Do State-Designated Enterprise Zones Work?

Frank D. Beck
Illinois State University
Department of Sociology-Anthropology

Introduction

        Enterprise zones are a response to the depressed economic conditions present in many urban and rural areas. More than 36 states initiated enterprise zone programs to deal with this issue and associated problems and hopefully gain access to some other monies should a federal zone program begin. While many researchers have analyzed zone effectiveness on a case by case basis, with mixed results (see Wilder and Rubin 1993), a thorough study of zones across the country had only been attempted once before (Erickson and Friedman 1991). This project, therefore, builds on this past work by answering some very important questions: 1) Have enterprise zones, from across the country, experienced growth?; 2) What incentives are responsible for this growth?; and 3) Have the socioeconomic conditions of local enterprise zone residents improved?

        To answer the first two questions, a survey of local zone coordinators from across the country was used to compare zones to each other. Local enterprise zone coordinators were asked about the growth of investment, firms, and jobs within the zone as well as what incentives were used to grow the economy. Fifty-one zones responded in enough detail to be included in this portion of the analysis. The third question regarding the effectiveness of enterprise zones at ameliorating depressive social conditions is addressed with an analysis of Census data for 421 zones. With these data two comparisons are made between zones and non-zone areas to see whether enterprise zones have experienced greater improvement in their socioeconomic conditions than places most like them. One such comparison is made between zones and their pure matches (areas most like the zone that are within the same state in a county that does not have a zone). The other comparison is between the zone and an adjacent match (areas proximate to the zone in question but wholly outside the zone boundaries). For specific information on the survey, data management, or data analysis methodologies employed for this project, please contact the author (information provided below). Interested parties could also acquire a copy of my dissertation cited in the references (Beck 1998).

For easy access to each of the tables, click here.  Links to the tables and summaries of the results are provided in the text below.  Some of the tables may be wider than your screen.  To see the tables in their entirety you can adjust the size settings of your display (this is within the Control Panel for Windows based operating systems) or print the page in "landscape" mode.

Enterprise Zones Have Experienced Growth

        Shown in Table 1 are some descriptive statistics on the number of firms and jobs present in enterprise zones when they were designated and when local coordinators responded to the survey in late 1996. Note that these data are for 51 zones--only those for which the zone coordinator returned a survey and provided data on firm and job growth.

        On average, there were about 88 companies and 1,573 jobs in these enterprise zones when they were designated. In 1996 there were 140 firms and 2,722 jobs. On average, enterprise zones experienced growth. They had 52 more firms and 1,149 more jobs than when they started. Some enterprise zones experienced massive growth. One zone grew by 1,150 firms and another increased its employment base by nearly 15,000 workers. Another place lost 215 companies and 4,537 jobs. The reader should note that the distributions of all of these variables are skewed to the right. For this reason, median values are also presented.

        Table 2 shows descriptive information for the number of key incentives used by enterprise zones to attract investment. Tax incentives are the most popular. On average, zones use over 9 of the 21 tax incentives asked about (contact the author for a copy of the survey itself). The most common tax incentives employed were property tax reduction for firms locating or expanding in the zone as well as tax exemptions for materials, equipment, and machinery. The least common form of tax incentives included general tax holidays and gross receipt exemptions. Zones use over 5 of the 16 different kinds of financial incentives I asked about. The most common form of these included business loans other than venture capital; the least common were preferences in procurement of contracts. On average, 5 of 12 infrastructural incentives were employed and those that did offer such incentives were far more inclined to focus on physical infrastructure than the upgrading of police and fire services or to accelerate zoning changes (the least common infrastructural incentive). Lastly, only 3 of the possible 11 service/quality of life incentives are used by enterprise zones. Most of the effort here was on job training for employees in the zone; less effort was put into day care for families of workers in the zone or regulatory relief for firms.

        Note that all 51 enterprise zones employed some tax incentives. One zone made all 16 of the tax incentives available to interested companies. Further, at another point on the survey, I asked what incentive coordinators thought most useful in the promotion of growth in their zone. Resoundingly, they responded that property tax abatements produced the most investment activity. As this project progressed I also had the opportunity to talk one-on-one with some coordinators. By and large they agreed—it is the tax incentives offered by enterprise zone that are most responsible for whatever growth enterprise zones produce. The reader should also note that some of these 51 zones employed no infrastructural or service/quality of life incentives.

        While not shown in the table, on average, these zones began providing incentives in 1986. Some zones had been designated in 1989 and others were set up as early as 1981. Seventy-one percent of the zone coordinators surveyed said their enterprise zones were publicly advertised or promoted in some way. While that is a large portion of the zones, it does mean that 29 percent of enterprise zones do not promote themselves. On average, these zones cover over 18,000 acres. This converts to about 29 square miles. Some of the zones are immense and others are quite small--only 20 acres. The median size is 1,952 acres; this converts to 3.05 square miles. On average, 46 percent of this acreage is zoned for industrial use as opposed to residential or commercial. It is also true that 46 percent of these zones are in metropolitan areas.

What’s Responsible for the Growth?

        Most directly, growth within a zone is attributable to growth in the local economy for which it is a part. Therefore, I separated zones into different categories based on whether their firm or job growth rate was higher or lower than the county in which the zone is located. The bottom of Table 3 indicates that 42 of the 51 zones had a firm growth rate higher than its respective county; 38 of the 51 zones generated jobs at a higher rate as well. Thus, now we know that zones were able to grow and they were, on average, able to do so at a rate higher than we would expect given their economic context. This is good news. Note, however, that the growth within any one zone could have come from other parts of the county, thereby producing a zero-sum gain.

        So, what are the causes of this pattern overall? What has caused enterprise zones to grow at a faster rate than what their economic context would predict? Table 3 shows the difference of means for each of the incentive variables across the two dependent variables of firm growth rate relative to county and job growth rate relative to county. Since the number of cases that performed worse than the county is so low, I could not test whether the differences between the groups are statistically significant. However, the means for each of these categories are in the expected direction. Those zones with firm and job growth rates higher than their respective counties used more incentives than those with lower growth rates. This is true for each of the four kinds of incentives across the two growth measures. There appears to be a relationship between incentives offered and growth of the zones.

        To test the relationships between these causal variables and growth I used standard cross-tabulation procedures. Table 4 shows the relationships between the incentive variables and whether or not the firm and job growth in the zones was higher or lower than the county. Remarkably, only one variable has a significant relationship with these relative growth rates and it is not the tax incentives measure. Only the number of service/quality of life incentives offered by places has a positive effect on the ability of zones to outperform their respective county. To construct this measure I asked local zone coordinators whether they offered job training, day care for families of zone employees, added support for public schools, crime prevention or reduction programs, and housing rehabilitation. Examination of the specific responses these 51 cases provided to this question is of little help. My initial thoughts were that job training would be a defining category of these incentives; that is, I thought places with higher relative growth rates were more likely to be offering job training to zone residents or zone employees. This is not the case. There is no pattern suggesting what kind of specific incentives fostered what kind of growth. I can say that only four zones indicated they provided day care to employees families. A number of places are trying to rehabilitate their housing, fight crime, and eliminate drug use. However, again, there is no pattern to the use of these specific activities and growth.

        The implication, therefore, is that an overall attention to the quality of life in a place through job training, housing rehabilitation, drug and crime prevention, etc. matters for enterprise zone growth. Indeed, attention to these issues matters more than the tax, financial, and infrastructural incentives offered.

Socioeconomic Characteristics of Enterprise Zones in 1980 (Prior to Designation)

        While the previous section showed that enterprise zones have experienced growth, it is as yet unclear if this growth has led to improvements in the socioeconomic status of local residents. Since enterprise zones were originally intended and are still touted as a solution to urban blight and rural economic depression, it is important to examine if zones are such a solution.  Indeed many scholars have noted that development that does not lead to jobs will not help local residents--it doesn't improve well-being (Wilkinson 1989).

        Using maps provided by local enterprise zone coordinators, I identified a census tract or block group that exists wholly within each zone. The physical space covered by these is all zone land; the tracts and block groups are areas taken from the middle of the zone. The residents in these spaces are, therefore, all zone residents; the data presented in the following tables for these spaces pertain to those residents. These zone areas are predominately urban as well. Some nonmetropolitan areas had no tract or block group designations in 1980. I, therefore, identified the town/place that the zone was within or closest to. Some of these are also industrial parks, but in essence the same conceptualization applies; they are places with an enterprise zone within them or next to them. The data, therefore, cover people affected by the existence of an enterprise zone but not necessarily residents of one. The averages and extremes for the 170 places under study will be less than those for the 187 tract and 64 block group areas.

        Table 5 summarizes the socioeconomic characteristics for these enterprise zones. Over 22 percent of zone residents in 1980, be it in tracts or block groups, were poor. This is more than twice the national poverty rate for 1980 of 11.6 percent (Bureau of the Census 1981). Almost 1 in 3 of these zone residents were within 125 percent of the poverty line. Surprisingly, only 6 percent of tract residents and 7 percent of block group residents were unemployed; this is comparable to national figures from the Census Bureau (1981). Forty-three percent in each area were not in the labor force. This would suggest that a larger than average portion of zone residents were elderly (i.e., populations not expected to work). For a variety of reasons other individuals of working age may not have been looking for work. At least 8 percent of the housing units in these areas were vacant in 1980, and more than half the units were rentals.

        The minimum and maximum values for these measures (not shown) demonstrate some extreme variability in the socioeconomic characteristics of these places. Some zone areas were quite well off with either 1.78 percent poverty, no unemployment, or no vacant housing. The residents of tracts and block groups at the other extreme suffered under 76 percent poverty, 27 percent unemployment, and a 39 percent vacancy rate.

        The average values for the place areas are lower than the tracts and block groups because the places contain enterprise zones and some zone residents, but not everyone in the place lives in conditions similar to the residents of tract and block group areas that are within enterprise zones. So, while the minimum value for the places were as low as those for tracts and block groups, the maximum values are not as high.

        Also presented in the table are the averages for the counties the zones are located within. On average, the socioeconomic health of the tracts and block groups within enterprise zones is considerably worse than the county the zone is located within. In some cases, particularly the two poverty measures, the rate for the zone is twice that of the county’s. It is clear that the decision on where to locate enterprise zones is informed by their level of socioeconomic distress. This is less clear within the "place" category. The relative difference between these places and their nonmetropolitan counties is smaller.

Comparing Zones to Non-Zone Areas on Outcomes: Socioeconomic Change from 1980 and 1990

        In order to see if enterprise zones have been effective at improving the socioeconomic conditions of the local residents, the tracts, block groups, and places just described were matched with other areas that did not receive zone designation. The zones were first matched to economically distressed areas most like themselves elsewhere in the state but not in a county with an enterprise zone. The matching criteria included socioeconomic, ecological, and demographic factors such as poverty rates, racial composition, urban location, etc. The comparison created through these matched sets, then, is a "pure" test of zone effectiveness because members of the control group could have been designated a zone by the state, but were not. The zone areas were also matched to areas adjacent and similar to themselves but wholly outside the zone boundaries. This matched set permits tests of whether enterprise zones provide additional economic activity to these adjacent areas (a multiplier effect) or if, as critics charge, they pull investment out of these areas and into the zone. The design employed, therefore, can be represented by the following:

where O represents an observation of zone characteristics in 1980 and 1990 and X represents the treatment effect of zone designation. The comparisons were then made between the treatment group and the control groups.

        The first comparison I made involves the population counts for these enterprise zone areas. The enterprise zone areas experienced some significant population change during the 1980s. Table 6 shows the percent change in population size for the zone areas, their pure matches, and their adjacent matches. The tracts and block groups that are wholly within enterprise zones experienced a 4.45 and 8.95 percent decline in their populations, respectively. The places lost population as well, but at a lower rate.

        The pure matches to these areas grew and the differences between the two groups are significant for all three areas. Indeed the pure block group matches grew by over 6 percent while the block group areas within zones experienced the major decline just described. The adjacent place and block group matches also experienced declines, but they are not significantly different from the zone areas. The adjacent tracts did experience some population growth and it was significantly different from the zone tracts. The extreme values in Table 6 are also interesting. At least one each of the tract and block group zones lost a lot of people. One block group lost 97 percent of its population. On the other hand, at least one each of the tracts and block group areas doubled its number of residents.

        It is possible that enterprise zone areas are losing population because these areas have been "marked" by the locality for non-residential growth. That is, enterprise zones are being put in places where people are leaving and whatever population that remains moves out as their neighborhood is transformed into an industrial/commercial district. Related to this is the argument that, over time, people who can move out of these areas are doing so—leaving behind an increasingly isolated and socially depressed population (Wilson 1987 and 1996). If this is the goal of enterprise zones to begin with, then they are meeting their goal. They appear to be a new form of urban renewal and neighborhood reclamation.

        Table 7 launches us into the analysis of enterprise zone effectiveness. The values presented are changes in the percent of each socioeconomic characteristic. Thus, the 3.17 poverty value for tracts within zones, means that, between 1980 and 1990, 187 zone areas experienced a 3 percentage point increase in their poverty rate. Given that the poverty rate for census tracts in 1980 was 23.38 percent (Table 5), the 3.17 figure means that these same areas had a poverty rate of 26.55 percent in 1990. Once you understand this, you should then look at all the values for the enterprise zone areas. Nearly all measures of adverse socioeconomic conditions increased over the 10 year period; conditions worsened over the decade. This is bad news. It is also true that most of the values for the zones increased at rates higher than their matches (both pure and adjacent). The poverty and unemployment rates in the zones, for example, either grew at rates higher than the matches or declined less than the matches. This pattern is not as clear for the housing measures.

        Since the matched areas were chosen because of their social and economic similarities to the zones, these urban and rural areas were suffering from similar social disruptions before one of them was designated as a zone. The implication then is that enterprise zones were not able to help distressed areas. Zone areas continued to decline during the 1980s and at a rate faster than other spaces. Enterprise zones were not successful in their fight to offset or reverse the deleterious trends faced by inner-city or rural residents.

        While not provided here, I did analyze these differences further. The differences seen in Table 7 remained constant or became more pronounced once taking other important factors into account. Using regression analysis, I controlled for the number of years the zone has existed, how bad it was to begin with (places with low unemployment to start did have a harder time reducing their unemployment than other areas), and its racial composition. Allowing for these factors did not alter the main conclusion.

Summary and Implications

        Enterprise zones have experienced growth. They had more companies and jobs in 1996 than they did when they were designated. Most of the zones also performed better than the economic context within which they were located; they outpaced the growth in their respective counties. The zones were able to attract firms and jobs.

        Surprisingly, it is not the tax incentives offered by enterprise zones that had this effect, but the service/quality of life incentives. Though the results are weak, concern by local and state governments with community needs such as crime and drug-abuse prevention, housing rehabilitation, and job training were partly responsible for enterprise zone growth. At a minimum, these incentives matter more than others. Further, this finding is important given the common assumption by local, state, and federal officials that reducing corporate tax burdens will produce growth. Indeed, many enterprise zone coordinators indicated on the survey or in conversations that tax breaks, particularly property tax abatements, were most useful in promoting growth. I am confused as to why these groups and even some casual observers of the issue would think this way. We know that, on average, taxes only make up two percent of a company’s costs (Keischick 1981). But all enterprise zones offered at least some tax incentives, while far fewer offered infrastructural or quality of life incentives. It is intriguing why practitioners place so much emphasis on this method of achieving economic growth.

        Discovering this pattern should also make researchers question their approach to the incentives issue. Prior research has largely focused on whether the incentives were working. Little effort has been put into determining which incentives lead to zone growth. Therefore, given the results of this project, I hope both researchers and practitioners spend more time examining these differences and acting on what they discover.

        Enterprise zones have lost population while some comparable areas have gained. This result alone seems to indicate that enterprise zones are serving an urban renewal function. That is, areas designated as enterprise zones, with their depressive social conditions, have been identified by local actors for economic renewal. Companies are moving into zones and local residents are moving out. This alone is certainly not a bad outcome, but it is unexpected. Some may consider me naïve, but I was hoping enterprise zones would improve the lives of residents in place. If this is not a goal and the intention of zone programs is to transform residential areas into industrial/commercial districts, this should be made more clear. State-designated zone programs should acknowledge this goal and have in place programs to assist those displaced by the economic transformation.

        Enterprise zones have not improved the socioeconomic conditions of local residents. Enterprise zones have not been able to reduce unemployment or poverty; they have not led to housing rehabilitation. Indeed comparing enterprise zones to areas very much like themselves and located in areas that do not have zones themselves shows that the socioeconomic health of the zones worsened during the 1980s at a rate faster and more extreme than that of the control group. This is especially true of the enterprise zones in smaller towns. The zones in these places are performing significantly worse than other places very much like them. The "pure" test of whether enterprise zones have done what they set out to do shows that they have not.

        In order to deal with a common critique of enterprise zones--that they pull investment from adjacent areas, I matched the zones under consideration to similar spaces nearby. Like the pure matches, the adjacent areas also fared better during the 1980s than the zones themselves. The largest difference occurred among the 187 tract areas. If anything, enterprise zones may be helping adjacent areas by employing their residents or producing multiplier effects from the companies locating or expanding within the zone.

        Since these comparisons are made between zones and areas socially and economically similar to themselves, I have demonstrated that it is not necessarily the depressive conditions that shape zone failure. Overall, the tracts, block groups, and places that are very similar to those in enterprise zones in 1980 did significantly better over the decade.

Grant Acknowledgments

This research was made possible by grants from the US Department of Housing and Urban Development (# H-5941SG) and the National Science Foundation (94-12762).

Contact Information

Frank D. Beck, Ph.D.
Department of Sociology-Anthropology
Campus Box 4660
Illinois State University
Normal, IL 61790-4660

fdbeck@ilstu.edu
(309)438-5850

References

Beck, Frank D. 1998. "Human Ecology, Pro-Growth Effort, and Community Development: The Case of Enterprise Zones." [unpublished dissertation]. The Pennsylvania State University.

Bureau of the Census. 1981. Statistical Abstract of the United States. Washington, DC: US Department of Commerce.

Erickson, Rodney A. and Susan W. Friedman. 1991. "Comparative Dimensions of State Enterprise Zone Policies." Enterprise Zones: New Directions in Economic Development. ed. by Roy E. Green. Newbury Park: Sage Publications.

Kieschick, Michael. 1981. Taxes and Growth: Business Incentives and Economic Development. Washington, DC: Council of State Planning Agencies.

Wilder, Margaret G. and Barry M. Rubin. 1993. "Rhetoric versus Reality: A Review of Studies on State Enterprise Zone Programs." Journal of the American Planning Association. 62:473-491.

Wilkinson, Kenneth P. 1989. "Community Development and Industrial Policy."  Research in Rural Sociology and Rural Development. William W. Falk and Thomas A Lyson (eds.) Greenwich, CT: JAI Press.

Wilson, William Julius. 1987. The Truly Disadvantaged. University of Chicago Press.

Wilson, William Julius. 1996. When Work Disappears: The World of the New Urban Poor. New York: Alfred A. Knopf.