This study examines the effects of sales taxes on food in Kansas counties from 2012 to 2013. Sales taxes that include sales of food for at-home consumption may be more stable due to their low price elasticity of demand. However, border effects caused by tax rate differentials across jurisdictions may create significant revenue losses for local governments. This study improves the reliability of existing empirical results by calculating an index of the relative tax burden of a county and all surrounding counties. Unlike existing studies where dummy variables are used to indicate whether the adjacent county has lower tax rate, we employ a measure of relative tax burdens as an indicator of sales tax policy differentials. Empirically, we employ a difference estimator on panel data with corrections for heteroscedasticity and autocorrelation. Our results suggest that food sales in Kansas are sensitive to tax rate differentials.
Keywords: sales taxes, state and local taxation, food taxes, tax incidence
The purpose of this study is to estimate the impact of food tax difference on food purchased to consume at home in Kansas counties during 2012 and 2013, especially in counties bordering on other states whose food taxes are lower. This study differs from existing studies in that it uses an all-dimensional tax rate difference in which a county's tax rate is compared with those of all counties it shares border with, instead of comparing the tax rate to a single adjacent county. This is because consumers can be mobile in many dimensions; they will choose to purchase food in any jurisdictions adjacent to their home county depending on after-tax price as well as taste and preferences. For Kansas data, this measurement seems to improve the results compared to those derived from the measurement strategy used in other papers. For Kansas counties, we find that after controlling for heteroscedasticity and autocorrelation, the cross border shopping effect due to tax rate differences is larger than found in previous studies. This suggests that in Kansas, the all-dimensional tax rate difference indicator can capture the effects of crossborder shopping better than the one-dimension indicator does, revealing a more pronounced effect of tax rate difference across jurisdictions.
As several states have considered cutting their income taxes, (1) sales taxes have become the focus of policymakers given that when income tax is cut, states will need to shift their reliance on revenue sources. (2) In Kansas, Governor Brownback's administration approved an income tax cut policy in 2013. The cut was across the board and income tax structure changed from four to two brackets, resulting in steady revenue decline from 2013 to 2015. The state's sales tax rate fluctuates; in 2013 it decreased from 6.3% to 6.15% and then increased to 6.5% in 2015, closing the budget gap in the fiscal year 2016.
Since 2013, the state has been struggling to find revenue sources to replace foregone income taxes; two major alternatives have been retail sales taxes (including those on food for at-home consumption--groceries) and sumptuary taxes (primarily on cigarettes). While taxes on groceries are a relatively stable revenue source since demand for food is price inelastic, state and local policymakers should be careful since sales taxes as a whole can fluctuate due to macroeconomic cycles. This situation can hamper long-run revenue growth paths (Hawkins, 2000). Additionally, since consumers can avoid food sales taxes by purchasing groceries in neighboring jurisdictions where taxes are relatively lower, food sales tax rates that are dramatically different than those of neighboring jurisdictions should be avoided. This study investigates this problem using Kansas counties as the unit of analysis.
The state's shift from reliance on income tax to reliance on sales taxes may create challenges to local governments because they tend to rely on sales tax revenue second only to property tax (American Legislative Exchange Council 2015). While local governments need to maintain their spending on public road, safety, and communities' amenities, state sales taxes can short-change local governments' capacity to raise revenue. A state's sale tax rate that is significantly different from that of its neighbors can cause local governments to lose their competitiveness in terms of economic development, which, in turn, defeats the state's reason for cutting income taxes for growth. In Georgia, one of the main obstacles for county governments to adopt local sales taxes is high state sales taxes (Zhao, 2005).
Furthermore, state sales taxes can add inefficiency to local government revenue systems, especially when grocery food is not exempted. Consumers will consider price level to decide where and how much to purchase grocery food, all else being equal. Assuming that before-tax prices for grocery food are similar between two adjacent counties and that travel cost is not high, the county that has a higher tax rate will lose grocery sales volumes and sales tax revenue to the adjacent county. This is because after-tax price for grocery food distorts the consumers' decision to purchase food in their home county. The reduced food consumption, which leads to decreased tax revenue, is an efficiency loss or 'excess burden" on society and occurs when consumers withdraw or reduce their consumption in response to higher tax prices applied to food. How much efficiency loss society suffers depends on demand elasticity to price (Fisher, 2004). The higher the demand elasticity to price, the more the consumer's response to price increases and the greater the efficiency loss for society.
Food tax exemptions may be more a result of political choice rather than policy-related choice by the state law makers. One previous study finds that the government share of the state economy is positively related to state's food tax exemption policy (Fletcher & Murray, 2006). In states like Kansas and Oklahoma, food sales taxes are not exempt from general sales taxes; local governments, especially those bordering on other states, could be more affected in terms of maintaining their revenue collection and avoiding inefficient tax systems. This is because consumers will choose to purchase their food on the side where food sales taxes are lower (e.g., Missouri) or do not apply (e.g., Colorado and Nebraska). In Kansas, of 105 counties, 30 share at least one border with other states including Nebraska, Colorado, Missouri and Oklahoma. Among these four states, Colorado and Nebraska totally exempt their food from sales taxes. Missouri passed laws to reduce food sales taxes to 1.225% in 1993. Oklahoma, like Kansas, does not exempt food from sales taxes. Thus, in general, Kansas's counties and cities that share their borders with Missouri, Colorado and Nebraska could face sales tax inefficiency, losing sales tax revenue that would otherwise be collectable. This, in turn, may force Kansas's counties and cities as well as the state itself to increase sales tax rates to generate adequate revenue, because when the tax base is narrow, tax rates have to increase. Furthermore, these counties and cities may become less competitive in terms of attracting new businesses and residents.
The remainder of this study is organized as follows. Section 2 discusses theoretical backgrounds for state and local sales taxes. Section 3 discusses Kansas food taxes and the inefficient food consumption in Kansas that might occur because of uncompetitive sales taxes. Section 4 presents the methodology, findings and discussion. Section 5 concludes.
For local governments, sales tax plays an important role in both economic development and financial management. One empirical study finds evidence that in non-metropolitan counties where the economic base is not diverse, a competitive sales tax rate is a magnet to attract new households and firms, holding constant the city's household amenities and individual firms' profits based on business type (Yu and Rickman 2012). From a financial management perspective, sale taxes, if set effectively and efficiently, can generate adequate revenue for local governments in addition to property taxes (Brunori 2006). According to economic efficiency and spatial distribution criteria, local sales taxes should not be higher than those of the neighboring jurisdictions, especially when the demands for goods and services are relatively elastic, since the tax rates will distort consumer decisions to avoid tax and hence the jurisdiction will lose sales tax revenue that would otherwise be collected (Winfrey 1998; Brunori 2006; Fisher 2007).
A number of studies document that discontinuity of tax treatment across state borders tends to create spatial distribution especially for cross-border shopping (Beard, Gant and Saba 1997; Luna, 2004), food (Tosun and Skid-more, 2007), alcoholic beverage consumption (Tosun 2003), state lottery sales volumes (Garett and Marsh 2002; Tosun and Skidmore 2004), housing establishment (Yu and Rickman 2012; Thompson and Rohlin 2012) and employment in the retail sector (Thompson and Rohlin 2012; Wong, 1996). Thus, if local governments want to have efficient and effective sales taxes that generate enough revenue, they should try to set a competitive rate--that is, a sales tax rate that is similar to those of neighbors given that tax avoidance behaviors of consumers could reduce local sales tax revenue.
Agrawal (2013) uses Nash equilibrium (3) theory to predict the behaviors described above. He postulates that according to the Nash equilibrium local governments will tend to smooth state sales tax discontinuity by setting local rates similar to those of the total sales tax rates set by their neighbors. Agrawal (2013) confirms this hypothesis by presenting evidence that, in the United States, for every 100 miles increase from a jurisdiction that has...