I DEVELOPMENT PA CT Robert W. Burchell David Listokin William R. Dolphin Lawrence Q. Newton Susan J. Foxiey with Robert M. Rodgers Jeffrey L. Greene Larry W. Canter David J. Minno WonsikShim Wansoolm
I DEVELOPMENT
PA C T
Robert W. Burchell David Listokin
William R. Dolphin Lawrence Q. Newton
Susan J. Foxiey
with
Robert M. Rodgers Jeffrey L. Greene Larry W. Canter David J. Minno WonsikShim Wansoolm
About ULI-: the Urban Land Institute
ULI-the Urban Land Institute is a nonprofit education and research institute that is supported and directed by its members. Its mission is to provide responsible leadership in the use of land in order to enhance the total environment.
ULI sponsors educational programs and forums to encourage an open international exchange of ideas and sharing of experience; initiates research that anticipates emerging land use trends and issues and proposes creative solutions based on this research; provides advisory services; and publishes a wide variety of materials to disseminate information on land use and development.
Established in 1936, the Institute today has some 13,000 members and associates from 46 countries representing the entire spectrum of the land use and development disciplines. They include developers, builders, property owners, investors, architects, public officials, planners, real estate brokers, appraisers, attorneys, engineers, financiers, academics, students, and librarians. ULI members contribute to higher standards of land use by sharing their knowledge and experience. The Institute has long been recognized as one of America's most respected and widely quoted sources of objective information on urban planning, growth, and development.
Richard M. Rosan Executive Vice President
Project Staff Senior Vice President, Research, Education, alld Publieations Rachelle L. Levitt
Vice President/Publisher Frank H. Spink, Jr.
Managing Editor Nancy H. Stewart
Manuscript Editor Carol E. Soble
Electronic Publishing Solutions of Annapolis, Maryland Book Design;Layout: David M. Williams Cover Design: Melinda S. Appel
Production Manager Diann Stanley-Austin
Word Processing Joanne Nanez
About the Center for Urban Policy Research, Rutgers University
The Center for Urban Policy Research at Rutgers University is among the nation's oldest and most active research organizations dedicated to the study of urban policy. CUPR has specialized in studies of housing, development impact, economic development, urban and suburban land use, transportation, and the environment. Founded in 1969, the center brings together a highly experienced full-time faculty of planners, economists, geographers, and computer and systems experts. The center also supports an in-house publications department.
The center has carried out more than $4 million in research grants, published more than 100 books, and organized many national conferences. CUPR has conducted research for federal, state, and local agencies, foundations, and private sector clients. Recent studies conducted by the center include the economic assessment of the New Jersey State Development and Redevelopment Plan; analysis of national housing mobility strategies; preparation of an urban transportation masterplan; development of a model subdivision and site plan ordinance; housing needs assessments for New Jersey and Westchester County, New York; and a multiyear investigation of community involvement in the siting of hazardous waste facilities.
Recommended bibliographic listing: Burchell, Robert W _/ David Listokin, et al. Development Impact Assessment Handbook. Washington, D.C.: ULI-the Urban Land Institute, 1994.
ULI Catalog Number: D86 International Standard Book Number: 0-87420-743-6 Library 01 Congress Catalog Card Number: 93-61040
Copyright © 1994 by ULI-the Urban Land Institute 625 Indiana Avenue, N.W. Washington, D.C. 20004-2930 0-'~ iod~ 0000 Printed in the United States of America. All rights reserved. No part of this book may be reproduced in any form or by any means, electronic or mechanical, including photocopying, recording, or by any information storage and retrieval system, without written permission of the publisher.
ii
Chapter 8
..
Introduction Fiscal impact analysis compares the public costs
and public revenues associated with residential and!or nonresidential growth (Burchell, Listokin, and Dolphin, 1991). If costs exceed revenues, a deficit is incurred; if revenues exceed expenditures, a surplus is generated. Fiscal impacts are projected for the public jurisdiction(s) where growth is taking place---the municipality, township, county, school district, and any special districts (Marcou and Tischler, 1978).
Fiscal impact analysis-or, as it has sometimes been called, cost-revenue analysis- has been part of the planning profession for over half a century. It started with a narrow application and over time has broadened to encompass a wide range of uses. The growing breadth of cost-revenue assessment is evident from the following historical synopsis.
Planners first employed fiscal impact analysis in the 1930s to examine the effects of the nascent public housing program. Expenditures for the program were justified on the basis that public housing projects yielded a net local fiscal surplus compared to the slums they replaced. In the early 1940s, fiscal impact analysis was similarly used in the context of urban renewal programs to demonstrate the local fiscal advantages of the new land uses that would replace the old.
Over time, the locus and context of fiscal impact analysis broadened (Levin, 1975). During the massive suburbanization movement of the 1950s, cost-revenue studies were effected to gauge the impact of new single-family detached homes and apartments on local school districts and municipal service providers. In the 1960s, supported by U.s. Department of Housing and Urban Development 701 planning assistance
FISCAL IMPACT ANALYSIS
funding, fiscal impact analysis was used to evaluate the economic effect of master plans. During this same period, cost-revenue projection was also applied to weigh the costs versus revenues of annexation tu both the annexing and annexed jurisdictions.
In the 19705, the techniquc emerged as an almost universal large-scale development accompanimenteither undertaken voluntarily by the deVeloper or required by the municipality (Levin, 1975; Cuthbertson, 1976). Public jurisdictions such as the Association of Bay Area Governments began to incorporate costrevenue assessment into their planning and other activities (Lewis and Hoffman, 1977). The 1970s also saw the rise of fiscal impact models. The Urban Institute was a pioneer in this regard (Muller, 1975); its work eventually led to computerized cost-revenue approaches such as the Municipal Impact Evaluation System (MUNlES) developed by Tischler and Associates, Inc.
The 1970s also marked the publication of The Fiscal Impact Handbook by Rutgers University (Burchell and Listokin, 1978). Building on earlier work conducted at Rutgers in a publication entitled Housing Development and Municipal Costs (Sternlieb et aI., 1972), the Handbook was developed as a baseline document for wide application by planners, developers, and others. The Handbook brought greater methodological consistency to a field that was characterized by divergent and often questionable approaches. Publication and dissemination of the Handbook led to Widespread use and acceptance of fiscal impact analysis.
By the 198Os, cost-revenue assessment had become a common, albeit not universal, element of development impact and planning assessment (Montasser and Tischler, 1980). A cost-revenue projection was
125
typically included in the socioeconomic section of environmental impact statements (see chapter 5); such a projection was required with respect to larger or in other ways atypical projects such as developments of regional impact (ORIs) in Florida. Annexation was often reviewed on its fiscal merits; rapidly growing communities ranging from Germantown, Tennessee, to Naperville, Illinois, routinely prepared technical guides to evaluate the fiscal effects of annexation on their borders (Tischler and Associates, 1988; Naperville, 1981). Planners considering changes to a community's master plan also factored in the various fi~cal consequences of land use altematives, along with other social and environmental effects. For instance, Newark, Delaware, developed a cost-revenue handbook to assess rezoning and other land use changes. Fiscal impact analysis was also increasingly used in a policy context ranging from decisions on the expenditure of economic development funds (i.e., whether the public subsidy would be recouped from the project-induced fiscal surplus) to the location of military facilities.
In the 1990s, fiscal impact analysis is finding application in still-emerging planning contexts. Emphasis on growth management has intensified interest in comparing the fiscal effects of development under a planned or managed land use system to the impacts of growth under sprawl or trendline conditions (American Farmland Trust, 1986). In Maryland, for instance, both Montgomery and Howard Counties have considered the fiscal as well as the traffic, environmental, and other effects of alternative future land use patterns (Montgomery County, 1989; Tischler and Associates, 1989). The state of New Jersey commissioned a study by Rutgers University to analyze the revenue implications of a state plan that will modify existing development patterns (Burchell, 1992). Communities are also recognizing that fiscal impact consequences should be factored into the determination of impact fees; growth resulting in a revenue surplus should receive a credit in the calibration of the impact charge. This fiscal impact-impact fee linkage has been incorporated into the formulation of impact fees for NaperVille, Illinois, and into the development of proffer charges in Virginia Beach, Virginia (Listokin, 1988; Burchell and Listokin, 1990).
Fiscal impact analysiS is also becoming an element in many community planning processes that traditionally have never incorporated cost-revenue considerations. An example is a capital facilities needs assessment. In Florida, for instance, state law mandates communities to prepare a capital improvement clement (CIE) that must include a projection of both future infrastructure needs and the revenues required to finance the identified capital improvements. Florida communities such as Venice have conducted fiscal impact studies of future growth to help satisfy their CIE planning requirement. Fiscal impact analysis is also being used in the context of personnel planning.
For instance, Plymouth, Minnesota, applied a communitywide cost-revenue assesslnent to determine, among other things, the level of staffing various city departments would need to accommodate expected growth.
In short, fiscal impact analysis has evolved from a technique narrowly applied to justify public housing and urban renewal spending to a tool that today is broadly used in a wide variety of planning contexts. Similarly, the fiscal impact assessment's analytic state of the art has evolved over the past half-century as noted in the discussion below.
Evolving State of the Art Over time, fiscal impact analysis has changed from
an ad hoc and overly simplistic projection that employed the same methodology in all cases to a more standardized and comprehensive assessment that encompasses different approaches suitable to varying applications. These changes make for a much more accurate analysis today than in years past.
In the mid-1970s, the authors considered the national state of the art of fiscal impact assessment (Burchell and Listokin, 1978). At that point, fiscal impact analysis favored direct, average-costing procedures that typically employed the per capita multiplier technique. Under this technique, service costs per unit of population (persons, pupils, and employees) would be derived and then applied to the devetopment-generated population (persons/pupils/ employees). These costs were then matched against growth-induced revenues to yield the net fiscal impact.
In assessing the state of the art as of the mid-1970s, the authors observed several characteristics. In most instances, the cost-revenue analysis was performed in a singular case-by-case fashion, and the per capita approach was applied arbitrarily. In addition to the absence of standardization, fiscal impact analysis typically focused on the end-state and was overly simplistic. It usually did not consider the effects of development over time as opposed to merely at final buildout. Neither did it factor in interactive effects such as the impact of local ratable additions on intergovernmental assistance.
While many of these methodolol;ical weaknesses continue today, noticeable improvements have emerged. In the movement to more standardized approaches, the per capita method-still the most common technique-is applied much more unifonnly. Other refinements are obvious as well. Even though average costing is still the most common application, it is tempered with an enhanced sensitivity to marginal impacts. Such impacts are often determined from case study interviews of local public officials knowledgeable of service needs and capacities. Another change is that the time frame of the analysis has shifted from exclusively end-state to periodic. As important as it is to define a development's fiscal consequences at buildout, it is likewise in
126
I r
sightful to trace its year-by-year effects on the way to completion--often through the econometric modeling application of cost-revenue analysis.
Finally, the base data employed in fiscal impact cost projections have been refined. Ongoing efforts attempt to derive accurate demographic multipliersthe average number of people and school children associated with different type and size configurations of housing units. Demographic multipliers-a key input in the estimation of public service costs-have been updated periodically by the authors in The Practitioner's Guide to Fiscal Impact Analysis series (Burchell and Listokin, 1980; Burchell, Listokin, and Dolphin, 1985) and by other sources such as the Illinois School Consulting Service.
The revenue side of the fiscal impact equation has also been marked by improved accuracy. Not only is the property tax more precisely estimated by considering actuallocal assessment practices, but the analysis gives greater consideration to nonproperty tax revenues such as local fees and charges and intergovernmental aid (Stein, 1976). These nonproperty tax revenues are important sources of income and may be enhanced and, at times, even reduced by proposed development. For instance, the introduction of a significant nonresidential ratable or certain types of residential development (such as an age-restricted housing complex) may, by increasing the local property valuation per pupil, reduce the state school aid received by a locality. Current application of fiscal impact analysis incorporates such changes into revenue flows.
Improvement is also evident in the analysis of nonresidential facilities. In addition to developing a better sense of how such development influences public revenues and affects local nonproperty and property tax collections and intergovernmental aid, researchers are continually searching for an improved definition of the public service cost consequences of nonresidentialland uses. To date, refined applications of both the per capita approach and intensive case study analysiS hold the greatest promise of enhancing the analysis of the fiscal effects associated with the development of nonresidential facilities.
In sum, fiscal impact analysiS has evolved from an ad hoc assessment to a more standardized and accurate discipline. In addition, the number of techniques available to the analyst has expanded. In years past, only one method was available for application-a simple version of the per capita multiplier approach. Today, the per capita method has been refined and supplemented by additional methodologies, indudlng the case study and econometric approaches.
Fiscal Impact Analysis: Methodology Fiscal impact analysis is a studied, technical under
taking. Step-by-step procedures to carry out the per capita, case study, and other methodologies are detailed
in a number of procedural guides. The most comprehensive guide is the Rutgers University publication entitled TIre Fiscal Impact Halldbook (Burchell and Listokin, 1978). The Halldbook has been synthesized and updated in a periodic series called The Practitioner's Guide to Fiscal Impact Analysis (Burchell and Listokin, 1980; Burchdl, Listokin, and Dolphin, 1985). In addition, a useful overview of the different techniques is provided in a 1988 report published by the International City Management Association and entitled"AnalYZing the Fiscal Impact of Development" (Tischler, 1988).
While it is not appropriate to duplicate the detail of the several procedural gUides, it is instructive to summarize the substantive content of cost-revenue assessment. Four basic procedures guide all fiscal impact methods as follows: • determine the population generated by growth
people, school-age children, and employees; • translate this population into consequent public
service costs; • project the revenues induced by growth; and • compare development-induced costs to revenues:
if costs exceed revenues, a deficit is incurred; if revenues exceed expenditures, a surplUS is realized. The three fiscal impact metllods-per capita, case
study, and econometric--<liffer mainly in the "translation" of population into public selvices and costs and, to a lesser extent, in terms of their revenue projections and comparison of costs to revenues. The differences will become evident from the following discussion of the procedures and methods of cost-revenue assessment.
Projecting Population It is important to identify the aspects of growth
that affect public service provision. For residential development, the population and pupil generation associated with different housing configurations is a major influence on municipal and school district operating and capital obligations. The housing types that arc most population-intensive place the greatest cost burden on the public sector in terms of accommodating growth.
The fiscal impact analyst uses demographiC multipliers to predict the poputalions that will result from new housing development. As discussed in chapter 6, multipliers calculate the number of the two principal users of local services: people, for municipal services; and school-age children, for school services. The multipliers for household size represent the average number of persons living in a housing unit and vary according to the type and size of housing units. Housing type refers to single-family (detached) homes, townhouses, garden apartments, high-rise units, etc.; size is expressed by number of rooms or bedrooms. As might be expected, detached single-family units are~ on average, associated with larger household sizes than are attached multifamily units while larger units house more household members. The sources
127
for demographic multipliers are detailed in Appendix II of this handbook and are summarized shortly.
The multipliers for school-age children represent the average number of children of school-attending age and are generally specified according to grade category (i.e., K-<i, 7--<l, 9-12). In conducting a fiscal impact study that considers effects on public services, the analyst typically focuses on public school-age children or the share of school-age pupils attending public schools.
The discussion thus far has focused on the attribute of residential growth-population (residents and public school-age children)-that prompts a public service response. While agreement is Widespread that demographics are the appropriate "unit" to be considered with respect to residential development, the debate over the corresponding nonresidential "unit" continues. Several studies, however, have concluded that it is the employment intensity of nonresidential development that prompts public service costs (Beaton, 1983). In other words, all things being equal, a nonresidential facility that introduces more jobs into a community will generate a more significant and costly public service response for both operating and infrastructure development outlays than a sister nonresidential project that is less labor-intensive.
Employment intensity is typically expressed in terms of the number of employees per 1,000 square feet of nonresidential space. It is higher for certain categories of use such as office development than for warehousing, for example. Data sources for the num
ber of employees per 1,000 square feet are discussed shortly.
The demographic multipliers for different type and size housing units and the employment-intensity levels for the different nonresidential land uses are applied against the matrix of the project pro forma to yield the anticipated project-induced populationpeople, pupils, and employees. An illustrative example is shown in Exhibit 8.1. The hypothetical mixed-use development encompasses 300 residential units and 150,000 square feet of nonresidential space. The residential component includes 200 three- and four-bedroom single-family detached homes and 100 two-bedroom townhouses; the nonresidential sector, 100,000 square feet of office space and 50,000 square feet of retail space. Applying household size and school-age children multipliers for the different type and size housing units planned for the project yields a development-induced population estimate of 891 people and 192 pupilS. (For the illustrative example, the school-age children projection of 192 students is not reduced to a public school-age pupil increment. In a comprehensive study, this last step would be included.) Applying a ratio of 3.0 employees per 1,000 square feet of office space and 2.5 employees per 1,000 square feet of retail space yields a projection of 425 employees from the nonresidential component of the mixed-use development. As in the social impact assessment, these population projections serve as the starting point for the identification of attendant effects.
Exhibit 8,1: EXAMPLE OF A PER CAPITA FISCAL IMPACT METHOD:
Number of Development Units (in Composition square feet}
Residential
Single-Family DC'tached
lhree-Bedroom 100
Four-Bedroom 100
Townhouse Two-Bedroom 100
Nonresidential
Office 100,000
Retail 50,DOD
Total
NUMERIC EXAMPLE
Popul<ltion (per unit or Project-Generated 1,000 square feel) Population2 Cost per Unit3 Project-Generated Cost
People1 Pupils! Employees People Pupils Employees reople Pupils Employees People Pupils Employees Total
3,09 0.67 NA 3.77 1.14 NA
2.05 !J.l1 NA
NA NA 3.0
NA NA 2.5
I
309 67 NA $581 $6,571 NA $179,529 $4~0,257 NA 377 114 NA $581 $6,571 NA $219,037 $749,094 NA
205 11 NA $581 $6,571 NA $119,105 $72,281 NA
NA NA 300 NA NA $247 NA NA $74,100
NA NA 125 NA NA $247 NA NA $30,875
891 192 425 $581 $6,571 $247 $517,671 $1,261,632 $104,975 $1,884,278
Noles
In this example, public school-age children are assumed for illustrative purposes to equal the school-age children count. In reality, the former will typically be 80 percent to 90 percenl of the latter. In a full, formal fiscal impact analysis, the public school-age children figure would be the basis for calculating public education costs. NA = Not applicable I Derived from the demographic multipliers obtained from the American Housing Survt'lJ (see Appendix II). 2 Equals number of units/square feet multiplied by the respective population/employee profiles. 3Cakulated as shown in Exhibit 8.2
.. 128
Projecting Costs Once the population introduced by growth is deter
mined, the next step is to translate the increment of people, students, and workers into added public services and costs. The Fiscal Impact Handbook, The Practitioner's Guide series, and the other reference volumes cited earlier all detaU several techniques for deriving the associated services and expenditures while the major methods are summarized below.
Per Capita Method The per capita method first determines current
public service costs on a per unit basis-per pupil for the school district and per capita and per employee for the municipality, township, village, county, and any special districts (i.e., fire, park, community college). The per student outlay is readily determined. Total school costs or the total school property tax levy is divided by the total pupil enrollment to yield the total cost or property tax expense, respectively, per student.
With noneducational services, however, it is incorrect simply to divide incurred outlays by the local population because such services benefit both residential and nonresidential land uses. Service costs must therefore be apportioned between these two types of development. The residential share of all residential and nonresidential service costs is estimated by dividing the residential property value and number of parcels by the residential and nonresidential property
values and the number of parcels, respectively. The calculation produces the residential percent of the residential/ nonresidential parcels and the residential percent of the residential/nonresidential property value. The two results are averaged, and the combined value is then applied to the total local mWlicipal costs to derive the estimated residential-associated share.
ThlS analysis can be illustrated by referring to the example presented in Exhibits 8.1 and 8.2. In the hypothetical community of 25,000 residents and 10,000 employees, municipal outlays totaled S17 million. The local tax base, comprising 10,000 parcels, amounted to $800 million. Of this total, 9,600 residential parcels were valued at $600 million, and 400 nonresidential parcels were valued at $200 million. The residential share of total valuation and total parcels was therefore 75 percent and 96 percent, respectively, for a combined average of 85.5 percent. The 85.5 percent figure is applied to the total municipal outlay of $17 million to yield estimated residential-associated expenditures of $14.535 million; the remaining $2.465 Il\illion is assigned to services associated with nonrestOentialland uses. With a local population of 25,000 and workforce of 10,000 employees, the municipal costs per capita total $581 ($14,515,000/25,000); the service outlay per worker amounts to $247 ($2,465,000/10,000).
Per pupil costs in the illu~trative community are calculated much more Simply. The local school district incurred outlays of $23 million to educate a student population of 3,500. The per student expense is therefore $6,571.
Exhibit 8.2: DETERMINATION OF FISCAL IMPACT COST PARAMETERS FOR THE PER CAPITA METHOD APPLICATION
1. Expenditures Total Municipal Expenditures
2. Parcels Total Parcels
Residential Parcels Residential Parcel Percentage
3. Assessed Value Total Assessed Value
Residential Parcel Assessed Value Residential Value Percentage
<1. Expenditure Parameters Estimated Share of Residential-Associated Expenditures
~stimated Municipal Residential-Associated Expenditures (lx4)
Total Local Population Municipal Expenditure per Capita Total School Expenditures Total School Population School Cost per Pupil Total Nonresidential-Associated Expenditures Total Local Employees Municipal Cost per Employee
$17,Goo,000
1O,0GO 9,600
96%
$80G,000,000 $600,000,000
75%
85.5% (96% + 75%) $14,535,000
25,000 $581 ($14.535 million + 25,000) $23 million 3,500 $6,571 $2.465 million ($17 million - $14.535 million) 10,000 $247 ($2.465 million + 10,000)
Source: Municipal and school district budgets.
129
Growth-induced public service costs are then factored by multiplying the per capita cost by the total number of people, employees, and pupils introduced by development. As described earlier, the illustrative mixed-use development was projected to add 891 people, 425 workers, and 192 students. At a per unit cost of $581 per capita, $247 per employee, and $6,571 per pupil, the per capita method would project costs of about $0.6 million to serve the new residents and added workforce and about $1.3 million to provide public education.
Service costs can also be expressed on a per unit basis. For instance, in the illustrative example, the singlefamily detached unit contains an average of about 3.43 people and 0.91 school-age children. (This discussion aggregates the three- and four-bedroom singlefamily detached units; a slightly different result is obtained when each type of single-family home is considered individually.) At a per capita and per pupil cost of $581 and $6,571, respectively, the single-family detached home generates $7,973 (3.43 x $581 + 0.91 x $6,571) in public service outlays. For the townhouse unit, which has a lower household size and schoolage children yield (2.05 and 0.11, respectively), public service costs amount to $1,191 (2.05 x $581) for municipal services and $723 (0.11 x $6,571) for schools for a total outlay of $1,914---less than one-quarter of the expenditure for the single-family detached home.
The per capita method is the classic average-costing approach. Service costs are projected at the average per unit outlay per capita, pupil, and employee. 'The technique is straightforward, relatively easily effected, and, in most cases, yields a quick handhold on the impacts of development. For these and other reasons, it is the most common technique employed in the field. As such, step-by-step procedures for implementing the per capita method are incorporated into the preview model.
Case Study Method The case study approach relies on intensive site-spe
cific interviews of public officials knowledgeable of local service conditions and capacities as the primary means of determining the effects of population growth on public services and costs. The interviews identify the anticipated marginal costs of growth given conditions of excess or deficient service capacity. In the case of excess capacity (capacity beyond that needed to accommodate the existing population at current service levels), growth will add to costs atlower-than-average per capita/ student/employee levels. In the case of deficient capacity (capacitY below that needed to accommodate the existing population at current service levels), growth will add to costs at higher-than-average per capita/student/employee levels.
The case study method thus departs from the per capita approach in both its assumption and approach. The per capita technique projects population-induced costs at the average per unit outlay, which is detennined ar
ithmetically by dividing local service costs by the current service population. By contrast, the case study approach focuses on defining the marginal response to growth through intensive field-level investigation.
These differences are illustrated by the application of the case study approach to the hypothetical mixeduse development. According to the per capita approach, the proposed development would require municipal outlays of about $0.6 million to serve new residents and employees and school outlays of about $1.3 million to accommodate new pupilS. These figures were based on the existing local costs per capita, employee, and pupil. According to the case study approach, expenditures would be ascertained from interviews that would detail the specific personnel and capital equipment necessary to accommodate the proposed growth. For the illustrative project, the case study approach estimates school costs at about $1.1 million-$0.2 million less than the per capita approach because of the system's excess capacity. The case study, however, reveals a higher municipal service outlay than that indicated by the per capita approach-$0.7 million compared to $0.6 million. The added cost reflects deficient municipal service capacity, especially with respect to the volunteer squads, which will require municipal outlays for capital equipment. The level of detail of anticipated effects is revealed only by application of the case study method.
Econometric Method Both the per capita and case study approaches
translate population into service costs in terms of a given set of service conditions and cost parameters determined either arithmetically or through site study, respectively. The econometric approach goes beyond the "one-snapshot-in-time" analysis to a "moving picture" of the community and its development profiles and tKe profiles' changing effects and interactions over'time. It employs a basic equation that relates a jurisdiction's public service expenditures to its revenue parameters-tax base, tax rate, etc. The cost and revenue parameters embodied in the equation typically start at the existing average levels----€.g., average costs per capita, average assessed value per unit, etc. To this equation are introduced historical and current data matrices that allow past steady-state and current development conditions (both the target development under study and other simultaneous development) to affect future projections. Data matrices may include local expenditure and revenue levels per capita; the development's pro forma schedules, including buildout by type of land use; value by category of land use; estimates of annual expenditure, revenue, and real property valuation change under steady-state conditions; and estimates of annual expenditure, revenue, and real property valuation change under target and other development conditions.
The econometric approach yields a picture of development impact at the end of the development period as
130
well as at multiple interim stages. nus technique also compares steady-state and other-than-target development conditions to those of the development under examination. In contrast to the typically fixed view afforded by the per capita and case srudy approaches, the econometric approach allows fiscal comparisons of growth with and without the target development and thus presents an unfolding and often-changing perspective on the fiscal consequences of growth.
The hypothetical example is useful in illustrating the differences among the three fiscal impact techniques. Given that the scltool district spent $23 million to educate 3,500 pupils, the per capita approach used a $6,571 average per pupil cost parameter to project growth-induced school outlays for the end-state of development. Thus, if the hypothetical project added 38 pupils a year over a five-year buildout, the development would at completion introduce 192 students and incur projected growth-induced school costs of $1.3 million (192 x $6,571). For its part, the case study identified some slack school capacity and, as such, projected total school outlays of$1.1 million at the deVelopment's completion or a $5,700 per pupil cost to educate the 192 project-induced students.
By contrast, the econometric approach starts with the eXisting average $6,571 per srudent cost factor but modifies the cost over time as the target development, other development, and steady-state conditions interact into the furure. By introducing a more efficient service population, the population "flows" that result from the incremental addition of new households might yield a $6,500 per srudent factor in the first year of the projection, a $6,400 per pupil outlay in the second year, and so on. The econometric approach, unlike the per capita and case srudy applications, would incorporate the year-byyear differences into its annual forecasts. Thus, first-year schaul expendirures would total $247,000 (38 x $6,500); second-year expendirures, $486,400 (76 x $6,400); and so on to the development's end-state. Similarly, the econometric approach would apply a year-by-year cost analysis to municipal outlays, in each case factoring how the target development affects local expendirures over and above steady-state trends and other development impacts.
Whatever tedmiquc is applied-per capita, case study, or econometric-the translation of population growth into public service costs yields one side of the fiscal impact calculation. These costs must then be matched against development-induced revenues.
Projecting Revenues Public service jurisdictions rely on revenues that in
clude both local and extralocal sources. Local sources comprise a variety of levies, while extralocal sources pertain to intergovemmental transfers from the state and federal governments.
The local levies are usually the more significant sources of income and encompass taxes and charges
and miscellaneous revenues. The most significant tax is commonly the levy on property. Other taxes include levies on personal property, utility use, consumer sales, and income. In addition to taxes, government jurisdictions receive income from fees and assorted revenues from interest earnings, permits, charges for services, fines and penalties, etc.
In considering growth-induced revenues, the full matrix of public revenues should be examined as follows: I Own-Source Revenues
A. Taxes 1. Real property 2. Personal property 3. Utitity 4. Income 5. Other
B. Charges and Miscellaneous Revenues 1. Interest, rents, royalties 2. Licenses and permits 3. Charges for services 4. Fines 5. Sales of fixed assets 6. Other
II Intergovernmental Revenues A. Slate Aid B. Federal Aid In projecting specifically how growth affects both
local and extralocal revenues, the fiscal impact analyst first considers the basis for each revenue source and then examines how development will affect each source. To illustrate, the property tax is a percentage levy on the value of land and improvements (real property). In some jurisdictions, a property tax is also imposed on personal property as well as on the value of automobiles. To project the property tax revenues from growth, the analyst first determines the assessed value of the development in terms of its property tax constituent components-real, personal, and automobiles. The assessed valuation is then multiplied by the prevailing property tax rate.
Similar step-by-step calculations shown in Exhibit 8.3 project revenue collections associated with the other revenue sources. To illustrate, many governments receive income from the sales tax. To calculate growth effects, the analyst estimates how development will add to local sales and then projects the concomitant gain in sales tax revenues based on the applicable sales tax rate. Another source of local income is interest earnings typically derived from investing major revenues before they are disbursed to pay for local expenses. The local interest earnings from growth can be determined by calculating how development adds proportionately to local revenue resources (i.e., property tax base) that can be invested.
In certain instances, revenue collections are related. to population size. For example, many local governments raise funds from fines/lioenses and pennits. As growth adds to the population base, the community will levy more traffic violations and overdue library charges, collect
131
Exhibit 8.3: FISCAL IMPACT ANALYSIS: MUNICIPAL REVENUE CALCULATION EXAMPLES
REVENUE SOURCE REVENUE FORMULA
I Own-Source Revenues I.A. Taxes I.A.I. Property Tax
Market value x Assessment factor Assessed val ue
Assessed value x Property tax rate Property tax revenue
I.A.2. Utility Tax Utility consumption Utility tax rate
X Utility tax revenue(Le., kWh consumption) (i.e., per kWh)
I.A.3. Sales Tax
Added local sales x Sales tax rate Sales tax revenue
I.A.4. Other Local Taxes I.A.4. a. Real Estate Transfer
Market value Transfer Annual turnover Real estate transfer of property
x tax rate
x percent sales tax revenue
I.A.4.b. Motor Vehicle Sticker Vehicle ownership by Vehicle sticker fee Motor vehicle sticker
type / size of housing unit x
per automobile tax revenue
I.A.4.c. Transient Occupancy Tax Daily Daily Nwnber Transient TransientNumber of
x occupancy x rate per x of rooms x occupancy occupancymotel rooms
rate room rented daily lax tax revenue
LB. Charges and Miscellaneous Revenue Assessed value of growth
+ x Existing interest income Interest revenue Existing community tax base
Population by type /size of x Per capita license License and housing unit and permit income permit revenue
I.B.I. Interest Earnings
I.B.2. Licenses and Pennits
I.B.3. Charges for Services
1.8.4. Fines and Forfeits (Examples)
II Intergovernmental Revenues State Assistance II.A. Income Tax Redistribution
II.B. Motor Fuels Redistribution
II.c. Other State Aid
Population by type/size of x housing unit
Vehicle ownership by type /si.ze of
housing unit
Population by type / size of
housing unit
Population by type / size of
housing unit
Population by type / size of
housing unil
Population by type / size of
housing unit
Per capita service charge income
x
Per capita automotive fine revenue
x Per capita
library fine
x
x
income
Per capita income tax
redistribution
Per capita motor fuels
redistribution
Per capita
Service charge revenue
Automotive fine
revenue
library fine
revenue
Income lax
revenue
Motor fuels
revenue
Other x "other" state aid
state aid revenue
more bicycle registration fees, etc. Therefore, local income from fines and licenses can be expected to increase as a proportionate share of population
Intergovernmental income is projected similarly. Thus, if a state granted $100 per pupil for textbooks, the development-associated income for such aid would be equal to the number of new students multiplied by $100. Where school aid fonnulas are more complex and, for example, incorporate ratios between the local versus state average equalized valuation per student, the analyst would then ascertain how these relationships would change with the onset of development and thus how state support fonnulas would be altered.
By applying the formulas depicted in Exhibit 8.3, the hypothetical example generates a total $2.2 million in revenues. Of this amount, the largest share, $1.4 million or about two-thirds, is attributable to the property tax; the remaining $0.8 million reflects the array of local nonproperty and intergovernmental aid sources described above. These figures would typically be disaggregated to the unit level. In the hypothetical example, the single-family detached homes (combining the three- and four-bedroom units) generate about $7,200 in total annual revenues; the lowerpriced townhouses yield $3,800 in total annual revenues.
In projecting growth-induced costs, fundamental differences may be observed among the per capita, case study, and econometric approaches. In determining revenues, the differences are more a matter of nuance. To illustrate, in the case study approach, the analyst typically spends the most time interviewing public officials responsible for revenue-municipal and school business administrators, the tax assessor, tax collector, etc. Further, in both the per capita and case study approaches, established tax parameters (i.e., a $2.00 property tax rate) are applied to forecast the growth-induced revenues at the end-state of devel
opment. By contrast, the analyst applying the econometric approach starts with a tax value such as the property tax rate but would modify the rate annually into the forecasting period as the target development, other development, and steady-state conditions affect the jurisdiction's fiscal posture and annual property tax levy. Thus, the first-year tax rate could be $2.00 but could decline to $1.90 in the second year, to $1.85 in the third year, and 50 on as ensuing development results in a fiscal surplus. This interactive relationship over time between development and the tax rate differentiates the econometric approach from the per capita and case study methods.
Comparing Costs to Revenues Once the growth-induced population is projected,
the next step is to determine the basic values for a fiscal impact assessment by translating the forecasted population into costs and estimating the developmentgenerated revenues. Costs are matched against revenues; if they exceed revenues, a deficit i~ incurred; if they fall short of revenues, a surplus is realized. When the per capita and case study approaches are applied, the cost-revenue comparison is typically expressed as an end-state value at development buildout; when the econometric technique is used, the cost-revenue comparison is usually presented as a series of annual net results. The results are typically expressed for both the overall project and the individual development categories.
In the illustrative example, application of the per capita approach yielded total annual service costs approaching $1.9 million versus annual revenues of about $2.2 million. The result is a net fiscal surplus of approximately $0.3 million. The result is a composite of the different fiscal impacts of the respective project components. Thus, as shown below, the single-family
SUMMARY OF THE FISCAL IMPACTS OF THE ILLUSTRATIVE EXAMPLE
Fiscal Impact per Unit (per residential unit! per 1,000 square feet)
Number Units/ Total Project Net Project Component 1,000 Square Feet Costs Revenues Net Impact Fiscal Impact Residential
Single-Family Detached1 200 $7,973 $7,190 ($783) ($156,600) Townhouses 100 $1,914 $3,804 $1,890 $189,000
Nonresidential Office 100 $740 $2,993 $2,253 $225,300
Retail 50 $616 $2,394 $1,778 $88900
$346,600
Rounded to $347,000
Note 1 Averages the results of three- and four-bedroom units, i.e., uses a 3.5 household size and 1.0 school-age child multiplier for "blended" 3.5 bedroom single-family detached units.
133
detached homes with high household and school-age children profiles produce a fiscal deficit, but the loss is more than offset by the surplus associated with the attached townhouse units and the project's office and retail uses.
The above discussion shows the fiscal impact result derived from the per capita method. With other methods, the finding would be somewhat different. For instance, wherever the per capita approach estimated annual service costs at about $1.9 million, the case study estimated a lower service outlay of $1.8 million-Dr $0.1 million less-due to the method's consideration of considerable excess school capacity. The net fiscal surplus would therefore be $0.1 million more by applying the case study method or a total of $0.4 million.
Selecting an Appropriate Fiscal Impact Method
Thus far, the three fiscal impact methods have been described in terms of their application. But it is important to consider the basis for their selection. Under what conditions would the analyst use the per capita, case study, Or econometric technique?
One consideration relates to the appropriateness of average costing. When the service system capacity bears a close relationship to service demand and the average cost of providing services to current users is a reasonable approximation of the cost to provide services to future users, the average-cost approach is most appropriate and dictates use of the per capita approach, which stresses average per unit service commitments. By contrast in jurisdictions with considerable slack or deficient service capacity, average per unit service costs would misstate the true effects of growth. The average would overestimate the needed service commitment in instances of slack capacity and would understate the expensive service response necessary under conditions of deficient capacity. Accordingly, the case study, which is most sensitive to local conditions and anticipated effects, would be appropriate.
The case study is also the best method for determining the impact of specialized nonresidential facilities, including military bases, energy plants, regional shopping centers, and other nonresidential development that does not lend itself to Simple average per unit cost projections. For similar reasons} the case study is suited to assessing the fiscal impacts of atypical residential development such as group homes or other facilities for special needs populations.
When is the econometric method applied? As an average-costing approach, the econometric method should not be used in situations where considerable slack or deficient capacity has been identified. Similarly, it is not appropriate for the specialized conditions best handled by the case study. The econometric
method is instead used in "standard" development cases where average per unit costing reflects public service realities. In other words, the same general conditions that dictate the selection of the per capita method apply as weB to the econometric method. The analyst, however, applies the econometric approach as opposed to the per capita technique to the analyses of unusually large developments with commensurately long build-out periods (e.g., a planned unit development comprising thousands of units that will enter the market over a decade). The very size of the developments might alter some of the basic conditions affecting a community's cost and revenue parameters (e.g., the local school-age population and reliance on local versus intergovernmental revenue). In addition, large-scale projects often produce different impacts over the course of their buildout. Because the econometric technique's equation incorporates a jurisdiction'S interactive and evolving cost-revenue effects over time, it is most suitable for ascertaining the impacts of large, multiyear projects.
A price, however, is attached to the use of the econometric method. The technique requires considerable setup time for the basic modeling and is therefore much more expensive than the other techniques discussed, especially the per capita method. Another consideration is that the advent of desktop computers permits the increasingly sophisticated application of the per capita method. As a result, that method can now incorporate some of the features that were once unique to the econometric technique. For instance, a per capita approach, once typically limited to showing only end-point results, may now incorporate intermediate fiscal outcomes as well.
In the field, the per capita method is by far the most commonly applied of all fiscal impact techniques. It therefore serves as the basis for the preview model. First, however, the discussion turns to requisite data sources for conducting a fiscal impact study.
Data Sources Whichever method is selected, several data ele
ments must be collected. The most critical information sources are discussed below.
Demographic Multipliers Demographic multipliers indicate the average num
ber of people and school-age children residing in different type and size housing units. They are available from national data sources and local surveys. The former include the Public Use Microdata Sample (PUMS) of the decennial census and the American Housing Survey (AHS). Local data are available from the municipality, schools, developers, and other sources.
134
Appendix II of this handbook contains the most current demographic information available from such macrolevel sources as the rUMS and AHS. It presents the average household size and average number of school-age children, the latter differentiated by grade level. The information is organized around the most common housing types and sizes (as expressed by number of bedrooms).
Another source of multipliers, typically of the school-age children profile, is local surveys usually conducted by or in cooperation with the school board or school superintendent and based on busing or attendance records. In fact, the fiscal impact analyst should usc a local survey to validate the multipliers obtained from the rUMS or AI-lS, particularly when a fiscal impact assessment is applied to specialized housing types. For instance, in projecting the number of school-age children generated by a proposed townhouse development in a ski resort, the analyst should consider the school yields of similar vacation-oriented housing developments in the area, not the general school-age children counts derived from the rUMS or AHS.
A local survey is conducted as follows. First, a list of housing developments similar to the project under study is compiled (i.e., recently completed ski-area vacation homes) along with information on the number of housing units and their bedroom distribution. Second, the number of school-age children (K-12) for each of the identified housing developments is obtained. The pupil counts are available from busing, attendance, and other records maintained by the local schools The third and final step is to divide the total number of school-age children by the total number of units to derive the average number of school-age children per unit for the project under examination (i.e., ski-area vacation homes containing two and three bedrooms).
The discussion thus far has focused on local pupil counts. [t is much more difficult to determine local household size. In some instances, demographic information on household size is locally available from renter and housing purchaser applications as well as from other marketing information maintained by developers. These sources can be used as a rough gauge to determine the average household yield per unit.
The use of locally derived demographic multipliers offers several advantages. First, with respect to fiscal impact assessment, demographic multipliers-used to determine the number of people, particularly school-age children, generated by growth-are often the single greatest source of controversy and thereby frequently call into question the credibility of the results of a fiscal impact assessment. Locally derived multipliers, however, can disarm much of the controversy. Second, local surveys typically provide the most current information and the most jurisdiction-specific data (e.g., for the same school district under study). By contrast, demographic multipliers from the PUMS or AHS may be dated by a
few years and/or may be available only on a larger geographic scale (e.g., a county, Metropolitan Statistical Area, or census region as opposed to a local commtmity or school district).
In practice, obtaining local demographic data is often problematic. Inlormation on the number of school-age children generated by specific types of development may not be recorded or may be held confidential and kept from the analyst. Even if school yields from completed projects can be obtained, the nwnber of units and bedroom distribution of the development in questionespecially the latter--are frequently difficult to ascertain. These hurdles are compounded with respect to the overall project-generated population because no counterpart to a school district keeps records on the number of people in different types of housing. Finally, a common problem with available sources of local information is insufficient sample size Of, worse, reliance on a biased sample.
In summary, despite the several advantages associated with locally derived demographics, the analyst usually has to rely on national sources-the rUMS or AHS. Whenever possible, however, the analyst should seek out local information. A local survey adds to the fiscal impact assessment's credibility and is especially compelling when considering atypical housing types such as vacation homes, units clustered around a spedal use (i.e., marina or golf course), and high- and low-end developments (i.e., estate homes and very low-income housing).
Nonresidential Multipliers Nonresidential multipliers indicate the number of
employees associated with different types of nonresidentialland uses such as office, retail, and industrial development. The multipliers are typically expressed as the worker count per l,OOO-square-foot module and are derived from the sources noted in conjunction with the calculation of the operation-phase direct employment impacts in the economic impact analysis (chapter 7). For instance, information on the number of retail employees per 1,OOo-square-foot module can be derived from the data contained in the Census ofRetail Trade. Office employment intensity is indicated in Trip Generation published by the Institute of Transportation Engineers. In addition to these published materials, the analyst may wish to contact trade and research organizations such as the Urban Land Institute and the International Council of Shopping Centers and local planning, economic development, and tax offices (e.g., certain jurisdictions impose a "head tax" per employee and thus keep records on employment intensity).
Population An accurate count of the public jurisdiction's cur
rent population is necessary for determining accurate
135
per capita/per pupil costs that then become the basis for the per capita and econometric methods. Information on population is available from the local planning department, the school district's vital statistics office, and/or from state, county, or regional planning agencies.
In using the population statistics, the analyst must take advantage of the most current data. Typically, school districts update their census of on-roll students on a quarterly or even more frequent basis. By contrast, figures on the total population are often outdated, especially in intercensal periods.
It is incumbent upon the analyst to update all population counts to the most current period by examining recent local housing production activity as indicated by the number of certificates of occupancy that have been issued and by pairing production with the household sizes indicated by the demographic multipliers (see AppendiX II). To illustrate, if a community reported an estimated 1990 population of 10,000 and added 100 units of single-family detached homes over 1991 to 1992, each with an estimated household size of 3.5, then the community's 1992 population would be estimated at 10,350. The current (1992) population would then be properly paired with current (1992) cost and revenue figures (e.g., total municipal expenditures and total nonproperty tax income related to population) to derive valid fiscal impact parameters (e.g., 1992 per capita municipal expenditures and per capita nonproperty tax revenues).
Costs and Revenues Cost information is obtained primarily from the
budgets of the municipality, school district, and other public jurisdictions affected by growth. TIle analyst must consider the full set of budgetary documents, including a summary volume/section as well as detailed backup materials on operating outlays (for staffing and support expenditures), capital costs (for major purchases), and debt service (for principal and interest repayment on preViously incurred liabilities). The analyst should also consult other documents that reveal the jurisdiction's spending proclivities and needs. Such documents range from the business administrator's annual message that usually highlights changes in the immediate past and upcoming fiscal years to the jurisdiction's official capital facilities plan.
Revenue information is similarly available from the municipal, school district, and county annual budgets as well as from the other documents mentioned above. It is important to go beyond merely identifying the jurisdiction's sources of income to understanding how these monies are allocated. Thus, information on how locally collected but state-distributed revenues (i.e., sales taxes) are transferred often must be obtained from the state treasurer's office. Similarly, information on the status, flow, and future distribution of state and federal intergove=ental transfers, such as
aid to education, must come from the state treasurer's office and/or from federal program sources.
Property Tax Rate/Equalization Ratios The analyst must obtain the property tax rates for
the relevant public jurisdictions and, equally as important, the assessment-to-sales or equalization ratios. These ratios are important for indicating the share of true value at which a property is assessed; a ratio of 0.5 means that a property is assessed for tax purposes at one-half its true value; a ratio of 0.25, at one-quarter of real value; and so on. To project revenues from the real property tax, the fiscal impact analyst multiplies the expected assessed value of the incoming development (as determined by the assessment ratios) by the property tax rate.
Information on the tax rate and equalization ratio is available from the clerk, assessor, or business administrator of the respective public jurisdictions affected by the development. In considering anticipated property tax income, the analyst must consider the actual practice of the tax assessor's office, which may not always comport with the nominal official guidelines. For instance, toward the end of the tax year, a "working" equalization ratio may differ from the published figure due to changing values. This working ratio must be incorporated into the fiscal impact calculation.
Trends/Projections in Expenditures, Revenues, and Populations
The econometric approach accounts for trends/projections in expenditures, revenues, and populations. Information is available from many of the source documents cited above as well as from interviews with public officials. For instance, the school superintendent should be queried with respect to anticipated enrollment trends and the receipt of state aid.
Preview and Quickway Models of Fiscal Impact Analysis
The preview model of fiscal impact analysis follows the per capita methodology-the most widely applied fiscal impact technique-and encompasses the basic steps of cost-revenue analysiS. First, the model projects the number of people/pupils/employees generated by growth. Second, it translates the population increment into attendant public service costs by multiplying the development-generated population by the per person/pupil/employee expenditure factors. Third, the preview model considers the revenues added by growth and, finally, compares costs to revenues to yield the net fiscal impact.
136
Quickway approaches the cost side of fiscal impact analysis by using model factors for the developmentgenerated population and employment. Population and employment arc converted to costs of develop~
ment by applying other model factors for average costs per capila, pupil, or employee. QUickway handles development-generated revenues by applying average equalized tax rates to the development's market value. Included. too, are average relationships between property taxes generated and other sources of revenue.
Quickway's results include a comparison of costs and revenues to the host municipality, school district, and county as well as a statement on the magnitude of the annual fiscal surplus or deficit: small, rnoder~
ate, or large. The level of the annual surplus or deficit is further presented as a share of total annual revenues raised.
The analysis in the preview model is effected through a series of inputs and outputs (see Exhibit 8.4). The inputs include base data either built into the model (i.e., demographic multipliers) or added by the user (i.e., the project pro forma and the local property tax rate). The outputs include the interim and final calculations of fiscal impact analysis such as the development-induced population, service costs, revenues, and the net fiscal effect.
The inputs and outputs of the preview model are discussed in greater detail below. The 23 input factors follow:
Population Faclors. To project the development-induced population and workforce, the demographic and nonresidential multipliers were already applied to the project pro forma in the social impact component of the model. The inputs therefore start with the public service expenditure profile.
Cost Factors (Inputs 1-9}.lnputs 1 through 9 encompass the public service costs for the different public jurisdictions affected by growth, including the municipality, county, school district, etc. To apportion municipal!county expenditures associated with residential and nonresidential uses, respectively, the analyst enters the valuation and number of parcels . contributed by the two respective land use categories. The data are readily available from the local business administrator, treasurer, and assessor.
Revenue Factors (Inputs 10-23). Inputs 10 through 23 include parameters for calculating the developmentinduced property tax, local nonproperty tax, and intergovernmental revenues. As with local costs, these factors are unique to each location; therefore, the specific local values must be entered into the model.
The values are readily available. For the property tax, the analyst obtains the applicable assessment-tosales ratio and the property tax rates from the assessor's office for entry into the model as inputs 10 through 15. For local nonproperty income and intergov€'mmental sources, revenues are expressed on either a per capita basis or a valuation-added basis
(i.e., per $1,000 of assessed value). Obtained from the existing local budget, these values are entered as inputs 16 through 23.
The preview model contains 16 outputs that are related to the input fields as follows:
Population Generation (Olltputs 1-3). For the development under exalnination, the model generates the number of people (output 1) as well as the number of children added, the latter differentiated by school grade level (K--{j, 7-8, 9-12) (output 2). In parallel, where nonresidential or mixed-use development is considered, the model yields the number of employees (output 3).
Cost Generation (Outputs 4-7). The model multiplies the project-induced population by the existing cost parameters contained in inputs 1 through 7 to generate the project-induced outlays required of the municipality, school district, and other public jurisdictions.
Revenue Generation (Outputs 8-12). The property tax determination involves two calculations. First, the assessed valuation (output 8) is computed by applying the assessment ratio to the market value of the project (both previously entered as inputs 10 and 11). The model then applies the property tax rate for the applicable public jurisdictions (inputs 12 and 13) to the assessed valuation to yield the development-induced property tax revenue by jurisdiction (output 9).
The model projects local nonproperty tax income (output 10) by multiplying the project-induced population and valuation (outputs 1, 2, and 8) by the previously determined nonproperty tax revenue per capita and per $1,000 valuation, respectively (inputs 16,17,20, and 21). The intergovernmental revenue generated by growth (output 11) is calculated through a similar procedure.
The sum of the development-induced property tax, nonproperty tax, and federal and state aid (outputs 9 through 11) yields the total income generated by growth (output 12).
Net Fiscal Impact/Effects (Outputs 13-16). The net fiscal impact is determined by comparing the development-induced costs versus revenues. The resulting surplus or deficit figures are indicated individually for all the affected public service jurisdictions as well as in aggregate (outputs 13 and 14). These cost-revenue outcomes are then placed in perspective by relating the figures to total revenues as well as to the property tax levy for each affected public jurisdiction (outputs 15 and 16).
Advantages and Limitations of the Preview Model
The fiscal impact preview model offers numerous benefits. First, it is patterned after the per capita approach, which is the most applicable and widely used cost-revenue method. Its data demands are not burdensome; the input factors are either built into the
137
Exhibit 8.4: PREVIEW AND QUICKWAY MODELS OF FISCAL IMPACT ANALYSIS
INPUT
DEVELOPMENT-INDUCED POPULATION (PREVIEW MODEL)
1. Development-generated residents 2. Development-generated school·age children
3. Development-generated employees
PUBLIC SERVICE EXPENDITURES (PREVIEW MODEL)
1. Total municipal expenditures 2. Total residential parcels
3. Total nonresidential parcels 4. Total re~identidl valuation
5. Total nonresidential valuation 6. Total municipal population
7. Total nonresidential employment 8. Total education expenditures
9. Total school enrollment
PUBLIC SERVICE REVENUES (PREVIEW MODEL) 10. Project value
11. Property·tax-assessment-to·sales rate 12. Municipal property tax levy (rate) 13. School property tax levy (rate)
14. Municipal property tax base
15. School property tax base 16. Municipal nonproperty taxes associated with
property value 17. Municipal nonproperty taxes associated with population
18. Municipal intergovernmental revenues associated with property values
19. Municipal intergovernmental revenues associated with population
20. School nonproperty taxes associated with value 21. School nonproperty taxes associated with population 22. School intergovernmental revenues associated with value 23. School intergovernmental revenues associated with
popUlation
OUTPUT
DEVELOPMENT-INDUCED POPULATION (PREVIEW MODEL)
1. Development-generated residents 2. Development-generated school-age children 3. Development-generated employees
PUBLIC SERVICE EXPENDITURES (PREVIEW MODEL)
4. Development-induced municipal expenditure 5. Development-induced municipal nonresidential
expendihlre 6. Development-induced school expenditure 7. Development-induced total expenditure
PUBLIC SERVICE REVENUES (PREVIEW MODEL)
8. Development~induced assessed valuation 9. Development-induced municipal property tax revenue
10. Development-induced municipal nonproperty taxes 11. Development-induced municipal intergovernmental
revenues 12. Total municipal development-induced public revenues 9. a. Development-induced school property tax revenues
10. a. Development-induced school nonproperty taxes
11. a. Development-induced school intergovernmental revenues
12. a. Total school development-induced public revenues
DEVELOPMENT NET FISCAL IMPACT (PREVIEW MODEL)
13. Development·mduced municipal net fiscal impact 13. a. Development-induced school net fiscal impact 14. Development~induced total net fiscal impact 15. Municipal fiscal impact perspective (total revenues)
(Quickway)
15. a. School fiscal impact perspective (total revenues) (Quickway)
16. I\.funicipal fiscal impact perspective (property taxes) 16. a. School fiscal impact perspective (property taxes)
138
model or otherwise readily obtainable. For instance, the assessment-to-sales ratio and property tax rates as well as the number and value of parcels by residential/nonresidential type are either posted in official public documents such as the annual tax roll or are available from a telephone call to or interview with the tax assessor. Similarly, the other cost and revenue parameters, including municipal and school expenditures, are also readily collected from published budgetary documents and/or from contacting the local business administrator and school superintendent.
Despite its modest input data demands, the model produces a full array of cost-revenue outputs, including a detailed breakout of the development-induced population; costs by public service jurisdiction; costs induced by the residential versus nonresidential components of a project; revenues by jurisdiction; revenues differentiated into property, other local source, and intergovernrnental categories; and, finally, the net fiscal effects by individual jurisdiction and the aggregate public impact. Thus, the model does not give a "black box" final result but arrays and shows all the calculations and intermediate products that lead to the calculation of the final fiscal impact.
Nonetheless, the model does have some shortcomings. As a per capita methodology, the model is widely applicable though not always best suited to all situations. To illustrate, in cases of severe excess or slack capacity, the case study approach provides a more accurate picture of the impacts of growth. In addition, the case study method offers a level of descriptive detail on local conditions and service responses that goes far beyond the preview model's strictly numeric outputs.
Another shortcoming concerns the manner in which the model calculates revenues. Revenue inputs are factored on a per capita or valuation-increment basis. While such an approach is generally acceptable, it oversimplifies certain revenues--especially intergovernmental transfers. Intergovernmental aid is often distributed according to complicated formulas that relate local to state parameters (local valuation per pupil compared to a state foundation level) whose values themselves change annually. The preview model is not structured to reflect such a high level of specificity and, as such, yields an approximation rather than the most accurate projection of development-induced revenues.
This discussion points to the appropriate use of the preview model. The model does not substitute for a comprehensive, report-length fiscal impact study that embodies case study detail, considers the nuances of each revenue source, and so on. The preview model does, however, provide a reasonably reliable depiction of the order-of-magnitude fiscal impact of growth. It is useful for quickly answering whether development will approximately break even or likely generate a large or small surplus or deficit.
Such order-of-magnitude information is useful in several situations. When a project is first considered, the developer often explores alternative development possibilities in terms of type, scale, and location. A quick preview analysis of the fiscal effects of the various scenarios can help the developer select the most promising options. The analysis would typically be undertaken in conjunction with parallel studies of economic feasibility and environmental and traffic impacts. (These other exploratory studies could be assisted through use of the preview model in the various substantive areas.)
Once a developer decides on a development option and applies to the appropriate public authorities for development approval, the fiscal impact preview model can proVide useful information. On a smaller project, say a subdivision of 20 to 30 homes, a full fiscal impact study may not be required or even appropriate. Nonetheless, it is often instructive to present the findings of the preview analysis-provided the limitations of the model are made clear. Even when an independently commissioned, standalone cost-revenue study is perfanned, a developer or consultant may wish to "run" the model to get a quick, in-house sense of the final results. This preview is useful because the full study may not be completed for many months.
Critiquing a Fiscal Impact Analysis A fiscal impact study is reviewed by several indi
viduals-the client paying for the report, planning/ zoning board members, the consultants retained by these boards, and the public at large. The following provides a checklist for such review:
General Considerations Inadequate Documentation
The "numbers" in a fiscal impact report-the development-induced residents and school-age children, cost figures, the new increment of property taxes, and so on-are often presented without sufficient background discussion and documentation. It is imperative to detail the assumptions, databases, and calculations that guided the analysis so that the study can be understood, followed, and replicated.
Unbalanced Presentation The cost-revenue equation must be appropriately
balanced across several dimensions. Most fundamentally, both revenues and costs must be presented. In a few lingering cases, an analyst presents only one side of the equation, typically the revenue side, and indicates that a "shopping center will generate $1 million in property taxes upon buildout." The revenue figures signify nothing in the absence of corresponding cost figures.
139
Even when both costs and revenues are indicated, it is important to avoid further imbalances. For instance, if the analysis considers only the costs supported by the property tax (i.e., the per capita expense determined by dividing the property tax levy by the local population), then it must focus exclusively on ad valorem revenue collections. If other growth-induced revenues (i.e., intergovernmental aid) are added, then the analysis is unfairly tilted to show a benefit from growth.
Large Order-of-Magnitude Impact A fiscal impact report is suspect if it shows a result
that is extremely large in terms of its order-of-magnitude effect on the community-whether a significant surplus Dr deficit with a marked effect on the local jurisdiction (e.g., either decreasing or increasing property taxes by 40, 50, 60 percent or more). While it is theoretically possible that a project can yield a significant result, it is more likely that a fiscal impact study is flawed by one of the above considerations (e.g., costs are not correctly paired with revenues) or by one of the technical errors outlined below.
Technical Checklist Demographic Multipliers
Are the multipliers current? Do they reflect the latest available information such as that contained in AppendiX II? Are the multipliers correct? For instance, to calculate a development's impact on the local school district, the analyst should use the number of public school-age children and not the more encompaSSing school-age children figures; the latter includes students attending nonpublic schools. Are the multipliers appropriate? As discussed, general demographic sources should not be used for specialized housing types. Are the demographics locally sensitive? Has every effort been made to obtain local information and/or feedback?
Projecting Costs Is the appropriate method applied-per capita,
case study, or econometric? Are the cost factors current? Do they reflect the latest budgets and population estimates? Are costs properly stated? For instance, public jurisdictions usually pay for infrastructure over time; consequently, development-induced capital costs should be shown on an amortized rather than lump-sum basis.
Projecting Revenues Are the appropriate revenues counted? As stated ear
lier, costs and revenues must be properly paired in terms of the property tax and nonproperty tax components. Are the revenue factors clIrrent? Do they reflect the latest assessed ratios, property tax rates, permit and fee schedules, and intergovernmental allocation formulas? Are the revenues locally sensilive? Do they
incorporate the field-level operating practices of the assessor and other tax officials? Are the revenues properly stated? For instance, certain revenues are generated annually, but others accrue on a periodic basis, Le., the time at which a property is sold and the recording tax paid. This difference in timing should be incorporated into the analysis. To illustrate, if properties tum over an average of once in five years and the transfer tax on a sale is $2,000, then the expected transfer tax income per year is $400. Applying the $2,000 amount per unit annually would overcredit the development-induced revenues.
The reviewer should also be sensitive to possible mathematical errors in the calculation of both costs and revenues. For example, a mismatch commonly occurs in pairing the property assessment with the tax rate. Applying an equalized tax rate to the assessed tax base results in an undercount of income; conversely, multiplying the market value by the nominal tax rate overcounts revenue. The correct procedure is to use the same basis for both rate and assessment.
In short, a fiscal impact analysis is a technical undertaking and, as such, many factors must be monitored in revieWing such a study.
Presentation The fiscal impact analysis report should contain
the following components:
A. Summary of Findings Brief presentation on the number of people, school
age children, total costs/revenues, fiscal surplus or deficit, and impact on local taxes.
B. Overview of the Report Outline of what was requested, what (if anything)
was added to the report, and-as important-what was not covered in the analysis.
C. Methodology Employed Basic technique employed-per capita, case study,
or econometric; what the procedure covers; the advantages associated with the chosen method.
D. Data and Data Refinement Information sources: published documents, com
puter tapes, site visits/personal interviews, and/or other local surveys.
E. Development Costs What are they? What do they include (operating
and capital)? How severe a test of the development are they? How do the estimated costs compare to the results from other studies conducted in either the same jurisdiction or elsewhere?
140
..
F. Development Revenues What are they? What do they include (local/inter
governmental transfers)? How conservative is the analysis of revenues? How do the report's projections compare to studies of development revenue flows conducted in the same locale or elsewhere?
G. Net Fiscal Impact Annual and!or cumulative annual difference be
tween costs and revenues. To be conservative, if the study demonstrates a significant positive revenue flow, the analyst must understate the surplus by adding a cushion for additional costs. If the study demonstrates a small net impact relative to earlier expenditures or the fiscal profile data, the analyst should not specify a trivial surplus or deficit but instead call the net impact a "break even."
H. Impact in Context What is the impact of the development in the con
text of total local expenditures or in the context of all revenues raised through the property tax? What will it mean to the tax rate? How does it compare to other tax increases or decreases experienced historically?
I. Rigor of the Analysis The analyst should emphasize the extreme care
taken to interpret the development and its impact and discuss how reliance on the chosen method-per capita, case study, or econometric-produced the most rigorous pOSSible analysis.
The public presentation of a cost-revenue report before a planning board, board of adjustment, and so on combines the technical nature of the written report with the art of oral communications. In general, the public presentation follows the report format described above. It should avoid jargon, obscure methodological discussion, and unnecessary numeric detail. The presentation should proceed as follows:
Introduction and Summary. Describe "fiscal impact analysis" in simple terms and present the findings for the case in question. Place the projected cost-revenue result in proper perspective by stating what it means to the local taxpayer.
Methodology. Outline the method used by relying on a general illustratiye example. Explain why the technique was chosen.
Projecting Population. Summarize the projected number of people, pupilS, and workers generated by growth. Given that projections often become a point of contention, describe in some detail how the development-induced population was determined. Indicate that the numbers were reviewed with local officials and in other ways checked against "real world" experience. Discuss the development-induced population within the context of the eXisting local population base.
Projecting Costs. Describe the cost projections in simple terms and proVide an illustrative example of the calculation. Describe briefly what the imputed costs will pay for-added staffing, equipment, and!or capital improvements. Relate the development-induced expenditures to existing local service outlays.
Projecting Revenues. Indicate the revenues that were projected-local property, local nonproperty, and intergovernmental-and give an illustrative example of lhe projection, especially with respect to the property tax. Most people are familiar with this revenue source. Summarize the dollar amounts generated from the different revenue sources and provide the totat; relate the total to the existing total revenue base.
Net Fiscal Impact. Restate the population-induced service costs, development-generated revenues, and the resulting net fiscal effect and immediately discuss the net fiscal effect in terms of what it signifies to the local budget and taxpayer (e.g., added/reduced annual property taxes of $50 per household).
It is also important to discuss, where appropriate, the conservative nature of the fiscal impact approach. The analyst should stress that fiscal impact analysis overstates the development-induced population and costs but undercounts revenues. To illustrate; the analyst might state, "While official school enrollment studies use a multiplier of 1.0 for single-family detached homes, the cost-revenue report factored a value of 1.1 derived from the PUMS. This means that an additional $15,000 in school outlays was attributed to the project." Similarly, the analyst should underscore the confidence level of the fiscal impact finding by stating, "Even if the development-induced costs are twice as high, a fiscal impact surplus will still result."
Summary. It is opportune at the concluding point in the presentation to summarize the interim and final demographic and fiscal results of the cost-revenue analysis and to discuss their significance. It is useful to prepare a summary sheet!board that arrays the critical project-induced effects against the community's current population and financial parameters. Such a presentation places the project-induced effects in perspective.
The art of effectively presenting the results of a fiscal impact study cannot be overstressed. A fiscal impact report cannot be presented in the abstract by merely stating the net result without an explanation of the intermediate calculations and base assumptions. Such an approach compromises credibility. On the other hand, too much emphasis on numeric and methodological detail will lose the audience. An effective presentation achieves a balance of appropriate technical detail and narrative that can be clearly followed by the audience.
One way of fostering effective communications is to use boards and!or handouts that illustrate some of the assessment's critical assumptions, calculations,
141
and findings, These materials should include graphics (i.e., bar and pie charts) that add vitality to and enhance the understanding of the presentation (see chapter 2 for further details).
Another suggestion is to anticipate common questions and address underlying concerns in the body of the presentation. To illustrate, when a fiscal impact report shows that attached housing yields a surplus, at least one member of the audience is likely to ask, "How can that occur? Don't such units cost less than detached homes?" Rather than waiting for the query, the analyst should include an appropriate explanation when such a fiscal impact finding is first presented. This explanation would point out that attached housing is often a good ratable because it introduces relatively small households with rew schoolage children. As such, attached units generate relatively modest serVice costs. Another audience member is likely to ask, "How can an expensive detached unit yield a deficit?" (The explanation is the converse of the one just given for attached units.) Similarly, the basis of the demographic multipliers and the reliability of the multipliers over time (see AppendiX II) are often the subject of still more questions.
In sum, thorough preparation is needed to ensure lhe effective public presentation of a cost-revenue study. The material is technical, and the public forum is often adversarial and controversial. The delivery of an accurate yet convincing presentation requires careful treading through a minefield of questions and challenges.
Future Directions This chapter began with a discussion of how fiscal
impact analysis has evolved over the past half-century and will continue to evolve during the 1990s. The follOWing changes are anticipated with respect to applications, data, and teChniques.
Today, fiscal impact analysis is most commonly commissioned by a developer to study an individoal project. In the future, the focus of the analysis will expand to an areawide level that encompasses several developments and overall land uses. (Such change parallels the shift in EIS evaluations frum project-specific to regional.) Increasingly, the public sector-agencies that prepare master plans, capital facilities projections, impact fee schedules, staffing need projections, and the like-will undertake fiscal inlpact assessments.
The mid- and late 1990s will also see improvements in data. Traditional fields of information such as demographic multipliers will be refined, AppendiX II of this handbook, for example, uses PUMS and AHS data. New types of multipliers will also be developed. For instance, interest is growing in the infrastructure consequences of growth. Accordingly, just as demographic cuunts are now associated with different type and size housing units, "counts" of average
water use, sewage generation, road use, and park and library facility needs will, in the future, be attributed to different categories of incoming growth. These "multipliers" will provide a more precise measure of the financial impacts of growth, specifically its capital improvement needs.
The development of infrastructure profiles has also been sparked by the growing interest in impact fees, As practiced today, fiscal impact analysis and impact fee determination are distinct fields. Fiscal impact analysis compares all development-induced costsboth operahng and capital-to all development-induced revenues, Impact fee determination focuses only on capital costs and credits a share of revenues against the development-induced infrastructure expenditures. Further, fiscal impact analysis is typically used exclusively as an evaluation technique, but impact fees go beyond planning considerations to a charging mechanism.
Given the relationship between fiscal impact and impact fee analysis as it exists today, the question is whether such a differentiation is appropriate. If the raison d'etre of the impact fee is that"growth should pay its way," then the findings of the broader fiscal impact equation are germane to the impact fee calculation. Some jurisdictions are already beginning to integrate the two fields of analysis, and the future will witness greater overlap (see chapter 10 for a discussion on strategies for integration).
The coming decade will also see changes in the technique of fiscal impact analysis. The per capito approach will slill dominate but will incorporate the best features of the other methods. The per capita technique, echoing the case study approach, will increasingly build in sensitivity to deficient and surplus service capacity conditions and will adjust development-induced service costs accordingly. The per capita technique will also begin to incorporate the multiyear, interactive study of the econometric method-a proce" facilitated by the growing use of the desktop computer.
A further modification in the per capita technique involves increased application of comparable city or community refinement. Heref the fiscal experience of jurisdictions as they change in size and direction of growth (e.g., gaining or losing population) is incorporated into the analysis as, for instance, in the specification of per capita costs. The experience of comparable commwlities offers important inSight into future fiscal changes and the impact of growth.
Finally, the discipline of fiscal impact analysis will have to confront the issue of reliabillty. The field to date does not test its results; the projected impacts of growth undprgo no evaluation after development to measure the actual costs incurred and revenues generated. While such a study is inherently technically difficult, the field of fiscal impact analysis will remain vulnerable to charges that its methods have not been empirically verified-at least until retrospective analysis becomes part of the process.
142