zoning and fiscal interdependencies

14
Journal of Real Estate Finance and Economics, 12:221-234 (1996) 1996 Kluwer AcademicPublishers Zoning and Fiscal Interdependencies MARYJANE LENON Department of Economics, Providence College, Providence, RI 02918 SAJAL K. CHATTOPADHYAY Department of Economics, U-63, Universityof Connecticut, Storrs, CT 06269 DENNIS R. HEFFLEY Department of Economics, U-63, Universityof Connecticut, Storrs, CT 06269 Abstract Although there are strong theoretical reasons to regard optimal zoning and fiscal policies as simultaneous deci- sions, most empiricalstudies have focusedon one type of decision or the other. Even fewerattemptshavebeen made to study interdependencies between competingtowns in their selection of zoning and fiscal policies. If these interdependencies are prevalent, autonomouslocal governmentsmay pursue zoning, taxing, and spending policiesthat are locally rationalbut socially inefficient. In this study, an optimization modelof the local economy is used to illustratethe natureof these interdependenciesand to specify an appropriateempiricaltest. Data from 164 Connecticut townships are used to estimatethe empirical model. Key Words: zoning, fiscal, interdependencies, property taxes, land-use controls, local government This study seeks to extend the theoretical and empirical analyses of local government behavior by incorporating the simultaneity of zoning and fiscal decisions within a town, as well as the interdependence of such decisions among neighboring towns. Government decisions regarding land-use and the provision and financing of local public goods affect the welfare of households and the profitability of firms. Depending on the zoning mechanisms employed and the degree of restrictiveness, zoning can influence land rents and housing prices, population density patterns, environmental quality, transportation costs, and even labor markets. Similarly, local tax and expenditure decisions affect the locations and microeconomic choices of households and firms, and, hence, market outcomes within local and regional economies. Taken separately, or together, these local zoning and fiscal deci- sions are potentially quite important and have been studied extensively. The literature on the theoretical and empirical aspects of zoning is rich, and interest in the topic remains keen: in recent years, two key journals have devoted entire issues to land-use controls (The Journal of Real Estate Finance and Economics, June 1990, and Land Economics, August 1990). A focus of theoretical studies has been the effects of zoning on land rents and housing prices. Ohls, Weisberg, and White (1974) show that aggregate land values are affected differentially by zoning, depending on the elasticities of demand in different submarkets. C ourant (1976) and Grieson and White (1981) demonstrate, respec- tively, that the effects of zoning on land values depend on whether the zoning is used by one municipality or throughout the metropolitan area, as well as on which particular zon- ing instrument is employed.

Upload: independent

Post on 23-Nov-2023

0 views

Category:

Documents


0 download

TRANSCRIPT

Journal of Real Estate Finance and Economics, 12:221-234 (1996) �9 1996 Kluwer Academic Publishers

Zoning and Fiscal Interdependencies

MARYJANE LENON Department of Economics, Providence College, Providence, RI 02918

SAJAL K. CHATTOPADHYAY Department of Economics, U-63, University of Connecticut, Storrs, CT 06269

DENNIS R. HEFFLEY Department of Economics, U-63, University of Connecticut, Storrs, CT 06269

Abstract

Although there are strong theoretical reasons to regard optimal zoning and fiscal policies as simultaneous deci- sions, most empirical studies have focused on one type of decision or the other. Even fewer attempts have been made to study interdependencies between competing towns in their selection of zoning and fiscal policies. If these interdependencies are prevalent, autonomous local governments may pursue zoning, taxing, and spending policies that are locally rational but socially inefficient. In this study, an optimization model of the local economy is used to illustrate the nature of these interdependencies and to specify an appropriate empirical test. Data from 164 Connecticut townships are used to estimate the empirical model.

Key Words: zoning, fiscal, interdependencies, property taxes, land-use controls, local government

This study seeks to extend the theoretical and empirical analyses of local government behavior by incorporating the simultaneity of zoning and fiscal decisions within a town, as well as the interdependence of such decisions among neighboring towns. Government decisions regarding land-use and the provision and financing of local public goods affect the welfare of households and the profitability of firms. Depending on the zoning mechanisms employed and the degree of restrictiveness, zoning can influence land rents and housing prices, population density patterns, environmental quality, transportation costs, and even labor markets. Similarly, local tax and expenditure decisions affect the locations and microeconomic choices of households and firms, and, hence, market outcomes within local and regional economies. Taken separately, or together, these local zoning and fiscal deci- sions are potentially quite important and have been studied extensively.

The literature on the theoretical and empirical aspects of zoning is rich, and interest in the topic remains keen: in recent years, two key journals have devoted entire issues to land-use controls (The Journal of Real Estate Finance and Economics, June 1990, and Land Economics, August 1990). A focus of theoretical studies has been the effects of zoning on land rents and housing prices. Ohls, Weisberg, and White (1974) show that aggregate land values are affected differentially by zoning, depending on the elasticities of demand in different submarkets. C ourant (1976) and Grieson and White (1981) demonstrate, respec- tively, that the effects of zoning on land values depend on whether the zoning is used by one municipality or throughout the metropolitan area, as well as on which particular zon- ing instrument is employed.

222 LENON, CHATTOPADHYAY, AND HEFFLEY

Empirical studies of the price effects of zoning also report mixed results. Studies by Hamilton (1978), Jud (1980), and Dowall and Landis (1982) lend support to the theoretical argument that zoning affects land and structure prices. Pollakowski and Wachter (1990) examine the effects of land-use constraints on housing prices and conclude that the collec- tive price effect of land-use controls is significantly greater than the effect of any one con- trol. Pertinent to our study, the authors also find that spillover effects cross political boun- daries. Other authors, including Renter (1973), Grether and Mieszkowski (1980), and Mark and Goldberg (1986), find less significant price effects of zoning in the particular com- munities which they studied.

The efficiency and welfare effects of zoning also have been studied. Like many forms of market intervention, zoning distorts economic choices. At the microeconomic level, it is easy to show that a binding constraint on minimum lot-size, for example, increases the developer's cost of producing a given quantity of housing services. On the other hand, zoning often is justified as a mechanism for reducing or avoiding certain externalities that raise the social cost of development. Thus, the net social benefits of zoning depend on the magnitudes of these opposing effects and the costs of zoning administration and enforce- ment. These efficiency and welfare issues have been examined by many authors. 1 Stull (1974), for example, demonstrates that zoning affects both the labor market (employment and wages) and the spatial pattern of land rents. Helpman and Pines (1977) extend Stall's model to show that attempts by local government to maximize aggregate land values can lead to a social optimum. According to Goetz and Wofford (1979), zoning redistributes wealth depending on the pre-zoning market structure: under monopoly conditions, zoning protects consumers by reducing externalities, while property owners reap economic pro- fits fron zoning imposed in more competititive markets. Wallace (1988) provides evidence that zoning by local governments tends to "follow the market" (i.e, does not distort) except in the case of large-lot zoning. Henderson (1985) examines the distortionary effects of zoning policies that regulate housing quality, while Miceli (1991) extends Henderson's model to include a public sector, and demonstrates that zoning policies designed to prevent free- riding in the consumption of local public goods must mandate minimum lot-sizes in excess of the average to achieve their goal. McEachern (1979) reaches a similar conclusion. While these latter studies attempt to capture the link between zoning and fiscal decisions within a community, as suggested in earlier studies by Hamilton (1975) and Brueckner (1981), they do not explicitly model the interdependence of policy making between neighboring communities. Such interdependencies are potentially important in a Tiebout setting where mobile households and firms comparison shop among rival towns.

A series of articles by McMiUen and McDonald (1989, 1991a, 1991b) treat zoning as an endogenous decision and find that: (1) land values and land-use are correlated, thus biasing estimates of land-value functions that assume exogenous zoning; and (2) zoning changes that control the allowable use of land can be predicted. In this latter study, the authors find that mixed land-use patterns are unlikely to persist over time; specifically, land at the edge of suburban areas is most likely to be converted from residential to nonresidential uses. Exclusively nonresidential or residential areas are less likely to ex- perience a change in allowable use.

In addition to its impact on efficiency and welfare, zoning changes may alter the size and composition of the local residential and business sectors, prompting changes in the

ZONING AND FISCAL INTERDEPENDENCIES 223

level and mix of public services. Hamilton (1975), Brueckner (1981), and others have ex- amined the relationship between zoning and fiscal policies, 2 but only within the last decade has this link been studied in a framework that allows property markets and local fiscal decisions to fully adjust to zoning changes (Heffley and Hewitt, 1988). Yet, even this latter general equilibrium approach treats zoning as a parameter and says little about why a town adopts a particular zoning policy in conjunction with its fiscal policies, or how these local policies might be affected by the policies of neighboring towns.

Other studies have focused on the zoning decision, taking the fiscal attributes of the com- munity as given. White (1978), McEachern (1979), Cooley and LaCivita (1982), Rolleston (1987), and Miceli (1992) approach the "causes" or motivation for zoning at the com- munity level. These studies identify fiscal, externality, and exclusionary motives for zon- ing, but do not fully model the interdependency of fiscal and zoning decisions that exists between neighboring towns. The issue of zoning interdependence is addressed by Burnell and Burnell (1989), but the empirical analysis lacks theoretical underpinnings and conse- quently may be subject to specification error. Lenon (1989) develops and tests a model of the representative household's demand for zoning, but does not include an optimizing public sector and, therefore, has no mechanism to handle the simultaneous nature of zon- ing and fiscal decisions within a town, or the zoning and fiscal interdependencies between towns.

While the preferences of residents certainly matter in the determination of zoning policy, those preferences are filtered through an elected local government which also must establish fiscal policies and which may be sensitive to the zoning and fiscal choices of neighborning towns. Casual observation suggests a zoning change in one town often leads neighboring towns to consider and possibly enact similar changes. For example, town i's fears that an increase in townj's minimum lot-size will divert development pressure to town i, straining i's local services, may cause town i to restructure its own zoning policies. Such in- terdependence may lead individual towns to adopt locally rational policies that are ineffi- cient from a social perspective.

Similar interdependencies may exist in local fiscal decisions. This has been discussed at length in the "fiscal competition" literature, but such studies typically ignore the rela- tionship between fiscal and zoning policies. Epple and Zelenitz (1981), Gerber and Hewitt (1987a, b), Zodrow and Mieszkowski (1986), Wilson (1986, 1987), Bucovetsky (1991), Wildasin (1991), and Henderson (1994), among others, offer varied models of competi- tion among jurisdictions with regard to tax and expenditure choices, but the implications of these models may be affected if fiscal and zoning decisions are simultaneously deter- mined. This study is an attempt to integrate zoning and fiscal decisions and, at the same time, to allow for the interdependence of these decisions among competing towns.

I. Zoning and fiscal behavior with interdependencies

In this section, we specify a model of a town's zoning and fiscal behavior. Nested within this model of government choice is a model of household choice and residential mobility. Households maximize utility and may locate in the town of interest or in other nearby towns, whose zoning and fiscal policies vary parametrically. 3 This "open city" framework not

224 LENON, CHATTOPADHYAY, AND HEFFLEY

only makes local population size endogenous, but serves to link one town's zoning and fiscal choices to prevailing policies in other towns. This model of household behavior, coupled with the open-city condition--that the town's representative household, with given income and tastes, must attain a level of utility available elsewhere--determines the local land rent. This equilibrium land rent depends on the town's own fiscal and zoning deci- sions, as well as those of competing towns. In keeping with the open-city literature, we further assume that the town selects zoning and fiscal policies so as to maximize aggregate land value within its borders. From the open-city submodel, however, aggregate local rents also depend on the zoning and fiscal policies of other towns. Consequently, parametric changes in external policies should prompt policy adjustments in the town of interest. The model is used to guide an empirical evaluation of the effects of a town's own attributes (income level, location, geographic size, etc.), as well as the influence of external zoning and fiscal decisions, on the town's optimal zoning, taxing, and spending policies (section 2). Cross-sectional data (1987-1988) for 164 Connecticut towns are used tO examine the implied linkages between communities' zoning, taxing, and spending policies. Possible policy implications of these findings are noted in section 3.

1.1. Households

Temporarily ignoring town-specific subscripts, we adopt the following notation:

U : utility y: income l: lot-size s: structure z: other goods r: price of land (rent)

m: price per unit of structure p: price of other goods k: travel costs per unit of distance x: distance from town to regional center g: household's consumption of a congestible local public good G: total local public expenditures t: residential property tax rate l: minimum lot-size

h: number of households

Each household derives utility from its consumption of land, structure, other goods, and its share of the local public good; or U(l, s, z; g). The latter component of utility is taken as given by any household choosing to locate in the town, since g is determined by total local public expenditures, the number of households in the town, and the con- gestibility of the public good. If units of structure, other goods, and distance are defined to allow m = p = k = 1.0, the representative household maximizes U subject to a budget constraint: 4

ZONING AND FISCAL INTERDEPENDENCIES 225

y - (1 + t ) ( r l + s) - z - x = 0; (1)

and a local lot-size restriction:

t - l _ > o . (2)

The interesting case, where both budget and zoning constraints bind, leads to demand functions of the general form; 5

I* = /; s*(y, r, x, /, t, g); z*(y, r, x, /, t, g). (3)

The zoning restriction, if binding, influences land consumption as well as the household's use of structure and other goods. The associated indirect utility function,

U*(y, r, x , / , t, g), (4)

reflects the importance of zoning, taxing, and spending decisions on household well-being. In a closed-city model, where households are immobile, a benevolent local government

might simply maximize U* subject to a fiscal constraint. In an open-city model, however, potential mobility ensures that land rents will adjust to equate U* to some outside level. This, of course, raises a question about what objective local policy makers might pursue within such an environment, and requires the formal introduction of a competing town.

1.2. Local government

Within an open-city framework, local government can do nothing to permanently influence the level of household utility (or profits, if there exist mobile firms) within a town. In such models, it is common to assume that local policies are set to maximize the value of the scarce and immobile resource controlled by the community--its land. Local governments in the United States have well-established powers to control land-use, to tax land and struc- ture, and to provide public services that may enhance land values. Given these land-related powers and the political influence of landowners, the assumption that local government may tailor policies to preserve and enhance the value of land is not far-fetched. Brueckner (1982, 1983) and others offer theoretical arguments and empirical results which support the land-value maximization hypothesis, at least with regard to fiscal decisions. Here, we regard both fiscal and zoning decisions as the instruments of land-value maximization.

Land value is directly related to land rent. Drawing on the structure of the indirect utility function (4), an expression for ri, the rent per unit of area in town i, is implicitly defined by the open city condition:

U*(y, ri, xi, li, ti, g~) - UT(Y, rj, xj, lj, tj, gj) = O. (5)

U* is the utility that a household with income y can get in town j ~ i. If U* > UT-, population flows fromj to i, causing ri to rise and ~ to ~ until equality is reestablished. 6

226 LENON, CHATTOPADHYAY, AND HEFFLEY

Opposite adjustments in population and rents occur if the inequality is reversed. Implicit solution for the value of r i that satisfies (5) gives a rent function of the general form:

r*(y, xi, li, ti, gi, rj, xj, l j , tj, gj). (6)

Land rents in town i thus depend on local zoning and fiscal policies (/i, ti, gi), as well as on the zoning and fiscal policies in competing towns (/j, tj, gj). Below, we denote the equilibrium rent function more compactly as r*.

The local government also must recognize the relationship between the individual household's consumption of the public good (gi), the town's total expenditures on public goods (Gi), and the number of resident households (hi). For simplicity, let:

gi = Gi/hi. (7)

This particular congestion function, which treats Gi as essentially private, is consistent with several empirical studies that find relatively high congestibility at the margin for a variety of local public goods (Borcherding and Deacon, 1972; Santerre, 1985).

If there is some given area of land zoned for residential use (Li), local decisionmakers also must be aware of the constraining effect of minimum lot-size (/i) on the number of households in the town. When binding, the lot-size restriction (//) determines the individ- ual household's demand for land, as in (3), while the area zoned for residential use deter- mines the local supply (assuming here, for simplicity, that no residentially zoned land is held vacant for speculative purposes). Clearance of the residential land market in town i requires that h il~ = Li, or:

hi = L~/~. (8)

Thus, given an area of residentially zoned land, larger minimum lot-size restrictions reduce the number of households which the town can accommodate. Note that (7) and (8) jointly imply gi = Gili/Li.

Finally, government officials must select zoning and fiscal instruments that will balance public revenues and expenditures, or:

R i + hit i [r ' l* + ms*] - Gi = O, (9)

where the first term (Ri) denotes exogenous revenues (grants from state or federal sources, local non-property tax revenues, license fees, etc.), and the second term reflects local revenues from the taxation of land and structure. Note that l* and s* are the representa- tive household's consumption of land and structure, and will be functions of the local govern- ment's zoning and fiscal policies, as in (3). These local policies also enter r*, which is given by (6), but r* also depends on the prevailing policies in town j.

Consider the problem faced by town i, which must select a minimum lot size (/i), a property tax rate (tl) and a total level of public expenditures (Gi). The town has some fixed area of land zoned for residential use (L~), and wishes to maximize the aggregate value

ZONING AND FISCAL INTERDEPENDENCIES 227

of this land subject to a balanced budget condition. 7 Thus, the problem is to maximize r*Li, where r* is given by (6), subject to (7), (8), and (9). First-order conditions for this problem implicitly define the optimal zoning (/*), taxing (t*), and spending (G*) behavior of town i, in terms of the parameters of the problem:

l~'(Y, Xi, Li, R~, xj, r> l~, tj, gj); (10)

r xi, ~,i, Ri, xj, rj, b, tj, gj); (11)

a~(y, xi, Li, gi, xj, rj, b; tj, gj). (12)

With minor modifications, noted below, we use this set of reduced-form equations to specify an empirical test of the importance of interdependencies in local zoning and fiscal policies.

2. Empirical analysis

The model in section 2 suggests that zoning and fiscal decisions within a given town i are not only jointly determined but depend on the zoning and fiscal policies of competing towns, as well as other parameters of town i. If such interdependencies are empirically important, then at a given point in time, cross-sectional observation of zoning and fiscal policies should reveal a nonrandom pattern, with systematic relationships between neighboring towns. To explore this matter, we use data for 164 Connecticut townships. Many of the data were provided by the Connecticut Public Expenditure Council (1989); data on zoning practices were drawn from a survey of local zoning authorities conducted by Lenon (1989). 8 The general structure of the estimated model is:

[*(y, x,, L,, R~, xj, l~, tj, ~ , Lj); (13)

r xi, Li, Ri, xj, b, t j, Cj, Lj); (14)

6*~y, xi, Li, Ri, xj, l~, q , t 9 (15)

The structure of (13-15) corresponds closely to the theoretical structure (10-12), with two modifications. First, if (7) and (8) hold i n j as well as in i, then gj can be expressed as a function of Gj, Lj, and 3" Second, since land rents in competing towns (rj) are likely to be strongly correlated with their site characteristics (xj, Lj) and their zoning and fiscal policies (lj, tj, Gj), we suppress rj in (13-15). 9

For each town in the sample, we define the following variables for estimation purposes:

l.* = prevailing minimum lot-size restriction in town i; _ l

t* = effective property tax rate (equalized mill rate) in town i; G* = total public spending in town i;

y = household income (a weighted average of renter and owner-occupied incomes);

228 LENON, CHATTOPADHYAY, AND HEFFLEY

xi = distance in miles from town i to the nearest metropolitan center; L i = land area of town i; R i = revenues from sources other than property taxes in town i; x/ = average distance to nearest metropolitan center for towns adjacent to town i; /j = average minimum lot-size for towns adjacent to town i;

Gj = average public spending for towns adjacent to town i; L/ = average land area for towns adjacent to town i.

This specification implies that a town's relevant competitors include all geographically ad- jacent towns. To limit the number of dependent variables and preserve degrees of freedom, we use the mean attributes of these adjacent towns to construct xj, lj, tj, Gj, and Lj. 1~ For each of the above variables, Table 1 gives the definition, source, mean, standard devia- tion, and range.

Equations (13-15) represent reduced forms of the model and may be estimated indepen- dently using OLS methods, since each equation contains the same set of independent vari- ables and no cross-equation parameter restrictions are imposed. Both linear and log-linear forms provide reasonable fits 11 and the signs and significance levels of most coefficients are robust to the specification. Results of a modified J test (the PE test), however, reveal that the log-linear form is superior in the zoning and expenditure equations (13, 15) and marginally preferred in the tax-rate equation (14). 12 The log-linear form also provides direct estimates of the partial elasticities and helps to reduce potential problems of hetero- scedasticity, common in this type of cross-sectional analysis. To preserve observations, zero-valued variables are set to arbitrarily small positive values; this occurs in 16 of the

Table 1. Descriptive statistics.

Standard Variable Source Mean Deviation Minimum Maximum

/* Lenon (1989) 1.327 .881 0.0 5.0

t* CPEC (1989) 11.088 2.702 4.0 20.1

G* CPEC (1989) 27160700 42182734 1329570 302874040

y CPEC (1989) 44149.7 15904.7 20218.5 113948.4

x i Lenon (1989) 13.134 7.810 0.0 36.0

L i CPEC (1989) 29.437 13.072 5.3 64.4

R i CPEC (1989) 9960150 19764460 283400 155033720

xj Lenon (1989) 13.035 5.384 0.0 30.0

/./ Lenon (1989) 1.313 0.563 0.4 3.4

tj CPEC (1989) 11.098 1.791 5.4 14.6

Gj CPEC (1989) 27652806 27781096 2416810 209415088

Lj CPEC (1989) 30.299 7.427 13.1 50.2

CPEC (1989):

Lenon (1989):

Connecticut Municipal Budgets, 1987-1988. Hartford: Connecticut Public Expenditure Coun- cil, Inc., May 1989. Lenon, M.J. "The Demand for Zoning: A Microeconomic Approach" Doctoral dissertation, Department of Economics, University of Connecticut, Storrs, CT, 1989.

ZONING AND FISCAL INTERDEPENDENCIES 229

1476 (= 164 • 9) data cells. Tests with alternative values of .001, .01, and. 10 indicate that the signs and significance levels of most coefficients are not very sensitive to the par- ticular value. However, . 10 yields the best fit for the zoning and expenditure equations (13, 15), and the fit of the tax-rate equation (14) is similar in all three cases (R 2 = .5046, .5020, .4973). Results for the. 10 case are reported in Table 2. In the discussion that follows, the estimated coefficients are described as elasticities and denoted ED.,I, where D* is the dependent variable, and I is the particular independent variable.

Of particular interest are the nine coefficients that reflect the degree of zoning and fiscal interdependence between town i and the set of adjacent towns. In each table, the direct cross-policy coefficients [El*.l., Et*. r, E~*~.,] are positive and statistically significant, indi-

t 7 z~/ t J cating that town i's zoning, taxing, and spending policies tend to be positively related to the corresponding "average policies" in neighboring towns. The elasticity of town i's minimum lot-size, with respect to the average minimum lot-size in surrounding towns, is .473. With an elasticity of .488, tax policy also is quite responsive to variations in the average effective tax rate in neighboring towns. Public spending in town i is less responsive to changes in the average spending level in adjacent towns (.216), but all three coefficients are significant at better than the 1% level.

Table 2. OLS estimation results (natural logs).

Equation (13) Equation (14) Equation (15) Variable log/* log t* log G*

constant -2.842458 1.674169 -2.283482 (-1.178) (2.103) (-2.260)

log y .512330 -.094151 ,396661 (2.318) (-1,291) (4.284)

log x i .055335 .002681 -,042101 ( 1.357) (. 199) (-2.464)

log L i .362980 -. 120239 .190879 (2.943) (-2.954) (3.695)

log R i -.296668 .127388 .839430 (-4.841 ) (6.299) (32.701 )

log xj -. 175654 .058646 .156718 (-1.610) (1.629) (3.429)

/ j .473300 -.064659 -. 183750 (2.753) (-1.140) (-2.552)

log tj .012644 .487606 -.934493 (.037) (4.283) (-6.467)

log Gj .052061 -.093396 .215917 (.446) (-2.423) (4.414)

log Lj .027981 .119343 -. 192822 (. 130) (1.685) (-2,145)

F ratio 20.636 16.928 433.042

adj R E .5202 .4679 .9598

DF 154 154 154

t-statistics in parentheses.

230 L E N O N , C H A T T O P A D H Y A Y , A N D H E F F L E Y

The results in Table 2 also reveal that, in addition to these direct interdependencies (zon- ing on zoning, taxing on taxing, spending on spending), there also appear to be some signifi- cant cross-policy effects, especially between taxing and spending policies. An increase in the mean tax rate in adjacent towns is associated with a decrease in town i's spending ( E ~ r = - .934) ; similarly, an increase in mean spending in the adjacent towns is negatively related to the tax rate established in town i(Et*.~. = -.093). In selecting op- t J timal fiscal and zoning instruments, within the confines of a budget constraint, local authorities must balance the rent-enhancing effects of more population against the adverse effects that additional population will have on each resident's consumption of congestible public goods. If neighboring towns raise their tax rates, for example, new residents will be attracted to town i. While this population pressure tends to increase the value of land in town i, it causes crowding in its local public services which, in turn, tends to reduce land rent. Consequently, the exogenous increase in tj not only allows town i to increase its tax rate to deter some of the influx (recall that Et~ ~ > 0) but also to reduce public spending (EG*t~ < 0). An analogous response seems to occur when the exogenous event is an increase m spending in surrounding towns; that is, not only is E~6. > 0, but JEt*. 6

�9 . ~ j i j

< 0. In such a case, town i makes its fiscal mix more attracUve to stem a population outflow and reduction in its land value caused by the increase in Gj.

In keeping with the focus of this article, we have discussed the coefficients that reflect policy interdependence among towns, but the empirical results also suggest that policy choices will reflect local attributes. For the zoning equation (13), minimum lot-size is positively related to per capita income (E.t*,y = .512) and land area of the town (Et*Li = .363), but negatively related to the level of other revenues (E:RI = --.297) in the local budget. Other factors held constant, large minimum lot-size restrictions are more likely to be adopted in high-income towns with ample land, but these lot-size restrictions can be relaxed if other (non-property-tax) revenues are increased to maintain local public ser- vices. The responsiveness of minimum lot-size to changes in other government revenues suggests that intergovernmental grants, appropriately structured, could be a useful tool for encouraging local governments to adopt less restrictive zoning ordinances.

Other revenues also have a significant positive relationship to both the tax rate and the level of public spending. The variable R i includes a variety of revenue sources, but the signs of these coefficients suggest that the effects of matching grants may be important, since increases in R i seem to induce local public expenditures (E~Ri = .839) and stimu- late local lax effort (JEt* R. = .127), rather than simply prompting a substitution of nonlocal revenues for local taxes: 13 The magnitude and statistical significance of these effects again indicate that intergovernmental grants may be an effective way to encourage decentralized local governments to select taxing and spending levels, as well as zoning policies, that re- main locally rational but are socially more efficient.

Taxing and spending policies are influenced by other local attributes. As is commonly observed, higher income towns are able to sustain higher public expenditures (EG*,y = .397) at lower effective tax rates (Et*y = - .094) , due to larger per capita property tax bases. "Land rich" towns, like higher income towns, also are able to spend more (Ec*Li = . 191) with lower tax rates (EtTL~ ---- --. 120). Distance from a metropolitan center only has a statistically significant effect on public spending (E~*,x ~ = - .042) , reflecting the higher public costs associated with urban living.

ZONING AND FISCAL INTERDEPENDENCIES 231

While local attributes such as income, geographic size, location, and revenues from sources other than the property tax are collectively important in explaining differences in zoning, taxing, and spending policies across towns, inclusion of the characteristics of neighboring jurisdictions is not only theoretically appropriate but statistically warranted. For each equation, an F-test on the subset of variables defining the policies and geographic attributes (lj, tj, Gj, xj, Lj) of neighboring towns leads to rejection of the null hypothesis that the associated coefficients are jointly equal to zero.

3. Final remarks

The goals of this study were: (1) to provide a theoretical basis for examining the relation- ship between zoning, taxing, and spending decisions within a town, as well as the in- terdependencies that may exist between towns in their choices of zoning and fiscal policies; and (2) to empirically test this model. Casual observation, anecdotal evidence, and intui- tion suggest that towns do not make such choices in a vacuum and are sensitive to the deci- sions of surrounding towns, as well as to local attributes. Yet, little systematic evidence of such effects, especially in zoning policies, has been offered.

A simple model of residential choice is imbedded within an optimization model of local government. Within this framework, a local government seeks to maximize the value of the community's residential land by selecting a minimum lot-size, a property tax rate, and a level of public spending, subject to a budget constraint, a public good congestion func- tion, and a residential land-clearance condition. The reduced-form structure of this model is used to guide an empirical analysis of zoning and fiscal policies within 164 Connecticut townships. Estimation results suggest that policy interdependencies do exist and may be rather important in determining the observed pattern of zoning, taxing, and spending levels among localities.

This pattern of zoning and fiscal policies may emerge from locally rational (land-rent maximizing) choices by town governments, but the failure to fully internalize the spillover effects of a town's decisions on neighboring towns may result in an equilibrium pattern of zoning and fiscal policies across towns that is socially inefficient. For example, zoning interdependencies may result in a group of towns adopting overly restrictive minimum lot- sizes. Each town might benefit from a reduction in minimum lot-size, provided that the easing of the restriction occurs simultaneously in all towns in the area. Thus, better coor- dination of these policies across towns might offer mutual benefits. Such coordination could be achieved in a variety of ways, including: (1) regionalization of zoning and fiscal policies; (2) zoning/fiscal "treaties"; or (3) a system of grants from a higher-level government that provides appropriate incentives for towns to adopt zoning and fiscal policies that are socially more efficient.

Given the strength of preferences for local autonomy and for having diverse options at the local government level, regionalization may be the most politically difficult corrective measure. Policy makers may find it easier to reduce inefficiencies associated with zoning and fiscal interdependencies by encouraging negotiated agreements between communities about their zoning and fiscal decisions, or by developing an appropriate structure of higher level grants. The "treaty" approach, of course, runs the risk of allowing local governments

232 LENON, CHATTOPADHYAY, AND HEFFLEY

in an area to behave as a cartel in sett ing zoning restrict ions, tax rates, and spending levels, thereby jeopard iz ing the more desirable features of Tiebout compet i t ion among towns, in- c luding jur isdic t ional diversi ty and pressure on local authori t ies to min imize the cost of

providing any chosen bund le of publ ic services. In tergovernmental grants, on the other hand, a l ready exist in mos t states. Typically, however, formulae used to de te rmine the size of the grants are based o n such things as popula t ion size, income, proper ty wealth, and

tax effort of the communi ty . Modi fy ing a par t icular formula to encourage and reward less

restrictive zoning ordinances seems relatively straightforward, bu t inevitable disagreements

about the appropriate measure of zoning restr ict iveness would need to be resolved. Efforts to resolve such disagreements could be worthwhile, however, since empir ica l results of this study suggest that external funds have measurab le and statistically s ignif icant effects

on local zoning and fiscal policies.

Acknowledgments

We are par t icular ly grateful to the Center for Real Estate and Urban Economic Studies (School of Business Adminis t ra t ion , The Univers i ty of Connect icut) for its support of this

study, and to El len Cromley (Depar tment of Geography, The Univers i ty of Connect icut)

for her technical assistance in const ruct ing a large bu t useful matrix. Thanks go to two

anonymous referees for their par t icular ly helpful comments .

Notes

1. Pogodzinski and Sass (1990) offer a comprehensive and critical review of zoning theory and these welfare issues. 2. The volume edited by Mills and Oates (1975) contains a number of articles which focus on the fiscal aspects

of zoning. 3. A more complete regional (as opposed to local) equilibrium model would allow the zoning and fiscal policies

of all communities to be endogenous. Such a model is analytically more complex and leads to a very dif- ferent, structural, econometric specification.

4. For the empirical analysis, there is no loss of generality in this normalization if m, p, and k do not vary across subsets of adjacent communities.

5. Hamilton (1975) and others have offered theoretical support for the notion that zoning constraints generally will be binding in a Tiebout enviroment, where mobile households sort themselves into communities with zoning and fiscal policies that roughly match their individual preferences.

6. In this discussion of the model, townj is treated as a single competing jurisdiction. For empirical purposes, policies and attributes ofj are represented by the mean policies and attributes of all towns bordering town i.

7. We abstract from nonresidential uses of land to accommodate the simultaneous determination of zoning and fiscal decisions within the town and the links between these decisions and the policies of neighboring towns. Heffley and Hewitt (1988) consider both residential and commercial land-use, but treat (allowable use) zon- ing parametrically and ignore interdependencies.

8. Connecticut includes 169 townships. Five observations are deleted due to missing data, including four towns which failed to respond to the zoning survey by Lenon (1989).

9. This deletion is also a practical matter. Consistent and reliable data on residential land rents in each of Con- necticut's 169 towns are generally not available. Reasonably accurate data on house sales do exist, but there remains the problem of separating land and structure prices.

10. In reality, fiscal and zoning interaction may not be limited to adjacent towns, but definition of each town's relevant set of competitors, based on other criteria, requires additional assumptions about the "community

ZONING AND FISCAL INTERDEPENDENCIES 233

shopping" behavior of households. Construction of the mean attribute values for adjacent towns was facilitated by the use of a 169 • 169 matrix depicting the pattern of adjacency (a unit element) and non-adjacency (a zero element) among the 169 Connecticut townships. We are much indebted to Ellen Cromley (Department of Geography, the University of Connecticut) for providing this matrix.

11. Adjusted R 2 values for linear forms of (13-15) are .3391, 4416, and .9202, respectively. 12. See MacKinnon, White, and Davidson (1983). Detailed results of this test are available from the authors. 13. Many empirical studies of the "flypaper effect" find that even lump-sum grants stimulate local public spend-

ing rather than being fully offset by local tax reductions.

References

Borcherding, T., and R. Deacon. (1972). "The Demand for the Services of Non-Federal Governments" American Economic Review 62, 891-901.

Brueckner, J.K. (1981). "Zoning and Property Taxation in a System of Local Governments: Further Analysis," Urban Studies 18, 113-120.

Brueckner, J.K. (1982) '~, Test for Allocative Efficiency in the Local Public Sector" Journal of Public Economics 19, 311-331.

Brueckner, J.K. (1983). "Property Value Maximization and Public Sector Efficiency" Journal of Urban Economics 14, 1-15.

Bucovetsky, S. (1991) "Asymmetric Tax Competition" Journal of Urban Economics 30, 167-181. Bumell, B., and J. Burnell. (1989). "Community Interaction and Suburban Zoning Policies," Urban Affairs Quarterly

24, 470-482. Connecticut Public Expenditure Council, Inc. (1989). Connecticut Municipal Budgets, 1987-1988. Hartford, Clay. Cooley, T., and C. LaCivita. (1982). ' ~ Theory of Growth Controls," Journal of Urban Economics 12, 129-134. Courant, P. (1976). "On the Effect of Fiscal Zoning on Land and Housing Values" Journal of Urban Economics

3, 88-94. Dowall, D., and J. Landis. (1982). "Land-Use Controls and Housing Costs: An Examination of San Francisco

Bay Area Communities," American Real Estate and Urban Economics Association Journal 10, 67-93. Epple, D., and A. Zelenitz. (1981). "The Implications of Competition Among Jurisdictions: Does Tiebout Need

Politics?" Journal of Political Economy 89, 1197-1217. Gerber, R., and D. Hewitt. (1987a). "Tax Competition and Redistribution Policy of Local Governments Com-

peting for Business Capital" Journal of Urban Economics 21, 69-82. Gerber, R., and D. Hewitt. (1987b). "Decentralized Tax Competition for Business Capital and National Economic

Efficiency" Journal of Regional Science 27, 451-460. Goetz, M., and L. Wofford. (1979). "The Motivation for Zoning: Efficiency or Wealth Distribution?" Land

Economics 55,472--485. Grether, D., and P. Mieszkowski. (1980). "The Effects of Nonresidential Land Uses on the Prices of Adjacent

Housing: Some Estimates of Proximity Effects" Journal of Urban Economics 8, 1-15. Grieson, R., and J. White. (1981). "The Effects of Zoning on Structure and Land Markets," Journal of Urban

Economics 10, 271-285. Hamilton, B. (1975). "zoning and Property Taxation in a System of Local Governments," Urban Studies 12,

205-211. Hamilton, B. (1978). "Zoning and the Exercise of Monopoly Power" Journal of Urban Economics 5, 116-130. Heffley, D., and D. Hewitt. (1988). "Land-Use Zoning in a Local Economy with Optimal Property Taxes and

Public Expenditures" Journal of Real Estate Finance and Economics 1, 373-391. Helpman, E., and D. Pines. (1977). "Land and Zoning in an Urban Economy: Further Results" American Economic

Review 67, 982-986. Henderson, J.V. (1985). "The Impact of Zoning Policies Which Regulate Housing Quality" Journal of Urban

Economics 18, 302-312. Henderson, J.V. (1994). "Community Choice of Revenue Instruments" Regional Science and Urban Economics

24, 159-183.

234 LENON, CHATTOPADHYAY, AND HEFFLEY

Jud, 13. (1980). "The Effects of Zoning on Single-Family Residential Property Values: Charlotte, North Carolina," Land Economics 56, 142-154.

Lenon, M.J. (1989). The Demand for Zoning: A Microeconomic Approach. Doctoral dissertation, Department of Economics, University of Connecticut, Storrs, CT.

MacKinnon, J., H. White, and R. Davidson. (1983). "Tests for Model Specification in the Presence of Alter- native Hypotheses: Some Further Results" Journal of Econometrics 21, 53-70.

Mark, J., and M. Goldberg. (1986). '9` Study of the Impacts of Zoning on Housing Values over Time" Journal of Urban Economics 20, 257-273.

McEachern, W. (1979). "Large-lot Zoning in Connecticut: Incentives and Effects" Center for Real Estate and Urban Economic Studies Report No. 29, The University of Connecticut, Storrs, CT.

McMillen, D., and J. McDonald. (1989). "Selectivity Bias in Urban Land Value Functions," Land Economics 65, 341-351.

McMillen, D., and J. McDonald. (1991a). "Urban Land Value Functions with Endogenous Zoning" Journal of Urban Economics 29, 14-27.

McMillen, D., and J. McDonald. (1991b). '9` Markov Chain Model of Zoning Change" Journal of Urban Economics 30, 257-270.

Miceli, T. (1991). "Free Riders and Distortionary Zoning by Local Communities" Journal of Urban Economics 30, 112-122.

Miceli, T. (1992). "Optimal Fiscal Zoning That Distorts Housing Consumption," Journal of Real Estate Finance and Economics 5, 323-331.

Mills, E., and W. Oates, Eds. (1975). Fiscal Zoning and Land Use Controls. Lexington: Lexington Books. Ohls, J., R. Weisberg, and M. White. (1974). "The Effect of Zoning on Land Value" Journal of Urban Economics

1, 428--444. Pogodzinski, J., and T. Sass. (1990). '9, Review of Zoning Theory" Land Economics 66, 294-314. Pollakowski, H., and S. Wachter. (1990). "The Effects of Land-Use Constraints on Housing Prices" Land Economics

66, 315-324. Reuter, F. (1973). "Externalities in Urban Property Markets: An Empirical Test of the Zoning Ordinance of

Pittsburgh," Journal of Law and Economics 16, 313-349. Rolleston, B. (1987). "Determinants of Restrictive Suburban Zoning: An Empirical Analysis," Journal of Urban

Economics 21, 1-21. Santerre, R. (1985). "Spatial Differences in the Demands for Local Public Goods" Land Economics 61,119-128. StuB, W. (1974). "Land-Use and Zoning in an Urban Economy" American Economic Review 64, 337-347. Wallace, N. (1988). "The Market Effects of Zoning Undeveloped Land: Does Zoning Follow the Market?" Journal

of Urban Economics 23, 307-326. White, M. (1978). "Self-Interest in the Suburbs: The Trend To~ard No-Growth Zoning" Po//cy Ana/ys/s 4, 185-204. Wildasin, D. ( 1991). "Some Rudimentary 'Duopolity' Theory" Regional Science and Urban Economics 21,

393-421. Wilson, J. (1986). '9` Theory of Interregional Tax Competition" Journal of Urban Economics 19, 296-315. Wilson, J. (1987). "Trade, Capital Mobility, and Tax Competition, Journal of Political Economy 95, 835-856. Zodrow, G., and P. Mieszkowski. (1986). "Pigou, Tiebout, Property Taxation, and the Underprovision of Local

Public Goods" Journal of Urban Economics 19, 356-370.