The Fantasy of Representative Government in the U.S.
Time to Face The Truth Vol II
I've said for some time that the political process in the U.S. was illusion designed by the ruling Elite to give cover of legitimacy to the Totalitarian State. Here is a Princeton University Study that statistically PROVES me right.
Have you noticed how powerless you are, how powerless congress seems to be, how the New World Order agenda keeps being enacted in D.C. and around the country no matter how many times you "vote your guy in" and whether it is a Democrat or Republican in the "Presidency" or majority in congress? In a "Representative Republic" how can this be? HERE is the stark reality of the New World Order = Corporate Ownership of the American Government.
The first-ever scientific study that analyzes whether the US is a democracy, rather than an oligarchy, found the majority of the American public has a “minuscule, near-zero, statistically non-significant impact upon public policy” compared to the wealthy.
The study, due out in the Fall 2014 issue of the academic journal Perspectives on Politics, sets out to answer elusive questions about who really rules in the United States. The researchers measured key variables for 1,779 policy issues within a single statistical model in an unprecedented attempt “to test these contrasting theoretical predictions” – i.e. whether the US sets policy democratically or the process is dominated by economic elites, or some combination of both.
"Despite the seemingly strong empirical support in previous studies for theories of majoritarian democracy, our analyses suggest that majorities of the American public actually have little influence over the policies our government adopts,” the researchers from Princeton University and Northwestern University wrote.
While “Americans do enjoy many features central to democratic governance, such as regular elections, freedom of speech and association,” the authors say the data implicate “the nearly total failure of 'median voter' and other Majoritarian Electoral Democracy theories [of America]. When the preferences of economic elites and the stands of organized interest groups are controlled for, the preferences of the average American appear to have only a minuscule, near-zero, statistically non-significant impact upon public policy."
The authors of “Testing Theories of American Politics: Elites, Interest Groups, and Average Citizens” say that even as their model tilts heavily toward indications that the US is, in fact, run by the most wealthy and powerful, it actually doesn’t go far enough in describing the stranglehold connected elites have on the policymaking process.
“Our measure of the preferences of wealthy or elite Americans – though useful, and the best we could generate for a large set of policy cases – is probably less consistent with the relevant preferences than are our measures of the views of ordinary citizens or the alignments of engaged interest groups,” the researcher said.
“Yet we found substantial estimated effects even when using this imperfect measure. The real-world impact of elites upon public policy may be still greater.”
They add that the “failure of theories of Majoritarian Electoral Democracy is all the more striking because it goes against the likely effects of the limitations of our data. The preferences of ordinary citizens were measured more directly than our other independent variables, yet they are estimated to have the least effect.”
Despite the inexact nature of the data, the authors say with confidence that “the majority does not rule -- at least not in the causal sense of actually determining policy outcomes.”
“We believe that if policymaking is dominated by powerful business organizations and a small number of affluent Americans, then America’s claims to being a democratic society are seriously threatened,” they concluded. http://rt.com/usa/us-democracy-oligarchy-policy-512/
April 9, 2014
Testing Theories of American Politics:
Elites, Interest Groups, and Average Citizens
Martin Gilens
Princeton University
mgilens@princeton.edu
Benjamin I. Page
Northwestern University
b-page@northwestern.edu
forthcoming Fall 2014 in Perspectives on Politics
For helpful comments the authors are indebted to Larry Bartels and Jeff Isaacs, to three
anonymous reviewers, and to seminar participants at Harvard and Rochester Universities.
Gilens and Page Testing Theories of American Politics 2
2
Abstract
Each of four theoretical traditions in the study of American politics – which can be characterized
as theories of Majoritarian Electoral Democracy, Economic Elite Domination, and two types of
interest group pluralism, Majoritarian Pluralism and Biased Pluralism – offers different
predictions about which sets of actors have how much influence over public policy: average
citizens; economic elites; and organized interest groups, mass-based or business-oriented.
A great deal of empirical research speaks to the policy influence of one or another set of actors,
but until recently it has not been possible to test these contrasting theoretical predictions against
each other within a single statistical model. This paper reports on an effort to do so, using a
unique data set that includes measures of the key variables for 1,779 policy issues.
Multivariate analysis indicates that economic elites and organized groups representing business
interests have substantial independent impacts on U.S. government policy, while average citizens
and mass-based interest groups have little or no independent influence. The results provide
substantial support for theories of Economic Elite Domination and for theories of Biased
Pluralism, but not for theories of Majoritarian Electoral Democracy or Majoritarian Pluralism.
Gilens and Page Testing Theories of American Politics 3
3
Who governs? Who really rules? To what extent is the broad body of U.S. citizens
sovereign, semi-sovereign, or largely powerless? These questions have animated much important
work in the study of American politics.
While this body of research is rich and variegated, it can loosely be divided into four
families of theories: Majoritarian Electoral Democracy, Economic Elite Domination, and two
types of interest group pluralism – Majoritarian Pluralism, in which the interests of all citizens
are more or less equally represented, and Biased Pluralism, in which corporations, business
associations, and professional groups predominate) Each of these perspectives makes different
predictions about the independent influence upon U.S. policy making of four sets of actors: the
Average Citizen or “median voter,” Economic Elites, and Mass-based or Business-oriented
Interest Groups or industries.
Each of these theoretical traditions has given rise to a large body of literature. Each is
supported by a great deal of empirical evidence – some of it quantitative, some historical, some
observational – concerning the importance of various sets of actors (or, all too often, a single set
of actors) in U.S. policy making. This literature has made important contributions to our
understanding of how American politics works and has helped illuminate how democratic or
undemocratic (in various senses) our policy making process actually is. Until very recently,
however, it has been impossible to test the differing predictions of these theories against each
other within a single statistical model that permits one to analyze the independent effects of each
set of actors upon policy outcomes.
Here – in a tentative and preliminary way – we offer such test, bringing a unique data set
to bear on the problem. Our measures are far from perfect, but we hope that this first step will
help inspire further research into what we see as some of the most fundamental questions about
American politics.
The central point that emerges from our research is that economic elites and organized
groups representing business interests have substantial independent impacts on U.S. government
policy, while mass-based interest groups and average citizens have little or no independent
influence. Our results provide substantial support for theories of Economic Elite Domination
and for theories of Biased Pluralism, but not for theories of Majoritarian Electoral Democracy or
Majoritarian Pluralism.
In what follows, we briefly review the four theoretical traditions that form the framework
for our analyses and highlight some of the most prominent empirical research associated with
each. We then describe our data and measures and present our results. We conclude by
Gilens and Page Testing Theories of American Politics 4
4
discussing the implications of our work for understanding American democracy and by
identifying some of the directions for future research that our findings suggest.
Four Theoretical Traditions
Each of the four theoretical traditions we are addressing has produced a body of literature
much too vast to review in detail here. We can only allude to a few central pieces of work in
each tradition. And we must acknowledge that a particular scholar’s work does not always fall
neatly into a single category. Some scholars work across – or independently of – our theoretical
categories, embracing multiple influences and complex processes of policy making. Here we
focus on ideal types of theory, for the purpose of outlining certain distinctive predictions that
those types of theory tend to make. Given the nature of our data, we focus on the societal
sources of influence that these theories posit, rather than on the mechanisms of influence that
they discuss.
Majoritarian Electoral Democracy. Theories of majoritarian electoral democracy, as
positive or empirical theories, attribute U.S. government policies chiefly to the collective will of
average citizens, who are seen as empowered by democratic elections. Such thinking goes back
at least to Tocqueville, who (during the Jacksonian era) saw American majorities as
“omnipotent” – particularly at the state level – and worried about “tyranny of the majority.”1 It
is encapsulated in Abraham Lincoln’s reference to government “of the people, by the people, for
the people,” and was labeled by Robert Dahl “populistic democracy.” 2
An important modern incarnation of this tradition is found in rational choice theories of
electoral democracy, in which vote-seeking parties or candidates in a two-party system tend to
converge at the mid-point of citizens’ policy preferences. If preferences are jointly singlepeaked
so that they can be arrayed along a single dimension, the “median voter theorem” –
posited verbally by Harold Hotelling, proved by Duncan Black, and popularized by Anthony
Downs in his Economic Theory of Democracy – states that two vote-seeking parties will both
take the same position, at the center of the distribution of voters’ most-preferred positions.
Under the relevant assumptions, public policy that fits the preferences of the median voter is not
only the empirically-predicted equilibrium result of two-party electoral competition; as the
“Condorcet winner” it also has the normative property of being the “most democratic” policy, in
the sense that it would be preferred to any alternative policy in head-to-head majority-rule voting
by all citizens.3
Subsequent “chaos” results by social choice theorists, starting with Kenneth Arrow, have
indicated that the median voter prediction follows logically only for unidimensional politics. If
citizens’ preference orderings are not unidimensional and are sufficiently diverse, majority rule –
Gilens and Page Testing Theories of American Politics 5
5
hence also two-party electoral competition – might not lead to any equilibrium outcome at all. 4
It is important to note, however, that what might theoretically happen will not necessarily ever
happen in practice. Real-world outcomes depend upon how institutions are organized and how
preferences are actually configured.
Despite the “chaos” results, and despite many criticisms of the median-voter theorem as
simplistic and empirically inapplicable or wrong,5 a good many scholars – probably more
economists than political scientists among them – still cling to the idea that the policy
preferences of the median voter tend to drive policy outputs from the U.S. political system. A
fair amount of empirical evidence has been adduced – by Alan Monroe; Benjamin Page and
Robert Shapiro; Robert Erikson, Michael MacKuen, and James Stimson (authors of the very
influential Macro Polity); and others – that seems to support the notion that the median voter
determines the results of much or most policy making. This evidence indicates that U.S. federal
government policy is consistent with majority preferences roughly two-thirds of the time; that
public policy changes in the same direction as collective preferences a similar two thirds of the
time; that the liberalism or conservatism of citizens is closely associated with the liberalism or
conservatism of policy across states; and that fluctuations in the liberal or conservative “mood”
of the public are strongly associated with changes in the liberalism or conservatism of policy in
all three branches of government.6
The fly in the ointment is that none of this evidence allows for, or explicitly assesses, the
impact of such variables as the preferences of wealthy individuals, or the preferences and actions
of organized interest groups, which may independently influence public policy while perhaps
being positively associated with public opinion – thereby producing a spurious statistical
relationship between opinion and policy.
Recent research by Larry Bartels and by one of the present authors (Gilens), which
explicitly brings the preferences of “affluent” Americans into the analysis along with the
preferences of those lower in the income distribution, indicates that the apparent connection
between public policy and the preferences of the average citizen may indeed be largely or
entirely spurious.7
The “electoral reward and punishment” version of democratic control through elections –
in which voters retrospectively judge how well the results of government policy have satisfied
their basic interests and values, and politicians enact policies in anticipation of judgments that
they expect will later be made by what V.O. Key, Jr., called “latent” public opinion – might be
thought to offer a different prediction: that policy will tend to satisfy citizens’ underlying needs
and values, rather than corresponding with their current policy preferences.8 We cannot test this
prediction because we do not have – and cannot easily imagine how to obtain – good data on
individuals’ deep, underlying interests or values, as opposed to their expressed policy
Gilens and Page Testing Theories of American Politics 6
6
preferences. But the evidence that collective policy preferences are generally rather stable over
time suggests that expressed collective policy preferences may not often diverge markedly from
subsequently manifested “latent” preferences. They may do so only under special circumstances,
such as economic recessions or disastrous wars.9 If so, the electoral-reward-and-punishment
type of democratic theory, too, predicts that most of the time public policy will respond to the
current policy preferences of the average citizen.
Economic Elite Domination. A quite different theoretical tradition argues that U.S.
policy making is dominated by individuals who have substantial economic resources, i.e. high
levels of income and/or wealth – including, but not limited to, ownership of business firms.
Not all “elite theories” share this focus. Some emphasize social status or institutional
position – such as the occupancy of key managerial roles in corporations, or top-level positions
in political parties, in the executive, legislative, or judicial branches of government, or in the
highest ranks of the military. Some elite theories postulate an amalgam of elites, defined by
combinations of social status, economic resources, and institutional positions, who achieve a
degree of unity through common backgrounds, coinciding interests, and social interactions.
For example, C. Wright Mills’ important book, The Power Elite, offers a rather nuanced
account of how U.S. social, economic, political, and military elites have historically alternated in
different configurations of dominance. Mills noted that his elites derived in substantial
proportions from the upper classes, including the very rich and corporate executives, but their
elite status was not defined by their wealth.10 Our focus here is on theories that emphasize the
policy-making importance of economic elites.
Analyses of U.S. politics centered on economic elites go back at least to Charles Beard,
who maintained that a chief aim of the framers of the U.S. Constitution was to protect private
property, favoring the economic interests of wealthy merchants and plantation owners rather than
the interests of the then-majority small farmers, laborers, and craft workers. A landmark work in
this tradition is G. William Domhoff’s detailed account of how elites (working through
foundations, think-tanks, and an “opinion-shaping apparatus,” as well as through the lobbyists
and politicians they finance) may dominate key issues in U.S. policy making despite the
existence of democratic elections. Philip A. Burch has exhaustively chronicled the economic
backgrounds of federal government officials through American history. Thomas Ferguson’s
analysis of the political importance of “major investors” might be seen as a theory of economic
elites. Most recently, Jeffrey Winters has posited a comparative theory of “Oligarchy,” in which
the wealthiest citizens – even in a “civil oligarchy” like the United States – dominate policy
concerning crucial issues of wealth- and income-protection.11
Gilens and Page Testing Theories of American Politics 7
7
Our third and fourth theoretical traditions posit that public policy generally reflects the
outcome of struggle among organized interest groups and business firms.12
Majoritarian Pluralism. The roots of what we can characterize as theories of
“majoritarian” interest group pluralism go back to James Madison’s Federalist Paper #10, which
analyzed politics in terms of “factions” -- a somewhat fuzzy concept that apparently
encompassed political parties and even popular majorities, as well as what we would today
consider organized interest groups, business firms, and industrial sectors. Madison argued that
struggles among the diverse factions that would be found in an extensive republic would lead to
policies more or less representative of the needs and interests of the citizenry as a whole – or at
least would tend to defeat “tyrannical” policies, including the much-feared issuance of
inflationary paper money that might cater to local majority factions of farmer-debtors but be
costly to merchant creditors.13
In the twentieth century, Arthur Bentley’s The Process of Government and then David
Truman’s monumental The Governmental Process put groups at the center of political analysis,
laying out a detailed picture of how organized interest groups might get their way. Truman
offered a comprehensive and still-interesting catalogue of lobbying techniques and other
methods of group influence. He also added an ingenious gloss to Madison that tends to increase
both the plausibility and the normative appeal of majoritarian interest group pluralism: the
assertion that all interests have at least a minimum of influence in group-dominated policy
making, because policy makers must (in order to avoid subsequent punishment) heed all
“potential” groups that would form if their interests were trampled upon.14
Robert Dahl’s analysis of New Haven city politics was Madisonian or Truman-like in its
insistence that many (all?) diverse interests were represented, though Dahl focused as much on
active members of the general public as on organized groups. Dahl’s analyses of American
politics in terms of “polyarchy” or “pluralist democracy” also come close to our ideal type of
majoritarian pluralist theory, since they imply that the wants or needs of the average citizen tend
to be reasonably well served by the outcomes of interest group struggle. Several contemporary
analysts of interest group politics likewise appear to accept (at least implicitly) a picture of group
struggle that results in more or less majoritarian results.15
A major challenge to majoritarian pluralist theories, however, is posed by Mancur
Olson’s argument that collective action by large, dispersed sets of individuals with individually
small but collectively large interests tends to be prevented by the “free rider” problem. Barring
special circumstances (selective incentives, byproducts, coercion), individuals who would benefit
from collective action may have no incentive to personally form or join an organized group. If
everyone thinks this way and lets George do it, the job is not likely to get done. This reasoning
suggests that Truman’s “potential groups” may in fact be unlikely to form, even if millions of
Gilens and Page Testing Theories of American Politics 8
8
peoples’ interests are neglected or harmed by government. Aware of the collective action
problem, officials may feel free to ignore much of the population and act against the interests of
the average citizen.16
Biased Pluralism. Olson’s argument points toward an important variant line of thinking
within the pluralist tradition: theories of “biased” pluralism, which posit struggles among an
unrepresentative universe of interest groups – characterized by E.E. Schattschneider as a
heavenly chorus with an “upper-class accent,” and more recently dubbed by Kay Lehman
Schlozman, Sidney Verba, and Henry Brady an “unheavenly chorus.” Theories of biased
pluralism generally argue that both the thrust of interest group conflict and the public policies
that result tend to tilt toward the wishes of corporations and business and professional
associations.17
Schattschneider suggested that policy outcomes vary with the “scope of conflict”: for
example, that business-oriented interest groups tend to prevail over ordinary citizens when the
scope is narrow and visibility is low. Grant McConnell added the idea that the actual
“constituencies” of policy implementers can consist of powerful groups. George Stigler
(articulating what some economists have scorned as “Chicago Marxism”) analyzed the politics of
regulation in terms of biased pluralism: the capture of regulators by the regulated. Charles
Lindblom outlined a number of ways – including the “privileged position” of business – in which
business firms and their associations influence public policy. Thomas Ferguson has posited an
“investment theory” of politics in which “major investors” – especially representatives of
particular industrial sectors – fund political parties in order to get policies that suit their
economic interests. Fred Block’s “neo-Polanyian” analysis emphasizes groups. Jacob Hacker’s
and Paul Pierson’s analysis of “winner-take-all-politics,” which emphasizes the power of the
finance industry, can be seen as a recent contribution to the literature of biased pluralism.18
Marxist and neo-Marxist theories of the capitalist state hold that economic classes – and
particularly the bourgeoisie, the owners of the means of production -- dominate policy making
and cause the state to serve their material interests. As the Communist Manifesto put it, “The
bourgeoisie has...conquered for itself, in the modern representative State, exclusive political
sway. The executive of the modern State is but a committee for managing the common affairs of
the whole bourgeoisie.”19 We cannot precisely test the predictions of such theories, because we
lack good measures of policy preferences by economic class. (In Marxist theory, neither income
nor wealth accurately signals class position.) We can note, however, that certain
“instrumentalist” Marxist theories, including the important version put forth by Ralph Miliband,
make predictions resembling those of theories of Biased Pluralism: that interest groups and
corporations representing “large scale business” tend to prevail.20
Gilens and Page Testing Theories of American Politics 9
9
As to empirical evidence concerning interest groups, it is well established that organized
groups regularly lobby and fraternize with public officials; move through revolving doors
between public and private employment; provide self-serving information to officials; draft
legislation; and spend a great deal of money on election campaigns.21 Moreover, in harmony
with theories of biased pluralism, the evidence clearly indicates that most U.S. interest groups
and lobbyists represent business firms or professionals. Relatively few represent the poor or
even the economic interests of ordinary workers, particularly now that the U.S. labor movement
has become so weak.22
But do interest groups actually influence policy? Numerous case studies have detailed
instances in which all but the most dedicated skeptic is likely to perceive interest group influence
at work. A leading classic remains Schattschneider’s analysis of the 1928 enactment of the
Smoot-Hawley tariff, an astounding orgy of pork-barrel politics.23 Still, many quantitatively
oriented political scientists seem to ignore or dismiss such non-quantitative evidence. There
have also been some efforts (particularly during the Cold War era, when unflattering depictions
of U.S. politics may have been thought unpatriotic) to demonstrate that interest groups have no
influence on policy at all. Raymond Bauer, Ithiel Pool, and Lewis Anthony Dexter argued that
business had little or no effect on the renewal of reciprocal trade authority. Lester Milbrath,
having conducted interviews with lobbyists and members of Congress, rated lobbyists’ influence
as very low. More recently, Fred McChesney has made the ingenious argument that campaign
contributions from interest groups may not represent quid pro quo bribery attempts by groups,
but instead result from extortion by politicians who threaten to harm the groups’ interests.24
Very few studies have offered quantitative evidence concerning the impact of interest
groups based on a number of different public policies. Important exceptions include the work of
Mark Smith and that of Frank Baumgartner, Jeffrey Berry, Marie Hojnacki, David Kimball, and
Beth Leech.25
Mark Smith examined 2,364 “business unity” issues – over a period of four decades – on
which the U.S. Chamber of Commerce (arguably a reasonable proxy for business groups as a
whole, on this particular set of issues where most businesses agreed) took a public stand for or
against. He then calculated six measures of the Chamber’s annual rate of “success” at getting the
action or inaction it favored from Congress.26 The Chamber’s average success rate in terms of
proportion of bills enacted or defeated appears to have been fairly high,27 but Smith did not argue
that such success necessarily demonstrates influence. (A batting-average approach to influence
would have to assume that stand-taking is unrelated to expectations of success. Further, in order
to gauge business’s independent impact and avoid spurious results, data on stands taken by other
actors would need to be included as well.) Instead, Smith devoted most of his effort to analyzing
the over-time correlates of high or low success, such as variations in the public “mood” and in
the partisan composition of Congress.
Gilens and Page Testing Theories of American Politics 10
10
Frank Baumgartner and his colleagues, in their meticulous examination of 98 cases of
congressional policy making in which interest groups were active, investigated whether the
magnitude of group resources that were deployed was related to outcomes across those cases. In
their multivariate analyses, Baumgartner et al. found a modest tendency for policy outcomes to
favor the side that enjoyed greater resources (PAC contributions, lobbying expenditures,
membership size, etc.).28
Prior to the availability of the data set that we analyze here, no one we are aware of has
succeeded at assessing interest group influence over a comprehensive set of issues, while taking
into account the impact of either the public at large or economic elites – let alone analyzing all
three types of potential influences simultaneously.
Testing Theoretical Predictions
What makes possible an empirical effort of this sort is the existence of a unique data set,
compiled over many years by one of us (Gilens) for a different but related purpose: for
estimating the influence upon public policy of “affluent” citizens, poor citizens, and those in the
middle of the income distribution.
Gilens and a small army of research assistants29 gathered data on a large, diverse set of
policy cases: 1,779 instances between 1981 and 2002 in which a national survey of the general
public asked a favor/oppose question about a proposed policy change. A total of 1,923 cases met
four criteria: dichotomous pro/con responses, specificity about policy, relevance to federal
government decisions, and categorical rather than conditional phrasing. Of those 1,923 original
cases, 1,779 cases also met the criteria of providing income breakdowns for respondents, not
involving a Constitutional amendment or a Supreme Court ruling (which might entail a quite
different policy making process), and involving a clear, as opposed to partial or ambiguous,
actual presence or absence of policy change. These 1,779 cases do not constitute a sample from
the universe of all possible policy alternatives (this is hardly conceivable), but we see them as
particularly relevant to assessing the public’s influence on policy. The included policies are not
restricted to the narrow Washington “policy agenda.” At the same time – since they were seen
as worth asking poll questions about – they tend to concern matters of relatively high salience,
about which it is plausible that average citizens may have real opinions and may exert some
political influence.30
For each case, Gilens used the original survey data to assess responses by income level.
In order to cope with varying income categories across surveys, he employed a quadratic logistic
Gilens and Page Testing Theories of American Politics 11
11
regression technique to estimate the opinions of respondents at the 10th income percentile (quite
poor), the 50th percentile (median), and the 90th percentile (fairly affluent).31
Here we use these policy preference data to measure – imperfectly, but, we believe,
satisfactorily – two independent variables posited as major influences upon policy making in the
theoretical traditions discussed above.
Policy preferences at the 50th income percentile – that is, the preferences of the medianincome
survey respondent – work quite well as measures of the preferences of the average
citizen (or, more precisely, the median non-institutionalized adult American), which are central
to theories of Majoritarian Electoral Democracy.32 In all cases in which the relationship between
income and preferences is monotonic, and in all cases in which there is no systematic
relationship at all between the two, the preferences of the median-income respondent are
identical to those of the median-preference respondent. In the remaining cases the two are very
close to each other.33
We believe that the preferences of “affluent” Americans at the 90th income percentile can
usefully be taken as proxies for the opinions of wealthy or very-high-income Americans, and can
be used to test the central predictions of Economic Elite theories. To be sure, people at the 90th
income percentile are neither very rich nor very elite; in 2012 dollars, Gilens’ “affluent”
respondents received only about $146,000 in annual household income. To the extent that their
policy preferences differ from those of average-income citizens, however, we would argue that
there are likely to be similar but bigger differences between average-income citizens and the
truly wealthy.
Some evidence for this proposition comes from the 2011 Cooperative Congressional
Election Study.34 Based on 13 policy preference questions asked on this survey, the preferences
of the top 2% of income earners (a group that might be thought “truly wealthy”) are much more
highly correlated with the preferences of the top 10% of earners than with the preferences of the
average survey respondent (r=.91 vs. .69).35 Thus, the views of our moderately high-income
“affluent” respondents appear to capture useful information about the views of the truly wealthy.
In any case, the imprecision that results from use of our “affluent” proxy is likely to
produce underestimates of the impact of economic elites on policy making. If we find
substantial effects upon policy even when using this imperfect measure, therefore, it will be
reasonable to infer that the impact upon policy of truly wealthy citizens is still greater.36
In order to measure interest group preferences and actions, we would ideally like to use
an index of the sort that Baumgartner and his colleagues developed for their ninety-eight policy
issues: an index assessing the total resources brought to bear by all major interest groups that
Gilens and Page Testing Theories of American Politics 12
12
took one side or the other on each of our 1,779 issues. But it is not feasible to construct such an
index for all our cases; this would require roughly twenty times as much work as did the major
effort made by the Baumgartner research team on their cases. Fortunately, however,
Baumgartner et al. found that a simple proxy for their index – the number of reputedly
“powerful” interest groups (from among groups appearing over the years in Fortune magazine’s
“Power 25” lists) that favored a given policy change, minus the number that opposed it –
correlated quite substantially in their cases with the full interest group index (r=0.73).37
Gilens, using a modified version of this simple count of the number of “powerful”
interest groups favoring (minus those opposing) each proposed policy change, developed a
measure of Net Interest Group Alignment. To the set of groups on the “Power 25” lists (which
seemed to neglect certain major business interests) he added ten key industries that had reported
the highest lobbying expenditures. (For the final list of included industries and interest groups,
see Appendix 1.) For each of the 1,779 instances of proposed policy change, Gilens and his
assistants drew upon multiple sources to code all engaged interest groups as “strongly
favorable,” “somewhat favorable,” “somewhat unfavorable,” or “strongly unfavorable” to the
change. He then combined the numbers of groups on each side of a given issue, weighting
“somewhat” favorable or somewhat unfavorable positions at half the magnitude of “strongly”
favorable or strongly unfavorable positions. In order to allow for the likelihood of diminishing
returns as the net number of groups on a given side increases (an increase from 10 to 11 groups
likely matters less than a jump from 1 to 2 does), he took the logarithms of the number of pro
groups and the number of con groups before subtracting. Thus:
Net Interest Group Alignment = ln(# Strongly Favor + [0.5 * # Somewhat Favor] + 1) - ln(#
Strongly Oppose + [0.5 * # Somewhat Oppose] + 1).38
Below we also report results for comparable group alignment indices that were computed
separately for the mass-based and for the business-oriented sets of groups listed in Appendix 1.
Our dependent variable is a measure of whether or not the policy change proposed in
each survey question was actually adopted, within four years after the question was asked. (It
turns out that most of the action occurred within two years). Of course there was nothing easy
about measuring the presence or absence of policy change for each of 1,779 different cases;
Gilens and his research assistants spent many hours poring over news accounts, government
data, Congressional Quarterly publications, academic papers and the like.39
In order to test among our theoretical traditions, we begin by considering all organized
interest groups together, not distinguishing between mass-based and business-oriented groups.
Within a single statistical model, we estimate the independent impact upon our dependent
variable (policy change) of each of three independent variables: the average citizen’s policy
Gilens and Page Testing Theories of American Politics 13
13
preferences (preferences at the 50th income percentile); the policy preferences of economic elites
(measured by policy preferences at the 90th income percentile); and the stands of interest groups
(the Net Interest Group Alignment Index).
Later, in order to distinguish clearly between Majoritarian Pluralism and Biased
Pluralism, we will use two separate measures of net interest group alignment, one involving only
mass-based interest groups and the other limited to business and professional groups. The main
hypotheses of interest, summarized in Table 1, follow fairly straightforwardly from our
discussion of our four ideal types of theory.
(INSERT TABLE 1 ABOUT HERE)
In their pure form, theories of Majoritarian Electoral Democracy (for example, rational
models of electoral competition that include no societal actors other than average citizens),
predict that the influence upon policy of average citizens is positive, significant, and substantial,
while the influence of other actors is not.
Theories of Economic Elite Domination predict positive, significant, and substantial
influence upon policy by economic elites. Most such theories allow for some (though not much)
independent influence by average citizens, e.g. on non-economic, social issues. Many also allow
for some independent influence by business interest groups – and therefore probably by interest
groups taken as a whole – though their emphasis is on wealthy individuals.
In general, theories of interest group pluralism predict that only organized interest groups
will have positive, significant, and substantial effects upon public policy. Influence proceeds
from groups, not from wealthy (or other) individuals. Depending upon the type of pluralist
theory, average citizens may or may not be well represented through organized groups, but they
do not have a great deal of independent influence on their own.
Theories of Majoritarian Pluralism predict that the stands of organized interest groups, all
taken together, rather faithfully represent (that is, are positively and substantially correlated with)
the preferences of average citizens. But since most political influence proceeds through groups,
a multivariate analysis that includes both interest group alignments and citizens’ preferences
should show far more independent influence by the groups than the citizens. Truman’s idea of
“potential groups” does, however, leave room for some direct influence by average citizens.
Theories of Biased Pluralism, too, see organized interest groups as having much more
influence than average citizens or individual economic elites. But they predict that businessoriented
groups play the major role.
Gilens and Page Testing Theories of American Politics 14
14
Recognizing the complexity of the political world, we must also acknowledge the
possibility that more than one of these theoretical traditions has some truth to it:
that several – even all – of our sets of actors may have substantial, positive, independent
influence on public policy. And we must consider the null hypothesis that none of these
theoretical traditions correctly describes even part of what goes on in American politics,
Influence upon Policy of Average Citizens, Economic Elites, and Interest Groups
Before we proceed further, it is important to note that even if one of our predictor
variables is found (when controlling for the others) to have no independent impact on policy at
all, it does not follow that the actors whose preferences are reflected by that variable – average
citizens, economic elites, or organized interest groups of one sort or another – always “lose” in
policy decisions. Policy making is not necessarily a zero-sum game among these actors. When
one set of actors wins, others may win as well, if their preferences are positively correlated with
each other.
It turns out, in fact, that the preferences of average citizens are positively and fairly
highly correlated, across issues, with the preferences of economic elites (see Table 2.) Rather
often, average citizens and affluent citizens (our proxy for economic elites) want the same things
from government. This bivariate correlation affects how we should interpret our later
multivariate findings in terms of “winners” and “losers.” It also suggests a reason why serious
scholars might keep adhering to both the Majoritarian Electoral Democracy and the Economic
Elite Domination theoretical traditions, even if one of them may be dead wrong in terms of
causal impact. Ordinary citizens, for example, might often be observed to “win” (that is, to get
their preferred policy outcomes) even if they had no independent effect whatsoever on policy
making, if elites (with whom they often agree) actually prevail.
(INSERT TABLE 2 ABOUT HERE)
But net interest group stands are not substantially correlated with the preferences of
average citizens. Taking all interest groups together, the index of net interest group alignment
correlates only a non-significant .04 with average citizens’ preferences! (See Table 2.) This
casts grave doubt on David Truman’s and others’ argument that organized interest groups tend to
do a good job of representing the population as a whole. Indeed, as Table 2 indicates, even the
net alignments of the groups we have categorized as “mass-based” correlate with average
citizens’ preferences only at the very modest (though statistically significant) level of .12.
Gilens and Page Testing Theories of American Politics 15
15
Some particular U.S. membership organizations – especially the AARP and labor unions
– do tend to favor the same policies as average citizens. But other membership groups take
stands that are unrelated (pro-life and pro-choice groups) or negatively related (gun owners) to
what the average American wants.40 Some membership groups may reflect the views of
corporate backers or their most affluent constituents. Others focus on issues on which the public
is fairly evenly divided. Whatever the reasons, all mass-based groups taken together simply do
not add up, in aggregate, to good representatives of the citizenry as a whole. Business-oriented
groups do even worse, with a modest negative over-all correlation of -.10.
Nor do we find an association between the preferences of economic elites and the
alignments of either mass-based or business oriented groups. The latter finding, which surprised
us, may reflect profit-making motives among businesses as contrasted with broader ideological
views among elite individuals. For example, economic elites tend to prefer lower levels of
government spending on practically everything, while business groups and specific industries
frequently lobby for spending in areas from which they stand to gain. Thus pharmaceutical,
hospital, insurance, and medical organizations have lobbied for more spending on health care;
defense contractors for weapons systems; the American Farm Bureau for agricultural subsidies,
and so on.
Initial tests of influences on policy making. The first three columns of Table 3 report
bivariate results, in which each of three independent variables (taking all interest groups
together, for now) is modeled separately as the sole predictor of policy change. Just as previous
literature suggests, each of three broad theoretical traditions – Majoritarian Electoral Democracy,
Economic Elite Domination, and interest group pluralism – seems to gain support. When taken
separately, each independent variable – the preferences of average citizens, the preferences of
economic elites, and the net alignments of organized interest groups – is strongly, positively, and
quite significantly related to policy change. Little wonder that each theoretical tradition has its
strong adherents.
(INSERT TABLE 3 ABOUT HERE)
But the picture changes markedly when all three independent variables are included in
the multivariate Model 4 and tested against each other. The estimated impact of average
citizens’ preferences drops precipitously, to a non-significant, near-zero level. Clearly the
median citizen or “median voter” at the heart of theories of Majoritarian Electoral Democracy
does not do well when put up against economic elites and organized interest groups. The chief
predictions of pure theories of Majoritarian Electoral Democracy can be decisively rejected. Not
only do ordinary citizens not have uniquely substantial power over policy decisions; they have
little or no independent influence on policy at all.
Gilens and Page Testing Theories of American Politics 16
16
By contrast, economic elites are estimated to have a quite substantial, highly significant,
independent impact on policy. This does not mean that theories of Economic Elite Domination
are wholly upheld, since our results indicate that individual elites must share their policy
influence with organized interest groups. Still, economic elites stand out as quite influential –
more so than any other set of actors studied here – in the making of U.S. public policy.
Similarly, organized interest groups (all taken together, for now) are found to have
substantial independent influence on policy. Again, the predictions of pure theories of interest
group pluralism are not wholly upheld, since organized interest groups must share influence with
economically elite individuals. But interest group alignments are estimated to have a large,
positive, highly significant impact upon public policy.
These results suggest that reality is best captured by mixed theories in which both
individual economic elites and organized interest groups (including corporations, largely owned
and controlled by wealthy elites) play a substantial part in affecting public policy, but the general
public has little or no independent influence.
The rather low explanatory power of all three independent variables taken together (with
an R-squared of just .074 in Model 4) may partly result from the limitations of our proxy
measures, particularly with respect to economic elites (since our “affluent” proxy is admittedly
imperfect) and perhaps with respect to interest groups (since only a small fraction of politically
active groups are included in our measure). Again, the implication of these limitations in our data
is that interest groups and economic elites actually wield more policy influence than our
estimates indicate. But it is also possible that there may exist important explanatory factors
outside the three theoretical traditions addressed in this analysis. Or there may be a great deal of
idiosyncrasy in policy outputs, or variation across kinds of issues, that would be difficult for any
general model to capture. With our present data we cannot tell.
The precise magnitudes of the coefficients reported in Table 3 are difficult to interpret
because of our logit transformation of independent variables. A helpful way to assess the
relative influence of each set of actors is to compare how the predicted probability of policy
change alters when moving from one point to another on their distributions of policy
dispositions, while holding other actors’ preferences constant at their neutral points (50 percent
favorable for average citizens and for economic elites, and a net interest group alignment score
of 0.) These changing probabilities, based on the coefficients in table 2, are line-graphed in
Figure 1 along with bar graphs of the underlying preference distributions.
(INSERT FIGURE 1 ABOUT HERE)
Gilens and Page Testing Theories of American Politics 17
17
Clearly, when one holds constant net interest group alignments and the preferences of
affluent Americans, it makes very little difference what the general public thinks. The
probability of policy change is nearly the same (around 0.3) whether a tiny minority or a large
majority of average citizens favor a proposed policy change (see the top panel of Figure 1).
By contrast – again with other actors held constant – a proposed policy change with low
support among economically elite Americans (one-out-of-five in favor) is adopted only about 18
percent of the time, while a proposed change with high support (four-out-of-five in favor) is
adopted about 45 percent of the time. Similarly, when support for policy change is low among
interest groups (with five groups strongly opposed and none in favor) the probability of that
policy change occurring is only .16, but the probability rises to .47 when interest groups are
strongly favorable (see the bottom two panels of Figure 1.)41
When both interest groups and affluent Americans oppose a policy it has an even lower
likelihood of being adopted (these proposed policies consist primarily of tax increases.) At the
other extreme, high levels of support among both interest groups and affluent Americans
increases the probability of adopting a policy change, but a strong status quo bias remains
evident. Policies with strong support (as defined above) among both groups are only adopted
about 56 percent of the time (strongly favored policies in our data set that failed include
proposed cuts in taxes, increases in tax exemptions, increased educational spending for K-12,
college support, and proposals during the Clinton administration to add a prescription drug
benefit to Medicare).
Majoritarian Electoral Democracy. What are we to make of findings that seem to go
against volumes of persuasive theorizing and much quantitative research, by asserting that the
average citizen or the “median voter” has little or no independent influence on public policy?
As noted, our evidence does not indicate that in U.S. policy making the average citizen
always loses out. Since the preferences of ordinary citizens tend to be positively correlated with
the preferences of economic elites, ordinary citizens often win the policies they want, even if
they are more or less coincidental beneficiaries rather than causes of the victory. There is not
necessarily any contradiction at all between our findings and past bivariate findings of a roughly
two-thirds correspondence between actual policy and the wishes of the general public, or of a
close correspondence between the liberal/conservative “mood” of the public and changes in
policy making.42 Our main point concerns causal inference: if interpreted in terms of actual
causal impact, the prior findings appear to be largely or wholly spurious.
Further, the issues about which economic elites and ordinary citizens disagree reflect
important matters, including many aspects of trade restrictions, tax policy, corporate regulation,
abortion, and school prayer, so that the resulting political losses by ordinary citizens are not
Gilens and Page Testing Theories of American Politics 18
18
trivial. Moreover, we must remember that in our analyses the preferences of the affluent are
serving as proxies for those of truly wealthy Americans, who may well have more political clout
than the affluent, and who tend to have policy preferences that differ more markedly from those
of the average citizens. Thus even rather slight measured differences between preferences of the
affluent and the median citizen may signal situations in which economic elites want something
quite different from most Americans and generally get their way.
A final point: even in a bivariate, descriptive sense, our evidence indicates that the
responsiveness of the U.S. political system when the general public wants government action is
severely limited. Because of the impediments to majority rule that were deliberately built into
the U.S. political system – federalism, separation of powers, bicameralism – together with
further impediments due to anti-majoritarian congressional rules and procedures, the system has
a substantial status quo bias. Thus when popular majorities favor the status quo, opposing a
given policy change, they are likely to get their way; but when a majority – even a very large
majority – of the public favors change, it is not likely to get what it wants. In our 1,779 policy
cases, narrow pro-change majorities of the public got the policy changes they wanted only about
30% of the time. More strikingly, even overwhelmingly large pro-change majorities, with 80%
of the public favoring a policy change, got that change only about 43% of the time.
In any case, normative advocates of populistic democracy may not be enthusiastic about
democracy by coincidence, in which ordinary citizens get what they want from government only
when they happen to agree with elites or interest groups that are really calling the shots. When
push comes to shove, actual influence matters.
Economic Elites. Economic Elite Domination theories do rather well in our analysis,
even though our findings probably understate the political influence of elites. Our measure of
the preferences of wealthy or elite Americans – though useful, and the best we could generate for
a large set of policy cases – is probably less consistent with the relevant preferences than are our
measures of the views of ordinary citizens or the alignments of engaged interest groups. Yet we
found substantial estimated effects even when using this imperfect measure. The real-world
impact of elites upon public policy may be still greater.
What we cannot do with these data is distinguish definitively among different versions of
elite theories. We cannot be sure whether we are capturing the political influence of the
wealthiest Americans (the top 1% of wealth-holders? the top 1/10th of 1%?), or, conceivably, the
less affluent but more numerous citizens around the 90th income percentile whose preferences are
directly gauged by our measure.
In any case, we need to reiterate that our data concern economic elites. Income and
wealth tend to be positively correlated with other dimensions of elite status, such as high social
Gilens and Page Testing Theories of American Politics 19
19
standing and the occupancy of high-level institutional positions, but they are not the same thing.
We cannot say anything directly about the non-economic aspects of certain elite theories,
especially those that emphasize actors who may not be highly paid, such as public officials and
political party activists.
Organized Interest Groups. Our findings of substantial influence by interest groups is
particularly striking because little or no previous research has been able to estimate the extent of
group influence while controlling for the preferences of other key non-governmental actors. Our
evidence clearly indicates that – controlling for the influence of both the average citizen and
economic elites – organized interest groups have a very substantial independent impact upon
public policy. Theories of interest group pluralism gain a strong measure of empirical support.
Here, too, the imperfections of our measure of interest group alignment (though probably
less severe than in the case of economically elite individuals) suggest, a fortiori, that the actual
influence of organized groups may be even greater than we have found. If we had data on the
activity of the thousands of groups not included in our net interest group alignment measure, we
might find many cases in which a group (perhaps unopposed by any other groups) got its way.
This might be particularly true of narrow issues like special tax breaks or subsidies aimed at just
one or two business firms, which are underrepresented in our set of relatively high-salience
policies. (Our data set includes only policies thought to be important enough for a national
opinion survey to ask a question about it.)
An important feature of interest group influence is that it is often deployed against
proposed policy changes. On the 1,357 proposed policy changes for which at least one interest
group was coded as favoring or opposing change, in only 36% of the cases did most groups favor
change, while in 55% of the cases most groups opposed change. (The remaining cases involved
equal numbers for and against.)43
Distinguishing between Majoritarian Pluralism and Biased Pluralism. Can we say
anything further about whether processes of interest group influence more closely resemble
Truman-like, broadly representative Majoritarian Pluralism, or Schattschneider-style “Biased”
Pluralism, in which business interests, professional associations, and corporations play the
dominant part?
We have already reported several findings that cast serious doubt upon Majoritarian
Pluralism. If the net results of interest group struggle were to help average citizens get their way
– with organized groups perhaps representing citizens more effectively than politically
inattentive Americans could do for themselves – we would expect that the net alignment of
interest groups would be positively and strongly correlated with the policy preferences of the
average citizen. But we know from Table 2 that they are not in fact significantly correlated at
Gilens and Page Testing Theories of American Politics 20
20
all. Interest group alignments are almost totally unrelated to the preferences of average citizens.
Moreover, there is no indication that officials’ anticipation of reactions from “potential groups”
brings policies in line with what citizens want.44 Empirical support for Majoritarian Pluralism
looks very shaky, indeed. We also know that the composition of the U.S. interest group universe
is heavily tilted toward corporations and business and professional associations.45 This fact
certainly points toward Biased rather than Majoritarian Pluralism.
To go a step further, theories of Majoritarian Pluralism predict relatively more
independent influence upon policy by mass-based interest groups than do theories of Biased
Pluralism. It may be useful, therefore, to distinguish between mass-based and business-oriented
interest groups and to investigate how much policy influence each group actually has.
Accordingly, we computed separate net-interest-group-alignment indices for businessoriented
and for mass-based groups (see Appendix 1 for lists of each) and included both of them
in a new multivariate analysis, along with the preferences of average citizens and economic elites
– dropping our previous measure of the net alignment of all interest groups.
The results of this analysis are given in Table 4. Clearly the predictions of Biased
Pluralism theories fare substantially better than those of Majoritarian Pluralism theories. The
influence coefficients for both mass-based and business-oriented interest groups are positive and
highly significant statistically, but the coefficient for business groups is nearly twice as large as
that for the mass groups. Moreover, when we restricted this same analysis to the smaller set of
issues upon which both types of groups took positions – that is, when we considered only cases
in which business-based and mass-based interest groups were directly engaged with each other –
the contrast between the estimated impact of the two types of groups was even greater.46
The advantage of business-oriented groups in shaping policy outcomes reflects their
numerical advantage within the interest group universe in Washington, and also the infrequency
with which business groups are found simultaneously on both sides of a proposed policy
change.47 Both these factors (numerical dominance and relative cohesion) play a part in the much
stronger correlation of the overall interest group alignment index with business groups than with
mass-oriented groups (.96 vs. .47, table 2). The importance of business groups’ numerical
advantage is also revealed when we rescale our measures of business and mass-oriented interest
group alignments to reflect the differing number of groups in each of these categories. Using this
rescaled measure, a parallel analysis to that in table 4 shows that on a group-for-group basis the
average individual business group and the average mass-oriented group appears to be about
equally influential. The greater total influence of business groups in our analysis results chiefly
from the fact that more of them are generally engaged on each issue (roughly twice as many, on
average), not that a single business-oriented group has more clout on average than a single massbased
group.48
Gilens and Page Testing Theories of American Politics 21
21
(INSERT TABLE 4 ABOUT HERE)
Taken as a whole, then, our evidence strongly indicates that theories of Biased Pluralism
are more descriptive of political reality than are theories of Majoritarian Pluralism. It is simply
not the case that a host of diverse, broadly based interest groups take policy stands – and bring
about actual policies – that reflect what the general public wants. Interest groups as a whole do
not seek the same policies as average citizens do. “Potential groups” do not fill the gap.
Relatively few mass-based interest groups are active, they do not (in the aggregate) represent the
public very well, and they have less collective impact on policy than do business-oriented groups
– whose stands tend to be negatively related to the preferences of average citizens. These
business groups are far more numerous and active; they spend much more money; and they tend
to get their way.
Table 4 also confirms our earlier findings about economic elites and median voters.
When the alignments of business-oriented and mass-based interest groups are included
separately in a multivariate model, average citizens’ preferences continue to have essentially
zero estimated impact upon policy change, while economic elites are still estimated to have a
very large, positive, independent impact.
American Democracy?
Each of our four theoretical traditions (Majoritarian Electoral Democracy, Economic
Elite Domination, Majoritarian Interest Group Pluralism, and Biased Pluralism) emphasizes
different sets of actors as critical in determining U.S. policy outcomes, and each tradition has
engendered a large empirical literature that seems to show a particular set of actors to be highly
influential. Yet nearly all the empirical evidence has been essentially bivariate. Until very
recently it has not been possible to test these theories against each other in a systematic,
quantitative fashion.
By directly pitting the predictions of ideal-type theories against each other within a single
statistical model (using a unique data set that includes imperfect but useful measures of the key
independent variables for nearly two thousand policy issues), we have been able to produce some
striking findings. One is the nearly total failure of “median voter” and other Majoritarian
Electoral Democracy theories. When the preferences of economic elites and the stands of
organized interest groups are controlled for, the preferences of the average American appear to
have only a minuscule, near-zero, statistically non-significant impact upon public policy.
Gilens and Page Testing Theories of American Politics 22
22
The failure of theories of Majoritarian Electoral Democracy is all the more striking
because it goes against the likely effects of the limitations of our data. The preferences of
ordinary citizens were measured more directly than our other independent variables, yet they are
estimated to have the least effect.
Nor do organized interest groups substitute for direct citizen influence, by embodying
citizens’ will and ensuring that their wishes prevail in the fashion postulated by theories of
Majoritarian Pluralism. Interest groups do have substantial independent impacts on policy, and a
few groups (particularly labor unions) represent average citizens’ views reasonably well. But the
interest group system as a whole does not. Over-all, net interest group alignments are not
significantly related to the preferences of average citizens. The net alignments of the most
influential, business oriented groups are negatively related to the average citizen’s wishes. So
existing interest groups do not serve effectively as transmission belts for the wishes of the
populace as a whole. “Potential groups” do not take up the slack, either, since average citizens’
preferences have little or no independent impact on policy after existing groups’ stands are
controlled for.
Furthermore, the preferences of economic elites (as measured by our proxy, the
preferences of “affluent” citizens) have far more independent impact upon policy change than
the preferences of average citizens do. To be sure, this does not mean that ordinary citizens
always lose out; they fairly often get the policies they favor, but only because those policies
happen also to be preferred by the economically elite citizens who wield the actual influence.
Of course our findings speak most directly to the “first face” of power: the ability of
actors to shape policy outcomes on contested issues. But they also reflect – to some degree, at
least – the “second face” of power: the ability to shape the agenda of issues that policy makers
consider. The set of policy alternatives that we analyze is considerably broader than the set
discussed seriously by policy makers or brought to a vote in Congress, and our alternatives are
(on average) more popular among the general public than among interest groups. Thus the fate
of these policies can reflect policy makers’ refusing to consider them rather than considering but
rejecting them. (From our data we cannot distinguish between the two.) Our results speak less
clearly to the “third face” of power: the ability of elites to shape the public’s preferences. 49 We
know that interest groups and policy makers themselves often devote considerable effort to
shaping opinion. If they are successful, this might help explain the high correlation we find
between elite and mass preferences. But it cannot have greatly inflated our estimate of average
citizens’ influence on policy making, which is near zero.
What do our findings say about democracy in America? They certainly constitute
troubling news for advocates of “populistic” democracy, who want governments to respond
primarily or exclusively to the policy preferences of their citizens. In the United States, our
Gilens and Page Testing Theories of American Politics 23
23
findings indicate, the majority does not rule -- at least not in the causal sense of actually
determining policy outcomes. When a majority of citizens disagrees with economic elites and/or
with organized interests, they generally lose. Moreover, because of the strong status quo bias
built into the U.S. political system, even when fairly large majorities of Americans favor policy
change, they generally do not get it.
A possible objection to populistic democracy is that average citizens are inattentive to
politics and ignorant about public policy; why should we worry if their poorly informed
preferences do not influence policy making? Perhaps economic elites and interest group leaders
enjoy greater policy expertise than the average citizen does. Perhaps they know better which
policies will benefit everyone, and perhaps they seek the common good, rather than selfish ends,
when deciding which policies to support.
But we tend to doubt it. We believe instead that – collectively – ordinary citizens
generally know their own values and interests pretty well, and that their expressed policy
preferences are worthy of respect.50 Moreover, we are not so sure about the informational
advantages of elites. Yes, detailed policy knowledge tends to rise with income and status.
Surely wealthy Americans and corporate executives tend to know a lot about tax and regulatory
policies that directly affect them. But how much do they know about the human impact of Social
Security, Medicare, Food Stamps, or unemployment insurance, none of which is likely to be
crucial to their own well-being? Most important, we see no reason to think that informational
expertise is always accompanied by an inclination to transcend one's own interests or a
determination to work for the common good.
All in all, we believe that the public is likely to be a more certain guardian of its own
interests than any feasible alternative.
Leaving aside the difficult issue of divergent interests and motives, we would urge that
the superior wisdom of economic elites or organized interest groups should not simply be
assumed. It should be put to empirical test. New empirical research will be needed to pin down
precisely who knows how much, and what, about which public policies.
Our findings also point toward the need to learn more about exactly which economic
elites (the “merely affluent”? the top 1%? the top 0.01%?) have how much impact upon public
policy, and to what ends they wield their influence. Similar questions arise about the precise
extent of influence of particular sets of organized interest groups. And we need to know more
about the policy preferences and the political influence of various actors not considered here,
including political party activists, government officials, and other non-economic elites. We hope
that our work will encourage further exploration of these issues.
Gilens and Page Testing Theories of American Politics 24
24
Despite the seemingly strong empirical support in previous studies for theories of
majoritarian democracy, our analyses suggest that majorities of the American public actually
have little influence over the policies our government adopts. Americans do enjoy many features
central to democratic governance, such as regular elections, freedom of speech and association,
and a widespread (if still contested) franchise. But we believe that if policymaking is dominated
by powerful business organizations and a small number of affluent Americans, then America’s
claims to being a democratic society are seriously threatened.
Gilens and Page Testing Theories of American Politics 25
25
Table 1. Theoretical Predictions Concerning the Independent Influence of Sets of
Actors upon Policy Outcomes
Sets of Actors
All Mass Business
Average Economic Interest Interest Interest
Citizens Elites Groups Groups Groups
Theory (ideal type)
Majoritarian Y n n n n
Electoral
Democracy
________________________________________________________________________
Dominance by
Economic Elites y Y y n y
________________________________________________________________________
Majoritarian y n Y Y Y
Pluralism
________________________________________________________________________
Biased Pluralism n n y y Y
________________________________________________________________________
n: little or no independent influence
y: some independent influence
Y: substantial independent influence
Gilens and Page Testing Theories of American Politics 26
26
Table 2. Correlations among Independent Variables
Average
citizens’
preferences
Economic
elites’
preferences
All
interest
groups
Mass public
interest
groups
Business
interest
groups
Average citizens’
preferences
Economic elites’ preferences
All interest groups
Mass public interest groups
Business interest groups
--
.78***
.04
.12***
-.10***
--
.05
.01
-.02
--
.47***
.96***
--
-.05
--
*** p<.001; n=1779.
Note: Entries are correlation coefficients corrected for measurement error as explained in
Appendix 2.
Gilens and Page Testing Theories of American Politics 27
27
Table 3. Policy Outcomes and the Policy Preferences of Average Citizens, Economic
Elites, and Interest Groups
Model 1 Model 2 Model 3 Model 4
Preferences of
average citizens
Preferences of
economic elites
Alignment of
Interest groups
.64 (.08) ***
--
--
--
.81 (.08) ***
--
--
--
.59 (.09) ***
.03 (.08)
.76 (.08) ***
.56 (.09) ***
R-sq
.031
.049
.028
.074
***p<.001
Note: All predictors are scaled to range from 0 to 1. The dependent variable is the policy
outcome, coded 1 if the proposed policy change took place within four years of the survey
date and 0 if it did not. Predictors are the logits of the imputed percent of respondents at the
50th (“average citizens”) or 90th (“economic elites”) income percentile that favor the
proposed policy change, and the Net Interest Group Alignment Index described in the text.
Standard errors are asymptotically distribution-free, and all analyses reflect estimated
measurement error in the predictors, as described in Appendix 2. The standardized
coefficients for model 4 in this table are .01, .21, and .16 for average citizens, economic
elites, and interest groups, respectively. N=1779.
Gilens and Page Testing Theories of American Politics 28
28
Table 4. The Separate Policy Impact of Business-oriented and Mass-based Interest Groups
Average citizens’ preferences
Economic elites’ preferences
Mass-based interest groups
Business interest groups
.05 (.08)
.78 (.08) ***
.24 (.07) ***
.43 (.08)***
N
R-sq
1,779
.07
***p<.001
Note: All predictors are scaled to range from 0 to 1. The dependent variable is the policy
outcome, coded 1 if the proposed policy change took place within four years of the survey
date and 0 if it did not. Predictors are the logits of the imputed percent of respondents at the
50th (“average citizens”) or 90th (“economic elites”) income percentile that favor the
proposed policy change, and the Net Interest Group Alignment Indices described in the text.
Standard errors are asymptotically distribution-free, and all analyses reflect estimated
measurement error in the predictors, as described in Appendix 2. N=1779.
Gilens and Page Testing Theories of American Politics 29
29
Figure 1. Predicted probability of policy adoption (dark lines, left axes)
by policy disposition; the distribution of preferences (gray columns, right axes)
Gilens and Page Testing Theories of American Politics 30
30
Appendix 1.
Business- and Mass-Based Interest Groups Included in Net Group Alignment Indices
Business and professional groups
Airlines
American Bankers Association
American Council of Life Insurance
American Farm Bureau Federation
American Hospital Association
American Medical Association
Association of Trial Lawyers
Automobile companies
Chamber of Commerce
Computer software and hardware
Credit Union National Association
Defense contractors
Electric companies
Health Insurance Association
Independent Insurance Agents of America
Motion Picture Association of America
National Association of Broadcasters
National Association of Home Builders
National Association of Manufacturers
National Association of Realtors
National Beer Wholesalers Association
National Federation of Independent Business
National Restaurant Association
Oil Companies
Pharmaceutical Research & Manufacturers
Recording Industry Association
Securities and investment companies
Telephone companies
Tobacco companies
Mass-based groups
AARP
AFL-CIO
American Federation of State, County, and Municipal Employees
American Israel Public Affairs Committee
American Legion
Christian Coalition
International Brotherhood of Teamsters
National Rifle Association
National Right to Life Committee
United Auto Workers union
Veterans of Foreign Wars of the U.S.
Not coded as either business or mass-based
National Education Association (includes a mass base of teachers but also university professors)
National Governors' Association (affected by interest groups rather than acting as an independent group)
Universities (unclear status as businesses or nonprofits)
Gilens and Page Testing Theories of American Politics 31
31
Appendix 2. Correcting for Measurement Error
Measurement error is unavoidable in social science data. Its consequences depend on
both the magnitude and the nature of error and, in particular, on whether the errors in
measurement of one variable are correlated with the errors of other variables in a given model.
The predictors used in our analyses come from two distinct data sources: aggregate
survey data, for estimating the policy preferencs of median- and high-income Americans; and
historical data on interest group issue engagement, for generating the Net Interest Group
Alignment Indices. Because our measures of interest group alignments are entirely distinct from
the preference measures in origin, there is no reason to expect that their measurement errors
would be correlated across policies with those for our other two independent variables. In the
case of interest group policy dispositions, then, we have only uncorrelated (random or
“classical”) measurement error to consider. We estimate this error based on inter-coder reliability
(Chronbach’s alpha) of .87.
But our two other independent variables come from a single data source, the same set of
national opinion surveys. Consequently, errors in our measures of the policy preferences of
average citizens and those of economic elites are likely to be positively correlated with each
other. Errors affecting both measures in the same way within a particular survey can arise from
such factors as sampling, question wording, question order, topics in the news at the time the
survey was in the field, and so on.51 Correlated errors on these preference measures tend to
produce upwardly biased statistical relationships between the two independent variables,
producing a higher uncorrected correlation between them (r=.94) than the corrected r=.78 shown
in Table 2.
In order to assess the magnitude of correlated errors, multiple measures of the same
underlying concept or attitude are needed. These are available for a subset of 387 of the 1,779
proposed policy changes in our dataset, for which which more than one survey question
addressed essentially the same policy issue and was asked in the same calendar year. Using the
subset of proposed changes with multiple measures, the error covariance between preferences of
median and high income Americans was estimated to account for 17 percent of the total
covariance between these measures.52
Using the measurement error estimates described above, we estimated structural equation
models in AMOS that purged of error the structural coefficients representing the associations of
the predictors with our outcome measure. The resulting coefficients are reported in tables 3 and 4
and are translated into predicted probabilities in figure 1.
Gilens and Page Testing Theories of American Politics 32
32
For comparison, appendix table A1 shows the results from an ordinary least squares
regression that parallels the multivariate model in table 3 but in which measurement error is not
taken into account. The implausible negative coefficient for average citizens and the impluasibly
large positive coefficient for economic elites suggests the presence of correlated measurement
error.53
As noted above, the corrections for measurement error described here can help reduce
flaws in our measures that affect the relationship between our indictors and their underlying
concepts. But these corrections cannot address the flaws that arise from the imperfect fit between
those concepts and the characteristics we would prefer to measure (such as our use of the 90th
income perenctile as a proxy for economic elites).
Gilens and Page Testing Theories of American Politics 33
33
Table A1. Ordinary Least Squares Analysis Parallel to the Structural Equation Model
Presented in Table 3.
Compare with table 3, model 4
Preferences of average citizens
Preferences of economic elites
Alignment of Interest groups
-.93 (.24) ***
1.66 (.24) ***
.46 (.07) ***
N
R-sq
1,779
.08
***p<.001
Note: All predictors are scaled to range from 0 to 1. The dependent variable is the
policy outcome, coded 1 if the proposed policy change took place within four years
of the survey date and 0 if it did not. Predictors are the logits of the imputed percent
of respondents at the 50th (“average citizens”) or 90th (“economic elites”) income
percentile that favor the proposed policy change, and the Net Interest Group
Alignment Indices described in the text.
Gilens and Page Testing Theories of American Politics 34
34
References
Achen, Christopher H. 1985. Proxy Variables and Incorrect Signs on Regression Coefficients.
Political Methodology 11: 299-317.
Arrow, Kenneth J. 1963. Social Choice and Individual Values. 2nd. ed. [orig. pub. 1951]. New
York: Wiley.
Bachrach, Peter, and Morton S. Baratz. 1962. Two Faces of Power. American Political Science
Review 56(4): 947-952.
Bartels, Larry M. 2008. Unequal Democracy: The Political Economy of the New Gilded Age.
New York: Russell Sage Foundation and Princeton University Press.
Bauer, Raymond A., Ithiel de Sola Pool, and Lewis Anthony Dexter. 1967. American Business
and Public Policy: The Politics of Foreign Trade. New York: Atherton.
Baumgartner, Frank R., Jeffrey M. Berry, Marie Hojnacki, David C. Kimball, and Beth L. Leech.
2009. Lobbying and Policy Change: Who Wins, Who Loses, and Why. Chicago:
University of Chicago Press, 2009.
Beard, Charles. 19`3. An Economic Interpretation of the Constitution of the United States.
Bentley, Arthur F. 1908. The Process of Government: A Study of Social Pressures. Chicago:
University of Chicago Press.
Berry, Jeffrey M. 1999. The New Liberalism: The Rising Power of Citizen Groups.
Washington, D.C.: Brookings.
Black, Duncan. 1948. On the Rationale of Group Decision-making. Journal of Political
Economy 56: 23–34.
Black, Duncan. 1958. The Theory of Committees and Elections. Cambridge:Cambridge
University Press.
Block, Fred. 1977. The Ruling Class Does Not Rule: Notes on the Marxist Theory of the State.
Socialist Revolution 7(3): 6-28.
Block, Fred. 2007. Understanding the Divergent Trajectories of the United States and Western
Europe: A Neo-Polanyian Analysis. Politics and Society (March): 1-31.
Burch, Philip H. 1980-81. Elites in American History. 3 vols. New York: Holmes & Meier.
Canes-Wrone, Brandice, Michael C. Herron, and Kenneth W. Shotts. 2001. Leadership and
Pandering: A Theory of Executive Policymaking. American Journal of Political Science
45: 532-50.
Dahl, Robert A. 1956. A Preface to Democratic Theory. Chicago: University of Chicago Press.
Dahl, Robert A. 1961. Who Governs? New Haven: Yale University Press.
Dahl, Robert A. 1989. Democracy and its Critics. New Haven: Yale University Press.
Davis, Otto A., Melvin J. Hinich, and Peter C. Ordeshook. 1970. An Expository Development of
a Mathematical Model of the Electoral Process. American Political Science Review 64:
426-48.
Domhoff, G. William. 2013. Who Rules America: The Triumph of the Corporate Rich. 7th. ed.
New York: McGraw-Hill.
Gilens and Page Testing Theories of American Politics 35
35
Downs, Anthony. 1957. An Economic Theory of Democracy. New York: Harper & Row.
Erikson, Robert S., Michael B. MacKuen, and James A. Stimson. 2002. The Macro Polity. New
York: Cambridge University Press.
Erikson, Robert S., Gerald C. Wright, and John P. McIver. 1993. Statehouse Democracy:
Public Opinion and Policy in the American States. New York: Cambridge University
Press.
Ferguson, Thomas. 1995. Golden Rule: The Investment Theory of Party Competition and the
Logic of Money-Driven Political Systems. Chicago: The University of Chicago Press.
Fiorina, Morris. 1981. Retrospective Voting in American National Elections. New Haven: Yale
University Press.
Gilens, Martin. 2011. “Two-thirds full? Citizen Competence and Democratic Governance.” In
New Directions in Public Opinion, ed., Adam Berinsky, pp.52-76. New York: Routledge.
Gilens, Martin. Affluence and Influence: Economic Inequality and Political Power in America.
2012. New York: Russell Sage Foundation and Princeton University Press.
Hacker, Jacob S., and Paul Pierson. 2010. Winner-Take-All Politics: How Washington Made the
Rich Richer – and Turned its Back on the Middle Class. New York: Simon & Schuster.
Hamilton, Alexander, James Madison, and John Jay. 1961 [orig. pub. 1787-88]. Ed. Clinton
Rossiter. New York: New American Library.
Hotelling, Harold. 1929. Stability in Competition. The Economic Journal 39: 41-57.
Isaac, Jeffrey. 1987a. Beyond the Three Faces of Power: A Realist Critique. Polity 20(1): 4-31.
Isaac, Jeffrey. 1987b. Power and Marxist Theory. Ithaca: Cornell University Press.
Jacobs, Larry R., and Benjamin I. Page. 2005. Who Influences U.S. Foreign Policy? American
Political Science Review 99: 107-123.
Lukes, Steven. 1974. Power: A Radical View. London: Macmillan.
Katznelson, Ira. 2013. Fear Itself: The New Deal and the Origins of Our Time. New York:
Liveright.
Key, V.O., Jr. 1961. Public Opinion and American Democracy. New York: Knopf.
Lindblom, Charles E. 1977. Politics and Markets. New York: Basic.
Lincoln, Abraham. 1863. Address at Gettysburg, Pennsylvania. In Speeches and Writings
1859-1865. The Library of America, New York, (1989), p.536.
McChesney, Fred S. 1997. Money for Nothing: Politicians, Rent Extraction, and
Political Extortion. Cambridge: Harvard University Press.
McConnell, Grant. 1966. Private Power and American Democracy. New York: Random
House.
McKelvey, Richard. 1976. Intransitivities in Multidimensional Voting Models
and Some Implications for Agenda Control. Journal of Economic Theory 12: 472-82.
Mansbridge, Jane. 2003. Rethinking Representation. American Political Science Review 97:
515-28.
Marx, Karl, and Friedrich Engels. 1972 [1848, 1888]. Manifesto of the Communist Party. In
The Marx-Engels Reader, ed. Robert C. Tucker (New York, W.W. Norton), pp. 331-362.
Gilens and Page Testing Theories of American Politics 36
36
See also http://www.marxists.org/archive/marx/works/1848/communistmanifesto/
ch01.htm
May, Kenneth O. 1952. A Set of Independent Necessary and Sufficient Conditions for Simple
Majority Decisions, Econometrica 20: 680–684.
Michels, Robert. 2001. Political Parties: A Sociological Study of the Oligarchical Tendencies of
Modern Democracy. Kitchener, Ontario: Batoche Books. Orig. pub. 1911.
Milbrath, Lester W. 1963. The Washington Lobbyists. Chicago: Rand-McNally.
Miliband, Ralph. 1969. The State in Capitalist Society. New York: Basic.
Mills, C. Wright. 1959. The Power Elite. New York: Oxford University Press (Galaxy edition),
orig. pub 1956.
Monroe, Alan D. 1979. Consistency between Public Preferences and National Policy Decisions.
American Politics Quarterly 7: 3-18.
Monroe, Alan D. 1998. Public Opinion and Public Policy 1980-1993. Public Opinion Quarterly
68: 6-28.
Olson, Mancur, Jr. 1965. The Logic of Collective Action: Public Goods and the Theory of
Groups. Cambridge: Harvard University Press.
Page, Benjamin I., Larry M. Bartels, and Jason Seawright. 2013. Democracy and the Policy
Preferences of Wealthy Americans. Perspectives on Politics 11(1): 51-73.
Page, Benjamin I., and Cari Lynn Hennessy. 2008. What Affluent Americans Want from
Politics. WP-11-08. Institute for Policy Research, Northwestern University.
Page, Benjamin I., and Lawrence R. Jacobs. 2009. Class War? What Americans Really Think
about Economic Inequality. Chicago: University of Chicago Press.
Page, Benjamin I., and Robert Y. Shapiro. 1983. Effects of Public Opinion on Policy.”
American Political Science Review 77: 175-90.
Page, Benjamin I., and Robert Y. Shapiro. 1992. The Rational Public: Fifty Years of Trends in
Americans’ Policy Preferences. Chicago: University of Chicago Press.
Poole, Keith, and Howard Rosenthal. 2007. Ideology and Congress. New Brunswick, NJ:
Transaction Publishers.
Przeworksi, Adam, and Michael Wallerstein. 1982. The Structure of Class Conflict in
Democratic Capitalist Societies. American Political Science Review 76:215-238.
Schattschneider, E.E. 1960. The Semisovereign People: A Realist’s View of Democracy in
America. New York: Holt, Rinehart and Winston.
Schattschneider, E.E. 1935. Politics, Pressures and the Tariff. New York: Prentice-Hall.
Schlozman, Kay Lehman, Sidney Verba, and Henry E.Brady. 2012. The Unheavenly
Chorus:Unequal Political Voice and the Broken Promise of American Democracy.
Princeton, N.J.: Princeton University Press.
Sen, Amartya K. 1970. Collective Choice and Social Welfare. San Francisco: Holden-Day.
Skocpol, Theda. 1992. Protecting Soldiers and Mothers: The Political Origins of Social Policy
in the United States. Cambridge, MA: Harvard University Press.
Gilens and Page Testing Theories of American Politics 37
37
Skocpol, Theda, and Kenneth Finegold. 1982. State Capacity and Economic Intervention in the
Early New Deal. Political Science Quarterly 97: 255-278.
Smith, Mark A. 2000. American Business and Political Power: Public Opinion, Elections, and
Democracy. Chicago: University of Chicago Press.
Stigler, George J. 1971. The Theory of Economic Regulation. The Bell Journal of Economics
and Management Science 2(1): 3-21.
Stimson, James A., Michael B. MacKuen, and Robert S. Erikson. 1995. Dynamic
Representation. American Political Science Review 89: 543-65.
de Tocqueville, Alexis. 2000 [1835, 1840]. Democracy in America. Trans. and ed. Harvey C.
Mansfield and Delba Winthrop. Chicago: University of Chicago Press.
Truman, David B. 1971. The Governmental Process. 2nd. ed. [first ed. pub. 1951]. New York:
Alfred A. Knopf.
Winters, Jeffrey A. 2011. Oligarchy. New York: Cambridge University Press.
Winters, Jeffrey A., and Benjamin I. Page. 2009. Oligarchy in the United States? Perspectives
on Politics 7(4): 731-51.
Wolfinger, Raymond E., and Steven J. Rosenstone. 1080. Who Votes? New Haven: Yale
University Press.
Zaller, John. 2003. Coming to Terms with V.O. Key’s Concept of Latent Opinion. In Electoral
Democracy, ed. Michael B. MacKuen and George Rabinowitz. Ann Arbor: University of
Michigan Press.
Gilens and Page Testing Theories of American Politics 38
38
Notes
1 Tocqueville 2000 [1835, 1840], 235-49. Tocqueville focused on state governments, which he
said “really direct[ed]” American society; he noted that the Founders had imposed limits on
direct democracy in the federal government (235, fn 1.) Yet he asserted in general terms that
“[t]he laws of democracy...emanate from the majority of all citizens” (222.)
2 Lincoln 1863. Dahl 1956, ch. 2, defined “populistic democracy” in terms of pure majority rule,
traced its theoretical roots to Aristotle’s political equality, Locke’s “majority power,” Rousseau’s
“general will” of the majority, and James Madison’s “republican principle,” and critically
analyzed its normative properties.
3 Hotelling 1929; Black 1948, 1958; Downs 1957. In his full-information ch. 2, Downs offers a
clever though somewhat indeterminate non-dimensional version of the theory; in ch. 8 he
explicates a variant of Hotelling’s single-dimensional version. For extensions to multiple
dimensions see Davis, Hinich, and Ordeshook 1970. May’s Theorem establishes that simple
majority rule is the only collective decision rule for choosing between two alternatives that
satisfies the Arrow-type democratic conditions of decisiveness, anonymity, neutrality, and
positive responsiveness to individuals’ preferences (May 1952).
4 Arrow 1963 [1951], McKelvey 1976. See Sen 1970.
5 A particularly trenchant critique is given in Ferguson 1995 (Appendix: “Deduced and
Abandoned”).
6 Monroe 1979, 1998; Page and Shapiro 1983; Erikson, Wright and McIver 1983; Stimson,
MacKuen, and Erikson 1995; Erikson, MacKuen, and Stimson 2002.
7 Bartels 2008, Gilens 2012. See also Jacobs and Page 2005, which indicates that the general
public may have little or no influence on U.S. foreign policy, when the preferences of business
leaders and other elites are taken into account.
8 Key 1961, ch. 11 and 472-76; Fiorina 1981; Zaller 2003. A variant on this logic that focuses
on the possibility of politicians “pandering” to current preferences under certain conditions is
given in Canes-Wrone, Herron, and Shotts 2001. “Latent” preferences are the collective policy
preferences that citizens would derive from their basic needs and values if they had accurate
information about the future unfolding of policy results. Such “latent” preferences are related to
the ideal, “authentic” preferences that some political theorists see as superior to, and more
worthy of governmental responsiveness than, the policy preferences that citizens actually
express. This can be seen as adding normative appeal to an electoral-reward-and- punishment
system of democratic control. See Mansbridge 2003.
9 See Page and Shapiro 1992.
10 Mills 1959, ch. 12, esp. p. 279. Robert Michels 2001 [1911], a founder of modern elite theory,
emphasized positions of leadership in political parties.
Some of Theda Skocpol’s early work (e.g., Skocpol and Finegold 1982) can be seen as
based on a state-centric elite theory that emphasizes public officials. Beginning with Protecting
Soldiers and Mothers (1992), however, Skocpol’s analyses of American politics have paid more
attention to non-state elites, social movements, organized interest groups, and the preferences of
the general public. Thus Skocpol’s work, like that of a number of other important scholars of
American politics (e.g., Katznelson 2013), does not fit neatly into our simple theoretical
categories.
Gilens and Page Testing Theories of American Politics 39
39
11 Beard 1913; Domhoff 2013; Burch 1980-1981; Ferguson 1995; Winters 2011; Winters and
Page 2009.
The boundary between elite theories that focus on economically elite individuals, and
interest group theories that focus on organized corporate interests (discussed below), is not
always a sharp one. Here we treat most theories that emphasize corporate organizations or
industrial sectors (e.g., Block 2007, Ferguson 1995) as primarily constituting “interest group”
rather than elite theories. We categorize self-identified elite theorists like Domhoff as such even
if they emphasize business elites and treat corporate organizations as important mechanisms of
influence. Winters 2011 may come closest to an exclusive focus on wealthy individuals rather
than organizations.
One might argue that the economic classes central to classical Marxist theories amount to
“economic elites.” But Marxist theorists see class position as only imperfectly related to wealth
or income, and their focus on ownership of the means of production suggests that business firms
and business associations may be the key political actors. Hence we discuss these theories below
in connection with corporations and organized interest groups, noting the kinship of
“instrumentalist” Marxist theories to theories of biased pluralism.
12 The term “interest group” tends to evoke images of membership groups like the League of
Women voters or the National Rifle Association, but many politically important “groups” are in
fact business corporations.
13 Madison, Hamilton, and Jay 1961 [1787-88], 77-84.
14 Bentley 1908; Truman 1971 [1951], esp. p.511.
15 Dahl 1956. 1989. Berry 1999 emphasizes the rising power of “citizen groups.”
16 Olson 1965.
17 Schattschneider 1960, p.35; Schlozman, Verba, and Brady 2012, ch. 10-14.
18 Schattschneider 1960, ch. 1; McConnell 1966; Stigler 1971; Lindblom 1977, Parts IV and V;
Ferguson 1995; Block 2007; Hacker and Pierson 2010.
19 Marx and Engels 1972 [1848]. This English-language text comes from the 1888 publication
edited by Engels.
20 Miliband 1969, ch. 6. Fred Block (1977) makes a critical distinction between
“instrumentalist” Marxist theories like Miliband’s, in which politically conscious members of the
ruling class use their economic resources to shape state action in their own material interests, and
“structural” theories, in which the capitalist economic system itself tends to shape state policies
and the preferences of its citizens – including workers, who are compelled to accept low wages
and high capitalist profits for the sake of future investment and growth. On alternative Marxist
theories of power, see also Isaac 1987b. For a formalized structural Marxist theory, see
Przeworski and Wallerstein 1982.
21 Again: our data can do little to distinguish among these or other mechanisms of political
influence. We focus on possible sources of influence among individuals and groups in society.
22 Schlozman, Verba, and Brady 2012, ch. 10-14, esp. pp. 321, 329, 356.
23 Schattschneider 1935.
24 Bauer, Pool, and Dexter 1963; Milbrath 1963; McChesney 1997.
25 Smith 2000; Baumgartner et al. 2009.
26 Smith 2000, ch.3.
Gilens and Page Testing Theories of American Politics 40
40
27 Numerical success rates are not reported in Smith 2000, but the “enactment scorecard” line in
Figure 4.1 (p.83) appears to show Chamber success on 60% or more of the bills in most years,
with very substantial variation from year to year.
28 Baumgartner et al. 2009, 233, 235. These multivariate results may be biased downwards
because the regressions include as independent variables congressional and executive branch
officials’ active support for (or opposition to) policy changes. Since officials’ behavior may well
have been influenced by the interest groups themselves, the inclusion of these predictors restricts
the estimates of group influence to direct effects, excluding any indirect impact that was
channeled through interest groups’ influence over officials.
On the other hand, the omission of other influential actors from the analysis could (if their
preferences were positively correlated with those of interest groups) produce spuriously inflated
estimates of interest group influence. A further complexity in assessing interest group influence
involves policy cases in which groups take no stand at all, which are not included in the analysis
by either Smith or Baumgartner et al. Inclusion of no-stand cases would be necessary if one
sought to assess the extent to which groups affect over-all policy results – though not for
assessing the extent of group influence in the cases where influence attempts are actually made.
29 Appreciation for their fine work in assisting with the colossal task of collecting and coding
these data goes to Marty Cohen, Jason Conwell, Andrea Vanacore, and Mark West at UCLA,
and Oleg Bespalov, Daniel Cassino, Kevin Collins, Shana Gadarian, Raymond Hicks, and Lee
Shaker at Princeton.
30 Arguments for the normative and empirical relevance of the “survey agenda” are discussed in Gilens 2012, 50-
56. Note that if (as we find) the public has little or no influence on our issues, which tilt toward high salience, it is
unlikely that the public has much influence on more technical or obscure matters.
31 In 2012, the 10th percentile of household income was about $12,200, the 50th percentile about
$51,000, and the 90th percentile about $146,000 (U.S. Census Bureau, “Income, Poverty, and
Health Insurance Coverage: 2012,” September 2013,
http://www.census.gov/newsroom/releases/pdf/20130917_ip_slides_with_plotpoints.pdf). For
further detail on these data, See Gilens 2012, 57-66.
32 Of course the average (median) citizen is not identical to the eligible or actual “median voter.”
But the generally small magnitude of differences between the policy preferences of voters and
those of all citizens (Wolfinger and Rosenstone 1980, 109-114; Schlozman, Verba, and Brady
2012, 120-21) suggests that our measure captures the spirit of median voter theories. To the
extent that differences exist, we consider the average citizen to be of greater normative – though
less empirical – importance.
33 To make an approximate assessment of the fit between the preferences of the medianpreference
respondent and those of the median-income respondent we calculated the median
preference within each of five income groups (at the 10th, 30th, 50, 70th and 90th percentiles).
For example, if, on a particular non-monotonic item, the five income groups had imputed median
preferences of .50, .60, .70 .65 and .55 (for the 10th to 90th income percentiles, respectively),
then the estimated over-all median preference would be .60 (in this case equaling the preference
at the 30th income percentile). In most cases the over-all median preference is the same as the
median income (50th percentile) preference. When it is not, the preference differences across
income levels tend to be small. Using this technique, the median over-all preferences and the
median-income preferences track each other very closely: r=.997.
Gilens and Page Testing Theories of American Politics 41
41
34 This study is one of the few surveys that meets the two criteria of having a sufficiently large
sample size (n=20,150) and a high enough top-coded income category to provide information on
the policy preferences of very affluent Americans.
35 Correlations of the policy preferences of the top 2% with those of the top 10% are based on the
76 CCES respondents who reported at least $350,000 in family income. Using the 179 CCES
respondents who reported at least $250,000 (roughly the top 4.5% of the income distribution),
the corresponding correlations are .97 and .76.
Some corroborating evidence comes from a comparison of the Survey of Economically
Successful Americans (Page, Bartels and Seawright 2013), based on a local sample of the
wealthiest 1% or 2% of Americans, and the Inequality Survey (Page and Jacobs 2009) which was
based on a representative sample of the American public. Eight policy preference questions that
were included on both surveys showed that the preferences of the top 25% of income earners
generally fell about half way between those of the average citizen and those of the wealthy. For
similar findings concerning the policy preferences of the top 4% or so of income earners, see
Page and Hennessy 2008.
36 We distinguish conceptually between two sorts of deficiencies in our measures: flaws that
affect the relationship between our indictors and their underlying concepts (such as the random
and correlated measurement errors we discuss in the appendix), and flaws that arise from the
imperfect fit between those concepts and the characteristics we would prefer to measure. For
example, the adjustments described in the appendix help us to improve our estimates of the
preferences of Americans at the 90th income percentile, but they cannot help to make those
estimates any more accurate as indicators of the preferences of the truly rich Americans whose
views we would prefer to include in our models.
37 Baumgartner et al. 2009, 225. We believe that our measure of net interest group alignment
(described below) is actually superior to the Fortune 25 proxy examined by Baumgartner et al.
because it includes industries that do not lobby through centralized trade organizations, it is
nonlinear in net number of groups, and it reflects the extent to which a particular issue is central
to the concerns of an interest group or industry.
38 For more detail on the Index of Net Interest Group Alignment, see Gilens 2012, 127-130.
39 On the measurement of policy change, see Gilens 2012, p.60 and endnote #18, p.284.
40 For correlations of individual groups’ positions with average citizens’ preferences see Gilens
2012, 156-57.
41 These particular values for low and high levels of support among affluent Americans and
interest groups were chosen because about fifteen percent of all proposed policy changes
generated either less than 20% or more than 80% support among the affluent, and about fifteen
percent of all proposed changes on which interests groups took a position generated a raw net
interest group score of either more than 5 groups strongly in favor or more than five groups
strongly opposed (counting “somewhat” favorable or opposed as one-half of a group).
42 See Monroe 1979, 1998; Page and Shapiro 1983; Stimson, MacKuen, and Erikson 1995.
43 Perhaps counterintuitively, it turns out that business groups tilted somewhat less toward
opposing proposed changes (33% opposed, 26% in favor) than mass-based groups did (38%
opposed, 20% in favor.)
44 Even if existing organized groups did not reflect the wishes of average citizens, officials’
anticipated reactions to unformed “potential groups” might in theory provide some
representation for average citizens, as David Truman argued they do. But our Table 3 finding of
Gilens and Page Testing Theories of American Politics 42
42
negligible independent influence by average citizens when existing organized interest groups’
positions are controlled for, leaves little room for potential groups (and officials’ anticipation of
them) to advance the preferences of ordinary citizens.
45 Schlozman, Verba, and Brady 2012, ch. 10-14, esp. pp. 321, 329, 356.
46 For the 369 out of 1,779 cases in which both business-based and mass-based interest groups
took a stand, the coefficients are just .09 (n.s.) for mass groups but .48** (p<.001) for business
groups.
47 For those proposed policy changes on which at least one business-oriented group took a
position, another business-oriented group was found on the opposite side less than five percent of
the time. Interestingly, mass-based groups were somewhat more likely to take stands on both
sides of an issue, for example pro-life and pro-choice groups on abortion, or the AARP which
opposed the Clinton health reform plan and the AFL-CIO that favored it.
48 For the analysis in table 4, both the business and mass interest group indices were scaled to run
from 0 to 1. When we rescaled these indices to reflect the differing numbers of business and
mass-based groups engaged on each issue, the standard deviation of the business alignment index
was .158 and the mass-oriented index .096; their associations with policy outcomes (analogous
to those shown in table 4) were almost identical, at .38 (p<.01) and .40 (p<.001), respectively.
49 Bachrach and Baratz 1962, Lukes 1974. See Isaac 1987a.
50 On the normative argument, see Dahl 1989, esp. ch. 7. For empirical evidence that its
conditions tend to be satisfied, see Page and Shapiro 1992 and Gilens 2011.
51 Since these factors do not affect the interest group coding we are less concerned about possible
correlated error between the mass and business interest group indices. See O'Brien, Robert M.
"Correcting Measures of Relationship between Aggregate-Level Variables." Sociological
Methodology 21 (1994): 125-65 and O'Brien, Robert M. "Correcting Measures of Relationship
between Aggregate-Level Variables for Both Unreliability and Correlated Errors: An Empirical
Example." Social Science Research 27, no. 2 (Jun 1998): 218-34.
52 See pp.62-66 and the appendix to Gilens 2012 for details on this calculation and the
comparability of the 387 survey questions with multiple meaures with the full set of 1,779 items.
53 For a discussion of diagnosing and correcting for correlated measurement error see Achen
1985 and Gilens 2012, pp.253-258.
