Book Search:  


Google full text of our books:


Why We Vote:
How Schools and Communities Shape Our Civic Life
David E. Campbell

Book Description | Reviews | Table of Contents

COPYRIGHT NOTICE: Published by Princeton University Press and copyrighted, © 2006, 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


On September 26, 1989, Traci Hodgson cast her ballot in Boston’s City Council election. It was the only vote cast in her precinct. For a number of reasons, the political science literature predicts that Traci should have been like the other 275 registrants in her precinct and not turned out to vote. She was only twenty-one, she had lived in Boston for less than two months, and she admitted that she was “not very familiar with the candidates running.” So why did she vote? When asked she replied, “I just think it’s important to vote. If you have the right, you ought to exercise it—whether you are going to make a difference or not” (Mooney 1989).

This book is about why Traci voted, and why she voted alone.


Our starting point in answering this question is not what did motivate Traci, but rather what did not. Clearly she did not vote to protect or advance her interests, as she admitted that she did not know enough about the candidates to select them on the basis of who best represented her. To someone with only a cursory familiarity with the study of American politics this may seem difficult to explain, as politics is typically described as a forum for the “clash of interests.” Equating politics with conflict underpins much of the political science literature and is an assumption shared by scholars working within many theoretical frameworks. I recall an introductory political science course in which a professor defined politics as the “scarcity of consensus.” In 1960, E. E. Schattschneider wrote simply, “At the root of all politics is the universal language of conflict” (2). Forty years later, Morris Fiorina and Paul Peterson note matter-offactly in their introductory textbook on American politics that “politics is fundamentally about conflict” (1998, xvi). For all the ink spilled by contemporary political scientists, however, no one has ever expressed this way of understanding politics better than James Madison in Federalist 10. Madison writes compellingly of how “the latent causes of faction are thus sown in the nature of man,” and that “the most frivolous and fanciful distinctions have been sufficient to kindle their unfriendly passions and excite their most violent conflicts” (1961, 79).

It does not seem, however, that Traci had her passions kindled as Madison describes. Instead, by her own account she was motivated by the glowing embers of obligation. She felt that she ought to vote, that it was her duty.

In invoking a sense of duty as a motivation to vote, Traci highlights a second, if more subtle, theme in both contemporary and classic writings on political engagement. If, in Federalist 10, Madison has written the quintessential statement on political participation as “protecting one’s interests,” then perhaps Tocqueville has written an equally quintessential description of political participation as driven by “fulfilling one’s duty.” In Democracy in America, Tocqueville observes that American political institutions lead citizens to see that “it is the duty as well as the interest of men to be useful to their fellows. . . . What had been calculation becomes instinct. By dint of working for the good of his fellow citizens, he in the end acquires a habit and taste for serving them” (1988, 512–13).

These two opposing views of politics are not merely the abstract statements of theoreticians, the political science rendition of how many angels can dance on the head of a pin. They also inform the writings and doings of political practitioners, with some of America’s founders as notable examples. Of course, Madison was one of the founders and, as noted, in his words we find a cogent description of how politics is inevitably defined by conflict. However, George Washington’s vision was of a republic free from strife among its citizens, in which citizens were involved in public affairs out of duty (Schudson 1998). Ironically, given their disagreements on so many other matters, Thomas Jefferson’s vision for the new republic mirrored Washington’s. Jefferson idealized a nation of small Tocquevillian communities, mentioning his ideal in almost every speech he gave (Morone 1990).

The ongoing debate between civic republican and liberal political philosophers over the nature of political life at the time of America’s founding underscores the two competing visions of democracy’s nature. In summarizing this extensive literature, James Morone writes,

[T]he dominant interpretation of liberal America focuses on the pursuit of self-interest. . . . In the republican view, the colonial and Revolutionary ideal lay, not in the pursuit of private matters, but in the shared public life of civic duty, in the subordination of individual interests to the res publica. Citizens were defined and fulfilled by participation in political community. (1990, 16, emphasis added)

The distinction between engagement driven by interests versus duty need not be seen as either/or. The very fact that evidence can be mustered to support both interpretations of the founders’ ideals suggests that neither one dominated to the exclusion of the other. Nor should we assume that the essential difference between engagement spurred by a threat to one’s interests and engagement motivated by a sense of civic obligation has faded over time. Indeed, the distinction between interest-driven and duty-driven engagement is at the core of this book, which is about contemporary patterns of public engagement in the United States.


While articulating that engagement in the public square is driven by Tocquevillian as well as Madisonian impulses underscores that there are two fundamentally different motivations for political activity, this observation alone does not provide much theoretical traction for empirical analysis. On its own, “fulfilling one’s duty” remains at best a tautological explanation for political participation. It is far more interesting to ask why Traci felt that voting was her duty.

A potential answer to that question was intuitively included in the newspaper article that told of Traci’s lonely ballot. Traci, the reader learns, had just moved to Boston from Kansas. She was born and raised in the town of Little River, population 693—a community that advertises itself as a “town with a lot of civic pride.”1 Significantly, in 1992 voter turnout in Little River was 67 percent, 12 percentage points higher than turnout nationwide, and 27 points higher than in Boston.2 Armed with this information about Traci’s hometown it seems plausible, even probable, that Traci voted because she hails from a community where voting is common. One might say that you can take the girl out of Kansas, but you can’t take the Kansas out of the girl.

There has long been a strain of research in political science that examines the role of place in understanding political behavior, although with the dominance of survey research in this literature, far more attention has been paid to individuals’ characteristics (King 1997). Loosely grouped together in a literature often labeled “contextual effects,” a small but growing set of studies have examined how the characteristics of the communities in which people live affect their political activity. Decades ago, Paul Lazarsfeld and his Columbia School colleagues took voters’ social environments seriously (Lazarsfeld, Berelson, and Gaudet 1948; Berelson, Lazarsfeld, and McPhee 1954), as have authors like Robert Huckfeldt and John Sprague (1995), Eric Oliver (2001), and James Gimpel, J. Celeste Lay, and Jason Schuknecht (2003) more recently.

The burgeoning literature on social capital has redirected political scientists’ attention to important differences among geographic units, whether they be Italian regions, U.S. states, or nation-states. As Putnam explains in Making Democracy Work, social capital consists of “trust, norms, and networks” (1993, 167) that foster collective action. Putnam demonstrates that Italian states have deep, longstanding historical differences in the extent to which their residents engage in various forms of collective action. In Bowling Alone, he employs a multi-item index of social capital for each of the fifty states, which includes voter turnout (2000).

Viewed through the lens offered by the social capital literature, then, we might be tempted to attribute Traci’s vote to her having lived in a place with a lot of social capital. The problem with this explanation is that it still provides little analytical leverage. Rather than “people vote because they feel they should,” now the explanation is “people vote because they live in a place where voting is the norm.” And how do we know that voting is the norm? Because many people vote there. This might be a small theoretical step forward, but not much more. The challenge is to find an explanation for why those norms are stronger in some places than in others.

The first section of this book deals precisely with just such an explanation. Presaging an argument that will be detailed further in chapters to come, I suggest that people vote out of both Madisonian and Tocquevillian motivations. In some communities more voters come to the polls in order to protect their interests, whereas in other places more of them cast a ballot because they feel it is their duty to do so. In these latter communities, civic norms are strong. Those norms are strong, in turn, because of consensus over values—what political scientists are more likely to call interests, and what economists call preferences. Where many people share the same values social norms are more easily enforced, specifically a norm encouraging civic participation. Diverse interests breed conflict, while uniformity fosters consensus; voter turnout can spring from both, but for different reasons. Furthermore, some forms of engagement in the community are more common in consensual communities, other forms in places better characterized as conflictual.

Assuming that I convince the reader that communities vary in the degree to which they can be characterized by their consensus over values, and thus in the extent to which their residents act in accordance with civic norms, Traci’s vote will nonetheless remain unexplained. Remember that when she cast her ballot, Traci was not actually living in homogeneous Little River, Kansas, but heterogeneous Boston, Massachusetts. Why would Traci vote out of a sense of duty once she had moved away from a place with strong civic norms? Tocqueville hints at the answer when in the quotation above he refers to people acting out of duty because it had become their “instinct,” having acquired a “habit and taste” for it. Traci had internalized the norm that voting is a duty and continued her duty-driven behavior even when she moved to a place where those norms would not be enforced. Critically, she spent her adolescence enmeshed in a community where civic norms were strong. During this particularly formative period of our lives, we are prone to developing habits that stay with us throughout our lives. In short, one can be socialized into acting out of a sense of duty, and an important (probably the most important) period of our lives for that socialization to occur is adolescence.

Again foreshadowing a more thorough discussion later, socialization is an important, if often implied, component of social capital theory. For example, while Putnam does not specifically raise the importance of childhood socialization in his discussion of civic traditions in Italy, it is certainly implied. How else would civic traditions perpetuate across time if the norms that constitute social capital were not transmitted across generations? Likewise, Putnam’s more recent stress on differences in civic engagement among generational cohorts implies that members of different generations have, collectively, undergone distinctive socialization experiences.

Why did Traci vote? Because she was raised in a community where she internalized the norm that voting is her civic duty. The simplicity of this statement, however, belies the complexity of the theoretical foundation upon which it rests. Implicit within it are a number of claims, none of which is necessarily conventional wisdom within political science. To make the case that the communities in which we spend our adolescence affect whether we vote in adulthood first requires establishing that

a. communities shape the civic and political engagement of the people who live within them, or what you do now depends on where you are now
b. the engagement of adolescents in particular is shaped by where they live, or what you did then depends on where you were then
c. adolescents’ engagement links to their engagement as adults, or what you do now depends on what you did then

Together these claims lay the foundation for the book’s central argument: the civic norms within one’s adolescent social environment have an effect on civic participation well beyond adolescence: what you do now depends on where you were then. These points all require theoretical justification and empirical evaluation, and so a section of the book is devoted to each.


Chapters 2–4 demonstrate that the communities in which we live shape the nature of what I will refer to as our public engagement. I use this particular term to describe what other authors generally call, interchangeably, civic or political engagement. I do this because I draw a distinction between engagement which is civic and that which is political. We will get into more formal definitions in chapter 2, but for now an intuitive example provides a sense of what I see as the critical difference between the two. Imagine that someone wanted to help the homeless population in her community. One way would be to volunteer at a soup kitchen that serves the homeless, an example of what I mean by civic, or civically motivated, engagement. Another way would be to volunteer for a candidate who has promised to enact policies to assist the homeless, or to march at the state capitol in support of a bill designed to address their needs. This second type of activity is what I define as engagement with a political motivation. Both types of engagement are public, in the sense that they have an effect beyond oneself and the people in one’s immediate sphere of influence. The fundamental distinction between them is not their ends—in both cases, the intention is to assist the homeless—but rather their means. Political participation has as its immediate objective to affect public policy, while its civic counterpart does not.

In places characterized by conflict, politically motivated public engagement is more common. Conversely, communities where there is relative consensus are more likely to host civic engagement, because these communities have strong civic norms encouraging engagement in publicly spirited activities. This, in a nutshell, is what I have labeled the dual motivations theory of public engagement. Chapter 2 applies the theory to voter turnout, one form of engagement that shares both motivations. Contrary to a strictly Madisonian perspective on politics, voter turnout rises where there is a relative absence of political conflict—where elections are blowouts—because voting has civic, as well as political, underpinnings. Madison, however, is redeemed by the observation that turnout also rises in communities where there is a high degree of political conflict. Chapter 3 then turns its attention to other evidence in support of the dual motivations theory. First, we see that voters are more likely to report voting out of a sense of duty in politically consensual communities, and more likely to vote to accomplish policy objectives in places with a high degree of conflict. Similarly, engagement on the civic side of the spectrum is more likely in consensual than conflictual communities, while politically motivated activity is more common in places known for conflict. Chapter 4 continues to test implications of the dual motivations theory, in this case by changing the focus of the inquiry from large-scale communities like cities and counties to personal social networks. Living in a homogeneous community makes it more likely that someone will have a homogeneous social network, which in turn spurs civically motivated public engagement.


Having established that place matters for adults, the next step is to examine whether it matters for adolescents. Is the collective action of youth affected by their social environment in the same manner as for adults? This is by no means assured, as we might think that adolescents are concerned only with the norms of their immediate peers. Chapter 5 demonstrates, however, that the political complexion of adolescents’ social environments does affect their degree of public engagement. As go the parents, so go the children. Young people in consensual communities are more likely to engage in civically motivated behavior, reflecting the strong civic norms in such places. However, the homogeneity one observes in such environments also has consequences that are normatively troubling, including lower levels of political tolerance among youth.


The reader might be willing to accept that place matters for both adults and adolescents, but might be skeptical that participation among adolescents has an impact on what they do as adults. Speculation that activity in one’s youth affects activity as one ages is common; evidence on the matter is less so. The data demands for such evidence are high, as respondents need to be interviewed and then re-interviewed years later, during and just following what is probably the most disruptive period in the average person’s life. Fortunately, a new archive of data that, to my knowledge, has previously been unexamined by political scientists is now available to explore this subject. These data track a representative sample of high school seniors over a ten-year period, with follow-up surveys administered every two years. With these data, it is possible to test for links between civic participation, namely community service, in one’s youth and both civic and politically tinged public engagement in early adulthood. Chapter 6 details the results of this analysis. In this chapter, we shall see that volunteering in one’s adolescence leads to voting later in life.


To this point, the reader will have seen evidence for separate strands of the explanation for why Traci voted. While each is necessary, however, in seriatim they are not sufficient. Chapter 7 weaves them together, with evidence that one’s social environment in adolescence has far-reaching influence on public engagement in adulthood. Testing such a claim requires data that combine both a longitudinal component—interviewing the same respondents first in adolescence and then as adults—as well as a contextual element—information about the communities in which respondents live. In particular, the analysis focuses on the social environment of what is arguably the greatest importance to adolescents: their schools. Specifically, chapter 7 demonstrates that strong civic norms in an adolescent’s high school lead to a greater likelihood of voting well over a decade following high school.


This book has as its primary objective understanding the motivations underlying voter turnout for the purpose of adding to the theoretical literature on voting. In particular, I hope to make the case that we should not ignore the civic dimension of turnout. When previous research has acknowledged that civic duty is a factor explaining why people vote, it is usually done grudgingly. Somewhat oddly, a factor that almost everyone agrees is important in explaining the vote has been virtually ignored in the political science literature. By shining more light on the simple question of why some people consider voting a duty of citizenship while others do not, we can advance our understanding of why some people turn out to vote while others do not.

The book concludes by proposing how theory might be put into practice. Understanding the forces that shape who votes is an intrinsically important endeavor, given voting’s central role in a democracy. It is especially imperative, however, given that we live in an age of low voter turnout among all Americans and declining turnout among young people especially. In the past, various reforms have been enacted in an attempt to reverse that decline, with varying degrees of success. At worst such reforms have had no effect. At best they have made a modest difference. I suggest that their underwhelming impact is because they have focused on relatively minor impediments to voting, like the rules governing registration. Since theory suggests that a sense of civic duty is a major influence on whether people vote, perhaps efforts to strengthen civic norms will have a greater participatory payoff. To that end, the evidence I have gathered suggests that schools are a lever to enhance young people’s sense of civic commitment. An investment made in enhancing the sense of civic responsibility taught in our schools has the potential to pay big dividends over the long haul in enhancing voter turnout and other forms of civically oriented public engagement. The final chapter, therefore, discusses some ways that this might be accomplished.


A theory is only convincing insofar as data exist to test it. Over the remaining chapters, a wide range of data will be employed. Since the basic research strategy is to see how living in a certain type of community affects individuals, I needed to combine data on individuals with data on the communities in which they live. Constructing these data sets has required me to compile data describing particular communities (usually counties but also metropolitan areas) and then merge them with data describing the individuals who participated in the survey. For example, I know whether an individual is an African American, as well as the percentage of African Americans within that individual’s community. To combine data in this manner requires identifying the geographic location of survey respondents, in some cases for surveys that are decades old, and then the relevant information about their communities. The specific sources of data3 I draw upon throughout the course of the book include (in order of their appearance):

  1. A data set combining information on U.S. counties from multiple sources, including the County and City Data Book, the U.S Census Bureau, and America Votes.
  2. Current Population Surveys in 1989, 1992, 1996, 2000, and 2002 with county and metropolitan area–level data merged with individual-level data.
  3. Citizen Participation Study, again with individual and aggregate data merged.
  4. Social Capital Community Benchmark Survey. Individual-level respondents were matched with the aggregate characteristics of their communities.
  5. The 1987 General Social Survey, combining individual-level and community-level data.
  6. Civic Education Study (U.S. component), conducted under the auspices of the International Association for the Evaluation of Educational Achievement. Individual-level responses were combined with data about each community included in the study.
  7. The 1996 National Household Education Survey (both parent and youth surveys). First, data from parental interviews were merged with data from their children. That file was then combined with county-level data.
  8. Monitoring the Future: A Continuing Study of American Youth. This study draws on longitudinal data collected as part of the Monitoring the Future series.
  9. Youth-Parent Socialization Study. This includes both the longitudinal component of the study, as well as the cohort data collected in the student panel members’ high schools. For both the 1965 and 1982 waves, community-level data were combined with the data on individuals.

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