AN ECONOMIC ANALYSIS OF ALTERNATIVE REVENUE
GENERATING POLICIES FOR STATE FINANCING
JAMES RONALD BAARDA
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL
OF THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
This dissertation is dedicated to my Mother and Father,
whose inspiration through teaching and example has been the
mainstay of my life.
The author wishes to express his appreciation to Dr. W. W.
McPherson, Dr. B. R. Eddleman, Dr. D. Jansma and Dr. C. Donovan
for their assistance and critical review of the manuscript. A
very special debt is owed to the author's wife for her tolerance
during the years of graduate school and for her excellent typing
of all drafts and the final preparation of the dissertation.
TABLE OF CONTENTS
ACKNOWLEDGEMENTS . . . . . . . . . . iii
ABSTRACT . . . . . . . . . . . . . vi
CHAPTER I. INTRODUCTION . ... . . . . .. 1
Statement of the Problem . . ... . . . . 1
Objectives . . . ... .. . . . ... 1
Plan of the Work ... . . . . . . . . 2
Taxation and Economics ......... .. .. 3
CHAPTER II. FLORIDA'S FINANCIAL SYSTEM . . . . . 9
The Financing Process . . . . . . . 9
The Basic Accounting Characteristics . . . 10
General Revenue Fund Taxes . . . . . ... 17
Taxes of Special Interest . . . . . ... 25
CHAPTER III. PREVIOUS WORK AND MATHEMATICAL FOUNDATIONS .. 38
Previous Work .............. . .. . 38
Procedure . . . . . . . . . . 43
CHAPTER IV. EMPIRICAL RESULTS . . . . . . .. 56
The Input-Output Model .... . . . .. . 56
Tax Extensions of the Model . . . . . . 64
Tax Intensity Coefficients . . . . . . 72
CHAPTER V. TAX EXTENSIONS OF THE MODEL . . . ... 75
Fitting Taxes Into the Model . . . . . . 75
External Information Requirements and Assumptions . 95
CHAPTER VI. TAX ANALYSIS .. . . . .........
Tax Response to Final Demand Changes .. . .....
Regression Analysis Comparison .....
Further Tax Analysis Measures .......
Summary of Tax Analysis Measures .. . . ......
Application of Tax Analysis Techniques .. . ....
Tax Policy Considerations ........
The Objectives of the Project ......
CHAPTER VII. SUMMARY . . . . . . . . . .
Summary of the Project .. . . . ........
Suggestions for Further Research .. . . ......
APPENDIX A. BASIC INPUT-OUTPUT MODEL CALCULATIONS . . .
APPENXIX B. ALLOCATION OF TAXES TO EXTENDED INPUT-
OUTPUT MODEL . . . . . . . . .
APPENDIX C. DEVELOPMENT OF TAX COEFFICIENTS . . . .
APPENDIX D. TAX PROJECTION TO 1972 . . . . . .
APPENDIX E. DEVELOPMENT OF TAX INTENSITY COEFFICIENTS . .
APPENDIX F. COMPOSITE OF TAX INTENSITY COEFFICIENTS
AND TAX FRACTIONS . . . . . .
APPENDIX G. SECTOR TAXES FROM TAX INTENSITY
COEFFICIENTS, 1967 AND 1972 PROJECTIONS .
SKETCH . . . . . . . . . . .
Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
AN ECONOMIC ANALYSIS OF ALTERNATIVE REVENUE
GENERATING POLICIES FOR STATE FINANCING
James Ronald Baarda
Chairman: Dr. W. W. McPherson
Major Department: Food and Resource Economics
The objectives of this study were (1) to develop an interindustry
model of the state economy to permit a study of the effects of changes
in the economy on the resulting changes in state revenue generated
from each industry sector and (2) to empirically use the model to
evaluate the impact of changes in each economic sector on state
revenues under existing and alternative tax revenue generating policies
A 44 sector input-output model of Florida's 1967 economy was used
as a base model. Taxes of interest were classified into three types
depending on their structural relationship to the model and its exten-
sions. The tax type was used as a guide for extension of the base
model, and taxes were allocated to sectors on the basis of final demand
values, gross sector output values and sector value added values.
From the extended model seven analytical measures useful for tax
analysis were developed.
"Tax coefficients" for each sector were developed showing what
proportion of gross sector output comprised tax revenue generation.
"Tax intensity coefficients" were defined as a set of numbers which,
when multiplied by any given set of final demand figures, yield the
resultant tax revenue generation for a given tax. These coefficients
were derived for each sector. "Tax fractions" were numbers performing
a similar function but for retail sales type taxes.
It was shown that a matrix of values could be calculated which
indicate the change in tax payment by any sector for any given change
in final demand for the output of any other sector. A set of per-
centages was developed to show what percent of total tax generated
by equal changes in final demand in all sectors was caused by each
sector's own final demand increase. An additional set of percentages
was developed to show, for a final demand change in a single sector,
the portion of the entire tax increase paid by the sector whose final
demand increased. Finally, ratios were developed to show the relation-
ship of total tax paid by sector and the total tax stimulated through-
out the economy by the final demand for that sector's output.
The methods of final demand projections were used to project
final demand vectors to 1972. Using the tax analysis techniques de-
veloped, 1972 tax values were estimated from the projected final
demands and were compared with reported tax revenue and tax estimates
from four regression equations.
The problems, assumptions, data requirements and significance of
tax classification and application of the extended interindustry model
were discussed in detail. Policy implications of the tax analysis
techniques developed in this study were also discussed.
Statement of the Problem
Governments make expenditures to achieve specific goals, generally
without regard to precisely how the expenditures will affect future
generation of revenues. The government depends on these revenues for
its ability to meet the needs of its citizens and the understanding of
how the generation of revenues is related to the economy of the politi-
cal units should be essential to wise fiscal policies. In general, how-
ever, this relationship is little understood. The problem that this
research project will investigate is that of determining relationships
between the economic structure of the state on the one hand, and revenue
generated in that economy on the other, with specific reference to the
State of Florida and Florida's economic structure. The specific problem
is that of determining how tax revenue generation is affected by changes
in economic activity brought about by changes in final demands for out-
The objectives of this project are twofold:
1. To develop an interindustry model of the state economy that
will permit a study of the effects of changes in the economy on the
resulting changes in state revenue generation from each industry sector.
2. To empirically use the model to evaluate the impact of changes
in each economic sector on state revenues under existing and alterna-
tive future tax-revenue generating policies and structures.
Plan of the Work
The following section discusses the place of this study in the
framework of the economic study of taxation with a summary of pertinent
economic concepts in public finance. Chapter II describes the public
finance system of Florida first, then focuses on particular taxes of
interest in this study.
Chapter III begins with the description of relevant previous work,
including summaries of the findings. The second part of the chapter de-
velopes the conceptual methods of the study in mathematical terms and
lays the foundation of specific empirical calculations. These calcula-
tions are described in detail in Chapter IV. The basic model used is
described, followed by three methods of-projecting final demand figures
from 1967 to 1972. Tax extensions of the model then show how tax data
are fitted into the extended input-output model,and how tax coefficients
and tax fractions are calculated. Tax intensity coefficients are de-
rived, and payment of tax by sector and response of tax revenue genera-
tion to sector final demand change is shown for the projection methods.
Chapter V is a detailed analysis of the choice of tax treatment for
each tax study and gives the basis for such choices and the data re-
quirements and assumptions necessary for a valid use of the input-output
model in tax analysis.
Chapter VI combines the mathematical formulations of the model from
Chapter III, the empirical work of Chapter IV and the tax treatment
choices of Chapter V to obtain a basic analytical framework for tax
analysis. Analytical measures are developed and the application of the
array of tax analysis techniques follows. Finally, the place of the tax
analysis techniques in the context of policy decisions in state public
finance is discussed.
Taxation and Economics
Taxation has been a subject of interest to economists for many
years, yet the diversiLy of these interests has left a considerable
number of gaps in the application of economic theory and research methods
to the evaluation of alternative systems for the collection of taxes.
One of these gaps, which will require considerable effort on the part
of economists, is that of integrating the theory and practice of govern-
ment finance as related to the generation of tax revenues from economies
comprising political units. This is not to imply that the considerable
interests by economists in the theory of taxation is unrelated to the
generation of tax revenue; rather, it is a difference of emphasis. In
fact, answers to tax revenue generation problems have most interesting
applications at all levels of economic theory that have to date been
concerned with taxation.
It is the purpose of this section on taxation and economics to
outline, though briefly, some ways in which economic theory and practice
have approached the subject of taxation. By outlining some past and
present work of a different nature on taxation, the results of this
particular study, though its purpose is specifically pragmatic, may add
to or suggest some significance to other bodies of thought. The place
of taxation in microeconomics, macroeconomics and welfare economics is
too far removed from the subject matter of this dissertation to con-
sider in detail. Theories of public finance come closer, and the
approach to tax revenue generation found in this discipline are sum-
Modern macroeconomics and fiscal policy are of course directly de-
pendent on taxation as a tool to achieve some economic goals. The
emphasis on these two bodies of economic concepts is toward the effects
of taxation on multipliers, equilibrium, etc., and not the effects of
economic change on tax revenue generation. A second and important reason
that macroeconomic and fiscal policy analyses of taxation are not im-
portant to this study is that the nature of taxation for purposes of
providing continual support for a state or local government is far re-
moved from that involved in the federal system, where deficits and sur-
pluses are useful tools for fiscal policy, not threats to the financial
soundness of the government.
Taxation is also an important part of welfare economics. It is
often used as a device for correcting the difficulties in economic
equilibrium caused by the presence of externalities. When these take on
the nature of social versus private cost, the domain of public finance
is approached, but, as with most other considerations of taxation, they
emphasize the use of taxation as a tool rather than as a means of
Public finance is the single large component of economics that has
come close to building a theory of taxation. It is, of course, not iso-
lated from macroeconomics or welfare economics, and draws heavily on
both for much of its structure including fiscal and monetary policy.
The theory of public finance, despite its name, is concerned with
numerous problems of the public household not directly related to taxa-
tion, and this is specially true in the setting of state, rather than
The theory of public finance has been only incidentally concerned
with problems of state taxation for two reasons. The first reason is
that the term "public finance" is something of a misnomer for the true
emphasis of those economists concerned with it. The most well known and
widely used book on public finance is Musgrave's work in which he states:
The complex of problems that center around the revenue expendi-
ture process of government is referred to traditionally as public
finance. Following this convention, the same term is used in that title
of this volume. But with much hesitation. Whole operations of the
public household involve money flows of concepts and expenditures, the
basic problems are not issues of finance. They are not concerned with
money, liquidity, or capital markets. Rather, they are problems of re-
source allocation, the distribution of income, full employment, and
price level stability and growth. [Musgrave (1959), p.3].
The second reason for the neglect of state taxation problems as
problems of revenue generation has to do with the nature of state fi-
nance, as opposed to a national financial structure. This can best be
seen by dividing the "public household" into three budgetary branches
[Musgrave (1959)]. These three branches are the allocation branch, the
distribution branch, and the stabilization branch. While state taxa-
tion is primarily, though not exclusively, concerned with the alloca-
tion branch, the primary interest in public finance theory rests on
the distribution and stabilization branches, and the great volume of
the theory of public finance is found on these two branches.
The distribution branch of the public household is concerned with
the "proper" distribution of economic benefits, on whatever criteria the
definition of "proper" rests. The stabilization branch is concerned
with activities aimed at stabilizing undesirable fluctuations of employ-
ment levels, wages, prices, etc. The taxing activities of a state
government deal very little with stabilization, and relatively more
with distribution. The function of a state government, as reflected in
its tax revenue financed budget, is the supply of public goods. Taxes
are not a means of redistributing income nor of stabilizing the economy.
They are a means of financing the supply of public goods. This results
in a strong connection with the allocation branch of the public house-
hold and little connection with the distribution and stabilization
branches. Though state governments are interested to varying degrees
in distribution and stabilization, the means of influencing them at the
state level is, even when present, not primarily budgetary.
The function of the allocation branch of the public household is
to provide for the satisfaction of public wants. The tax structure
of state government is based on the need to fund these functions, and
it is the objective of this project to investigate certain characteris-
tics of this funding process. It is in this function of taxes--as
sources of revenue to satisfy public wants--that the theory of public
finance deals with principles of taxation in the way most relevant
to a state taxing system, though the particular approach to the tax-
ing process taken in this project differs in essential ways from the
Over the many years of discussion of how to provide for public
wants, two distinct points of view have arisen. One is called the
"benefit" approach to satisfying public wants, and the other the
"ability-to-pay" approach [Musgrave (1959); Musgrave and Peacock
(1958)]. The general nature of each approach is suggested by their
titles. The benefit approach takes a "market" view of public goods
and services, so that those who benefit from these goods and services
must pay in some proportion to the benefit received from specific
expenditures of tax revenue.
There is a crucial difference between the approach to taxation
taken in either the benefit or ability-to-pay views of public want
satisfaction and the view taken in this project. The underlying purpose
of the public finance theory dealing with satisfaction of public wants
and taxation is to investigate the relationship between the burden of
taxation and the supply of public goods and services. The approach of
this project, on the other hand precisely, takes the supply of public
goods and services as given, and investigates the relationship between
the burden of taxation and the state's economy, treating the economic
system as independent of the use of the collected taxes in supplying
public goods and services.
Though considerable theoretical work has been done on the charac-
teristics of taxation which are of interest to the theory of public
finance, and although these, for the most part, are not directly re-
levant to this project because of the differing purposes of this project
and past theoretical work, a number of attributes of a taxing system
still remain to be studied by the economist. If all taxation criteria
that are involved in achieving goals of the distribution and stabiliza-
tion branches of the public household are eliminated, and if the
questions of how the tax burden should be apportioned among the
populous of a state for purposes of meeting "ability-to-pay" or "benefit"
criteria (whatever they might be) for funding the allocation branch
are also eliminated, there still remain some important questions of
taxation. These remaining questions are of extreme interest to public
policy makers involved with financing state operations. They are com-
monly divided into three--tax adequacy, tax flexibility, and tax
sensitivity or stability.
The adequacy of tax refers to its ability to meet the financial
needs of the government. Though the flexibility and sensitivity of a
tax can be viewed as part of the adequacy of a tax, they are separated
here because they will he emphasized in this project to the exclusion
of general adequacy. The general adequacy of a tax will not be direct-
ly studied for two reasons. First, the adequacy of a tax does not
treat expenditures as independent of tax revenue generation, since
expenditures are made partly on the basis of how much money will be
available from the revenue of taxes. Secondly, Florida's present fi-
nancial structure is on the whole balanced, so that a tax system is
not being studied as if it were de novo, but as a presently adequate
and continuing system. The adequacy of the tax in the future thus de-
pends on the changes that will take place in the present taxing and
economic structures, changes best described as problems of flexibility
The flexibility of a tax is its ability to provide changing re-
venue as a function of changes in the tax structure, such as a change
in the rate of a particular tax. The sensitivity of a tax as the term
is commonly used, is its ability to provide changing revenue in re-
sponse to increased total or per capital income. Tax stability is
actually the same as tax sensitivity, stated in terms of a tax's
ability to provide constant revenue in spite of fluctuating economic
conditions. Sensitivity has replaced stability in emphasis in past
years because of the lack of large economic fluctuations which in the
past threatened the financial soundness of state government. More
interest is now shown in tax sensitivity to take advantage of long-run
trends in state economies. This project deals with refinements in tax
flexibility and sensitivity measures, and particularly with tax
FLORIDA'S FINANCIAL SYSTEM
To understand the place which the subject matter of this project
occupies in the total financial system of Florida and to understand how
to make a number of decisions that must be made in the process of com-
pleting this project, familiarity with the overall structure of
Florida's financial system is necessary. This and the following sec-
tion summarize the financing process, describing some basic accounting
characteristics of the system, summarizing the taxing system of inter-
est, and finally describing in more detail the taxes of special inter-
The Financing Process
The budget planning cycle is a detailed process, and a scenario of
the cycle for the 1972-73 fiscal year (July 1972-June 1973) is summa-
rized [Governor's Budget Report (1972-73)]. The process started in
July 1971 when the Department of Administration issued instructions to
all state agencies asking for submission for budget requests and pro-
posed plans. The documents were submitted to the agency from September
to November 1971. With this information the Department of Administra-
tion prepared an analysis by December, 1971. In November the Governor's
budget hearings took place, with public hearings on the agencies'
In January of 1972 the Governor's recommended budget and proposed
agency plan was submitted to the legislature. The legislature considered
the recommendations and agency plans during February to April, 1972,
and an appropriations act resulted. This appropriation was considered
by the Governor and appropriate action was taken. In May of 1972 the
Division of Planning and Budgeting furnished each state agency with an
approved budget under which the agencies operated in the 1972-73
Basic Accounting Characteristics
Though an appreciation of the budget process is useful, even more
important is the general accounting structure of Florida. A brief view
of the structure can be divided into two parts. First the replenishment
of funds from revenue sources, and, second, the use of the funds to
carry out the functions of government. Because the generation of re-
venue is of greater interest in this project, funding of government
functions will be mentioned first and only briefly.
No money can be used from the state treasury except in pursuance of
appropriations made by law [Florida Constitution, Article VII, Section 1
(c)]. Making these appropriations of public funds is the duty of the
legislature. There are four ways in which appropriations, as spending
authority, may arise:
1. General Appropriations Act. This is the single enactment of
law making provision for most state agency spending during the
coming fiscal period. It is the principle source of spending
authority during each fiscal period, and includes both general
revenue fund and some trust fund spending.
2. Special Appropriations Acts. An act separate from the general
appropriations act may fund a particular program. There may
be a number of these special appropriations acts.
3. A statutory provision may direct that a particular purpose
be funded in the "amount necessary."
4. A general law [Florida Statutes 215.32(3)] specifies that
trust fund money shall be appropriated for the purpose for
which the money was received. This appropriation comprises
most of the spending power not given by the general appro-
All money received by the state is deposited in the State Trea-
sury unless otherwise provided by state law [Florida Statutes 215.32].
The money so received is divided into three funds--the general revenue
fund, the trust fund accounts, and the working capital accounts.
1. General Revenue Fund. This fund consists of all receipts of
a general, non-dedicated, and non-earmarked nature. Almost
all of these funds are expended pursuant to general revenue
funds appropriations acts.
2. Trust Fund. This account consists of all money received by
the state which, under law or trust agreement, is segregated
for a purpose authorized by law. The state agency collecting
or receiving these funds is responsible for their proper ex-
penditure as provided by law.
3. Working Capital Fund. This fund is a special fund of not more
than $50,000,000 transferred from the general revenue fund sur-
pluses remaining at the end of the fiscal year. It is pri-
marily for temporary working capital needs.
It is necessary to narrow the scope in this project from a study of
total receipts by the state government to a reasonable number of tax
receipts. A first step in the selection process is a choice of fund
receipts, making a choice of tax receipts contributing to the general
revenue fund, the trust fund, or the working capital fund. The working
capital fund is relatively minor in amount, and is a transfer only.
It was thus eliminated from consideration. A choice between the general
revenue fund and the trust fund remains, and taxes contributing to the
general revenue fund have been chosen for a number of reasons. One of
these reasons concerns the general approach of this project and the
others concern the nature of the fund receipts in relation to taxes.
As has been pointed out previously, the focal point of this pro-
ject is the collection of taxes as a source of state revenue, and any
relation between tax revenue generation and expenditures is specifically
excluded as a point of study. The supply of public goods and services
is not considered, only the financing of the means of providing such
goods and services. The very nature of the trust fund accounts directly
ties taxes collected with expenditures made, since it is precisely
this property of earmarking that sets aside the trust fund as a
separate accounting unit.
On the other hand, the general revenue fund account is expended by
general appropriations acts with no regard to the specific source of
funds. No connection exists between a particular tax and expenditures
for satisfying public wants other than that tax's general contribution
to the amount of money available for expenditures on public goods and
A second and related reason makes the tax collection of the gener-
al revenue fund more appealing than that of the trust fund. As in-
dicated in the statement of the problem the interest in this project
is that portion of revenue which supports public expenditures that are
made without regard to the specific source of funds. In general,
though by no means always, a trust fund revenue source arises directly
with the provision of some public good or service. In many cases, the
law which provides for these expenditures concomitantly provides a
specific source of earmarked funds. Thus the manner in which trust
funds arise makes the trust fund less appealing in a study of unre-
stricted general taxation.
Trust fund receipts make up a larger portion of total receipts than
general revenue fund receipts. Figures for the 10-year period ending
with the fiscal year 1972-73 are given in Table 1. The percentages of
total revenue receipts contributed by the general revenue fund and by
the trust fund are given for the same period in Table 2.
The information in Tables 1 and 2 would seem to indicate that con-
centration of general revenue funds taxation is concentration on a re-
latively minor part of state financing. This is not in fact the case.
Table 3 gives a slightly more detailed breakdown of funds and receipts
for fiscal year 1972-73.
The differences between totals for each fund and the fees,
licenses, and tax sources are made up of the following categories:
Aids and Donations
Counties and cities
Other Direct Revenue
Table 1. Total Receipts, by fund, for the period 1963-64 to 1972-73.
1972-73 1971-72 1970-71 1969-70 1968-69
General Revenue Fund $1,B29,1S3,058.08 $1,522,894,809.92 $1,247,043,299.32 $1,073,027,473.10 $ 933,638,297.55
Trust Fund 4,775,599,482.47 3,500,994,956.66 2,697,270,995.27 2,152,598,131.10 2,130,936,236.62
Working Capital Fund 28,255,545.88 21,744,454.12 49,97/,579.98 23,802.12 61,663,665.39
TOTAL RECEIPTS $6,633,043,086.43 $5,045,634,220.70 $3,994,288,874.57 $3,225,649,406.32 $3,131,238,249.56
1967-68 1966-67 1965-66 1964-65 1963-64
General Revenue Fund $ 664,534,661.96 $ 379,932,494.97 $ 546,211,163.22 $ 515,600,856.68 $ 471,156,581.43
Trust Fund 1,789,473,433.57 1,680,659,438.19 1,423,922,717.36 1,356,789,184.13 1,121,845,423.63
Working Capital Fund 64,720,793.83 86,192,000.00 50,844,171.61 6.053,000.00 2,997,854.40
TOTAL RECEIPTS $2,518,728,889.36 $2,346,783,933.16 $2,020,978,052.19 $1,878,443,040.81 $1,595,999,859.46
Source: Florida Comptroller, Annual Report, various years.
Table 2. General Revenue Fund (GRF) and Trust Fund
(TF) as percentage of Total Gross Receipts,
by year, for the period 1963-64 to 1972-73
Year GRF TF
Source: Florida Comptroller, Annual Report,
various years, by calculation.
Table 3. Partial breakdown of total receipts by fund, fiscal years 1967-68 and 1972-73.
General Revenue Fund $1,829,188,058.08 $24,974,999.24 $117,946,268.29 $1,489,951,358.00
Trust Fund 4,775,599,482.47 52,997,360.50 65,613,301.41 831,312,483.43
Working Capital Fund 28,255,545.88
TOTAL ALL FUNDS $6,633,043,086.43 $77,972,359.74 $183,559,569.70 $2,321,263,841.43
Fund Total Receipts Fees Licenses Taxes
General Revenue Fund $ 664,534,661.96 $11,680,856.08 $65,679,882.49 $53 .334,973.65
Trust Fund 1,789,473,433.57 20,634,161.10 33,526,188.17 388,844,159.11
Working Capital Fund 64,720,793.83
TOTAL ALL FUNDS $2,518,728,889.36 $32,315,017.18 $99,206,070.66 $921,179,132.76
Source: Florida Comptroller, Annual Report, 1967-68 and 1972-73.
Sale of investments
Sale of revenue certificates
It can be seen from Table 3 that taxes going into the trust fund
comprise only 35.81 percent of total taxes collected with taxes to the
general revenue fund providing 64.19 percent of total taxes and 81.45
percent of the total general revenue fund receipts for the fiscal year
1972-73. Corresponding figures for fiscal 1967-68 are 42.21 percent,
57.79 percent and 80.11 percent,respectively. Detailed breakdowns of
the taxes going into the general revenue fund and the trust fund are
shown in Tables 4 and 5, respectively. Table 6 gives taxes contributing
to the general revenue fund as a percentage of total state funds and of
total general revenue funds for the 10-year period ending with fiscal
General Revenue Fund Taxes
A more detailed understanding of the tax revenues contributing
it the general fund is necessary. The breakdown of tax receipts as
reported by the comptroller's reports for the 10-year period ending
fiscal year 1972-73 is given in Table 7 including percentages of total
General Revenue Fund receipts for each classification. The consider-
able differences in classification and amounts between the general
revenue fund breakdown by taxes in Table 4 and the breakdown of total
funds in Table 6 indicates the difference between a general system of
classification used for descriptive purposes and the more specific
Table 4. Taxes directly contributing to the general revenue fund, fiscal
years 1972-73 and 1967-68
Corporation charter tax
Corporation commission tax
Documoi.'iry stnmp tax
Tax certificate fund
Tax redemption fund
Pullman Company gross receipts tax
Inheritance and estate tax
Water and sewer tax
Pari-mutuel 5% tax
Pari-mutuel breaks tax
Breaks under Chapter 29810
Sales and use tax
Corporate capital stock tax
Elections commissions tax
Corporate privilege tax
Corporate income tax
Source: Florida Comptroller, Annual Report, 1967-68 and 1972-73.
* Indicates the most important tax source on the basis of dollar
volume of tax collected.
** Indicates no tax or a changed classification.
Table 5. Taxes directly contributing to the trust fund, fiscal years
1972-73 and 1967-68.
Intangible tax trust fund $
OiL and gas tax trust fund
Land acquisition clearing trust fund
State fire marshall trust fund
Municipal fireman's pension trust fund
Municipal police officers' retirement
Florida harness horse racing promotion
Cigarette tax collection trust fund
Racing commission operating trust fund
Additional harness and dog track tax
Florida horse racing promotion trust
Alcoholic rehabilitation trust fund
Auto transportation road tax clearing
PSC regulatory trust fund
Gasoline tax clearing trust fund I.
Special motor vehicle fuel tax clear-
ing trust fund
Sporting goods tax clearing trust fund
Recreational land acquisition trust fund
Citrus advertising trust fund
Additional grapefruit tax trust fund
30,082,700.41 $ 91,400,131.57
Table 5 continued.
Tax 1967-68 1972-73
Citrus advertising, emergency trust fund
Florida orange products advertising trust
Brand advertising reserve trust fund
Special brand promotion trust fund
Racing scholarship trust fund
Educational survey trust fund
Board of Education capital outlay bond
Workman's compensation administration
Workman's compensation special dis-
ability trust fund
Unemployment compensation clearing
Quarter horse racing trust fund
Bicentennial Commission trust fund
Gross receipts tax--express companies
2% premium tax clearing trust fund
Severance tax, solid mineral trust fund
State Park trust fund
Source: Florida Comptroller, Annual Report, 1967-68 and 1972-73.
* Indicates no tax or changed classification.
Table 6. Taxes in the General Revenue Fund as percentages of
total General Revenue Fund (% GRF) receipts and total
state revenue (% TOTAL) receipts, fiscal years 1963-64
Year % TOTAL % GRF
1972-73 22.46% 81.45%
1971-72 23.29 77.17
1970-71 23.38 74.90
1969-70 26.58 79.90
1968-69 24.81 82.76
1967-68 21.14 80.11
1966-67 19.45 78.71
1965-66 21.24 78.59
1964-65 21.15 77.04
1963-64 22.69 76.85
Source: Florida Comptroller, Annual Report, various years,
Table 7. Revenue from the Florida Tax Dollar to
1963-64 to 1972-73.
Revenue Source 1972-73
Sales and Use Tax $1,041,145,003
% Total Revenue 56.92
Beverage Tax, Licenses, Fees 150,509,948
% Total Revenue 8.78
Motor Vehicle Licenses, Fees 116,503,159
% Total Revenue 6.37
Documentary Stamp Tax 81,374,669
% Total Revenue 4.44
Insurance Premium Taxes 30,581,649
% Total Revenue 1.67
Racing Taxes 28,266,065
% Total Revenue 1.55
Cigarette Taxes 33,419,837
% Total Revenue 1.83
Corporation Income Tax 147,708,344
% Total Revenue 8.07
Intangible Property Tax 39,306,466
% Total Revenue 2.15
Other Taxes, Licenses, Fees 150,307,746
% Total Revenue 8.22
TOTALS S1.829 122 866
Comptroller's classification, fiscal years
the General Revenue Fund,
$ 975,767,559 $
Table 7 continued.
Sales and Use Tax
% Total Revenue
Beverage Tax, Licences, Fees
% Total Revenue
Motor Vehicle Licenses, Fees
% Total Revenue
Documentary Stamp Tax
% Total Revenue
Insurance Premium Taxes
% Total Revenue
% Total Revenue
% Tocal Revenue
Corporation Income Tax
% Total Revenue
Intangible Property Tax
% Total Revenue
Other Taxes, Licenses, Fees
% Total Revenue
Source: Florida Comptroller, Annual Report, various years.
tax classifications required for analytical purpose. Though no
detail justification of the differences in the two tables is necessary,
mention of some important points is helpful.
Concentrating on 1972-73 figures, the following observations can
be made. The sales and use tax is a total figure, with no other contri-
butions to that category than the tax itself. Most of the beverage tax,
license and fee category is made up of the beverage tax, with the net
consisting of the beverage license transfer. Motor vehicle licenses
and fees funds come entirely from licenses and fees with no tax contri-
bution. Documentary stamp tax funds come entirely from the documentary
stamp tax. The insurance premium 2 percent tax is a transfer required
by law and is accounted for as a trust fund tax collection transferred
to the General Fund. Racing taxes are composed of three racing taxes
listed in Table 4. Cigarette taxes are, like insurance premium taxes,
a trust fund collected to the General Revenue Fund by law [Florida
Statutes 210.20 and Florida Statutes 215.22 (19)]. Corporation in-
come tax is a direct collection, collected for the first time in 1971-
72. The intangible tax is a transfer required by law [Florida
Statutes 199.292] from a trust fund account and was shown for the
first time in the accounts in 1971-72.
Revenues studied in the project are limited to taxes, and only
those taxes collected directly into the General Revenue Fund. A fur-
ther selection was made of the major taxes consisting of those taxes
contributing over $1 million each to the General Revenue Fund in the
fiscal year 1972-73. These taxes are marked by single asterisks in
Table 4. One source of the corporation privilege tax contributes al-
most all of that tax, and that portion only is included in tables other
than Table 4.
The selected taxes are given in Table 8 for the 10-year period
ending fiscal 1972-73. Each of these taxes is given as a percentage of
total General Revenue Fund taxes, total General Revenue Fund receipts
and total of all state receipts in Tables 9, 10 and 11 respectively.
Total percentages for selected taxes are given in Table 12.
Taxes of Special Interest
The general statutory nature of each tax in the group of selected
taxes is described in the following paragraphs. The descriptions are
condensed considerably for summary purposes. These taxes have not re-
mained constant over the last 10 years either in rates of other sub-
stantive provisions, and the descriptions of focus on the 1972-73 period.
Pari-mutuel Wagering Tax
Florida Statutes 550.16 (8) provides for a tax of 4 1/2 percent
to be levied on the contributions to pari-mutuel wagering pools which is
to be paid into the general fund. The breaks tax [Florida Statutes
555.26] is the tax on the difference between the total amount contri-
buted to a pari-mutuel pool and the total of the commission and the sum
actually redistributed to the contributors. The tax is the entire
amount of the break. For most of the empirical work the two taxes are
combined and designated wagering taxes.
The excise tax on alcoholic beverages is paid by the manufacturer,
distributor or vendor [Florida Statutes 261.46]. The rates of tax are
dependent on the type of beverage, its alcoholic content and whether or
not it was manufactured with Florida agricultural products. The rates
are as follows:
Table 8. Tax collection into the General Revenue Fund, selected tax sources, for fiscal years 1963-64 to 1972-73.
Sales and Use
Corp. Capital Stock
Inheritance & Estate
Table 8 continued.
Tax 1967-68 1966-67 1965-66 1964-65 1963-64
Sales and Use
Corp. Capital Stock
Inheritance & Estate
Source: Florida Comptroller, Annual Report, various years.
Table 9. Selected taxes as percent of total General Revenue Fund taxes, fiscal years 1953-64 to 1972-73.
Tax 1972-73 1971-72 1970-71
1969-70 1968-69 1967-68 1966-67 1965-66 1964-65 1963-64
Sales and Use
Corp. Capital Stock
Tnheritance & Escate
1.87 1.88 2.51 2.47 3.68 2.61 4.25
Source: Florida Comptrcller, Annual Report, various years, by calculation.
Table 10. Selected taxes as percent of total General Revenue Fund receipts, fiscal years 1963-64 to 1972-73.
Tax 1972-73 1971-72 1970-71 1969-70 1968-69 1967-68 1966-67 1965-66 1964-65 1963-64
Corp. Char-er 0.10% 0.08% 0.08% 0.13% 0.13% 0.13% 0.09% 0.10% 0.10% 0.11%
Wagering 5% 1.35 1.46 1.69 1.72 1.87 2.39 2.53 2.66 2.66 2.66
Wagering Breaks 0.16 0.19 0.24 0.26 0.21 0.27 0.30 0.33 0.32 0.34
Beverage 8.49 9.21 9.69 11.46 11.77 13.57 13.62 13.59 13.51 13.80
Sales and Use 56.92 57.51 57.35 61.34 61.13 53.82 51.88 51.82 50.53 48.83
Documentary Stamp 4.45 3.85 3.20 2.19 3.27 3.99 3.78 4.23 4.17 4.26
Corp. Capital Stock 0.01 0.20 0.37 0.40 0.46 0.63 0.70 0.70 0.76 0.23
Corp. Privilege 0.04 0.65 -
Corp. Income 8.08 1.83 -
Inheritance & Estate 1.08 2.06 1.48 1.49 1.55 2.01 1.94 2.89 2.01 3.27
Source: Florida Comptroller, Annual Report, various years, by calculation,
Table 11. Selected taxes as percent of total all state receipts, fiscal years 1963-64 to 1972-73.
1972-73 1971-72 1970-71 1969-70 1968-69 1967-68 1966-67 1965-66 1964-65 1963-64
Sales and Use
Corp. Capital Stock
Inheritance & Estate
0.46 0.50 0.47 0.53 0.48 0.78 0.55 0.96
Source: Florida Comptroller, Annual Report, various years, by calculation.
Table 12. Total selected taxes as percentage of General Re-
venue Fund taxes (% GRFT), total General Revenue
Fund receipts (% GRFR) and total all state receipts
(% SR), by year, fiscal years 1963-64 to 1972-73.
Year % GRFT % GRFR % SR
1972-73 99.94% 81.40% 22.45%
1971-72 99.82 77.03 23.24
1970-71 99.90 74.08 23.36
1969-70 98.86 78.99 26.56
1968-69 97.16 79.49 24.10
1967-68 95.89 76.81 20.26
1966-67 95.11 74.84 18.50
1965-66 97.11 76.32 26.63
1964-65 96.12 74.06 20.34
1963-64 93.88 73.70 21.75
Source: Florida Comptroller, Annual Report, various
years, by calculation.
Malt beverages contain more than 1 percent of alcohol by weight--
$0.32 per gallon when in bulk or kegs, $0.04 per pint or fraction
when sold in quantities less than a gallon. No tax is required of
malt beverages containing less than 3.2 percent alcohol when sold on
military reservations within the state.
Beverages including wines, except natural sparkling wines and malt
beverages, containing between 1 percent and 14 percent alcohol by
weight not manufactured in Florida with Florida grown products--$1.15
Beverages including wines, except natural sparkling wines and
malt beverages, containing between 1 percent and 14 percent alcohol by
weight manufactured in Florida with Florida agricultural products--
$0.23 per gallon with a yearly increased schedule to 1975.
All wines, except natural sparkling wines containing more than
14 percent alcohol by weight, not manufactured in Florida with Florida
agricultural products--$1.60 per gallon.
All wines, except natural sparkling wines, containing more than
14 percent alcohol by weight manufactured in Florida with Florida
agricultural products--$0.35 per gallon with a yearly increase schedule
Natural sparkling wines not manufactured in Florida with Florida
products--$2.30 per gallon.
Natural sparkling wines manufactured in Florida with Florida pro-
ducts--$0.46 per gallon with a yearly increased schedule to 1975.
Beverages, except wines, containing between 14 percent and 48 per-
cent alcohol by weight not manufactured in Florida with Florida products--
$3.75 per gallon.
Beverages, except wines, containing between 14 percent and 48 per-
cent alcohol by weight manufactured in Florida with Florida products--
$0.958 per gallon with a yearly increase schedule to 1975.
Beverages containing more than 48 percent alcohol by weight not
manufactured in Florida with Florida products--$7.52 per gallon.
Beverages containing more than 48 percent alcohol by weight, manu-
factured in Florida with Florida products--$1.896 per gallon with a
yearly increase schedule to 1975.
Sales and Use Tax
As indicated in the tables the sales and use tax item is by far the
largest contributor to the General Revenue Fund, and has been so for a
number of years. It is not necessary to give a detailed description of
the sales and use tax, but its basic character and some special features
of interest are useful.
Sales and use tax [Florida Statutes, Chapter 212] is a tax on the
retail sale or use of tangible personal property in Florida and includes
the rental or lease of such items and the provision of a number of ser-
vices. The sales tax is the term used when the sale or the rental of
property is taxed and the use tax applies when the article is not sold
but is brought into Florida for use, consumption, distribution or stor-
age. The term "sale" includes the sale or exchange of tangible personal
property, rental of rooms and facilities for transients, the process of
manufacturing for a consumer using the consumers' goods, and sale where
title is retained as security. A retail sale is a sale of an item for
other than resale purposes, and includes goods and taxable services
consumed in the manufacture of consumer goods, but not the materials
fabricated into the final consumer product.
The tax rate is 4 percent of the sales price on retail sales, col-
lectable by the dealer and added to the sales price. The rate is 4 per-
cent of the cost price for use, consumption, distribution or storage,
and 4 percent of the gross proceeds of lease or rental of tangible
personal property, and of charges for telephone, cable TV service and
electrical power. The rate is also 4 percent on the sale, use con-
sumption or storage for use in Florida of machines and equipment and
their parts and accessories used in manufacturing, processing compound-
ing, producing or mining personal property for sale, or to be used in
furnishing communications, transportation or public utility services.
A number of important exemptions apply to the sales and use tax.
Some of these exemptions are: groceries, medicine, some items bearing
other excise taxes (though not alcoholic or malt beverages), a number
of items used in fishing and agriculture, sales to political sub-
divisions, and other miscellaneous exemptions. A reduced tax of 3 per-
cent applies to certain farm equipment, and other partial exemptions
apply to specific situations.
Documentary Stamp Tax
The documentary stamp tax [Florida Statutes, Chapter 201] is an
excise tax on documents evidencing certain enumerated legal agreements
or legal status. The tax is collected in the form of document stamps
which must be attached to the documents in the amount required for the
tax, similar to the affixing of a postage stamp to pay for postal ser-
vice. These taxes are assessed at the creation of the document and are
non-recurring. Following is a partial list of the documents requiring
the stamp and the amount of the tax imposed:
On deeds and other instruments relating to land--$0.30 for each
$100 of consideration.
Agreements for sale, transfers of legal title, or rights to
subscribe to shares or certificates of stock of a corporation--$0.15
for each $100 of stock of certificate face value if par stock, and
$0.15 for each $100 of actual value if no par stock.
On each original issue of corporate stock--$0.15 for each $100
On bonds, debentures and certificates of indebtness--$0.15 for
each $100 of obligation.
On promissory notes, non-negotiable notes, written obligations to
pay money, and assignments of wages--$0.15 for each $100 of obligations.
Certificates of deposit issued by banks are exempt.
Corporate Capital Stock Tax
The corporate capital stock tax [Florida Statutes 608.33] was a
yearly tax imposed on the corporation based on the investment capital
represented by shares of stock outstanding. The tax did not apply to
railroad, pullman, telephone, telegraph, insurance banking and trust
companies, building and loan associations, cooperative marketing
associations and corporations not for profit. The tax rates were as
Capital Stock Tax
Less than $10,000 $20.00
$10,000 to $25,000 50.00
$25,000 to $50,000 100.00
$50,000 to $100,000 150.00
$100,000 to $200,000 200.00
$200,000 to $500,000 400.00
$500,000 to $1,000,000 1,000.00
$1,000,000 to $2,000,000 1,500.00
Over $2,000,000 2,000.00
The corporate stock tax section was repealed by Florida Laws,
1971, 71-359, in relation to the new corporate income tax.
Inheritance and Estate Tax
The inheritance and estate tax [Florida Statutes, Chapter 198] is
a tax that dovetails with the federal estate tax [Internal Revenue Code,
Subtitle B, Chapter 11, Subchapter A]. The federal tax provides a
credit to payment of federal estate tax up to specified values of in-
heritance and gift taxes paid to the states. Florida, in response to
the limit, imposes an inheritance and estate tax equal to that credit
limit, thus imposing the maximum tax possible without causing double
taxation. Calculation of the Florida tax depends on the calculation of
the tax base for federal estate tax purposes.
Corporate Income Tax
A change in the Constitution of Florida has recently made possible
the imposition of an income tax on corporate income though not on in-
dividual income. The statutory basis for the income tax is found in
Chapter 220 of the Florida Statutes, with the tax becoming effective in
1971. The institution of the income tax was accompanied by changes in
the corporate capital stock tax and the corporate privilege tax. As of
the fiscal year 1972-73 the change over was still effecting the tax
It was the intent of the Florida Legislation, expressed in Florida
Statutes 220.2, to follow as closely as possible the Federal system of
income taxation for corporations (Internal Revenue Code, Subtitle A).
The Federal income tax is basically a tax on the net income of a cor-
poration. From the Federal tax base a number of additions and subtrac-
tions are made for purposes of the state tax. When a corporation is do-
ing business in states additional to Florida, the adjusted Federal
income is apportioned to Florida on the basis of the property owned,
payroll outlay and sales receipts of the corporation attributed to
Florida. The proportions of property, payroll and sales attributed
to Florida are weighted 25 percent, 25 percent and 50 percent, respec-
tively and the final fraction is the fraction of total corporate ad-
justed Federal income to which the Florida tax rate is applied. The
tax rate is a constant 5 percent of the apportioned adjusted income.
PREVIOUS WORK AND MATHEMATICAL FOUNDATIONS
The empirical work of interest on state tax measurement is con-
tained in seven articles, the first of which was published in 1952
[Groves and Khan (1952)]. Groves and Kahn and succeeding authors empha-
sized the elasticity concept as a useful measure on which to base ob-
servations about state tax criteria. They estimated the simple re-
log T = log a + e log Y
where T is the tax of interest, Y is the aggregate personal income in
the state unit, and e is the revenue--personal income elasticity co-
Davies (1962) concentrated on consumption taxes and refined some
of the measurements made by Groves and Kahn. Using a similar though not
identical regression equation Davies analysed the income elasticity co-
efficients for a number of states to find an overall estimate and to
determine if a different elasticity coefficient was evident during
economic upswings and downswings. The work also refined definitions
of stability to account for changes in the value of the dollar. An
elasticity coefficient of 1.408 was obtained for the Florida sales tax.
Wilford (1965) reviewed the criteria used by Groves and Kahn and
concluded that they were inadequate, over-emphasizing the stability of
a tax rather than characteristics of a tax which make it flexible or more
sensitive. Wilford extended the regression model to include the tax
rate for each tax, and applied the following equation to the Texas tax
log T = log a + e log Y + f log r
where r is the rate and f is the rate-revenue elasticity coefficient.
Wilford also estimated the same equation using per capital personal in-
come rather than aggregate personal income. The elasticities were
significantly larger than those estimated with aggregate personal in-
come, and led Wilford to the conclusion that the taxes were less stable
than had appeared when the aggregate measure of personal income was
used. Elasticities estimated by Wilford are summarized in Table 13.
The "revenue function":
R. = A.Yel r.e2,
1 1 1
where i is the particular tax, used by Wilford, was extended by Legler
and Shapiro (1968) to the form:
0 e e e3 e 5
R = A(y) (N) (rl)3 (r2) (p)
R = total tax revenue
A = constant
y = per capital income
el = income--revenue elasticity coefficient
N = state population
e2 = population--revenue elasticity coefficient
rI = income tax rate
e3 = rate--revenue elasticity coefficient for income tax
r2 = sales tax rate
e4 = the rate--revenue elasticity coefficient for sales tax
p = before tax sales price
Table 13. State taxes classified by sensitivity to changes in
aggregate and per capital state personal income and
rate changes, Texas.
Estimated Per Capita
Tax and Category Elasticity Income--Revenue
Low Income Elasticity
Alcoholic beverages 0.51 0.68
Cigarette 0.48 0.69
Medium Income Elasticity
Ad valorem property 1.01 1.61
Net motor fuels 1.11 1.40
Motor vehicle sales 1.15 1.73
Interest and penalties 1.37 1.99
High Income Elasticity
Inheritance 1.46 2.15
Corporation franchise 1.63 2.16
Insurance occupation 1.69 2.47
Motor vehicle licenses 1.75 2.96
Low Rate Elasticity
Medium Rate Elasticity
Motor vehicle licenses 0.54
Net motor fuel 0.67
High Rate Elasticity
Motor vehicle sales 0.74
Alcoholic beverages 0.86
Corporation Franchise 0.91
Source: Compiled from Wilford (1965).
e5 = price--revenue elasticity coefficient for sales tax.
The relative price of taxed versus untaxed goods was dropped from
the regression equation used to estimate the coefficients on the basis
of the assumption that the relative prices had not changed over the
sample period. The remaining coefficients were estimated statistically
by multiple regression using the equation
In R = a + e In y + e21n N + Z e In r + e In r2j + Ut
1 21 3i pj 4j 2j t
a = antilog of the constant A
i = the particular income tax
n = the number of income taxes
j = the sales related tax
m = number of sales related taxes
U = In C where E is the random error.
t t t
With this equation the 2 + n + m elasticity coefficients were estimated.
Results of this estimation are shown in Table 14.
Two recent papers [Friedlaender, et al. (1973) and Shapiro and
Legler (1973)] discuss problems in the model formation. Friedlaender,
et al. restrict the tax studies to sales tax. They used the estimation
log Rs = B log y + B? log N + B3 log rs + log u
R = revenue from sales tax
B = income--revenue elasticity coefficient
y = per capital income
B2 = population--revenue elasticity coefficient
Table 14. Elasticity coefficients of state tax structures, selected states.
Tax Rate Per
Motor Alcoholic Personal Corporate Capita
States Sales Fuels Beverages Tobacco Income Income Income Population
California -0.2080 0.3022 -0.0015 -0.1676 0.4663 2.1026 0.1893
Colorado 0.4492 0.1756 0.2977 0.0391 0.5622 1.2657
Connecticut 0.0494 0.5373 0.0164 0.1652 0.0202 1.0630 0.9092
Illinois -0.2452 0.0202 0.2784 0.3181 0.5981 3.5345
Iowa 0.4027 0.1372 0.0936 0.4153 -0.1732 0.4154 3.8565
Maryland 0.1047 -0.0991 -0.4284 0.4958 -0.0707 0.8121 1.4034
Michigan 0.4884 -0.0552 -0.0674 0.6891 2.0964
Ohio -0.371 0.1468 -- 0.8001 2.5109
Source: Compiled from Legler and Shapiro (1968).
N = population
B3 = rate--revenue elasticity coefficient
r = sales tax rate
u = random disturbance term.
The thrre elasticities for each state are given in Table 15.
Bahl and Shellhammer (1965) applied an input-output model approach
to an evaluation of business taxes in West Virginia. Their technique
parallels that of this project in some ways and will be referred to in
the course of the discussion where appropriate.
Basic Input-output Model
Battison and Jansma (1969) describe the basic input-output model
and an extension that will be used as the primary model in this project.
A diagram dividing the model into quadrants is given in Figure 1, and
the following mathematical summary of the basic model and its extension
describes the diagram in more detail.
The input-output model divides the economy by sector, with the
total output of each sector composed of output going into intermediate
demand and output going to autonomous final demand.
(1) X. = E x.. + Y., i = 1, ..., n
i j=l 1j I
where X. = total output of sector i
x. = output of sector i going to sector j to produce total
Y. = output of sector i going into autonomous "final
n = number of sectors.
Table 15. Rate, income and population elasticities for selected states.
Per Capita Sales Tax
Income Population Rate
State Elasticity Elasticity Elasticity
Alabama 0.67 2.60 0.99
Arizona 1.10 0.90 0.85
Arkansas 0.97 0.54 0.77
Colorado 0.84 0.76 0.90
Florida 0.90 1.73 1.18
Illinois 1.30 1.12 0.93
Iowa 0.91 0.91 1.01
Kansas 0.73 0.78 1.13
Maine 1.33 1.45 0.71
Maryland 0.78 1.87 0.79
Michigan 1.07 0.54 0.71
North Dakota 0.63 0.27 1.07
Rhode Island 1.18 0.97 0.80
Tennessee 0.96 1.90 0.82
Washington 0.69 1.45 1.07
Source: Compiled from Friedleander, Swanson and Due (1973).
Primary inputs and imports
Figure 1. The Input-Output Model Divided into Quadrants.
Direct input coefficients, ai., represent the cents worth of
purchase by sector i from sector j to produce a dollar's worth of total
output X.. They are calculated by
since x.. = a.. X. equation (1) can he written as
1j i] J,
(3) X. = a.E X. + Y; i = 1 .. n
i i 13ij 3 1
or, in matrix form,
(4) X AX = Y
where X = n x 1 matrix of total sector output, X.
A = n x n matrix of input coefficients, a..
Y = n x 1 matrix of final demand output, Y..
Writing equation (4) solve for X, equation (5) is obtained.
(5) X = (I-A)-1 Y
where (I-A)- is the inverted Leontief matrix. The elements of (I-A)
are denoted c.., and are the interdependence coefficients representing
the direct and indirect requirements from the i sector to support a
dollar's worth of output of the jth sector going to final demand.
Total output of sector i can be written in terms of the interdependence
coefficients and the final demand,
(6) X. c.. Y
1 .1 13
Quadrant III is the one of most interest for taxation analysis.
It traditionally consists of the primary inputs, including inventory de-
pletion, imports, net wages and salaries, rentals, net returns to
entrepreneurs, and government revenue. Payments to government are
treated as a purchase of the pritnary input government services.
Modification of Input-output Model for Tax Analysis
Modification of the basic model for tax analysis consists of the
extension of Quadrant III. A new row is added for each tax to be in-
dividually studied, to give a new set of m rows for m taxes. Let each
cell value be represented by tkj, where k = 1, ..., m, and j = 1, ...
n, m being the number of taxes.
Tax coefficients wkj are calculated by
(7) w = -ki, k = 1, ..., m; j = 1, ... n.
Total tax collection T. is the summation
(8) T E t E wk X..
k J kj j, kj
In matrix form
(9) T = W-X
where T = m x 1 vector of tax totals
W = m x n matrix of tax coefficients
X = n x 1 vector of sector total output.
Writing Tin terms of final demand rather than total output gives
(10) T = W- (I-A)1Y.
Letting Z be the m x n matrix representing W-(I-A.)- Equation (10) now
(11) T = Z-Y.
The Z values, zkj, can be called "tax intensity coefficients". They re
present the payment by sector j to the tax k for a dollar's worth of
output of sector j going to final demand. For any set of final demand
values, tax yields can be calculated by premultiplying the final demand
matrix by the Z matrix.
(12) zj = tki cij;
total tax collection of tax K is
A It A
(13) Tk = zkj Y3 = ij t.ki cij Y1 .
Determination of Tax Intensity Coefficients.
The simplicity of the extension of the basic model belies the
practical and conceptual difficulties of relating tax collection to tkj
values. The practical problems are the ever present sufficiency of
data problems. Even if data on taxes were sufficient to allocate the
taxes collected among the sectors, the legitimacy of doing so by simple
assignment of tkj values does not necessarily follow. The relation of
tax collection to final demand through the use of tax intensive co-
efficients is the key to the final result, and the development of wkj
values is the crucial step of achieving the goal, whether or not tkj
values are relevant. The following paragraphs establish three hypo-
thetical taxes to demonstrate three tax types.
Assume a tax Tk, k = 1, which is a tax based solely on the total
final demand in the input-output model. This tax would be a form of
sales tax in which a tax was collected as proportion, say rl, of all
goods and services leaving the interindustry sectors to the final demand
sectors. The amount of this tax would then be rl FD, where FD (instead
of the more traditional Y) is the total entry into final demand of these
goods and services. This tax could be divided into sectors by noting
that the total tax T1, would be a summation of tax collected from each
sector's contribution to final demand, as in Equation (14).
(14) T = Er FDi = r > FD.
1 l il 1 dr i
where FD. is the purchase from sector i to final demand.
The tax contributed by sector i final demand purchases is given as
(15) Rli = rl fd FD
where fd. is the proportion of final demand accounted for by purchases
from sector i,
(17) FD. = fd. FD.
The important point about this tax is the fact that Tli in Equa-
tion (14) is not the same as tkj in Equation (7) where i = j and k = 1,
and the calculation of the coefficient of interest, wkj, cannot proceed
as in Equation (7), and the calculation of a tax intensity coefficient
does not proceed as in Equations (10) and (11).
Since the direct relationship between final demand and the tax
collection is known, tax intensity coefficients are not necessary to de-
fine this relationship. The usefulness and power of the input-output
model is not needed for this case, and adds nothing to Equation (14)
except the definition of sectors and final demand. For consistency,
and because the formulation of a tax such as TI will never be found,
w.. terms can be developed.
Rewriting Equation (10) for one tax, it becomes
(18) T = W1.(I-A)- FD
where W = 1 x n vector of w coefficients for tax 1.
By definition of T1 in Equation (14),
(19) T1 = rl.FD
where r = 1 x n vector of rl values
FD = n x 1 vector of FD values.
(20) W *(I-A)-*FD = r I FD.
Postmultiplying Equation (20) by FD- and (I-A),
(21) W = rl.(I-A).
With this formulation of W1, the W matrix can be used directly to find
the tax intensity coefficients in Equation (11). This will, of course,
result in a tax intensity coefficient identical to the rate of tax.
Several variations in a tax of type T1 can be fitted into this
series of equations. Equation (14) will not apply if the tax rate, rl,
varies depending on which sector the final demand purchases are made
from,nor if some portions of the goods and services purchased are not
taxed. In these two cases the only modification that is needed is a
change in vector rl in Equation (21). Instead of being a vector of
identical values, the rlj values will depend on either varying rates
among sector purchases or with the proportion of final demand goods and
Another variation is important. The existing input-output analysis
of Florida's economy [Farler and Tyner (1973)] divides final demand into
personal consumption, government and net exports. Given that the sales
tax T1 applies only to purchases for personal consumption, modification
of Equation (21) will follow the same course as the rate variations, re-
sulting in a modified rl vector depending on the proportion of final
demand in each sector going to personal consumption.
T was a tax based on the purchases from the intermediate sectors.
Suppose now that T2 is a tax based solely on the purchases by the
intermediate sectors from a primary input row, say, net payments to
labor. Further assume that this tax is in some fixed proportion r2 to
the payments to the labor sector L. The total tax is then calculated as
(22) T2 r2L = r2 L.
where L. is the payment to labor by sector j. L. will be related to
final demand by the equation
(23) L = J1(I-A) --FD
where ( is a 1 x n vector of payments to labor coefficients calculated by
(24) ^j .
Since r2 is a proportion of Lj it is also
the tax T2 is related to final demand by
a proportion of I., and
(25) T2 = r2.1-(I-A) -FD
where r2 is a constant. The w2j values are in this case r2 Ij, and with
the w2j values the tax intensity coefficients can be calculated by
Equations (10) and (11).
Variations of the tax T2 can also be incorporated into this series
of equations. For example, if payments to some kinds of sector related
labor were not taxed or taxed at a different rate, the w values would
be calculated by r2j Aj. This vector of w2j values could be used direct-
ly to arrive at the tax intensity coefficients.
Now turning to the simplest case, suppose T3 is a tax imposed at a
rate of r3 on all purchases from the primary sectors and all intermediate
purchases. The tax coefficients would be simply the wkj values calcula-
ted by Equation (7) and the derivation of the tax intensity coefficients
zkj would follow directly in Equations (10) and (11). The wkj values
would be the r3j values.
Needless to say, seldom if ever does a particular tax follow
precisely the patterns of these tax types discussed. Each tax has
characteristics of each type and the combination will depend on many
factors. Classification of tax by type, as above, or category, as
follows, is for clarification of concepts, and is in no way a rigid
pattern into which everything must be forced.
Fitting the Taxes into the Model
It has been assumed that the tax rate is a linear proportion of
the taxed entity, whether that entity is a final demand sector as with
TV, a primary input sector as with T2, or total inputs to a sector as
with T3. Maintaining this assumption, there are two attributes about a
tax that will determine the assignment of wkj values. The first is the
allocation of a tax among the sectors, and the second is the assignment
of types of tax collection among the type of tax outlined by T1, T2 and
T tax types to assure the proper relationship of tax collection and
final demand. These two attributes are closely related and determination
of tax intensity coefficients requires consideration of both.
Allocation of taxes among sectors is, like the determination of the
tax intensity coefficients, not a simple task, either practically or
conceptually. Conceptually, taxes might be divided into four categories:
product related, sector related, non-industrial transaction, and non-
industrial event. Within each of these categories the tax intensity co-
efficient calculation might be of type 1, type 2 or type 3. Each
statutory tax might have characteristics of one or more of the tax
categories and one or more of the tax types.
A product related tax is one that is associated with a good or
service as it is bought and sold within the input-output model framework.
Examples would be a sales tax that is attached to the sale of any
specified group of products, and a beverage tax attached to a specific
product. This does not necessarily mean that the sector allocation
problem is solved by assigning a tax to the sector or sectors associ-
ated with the production of that product, and an example is given in
the paragraph describing the non-industrial transaction category.
A sector related tax is associated with a sector though not pri-
marily through a flow of goods associated with that tax. Examples
might be a corporate charter tax, where the tax is based on the act of
incorporation, not a good or service provided for by that corporation.
Allocation of the tax would be by number of incorporations, and capital
stock values for each sector. An income tax would fall into this cate-
gory, with the associated payments to households by sector.
The non-industrial transaction category is a group of taxes relat-
ed to the interindustry system to the extent that the tax can be relat-
ed to the system only indirectly through some interindustry criteria.
An example of this tax might be the wagering tax. Wagering might best
be related to household income, and if there is no reason to believe
that a significant difference exists among sector-related income and
income-related wagering, the allocation of the tax would be in some pro-
portion to the payments to households by each sector. The documentary
stamp tax might also fall into this category. The consumption of
alcoholic beverages might also be related to payments to households with
a resulting proportional tax allocation among sectors.
The non-industrial event is similar to the non-industrial transac-
tion category but the taxable event is even less related to inter-
industry type transactions. The documentary stamp tax may fall in this
group, but the best example is the inheritance tax. The inheritance tax
is not associated with the model by product, sector or transaction but
may be allocated among sectors based on relationships in the model. For
example, inheritance tax may be related to wealth, and wealth may be
further related to income to households by sectors.
Selection of tax type and category to apply to each tax cannot be
based only on statutory characteristics of the tax and economic
characteristics of the input-output model, if a conceptually correct
treatment of a tax is impossible because of data problems, for example,
it might be necessary to place it in a different category to make
allocation possible. Conversely, if data allow a detailed allocation
of tax among sectors, judgement may require that a different allocation
be used to conform to the purposes of the project. A possible classifi-
cation scheme for taxes of interest in this study plus the income tax
is shown in Table 16.
Table 16. Selected Taxes Tentatively Classified by Type and Category.
Tax Product Sector Transaction Event
Wagering Type 1 Type 2
Beverage Type 1 Type 1,2
Documentary Stamp Type 1,3 Type 2
Inheritance Type 2
Corporate Capital Stock Type 3
Sales and Use Type 1,3
Income Type 2
The Input-Output Model
The 1967 Model
The basic input-output model used for modification and manipula-
tion is that developed by Farler and Tyner (1973) at the University of
Florida. It is a 44 sector Florida State model which used estimated
1966 national coefficients to formulate a 1967 input-output model. The
sectors into which the economy is divided are shown in Table 17. The
inverted Leontief matrix (I-A)-1 (the matrix of direct and indirect co-
efficients for the open model) used by Farler and Tyner is used direct-
ly as calculated in the published material. The estimates of final de-
mand sectors (consisting of personal consumption, exports and government
purchases) were also used directly as given for the 1967 model and are
shown in Table A-1. Pre-multiplication of the final demand totals
matrix by the (I-A)- matrix yields the total output vector shown in
Table A-2. This basic model is used as a starting point for modifica-
tion and is also used as the "base" for projection for 1972 figures as
described in more detail below.
There are two reasons for not estimating a new set of technical co-
efficients for 1972. State models are constructed using national coef-
ficient estimates as a starting point for modification and aggregation.
The most recent national coefficients available are the 1968 estimates
published in the first part of 1974. Since technical coefficients are
normally acceptable for about 10 years, the two-year update on national
Table 17. Economy Sectors in the Basic Input-Output Model.
1 Livestock and livestock products
2 Citrus products
3 Vegetable products
4 Other Agricultural products
5 Forestry and fisheries
6 Agr. forestry, and fisheries services
7 Metal ores, crude petroleum and natural gas
8 Stone and clay mining and quarrying
9 Chemical fertilizer and fertilizer material mining
10 Maintenance and repair construction
11 Ordnance and accessories
12 Canned, cured, frozen veg., fruits, and seafoods
13 Other food and kindred products
14 Tobacco manufactures
15 Textile mill products
17 Lumber and wood products
18 Furniture and fixtures
19 Paper and allied products
20 Printing and publishing
21 Chemical and allied products
22 Petroleum products
23 Rubber and plastic products
24 Leather and leather products
25 Stone, clay, and glass products
26 Primary metal industry
27 Fabricated metal products
28 Machinery (except electrical)
29 Electrical machinery
30 Transportation equipment
31 Instruments and related products
32 Miscellaneous manufacturing
33 Wholesale and retail trade
34 Transportation and warehousing
36 Electric, gas, water, sanitary services
37 Finance and insurance
38 Real estate and rental
39 Hotel and lodging
40 Personal and repair services
41 Business services
42 Amusements, recreation and motion pictures
43 Medical, educ., legal, non-profit organizations
44 Other economic activity and new construction
estimates does not justify an attempted re-estimation of state co-
efficients. A second drawback in re-estimating coefficients is that
the results of the 1972 censuses of manufactures and businesses are not
yet available and will not be for some time. The information from the
census results is an important part of state model formulation, and re-
sults obtained in their absence would be questionable improvements over
estimates using census data from 1967.
On the other hand considerable advantage exists in using identical
coefficients for both 1967 and 1972. Changes in the tax structure can
be related directly to a stable model, with final demand changes the
only other variable to be considered. In addition, changes in final de-
mand and the accompanying changes in primary inputs, intermediate ex-
changes and total sector output can be compared directly with the tax
changes for corresponding final demand changes, which provides a
possible conditional evaluation of tax allocation to sectors and struc-
tural tax allocation.
The 1967 model is the base period both for the tax coefficient
estimations and for the projection to 1972. Several ways exist for
going from the 1967 model to the 1972 model. Given the use of the same
coefficients and the emphasis on relating tax revenue generation to
final demand changes, the most appealing method is the projection of
the 1967 final demand figures to 1972, and the re-calculation of total
output figures for sectors by pre-multiplication with the (I-A)- matrix.
This method is used in three ways.
Projection to 1972
Projection I uses the final demand figures from 1967 as base figures
and projects each of these three final demand sectors individually.
Final demand for each of the 44 intermediate sectors is kept in the
same proportion as the 1967 values as for the final demand sectors.
Projection of personal consumption is based on two factors--
aggregate personal income and the overall U.S. aggregate personal con-
sumption and expenditure pattern for the 1967 to 1972 interval. The
aggregate personal consumption to aggregate personal income ratios for
Florida were taken as the same for the U.S., and the increase in
Florida aggregate personal consumption was calculated from the follow-
where subscripts indicate year and
FC = Florida aggregate personal consumption
FI = Florida aggregate personal income
USC = U.S. aggregate personal consumption
USI = U.S. aggregate personal income.
Solving for the ratio FC72/FC67 gives a ratio of 1.5635. Multiplying
this ratio by personal consumption from each sector in 1967 yields the
new vector of personal consumption for 1972 as shown in Table A-3.
The export sector is projected in the same proportion as the 1967
to 1972 total personal consumption increase in the U.S. as a whole.
The ratio used is:
where subscripts indicate years and
EX = Florida exports
USC = U.S. aggregate personal consumption.
This ratio is 1.4760 and the 1972 vector of net exports is shown in
Table A-4. The government sector is projected in the ratio of 1967
and 1972 aggregate state and local government consumption for the
U.S. as a whole using the formula:
F(28) C72 USSLG72
where subscripts indicate year and
FG = Florida government expenditures
USSLG = State and local government expenditures for the U.S.
as a whole.
This ratio is 1.7578 and the 1972 vector of government expenditures is
shown in Table A-5.
The ratio of 1972 to 1967 total final demand using these projec-
tions is 1.5443. Total final demand by sector for 1972 from projection
I is shown in Table A-6, along with the subtotal of personal consump-
tion and net export, to be used later.
Projection II also uses the final demand figures from 1967 as base
figures and projects each of the three final demand sectors individual-
ly. Final demand for the 44 sectors for 1972 are not proportional to
those of 1967. The projection total is made to total that of projec-
tion I by proportional reduction of sector figures obtained from
initial projection calculation. The initial calculations are based on
the method used by Eddleman(1973) to make more long range projections.
In his projection Eddleman needed to use projections of population
and per capital income to his projection date. However values of per-
sonal per capital income and population for 1972 are now available for
direct use in the 1972 projection.
Projection of personal consumption is based on three factors rather
than two--personal per capital income, population and expenditure
elasticity estimates for sector products. The formula used is as
E72 [ c( -1) + 1] E
i72 PI67 167 N67
where E72 = Household purchases from sector i in 1972
ci = Expenditure elasticity for sector i
PI72 = 1972 per capital personal income
PI67 = 1967 per capital personal income
E 67 Household purchases from sector i in 1967
N7 = 1972 population
N67 = 1967 population.
Expenditure elasticities for the purchase of the sector products
are taken from Eddleman (1973). Those for the primary agriculture
sectors (1-4), agricultural processing sectors (12, 13), non-agricultur-
al manufacturing sectors (21, 22, 26-30, 32), trade sector (33),
utilities sector (36) and services sectors (40, 41) are taken directly
or derived from approximations for the U.S. from numerous data sources.
The other sectors are assigned expenditure elasticies of unity.
Elasticity figures used are shown in TableA-7. Projected personal con-
sumption for 1972 from the initial unadjusted calculations are shown
in Table A-8.
Estimation of export figures follows the same formula with PI now
being per capital personal income in the U.S. excluding Florida and N
the U.S. population excluding Florida. E is of course export rather
than household expenditures. The projected exports for 1972 from the
initial unadjusted calculations are shown in Table A-9.
The projections of government expenditures are in direct propor-
tion to the change in total Florida personal income, with the ratio
being 1.82. Expenditure elasticities are not a part of the calcula-
tion. Government sector projections, unadjusted, are shown in Table
The unadjusted total final demand for all sectors is
$27,880,758,000 for the initial projection II. The corresponding total
for projection I is $24,920,668,000. To bring the projection II total
into conformity with the projection I total, all sector values for
projection IT are reduced by the ratio 0.8938. The adjusted projection
II figures, hereinafter referred to only as projection II figures, are
given in Table A-1l. The total final demand by sector for 1972 from
projection II is shown in Table A-12, along with the subtotal of per-
sonal consumption and net export to be used later.
Two facts suggest that projections I and IT might be under-
estimating final demand. First, any increase in tourist income will
not be reflected in projections I and II since the projections are
based only on income-consumption ratios. The large increase in pur-
chases by tourists from 1967 to 1972 is thus neglected. The second
reason for suspecting an underprojection of final demand for 1972 is
the corresponding underestimation of sales and use tax revenue based
largely on retail sales.
In projection III, personal consumption and net exports are in-
creased in the same proportion as total taxable retail sales for Florida
increased from 1967 to 1972. The ratio of retail sales in 1972 to re-
tail sales in 1967 is 2.1572. The government sector is projected as in
projection I. The final demand figures in projection III are shown in
The utilization of three different projection methods achieves
several useful results. First, the projections give two different
proportions for each of the 44 sectors. One, from projections I and III,
maintains the proportions of the 1967 base vector, while the other,pro-
jection II, reproportions the contribution of purchases from each sector.
This is true for both personal consumption and net exports, though both
methods retain 1969 proportions for the government expenditure sector.
Second, the three projection methods give three differing results
in the percentages in total final demand accounted for by personal con-
sumption,net export and government. In 1967 personal consumption, net
export and government contributed 55.86 percent, 37.23 percent and 6.91
percent respectively. Projection I gives corresponding figures of
56.56 percent, 35.58 percent, and 7.86 percent, projection II results
in figures of 60.06 percent, 32.12 percent and 7.82 percent, and pro-
jection III results in figures of 56.59 percent, 37.71 percent and
The adjustment downward by about 10 percent of the initial pro-
jection II figures gives an equal total final demand for both projections
I and II. This permits some isolation of the effects of the differences
pointed out in the two previous paragraphs while conforming the gross
differences in the two projection methods.
It should be noted that these methods of projection are not in-
tended to be the best possible estimates of 1972 final demand. Esti-
mation of final demand figures for 1972 might well be more accurate us-
ing more specific and relevant 1972 data than that used in the pro-
jection. Three reasons justify the projection methods. First is its
ease of use. Second, three different results give a variety of figures
which will shed more light on what the tax calculation results might
mean. Third, and most important, because of the data lag input-output
models are consistently at least five years in the making. The useful-
ness of tax analysis utilizing such vintage information is somewhat
limited. Using final demand projection as an estimation technique al-
lows more current analysis, and even a possible projection of future
dates, though it is recognized that the original interdependency
matrix is still retained as a crucial part of the new structure.
Total Output Calculations
Total output figures for the 44 sectors on the 1967 model were
given in Table A-2. Pre-multiplication of the projected final demand
figures for projections I, II, III by the (I-A)- matrix yields the
total sector output for each sector for 1972. Projected total gross
outputs from the three projection methods for total final demand are
given in Table A-14.
Tax Extensions of the Model
Summary of Tax Data Used
This summary of tax data used brings together the tax figures of
specific interest in the modification of the input-output model. Taxes
of prime interest in the years 1967 and 1972 are shown in Table B-1,
sales and use taxes are excluded. Corporate capital stock tax changes
in conjunction with the new corporate income tax occurred between
fiscal years 1967-68 and 1972-73 and resulted in the elimination of
corporate capital stock tax as a significant revenue generating tax.
The corporate income tax was non-existant in fiscal 1967-68 and appears
only in the 1972-73 column.
The sales and use taxes pose a particular problem. Between
fiscal 1967-68 and 1972-73 the sales tax rate changed from 3 percent
to 4 percent. To make the sales and use tax analysis reasonable, two
methods were used to adjust the reported 1967-68 amount to an imaginary
value reflecting the higher rate. The first method uses an elasticity
concept and the second uses an intercept shifter concept.
The rate elasticity is estimated as an exponent in an equation
of the Cobb-Douglas form with aggregate Florida personal income and
Florida population as the independent variables in addition to tax rate.
The equation is:
(29) Tax = a ybl pb2 rb3
where Tax = Sales and use tax revenue
a = constant
Y = Florida aggregate personal income
b1 = income-revenue elasticity
P = Florida population
b2 = population-revenue elasticity
r = tax rate
b3 = rate-revenue elasticity.
The equation used for estimation of the elasticities is the log linear
(30) In Tax = In a + b InY + b2 InP + b3 Inr.
Using a 20-year time series data set the estimated rate elasticity is
1.14374, with an R2 for the linear equation of 0.99. More information
on regression analysis of sales and use tax data is contained in
Using the elasticity of 1.14374, the change from a 3 percent to
4 percent tax rate increases the 1967-68 sales and use tax value to
$494,076,632. This figure is designated hereinafter as sales and
Calculation of a 1967-68 sales and use tax hypothetical value
using an intercept shifter involves estimation of a dummy variable as
shown in the following linear equation:
(31) Tax = A + blY + b2P + b3rd'
where b3 is the intercept shifter and rd is the dummy variable assigned
the value 0 in years when the 3 percent rate was effective and 1 in
years when the 4 percent rate was effective. Using a 20-year time
series data set the value b3 is 158,499,158 with an equation R of 0.99.
The shift in intercept gives a hypothetical value of $516,181,844
for the 1967-68 sales and use tax revenue figure, hereinafter designated
sales and use(S).
Three values are used for most calculations dealing with the sales
and use tax--the original value of $357,682,686, the value $494,076,632
obtained from the application of the elasticity measure and $516,181,844
obtained from the use of the intercept shifter.
Summary of Three Tax Types
In Chapter III the considerations of fitting taxes into the basic
model were discussed, and a general conclusion reached was that taxes
might logically be divided into three types--those that relate to
total sector output, those that relate to some primary input sector
and those that relate to final consumption of sector output. As an
introduction to the following sections these concepts are summarized.
Taxes that relate to total sector output are allocated among the
sectors in the same proportion of total tax as sector output is of the
total output for the 44 sectors. By dividing the tax thus allocated
to each sector by the total output of the corresponding sector the
tax coefficients described as w3 and r3 in Chapter III are obtained.
Taxes relating to some sector of quadrant III are allocated among the
sectors in the same proportion as the proportion of each sector value
to the quadrant III total for that row. The tax coefficient, desig-
nated w2 in Chapter III, is then calculated through division by the
total sector output. A different approach must be taken toward the tax
based on final demand. The total tax is allocated among the sectors
of final demand in the proportion that each sector represents of total
final demand. There are further modifications for the tax based on
less than all columns in quadrant I comprising final demand. As ex-
plained in Chapter III for rl, tax coefficients are calculated by total
sector output division. A "tax fraction" is calculated by dividing
sector final demand into the tax for each sector, whether total or
partial final demand figures are used.
A special note should be made concerning the use of the "r" terms
(rl, r2, r3). In Chapter ITT they are called "rates". It is empha-
sized at this point, with the introduction of real taxes, that r is a
"derived" rate only and should not be confused with the statutorially
defined tax rate. The r values are derived from calculations of
actual tax collection ex post facto and are rates only in the sense of
being the quotient of a numerator and a denominator.
Proportional Assignment of Taxes to Sectors
In keeping with the projection methods used in the present study,
1967 is the base model. Proportion calculations are made on the basis
of the values found in the base 1967 model and on tax data for the 1967-
68 fiscal year. Exceptions to this procedure are the use of assumed
sales and use tax values taking into account the changed tax rate.
The first method of assigning taxes is on the basis of the dis-
tribution of total gross output among Phe 44 economy sectors. The pro-
portion of total gross output assigned to each sector is calculated
with the following equation:
(32) p j = 1,44
where p. is the proportion, X. is the sector total output and Z X. is
the total gross output. Multiplication of p. by tht total tax collec-
tion for each tax gives the tax allocation sought. The p. values
derived from the 1967 base model are given in Table B-2 and the tax
allocation for all taxes except corporate income tax is presented in
Tax allocation to sectors on the basis of a quadrant III row is
accomplished by the determination of the proportion p. from
(33) pj = j =1,44
where p. is the proportion, r. is the purchase of row output used in
producing sector j output and L r. is the total use of row output by
As an example of a primary input row allocation, the total value
added row from the 1967 base model is used. This row is one of two
rows that comprise total quadrant III inputs; the other row is imports.
The value added row does not isolate income from primary resource
inputs and thus is defective as a measure of taxes based on income.
It is used here, however, as an example of quadrant III proportioning.
The proportion values resulting from total value added allocation are
presented in Table B-4 and the corresponding tax allocation figures
are shown in Table B-5.
When taxes are based on final demand it is necessary to allocate
taxes among final demand sectors. This allocation is made in two
ways. The first method is an allocation based on total final demand.
The second method, assuming that final demand consumed by the govern-
ment sector is not taxed, bases the taxation on only personal con-
sumption and net exports, the two remaining contributors to final
demand in the model. Net exports may not be taxed as domestically
consumed products in the type 1 tax situation, but tax types 2 and 3
should be fully captured.
Proportions for total final demand are calculated by
(34) Pi = D i = 1,44
where FD. is the total sector final demand for sector i and L FD. is
1 i 1
the total final demand for the 44 sectors. Proportions obtained from
Equation (34) are given in Table B-6 and the tax allocation from multi-
plication by 1967 tax totals are shown in Table B-7.
Excluding the government sector and allocating taxes in proportion
to personal consumption and net export alone yields proportions shown
in Table B-8 and tax allocation shown in Table B-9.
Derivation of Tax Coefficients
Tax coefficients are obtained by division of the tax k allocated
to each sector j by the total sector output for that vector, X., as
shown in Equation (7) from Chapter III, repeated here as Equation (35)
(35) Wkj "
Tax coefficients based on total sector output allocation are con-
stant among sectors for each tax, that is, wkj = wk for all j and
k. These coefficients are shown in Table C-l.
Tax coefficients using allocation other than total sector output
proportioning are not identical, though calculation is also done
according to Equation (35). Calculated tax coefficients for tax
allocation based on total final demand are shown in Table C-2.
Final Demand Fractions
As suggested before, calculation of tax coefficients for taxes is
based on the final consumption of goods and services. Division
is carried out with sector final demand figures rather than total
sector output figures as divisors. The separation of final demand into
the total and partial parts dictates two sets of calculations. The
first involves the division of tax allocation on the basis of total
final demand by total final demand for each sector. The partial method
suggests division of tax figures allocated on the non-government final
demand basis by the non-government final demand for each sector, with
the results for both sets of calculations shown in Table C-3. The
method used for partial final demand figures meets the requirements of
Equation (17) for the calculation of fractions.
Projection to 1972
Having calculated tax coefficients and tax fractions using the basic
input-output model and 1967 tax data, and having also calculated input-
output information for 1972 using various projection methods it is now
possible to apply the tax coefficients and tax fractions to the 1972
results to obtain 1972 tax information. Three separate sets of results
will be derived from the three different proportioning calculations--
total gross output proportioning, total value added proportioning, and
final demand proportioning. For each proportioning method the three
different projection methods will yield three sets of results.
Total gross output proportioning. Tax revenue generation for 1972
using the methods developed thus far is calculated by multiplying the
total gross output for 1972 by the tax coefficients as indicated in
Equation (8),repeated here as Equation (36)
(36) Tk = tkj = kj Xj, n = 44.
Each Tk is a summation of tkj values for each sector. The result of
this equation for each projection for each tax is shown in Tables D-1
Total value added proportioning. Tax estimation for 1972 is
achieved through an operation similar to that shown in Equation (36)
though of course the tax coefficients for each sector are different.
Results are shown in Tables D-4 through D-6.
Total final demand fractions. Equation (14) reproduced as Equation
(37) shows the method to be used to determine 1972 taxes from tax
fractions and final demand figures for 1972 previously calculated.
(37) TI = E rI FD = r Z FDI, n = 44.
The tax estimates for 1972 using final demand tax fractions result from
a direct multiplication of each previously determined final demand tax
fraction by the new total final demand. Results are shown in Tables
D-7 through D-9.
Partial final demand fractions. Tax estimation for 1972 from
partial final demand figures are similar to the total calculations
except that only the personal consumption and export components of 1972
final demand projections are used, and the partial tax fractions are
used. Results are shown in Tables D-10 through D-12.
Tax Intensity Coefficients
The final set of basic calculations in this chapter deals with
"tax intensity coefficients." In the work of Battison and Jansma (1969),
from which the analogy of tax intensity coefficients is derived, the
term factor intensity coefficients was used in relation to primary
inputs in quadrant III only, and were results of multiplication of the
factor coefficients and the (I-A)- matrix as shown in Equation (10)
and (11) of Chapter III, reproduced here as Equation (38) and (39).
(38) T = W.(I-A)-1Y.
(39) T = Z-Y.
A simple definition can be formulated--tax intensity coefficients are
those numbers which, when multiplied by any given set of final demand
figures, yield tax revenue generation values for a given tax.
Total Gross Output Proportioning
Equation (38) shows that the calculation of the tax intensity co-
efficient for a total gross output proportioned tax is the premulti-
plication of the inverted Leontief matrix by the row of tax coefficients.
In this case the tax coefficients are identical among sectors for each
tax. The result of the multiplication is a one row vector for each
tax, given in Table E-l.
Total Value Added Proportioning
Total value added proportioning follows a similar calculation, but
of course the tax coefficients are different for each sector. The
results of this calculation are shown in Table E-2.
Final Demand Proportioning
The vector which, when post-multiplied by final demand, yields
tax revenue generation in the case of final demand proportioning is
simply the tax fraction vector. This result is shown by Equation (39)
where Z, the tax intensity coefficient, is the tax fraction vector
transposed for matrix multiplication purposes. The tax fractions
have been previously shown in Table C-3 and are vectors of a constant
in a 1 x 44 vector.
Projection to 1972
Multiplication of the tax intensity coefficient vector by the
final demand vector yields a vector of tax values, each element of
which indicates the tax revenue generation resulting from the final
demand for the product of each sector. Sales and use tax is treated
differently in later analyses and projections are not made for that
tax. For the other taxes, tax revenue generation resulting from
sector final demand for 1967 is shown in Tables G-1 and G-2. For
projections to 1972 the values are shown in Tables G-3 through G-8.
TAX EXTENSIONS OF THE MODEL
Fitting Taxes into the Model
There are a number of conceptual methods of using the informa-
tion contained in an input-output model to obtain information about tax
revenue generation. The choice of methods is important to the core
purposes of using the input-output model for tax analysis, and the
choice deserves further discussion. This section looks first at some
general principles, then at each specific tax to determine what the
nature of the tax suggests for method choice. Finally, a composite
model is presented.
The Two Approaches
There are two basic approaches to tax analysis through input-
output models. The first is a restrictive approach and the second is
a more general approach. The appropriateness of each method for a given
tax will depend on the nature of the tax, so that the final tax analy-
sis contains elements of both the restrictive and general approaches.
The restrictive approach treats the tax revenue as proceeding from
a single sector or from very few sectors. This may be exemplified by a
beverage tax which is of course a tax on a limited number of specified
items. If the taxed beverages are products of a single sector, it
is possible to view the beverage tax as a percentage of a single
sector's sales. All information about the tax available from the
input-output analysis is contained in the behavior of a single sector
as it responds to economic changes reflected by changes in final demand.
Further, if sales are taxed only at final demand levels all information
available would be contained in one or two cells of the entire model--
those which show final demand purchases from the beverage sector.
The generalized approach allows for somewhat more flexibility and
is resorted to in cases where the restrictive approach is inappropriate
or where its use yields more information. The generalized approach
treats taxes as related in some way to figures found in various input-
output model cells which show the response of more than a single sector
to a change in final demand. To continue the example with the beverage
tax, it can be hypothesized that taxed beverage consumption is some
proportion of total personal consumption. With this formulation of the
tax the relevant set of cell figures is the entire personal consumption
vector. It can alternatively be hypothesized that the consumption of
taxed beverages is dependent upon personal income, so the useful portion
of the input-output model is found in quadrant III.
The choice of using a restrictive or general approach is more a
result of other decisions about how to treat an individual tax than a
preliminary choice. Therefore it is more convenient to look closely
at each tax and decide first how it must be analyzed, then observe
whether a restrictive or general approach is most appropriate.
Tax Treatment Consideration
The mathematical formulationsof Chapter III suggest that there
are at least two basic considerations involved in analysing a tax with
the input-output model. One of these is the placement of the tax
within the model to assure the proper relationship of tax collection
and final demand. This decision is resolved by choosing a tax type
T1 T2 or T3 which best fits the individual tax. Allocation of taxes
among the sectors is also dependent on the character of each individual
tax study. The assignment of taxes to categories was suggested in
Chapter III, the categories being product related, sector related,
non-industrial transaction and non-industrial event.
The discussion of each tax in the following sections will consider
each of the choices that must be made about tax treatment. Given the
data available, a decision will be made as to how best to treat the
tax in the input-output framework being used. A short discussion will
then be given of some of the assumptions and problems associated with
the resulting tax analysis figures.
Each tax is considered by taking the tax description given in
Chapter II, applying the concepts from Chapter III, and drawing from
the appropriate empirical results found in Chapter IV.
The wagering tax is a combination of taxes on a contributed pool.
It depends directly on the amount of contribution to the betting pool.
A restrictive approach to wagering tax analysis would suggest a
separate sector whose output is the service of providing wagering
facilities. The tax would be related to this single sector and would
be determined by establishing a relationship between purchases of the
service and tax collection. The single cell of the input-output model
indicating final demand purchases of the service would give all the
information available from the input-output model, but would be of
little interest from an analytical point of view. A determination of
the whole vector of final demand would be irrelevant with the exception
of that cell.
A general approach is more appealing. When the general approach is
used, the activity of wagering, on which the tax is based, is not a pro-
duct. Neither is it related specifically to one sector. It is, how-
ever, an economic transaction rather than a fortuitous or non-economic
occurrence. It would most logically be classified as a non-industrial
When treated generally as a non-industrial transaction a connection
must be found with the input-output model. Rather than consider wager-
ing as a proportion of all final demand purchases or as a proportion of
total gross output, it would seem that wagering might coincide most
closely with personal income. (This is an acceptable assumption only in
its restricted context.) This suggests treatment as a type 2 tax based
on the quadrant III row giving payments to individuals through wages
and salaries, with total value added proportioning material from
Chapter IV providing the appropriate information.
The initial 1967 allocation to sectors is taken from Table B-5
and the tax coefficients calculated from this allocation are from Table
C-2. The tax projection to 1972 using final demand projections I, II
and III are from Tables D-4, D-5 and D-6 respectively. The tax in-
tensity coefficients are taken from Table E-2.
The quadrant III values upon which the wagering tax is based are
of course determined by the final demand. The tax collection is de-
termined by the wagering activity of individuals. There is a gap
between the quadrant III figures (the proxy for income) and the wager-
ing activity. This gap is not explained by the input-output model.
Three major factors must be recognized as making the step from
income paid by each sector and wagering expenditures. The first is a
total wage figure (one for each sector) as contrasted to wages received
by the individual--wages received by the individual being the basis for
individual wagering expenditures. It is possible, for example, that
the same total wage figures for two sectors would be paid to more
relatively low income individuals than the other sector, or that the
recipients of wages and salaries from one sector would have greater
wagering expenditures than the other.
A second factor between the quadrant III payments to individuals
and wagering expenditures are the income-wagering expenditure relation-
The third factor is one which is not reflected in the quadrant III
row but may well have an important part in determining wagering ex-
penditures. These are the wagering expenditures by individuals whose
income is not reflected in the quadrant III row, the most outstanding
example of this being the wagering expenditure of tourists.
With knowledge of income characteristics of each cell in the
payments to individuals row, the income-wagering relationship, and the pre-
sence and characteristics of the wagering expenditures from income
not reflected in the input-output model, a clear relationship could be
developed between final demand and wagering tax revenue generation.
A converse view of the items filling the gap between the in-
formation provided by the basic input-output model and wagering tax
revenue is obtained by accepting the results of the empirical study
and listing these items as assumptions necessary to make the results
valid. Knowledge of final demand will yield valid wagering tax revenue
results through the use of the calculated tax intensity coefficients
as calculated in Chapter IV, assuming (1) there are no latent differ-
encesamong sectors in the individual income and expenditure character-
istics which make tax allocation among sectors on the basis of total
personal income invalid, (2) the personal income-wagering expenditure
relationship is such that wagering tax revenue is a linear function
of the total income figures for each sector and the coefficient is
that of the calculated tax coefficient, and (3) there are no
influences on wagering tax revenue generation from sources not
captured in the basic input-output model formulation. It should be
noted that assumption (2) is an assumption concerning the combined
effects of the personal income-wagering expenditure relationship and
the wagering expenditure-tax revenue generation relationship. The
three assumptions are of course additional to the entire set of
assumptions underlying both the basic input-output model and the use
of total value added instead of specific payments to individuals, both
of which are discussed later in this chapter.
The beverage tax is a tax on the sale of enumerated kinds of
alcoholic beverages sold in Florida, with varying rates depending
on the type of beverage and its source. Its treatment is similar
to that of the wagering tax except it is of course a good rather than
The restrictive approach would follow the same pattern as with
the wagering tax, including the confinement of beverage purchases to
an isolated sector representing taxable beverage distribution and sale,
and the application of tax information to the basic input-output model
at single or minimum cell points. Two additional problems arise when
using this approach for the alcoholic beverage tax. One is the nature
of the beverage industry, with a large portion of beverages imported
from other states and from abroad. The only entrance into the inter-
industry model would be a large import cell, some smaller interindus-
try purchases cells, and the sales to final demand. Second is the
fact that there is a rather detailed tax rate breakdown depending
on type of beverage and on whether or not the beverage is manufactured
with Florida products. A restrictive approach would suggest, though
not necessarily require, a separate final demand cell for each
beverage carrying a different tax rate.
A general approach would suggest the use of some information
available in the input-output model which would reflect the more com-
plex relationships within the economy, with the concommitant
difficulty of removing the beverage tax revenue generation from a
direct relationship with the input-output cell values. The most de-
sirable choice would be that of relating beverage consumption to per-
sonal income by tying beverage tax revenue generation to the quadrant
III row indicating payments in wages and salaries. This places the
beverage tax in the category of a non-industrial transaction instead
of a product related tax, and also labels it a type 2 tax.
The initial 1967 allocation to sectors is taken from Table B-5
and the tax coefficients calculated from the allocation are from Table
C-2. The tax projection to 1972,using final demand projections I, II
and III, are from TablesD-4, D-5 and D-6,respectively. Tax intensity
coefficients are taken from Table E-2.
There exists a gap between the actual figures given by the basic
input-output model and the beverage tax revenue generation that was
found in the case of the wagering tax. This step from the input-output
model to the tax revenue generation can be filled by knowledge similar
to that required by the wagering tax, including (1) a breakdown within
the total payments to individuals figures to allow a detailed applica-
tion of (2) a measure of the income-taxable beverage expenditure re-
lationship in the (3) absence of influence by purchases not accounted
for in the quadrant III rows. The alternative approach, accepting
the results of the analysis as valid, requires that the necessary sub-
steps be embodied in assumptions similar to those for the wagering tax
analysis. The beverage expenditure-beverage tax revenue generation
relationship is of course complicated by the multiple rate system
Documentary Stamp Tax
Documentary stamp taxation is a form of tax on documents creating
some form of obligation, generally indebtedness, including equity in-
debtedness such as corporate securities, and on instruments evidencing
transfers of interest in land. The nature of the tax makes analysis
a particularly nebulous undertaking because of the difficulty of de-
termining to what information in the basic input-output model the tax
can be connected which will allow inferences about tax revenue genera-
tion from that information.
The restrictive approach does not seem to be useful, since there
is no part of the basic input-output model that can be isolated as
being closely associated with documentary stamp tax revenue generation.
Indebtedness of equity or non-equity types and transfers of interest
in land are not unique to any sector, product or transaction as re-
flected in the input-output model.
The association with the basic input-output model required by a
general approach is less specific, nevertheless it must permit infer-
ences to be drawn about tax revenue generation given observable values
in the model. How best to treat the tax to make the inferences valid
is more difficult for the documentary stamp tax than for the wagering
and beverage taxes previously considered for the same reason that makes
the restrictive approach inappropriate--the lack of a firm relationship
between the input-output information and the tax revenue generation.
The documentary stamp tax is not associated with any product or
group of products in such a way as to make the information about those
products available from the input-output model useful for tax analysis.
Part of the tax deals with the sale of property which might in some way
be associated with the input-output model, and the tax does apply to
secured transactions under the Uniform Commercial Code pertaining to
the sale of goods. The taxation of initial stock issues and stock
transfers is not related to any particular sector good or service.
Thus treatment of the tax as product related or sector related is not
desirable given the nature of the tax.
Table 16 lists the documentary stamp tax tentatively in two places--
as a non-industrial transaction and a non-industrial event related tax.
The tax is associated, however, with voluntary non-fortuitous economic
transactions and a non-industrial transaction seems to be the more
reasonable classification. As a non-industrial transaction associated
tax, the documentary stamp tax can be treated as a type 1 tax associated
with input-output information found in the final demand sectors, or a
type 2 tax associated with quadrant III information, in this case the
value added row, or a type 3 tax related to the total economic
activity of a sector as reflected by the total gross output figures in
the input-output model.
The major features of the documentary stamp tax can be roughly di-
vided into three categories. First, the tax on indebtedness of the non-
equity type, specifically on secured transaction, might be associated
with final demand; an increase in total purchases would be associated
with a higher level of indebtedness and thus the greater the tax revenue
generation. This would suggest a type 1 tax treatment. Second, since
interindustry purchases generating indebtedness are also subject to the
tax, type 3 treatment might be suggested. The property transfers on
which the tax is imposed will also suggest either type 1 or type 3. The
third feature of the tax is its application to stock issue and transfers.
This could be associated most closely with information from the input-
output model gross output cells, since the total activity of the sector
might reflect the stock transfer activity and new issues of industries
in the sector. This would suggest a type 3 tax.
Since the type 3 tax treatment is common to all three features of
the tax, and the feature of the tax suggesting a type 1 treatment
is not exclusive of type 3 treatment, the documentary stamp tax will
be used in a type 3 framework, associated with the information in
the input-output model represented by the total gross output figures.
The allocation of the tax in the 1967 base model is taken from Table
B-3 and the tax coefficients calculated from this allocation are found
in Table C-l. Projected 1972 documentary stamp tax sector figures
are taken from Tables D-l, D-2 and D-3 for projectionsI, II, and III
respectively. The tax intensity coefficients are found in Table E-l.
The relationships which permit making the step from total gross
output figures in the basic input-output model to final tax collection
are now considered, dividing the discussion in three parts correspond-
ing to the non-equity indebtedness, stock issue and transfer, and real
property transfer aspects of documentary stamp tax. Basing tax
analysis on the total gross sector output figures requires an apprecia-
tion of the two different meanings of the total gross output numbers.
Calculation of a total gross output figure for sectors gives the
total of all output going to all other industries and to final demand.
It also gives the total purchases of that sector from all other
industries and from quadrant III sectors. It seems to be more useful
to emphasize the sales concept rather than the purchases concept and
this usage will be consistent.
The portion of the tax which relies on indebtedness instruments
must show the response of the documentary stamp tax to cell values in
the basic input-output model representing sales of the goods or service
of each sector to every other sector and to final demand. The previous
discussions of wagering tax and beverage tax have suggested three parts
to the total relationship of input-output information to tax revenue
generation--the characteristics of the information in the input-output
model itself, the "gap" between what is shown in the input-output model
and the specific economic activity taxed, and, finally, the relationship
between that economic activity and the actual revenue generation from
the tax. The indebtedness feature of the documentary stamp tax requires
an especially close look at the first of these, primarily because of the
restrictions on individual cell value relationships. To use gross
output as the measure in the first step in tax analysis, it is necessary
to accept the fact that the taxable sales must be treated equally for
all cells in the intermediate sectors and in all final demand sectors.
This also means that sales represented by cell values must be treated
alike for all goods and services from whatever sector.
Given the requirements of the structure of the input-output model
the next step in tax revenue determination requires information about
the relationship between sales indicated by cell values and the
financing arrangements surrounding the sale. The financing arrangement
must be known to show the relationship between the total sale figure
and the indebtedness which yields tax revenue. Finally, the relation-
ship between the indebtedness and actual tax revenue generation must
Placing the required information in the context of assumptions,
the proposed treatment of the indebtedness feature of the documentary
stamp tax will give information on tax revenue generation if the
following assumptions are made: (1) all sales information from the
intput-output model gives rise to proportional indebtedness, requiring
that (a) the amount of indebtedness is some given proportion of sales
values, (b) all goods and services are treated alike, and (c) there is
no distinction between inter-industry sales and sales to final demand,
and (2) the relationship between indebtedness and tax revenue genera-
tion is linear and identical for all cell values. Finally, the com-
bination of the above assumptions must lead to an overall relationship
between total gross sector output and tax revenue generation given by
a single linear relationship quantified by the tax coefficient.
The feature of the documentary stamp tax dealing with stock issue
and transfer presents a different set of information requirements when
treated as a type 3 tax. Total output of a sector is used as a measure
of economic activity of industries within the sector, and economic
activity is in turn used as an indicator of stock issue and transfer
activity upon which the tax is based.
The use of the gross output figures to gain information about
documentary stamp tax revenue is based on four steps. First, the total
gross output for each sector is used as a measure of the activity of
the sector, and is based on sales to other industries and final demand
alike. It is, therefore, a measure of total sales. The next relation-
ship of importance, and the one giving the most difficulty, is that
relating economic activity, as measured by total sector sales, to
stock issue and transfer.
Stock issue may be dependent on sector total output in several
ways. First, as output increases there are a large number of incor-
porations because of an increase in the number of firms producing the
output. Second, those firms which are already incorporated may wish
to improve their financing position with additional stock issue. Both
of these causes for stock issue are based on a change in total output
rather than the rate of total output. The relationship between a
steady sector output level and stock issue occurring in the course of
normal business turnover must also be known.
As a further step from input-output to tax information, the
transfer of stock must also be related to economic activity as repre-
sented by total output. As with stock issue, transfer may be related
to a change in output levels, and this relationship must be related to
a change in output levels, and this relationship must be known. Also
required is information on the stock transfer activity at each steady
level of output.
The total step showing a level of stock issue and transfer is a
combination of the four relationships mentioned above--issue and trans-
fer activity for both rate change and steady state in gross output.
When the overall combined relationship of total sector output to total
stock issue and transfer is known, the final step requires the addition-
al determination of the relationship between the issue and transfer and
final documentary stamp tax revenue generation. Two features of the
tax should be recognized at this point. First, the tax is based on
the value of the issue or transfer, not the number of issues or trans-
fers. Second, since the tax occurs only at issue or transfer, increases
or decreases in the value of stock are not accounted for until a
transfer is made.
In the context of required assumptions, those necessary for stock
issue tax analysis are: (1) stock issue, whether from new incorporation
or new issue of existing corporations is a function of the level of
economic activity as shown by the total gross sector output from the
input-output model, and (2) the function is linear and given by the tax
coefficient for documentary stamp tax. For the transfer tax the assump-
tions are analogous.
The third feature of the documentary stamp tax concerns the trans-
fer of property. Like stock transfer, the amount of tax is based on
the consideration paid for the property at the time of its transfer.
Following the pattern of analysis used for the indebtedness feature of
the tax, sales to all sectors and to final demand must be related to
the value of real property sales.
The assumptions to derive the needed tax information from the
total gross sector output figures are analogous to those for the in-
debtedness analysis: (1) all sales information from the input-output
model must give rise to proportional land sales, requiring that the
value of property transfer is some given proportion of sales value, and
(2) the relationship between property transfer values and revenue gen-
eration is linear and identical for all cell values.
The Florida inheritance tax is based upon the value of the estate
at the decedent's death and is patterned after the federal estate tax.
The rate is a progressive rate with exceptions and exclusions that
eliminate the tax on estates of less than $100,000.
The restrictive approach to estate taxation is not possible. The
imposition of the tax occurs only on the happening of a fortuitous
event and classification as a non-industrial event is suggested. The
tax is also personal in nature applying only to individuals. The tax
can therefore best be analyzed as a type 2 tax based on quadrant III
information, specifically the personal income information row.
The empirical results of Chapter IV give the information available
for inheritance tax using the type 2 analysis. The allocation of
inheritance tax to each sector is taken from Table B-5 and the tax
coefficients calculated from that allocation are found in Table C-2.
Tax revenue generation for 1972, using final demand projection methods
I, II and III, are taken from Tables D-4, D-5 and D-6, respectively.
Tax intensity coefficients for inheritance tax are from Table E-2.
The information necessary to bridge the gap between the quadrant
III values and the inheritance tax revenue is somewhat analogous to
previous type 2 tax analyses, but there are significant and interesting
differences. The first step requires a consideration of the character-
istics of the income payments to individuals and how these characteris-
tics vary among sectors.
The second step, that of determining a relationship between income
and the value of an estate, requires some relationship between the level
of income as reflected in the quadrant III information from the input-
output model and the accumulated personal wealth of each sector. Then
it is necessary to find a relationship between personal wealth accumu-
lation and taxable estate value.
The determination of tax revenue generation from general estate
value information is a more difficult task than that of determining
tax revenue generation from the taxes considered previously for two
major reasons. Taxable estates are not a homogeneous component of all
estates and the proportion is not easy to determine. Second, the tax
rate varies, the great majority of estates escape taxation entirely and
the taxed estates have a wide range of tax rates and payment.
One of the two special problems with the information necessary to
fill the gap from gross quadrant III values to estate tax revenue gen-
eration has been alluded to above--that of getting enough information
on the character of income and estates to permit an application of a
single relationship to the gross sector income-estate tax revenue step.
The relationship is made crucial because of the disparate tax imposition
on estates of different values, including the elimination of tax for
"smaller" estates. A second special problem concerns the time factor.
An estate is the result of an accumulation process, and even if the tax
revenue generation from an estate could be tied directly to the income
which gave rise to the estate, a time lag of many years between the
income and the final accumulated estate would exist.
In the context of assumptions, the results obtained from the
empirical work of Chapter IV will be a valid use of the input-output
information for estate tax revenue generation only if the following
assumptions are made: (1) characteristics of each sector quadrant III
cell are such that all sectors give a contribution to the value of the
estates in exact proportion to their total value, and (2) the contri-
butions to estates are such that the tax revenues from each sector's
estates are a linear function of the quadrant III values. Assumption
(2) is a combination assumption that the quadrant III contributions to
estate values are such that the values of the estates are proportional
to the cell values and the tax revenue generation from the estates is
a linear relationship, the combination yielding the linear function of
assumption (2). The total result of the assumptions must be a linear
relationship between the quadrant III cells and inheritance tax revenue
with the value of the tax coefficient.
Corporate Capital Stock Tax
The corporate capital stock tax was a yearly graduated tax based
on the book value of a corporation's securities. It was phased out
with the advent of the corporate income tax.
The general approach seems to be the better approach to analysis
of this tax. The tax is associated directly with the economic activity
reflected in the input-output model, though not with any particular
product, yet it is not non-industrial, thus it can best be classified
as a sector related tax.
The collection of the tax falls directly on the corporate entity
on the basis of the book value of its outstanding stock. The capital
value of corporations is not associated in any direct way with a quad-
rant III row indicating corporate purchases of primary input. It seems
more reasonable to tie the corporate capital position to its sales.
Since no distinction between sales to final demand and interindustry
sales is justifiable, the total gross sector output measuring gross
sales is the choice for tax type classification, which gives the corpor-
ate capital stock tax a type 3 classification.
Allocation of the capital stock tax to sectors based on type 3
analysis for the 1967 base model is taken from the empirical work shown
in Table B-3. Tax coefficients calculated from this allocation are
found in Table C-l. Sector tax projections for 1972 for final demand
projection methods I, II and III are from Tables D-l, D-2 and D-3, re-
spectively. The results of tax intensity coefficient calculations are
from Table E-l.
The total gross output figure for each sector provided by the
basic input-output model shows only the total interindustry and final
demand sales. Several relationships must be defined to reach the
point where that information will yield tax revenue generation figures.
The first of these must indicate how sector total sales are related to
the capital structure of the corporate portion of all producers in the
sector, and how this relation varies by sector. Also necessary, of
course, is information about the portion of total output attributed
to incorporated entities.
The nature of the tax prevents establishment of a simple relation-
ship between corporate sales and the market value of corporate stock. The
value of stock which is taxed is book value, so the important aspect of
the securities financing of corporations is not that of sales and sale
value but of sales and value of stock at issue, which includes changed par
or non-par stock values and number of stocks issued,a combined contri-
bution both by new issue of existing corporations and by new incor-
porations. In opposition to the stock transfer and issue feature of
the documentary stamp tax, the taxable stock value depends on existing
book value alone and requires no knowledge about the transfer activity,
only the balance. Additionally, the same book value will yield the
same tax each year without regard to issue or transfer activity.
Finally, the stock value must be related to tax revenue generation.
Because of the graduated nature of the rate, a linear function is not
Sales and Use Tax
The Florida sales and use tax is based largely on retail sales
of enumerated goods and service. It is not restricted to sales to
final demand sectors, however, and there are some features of the tax
which are aimed specifically at certain types of services, such as
short-term lodging facilities.
Because the sales and use tax is so closely associated with a
specific class of transactions, the more restrictive approach to tax
analysis is acceptable. The tax is not associated with a single sector's
output,however, and cannot be tied to a few sectors. Some output of
most, if not all, sectors is subject to the tax. The tax is directly
connected to the sale of a product and is best classified as a product
related tax where the term product includes the whole array of goods
and services produced in the economy.