Nguyen, Tanette N., "Comparing Tactical and Strategic Modeling Techniques in the Field of Public School Finance", 1981

SUMMARY

COMPARING TACTICAL AND STRATEGIC MODELING TECHNIQUES
IN THE FIELD OF PUBLIC SCHOOL FINANCE

‘Tanette N. Nguyen
Graduate School of Public Affairs
Department of Public Administration
State University of New York at Albany
~ Albany, New York 12222
PURPOSE OF THE STUDY
The purpose of the research is to compare two computer similation
modeling techniques regarding the impact of the implementation of a
cost~of-education index in the New York State aid formula for education.

Throughout the research, an attempt is made to assess the advantages and

disadvantages of the two simulation methods and to examine the utility of

combining both approaches in the analysis of school finance issues. Using the
two computer techniques, the study evaluates the impact of incorporating a
cost-of-education index in the state aid formula in terms of equalizing per
pupil expenditures throughout the state. Although the issues being discussed
in the study are generic to most states, the research is based on the

experience in New York State.

BACKGROUND AND PROBLEM
Essentially a local responsibility during the eighteenth and nineteenth

centuries, elementary and secondary education has become increasingly a matter

of concern for the state over the last century. Billions of dollars are
earmarked every year for education by state governments. In many states,
funds are allocated among individual school districts based on equalizing
formulas. Over the past decade, court rulings throughout the nation have
questioned the ability of current state aid formulas to alleviate the

disparity in schooling expenditures among various localities. In late 1971,
-2-
in Serrano y. Priest, the California Supreme Court ruled that the state's
method of funding education was in violation of the equal protection of the
lath Amendment to the United States Constitution. The Court set up. the

"fiscal neutrality" standard whereby the quality of a child's education must

not be a “function of the wealth of his parents and neighbors."

A number of states across the nation has followed the Serrano example in
challenging the constitutionality of their education finance system. As a
result of this reform movement, the legislatures in many states have attempted
to redesign the school finance system in order to make local tax burden less
dependent on local wealth, and guarantee all children a more equitable level
of education. Legislators and fiscal agencies have been under much pressure
to develop alternative methods of funding education and to support their
recommendations with detailed analyses.

As early as the 1950's the New York State Education Department developed
some district-by-district analyses. In the 1960's, thanks to improvements in
computer technology, several models were built for the analysis of school aid

. formulas. In 1962, Cornell University produced district-by-district analyses
for the New York State Joint Legislative Committee headed by Charles
Diefendorf. Such models were not, however, widely used.. Overall, traditional
processes in the area of public school finance have remained rudimentary in a
number of states. They usually involve time-consuming hand calculations, with
a large margin of error and little in-dépth analysis. The inadequacy of such
methods to address the compelling set of issues raised by the court cases,
together with the increasing complexity of state aid formulas and the growing
volune of data to be processed, has prompted a more widespread development of
computer simlations in the area of public school finance. These computer

models are essentially tactical by nature. They show the decision maker the

336

detailed short-run impact of proposed state aid packages on individual school
districts as well as at the state level and suggest a course of action.

Reform was not limited to the area of school finance. A series of
concurrent public referenda, judicial decisions, and federal mandates in the
general realm of public finance has had some drastic impact in the field of
education finance. The passage of Proposition 13 in California in June 1978
has limited the ability of localities to raise revenues, and placed a cap on
state and local expenditures. Subsequently, similar tax or spending
Limitation proposals have been initiated in several other states. At the same
time, court cases in many states have mandated full value assessment of the

a basis for the financing of the local share of

property which serves

educational costs (e.g., Hellerstein v. Assessor of Town of Islip in New York

State, 1975 ). Urban school districts have also been restricted in their
capacity to borrow funds in order to meet present and long-term expenditure

needs (e-g., Hurd vs City of Buffalo in New York State, 1974).

In an era of inflation, economic stagnation, and mounting pressure for
more government expenditures at the state level coupled with taxpayer revolt |
and widespread reform, the field of public school finance is becoming
increasingly interconnected and complex. Both traditional methods of analysis
and tactical simulation models are static by nature and involve short-run and
precise projections on a district-by-district basis. These models are not
adequately equipped, however, to examine in depth the intricacies and
Amplications of the current system. In addition, they are unable to foresee
the long-range ramifications of policy changes. Finally, they contain no
mechanism concerning the behavioral responses of the localities to court
mandates and to the recommendations proposed by the decision maker.

There existe currently another class of simulations which examine overall

models, referred to

policy-related issues at a more conceptual level. The:
with detailed financial information.

a=

strategic models, are not concerned with individual school districts and

They focus mainly on aggregate key

policy variables at the state level, and on long-term effects of these

variables on the system.

Strategic models are exploratory by nature. They

search for unforeseen and sometimes unintended consequences of policy actions.

Very Little strategic modeling, however, is presently being done by state

government agencies because the utility of this kind of simulation is not

immediately apparent to the political and bureaucratic decision makers.

Moreover, most states do not have the technical expertise to conduct the

simulations in-house.

PROPOSITIONS AND HYPOTHESES

Tactical and strategic modeling techniques are very different in terms of

their basic assumptions, approaches and outputs.

Table I presents some

general dimensions which will be used in the study to compare the two types of

models and the expected application of those dimensions to the two types of

models.
DIMENSION OP
COMPARISON KEY
Purpose
Boundary

Time Horizon

output

Level of
Aggregation

TABLE I

TACTICAL
MODEL,

“Evaluate specific state aid

proposals for current action.

Limited to the Education Law.
Use of relatively few
variables.

Short term (1 to 5 year
non-dynamic projections}.

Precise impact of state aid
proposal on each achool.

Disaggregation on a
district-by-district level.

STRATEGIC
MODEL

Explore overall policy issues
and their likely impacts.

Broadened to include general
public finance issues. Use
of more heterogeneous data.

Long term (1 to 20 year
dynamic projections).

More concerned with
long-range behavior of the
system. '

Highly aggregated sectore
(urban v. suburban v. rural
localities).

337

“5+

‘The purpose of the research is to convey the notion that in fact tactical
and strategic simulations are not competing against one another, but rather
can be seen as complementary techniques, reinforcing one another. The output
of the tactical simulation, for instance, could serve as the input for the
strategic model. The result of a concurrent use of tactical and strategic
models, it is hoped, will be better analytical capability in the decision
making process, without losing the detailed and precise information much
needed by the decision maker.

The implementation of a cost-of-education index in the New York State aid
formula for education will serve as an illustration of the proposition and
hypotheses stated above. Essentially the purpose of the cost-of-education
index is to adjust for educational cost differentials among school districts.
Theoretically, the index should help the state move toward a more equitable
allocation of education by compensating localities which have to pay a higher
price for the same standard education resource relative to the state average

price of that resource. A tactical model will show how, in the short run,

such adjustment affects individual school districts. Gainers as well
losers can be easily identified in the simlation. A strategic model, on the
other hand, will provide some insight on the long range impact of implementing
the cost-of-education policy. It will help assess patterns of local responses

to this equalization proposal.
STAFFGROUP STRATEGIC SURVEYS TNO CONTENTS

MOBILITY OF RESEARCHERS IN THE NETHERLANDS page
(summary)
1. Background J
2. This research project 2
3. Conclusions 7
f 3.1. General conclusions 7

3.2. Conclusions based on the labour

2
+ pos
Rap market model 10
wae

3.3. Conclusions based on the R & D model 4

3.4. Conclusions concerning the need of data

+) and further research 7

PN 4, Recommendations for the measures to be taken 19

(,

D.R. Davies

REHM. Smits

R. Tweehuysen

W. Wiis December 1980
COMPARING TACTICAL AND STRATEGIC MODELING TECHNIQUES
IN THE FIELD OF PUBLIC SCHOOL FINANCE
Tanette N. Nguyen
Department of Public Administration
Graduate School of Public Affairs

L State University of New York At Albany
3 Albany, New York 12222

ABSTRACT

This study proposes to compare two types of computer simulation
techniques, namely tactical and strategic simulations. It explores the
advantages and disadvantages of the two methods and stresses the importance of
the insight to be gained by combining both approaches in the evaluation of
public policies. A school finance reform policy is presented as a case study.
More specifically, the research evaluates the implementation of a
cost-of-education index (a mechanism to adjust for disparities in educational
costs among school districts in a state) in the New York State aid formula.
The study investigates, using the two computer simulation techniques, the

impact of this policy in terms of equalizing per pupil expenditures.

INTRODUCTION

Purpose of the Study

The purpose of the research is to compare two computer simulation
modeling techniques regarding the impact of the implementation of a
cost-of-education index in the New York State aid formula for education.
Throughout the research, an attempt is made to assess the advantages and
disadvantages of the two simulation methods and to examine the utility of
combining both approaches in the analysis of school finance issues. Ur'ng the

two computer techniques, the study evaluates the impact of incorporating a
“ow
cost-of-education index in the state aid formula in terms of equalizing per
pupil expenditures throughout the state. Although the issues being discussed
in the study are generic to most states, the research is based on the
experience in New York State.
Essentially the study:
1) Surveys the use of tactical and strategic simulation modeling
techniques in the field of public school finance
2) Illustrates the application of a tactical model by implementing, as
an example, a cost-of~education in the New York State aid formula
3) Illustrates, again using a cost-of-education index, a strategic model
4) Compares the two computer techniques in terms of their usefulness in
the decision making process, and within the context of policy

analysis.

Definitions

A model is an analytical representation of selected features and
relationships of a real-world entity. The entity being represented is also
referred to as the “reference system" (Greenberger, crenson, and Crissey,
1976:49 ). The development of computer technology has led to the widespread
use of formal models as explicit devices to help understand or improve the
reference system.

A Sharaelon is the method of developing a model of a real situation and
then performing experiments upon the model to test the accuracy of its

behavior under varying conditions (Fogarty, 1976:267).
eRe
Public Policy Models

Modeling has become a rather commonplace activity in the public sector.
A multitude of different types of formal models has been built in a wide
variety of policy areas to help the decision maker better understand the
intricacies of socio-economic systems (Greenberger, Crenson, and Crissey,
1976:xiv). In addition, computer modeling is a potentially powerful tool to
communicate ideas and focus debate around specific policy issues. The
proliferation of models in public agencies is not, however, a reliable gauge
of the actual impact of modeling in the decision making process. A large
fraction of models that have been developed has never been put to use. In
many instances, a number of characteristics inherent to the models has
hindered their usefulness to the decision maker. Complicated, large-scale
models are difficult to understand. They often produce only generalized
results that hold only limited interest to public officials confronted with
the intricacies of specific problems. ‘Tenuous assumptions, crudely
represented relationships, and inadequately calculated variables further
undermine the validity of the models' output (Greenberger, Crenson, and
Crissey, 1976:23-27). As a general rule, the political setting and the
organizational framework within which a given model is developed and applied
are crucial determinants of its usability (Greenberger, Crenson, and Crissey,

1976:20).

SIMULATION MODELS IN THE FIELD OF PUBLIC SCHOOL FINANCE

The study focuses on the field of public school finance to illustrate
the application of models in the public sector and the difficulties they

encounter.
-4-

Tactical and Strategic Models: Definitions

Keen and Clark (1978) distinguish between two broad types of computer
simulation models, in the field of public school finance. They are tactical
and strategic models. According to the Webster New collegiate Dictionary, the
word 'tactical' is a synonym for 'short range’ and tactical decisions are
decisions made or carried out with only a limited or immediate end in view.
They involve actions of less magnitude than those of a strategy. Within the
context of this study, tactical simulation models are computer models which
involve the creation and short-run evaluation of specific public policy
proposals, at a detailed level of disaggregation. As a general rule, they
have a limited boundary. More specifically, they involve narrowly focussed
analyses, and deal with a selected range of variables. The output from a
tactical simulation is generally presented in the form of a series of tables
showing the detailed and precise impact of a given proposal on each school
district and on the entire state under study. Some more sophisticated models
can also provide simple statistics such as mean, median, and ranges, and
perform advanced multi-regression analyses. Tactical simulations are most
often used as planning devices to assist the decision making process. 1

Strategic models on the other hand, entail analyses that are long-run,
historical, evaluative, and conceptual. More specifically strategic
simulations focus on broad policy alternatives and on long-term relationships
among variables in the reference system. They examine more qualitative policy
issues such as behavioral response of the system to changes in parameter
values or in the model's structure. Their concern is not on short-run and
detailed information on individual school districts but on aggregate key

policy variables. The purpose of strategic research to a great extent is to
<5 =
explore policy alternatives, analyze and explain their outcomes, and generate
insights about important variables and relationships underlying the reference
system. Strategic models are not directly linked to the policy making
process. They are mainly used in the field of academic research.
Consequently they are more independent of the decision maker's immediate

concern with detailed policy outcomes.

Background: School Finance Reform in the United States

Essentially a local responsibility during the eighteenth and nineteenth
centuries, elementary and secondary education has become increasingly a matter
of concern for the state over the last century. Billions of dollars are
earmarked every year for education by state governments. In many states,
funds are allocated among individual school districts based on equalizing
formulas. Over the past decade, court rulings throughout the nation have
questioned the ability of current state aid formulas to alleviate the
disparity in schooling expenditures among various localities. In late 1971,
in Serrano v- Priest, the California Supreme Court ruled that the state's
method of funding education was in violation of the equal protection of the
14th Amendment to the United States Constitution. The Court set up the
"fiscal neutrality" standard whereby the quality of a child's education must
not be a "function of the wealth of his parents and neighbors." In San

Antonio Independent School District v. Rodriguez, (1973), however, the United

States Supreme Court held that the Texas system of financing public education,
despite its inequities, does not violate the equal protection clause since
education is not guaranteed by the federal constitution and therefore cannot

be considered as a fundamental right. The decision of the United States
Supreme Court blocked the federal constitution as a legal route
to school finance reform. The arena for school finance litigation was then
shifted back to the state courts.

A number of states across the nation has followed the Serrano example in
challenging the constitutionality of their education finance system. As a
result, the pace of school finance reform has accelerated rapidly over the
past few years (Odden and Augenblick, 1980; and Lawyers' Committee for Civil
Rights Under Law, 1980). The legislatures in many states have attempted to
redesign the school finance system in order to make local tax burden less
dependent on local wealth, and guarantee all children a more equitable level
of education. Legislators and fiscal agencies have been under much pressure
to develop alternative methods of funding education and to support their
recommendations with detailed analyses.

Reform was not limited to the area of school finance. A series of
concurrent public referenda, judicial decisions, and federal mandates in the
general realm of public finance has had some drastic impact in the field of
education finance. The passage of Proposition 13 in California in June 1978
has limited the ability of localities to raise revenues, and placed a cap on
state and local expenditures. Subsequently, similar tax or spending
limitation proposals have been initiated in several other states. At the same
time, court cases in many states have mandated full value assessment of the
property which serves as a basis for the financing of the local share of

educational costs (e.g., Hellerstein v. Assessor of Town of Islip in New York

State, 1975 ). Urban school districts have also been restricted in their
capacity to borrow funds in order to meet present and long-term expenditure

needs (e-ge, Hurd v. City of Buffalo in New York State, 1974).
a7
In an era of inflation, economic stagnation, and mounting pressure for
more government expenditures at the state level coupled with taxpayer revolt
and widespread reform, the field of public school finance is becoming
increasingly interconnected and complex and no longer can be treated in

isolation.

Tactical Simulations in Public School Finance

As early as the 1950's the New York State Education Department developed

somd district-by-district analyses. In the 1960's, thanks to improvements in
computer technology, several models were built for the analysis of school aid
formulas. Later in 1962, Cornell University produced district-by-district
analyses for the New York State Joint Legislative Committee headed by Charles
Diefendorf. Such models were not, however, widely used. Overall, traditional
processes in the area of public school finance have remained rudimentary in a
number of states. They usually involve time-consuming hand calculations, with
a large margin of error and little in-depth analysis. The inadequacy of such
methods to address the compelling set of issues raised by the court cases,
together with the increasing complexity of state aid formulas and the growing
volume of data to be processed, has prompted a more widespread development of
computer simulations in the area of public school finance. These computer
models are essentially tactical by nature. They show the decision maker the
detailed short-run impact of proposed state aid packages on individual school
districts as well as at the state level and suggest a course of action.

A number of tactical simulation models has been developed over the past
decade. Early efforts were made to build generalized calculation models that

could be adapted to any state. Sklar and Ioup (1971), under sponsorship of
-3 -
the President's Commission on School Finance, developed a Prototype National
Educational Finance Planning Model (NEFP) to simulate the nation's future
educational needs and resources. During 1972-74 the model was refined and
made operational in several states. The NEFP model is a powerful tool. It
has almost unlimited potential because of the infinite variety of data which
it can accommodate and the new decision options it can make available to the
decision maker (Boardman, et al., 1973; and Boardman, 1974). The model was
adopted by several states and is still in use in New Mexico (Keen and Clark,
1979). Another generalized School Finance Equalization Management System
(SFEMS) model was developed by staff at the Educational Testing Service and
set up in several states around the nation (Keen and Clark, 1979).
Generalized models however, are cumbersome and hard to operate. In addition,
they cannot be used for a specific state without extensive modifications. As
a result, most states have chosen to build their own tactical capabilities
in-house. It is often easier to develop a simulation de novo rather than
force-fit a specific state's formula into a generalized structure (Keen and
Clark, 1979).

Aside from a detailed survey conducted by Keen and Clark (1979), the main
source of information on tactical modeling techniques in the field of school
finance is provided by the user's manuals for the models used by specific
states (LEAP, Washington, 1978; LEGICOM, Michigan, 1977; PASSS, Pennsylvania,
1978; SIMULBUD, New York, 1978; SSF, Oregon, 1975). Other studies available
consist of developing or evaluating computer-based school finance simulations
for specific states (Bishop, 1975, for Texas; Bookman, 1977, for West
Virginia; Huxel, 1973, for New Mexico; Keen, 1978, for California; Mayfield,

1973, for Georgia; Odden and Vincent, 1976, for Missouri; Oregon State
-9 -
Legislature: Committee on Equal Educational Opportunity, 1974A, 1974B; Pierce
et al., 1975, for Oregon; South Dakota State Division of Elementary and

Secondary Education, 1977; and Wegryn, 1977, for Michigan).

Strategic Simulations in Public School Finance

There exists currently another class of simulations which examine overall
policy-related issues at a more conceptual level. These models, referred
to as strategic models, are not concerned with individual school districts and
with detailed financial information. They focus mainly on aggregate key
policy variables at the state level, and on long-term effects of these
variables on the system. Strategic models are exploratory by nature.

They search for unforeseen and sometimes unintended consequences of policy
actions.

A survey of the literature shows that strategic research performed in the
field of public school finance encompasses mostly econometric studies with the
exception of a few system dynamics simulations. Econometric models have been
essentially cross-sectional multi regression analyses of the impact of various
state aid formulas as devices to neutralize the effects of local wealth
difference among school districts. In response to various court rulings on
the unconstitutionality of present methods of funding public education through
local property tax, Stern (1973) built a prototype econometric model of
current expenditures by local school districts in Massachusetts. Stern
simulated alternative formulas for distributing general purpose state aid and
came to the conclusion that a District Power Equalizing (DPE) formula (which
assures that districts producing the same tax rate on local property will

receive equal revenues through a combination of local and state funds) adjusts
~10=

expenditure disparities that are due to property value. The formula does not,
however, reduce differences associated with socio-economic status (measured in
terms of income).

In a regression analysis of 105 towns in Massachusetts, Feldstein (1975),
also shows how the DPE form of aid fails as a device to fully neutralize the
effect of local wealth differences among school districts. In addition,
Feldstein distinguishes between matching grants and block grants and concludes
that a matching grant system, where the state matches the local effort, has
superior incentive features. Ladd's (1975) model of seventy-eight communities
in the Boston SMSA looks at the implication of the composition of the local
property tax base on educational expenditures. She suggests that a school
district's educational expenditures are closely related to the size of its
total property tax base and that the composition of the base into commercial,
residential, and industrial property affects local decisions to provide
educational services. Hence the separate components of the tax base deserve
greater attention in the determination of local fiscal capacity for education.

Drawing from the results of a study of Vermont, Gatti and Tashman (1976,
1978) suggest a proposed solution to redress the flaws of the DPE formulation.
Essentially they advocate the inclusion in the state aid formula of an income
component as well as a measure of the district's ability to export school
taxes since both are highly significant determinants of school districts'
outlays on public education. In an analysis of the Illinois school system,
Friedman and Wiseman (1978) have looked at the impact of legislative reform on
wealth-related disparity in expenditures among pupils. They stress the
importance of distinguishing between immediate effects of the reform on the

distribution of expenditures per pupil, intermediate impacts, which occur
<1
after voters have responded to the new formula, and long-run effects brought
about by shifts in tax rates. Grubb and Michelson (1974) recognize that both
state taxes and other local taxes may be important determinants of school
district expenditures. Their study consists of an evaluation of three
alternative state aid formulations on a sample of 159 school districts in
Massachusetts. In addition, Inman and Wolf (1976) and Inman (1978) have built
a general equilibrium model of a typical U.S. metropolitan’ economy. The model
is unique in that it includes a mechanism depicting the communities‘
behavioral reaction to school fiscal reform. The empirical specification of
this model was applied to New York City and fifty-eight Long Island school
districts. Finally, Greene (1979) has presented a detailed review and
evaluation of past econometric models in the field of school finance.

Aside from Inman's powerful model which explicitly incorporates local
behavioral reaction to reform, insufficient attention has been devoted to the
dynamics of school finance. As a result very little is known about the
ultimate effects of equalization proposals. Knickman and Reschovsky (1980)
have called for the explicit inclusion of localities' behavioral assumptions
in analyses of school finance policies. Treacy and Frueh (1974) advocate the
use of time series data and demographic projections incorporating plausible
estimates of migration behavior. These dynamic factors should be incorporated
when assessing the effects of policy changes in school financing. - Unexpected
and undesired changes in school finance systems, they argue, will occur so
long as reforms are made under erroneous assumptions concerning the structural
relationships existing within the state.

System dynamics models are specifically geared to address those critical

dynamic issues which have been overlooked by past studies. System. dynamics
=12-

was developed in the 1950s by Jay W. Forrester at the MIT Alfred P. Sloan
School of Management. In a nutshell, system dynamics is a method for
understanding and solving complex problems using the concept of dynamic
‘feedback' structure. A feedback system consists of a closed loop structure
that brings results from past actions of the system back to control its future
actions. In the field of public school finance, for instance, the amount of
money spent by a given school district for educational purposes over the long
run is not determined by the present true condition of the school district.
Instead, it is conditioned by the past circumstances that have been observed,
analyzed, and digested by the community.

Very little system dynamics modeling has been performed in the area of
education finance. In a study of the funding for special education in
Massachusetts, Andersen (1977, 1979, and 1980) has. demonstrated that by
ignoring behavioral responses of local school districts, traditional tactical
models have failed to analyze patterns of expenditure growth. Asa
consequence, they have produced erroneous cost estimates of reform proposals.
Chen (1980) has discovered that policies designed at equalizing school
expenditures may work in the short run, but are likely to be reversed in the
long term, because of local communities' reaction to incentives built into the
reforms themselves (see also, Chen, Andersen and Nguyen, 1980). In further
exploratory work, similar patterns of reversal in policy conclusions with
respect to strategic simulations of selected issues involved in financing
special education in New York State were found (Nguyen, Andersen, and Chen,
1980). In addition, an interactive model has been developed at MIT to analyze
Massachusetts' funding system and to explore complex strategic issues in the

field of school finance (Stabell, Growchow, and Haan, 1972).

=13-

Propositions and Hypotheses

Tactical and strategic modeling techniques are different from each other.
They serve different purposes and strive in different environments. Tactical
models are usually built in-house to meet the needs of the decision maker.
Their usefulness stems directly from the fact that they can respond to the
demands and the constraints of the political realities which call for
reliable, easy-to-interpret redalte on a district-by-district basis. The
numerical value of the parameters entered in tactical simulations is highly
accurate. No controversial assumptions are made concerning the structure of
the model since simulations, which involve essentially tinkering with and fine
tuning the current state aid formula, are based on the Education law.
Furthermore, the output from simulation runs is highly disaggregated so the
decision maker can assess at a glance how individual school districts fare
under various state aid packages. The time horizon of tactical simulations is
limited (one year to five years). Projections, when made, are usually
straightforward linear extrapolations from the first year's output. Overall,
tactical models are easy to understand and reliable.

Because of their simplicity, however, tactical. models are also limited in
scope. Decisions on school finance issues do not occur in a vacuum. They are
part of the overall public finance system. By dealing exclusively with the
state aid formula, tactical models ignore the fact that the choice of any
education package will affect the amount of remaining resources that are
available for other categories of services. In addition, they contain no
mechanism concerning the behavioral responses of the ‘localities to court
mandates and to the recommendations proposed by the decision maker (Odden et

al., 1977). In other words, they are inadequately equipped to assess the long
-14-

range impact of local response to formula changes on the overall goal of
equalization. This class of issues is better addressed by strategic
simulation models.

Strategic models encompass a broader boundary and a longer time horizon
than tactical models. They provide a more comprehensive and realistic picture
of the overall field of public finance. They allow, for instance, the public
official to gain some useful insight as to how some policy changes in one
area, say education, might affect other services (such as social welfare, or
transportation). Strategic models use heterogeneous data, ranging from
individual school districts' property values, and pupil counts, to state
income and sales taxes and levels of expenditures for non-educational
services. The range of the variables in the simulation output is also
diversified.

Up to the present time, however, strategic simulations have remained with
the realm of academic research. Very little strategic modeling is currently
being done by state government agencies for school aid purposes. The utility
of this kind of simulation is not immediately apparent to the political and
bureaucratic decision makers. Indeed, strategic simulations do not seek for
the detailed numerical accuracy of their tactical counterparts. First of all,
output is not disaggregated on a district-by-district basis. Instead,
strategic models group school districts which share some characteristics in
common into sectors (e.g. metropolitan, urban, rural sectors). Analyses are
made at this sector level. In addition, strategic simulations are not
concerned with what will happen next year, but rather with the long range
behavioral response of various sectors under different scenarios and policy

changes. Given the nature of the political process, long run projections are
=15-

not of direct value to the elected official who is more interested in the
immediate ramifications of his decisions. Lack of interest in long term
prediction stems from the fact that the legislative operates in a
muddling-through mode. Policy changes, it is argued, are always possible if
and when the situation starts to deteriorate. This mode of operation makes
long range projections obsolete. Another major drawback of academic models is
the teneousness of many of their underlying assumptions. Many assumptions
built within the structure of a strategic model (e.g. the interaction between
various variables) are based upon the modeler's own perception of the reality.
Such assumptions, however carefully devised, constitute ground for
controversial debates and contribute to lower the model's overall validity in
the eyes of public officials. Finally, strategic models are difficult to
conceptualize. To some extent, the decision maker looks upon the complex
structure of a strategic model as a black box which he does not understand nor
have any control over, and which he consequently distrusts.

The purpose of the research is to convey the notion that in fact tactical
and strategic simulations are not competing techniques exclusive of one
another, but rather they can be seen as complementary techniques, reinforcing
one another. The output of the tactical simulation, for instance, could serve
as the input for the strategic model. Conversely, strategic research
generates insights that might be directly applied into a tactical simulation.
The result of a concurrent use of tactical and strategic models, might lead to
better analytical capability in the decision making process, without losing

the detailed and precise information much needed by the decision maker.
-16-
Future Developments

Tactical and strategic simulation modeling techniques will be compared
regarding the impact of the implementation of a cost-of-education index in the
New York State aid formula for education. Essentially, the purpose of the
cost-of-education index is to adjust for educational cost differences among
school districts. Theoretically, the index should help the state move toward
a more equitable allocation of education by compensating localities which face
higher costs for the same amount of education relative to the state average

price of that resource. o
The impact of the cost-of-education index in New York State, will be
analyzed using two different computer models. The first is SIMULBUD
(Simulation for Budgeting). SIMULBUD is a tactical modeling technique built
as a device to simulate alternative school finance formulas and to study the
detailed distributive impact of various policy proposals on a
district-by-district basis. Modifications and cofrections can be effected on
an interactive mode. Results appear instantaneously on the terminal under the
form of summaries, lists, totals, tables, and correlation coefficients.
Appendix A provides an example of an output run from a SIMULBUD simulation.
The output from a tactical model shows how, in che short run, such adjustment
affects individual school districts. Gainers (ies, those school districts
which register an increase in state aid as a result of the implementation of a
cost-of-education index) as well as losers can Be easily identified in the
simulation. A strategic model, on the other hand}. will provide some insight

oy
on the long range impact of implementing the cost*of~education policy. It

will help assess patterns of local responses to this equalizing proposal. A

system dynamics model of the New York State school finance system has been

“17

built, to replicate the general patterns of interrelationships and structural
properties in the state. Output of the model is presented in the form of
graphs plotting the behavior of parameters against time (see Appendix B or a
sample run). Unlike SIMULBUD, which strives for the numerical accuracy of
simulation results, system dynamics concentrates on trying to formulate

general patterns of behavior of the system under alternative policies.

BIBLIOGRAPHY

Books and Articles

Andersen, David F. 'Mathematical Models and Decision Making in Bureaucracies:
A Case Study Told from Three Points of View.' Ph. D. Dissertation, Sloan
School of Management, MIT,1977.

Andersen, David F. 'A System Dynamic Simulation of Special Education
Reimbursement Policies.' Albany, New York: Graduate School of Public
Affairs, State University of New York at Albany, 1979.

Andersen, David F. ‘Using Feedback to Test Educational Finance Policies.’
Policy Sciences. Vol. 12, Nos 3, Oct. 1980. pp. 315-331.

Bishop, Terry Neal. Development and Evaluation of a School Simulation
Planning Model.' Ph. D. Dissertation, The University of Texas at Austin,
1975, Dissertation Abstracts International. Vol. 36-A, No.2, Aug. 1975.
p. 621-A.

Boardman, Gerald R., et al. NEFP Decision Process: 'A Computer Simulation
for Planning State School Finance Programs.' User's Manual. Gainesville,
Florida: National Educational Finance Project, 1973.

Boardman, Gerald R. 'A School Finance Computer Simulation Model.' AEDS
Journal, Vol. 8, No. 2, Winter 1974. pp. 39-50.

Bookman, Michael K. ‘Analysis of Equality of Educational Opportunity and
Taxpayer Equity Through the Modeling and Testing of a Computer-Based
Public School Finance Simulation for a Selected State.’ Ph. D.
Dissertation, The University of Florida, 1977. Dissertation Abstracts
International. Vol. 38-A, No. 11, May 1978. p. 6421-A.

Chen, Fiona F. 'A Broad Quantitative View of Public School Financing.’
Albany, New York: Graduate School of Public Affairs, State University
of New York at Albany, 1980. (Mimeographed).
-18-

Chen, Fiona F.; Andersen, David F.; and Nguyen, Tanette N. 'A Preliminary
System Dynamics Model of the Allocation of State Aid to Education.'
Proceedings of the Eleventh Annual Pittsburgh Conference. Pittsburgh,
Pennsylvania, 1980.

Feldstein, Martin S. 'Wealth Neutrality and Local Choice in Public Education.'
The American Economic Review. Vol. 65, No. 1, March 1975. pp. 75-89.

Friedman, Lee S., and Wiseman, Michael. ‘Understanding the Equity
Consequences of School Finance Reform.' Harvard Educational Review.
Vol. 48, No.2, May 1978. pp. 193-226.

Gatti, James F., and Tashman, Leonard J. 'Equalizing Matching Grants and the
Allocative and Distributive Objectives of Public School Financing.’
National Tax Journal. Vol. 29, No. 4, Dec. 1976. pp» 461-476.

Gatti, James F.; Tashman, Leonard J.;and Sweet, Jonathan K. 'The Wealth
Neutrality of District Power Equalizing Grants in Public School
Financing: Additional Evidence.' Journal of Education Finance. Vol. 4,
No. 2, Fall 1978. pp. 213-224.

Greenberger, M.-; Crenson, M.A.; and Crissey, B.L. Models in the Policy
Process. New York: Russell Sage Foundation, 1976.

Greene, Kenneth V. 'The Equalizing Effects of District Power Equalization: A
Review of the Economics Literature.' Journal of Education Finance,
Vol. 5, No. 2, Fall 1979. pp. 187-214.

Grubb, W. Norton, and Michelson, Stephan. States and Schools. Lexington,
Mass.: D.C. Heath, 1974.

Huxel, Lawrence Le 'A Computed Based Simulation Model for Public School
Finance in New Mexico.' Ph.D. Dissertation, University of New Mexico,
1973. Dissertation Abstracts International. Vol. 34A, No.6, Dec. 1973.
ps 2969-A.

Inman, Robert P., and Wolf, Doug. ‘SOFA: A Simulation Program’ for Predicting
and Evaluating the Policy Effects of Grants-in-Aid.' Socio-Economic
Planning Science. Vol. 10, June 1976. pp. 77-88.

Inman, Robert P. ‘Optimal Fiscal Reform of Metropolitan Schools: Some
Simulation Results.' American Economic Review. Vol. 68: Noe 1, March
1978. pp. 107-122.

Keen, Peter G.W. The California School Finance Model: A Case Study of
Effective Implementation. Research Paper No. 467. Stanford, California:
Stanford University, Graduate School of Business, Jan. 1978.

Keen, Peter G.W., and Clark, David G. ‘Computer Systems and Models for School
Finance Policy Making: A Conceptual Framework.' Auge 1978.
(Mimeographed).
=19-

Keen, Peter G.W., and Clark David G. Simulation for School Finance: A Survey
and Assessment. Denver Colorado: Educational Commission of the States,
1979.

Knickman, James R.; and Reschovsky, Andrew. 'The Implementation of School
Finance Reform.' Policy Sciences+ Vol- 12, No. 3, Oct. 1980. pp. 299-
314. :

Ladd, Helen F. ‘Local Education Expenditures, Fiscal Capacity and the
' Composition of the Property Tax Base.-' National Tax Journal. Vol. 28,
No. 2, June 1975. pp. 145-158.

Lawyers’ Committee For Civil Rights Under Law. School Finance Project.
Update on State-Wide School Finance Cases. Washington, D.C.,

Apr. 1980.

LEAP, Legislative Evaluation and Accountability Program Application Briefs.
Olympia, Washington: LEAP Program Committee, 1978.

LEGICOM Operating Manual. Lansing, Michigan: Senate Fiscal Agency, Aug. 1977.

Mayfield, Walter D. 'A Computer Simulation of the Distribution of Tax Dollars
through Selected Funding Models to School Districts in Georgia." Ph.D.
Dissertation, Georgia State University, 1973.

Miner, Jerry, and Sacks, Seymour. Study of Adjustment of New York State
School Aid Formula to Take Account of | Municipal Overburdens.
Metropolitan Studies | Program Maxwell School of Citizenship and Public
Affairs. Syracuse University, May 1980.

New York State Executive Department, Division of the Budget. SIMULBUD-L
User's Instruction Manual. Albany, New York. Revised Oct. 1978.

Nguyen, Tanette N.; Andersen, David F.; and Chen, Fiona F. 'The Dynamics of
State Aid to Education: Interactions Between Special Education , Regular
Education, and Non-Schooling Expenditures.’ Proceedings of the 1980 _
Conference on Cybernetics, Man, and Society. Cambridge, Mass., ‘1980.

Odden, Allan, et al. School Finance Computer Simulations. Denver, Colorado:
Education Finance Center, Education Commission of the States, Nov. 1977.

Odden, Allan; and Augenblick, John. School Finance Reform in the States:
1980. Denver, Colorado: Education Finance Center, Education Commission
of the States, Apr. 1980.

Odden, Allan, and Vincent, Phillip E. An Analysis of the School Finance and
Tax Structure of Missouri. Denver, Colorado: Education Finance Center,
Education Commission of the States, 1976.

Oregon State Legislature: Committee on Equal Educational Opportunity.
Alternative School Finance Plans for Oregon: A Staff Report. Salem,
Oregon: School Finance Project, Oct. 1974.

=20-

Oregon State Legislature, Committee on Equal Educational Opportunity. A Local
Guaranteed Yield Plan for Oregon: A Second Staff Report. Salem, Oregon:
School Finance Project, Nov. 1974.

Pennsylvania Department of Education, Bureau of Information Systems. PASSS:
Pennsylvania School Subsidy Simulation. Summary Report by Cho, S.H., and
Ju, SeS. Pennsylvania, June 1978.

Pierce, Lawrence C., et al. State School Finance Alternatives: Strategies
for Reform. Eugene, Oregon: Center for Educational Policy and
Management, University of Oregon, May 1975.

Sklar, Sigmund L., and Ioup, William E. A Prototype National Educational
Finance Planning Model. Projections of Educational Needs, Resources, and
Disparities under Various Forecasting and Policy Assumptions. The
President's Commission on School Finance. Washington, D.C.: U.S.
Government Printing Office, Dec. 1971.

South Dakota State Division of Elementary and Secondary Education. Financing
the Public Schools of South Dakotas Gainesville, Florida: National
Educational Finance Project, 1973.

South Dakota State Division of Elementatry and Secondary Education. 'Proposed
State Aid Formula for South Dakota.’ Denver, Colorado: School Finance
Equilization Study Workshop, March 3-4,1977.

Stabell, Charles B; Growchow, Jerrold M.; and Haan, Anders. 'The Equalization
of School Expenditures in Massachusetts.' Working Paper No. 606-72.
Cambridge, Massachusetts: MIT Sloan School of Management, June 1972.

Stern, David. 'Effects of Alternative State Aid Formulas on the Distribution
of PUblic School Expenditures in Massachusetts.' Review of Economics and
Statistics. Vol. 55, No. 1, Feb. 1973. pp. 91-97.

Treacy, John J., and Frueh, Lloyd W.,IILl. 'Power Equalization and The Reform
of Public School Finance.' National Tax Journal. Vol. 27, No. 2, June
1974. pp. 285-299.

Wegryn, James. An Assessment of the Legislative Computerized Budget Project
LEGICOM. Lansing, Michigan: Senate Fiscal Agency, Jan. 1977.

Court Cases

Serrano v. Priest, 5 Cal. 3d 584, 96 Cal. Rptr. 601, Pe2d 1241 (1971).

San Antonio Independent School District v. Rodriguez, 411 U.S.1 (1973).

Hellerstein v- Assessor of Town of Islip, 37 N.Y¥.S. 2d 1 (1975).

Hurd ve City of Buffalo, 34 N.Y.S. 2d 628 (1974).
APPENDIX A

SAMPLE OUTPUT OF A TACTICAL SIMULATION”

*source: Description of Educational Improvement Index. New York State
Division of the Budget, January 1980.

DISTRICT EDUCATION IMPROVEMENT INDICES PROPOSED FOR USE IN OPERATING AID
FORMULA 1980-81 SCHOOL YEAR, BY COUNTY

Educational Educational
Dist. Improvement Dist. Improvement
Code District Name Index Code District Name Index
10100 ALRKANY 1.01894 i 40101 ALLEGANY 1.01145
———""TORO1” BERNE-KNOX-WI 9S 7851 402047 WEST “VALLEY 79670667
10306 KETHILEHEM 1994479 40301 LIMESTONE 1.07204
TOAOZ-RAVENACORYMAN 7970313" |  ——4OPOT-ELE TOOT TUTE 107542!
10500 COHOES 1.01743 41101 FRANKLINUILLE 1.01154
———1 0601 SOUTH COLONTE 897149" | —4T401-HINSTIALE F95AZ25 |
10605 NORTH COLONIE 1.91743 : 41801 LITTLE VALLEY .980231 :
———" JOST MENANTIST CSO, TTOTsOB™ A2QIOL CATTARAUGUS ~-- 1.00128
10622 MAFLEWOON CSN 1.02402 42400 OLEAN 1.03194 ;
————"107 01- GREEN TSLAND 17 09840 —FESCT BOWANTIA TT009T9 i
10802 GUILTIERLANI! 1 PO5 654 : 42901 PORTVILLE sO71L577 !
TI 0OS-VOORHEESVILEE 1007037 S300 RANTIOLFH 988215 7 7
11200 WATERVLIET +9B1301 43200 SALAMANCA

ZOLOT ALFREN-ALMOND” 968548 —

11,006

“TAZROL YORMSHIRE-FION 1, O03Te

20501 BELMONT 978249 50100 AUBURN
20607 -ANTOVER’ TO9ITS4 SOSOT-WEERSPORT
20701 ANGELICA 1942560 50401 CATOUMERTDIAN «6
uum 2OBOT BELFAST TV0S385 ~~ —“SO7OT SOUTHERN CAYIIG 1. 001 96
21001 BOLIVAR 1.00827 51101 FORT BYR
"211 O2-CANASERAGA T9E4IIS—— CST SOTO NOR UTS
21501 CUBA 989161

2YSOT FRIENDSHIP 1714369
22001 FILLMORE 1.02095
~ 221077 WHITESVILLE 1715879

541901 UNTON SPRINGS

60201 SOUTHWESTERN 1.004387
60301 FREWSRURG +2

01
22501, RUSHFORU 966521 —SOAOI-CASSAIIAGA VALI 13 02462
. Serer BEE z 1086rs 60501 MAYVILLE 97208
22601. WELLSV ELLE «977491 ——SOS02- CHAUTAUQUA TTOSS2
22901 RICHBURG POLST 60801 PINE VALLEY 1.02142
60701 CLYNER LVO3959
30101 CHENANGO FORKS .969494 60800 TUNKTRK 996280
30200-BINGHAMTON + 8 72307 — STOOL REMUS FOINT™ 2 9990977
FO501 HARFURSUTLILE 1.0 61101 FAL 1.03874

SO601 SUSQUEHANNA VA~- 9628287
ZO701 CHENANGO-VALLE «979196

——TF IP OITMATNEFENTUELIN —. 99S 156
31301 DEFOSTT 1.02079

314017WHITNEY-FOTNT Te02256 7
31501 UNTON-ENDTCOTT «974006 ‘

T1502 JOHNSON (CITY 17 03932 °°
31601 VESTAL. 1P9S5AO4

“= —"31701 WINDSOR “7 Ve8745a7

TSTSOT- SILVER CREEK” 6987997 ~

61503 FORESTVILLE 2950372
S1S01 PANAMA TT 00T8S6
61700 JAMESTOUN 1.01212
62201 FREDONIA 40081877
62301 RROCTON + 976238
624017 RIPLEY 777 7. 002027
62601 SHERMAN 1.03960
——SPPor WESTFIELT 29740847
70600 ELMIRA + 995884
"70901 HORSEHEATIS 1.00830

70902 ELMIRA HETGHTS .9469744
APPENDIX B

SAMPLE OUTPUT OF A STRATEGIC SIMULATION*

“The Dynamics of State Aid to Education: Interactions Between
Special Education, Regular Education, and Non-Schooling
Expenditures." Tanette Nguyen, David Andersen, and Fiona Chen,
Graduate School of Public Affairs, State University of New York at

Albany, 1980.

. E
Source:
DISTRICT EDUCATION IMPROVEMENT INDICES PROPOSED FOR USE IN OPERATING AID
FORMULA 1980-81 SCHOOL YEAR, BY COUNTY

Educational Educational
Dist. Improvement Dist. Improvement
Code District Name Index Code District Name Index
“SOICT AFTUR Sa ae7 130200 BEACON 1.00856
80201 RAINERINMGE-GU. 1. 00891 130502" DOVER ~~~ "17086247
“80601 GREENE . ™ 130801 HYDE FARK 6995636 :
80701 MOUNT UPTON . [TS LTO NORTHEAST 14010867;
~~~ ST001- NEW RERLIN "13 06375 131201 FAWLING «99SROP
81002 S. NEW RERLIN 1.02800 [T2130 I- PINE FLAINS —_
"81200 ° NORWICH ~ 789877 048- “! 131500 FOUGHKEEFPSTIE
81401 GEORGETOWN 1.02929 —TSTSOI ARLINGTON "1700081
—~B1 501 OXF ORT! 1.0595 : 131602 SPACKENKILL «9824627
82001 SHFRBURNE-EARI. 1.00864 T3170 RED Hao + P7STS2
131801 RHINEBECK 1.00094
90205 AUSABLE” VALLEY 170693571 3201 UAPPINGERS 999327—

90301 KEEKMANTOMWN 004669 i 132201 MILLEROOK 1.01551

~BOE0I NORTHEASTERN I702820
aed YON as p80458 “THOTT ALDEN sire
20801 HGNNEMOR: T4a765 140203 WILLTAMSVILLE .993470
90901 NORTHERN ATIIRO 1.00447 CO eb ES HEME “eae
SSTOIT PERU nc nr eee 7eoee

213 EGGERTSUTL, 20172
$1200 PLATTSBURGH tis psQets EGGE TE a OR ee

01709

_ eons — -- TTAORTBTAMHERSTHSNYDER 1700178
91401 SARANAT 1962574 140251 AMHERST CHS 999718
120301 ET AURORA OTT a
100501 COPAKE-TACONTC .979024 140600 BUFFALO 16417944
—" 400902 GERMANTOUN TVO0827 1207 OI-CHEEKTOWAGAT TOU247T a
101001 CHATHAM 1.01954 140702 CHEEK-MARYVALE .99018°
—~10T300-HUT'SOW Toos4i” ~~; “Y40703 “CLEVELAND “HILT: 984589 —7>
101401 KINIERHOOK 1994165 140707 DEFEW 19463434
——jO1601 “NEW-LEBANON 77 979542 ~~ —T40709-THEEK=SLOAN Th 023347
140801 CLARENCE L99TS6R
410101 CINCINNATUS 1.02788 EATLOT-SPRINGUILEE a OR9 72
—}10200- CORTLAND 96asBS 14201 EDEN 101800

iT41 3O1-TROQUDIS—— 4.017 49-77 ~
-——— 141401 EVANS-BRANT LS 1.01928

110304 MCGRAW 6957713
PL O7 O17 HOMER OD -

. 18 ““T41501 GRANT “ISLAND "594917077
110901 MARATHON 83 141601 HAMBURG 1.02300
_* —TV31604-FRONTIER

~ L2OLO2 “ANTES ™ 7" >", 860787 141701 HOLLANT! 1. 02302

ox! $01 TOWNSVILL 141 BOO"LACKALIANNA "7" 1 601542

OT CHARLOTTE | 141901 LANCASTER 1.00948
120501 DELHI _ TT Ta2OrTaKRON” ~~ 977961 ~

“420701 FRANKLIN
120906 HANCOCK L
121401 MARGARETUTLLE 41.022

142201 NORTH COLLINS 1.03553
—T42301-ORCHARN PARK “1 00749 ~~

142500 TONAWANTIA 996981
121501 GRAND GORGE ‘ 142601 KENMORE-TONAWA 1.01796
121502 ROXBURY ~~~ “4 i 142801 UFST SFNFECA 1.00159
121401 STIINEY 1.00590 1
121701 STAMFORT 1.00950

121702 S. KORTRTIGHT 1.042460
1279014 bal TAN

‘Sector Ut: Contra! Citres

see Siete

1 sac 1 wy woot nates
acta Mogi oat iets

Sector Hl

Seetar WY baw wealth Districts

tese00

A. Percentage Change of State Aid Fraction B. Percentage Change of Tax Rates
versus time lyrs.) versus time [yrs.|

seen anita basticts

: : :

€. Adequacy of Education D: Adequacy of Local Other Services
versus time [yrs.| + versus time [yrs.}

E. Adequacy of State Expenditures
versus time |yrs.|

Changes in Both Stace and Local Sectors Resulting from 2 20% step increase in Needed
Educational Expenditure per pupil in local tactor It (Cestral Cities).