Dynamic Sales Tax Competition: Evidence from Panel Data at the Border 
Interest in attracting firms, the jobs they generate, and the tax revenues associated with this economic activity can foster tax competition in which local governments adjust their tax rates in response to rate changes in nearby jurisdictions (horizontal tax competition) and to changes in their own state (vertical tax competition). Thus tax competition can potentially affect not only where consumers shop and firms locate but also funding for local public services.
This paper examines both vertical and horizontal tax competition over time by studying the strategic response of county sales taxation to own-state sales taxes and to the combined state and county taxes of cross-border neighboring municipalities. Using county and state sales tax data from 2003 through 2009, the authors employ both static and dynamic panel analysis as well as an instrumental variables approach in combination with a border analysis.
Key Findings
- In the static analysis, a jurisdiction is likely to have high county sales taxes if its cross-border neighbor has high combined state and county sales taxes, whereas a jurisdiction is likely to have low local sales taxes if its own-state sales taxes are high.
- When examining tax competition over time, the authors find that a county is more likely to pass a sales tax rate increase when its cross-border neighbor passes a sales tax increase, but it is much less likely to pass a sales tax rate increase when its own state passes a sales tax increase.
- In the dynamic panel analysis, the likelihood of a county passing a sales tax rate decrease plummets when its cross-border neighbor increases its sales tax rate 1.0 percentage point, whereas the likelihood of the county passing a sales tax rate decrease rises when its own state passes a sales tax rate increase. More specifically, when a county’s own state increases the sales tax rate 1.0 percentage point, the county is 85 percent more likely to decrease its tax rate.
- In terms of the size of the sales tax increase, in the dynamic panel analysis, a county responds to a 1.0 percentage point increase in its own-state tax rate by decreasing its local rate 0.02 percentage point.
- Static and dynamic panel analyses find that horizontal tax competition has little to no effect on the size of a sales tax increase.
Implications
In the dynamic panel analysis, the authors find that vertical and horizontal tax competition may affect the decision to pass a sales tax rate change, but it is unlikely that either has any influence on the size of that change. Furthermore, the authors find evidence of vertical tax competition in the dynamic panel analysis, but the magnitudes of the implied tax changes in the strategic response are very small. This finding suggests that most of the reported correlation between taxes in neighboring jurisdictions or between levels of government within a state that has been labeled tax competition in previous cross sectional analyses is likely due to selection, not causation. The policy implications of the dynamic analysis is that tax competition may be a much less important consideration for local policymakers than what is implied by earlier cross-sectional studies.
Abstract
We examine both vertical and horizontal tax competition over time by studying the strategic response of county sales taxation to state sales taxes and to cross-border neighboring municipalities’ combined (state and county) taxes. Using county and state sales tax data from 2003 through 2009, we employ both static and dynamic panel analysis as well as an instrumental variables approach in combination with a border analysis. Our results confirm the presence of tax competition in the cross section, as previous studies have found. Results from the fixed-effects and dynamic panel analysis also indicate the presence of vertical competition, though quite small, as counties are consistently responsive to changes in their own state sales tax level across all models and specifications. However, the panel findings suggest little to no horizontal tax competition. Following Parchet (2019), we address additional concerns about endogeneity by instrumenting the neighboring-county sales tax rate with the state-level sales tax rate of the neighboring state. Results from instrumental variables analysis reinforce the presence of a small vertical tax competition between local and state sales tax policies. Interestingly, our results, like those of Parchet (2019), indicate that cross-border local sales tax rates act as strategic substitutes.