Expenditure Tradeoffs in the American States: a Longitudinal Test, 1948-1984

• Published date: 01 December 1991
• Date: 01 December 1991
• DOI: 10.1177/106591299104400407
• Subject Matter: Articles

EXPENDITURE TRADEOFFS IN THE AMERICAN
STATES: A LONGITUDINAL TEST, 1948-1984
JAMES C. GARAND, Louisiana State University
and
REBECCA M. HENDRICK, University of Wisconsin, Milwaukee
ver
the years, numerous scholars have explored the degree to
~
which there are tradeoffs between defense and domestic (mostly
welfare) spending in the United States and other Western
democracies (Russett 1970, 1982; Perroff and Podalak-Warren 1979;
Domke, Eichenberg, and Kelleher 1983; Fischer and Kamlet 1984;
Kamlet and Mowery 1987; Mintz 1989). While there is substantial
scholarly disagreement about how tradeoffs among expenditure cate-
gories are best measured in the first place (cf., Berry and Lowery
1990), there seems to have emerged in the extant literature a consen-
sus that calls into question the degree to which systematic tradeoffs
exist between defense and domestic expenditures. Such a finding would
seem to be at odds with some journalistic and historical accounts that
claim evidence for tradeoff effects in the decision process for defense
and domestic spending. Scholars have suggested that the lack of an
observed pattern of tradeoffs may be due to the ability of national gov-
ernments to avoid making hard decisions about trading spending for
one expenditure area off against spending for another by financing
new spending through either increased taxes or larger budget deficits
(Domke et al. 1983; Mintz 1989).
To date, there has been virtually no attempt to examine the degree
to which tradeoffs occur among spending categories within the Amer-
ican states. To be sure, there are no spending categories analogous to
defense and domestic spending found in the American states; it is
clear that one cannot frame the question of expenditure tradeoffs in
Received: January 11, 1990
Revision Received: October 26, 1990
Accepted for Publication: November 1, 1990
NOTE: This is a revised version of a paper presented at the 1988 annual meeting of the
American Political Science Association, September 1-4, 1988, Washington, D.C.
The order of the coauthors is determined alphabetically, with each author sharing
equally in the preparation of the final manuscript. We gratefully acknowledge the
assistance of Martin A. Miller of the Social Science Research Facility, University
of Wisconsin, Milwaukee, though we remain responsible for any errors contained
herein.

---

916
the American states as a &dquo;guns-versus-butter&dquo; debate, as is often done
in research on expenditure tradeoffs at the national level. Nonetheless,
there are important reasons for testing models of expenditure tradeoffs
in the American states. First, as Elling (1983: 269) suggests, the bud-
getary process is the &dquo;single most important policy statement that a
state government makes from one year to the next,&dquo; with spending
decisions serving as &dquo;the most tangible indicators of a given governor’s
or legislature’s policy priorities.&dquo; In contrast, the federal government
has a wide variety of nonbudgetary mechanisms (e.g., regulatory
authority, symbolic power, etc.) available for establishing and chang-
ing policy priorities. Although state governments also have some of
these powers, their discretion in these matters is generally more lim-
ited than that of the federal government. Thus, the determination and
change of policy priorities are more likely to be reflected in state bud-
get and expenditure patterns.
Second, there are important structural reasons to believe that spend-
ing decisions in the states may be much more of a zero-sum game than
spending decisions at the national level, meaning that expenditure
tradeoffs will be more likely to occur in the states. Because many state
governments face legal or constitutional restrictions on increasing rev-
enues and/or deficits, they do not have the same degree of flexibility to
finance expenditure increases through new taxes or debt. Therefore,
shifts in policy priorities at the state level may necessitate the with-
drawal of funding from one or more spending areas to increase spend-
ing in another area. Even in the absence of the guns-versus-butter
spending areas, there are equally pivotal policy areas in the states that
could serve as targets for tradeoffs. The most likely areas are high-
ways, welfare, education, and health expenditures, which together
comprise as much as 85 percent of expenditures in most states (Hanson
1990).
The purpose of this paper is to assess the empirical validity of the
tradeoff hypothesis in the American states during the post-World War
II era. We utilize data on state expenditures in the key spending areas
of education, highways, welfare, and health and hospitals for the years
1948 to 1984 to construct a model of the degree to which tradeoffs in
priorities among expenditure categories can be observed within each
of the American states. Like Domke et al. (1983), we focus our efforts
on systematic, short-term patterns of tradeoffs from one year to the
next. If expenditure tradeoffs exist, one would expect spending prior-
ity changes in one area to be inversely and systematically related to

---

917
spending priority changes in other areas, once the confounding effects
of control variables are taken into account. In this paper we chart pat-
terns of tradeoffs in expenditure priorities among various spending
areas in each state as a means of assessing the empirical validity of the
tradeoff model across states and across a range of expenditure areas.
Ultimately, this effort may provide greater insight into state spending
outcomes and decisions across a range of policy areas.
THE CONCEPT OF EXPENDITURE TRADEOFF
Before one can address the issue of whether expenditure tradeoffs
exist in the American states, one must first define the concept and
describe in precise terms the criteria utilized to infer the existence of
such tradeoffs. For our purposes, an expenditure tradeoff is defined as
a systematic pattern of direct shifts in spending priorities from one
spending category to another, implying a pattern of decisions to accept
lower relative spending levels in one or more specific expenditure areas
in order to fund higher relative spending levels in another expenditure
area. The concept of expenditure tradeoff implies a dynamic policy
process, with spending priorities at some time (t) shifting to a new set
of spending priorities at time t + 1 as a result of a willingness by policy
makers to increase relative spending levels in one category at the expense
of another. Alternatively, we define mutual changes as the tendency of
positive (negative) changes in one spending area to be associated with
positive (negative) changes in one or more other spending areas. Mutual
change relationships occur when there is a tendency to increase or
decrease simultaneously the relative yearly spending levels among two
or more expenditure categories. This phenomenon suggest no shifting
of priorities, and corresponds to Wildavsky’s (1964) concept of &dquo;fair
share&dquo; budgeting.
Moreover, we define an expenditure priority as the relative weight
given by policy makers to an expenditure category when making deci-
sions to allocate society’s resources. Priorities not only involve the deci-
sion to allocate resources to one govermental expenditure category
rather than another, but also the decision to allocate resources from
the private sector to the public sector. For example, the welfare prior-
ities for a political system are defined inadequately in terms of raw
welfare spending levels or welfare spending as a proportion of total
government spending. Instead, welfare priorities can best be defined
in terms of welfare spending as a proportion of some equivalent to

---

918
Gross National Product (GNP) or Gross Domestic Product (GDP),
both of which represent, at the extreme, the total of resources avail-
able to policy-making bodies for welfare spending. Such a formulation
recognizes that two political systems may both allocate 25 percent of
their government expenditures to welfare spending, but the political
system that allocates a higher share of GDP to government expendi-
tures (and hence, a higher share to welfare spending) can be viewed as
according welfare spending a higher priority.
To be more specific about how tradeoffs and mutual changes are
determined, consider the following hypothetical data that illustrate
two patterns of spending as a proportion of GNP for education and
welfare expenditures over seven time periods. Although the examples
represent different hypothetical relationships among only two public
goods, they demonstrate how tradeoffs and mutual changes occur among
any number of expenditure categories. For expository purposes, the
examples illustrate extreme tradeoff and mutual change conditions.
An examination of the pattern of GNP proportions in case A
reveals increasing proportions of GNP allocated for both education
and welfare. However, education proportions are increasing at a decreas-
ing rate, while those for welfare are increasing at an increasing rate.
As such, the pattern in case A represents a tradeoff situation in which
priorities for welfare are increasing relative to education; the higher
the increase in welfare, the lower the increase in education, even
though spending for both is increasing. In this case it would appear
that increase in welfare spending priorities are coming at the expense
of larger increases in education priorities. Over time, political and
other circumstances may reverse the trend by placing a higher relative

---

919
priority on education. In this...