Book Search:  


Google full text of our books:


How Policies Make Citizens:
Senior Political Activism and the American Welfare State
Andrea Louise Campbell

Book Description | Reviews | Table of Contents

COPYRIGHT NOTICE: Published by Princeton University Press and copyrighted, © 2003, by Princeton University Press. All rights reserved. No part of this book may be reproduced in any form by any electronic or mechanical means (including photocopying, recording, or information storage and retrieval) without permission in writing from the publisher, except for reading and browsing via the World Wide Web. Users are not permitted to mount this file on any network servers. Follow links for Class Use and other Permissions. For more information, send e-mail to

This file is also available in Adobe Acrobat PDF format

Chapter 1



SOME GROUPS participate more than others--the affluent more than the poor; the educated more than the uneducated; whites more than blacks and Latinos; the elderly more than the young. Does it matter? Do high-participation groups get more of what they want from the government? Do participatory inputs shape policy outputs? Critics allege that the American system of government inadequately represents the interests of the underprivileged, to their detriment.1 Indeed, one motivation driving researchers to measure participatory differences across groups is an assumption that they lead to unequal policy outcomes. Do they?

And why are some groups more politically active than others? We know part of the story--some individuals possess more politically relevant resources, like income and education, than others, some are more interested in public affairs, and some are more likely to be recruited to participate. And these factors arise from early socialization at home and in school and from affiliations with voluntary associations, workplaces, and religious institutions.2 But what about the role of public policy? Might government policy also be a source of political resources, a sense of stake in the system, even a mobilizing factor?

This book focuses on the reciprocal relationship between political participation and public policy. I show that mass participation influences policy outcomes--the politically active are more likely to achieve their policy goals, often at the expense of the politically quiescent. And the ability of the politically active to do so is in part a legacy of existing public policy--policy influences the amount and nature of groups' political activity, often exacerbating rather than ameliorating existing participatory inequalities. Public policies can confer resources, motivate interest in government affairs by tying well-being to government action, define groups for mobilization, and even shape the content and meaning of democratic citizenship. These effects are positive for some groups, like senior citizens, raising their participation levels. For other groups, government policy can have negative effects. Because of the difficult and demeaning experience of obtaining welfare, for example, its recipients participate at rates even lower than their modest socioeconomic situations would predict.3 These effects feed back into the political system, producing spirals in which groups' participatory and policy advantages (or disadvantages) accrue. Citizens' relationships with government, and their experiences at the hand of government policy, help determine their participation levels and, in turn, subsequent policy outcomes.

With its influences on participation, public policy affects the basic mechanisms of democratic government. An important function of government is the allocation of societal goods. In a democracy, the people are supposed to voice their preferences through their political participation. Indeed, democratic theory is predicated on the equal ability of citizens to take part in this way.4 But policy design--who gets benefits, how generous they are, and how they are administered--affects groups' participatory capacities and interests. The distributional consequences are profound. Policy begets participation begets policy in a cycle that results not in equal protection of interests, but in outcomes biased toward the politically active. Thus the very quality of democratic government is shaped by the kinds of policies it pursues.

Senior citizens in the United States and their activity in relation to Social Security form the empirical basis for this study. Social Security and Medicare transfer 40 percent of the federal budget to seniors, with significant effects on their political behavior. They are the Über-citizens of the American polity, voting and making campaign contributions at rates higher than those of any other age group. They also actively defend their programs, warning lawmakers through their participation not to tamper with Social Security and Medicare. The result is continued program growth, even as programs for the poor are cut.

Seniors were not always so politically active. In the 1950s, when Social Security benefits were modest and covered only a fraction of seniors, the elderly participated at lower rates than younger people. The growth of the program in part fueled the increase in senior participation over time. Social Security provided the once-marginalized senior population with politically relevant resources like income and free time. The program increased seniors' engagement with politics by connecting their fortunes tangibly and immediately to government action. It fashioned for an otherwise disparate group of people a new political identity as program recipients, which provided a basis for mobilization by political parties, interest groups, and policy entrepreneurs. And Social Security incorporated seniors into the highest level of democratic citizenship, their relationship with the state marked by full social and political rights and privileges, including the right to fend off proposals for program change that they find objectionable. Indeed, Social Security (and Medicare) have provided two stimuli: the growth of the programs, which has enhanced seniors' participatory capacity over time, and threats to the programs, which have inspired participatory surges that lawmakers heed and that protect the programs from retrenchment efforts. The combination of these stimuli produces a loop in which senior welfare state programs expand: the programs enhance seniors' participatory capacity so that when their programs are threatened they are able to defend them; thus protected, the programs continue to augment seniors' participatory capacity (figure 1.1). In short, Social Security helped create a constituency to be reckoned with, a group willing, able, and primed to participate at high rates, capable of defeating objectionable policy change.

Welfare recipients are not so fortunate. In contrast to Social Security, welfare has negative effects on its clients. Welfare recipients do have an interest in public affairs arising from their dependence on government action, but this level of interest is lower than that of Social Security recipients and cannot by itself enhance participation levels--adequate political resources and mobilization are needed too, and here welfare recipients are severely disadvantaged. Indeed, the design of welfare, which requires recipients to meet with caseworkers who ask probing personal questions and who appear to have great discretion over benefits, undermines recipients' feelings of efficacy toward the welfare system and the government in general, reducing their rates of political participation.5 The lack of positive policy feedbacks helps make welfare an easier fiscal target and contributes to greater retrenchment in that policy area.


The participation literature has made impressive empirical and theoretical gains in explaining who is politically active, but has largely neglected the crucial question of differential policy outcomes. As the dean of participation scholars, Sidney Verba, notes, there has been much emphasis on equal capacity and equal voice across demographic groups, but understanding the real result of participation requires consideration of equal outcomes. The literature assesses what individuals do rather than what effect their activity has, which is a far more difficult and complex task.6

Anecdotally we can think of high-participation groups prevailing in the policy arena and low-participation groups suffering: during the 1990s the wealthy saw capital gains tax rates reduced while welfare lost its entitlement status. But linking participatory inputs to policy outcomes is difficult, and has rarely been done systematically.

There have been some attempts to connect participation and policy. Frances Fox Piven and Richard Cloward argue that low turnout rates among the poor result in public policy that favors higher-status voters.7 Kim Quaile Hill and Jan Leighley find a correlation between turnout among the poor and levels of welfare spending among the fifty states.8 But these studies use voter turnout, a "blunt" instrument of participation for which issue content must be assumed.9 That the poor participate less is compatible with the notion that the highly participatory prevail in the policy arena, but does not itself make the link. No single study has yet connected the dots and shown conclusively that differential participation rates across groups influence policy outcomes.


Thanks to advances in polling technology over the past half century, we know a great deal about who participates in terms of education, income, race, ethnicity, and gender. And increasingly sophisticated models tell us why some individuals and groups participate more than others. One factor past work leaves out is the effect of government policy.

In their seminal 1972 work, Participation in America, Sidney Verba and Norman Nie showed that participation of all forms, from voting to protesting, is more common among individuals of higher socioeconomic status (SES), which they measured with an index combining education, income, and occupation. In Who Votes? (1980) Raymond Wolfinger and Steven Rosenstone took the SES model one step further, disaggregating SES into its three components. They found that education has the most profound effect on voting.

In the most recent major work on political participation, Voice and Equality (1995), Verba, Kay Lehman Schlozman, and Henry Brady develop a Civic Voluntarism Model to explain the mechanisms by which SES influences participation. Political activity is fueled by three "participatory factors": resources, like income and politically relevant skills; engagement with politics, including political interest and knowledge; and mobilization. These factors arise from personal characteristics, preadult experiences at home and school, and affiliations with the workplace, religious institutions, and voluntary organizations.

Although these developments in the participation literature have considerably enhanced our understanding of participatory differences, they have largely ignored policy influences on would-be participators--how government policies might influence these participatory factors. The policy feedbacks literature in American and comparative politics suggests that policy effects are central. The work of historical institutionalist scholars such as Theda Skocpol, Paul Pierson, Sven Steinmo, and Peter Hall shows, at the macro level, how existing policy structures constrain subsequent policymaking. Current policies foreclose some possibilities, preordain others.10 For example, Skocpol argues that the extensive system of pensions for Civil War veterans delayed the broader implementation of old-age pensions for decades, because of their association with corrupt patronage politics.

Like Skocpol's work, most research in this area examines policy influences on states and elites. But Pierson suggests that policy influences on mass publics, while understudied, are among the most important.11 He further asserts that they take two forms, material and cognitive. Indeed, the development of Social Security fulfills Pierson's predictions about the influences policies might have on client groups. By conferring politically relevant resources, Social Security has had tremendous material effects, fundamentally enhancing seniors' participatory capacity above what they could have achieved in the absence of the program. The program's cognitive effects may be even more significant. Cognitive feedback effects provide otherwise scarce and costly information to individuals. According to Pierson, these cues "may influence individuals' perceptions about what their interests are, whether their representatives are protecting those interests, who their allies might be, and what political strategies are promising."12 Social Security's cognitive effects have fostered senior interest in public affairs and enhanced their feelings of political efficacy as they achieved the notice of elected officials.

The political learning literature suggests an additional mechanism by which the cognitive effects operate--that the manner in which government policies treat clients instills lessons about groups' privileges and rights as citizens. Policy design sends messages to clients about their worth as citizens, which in turn affects their orientations toward government and their political participation. Policy experiences convey to target populations self-images and outgroup images (who is "deserving" and who is not). Program recipients learn what actions are appropriate for their group. "Advantaged" groups like Social Security recipients consistently hear messages that they are "good, intelligent people" who have legitimate claims with the government.13 Social Security fosters seniors' participation not only by enhancing their participatory capacity and giving them a compelling reason to pay attention to public affairs but also by affirming their rights as citizens to defend their benefits.

The profound effect of policy design on democratic citizenship is perhaps most evident with low-income seniors. Decades of participation research have shown that high-income individuals are more politically active. The income-participation gradient is particularly steep in the United States, which lacks lower-class mobilizing agents like leftist parties or strong unions. As we will see in chapter 3, however, low-income seniors are more likely than high-income seniors to participate with regard to Social Security. They are, for example, more likely to write a letter to a government official about the program, so that the usually positive relationship between income and participation is reversed. The difference in this policy area is that low-income seniors are more dependent on the program: Social Security makes up a larger share of their income, and so they are rationally more active. Furthermore, as a universal, non-means-tested program, Social Security includes low-income seniors in its "advantaged" recipient group. Unlike other low-income groups, poor seniors absorb from their policy experiences the same positive citizenship lessons their more affluent counterparts learn, including the message that defending their self-interests through political activity is legitimate.

Thus Social Security has implications for self-interest as a political motivation. The program ties seniors as citizens to state functions in an immediate way. Their engagement with public affairs is enhanced because their self-interest is so significantly and obviously implicated. Self-interest is seldom a guide to behavior, since people often do not recognize their interests or may choose to act in an altruistic manner.14 But seniors' interest in Social Security is so immediate, quantifiable, and tangible that it influences their activities. The political gerontologist Neal Cutler predicted twenty-five years ago that the life changes associated with aging and retirement--reduced income, dependence on Social Security, and need for medical care--would produce a new set of self-interests that are important and immediate enough to shape behavior.15 Social Security's design empowers both high-and low-income seniors to defend these new interests.

And Social Security fundamentally alters the nature of citizenship for seniors, shaping the manner and extent to which they are included as members of the political community.16 According to T. H. Marshall, citizenship has three components: civil, social, and political.17 As American citizens, seniors enjoy the civil rights of liberty, property, freedom of speech, and equality before the law. As senior citizens receiving Social Security, they enjoy, more than most other groups, the social rights of security and welfare that are crucial aspects of twentieth-century notions of citizenship.18 And in part because of their full enjoyment of social rights, seniors have realized political rights--access to democratic decision making via participation--to a greater degree than many other groups, and more completely than they would have in Social Security's absence.


According to the comparative politics literature, social movements and interest groups shape welfare state outcomes.19 Welfare state programs supposedly spawn large, active constituencies that exert political pressure to maintain or expand their programs, even in an era of retrenchment. Nowhere does this seem as likely as with Social Security and Medicare, the two largest social welfare programs in the United States.

This theoretical claim is plausible enough that political commentators and pundits have seized upon it. Journalistic accounts of "greedy geezers," "golf-cart grannies," and "the pampered elderly" abound.20 Media coverage of election outcomes often refers to a powerful senior voting bloc.21 Such accounts correctly note that senior citizens vote at high rates, but writers often make unsubstantiated claims about seniors' policy preferences and cohesiveness.

Political scientists have left these assumptions largely unexamined, and even make such assumptions themselves. American government textbooks, for example, often cite seniors as a powerful voting bloc or as an example of an effective interest group, with little evidence: "The political risks associated with Social Security cutbacks are too great for most politicians to bear. As a group, older Americans exercise enormous political power" asserts one prominent textbook, a plausible but undocumented statement.22 Even scholarly treatments make assumptions about seniors' political power.23 Basic questions about senior participation have been left unanswered: the level of senior participation beyond voting; changes in senior participation over time; the meaning of that participation; and the effect of senior participation on the policymaking process.24

Demographic and fiscal trends suggest the normative importance of a study of senior participation and its effects. Seniors25 are a large and growing portion of the U.S. population. The senior segment has grown from 8.1 percent of the U.S. population in 1950 to 12.7 percent in 1999, and is projected to reach 20 percent in 2030.26 There are now more than 34 million Americans aged 65 and over, nearly three times as many as there were in 1950. Seniors form an even larger proportion of the electorate. In 1996, when seniors were 12.7 percent of the total population, they made up 17.2 percent of the votingage population and 24 percent of all voters in the presidential election.27 As the senior population has grown, so have welfare state programs conferred on the basis of age. Social Security and Medicare together totaled $555 billion in 1997, almost 35 percent of federal budget outlays and 40 percent of outlays excluding interest on the national debt.28 This is twice the size of the defense budget and twenty-seven times federal and state spending on welfare.

Fortunately, given its theoretical and normative importance, the senior citizen case study offers analytical tractability as well. To link participation and policy in a given issue area it is necessary to know the participation rate of the "issue public," that is, the participation rate among people interested in the issue.29 This is calculated by dividing the number of issue public members who participate by the total number of issue public members. These data usually come from surveys of the mass public. The difficulty in many issue areas is that we do not know the denominator, the size of the issue public. Without the appropriate identifying questions in surveys, the researcher cannot determine who is interested in the environment or gun rights or abortion. Longitudinal research is even more difficult, since chances are slim that identifying questions were repeated consistently over time.

Issue publics must also be large enough for statistical analysis. There are so many issues in the public realm that even with the appropriate identifying questions, few would yield publics large enough for analysis in a typical survey of 1,500 or 2,000 respondents.

In addition, we often know from surveys whether respondents participated but not the content of their participation. We might know that a survey respondent wrote a letter to an elected official but we do not know the issue mentioned or the position the respondent took. It is difficult in these circumstances to determine the meaning of respondents' participation.

The senior citizen case study avoids these problems. Well over 90 percent of senior citizens receive Social Security and Medicare, and senior citizens are identifiable in virtually all relevant surveys, including longitudinal datasets. Constituting such a large proportion of the votingage population, the elderly provide plenty of cases for statistical analysis in typical national survey samples.30

Further facilitating analysis is the fact that the vast majority of seniors are on the same side of the Social Security issue. Opinion data show that senior citizens are virtually unanimous on Social Security: they don't want their programs cut. When asked in the 2000 National Election Study (NES) whether they wanted federal spending on Social Security increased, decreased, or kept the same, only 2 percent said decreased; when asked about Social Security benefit levels in the 1992 NES, just 2 percent said benefits were too high. Seniors want either to preserve the status quo or to increase spending on their programs. That seniors are on one side of the issue greatly aids interpretation of two key data sources. The 1990 Citizen Participation Study includes program-specific participation questions, asking respondents whether they have voted, made a campaign contribution, or written a letter about Social Security. Seniors' near unanimity on this issue means we can be assured of the pro-Social Security direction of that participation. A new dataset assembled for this study--the Roper Social and Political Trends Archive--consists of surveys administered ten times per year over a twenty-year period and includes questions on participatory activities such as writing letters to Congress. These data can reveal participatory reactions to specific policy events, and so the timing of the letters makes evident the issue content. This overcomes the problem that plagued earlier attempts to link participation and policy--having to assume the issue content of the vote. And near unanimity in support of Social Security means we can interpret a surge in senior letter writing in response to proposed cuts as a defensive move.

From this study a complex picture of senior participation and policy emerges. One aspect is that, as with many stereotypes, there is a grain of truth in the "greedy geezer" image. To an extent, seniors defend their own programs at the expense of policies benefiting others. Economists have shown, for example, that the proportion of elderly residents in a location is associated with a significant reduction in per child education spending.31 By contrast, there was a positive association between elderly population share and school spending in the early 1900s, before Social Security's enactment.32 Government policy seems to have shifted seniors' self-interests away from supporting the education of younger people upon whom they were once directly dependent, toward defending the government benefits upon which their livelihoods now depend.

But if this crowding-out effect is not desirable from the standpoint of democratic governance, other age-related policy effects are. Social Security has democratized senior participation, reducing participatory inequality within the senior constituency. Hence seniors' welfare state programs have exacerbated political inequality between age groups, but have moderated it within the senior population. Social Security has both deleterious and laudable effects on American democracy. The reality of "senior power" is therefore more ambiguous--and the development of the senior constituency considerably more complicated--than previous accounts suggest.


Chapter 2 describes two historical trends: the rise of senior participation since the 1950s and the development of Social Security and Medicare during the same period. Over time seniors eclipsed nonseniors in the participatory arena, and now are disproportionately active. At the same time, Social Security and Medicare coverage expanded and benefits grew, with profound effects on seniors' empirical and attitudinal well-being. This chapter hints at the causal connections between these two phenomena, which are explored systematically in chapter 3.

Chapter 3 shows the individual-level mechanisms behind senior participation in cross-sectional data. Like younger people, seniors with higher levels of participatory resources, political engagement, and mobilization opportunities are more likely to participate in politics. We also see that Social Security has some unusual participatory effects. Although general engagement with politics increases with income among seniors (as among nonseniors), interest in Social Security is greater among lower-income seniors, who are more dependent on the program. This boosts the participation of low-income seniors above what we would expect for persons of such low socioeconomic status and helps explain the high level of senior participation overall. This effect is reinforced by the pattern of mobilization by political parties, which are more likely to mobilize low- than high-income seniors.

Chapter 4 elaborates on the participation-policy cycle depicted in figure 1.1, showing how the factors identified as important in senior participation grew over time and, in particular, examining the role of senior welfare state programs in enhancing those factors. Among the trends the chapter explores are: the rise in senior resources, the rise in senior dependence on Social Security as coverage and benefits levels increased, the politicization of senior programs and senior interest groups, and changes in party strategies.

Chapters 2 through 4 show how senior participation increased over time and how Social Security contributed to that increase--in short, how policy influences participation. Establishing the other half of the reciprocal relationship--the influence of participation on policy--is the subject of chapters 5 and 6. Making the link between participation and policy requires three steps: showing that the group in question participates at high rates; determining that the high participation sends a distinctive message to the government; and demonstrating that lawmakers hear and react to that distinctive message. That a group participates at disproportionately high rates is of no consequence if the participation does not send a unique message to the government. That unique message is in turn meaningless if lawmakers fail to take it into account in their policymaking.

The earlier part of the book shows that seniors participate at high rates; chapter 5 demonstrates the distinctive message behind that participation. Seniors are not just interested in Social Security, but also react to Social Security policy events. Using the new Roper data, I show that seniors respond to threats to their programs with surges in letter writing to Congress, their contacting rates rising by as much as 50 percent. Senior participation is high and sends a message to lawmakers: do not tamper with Social Security and Medicare. Social Security has been called the "third rail of American politics"--touch it and you die--not because seniors are interested in politics generally, or even because they are vigilant about Social Security, but because they act when their programs are threatened. Politicians see these surges in senior participation in reaction to policy events and come to interpret high participation by seniors in general as focused on their welfare state programs, an indication that to propose program changes or cuts is to invite electoral retribution.

Chapter 6 shows how members of Congress react to senior participation. Senators and representatives who hear more from their senior constituents are more protective of senior programs. In some dramatic cases, lawmakers even switch their votes on age-related policy in the face of protests by their senior constituents. Participation shapes policy outcomes.

Chapter 7 looks briefly at the participation of others who receive government benefits--veterans and welfare clients--to begin to explore the conditions under which government programs contribute to upward or downward participation-policy spirals. Like Social Security, veterans' benefits confer resources that enhance participation, foster interest in public affairs, endow recipients with a political relevance that invites mobilization by interest groups and parties, and enhance recipient feelings of government responsiveness. These program recipients participate at higher levels than they would in the absence of the programs. Welfare recipients, by contrast, participate at even lower levels than their already modest participatory capacities would predict, largely because of the disengaging aspects of program design that relegate them to a lower tier of democratic citizenship.33

Chapter 8 summarizes the study's findings about the reciprocal relationship between policy and participation. I consider the political implications of proposed changes in Social Security's design, such as the introduction of individual accounts (Social Security "privatization"), and discuss the degree to which Social Security is likely to retain its "third rail" reputation in the future. Privatization threatens to demobilize the poor elderly in particular, by breaking the tie between their well-being and government action. Their resource levels may also be reduced if administrative costs or the vagaries of the market produce lower returns than Social Security's current defined benefit structure. Whether through diminished engagement or reduced benefits, privatization could exacerbate political inequality among seniors, erasing Social Security's democratizing effect. Indeed, proposals to change the program's structure could expose differential class interests among the elderly, heretofore hidden by Social Security's universal, if somewhat redistributive, design. Cleavages could open in this population that until now has successfully maintained cohesion around its welfare state programs. In short, privatization could break the back of the senior lobby.


This study uses mass survey data to evaluate the participation and attitudes of seniors and nonseniors. These data come from a variety of cross-sectional and time-series sources.

The National Election Study time series is crucial for determining the participation rates of seniors and other age groups since the early 1950s (chapters 2 and 4). Although the NES participation time series does not begin before the implementation of Social Security in 1940 (which would be ideal), it does extend back to 1952, when only 16 percent of seniors received benefits and when the real value of the average benefit was just over one-third of today's benefit. I also use some of the recent NES cross-sectional surveys to perform multivariate analyses of the influences behind senior participation (chapter 3).

The American Citizen Participation Study is the most comprehensive survey of participatory activity currently available.34 It contains detailed questions about political engagement and mobilization unavailable in other datasets, and also includes Social Security-specific participation items that are crucial to this study. The survey was conducted in two stages. First, 15,053 Americans were interviewed by phone about their political and voluntary activities and demographic characteristics. A longer, in-person follow-up interview was conducted with 2,517 of the original respondents. I chiefly use data from the follow-up interview, along with select variables from the initial screener interview (chapter 3).

The heart of the book's argument is based on a time series-cross sectional dataset assembled for this project: the Roper Social and Political Trends Archive, a series of approximately 200 national cross-sectional surveys conducted ten times per year from September 1973 through October 1994. The Roper surveys contain a battery of twelve participation questions as well as basic demographic information. Each of the 200-plus surveys has approximately 2,000 respondents, so that the concatenated dataset contains over 400,000 cases.35 One limitation of the NES data is that the participation battery consists mostly of electoral items, and the survey itself is administered biennially in tandem with the election cycle. The Roper surveys contain nonelectoral participation items like letter writing to Congress, and because the surveys were conducted approximately monthly, the concatenated dataset can be used to assess participatory reaction to specific policy events (chapter 5). Occasionally the Roper Organization also asked questions about interest in Social Security and other matters relevant to the study of age-related policy, and I use several of the individual surveys for cross-sectional analysis as well (chapter 3).

Finally, in chapter 6 I combine the Roper surveys with congressional roll-call data to assess lawmaker responsiveness to senior constituents.

Return to Book Description

File created: 8/7/2007

Questions and comments to:
Princeton University Press

New Book E-mails
New In Print
PUP Blog
Princeton APPS
Sample Chapters
Princeton Legacy Library
Exam/Desk Copy
Recent Awards
Princeton Shorts
Freshman Reading
PUP Europe
About Us
Contact Us
PUP Home

Bookmark and Share