III. Previous Work
The literature review covers the modeling of fiscal impacts due to rural residential development. Burchnell and Listokin (1978) identify two broad types of fiscal impact methodologies: average cost and marginal cost analyses. Average cost approaches involve the use of ratios or multipliers per unit of service extended. These approaches assume future costs are approximately the same as current costs. They do not account for deficient or excess service capacity, AFT (1999). Marginal cost approaches typically involve calculating specific impacts through a case study or the use of statistical models, (Smith, Propst, and Abberger 1991).
Average cost approaches
The American Farmland Trust COCS is an average cost approach to address the fiscal impact of residential development, AFT (1999). The procedure for implementing a COCS study is straightforward. First, the analyst decides on the land use categories that are relevant for the policy issue at hand. This means residential and agricultural among others in the case of farmland conversion. Typical AFT type studies include three to as many as five types of land uses: residential, agricultural, forestland, commercial, and industrial. Then using what information is available at the local level, expenditures and revenues are allocated to each land use category. This may be done either through records where land use categories can be identified or on a simple proportional basis. A ratio of expenditures to revenues is then calculated for each category. A ratio of expenditures to revenues of less than one suggests the land use is a net fiscal benefit to county government. A ratio of expenditures to revenues of greater than one suggests that the particular land use is a net fiscal loss to county government.
A summary by Heimlich and Anderson (2001) identified 88 cost of community service studies. The studies indicated expenditure to revenue ratios for residential development that exceeded one while those for farm and timberland fell below one across all studies. The average ratio of expenditures to revenues for rural residential development equaled 1.24 compared to 0.38 for farmland and open space. It is important to point out that AFT cost of community services studies generally do not distinguish between urban and rural residential development.
The AFT methodology clearly implicates residential development as a net fiscal loss to local governments. Yet, cautions have been voiced concerning the use of the ratios developed by this approach. The ratio of fiscal expenditures to fiscal revenues generated for each land use is a snapshot of the financial relationships between users and providers at a point in time. The net fiscal impacts of a specific change in land use may not necessarily change in the proportion as indicated by the ratio calculated. Both numerator and denominator are endogenously determined in this case.
The problem with average cost approaches is that they are poor indicators of general results. Taylor (1999) used the AFT approach for Sublette County, Wyoming. A ratio of expenditures-to-revenues was estimated at $2.35 for residential development in the case where there is a resident who is consuming services but is not employed locally. The ratio falls to 1.27 if the resident is employed. This revised ratio takes into account taxes paid by the local employer (commercial property taxes and use taxes). The estimated ratio from an average cost approach then is driven by the assumptions concerning the resident.
Kelsey (1996) critiqued the AFT methodology as used for six Pennsylvania counties. He concluded that cost of community services studies do provide useful information to communities. The author identified a series of limitations in interpretation that local policy makers must be aware of to use the ratios correctly. The first is that the ratio is primarily a reflection of the proportion of local spending for schools. The second is that the ratio is averaged across land types and therefore differences among uses within the same category are lost. The third is that the starting point or unit basis used to measure the ratio can affect the fiscal ordering of land uses. Deller (1999) added another assumption being that different land uses are independent, e.g. a rural resident may be employed by a local business paying the commercial tax rate. That employee then can be viewed as contributing to the commercial tax revenue of the business. An AFT ratio does not capture this relationship.
Marginal cost approaches
Other work in fiscal impact modeling has stronger ties to concepts derived from economic theory. Grosskopf, Hayes, and Hirshberg (1995) use an economic distance function approach to estimate efficiencies in the provision of public law enforcement services.1 The distance function approach has the advantage that it allows for multiple outputs and completely describes the technology. Coefficients then become a direct measure of efficiency changes for a particular policy scenario.
Heikkila and Craig (1991) and Heikkila (2000) present an approach to fiscal impact modeling that draws upon the economic theory of the firm. Levels of government service are a function of inputs (mostly labor and capital) and neighborhood or community characteristics. Once public service impacts are estimated, a welfare analysis is used to measure the change in benefits associated with population changes.
Marcoullier, Deller, and Green (2000) estimate a fiscal impact model using a generalized econometric approach to evaluate the fiscal effects of second home development. The fiscal impact of recreational housing development, on a variety of public services, is analyzed. Public service expenditures are regressed against tax, demographic variables, and recreational houses per capita. The authors report results suggesting that recreational housing just pays for itself.
The study summarized here focuses on measuring the net impacts of rural residential development on the fiscal structure of Wyoming county governments and school districts. The model departs from the approach used by American Farmland Trust (AFT), which is primarily a categorization of rural and urban residents (AFT, 1999). The analysis provided below presents estimates of the fiscal impacts of rural residential development using an econometric model of county revenues, county expenditures, school district revenues, and school district expenditures. This modeling approach reveals marginal as well as average costs, and can make possible projections about cost and revenues of future development. The modeling framework allows for analyzing specific scenarios and can be used to test specific assumptions that are implicit in the AFT approach. This analysis is useful for evaluating the fiscal impacts of rural residential development in the aggregate. The model is used to derive AFT type ratios for specific Wyoming counties.