The "Politics" of Redistribution

• Author: Eric M. Uslaner,Ronald E. Weber
• DOI: 10.1177/1532673X7500300202
• Date: 01 April 1975
• Published date: 01 April 1975
• Subject Matter: Articles

THE "POLITICS" OF REDISTRIBUTION
Toward a Model of the Policy-Making
Process in the American States
ERIC M. USLANER
University of Florida
RONALD E. WEBER
Indiana University
Students of policy-making in the American states have been
engaged in an almost endless debate over the determinants
of policy outputs. Do &dquo;socioeconomic&dquo; variables, typically
represented by measures of per capita income and other
demographic factors such as urbanization and industrialization,
have a greater impact on policy formation than &dquo;political&dquo;
variables, such as the extent of two-party competition, the level
of turnout in elections, legislative professionalism, and appoint-
ment (see Dawson and Robinson, 1963; Hofferbert, 1966; Dye,
1966; Sharkansky, 1968a; Sharkansky and Hofferbert, 1969)?
Author’s Note: We are grateful to Leroy N. Rieselbach, Anne L. Schneider, John L.
Sullivan, Lawrence C. Dodd, Charles W. Ostrom, Jr., Barbara Salert, and John V.
Gillespie for their comments on earlier drafts of the paper, and to Carol Jansen for
her assistance on the data collection. Computations for this paper were performed at
the Indiana University Research Computing Center with support from the Research
Computing Fund of Indiana University. The paper was prepared using the facilities of
the Political Science Laboratory and Data Archive.
AMERICAN POLITICS QUARTERLY, Vol. 3 No. 2, April 1975
@1975 Sage Publications, Inc.
[130]

---

[131]
We doubt that any &dquo;final test&dquo; of this question is possible.
(For a most convincing argument that &dquo;crucial experiments&dquo;
cannot lead to the rejection of a theory, see Lakatos, 1970.)
Rather, we maintain that the analyses of policy outputs which
have dominated previous research have served to obfuscate
rather than to clarify the structure of the conversion process in
state policy-making. In these studies, the concern has been with
the determinants of policy outputs rather than with the actual
factors which state decision-makers employ in formulating these
outputs. Environmental factors-be they economic, social, or
political-undoubtedly have a great impact on the decision-
making process. Previous studies, however, have not attempted
to link the environment to the decision-making process in state
houses, legislative bodies, and the bureaucracies. The missing
element is the decision-maker himself. We view the various
environments of the political system as variables representing
either positive or negative reinforcements of values held by the
decision-makers-and not as &dquo;causes&dquo; of policy outputs. Thus,
without denying the importance of environmental factors in the
study of state policies, we concentrate instead on the &dquo;polit-
ical&dquo; aspects of the conversion processes in an attempt to make
inferences about policy-making in the states.
Our perspective on policy formation introduces another
element into the study of comparative state politics. Studies
stressing the importance of environmental variables have tended
to concentrate on gross indicators such as expenditures (either
total or per capita) for specific policy areas such as welfare,
education, highways, and so forth). The decision-making arena,
however, requires a more comprehensive approach. Decision-
makers do not consider each request for funds in isolation.
Instead, they face the problem of how to distribute the
resources of a state’s tax base among various groups competing
for a share of the total pie. Every dollar spent on highways
cannot be expended on public welfare. The question at issue is
not simply one between incrementalism in budgeting and
program budgeting. Rather, it is one of policy impact. Which
groups in a population reap the most benefits from the
policy-making process and which groups gain the least?

---

[132]
Hence, the study of policy formation in the states involves
three basic questions: (1) who benefits from a state’s allocation
of its resources? (2) what criteria do decision-makers use in
distributing benefits and costs? and (3) what constraints or
positive reinforcements in a state’s political system affect the
decision-making process? Research on comparative state politics
has tended to ignore the first and second questions almost
entirely and to treat the third as the method of explaining
variations in state expenditures. A significant advance recently
has been made on the first question: Fry and Winters (1970)
have formulated an index of &dquo;redistribution&dquo; of state resources.
Distributive policies are defined as &dquo;those perceived to confer
direct benefits upon one or more groups. Typically such policies
are determined with little or no conflict over the passage of
legislation, but only over the size and the specific distribution
of the shares&dquo; (Salisbury, 1968: 158). Redistributive policies,
which also confer direct benefits on groups within a state, on
the other hand, are perceived to impose costs on other groups
(Salisbury, 1968: 158-159). Thus, the Fry-Winters index of
redistribution is based on a ratio between the expected benefits
the lowest three income classes in a state obtain from a wide
range of fiscal policies and the share of the tax burden these
groups contribute (Fry and Winters, 1970: 513-515).
Unfortunately, Fry and Winters addressed themselves only to
the third question. Their data analysis follows the traditional
pattern of using the socioeconomic and political variables
discussed above to explain variations in state redistribution
ratios. They conclude that political variables have a stronger
independent impact on redistribution than socioeconomic
variables. But Sullivan (1972) has suggested that their findings
may be an artifact of the large number of political variables
relative to the number of socioeconomic variables employed in
their analysis. Again we see an example of the traditional debate
in the study of policy formation in the American states, and the
discipline is no closer to a resolution of this debate now than it
was fifteen years ago.
We believe, however, that the work begun by Fry and Winters
is important, because it comes directly to grips with our first

---

[133]
question. In this paper, we propose some first steps toward a
conversion model of the policy-making process in the American
states and assess its predictions. The explanatory power of the
model lies not only in its ability to produce relatively high
multiple correlation or regression coefficients, but also in its
focus on the decision-makers’ actual role in policy formation
and on the constraints and positive reinforcements in the
political environment that affect these decision-makers.
FORMULATION OF THE MODEL
The most important assumption we make is that &dquo;policies are
made in response to someone’s preferences&dquo; (Weber and
Shaffer, 1972: 684). The preferences we posit to have the
greatest immediate impact on the determination of public
policies are those of the men and women actually charged with
making the decisions on such policies (see Figure 1 ). Given the
paucity of information currently available on the preferences of
governors and bureaucrats (for exceptions, see Wright 1967;
Beyle, 1968; Grupp and Richards, forthcoming), we have
chosen state legislators as the central decision-makers in our
model. The legislator as an individual has preferences on public
policies. These preferences, together with the cues he gets from
the environment and his fellow policy-makers, are assumed to
account for his decision-making behavior. The cues from the
environment include his own perceptions of the need for such
policies, the preferences of his constituents, and the demands
made by groups particularly concerned with a specific set of
policies. At the state level, however, a legislator will not be able,
in general, to gauge public opinion on a given policy in his
district. (For a report on a methodology for estimating public
opinion in state legislative districts, see Schneider, forth-
coming.) Consequently, we do not assume that public opinion
on a given policy has a direct impact on policy formation at the
state level. Rather, the impact is indirect; only mobilized public
opinion is effective.1 To the extent that citizen preferences can
be mobilized, either political parties or interest groups will be
the primary agents of such mobilization. With respect to

---

[134]
Figure 1: A
CONCEPTUAL
MODEL FOR THE ANALYSIS OF POLICY-MAKING
This model is comparable to those presented in Richard 1. Hofferbert, &dquo;Elite
Influence in State Policy Formation,&dquo; Polity, Vol. 2 (Spring, 1970), pp. 316-344; and
Stuart H. Rakoff and Guenther F. Schaefer, &dquo;Politics, Policy, and Political Science:
Theoretical Alternatives,&dquo; Politics and Society, Vol. 1 (Nov., 1970), pp. 5 1-77. The X
symbol suggests that a facilitative or mediating effect is taking place with the
Governmental Institutions Characteristics and the Electoral and Party System
Characteristics influencing the linkage of one set of preferences with another set of
preferences.
redistribution, the party system is a more powerful inter-
mediary force than interest groups. The latter generally are
more concerned with questions of regulation than of redistri-
bution (see Lowi, 1964; Salisbury, 1968; Salisbury and Heinz,
=
1970). Furthermore, the groups in a state which benefit most
from redistributive policies (the three lowest income classes)
rarely possess the resources to form powerful interest groups.
Political parties in the United States, however, are not
primarily agents which...