- State and federal aid payments to local governments have grown explosively
over the last fifteen years. One reason for this growth is
that the donor governments can alter a local government's budget choices
through grants. Grants have not remedied social problems, however, because
the local governments' responses to grants were not anticipated.
Economists have broadened our understanding, but debates remain about
the proper modeling of local buur grant impact knowledge. Accordingly,
this study examined the effects of thdgeting and, because researchers used
aggregated data, gaps exist in oree types of mental health grants
on the budget allocations for Oregon counties.
The theoretical literature on the expenditure effects of grants
was first reviewed. The constrained maximization, median voter, and
budget maximization models of local fiscal decision-making were described
and then compared as to their predictions about the effects of different
grants. It was concluded that too little was known about budgetary
processes to use or compare the models' predictions.
Empirical studies were then reviewed. Researchers, largely using
demand frameworks, found that grants significantly affected local
spending and that different grants induced different spending responses. Their estimates of the stimulus differed, though. Moreover, little or
no research was undertaken on the employment, wage, and output effects
of grants. The theoretical and statistical problems with these studies
were examined. These problems were: (1) the misspecification of aid
variables; (2) the aggregation of government units and public services;
(3) the lack of institutional and political realism.
A theoretical model of Oregon counties' expenditure and production
decision-making for mental health services was developed based on the
insights and criticisms of existing models. The model consists of
eleven equations; some describing the "expenditure stage" of the budget
process, others describing the "output stage". It was argued that
county commissioners make the expenditure decisions, and that mental
health administrators make the production decisions. The framework
allowed us to examine the effects of mental health grants on expenditures,
wages, staff numbers, patient numbers, and output and to study
the determinants of grant participation.
Using regression analysis, the equations were estimated from the
observations for 31 Oregon counties in fiscal year 1975-1976. Ordinary
least squares was used in the expenditure and grant participation equations.
Two-stage and three-stage least squares were used in the rest.
Regressions were run for western and eastern Oregon counties when possible.
For all observations, the major findings suggested that a dollar of
state matching mental health aid per capita stimulated per capita mental
health expenditures by $1.37, increased the professional staff by .556
to .762 persons per 10,000 county residents and increased average professional
salaries by $2,173. A dollar of federal matching aid per
capita appeared to have an expenditure effect of $1.03, an employment effect of .722, and no salary effect. A dollar of non-matching aid per
capita had an estimated expenditure effect of $1.00, an estimated employment
effect of .35, and no salary effect.
In eastern Oregon, the major findings indicated that the marginal
expenditure effect of federal aid was $1.41, the marginal expenditure
effect of non-matching aid was $.96, and that state matching aid had no
In western Oregon, a dollar of state matching aid per capita had
an estimated expenditure effect of $2.23, a professional employment
effect of 1.25, and no significant salary effect. A dollar of non-matching
aid per capita had an estimated expenditure effect of $1.67,
and no significant employment or salary effects.
In all regressions, the mental health grant estimates were not
statistically different from one another. Finally, a production function
for mental health services was unsuccessfully estimated and discussed.