1. State and Local Public Finance Professor Yinger Spring 2016 LECTURE 3 VOTING

2. Class Outline  The Median Voter Model o How does it link voting and demand? o Why it is useful? o What are its limitations?  Property Tax Limitations o What are they? o How do they shed light on voting? State and Local Public Finance Lecture 3: Voting

3. Voting and Demand  In most cases, citizens cannot directly express their demand for local public services.  So they express their demand through their voting, either for public officials or on referenda.  In this class, we explore how an understanding of demand helps us understand the choices local governments make. State and Local Public Finance Lecture 3: Voting

4. The Median Voter  The median voter always votes on the winning side. o Line voters up by the strength of their preference for public services. o Identify the voter in the middle—the median voter. o A majority vote must include the median voter. -------------------------M----------------------------- weakest strongest preference State and Local Public Finance Lecture 3: Voting

5. Identifying the Median Voter  But voters do not line up, so the median voter is not identified!  The median voter model shows how to identify the median voter assuming preferences are driven by demand factors. o The median voter has the median income and the median tax price. o This model places certain restrictions on preferences. o This model assumes political institutions are neutral. State and Local Public Finance Lecture 3: Voting

6. The Median Voter Model TP = MC( ) D = MB S S* $ State and Local Public Finance Lecture 3: Voting S = Local public service quality D = Median voter’s demand MB = Median voter’s marginal benefit from S TP = Median voter’s tax price MC = Marginal cost of S = Median voter’s tax share S* = Median voter’s preferred S

7. Using the Median Voter Model The median voter model is widely used because:  It explains community decisions based the demand function for a single voter—the median voter.  It makes use of widely available data at the community level: spending (or performance), median income, median tax price (median divided by mean house value).  It works! State and Local Public Finance Lecture 3: Voting

8. Example: Eom, Nguyen-Hoang, Duncombe & Yinger, EF&P, 2014  1991-2011 data for over 600 school districts in New York State  Service quality measured by an index of high school exams and drop-out rates  Income elasticity = 0.17 (based on income per pupil)  Price elasticity = -0.22 (based on median tax price)  Significant preference variables include share of housing that is owner- occupied and share of population age 65 or older. State and Local Public Finance Lecture 3: Voting

9. Other Examples  The median voter model works well for local governments, particularly school districts.  Recent applications include: o Massachusetts (Nguyen-Hoang and Yinger, Journal of Education Finance, Spring 2014). o California (Duncombe and Yinger, International Tax and Public Finance, June 2011). State and Local Public Finance Lecture 3: Voting

10. Preference Assumptions  Median voter models need “single-peaked” preferences.  Consider two communities with a high demand for education, one (City A) with a good private school and the other (City B) without.  In City A, the high demand for education may not show up as a high demand for public education.  In this case, which is extreme, the median voter model breaks down. State and Local Public Finance Lecture 3: Voting

11. The Role of Institutions  Scholars disagree about the importance of institutions.  Some scholars believe institutions are fairly neutral in most cases, particularly outside large cities.  Other scholars believe bureaucrats can have a large impact on spending and service quality—the leviathan view.  In some cases, it is possible to control for institutions, such as whether a city as a mayor or a city manager. State and Local Public Finance Lecture 3: Voting

12. Tax Limits  Tax limits challenge the median voter model: Why do voters need to limit taxes if they control spending anyway?  Some scholars believe that most tax limits make a political point with little impact on spending.  Others believe that tax limits reveal voters’ belief in leviathan—and their desire to reign in bureaucrats. State and Local Public Finance Lecture 3: Voting

13. Evidence about Tax Limits  Studies of tax limits in the U.S. lead to four main conclusions.  First, property tax limits shift the tax burden away from property taxes to other revenue sources. o This shift was very large right after Proposition 13 in CA and Proposition 2 ½ in MA. o But the shift has moderated since then. State and Local Public Finance Lecture 3: Voting

14. Evidence about Tax Limits, 2  Second, strong tax and expenditure limits (TELs) limit but do not stop the growth in state government revenue. o A large share of this effect comes from a drop in state aid to local governments o And hence involves a higher burden on local tax sources. o The magnitude of these impacts depends on the design of the limit (see Kioko, PB&F, 2011). State and Local Public Finance Lecture 3: Voting

15. Evidence about Tax Limits, 3  Third, there is no compelling evidence that TELs have boosted the efficiency of state and local governments. o States with strong TELs have experienced a decline in the quality of public services. o The performance by California’s students on national tests has dropped significantly, for example, since Proposition 13. State and Local Public Finance Lecture 3: Voting

16. Evidence about Tax Limits, 4  Fourth, voter demand factors still matter even with strong TELs.  California has severe limits, for example. o But Duncombe and Yinger (ITPF, 2011) find that voters with higher incomes or lower tax prices still pick higher school quality— o Through active monitoring of school officials or the use of secondary revenue sources, such as parcel taxes or private foundations. State and Local Public Finance Lecture 3: Voting

17. The New Tax Cap in NY  A couple years ago Governor Cuomo proposed and the legislature passed a cap on tax levy increases for all local governments.  The cap is the lesser of 2% and inflation, with exceptions.  The cap exempts new construction (to preserve incentives for development and growth). State and Local Public Finance Lecture 3: Voting

18. The New Tax Cap in NY, 2  The legislation allows jurisdictions to override the cap with a 60% vote.  But in a bizarre twist to the story, NY added a tax-freeze credit in 2014, which rebates any property tax increase below the cap.  As of this year, the credit is provided only if a district not only meets the cap, but also implements an “efficiency plan.” State and Local Public Finance Lecture 3: Voting

19. The New Tax Cap in NY, 3 State and Local Public Finance Lecture 3: Voting  Few school districts voted to override the original cap.  With the addition of the “tax-freeze credit” in 2014, the number dropped still more.

20. The New Tax Cap in NY, 4  Nobody should expect this cap to make local governments more efficient.  Without more state aid (unlikely) or many overrides (more likely, at least in the long run) this cap will cut local services. State and Local Public Finance Lecture 3: Voting

21. The New Tax Cap in NY, 5  This cap is likely to increase disparities across jurisdictions in education and other public services.  Richer school districts are more likely to override the cap.  Poorer school districts need larger percentage increases in revenue just to keep up with richer districts. State and Local Public Finance Lecture 3: Voting

22. The New Tax Cap in NY, 6  Some wealthy school districts raise $30,000 in property taxes per pupil; Buffalo, Rochester, and Syracuse (the Upstate Big 3) raise $3,700.  In 20 years, the allowable annual revenue will go up by $30,000×[(1.02) 20 – 1] = $14,578 in these rich districts,  But by only $3,700×[(1.02) 20 – 1] = $1,798 in the Upstate Big 3. State and Local Public Finance Lecture 3: Voting