You are on page 1of 25

Buying Time: Moneyed Interests and the Mobilization of Bias in Congressional Committees

Author(s): Richard L. Hall and Frank W. Wayman

Source: The American Political Science Review, Vol. 84, No. 3 (Sep., 1990), pp. 797-820
Published by: American Political Science Association
Stable URL:
Accessed: 22-11-2016 21:09 UTC

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted
digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about
JSTOR, please contact

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at

American Political Science Association, Cambridge University Press are collaborating with
JSTOR to digitize, preserve and extend access to The American Political Science Review

This content downloaded from on Tue, 22 Nov 2016 21:09:36 UTC
All use subject to
University of Michigan
Ann Arbor
University of Michigan

Over the last two decades institutional critics have increas-

ingly charged that moneyed interests dominate the legislative process in Congress.
Systematic research on campaign contributions and members' floor voting, however,
provides little supporting evidence. We develop a view of the member-donor relation-
ship that questions the theoretical underpinnings of the vote-buying hypothesis itself and
suggests two alternative claims: (1) the effects of group expenditures are more likely to
appear in committee than on the floor; and (2) the behavior most likely to be affected is
members' legislative involvement, not their votes. In order to test this account, we
specify a model of committee participation and estimate it using data from three House
committees. In contrast to the substantial literature on contributions and roll calls, our
analysis provides solid support for the importance of moneyed interests in the legislative
process. We also find evidence that members are more responsive to organized business
interests within their districts than to unorganized voters even when voters have strong
preferences and the issue at stake is salient. Such findings suggest several important
implications for our understanding of political money, interest groups, and the represen-
tativeness of legislative deliberations.

A t least since the interests of U.S. democracy. Such

Madison railed about the mischiefs of fac-concerns have hardly abated thirty years
tion, critics of U.S. political institutions since the publication of Schattschneider's
have worried about the influence of orga- essay. In particular, the precipitous
nized interests in national policy making. growth in the number and financial
In this century, one of the most eloquent strength of political action committees has
critics of the interest group system was E. refueled the charge that moneyed interests
E. Schattschneider, who warned of the in- dominate the policy making process. The
equalities between private, organized, current Congress is The Best Congress
and upper-class groups on the one hand Money Can Buy according to one critic
and public, unorganized, and lower-class (Stern 1988), one where Honest Graft is
groups on the other. The pressure system, an institutional imperative (Jackson 1988;
he argued in The Semisovereign People see also Drew 1982; Etzioni 1984). "The
(1960), "mobilized bias" in national policy rising tide of special-interest money," one
making in favor of the former, against the close observer concludes, "is changing the
interests of the latter, and hence against balance of power between voters and



This content downloaded from on Tue, 22 Nov 2016 21:09:36 UTC
All use subject to
American Political Science Review Vol. 84

donors, between lawmakers' constitu- tives are viewed as parties to an implicit

tional constituents and their cash con- cooperative agreement, but the con-
stituents" (Jackson 1988, 107). straints on member behavior and the ra-
Despite the claims of the institutional tional calculations of group representa-
critics and the growing public concern tives limit the extent to which votes
over PACs during the last decade, the become the currency of exchange. In-
scientific evidence that political money stead, we advance two hypotheses about
matters in legislative decision making is the effect of money on congressional deci-
surprisingly weak. Considerable research sion making.
on members' voting decisions offers little First, we suggest that in looking for the
support for the popular view that PAC effects of money in Congress, one must
money permits interests to buy or rent look more to the politics of committee
votes on matters that affect them. Based decision making than those of the floor.
on an examination of 120 PACs in 10 This view, of course, is neither original
issue areas over four congresses, one re-nor remarkable. Students of Congress
cent study concludes flatly that PAC con- have long contended that interest group
tributions do not affect members' voting influence flourishes at the committee
patterns (Grenzke 1989a). Another study, level, and recent students of PAC influ-
designed to explore the "upper bounds" of ence invariably advocate that work move
PAC influence on House roll calls, in this direction (e.g., Grenzke 1989a, 18;
emphasizes "the relative inability of PACs Schlozman and Tierney 1986, 256). To
to determine congressional voting" date, however, systematic studies of
(Wright 1985, 412). Other studies have PACs and committee decision making
come to similar conclusions (see e.g., have been altogether rare (for an impor-
Chappell 1982; Wayman 1985; Welch tant exception, see Wright 1989). We
1982), though there are also dissenting focus here at the committee level and em-
voices (e.g., Kau and Rubin 1982; Silber- phasize the theoretical reasons for doing
man and Durden 1976). On the whole, so.
then, this literature certainly leads one to Second, and more importantly, our
a more sanguine view of moneyed inter- account of the member-donor exchange
ests and congressional politics than one leads us to focus on the participation
gets from the popular commentaries. of particular members, not on their
Does money matter? votes. This variable, we believe, is a cru-
Our approach to this question is two- cial but largely neglected element of con-
pronged. In the first two sections, we gressional decision making. It is especially
revisit the question by developing a theo- important in any analysis of interest
retical account of the constrained ex- group influence in a decentralized Con-
change between legislator and donor quite gress. In their famous study of lobbying
different from the one evident in the sub- on foreign trade policy, for instance,
stantial literature cited above. In particu- Bauer, Pool, and Dexter concluded that a
lar, we adopt the premise that PACs are member's principal problem is "not how
rational actors, seeking to maximize their to vote but what to do with his time, how
influence on the legislative outcomes that to allocate his resources, and where to put
affect their affiliates; but we take issue his energy" (1963, 405). More recently,
with the standard account of PAC ration- Denzau and Munger (1986) have modeled
ality. Our approach does not lead us to the interest group-member relationship as
predict a strong causal relationship be- an exchange of contributions and elector-
tween PAC money and floor votes. House al support for legislative services or effort.
members and interest group representa- If money does not necessarily buy votes


This content downloaded from on Tue, 22 Nov 2016 21:09:36 UTC
All use subject to
Bias in Congressional Committees

or change minds, in other words, it can serving as "investments" that will pay
buy members' time. The intended effect is dividends in legislative support at some
to mobilize bias in congressional commit- later date (e.g., Chappell 1982; Jacobson
tee decision making. 1980, 77, 82). The scientific evidence that
We then develop and estimate a model such effects appear only infrequently may
of committee participation that permits a be cause for relief among critics of the sys-
direct test of whether moneyed interests tem, but it is puzzling to theorists of insti-
do mobilize bias in committee decision tutional behavior. Why should PACs
making. Analyzing data from three House flourish, both in number and financial
committees on three distinct issues, we strength, when their legislative efficacy is
find that they do. In the final section we so low? The payoffs would appear inade-
briefly discuss the implications of the quate to sustain the cooperative relation-
findings for our understanding of money, ship.
interest groups, and representation in One possible explanation is that PACs
Congress. raise and disburse money with local con-
gressional elections, not specific legisla-
tive ends, in mind. Wright (1985) argues,
The Rational PAC Revisited in fact, that the decentralized nature of
most PAC organizations inclines them to
The interdependencies of legislators do just that. But this account simply
and moneyed interests have been widely moves the issue of PAC rationality to a
discussed by political scientists and wide- second, institutional level. Why would
ly lamented by critics of pluralism (see PACs organize in this way? Wright sug-
esp. Hayes 1981). The basis for politi- gests that the typical national PAC office
cal exchange is clear. Each depends at permits local officials substantial discre-
least partially on the other to promote itstion because it wants to encourage them
goals. Interest groups seek, among otherto continue raising funds. But the organi-
things, favorable action on legislation zation's fund-raising and disbursement,
that will affect them; members of Con- presumably, are intended for some more
gress seek financial and political support ultimate purpose, namely, to increase the
from particular groups. Like the relation- net political benefits associated with gov-
ship between legislators and bureaucrats, ernmental action (or inaction) on issues
however, the relationship between legis- that affect it. On the whole, using money
lators and interest groups is one of im- solely to affect election outcomes is not
plicit exchange: the actors "trade specula- likely to be a rational means to this end.
tively and on credit" (Arnold 1979, 36; see The probability that any single group's
also Denzau and Munger 1986; Hayes contribution will affect the outcome of a
1981). Contributions are marked some- congressional election-in which a wide
where in the invisible ledger, and a range of more powerful forces are at
group's political strategists presumablywork-is almost certainly slight. In the
can use them to their momentary legisla- aggregate it might affect the organiza-
tive ends. tion's political support within Congress
This account of the legislator-interestby only a member or two (Wayman
group relationship underpins the now- 1985). While organizational arrangements
considerable literature on contributions may create some inefficiency in the way
and roll call voting. The working hypoth- PACs employ funds to promote their
esis is that contributions influence legisla- political ends, one should still expect to
tive outcomes by "purchasing" the votes find systematic patterns of allocation that
of particular members or, less directly, by are driven by legislative considerations,


This content downloaded from on Tue, 22 Nov 2016 21:09:36 UTC
All use subject to
American Political Science Review Vol. 84

even among PACs that are highly decen- man and Tierney, 1986). But this solution
tralized (and especially among those that only provokes a second query: If money
are not). Indeed, there is growing evi- buys access, what does access buy? (see
dence that this is the case (Grenzke esp. Herndon, 1982, 1017). Presumably,
1989b). it gives the representatives of contributing
If the principal value of contributionsgroups important opportunities to direct-
lies in their potential to affect floor rollly lobby and potentially persuade legisla-
calls, however, a second puzzle appears. tors to the group's point of view. In this
One would expect to find contribution scenario the language of access may serve
strategies that favor the swing legislators symbolically to launder the money going
in anticipated floor battles, since these are from group to roll call vote, but the effect
the cases where the marginal utility in of the group on the vote should still ap-
votes purchased per dollar spent is likely pear in systematic analysis (Grenzke
to be greatest (Denzau and Munger 1986). 1989a). As we note above, it does not.
Money allocated to almost certain sup-
porters (or almost certain opponents)
should be counted as irrational behavior, The Rational PAC Revised
evidence of scarce resources wasted. In
fact, however, the evidence suggests that The literature on PAC contribution
such "misallocations" systematically oc- strategies and members' roll call voting
cur. The Business-Industry Political Ac- behavior thus suggests two puzzles. First,
tion Committee (BIPAC) and the Nation- if group strategists are reasonably ration-
al Chamber of Commerce give over- al, why would they continue to allocate
whelmingly to conservative Republicans scarce resources to efforts where the ex-
(Kau and Rubin 1982, 88; Maitland 1985).pected political benefits are so low? Sec-
Labor PACs such as the AFL-CIO's Com- ond, if PAC allocation strategies are de-
mittee on Political Education give over- signed to influence members' votes, why
whelmingly to incumbent Democrats do they contribute so heavily to their
loyal to labor's agenda (Chappell 1982; strongest supporters and occasionally to
Grier and Munger 1986; Jacobson 1980). their strongest opponents? Is it the case
Oil PACs give to conservative incum- that PACs are systematically irrational
bents regardless of party and to friends(e.g., Welch 1982, 492) and, by extension,
regardless of ideology (Evans 1988). In that claims about the influence of money
general, PACs are prone to reward their on legislative process almost certainly ex-
friends-even when their friends are not aggerated? We believe that the premise of
in danger of defeat. In a specific test of the rationality need not be rejected but that
swing hypothesis, in fact, Welch found theoretical work in this area requires a
that if anything, dairy PACs were less more complete account of rational PAC
likely to contribute to swing legislators on behavior. We extend here an account
dairy issues, all other things being equal developed formally in Denzau and
(1982).' On the whole, it would seem that Munger's model of a supply price for
if, as Schattschneider (1960) said, public policy (1986). Simply put, interest
moneyed interests sing with an upper- group resources are intended to accom-
class accent, they also spend a good deal plish something different from, and more
of effort singing to the choir. than, influencing elections or buying
One oft-mentioned solution to these votes. Specifically, we argue that PAC
puzzles is that contributions buy not votes money should be allocated in order to
but "access" to members and their staffs mobilize legislative support and demobil-
(e.g., Berry 1984; Gopoian 1984; Schloz- ize opposition, particularly at the most


This content downloaded from on Tue, 22 Nov 2016 21:09:36 UTC
All use subject to
Bias in Congressional Committees

important points in the legislative proc- dollars spent, especially when individual
ess. PAC contributions are limited by the Fed-
This argument turns directly on what eral Election Campaign Act to ten thou-
we already know about the nature of sand dollars-a slight fraction of the cost
legislators' voting decisions from a very of the average House race. Individual
rich literature. The simple but important votes, that is, simply aren't easy to
point is that a number of powerful factors change; and even if some are changed, the
exist that predispose a member to vote a utility of the votes purchased depends on
certain way, among them party leaders, their net cumulative effect in turning a
ideology, constituency, and the position potentially losing coalition into a winning
of the administration (Fiorina 1974; Jack- one. For the rational PAC manager, the
son 1974; Kingdon 1981).2 Kingdon notes, expected marginal utility approximates
moreover, that members' votes on partic- zero in most every case. All other things
ular issues are also constrained by their being equal, scarce resources should be
past voting histories (1981, 274-78). allocated heavily elsewhere and to other
Members attach some value to consis- purposes.
tency, independent of the other factors How, then, should the strategic PAC
that influence their voting behavior. A distribute its resources? The first principle
third and related point is that the public, derives from the larger literature on inter-
recorded nature of the vote may itself est group influence in Congress. Well
limit the member's discretion: a risk- aware of the decentralized nature of con-
averse member may fear the appearance gressional decision making, interest
of impropriety in supporting major cam- groups recognize that resources allocated
paign contributors in the absence of some at the committee stage are more efficiently
other, legitimate force pushing her in the spent (e.g., Berry 1984; Grier and Munger
same direction. Finally, the dichotomous 1986; Kingdon 1981, 170-71). Interest
nature of the vote acts as a constraint. group preferences incorporated there
Money must not only affect members' at- have a strong chance of surviving as the
titudes at the margin but do so enough to bill moves through subsequent stages in
push them over the threshold between the sequence, while provisions not in the
nay and yea. In short, the limits on committee vehicle are difficult to attach
member responsiveness to messages later. Second, the nature of the committee
wrapped in money are substantial, per- assignment process increases the probabil-
haps overwhelming, at least insofar as ity that organized interests will find a
floor voting is concerned. sympathetic audience at the committee or
Of course, almost all studies of PAC subcommittee stage. Members seek and
contributions and roll calls acknowledge often receive positions that will permit
the importance of such factors and build them to promote the interests that, in turn,
them into their statistical models of the help them to get reelected (Shepsle 1978).
voting decision. But it is also important to Finally, the less public, often informal
consider the implications of these findings nature of committee decision making sug-
for the vote-buying hypothesis itself. In- gests that members' responsiveness to
terest group strategists tend to be astute- campaign donors will receive less scru-
enough observers of the legislative proc- tiny. Indeed, a long tradition of research
ess to appreciate the powerful constraints on subgovernments emphasizes that such
that shape members' voting behavior. To clientelism flourishes at the committee
the extent that this is true the rational stage (e.g., Ripley and Franklin 1980;
PAC should expect little in the way of Shepsle 1978, chap. 10; but see Gais,
marginal benefits in votes bought for Peterson, and Walker 1984). In short,


This content downloaded from on Tue, 22 Nov 2016 21:09:36 UTC
All use subject to
American Political Science Review Vol. 84

groups will strategically allocate their simply the direction of legislators' prefer-
resources with the knowledge that invest-ences but the vigor with which those pref-
ments in the politics of the appropriateerences are promoted in the decision mak-
committee or subcommittee are likely to ing process. Such strategies should take
pay higher dividends than investments the form of inducing sympathetic mem-
made elsewhere. Indeed, this principle is bers to get actively involved in a variety
especially important in the House, where of activities that directly affect the shape
the sheer size of the chamber's member- of committee legislation: authoring or
ship, the greater importance of the com- blocking a legislative vehicle; negotiating
mittee stage, and the frequent restrictions compromises behind the scenes, especially
on floor participation recommend a more at the staff level; offering friendly amend-
targeted strategy (see, esp., Grenzke ments or actively opposing unfriendly
1989b and Grier and Munger 1989). ones; lobbying colleagues; planning strat-
If PACs concentrate at the committee egy; and last and sometimes least, show-
level, what, specifically, do they hope to ing up to vote in favor of the interest
gain there? Purchasing votes is one possi- group's position. The purposes of PACs
bility; and, in fact, the rationale for allo- in allocating selective benefits, then, are
cating campaign money to buy votes in analogous to the purposes that Arnold at-
committee is somewhat stronger than for tributes to legislatively strategic bureau-
vote-buying on the floor. But even within crats: the goal is not simply to purchase
committee, PACs still tend to give to their support but to provide incentives for sup-
strongest supporters. In addition, com- porters to act as agents-at the extreme,
mittee votes, like floor votes, are dichoto- to serve as "coalition leaders" on the prin-
mous decisions. And despite the lower cipal's behalf (see Arnold 1979, 40-42 and
visibility of committee decision making, esp. 98-100).
the factors of constituency, ideology, par- Several arguments support this view.
ty, and administration are almost certain- First, participation is crucial to determin-
ly at work. In fact, while research on ing legislative outcomes; and voting is
PACs and committee voting is just now perhaps the least important of the various
beginning to emerge, there is little evi- ways in which committee members partic-
dence that contributions influence votingipate (Hall 1989; Mayhew 1974, 95). Sec-
in committee any more than they do ond, while members' voting choices are
voting on the floor (Wright 1989). highly constrained, how they allocate
The alternative hypothesis that we their time, staff, and political capital is
test here is that political money alters much more discretionary (Bauer, Pool,
members' patterns of legislative involve- and Dexter 1963, 406-7). At any given
ment, a point that emerges from an moment, each member confronts a wide
older literature on interest group in- range of opportunities and demands, the
fluence in Congress (e.g., Bauer, Pool, response to any subset of which will serve
and Dexter 1963; Matthews [1960] 1973, one or more professional goals. To be
esp. 192-93) but is given its fullest theo- sure, the member must choose among
retical expression in the recent work of them. Legislative resources are scarce,
Denzau and Munger (1986). Denzau and and their allocation to one activity results
Munger suggest that interest groups pro- in other beneficial opportunities foregone
vide political resources in an implicit ef- (Bauer, Pool, and Dexter 1963; Hall 1987;
fort to purchase policy-relevant "services"Matthews [1960] 1973, 182-93). But for
from members or their staffs. Stated the most part, the purposive legislator is
somewhat differently, the object of a ra- free to choose among the abundant alter-
tional PAC allocation strategy is not natives with only modest constraints im-


This content downloaded from on Tue, 22 Nov 2016 21:09:36 UTC
All use subject to
Bias in Congressional Committees

posed by constituents, colleagues, or is widely perceived as a legitimate one: in

other actors. Hence, the member's level of asking for help, the group is encouraging
involvement is something that a strategic members to do precisely what they would
PAC can reasonably expect to affect. The do were resources plentiful. In contrast,
contribution need not weigh so heavily in contributions to opponents are meant to
a member's mind that it changes his or her encourage them to go against their predis-
position in any material way; it need only positions: the implicit message is to "take
weigh heavily enough to command some a walk" on an issue that they may care
increment of legislative resources. The about. In short, the expected effects are
minimum threshold that must be passed is not symmetric; the mobilization hypothe-
thus a fairly modest one, and the potential sis is on stronger theoretical ground.
effect of contributions on behavior is one A final advantage of the view of ration-
of degree. Specifically, the member will al action employed here is that it. renders
allocate scarce legislative resources on the the matter of access more comprehensi-
group's behalf so long as the marginal ble. We have already noted that accord-
utility of the contribution to the membering to the standard account of PAC be-
exceeds the expected marginal utility of havior, the importance that both legisla-
the most valuable remaining use of the tors and lobbyists attach to the money-
member's resources (see also Denzau and access connection makes little sense, given
Munger 1986). the evidence that money has little ultimate
A third advantage of this view is that effect on votes. In light of the theory
it explains the ostensibly anomalous tend- sketched here, however, access becomes
ency of PACs to contribute so heavily to an important, proximate goal of the inter-
members who are almost certain to win est group pursuing a legislative agenda.
reelection and almost certain to support Access is central to stimulating agency. It
the group's point of view. Such behavior gives the group the opportunity to let
now appears quite rational. It is precisely otherwise sympathetic members (and
one's supporters that one wants to mobil- their staffs) know that some issue or up-
ize: the more likely certain members are coming activity is important to them. The
to support the group, the more active it ideal response they seek is not simply "I'll
should want them to be. Furthermore, support you on this" but "What can I do
this view of purposive PACs makes sense to help?" Perhaps more importantly, ac-
of the evidence that PACs sometimes con- cess refers to the reciprocal efforts of the
tribute to members who will almost cer- group. It is the pipeline through which the
tainly oppose them and whose involve- group effectively subsidizes the consider-
ment in an issue stands to do the group able time and information costs associ-
harm. The PAC may have no hope of ated with their supporters' participation
changing the opponent's mind, but it in the matters the group cares about. As
may, at the margin at least, diminish the various accounts reveal, group represen-
intensity with which the member pursues tatives often serve as "service bureaus" or
policies that the organization does not adjuncts to congressional staff (e.g.,
like. The intent of the money, then, is notBauer, Pool, and Dexter 1963, chap. 24;
persuasion but demobilization: "We Kingdon 1981, 154-55). They provide
know you can't support us, but please technical information and policy analysis;
don't actively oppose us." However, we they provide political intelligence; they
should not expect the demobilizing effect draft legislation and craft amendments;
of money to be nearly so strong as the they even write speeches or talking points
mobilizing effect. The message provided that their supporters can employ in efforts
through contributions to one's supporters on their behalf. Such subsidies to the


This content downloaded from on Tue, 22 Nov 2016 21:09:36 UTC
All use subject to
American Political Science Review Vol. 84

"congressman-as-enterprise" (Salisbury downward by as much as a dollar per

and Shepsle 1981) do not necessarily per- hundredweight over two years, creating
suade, but they should affect the patterns budget savings of 4.2 billion dollars for
of activity and abdication that have a fiscal years 1983-85 and decreasing the
direct bearing on legislative deliberations profitability of milk production by as
and outcomes (Hall 1987, 1989). much as 30% for the typical dairy farmer.
In each case, then, evidence of the influ-
ence of PAC money on congressional
The Data: decision making can hardly be counted
Money and Mobilization on narrow or trivial. The deliberations in
Three Committees each case bore in significant ways on
major interests, both public and private.
The data for this investigation are A second feature relevant to this inves-
drawn from staff interviews and markup tigation follows from the economic
records of three House committees on importance attached to these issues. All
three issues: (1) the Dairy Stabilization three were salient among actors other
Act, considered by the Agriculture Com- than the private groups immediately af-
mittee in 1982; (2) the Job Training Part- fected, a feature that the considerable re-
nership Act (JTPA), considered by Educa- search on roll call voting suggests should
tion and Labor in 1982; and (3) the depress the effect of PAC contributions
Natural Gas Market Policy Act, consid- on congressional decision making (see,
ered by Energy and Commerce during esp., Evans 1986). This was especially
1983-84. true for the natural gas and job training
Several features of these cases make bills. While the Natural Gas Market
them particularly appropriate for explor- Policy Act never received action on the
ing the effects of money on the participa-House floor in the 98th Congress, it was a
tion of committee members. First, all were highly visible issue while still in commit-
highly significant pieces of legislation, thetee. Consumer interest in the issue of
stakes of each measuring in the billions ofnatural gas pricing was unusually high.
dollars. At issue in the Natural Gas Gas heating costs had been climbing
Market Policy Act was the deregulation quickly in much of the country despite a
of natural gas prices, a proposal that substantial surplus of domestic natural
would transfer billions of dollars from gas (Davis 1984; Murray 1983; Uslaner.
one region to another, from consumer to 1989, chap. 5); and this fact was wide-
industry, and within the industry from in-ly publicized through the efforts of
terstate pipelines and distributors to the the Citizen/Labor Energy Coalition
major natural gas producers (Uslaner (Pressman 1983). The Washington Post,
1989, chap. 5; Maraniss 1983). Annual in turn, gave Commerce Committee delib-
spending on the Job Training Partnership erations front-page coverage, and the
Act was expected at the time of its passage issue was a high priority for the Reagan
to be in the four-to-five-billion-dollar administration. The job training bill, like.
range (Donnelly 1982, 1035), and it re- wise, was one of the most important
placed one of the most important domesticdomestic initiatives of Reagan's first term
programs of the 1970s (Franklin and Rip- and received considerable media atten-
ley 1984). While more narrow than these tion. The principal purpose of the bill was
in scope, the Dairy Stabilization Act also to replace the much maligned but widely
entailed significant economic effects. The used public jobs program, the Compre-
principal purpose of the act was to adjust hensive Employment and Training Act
the scheduled support price for milk (CETA), at a time when the national un-


This content downloaded from on Tue, 22 Nov 2016 21:09:36 UTC
All use subject to
Bias in Congressional Committees

employment rate threatened to exceed each of the cases here commanded the at-
10% for the first time in four decades. To tention of a wide range of political actors.
a lesser degree, finally, the 1982 dairy bill Second, past research suggests that we
was also salient among actors off the com- will find little PAC influence in precisely
mittee and outside the industry. While thethese three policy areas. Should we find
interest of the general public in dairy support for the hypothesis that money
policy was slight, the burgeoning budget mobilizes support (or demobilizes opposi-
deficit loomed large on Capitol Hill, and tion) at the committee level, we should be
it clearly motivated the decision to change on reasonably solid ground to conclude
dairy policy only one year after passage that (1) the results of this exploration are
of an omnibus farm bill (Wehr 1982a, apt to generalize to other committees and
1982b). Indeed, relative to other domestic other issues and (2) the null results of past
nonentitlements, dairy subsidies were research are more likely to be artifacts of
widely perceived as a major budget of- the legislative behavior and the legislative
fender. The administration thus counted stage studied than evidence that moneyed
the price adjustments a high priority, one interests do not matter in congressional
that commanded considerable attention decision making.
from Budget Director David Stockman,
and the House Budget Committee was
involved at every stage of the process. The Model
Finally, each of the policy areas we ex-
amine here has received the attention of The model of participation we use to
previous scholars studying PAC contribu- test for the hypothesized effects is adapted
tions and floor roll calls; and in each case from Hall 1987.3 The model begins from
the effects of PAC money were found to the same motivational premise that we
be slight. In a study of dairy legislation employed in our discussion of PAC con-
considered in the House in 1975, for in- tribution behavior. Members of Congress
stance, Welch (1982) concluded that dairy are purposive actors who allocate their
PAC contributions were the least impor- time, staff, and other legislative resources
tant determinant of voting on milk price in such a way as to advance certain per-
supports and that their effect on the legis- sonal goals or interests. There are several
lation was negligible (see also Chappell goals that commonly figure in these calcu-
1982). Grenzke (1989a) estimated a dy- lations. The one most prominently cited
namic model of members' voting behavior in the literature on legislative behavior is
over four congresses and found that labor reelection or, more generally, service to
union contributions had either a negli- the district (see, esp., Mayhew 1974); but
gible or a negative effect on members' we report elsewhere that the relevance of
propensity to take prolabor positions on any particular goal to a member's partici-
the House floor (but see Wilhite and pation depends directly on the nature of
Theilmann 1987). And Wayman and the issue and the legislative context (Hall
Kutler (1985) found no effect of natural 1987). To use language borrowed from
gas industry campaign contributions onKingdon (1981), goals are "evoked." Any
members' votes during House considera- particular issue may evoke several goals
tion of natural gas deregulation in 1975. simultaneously or may evoke none at all.
At two levels, then, past research indi- In the latter case, a member is simply un-
cates that our selection of cases is biased interested, the expected benefits of partic-
against our argument. It suggests that ipation slight; in the former, the level of
high salience issues should exhibit little interest is intense, the expected benefits of
PAC influence on legislative behavior, yet participation high.


This content downloaded from on Tue, 22 Nov 2016 21:09:36 UTC
All use subject to
American Political Science Review Vol. 84

In the three cases under study here, greater

in staff allocations that these posi-
fact, several goals were probably at work
tions bestow, the procedural control over
in the resource allocation decisions of the agenda, and the central place in the
most committee members. For instance, committee communication network
the natural gas bill raised issues of govern- diminish the time and information costs
ment intervention in the economy and the associated with meaningful involvement
country's long-term dependence on in the issue at hand. Finally, freshman
foreign energy sources. The budgetary status tends to increase the information
implications of the dairy bill undoubtedly costs and diminish the opportunities or
evoked some committee members' con- resources a member enjoys for any partic-
cerns about good fiscal policy and its ular bill.
macroeconomic consequences. The Job The variable of greatest interest in this
Training Partnership Act concerned the investigation, however, is the level of
government's obligation to redress ine- contributions each member receives from
qualities of economic opportunity result- PACs interested in the issue at hand. To
ing from inadequate or outdated job what degree, that is, does money affect
members' decisions regarding whether
skills. But the goal most consistently evi-
dent in staff interviews, markup debates, and to what extent they will participate in
and secondary accounts of the three bills the committee deliberations? Two points
require emphasis here. First, the foregoing
was promoting or protecting district inter-
ests. For the purposes of this analysis, discussion suggests that the effects of
then, we adopt the simpler and more trac- money on participation should not be
table motivational assumption common simply linear. The positive effect of con-
to most models of legislative behavior.4 In tributions on participation should be con-
deciding whether and to what extent to tingent on probable support; this is the
participate on a particular issue, the mem- mobilization hypothesis. To the extent
ber estimates both the expected benefits that contributions are given to probable
and expected costs, where benefits are a opponents, on the other hand, they
direct function of the issue's economic should diminish participation; this is the
relevance to the districts demobilization hypothesis.
If the interests of one's constituents Second, contributions may well be
motivate a member to become involved, related to other activities that moneyed
the costs of participation are also impor- interests employ to further their legisla-
tant and highly variable: resources are tive aims, making it difficult to isolate the
scarce, and the allocation to one activity effects of any particular part of their ef-
results in other profitable opportunities fort (Rothenberg 1989; Wright 1989). For
foregone. Several factors affect the re- instance, it may be the case that those
sources available to particular members groups that organize PACs for the pur-
on particular issues. First, assignment to pose of channeling money to candidates
the subcommittee of jurisdiction provides are also the most active in developing
members both with greater formal oppor- grass roots campaigns or direct lobbying
tunities to participate and access to an efforts. While there is evidence to suggest
earlier stage of the sequential process. It that the correlation among these activities
also gives the member greater access to is modest for the cases under study here,6
staff and to lines of communication with our data on interest group activity are
other interested actors both on and off the limited to political action committee cam-
committee. For similar reasons, a commit- paign contributions. Hence, while our
tee or subcommittee leadership position model tests for the effect that money has
subsidizes participation even more. The on committee behavior, one might more


This content downloaded from on Tue, 22 Nov 2016 21:09:36 UTC
All use subject to
Bias in Congressional Committees

accurately characterize our results as cap- creases and census data regarding con-
turing the effect of the several resources gressional district natural gas use.10 If
that moneyed interests employ.7 high production and high inflation cap-
The dependent variable is the participa- ture dimensions of intradistrict salience,
tion of member i on bill j, where partici- however, the presence of both at once
pation refers to a member's activity both should produce intradistrict conflict. The
during formal committee markups and member is torn between two significant
committee action behind the scenes. Our economic interests, and activity on behalf
data on activity are drawn from two of one may alienate the other. Indeed,
sources: semistructured interviews with Fiorina (1974) suggests that unrequited
both the majority and minority staffers constituents are likely to punish more
assigned to cover each bill and the largely than the requited are to reward. As intra-
unpublished but meticulously kept com- district conflict increases, in any case, the
mittee and subcommittee markup expected benefits of activity on the issue
records. The summary measure of partici- should diminish, ceteris paribus. In the
pation that we use for the purposes of natural
this gas case, then, intradistrict con-
exploration is a simple scale score derived flict occurs as the production and infla-
from a factor analysis of six activities: at- tion variables both approach their upper
tendance; voting participation; speaking; limits. We measure this condition as the
offering amendments during committee product of two terms: "high production"
markups; role in authoring the legisla- is the extent to which natural gas produc-
tive vehicle or an amendment in the tion in the district exceeds the mean dis-
nature of a substitute; and negotiating trict production for all members of the
behind the scenes at either the member orcommittee; similarly, "high inflation" is
the staff level.8 The measurement of the the extent to which the district inflation-
independent variables, in turn, follows ary effect exceeds the mean for all com-
directly from the preceding discussion. mittee members. When either district gas
Members' institutional positions and production or inflationary effect is below
status are measured with dichotomous the committee mean, then, intradistrict
variables that are set at zero except as the conflict is zero.
following conditions hold: subcommittee In the other two cases the measurement
membership takes a value of one if a of district interest is uncomplicated by
member sat on the subcommittee with potential conflicts within members' geo-
jurisdiction over the bill; leadership posi- graphic constituencies. In the dairy stabil-
tion takes a value of one if a member was ization case district relevance is directly.
chair or ranking minority member of related to the importance of dairy farm-
either the full or subcommittee; and fresh- ing, measured simply by the total number
man status takes a value of one for mem- of dairy cows in the member's district as
bers in their first term in the House. reported by the United States Department
In measuring the relevance of each issue of Agriculture biennial census. Given that
to committee members' districts, we milk prices were not a salient consumer
assume that relevance is primarily issue per se and that the Dairy Stabiliza-
economic in nature. In the natural gas tion Act was not likely to affect retail
case, this takes two quite different forms: prices in any significant way, we do not
total district-level natural gas productions assume a more general public concern
and the economic effect of gas price in- with this issue. For the Job Training Part-
creases on residential consumers in the nership Act, likewise, district relevance is
member's district, which we measure us- directly related to the importance of fed-
ing industry data on natural gas price in- eral jobs programs in addressing struc-


This content downloaded from on Tue, 22 Nov 2016 21:09:36 UTC
All use subject to
American Political Science Review Vol. 84

tural unemployment, which we measure cannot assume that a particular industry

as the current level of CETA expenditures is necessarily unified, however: one seg-
in the member's district." This variable ment of an industry may have different
not only taps the district-specific eco- interests and work in ways that offset
nomic benefits of clients of the expiring some other segment. In the case of the
job training program but (given that federal dairy legislation, no such split
CETA allocations were directly tied to within the industry was apparent among
local unemployment rates) also captures the principal actors, thus permitting the
the severity of structural unemployment fairly simple specification described
in the district. above. But in general-and in the natural
Consistent with the preceding theoreti-gas case in particular-an industry may
cal discussion, we estimate the effect of not be so easily simplified. While the gas
group expenditures on participation by producers were by far the most visible
including pairs of interactions between and most vigorous among the corporate
group contributions (measured as the actors and gave by far the most money in
amount contributed during the two-year campaign contributions among energy
election cycle prior to committee action) PACs, the natural gas industry was seri-
and indicators of probable support or op- ously divided (Pressman 1983; Uslaner
position. For each case, the exact specifi- 1989, chap. 5), a feature that we attempt
cation of the interactions is straightfor- to capture. The alignments were by no
ward. In the dairy stabilization case, we means perfect, but the principal issues at
measure probable support or opposition stake in the legislation before House
using the ratings of the National Farmers' Energy and Commerce pitted the major
Union (NFU),12 an organization that gas producers. and intrastate pipelines
strongly supports federal intervention in against the interstate pipelines and distrib-
the agricultural economy to control sup- utors. As a result, different segments of
ply and support the commodity prices the industry were likely to target different
paid to farmers. Given that we expect members to serve as legislative agents and
very different effects for contributions on identify different members as their likely
the behavior of likely supporters and op- opponents. Our first task therefore was to
ponents, however, the model requires two distinguish the various energy PACs ac-
separate interactions: Money to support- cording to the principal business activities
ers is the product of contributions'3 and of their affiliates. Using the detailed
the member's distance from the mean descriptions of individual companies pro-
NFU score where the members' rating isvided by Moody's Investor Service (1983a
greater than the mean; the money-supportand 1983b), we classified each affiliate
term is zero otherwise. Money to oppo- according to its principal interests in the
nents is the product of contributions and natural gas area.'4 We then divided the
the member's distance from the mean contributions a member received accord-
NFU score, where the member's rating is ing to whether they came from producers
less than the mean; the money-opposition or intrastate pipelines on the one hand
term is zero otherwise. Following the and interstates or distributors on the
theoretical reasoning of the last section, other. The measure of contributions that
then, the expected effect on participation we employ, then, is the producer-intra-
is positive for money to supporters. The state contributions minus the interstate-
expected effect is negative for money to distributor contributions, the value of
opponents in each case. which was positive in almost every case.
Any attempt to model the effect of con- The operationalization of the interac-
tribution activity on legislative behavior tions tapping the net producer-intrastate


This content downloaded from on Tue, 22 Nov 2016 21:09:36 UTC
All use subject to
Bias in Congressional Committees

effects, in turn, was handled in a fashion favored an expanded role for private in-
analogous to the dairy stabilization case. dustry councils so that federally subsi-
In the natural gas case, however, mem- dized training would be tailored to meet
bers' Americans for Democratic Action the changing needs of the private sector
(ADA) scores were more appropriate as (Baumer and Van Horn 1985, 173). As in
an indicator of likely support or opposi- the natural gas case, we thus employ a net
tion. For the producer-intrastate segment contributions variable, which takes the
of the gas industry at least, the issue of value of the member's total labor contri-
greatest concern was the extent to which butions less the total contributions re-
the government continued its intervention ceived from national business organiza-
in the natural gas market by controlling tions.16 As in the other two cases, like-
the price of old gas. The ADA score wise, the indicator of probable support or
should tap members' historical tendency opposition was constructed using the ap-
to support such federal interventions propriate group rating, in this case, the
quite well. Money to supporters, then, is AFL-CIO's Committee on Political Educa-
the product of net producer-intrastate tion (COPE) score. Money to supporters,
contributions and the member's distance then, is the product of net labor contribu-
from the mean ADA score where the tions and the member's distance from the
member's rating is less than the mean; the mean COPE score, where the member's
money-support term is zero otherwise; rating is greater than the mean; and
and money to opponents is the product of money to opponents is the product of
contributions and the member's distance contributions and the member's distance
from the mean ADA score where the from the mean COPE score where the
member's rating is greater than the mean; member's rating is less than the mean.
the money-opposition term is zero other-
Unlike the dairy and natural gas cases, Results and Interpretations
finally, the job training bill did not in-
volve issues specific to a particular indus-In estimating the model of participa-
try. The organized interests most con- tion, we explicitly account for the possi-
cerned with CETA and its prospective bility that contributions are effectively en-
replacement were the national labor dogenous, that is, that in allocating con-
unions: public service employment and tributions to committee members during
training programs were at the top of the previous election cycle, a group may
labor's agenda, especially in 1982, when attempt to anticipate who the principal
unemployment was approaching postwar players will be on issues it cares about.17
records. Moreover, labor unions were one To the extent this is true, at least, the er-
of the single largest categories of contrib- ror term will be correlated with contribu-
utors to congressional campaigns and tions and the ordinary least squares coef-
gave to five-sixths of the members of ficient on the latter will be upwardly bi-
House Education and Labor. It is the ef- ased. We thus estimate the participation
fect of these contributions on committee model using two-stage least squares, with
behavior with which we are primarily the second stage results reported in the
concerned. This is not to say, however, tables.18 In each of the three cases, the
that labor unions were the only groups in- model performs quite well, explaining
terested in mobilizing support on this over 55% of the variance in participation.
bill.15 On the business side, national busi- More importantly, the analysis provides
ness associations generally opposed any solid support for the principal hypothesis
public service employment provisions and of this study, that moneyed interests


This content downloaded from on Tue, 22 Nov 2016 21:09:36 UTC
All use subject to
American Political Science Review Vol. 84

mobilize bias in committee decision mak-insignificant, its size is substantively non-

ing. trivial, and the negative sign is consistent
with the demobilization hypothesis. In
This finding is clear for all three cases. 19
short, the more money a supporter re-
The campaign contributions that dairy in-
ceived from the dairy PACs and the
dustry PACs gave to their likely support-
stronger the member's support, the more
ers significantly increased their participa-
likely he or she was to allocate time and
tion, even when we controlled for the im-
portance of the issue to individual mem-effort on the industry's behalf (e.g., work
bers' districts, whether they sat on thebehind the scenes, speak on the group's
subcommittee of jurisdiction, and behalf, attach amendments to the com-
whether they held a leadership position mittee vehicle, as well as show up and
(Table 1). Such factors are reported else-
vote at committee markups). Alternative-
where to be strong determinants of com- ly, money may have diminished the inten-
mittee participation (Hall 1987), and each sity of the opposition. The effect of
money on decision making in the House
is also likely to affect contributions since
interest groups tend to concentrate their Agriculture Committee, then, was to en-
resources on members who hold positions courage industry supporters to be active
of institutional power (e.g., Grenzke and, if anything, to encourage industry
1989b; Grier and Munger 1986, 1989), as opponents to abdicate.
well as on members who have a district The results of the job training case are
stake in their industry. That the mobiliza- also clear, and the specific estimates are
tion coefficient remains positive and sig- striking in their similarity to the dairy
nificant in the face of the multivariate stabilization case. As Table 2 shows, the
controls reinforces the interpretation that contributions that labor groups made to
the connection between group resources their supporters had a substantial, statisti-
and mobilization is causal. When dairy cally significant effect on participation
PACs did give to their probable oppo- during Education and Labor delibera-
nents, moreover, there is some evidence
tions. Remarkably, the unstandardized
that the contributions diminished partici-coefficient for the money support variable
pation. While the coefficient on the is almost identical in size to the analogous
money-opposition variable is statistically coefficient in the dairy stabilization model

Table 1. PAC Money and Committee Participation:

1982 Dairy Stabilization Act

Independent Variables 2SLS Coefficient t-statistic

Intercept .01 .05

Number of dairy cows in district .27** 2.21
Dairy PAC contributions to supporters .26** 2.42
Dairy PAC contributions to opponents -.11 -.61
Membership on reporting subcommittee .17** 3.54
Committee or subcommittee leadership position .35** 4.50
Freshman status -.02 -.31

Note: Adjusted R-squared = .6

The contributions term is the p
**Statistically significant at .05


This content downloaded from on Tue, 22 Nov 2016 21:09:36 UTC
All use subject to
Bias in Congressional Committees

Table 2. PAC Money and Committee Participation:

1982 Job Training Partnership Act

Independent Variables 2SLS Coefficient t-statistic

Intercept .13 .77

CETA expenditures in district .03 .23
Labor union net contributions to supporters .25* 1.62
Labor union net contributions to opponents -.18 -.80
Membership on reporting subcommittee .19** 2.61
Committee or subcommittee leadership position .47** 4.55
Freshman status -.05 -.51

Note: Adjusted R-squared = .5

The net contributions term is
*Statistically significant at .10
**Statistically significant at .0

despite the fact that the two cases are likely to mitigate the efficacy of interest
drawn from different committees with group efforts, and they complicate the
qualitatively different jurisdictions and measurement of anticipated support and
policy environments (Smith and Deering opposition. Still, the mobilization
1984). In each case, a change in the money hypothesis finds strong support in the
support variable from its minimum to its behavior of Energy and Commerce mem-
maximum value moves a member approx- bers. While the size of the unstandardized
imately one-fourth of the way along the coefficient for the money support variable
participation scale, almost exactly one is somewhat smaller than for the other
standard deviation. In both cases, like- two cases, it is still substantial and statisti-
wise, this coefficient is greater than thatcally significant at the .05 level. A change
for subcommittee membership, a variable in the money support variable from its
generally considered central to under- minimum to its maximum moves a Com-
standing participation in the postreformmerce Committee member approximately
House. As in the dairy stabilization case, one-sixth of the way along -the participa
finally, the Education and Labor bill pro- tion scale. By way of illustration, this
vides some support for the demobilization amounts to the difference between Minne-
hypothesis. While it fails to meet conven- sota Representative Gerry Sikorski, who
tional levels of statistical significance, the did little more than faithfully attend and
size of the money-opposition term proves vote during formal markups, and
negative and substantively significant, Alabama Representative Richard Shelby,
nearly matching the size of subcommittee whose staff participated in behind-the-
membership. scenes negotiations and who offered two
The results regarding moneyed interests substantive amendments during subcom-
and mobilization are only slightly less mittee markup, both of which passed.
compelling in the natural gas case, a case As Table 3 shows, finally, the demobil-
complicated both by divisions within the ization hypothesis is not supported in the
industry and the apparent importance of natural gas case. While the coefficient on
both organized and unorganized interests. the money opponents interaction is slight,
As we note above, such conditions are its positive sign is inconsistent with our


This content downloaded from on Tue, 22 Nov 2016 21:09:36 UTC
All use subject to
American Political Science Review Vol. 84

Table 3. PAC Money and Committee Participation:

1984 Natural Gas Market Policy Act

Independent Variables 2SLS Coefficient t-statistic

Intercept .08 .40-

Natural gas production in district .32* 1.65
Natural gas price increase effect on district .17* 1.35
High production/high inflation interaction -.18 -1.28
Producer-intrastate net contributions to supporters .17** 1.69
Producer-intrastate net contributions to opponents .01 .06
Membership on reporting subcommittee .23** 3.17
Committee or subcommittee leadership position .54** 4.77
Freshman status .13* 1.31

Note: Adjusted R-squared - .

The net contributions term is t
*Statistically significant at .10
**Statistically significant at .0

prediction. The foundation for the Gas Market Policy Act. Indeed, a change
demobilization hypothesis being theoreti- in gas production from its minimum to its
cally weaker, however, the null result maximum corresponds to a 32% change
here, as well as the weak results in the along the participation scale, the differ-
dairy and job training cases, are not alto- ence between simply showing up and
gether surprising. The theoretically being a major player on the bill. By com-
stronger hypothesis, that money mobil- parison, however, the effect of natural gas
izes a pro-PAC bias at the committee price increases on district consumers ap-
level, is confirmed in all three. pears smaller by half. And the importance
For the most part, the other variables inof structural unemployment and program
the model also perform as predicted and spending in the districts of Education and
suggest interesting implications for the Labor members had at best a slight effect
politics of representation in a decentral- on their involvement in the Job Training
ized Congress. The relevance of an issue Partnership Act.
to the member's district enhances member Pending better measurement of unorga-
participation in two cases, providing evi- nized constituents' interest at the district
dence that Agriculture and Commerce level, of course, we cannot draw unquali-
members purposively allocate their legis-fied conclusions regarding their impor-
lative time and resources to promote the tance in shaping committee behavior.
interests of their constituencies. On House Should such patterns hold up under sub-
Agriculture, the more important dairy sequent analysis, however, the implica-
farming was to the member's district, tions for member responsiveness to indus-
the more likely he or she was to partici- try interests and industry money relative
pate in committee deliberations. Likewise, to more general constituency concerns
the greater the presence of natural gas would be several and important. If mem-
production in the district, the more likely bers allocate their scarce legislative time
the Energy and Commerce member was to and resources with district interests in
participate in deliberations on the Natural mind, they perceive their districts in terms

8 1

This content downloaded from on Tue, 22 Nov 2016 21:09:36 UTC
All use subject to
Bias in Congressional Committees

of different constituencies; and these per-coefficient was unchanged.

ceptions affect their behavior as represen- Finally, most of the variables that tap
tatives (Fenno 1978). In part, the results members' institutional positions prove to
presented here suggest that organized eco- be strong determinants of committee par-
nomic interests within districts figure ticipation. While the coefficients on fresh-
more prominently in the psychology of man status differ in sign, both subcom-
representation than the diffuse and unor- mittee membership and leadership posi-
ganized interests of rank-and-file voters. tion are positive, statistically significant,
Such was the charge that Schattschneiderand substantively large in all three cases.
made thirty years ago, one which critics Even on issues that are widely perceived
of pluralism have echoed repeatedly among the committee membership to be
since. important, issues where the organized
At the same time, however, the findings interests in the policy environment are
in the natural gas case also suggest that themselves active, the opportunities and
the preferences of unorganized interests resources provided by formal institutional
sometimes constrain the responsiveness ofposition are major factors in determining
members to organized groups, thus con-who makes the laws at the committee
firming the thesis of Denzau and Munger stage. Such findings are generally consis-
(1986) regarding how unorganized inter-tent with findings from other committees
ests get represented. Beyond the positive and larger samples of issues (Evans n.d.;
coefficient for the inflationary effect vari-Hall 1987, 1989; Hall and Evans 1990) and
able, this is evident in the size and signifi-
reinforce the assumption that the model
cance of the coefficient on the high pro- of participation employed here is specified
duction-high inflation interaction. Even ifcorrectly.
members are inclined to respond to pro-
ducer interests, in short, this tendency is
mitigated when consumer interests are Conclusion
also high. However, we should point
out two things. First, the simultaneous We have elaborated a theory of the
occurrence of both strong producer inter-member-group exchange relationship that
ests and high consumer-voter salience is comprehends the general patterns of PAC
rare. Indeed, this distinguishes the natu- contributions reported in the literature.
ral gas issue from most of the issues withHouse members and interest group repre-
which members of Congress typically sentatives are parties to an implicit coop-
deal, and even in this case only 4 of the 42 erative agreement, but the constraints on
members of Energy and Commerce were member behavior and the rational calcu-
seriously cross-pressured. Second, we lations of group strategists limit the extent
found no such constraint on the behav- to which votes become the basis for ex-
ioral effect of producer contributions. change. This view suggests expectations
One might expect, for instance, that the about the effects of money on congres-
mobilizing effect of producer contribu- sional decision making quite different
tions would be diminished for a member from the ones that motivate the substan-
who also represents a high inflation dis- tial research on the subject. We should
trict. In one variant of the model tested find little causal connection between con-
here we included an interaction between tributions and votes, especially on the
the money support and high inflation floor-an expectation generally sup-
variables, with the result that the coeffi- ported, although not adequately ex-
cient was correct in sign (negative) but plained, in the literature. We should ex-
very near zero and the money-support pect to find an important connection be-


This content downloaded from on Tue, 22 Nov 2016 21:09:36 UTC
All use subject to
American Political Science Review Vol. 84

tween contributions and the legislative in- Mayhew has suggested (1974, 95), parlia-
volvement of sympathetic members, espe- mentary suffrage gives a member rela-
cially in committee-a relationship that tively little leverage over the shape of
empirical research to date has altogether legislation, especially at the committee
ignored. stage. Only a small fraction of the deci-
In order to test this view of moneyed in- sions that shape a bill ever go to a vote,
terests and congressional decision mak- either in committee or on the floor. The
ing, we investigated the participation of vast majority are made in authoring a
House members on three issues in three legislative vehicle, formulating amend-
committees. In each case, we found solid ments, negotiating specific provisions or
support for our principal hypothesis: report language behind the scenes, devel-
moneyed interests are able to mobilize oping legislative strategy, and in other ac-
legislators already predisposed to support tivities that require substantial time, in-
the group's position. Conversely, money formation, and energy on the part of
that a group contributes to its likely oppo- member and staff. While such efforts by
nents has either a negligible or negative ef- no means guarantee that a particular
fect on their participation. While previous member will influence the final outcome,
they are usually a precondition for such
research on these same issues provided lit-
tle evidence that PAC money purchased influence (Hall 1989).
members' votes, it apparently did buy the A second and related implication of this
marginal time, energy, and legislative investigation, then, is that empirical re-
resources that committee participation re- search should expand its view of the legis-
quires. Moreover, we found evidence that lative purposes of political money and the
(organized) producer interests figured other group resources that may accom-
more prominently than (unorganized) pany it (see also Salisbury 1984, esp.
consumer interests in the participation 70-72). We focus here on committee par-
decisions of House committee members- ticipation; but the more general implica-
both for a case in which the issue at stake tion is that group expenditures may do
evoked high district salience and one much more than buy votes, or they may
where it did not. And we found little evi- buy votes under certain conditions and af-
dence that committee members respond to fect other forms of legislative behavior
the interests of unemployed workers ex- under others. Such a suggestion, of
cept insofar as those interests might be course, usually appears in the various
represented in the activities of well- studies that examine the relationship be-
financed and well-organized labor unions. tween contributions and floor roll calls,
Such findings suggest several implications but it needs to be elevated from the status
for our understanding of political money, of footnote or parenthetic remark to a
interest groups, and the legislative central element of future research designs.
process. Even for a small set of issues and a single
The first and most important implica- group, the legislative strategies available
tion is that moneyed interests do affect theare several, sometimes mixed. To specu-
decision-making processes of Congress,late beyond the research reported here,
an implication that one does not easily for instance, we believe groups allocate
derive from the existing political science their various resources (1) to mobilize
literature on contributions. In fact, it mat- strong supporters not only in House com-
mittees but also on the Senate floor, ir
ters most at that stage of the legislative
process that matters most and for a form dealings with executive agencies, and ir
of legislative behavior likely to have a various other decision-making forum:
direct bearing on outcomes. As David relevant to the group's interests; (2) t(


This content downloaded from on Tue, 22 Nov 2016 21:09:36 UTC
All use subject to
Bias in Congressional Committees

demobilize strong opponents; and (3) to legislative deliberations; and the resources
effect the support of swing legislators. We of moneyed interests at least partly deter-
require greater knowledge of the frequen- mine the weights.
cy and efficacy of such strategies, in any The extent to which such efforts are
case, before we denigrate the role of damaging to representative government,
moneyed interests in Congress, especially as Schattschneider claimed, depends in
when the overwhelming weight of the evi- part on the balance of interests and re-
dence provided by Washington journal- sources apparent in the relevant set of
ists and political insiders suggests that groups that are organized for political ac-
they matter a great deal. tion. On any given issue, the efforts of
Finally, the argument presented here one interest to mobilize supporters in
provides a very different slant on the role Congress may be at least partially offset
of interest groups as purveyors of infor- by the efforts of some competing group to
mation in the deliberations of representa- mobilize its own supporters; indeed, there
tive assemblies. A common defense of is some evidence that such countervailing
group lobbying activity, in fact, is that itefforts occurred in the natural gas case.
provides ideas and information although But for those who believe that money is
its effect on member preferences is slight. an illegitimate resource in such efforts-
Members (and their staffs) tend to con- that pluralism requires something more
sume information selectively, relying on than a competition among moneyed inter-
sources with whom they already agree ests-the results of this study can only be
and discounting sources with whom they disturbing.
usually disagree (e.g., Milbrath 1963).
The view that we have advanced here sug-
gests that while this may in fact describe
how such information is used, it does not This research was supported in part by the Na-
render it inconsequential. In light of the tional Science Foundation under Grant
extraordinary demands on each congres- SES-8401505. For assistance or comments at vari
stages of this paper, we are indebted to Severin
sional office, information-gathering it; Borenstein, John Chamberlin, Cary Coglianese,
analyzing it; turning it into speeches, David C. King, John Kingdon, Tim McDaniel, Mike
amendments, and bills; using it to develop Munger, Ken Organski, Randall Ripley, Robert
legislative strategy-can be very costly. Salisbury, Eric Uslaner, Carl Van Horn, Jack
Wright, and participants in a faculty seminar at the
Such costs, more than anything, limit the
Institute of Public Policy Studies, University of
extent to which a nominal member will be Michigan, Ann Arbor. For assistance in collecting
a meaningful player in the decision- and coding data, we thank Nick Greifer, Ed Kutler,
making process on a particular bill. At the
Gary Levenson, and Dan Polsky. An earlier version
very least, then, money-induced activityof this paper was presented at the 1989 meeting of
the Midwest Political Science Association, Chicago.
will distort the "'representativeness of1. Rothenberg (1989) finds that in the allocation of
deliberations," a standard that democratic
lobbying resources on the MX missile issue Common
theorists since John Stuart Mill have used
Cause did concentrate more on likely "fence strad-
to evaluate the legitimacy of legislative
dlers." By extension, his analysis provides an excel-
lent guide for modeling the effect of expected voting
assemblies (Chamberlin and Courant
behavior on contributions. See Smith 1984 for an
1983). But it may also affect the "repre- important formulation of this argument.
sentativeness of decisions." By selectively 2. Kingdon found that there was no conflict in the
subsidizing the information costs associ- member's "field of forces" in almost half of the im-
ated with participation, groups affect the portant votes that members cast on the House floor.
In an additional 33% of the votes all of the personal
intensity with which their positions are goals that were relevant to a vote pointed the
promoted by their legislative agents. In member in the same direction (1981, 255). While his
short, not all preferences weigh equally in study was conducted before the precipitous rise of


This content downloaded from on Tue, 22 Nov 2016 21:09:36 UTC
All use subject to
American Political Science Review Vol. 84

PAC contributions, Kingdon found little evidence egy was only feasible in districts that had a signifi-
of group influence on members' voting decisionscant number of dairy producers, and the correlation
(chap. 5). between district dairy production and dairy industry
3. In adapting our model from Hall 1987, we re- PAC contributions was only .09 for the period 1979-
tained only the variables that were found to be con- 80. Likewise, while the dairy industry gave a great
sistently significant and collapsed the several specific deal of money to some House Agriculture members
leadership positions into a single dichotomous and none to others, the National Milk Producers by
variable. themselves contacted every committee member
4. This assumption is especially valid for the dairy regarding the dairy stabilization bill, either through
and natural gas cases, though somewhat problem- letters to the member's Washington office or meet-
atic for the JPTA. Like so many of the issues thatings with the member or the member's staff.
come before the House Agriculture Committee, the 7. On this point, we are especially indebted to
dairy program is a classic constituency issue. If conversations with Jack Wright.
anything, the administration's assault on the price 8. Data on the first four activities were taken
levels intensified such interests in the minds of the directly from the committee and subcommittee
legislators. Likewise, the Natural Gas Market Policymarkup minutes and transcripts. Indexes of author-
Act evoked strong sentiment among consumers, dis-ship role and behind-the-scenes participation were
tributors, and pipelines in some states and producerscoded on four-point scales from semistructured in-
in others, sentiments that were loudly communi- terviews with both minority and majority staffers
cated to their representatives in Washington (Mur- who had primary responsibility for staffing the bill
ray 1983). The resulting regional split within the under study. On the collection and coding of these
committee was noted at length in virtually every data, see Hall 1987, though the data-reduction tech-
account of its deliberations (see e.g, Maraniss 1983; nique used here loses less information than the Gutt-
Murray 1983). man scale scores and the informal participation in-
5. In attempting to capture the representation of dexes that were used in that analysis. The factor
constituency interests, however, we necessarily analysis that generated the scores retained only one
neglect constituents' preferences regarding the publicfactor using conventional methods, the weights
goods dimensions of each of these bills. On the im- assigned were similar across the three cases, and
portance of public goods preferences to political (most importantly) the ordinal ranking of the
representation, see J. Jackson and King (1989). weights for each activity were precisely those
6. The principal grass roots campaigns in the hypothesized in Hall 1987. In addition to the results
natural gas case were conducted by the Citizen/ reported here, however, we also estimated the model
Labor Energy Coalition (CLEC) (which conducted using both the Guttman scales and the informal par-
door-to-door efforts in a number of states) and the ticipation index as well as a simple summary of the
public utility companies (who used inserts in month- two. These several measures of participation are all
ly utility bills to encourage their customers to write highly correlated, and various estimates of the
letters to their representatives). Both were also ac- model using them generally confirmed the findings
tively engaged in lobbying members of the Energy that we report here. Problems of measurement un-
and Commerce Committee. (Indeed, the CLEC was doubtedly remain, however; and addressing them is
one of the most vigorous in this respect; see Press- an important matter for future work.
man 1983.) Neither of the two were major campaign 9. The measure of district natural gas production
contributors, however. The CLEC did not have an was constructed from county-level data acquired
organized PAC, and of the various segments of the directly from state departments of natural resources.
gas industry the utility companies contributed rela- Where counties were not wholly contained within a
tively little money. (The major gas producers, for in- single district, the proportion of natural gas produc-
stance, contributed more than the distributors by a tion credited to particular districts was estimated by
factor greater than seven to one.) Similarly, there comparing congressional district maps with the
were dozens of groups active in lobbying on the Job geologic surveys showing the geographic location of
Training Partnership Act that contributed little or natural gas production within counties. The produc-
nothing in the way of campaign money, including tion data are for the year 1983, the year in which the
various public interest groups, state and local offi- Energy and Commerce Committee began considera-
cials, education organizations, and the National tion of the Natural Gas Market Policy Act.
Governors' Association (Baumer and Van Horn 10. District-level data on natural gas price changes
1985). The correlation between contributions and were not immediately available, but the intrastate
other interest group activities is probably higher for variations should be sufficiently small as to make the
the dairy stabilization case, but even here it should state-level data reasonable approximations of the in-
be fairly modest. The various dairy organizationsflation in district natural gas prices. However, there
were in fact active in getting local dairy producers are to dramatic variations in the use of natural gas
write letters and meet with their representatives dur-from one district to the next, so that the economic ef-
fect of a given price increase on residential energy
ing visits to the district. But such a grass roots strat-


This content downloaded from on Tue, 22 Nov 2016 21:09:36 UTC
All use subject to
Bias in Congressional Committees

consumers may vary dramatically across disticts their natural gas revenues. Such information permit-
within a state (e.g., many rural districts that de- ted an unambiguous classification in almost every
pended primarily on fuel oil for home heating were case. Natural gas peak associations were categorized
virtually unaffected by major increases in the price according to the nature of the businesses they repre-
of natural gas). Fortunately, however, district-level sented. In addition, some of the classifications were
information about household fuel use is available. In checked against similar classifications made by Eric
order to create a district-level indicator of consumer Uslaner using both interview and archival data. In
interest, then, we simply multiplied the state-level every case where our data overlapped, our classifi-
price increase for 1981-82 times the percentage of cations matched his (see Uslaner 1989).
households in the district that used natural gas for 15. Other groups interested in the legislation in-
their home heating. State data on the average price cluded the National Governors' Association, na-
of natural gas delivered to consumers were taken tional education groups, city and county officials,
from the Natural Gas Annual 1982 (vol. 1, Table 17) and the various organizations that represented them,
and the Natural Gas Annual 1983 (vol. 1, Table 18). such as the National League of Cities (Baumer and
District-level data on household energy sources were Van Horn, 1985). Of these, however, only the
taken from the U.S. Bureau of the Census 1981. education groups contributed money; and they
11. District-level data on CETA expenditures were tended to align with, and contribute to, the same
calculated from The Employment and Training members as organized labor. The education
Reporter (1980), which lists the 1981 allocations to contributions were very small in any case, and the
counties, cities, or other "prime sponsors" located alternative strategy of adding them to the labor PAC
within members' districts. In cases where a prime total had no effect on the coefficients.
sponsor was located in more than one congressional 16. Included in this category were the American
district, the expenditure for that sponsor was allo- Business Association, the Business Industry Council,
cated equally among the several districts in which the Chamber of Commerce, the National Associa-
the sponsor administered its program. tion of Manufacturers, and the National Federation
12. It is important to note, however, that we are of Independent Businesses.
not assuming that NFU, ADA, or any other voting 17. We believe that there is far less reason a priori
index measures members' personal ideology (much to believe that PAC contributions should be consid-
less their true preferences); a number of factors com- ered endogenous in modeling members' participa-
bine to determine these voting patterns, ideology tion than in modeling their roll-call voting behavior.
being only one. (See Jackson and Kingdon 1990; While it is likely that PACs will give disproportion-
Carson and Oppenheimer 1984.) Rather, we simply ately to members with important committee posi-
assume that the rating summarizes members' past tions, there is little evidence to suggest that the an-
voting behavior, which in turn form the basis for ticipated participation of member i on some particu-
particular groups' expectations about what positions lar bill j (independent of what one would anticipate
members will take in the future. Indeed, one of the given the member's institutional position or posi-
principal reasons that groups construct their own in- tions, seniority, and interests-factors that are built
dexes is to help them distinguish between friend and into our model) figures prominently in PAC alloca-
foe, and raters themselves report that the ratings tion decisions. Such calculations, at least, have been
"have their greatest impact on the distribution of nowhere evident in the considerable political science
campaign funds, because they provide a simple test or journalistic literature on this subject. Hence, we
of support or opposition" (Fowler 1982). The NFU also estimated the equations for both cases using or-
scores were taken from the National Farmers Union dinary least squares. The parameter estimates from-
Newsletter (1982a, 1982b). the ordinary least squares and two-stage least
13. We measured dairy industry contributions for squares (2SLS) were very similar, with the exception
each member as the summary of contributions from that the magnitude of the 2SLS mobilization coeffi-
the three main dairy PACs during the previous elec- cients were somewhat smaller in the natural gas and
tion cycle: Committee for Thorough Agricultural job training cases. By presenting the 2SLS results,
Political Education of Associated Milk Producers; then, we address the potential endogeneity problem
Mid-America Dairymen; and Dairymen Special and, as it turns out, slightly bias our results against
Political Agricultural Community Education. our main conclusions.
14. The Moody's entry included a brief descrip- 18. While the first-stage results are not relevant to
tion of each business's activities that usually in- our substantive interests here, they do bear on the
dicated whether it belonged primarily in one confidence of the second-stage results and thus war-
category or another. Where that description men- rant some attention. In estimating the first stage, we
tioned interests in more than one category, we went adapted the contributions model from the substan-
to the financial statements or audit summaries pro-tial literature on the allocation strategies of national
vided in the Moody's entry and classified businessesPACs (e.g., Evans 1986; Grenzke 1989b; Gopoian
as producer, interstate pipeline, intrastate pipeline,1984; Grier and Munger 1989), including three vari-
or distributor according to the principal sources ofables that qualified as instruments: party, the rele-


This content downloaded from on Tue, 22 Nov 2016 21:09:36 UTC
All use subject to
American Political Science Review Vol. 84

vant voting index, and the marginality of the tive Decisions: Proportional Representation and
district. In all three cases the first-stage results were the Borda Rule." American Political Science
satisfactory. The adjusted R-squared was .34 for the Review 77:718-33.
model of dairy industry contributions to Agriculture Chappell, Henry. 1982. "Campaign Contributions
Committee members, .58 for the model of net pro- and Congressional Voting: A Simultaneous
ducer contributions to Commerce members, and .48 Probit-Tobit Model." Review of Economics and
for the model of net labor contributions to Educa- Statistics 62:77-83.
tion and Labor members. More importantly, in Davis, Joseph A. 1984. "House Energy Committee
every case the coefficient on at least one of the three Approves Natural Gas Bill." Congressional
instruments was large, correct in sign, and statisti- Quarterly Weekly Report 14 April:888-89.
cally significant at the .05 level. Checks for multi- Denzau, Arthur, and Michael C. Munger. 1986.
collinearity among the independent variables in the "Legislators and Interest Groups: How Un-
second stage equations likewise provided little cause organized Interests Get Represented." American
for concern. Regarding the appropriateness and im- Political Science Review 80:89-106.
plementation of the two-stage least squares estima- Donnelly, Harrison. 1982. "Job Training Bills: No
tion procedure, see Hanushek and Jackson 1977, 'CETA' Revisited." Congressional Quarterly
chap. 9; Pindyck and Rubinfeld 1981, 328-31. Weekly Report 8 May:1035.
19. As a check on the results reported here, we Drew, Elizabeth. 1982. "Politics and Money." The
also estimated for the effects of contributions on par- New Yorker 6 December:54-149.
ticipation without interacting them with anticipated Etzioni, Amitai. 1984. Capital Corruption: The New
support or opposition, and, as our theory would Attack on American Democracy. New York:
predict, the effects were consistently weaker. Note, Harcourt, Brace, Jovanovich.
secondly, that we do not include the relevant group Evans, C. Lawrence. N.d. "Participation in U.S.
support score separately in the model, a variable Senate Committees." Political Science Quarterly.
that has proven important in estimating the effects Forthcoming.
of contributions on roll calls. Even if ideology is Evans, Diana. 1986. "PAC Contributions and Roll-
what the voting scores capture (see Jackson and Call Voting: Conditional Power." In Interest
Kingdon 1990), there is no theoretical reason to ex- Group Politics, 2d ed., ed. Allan J. Cigler and
pect that liberals will be more active than conserva- Burdett A. Loomis. Washington, DC: Congres-
tives (or vice versa) or for that matter that ideologi- sional Quarterly Press.
cal moderates will be less active than either conserv-Evans, Diana. 1988. "Oil PACs and Aggressive
atives or liberals. In any case, we tested for the effect Contribution Strategies." Journal of Politics
of past voting behavior on the participation of 50:1047-56.
members in each case. For all three, the t-statistics Fenno, Richard F. 1978. Home Style: House
for the voting score coefficients were less than .5. Members in Their Districts. Boston: Little,
Fiorina, Morris P. 1974. Representatives, Roll Calls,
References and Constituencies. Lexington, MA: D.C.
Arnold, R. Douglas. 1979. Congress and the Fowler, Linda L. 1982. "How Interest Groups Select
Bureaucracy: A Theory of Influence. New Issues for Rating Voting Records of Members of
Haven: Yale University Press. the U.S. Congress." Legislative Studies Quarter-
Bauer, Raymond, Ithiel de Sola Pool, and Lewis A. ly 7:401-13.
Dexter. 1963. American Business and Public Franklin, Grace A., and Randall B. Ripley. 1984.
Policy. Chicago: Aldine & Atherton. CETA: Politics and Policy. Knoxville: University
Baumer, Donald C., and Carl E. Van Horn. 1985. of Tennessee Press.
The Politics of Unemployment. Washington: Gais, Thomas, Mark Peterson, and Jack Walker.
Congressional Quarterly Press. 1984. "Interest Groups, Iron Triangles, and
Berry, Jeffrey M. 1984. The Interest Group Society. Representative Institutions in American National
Boston: Little, Brown. Government." British Journal of Political Science
Bureau of National Affairs, Inc. 1980. "FY 1981 14:161-85.
CETA Allocations and Government's Grants." Gopoian, J. David. 1984. "What Makes PACs Tick?
The Employment and Training Reporter. Oct. An Analysis of the Allocation Patterns of
29. Economic Interest Groups." American Journal of
Carson, Richard A., and Joe Oppenheimer. 1984. Political Science 28:259-81.
"A Method of Estimating the Personal Ideology Grenzke, Janet M. 1989a. "Shopping in the Congres-
of Political Representatives." American Political sional Supermarket: The Currency Is Complex."
Science Review 78:163-78. American Journal of Political Science 33:1-24.
Chamberlin, John R., and Paul N. Courant. 1983. Grenzke, Janet M. 1989b. "Candidate Attributes and
"Representative Deliberations and Representa- PAC Contributions." Western Political Quarter-


This content downloaded from on Tue, 22 Nov 2016 21:09:36 UTC
All use subject to
Bias in Congressional Committees

ly 42:245-64. Moody's Investor Service. 1983a. Moody's Indus-

Grier, Kevin B., and Michael C. Munger. 1986. "The trial Manual. New York: Moody's Investor
Impact of Legislator Attributes on Interest- Service.
Group Campaign Contributions." Journal of Moody's Investor Service. 1983b. Moody's Public
Labor Research 7:349-61. Utilities Manual. New York: Moody's Investor
Grier, Kevin B., and Michael C. Munger. 1989. Service.
"Committee Assignments, Constituent Pref- Murray, Alan. 1983. "Pressure from Consumers
erences, and Campaign Contributions to House Pushes Congress into Action on Pricing of
Incumbents." Typescript. Natural Gas." Congressional Quarterly Weekly
Hall, Richard L. 1987. "Participation and Purpose in Report 5 March:443-47.
Committee Decision Making." American Politi- National Farmers Union. 1982a. "1981 Voting Rec-
cal Science Review 81:105-27. ord-House." National Farmers Union Washing-
Hall, Richard L. 1989. "Committee Decision Making ton Newsletter 5 February:4-8.
in the Postreform Congress." In Congress Recon- National Farmers Union. 1982b. "1982 Voting Rec-
sidered, 4th ed., ed. Lawrence C. Dodd and ord-House." National Farmers Union Washing-
Bruce I. Oppenheimer. Washington, DC: ton Newsletter 15 October:4-8.
Congressional Quarterly Press. Pindyck, Robert S., and Daniel L. Rubinfeld. 1981.
Hall, Richard L., and C. Lawrence Evans. 1990. Econometric Models and Economic Forecasts. 2d
"The Power of Subcommittees." Journal of Poli- ed. New York: McGraw-Hill.
tics 52:335-55. Pressman, Steven. 1983. "Lobbying Free-for-all
Hanushek, Erik A., and John E. Jackson. 1977. Sta- Opens as Congress Begins Mark-up of National
tistical Methods for Social Scientists. Orlando: Gas Pricing Bills." Congressional Quarterly
Academic. Weekly Report. April 23: 793-97.
Hayes, Michael T. 1981. Lobbyists and Legislators. Ripley, Randall B., and Grace Franklin. 1980. Con-
New Brunswick: Rutgers University Press. gress, the Bureaucracy, and Public Policy. Rev.
Herndon, James F. 1982. "Access, Record, and ed. Homewood, IL: Dorsey.
Competition as Influences on Interest Group Rothenberg, Lawrence S. 1989. "Do Interest Groups
Contributions to Congressional Campaigns." Make a Difference? Lobbying, Constituency In-
Journal of Politics 44:996-1019. fluence, and Public Policy." Presented at the an-
Jackson, Brooks. 1988. Honest Graft. New York: nual meeting of the Midwest Political Science
Knopf. Association, Chicago.
Jackson, John E. 1974. Constituencies and Leaders Salisbury, Robert H. 1984. "Interest Representation:
in Congress: Their Effects on Senate Voting The Dominance of Institutions." American Polit-
Behavior. Cambridge: Harvard University Press. ical Science Review 78:64-76.
Jackson, John E., and David C. King. 1989. "Public Salisbury, Robert H., and Kenneth A. Shepsle.
Goods, Private Interests, and Representation." 1981. "U.S. Congressmen As Enterprise." Legis-
American Political Science Review 83:1143-64. lative Studies Quarterly 6:559-76.
Jackson, John E., and John W. Kingdon. 1990. Schattschneider, E. E. 1960. The Semisovereign
"Ideology, Interest Group Scores, and Legislative People. Hinsdale, IL: Dryden.
Votes." University of Michigan. Typescript. Schlozman, Kay L., and John T. Tierney. 1986.
Jacobson, Gary C. 1980. Money in Congressional Organized Interests and American Democracy.
Elections. New Haven: Yale University Press. New York: Harper & Row.
Kau, James B., and Paul H. Rubin. 1982. Con- Shepsle, Kenneth A. 1978. The Giant Jigsaw Puzzle.
gressmen, Constituents, and Contributors: Chicago: University of Chicago.
Determinants of Roll Call Voting in the House Silberman, Jonathan, and Garey C. Durden. 1976.
Representatives. Boston: Martinus Nijhoff. "Determining Legislative Preferences on the
Kingdon, John W. 1981. Congressmen's Voting Minimum Wage: An Economic Approach." Jour-
Decisions, 2d ed. New York: Harper & Row. nal of Political Economy 84:317-29.
Maitland, Ian. 1985. "Interest Groups and Economic Smith, Richard A. 1984. "Advocacy, Interpretation,
Growth Rates." journal of Politics 47:44-58. and Influence in the U.S. Congress." American
Maraniss, David. 1983. "Power Play: Chairman's Political Science Review 78:44-63.
Gavel Crushes Gas Decontrol Vote." Washing- Smith, Steven S., and Christopher J. Deering. 1984.
ton Post 20 November:Al. Committees in Congress. Washington, DC: Con-
Matthews, Donald R. [1960] 1973. U.S. Senators gressional Quarterly Press.
and Their World. Reprint. New York: W. W. Stem, Phillip M. 1988. The Best Congress Money
Norton. Can Buy. New York: Pantheon.
Mayhew, David R. 1974. Congress: The Electoral Uslaner, Eric M. 1989. Shale Barrel Politics. Stan-
Connection. New Haven: Yale University Press. ford: Stanford University Press.
Milbrath, Lester M. 1963. The Washington Lobby- U.S. Bureau of the Census. 1981. Fuels and Financial
ists. Chicago: Rand McNally. Characteristics of Housing Units: 1980, Congres-


This content downloaded from on Tue, 22 Nov 2016 21:09:36 UTC
All use subject to
American Political Science Review Vol. 84

sional Districts in the 98th Congress. Washing- Top." Congressional Quarterly Weekly Report
ton: GPO. 21 August:2050-51.
Wayman, Frank W. 1985. "Arms Control and Stra- Welch, William P. 1982. "Campaign Contributions
tegic Arms Voting in the U.S. Senate: Patterns of and Legislative Voting: Milk Money and Dairy
Change, 1967-1983." Journal of Conflict Resolu- Price Supports." Western Political Quarterly
tion 29:225-51. 35:478-95.
Wayman, Frank W., and Edward Kutler. 1985. "The Wilhite, Allen, and John Theilmann. 1987. "Labor
Changing Politics of Oil and Gas: Ideology, PAC Contributions and Labor Legislation: A
Campaign Contributions, and Interests." Pre- Simultaneous Logit Approach." Public Choice
sented at the annual meeting of the American 53:267-76.
Political Science Association, New Orleans. Wright, John R. 1985. "PACs, Contributions, and
Wehr, Elizabeth. 1982a. "Dairy, Grain Proposals Roll Calls: An Organizational Perspective."
Draw Administration Fire." Congressional American Political Science Review 79:400-14.
Quarterly Weekly Report 24 July:1751. Wright, John R. 1989. "Contributions, Lobbying,
Wehr, Elizabeth. 1982b. "New Farm Support Plans, and Committee Voting in the U.S.House of Rep-
Food Stamp Changes Push Savings Totals Over resentatives." University of Iowa. Typescript.

Richard L. Hall is Assistant Professor of Political Science, University of Michigan,.

Ann Arbor, MI 48109.
Frank W. Wayman is Professor of Political Science, University of Michigan, Dear-
born, MI 48128.


This content downloaded from on Tue, 22 Nov 2016 21:09:36 UTC
All use subject to