Before the Roll Call: Interest Group Lobbying and Public Policy Outcomes in House Committees

• Author: Diana Evans
• DOI: 10.1177/106591299604900203
• Published date: 01 June 1996
• Date: 01 June 1996
• Subject Matter: Articles

Before the Roll Call: Interest
Group Lobbying and Public
Policy Outcomes in House
Committees
DIANA EVANS, TRINITY COLLEGE
This research examines the relationship between interest group activities
and the public policy decisions of congressional committees. Unlike the
conventional approach, which estimates the relationship between interest
group resources and legislators’ roll-call votes, the technique used here
treats as the dependent variable interest group success at getting what they
want
from committees on their individual policy preferences. This approach
has the unusual advantage of allowing systematic investigation of deci-
sions made behind the scenes which never come to a formal vote. It also
facilitates estimation of the relationship between policy outcomes and lob-
bying directed at committee staff as well as members, and allows evalua-
tion of the effect of contextual variables, especially conflict. A model of
interest group success on dozens of provisions of two complex bills was
estimated. Although conflict among groups dampened interest group suc-
cess, lobbying committee staff was associated with group success. How-
ever, even under the decision-making condition expected to be most
favorable to interest groups, virtually complete public invisibility, interest
groups did not always get their way with congressional committees.
The central question in the literature on Congress and interest groups con-
cerns the degree of interest group influence on the public policy decisions
made by members of Congress. Although some theoretical literature on inter-
est groups suggests that under specific conditions they are virtually certain to
get what they want (for example, Hayes 1981; Lowi 1979; Wilson 1974),
individual-level studies of congressional decision making most often have failed
NOTE: The author thanks Richard Smith, Robert Salisbury, Adam Grossberg, and two
anonymous reviewers for their advice on previous versions of this article. Thanks
also to Bill Adams and Jared Haller for research assistance. This research was
funded by grants from the Dirksen Congressional Center and Trinity College.
287

---

to find a strong interest group impact, particularly for PAC contributions, on
Congress members’ roll-call voting (Chappell 1982; Grenzke 1989; Nelson
1982; Saltzman 1987; Wayman 1985; Welch 1982; and Wright 1985)1
However, several studies indicate that the impact of PAC contributions
varies with the context in which legislation is considered, especially the vis-
ibility of the issue (Evans 1986; Schroedel 1986; Jones and Keiser 1987;
Neustadtl 1990). The policymaking process normally is far less visible in com-
mittee, where legislation is written, than on the floor of the legislative cham-
ber ; thus, interest groups are likely to have the greatest success in committee,
where they frequently enjoy long-standing relationships and a coincidence of
interests with legislators and staff (for example, Fenno 1973; Kingdon 1989:
154-62; Maass 1951; Cater 1964; Hamm 1983; Ripley and Franklin 1987;
Shepsle 1978). Although some recent research has focused on decision mak-
ing by committees (Schroedel 1986; Hall and Wayman 1990; Wright 1990),
most studies of the effect of PAC contributions model their impact on roll-call
voting on the floor of the House or Senate.
The present research examines interest groups’ activities at the committee
stage using a novel unit of analysis, the groups’ policy preferences; the depen-
dent variable is interest group success on their policy preferences. The ap-
proach developed here allows estimation of the impact on interest group success
of a number of important variables that normally cannot be included in con-
ventional roll call analysis. Moreover, unlike roll call studies, this design al-
lows analysis of publicly invisible decisions made behind the scenes in
congressional committees.
DESIGN OF THE ANALYSIS
The argument here is that interest group success at gaining favorable commit-
tee responses to their policy requests is a function partly of the group’s de-
ployment of its resources, especially campaign contributions and lobbying,
and partly of contextual factors over which it has little or no control, most
importantly the degree of conflict over the issue. In this section I delineate a
method for estimating the impact of such variables on the policy outcomes of
private committee negotiations and the shortcomings of roll-call analysis in
this regard. In the following sections I demonstrate the efficacy of this tech-
nique using two case studies of bills from the U.S. House of Representatives,
one of which is a highway reauthorization bill reported by the Public Works
1 A minority of scholars have found that campaign contributions from interest groups
have a substantial impact on roll-call voting (for example, Brown 1983; Fleisher 1993;
Frendreis and Waterman 1985; Kau, Keenan, and Rubin 1982; Wilhite and Theilman
1987).
288

---

and Transportation Committee and the other a bill to regulate food package
labeling from the Energy and Commerce Committee.
The unit of analysis is the individual policy preference (which takes the
form of a request for a particular policy output) expressed by each group that
testified in subcommittee hearings. The dependent variable consists of inter-
est group i’s success on its policy request j as indicated by the committee
report on the bill in question. Each group normally makes more than one
policy request, and there is a separate observation for every preference ex-
pressed by each group. Thus, the number of observations equals the sum of
all the policy preferences expressed by all the groups? This produces a focus
on the policy outcomes of committee decisions rather than individual mem-
bers’ decisions, like most studies of interest group influence.
Using interest group policy preferences as the unit of analysis produces
several unusual analytic opportunities. Although interest groups are likely to
take positions on numerous provisions of complex bills, typically few of those
individual provisions ever come to a vote. Clearly, the conventional method
of analyzing the roll-call votes of members of Congress to determine the im-
pact of interest groups on their policymaking limits the range of issues on
which the impact of interest groups can be analyzed. By contrast, the research
design used here allows the examination of interest group participation in
committee decisions that do not come to a recorded vote, as was true of virtu-
ally all of the decisions considered here. Rather, deals were negotiated behind
the scenes in settings so private that for all practical purposes they were not
visible to the general public until well after the fact, if ever. In such conditions
of virtual invisibility, deals struck for the benefit of organized interests can be
cloaked in terms of the public interest more convincingly than decisions that
involve public debate (Edelman 1964; Mayhew 1974: 134-35), as the impli-
cations of the latter are more likely to have been exposed by expanded con-
flict (Schattschneider 1960). Thus, behind-the-scenes settings are ones in which
we expect resource-rich interest groups to have the maximum impact.
In addition, this method facilitates estimation of the relationship between
both a variety of lobbying techniques and contextual factors on one hand and
interest group success on a large number of policy decisions on the other. The
impact of lobbying most often has been neglected in systematic analyses of
congressional decision making (for exceptions, see Langbein and Lotwis 1990;
Smith 1984; and Wright 1990). Using this method it is possible to assess the
’ John Kingdon (1989) used a somewhat similar design in his analysis of House mem-
bers’ voting decisions, interviewing members about multiple voting decisions per member
and analyzing the determinants of those decisions in a single data set (Kingdon 1989:
13-16 and Appendix A).
289

---

relationship between policy outcomes and lobbying directed not only at rep-
resentatives but also at congressional staff, whose central role in the process is
well established (Browne 1988; Berry 1989; Fox and Hammond 1977; Malbin
1980). Moreover, intergroup conflict damages a group’s chances of success
(for example, see Oppenheimer 1974; Price 1975 and 1978; Ripley and
Franklin 1987). This is true in part because conflict is more likely to occur on
issues that inherently involve winners and losers, specifically regulatory and
redistributive issues, than on distributive issues where there are more likely
no clear losers (Lowi 1964; Hayes 1981). Similarly, when members hear only
one side of an issue, it is somewhat less likely that they will see a basis in
terms of their public policy or influence goals for opposing the groups’ posi-
tion. The technique developed here makes it possible to asses the impact of
conflict on interest group success?
In conventional roll-call studies, where the usual approach is to model
each vote separately, analysis of the impact of such explanatory variables is
difficult or impossible. Even if a contextual variable such as the level of con-
flict is different on each of several separately analyzed votes, it is impossible to
make reliable generalizations about its impact, as the level of analysis does not
allow it sufficient variance. The same is true for lobbying, as a group may
lobby at different levels for different proposals.
The other common approach to roll-call analysis is to analyze individual
members’ votes on many issues by constructing an...