No. 8 November 1997
The Whitman Tax Cuts: Real Gains For New Jersey Taxpayers
Timothy J. Goodspeed is a Professor of Economics at Hunter College
Governor Christine Todd Whitman spearheaded a successful effort to cut the state income tax in 1994, the most recent development in an ongoing political struggle over the financing of public schools in New Jersey. The debate stems from 1973, when the New Jersey Supreme Court ruled that the practice of relying entirely upon local property taxes to fund schools violated a provision of the New Jersey State Constitution stipulating that all children are entitled to a “thorough and efficient” education. In response, then-Governor Brendan Byrne proposed a statewide income tax to raise funds for aid to local school districts. The income tax became law in 1976 only after the New Jersey Supreme Court closed public schools in response to the Legislature’s initial failure to approve the tax.
Since that time, the judicial and legislative branches of the New Jersey state government have continued to wrestle over the issue of school financing. A large state tax increase and a restructuring of aid under former Governor Jim Florio also came in response to a State Supreme Court decision, and set the stage for Governor Whitman’s subsequent cut.
Under current law, revenue from the state income tax is dedicated solely to the relief of local property taxes, more than 80 percent of whose revenue is used for aid to local school districts. (The other 20 percent is split about evenly between municipal aid and the homestead rebate.) For that reason, there is a link between the state income tax and school district property taxes.
This link has led to the fear that the Whitman tax cut would simply result in a dollar-for-dollar rise in local property taxes, thus negating any savings that taxpayers might realize. The report examines this proposition in three ways: Section I is an analysis of the latest publicly available aggregate data on New Jersey income and property taxes. Section II analyzes studies of how local government spending changes when state or federal aid has changed. And section III is a look at the relevant economic theory.
I. Did the State Income tax cut yield a lasting benefit to taxpayers?
The relationship between the state income tax and local property taxes has been the subject of ongoing debate, revolving around a simple question: Does a reduction in state income taxes yield a real and lasting benefit for taxpayers, or does it merely shift the burden from one category of taxes to another? Figure 1 shows the real rate of growth of New Jersey income and property taxes, based on the 1996 Annual Report of the Division of Taxation of the New Jersey Department of the Treasury. The Annual Report for 1996 is published in January 1997 and represents the most recent information available. Property tax figures are from 1995, the most recent year for which this data is available.
The nominal dollar figures are divided by the consumer price index to yield real figures. The rate of growth is computed as the difference in revenue from one year to the next, divided by the revenue figure from the first year. The graph takes into account all three governmental levels that levy property taxes: counties, municipalities, and school districts.
The most obvious conclusion that can be drawn is that the aggregate statistics do not reveal much of a pattern. The real rate of growth of property taxes did drop initially in 1991 when then-Governor Florio raised the state income tax. However, property taxes experienced a fairly steady real increase of 1 - 2 percent per year during fiscal years 1992 to 1995, despite the fact that the income tax revenue both increased and decreased at different times during that same period: fiscal year 1992 saw high real growth in the income tax, again reflecting the Florio tax increase, while 1993 and 1994 saw modest (1 to 2 percent) real growth, and 1995 was a year of negative real growth, reflecting the Whitman tax cut. Over the twelve-year period, property tax growth has ranged from 4 to 13 percent (-2.7 percent to 9 percent in inflation-adjusted terms).
However, since more than 80 percent of income tax revenue goes to school districts, it is more relevant to compare the income tax vs. school district property taxes. That comparison is charted in Figure 2. (Information for this graph also comes from the Annual Report, Appendix B.) This figure reveals a pattern similar to the one above. Namely, school district property taxes increased steadily and moderately from 1991 to 1995 while tax revenue both increased and decreased at different times in the same period.
An ongoing study by Goodspeed (1997), uses regression analysis to analyze data on individual New Jersey school districts for the years 1991 to 1995. Property taxes neither decreased enough to offset an aid increase or increased enough to offset an aid decrease. (One explanation for this may be what is termed the “flypaper effect,” which will be discussed below.)
Results so far also reveal that responses to the income tax cut have varied widely from school district to school district, and that the state tax cut has not led to across-the-board property tax raises. The evidence does indicate that districts experiencing greater increases in enrollment tend to have higher property taxes. The fact that higher income school districts tended to be losers under the Florio tax increase (according to the study by Bogart, Bradford, and Williams, 1992) and are likely winners under the Whitman tax decrease also influenced changes in property taxes in those districts. The study indicates that these higher-income districts tended to decrease their property taxes by more than other districts after the Florio increase, and tended to raise their property taxes by more than other districts after the Whitman tax cuts.
These results suggest that the combined impact on an individual taxpayer resulting from the changes in the income and property tax will vary depending on such factors as income and their particular school district.
Overall, however, it is clear that the cuts have yielded significant aggregate savings. The Whitman tax cuts reduced rates to 1.9 percent - 6.65 percent effective January 1, 1994; to 1.7 percent - 6.58 percent effective January 1, 1995; and to 1.4 percent - 6.37 percent effective January 1, 1996. Through 1995, this represented a reduction of 15 percent for taxpayers in the lowest tax bracket and 6 percent for those in the highest. Total revenue for 1995 was $4.54 billion. Assuming no supply-side response, the savings for 1995 was between $272 million and $681 million (the lower figure assumes the 6 percent savings that would result if everyone were in the highest tax bracket while the higher number assumes the 15 percent savings that would result if everyone were in the lower tax bracket.)
The New Jersey Treasury’s Division of Taxation estimates the revenue loss at $312 million. Since about 80 percent of this goes as aid to local school districts, we can assume that aid was reduced by between $217 and $545 million; according to the New Jersey Treasury Department’s estimate, this figure would be $250 million. Total property taxes for school districts rose from $5.985 billion in 1994 to $6.234 billion in 1995, an increase of $249 million.
It is difficult to know how much property taxes would have risen if income taxes had not been cut. However, the average real increase in property taxes in 1992 to 1993 was 1.17 percent while the increase in 1994 to 1995 was 2.265 percent. It is, therefore, reasonable to guess that property taxes might have risen by about half of their actual increase in 1995, or about $125 million.
One reaches a similar estimate by applying a rule of thumb derived from other studies, namely, that property taxes tend to rise by about $.50 for each $1 cut in aid to an estimated cut in aid equal to 80 percent of the loss in income tax revenue estimated by the New Jersey Treasury Department. This means that on the whole property taxes rose by about $125 million more than they would have in 1995 while income tax payments were lower by between $217 and $545 million. Again, even assuming the lowest possible estimate, taxpayers on the whole saved money.
Of course, the combined impact for an individual could be positive or negative depending on a variety of factors, including their income tax bracket, the actual loss in aid for their school district, changing enrollment of the school district, and any choice on the part of the individual to vote for greater spending on education.
II. How Local Governments Respond to Changes in State Aid
A large empirical literature, developed in the 1970s and 1980s, addresses the question of how lower-level governments respond to changes in aid from higher levels of government. Bradford and Oates (1971), who provided the theoretical background for much of the work that followed, theorized that local governments would view lump-sum grants as increases in income. This implies, for example, that if a local government would spend about five cents of an extra dollar of income on schools, they should also be expected to spend the same portion of an additional dollar of grant money for schools, and the other $.95 would be returned to taxpayers in the form of lower property taxes.
In fact, as Oates (1994, p. 135) notes, according to a survey of studies on intergovernmental grants by Gramlich (1977), “something like 40 to 50 cents of such grant dollars manifest themselves in additional public spending.” Since the empirical findings did not correspond to the theoretical prediction of Bradford and Oates, the phenomenon was dubbed the “flypaper effect” because money from the state appeared to stick to the local government rather than being returned to taxpayers in the form of lower property taxes.
As more and more studies continued to find the flypaper effect, the empirical literature in turn inspired other papers that tried to explain the phenomenon. One explanation is that voters are subject to “fiscal illusion.” For instance, Oates (1979) and Courant, Gramlich, Rubinfeld (1979) suggest that grants may contribute to voters’ confusion over average and marginal costs of public services, such as education. When grant money flows to education, voters may think (falsely) that there has been a drop in the price of providing schooling and decide they wish to spend more. Another fiscal illusion model emphasizes efforts by a local bureaucracy to maximize its budget by concealing grant funds (Filimon, Romer, and Rosenthal, 1982). The result is that voters are again fooled into thinking that the price of education has fallen and vote to increase the size of the budget, achieving the goal of local officials. Another twist is that of Romer and Rosenthal (1980), who use the idea of a reversion level discussed in Romer and Rosenthal (1979). In this model, budget maximizing bureaucrats offer a diminished budget as the only alternative to a bloated budget, with the likely outcome being that voters choose the large budget as the lesser of two evils. Further, Romer and Rosenthal (1980) argue that lump-sum grants will increase the reversion level, which, under certain conditions, leads to a higher level of spending than a simple increase in income. A third explanation, that of Fisher (1979), suggests that differing tax prices for an individual between higher and lower levels of government can lead to a flypaper effect.
More recent studies examine whether local governments respond differently to decreases —as opposed to increases—in aid. This literature was inspired by the cuts in federal aid that took place during the 1980s. Two papers theorize that increases and decreases prompt different reactions. Gramlich (1987) suggests that local governments may have difficulty in cutting back on established programs when aid is cut. Alternatively, Stine (1994) suggests that governments may tighten their belts when aid is cut.
Very few empirical studies examine this question. Those that do either find no difference between how local governments react or that decreases prompt a more pronounced flypaper effect than increases. Stine (1994) finds that the flypaper effect is so strong that Pennsylvania counties actually decreased taxes when aid was cut. Ghamkar and Oates (1997) find state and local governments behave similarly in the wake of both increases and decreases.
In summary, studies of the behavior of local governments to changes in aid generally find a flypaper effect. Moreover, the flypaper effect appears to be at least as strong when aid is decreased as when it is increased. Based on that hypothesis, one would predict that New Jersey property taxes are unlikely to rise dollar for dollar when state aid is cut.
III. The Effect of the New Jersey Income Tax Cut on Local Property Taxes
We can also use economic theory to understand what is happening to New Jersey property taxes, keeping in mind the particular circumstances that exist in the state.
First, it is important to note that the state income tax in New Jersey is primarily a response to court mandates concerning aid for education, and the aid is, therefore, essentially redistributive in nature. Moreover, beginning in 1991, aid to wealthier school districts was phased out and while poorer districts received increases. Consequently, property taxes were likely to have risen in wealthier districts for reasons quite apart from income tax cuts. Conversely, the increased aid to poorer districts was likely to result in smaller property tax increases there. Indeed, one might argue that the election of a governor who promised to cut income taxes represented voters’ response to the court mandate. In this sense, the constant debate over New Jersey income and property taxes is part of a larger back and forth debate between voters and the court.
Two other factors are also important to keep in mind. First, property taxes may increase because of changes in school districts’ enrollment. If enrollment increases relative to the taxpaying population, each household will have to bear a larger share of the costs, and property taxes will rise.
In addition, income tax cuts naturally leave households with more disposable income. Consequently, these households are likely to demand additional goods—including services provided by local governments. Property taxes are likely to rise for this reason. But a distinction should be made between this phenomenon and tax increases not precipitated by consumer demand.
In summary, property taxes will tend to rise more in those school districts with a greater enrollment relative to the taxpaying population, in those districts whose aid has been phased out, and in those districts that are the greatest beneficiaries of the income tax cut that leaves them with more disposable income. A fourth factor, the flypaper effect mentioned in the previous section, will tend to dampen any increase in property taxes that results from reduced aid.
Economic theory suggests, therefore, that increases in property taxes may be spread unevenly across school districts. Although previous studies indicate that a $1 aid reduction is likely to translate to an average property tax increase of only $0.50, factors such as increased school enrollment and the degree to which the residents of a district have benefited from the income tax cut suggest that the actual increase in property taxes is likely to be uneven.