The Local Response to Tax Limitation Measures: Do Local Governments Manipulate Voters to Increase Revenues?

This paper provides evidence that some cities subject to a statewide tax limit manipulate their mix of productive and administrative services in an attempt to get voters to override the statewide limit. When a statewide limit reduces a city’s budget, one manipulative response is to cut “service” inputs (for example, teachers or uniformed police officers) by a relatively large amount, while cutting administrative inputs by a relatively small amount. This approach reveals a relatively large trade-off between public and private goods, and the severe consequences from a tax limit may encourage local voters to override the statewide limit. We provide evidence that cities with local-override options tend to adopt this approach. Manipulation is most prevalent among cities run by city managers (as opposed to strong mayors). This paper provides evidence that some cities subject to a statewide tax limit manipulate their mix of productive and administrative services in an attempt to get voters to override the statewide limit. When a statewide limit reduces a city’s budget, one manipulative response is to cut “service” inputs (for example, teachers or uniformed police officers) by a relatively large amount, while cutting administrative inputs by a relatively small amount. This approach reveals a rel* We are grateful to two anonymous referees, Andrew Austin, Bob Helsley, Bill Hoyt, John Lott, John Matsusaka, Sam Peltzman, Kim Rueben, Steve Sheffrin, seminar participants at Indiana University, Princeton University, and the Universities of British Columbia, California-Davis, Chicago, Florida, Kentucky, Oregon, and Southern California, and participants at meetings of the American Economic Association, Econometric Society, Public Choice Society, and Vencontro Nacional Emprego e Deservolvimento Regional in Coimbra, Portugal, for helpful comments and suggestions. Figlio acknowledges research support from the National Science Foundation and the U.S. Department of Education. Much of the data used in this analysis were provided either by the Inter-university Consortium on Political and Social Research or the U.S. Department of Education. All remaining errors are our own. 234 the journal of law and economics atively large trade-off between public and private goods, and the severe consequences from a tax limit may encourage local voters to override the statewide limit. We provide evidence that cities with local-override options tend to adopt this approach. Manipulation is most prevalent among cities run by city managers (as opposed to strong mayors). Previous studies have shown that voters’ predictions of the consequence of tax limits differ from the actual consequences. In anticipation of the vote on Proposition 13 in California, 38 percent of voters believed that state and local governments could absorb a 40 percent cut in tax revenue without cutting services. In both Massachusetts and Michigan, a majority of citizens believed that tax limitation measures would have no effect or a very small effect on public services. But these forecast effects are inconsistent with the actual outcomes of tax limits. In terms of actual consequence, Thomas Downes and David Figlio show that tax limits increased student-teacher ratios, decreased teacher salaries, did not affect administrative spending, and reduced test scores; Figlio and Kim Rueben show that tax limits are associated with reduced teacher quality. Maura Doyle shows that tax limits reduced the quality of municipal fire service. Downes, Richard Dye, and Therese McGuire show that tax limits in the Chicago area decreased math scores by a small amount but did not affect reading scores. Our simple theoretical model suggests that a city may manipulate the tradeoff between public and private goods to encourage voters to override a statewide limit. To test this hypothesis, we use balance-sheet data from 5,150 U.S. cities. We measure the service ratio as the ratio of spending on police and fire protection to spending on general administration. In the years following the adoption of a statewide tax limit, the service ratio decreases by a relatively large amount in cities that have a local-override option. Among cities with override options, the largest reductions in the service ratio occur in cities whose citizens have the least interjurisdictional mobility and in cities run by city managers. These results are not driven by population differences across the cities and are apparently not driven by endogeneity problems. Data from within single spending categories 1 Jack Citrin, Do People Want Something for Nothing: Public Opinion on Taxes and Government Spending, 32 Nat’l Tax J. 113 (1979). 2 See Paul Courant, Edward Gramlich, & Daniel Rubinfeld, Public Employee Market Power and the Level of Government Spending, 69 Am. Econ. Rev. 806 (1979); and Helen F. Ladd & Julie Boatright Wilson, Why Voters Support Tax Limitations: Evidence from Massachusetts’ Proposition 21⁄2, 35 Nat’l Tax J. 121 (1982). 3 Thomas A. Downes & David N. Figlio, Do Tax and Expenditure Limits Provide a Free Lunch? Evidence on the Link between Limits and Public Sector Service Quality, 52 Nat’l Tax J. 113 (1999); David N. Figlio, Short-Term Effects of a 1990s Era Property Tax Limit: Panel Evidence on Oregon’s Measure 5, 51 Nat’l Tax J. 55 (1998); David N. Figlio, Did the “Tax Revolt” Reduce School Performance? 65 J. Pub. Econ. 245 (1997); David N. Figlio & Kim Rueben, Tax Limits and the Qualifications of New Teachers, 80 J. Pub. Econ. 49 (2001). 4 Maura Doyle, Property Tax Limitations and the Delivery of Fire Protection Services (working paper, Fed. Reserve Board Governors 1994). 5 Thomas A. Downes, Richard F. Dye, & Therese J. McGuire, Do Limits Matter? Evidence of the Effects of Tax Limitations on Student Performance, 43 J. Urb. Econ. 401 (1998). tax limitation measures 235 provides corroborating evidence: school districts with override options decrease teacher-administrator ratios by larger amounts, and police departments in large override cities decrease officer-administrator ratios by larger amounts. I. A Model of Manipulative Local Government There has been some research on manipulative government and on the local government responses to tax limits. Thomas Romer and Howard Rosenthal show that if the reversion budget (the budget adopted if voters do not approve a proposed budget) is sufficiently far from the voters’ preferred budget, voters will approve a budget that exceeds their preferred budget. In the models of Kangoh Lee and Eugenia Toma and Mark Toma, bureaucratic effort is endogenous, and bureaucrats can respond to a tax limit by shirking more, thus reducing the impacts of a tax limit. Downes suggests that Proposition 13 caused some bureaucrats (who care about both output and staffing levels) to act in a manner more consistent with the preferences of the median voter (who cares about output). To motivate our model of a manipulative local government, consider the response of a city government to a budget cut caused by a statewide tax limit. The city produces a local public good with two inputs, an administrative input (a) and a service input (s), for example, teachers or uniformed police officers. City officials have a payoff function , which is not necessarily the same p[a, s] as the production function . One response to a budget cut is to simply g[a, s] scale back both inputs to maximize the officials’ current payoff, taking the smaller budget as exogenous. The second option is to deviate from this year’s payoffmaximizing input bundle in an attempt to persuade local voters to override the statewide limit. A relatively large cut in the service input will cause a relatively large decrease in public output, meaning that voters may be more likely to override the limit. A city with an override option may sacrifice part of its current payoff in anticipation of a bigger budget and a higher payoff in a later period. Consider a city with a fixed population. The city’s budget constraint is b p , where is the budget and the prices of both inputs are fixed at one. The a s b representative citizen has a fixed income y and preferences over a private good (x) and the two public inputs. The budget constraint and utility function of the representative citizen are, respectively, and . y p x b v p x u(a, s) Figure 1 shows the solution to the city’s input-choice problem for an exogenous budget . At the payoff-maximizing bundle (shown by point j), the price ratio b0 equals the ratio of the marginal payoffs. We assume that the city has a bias toward administrative inputs: the city’s decision makers, who are of course ad6 Thomas Romer & Howard Rosenthal, Bureaucrats versus Voters: On the Political Economy of Resource Allocation by Direct Democracy, 93 Q. J. Econ. 563 (1979). 7 Kangoh Lee, Bureaucrats and Tax Limitation, 34 J. Urb. Econ. 24 (1993); Mark Toma & Eugenia Toma, Bureaucratic Responses to Tax Limitation Amendments, 35 Pub. Choice 333 (1980). 8 Thomas A. Downes, An Examination of the Structure and Governance in California School Districts before and after Proposition 13, 86 Pub. Choice 279 (1996). 236 the journal of law and economics