Abstract
This research explores the relative effectiveness of a comprehensive set of local economic development incentives and focuses on two questions: What contributions do common development tools make to the economic health of municipalities?; and, Are there other types of local activities, not typically considered as development tools, that might be more effective in contributing to local economic prosperity? It finds that the factors most consistently and positively related to economic health are investments in the downtown, spending on basic local public services, and using no economic development incentives at all. These findings suggest one primary policy recommendation: the wisest course of action for most cities would be to eschew particularized development incentives, especially those that require tax expenditures.
The old saying “if all you have is a hammer, every problem looks like a nail” is a good metaphor for how local economic development policy has been implemented in cities across the nation. Conceptions of what constitutes an economic development tool tend to be narrow. Most commonly, they focus on efforts to offset perceived disadvantages of a location (or to make an already attractive place more so). Combinations of subsidies and abatements to lower production costs for businesses constitute many economic development packages. 1 As a result, economic development policies tend to be highly path dependent; older tools continue to be used even while new ones are added.
National surveys of local economic development practice over time, conducted by the International City/County Management Association (ICMA), show a relatively narrow list of tools and activities to be most common: collaboration between local governments and chambers of commerce; business surveys and calls on individual businesses; streamlined zoning and permitting processes; promotional and marketing materials; infrastructure investment; tax increment finance districts; and tax abatements (see Reese & Sands, 2012; Zheng & Warner, 2010). Thus boosterism, business input, and bonuses have long been the hammers of local development toolkits.
Much research has been devoted to assessing and evaluating 2 the extent to which these types of development policies achieve the goal of increased local economic prosperity, often raising questions about their effectiveness. There are several gaps in this literature, however. First, evaluations have tended to focus on one tool at a time—tax abatements, tax increment finance authorities (TIFA), enterprise zones, and so on. While enterprise zones, for example, typically include the application of a number of different tools, it is not possible to evaluate the effectiveness of each of the components parts in relation to each other. Second, narrow conceptions of what constitutes a development tool have left unexplored many local policies that might also hit the nail of economic prosperity. Tax expenditures such as abatements or tax increment finance authorities have opportunity costs; if a certain percentage of local taxes are not being collected, then revenue is lost that could be used to pay for public services such as police or parks and recreation.
This article asks the following two research questions: (1) What contributions do common development tools make to the economic health of the residents of municipalities particularly in relationship to each other?; and, (2) Are there other types of local policies, not typically considered “economic development,” that might have a stronger relationship to local residential health? In addressing these questions the research makes the following contributions to the literature on local economic development policy:
Considers economic development tools in comparison to spending for other public sector functions in relation to the economic health of local residents;
Compares creative class policies to more traditional supply-side approaches; and
Explores these relationships for a large number of cities, of all sizes, in a single state environment over time.
Assessing the Standard Tool Box
Although a wide array of local economic development tools are enabled by state governments, those most commonly used are more limited, as suggested by the national surveys conducted by the ICMA. This research focuses on five major economic development tools in the State of Michigan—Industrial Facilities Tax Abatements; Renaissance Zones; Tax Increment Financing Districts; Cool Cities Grant and Planning Programs; and MEGA (Michigan Economic Growth Authority) grants. It assesses their relative relationships to local residential fiscal health while explicitly considering the potential benefits of forgoing such tax expenditures at the local level and spending for basic local services.
With the exception of the Cool Cities initiatives, these development tools focus on the supply-side as defined by Eisinger (1988). Specifically they are directed at lowering the costs of production for the private sector and are aimed at attracting or expanding businesses (Blair & Reese, 1999); they have, as their defining goal, the reduction of costs for businesses in specific communities or states. It has been suggested that supply-side tools are based on four central assumptions or premises: growth results from the attraction or expansion of mobile firms; each city or state is in competition with others (variously defined by proximity or type); cities or states should try to attract/retain all possible firms; and, economic development policies should be directed at the reduction of costs of doing business (Reese, 1992).
Cool Cities Initiatives, on the other hand, are directed at the creation of new business establishments in downtown areas (all tools are described in detail in the following section). Thus they are demand-side policies that are designed to encourage and foster the development of new businesses, capital, research and development, and the like (Eisinger, 1988). Demand-side policies have as their central goal the development of new business and the stimulation of the local market to facilitate business development and expansion. Thus the government takes a more active role in identifying markets and aiding in the creation or expansion of local firms, often focusing on the start-up phase (Reese, 1992). The following sections discuss the extant evaluative literature on the five development policies explored in this study as well as that related to spending for local public services.
A critical element of evaluations of local economic development policies is the determination of the dependent variable. The obvious question in judging the effectiveness of policy is effectiveness in what? Selection of a dependent variable is not straightforward in this case for two primary reasons: (a) there is little agreement in either the academic or political spheres on the appropriate definition of success for economic development policies; and (b) there are significant (and common) data availability problems. Development “success” has been measured in a variety of ways including employment growth, job generation, income growth, increased tax base, dollars of investment, and numbers of businesses attracted (see Abravanel, Pindus, & Theodos, 2010, and, Reese & Fasenfest, 1997, for summaries of this literature). Yet these indicators do not mean that a good outcome for businesses leads to other good outcomes such as improved economic conditions for residents (Beauregard, 1999; Felsenstein & Persky, 1999; Reese & Fasenfest, 1999).
In this study, the dependent variable is the economic health of the residents of a particular city. Aside from data availability issues (discussed more fully in the method section), this variable posits a slightly different conceptual question about economic development tools. As opposed to asking whether they benefit businesses or contribute to the local tax base, the focus here is on residents of communities. In other words, do the tax expenditures inherent to the programs explored here contribute to the fiscal well-being of residents? Are there other public expenditures that could do a better job of enhancing residential fiscal health? This variable has been used in a number of other studies to represent the economic health or well-being of a city’s population (Feiock, 1992; Fleischman et al., 1992; Moss, 1997; Reese & Rosenfeld, 2002; Reese & Ye, 2011; Sands & Reese, 2008, 2012; Sharp, 1991; Wolman, 1996). Again, as noted in the introduction, the purpose of this analysis is to compare traditional supply-side tools (and one demand-side tool) to spending for local government services in their contribution to the economic health of residents. The conceptual premise of economic development policies focusing on the private sector is that benefits will trickle down to local residents. Yet spending on local services would appear to benefit residents more directly and thus may represent a better investment of local revenue if the goal is to improve the quality of life in the community (Fasenfest, Ciancanelli, & Reese, 1997; Reese & Ye, 2011).
Evaluation Literature
Special development zones
Special development zones such as Enterprise Zones (EZs) reduce business costs of operation by eliminating property and other local taxes, or by providing workforce training programs, onsite infrastructure and land assembly, lenient permitting and regulations, tax credits for job creation, and in some state versions like the Michigan case examined here, abatement of property taxes for residents (Elvery, 2009; Peters & Fisher, 2002). Much has been written about the effectiveness of EZs and similar development zones. Generally, research has suggested that zones tend to move firms around rather than creating new enterprises, displace current residents and businesses due to increases in land values, and do not create economic growth commensurate with the inherent level of tax expenditure (see Peters & Fisher, 2002; and Wilder & Rubin, 1996, for reviews). However, some research has found that businesses located within zones experience more rapid growth than comparable firms outside them, at least initially (Bondonio & Greenbaum, 2007; O’Keefe, 2004). There has been some research suggesting that enterprise zones may reduce resident unemployment somewhat but any changes do not reach statistical significance (Greenbaum & Engberg, 2000; Rogers & Tao, 2004).
Many studies have been less positive. It appears that zones do not create significantly more employment than areas not designated (Greenbaum & Engberg, 2000) although firms located in zones may increase employment at a somewhat faster rate (Bondonio & Engberg, 2000). The incentives offered in the zones do not appear sufficient to offset the locational disadvantages of distressed areas and are not a significant factor in business location decisions (Dabney, 1991).
An evaluation of Louisville’s enterprise zones found that few development goals were met, particularly given the high cost of the program; for areas that did experience benefits, the success did not appear to be the result of the zones (Lambert & Coomes, 2001). In particular, there does not appear to be any significant job benefits for residents of zones since most of the workers commute from other areas (Elvery, 2009; Peters & Fisher, 2002), nor any significant revitalization for either specific neighborhoods or the community in general (Lambert & Coomes, 2001). Furthermore, while new businesses in zones appear to create additional jobs, their numbers are not sufficient to offset the negative effects on firms located in the area prior to zone designation (Greenbaum, 1999). Research on other countries such as Great Britain have found enterprise zones to be even less effective than those in the United States (Rubin & Richards, 1992)
Michigan’s Renaissance Zones (RZs) represent a more radical option in terms of foregone revenues than the federal Enterprise Zone program. They are aimed at generating neighborhood or targeted area redevelopment. Early assessment of the then 11 Renaissance Zones in the state indicated some visible success although some zones were more successful than others (Sands, 2003). While all of the zones reported some development activity, 16 of the 41 subzones had seen little or no private investment in the first three years of their existence. And, the zones appeared to have had little effect on statewide revitalization, only mixed results in promoting area redevelopment, and generated virtually no spillover effects (Sands, 2003).
Tax abatements
Aggregate trends in abatement use across the United States indicate that cities have offered them at a relatively static rate between 1994 and 2009—around 42%, on average (Reese & Sands, 2012). Proponents of tax abatements contend that the new investment and additional indirect and induced spending produce total revenues sufficient to improve the overall economic welfare of citizens (Goss & Phillips, 2001) and that abatements can address underemployment in slack labor markets (Gramlich, 1997). In one evaluation, tax abatements appear to have had a positive and statistically significant impact on economic growth for low unemployment counties but no significant impact on growth in high income ones (Goss & Phillips, 2001). Surveys of recipients of tax abatements have generally reported that the incentives were effective in influencing the location decisions of firms (Calzonetti & Walker, 1991; Premus, 1982; Rubin, 1991). This suggests that tax incentives can be used to more evenly spread development across jurisdictions within a state, but only if targeting is part of the incentive legislation (Goss & Phillips, 2001). Bartik (1991) has provided similar findings, measuring the effect of state and local fiscal variables on economic development, concluding that local taxes and expenditures exert a significant influence on the level of economic activity observed in a region.
Other assessments, however, have concluded that abatements were basically ineffective in stimulating local economic growth (Fullerton & Aragones-Zamudio, 2006; Sands & Reese, 2012; Schmenner, 1982). Generally, critics contend that few measurable benefits accrue from tax abatements and they find little support for the supposition that any new jobs are “created” (Fisher & Peters, 1998; Peters & Fisher, 2004; Wassmer & Anderson, 2001). Rather, while tax incentives may influence the choice of a particular site, overall they tend to move jobs from one location to another, with little or no net increase in total job opportunities occurring as a result (Hovey, 1986).
On a more narrow scale, it appears that abatements fail to have an additive impact on overall business activity, and any positive effects are short-lived, occurring only in the time period immediately after the tax incentive program begins (Wassmer, 1994). For example, in a Detroit-area study, manufacturing property tax abatements were found to increase property values in only one year (the earliest, 1977) of a longitudinal study. By 1984, 1987, and 1992, there were no salutary effects of manufacturing property tax abatements, and commercial tax abatements had no effects in any time period (Wassmer & Anderson, 2001). In their meta analysis of tax abatement studies, Peters and Fisher note, “the best case is that incentives work about 10 percent of the time, and are simply a waste of money the other 90 percent” (2004, p. 32).
The Michigan Economic Growth Authority (MEGA) grants examined here are a type of tax abatement policy specifically focused on tax credits (see the following section for more detail on the program). An econometric analysis of this program explored its effects on aggregate income, employment, unemployment, and activity in the manufacturing, wholesale, and construction sectors. Findings indicated short-term increases in construction employment but no impacts on any of the other indicators (Hicks & LaFaive, 2011). Thus the authors concluded that, from 1995 at the beginning of the program to 2002, there had been “little, if any, benefit accruing to the counties in which the projects were located or in surrounding counties” (Hicks & LaFaive, 2011, p. 203). Research on a similar program in Ohio also found no positive outcomes in terms of job growth (and that firms significantly overstated growth in self-reports of employment, see Gabe and Kraybill, 2002).
In summary research on property tax abatements has raised concerns that they
are only effective at the margins in business location decisions;
serve to increase the “zero sum” aspect of local development;
tend to redistribute public sector revenues to private sector interests;
are used primarily by healthy cities that can “afford” to forgo the potential tax revenues;
tend not to produce jobs and tax base benefits commensurate with the loss of local revenues;
have not achieved the levels of growth desired and have negative secondary impacts;
are essentially useless because firms would have remained in place, or even expanded, absent the incentives, thus cities and states “pirate” jobs from each other;
fail to have an additive impact on overall business activity and have only very short-lived positive effects; and,
are not large enough to counterbalance negative attributes of otherwise high cost or undesirable locations. (Dabney, 1991; Fisher & Peters, 1998; Hood, 1994; Lynch, Fishgold, & Blackwood, 1996; Peters & Fisher, 2002; Sands & Reese, 2008, 2012; Schwarz & Volgy, 1992; Wassmer, 1994).
Tax increment financing (TIF)
Tax increment financing is often an attractive tool because it allows municipalities to undertake important and costly improvements (particularly when bond financing is required) without levying new taxes. 3 Once a TIF district has been established, any subsequent growth in aggregate property tax revenues, either as a result of new construction or rising values of existing properties, is “captured” by the district and used for investments within. The property tax rates are the same within the tax increment finance district as elsewhere in the community; the incentive for being included in the district is that taxes paid result directly in public investments in the district.
Initially employed to offset declining federal assistance for urban redevelopment projects, tax increment finance districts are now commonly used in urban areas across the United States (Goshorn, 1999, p. 924). Most states enabled local governments to use TIF authorities by the turn of the century (Johnson & Mann, 2001); the majority allow the use of TIF districts for either commercial or industrial projects and 30 states allow tax increment financing for residential uses. Tax increment financing is also increasingly used to fund brownfield remediation. Over time, then, TIF has become one of the most popular economic development tools, with steadily increasing use since the 1970s (Forgey, 1993). In part, its popularity lies in the presumptive development benefits: municipalities do not have to pay for improvements via increased taxes; TIF creates the potential for increased revenue growth without an increase in taxes; for business owners in the TIF area higher levels of services can be provided without paying higher taxes; and, they can aid redevelopment in targeted areas of blight or financial need (Sands et al., 2007; Weber, 2003).
Criticisms of tax increment financing include
Even when property values rise and TIF revenues grow, the funds are not available to support general services. Thus municipalities may need to raise citywide tax rates to ensure that necessary services are provided outside of the district or decrease services. This raises equity issues as potential public revenues are diverted to business districts (Sands et al., 2007; Weber, 2003).
The tax increment finance district may divert substantial revenues from local school or park districts. Such jurisdictions may also be required to provide additional services as a result of development in the TIF (Huddleston, 1986; Lehnen & Johnson, 2001; Stinson, 1992; Weber, 2003).
A concern that designation is used arbitrarily, without meeting rigorous standards for the allocation of tax revenues, that is, blight and necessity (“but for”). Flexible standards in the definition of blight can result in private subsidies rather than true economic development (Dye & Sundberg, 1998; Luce, 2003).
TIF can be problematic if sunset provisions are not included.
TIF districts are governed by boards that lack transparency and accountability.
TIF is only likely to be effective when new development does in fact occur or where property values are increasing, thus the original designation should be done carefully (Sands et al., 2007).
Tax increment finance districts can take a number of different forms (Sands et al., 2007; Weber, 2003). They may provide the basis for bond financing of a specific public improvement, such as a water or sewer line or be used to meet the cost of remediation of brownfield sites. Downtown Development Authorities may use revenue from TIF districts for a range of activities, including ongoing operating expenses, infrastructure improvements (parking decks and street improvements for example), subsidies to firms or community events. Michigan law allows seven different types of tax increment finance districts/authorities, the most common of which are Downtown Development, Local Development Finance, Brownfield Redevelopment, and Tax Increment Finance Authorities.
Creative class, arts, and culture
The interest in promoting arts and culture as an economic development strategy is often attributed to Florida’s book on the creative class (2002). The Cool Cities Initiatives in Michigan are based on creative class arguments and represent the only demand-side tool examined here. According to “creative class” proponents, successful local economies will need to rely on information and creativity for their well-being and as a source of economic vitality (Florida, 2005; Glaser & Mare, 2001; Ley, 2003). The literatures on the creative class and culture and the city argue that three goals can be achieved by a focus on these issues: economic development, regeneration or revitalization, and cultural effects (Markusen & Gadwa, 2010). Of these, however, economic development goals often take the fore (Grodach, 2012). This is evidenced by research showing many cities moving to place cultural affairs, education, and recreation within their economic development functions (Grodach & Loukaitou-Sideris, 2007) and local officials indicating that their primary motivation for investment in artist housing is economic development (Strom, 2010). Michigan’s Cool Cities grant initiatives were designed to use culture and art as an economic development stimulus through business creation and retention of creative class populations.
Although an attractive argument on paper (the creative class concepts have been described as “politically seductive”; Peck, 2005, p. 766), extant research has questioned many of the operational components and, more importantly, the effectiveness and efficiency of creative class economic development policies. The connections and processes required to support the notion that creativity, or “creatives,” lead to economic prosperity have not been sufficiently tested empirically, the assumptions embedded within creative class arguments have raised many questions among academics and other policy evaluators (Ley, 2003; Peck, 2005; Scott, 2006; Markusen & Gadwa, 2010; Thomas & Darnton, 2006; and many others), and much policy activity has proceeded robust evaluations.
There is a growing body of research exploring whether the creative class actually leads to, or is even correlated with, economic growth. Indeed, there appears to be no discernible relationship between economic growth and any of the commonly used creative class indicators (Hoyman & Faricy, 2010; Sands & Reese, 2008). Recent work of this nature suggests that high tech employment, in particular, is unrelated to economic health or, in some cases, appears to be negatively correlated with economic growth (Hoyman & Faricy, 2010). Research has found entertainment employment to be negatively related to health and only finance, insurance, and real estate (FIRE) employment to be positively related; most typical creative class employment categories do not remain significantly correlated to health in multiple regression (Reese, 2012). Other research indicates that education and skill development appear more important than culture or amenities in economic growth (Glaeser, 2005; Glaeser & Mare, 2001) and, indeed, that innovation appears just as likely in older manufacturing centers as newly creative cities (Chapple, Markusen, Shrock, Yamamto, & Yu, 2004). Of all the aspects of the creative class arguments, having a highly educated population appears important; however, graduation rates from local schools (a possible proxy for the quality of the local school system), appear more critical than higher education (Reese & Ye, 2011).
But the individuals within the creative class are only a small part of the creative economy arguments as noted earlier. Investments in physical art spaces and artist housing have been found to have many beneficial economic and community-building outcomes including acting as neighborhood anchors (Seifert & Stern, 2010; Strom, 2010) allowing often marginalized groups to have a space for activities or ventures (Borrup, 2006), serving as incubators for artists (Montgomery, 2007), and creating a place to build social networks (Grodach, 2010). Artists and community cultural activity have been argued to revitalize neighborhoods (Lloyd, 2005) and create bridges between classes and cultures (Alvarez, 2005; Wali et al., 2006).
Public Spending and Investment
The whole premise of economic development policy, from location incentives through human capital development, to the creative class, is based on the assumption that public policy matters. But a variety of local policies can have effects on the economic health of a community beyond traditional economic development tools including infrastructure, education, and a host of other local amenities and services. While not commonly considered to be economic development “tools” per se, investment in the quality and quantity of local services can make a significant contribution to the economic health of residents and serve as an attractive feature for businesses and entrepreneurs considering alternative locations (Besser, Recker, & Parker, 2010; Florida, 2002; Gottlieb, 1994; Liu, Kolenda, Fitzpatrick, & Todd, 2010; Reese, 2012; Reese & Ye, 2011; Trip, 2007). Investments in public safety have been found to be related to economic health (Reese & Ye, 2011) and crime has consistently been found to be a drag on economic growth (Bowes, 2007; Erickcek & McKinney, 2006; Gottlieb, 1994; Liu et al., 2010; Trip, 2007).
To assess the relationship between community economic health and these quality of life contextual features, a number of public policy variables are considered here including economic development policy, public spending for a variety of services, and education spending and performance. Investment in education and public services has been shown in past research to contribute to economic prosperity. Specifically, investment in local schools has been suggested as a driver of economic growth (Gottlieb, 1994; Wrigley & Lewis, 2002). On a more macro scale, research has indicated causal connections between human capital accumulation and economic growth (Krueger & Lindahl, 2001; Lucas, 1988; Toya, Skidmore, & Robertson, 2010, for example). Similarly, scholars have argued that public services or investment in amenities such as recreational opportunities can contribute to a local economy (Deller, Tsai, Marcouiller, & English, 2001; Goe & Green, 2005).
Method
Evaluating Economic Development Policy
At the outset it is important to be clear on what the available data from Michigan allow for in the way of policy assessment. Bartik and Bingham (1997) raise a number of concerns about evaluations of economic development incentives finding serious methodological limitations in most evaluation studies. When evaluations are based on surveys of recipients, there is often an incentive to lie about the importance of incentives in the location or investment decision, since necessity (“but for” requirements) is often a prerequisite for receiving benefits. Studies that focus on changes in employment levels typically do not consider the context in which these changes occur. For example, abatements granted at the low point of the business cycle are more likely to exhibit positive results than a similar incentive implemented just as the local economy enters a period of decline. Bartik and Bingham ultimately suggest that a more comprehensive community impact study is required to accurately assess the effectiveness of a particular economic development program.
Evaluations of economic development policies are difficult to design and conduct because of the inability to create counterfactuals and measure program outcomes as opposed to processes (Abravanel et al., 2010; Bartik & Bingham, 1997; Reese & Fasenfest, 1999). Other challenges to economic development policy evaluations include the fact that cities are complex systems making it impossible to control for all pertinent factors; programs are usually small relative to the size of cities and/or their problems; different metrics and measures have been employed across studies; and the time needed for any result to take hold may be beyond the scope of the evaluation period (Abravanel et al., 2010; Hollister, 2007; U.S. Government Accountability Office [GAO], 2009).
The available data for the Michigan economic development tools examined here have many of the limitations just noted. Complete data are not available over time for most of the programs, and outcome data, in terms of jobs created or retained, are not collected either on a statewide basis or in most cities. Thus policy effort in terms of abatements granted or tax increment districts created is what can be measured (more on this point follows). There is no evident event to act as an instrumental variable to rule out endogeneity between residential economic health and the application of economic development policy. Because this study focuses on all cities in the State (N = 286) what it gains in breadth and generalizability, it loses in the ability to measure uniform control variables for the whole population. Thus the study does not constitute a formal evaluation of each economic development policy. Language used in the article refers to assessment or correlation for this reason. Yet the study is still valuable. It differs from extant research in that it focuses on five economic development tools used in a broad range of cities. It compares the relationship between effort across the tools to local government spending for a variety of common services and functions. Hence it addresses the important question of what expenditure of local government resources is most often correlated with the economic health of residents—commonly used supply-side policies, a creative class demand-side policy, or regular expenditures for local government services.
Case and Data
The population of cities in the State of Michigan provides the data for this study (N = 286); thus the unit of analysis is individual cities as opposed to metropolitan areas. The Michigan case is both exemplary and cautionary. As one of the states longest and hardest hit by the Great Recession, current local fiscal conditions present a worst case scenario. Realistic hopes for successful outcomes of economic development activities are arguably most challenging (although some communities have withstood the recession in relatively good shape). Michigan municipalities have also evidenced long-term and consistent use of many common economic development tools (Reese & Sands, 2007). While the findings may not apply to states and cities with more robust economies, or where population and economic growth are occurring, they do speak to conditions in many deindustrializing regions.
Budget and revenue data for Michigan’s cities beginning in 2005 and ending in 2010 serve as the root of the database. While municipalities in Michigan are required to file Annual Local Unit Fiscal Reports with the state, the budget and revenue data therein were not available in a usable and standardized manner; thus, systematic analysis of local government finances has been limited. Michigan State University Extension (MSUE) worked with the State of Michigan Department of Treasury to develop a web-based local government financial data management system. An array of local fiscal data is now available and has been employed here (see Skidmore & Scorsone, 2011, and Michigan Department of Treasury F65 Government Fiscal Data Portal, http://f65.mitreasury.msu.edu/). 4
Selected census data from 1980 to 2010 were added to the data set, including demographic composition, employment, and economic conditions (poverty, income). 5 Finally, data from five of Michigan’s local economic development programs were included: Industrial Facilities Tax Abatements; Renaissance Zones; Tax Increment Financing Districts; Cool Cities Grant and Planning Programs; and MEGA (Michigan Economic Growth Authority) grants. These specific tools are the focus of the research for several important policy and theoretical reasons:
Their widespread use;
Their availability to a broad spectrum of Michigan municipalities; and,
Statewide data availability.
In addition to the five economic development tools enabled by the state, local spending for a variety of basic government functions, including economic development, is also considered. These data provide a local context for the programmatic assessment and broaden the definition of what constitutes a local development tool. The respective data are described below; more detailed program descriptions can be found at www.gusp.msu.edu.
Independent Variables: Local Economic Development Programs and Spending
The five economic development tools and government spending variables are described below. It should be noted that, with the exception of the last variable—government spending—and one of the creative class programs (Neighborhoods in Progress), dollar values devoted to the program or value of taxes abated per community are not available. Thus program effort is indicated by numbers of abatements or numbers of development zones. This is, of course, not ideal (see Wolman, 1996, for an extended discussion). It is possible that greater effort as indicated by the value of dollars abated or tax increment dollars achieved would reap more positive outcomes than a simple count of abatements or districts would indicate. As the program descriptions below will show, almost all of the tools examined, with the exception of the Cool Cities Initiative, involve primarily tax expenditures rather than up-front revenue invested. As such it is not clear that greater effort in terms of taxes forgone would lead to better outcomes for local residents—indeed the reverse is hypothesized to be the case.
Tax Abatements: The tax abatement process under Michigan Public Act 198 of 1974 allows a local government unit to establish a plant rehabilitation district, an industrial development district, or both, if it levies taxes that equal or exceed 30 mills. This criterion was met by virtually every one of Michigan’s municipalities. The tax abatement data come from the files of the Michigan Economic Development Corporation (MEDC) and its predecessor agencies. These data are available for all abatements awarded from 1980 to 2006 and include the number of abatements, projected real and personnel property investment, and projected retained and created jobs. 6
The Industrial Facilities Tax Abatement program has been popular with both firms and local governments. From its inception in 1974 through 2005, Michigan local governments granted a total of almost 18,600 tax abatements. Tables 1 and 2 summarize the number of abatements granted by Michigan cities. 7 While abatement use by all municipalities in the state (cities, townships, and villages) began slowly and increased over time, abatement use by cities has been high and relatively steady over the 26-year period of data here (Sands and Reese, 2012).
Total City Tax Abatements 1980-2006.
Note. Drawn from: Citizens Research Council (1986); MEDC.
Average City Tax Abatements 1980-2006.
Note. Drawn from: Citizens Research Council (1986); MEDC.
While there are a handful of cities that have granted large numbers of abatements, mean abatements per community are much lower, ranging from 6 to 11 in each period. Although based on estimates contained within the abatement applications, it is evident in both tables that personal property (equipment) exceeds real property (new facilities) investment and that estimated retained jobs far exceed new ones. It is important to remember that these are projections not actual outcome data.
Renaissance Zones: PA 376 of 1996 grants all occupants of Renaissance Zones exemptions from a dozen different state and local taxes, including state and local income taxes and most property taxes; the state sales tax, and any ad valorem taxes pledged to the repayment of bonded indebtedness are the only non-Federal taxes that continue to be collected. Renaissance Zone data were also obtained from the Michigan Economic Development Corporation. RZs in Michigan are significantly different from the national Enterprise Zone program in that both businesses and residents are the beneficiaries of the tax incentives, thus they are designed to avoid some of the negative aspects of the national program found by other researchers, that is, that local residents are disadvantaged in relationship to or displaced by new businesses (Peters & Fisher, 2002).
By design, the communities with early Renaissance Zones, from program inception in 1996 to just prior to program extension in 2000, were fiscally stressed; Benton Harbor, Detroit, Flint, Lansing, and Saginaw had early zones. However, as different types of zones became eligible for the program, the number and range of municipalities with zones expanded to include smaller and arguably healthy communities. Although researchers have suggested that such expansion of the national Enterprise Zone program changed its nature from a spatially targeted one into a program to improve state economic competitiveness, that a wider number of residents enjoy tax breaks mitigates this effect in Michigan (Turner & Cassell, 2007).
Over the course of the program, the 10 cities making the greatest use of Renaissance Zones (each with more than 12 zones) are Detroit, Grand Rapids, Muskegon, Saginaw, Flint, Kalamazoo, Jackson, Benton Harbor, Alpena, and Lansing. Table 3 provides descriptive data on Renaissance Zones from 2005 to 2010. Although types of zones eligible have increased over time, the absolute number of zones has declined somewhat as early zones have expired. Job and investment data are, as with tax abatements, estimates drawn from applications. Again, projected retained jobs far exceed new ones.
Total Renaissance Zone Activity 2005-2010.
Tax Increment Financing Districts: PA 450 of 1980 allowed municipalities broad use of tax increment financing; the program has been closed to new applicants since 1987 replaced by eight separately enabled programs. The original Tax Increment Financing Authority (TIFA) legislation allowed for any type of land use including commercial, residential, and industrial. The State of Michigan does not compile lists of tax increment finance districts; data for districts were assembled from the Michigan Departments of State and Treasury, individual city websites, and then through snowball sampling as part of a statewide population survey of TIFAs (Khan, 2012). Most tax increment finance districts in the state are believed to be included in the data set; however, it is possible that some are missing. Five of the most common TIF programs are explored here—Tax Increment Finance Authorities (TIFA), Brownfield Redevelopment Authorities (BFRA), Corridor Improvements Authorities (CIA), Downtown Development Authorities (DDA), and Local Development Finance Authorities (LDFA). Brief descriptions are provided below. Ninety-one cities in the state have Tax Increment Financing Authorities or 32% of all cities; 132 cities have Brownfield Redevelopment Authorities (of these, 13 have two and one city, Kalamazoo, has five) for 46%; only four cities have a Corridor Improvement Authority; 203 (74%) cities have at least one Downtown Development Authority with 10 having two; and, 26% of cities have Local Development Finance Authorities (75).
Brownfield Redevelopment Authorities (BFRA) were authorized in 1996; the program allows municipalities to establish BFRAs and use tax increment financing for environmental remediation of brownfield sites. Corridor Improvement Authorities were enabled by PA 280 in 2005. Under the program municipalities may establish one or more authorities that use tax increment financing to make capital improvements within an established commercial district. Downtown Development Authorities (DDAs) were enabled in Michigan in 1975 through PA 197. As is common, DDAs can raise revenue for physical infrastructure improvements, property acquisition, marketing, and operations through tax increment capture for a designated downtown area. Local Development Finance Authorities were enabled by the State of Michigan in 1986 through PA 281. These are designed to allow local governments to target industries by type with eligibility limited to manufacturing or processing of goods or materials by physical or chemical change; agricultural processing; high technology activities; energy production, primarily from biomass or wood waste; or business incubators (Citizen’s Research Council of Michigan [CRC], 2007, p. 83).
Cool Cities Initiative: This program, under which a number of separate programs were consolidated in 2004, was a clear effort to address Florida’s contention that a focus on the creative class would bring economic growth benefits to communities. Data for the complete history of the Cool Cities Program came from reports of the Senate Fiscal Agency and from files supplied by the former Cool Cities program director now with the Michigan State Economic Development Authority (MEDA), as well as information drawn from http://www.coolcities.com/main.html. Four initiatives were offered; The Neighborhoods in Progress program was the only piece of the Cool Cities initiative that provided significant funding for local projects through state matching money. The program focused on investments in downtowns that would attract and retain city center residents. The Main Street program was directed at preserving and managing historic downtowns by providing customized technical training. Blueprints for Downtowns and Blueprints for Neighborhoods were directed at creating plans for the revitalization of downtowns and of neighborhoods contiguous to a traditional downtown through support for a consultant (for the former) and a reward of eligibility for community development block grant money (for the latter).
For the Neighborhoods in Progress program, 31 municipalities received grants, 25 of them for a single project. Four cities received four grants each; Flint, Kalamazoo, Lansing and Saginaw. Grand Rapids received 5 grants and Detroit received 10. Each grant was either for the maximum US$100,000 or very near that amount (one was for US$90,000 and another for US$99,000). Thirteen municipalities received Main Street support, 44 Blueprints for Downtowns, and six Blueprints for Neighborhoods.
MEGA Grants: The state MEGA grant program was enabled in 1995 and discontinued in 2011 although benefits continue for businesses already receiving awards. The program offered single business and income tax credits targeting firms making large-scale investments and creating jobs. Data for all MEGA grants for the history of the program were provided by the Michigan Economic Growth Authority, a division of the MEDC. Under half of the cities in the state are the location of a firm receiving a MEGA grant (45%). And, numbers of MEGAs per community are quite small; 79% of those with MEGAs have had five or less, 60% have had only one or two. The highest number of MEGA projects (47) has been in Troy, while Ann Arbor and Auburn Hills have had 43, Detroit 28, and Grand Rapids and Southfield 25. The proposed mean number of created jobs per city is 567 and the mean of retained jobs is 645.
Local Spending: Local government spending data were obtained from the Department of Treasury Fiscal Data set. The reporting form for local spending includes seven categories: general government (salaries of elected officials, finance and tax administration, buildings and grounds); public safety (police and fire services, jails, building inspection); public works (roads, sewers, water, public transportation); health and welfare (health departments, medical examiner, emergency services); community and economic development (economic development, redevelopment and housing, planning and zoning); recreation and culture (parks and recreation, libraries, culture); and “other” (fringes and benefits, debt service, capital outlay). Average expenditures are grouped by the seven spending categories indicated in Table 4.
Mean Spending By Category. a
Data represent total spending not adjusted for population size.
The data in Table 4 indicate a general pattern of stability in local government expenditures. The percentage spending on all categories is quite static over time, except for a slight increase in health and welfare and community and economic development over recreation and culture spending in the 2009-2010 period. Spending levels were flat during the Great Recession in 2008 and 2009 but increased in 2010. Public works spending is highest in all years, followed by other and public safety. Much less is spent on average for the remaining categories. Health and welfare spending appears the most volatile over time but by 2010 is similar to general government along with community and economic development. Recreation and Culture, and general government are quite flat over time. Community and economic development shows the same pattern until an uptick in 2010.
Dependent Variable: Residential Economic Health
The focus of the project is on the relative relationship of five common economic development tools and government spending to residential fiscal health. This is the case for several reasons. The first relates to data availability. For tax abatements and MEGAs, relevant outcome measures are contained in the applications: investment and amount of new and retained jobs. However, these figures are only estimates used for the purposes of the application; in most cases, actual data for these variables do not exist. Firms in most communities are not required to report actual performance on these indicators; there are no requirements that municipalities report these data to the state. 8 For Renaissance Zones, municipalities may voluntarily report annually on new and retained jobs and private investment. Data on outcomes for the Cool Cities program and various tax increment finance programs are similarly nonexistent.
While the data availability just discussed is specific to the Michigan case, the absence of reliable outcome data is common in evaluations of economic development policies more generally. Thus, assessments of economic development policies must often use a variety of proxy indicators. The indicator of “success” used in this study is one such commonly used proxy. The dependent variable, residential economic health, is an index comprised of three items drawn from the census: median household income, unemployment, and poverty. 9
A general caveat about issues of causation or time ordering is important at this point. Because of the nature of the available data, it is impossible to establish with certainty that a particular economic development tool caused a particular level of residential economic health. There are simply too many variables for which data are unavailable. However, because data over time are available for many indicators, the sequencing of variables in time can be specified. Census data from four decades provide a sense of trends as past demographics and economic conditions emerge as current ones. In addition, for most of the development tools, it is possible to compare past use to future residential economic health. Thus, while these lagged relationships are not conclusive, they are considered indicative of time ordered relationships. Given this caveat, the analysis proceeds to identify the correlates of economic health and also the temporal ordering of relationships where possible.
Correlation Analysis
The assessment of development incentives begins with correlational analyses to identify which tools are related to economic health, then proceeds to regression models to assess relative effects. 10
Tax Abatements
For ease of presentation, the use of tax abatements has been grouped into five time periods corresponding to economic cycles in the State. Table 5 presents the correlation coefficients of number of abatements granted and resident economic health for each time period. The obvious observation is that there is no significant correlation between tax abatements and economic health or change in health over time.
Tax Abatements and Health.
Note. N = 221 independent variable is number of abatements granted in each period.
indicates a significant correlation at .05.**indicates a significant correlation at .01.
Tax Increment Financing
Similar to tax abatements, there are very few significant correlations between the number of different types of tax increment finance districts and residential economic health. Only Downtown Development Authorities are significantly correlated with health. The negative correlations in early years suggest that DDAs were more likely to be used in more fiscally stressed communities. Continued use is not correlated with better health, however (Table 6).
Tax Increment Financing Programs and Health.
Note. N = 223 independent variables are the numbers of each type of district.
indicates a significant correlation at .05. **indicates a significant correlation at .01.
Renaissance Zones
The number of Renaissance Zones in a community is consistently and negatively correlated with economic health. 11 Cities that were stressed early on are more likely to have zones and, (at least in a static sense), the zones continue to be negatively correlated with health at later points in time (Table 7). However, cities with improving health between 2000 and 2010 have more Renaissance Zones. It is impossible to establish with certainty the causal ordering here, but it could be the case that Renaissance Zones need time to become established and contribute to the economy of a city. The negative correlations with early economic growth, the lack of correlations from 1990 to 2000, and then later positive correlations with health change from 2000 to 2010 suggest that this may well be the case. Only limited data were available on the start dates of the Renaissance Zones so it is not possible to specifically pinpoint establishment and future effects.
Renaissance Zones and Health.
Note. N = 273 independent variable is the number of RZs in each year.
indicates a significant correlation at .05. **indicates a significant correlation at .01.
Cool Cities Initiatives
Table 8 presents the same data for the Cool Cities Initiatives; here, there are some significant correlations with residential economic health. Again the number of grants used is the independent variable measured with the exception of the dollars of Neighborhoods in Progress funds granted. Most of the Cool Cities Initiatives, by definition, were directed to more financially stressed communities and that pattern is evident in the data. Again, the programs were funded from 2004 to 2008. Thus cities that were more economically stressed in 1980 and 1990 received more Cool Cities support. The negative relationship between Cool Cities Initiatives and health did not go away, however, as cities moved into the 2000s.
Cool Cities Initiatives and Health.
Note. N = 273 independent variables are number of grants awarded with the exception of NIP US$.
indicates a significant correlation at .05. **indicates a significant correlation at .01.
Both the Neighborhoods in Progress and Blueprints for Michigan Downtowns cities appear to have experienced residential health improvements between 2000 and 2010. These two programs are significantly and positively related to health improvements in this time period. Again, there are too few cities receiving grants to posit a conclusion that the grants were successful or caused the economic improvement. But, unlike tax abatements, a significant positive correlation is present for these two programs (and for Neighborhoods in Progress funding as well, which simply mirrors the award of an NIP grant since all cities received the same funding amount).
Michigan Economic Growth Authority (MEGA)
There are no significant relationships between the number of MEGAs granted and residential health or health change over any period of time (Table 9). Up until the 2004-2007 period, insufficient numbers of MEGAs had been granted to be able to run correlations. 12 It seems clear from this analysis that MEGAs alone have no relationship to the well-being of city residents. This is not to say, however, that when used in combination with other economic development incentives MEGAs were ineffective. However, other research on the MEGA program also found no evidence of positive outcomes other than a temporary increase in construction employment (Hicks & LaFaive, 2011).
MEGA and Health. a
Note. N = 273.
1995-2003 numbers are insufficient to calculate correlations; the lack of correlations may be due in part to the few MEGAs granted. Independent variable is the number of MEGAs granted in each time period.
indicates a significant correlation at .05. **indicates a significant correlation at .01.
By way of summary, Table 10 presents the relationship between the total number of different types of economic development tools used and economic health. Total effort is a sum of the number of individual incentive tools used at any time over the full course of the data included in this study; what is measured is aggregate use over time. Generally the data suggest that communities with poorer economic health employ more development tools overall. The lack of a significant correlation between cumulative policy use and economic change between 2000 and 2010 implies that greater effort is not related to change in residential economic health.
Total Economic Development Tool Use and Health.
Note. N = 287 independent variable is total number of tools used over time.
indicates a significant correlation at .05. **indicates a significant correlation at .01.
Government Spending
This part of the analysis is based on per capita spending. There are only three categories of spending with significant correlations with health; general government, public safety, and parks and recreation (see Tables 11-13). 13 There are no significant correlations between spending on economic and community development and health.
General Government Spending and Health.
Note. N = 273 independent variable is per capita spending.
indicates a significant correlation at .05. **indicates a significant correlation at .01.
Public Safety Spending and Health.
Note. N = 273 independent variable is per capita spending.
indicates a significant correlation at .05. **indicates a significant correlation at .01.
Recreation and Culture Spending and Health.
Note. N = 273 independent variable is per capita spending.
indicates a significant correlation at .05. **indicates a significant correlation at .01.
As noted earlier, general government spending includes expenditures for the essential functions of running the city as well as for upkeep of public facilities such as government buildings and grounds. Table 11 presents the correlations between general government spending and residential economic health. Two patterns are clear. First, past and current spending on general government is significantly and positively related to future economic health. Second, cities that were improving in residential health between 1990 and 2000 spent more on general government in later years. This suggests a pattern where economically healthy cities have been able to spend more on civic infrastructure and this spending potentially fostered further economic health.
Spending on public safely appears quite strongly related to residential economic health (Table 12). Again, two patterns are evident. Cities that have prospered in the past are able to spend more money on public safety; past and current spending on public safety is positively and significantly correlated with future economic health.
Finally, the data in Table 13 indicate significant and positive correlations between spending on recreation and culture and residential economic health although the relationships are not as consistent as those for public safety and general government. Cities that were healthier in 1990 spent more in all future time frames. However, health in 2000 does not appear to have the same effect. Spending on recreation and culture in 2005, 2006 and 2009 is significantly correlated to economic health in 2010. The correlations for 2007 and 2010 spending are close to significance with health in 2010.
While it cannot be concluded from these data that spending on these three services causes economic health, past spending in these areas is correlated with future health. Absent any significant correlations between public works, health and welfare, and community and economic development and residential health, it seems safe to conclude that spending in these areas will not measurably contribute to improving the economic health of residents. Past economic health appears to allow cities to spend more on the three critical categories. While in one sense this suggests a “rich get richer” dynamic, it also indicates that greater local spending in these areas may well be a good investment strategy for local government.
Comparing the Relative Effects of Development Tools
To explore the relative effects of the various economic development tools, regression analyses were run with 2010 health index scores as the dependent variable and all of the development tools and spending categories that were significantly correlated with 2010 health as independent variables. These include the following: Neighborhoods in Progress (part of the Cool Cities Initiatives); Brownfield Redevelopment Tax Increment Finance Authorities, Downtown Development Authorities, Renaissance Zones, and per capita spending for general government services, public safety, and recreation and culture. 14 Spending for the three categories of services is combined into an index because of significant multicollinearity. Examination of correlations among independent variables indicated a strong relationship between Neighborhoods in Progress and total Renaissance Zones (.77), thus these were also combined into an index based on factor scores. The full equation is described below.
where:
Economic health = f-score on index of residential economic health
Spending = f-score on index of total spending for public safety, recreation, and general government
NIP/RZ = f-score on an index of total Neighborhood in Progress grants and RZs
BFRA = total number of brownfield redevelopment authorities
DDA = total number of downtown development authorities
Table 14 provides the results of the regression analysis including all of these variables. Of the variables included in the model (based on their bivariate correlations), all but one remain significantly correlated with health in multiple regression (Brownfield Redevelopment Authorities). A reduced model is presented in Table 15 excluding this variable. The use of Downtown Development Authorities, Neighborhood in Progress grants, and Renaissance Zones are negatively correlated with 2010 residential health. Spending on services is the best (and only positive) predictor of health. It should be noted that the R2 values for both equations are quite low. The reduced model accounts for only 10% of the variation in economic health. Clearly there are factors, either development policies or, more likely, other characteristics of individual cities that are necessary for a complete accounting of residential economic health.
Regression Results: Economic Development Policy and Health.
Note. R2 = .10 Max VIF = 1.04. a
BFRAs were correlated with DDAs and RZs at .17 and .14 respectively and spending and RZs are correlated at –.17. These are below the levels that raise concerns about multicollinearity.
Reduced Model.
Note. R2 = .10 Max VIF = 1.04.
These findings beg the question: If traditional and widely used economic development tools do not account for variations in residential economic health among cities in Michigan, what does? While the focus of this analysis is not to explain economic health, it is useful to consider this question to place economic development activities in the larger context of what forces are related to residential economic health. Extant literature has suggested that two factors, supported by local service spending, are related to economic health: crime and spending for, and quality of, local public schools (see Reese, 2012, and Reese & Ye, 2011, for a more detailed discussion of this literature). Thus crime rates, per student spending on instruction and support services, and graduation rates were introduced into the model to gain insights into the role of basic local services in health.
Table 16 provides the best fitting results of a regression analysis including public service spending, crime rates, mean graduation rates from 1998 to 2008, and per student spending on instruction and support services from the 2009/2010 school year. 15 Correlations indicated that service spending and graduation rates were significantly correlated (positively) as were spending and crime (positively), and crime and graduation rates (negatively). Based on factor analysis, these variables were combined into an index. 16 The full equation is described below:
where
Economic health = f-score on index of residential economic health
Spending/crime/graduation = f-score on index of total spending for public safety, recreation, and general government, crime rates, and graduation rates
Instructional support per student = dollars spent per student.
Best Fitting Model for Economic Health 2010.
Note. R2 = .35 Max VIF = 1.00.
Adding education and crime data increased the R2 to .35. Cities are healthier if they spend more on general government services, recreation and culture, and public safety, have lower crime rates, have higher graduation rates from the local public schools, and spend more on instruction and support services in the schools. This further reinforces the sense that the most effective use of local government revenues may well be in spending for basic public services, including education, with goals being reductions in crime and higher graduation rates.
Conclusions and Recommendations
Conclusions
By way of conclusion, it is useful to revisit the primary contribution of this research noted at the outset of the article. Specifically the research considers economic development tools in comparison to spending for other public sector functions in relation to the economic health of local residents—a comparison that includes consideration of creative class policies and more traditional supply side approaches. Given the complexity of the forgoing analysis, it is useful to provide an “executive summary” of the findings as a way of moving the discussion to logical policy recommendations (Table 17). The summary below is organized by the various economic development tools and government spending explored here and is followed by revisiting the research contributions.
Tool Summary.
Tax Abatements: There are no significant relationships between residential economic health and the use of tax abatements, regardless of which factor is considered the independent variable. In other words, consistent with extant research on Michigan municipalities (Sands & Reese, 2012), there is no relationship between the economic health of a city’s residents and its use of tax abatements; there is no relationship between past use of tax abatements and future economic health or changes in health. In short, there are no statistically significant relationships between abatements and economic health as measured in this analysis.
Tax Increment Finance Authorities: In most cases, there are no significant relationships between the use of tax increment financing authorities and residential economic health. Poorer cities exhibit greater use of Downtown Development Authorities; however, use of DDAs does not appear to have eased the economic stress in these cities. For the other tax increment financing programs, there is no relationship between program use and health.
Renaissance Zones: Cities with poorer economic health have used Renaissance Zones to a greater extent and this is related to improved economic health over time.
Cool Cities Grants: There are several significant relationships between the use of Cool Cities grants and resident economic health. First, more economically stressed cities received more Neighborhoods in Progress and Blueprints for Downtowns grants. And, the use of these grants is significantly correlated with improvements in economic health over time.
MEGA: there are no relationships between the use of MEGA and economic health.
Government Spending: There are a number of significant and positive relationships between government spending and residential economic health. Spending for public safety, recreation and culture, and general government services is positively correlated with economic health. Additionally, investment in instruction and support services in local public schools also shows a positive relationship to economic health in multiple regression.
There are considerable literatures on the effectiveness of tax abatements, tax increment finance districts, and special development zones. As suggested by the literature review, there are mixed findings regarding all of these tools. The findings here regarding Renaissance Zones are in line with extant research on Enterprise Zones suggesting that any growth is likely not significant or commensurate with the inherent level of tax expenditures. More specific to Michigan, the data here suggest little short range but potential longer range benefits of Renaissance Zones in line with early assessments of the program (Sands, 2002). The findings that tax abatements are not correlated with improvements in residential economic health are also in line with much previous research (Fisher & Peters, 1998; Peters & Fisher, 2004; Sands & Reese, 2012; Wassmer & Anderson, 2001).
These findings beg the question of why local officials continue to embrace these programs given a significant body of research and evaluation that suggests that they are largely ineffective. While this question is beyond the focus of the present study, scholars have pointed to a number of dynamics that lead to this phenomena including (see Wolman, 1996, for a review of this literature)
The path dependent nature of economic development incentives, that is, once they are offered, they come to be expected, and so continue into the future even as new policies are added (Anderson & Wassmer, 1995; Byrnes, Marvel, & Sridhar, 1999; Reese, 2006)
The corporate surplus dynamic whereby local officials offer inducements beyond what businesses really need to locate because they know they are in competition with other communities but do not know either their offers or what businesses really want (Jones & Bachelor, 1984)
A sense among local officials that voters will punish them at the polls if incentives are not offered and jobs are lost (Feiock & Clingermayer, 1986; Pagano & Bowman, 1995; Schneider & Teske, 1993)
Lower levels of professionalism of and resources devoted to the economic development function (Fleischmann, Green, & Kwong, 1992; Pelissero & Fasenfest, 1989; Reese & Rosenfeld, 2002; Rubin & Rubin, 1987)
The inclusion of creative class policies and government spending on public services in relation to traditional supply-side development tools adds to this literature and again raises the question of opportunity costs. As the only demand-side policy considered here, it appears that the cool cities initiatives have a more positive relationship with residential economic health than the supply-side tools considered.
If local officials were to reduce the tax expenditures implicit in tax abatements and special development zones and districts, what might be done with that revenue? The research here indicates that spending on local public services related to public safety, public schools, recreation and culture, and general government services is significantly and positively correlated with residential economic health. In other words, such expenditures appear to have a more positive relationship to the quality of lives for local residents than traditional supply-side economic development tools
Policy Recommendations
The factors most consistently and positively related to economic health are investments in the downtown (via the Cool Cities Initiatives but not Downtown Development Authorities) and spending on basic local public services. These findings suggest one primary policy recommendation: the wisest course of action for most cities would be to eschew particularized development tools, especially those that require tax expenditures. Instead, they should husband local revenues to be able to better support basic services to residents. Using municipal revenues to provide high quality local services, particularly in the areas of public safety, education, recreation, and the arts appears to be the most promising economic development strategy. Of course, as noted previously, there are a number of political and structural reasons that these common development tools continue to be used.
Documenting that spending on services has a more consistent and positive relationship to the economic health of local residents is not likely to be enough to change local practice. Again, there are limitations to this study, most importantly, the lack of controls to determine time ordering and to rule out other local forces that logically effect residential health. It is possible, however, that the problem with incentives lies not in the use of them but in their application. In other words, if incentives were used differently, greater effectiveness might result. Indeed, research on Michigan’s tax abatement and Renaissance Zone programs, as well as studies of tax increment financing authorities, have highlighted how such incentives can be used to greater effect (see Sands, 2003; Sands & Reese, 2012; Weber, 2003).
At root, changes in state enabling legislation for these economic development tools would be necessary to lessen their use among cities. Yet, wholesale elimination of economic development tools at the state level is also not likely. Michigan state government has eliminated the Cool Cities Initiatives (probably a mistake) and the MEGA program (probably a good move) so there has been recent willingness to change the enabling framework at the state level. Short of elimination, there are other actions, supported by extant research, that could be taken to limit the use of supply-side tools, potentially lessening local tax expenditures and increasing the resources available to provide services. Generally, potentially beneficial revisions in enabling legislation would focus on targeting abatements and zones to areas of greatest need, targeting incentives to a more limited range of firms, and improving monitoring so that the tools can be better evaluated and distributed (Peters & Fisher, 2004; Sands, 2003; Sands & Reese, 2012; Weber, 2003). More specifically, state legislation should
Limit use of abatements, tax increment finance districts, and Renaissance Zones based on need of the local unit or area, type of investment proposed, the likelihood of verifiable new jobs, and industry of the firm.
Incorporate limits on the length of time periods for use (including sunset legislation), the number of abatements or subsidies received by the same firms, and on the number of the same jobs supported by abatements, which have a particularly high recidivism rate.
Link benefits to performance whereby tax relief is in proportion to the achievement of specific targets involving jobs and investment.
Include evaluation of requests based on necessity and consistency with local economic objectives, and
Include monitoring of results including mandatory reporting of investment and job creation.
In summary, these findings from Michigan cities suggest that economic development tools in the form of tax abatements, tax increment financing arrangements, and the most extreme tax remission, Renaissance Zones, do little to change local residential economic fortunes either for better or worse, at least as typically implemented. While it is tempting to suggest that these types of tools should be “disenabled” at the state level, it is unlikely that this would be a politically feasible solution, given their widespread use and long history. But, it is just as unreasonable to expect that local governments will curtail their use voluntarily even in the face of negative evaluations. Again, “if all you have is a hammer, then every problem looks like a nail.” Unless limitations are built into state enabling legislation, then municipalities will continue to use these hammers because they are what are most readily available.
A broader understanding of the process and goals of economic development and greater appreciation for the limitations of particularized development tools may foster an environment where local officials look to other ways of promoting fiscal prosperity. Recognizing investment in local services, including public schools, as a potentially effective economic development tool is a critical first step. Making clear the trade-offs between tax expenditures and the ability to provide high quality local services is another.
Footnotes
Acknowledgements
Thanks are due to the following individuals for providing data, analysis support, database assistance, and general research assistance: Michael Crawford, Jessica Faist-Whitt, Kathryn Huttenga, Megan Johnson, Dr. Heather Khan, Dr. Josh Sapotichne, Dr. Mark Skidmore, Eric Walcott, and Minting Ye. The input of Dr. Gary Sands was also critical to the project.
Declaration of Conflicting Interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author disclosed receipt of the following financial support for the research, authorship, and/or publication of this article:This research has been supported by grants from The Institute for Public Policy and Social Research (IPPSR) at Michigan State University.