This report will show that the local property tax is an efficient method of funding high-quality public education in New Hampshire when compared to methods that are more centralized, such as a state income or sales tax. A system of finance is more efficient than another if, with the same economic cost, the system produces higher-quality results than another. New Hampshire's public school system, which relies largely on local property taxation, is efficient because it supports at a reasonable cost public schools whose students do very well by national standards.Overview of report
There is much economic evidence in support of the efficiency of local property taxation for schools. Local property taxation is more efficient because it provides a connection between school spending and the value of residential property. The manifestation of this connection is called "capitalization." When a home buyer pays more for a house because the property tax rates in the community are low, economists say that the lower tax rates are "capitalized" in the price of the house.
Capitalization of local property taxes has two important implications. First, the claim by petitioners' experts that tax base and tax rates are the principal measures of ability to pay and financial sacrifice, respectively, is shown to be wrong. Their claim ignores that high tax rates are offset by lower housing costs. Petitioners' experts entirely overlook the economic principle of capitalization, and, as a result, their conclusions about the burden of the tax and related conclusions about the school finance system are misleading.
Supposedly excessive tax burdens are offset by lower housing costs. Second, capitalization provides an incentive for all residential property owners to demand that their local school boards supply good schools without wasteful spending. Higher-quality schools raise the value of residential property in any school district. Having to pay for school improvements out of property taxes fixes the attention of resident voters. They will ask whether the higher spending on schools will add more to their property values than the higher tax rates will detract from those values. The local property tax system, unlike any statewide system of finance, thus has built-in incentives that promote efficiency. Local voters will support policies when an extra dollar spent on schools is worth more than the extra dollar of taxes they must pay. States that have moved away from local property taxation have sacrificed this efficiency discipline, and their school systems have suffered as a result.
1.The property tax is an administratively efficient method of apportioning individual tax burdens.
The property tax is not a tax on physical property itself. It is a tax on the owners of property in proportion to the value of their ownership share in the local tax base. Property that has no identifiable owner cannot be taxed. It may exist as a physical object, but it cannot be made to disgorge economic resources without an owner. Land does not spew forth tax dollars; houses do not exude revenue from their foundations. When this report refers to the property tax, it means a tax on owners of property in proportion to their share in the overall value of the district that comprises the tax base.
Once the share of ownership in the tax base is established by assessors and the budget for local expenditures is determined by local governing bodies, the property tax rate itself is a residuum. Suppose, for example, a family owns a house valued at $100,000. If the tax rate is 3.0 percent, their tax bill is $3000. Should the nominal value of all taxable property in the district double, the family's house value is now $200,000. If the amount of tax revenue to be raised by the district stays the same, the tax rate will now be 1.5 percent, and the family's tax bill is still $3000. Should the nominal value of all property be cut in half, the family's assessment is $50,000 and the tax rate is raised to 6.0 percent, and the family's tax bill is again $3000. The foregoing arithmetic illustrates an important difference between local property taxes and statewide taxes such as sales and income taxes. With a sales or income tax, a rate is typically set by the legislature and held constant for several years. The annual revenue thus becomes the product of the tax rate and the amount of gross sales or taxable income in the state. As a result, state expenditures and revenues do not always match up, and budgetary adjustments must constantly be made throughout the year. In contrast to this uncertain means of budgeting, the annual property tax levy is determined from the budget of the local district, and it is then apportioned over a predetermined tax base. The tax rate is simply the ratio of the approved budget to the tax base.
This feature of the property tax is advantageous for funding local schools. Expenditures can be determined without having to anticipate the level of economic activity that would produce income or sales tax revenues. The property tax is thus a more stable and predictable source of revenue. A temporary recession that lowers all property values does not automatically reduce school budgets. The property tax rate is merely a means of allocating to individual property owners the amount they must pay once the budget and their relative share of the tax base have been determined. Unlike statewide property, sales, or income taxes, differences in local property tax rates are not indicators of differences in economic burdens of the tax.
2.The economic burden of the property tax is altered by capitalization.
Economic analysis reveals that those persons who are legally obligated to pay the tax are often different from those who bear the ultimate burden of the tax. (Burden of the tax means the amount of goods and services that people must give up as a result of the taxes.) A generally accepted example of this principle is the social security taxes that are nominally paid by employers. Economists have concluded that this burden is usually "shifted" to employees (Musgrave and Musgrave 1989, p. 441). The employees do not actually get the employer's tax bill, but they bear much of the burden of the tax because employers pay them lower wages to make up for the higher taxes.
Such shifting also occurs with the local property tax through an economic phenomenon called "capitalization." Capitalization focuses on buyers of property. When a family seeks to purchase a house in a geographic area containing more than one school district, it evaluates possible purchases on the basis of the fiscal advantages as well as the physical qualities of the house. Holding the quality of schools constant, the family will be willing to pay more for a house in a district with low property taxes than it would for an otherwise similar house in a district with high property taxes.
When economists find that a lower-than-average tax rate increases the purchase price of a house, they say that the lower taxes are "capitalized" in the price of the house. Similarly, when purchasers pay a lower price for a house because it is in a high-tax district, economists say that the higher tax rates have been capitalized. Higher-than-average tax rates will be capitalized by lowering house prices, and lower-than-average tax rates will be capitalized by raising house prices. (An example that works out the arithmetic of this principle is in Appendix A, and an example that explains it by analogy is in Appendix B.)
Petitioners' experts claim that residents of a district with a larger-than-average tax base per pupil have more resources to pay for schools. [footnote 1] Capitalization explains why this facile analysis is not necessarily true. Buyers of houses in districts with a large tax base per pupil must pay a higher price for their houses. The larger tax base is "capitalized" in the price of the house they buy. As a result, home buyers in relatively large tax-base districts will have to incur a larger annual mortgage cost.[footnote 2] Because of these higher housing costs, they have less money each year to pay their (lower) property taxes.
On the other side of the capitalization coin, the buyers of houses in districts with a smaller tax base per pupil pay less for their houses. As a result, their annual housing costs (the mortgage and related payments) are lower than in the district with the larger tax base per pupil. Because of the lower housing costs, house buyers in smaller tax-base districts will be able to pay their higher property taxes. In sum, the rule is this: Taxes up, house costs down. Taxes down, house costs up.
Because it is people who pay property taxes, the economic burden of the tax must be evaluated in terms of its effect on people's incomes. When complete capitalization occurs, the burden of the property tax on families of a given level on income will be same among all communities, even when large differences in tax rates exist. In a high-tax-rate, low-tax-base community, the economic burden of the property tax will be in the form of a larger annual tax bill and smaller annual mortgage and other housing costs.
The example from Appendix A can be used to illustrate this fact. This example compares two districts with identical houses and school expenditures. The annual tax payment in the low property-tax-base district is $5000, and the corresponding annual mortgage is $10,000. For a family with $50,000 in gross income, these two payments leave them with $35,000 a year to allocate to other expenditures or savings.
In the high property-tax-base district with the lower tax rate both taxes and mortgages are different. Because the lower tax rate raises the value of the otherwise identical house to $122,475, the annual mortgage payment (calculated at 10 percent of value) is $12,247.50 and the annual tax payment is $2752.50. The sum of these two payments is $15,000, which is the same as the sum in the low tax-base community. For a family in the high tax-base district with $50,000 in gross income, the amount remaining after taxes and mortgage payments is $35,000, the same as it was for the family in the low-tax-base community. It might be argued that the mortgage payment is a private decision and should be disregarded in comparing the burden of taxes. But this argument is illogical, because the only thing that makes the mortgage payments vary in this example are the property tax payments. Analyzing property taxes alone ignores a significant part of the true economic cost because schools and housing are inextricably tied.
To determine why taxes and housing costs must be evaluated as a whole, consider two private boarding schools, both of which require that students live in school-owned housing. One school charges $10,000 in tuition and $10,000 for room and board. The other charges $12,000 for tuition but only $8,000 for room and board. In comparing the cost of going to each school, it would be illogical to look only at tuition. It is the total cost--$20,000--that parents will look at in comparing the two schools.
When capitalization occurs, the petitioners' claim that tax base per pupil is the principal measure of ability to pay for schools is shown to be wrong. In the previous example, tax bases differed considerably, but each homeowner's ability to pay those taxes was the same. Thus tax bases are not a measure of local residents' ability to pay for schools. The petitioners also claim that higher tax rates indicate greater economic effort to pay for schools. But in the previous example, the total economic efforts were the same despite differences in tax rates.
It also follows from this analysis that there is no such thing as a "free" tax base that the state can simply transfer from one district to another. The state cannot increase tax rates in high property-tax-base districts and transfer revenues to other districts without causing a capital loss in the high property-tax-base districts. The loss would fall on all homeowners in the high-base district, regardless of their incomes. Their property taxes would rise, but their mortgage payments, which are based on purchase price, would not fall accordingly. Only after they sold their homes would the mortgage payments of the new buyers diminish. The combined effect of higher taxes and constant mortgage payments would be a loss to current taxpayers. Redistribution of the tax base is not possible without cost.
The claim that larger-than-average local tax bases can be exploited by the state without economic consequences also runs counter to economic logic. If there were unclaimed $50 bills lying on the ground, economists would expect that they would not lie there for long. People would pick them up in short order. If it were true that local tax bases were a costlessly obtainable tax resource, state legislators would have noticed it a long time ago. Like those who stumble upon $50 bills, the legislators would have picked up the "free" tax resources. But, of course, the legislature has had continuous opportunities to change the system, and it has declined.
The Legislature's unwillingness to tap into local property tax bases has nothing to do with the alleged "disproportionate influence" that so-called "property-rich" towns have in the legislature.[footnote 3] It is because the majority of voters of New Hampshire do not wish to change the system of local control of property taxes. On balance voters perceive greater benefits in keeping property taxes within their own jurisdiction, even when taxing property in other jurisdictions might appear to lower their tax bills.
The effect of capitalization in mitigating the burdens of public policies on property owners has been indirectly recognized by the courts in other contexts. In Claridge v. New Hampshire Wetland Board, 125 N.H. 745, 485 A.2d 287 (1984), the New Hampshire Supreme Court pointed out that landowners who bought wetlands property after the regulations were put into effect had no reason to complain of property devaluation (Claridge, 485 A.2d at 292). The buyers had notice of the regulation and, one presumes, offered a lower price for the land, knowing that economic uses would be limited. To provide compensation for the wetland restrictions long after they had been in place would reward purchasers of property for a risk they had already been compensated for in the form of lower prices. This is a principle that has long been recognized by the courts. [footnote 4]
3. Studies show that property taxes are capitalized.
The analysis and illustrations in the previous section were derived from accepted economic principles. Because it undermines the petitioners' claim that tax base is the proper measure of economic resources and tax rates measure economic effort, it is important to establish that tax capitalization exists. This section establishes the existence of tax capitalization by describing three sources: (a) published econometric studies that have examined the effect of property taxes on housing prices; (b) an econometric study of 73 New Hampshire school districts that shows that school tax rates are significant determinants of variations in house values; and (c) specific examples of tax capitalization from New Hampshire.
The evidence that higher property taxes are capitalized in housing prices is established by an overwhelming number of econometric studies. A book by Yinger, Borsch-Supan, Bloom, and Ladd (1988, pp. 11-47) reviewed in detail 30 published studies of tax capitalization by professional economists. These studies used a variety of samples from states and metropolitan areas around the country in which property taxes were the main means of financing local schools. All but three of the studies show capitalization of property taxes. The studies showing significant capitalization examined samples of local governments in California, Connecticut, Kentucky, Missouri, Montana, New Jersey, New York, and Pennsylvania. Of the three that showed no capitalization, two used samples from Canada, whose tax laws differ from those of the U.S.
The Yinger et al. book added to these studies a detailed analysis of tax capitalization in cities and towns around Boston. This analysis found that lower property taxes clearly raised home values. In a subsequent evaluation of school finance issues specifically, two of the authors of this study stated that "In the presence of capitalization, which has been documented by many empirical studies, increases in real income associated with higher service quality or lower taxes may be partially or totally offset by higher rents or housing prices" (Ladd and Yinger 1994, p. 218; notes omitted).[footnote 5]
The second demonstration of capitalization is the econometric study that is described in detail in Appendix C. This study was undertaken to determine whether the evidence from other states also applies in New Hampshire. To answer this question, the econometric method common to most other studies, multiple regression analysis, was applied to a New Hampshire sample of school districts. The sample was limited to 73 school districts because of data limitations described in the appendix, but the same conclusions reached by this study would apply to all districts.
The regression analysis found that variations in the median value of owner occupied homes among districts were accounted for by the number of rooms in the structure, the distance from the southeast part of the state (measured as airline distance from the Boston area), the average age of the houses, the school tax rate, and the average composite score on fourth-grade California Achievement Tests (CAT). The regression shows that lower school tax rates are associated with higher house values even after characteristics of the house (number of rooms and age) and location within the state (distance from Boston) are taken into account.
The school tax variable is highly significant by generally accepted statistical methodology. There is a less than one-tenth of one percent probability that a repeated test using a different sample would not lead to the same conclusion. There can be no doubt that school property taxes are capitalized in New Hampshire towns and cities. The size of the estimate indicates that the extent of capitalization is nearly 100 percent.
Other examples of property tax capitalization are given in the report to the state by Robert Estey. Rather than use average house values of communities across the state, as in the previously-discussed statistical test, Mr. Estey, an experienced assessor in Hampton, compared specific properties from the towns of Hampton and neighboring Seabrook. These are especially apt examples because the two towns send their children to the same high school. Thus the buyers of homes in the two towns receive similar public educational services. Mr. Estey found that a house in Seabrook sold for about $20,000 more than the physically comparable houses in nearby Hampton because Seabrook has a lower tax rate, due largely to the nuclear power plant. Thus buyers of houses in Seabrook submit to higher annual mortgage payments that entirely offset the advantages of lower annual taxes.
Mr. Estey also found that assessors in the towns of Hampton, Exeter and Rochester valued a standard-quality home (a three-bedroom Cape) at different rates that reflected differences in property taxes. In the higher-tax-rate towns, the same property would cost a buyer less (according to the local assessors) than it would in the lower-tax-rate town. These estimates of capitalization are less exact because they do not account for possible differences in local school costs and location advantages, but they are nonetheless consistent with the proposition that property taxes are capitalized.
A comparative study conducted by the New Hampshire Department of Revenue Administration also found evidence of capitalization. The study compared recent sale prices of houses that are nearly identical (as shown by photographs) except that they are located in communities with different tax rates. The comparisons of four houses in Concord to four houses in Bow are most appropriate for determining the existence of capitalization because Concord and Bow are adjacent and their students go to the same high school. Tax rates in Bow are significantly lower than in Concord. But in each of the four paired comparisons, the price of the house in Bow is significantly higher. A buyer who thought she would get a good deal by purchasing the house in low-tax-rate Bow would be disappointed. She would have to pay between $10,000 and $20,000 (approximately) more for the Bow house that is otherwise similar to the Concord house. The Bow house's addition to her annual mortgage payments would largely offset the lower annual tax payments.
Capitalization is also suggested in evidence introduced by the petitioners. Three of the five taxpayer plaintiffs in the current litigation indicated that the reason they bought housing in their high-tax-rate communities was because its price was lower than elsewhere. Interrogatory #51 (First Set, p. 49), asked about "considerations involved in each taxpayer plaintiff's decision to reside in his or her current place of residence." "Affordable housing" was mentioned by Viar of Allenstown, Fitzgerald of Lisbon, and Elliot of Claremont. Capitalization of the property tax implies that if the tax rates in those places had been lower, the houses would have been less affordable in terms of their purchase price.
4. Capitalization promotes higher quality schools without wasting resources.
The previous sections dealt with capitalization of property-tax differences, holding school quality (and, implicitly, other public services) constant. The lesson there was that buyers of houses get what they pay for. Low-tax districts are more expensive to live in. The present section considers the other side of the capitalization coin. Rather than reviewing what happens to buyers of property, this section analyzes the economic motivation of a district's present owners, most of whom expect to be sellers of their houses sometime in the future. These citizens want to see their home values increased because it raises their own wealth. This section will show that this financial interest is a valuable supplement to collective efforts to provide good schools.
The basic premise is this: Because better-quality schools increase property values and higher tax rates reduce property values, voters have a financial incentive to make sure that money spent on schools actually improves education. Concern with the value of one's property makes even childless voters willing to spend local money for schools when the money results in better education. The principle that better schools raise house values is subtle, since there is no immediate cash flow (as there is in taxes) that measures differences in school quality. What makes a school system attractive to some families may not mean much to others. (This demonstrates the advantage of a multiplicity of districts, which allows for greater choice and experimentation in public education.) Nonetheless, there appear to be school characteristics that are sufficiently valued by a majority of home buyers that can cause positive capitalization. Districts whose schools add to these qualities will have higher house values than they had before. For example, the regression analysis in Appendix C showed that higher achievement test scores have a positive impact on the value of homes in New Hampshire towns and cities.
The connection between, on the one hand, house values and, on the other hand, local taxes and local school quality puts in place a powerful benefit-cost mechanism. Voters in every district have an incentive to approve efficient spending on schools. A spending proposal that raises school quality in ways that appeal to prospective home buyers will raise the value of existing homes. The increase in local taxes to finance that spending will lower the value of those same homes.
Thus not all spending proposals will be approved. Only those school spending proposals in which the net effect on home values is positive will be approved. This mechanism--the local property tax system--is superior to others because of the close match between those who benefit from school spending and those who bear the costs. For a public service in which ideas about quality vary, it is highly desirable to match up benefits received and costs paid. Local property taxation makes that match for schools better than a more centralized system. Economists have attempted to measure capitalization of school quality in two ways. Several studies have shown that, where there are a number of different school districts to choose from,[footnote 6] those with higher-spending per pupil have higher house values.[footnote 7] Other studies have attempted more subtle measures of school quality, such as student success on standardized tests. Higher scores have been shown to increase the house values of communities in which the students go to school.[footnote 8] A recent study by Bradbury, Case and Mayer (1995) found significantly greater housing values in Massachusetts towns and cities whose school spending and test scores increased more than average in the 1990-1994 period. That better schools cause higher housing values has long been established in many studies throughout the United States and confirms the conventional wisdom among real estate professionals.
Capitalization of school qualities provides a reason for all homeowners to have a financial interest--above and beyond their general interest in civic affairs and education--in the quality of schools in their towns. Better schools increase the value of their homes, regardless of whether the homeowner has children in school. The numerous studies that find that better school qualities increase home values do not inquire whether the homeowner has school-age children. The value of any home with room for children will increase when schools are improved. Prospective buyers of those homes include people who have or expect to have children eligible to attend local schools. A childless family that owns a home has a selfish reason to be willing to pay for efficient local education.
The capitalization of school quality helps explain what would otherwise be a paradox. In most communities, voters who do not currently have children in public schools outnumber those who do. If schools were of interest only to those who had children in school, one would expect that other voters would constantly defeat school budgets above the state-mandated minimum. Although some budgets are reduced every year, most are passed without modification. Most childless voters have no financial reason to oppose efficient school spending.
Evidence from around the country supports the foregoing contention. A professional poll of a suburban Illinois town asked whether voters would favor increased property taxes to increase funding for local schools. According to two social psychologists who analyzed the poll, the most consistently significant result was that voters "who thought that property values in the neighborhood would decrease if the schools deteriorated were likely to favor a tax increase" (Rasinski and Rosenbaum 1987, p. 1000). An analysis of a California community's 1970 school tax referendum by two economists found that "the larger is the average expected increase in property values in a precinct, the more likely it is that voters in that precinct will support the referendum" (Sonstelie and Portney 1980 [Sept.], p. 194).
The minutes of the Claremont, New Hampshire, School District meeting of March 11, 1995, also provide evidence of this concern. Allen Whipple spoke in favor of a bond issue that would raise taxes for a new school. He invoked the Sullivan County Citizens for Tax Relief in support of the bond issue: "Their goal is property tax relief. The goal is more than just cutting budgets. The goal is to make city hall and the schools more efficient. An environment must be created that will increase the tax base and the average pay of a worker in Claremont. Part of this will be accomplished by having an efficient education system....The school facilities will play a major role in attracting new business to Claremont" (p. 4).
The other side of the positive effect of school spending on house values is the negative effect of higher property taxes on house values. The adverse effect of taxes needed to finance more school spending gives all voters--those with school children as well as those without--an incentive to make sure that school taxes are efficiently used. A proposed school tax increase that would not improve school quality (as judged by local voters) would have a net negative effect on housing values. The higher tax rates would reduce home values, but the reduction would not be offset by higher quality schools.
The process by which local voters and their local representatives evaluate the efficacy of school spending also explains why school expenditures per pupil can vary across the state. Costs of educational inputs and the efficacy of various educational measures can vary, and these costs and benefits are often comprehensible only to those who live in the district. The outcome of a perfectly efficient balancing of educational costs with its benefits would, under such conditions, always result in variation in per pupil expenditures.
5. Shifting school finance to the state undermines efficiency and educational quality.
The balancing of the two capitalization effects--more school spending raises house values; higher taxes reduce house values--offers an explanation for why New Hampshire public schools perform so well on national tests (section 7 below). The capitalization effects are effective in New Hampshire because the state relies on the property tax to finance most school expenditures. Local voters perceive both the benefits and burdens of higher and lower spending decisions, and they and their representatives on school boards have a strong incentive to weed out inefficient spending proposals from those that are effective.
The administration of the property tax also promotes efficient decisions by voters. When a local school district decides to raise its budget by ten percent, it is easy for each voter to calculate how much additional taxes he or she will pay. Voters can then decide whether these higher taxes are worth enduring in light of the improvements in the school system. Such a mechanism is almost entirely absent at the state level. The impact of higher state taxes for a specific school spending proposal is harder to calculate, and, even if they were calculated, the benefits of the proposal have little impact on voters who do not have children in public schools. While the latter voters may support education in principle, only the local property tax provides them with the additional reason to support efficient spending proposals: It will increase their own wealth.
Removing responsibility for raising school revenues from the local property tax to a statewide tax (regardless of whether such a tax is on sales, income or property) removes the connection between school taxes and house values, because the tax revenues would no longer be controlled by locally-elected school boards. Local control requires control over one's own tax base. [footnote 9] As a political scientist who served on a school board reported, "The effective place for citizen control is the budget" (Pomper 1984, p. 222). Without such control, he pointed out, school administrators cannot be prodded to make rational choices between a new math course and an additional secretary for the principal's office. A locally elected school board, responsible to the taxpaying electorate, can best monitor school administrators who are naturally loathe to deny any spending request within their schools.
Capitalization of school quality in house values occurs only under certain circumstances. Potential buyers of the houses must have several choices of school districts for them to register their preferences for houses in one district over another. If schools were the same in every community, or if there were only one community in which buyers could live, there would be no reason to expect that the quality of that single school system would have any effect on home values. A necessary condition for capitalization is that home buyers can select among communities whose public services match up with their varied preferences.
Statewide financing of schools greatly reduces the capitalization effect. While certain differences in schools may persist under full state funding because of differences in the population, the essential link between local taxes and local spending is broken. Local voters can no longer expect to benefit from improved schools through increases in their property values. In contrast to the large number of economic studies that find capitalization of local school quality measures, there are no studies that find interstate capitalization of school spending or test scores.
6. State aid to local districts should focus on school spending, not property tax relief.
The capitalization of both school benefits and property tax costs means that buyers of homes in any community have to pay for what they get. Their tax burdens are always "tolerable" because their mortgage payments offset any unusually high tax payments. The same phenomenon also frustrates attempts to make communities "richer" by offering them state aid. Increases in state aid to districts will be capitalized in housing prices, making them more expensive to live in for potential buyers. A new state aid program provides a one-time benefit for owners of property in the favored districts. Subsequent buyers get no special benefits, because they have to pay more for their housing (Bogart, Bradford and Williams 1992).
Simply because capitalization frustrates redistribution of income, however, does not deny a role for state aid to school districts. Generally speaking, districts whose residents have low incomes tend to spend less on schools, and state aid is one way to induce greater spending than they might otherwise choose. Likewise, having a large amount of commercial and industrial property often makes it less costly for residents to raise taxes.[footnote 10] If a quarter of the taxes are paid by owners of nonresidential property, the residents will perceive that a $1 increase in local school spending will cost them only 75 cents. In economists' jargon, both state aid (if offered as a matching grant whose receipt depends on additional local revenues) and nonresidential property lower the "tax price" of local schools.[footnote 11] Economists generally predict that lower prices induce people to buy more of any good, and schools are no exception.
Once benefits of lower tax prices (via nonresidential property or state aid) are capitalized in house prices, home buyers get what they pay for. As a result, even in districts with substantial amounts of nonresidential property, homeowner-voters will still perceive a trade-off that affects their property values. If they are presented with a proposal to increase school spending and school taxes, they will consider whether the proposal raises or lowers their property values. Inefficient proposals will lower their home values and will be opposed, even if much of the taxes are paid by nonresidents. A state that wished to conserve on state funds while stimulating local school spending would want to focus on two characteristics of communities. One would be low incomes of residents, and the other would be lack of nonresidential tax base. A combination of these conditions could be associated with low school spending. The reason is not that property taxes are too high in such places, since high taxes are offset by lower housing costs, but rather that personal incomes are low.
Because many low-income communities in New Hampshire are in a position to tax nonresidential property, their spending on education can rival or exceed that of towns with above-average incomes. Thus, the major issue that the state may want to deal with is the spending of communities with disproportionately large numbers of low income people that lack a nonresidential tax base.
There is need for caution in supplying state aid, however. Katherine Bradbury undertook a study of the effects of state aid on the spending levels of Massachusetts and Rhode Island school districts in 1990-91. She concluded: "Although school aid funds must be spent on schools, districts can implicitly divert some of the revenue to other public or private purposes by choosing to raise less money for schools from local sources than they would in the absence of the aid" (Bradbury 1994, p. 52). Her statistical examination of Massachusetts in 1991 found that only about a third of the aid given to property-poor districts found its way into increased school spending. The rest went to lower property taxes (p. 56). In Rhode Island, whose state-aid formula was designed to stimulate spending, Bradbury concluded that only about two-fifths of the state money found its way into increased school spending (p. 58). The rest went for lower local taxes.
These examples suggest that state aid should be offered selectively so as not to turn into tax relief. Tax relief may be desired by local voters because it would make their homes more valuable, but it would not improve the quality of education. (The local tax relief would, of course, be offset by higher state taxes, so there is no reason to believe that overall tax burdens would decline.)
New Hampshire's state aid formula is targeted in such a way that low income communities with small nonresidential tax bases (per student) get the largest amount of aid. The state's formula does not specifically single out nonresidential property as a criterion for aid. It instead uses both per capita income and equalized valuation per weighted pupil. Equalized valuation, however, consists of the sum of residential and nonresidential property. Because low income people typically live in lower-valued houses, the equalized property valuation per student is large only in communities with either high income households (who live in higher-valued houses) or with large amounts of nonresidential property. Aid is denied communities that have the wherewithal to pay for education by either high personal incomes or low tax prices. This is an efficient way to allocate school aid insofar as it does not undermine local incentives to provide higher-quality schools from their own sources. One reason that New Hampshire has relatively little need for state aid is because its communities are large enough in area to encompass a wide range of land uses. Some states, such as New Jersey, have developed local governments on very small land areas. When these small municipalities become inhabited by low-income people, there is little area left to attract nonresidential property to lower tax prices. This problem has not arisen in New Hampshire because the towns and cities have a substantial land area to accommodate a variety of land uses. For this reason, most low-income people in New Hampshire live in communities that contain both higher-income sections of town and/or substantial commercial and industrial tax base. Most students from low-income families thus do not need state aid for schools.
7. Comparisons with other states indicate that New Hampshire public schools are successful.
This section shows that New Hampshire does well in the level and growth of school financing and that New Hampshire students do well in comparison to students in other states. The next section will offer evidence that an important reason for New Hampshire's success is its extensive reliance on local property taxation. This evidence comes from states which have shifted away from local property taxation to state funding of education.
By any measure of support for education, New Hampshire does well when compared with other states according to the Report Card on American Education (Feistritzer 1993).[footnote 12] In 1992-93, the state spent $5653 per pupil, which was 10 percent above the national average, and which ranked it 21st among the states (Table 15, p. 38). More impressive is the relative gain in spending. Between the school years 1972-73 and 1992-93, New Hampshire spending per pupil increased by 542 percent, the 8th largest increase among all of the states (Id.).
New Hampshire's twenty-year growth in per pupil spending was accomplished largely by property taxation. In 1972-73, New Hampshire schools got 87.5 percent of their funds from property taxes. By 1992-93, the corresponding figure was 89.0 percent (Feistritzer 1993, Table 25, p. 48). By this measure, the property tax is a more than adequate source to finance growth in educational funding.
New Hampshire' growth in spending over the last two decades is even more remarkable when its 8 percent enrollment increase during that period is taken into account. Enrollment growth by itself tends to decrease spending per student (Gustman and Pidot 1973, p. 13; Silva and Sonstelie 1995, p. 210). The seven states whose growth in spending per pupil exceeded that of New Hampshire all had substantial declines in enrollment (Feistritzer 1993, Table 27, p. 53) .[footnote 13] New Hampshire managed to run counter to the usual trend of smaller increases in spending per pupil when there are increasing enrollments.
Interstate measures of educational accomplishment are difficult to make because of the lack of a consistent standardized test. Simple comparisons by the most widely taken test, the Scholastic Aptitude Test (SAT), are flawed because high participation rates in some states (such as New Hampshire) tend to reduce the states' average scores. However, recent econometric studies have adjusted for participation rates and other differences among states. Once these factors are taken into account by multiple regression analysis, New Hampshire seniors in both public and private schools rank very high.
Two studies in the Economics of Education Review (1993) used statistical techniques to control for the participation rate bias and other sources of interstate differences. The new studies arrived at an adjusted ranking of states according to SAT scores. After accounting for the characteristics of individual test takers such as race, gender, family income, and parental education, Amy Graham and Thomas Husted (1993, p. 201) ranked New Hampshire eighth on combined SAT math and verbal scores among the 38 states in their sample.
The other study ranked New Hampshire students even higher. After accounting for differences among states such as teacher salaries, spending per pupil, race, household size, poverty rates, class size, and the percent of private school students, Mark Dynarski and Philip Gleason (1993, p. 208) ranked New Hampshire third in SAT math scores in a sample of 38 states and the District of Columbia. Results from the National Assessment of Educational Progress in 1992 and 1994 rank New Hampshire public school students similarly high (Mullis et al. 1993; Williams et al. 1995). New Hampshire's system of educational finance supports public schools whose students compete successfully on academic grounds with students from other states.
8. Court-induced shifts from local to state funding may decrease support for education.
A method of evaluating the adequacy of New Hampshire's school finance system is to examine the experience of other states. Courts in other states have ruled on the issue of the constitutional validity of reliance of local property taxes to fund education. Because these suits began in the early 1970s, social scientists have had an opportunity to examine the "laboratory of the states," as Justice Brandeis put it. This section addresses what social scientists have learned from states in which courts have caused the state to assume a larger share of funding for education.
The most widely agreed-upon point is that plaintiff victories in school-finance litigation have shifted the financing of schools to the state level and reduced the local government's share. The experience in other states shows that this shift does not necessarily increase total resources available for education. Bahl, Sjoquist, and Williams (1990) found that eight of the nine (at that time) court rulings that overturned existing school-finance systems had resulted in at least a 10 percentage point increase in the share of funding coming from state sources (p. 165). However, they found that the increased state share always reduced reliance on local taxes (p. 169). Shifting to state revenues came at the expense of local financing.
The state whose court undertook the earliest and most sweeping remedy is California. As a result of the Serrano decisions,[footnote 14] California moved from a system in which a majority the funds came from local government in 1970-71 to one in which "school district revenues have been determined almost entirely in Sacramento" (Picus 1991, p. 52). Although about a quarter of funds still come from property taxes, California analysts regard these funds as largely controlled by the state legislature (O'Sullivan, Sexton and Sheffrin 1995, p. 139). California is a good example of what happens when responsibility for school finance is shifted almost entirely to the state government.
By almost every measure, California's school system has declined in quality and is failing low-income students especially. Among the chroniclers of California's failure was Jonathan Kozol, who has been an advocate of school finance litigation. He wrote in his 1991 book: "Today, in all but 5 percent of California districts, funding levels are within $300 of each other. Although, in this respect, the plaintiffs [in Serrano] won the equity they sought, it is to some extent a victory of losers. Though the state ranks eighth in per capita income in the nation, the share of its income that now goes to public education is a meager 3.8 percent--placing California forty-sixth among the 50 states. Its average class size is the largest in the nation" (Kozol 1991, p. 221).
California's relative decline among the states is sometimes said to be due to Proposition 13, the 1978 voter initiative. Proposition 13 required that the rate of tax on all property throughout California had to be set at no more than one percent of its 1975 value, with no more than a 2 percent annual assessment increase, except that property could be reassessed to market value upon sale. This created a statewide, uniform tax rate applicable to every school district in the state. A thorough economic examination of its consequences found that "Proposition 13 has effectively converted the property tax from a local to a statewide tax. This process actually began with a series of court decisions concerning equalization of school finance across districts, and indeed some commentators have suggested that the loss of control of local school financing was the ultimate cause of Proposition 13" (O'Sullivan, Sexton and Sheffrin 1995, p. 137).
According to almost all observers, a drastic decline in support for public schools followed Proposition 13. Nothing in Proposition 13, however, prevented the California state legislature from raising other taxes to make up for the loss in local school revenues that occurred in almost all jurisdictions. California's experience shows that a statewide property tax (the result of Proposition 13) does not necessarily provide greater support for education and that other state taxes cannot be relied upon to replace its shortfall. Econometric work by Fabio Silva and Jon Sonstelie (1995) confirmed that half of California's relative decline in school spending in the 1970s and 1980s could be attributed to the state's response to Serrano v. Priest, not simply to Proposition 13. There is no reason to expect that overall support for education will increase if school finance is shifted from local to statewide sources of revenue.[footnote 15]
9. Centralized financing has not improved education in other states.
The previous section established that court decisions favoring plaintiffs can result in a substantial declines in overall support for education. It might be asked, however, if educational quality has changed in response to a shift from local to state funding, whether as a result of litigation or legislation. The petitioners regard the large share of education financed by local property taxes as prima facie evidence of educational inadequacy.[footnote 16] One would like to know if states with larger state shares do better.
The most important point is the complete absence of evidence that educational quality has improved in states whose financing of education has become more centralized. The petitioners cite no studies to demonstrate that shifting education finance from local taxes to statewide taxes improves educational outcomes, as measured by test scores, dropout rates, graduation rates, college attendance, or job prospects.
Perusal of the Journal of Education Finance for the last ten years reveals no articles that purport to show that average test scores or other indicators of measurable educational quality have improved as state revenues replaced local revenues. There was no of lack of articles examining the distributional results of school finance reforms. Almost every issue contains at least one article offering a perspective on various experiences with an increased state role in school finance. None of those articles reported that test scores or other statewide indicators of educational quality has risen as a result. A ten-year search of another journal in this area, the Economics of Education Review, was similarly fruitless.[footnote 17]
Among the few articles in the literature to have examined the relationship between state funding and school quality issue is Sam Peltzman (1993). He found a modest but statistically significant relationship between 1972-1981 increases in the state's share of funding for education and declines in statewide SAT scores (pp. 353, 355). Another study by Thomas Husted and Lawrence Kenny (1995) indicates that states with a larger fraction of education financed by the state had lower SAT scores. They found that states whose supreme courts had previously ordered reforms had especially low scores (p. 13). Although these two studies do not confine their analysis to states with court-ordered reforms, they do indicate that the higher state share desired by the petitioners does not necessarily lead to improved education.
The State of Hawaii offers another experience. Its schools have long been financed entirely by the state, and the school system is regarded as one big district. In comparing Hawaii's system to other states, Hawaii-based John Thompson (1986, p. 293) concluded that "...there seems to be little evidence that full state funding provides more money for the school system than when significant proportions are provided by locally raised taxes....Also, student achievement does not appear to be affected by either the funding or the statewide organization of the schools." In a reconsideration of Hawaii in a later article, Thompson was even less sanguine about student achievement. He reported that Hawaii's public school students performed well below the average for most other states in standardized mathematics tests (1992, pp. 299-300). 10. Petitioners' experts ignore social science methodology as well as economic factors favoring the local property tax.
This section comments on the petitioners' expert reports. The method by which the petitioners' experts seek to show that New Hampshire's school financing is inadequate is not in accord with generally accepted social science methods. A necessary condition for any science is that its results can be replicated. That means that another scientist with similar training could examine the data and come to the same conclusion. The comparison of different schools by Mueller and Schultz uses methods that are entirely subjective. Only by chance could another social scientist come up with the same pattern of "+", "++", "-", and "- -" that he attached to various aspects of schools.[footnote 18]
This is not to say that Mueller and Schultz misreported their impressions. It is to say that such impressions cannot be called scientific. As an editorial board member of two social science journals and a referee for many others, I have never seen a submission, let alone a published article, based on such subjective methods.
A second aspect of the petitioners' endeavor that is scientifically suspect is its choice of sample. Petitioners attempt to extrapolate their comparisons of the five plaintiff districts and the five "paired" districts to the entire state. Aside from the subjective basis of the comparisons, the selection of just one comparison district for each plaintiff district is inadequate to support their conclusions. One would need to know if the fiscal differences one sees between, say, Claremont and Lebanon are unique to Lebanon, or whether they would apply to other districts as well. To draw conclusions from such a limited sample would be like reporting that five supposedly homeless people actually have places to stay and thereby concluding that homelessness is a myth.
Because districts differ in many respects that might affect their school spending, one would need a large sample of districts and apply modern statistical methods to determine whether differences in spending are the result of differences in personal income, the composition of the tax base, or factors not related at all to fiscal differences. The petitioners' experts have not controlled for other sources of differences between the districts. As a result, the conclusions that they reach can be regarded as no more than anecdotes. Anecdotes can be useful to illustrate quantitative evidence, but they cannot by themselves have probative value.
A related offense is the display of data in the Alexander/Salmon report which purports to show differences in per pupil spending by district (Chart II). The segregation of Nashua and Manchester, the two districts containing one of every seven students in the state, is without any scientific justification. Likewise, their bar graphs in Chart III showing "State and Local Revenue Per Pupil," where the beginning point is $2000 rather than zero, might well be taken from the pages of the old manual, How to Lie with Statistics.[footnote 19] To systematically disregard critical information and to display data in ways that are misleading is not consistent with disinterested scientific inquiry. The petitioners' experts also make claims about economic phenomena without undertaking any economic analysis. They claim that tax rates measure economic burdens and that tax base is the main measure of school finance resources. Yet they admit to undertaking no economic analysis.[footnote 20] The economic analysis in the present report shows that the petitioners' experts focus on these variables is misleading. Both economic theory and statistical evidence from professional, peer-reviewed economic research show that tax rates and tax bases are not valid measures of economic burdens or resources. For this reason, the tabular comparisons of differences in tax base in sections II and III of the Alexander/Salmon report are meaningless. They do not show what they purport to show. The comparisons of districts arrayed according to size (Chart III) show "wealth" and "tax effort" without any accounting of the differences in house prices that result.
11. Conclusion: New Hampshire chooses the property tax for sound economic reasons.
In a recent article, Campbell and Fischel (1995) show that New Hampshire voters have in recent years been presented with alternatives to the local property tax as a means of funding education. The voters have rejected them. The present report concludes that New Hampshire's revealed preference for a system of school finance that relies largely on local property taxation is consistent with sound economic reasons. The supposed inequities of school property tax rates are eliminated by capitalization. Supposedly burdensome tax rates are offset by lower housing costs, and the alleged benefits of low tax rates are offset by higher housing costs.
The same capitalization of taxes in home values induces voters and their representatives to govern their local schools in economically efficient ways. A desire to increase their home values encourages voters to accept school expenditures that are cost-effective and reject those that are wasteful. A state-funded system necessarily foregoes this efficient incentive.
New Hampshire voters and legislators can rationally look at other states that have shifted from local property taxation to statewide funding of education. The nation's two-decade experience with court-induced reforms shows that support for education does not necessarily increase when funding is shifted to the state. In several states, financial support for schools has declined after local taxation was supplanted by the state. In no state is there any evidence that the quality of education has improved as a result of statewide financing, and several studies point to lower test scores in states with centrally-financed schools. The available national comparisons indicate that New Hampshire has done exceptionally well for its public school students.
Appendix A
A Numerical Example of Capitalization.
Here is how tax capitalization works. Suppose there are two taxing districts A and B, each with a single household.[footnote 21] The households in A and B each have one child in the school system. The two taxing districts share the same school, which costs $5000 per year per pupil. Each household owns a house that is initially worth $100,000. The mortgage rate (here to include all costs of ownership except property taxes) is 10 percent, so that each house pays the bank $10,000 annually. The tax rate in each district is $5000/$100,000 = .05 percent. The total annual cost of owning a house in each community is thus $15,000, which is the sum of mortgage payments ($10,000) plus annual property taxes ($5000).
Now let community A add to its tax base a $100,000 office building.[footnote 22] The building imposes no fiscal costs and has no adverse environmental effects (pollution, traffic, crime) on the house that would detract from the house value. Thus the tax base initially (before market adjustment) swells to $200,000 in district A, and it would appear that the tax rate would fall to .025 (=$5000/$200,000) and the tax payment of the household in A would fall to $2500 per year.
But this cannot last in an efficient housing market. When the house in district A is sold, it will have to go up in price. Buyers will see that district A has a lower tax rate and hence a lower tax payment than that of an identical house with identical schools in district B. Competition among buyers will bid up the price of the house in district A until the financial advantages of living in A are the same as in B. (It is assumed that there are many other districts just like B in this situation, so house value in B does not fall.) How high will the price of the house in A rise? It will go up to $122,475 in this example. At this value, the total tax base in A will be $222,475 (that is, the value of the house, $122,475 plus the value of the $100,000 office building). With this tax base, the tax rate for schools in A will be $5000/$222,475 = .022474. This tax rate times the value of the house (i.e., .022474 x $122,475) will yield an annual property tax payment by the household in A of $2752.50. The 10 percent mortgage on the house in A will cost the homeowner $12,247.50 annually. The sum of these two unwieldy numbers ($2752.50+$12,247.50) is $15,000.
One might reasonably ask how the number $122,475 was selected as the purchase value of the house in A. It was chosen by algebraic calculation precisely to yield the result that the sum of taxes and mortgages will be the same in each community.[footnote 23] This might seem like it is cooking the result, but it is not. Participants in the housing market will, by trial and error, eventually settle on this number. If the price of the house in A were higher than $122,475, then too much would be paid in mortgage payments plus taxes in A. As a result, buyers would select instead the house in community B, and the price in A would have to fall. If the house price in A were much lower than $122,475, then the sum of mortgage and taxes would be lower than in B, and buyers would flock towards A, thus bidding up the price of the house there. The sum $122,475 is thus the market-clearing equilibrium price in this situation.[footnote 24] It gets there because of the economic behavior of house buyers, not because the algebra happens to set it there.
Reflect now on how the two communities differ after market equilibrium is achieved. In district B, the tax rate is still 5 percent. In district A, the tax rate is only about 2.25 percent. In district B, the household's tax payment is still $5000, while in district A, the household's tax payment is only $2752.50. Yet the financial cost of being a resident in either community and sending one's children to school is the same. In B, the financial cost is the annual taxes plus annual mortgage payments: $5000 + $10,000 = $15,000. In A, the financial cost is the annual taxes plus annual mortgage payments: $2752.50 + $12,247.50 = $15,000. Consider the position of the petitioners' experts. They see districts A and B as highly unequal because tax rates are lower in A. They claim that the ability of households to pay taxes varies in direct proportion to their district's tax base. With a tax base of $222,475 per pupil, the resident in district A seems more than twice as able to finance school spending as those in B. If a prospective resident were to ask the petitioners' experts which district he should buy a house in, they would obviously steer the buyer towards A.
But the petitioners' experts would be wrong because they neglect the workings of the property market. There are no $50 bills lying around for people to pick up. The prospective home buyer would see that tax rates in A were lower, and that tax payments on the house were also lower. But he would also see that the annual mortgage payments he would have to make in A were considerably larger than on the house in B. If he did buy the "tax bargain" in A, he would find he was house-poor because of exactly-offsetting, higher mortgage payments. Thus he might as well choose the house in district B. The taxes are higher in B, but he can afford them because the mortgage payment in B is lower. He has the ability to pay the higher taxes in B because of the lower house price, with its lower annual mortgage payments.
Appendix B
Capitalization Illustrated by Analogy.
Here is how capitalization works. Suppose you own a single-family home. The value of the home will reflect the services that it is expected to provide. If it has eight rooms and a half-acre lot, the value will reflect not simply their present existence but the expectation that such things will persist into the future. Future persistence of the qualities of a house is critical to present value. An eight-room home that is owned in fee simple may be valued at $150,000. The value of a leasehold to the same house will have a negligible value if the leasehold is expected to terminate tomorrow. Expected future uses are critical to capitalized value.
When some event adds to or reduces the value of a house (or any other durable asset), economists say the event is "capitalized" in the value of the house. Suppose it were just discovered that the newly built house had a high-quality, reliable artesian well that made connection with an off-site water supply unnecessary. The free water could be used instead of the town or commercially-supplied water, and the owner would not have to pay water bills. If the expected water bills would have been $250 per year, the present value of the free well would be on the order $5000.[footnote 25] Thus a buyer of the house would not pay $150,000, but the higher price of $155,000, reflecting the special value of the well that was attached to the house. The expected benefits of the well are said to be "capitalized" into the price of the house.
Consider now a public service instead of a private well. Suppose there are two adjacent towns (dubbed "Paytown" and "Freetown," to keep them straight) with public libraries of equal quality. In Paytown, the library is funded by a property tax levy that amounts to $250 per year on the average house. In Freetown, a local philanthropist has built the library and granted the town an endowment to pay for maintenance and additions to the collection. In Freetown, the reduced taxes that resulted from the philanthropist's gift will be capitalized in the value of the houses. A house that would sell for $150,000 in Paytown (the town whose library required continuous public tax support) would sell in Freetown for $155,000. The present value of $250 annual tax saving is about $5000, so the average house value in Freetown would rise by $5000.
This example illustrates a simple lesson about the ability to pay for public services. Those who financially gained from the philanthropist's gift of the library were people who owned homes at the time the gift to the library became known. Later buyers of homes surely will enjoy a fine, "free" library, but they will have paid for that enjoyment in advance in the form of higher housing prices. To vary an economics saying, there is no such thing as a free library. Now look at the position of homeowners in Paytown, whose library is tax supported. They have to pay $250 per year for the same level of library services as in nearby Freetown. But the Paytown residents have paid $5000 less for their homes in the first place. As a result, they have more money left in the bank (or more in their paycheck after the monthly mortgage payment) with which to pay the taxes. Because the higher property taxes were known when they purchased their homes, they are on no worse financial footing--and have no less ability to pay for libraries--than those who bought homes in the community with the "free" library.
The library example can be extended to public schools and the local taxes needed to pay for them. The school tax base will be dealt with in the balance of this section. Suppose two towns ("Riverside" and "Hilltop") share the same schools but have different tax bases. The cooperative agreement between the two towns allocates the cost of schooling on a per pupil basis rather than on a combined property valuation basis. Riverside has a power dam that constitutes about half of its tax base and requires no offsetting municipal expenditures. The rest of Riverside's tax base, and all of Hilltop's tax base, is composed of owner-occupied housing [of equal value], and public school populations are the same in each town. Riverside will have a lower school property tax rate (because the power dam pays about half of it), and residents of similar houses in Riverside will be subject to a lower property tax rate than residents of Hilltop.
This example looks like the classic fiscal disparity that the petitioners would like the court to cure. On its face, it looks like homeowners in Hilltop are paying twice as much for the same schools. But this is wrong once capitalization is considered.[footnote 26] Prospective buyers of homes in the two towns are aware of the differences in taxes on their properties. For an otherwise similar home, taxes in Riverside are less than in Hilltop. As the result of competition among buyers, the buyer in Riverside must pay the capitalized value of the difference for her home in Riverside than for an otherwise similar home in Hilltop.
On the other side of the coin is the buyer of the house in Hilltop, the one with the higher tax rate. He paid less for his house than the buyer of the house in Riverside. Thus he has a lower annual mortgage payment to make. He can use this saving in housing costs to pay his higher annual property taxes. This illustrates the primary implication of the principle of tax capitalization: The sum of housing costs and property tax payments will be the same in both communities.
Appendix C
The New Hampshire Capitalization Regression.
This appendix describes the method by which it was established that capitalization, as described in sections 2 and 3 above and in Appendices A and B, occurs in New Hampshire school districts. The sample consists of the 73 New Hampshire towns and cities whose population was at least 2500 in 1990 and which were not part of an elementary-school cooperative school district. The minimum population was necessary because U.S. Census data for the median number of rooms in owner-occupied housing were not available for places whose population was smaller than 2500. Cooperative school districts were excluded because test scores might reflect students from other towns and because financial arrangements among cooperative districts are complex. These exclusions do not mean that the conclusions of this study would not apply to smaller towns and cooperative districts.
The statistical technique is linear regression. In this method, variations in the dependent variable (the 1990 median value of owner-occupied homes in a district) are accounted for by variations in independent variables. The independent variables in the regression are: school mill = the school tax rate per $1000 equalized value for the district for the school year 1990-91. Source: New Hampshire State Department of Education, Computer and Statistical Services, Fact Sheet CAT sum = the sum the two elements of each district's "Total Battery" percentile scores on the California Achievement Test given to fourth-graders in the district in the school year 1990-91. These consist of the "ANP" or anticipated national percentile and the "NP" or actual national percentile. New Hampshire State Department of Education, New Hampshire Statewide Testing Program, Summary of California Achievement Test Program for 1990-91 School Year.
rooms = median number of rooms in owner-occupied houses from the 1990 Census. Source: U.S. Bureau of the Census, 1990 Census of Population and Housing, New Hampshire.
miles N = airline distance in miles from town or city to a single point on a map north of Boston, Mass. (approximately at the intersection of I-93 and I-95). Source: measured from map of New Hampshire.
house age = median age in years of houses in the community in 1990, from the U.S. Census. Source: U.S. Bureau of the Census, 1990 Census of Population and Housing, New Hampshire.
The results of the regression show that the independent variables (including the intercept) account for 86 percent of the variation in home values in the sample. This is inferred from the statistic "R-Square," which is a commonly used summary measure. An R-Square of 0.86 indicates a very good fit. The highest possible value is 1.00, which is a perfect fit, and the lowest possible value is 0.00, which would indicate no relation at all between the independent variables and the median value of homes.
The coefficients are estimates of how much the independent variables affect home values, and the P-values, which are derived from the t-statistics, measure the confidence with which one can be sure that each coefficient is greater than zero. A P-value of .05 or lower is regarded as "statistically significant" in empirical studies. All of the coefficients in this regression are significant at least at the 5 percent level of confidence.
The variables most relevant to the present study are school mill and CAT sum. Their estimated coefficients, evaluated at the mean of the sample, imply the following: A one point increase in CAT sum raises the value of a house by $229.85 and a one point increase in school mill lowers the value of a house by $2685.23. Both variables are highly significant and accord with those of other studies discussed in section 3 above.
The estimates herein show with a high degree of confidence that school tax rates and test scores are capitalized in the value of owner occupied housing values. The degree of tax capitalization by this estimate is nearly 100 percent. A one mill increase on the value of the sample's mean-value house ($131,401) would annually yield extra taxes of $131.40. If a discount rate of 5 percent is applied to $131.40 over an indefinite time horizon (see footnote 25 in Appendix B above), the present value of the extra taxes is $2628. This is 98 percent of the estimated coefficient ($2685.23), which implies that capitalization is almost complete.
The other variables are included to control for other factors that influence housing values in New Hampshire and which have been used in studies similar to this one. All of these variables are significant.
The rooms variable is a measure of the size of the house. It indicates that at the mean, an additional room would add $41,331.60 in value to a house. This may overstate the influence of rooms, since larger houses are often on larger lots, for which no data were available.
The variable miles N (airline distance from the Boston area) was negative and significant. It indicates that, other things equal, homes in the southeastern part of the state are more valuable than in the northern part. This accords with urban economic theory, which holds that people will pay more to live in places where there is a higher density of jobs and closer proximity to work. The southeastern part of New Hampshire is where most of the employment in the state is located, and it offers most convenient access to jobs and services in the Boston area. The large coefficient of -$372.89 per mile may also reflect the effects of the Boston area's unusual rise in prices in the 1980s, which spilled over to southern New Hampshire.
The variable house age is a proxy for housing depreciation and obsolescence. Except for those with special antique appeal, older houses generally sell for less. In this case, the estimate suggests that at the mean of the sample, an additional year of age subtracts $535.38 from the average house's value.
Regression Results:
Dependent variable: Median Value of Owner-Occupied Housing in 1990.
Regression Statistics:
Dependent variable mean value 131401.37
R-Square 0.862695
Number of Observations 73
variable names Coefficients standard error t -statistic P-value var. mean school mill -2685.23 447.04879 -6.006573 7.1E-08 12.8509 CAT sum 229.8488 109.10801 2.106617 0.03864 111.657 rooms 41331.6 3340.971 12.371135 1.7E-19 6.105 miles N -373.885 71.516556 -5.227945 1.6E-06 54.72 house age -535.376 208.22407 -2.571151 0.0122 27.287 Intercept -77034.1 20858.4 -3.693191 0.00043
Explanations of other recorded variables: The intercept is the value the dependent variable would take if all of the dependent variables were zero. It is purely an artificial construct in this instance, so its negative value has no special meaning. The standard error of the coefficients summarizes how far the estimated values are from the true values. The coefficient divided by the standard error gives the t-statistic. The variable means are for the towns and cities in the sample, not for the state as a whole.Data for the Regression study:
Town House Val rooms CAT sum miles N school mill house age
Allenstown town 107400 5.2 102 48 17.48 21 Alton town 127500 5.8 118 63 5.59 23 Amherst town 197700 7.7 134 44 14.83 22 Auburn town 139500 6.4 131 38 14.15 22 Barnstead town 98800 5.7 98 56 14.68 22 Barrington town 121400 5.6 94 44 15.44 20 Bedford town 199500 7.6 123 41 12.42 19 Berlin city 67700 5.7 117 130 11.65 56 Bow town 164900 7.1 126 51 11.84 20 Brentwood town 169400 6.7 118 29 16.66 24 Candia town 135700 6.3 110 39 16.59 23 Chester town 149600 6.6 120 32.5 13.69 23 Chesterfield town 134100 6.1 109 85 9.6 28 Claremont city 82100 5.9 89 91.5 15.18 44 Concord city 112400 6 115 54 14.64 32 Conway town 116500 5.6 114 95.5 9 23 Deerfield town 139400 6.3 105 42 15.33 22 Dover city 119900 6 112 41.5 11.43 37 Epping town 123300 5.6 99 32.5 16.55 22 Epsom town 115100 5.8 110 48 12.57 20 Exeter town 154000 5.8 140 27 11.99 31 Farmington town 91500 5.7 82 56 10.36 31 Franklin city 90000 6 79 70 13.2 43 Fremont town 133600 5.9 121 31 13.93 18 Gilmanton town 106900 5.8 95 64.5 13.23 21 Gorham town 78500 5.5 138 123 10.97 35 Greenland town 168100 6.5 113 31 8.87 26 Hampstead town 185300 6.7 143 26 12.88 16 Hampton town 162500 6.2 118 24 7.56 29 Hanover town 215500 7.4 139 104 10.96 35 Haverhill town 81200 6 83 115 17.27 53 Henniker town 124000 6.1 111 63 14.78 22 Hinsdale town 89800 5.5 91 84 15.73 27 Hollis town 229800 7.5 128 40 11.06 20 Hooksett town 128400 5.8 118 46.5 10.61 20 Hopkinton town 144900 6.9 130 59 15.47 27 Hudson town 145000 6.3 133 32 12.4 18 Keene city 113000 6.2 116 76 15.66 41 Laconia city 108500 6.2 99 71.5 10.15 44 Lebanon city 117200 6.1 113 100 16.38 38 Litchfield town 147500 6.5 116 36 18.39 19 Littleton town 90200 5.9 87 126 12.08 40 Londonderry town 152900 6.6 139 32 17.46 18 Manchester city 118600 5.9 99 40 11.77 42 Merrimack town 139700 6.4 121 37 11.4 17 Milford town 137600 6 124 44 12.02 22 Milton town 101700 5.4 105 58 16.22 24 Moultonborough 145200 5.8 99 84 4.19 22 Nashua city 138800 6.1 114 33 10.81 26 New Boston town 148700 6.3 102 50 12.26 15 Newmarket town 131500 5.9 112 33.5 15.35 18 Newport town 83100 5.7 83 84.5 14.74 38 North Hampton town 187400 6.6 131 27.5 8.26 32 Northwood town 117600 5.6 114 44.5 14.89 24 Nottingham town 128500 5.8 160 39 15.63 22 Pelham town 160500 6.6 109 25.5 13.94 27 Pembroke town 120000 6.1 83 49.5 13.35 25 Pittsfield town 97500 5.9 95 53 17.73 28 Plymouth town 106500 6.1 132 89.5 14.12 23 Raymond town 118600 5.4 96 35 17.91 19 Rochester city 105200 5.5 90 49.5 14.04 29 Rollinsford town 116900 6.1 104 43 11.32 43 Rye town 214100 6.8 114 30 4.69 41 Salem town 150300 6.1 119 24 9.97 27 Seabrook town 145500 5.1 98 20 1.5 23 Somersworth city 109600 5.7 89 46 15.27 33 Strafford town 122500 6.1 113 48 18.58 22 Stratham town 177700 6.6 119 31 14.23 13 Sunapee town 137300 5.9 122 81.5 6.14 42 Wakefield town 106200 5.7 119 67 6.6 26 Weare town 124000 5.8 107 56 15.97 16 Wilton town 134400 6.4 117 49 11.92 34 Winchester town 86900 5.5 85 79.5 12.59 24
ReferencesFootnotes
Alexander, Kern. "The Common School Ideal and the Limits of Legislative Authority: The Kentucky Case." Harvard Journal on Legislation 28 (1991): 341-366.
Bahl, Roy, David Sjoquist, and W. Loren Williams. "School Finance Reform and Impact on Property Taxes." Proceedings of the Eighty-Third Annual Conference on Taxation (pp. 163-171). Columbus: National Tax Association-Tax Institute of America, 1990.
Blair, John P., and Samuel R. Staley. "Quality Competition and Public Schools: Further Evidence." Economics of Education Review 14 (No. 2, 1995): 193-198. Bogart, William T., David F. Bradford, and Michael G. Williams. "Incidence Effects of a State Fiscal Policy Shift: The Florio Initiatives in New Jersey." National Tax Journal 45 (December 1992): 371-388.
Bradbury, Katherine L. "School District Spending and State Aid: Why Disparities Persist." New England Economic Review (January-February 1994): 50-68. Bradbury, Katherine L., Karl E. Case, and Christopher Mayer. "School Quality, Local Budgets and Property Values." manuscript, Federal Reserve Bank of Boston, September 1995.
Campbell, Colin D., and William A. Fischel. "Preferences for School Finance Systems: Voters versus Judges." working paper, Department of Economics, Dartmouth College, October 1995. (forthcoming in the National Tax Journal.) Downes, Thomas A., and Mona P. Shah. "The Effect of School Finance Reforms on the Level and Growth of Per Pupil Expenditures." Discussion Paper 95-05, Department of Economics, Tufts University, June 1995.
Dynarski, Mark, and Philip Gleason, "Using Scholastic Aptitude Test Scores as Indicators of State Educational Performance," Economics of Education Review 12 (No. 3, 1993): 203-211.
Feistritzer, C. Emily. Report Card on American Education: A State-by-State Analysis 1972-73 to 1992-93. Washington, DC: American Legislative Exchange Council, September 1993.
Fischel, William A. "Discounting in One Lesson." The Practical Litigator 2 (September 1991): 27-36.
Goodman, Allen C. "Capitalization of Property Tax Differentials Within and Among Municipalities." Land Economics 59 (May 1983): 211-219.
Graham, Amy E., and Thomas A. Husted. "Understanding State Variations in SAT Scores." Economics of Education Review 12 (September 1993): 197-202.
Grether, David M., and Peter Mieszkowski. "Determinants of Real Estate Values." Journal of Urban Economics 1 (April 1974): 127-146.
Gurwitz, Aaron S. "School Finance Capitalization." Rand Note (N-1312FF) prepared for the Ford Foundation. Santa Monica, CA: Rand Corporation, January 1980.
Gustman, Alan L., and George B. Pidot, Jr. "Interactions between Educational Spending and Student Enrollment." Journal of Human Resources 8 (Winter 1973): 1-23.
Hamilton, Bruce W. "Capitalization and the Regressivity of the Property Tax: Empirical Evidence." National Tax Journal 32 (June Supplement 1979): 169-180.
Heinberg, J. D., and Wallace E. Oates. "The Incidence of Differential Property Taxes on Urban Housing: A Comment and Some Further Evidence." National Tax Journal 23 (March 1970): 92-98.
Holcombe, Randall G. "The Median Voter Model in Public Choice Theory." Public Choice 61 (1989): 115-125.
Hoxby, Caroline Mintner. "Markets and Schooling: The Effects of Competition from Private Schools, Competition Among Public Schools, and Teachers' Unions on Elementary and Secondary Schooling." Proceedings of the Eighty-Seventh Annual Conference on Taxation (pp. 124-141). Columbus: National Tax Association, 1994.
Huff, Darrell. How to Lie with Statistics. New York: W.W. Norton, 1954.
Husted, Thomas A., and Lawrence Kenny. "Evidence from the States on the Political and Market Determinants of Efficiency in Education." manuscript, Department of Economics, The American University, June 1995.
Jud, G. Donald. "A Further Note on Schools and Housing Values." AREUEA Journal 13 (Winter 1985): 452-462.
Jud, G. Donald, and J. M. Watts. "Schools and Housing Values." Land Economics 57 (August 1981): 459-470.
Kozol, Jonathan. Savage Inequalities. New York: Crown Publishers, 1991.
Ladd, Helen F., and John Yinger. "The Case for Equalizing Aid." National Tax Journal 47 (March, 1994): 211-224.
Li, Mingche M., and H. James Brown. "Micro-Neighborhood Externalities and Hedonic Housing Prices. Land Economics 56 (May 1980): 125-141.
McDougal, Gerald S. "Local Public Goods and Residential Property Values: Some Insights and Extensions." National Tax Journal 29 (December 1976): 436-447.
Mullis, Ina V.S., Jay R. Campbell, and Alan E. Farstrup. Executive Summary of the NAEP 1992 Reading Report Card for the Nation and the States. Washington: U.S. Department of Education, September 1993.
Musgrave, Richard A., and Peggy B. Musgrave. Public Finance in Theory and Practice, fifth edition. New York: McGraw Hill, 1989.
Oates, Wallace E. "The Effects of Property Taxes and Local Public Spending on Property Values: An Empirical Study of Tax Capitalization and the Tiebout Hypothesis." Journal of Political Economy 77 (November 1969): 957-971.
Orr, Larry L. "The Incidence of Differential Property Taxes on Urban Housing." National Tax Journal 21 (September 1968): 253-262.
O'Sullivan, Arthur, Terri A. Sexton, and Steven M. Sheffrin. Property Taxes and Tax Revolts: The Legacy of Proposition 13. Cambridge: Cambridge University Press, 1995.
Peltzman, Sam. "The Political Economy of the Decline of American Public Education." Journal of Law and Economics 36 (April 1993): 331-370.
Picus, Lawrence O. "Cadillacs or Chevrolets? The Evolution of State Control over School Finance in California." Journal of Education Finance 17 (Summer 1991): 33-59.
Pomper, Gerald M. "Practicing Political Science on a Local School Board." PS 17 (Spring 1984): 220-225.
Rasinski, Kenneth A., and Susan M Rosenbaum. "Predicting Citizen Support of Tax Increases for Education: A Comparison of Two Social Psychological Perspectives." Journal of Applied Social Psychology 17 (November 1987): 990-1006.
Reinhard, Raymond M. "Estimating Property Tax Capitalization: A Further Comment." Journal of Political Economy 89 (December 1981): 1251-1260.
Rosen, Harvey S., and David J. Fullerton. "A Note on Local Tax Rates, Public Benefit Levels, and Property Values." Journal of Political Economy 85 (April 1977): 433-440.
Silva, Fabio, and Jon Sonstelie. "Did Serrano Cause a Decline in School Spending?" National Tax Journal 48 (June 1995): 199-216.
Sonstelie, Jon C., and Paul R. Portney. "Gross Rents and Market Values: Testing the Implications of Tiebout's Hypothesis." Journal of Urban Economics 7 (January 1980): 102-118.
Sonstelie, Jon C., and Paul R. Portney. "Take the Money and Run: A Theory of Voting in Local Referenda." Journal of Urban Economics 8 (September 1980): 187-195.
Staley, Samuel R., and John P. Blair. "Institutions, Quality Competition and Public Service Provision: The Case of Public Education." Constitutional Political Economy 6 (Winter 1995): 21-33.
Theobald, Neil D., and Lawrence O. Picus. "Living With Equal Amounts of Less: Experience of States with Primarily State-Funded School Systems." Journal of Education Finance 17 (Summer, 1991): 1-6.
Thompson, John A. "Funding and Spending in Paradise: Notes on the Hawaii Model of Educational Finance." Journal of Education Finance 12 (Fall 1986): 282-294.
Thompson, John A. "Notes on the Centralization of the Funding and Governance of Education in Hawaii." Journal of Education Finance 19 (Spring 1992): 288-302.
Williams, Paul L., Clyde M. Reese, Jay R. Campbell, John Mazzeo, and Gary W. Phillips. 1994 NAEP Reading: A First Look. Washington: U.S. Department of Education, 1995.
Yinger, John, Howard S. Bloom, Axel Borsch-Supan, and Helen F. Ladd. Property Taxes and Housing Values: The Theory and Estimation of Intrajurisdictional Property Tax Capitalization. Boston: Academic Press, 1988.
- See Alexander deposition of October 26, 1995, pp. 95-100, 130 (saying equalized property tax base is measure of fiscal capacity to raise revenue), 118, 147 (equating tax effort with tax rate).
- Mortgage costs are used as an example because they are the largest single cost for most buyers. The term used in the text, housing costs, includes the annual interest payments on a loan to buy the house plus all additional economic costs entailed with a larger purchase price. These include the opportunity costs of savings used as a down payment as well as other financing costs, such as mortgage points, that rise with the price of a house. In the text, these aggregate costs will be referred to as housing costs. Housing costs do not include taxes.
- The claim of disproportionate influence by high-property-wealth districts has been made by Kern Alexander (1991), though he presented no empirical evidence for it. Campbell and Fischel (1995) have assembled evidence from New Hampshire as well as other states that shows that there is no such disproportionate influence.
- In denying compensation to a homeowner whose house was made less valuable by a change in the grade of a street, the Massachusetts Supreme Judicial Court said: "Those who purchase house lots bordering upon streets are supposed to calculate the change of such elevations and reductions as the increasing population of a city may require... and as their purchase is always voluntary, they may indemnify themselves in the price of the lot which they buy, or take the chance of future improvements, as they shall see fit. The standing laws of the land giving to surveyors the power to make these improvements, every one who purchases a lot upon the summit or on the decline of a hill, is presumed to foresee the changes which public necessity or convenience may require, and may avoid or provide against a loss." Callender v. Marsh, 18 Mass. (1 Pick.) 418, 431 (1823).
- See also Gurwitz (1980, pp. 23-24.): "Only if there is no measurable capitalization do fiscal disparities constitute prima facie evidence of horizontal taxpayer inequity."
- Having numerous independent school districts, a condition that characterizes New Hampshire, has itself been shown to improve school quality. See Hoxby (1994), Blair and Staley (1995), Staley and Blair (1995), and Husted and Kenny (1995) for studies that confirm the beneficial effect of competition among public school districts.
- The pioneering capitalization study was Wallace Oates (1969). Using a 1960 sample of northern New Jersey communities, Oates concluded that "if a community increases its tax rates and employs the receipts to improve its school system, the [statistical] coefficients indicate that the increased benefits from the expenditure side of the budget will roughly offset (or perhaps even more than offset) the depressive effect of the higher tax rates on local property values." (p. 968). Before one concludes from this that communities can spend themselves rich, it must be pointed out that Oates assumed that the increased local school expenditures were perceived by parents as cost effective. Id, p. 962, n. 7. Later studies confirmed Oates's results in California. Using a 1970 sample in San Mateo County, California, Jon Sonstelie and Paul Portney (1980 [Jan.]) found that "The annual gross rent of our median house is increased by about $52 for each additional month of average reading improvement achieved by students in the elementary school district. Each additional dollar of per-pupil expenditures on elementary education increases the annual gross rent of the median house by more than 90 cents" (p. 114). Raymond Reinhard (1981) applied an improved econometric method to the data from Oates and from Sonstelie and Portney and found even larger capitalization effects from school expenditures and test scores. Other studies finding capitalization of per pupil spending include Hamilton (1979) [Toronto area in 1961], Heinberg and Oates (1970) [Boston area in 1959], and Orr (1968) [Boston area in 1959].
- Studies finding capitalization of test scores [listed with sample area and dates] are Goodman (1983) [New Haven, Connecticut, area in late 1960s], Grether and Mieszkowski (1974) [New Haven, Connecticut, area in mid 1960s], Jud (1985) [San Francisco and Los Angeles area in 1980], Jud and Watts (1981) [Charlotte, North Carolina in 1977], Li and Brown (1980) [Boston area in 1971-72], McDougal (1976) [Los Angeles area in 1970], and Rosen and Fullerton (1977) [northern New Jersey communities in 1960 and 1970]. The positive effect of test scores on home values in New Hampshire was confirmed in the regression in Appendix C.
- An illustration of how local control is undermined by having the state take responsibility for school finance is the recent experience of New Jersey. "New Jersey officials, under court order to equalize spending between rich and poor school districts, today proposed that the State Legislature, rather than local voters, set the basic school budget for all districts." New York Times, November 22, 1995, p. A1, col 1.
- This must be qualified to the extent that commercial and industrial property require additional local municipal expenditures. In this case, the fiscal gain to the district is partly offset by the increased expenditures to service the nonresidential property or to deal with spillover effects such as traffic. In addition, the tax price effect is attenuated if higher taxes induce industry to move away.
- The tax price is a concept used by economists to characterize the financial position of the median voter (Holcombe 1989). This voter stands in the middle of the distribution of home values in the community and is regarded as the representative of the majoritarian political process. In a purely residential community, the tax price faced by the median voter is unity: A one dollar increase in local taxes for schools will cost the median voter one dollar. The only way to reduce this tax price is to provide intergovernmental subsidies or have some nonresidential property that shares in the expense but does not contribute to costs.
- This source is used here rather than the official New Hampshire statistics because Feistritzer's Report Card offers a convenient means of comparing New Hampshire to all other states. Because of differences in sources, methods and objectives, data in the Report Card may differ in some respects with those presented elsewhere by the state.
- The seven states whose spending per pupil increased faster than N.H. are listed as follows, with their 1972-1992 enrollment changes in parentheses: Kentucky (-10%), Maine (-14%), Indiana (-22%), West Virginia (-22%), New Jersey (-25%), Ohio (-27%), and Connecticut (-28%). During this same period, New Hampshire enrollments grew 8%, which was the only positive growth by any state in the northeastern United States (Feistritzer 1993, Table 27).
- Serrano v. Priest, 5 Cal. 3d 584, 487 P.2d 1241, 96 Cal. Rptr. 601 (1971) ("Serrano I"); 18 Cal. 3d 728, 557 P.2d 929, 135 Cal. Rptr. 345 (1976) ("Serrano II")
- See also Theobald and Picus (1991), who document the decline of California and Washington State school finances after court victories by plaintiffs. Recent work by Thomas Downes and Mona Shah (1995) indicates that California's response to court-ordered school finance reform is not unique. Their econometric work indicates that other states whose courts have ordered similar reforms have also had slower growth in education funds. This is not a uniform result for all court-ordered reforms. Downes and Shah concluded that events peculiar to individual states affected the outcome. Their work proves that there is no reason to believe that court-ordered reform will increase total resources available for education.
- Petitioners suggest that an important problem with New Hampshire's system in that the state's share of funding is too small, and that a desirable share would be on the order of fifty percent. (First interrogatories, #9, pp. 21-22). See also Alexander deposition of Oct. 26, 1995, p. 117 (saying that New Hampshire's 10 percent state share of funding is "irrational.")
- My investigation went back only to 1985 because one would want about ten years to evaluate the effects of reforms, and there were only a few reforms implemented prior to 1975. The Spring 1987 issue of the Journal of Education Finance was devoted to a retrospective of the field, and its survey articles also made no mention of improvements in quality indicators over the period.
- Mueller/Schultz report, June 15, 1995, pp. 31-45.
- See Darrell Huff (1954), chapter 5, "The Gee-Whiz Graph," which shows how truncating graphs can make differences seem larger than they are.
- Mueller deposition, October 7, 1995, quotes Dr. Mueller at 333: "We did no economic analysis." On p. 336, he states: "Tax rates are economic rates."
- The single household is only for expository convenience. However, it is conventional in economic characterizations of municipal finance to use a single, representative voter and taxpayer in these situations (Holcombe 1989).
- The value of the office building will not change with tax rates because it is also assumed that the $100,000 valuation was established in anticipation of the final, equilibrium tax rate.
- This involves solving for value of a house, V, in A by using two equations with two unknowns: .10V+tV=$15,000 and t=$5000/(V+$100,000) in which .10 is the mortgage interest rate, t is the tax rate in A, $15,000 is the annual cost of housing and taxes in community B, $5000 is annual school expenditures in A or B, and $100,000 is the value of the nonresidential tax base. In this particular example, the algebraic solution is V2=$15 billion, so that V=$122,475. Other numbers can be assumed for those given without changing the economic result in the text.
- Thus the hypothetical comparison in the New Hampshire School Boards Association Data Booklet (appended to Mueller/Schultz interrogatory report, p. 89) to two pictures showing an identical "$100,000 House" in Pittsfield and Moultonboro is entirely misleading. The differences in taxes would change the dollar value of the houses.
- The $5000 figure was determined by dividing the annual savings, $250, by an interest rate of 5 percent. The division is done in order to account for future years' benefits from the well. For an explanation of why this simple formula gives a reasonable approximation of present value for long-lasting annual flows, see Fischel (1991). The simple formula is V=R/i, where V is capitalized value (that is, the current sale value of the asset), R is the current annual value, and i is the "real" interest rate, the rate that would prevail in the absence of any expected inflation.
- As the example in Appendix A above demonstrates, the house in the lower-tax rate district (Riverside) will appreciate in value, thus raising its tax payment, so that the tax payments in Hilltop will be less than twice those of Riverside. The critical point in the text, however, is that the financial position of homeowners with the same personal income in Hilltop and Freetown will be exactly the same, despite different tax rates and tax payments.
Other Interesting Educational Links New Hampshire Politics
P.O. Box 1120
Merrimack, NH 03054
Full T1 direct Internet Backbone connection and Internet Services
provided by Mainstream EIS