Abstract
This empirical research focuses on three major local taxes—property tax, sales tax, and telecommunications tax—to examine their impacts on local economic development in the six-county Chicago metropolitan area. The statistical results indicate that these three major local taxes have significant negative effects on business employment. The study implies that it may be counterproductive for local governments to raise tax levels in order to address some immediate revenue shortfalls in the aftermath of the recent economic recession.
Introduction
In the aftermath of the recent economic recession, state and local governments are and will be confronted with a difficult challenge: a record-high unemployment rate that requires more government spending and a substantial decline in revenues as a result of the shrinking economy. The reduction in taxable economic activities may drive states and localities to increase tax rates in order to support necessary expenditures. However, the rise in government taxes will likely have a negative impact on economic development and may slow down the pace of economic recovery.
One important policy question is to what extent higher tax levels may stymie or even reduce business development if government tax policy does have a significant impact on economic development. The basic economics answer is obvious: Government taxes do matter, because taxes are cost factors that reduce business profitability. The empirical evidence, however, has been mixed with regard to the effects of government taxes on economic development. Some scholars do not find a statistically significant effect of government taxes on economic development. 1 Although many empirical studies report significant results, they differ in the magnitude of the effect (Bartik, 1991; Wasylenko, 1997).
The current economic challenge and the mixed empirical evidence call for more empirical research in this field. This study focuses on three major local taxes―property tax, sales tax, and telecommunications tax—to examine their impacts on local economic development in the six-county Chicago metropolitan area. Municipal telecommunications tax is new in the literature of economic development. The state of Illinois simplified its fairly complex system of taxes and fees on telecommunications in 2003, providing an opportunity to incorporate this new local-option tax in the research. This also explains why 2003 to 2008 is selected as the period of study.
This research also provides some additional empirical evidence with regard to the debate of Cook County’s one-percentage increment in sales tax enacted in 2008. In the city of Chicago, the county tax, plus city and state taxes, mean consumers paid a total sales tax of 10.25%—the highest of any major city in the country after the hike of the county sales tax rate in 2008. The Cook County Board of Commissioners has voted to phase out the countywide increment of one-percentage sales tax by 2013. However, the repeal was the result of a fierce fight between county leaders and would not have happened had the state government not intervened in the end. 2 It is important to understand how that much of a change in local sales tax rates may affect the local business community. The statistical results indicate that these three major local taxes likely have significant negative effects on business employment. The findings of this study lend some support to the repeal of the Cook County sales tax increase enacted in 2008.
This article is organized as follows: The second section presents a review of empirical studies in the field. The third section describes demographics and employment growth of the region as well as some selected municipalities. The next two sections discuss modeling and data issues relevant to this research and present and discuss the statistical results, followed by a concluding section.
Empirical Studies on Impacts of Tax Rates on Economic Development
In state and local public sectors, economic development is generally defined as “changes that affect a local economy’s capacity to create wealth for local residents” (Bartik, 1995). A variety of economic indicators have been used as benchmarks in state and local economic development programs, including (but not limited to) new plant openings, new branch plants, employment and population growth, foreign direct investment, changes in personal income, and so on. It is noteworthy that, for most development programs, business employment expansion remains the top priority for policymakers.
An effective economic development program requires a thorough understanding of the factors that may affect location and expansion of businesses. Economic logic suggests that any factors that alter business profitability are likely to affect location of business activities (establishment, employment, etc.). In the literature, two broad types of factors are identified. The first type are nonfiscal factors, including market factors such as market demand, energy prices, labor quality and cost, and other location-specific factors such as climate, weather, and natural resources. The second type are fiscal factors, including government taxes on businesses. While the second type is under the control of governments, the first type is primarily determined by the general economic trends and is outside government reach.
Although the package of policy tools in economic development has been changing over time, business tax incentives have been the primary instruments for attracting firms to locate and expand in a jurisdiction. Policymakers in state and local governments have had a sustained interest in the role that taxation plays in economic development. The empirical evidence for whether state and local taxes contribute to interjurisdictional reallocation of economic activity, however, has remained mixed. Bartik (1991) suggested that the average elasticity is −0.3 for the tax responsiveness of location and economic growth for states or regions and that the range of the elasticity estimates is between −0.1 and −0.6. In a later review, Wasylenko reported that a large share of the elasticity estimates indicate responsiveness less than −0.3, and the variance of elasticity estimates is much larger than the previous studies (Wasylenko, 1997).
The literature also suggests that the impact of tax differentials is generally larger in intraregional than interregional empirical studies because many other factors relevant to economic development can be better controlled in an intraregional than an interregional setting. Bartik (1991) noted that, for the intraregional elasticity of economic activity with respect to taxes, −2.0 should be a reasonable estimate. However, elasticity estimates differ substantially; some of them are even statistically or economically insignificant.
Some later empirical studies also provide evidence on the effects of local taxes on intraregional economic growth. Luce (1994) reported that local property and wage taxes have significant and negative effects on job location in the Philadelphia metropolitan area, with elasticity estimates of −0.371 and −0.204, respectively. One empirical study of the D.C. metropolitan area examined the effects of local taxes, public spending, and quality-of-life variables on population and employment growth (Mark, McGuire, & Papke, 2000). Using a panel of nine areas of Maryland, Virginia, and D.C. during the period 1969 to 1994, they reported that higher rates of two business taxes—sales and personal property—reduce annual employment growth by a significant amount. They did not, however, find a significant effect of local property tax on either residential or business location choice. One year later, Dye, McGuire, and Merriman (2001) published a study that examines the effect of local property taxes on economic activity in the Chicago metropolitan area. The findings suggest that a high effective property tax rate is a deterrent to business location decisions and economic activity.
To date, the literature has been far from reaching an empirical consensus about the role of taxation in economic development. In addition to some methodological issues, the inconsistent estimates likely result from the heterogeneity with regard to how tax policy works under different circumstances. This research is intended to contribute to the body of literature in two aspects. As an update on the impact of local taxes on economic development, the results provide new evidence to inform local government officials of the negative effects of government tax hikes that are likely to be pursued as a means of coping with revenue shortfalls. Second, local telecommunications tax has never been addressed in the literature. This is probably the first empirical examination of whether local telecommunications tax has a significant negative impact on business development, especially in sectors heavily dependent on telecommunication services, as claimed in some industry reports (Cline & Phillips, 2005).
Northeastern Illinois Municipalities: Demographics and Private Employment
This study focuses on the northeastern Illinois area that includes six counties: Cook, Du Page, Kane, Lake, McHenry, and Will. This six-county area covers the majority of the Chicago-Naperville-Joliet Metropolitan Division. 3 They are six of the seven largest counties in the state of Illinois and also the six fastest-growing counties by net change of population from 1990 to 2000. 4 As a whole, this area has about 65% of the state of Illinois’ population. There is also a fairly high concentration of business presence in this area. In 2008, it was the location of about 59% of all private employment in Illinois. Its shares of employment in Manufacturing (with two-digit North American Industry Classification System [NAICS] Codes 31-33), and Wholesale and Retail Trade, Transportation and Warehousing (with two-digit NAICS Codes 42-49) are 62% and 64%, respectively. The share of employment in other service sectors is also more than 60%, except Education Services. 5
The economic landscape and industrial structure have undergone some significant changes in recent decades. The goods-producing industries, especially the manufacturing industries, have been declining, and the service-providing industries have been on an expansionary path. The six-county area is no exception to this general economic trend. The period 2003 to 2008 saw negative growth in the four goods-producing sectors. In particular, the number of employees in the manufacturing sector declined by about 10% in the five-year period. Meanwhile, most of the service-providing sectors grew. The growth rates of employment are about 13% and 10% for the Accommodation and Food Services sector and the Health Care and Social Assistance, respectively.
The overall trend in business development may disguise substantial variation at the local level that may result from different policy choices important to business location and expansion. Table 1 presents a list of the nine biggest winners and losers in terms of percent change in private employment during the period 2003 to 2008. The gap is obvious and substantial: The nine winners increased their number of private employees by more than 33%, whereas the same economic indicator declined by more than 13% in the nine municipalities that are at the bottom of the list.
Big Winners and Losers in Employment Growth, 2003-2008
There is a long list of factors that may lead to varied development in business employment across municipalities. This research is intended to examine the tax factors that may have affected employment growth of major municipalities in this six-county area from 2003 to 2008. Like business employment, tax levels on businesses also differ significantly by locality. With the assumption that government taxes add business costs and, hence, negatively affect their profitability, I expect that businesses are likely to locate their new establishments and expand the existing ones in localities that have relatively low levels of business taxes. On the other hand, business development in high-tax jurisdictions is likely to stagnate or even decline.
Three major local tax rates for the 18 municipalities are presented in Table 1. In Illinois, local governments are authorized to collect property, sales, and telecommunications taxes. Although local governments are generally restricted from levying these local-option taxes beyond certain caps, they have substantial flexibility in deciding whether to levy any of the taxes, and tax rates are subject to state-imposed caps.
Preliminary analysis lends some support to my hypothesis that local tax rates may affect private employment in a way that is consistent with basic economic theory. The mean values of aggregate local property tax rate, aggregate local sales tax rate, and municipal telecommunications tax rate of the biggest winners are significantly lower than those of the biggest losers. This indicates that tax levels may play an important role in intrajurisdictional variation in economic development.
Model and Data
Similar to other empirical studies in this field, a log-linear variation of the model proposed by Steinnes and Fisher (1974) is employed:
The variable EM is employment and EM−1 is employment with one-period lag; PT, ST, and TT represent local property tax, local sales tax, and local telecommunications tax; PP and IN refer to population and per capita income; and ϵ is the error term, which is assumed to be normally distributed. All variables in the model are in logarithm.
The measure of employment is the number of employees in all industries or by major industry sectors. The data source is Where Workers Work, a series of annual publications by the Illinois Department of Employment Security. Based on administrative data collected under the Illinois Unemployment Insurance Act, Where Workers Work reports basic private-sector employment totals for six northeastern Illinois counties, as well as major communities in each of these counties, including the City of Chicago. 6 The employment data are broken down into major industry sectors indentified by a two-digit NAICS code.
Although the Illinois Unemployment Insurance Act does not require employers to report their employment by work location, the data do reflect actual work locations to the extent possible since 1991. Because the time frame is 2003 to 2008, the employment data used for this study reflect employment by the place of work. In Steinnes and Fisher’s (1974) model, employment by place of residence is one of the independent variables. This variable is not included because there are no data at the local level for employment by place of residence in the period of study. The exclusion of employment by residence is probably not a major concern for two reasons. First, the place of work may not be closely related to the place of residence because northeastern Illinois has a large number of local jurisdictions. It is easy and common for workers to commute fairly long distances from their home communities to work. Second, employment by place of residence is highly correlated with municipal population, at least for this specific area and period. 7 Given that municipal population is one of the independent variables, there is no need to include employment by residence as another independent variable.
The model regresses the municipality’s employment in each year after 2003 to 2008 on employment in 2003, local property tax in 2003, local sales tax in 2003, municipal telecommunications tax in 2003, and other relevant control variables. As in other states, municipal governments, county governments, and school districts in Illinois levy property taxes. In addition, a variety of other special districts also collect property tax to finance the delivery of special services such as fire protection, parks, sanitation, and libraries. Because the property tax burden encompasses property taxes levied by all local governments and authorities, the aggregate local property tax rate that is the sum of property tax rates levied on taxpayers is used. Considering that the statutory property tax rates may not be comparable across municipalities because of varied assessment ratios, effective tax rate instead of statutory tax for the local property tax variable in the model is. The effective tax rate is the statutory tax rate multiplied by the countywide median assessment ratio. 8
The state of Illinois allows its home-rule municipalities to levy a local-option sales tax and to adjust their sales tax rates within limits. Non-home-rule municipalities may also collect local sales taxes but they are restricted in adjusting the tax rate. Like property tax, sales tax is collected by multiple local governments, including counties, municipalities, and other public authorities such as the regional transportation authority. The measure for the local sales tax in the model is the aggregate sales tax rate of county government, municipal governments, and other regional public authorities.
Illinois enacted the Simplified Municipal Telecommu-nications Tax Act, which took effect on January 1, 2003. The act replaced three municipal-level taxes and fees on telecommunications services with a single municipal-option tax called the simplified municipal telecommunications tax. 9 Municipalities are authorized to levy a single tax rate up to 6% on the consumption of telecommunications services within their jurisdictions. The city of Chicago is allowed to levy a rate of 7%. Because only municipal governments are authorized to levy this tax, municipal telecommunications tax rate is used as the measure for this variable.
In addition to the three tax variables, the model includes municipal population and per capita income as two control variables. Municipalities with a larger population are likely associated with more private employment because there is more local demand for consumption and a larger available workforce. Per capita income, as a measure of a community’s wealth, may help control for some relevant factors such as quality of life, local schools, and so on, because a wealthier locality is likely to have better schools and higher quality of life. In addition, wealthier communities are more attractive to businesses because they tend to have stronger consumer demand for private commodities and services.
The data on three tax rate variables are from the Illinois Department of Revenue (IDOR). Local sales tax rates and municipal property tax rates are derived from the machine readable files on the IDOR’s website. The data on statutory property tax rates and assessment ratios are from tax statistics published by IDOR. The data of population and per capita income are obtained from the Census 2000 summary file 3. Table 2 presents variable definition and data source.
Variable Definition and Data Source
Due to the availability of data, this study focuses on municipalities with populations more than 10,000 in the six-county northeastern Illinois area. Because some of these major municipalities are primarily residential, without substantial business presence they are not separately listed in Where Workers Work. Therefore, employment data for some of the major municipalities in this area are unavailable. According to Census 2000, 139 municipalities in this area had a population of more than 10,000. Employment data for all industries for 112 municipalities is used, resulting in a sample that includes slightly more than 80% of all major municipalities in northeastern Illinois. The samples of individual sectors may be smaller because some sector-specific employment data are withheld because of confidentiality concerns.
Some descriptive statistics are provided in Table 3. The descriptive statistics show a substantial variation in private employment across municipalities. The range of employment data is quite large for all industries and individual sectors. This provides a good opportunity to explore some policy factors that may explain such a vast variation. The descriptive statistics also show significant variances in the three tax-rate variables. For instance, local property tax rates range from $1.10 to $4.61 per $1,000 of property value. The maximum local sales tax rate is 2.75%, 11 times the minimum tax rate 0.25%. The municipal telecommunication tax rate ranges from 1% to 7%.
Descriptive Statistics
Results and Discussion
The model is estimated using the ordinary least squares (OLS) procedure with a robust estimate of variance. 10 The option of robust estimate of variance allows the possible correlation between some explanatory variables and the error term and can produce a consistent estimate of standard errors, even if the observations are not independent. The model is initially run on a sample of 112 observations for employment in all industries. The OLS procedure is implemented with employment each year after 2003 as the dependent variable. The statistical results are reported in Table 4.
Statistical Results—All Industries
Note. The dependent and independent variables are all in logarithms. The model is ordinary least squares. Robust standard errors are in parentheses.
Significance level <1%. **Significance level <5%. *Significance level <10%.
The study of municipalities located within a metropolitan area is basically an intraregional empirical investigation in which many unspecified factors are constant or at least less variable than interregional studies. The hypothesis is that the differences in major local tax rates are important explanatory factors for the variation in employment across municipalities. The statistical results confirm expectations. The values of R2 are quite large in the estimations of all years, suggesting that this simple econometric model explains most of the cross-sectional variation in employment. As expected, lagged employment plays a very significant role in explaining the variation in employment one or more years later. As for magnitude, the estimated coefficients of employment with different lagging periods range from 0.93 to 0.96. It is noteworthy that the impact of employment in 2003 declines slightly as the dependent variable changes from employment in 2004 (one year after 2003) through employment in 2008 (five years after 2003).
All three local taxes show a statistically significant and negative effect on employment after a short immediate period. Starting in 2005, the 2003 local sales tax rate presents a significant effect on employment, with the estimated coefficient rising from −0.0288 in 2005 to −0.0613 in 2008. This indicates that the local sales tax rate does affect employment, and its impact is accumulative over time. A similar pattern is found in local property tax and municipal telecommunications tax. The local property tax and municipal telecommunications tax in 2003 start to show a statistically significant effect on employment in 2006 and 2007, respectively. It is interesting that local businesses respond to different local taxes at varied paces―their response to local sales tax is the most prompt, whereas it takes one to two more years to respond to the variation in local property or municipal telecommunications taxes. The results also indicate that none of the taxes has an immediate impact on total private employment in the following year.
Long-term elasticity of employment to the variation in local tax rates is calculated by dividing the estimated coefficients of individual tax rate variables by λ that is derived from the coefficient of the lagged employment variable. The results for employment in 2007 and 2008 are only used because they are relatively long term in this study. The value of λ is 0.0575 for 2007 or 0.0741 for 2008. The elasticity of local property tax, local sales tax, and municipal telecommunications tax is −2.21, −0.79, and −0.42, respectively. 11 The estimated elasticity is in line with other empirical studies in the field. The elasticity of employment to local property tax is very close to the elasticity estimate of −2.0 Bartik believed reasonable for the intraregional studies. The elasticity to local sales tax or municipal telecommunications tax is lower than what Bartik summarized. This is probably because sales and telecommunication taxes cost relatively less than property tax for the business sector as a whole.
Economic significance matters more to policymakers than statistical significance. The statistical results suggest that a 1% lower local property tax rate may result in about 2.2% higher employment in the long term. Given that mean employment in 2003 was 25,430, a 1% decrease in the local property tax rate could have created about 560 more private jobs, everything else being equal. In addition, the study estimates that a 10% lower local sales tax rate (municipal telecommunications tax rate) may result in about 7.9% (4.2%) higher employment in the long term, equivalent to about 2,009 (1,068) industry jobs. It seems that the jobs created (lost) due to rise (decline) of major government tax rates are significant.
The model on employment of individual major sectors is also estimated. Tables 5 and 6 present statistical results for eight sectors―Wholesale Trade; Retail Trade; Finance and Insurance; Real Estate and Rental and Leasing; Professional, Scientific and Technical Services; Health Care and Social Assistance; Accommodation and Food Services; and Other Services. 12 The results show varied responses to the three government taxes in different sectors. For instance, local property tax shows a significant effect on employment in all sectors except Wholesale Trade and Retail Trade. 13 The estimate of local sales tax rate is statistically significant in all the sectors other than Finance and Insurance, and Real Estate and Rental and Leasing sectors. 14 The estimate of municipal telecommunications tax is only significant in sectors of Accommodation and Food Services, and Other Services. In general, the sector-specific results make sense. For instance, the Wholesale Trade and Retail Trade sectors are sensitive to local sales tax but not to local property or telecommunications taxes. Local property and sales taxes have a much broader impact than telecommunications tax, given that the former two taxes are more visible and more broadly based than the latter one.
Statistical Results—Selected Sectors (1)
Note. The dependent and independent variables are all in logarithms. The model is ordinary least squares. Robust standard errors are in parentheses.
Significance level <1%. **Significance level <5%. *Significance level <10%.
Statistical Results—Selected Sectors (2)
Note. The dependent and independent variables are all in logarithms. The model is ordinary least squares. Robust standard errors are in parentheses.
Significance level <1%. **Significance level <5%. *Significance level <10%.
In addition to government taxes, municipal population and per capita income have some limited effects on private employment. The estimate of population is statistically significant in regressions of all industries, as well as five of the eight sectors. The estimates are positive, indicating that businesses tend to locate in larger municipalities. The estimated coefficient of per capita income is positive but only significant in two sectors―Finance and Insurance, and Real Estate and Rental and Leasing. This suggests that industries highly dependent on residents’ wealth are likely to locate their establishments in communities with higher incomes.
Conclusion
This empirical study is motivated by a nationwide dilemma between raising taxes to support necessary government spending and stimulating employment growth through lowering taxes on businesses, as well as a local debate on Cook County’s one percentage point sales tax hike. Updated empirical evidence is needed to address the effects of government taxes on economic development to better inform local governments when they consider any change in local-option taxes.
The substantial variation of employment across municipalities and significant differences in the level of three major local taxes in the six-county northeastern Illinois area provide an opportunity to implement this cross-sectional empirical research. The results suggest that lower local property tax, local sales tax, and municipal telecommunications tax may have played a significant role (both statistically and economically) in the employment growth of all industries. Moreover, local property tax and sales tax have a broader impact than municipal telecommunications tax because they present a statistically significant and negative effect on employment in more major industry sectors. It corroborates findings from many other empirical studies and surveys of businesses that government taxes are still a significant cost factor to business sectors when they decide where to locate or expand their establishments in a metropolitan area.
The policy implication for local governments is that reductions in major local-option taxes could be employed as an effective policy instrument in the growth of private employment that is critical to the economic recovery and prosperity in the aftermath of the recent economic recession. It also implies that it may be counterproductive for local governments to raise tax levels in order to address some immediate fiscal needs.
Based on the estimated effects of local taxes on private employment, Cook County’s one percentage point sales tax increase would have a quite significant long-term effect on local economic development if it were not repealed. A tax rate hike of such a magnitude would likely lead to a significant loss of private employment in the long-run within the county and the municipalities in the county, and the additional sales tax revenues may not offset such an economic cost.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.