whether job training is adequate to allow firms to hire disadvantaged workers. Collecting data on these indicators would require use of administratively collected data, surveys, and key informant interviews. The evaluation of comprehensive economic development programs will always be problematic, but these suggestions will help provide a clearer picture of program achievement, particularly with regard to equity objectives.

Concluding Discussion
The policy debates on enterprise zones tend to underestimate and obscure the costs of overcoming economic disadvantage. There are a lot of political questions to be asked about enterprise zones, like who should benefit from enterprise zones to justify the unequal tax treatment, and how aggressively should benefits be targeted toward meeting the goal of minority equity. This paper is premised on the claim that inequity along lines of race or minority status is detrimental to the well-being and stability of the nation and that as a primary component of contemporary urban policy, enterprise zones should seek to address existing disparities.
There is little reason to believe that enterprise zones programs that simply emphasize investment and job creation will have any significant impact on the concentration of poverty in central cities or on the long-term economic status of minority populations. This study has found a lack of attention to the question of who benefits. Many authors have posed this question in a variety of ways, but the existing evaluation research holds only generally negative, albeit inconclusive, evidence that enterprise zones benefit urban minorities or the urban poor. This picture pieced together from diverse sources is one of doubt about the potential effectiveness of state enterprise zone programs. As tax incentive programs, enterprise zones are not designed to combat the many facets of urban poverty including inequality, segregation, and discrimination. Enterprise zones are a tool for increasing local economic prosperity in a targeted area without a clear connection to any economic model of urban poverty areas. The existence of profound inequality speaks to the fact that even if enterprise zones were successful at increasing investment and jobs in targeted areas, they would not necessarily benefit people of all backgrounds or races equally. This could manifest itself as the difference between improving the long-term earning potential of disadvantaged workers and gentrifying a particular neighborhood.
According to the economic literature, ghetto development is a multifaceted artifact of the economy. The literature reviewed in Chapters Three and Four suggests that reversing economic decline in areas of concentrated poverty will require intervention on many fronts including investment in public services, education, and job networks.
In conclusion, enterprise zones are not the correct tool to work toward a solution to America's urban crisis. They have equivocal effects on the poor and disadvantaged. They focus narrowly on jobs when geographical location of jobs is not the only issue. The lack of conclusive evidence about benefits in and of itself is not a compelling reason to abandon the urban enterprise zone approach. But when we put all the fragments together, the story that emerges is that enterprise zones, in their common form, are the wrong tool for fighting concentrated urban poverty. The type of interventions incorporated in state enterprise zone programs do not effectively intervene in the processes of ghetto formation as we currently understand them. Targeting a small area for revitalization is a valid way to maximize visible impact, but we need a more comprehensive set of tools to deal with the larger issues of concentrated urban poverty.