Verba
Verba
Verba
POLITICAL EQUALITY
Sidney Verba
Harvard University
Je participe
Tu participes
Il participe
Nous participons
Vous participez
Ils profitent1
1
Epigraph from Verba and Nie, Participation in America (1972
2
Of the various ways in which citizens in the United States can be unequal, political
inequality is- one of the most significant and troubling. By political equality we refer to the
extent to which citizens have an equal voice over governmental decisions. One of the bedrock
principles in a democracy is the equal consideration of the preferences and interests of all
citizens. This is expressed in such principles as one-person/one-vote, equality before the law,
and equal rights of free speech. Equal consideration of the preferences and needs of all citizens
is fostered by equal political activity among citizens; not only equal voting turnout across
significant categories of citizens but equality in other forms of activity. These activities include
work in a political campaign, campaign contributions, activity within one=s local community,
direct contact with officials, and protest. Equal activity is crucial for equal consideration since
political activity is the means by which citizens inform governing elites of their needs and
preferences and induce them to be responsive. Citizen participation is, thus, at the heart of
political equality. Through their activity citizens in a democracy seek to control who will hold
public office and to influence what the government does. Political participation provides the
mechanism by which citizens can communicate information about their interests, preferences,
2
This paper draws on previous work by myself and my collaborators, in particular on
Verba (2000), Verba, Schlozman and Brady (1995), Verba and Nie (1972), Verba, Nie and Kim
(1979), and Burns, Schlozman and Verba (forthcoming, 2001).
3
Equality in all domains of social and political life is complex. It can be about many
different valued goods (income, education, health, etc.), it can be across individuals or groups, it
can be calculated with different measures, and on the basis of different criteria. There are, as the
title of Douglas Rae=s book, Equalities, makes clear, many forms and versions of it. And in most
of its forms, it is something of mixed value. For most valued things -- income or education or
health or respect or political influence -- gross inequalities are something we dislike. But
process of attaining it would be too costly, or its consequences would be negative. It thus may be
useful to consider why political equality -- defined roughly as equal influence over government
policy across citizens -- might be desirable within the framework of democratic governance.
C Political equality is a valued good per se. The ability to express one=s political views is
Put another way: There are some who denigrate the importance of voting, since voting rights
and voting participation, when achieved, (as among blacks in the American South or in South
Africa) does not bring with it the solution to all or even most problems. But those who denigrate
the importance of the vote are almost certainly people who already have the right to vote.
C Political equality builds community: societies are bound together by cooperative activity
toward shared goals. This is how that precious commodity of social capital is formed.
election outcomes by the losing side, etc. That the laws or electoral outcomes one might
not favor deserve respect derives from the fact that they were selected, through proper
procedures, by the people (or at least the larger number of the people.)
C Political participation is educative: people learn about politics and about democracy and
about their own needs and preferences through participation for all. Equality in political
speaking up or in other ways -- are more likely to have government policies that pay
attention to their needs and preferences. In this sense, political voice represents a general
capacity to achieve many goals. Equality in such general capabilities is, as Amartya Sen
has pointed out, a basic form of equality. Democracy implies equal consideration of the
needs and preferences of all citizens. This instrumental aspect of political equality -- the
ability to inform the government of one=s needs and preferences and to pressure the
Note that some of these reasons for wanting equal citizen participation rest more heavily
on the participation part of the term and others on the equality part. Conveying a sense of
membership or building community or creating legitimacy would wiulkd seem to depend on the
level of participation. Democracy seems weaker if few take part. On the other hand, the equal
protection of interests depends more on who participates. The the voice of the public were
conveyed via a random sample of the public, it might suffice. (Why sample surveys, which can
approximate a randomly selected voice, do not not quite do this, is an interesting issue which I
5
will not pursue here.) (But see, Verba, 1995 and Brehm, xxxx)
True political equality, where all ordinary citizens (i.e., those not in governmental
decision making positions) have equal influence, would be impossible to attain and probably very
bad. In considering the data on political equality and how one might respond to those data, it is
important to keep this in mind. In brief form, here are some reasons why that is the case.
C Achieving actual equal voice would be very difficult, involve severe governmental
and a ceiling on activity. A floor would require all to be active at some defined level --
compulsory. Voting is compulsory (usually with minor penalties for failure to vote) in
some countries. It is innocuous, perhaps, but would go against the grain in America and
not add much to political equality. Compulsory attendance and contributions smacks
C Ceilings are difficult: A ceiling on political activity exists in relation to the vote -- one
person, one vote. But ceilings on money have run up against court interpretations of the
First Amendment, and any limitation on other activities -- on writing letters, protesting,
The above refer to governmental attempts to equalize the political clout of individuals.
But what about political equality achieved -- if it could be -- without government intervention.
C Citizen incapacity: One concern might relate to the incapacity of the citizenry to make
the informed and wise judgements needed to guide policy. One consequence of unequal
activity is that the better educated tend to be more active when there is unequal activity.
If all were active, there would be government by the less well informed, the less
reasonable citizens. This is a powerful argument, that can only be mentioned here.
There is a great deal of literature on citizen knowledge and competence.3 Much shows
how ill-informed the average citizen is. But much also stresses how citizens can Aget by@
survive a citizenry that is not made up of philosopher kings -- or even princes or dukes.
But would things get worse if there were actual equality among citizens of any Aquality@?
C Citizen support for democracy: The active minority in America is more committed to
social tolerance and political freedoms than is the less active (some might argue because
the former are more active.) A non-stratified participant population might be more
C Equal political voice is inevitably equally weak voice: If each citizen has an equal voice -
- one vote, one small contribution -- it gives great power to governing officials who can
3
Lippmann (1927), Della Carpini and Teeter (1996), Popkin (1991), Sniderman (1991)
and others.
7
C Surrogates can speak for you: Equal voice isn=t necessary because others can speak for
you. It is not necessary for someone to be active to have their interests represented.
Childen, the mentally handicapped, etc. are represented by others in the political arena.
More and more interest and advocacy groups have few members but speak for
disadvantaged groups and other social causes. (Berry, 1999; Skocpol). Affluent liberals
contribute money to support policies that are not in their narrow self-interest.
The discussion above highlights one of the great democratic dilemmas: the possible
conflict between the equality of instrumental participation and the quality of citizen participation.
As I will illustrate in the data section of this paper, political participation that is stratified in
favor of the more advantaged members of society -- in particular the more educated -- is likely to
be of higher Aquality@. The participants will be better informed, can make more competent
political judgements, and they will be more tolerant of alternative political positions -- a
fundamental requirement in a democratic polity. But at the same time, they will have needs and
4
Stouffer (1955), Sullivan et al (1992).
5
Kornhauser (1959)
8
preferences that differ from the less advantaged members of society. And, thus, a participatory
system dominated by the participants of Ahigher quality@ will violate the democratic ideal of
These two words are often linked when it comes to discussions of democracy and they are
often seen as contradictory one with the other. Liberty creates the opportunity to become
unequal. In the work of my colleagues and me, we have stressed the way in which the strong
ideal of equality in the political domain (an ideal of substantive equality with each person having
equal voice just as they have an equal vote) is undercut by equality of opportunity (a close cousin
of liberty) in the economy. (Verba and Orren, 1985, Verba, Schlozman, and Brady, 1995) The
latter created inequality of resources which then are used in the political sphere. Two opposite
ways of thinking of their relationship have appeared in recent literature. In a recent book, John
Mueller (Mueller, 2000) argues that liberty (the right to participate) is, in fact, the key form of
political equality. Any other kind of political equality is superfluous and harmful to democracy.
If people are free to develop their political capacity, the right kind of equality of citizen voice
emerges -- one that (following the title of his book: Democracy and Ralph=s Pretty Good Grocery
Store) is pretty good and certainly good enough. Amartya Sen in a recent book (Sen, 2000) turns
the issue upside down. Equality, he argues, is crucial to create liberty. Unless individuals have
the right and the capability to control their own lives (and that includes a substantially equal
voice in political matters) they are not free. The difference between the two views lies, I believe,
in the distinction between the right to equal participation and the capability, in fact, to participate
effectively. Sen stresses the need for basic resources (literacy, health, enough income) so that
9
What is Participation?
There are many definitions of political participation. It can involve a variety of acts,
pursued more or less intensely and effectively, directed at a range of government officials.
Voting is the most common act and the most studied, but participation can also working for a
political candidate or cause, writing letters to officials, taking part in community activity,
protesting, etc. It can involve giving time or giving money. These acts have in common that
they are aimed at influencing the actions of government officials or their identity (that is, who
gets into an official position.) The action of officials that the participation seeks to influence may
range from influencing who gets into office (without much of an additional agenda), to the
creation of public policies, to the provision of pork and special selective favors. If
representatives are to do their job of representing the preferences and interests of their
constituents, they need information about those preferences and interests and there must be
inducements to make them carry out their representative functions. Acts of political participation
are crucial to this process. They convey information about the public (or at least the participant
part of the public) and/or convey inducements that might motivate the recipient of the message to
C Equal rights to participate. This is, of course, the main democratic requirement: universal
adult suffrage, freedom of speech for all, etc. As far as the U.S. is concerned, these are
pretty much in place since the 19th amendment and the Civil Rights Acts.
C Equal capacity: As the old saying goes, freedom of the press is a wonderful right if you
10
own a press. Various rights require resources if one is effectively to use them. Here is
where inequality rears its head in the context of equal rights. And this is what links
political equality to equality in the other major spheres of life. If an equal ability to
influence governmental outputs through one=s political activity is what we are looking for,
then having the legal right to vote or speak up or take part in a political campaign or
the skills, the resources (these days, especially the money) to do what one has to do. The
extent of political equality, this means, is deeply dependent upon equality in other
domains that foster that capacity -- on the extent of equality in education, in income, in
C Equal voice: One may have the right and the resources to be active but, for various
reasons -- from the absence of opportunities, to ignorance, to indifference --, one may
choose not to participate. So equal rights and equal capacity do not necessarily mean
equal voice.
In this paper, I will not deal with the first issue: rights to participate. Most of the data I
will present refer to the second two issues above: equal capacity and equal voice -- that is, once
rights are secured who has the capacity to use those rights and who actually uses them. The
literature on these issues is substantial. But the real result of participation requires the
consideration of three additional issues -- issues that are less well researched.
C Equal attention: Two citizens may be equal in their voice (they both vote or they both
write frequent and compelling letters to government officials or they both attend town
meetings regularly), but one may (for varied reasons) receive more attention than another.
11
C Equal output: The ultimate purpose -- or, at least, one of the main purposes -- of political
favorable policy, or a political favor. Political equality in its fullest sense would be equal
output.
C Equal outcomes: Policies (outputs) do not always accomplish what they were intended to
do. Full equality of treatment might be a set of policies, the results of which treated all
citizens equally.
In some sense, these last three aspects of political equality represent its Atrue@ meaning
and are the ultimate payoff from equal participation and equal voice. They are difficult to
observe and measure. Indeed, given differences in needs and preferences among citizens, the last
two are probably impossible to achieve. Most of the data I will present refer to the first three
Let me pursue the distinction between conveying information and motivating a response
through communicating inducements a bit further. How would one recognize and measure
political inequality in sending the message and in receiving a response.? The firsdt task is crucial
to the second. It is not an easy one. The second task is much harder. Most of my own work has
focussed politics from the bottom up: on the ordinary citizens to see what they do, how what
they do differs across individuals and groups, and why it is that there are these differences. But
from a broader perspective, one ought to look at participation as an interaction between the
senders of messages and the recipients. For both the activist senders of messages and the
government officials or political candidates who are the targets of these messages, we can look at
12
them as people with preferences and motivations -- people who want by their actions to achieve
some state of the world. Citizens may seek the enactment of some policy that benefits them or
perhaps just fits into a general ideology or philosophy (our data show that many citizens are
active for that purpose despite what the literature says they should do if they are rational), or they
may seek some particular selective benefit, or they may act with no particular response in mind
(to please a friend or relative or fulfill a civic obligation.) Similarly, the recipients of the
message may act out of varied motivations: they may be motivated to respond to the preferences
expressed by the message senders (because they think this the right thing to do); or to respond
not to what the message senders want but what will be better for them (perhaps as normatively
committed trustees); or the officials and candidates may be motivated by their selective self-
interest (to obtain the perks of office or perhaps the adulation of the people so they will respond
to those who can offer these selective benefits); or they might be motivated by that middle and
complex motivation that we political scientists focus on, the desire to achieve office and remain
there (which they might calculate would be served by following the expressed preferences of the
activists or perhaps by meeting their needs even if they are not aware of them -- under the
assumption that the citizens will reward beneficial outcomes rather than seemingly responsive
policies). In sum, citizen participants and governing officials can be understood as people with
This indicates, I think, why it is difficult to develop a measure of the degree to which any
two citizens are equal or unequal in their political voice, or any two groups of citizens (rich or
poor, white or black, women or men) are, on average equal in their political voice.
Even if we limit ourselves to the easier part of the picture -- as I and my colleagues have done in
13
much of our work on participation ( that is to the input deriving from what citizens do)- it is hard
to compare the voice of one citizen to another because they can express it in so many ways.
Perhaps the vote is the most comparable act because it can be cast in a very limited way -- it does
not matter with what vigor or skill you pull the lever.6 But for other acts, there are many
variations -- and variations that may be systematic across groups. Political work in a campaign
or a contribution can be limited in amount or large in amount. A letter can be more or less
convincing. And given the bundle of things some one can do, there is no easy metric to compare
the significance or political value of a letter with a contribution; of a vote with attendance at a
political meeting.
If we move from the task of comparing what individuals do to what effect it has, the task
of comparing across individuals and groups is more complex yet. The vote ceases to be so
equal. Voter skill becomes important: has he or she made the most appropriate choice given the
motivation behind the vote. I suppose it does not matter much if we are talking of the individual
voter since the individual=s vote is not going to make much of a difference. But it might matter
for groups of voters whose aggregate votes might swing an election. In addition, the structure of
the electoral system affects the equality of the vote. Are votes counted honestly? Are district
6
This was obviously written before the 2000 election crisis over punch cards, dimpled
chads and hanging chads. The general point still holds.
14
sizes equal? What form of PR is used with what consequence? Political science has an entire
industry working on differing electoral rules and how they affect the value of the individual=s
vote. What about permanent minorities: groups whose votes never lead to the election of
representatives because they are too small in number to be successful under the particular
electoral rules? Whether that is a serious violation of the equality of participation might depend
on another contextual characteristic -- the seriousness of the basis for their minority status. A
racial or economic or religious minority in a society where many of the main issues revolve
around that fault line is deeply deprived of political voice if their votes lead to no success -- even
if they have the equal right to cast it and do so at the same rate as other groups.
And, of course, the situation is more complex when one goes beyond the vote. Letters,
contributions, protest marches are all part of a participatory repertory. But their effectiveness
also depends on many contingencies: is the target of the act a target that will pay attention? Is it
a target that can act on the issue? Will it act on the issue? I write a letter to Ralph Nader
commenting on one of his positions? Will he get it and read it? It may not reach him since he is
traveling so much and is bombarded with communications? And if he reads it, will it move him?
And even if it moves him, so what? He is not going to be elected anyway. I may still get the
gratification of having sent a message to someone I admire in relation to a cause or issue I hold
important. But if we are considering the impact on the policy I was concerned with, the act is
less effective.
One more problem: what is a responsive output? For an election, it would seem to be
having your party win the election. But it might not be that simple a dichotomy. The winning
candidate may not deliver because she is in there for the selective benefits of office. Or she may
15
deliver what she thinks will be good for you in the long run, not what you want. She may deliver
something designed for those who voted against her -- and that for varied reasons. And if the
outcome is not an election result but a policy, the calculation becomes harder still. Policies are
complex. The citizen input may shape a policy at the margin or in relation to a particular feature.
Even if the overall shape of the output seems non-responsive there may be some important
The problem of comparing across political acts can be illustrated by considering the vote
and a letter written to some governing official. The individual vote is a very weak and a very
blunt instrument. My vote has no chance of affecting the election and my vote carries little
information about what I want to get out of it -- I vote for George Bush but my vote does not say
why. The mandate is always uncertain. Yet the aggregate of the votes determines whether Bush
or Gore becomes president. Add them up and it counts. A letter is different. It can give
(reasoned or unreasoned) arguments for a position. If read it can move a person. But the odds
are slim that it will be read carefully given the volume of mail into the White House or into a
Congressional office. Yet, a reasoned letter can have an effect. (Jacobs and Shapiro, 2000). If
you are lucky and your letter is pulled for notice, you might have an effect -- even though letters
in the aggregate do not have the final authority that the aggregate of votes does.
So which group has more voice: a group with high voting turnout but not much letter
writing or a group with prolific letter writers who do not go to the polls so much. (The two
forms of activity are correlated, so the example is unrealistic.) As for the vote: each vote is
trivial and conveys little information but each vote is counted and the sum of the votes is
dispositive. The letters may contained compelling arguments but there is no law that says they
16
must be read, let alone followed. And the odds that a particular letter will be read may be very
small. Yet if there is a letter writing group -- it sends a lot of thoughtful letters -- its clout may be
great.
The difficulty of measuring the extent of political inequality can be seen if we compare it
to another, more fully discussed and studied, aspect of inequality: income inequality. To begin
with, of course, the metric is simpler. We can more easily compare incomes across individuals
than we can political. A dollar is a dollar, but even a vote is not a vote -- and who knows how a
vote compares to other acts. The value of income is, of course, not measurable simply in terms
of the dollar amount. Individuals differ in how skilled they are in using their dollars. Some are
better located to be effective consumers -- they have access to good stores with decent prices, etc.
But the variation in the effectiveness of a dollar to provide utility is small compared withe
The biggest difference between income and political equality is that income is an
individually held resource than is used largely to purchase selective, individual goods. You pay
your money in a store and you take out the bundle of goods. Political activity is, at least in good
part, aimed at collective outcomes. Thus, whatever you have in political voice, it does not mean
you will get political benefits. I have spelled out a lot of that above.
That one is using the resource of political activity -- the allocation of time or money to it -- to
achieve a collective goal, means that a lot depends on what others do. You spend your money
wisely, and you get what you buy. Political activity faces much greater coordination problems.
You spend your politically resources wisely -- you contribute with careful calculations combining
your money with messages as to what you want, you communicate in a compelling manner with
17
officials and they hear what you say -- but much depends on others including the opposition. If
you increase your contributions, but the other side increases theirs= more, you have not gained but
lost. This leads to an important distinction between income and political inequality. The latter is
much more a matter of relative standing. True, if envy rules the world, then income inequality is
relative too -- my income my go up but it will not make me happier if yours goes up too.
Nevertheless, both of us can get richer and happier. But in political life, it is more of a
One reaction I sometimes get from cynical non-social scientists and from social scientists
of various stripes when I tell them that I study political participation is: why bother? It doesn=t
make any difference. It=s all decided on the inside by the politicians, the interest groups, and the
bureaucracy. They agree that it is Aall decided around the green table@ as my landlady in Vienna
I don=t believe it. I don=t believe it, in part. because I don=t want to. Those of us who
study various aspects of politics develop unprovable commitments to one aspect of politics or
another. And I believe that all the calculation of how to appeal to the public cannot be mere
show. Nor is all the activity by citizens mere futile gesturing -- and there is a great deal of all
kinds of activity at all levels even in an era when Americans are spending all their time by
But we don=t have any clear estimate. The problem is that everything is endogenous.
Preferences surely are. And this involves a preference for political activity as well as preferences
18
for outputs that such activity might foster. Citizens send messages about their preferences to the
government, but their preferences are in part shaped by elites, perhaps mediated by the media.
Are they sending merely echoes of what the leaders want to hear? Some, like Ben Ginsberg,
manipulate citizens. John Zaller also stresses elite shaping of opinion (Ginsberg (1986), Zaller
(1993)). Zaller, however, found a good deal of public autonomy when it came to Monica
Lewinsky (Zaller, 1999); apparently there are some subjects about which the public knows what
it thinks.
Jacobs and Shapiro, in their new book (2000) on political pandering (or rather its
absence) give an interesting mixed picture. Politicians pay attention to the public -- they read
polls, they read the mail, they read the various tea leaf indications of public opinion. But they do
not do it in order to follow the public. Rather they do it in order to figure out how -- through
what the authors call crafted talk -- to bring the public around to favor the policies they want to
pursue. It=s something like elite manipulation, but not completely -- or at least not as cynically as
some might have described it. And, it is a form of citizen control. They need to convince the
citizenry -- and that is certainly suggests a flow of influence from the bottom up. The public is
not ignored. And they have to craft communications that appeal. Nor does it violate our notion
The issue here is paralleled by a complex issue when it comes to political mobilization.
One of the main sources of citizen activity is mobilization. Rosenstone and Hansen assign a
good deal of the cause of the decline in activity to a decline in party mobilization (1993). In
Voice and Equality and subsequent work, Brady, Schlozman and Verba (1995, 1999) show how
19
mobilization shapes the amount as well as the distribution of political activity. Input from below
is mobilized by stimuli from above. Is that democratic responsiveness or not? When Mayor
Daley (the real, earlier, one) brought out the vote on the South and West sides of Chicago, was he
mobilizing and manipulating them for his purposes, or was he making them into real actors in the
political process? Is mobilization more like Albania in the old days when it led the world in the
proportion who voted, or more like the civil rights movement where blacks were brought into
politics through mobilization, but to pursue goals to which they were committed. Probably it
is some of each. The two parties are now trying to bring out the vote -- in order to elect
themselves and reap the selective benefits of office. That=s part of it. But they are doing do to
pursue an agenda that they believe in the interest of the people they are mobilizing. They need
healthy democracies in which there is real control by citizens over the government -- as there is
in the U.S. though it is hard to say how much -- are better societies. They do not commit
massive crimes against their own citizens -- or are at least less likely to and they will be less
massive. Amartya Sen has made the major and very striking point that famines do not take place
in democracies -- since famines come less from a sheer absence of food than from an inadequate
distributional mechanism. (Sen, 2000). And democracies are under some pressure to provide
such mechanisms. Political participation -- open and free -- makes the state of the society clear;
that=s one step to the government being responsive. And it makes clear that there are negative
consequences for the government if it does not respond. That=s the second step to
20
responsiveness.
We cannot get precise estimates, but it seems fairly clear that the citizens who are worse
off are those who are not active and whom nobody wants to try to mobilize. If a group gets into
the system -- by this I mean that they come to be active autonomously, or more likely some
political entrepreneurs (in the good sense of the word) decide to try to arouse them, or more
likely still both happen in an interactive fashion -- it will begin to see its needs taken into
account. The biggest revolution in recent American political history is the democratization of
the American South since the sixties. Once the legal barriers were broken, African-Americans
became potential participants -- combining their own motivation to be active as voters and in
other ways with the mobilization activities of black leaders who wanted to see the system change
and with candidates for office, black and white, who wanted to get elected. And who can deny
that their participation changed their place in society and changed society. Or consider the
elderly and social security. California will be different now that Latinos are moving from a
The gradient of income or education in relation to political activity is, for a variety of
reasons, steeper in the United States than in almost any other industrialized democracy. (Verba,
Nie and Kim, 1979) It is likely no accident that the United States ranks at the top or near the top
among such nations in the degree of income inequality. Most directly related to the political
effects on income inequality is that nations are more similar in the degree of inequality of income
before government intervention and quite different after governmental taxes and transfers are
taken into account. Kenworthy (in Solow, 2000) shows that in the U.S., what the government
does -- its tax policies, its social service policies -- produces a lot less equalization than in other
21
comparable countries. Is this connected to the greater political clout of those lower on the socio-
economic scale in these other countries? It is hard to get a definitive answer -- but it is not
unlikely that that is the case. In an interesting recent paper, Kelly and Stimson tentatively --
tentative because they are just beginning work on this -- show a connection across the usually
unconnected phenomena of input, output, and outcome. (Kelly and Stimson, 2000). They show
that public mood is connected to government welfare and spending policies which in turn affect
income inequality.
Also, there are social judgements we make on the basis of less that precise criteria.
Despite what I have said about the greater ease of measuring income inequality, the overall
question of whether there is too much inequality in the income in the United States is not
general evaluation of the kind of society one wants. Somewhere between the impossible and
undesirable goal of complete equality of income and a complete caste like bifurcated system into
the haves and the have nots -- where he haves have it all and the have-nots have nothing -- are
many distributions. Thomas Scanlon in his recent book, What We Owe to Each Other, (1998) a
book with a Rawlsian scope and richness, says that we ought to reject discrepancies across
groups that a reasonable person would reject. This is vague, but useful. A lot depends on who is
reasonable. I, of course, tend to find some people more reasonable than others -- they are the
ones who share my social values and philosophy. The reasonableness definition is difficult -- but
useful.
In the end, a concern with political equality is a moral concern. Even if we go about it as
we should as social scientists; even if we understand the difficulties of studying or measuring it;
22
even if we are aware that political equality is an unattainable goal; even if we understand that it is
not a fully desirable goal, indeed, that a system in which all had equal voice would be
unworkable and perhaps dangerous to democracy; we recognize that it is a goal worth striving
for. Gross inequalities of political influence across social groups -- the existence of group with
little voice and little hope of gaining it -- is dangerous to our democracy. It violates that sense we
have that each citizen is a worthy participant, that he or she is -- to go to the deepest
the good; and is, therefore, a person who should have equal voice in American democracy.
As the previous discussion makes clear, there is no simple answer to questions about the
state of political equality in the United States. There are different kinds of inequality. In the
following sections, I will present data on various aspects of political equality. I will look at
several issues:
C Inequality among whom: I will focus mostly on differences across groups distinguished
by social class type characteristics (largely income and education) but will present a bit of
data on differences by race/ethnicity and gender. Much more could be said about other
bases of inequality such as age, or region, or political ideology, but I=ll focus on the class
items.
The analysis will look at the following themes in relation to the distinctions above:
In comparative perspective.
In politics itself.
It is clear that there is much to cover, I will discuss some of the data in a rather cursory
fashion.
As has been made clear, one cannot discuss the amount of political inequality without
first answering: Ainequality in what?@ and Aamong whom?@ Public discussions of political
inequality often -- and quite appropriately -- focus on differences in voting turnout between rich
and poor or whites and minorities or other groups. And there is, of course, much discussion of
the role of money in American politics where the mode of activity is contributions and the group
difference is between rich and poor. Each of these discussions highlights an important fact about
American politics: despite the fact that all are equally free to participate (there are no significant
limitations on universal suffrage and all citizens are free to contribute to causes and campaigns
with any legal limitations fostering equality rather than inequality), there are sharp differences in
24
the activity rates across income groups or educational groups. And this runs across many
Figures 1 and 2 report data aggregated over a number of years on the relationship between
income and various political activities, and education and the same activities.7 The story is
straightforward and clear. The advantaged are more active than the less advantaged.
To elaborate:
additive scale of political activity. The scale is a summary measure of the number of political
activities in which the individual has engaged. It is an additive scale with each of the following
counted as one act: voting in the 1988 election, working in a campaign, contributing money to a
campaign, contacting an official, taking part in a protest, working informally with others in the
at the meetings of such a board, and membership in an organization that takes political stands.
The increase in political activity with income is clear. Those in the lowest level of income
average one act (usually voting); those at the top of the income scale average more than three
acts.
There are numerous and significant differences across political acts. Because the
7
Data sources are listed on the Figures and Tables.
25
summary measure obcures differences among acts, we can decompose it into its components.
Figure 4 compares two income groups at the extremes--those with family incomes below
$15,000, that is, below or very close to the poverty line, and those with family incomes $75,000
and above--with respect to the proportion who take part in various political activities. For each
kind of participation, affluence and activity go together. Of the various acts, voting is perhaps
the most egalitarian. Turnout is much higher among the wealthy than the poor, but voting is the
only act for which the affluent are not at least twice as likely to be active. Those with the lowest
family incomes are less well represented among those who take a more active role in political
campaigns. The poor are one quarter as likely as the affluent to do campaign work and about one
There is also a participation gap among income groups with respect to contacting, a mode
of activity that might be of special relevance to the disadvantaged who depend upon government
programs; informal community activity; and, especially, serving on a local governing board.
Moreover, although the disparity is somewhat smaller than for other modes of activity, the poor
are less likely to attend protests, a form of activity that is often described as the "weapon of the
weak" because it is available to those with few economic resources. In short, the affluent are
overrepresented among activists to an extent that varies from act to act: this overrepresentation
is least pronounced when it comes to voting and, not unexpectedly, most pronounced when it
Time and Money, and the Volume of Activity: Voting is a unique political act in that
individuals are limited to one and only one act -- one person, one vote. For other activities,
people who are active can be more or less active. Thus, one can give a little time to a political
campaign or a lot; and the same with money. We can learn more about the representation of
income groups through various forms of political activity by considering the volume of activity
they produce. Figure 5 gives us a politician's eye view of what the citizenry would look like if
each income group's visibility depended on the amount of its political activity. The upper-left
section of Figure 5 presents as a baseline the distribution of various family income groupings
within the population. The other graphs show the proportion of the population that falls in
various income categories weighted by the amount of activity produced by that income group: by
the votes cast; the number of hours worked in campaigns; the number of dollars contributed to
candidates, parties and campaign organizations; the number of contacts produced; and the
number of protests attended. Thus, the upper-right graph shows the proportion of votes coming
from the various income groups in the 1988 presidential election. A citizen who voted is
weighted as one; a non-voter is weighted as zero and does not appear on the figure. The other
parts of Figure 5 show the proportion of campaign hours or campaign dollars coming from each
Consider first the electoral arena. Those at the top of the income hierarchy produce more
than their proportionate share of votes, campaign hours and campaign dollars. However, the
distortion is much less pronounced for votes than for campaign time and, in turn, much less for
campaign time than for campaign money. The 3 percent of the sample with family incomes over
$125,000 are responsible for 4 percent of the votes, 8 percent of the hours devoted to
campaigning, and fully 35 percent of the money contributed. Indeed, when it comes to campaign
dollars, the top two income groups, which together account for less than 10 percent of the
population, donate more than half of the money. At the other end of the family income scale are
those with family incomes under $15,000, who form 19 percent of the sample. They are
campaigners. When it comes to making electoral contributions, however, they are barely visible-
Contacting and protesting are of particular concern both because they are modes of
activity that permit the transmission of relatively specific messages and because they are
especially important for the less well off--contacting because contacts permit the addressing of
specific individual problems and protesting because protests require little in the way of resources.
Reflecting patterns presented above, the bottom section of Figure 5 shows that the affluent
produce more than their proportionate share, and the poor less than their share, of contacts and
protests. In terms of the volume of activity, the poor are not as underrepresented as they are with
respect to campaign contributions, but their share of the contacts and protests is lower than their
We have focused on a comparison of the affluent and the poor. It is important to note,
33
however, that the difference among the modes of activity affects the representation of the middle
class as well. Consider those in the $35-50,000 income bracket. They are somewhat
comes to giving money is even stronger for the group that is one step further down on the
economic ladder (those with family incomes between $15,000 and $35,000) who might be
thought of as at the border of the poor and the middle class. The special inequality associated
with monetary contributions affects the poor most strongly, but it also means that most of the
In one sense, these data merely underline the obvious: that the rich have more money.
Nonetheless, the distinction between votes, time, and money matters for politics. As political
campaigns, and other aspects of political life, have become increasingly professionalized and
dependent upon financial contributions from citizens, the stratification of political life has,
presumably, become more severe. As checkbook participation has become the more common
form of participation, the stratification of American politics has probably been exacerbated as
well.
Measuring participatory inequality is more difficult still because of the multiple ways in which
people can be active, the lack of common metrics, and the absence of comparable data sets. As is
well know, voting turnout in the United States is lower than it is elsewhere, though the U.S. does
not lag behind-- indeed, tends to be ahead -- in many other forms of political activity. When it
34
comes to the equality of political activity, the best evidence is that there is more inequality in the
U.S. than in other comparable democracies. (Verba, Nie and Kim, 1979) Table 1 shows the
correlation between three political scales (a scale of voting, a scale of campaign activity, and a
scale of political interest) and a socio-economic scale for Austria, the Netherlands, and the U.S.
For voting turnout as well as campaign activity, the relationship of activity to SES is stronger in
the U.S. than elsewhere. Political interest offers an intriguing contrast. The relationship to
SES is stronger and closer across all three countries. This may offer a clue to the difference in
the SES-activity relationship among the nations. Political activity, we assume, is more
dependent on the institutional context in which it takes place than is political interest. Political
parties can mobilize individuals to political activity in ways that exceed the extent to which they
want to or can get them interested. Conversely, they can Alock-out@ individuals and groups who
find no party for which to vote or work. But they cannot as easily influence political interest.
Thus, the data suggest that political interest grows with socio-economic status (in particular with
education), but political activity -- particularly in the electoral arena -- may depend more upon
institutional structures.
35
36
Time and Money: There has been much recent discussion of the decline of social activity
in the United States (Putnam, 2000). Much of the focus has been on the non-political domain --
on organized and unorganized social activities. There is evidence that political activity has
declined as well. The decline in voting has been dramatic since about 1960, as has the decline in
active participation in political campaigns. One area in which there has not been a clear
Americans seem to be participating more by giving money than by giving time. (The data are
not unambiguous, but there does seem to be a clearer decline in campaign work than in campaign
contributions -- Figure 6.) Money plays a bigger role in politics, relative to time, for two reasons:
supply and demand. The supply of time has diminished as more families have become two
career families, but the supply of money has not diminished as much. And money is more in
demand by those running political campaigns and promoting political causes. They no longer
want volunteers to stuff envelopes as much as they want money to purchase equipment for mass
professionals.
37
38
inequality. Money is much more stratified than time. The rich have more money than the poor,
as Hemingway noted. But so do whites compared with minorities, men compared with women,
people in high status jobs compared with those in lower status jobs. This obvious point has
important implications. Time, as a resource, is not as stratified. The main characteristics that
determine how much time a person has are not the usual variables associated with stratification
(income, education, race, etc.) but are life circumstances characteristics (whether one is working,
whether one has children at home etc). (Verba, Schlozman, and Brady, 1995, Chap. 10). Thus,
as time is replaced by money as the main mode of political participation, the participatory
system becomes skewed in the direction of those at the top of the stratification hierarchy who
have different preferences and interests. Politics does not involve many issues between the
harried and the leisured, it involves many issues between the affluent and the poor.
There is an interesting difference between time and money as an input into the political
process. The affluent are more likely to make a political contribution. In addition, the higher
their income, the more they give. Interestingly that is not the case for time. The affluent are
somewhat more likely to give time to political campaigns than the less well off, but if they decide
to give time, rich and poor are not different in how much they give. This is seen on Figure 7
which presents the average number of hours given by campaign workers and the average number
did campaign work, the amount of time given is relatively uniform across the income groups,
with the only deviation being the higher number of hours given by those in the lowest income
group. The data for dollar contributions could hardly provide a sharper contrast. For campaign
39
contributions, the amount given increases with family income. The increase in the amount
donated is relatively gradual through the middle incomes and quite steep in the higher-income
categories with the result that, among contributors, the average donation from an affluent
The data suggest that time and money function differently as inputs for political activity.
As we have seen, for all forms of activity, the affluent are more likely than the poor to become
active. Once active, however, the poor are as generous with their time as those who are better off
financially. Their lack of financial resources does not appear to act as an impediment to the
investment of sweat equity. Since nobody's day contains more than twenty-four hours, the well-
heeled can give only so much time--apparently not more, on average, than the poor. No such
leveling occurs when it comes to money. Not only are those with higher family incomes more
likely to make donations, but they give more when they do. If time is less important in politics,
Overall Activity Measured Over Time: In recent work, my colleagues and I have tried to
trace changes in the equality of overall political activity. (Brady, Schlozman, Verba and Elms,
forthcoming, 2001) Using the same Roper data used by Robert Putnam, but using it to trace the
trend in political activity, we considered whether the difference between the top and bottom
quintiles in socio-economic status (essentially measured by income and education) in their rate of
political activity has changed over time. Figure 8A shows the scores of the top and bottom
quintiles on a composite scale of political activity -- the sum of twelve political acts about which
the Roper organization asked -- as they have changed over time. The data are not as good as one
would like but the best available. They clearly reflect the general decline in political activity. On
the other hand, they show no clear increase or decrease in political inequality. (This is seen more
clearly in the ratios of the top quintile to the bottom reported on Figure 8B) The data are not
necessarily inconsistent with the discussion above on money and politics, since political
contributions represents but one of the twelve measures of inequality. (As an interesting aside: in
the massive amount of attention given to the decline of civic engagement in America, almost no
It is difficult in this overview of political inequality to give a full account of the literature
on the sources of political inequality. A brief account might highlight three books by my
collaborators and me over the past quarter century. (There is, of course, a larger literature than
this -- and I will cite some of it -- but the main research in the field has, I believe, revolved
around these works.) In 1972, Verba and Nie (Participation in America) introduced the ASES
model,@ in which income and education were shown to be the principal predictors of activity.
42
This analysis clarified the nature and extent of stratification in American political participation.
However, this model was as much an accounting model as an explanatory model, demonstrating
the strong association between socio-economic status and political activity without explaining
The Civic Voluntarism Model, developed in the early 1990's by Verba, Schlozman, and
Brady (Voice and Equality), filled the explanatory gap by specifying the causal links between a
number of participatory factors and political activity. These participatory factors are motivation
(interest in politics, a sense of political efficacy, a belief in the duty of a citizen to be active,
partisan commitments, as well as specific preferences for governmental outputs), resources (time,
money, and civic skills -- the latter being those communications and organizational skills that
which one can be recruited to political activity). By demonstrating how these participatory
factors, in turn, derive from socio-economic status -- from one=s educational background, the
kind of job one has, the number and kind of organizations one belongs to --, they were able to
explain why socio-economic status and participation are connected. In particular, they explained
the primacy of educational attainment by showing that education has not only a direct impact on
political activity but more important indirect effects by enhancing nearly every other participatory
factor: the well educated are more affluent, more likely to exercise civic skills and to receive
requests for political activity, and more politically interested and knowledgeable. Table 2
presents an analysis that summarizes the effects of these various factors on political activity.
43
44
In a The Private Roots of Public Action (forthcoming 2001) Burns, Schlozman and Verba
take our understanding one step further by specifying the processes by which participatory factors
institutions and the treatment within institutions and by demonstrating the connections between
institutions, they spell out more fully how institutional processes work, thus deepening our
understanding of the links between non-political and political life. In particular, this approach
allows us to specify further the single most important source of participatory inequality: the
cumulative effect of education. Those who are born to parents of limited educational attainment
begin life with a participatory disadvantage that is difficult to overcome. The effect of parents=
education is felt on the child=s education. And educational differences that exist upon entering
adulthood will be magnified thereafter by the institutions of everyday life. The non-political
institutions of adult life grant a significant advantage to the well-educated with respect both to
institutional access and to the acquisition of participatory factors within institutions. Educational
attainment plays a particularly powerful role with respect to the workplace, a domain that is a
rich lode of participatory factors. For both men and women, education stratifies who goes to
work, what kinds of jobs they get, and whether they develop civic skills and receive requests for
political activity on the job. (A significant exception to the pattern of association between
The main theme of this analysis is that participatory inequality is deeply embedded in the
basic institutions of American life -- which provide citizens with the resources and motivation
and connections that foster activity. It shows the close link between the non-political and the
political. And it shows how the advantage of the better-off in political activity is cumulative. A
good education fosters political resources such as skills as well as motivation such as political
interest and a sense of civic duty; it places one in networks of people who stimulate your activity;
it channels one into occupations where skills and motivation are reinforced, and into networks
that reinforce the tendency towards activity, and on and on. Thus, the non-political institutions of
Politics and political inequality: The roots of political participation, in this analysis, are
found outside of political life. But political life itself plays a role. There are no longer legal
the fact that so much of political life is dominated by white males may send a message to
potential participants that the game of politics is not for them. There is some evidence that this
is the case. African-Americans are more politically efficacious and more active in cities in
which there are elected black officials, although there is some issue as to which comes first
(Bobo and Gilliam, 1990); African-Americans are more likely to contact political officials if the
officials are black (Verba and Nie, 1972, Gay, 1997), and women are more likely to be interested
in and informed about politics when they are in states in which women hold statewide or national
office. (Burns, Schlozman, and Verba, forthcoming, 2001). (Figure 9 and Table 3.)
Does It Matter?
political participation makes a difference. It depends on what the political activity is all about.
If we think of the outcome of such activity as a set of selective benefits (a Christmas turkey or a
good city job as with the old style political machine, or a government contract or favorable tax or
tariff decision as with newer politics, or various varieties of pork for the local district and its
leaders and residents), then it does matter who is active. If we think of the outcome as a more
collective good -- a general policy -- then the connection is less clear, and this for several
reasons. For one thing, it would not make a significant difference if one group were more active
than another if there were little or no differences across the groups in their preferences or needs.
Participatory inequalities are of significance only if they are among Apolitically relevant groups@ -
- groups for whom there would be political outcomes of relevance to them. Note, however, that
these politically relevant differences can take various forms: differences in policy preferences in
relation to issues on the political agenda, differences in what issues are likely to be put on the
agenda by one group or another, and differences in the objective needs of a group. Let=s look
more closely.
48
49
50
Wolfinger and Rosenstone in their study of voting turnout make an important point that
has become part of political science common wisdom: although voters differ from non-voters in
their demographic composition (their level of education, their income, etc.) they differ little in
their policy preferences as measured by their responses to a series of survey questions measuring
attitudes government welfare and other policies. In our research (Verba, Schlozman, and Brady,
1995), we replicated the Wolfinger and Rosenstone analysis and confirmed that there was little
attitudinal difference between the voters and the non-voters. (See Table 4). (See also, DeNardo,
1980 and Shaffer, 1982) Others have found differences but they are usually not large. (Bennett
and Resnick, 1990). In addition, various analysts have shown that the outcome of an election
would rarely be changed if one eliminated turnout differentials among groups; if, for instance,
everyone voted. (Teixeira, 1992) (Petrocik, however, finds that it might have made a difference
in 1980 and 1984, Petrocik, 1987) This is a corrective to a common view that an increase in
voting turnout is coterminous with an increase in the proportion of the voting population from
those who are less well off. This may not always be a valid assumption. (See Hill and Leighley,
1992) In sum, those who believe that an equalization of voting turnout among those who
currently are less likely to vote would produce a major change in the electoral mandate (see, for
instance, Piven and Cloward, 1988) might well be disappointed by the results of massive vote
mobilization.
51
However, since the existence of politically relevant differences between the active and
inactive citizens is at the heart of the issue of the significance of participatory inequality, it is
useful to look a this more closely -- using the range of information available in the
CitizenParticipation Study.
The similarity in many policy positions of voters and non-voters is an important point, but
its significance can be overstated. The absence of difference refers to a comparison of voters and
non-voters in their responses to standard National Election Studies questions on a set of political
52
issues. If one expands the scope of inquiry, one gets different results. First, we must take a
broader view of activity, moving beyond electoral turnout to include various kinds of political
activity that can convey more precise messages to policymakers and can be multiplied beyond the
enforced equality of the single vote. In addition, we must take a broader view of politically
relevant attributes, encompassing not only demographics and policy positions as expressed in
response to survey questions but also policy-relevant circumstances and the actual content of
participatory input.
characteristics and their policy preferences--the same pair of attributes Wolfinger and Rosenstone
use to compare voters and non-voters -- with respect to their overall level of political activity as
measured by the additive scale of political acts. Figure 10, which considers several demographic
questions, shows the overall activity rates of demographic and attitudinal groups. Even when we
enlarge our understanding of political activity, the distinction that Wolfinger and Rosenstone
make between demography and attitudes holds up. Disparities in activity across socio-economic
across attitudinal groups. Those with high family incomes score much higher in overall
participation than those with incomes at the poverty line, and professional and managerial
workers score higher than unskilled and service workers. The gap in overall participation
between college graduates and those who never finished high school is especially wide. In
associated with less variation in political activity. There is, however, a small but consistent
53
tendency for those who are more conservative in attitude or Republican in party identification to
be more active. Although the differences are not great, they are more substantial than those
The contrast between the similarity in the amount of activity across groups defined in
terms of attitudes and ideological positions and the difference in activity among groups defined
terms is not matched by distortion in substantive representation. The activists who carry the
voice of the people into the political process may not look like the rest of the public, but they do
not differ much in what they want. However, policy preferences are not the only politically
relevant characteristics. Figure 11 presents data on the activity of respondents who have varying
their families living with them. The Citizen Participation Survey asked whether, in order to
make ends meet, the respondent or any immediate family member living in the household had to
"put off medical or dental treatment," "cut back on the amount or quality of food," or "delay
paying the rent or making house payments." It asked as well about two other strategies for
making ends meet: did anybody "cut back on spending on entertainment or recreation" or "work
extra hours or take an extra job?" In addition, we asked whether the respondent or a member of
the immediate family living in the household received means-tested government benefits (food
veterans' benefits, Medicare, or educational loans). We were, thus, able to identify those who
54
have real financial needs or depend upon government programs. These measures are, of course,
closely related to income, but they give additional indications of respondents' potential interests
As shown in Figure 11, there are substantial differences in rates of political activity across
groups distinguished by financial need or by the receipt of government benefits. Consider the
activity of those who report some financial pinch. They are slightly less active than the
population as a whole, with the divergence from the average for all citizens increasing with the
severity of the financial squeeze. Those who report the relatively mild--and quite widespread--
need to cut back on recreation do not differ from the population as a whole very much. Those
who had to cut back on spending for food or who delayed paying the rent, however, are
substantially less active. Clearly, those with real needs are less visible in the participatory
system.
This pattern is even more pronounced if we consider those who reported that they or a
member of the immediate family receive one of a number of government benefits. The receipt of
benefits per se does not imply a low level of activity. Those who receive non-means-tested
benefits such as student loans or veterans' benefits, Medicare, or Social Security are at least as
active as the public as a whole. In contrast, those who receive means-tested benefits such as
AFDC, Medicaid, food stamps, or subsidized housing are substantially less active than the public
as a whole. The differences imply that those who would be in most need of government
response--because they are dependent on government programs--are the least likely to make
It is useful to decompose the data in Figures 10 and 11, in which activity is measured by a
summary of several participatory acts, in order to consider the kinds of participation separately.
As we have seen, voters do not differ much from non-voters when it comes to policy preferences.
But what about in terms of their actual needs or their actual dependence on government
programs? And, do those who do more than just vote--those who engage in participatory acts
that may involve a greater volume of activity or convey more information--differ from the
population as a whole? Figure 12 reports for various kinds of activity the Logged Representation
attitudes on economic issues, as well as by their objective economic circumstances. The scale is
defined in Verba, Sclozman and Brady, Chapter 4.) The degree of over- or underrepresentation
is measured for several different activist groups: voters, campaign workers, campaign
varying both the characteristics of activists and the activities through which they can be over- or
underrepresented, Figure 12, thus, tells us something about both parts of our puzzle: what
characteristics of citizens are better represented through activity and what activities better
represent citizen characteristics. It also tells us something about the extent of under- and
overrepresentation. For instance, the top section of the figure tells us that, among voters, liberals
are very slightly underrepresented compared to their proportion in the population as a whole;
those who said they had to cut necessities are a little more underrepresented; and those who
Across all types of activities, those who take part represent more accurately the attitudes
of individuals as measured by the standard survey questions on economic policy than they
59
represent individuals' actual needs, as measured by the need to cut necessities, by income, or by
proportionately among activists across the various activities. When we consider the
representation index for differences based on the actual needs of citizens and, even more so, for
differences based on their receipt of means-tested benefits, we find that the disparities are much
greater.
There are also significant variations in terms of types of activity. Considering activities
associated with elections, the voting population and the population of campaign workers are, as
mentioned, fairly representative of the population at large with respect to attitudes, and only
somewhat unrepresentative of those who have had to cut back on necessities. However, the poor
and those who receive means-tested benefits are substantially underrepresented among voters
and, especially, campaign workers. It is, however, among campaign contributors that the
underrepresentation of the poor and needy is most pronounced. This is hardly surprising. One
would not ordinarily expect those who have severe economic problems to be campaign
contributors. But that does not change the substantive implication of that fact for the messages
that are received through participatory channels--in this case, one of the most effective and
salient channels. Compared with campaign contributors, the underrepresentation of those who
have had to cut back and the poor in other activist populations is more moderate. The
differences, though, are still substantial compared with the representation of policy attitudes.
60
61
For three modes of activity that might be particularly relevant for needy citizens--getting
in touch with public officials, attending protests and demonstrations, and being active in the
community--those who felt financially strapped are fairly accurately represented in the activist
population. However, for those who received means-tested benefits and for those in poverty,
there is fairly severe underrepresentation. Contacting is, presumably, especially important for
citizens who receive government benefits, since ensuring the flow of benefits may entail the need
underrepresented among the contactors. Although protest should be particularly important for
disadvantaged groups that lack financial resources or connections, those receiving a means-
tested benefit are underrepresented in that participant population as well. And the recipients of
means-tested benefits are about half as likely to be community activists as their proportion in the
Sending a Message
Participatory acts differ not only in the extent to which their volume can be multiplied but
also in the extent to which they can convey to policymakers detailed information about citizen
concerns. The Citizen Participation Study data allow us to investigate in several ways this
largely uncharted, but crucial, aspect of citizen activity. For one thing, we can consider the
specific message: that is, who combine a significant amount of activity with information about
what they want in response. In addition, we can look at those who are active in relation to
particular government benefit programs. And, finally, we can consider the actual messages sent
by activists who engage in modes of participation that permit the communication of information.
62
These data greatly enrich our understanding of what is represented through participation.
Let us begin by considering one group of activists who join an activity of substantial
dimensions with an explicit message: those who contributed $250 or more to a political
campaign and who reported that they "communicated to the candidate or to someone involved in
running the campaign [their] views on an issue of public policy--for example, about what [they]
wanted the candidate to do when in office." Figure 13 indicates how well various sub-groups
defined by their attitudes and economic circumstances are represented within this group of
participants whose activity is high in its potential both for generating pressure and conveying
information to candidates. As expected, those who make major campaign donations are more
representative in terms of their attitudes on economic issues than in terms of their actual
economic needs. The affluent are nearly five times as likely--and those who have had to cut back
on necessities are roughly half as likely--to appear in this politically potent group as is their
proportion of the population. The most marked underrepresentation is found among the poor
and those in households where a family member receives a means-tested benefit. These data
echo the results reported earlier as to who is and who is not represented, but the relationships are
even stronger. When it comes to an act that communicates a good deal of information and has
subsidized housing are much less active than those receiving such non-means-tested benefits as
Social Security, veterans' benefits, or Medicare. We can take the analysis one step further,
however, by probing whether their activity is in any way directly connected to these benefit
programs and whether, therefore, the government receives more messages about benefits
benefits. In the Citpart survey we inquired whether recipients of various government benefits
had been active in relation to that benefit. For each government program for which the
respondent (or a family member living in the household) was a recipient we asked about the
following activities: Had they taken that program into account in deciding how to vote? Had
they given a campaign contribution based, at least in part, on concern about it? Had they
contacted an official to complain about the program? Did they belong to an organization
concerned about that program? The data in Figure 14 on the proportion of the recipients of each
benefit who reported an activity related to the benefit program are consistent with what we know
about the overall activity levels of the recipients of government benefits. For each kind of
means-tested benefits are more likely to have been active than recipients of means-tested
benefits.
associated with the benefit--with the AARP and veterans' organizations presumably playing a
major role--and campaign donations. However, it applies as well to the considerations that enter
65
into voting decisions. The data on contacting are interesting. We might expect that inclusion in
the non-means-tested programs would be more or less automatic and, thus, would require fewer
contacts. Nevertheless, Medicare recipients are more likely than Medicaid recipients to contact
about their medical benefits; Social Security recipients are more likely than AFDC recipients to
contact about their benefits. Clearly, the government hears more from those on some programs
than on others, and the ones it hears from are the more advantaged citizens.
66
67
Votes are limited in their ability to transmit precise information about citizen priorities and
preferences, while contacts can convey a lot more. We asked respondents about the issue or
problem that animated their activity -- in relation to various kinds of activity. Table 5
summarizes the subject matter behind the political activity in which a codeable issue concern was
expressed and compares advantaged and disadvantaged respondents with respect to the issue
concerns that animate their participation. In order to ensure that what is on people's minds is
which an explicit message can be sent: contacting, protesting, campaign work or contributions
board. The issue-based political act is the unit of analysis, and the figures represent the
proportion of all issue-based activities for which the respondent mentioned, among other things,
Although the advantaged and disadvantaged are similar in having wide-ranging policy
concerns, they differ in the distribution of their concerns. Compared with the issue-based activity
of the advantaged, the activity of the disadvantaged is more than twice as likely, and the activity
of respondents in families receiving means-tested benefits four times as likely, to have been
animated by concerns about basic human needs--poverty, jobs, housing, health, and the like.
Moreover, the activity of the disadvantaged is more likely to have been motivated by concern
about drugs or crime. The activity of the advantaged, in contrast, is more likely to have been
inspired by abortion, the environment, or such economic issues such as taxes, government
When we consider the actual number of communications, however, a very different story
emerges. Because the disadvantaged are so much less active than the advantaged, public officials
actually hear less about issues of basic human need from the disadvantaged than from the slightly
smaller group of advantaged respondents--even though references to basic human needs occupy
relatively greater space in the bundle of communications emanating from the disadvantaged.
These findings might suggest that, although the disadvantaged are underrepresented with
respect to participatory input, their concerns and needs are, nonetheless, being expressed by
others. When the disadvantaged speak for themselves on issues of basic human need, however,
their communications differ in two fundamental ways from those sent by others. First, when the
disadvantaged communicate with public officials about basic human needs, they are much more
likely than the advantaged to be concerned about problems that affect them personally. Even
affluent citizens may have need for government assistance with respect to basic human needs:
they may have health problems or a handicapped child in school; if elderly, they receive
70
Medicare and Social Security. Still, a much larger proportion of the messages about basic human
needs from the disadvantaged involve particularized communications about problems specific to
themselves or their families--a question about eligibility for Social Security, a complaint about
Such particularized concerns were behind 56 percent of the issue-based activity in which
human needs issues were mentioned by the disadvantaged, but only 8 percent of that in which
human needs issues were mentioned by the advantaged. Even when the human needs issue was
framed as a policy issue rather than a particularized concern, the disadvantaged were much more
likely to report that the problem is one that affects themselves or their families as well as others
in the community. All in all, of those who communicated to public officials about issues of
basic human needs, 71 percent of the disadvantaged, but only 29 percent of the advantaged were
participation. It may be useful to explore this further. The best predictor of political
activity is education. In our research on participation, my colleagues and I tried to answer the
question: why is education so potent a predictor of political activity? Was it that it made people
more informed? Made them feel more efficacious? Made it easier for them see connections
between their values and preferences and governmental action? Was it that it inculcated the
values of citizenship -- that one ought to be participant? Did it increase store of resources that
people had -- skills that made one an effective participant or income useful in making political
71
contributions? Did it put the individual into networks so that they were surrounded by others
who were active and by others who could help them act effectively? The answer was -- to use
the forced choice language of tests and surveys -- all of the above. Education fosters activity
though its effect on information, skills, values, resources, networks, and more. No wonder it is
The educated citizen is not only more active; she is a better citizen. The educated citizen
is more informed, has a more coherent or consistent set of political values, and can make better
connections between means and ends. Educated citizens have other virtues: they are in general
more supportive of the rules of democracy, more tolerant of unpopular voices, more committed
to communal rather than individualistic goals. Some simple data from our studies will make this
clear. Compare individuals with no college education with college graduates. The latter are
more active in any of the activities we measure, and by a long shot. They are, for instance, four
times as likely to have contacted a government official (a good way to express political voice)
than is someone with only a high school education. They get twice as many information items
right. When they contact, they are four times as likely to deal with an issue that has broad
relevance, rather than some issue limited to themselves and their family; they have, thus, more
civic concern. And in other ways they are exemplary. They are five times as likely to support
the right of someone who opposes religion to speak in the community. And on and on. They are
better citizens.
I don=t want to overstate this case. There is the possibility that we give the educated
citizen more of the benefit of the doubt when it comes to civic virtue that they deserve. And this
may be because they speak our language. Some early research on the civic mindedness of
72
citizens found that better educated citizens were more likely to have a long time horizon and to
think of the public good when contemplating social and economic issues. Less well educated
citizens, newer immigrants had a shorter time horizon and thought of policies from the point of
view of their immediate impact on themselves. I think there is some truth to this, but many years
of looking at what respondents say about public issues has also made clear that those who are
more articulate have little difficulty in expressing our own self-interest in broad social terms. In
our research on unemployment we found that less well educated people need a job and ask for a
job. More educated respondents need a job and discuss the problem of unemployment -- while
also asking for a job. (Schlozman and Verba). Neither myself or my colleagues who have sent
in research proposals to the National Science Foundation ever argued that the research funding
would allow us to do what we like better to do (research) than what we might otherwise have to
do (teach); or that the funding would increase the prestige of our research institute etc. But it
may be (just may be) that some of us have such things in mind -- as well as having a sincere
commitment to the substance of the research and its value to scholarship and the understanding
of society.
Thus, it is to the good that the educated are more active. That education fosters activity
means it fosters better democratic participation. But the educated do not differ from the rest of
the citizenry only in their greater competence and commitment to democracy. They are also
wealthier, more likely to be male, more likely to come from the dominant race and ethnic groups.
They are less likely to support spending on programs to aid the poor. More important, perhaps,
they are less likely to face the deprivations faced by those with less education. On a large
number of measures of need -- the need to put off medical treatment, the need for better housing,
73
etc. -- they differ substantially from those who did not attend college. In sum, educated activists
are more civically competent which makes for a more enlightened input into the democratic
policy process; they are wealthier and more advantaged which means they have policy
preferences and needs different from the population as a whole, and that makes for a more biased
input into the policy process. Insofar as the participation of the educated is driven by the
democratic values of tolerance or by the civic beliefs they acquire in school, this creates a better
citizenry. Insofar, as their activity is driven by the components of their social class position (their
income, the networks of connections that come with various jobs) this creates the more biased
polity. We may want an educated citizenry, but we wind up with a wealthy one.1
Linkage between public preferences and governmental response is, of course, the crux of
the matter. And, as discussed above, it cannot be assumed in any simple manner. Citizen
preferences may be ignored by the incumbent officials; the latter may respond to the long term
needs of the public (as the incumbents see them) rather than their expressed preferences (is that
responsiveness?), and so on. Preferences (of both the citizens and elites) are endogenous. All in
If one considers the public as a whole, there is evidence of similarity between public
preferences and governmental response. (Page and Shapiro, Stimson, et al) The direction of
causality is, however, not clear. Do elites manipulate public opinion to fit their views (Zaller,
1993, but see Zaller, 1999; Ginsberg)? Or do elites pick up cues from a somewhat autonomous
public (Page and Shapiro, Stimson et al)? Or is it an interaction in which elites try to shape the
supporting positions they desire? (Jacobs and Shapiro). There is a connection between public
74
opinion and government action. Whether the two are connected because of top-down causality
or bottom-up causality (or some combination) is not clear. The U.S. has one of the more unequal
income distributions among industrialized democracies, and the role of government appears
closely related to that fact. As Table 6 indicates, the degree of income inequality before
government taxes and transfers is much more similar across nations than afterwards. In most of
the nations studied, government intervention reduces inequality much more than does
government policy in the United States. The connection to participatory inequality has not been
If one turns to the more specific question of the relationship between the inequality of
political activity and governmental response, there is even less to go on. Does it matter if a
particular group participates or not? It is hard to think that it does not. Surely the responsiveness
of the government to the elderly is not unrelated to the rise of political activity in that group.
Nor is the change in the status of African-Americans in southern politics since the voting rights
acts unrelated to their enfranchisement and the other activities that have followed from it. Works
on welfare spending in the American states present some of the only linkage evidence on the
relation between lower class turnout and governmental support for programs to aid the poorer
members of society. (See Table 7 for data on the relationship between estimates and lower class
voting by state and AFDC payments.) (Hill and Leighley, 1992, Hill, Leighley and Hinton-
Andersson, 1995).
75
76
77
For many reasons, many of them spelled out in the beginning of this paper, this is a
difficult subject to study systematically. It is, nevertheless, an important topic in need of further
research.
Equalizing Participation
Are there ways in which group differences in political activity might be equalized?
Equalizing by Law
Making participation easier: There is a variety of ways in which laws can equalize
participation. One might be to make political activity easier. In the past several decades voting
has been made easier with the easing of registration rules and residency requirements and the
motor voter bill. Whether this -- and other changes like Sunday voting -- would reduce the
Aspeaking out@ smack more of authoritarian systems than democratic ones. Required voting --
which exists in some democracies, usually with small penalties for failure to vote -- would
equalize turnout. Figure 15 shows the relationship between educational level and voting turnout
in the Netherlands under an older system of compulsory voting and after that compulsion was
removed. Under compulsory voting, as one would expect, turnout in all educational groups was
higher. But for our purposes, the main point is that the line is flatter under compulsory voting. It
has its biggest impact on those who would otherwise be least likely to vote -- those with lower
78
levels of education. Thus, compulsory voting might make a difference. It is, however, unlikely
to be a policy option in the United States where the voluntary nature of political activity (and lots
of other things) is highly prized. And it would only work for voting.
79
voting -- each voter has one and only one vote. The other area where this might work is in
relation to monetary contributions. This, of course, is what campaign finance reform is all about.
I cannot get into this field in this paper but to comment that limitations on campaign
C The First Amendment since the courts have considered some contributions to be a form
C The fact that incumbents and candidates are not sure they like it and will not likely pass
effective legislation.
C The fact that there are too many smart and well-paid consultants and lawyers in America
Social conditions
income generating systems, one way to equalize participation is to equalize these systems. If
money is the coin of politics, one way to change this is not by limiting the use of money but
equalizing the availability of money through income equalization. Or if education is the key to
activity in general, more equal schools should equalize participation. Educational equalization
or income equalization represent topics beyond the scope of this paper and of my competence. I
just note that the very inequality in political activity that is the subject of this paper militates
Thus, the issue may be: how to get people whose social circumstances (their incomes,
educations, job environments etc.) would not be conducive to be active in politics, to become
active. How do you get those who would otherwise not take part to take part. Two answers are
81
Consciousness
Disadvantaged groups often may increase their activity by becoming aware of their
disadvantage, by seeing a connection between politics and that disadvantage and, newly
conscious, become active. This seems to have played a big role among African-Americans in the
civil rights era (Vereba and Nie, 1972); similar things may be happening to Latinos and newer
ethnic groups. It is a large and complex area of research in which the results are not clear. (See,
for example, Shingles, 1981, Miller et al, 1981, Wilcox, 1990, and Hardy-Fanta, 1997) In part it has
with the direction of causality -- since consciousness may generate activity, but activity is a good
way of generating consciousness. I will not pursue this aspect of the issue here, but rather turn to
a more concrete issue closely related to consciousness: mobilization. Where people become
politically conscious, it is often because there is some set of leaders or organizations that both
foster that consciousness but provide the institutional framework whereby it can be converted
Political activity is not always -- perhaps not often -- spontaneous. It often takes place in
response to requests by others to take part. Thus, political equality may depend upon who is
mobilized.
Parties and unions: In many democracies, the institutions that mobilize those lower on the
socio-economic scale are unions and working class parties. (Powell, Jackman, Lipjhart) Earlier,
we saw data on the relative equality of political activity across social groups in the United States
82
and some European democracies. One of the main reasons for the relative political equality in
the Netherlands or Austria -- the two European democracies where data were available is the role
of parties and working class organizations to mobilize people who would not otherwise be active
to political activity. Figures 16A and 16B illustrate this. They show the relationship between
educational level and political interest (Figure 16A) and between education and voting turnout
(Figure 16B). And they show that relationship for those who have various levels of affiliation
with a political frpom the strongly affiliated to the unaffiliated. As one can see from Figure 16A,
the U.S. and the other two countries look similar. Whether or not you are affiliated with such an
institution, you are more likely to be interested in politics if you are more educated. All the lines
slope upward. The issue is: how do you get the less politically interested people with lower
institution like a party. In Austria and the Netherlands, party attachment dominates voting
turnout. If you are strongly affiliated with a party, you vote, whatever your educational laevel. If
you are unaffiliated, you do not. In the U.S., the individual=s level of education plays a bigger
The possibility of mobilizing inactive in the U.S. is hindered by the weakness of the
parties as mobilizers, a function of the general weakness of party organization and the absence of
parties that are closely identified with the working class. There is evidence, that parties are
becoming less effective as mobilizers -- and this is one of the major sources of declining voter
turnout. (Rosenstone and Hansen). In addition, there is evidence that union strength is directly
related to voter turnout (across nations as well as across the states in the U.S.) and that the
decline in union membership is related to a decline in voter turnout. (Table 8) (Radcliff, Davis).
83
Churches: Of the various institutions in the United States that could mobilize socio-
84
economically disadvantaged people, churches are probably the most potent. Americans are, in
general, church going. And, unlike many institutions in the United States, they are not stratified -
- i.e., the proportions of people attending church are (unlike the proportions belonging to secular
organizations) are not tilted towards the better off. Churches can provide individuals with
resources such a civic skills not otherwise available to them. Churches have done this most
effectively for African-Americans (Verba, Schlozman and Brady, 1995, Harris, 1999,). The
potential for a church based effect can be seen if we compare the degree of church affiliation with
union affiliation for American blue collar workers. Sixty-seven percent of the blue collar
workers in the Citpart sample reported church membership compared with 26 percent who
reported membership in a union. And church members are much more likely to attend church
services than are union members to attend a union meeting. (52% of blue collar worker union
members report having attended a union meeting in the past year, 94 percent of the blue collar
church affiliates report attending a church service in the past year --67% attend nearly every
week).
Churches can mobilize those of lower socio-economic status to a level of political activity
beyond that which would be expected given their social characteristics. In this they can play the
role that workers= parties or unions have traditionally played in some other industrialized nations.
The difference may be the issue agenda in relation to which they do the mobilization. The
African-American church, especially during the civil rights era, mobilized lower status blacks for
a racial and social welfare agenda. Today, many churches mobilize their lower status members
the mobilization of groups that would otherwise be quiescent. In so doing, the study these
groups -- minorities, women, the handicapped, the elderly, etc. -- to see how mobilization affects
their activity. But if one takes a wider view of mobilization, looking at the overall picture of
mobilization attempts -- at all the telephone calls we get a dinner to support this or that candidate
or cause, or that fill up our mailboxes -- we find that rather than political mobilization being
counter-stratificational, it reinforces the stratification patterns found in the public already. (See
Brady, Verba, and Schlozman, 1999). Political recruiters are, our analysis shows, Arational
prospectors@. They seek out people who are likely to participate and who, if they participate,
will do so effectively. They seek out those who have participated before and have the resources
for participation. As Figure 17 shows, lots of people never are contacted to support a candidate
or cause; and those outside of such networks are disproportionately poor. The result is that
mobilization -- looked at most generally -- only makes the participation input from the American
public more skewed towards the better off. This is especially the case when it comes to political
contributions. Recruiters, with good reason, seek out the affluent. The result is that
contributions that are mobilized tend to come from the affluent more than is the case for
spontaneous contributions (Table 9), and they are larger. Thus, mobilization does not reduce
REFERENCES
Bibliography
Allen, M.P., and John L. Campbell, "State Revenue Extraction from Different Income
Groups:Variations in Tax Progressivity in the United States, 1916-1986", American Sociological
Review (1994) 59, 169-186.
Bennett, S.E., and David Resnick, "The Implications of Nonvoting for Democracy in the United
States", American Journal of Political Science (1990) 34:3, 771-802.
Bobo, L and F Gilliam, "Race, Socio-Political Participation and Black Empowerment", American
Political Science Review (1990) 84:2, 377-393.
Brehm, John, The Phantom Respondents: Opinion Surveys and Political Representation (Ann
Arbor: University of Michigan Press, 1993).
Browning, R.P., D.R. Marshall, and D.H. Tabb, Protest is Not Enough: The Stuggle of Blacks
and Hispanics for Equality in Urban Politics (Berkeley: University of Calfornia Press, 1984).
Cohen, C.J., and Michael C. Dawson, "Neighborhood Poverty and African-American Politics",
American Political Science Review (1993) 87:2, 296-302.
Dahl, Robert Alan, Democracy and its Critics (New Haven: Yale University Press, 1989).
Delli Carpini, Michael X. and Scott Keeter, What Americans Know About Politics and Why it
Matters (New Haven: Yale University Press, 1996).
DeNardo, James, "Turnout and the Vote: The Joke's on the Democrats", American Political
Science Review, (1980) 74, 406-20.
Galen Irwin, "Compulsory Voting Legislation: Impact on Voter Turnout in the Netherlands, "
Comparative Political Studies (1974), 7, 292-315.
Gay, Claudine, Taking Charge: Black Electoral Success and the Redefinition of American
Politics, a thesis presented by Claudine Gay, 1997.
Ginsberg, Benjamin, The Captive Public: How Mass Opinion Promotes State Power (New York:
Basic Books, 1986).
90
Hill, Kim Quaile, Jan E. Leighley, and Angela Hinton-Andersson, "Lower-class Mobilization and
Policy Linkage in the U.S. States", Journal of Politics (1995) 39:1, 75-86.
Hill, Kim Quaile and Jan E. Leighley, "Mobilizing Institutions and Class Representation in U.S.
State Elections", Political Research Quarterly (1994) 47, 137-50.
Hill, Kim Quaile and Jan E. Leighley, "The Policy Consequences of Class Bias in State
Elections", American Journal of Political Science (1992) 36, 351-65.
Jackman, R., "Democratic Institutions and Voter Turnout in the Industrialized Democracies",
American Political Science Review (1987) 81, 405-434.
Jacobs, Lawrence R. and Robert Y. Shapiro, Politicians Don't Pander: Political Manipulation and
the Loss of Democratic Responsiveness (Chicago: University of Chicago Press, 2000).
Kornhauser, William, The Politics of Mass Society. (Glencoe, IL: Free Press, 1959).
Lippmann, Walter, The Phantom Public: a Sequel to "Public Opinion" (New York: Macmillan,
1925).
Miller, Arthur H., Patricia Gurin, Gerald Gurin, and Oksana Malanchuk, "Group Consciousness
and Political Participation", American Journal of Political Science (1981) 25, 494-511.
Mueller, John E., Capitalism, Democracy, and Ralph's Pretty Good Grocery (Princeton, NJ :
Princeton University Press, 1999).
Petrocik, John, "Voter Turnout and Electoral Preferences: The Anomalous Reagan Elections", in
Kay L. Schlozman ed., Elections in America (New York: Allen and Unwin, 1987).
Piven, Francil F., and Richard Cloward, Why Americans Don't Vote (New York: Pantheon,
1988.
Popkin, Samuel L., The Reasoning Voter: Communication and Persuasion in Presidential
91
Putnam, Robert D., Bowling Alone: the Collapse and Revival of American Community (New
York: Simon & Schuster, 2000).
Radcliff, B., and P. Davis, "Labor Organization and Electoral Participation in Industrial
Democracies", American Journal of Political Science, (2000) 44, 132-141.
Rosenstone, S, and J.M. Hansen, Mobilization, Participation, and Democracy in America (New
York: Macmillan, 1993).
Scanlon, Thomas, What We Owe to Each Other (Cambridge: Belknap Press of Harvard
University Press, 1998).
Schlozman, Kay Lehman and Sidney Verba, Injury to Insult: Unemployment, Class, and Political
Response (Cambridge: Harvard University Press, 1979).
Shaffer, Stephen "Policy Differences Between Voters and Non-voters in American Elections",
Western Political Quarterly (1982), 35, 396-410.
Shingles, Richard, "Black Consciousness and Political Participation: The Missing Link",
American Political Science Review (1981) 75, 76-91.
Sniderman, Paul M., Richard A. Brody, and Philip E. Tetlock, editors, Reasoning and Choice:
Explorations in Political Psychology (New York: Cambridge University Press, 1991).
Stimson, James A., Michael B. MacKuen, and Robert S. Erikson, "Dynamic Representation",
American Political Scence Review (1995), 89, 543-65.
Stouffer, Samuel Andrew, Communism, Conformity, and Civil Liberties: a Cross-Section of the
Nation Speaks its Mind (Garden City, NY: Doubleday, 1955).
Sullivan, John L., James Piereson, and George E. Marcus, Political Tolerance and American
Democracy (Chicago: University of Chicago Press, 1982).
Teixeira, Ruy A., The Disappearing American Voter (Washington, D.C.: Brookings Institution,
1992).
92
Thomas, J.C., and J. Melker "Explaining Citizen Contacts with Municipal Bureaucrats", Urban
Affairs Review (1999) 34:5, 667-690.
Verba, S., "The Citizen as Respondent", American Political Science Review, March, 1996.
Verba, S., and Norman Nie, Participation in America (New York: Harper and Row, 1972).
Verba, S., Norman Nie, and Jae-on Kim, Participation and Political Equality: A Seven Nation
Comparison (New York: Cambridge University Press, 1979).
Verba, S., and G. Orren, Equality in America (Cambridge: Harvard University Press, 1985).
Verba, S., K.L. Scholzman, and Henry E. Brady, Voice and Equality (Cambridge: Harvard
University Press, 1995).
Verba, S., K.L. Schlozman, and Nancy E. Burns, "Knowing and Caring About Politics", Journal
of Politics [insert volume number and page numbers]
Walton, Hanes, Jr., Invisible Politics: Black Political Behavior (Albany: State University of New
York Press, 1985).
Wilcox, Clyde, "Race, Gender Role Attitudes and Support for Feminism", The Western Political
Quarterly (1990) 43, 113-121.
Wolfinger, R and S. Rosenstone, Who Votes? (New Haven: Yale University Press, 1980).
Zaller, John, The Nature and Origins of Mass Opinion (New York: Cambridge University
Press, 1992).
1.