Political Knowledge, Political Engagement, and Civic Education PDF

P1: FXY April 17, 2001 9:32 Annual Reviews AR130-09 Annu. Rev. Polit. Sci. 2001. 4:217–34 c 2001 by Annual Reviews. All rights reserved Copyright ! POLITICAL KNOWLEDGE, POLITICAL ENGAGEMENT, AND CIVIC EDUCATION Annu. Rev. Polit. Sci. 2001.4:217-234. Downloaded from www.annualreviews.org by Universit...

P1: FXY April 17, 2001 9:32 Annual Reviews AR130-09 Annu. Rev. Polit. Sci. 2001. 4:217–34 c 2001 by Annual Reviews. All rights reserved Copyright ! POLITICAL KNOWLEDGE, POLITICAL ENGAGEMENT, AND CIVIC EDUCATION Annu. Rev. Polit. Sci. 2001.4:217-234. Downloaded from www.annualreviews.org by University of Michigan - Ann Arbor on 09/20/10. For personal use only. William A. Galston School of Public Affairs, University of Maryland, College Park, Maryland 20742; e-mail: [email protected] Key Words political socialization, participation, democracy, service learning, higher education Abstract After decades of neglect, civic education is back on the agenda of political science in the United States. Despite huge increases in the formal educational attainment of the US population during the past 50 years, levels of political knowledge have barely budged. Today’s college graduates know no more about politics than did high school graduates in 1950. Recent research indicates that levels of political knowledge affect the acceptance of democratic principles, attitudes toward specific issues, and political participation. There is evidence that political participation is in part a positional good and is shaped by relative as well as absolute levels of educational attainment. Contrary to findings from 30 years ago, recent research suggests that traditional classroom-based civic education can significantly raise political knowledge. Service learning—a combination of community-based civic experience and systematic classroom reflection on that experience—is a promising innovation, but program evaluations have yielded mixed results. Longstanding fears that private schools will not shape democratic citizens are not supported by the evidence. INTRODUCTION One of the oldest topics in political theory, civic education is once more on the radar screen of contemporary political science. Compared with previous generations, scholars today are more likely to agree that well-designed institutions are not enough, that a well-ordered polity requires citizens with the appropriate knowledge, skills, and traits of character (Galston 1991:Ch. 10). And it is reasonably clear that good citizens are made, not born. The question is how, by whom, to what end? Since Plato and Aristotle first discussed the matter, it has been clear that civic education is relative to regime type. Democracies require democratic citizens, whose specific knowledge, competences, and character would not be as well suited to nondemocratic politics. There is an additional level of complexity: How we think about the formation of democratic citizens depends on the specific conception of 1094-2939/01/0623-0217$14.00 217 P1: FXY April 17, 2001 Annu. Rev. Polit. Sci. 2001.4:217-234. Downloaded from www.annualreviews.org by University of Michigan - Ann Arbor on 09/20/10. For personal use only. 218 9:32 Annual Reviews AR130-09 GALSTON democracy we embrace (see March & Olsen 2000:148), and this is a matter of considerable debate. What balance is to be struck between representation and direct participation; between self-interest and public spirit; between rights and responsibilities; between liberty and equality; between reasoned deliberation and passionate mobilization; between secular and faith-based foundations of civic discourse and action; between unity and diversity; between civic loyalty and civic dissent? In turn, these theoretical debates have implications for the content and conduct of democratic civic education—the relationship to be established between classroom instruction and community-based experience, for example. Practitioners guided by Barber’s conception of “strong democracy” will measure the performance of civic education along dimensions that include, but go beyond, the skills required of average citizens in representative systems (Barber et al n.d.). Despite these differences, the contours of a rough-and-ready overlapping consensus are now coming into view. This consensus typically replaces either/or choices with both/and propositions. The skills needed to judge the deeds of representatives and to initiate action oneself are both important; civil discourse need not lack passion; the emphasis on the ability to make reasoned public judgment does not give secular reason pride of place over faith; classroom study and community practice both play a role in forming citizens; and so forth. The burgeoning “servicelearning” movement discussed below is one indication of this emerging synthesis. Another key question raises both normative and empirical issues. What degree of civic and political knowledge is required to be a competent democratic citizen? The traditional normative view was that knowledge requirements are high for democratic citizens. The discovery earlier in this century that most Americans have a low level of public knowledge created shock waves among social scientists and sparked a range of revisionist responses. Some argued for elite or expertcentered conceptions of democratic governance; others claimed that even if individual citizens have not mastered the details of public policies and institutions, citizens in the aggregate display well-grounded and stable judgments; still others offered accounts in which citizens with low levels of information are able to use shortcuts, heuristic devices, and cues to make reasonable judgments. More recently, these responses have themselves evoked sharp criticism (e.g. Hoffman 1998, Somin 1998), and many of the revisionists have responded by clarifying their views. Shapiro (1998:524–25), one of the chief proponents of the aggregate rationality thesis states explicitly that the stability of public opinion over time is no guarantee of the quality of those attitudes. Popkin & Dimock (1999), architects of the low-information rationality thesis, show that citizens with low levels of information cannot follow public discussion of issues, are less accepting of the give and take of democratic policy debates, make judgments on the basis of character rather than issues, and are significantly less inclined to participate in politics at all. Here again, there are signs of an emerging consensus. Competent democratic citizens need not be policy experts, but there is a level of basic knowledge below which the ability to make a full range of reasoned civic judgments is impaired. Moreover, a broad-based discussion during the 1990s has yielded substantial agreement P1: FXY April 17, 2001 9:32 Annual Reviews AR130-09 Annu. Rev. Polit. Sci. 2001.4:217-234. Downloaded from www.annualreviews.org by University of Michigan - Ann Arbor on 09/20/10. For personal use only. CIVIC EDUCATION 219 on the content of this knowledge, which in turn has served as the basis for constructing the Civics Assessment of the National Assessment of Educational Progress (Cent. Civic Educ. 1994, Natl. Assess. Gov. Board 1996). A final introductory issue concerns the role of formal civic education in the political socialization of young people. Although citizens are made rather than born, it does not follow that civic education is the key formative mechanism. To begin with, all education is civic education in the sense that individuals’ level of general educational attainment significantly affects their level of political knowledge as well as the quantity and character of their political participation. In addition, noneducational institutions and processes—families, ethnic groups, voluntary associations, and concrete political events, among others—are crucial influences on civic formation. Indeed, the conventional wisdom for the past three decades has been that formal civic education plays an insignificant role in the overall process of civic formation. It is only in the past few years that the pendulum has begun to swing back. The renewed attention to civic education is more than an academic trend. It reflects as well broader concerns about the condition of US civic culture, especially among the young. To be sure, anxiety about the civic engagement of young adults is nothing new, and its persistence is easy to understand. As far back as solid evidence can be found, at any given historical moment, young adults have tended to be less attached to civic life than are their parents and grandparents. It is not difficult to explain this gap. Civic attachment is linked to factors such as professional interests (and self-interests), stable residential location, home ownership, marriage, and parenthood, all of which are statistically less characteristic of younger adults. In every generation, the simple passage of time has brought maturing adults more fully into the circle of civic life. If the only significant differences were cross-sectional, today’s heightened concern would be myopic. But there are also disturbing trends over time. If we compare generations rather than cohorts—that is, if we compare today’s young adults not with today’s older adults but with the young adults of the past—we find evidence of diminished civic attachment. Some of the basic facts are well known. In the early 1970s, about half of the 18–29-year-olds in the United States voted in presidential elections. By 1996, fewer than one third did. The same pattern holds for congressional elections— about one third voted in the 1970s compared with fewer than one fifth in 1998. Less familiar are the trends charted by the annual survey from the University of California, Los Angeles, conducted since the mid-1960s and involving roughly 250,000 matriculating college freshmen each year. Over the more than three decades since the initiation of this survey, every significant indicator of political engagement has fallen by at least half. Only 26% of freshmen think that keeping up with politics is important, down from 58% in 1966. Only 14% say they frequently discuss politics, down from 30%. Acquisition of political knowledge from traditional news sources is way down, and relatively few young people are using the new media to replace newspapers and network TV news as sources of political information (Bennett 1997; Sax et al 1998, 1999; Natl. Assoc. Secr. State 1999; Rahn 1999). P1: FXY April 17, 2001 Annu. Rev. Polit. Sci. 2001.4:217-234. Downloaded from www.annualreviews.org by University of Michigan - Ann Arbor on 09/20/10. For personal use only. 220 9:32 Annual Reviews AR130-09 GALSTON There are signs that these trends have continued unabated throughout the 1990s. For example, a Pew Research Center poll of voters in their late teens and twenties found that fewer than half were thinking “a great deal” about the 2000 election, versus two thirds at the comparable point in 1992. Four in ten said it does not matter who is elected president, twice as high a percentage as in 1992 (Mason & Nelson 2000). It would be wrong to infer that young adults are retreating into pure privatism. Today’s entering freshmen are reporting significantly increased levels of volunteering in their senior year of high school, a trend that seems to be carrying over to their early college years (Sax et al 1998, 1999). But only a fraction of today’s young volunteers believe that they will continue this practice through their college years and into the paid workforce. And even if they did, there is no evidence that it would lead to wider political engagement. On the contrary, most young people characterize their volunteering as an alternative to official politics, which they see as corrupt, ineffective, and unrelated to their deeper ideals. They have confidence in personalized acts with consequences they can see for themselves; they have no confidence in collective acts, especially those undertaken through public institutions whose operations they regard as remote, opaque, and virtually impossible to control (Hart-Teeter 1997, Natl. Assoc. Secr. State 1999, Medill News Serv. 2000). In a recent survey conducted for the Kennedy School’s Institute of Politics, 60% of students polled said they were actively involved in community service, versus only 7% who had been involved, or planned to get involved, in a political campaign (Mason & Nelson 2000). To be sure, the interpretation of these trends is contested. Libertarians may well regard the retreat from the public sphere as healthy. Many determined partisans of civil society welcome volunteering as a substitute for government programs. But from political stances ranging from traditional liberal to compassionate conservative, and from a range of normative/theoretical perspectives as well, the attenuation of political knowledge and engagement is worrisome. Even if one rejects the philosophical proposition that active citizenship is essential to human flourishing, or the civic-republican view that public-spirited action is intrinsically superior to self-regarding pursuits, it is hard to avoid the hypothesis that at some point the withdrawal from public engagement endangers the healthy functioning of democratic polities. At the very least, if the tendency to withdraw is asymmetrically distributed among population groups, then the outputs of the political system are likely to become increasingly unbalanced. And if those who withdraw the most are those who have the least, the system will become even less responsive to their needs. Political engagement is not a sufficient condition for political effectiveness, but it is certainly necessary. The principal purpose of this review, however, is not to expand on these classic themes of normative democratic theory. It is rather to explore the recent scholarship that has renewed interest in the impact of citizen knowledge on the exercise of citizenship and in formal civic education as a component of political socialization. I focus especially on three major academic contributions (Delli Carpini & Keeter P1: FXY April 17, 2001 9:32 Annual Reviews AR130-09 CIVIC EDUCATION 221 1996, Nie et al 1996, Niemi & Junn 1998), as well as evaluations of specific programs and strategies of civic education. I examine what this literature shows about the power and the limits of political knowledge; about civic education as a way of acquiring politically relevant knowledge, skills, and attitudes; and about the most effective ways of organizing and conducting this education. Annu. Rev. Polit. Sci. 2001.4:217-234. Downloaded from www.annualreviews.org by University of Michigan - Ann Arbor on 09/20/10. For personal use only. THE LEVEL AND DISTRIBUTION OF CIVIC KNOWLEDGE The most comprehensive recent study of US citizens’ attainment of civic knowledge is What Americans Know about Politics and Why It Matters (Delli Carpini & Keeter 1996). The authors assemble more than 50 years of survey data drawn principally from the Roper Center, American National Election Studies (ANES), and their own surveys—in sum, more than 2000 factual knowledge questions concerning political institutions and processes, leaders and parties, and public policies. The public’s knowledge of institutions and processes is significantly higher than its knowledge of people and policies, perhaps because the former are more stable over time and require less regular monitoring. Along no dimension does the median score of correct answers top 50% (Delli Carpini & Keeter 1996:68). In the aggregate, political knowledge describes a normal distribution around the median, with a large “middle class” and smaller knowledge-rich and knowledge-poor groups (Delli Carpini & Keeter 1996:153–54). As the authors recognize, these raw statistics do not permit normative conclusions about the their adequacy or inadequacy for informed citizenship; the glass of knowledge can be regarded as half empty or half full (Delli Carpini & Keeter 1996:133). They suggest plausibly that “all things being equal, the more informed people are, the better able they are to perform as citizens” (Delli Carpini & Keeter 1996:219). But how much knowledge is enough? What are reasonable expectations for the majority of citizens? The National Assessment of Educational Progress (NAEP) Civics Assessment helps clarify this issue. Each NAEP subject-matter assessment is divided into four achievement levels: “below basic,” which means little or no demonstrated knowledge of the subject; “basic,” which indicates partial mastery; “proficient,” the level representing a standard of adequate knowledge; and “advanced.” These achievement levels represent absolute thresholds, not percentiles. In principle, every student could reach the level of proficiency. Within this framework, the recently released results of the 1998 Civics Assessment are not encouraging. Thirty five percent of high school seniors tested below basic, indicating near-total civic ignorance. Another 39% were only at the basic level, less than the working knowledge that citizens are deemed to need (Lutkus et al 1999:23). To be sure, the specification of the four achievement levels can be challenged as not grounded in evidence linking them to specific acts and skills of citizens. Still, the 1998 NAEP Civics Assessment is the fruit of nearly a P1: FXY April 17, 2001 Annu. Rev. Polit. Sci. 2001.4:217-234. Downloaded from www.annualreviews.org by University of Michigan - Ann Arbor on 09/20/10. For personal use only. 222 9:32 Annual Reviews AR130-09 GALSTON decade of intellectual spade work and nationwide consensus building (Cent. Civic Educ. 1994, Natl. Assess. Gov. Board 1996). It represents the most plausible judgment we have concerning the knowledge required for civic competence. Moreover, about one quarter of all students meet or exceed the standard of proficiency and, as Delli Carpini & Keeter (1996:219) rightly remark, “the top quartile... is not composed of superhumans.” There is no evidence that overall levels of civic knowledge have altered much over time. A recent study comparing the responses to questions that were asked in both the 1988 and 1998 NAEP Civics Assessment found that percentages of correct answers had hardly changed over the decade between the two assessments. Fourth graders did slightly better, eighth graders did slightly worse, and twelfth graders showed no significant change (Weiss et al 2000). More broadly, Delli Carpini & Keeter (1996) find that overall levels of political knowledge have hardly budged over the past half century. This is a remarkable finding in light of the fact that political knowledge is highly correlated with levels of formal education. For example, an analysis of 1992 ANES data shows that on a seven-point scale of political knowledge, almost 40% of all college graduates were in the top two categories, compared with less than 10% of high school graduates. Conversely, 25% of high school graduates were in the bottom two categories, compared with only 4% for college graduates (Popkin & Dimock 1999:128). Yet the percentage of Americans with college degrees is vastly higher than it was 50 years ago. How can it be that political knowledge has failed to increase? Closer analysis shows that for many (though not all) categories of political knowledge, today’s high school graduates are roughly equivalent to the high school dropouts of the late 1940s, and today’s college graduates are roughly equivalent to the high school graduates of that earlier epoch (Delli Carpini & Keeter 1996: 197–98). Over the past half century, decreased civic achievement at each level of formal education has been counterbalanced by the changed distribution of students among the levels. One interpretation of these data is that school-based civic instruction is less effective than it once was; another is that formative processes outside of school have weakened (Delli Carpini & Keeter 1996:110, 199). Delli Carpini & Keeter (1996) also document large differences in political knowledge between subgroups. Not only education but also race, gender, and self-reported levels of political interest are strongly correlated with all dimensions of political knowledge. The regular use of newspapers and radio is correlated with several dimensions, as is the regular discussion of politics with friends and family. Controlling for other measures of media use, watching TV news is negatively correlated with all types of political information (Delli Carpini & Keeter 1996:144–45). As Delli Carpini & Keeter interpret their multidimensional data, most citizens are political generalists rather than specialists; “people who know a lot about one aspect of national politics also know a lot about others” (Delli Carpini & Keeter 1996:151). [This conclusion has been challenged on methodological grounds (Krosnick 1998:188).] P1: FXY April 17, 2001 9:32 Annual Reviews AR130-09 CIVIC EDUCATION 223 THE SIGNIFICANCE OF CIVIC KNOWLEDGE Annu. Rev. Polit. Sci. 2001.4:217-234. Downloaded from www.annualreviews.org by University of Michigan - Ann Arbor on 09/20/10. For personal use only. Intuitively, it may seem implausible that civic knowledge is central to democratic citizenship. Why does it matter whether young people can identify their senators or name the branches of government? Surprisingly, recent research suggests important links between basic civic information and civic attributes we have reason to care about. The major findings may be summarized as follows: 1. Civic knowledge helps citizens understand their interests as individuals and as members of groups. The more knowledge we have, the better we can understand the impact of public policies on our interests, and the more effectively we can promote our interests in the political process. Delli Carpini & Keeter (1996:238–64) offer a wealth of evidence that political knowledge fosters citizens’ “enlightened self-interest”—the ability to connect personal/group interests with specific public issues and to connect those issues with candidates who are more likely to share their views and promote their interests. Political knowledge, then, is a key determinant of instrumental rationality (see also Zaller 1992). 2. Civic knowledge increases the consistency of views across issues and across time. Utilizing panel surveys from ANES, Delli Carpini & Keeter (1996:232–34) find a strong linear relation between political knowledge and the stability of political attitudes. They also find that more knowledgeable voters display much higher levels of ideological consistency (as measured along a unidimensional liberal-conservative axis) between issues than do the less well informed (Delli Carpini & Keeter 1996:236–38). 3. Unless citizens possess a basic level of civic knowledge—especially concerning political institutions and processes—it is difficult for them to understand political events or to integrate new information into an existing framework. (By analogy, imagine trying to make sense of the flow of events in a sports competition for which one does not know the rules of the game.) Popkin & Dimock (1999) distinguish between “personal character” and “political character” (conduct judged in the specific context of political roles, institutions, issues, and responsibilities). They show that low-information citizens are much more likely to judge officials according to their perception of noncontextual personal character. “Without knowledge of how government works, it is difficult to assess the true priorities of a legislator in the American system.... Voters less able to use these political cues will rely on estimates of personal character instead of attitudes about parties and issues.... [R]eliance on personal character as a proxy for political character is related to uncertainty, and uncertainty is related to a lack of understanding about politics” (Popkin & Dimock 1999:125, 127). P1: FXY April 17, 2001 Annu. Rev. Polit. Sci. 2001.4:217-234. Downloaded from www.annualreviews.org by University of Michigan - Ann Arbor on 09/20/10. For personal use only. 224 9:32 Annual Reviews AR130-09 GALSTON 4. General civic knowledge can alter our views on specific public issues. For example, the more knowledge citizens have about civic matters, the less likely they are to fear new immigrants and their impact on our country (Popkin & Dimock 2000). 5. The more knowledge citizens have of civic affairs, the less likely they are to experience a generalized mistrust of, or alienation from, public life. Ignorance is the father of fear, and knowledge is the mother of trust. One possible explanation for this relationship is the phenomenon of attribution error. More knowledgeable citizens tend to judge the behavior of public officials as they judge their own—in the context of circumstances and incentives, with due regard for innocent oversights and errors as well as sheer chance. By contrast, less knowledgeable citizens are more likely to view public officials’ blunders as signs of bad character (Popkin & Dimock 1999:127–29). Moreover, low-information citizens encountering vigorous political debate with its inevitable charges and countercharges are more likely to conclude that there are no white knights and adopt a “plague on both your houses” stance. For those who understand politics, debate can be as clear as a tennis match; for those who do not, it more closely resembles a food fight (Popkin & Dimock 1999:134). 6. Civic knowledge promotes support for democratic values. For example, the more knowledge citizens have of political principles and institutions, the more likely they are to support core democratic principles, starting with tolerance. Delli Carpini & Keeter (1996:221–24) explore three possible explanations for this linkage and find substantial support for the “social learning” hypothesis that specific knowledge of civil rights and civil liberties increases tolerance for unpopular minorities. Nie et al (1996:71–72) find direct paths from education to both knowledge of democratic principles and tolerance. 7. Civic knowledge promotes political participation. All other things being equal, the more knowledge citizens have, the more likely they are to participate in public matters. For example, the regression analysis of Delli Carpini & Keeter (1996:226–27) shows a highly significant independent effect of political knowledge on the probability of voting. Popkin & Dimock (1996) agree: “The results of our model highlight the strong and independent influence of contextual knowledge on turnout. Controlling for correlated measures of sophistication, knowledge about politics stands out as a consistently strong factor shaping the decision to vote.” Their multivariate analysis leads them to conclude, “The dominant feature of nonvoting in America is lack of knowledge about government; not distrust of government, lack of interest in politics, lack of media exposure to politics, or feelings of inefficacy” (Popkin & Dimock 1999:142). Political knowledge affects participation, not only quantitatively but also qualitatively. Holding socioeconomic status constant, Delli Carpini & Keeter (1996: P1: FXY April 17, 2001 9:32 Annual Reviews AR130-09 Annu. Rev. Polit. Sci. 2001.4:217-234. Downloaded from www.annualreviews.org by University of Michigan - Ann Arbor on 09/20/10. For personal use only. CIVIC EDUCATION 225 259–60) find that more knowledgeable voters are more likely to vote on the basis of national economic conditions than personal economic circumstances. In this important issue area (which may well be generalizable to others), political knowledge makes it more likely that citizens will ask not only “How am I doing?” but also “How are we doing?” Like other analysts, Nie et al (1996) emphasize the link between absolute years of formal education and the development of prodemocratic principles and attitudes, such as tolerance for unpopular groups. But they also try to demonstrate, with some success, that because key dimensions of political engagement are an inherently scarce “positional good,” engagement is linked to relative rather than absolute levels of educational attainment. Education serves as a sorting mechanism; whatever educational attainment may be median at a given time, individuals significantly above the median will tend to be members of social networks that can connect their views more effectively to political leaders and institutions. This matters because the capacity of the political system to pay attention to inputs is inherently limited. (For example, as information technology permits more and more citizens to communicate with their representatives, the impact of each message will decrease.) If the people as a whole are the principal, their elected and appointed agents will always need mechanisms for allocating their time and attention, and position in social networks is one such mechanism. As education levels rise across society, the positive effects of absolute increases in knowledge and understanding are counterbalanced by the negative effects of increased competition for scarce positions of social centrality. These countervailing forces explain the apparent paradox that rising education levels over the past generation have not yielded commensurate—indeed, any—gain in political engagement (Nie et al 1996:Ch. 7, 8). This finding is of particular importance today. Since a generation ago, education has become a more significant sorting mechanism while others have weakened. This is not an argument against the overall civic rationale for public investment in education. In addition to promoting support for democratic principles, education increases verbal cognitive proficiency and related intellectual skills, which improves an individual’s ability to understand political events and act in an instrumentally rational manner (Nie et al 1996:194). Still, if the argument of Nie et al (1996) is correct, some of the traditional expectations for civic education cannot be fulfilled. In particular, there is no reason to predict (or hope) that rising levels of education will translate into increased civic engagement or diminished inequalities of engagement among different groups (Nie et al 1996:190–92). But is their case compelling? Delli Carpini (1997) notes that although Nie et al conceptualize voting as an attribute of both civic enlightenment (affected by absolute education) and civic engagement (affected by relative education), their analysis of voting trends over the past two decades finds that voting is a function of relative rather than absolute education levels. This result is troubling because voting does not “clearly fit the logic of social network centrality, which assumes limited access to those in power and politics as a zero-sum game” (Delli Carpini P1: FXY April 17, 2001 Annu. Rev. Polit. Sci. 2001.4:217-234. Downloaded from www.annualreviews.org by University of Michigan - Ann Arbor on 09/20/10. For personal use only. 226 9:32 Annual Reviews AR130-09 GALSTON 1997:972). Although the content of my vote may cancel out yours, my act of voting does not compete with yours; my ballot does not make it more difficult for your ballot to be counted nor for its effects to be felt by those in power. More broadly, the vision of politics as zero-sum overlooks its more collective, consensual dimensions. And within the competitive arena, there is reason to wonder whether leaders and institutions are always overloaded with voices struggling against one another for scarce attention (Delli Carpini 1997:972). Especially at the local level, public officials complain that citizens often fail to participate in key events—public hearings, town hall meetings, candidate forums, and others—that give them opportunities to hear and be heard directly. If citizens in fact underutilize rather than overwhelm public institutions, then there may be hope that education, properly conducted, could increase political engagement—especially if (as we have seen) the number of years of formal education is a poor predictor of absolute levels of political knowledge and if there is reason to question the measure of political knowledge that Nie et al (1996) employ (for a discussion of these doubts, see Torney-Purta 1997:451, 453). Indeed, a recent study (NH Nie, DS Hillygus, unpublished data) demonstrates a remarkable correlation between verbal ability and political participation among college graduates of the same age, which suggests that the distinction between years of formal education and actual educational attainment is of considerable civic significance. THE ROLE OF CIVIC EDUCATION IN POLITICAL SOCIALIZATION In the end, we do not have a compelling reason to doubt that civic knowledge affects civic competence, character, and conduct. But what affects knowledge? Classroom-Based Civic Education In a generally admiring review, Torney-Purta (1997:447, 453, 456) notes that the books by both Delli Carpini & Keeter (1996) and Nie et al (1996) suffer from a “missing link.” Neither really unpacks the mechanism by which formal education affects political knowledge and understanding. For three decades, the scholarly consensus has been that formal, classroom-based civic education has no significant effect on civic knowledge (Langton & Jennings 1968). Recent findings challenge this consensus and begin to provide insight into both the overall effects of civic education on political knowledge and the specific pedagogical strategies that effectively foster political understanding. Some of these findings reflect evaluations of individual civic education programs. For example, several research studies conclude that “We the People... The Citizen and the Constitution,” a nationwide program of civic education administered by the Center for Civic Education, is especially effective in improving the civic knowledge of elementary, middle, and high school students relative to P1: FXY April 17, 2001 9:32 Annual Reviews AR130-09 Annu. Rev. Polit. Sci. 2001.4:217-234. Downloaded from www.annualreviews.org by University of Michigan - Ann Arbor on 09/20/10. For personal use only. CIVIC EDUCATION 227 students in comparison groups. In addition, participants develop a stronger attachment to democratic attitudes and principles and an enhanced sense of political interest and effectiveness (Leming 1996). Other research is broader based. In a study of political socialization of young people in four communities, Conover & Searing (2000) explore the role of high schools in fostering civic understanding and practice. They focus on four elements of the school experience: the sense of the school as a community; the students’ level of civic engagement in school and extracurricular activities; the level of political discussion in school; and the formal academic curriculum. They find that all four elements significantly affect young people’s civic consciousness and practice, albeit in different ways. Remarkably, the informal civic education that occurs in such non-civics courses as English literature may be more effective than civic education as currently taught (Conover & Searing 2000:111–13). Conover & Searing (2000:108) regard the overall result of these formative processes as inadequate, even disturbing: “While most students identify themselves as citizens, their grasp of what it means to act as citizens is rudimentary and dominated by a focus on rights, thus creating a privately oriented, passive understanding.” In a major study based on data from the 1988 NAEP Civics Assessment, Niemi & Junn (1998) find significant effects from the amount and recency of civic course work, the variety of topics studied, and the frequency with which current events were discussed in class. These course effects are independent of such background variables as gender, ethnicity, and home environment, as well as interest in government and academic aspirations. Classroom effects are smaller for Hispanics than for Caucasian students, and smaller for African-Americans than for Hispanics. (Class discussion is the only classroom variable that yields significant results for African-Americans.) Differences between girls and boys are small, although boys are more strongly affected by their classroom experiences and home background. Although formal classes are significant for all dimensions of civic knowledge, not surprisingly they have somewhat smaller effects in areas, such as citizens’ rights, in which non-school sources are likely to provide relevant information. (Young people’s familiarity with the details of criminal suspects’ Miranda rights is stunningly high.) Niemi & Junn (1998) find far less significant effects of classroom experience on key dimensions of trust in government, and the overall explanatory power of their multivariate model for trust is small (adjusted R2 = 0.05 compared with 0.31 for civic knowledge). Finally, they conduct a parallel analysis, based on NAEP data, of the civic knowledge effects of American history courses, with results they rightly describe as “strikingly parallel” to civics courses (Niemi & Junn 1998:142). Niemi & Junn (1998) offer an explanation for the divergence of their findings from those of scholars a generation ago, best exemplified by the work of Langton & Jennings (1968). First, Langton & Jennings did find some effects of civic education on knowledge, which they downplayed in an analysis heavily weighted toward attitudinal items. Second, Langton & Jennings did not take into account the grade in which students took civic education classes; Niemi & Junn show that twelfth-grade P1: FXY April 17, 2001 Annu. Rev. Polit. Sci. 2001.4:217-234. Downloaded from www.annualreviews.org by University of Michigan - Ann Arbor on 09/20/10. For personal use only. 228 9:32 Annual Reviews AR130-09 GALSTON classes have more impact than those taken earlier. Third, Langton & Jennings did not include discussion of current events in their analysis, and there are good reasons to believe that these discussions are more likely to provide nonduplicative civic knowledge than are other classroom activities. Finally, the Langton & Jennings measure of political knowledge was technically flawed, including items unlikely to be emphasized in standard civics courses as well as common-knowledge items for which the range of variation between students (hence the effects of formal courses) were bound to be limited. To be sure, questions of methodology and interpretation can be raised about the analysis of Niemi & Junn (1998). For example, Torney-Purta (1999:258–59) points out that their model leaves about 70% of the variance in knowledge among students unexplained, and 95% of the variance in trust; that nonclassroom variables such as college attendance plans are more significant; and that possible correlations between individual predictors are left unexplored. In a reanalysis of the Niemi & Junn data set, Greene (2000) challenges their specification and interpretation of two of the three classroom variables. He finds that the only significant timing effect is whether a student was enrolled in a civics course at the time the NAEP test was taken, which raises doubts about whether the civic knowledge gained will persist over time. He suggests that, as implemented, the “range of topics” variable may itself test for civic knowledge, rendering it endogenous to what it seeks to predict. He finds the third classroom variable—class discussion—impervious to methodological objection, however (Greene 2000:696–97). In the end, neither he nor Torney-Purta rejects the broad thrust of Niemi & Junn’s challenge to conventional wisdom on the effects of civic education. Niemi & Junn (1998) emphasize the real-world, not merely statistical, significance of their findings. By itself, civic course work raises overall political knowledge by 4%; when combined with the study of a wide range of topics and regular discussion of current events in the classroom, this figure rises to 11%. [Niemi & Junn’s conclusions are bolstered by NH Nie & DS Hillygus (unpublished chapter), who show that the content of the college curriculum—in particular, the number of social science courses taken—has a statistically significant impact on political participation among college graduates.] Niemi & Junn’s findings suggest the need for improved instruction in such academic areas as basic democratic theory and knowledge of non-American political structures and in such practical skills as the ability to decode simple charts and tables. Finally, although fully recognizing the pedagogical and political obstacles, they recommend a shift away from a national-level emphasis toward local issues, and away from anodine institutional and historical rote work toward discussions of contemporary political controversies (Niemi & Junn 1998:Ch. 7). [The growing evidence against the efficacy of rote and memorization-based civic pedagogy is stressed by Torney-Purta (1997:53–54, 1999:258).] The alternative, although less likely to create community pressures on teachers and school administrators, is more likely to increase cynicism and alienation by painting a picture of conflict-free politics at odds with everyday experience (Frazer 2000:124–25). P1: FXY April 17, 2001 9:32 Annual Reviews AR130-09 CIVIC EDUCATION 229 Annu. Rev. Polit. Sci. 2001.4:217-234. Downloaded from www.annualreviews.org by University of Michigan - Ann Arbor on 09/20/10. For personal use only. Service Learning Thus far, the analysis has focused on traditional classroom-based civic education. But over the past decade, the most rapid growth has occurred in a different form of civic education, called service learning. The National Center for Educational Statistics defines service learning as “curriculum-based community service that integrates classroom instruction with community service activities.” The service must be organized in relation to an academic course or curriculum, must have clear learning objectives, and must address real community needs over a sustained period of time; the learning occurs through both community-based practice and regularly scheduled critical reflection on that practice (Skinner & Chapman 1999:3). As of academic year 1998–1999, 32% of all public schools had incorporated service learning into their curricula, including a remarkable 46% of high schools (versus just 9% of high schools in 1984). Encouraging students to participate more actively in their communities and encouraging them to improve their knowledge of those communities are the most frequently cited goals for service learning (Skinner & Chapman 1999:17). Parallel developments have occurred at the college level, as such organizations as Campus Compact and such scholar-activists as Benjamin Barber, Richard Battistoni, and Harry Boyte have worked to revive the longneglected civic mission of higher education (see Battistoni 2000). The Corporation for National and Community Service has supported roughly 100 service learning programs each year since 1995 (Gray et al 1999). As might be expected, this dramatic expansion has sparked a flurry of program evaluations of widely varying quality. Billig (2000), in the most recent survey of the evaluation literature, remarks, “Research in the field of service-learning has not caught up with the passion that educators feel for it.” She goes on to catalogue the deficiencies of this research. Few of the studies used control groups; few tracked whether short-term impacts were sustained over time; many relied on self reports; few specified theoretical models or tested hypotheses clearly linked to these models (Billig 2000:660). A wealth of evidence supports the proposition that students participating in community-based service activities are far more likely than others to participate and lead later in life (Youniss et al 1997), but relatively few studies are structured to distinguish the effects of youth participation from the effects of preexisting civic behaviors and attitudes. The ones that do tend to find that service learning has an additional independent effect (Giles & Eyler 1998:67). One of the methodologically strongest studies finds that even if key background variables are held constant, patterns of service activity during college have a substantial effect on the amount of service performed five years later. Indeed, after controlling for the amount of service performed during college, the amount performed during high school has an independent effect on the activities of young adults a decade later (Astin et al 1999:195–96). A massive study of more than 22,000 college students by Astin et al (2000) finds that the positive effects of service by individual students are amplified by discussion of service experiences among students. Service learning is especially effective in generating procivic attitudes and activities P1: FXY April 17, 2001 Annu. Rev. Polit. Sci. 2001.4:217-234. Downloaded from www.annualreviews.org by University of Michigan - Ann Arbor on 09/20/10. For personal use only. 230 9:32 Annual Reviews AR130-09 GALSTON because, compared with community service, service learning is “much more likely to generate such student-to-student discussions” (Astin et al 2000:2). Overall, the literature reveals mixed but encouraging results. One study of 369 middle and high schools students from 10 different service-learning programs found negligible effects on the development of a sense of civic responsibility and engagement (Blyth et al 1997:47–49). A rigorous evaluation of 17 middle and high schools with programs supported by the Corporation for National and Community Service found that a year after the end of the initial experience, most of the positive impacts had disappeared (Melchior et al 1999:15). A RAND evaluation of college-level service learning programs noted the strong influence of self selection on student outcomes and found that the increased civic responsibility manifested by participants was restricted to service activities and did not extend to wider political involvement (Gray et al 1999:55). On the other hand, a study of 3450 undergraduates conducted by Astin & Sax (1998:255–56) found a significant impact of service on 12 civic responsibility measures, including the disposition to participate in politics and bring about social change—results that were confirmed and strengthened by students’ responses to seven items pretested when students entered cllege as freshmen. A comprehensive review of service learning evaluations from kindergarten through twelfth-grade students found “inconsistency in virtually all outcome areas.” On balance, however, the evidence suggests that students who participate in high-quality programs that integrate community service with systematic reflection on their experience are more likely to develop an understanding of political context and governing institutions, to think of themselves as politically efficacious, and to become civically and politically engaged (Billig 2000:661, Melchior et al 1999:11). One study finds that as high school students’ voice in the selection and definition of community engagement increases, so does the effectiveness of service learning in improving students’ sense of efficacy, political participation, and tolerance for “out-groups” (W Morgan, M Streb, unpublished paper; see also Hildreth 2000). Another study, in which college students were randomly assigned to either service learning or traditional sections of introductory government courses, found significant effects of service learning in the self-reported importance students attached to “working toward equal opportunity for all U.S. citizens,” “volunteering my time helping people in need,” and “finding a career that provides the opportunity to be helpful to others or useful to society” (Markus et al 1993:413). Studies conducted by the Walt Whitman Center employing an innovative measure of civic skills found that college-level service learning significantly increases civic and political leadership skills (Barber et al 1997, n.d.). A comprehensive evaluation of California programs for kindergarten through twelfth-grade suggests “cautious optimism” about service learning and emphasizes the importance of program quality (Weiler et al 1999:ix). This emphasis is consistent with other studies (Wade & Saxe 1996:343). A range of evidence suggests that service learning is significantly more effective in the late high school years than earlier (see especially Melchior et al 1999:11, 17), a result that parallels findings for traditional classroom-based civic education. P1: FXY April 17, 2001 9:32 Annual Reviews AR130-09 CIVIC EDUCATION 231 Annu. Rev. Polit. Sci. 2001.4:217-234. Downloaded from www.annualreviews.org by University of Michigan - Ann Arbor on 09/20/10. For personal use only. Niemi & Junn (1998:144, 156) suggest the plausible interpretation that by twelfth grade, students have a better general understanding of politics and society as a matrix into which new information and concepts can be integrated and that they are “close enough to formal adulthood that civics lessons have a degree of meaningfulness lacking in earlier years.” A recent survey of college-level service learning reinforces this finding: Courses for this age group can be effective, but only when the service is clearly related to the academic course work and lasts long enough for students to develop a sense of ownership of the project, and when substantial classroom time is devoted to reflection on community-based experiences (Hepburn et al 2000). Public Versus Private Schools as Civic Educators Education for citizenship was one of the major motives for the creation of US public schools, which began a century and a half ago. Ever since, public schools have been regarded as the most appropriate sites for forming citizens, whereas private schools have been regarded with suspicion as sources of separatism, elitism, and antidemocratic principles. Recent research casts doubt on this long-held view, however. Niemi & Junn (1998:84) find only small and inconsistent differences in civic instruction across public and private schools. A study by Campbell (2000) finds that even after correcting for differences of family background among students, private schools were at least as effective as public schools in conveying democratic civic knowledge and principles, with Catholic schools leading the pack. After correcting for a wide range of demographic variables, Wolf et al (2000:21) report, “College students who received most of their prior education in private [elementary and secondary] schools exhibit higher levels of political tolerance than comparable publicly educated students.... [I]t seems unlikely that the private school tolerance advantage is merely a selection effect.” If future research confirms these results, we will be compelled to rethink some long-held beliefs about sources of civic unity in the United States, and to reflect anew on the relationship between the ethos of individual schools and the civic purposes of education. CONCLUSION Niemi & Junn (1998:157) speak of the “near-abandonment” of work on political socialization, and Conover & Searing (2000:91) characterize political socialization as a field in a “state of disarray” (see also Campbell 2000:5). Niemi & Junn trace this situation to the failure of research during the 1960s and 1970s to establish significant links between what young children think about politics and their views as mature adults (see also Owen 2000). But they point out that what is true of children may not apply to older youth (Niemi & Junn 1998:157–58). Journalists and pundits have long observed what they take to be distinct political generations. Now, research-based evidence is emerging that political stances shaped during the midto-late teen years persist throughout adult life (Nie et al 1996:138). Summarizing P1: FXY April 17, 2001 Annu. Rev. Polit. Sci. 2001.4:217-234. Downloaded from www.annualreviews.org by University of Michigan - Ann Arbor on 09/20/10. For personal use only. 232 9:32 Annual Reviews AR130-09 GALSTON the results of a three-wave survey of American twelfth graders begun in 1965 and continued through two decades, Jennings (1996:249) states, “What each cohort brings into political maturity has a good deal of continuity and provides a certain degree of stability in terms of what that cohort is likely to draw on as it moves through the rest of the life cycle.” Putnam (2000) offers evidence depicting a “long civic generation,” born roughly between 1910 and 1940, whose young-adult patterns of exceptional civic interest and engagement have persisted up to the present. By contrast, generations born after 1940 have demonstrated persistent patterns of deepening disengagement (Putnam 2000:Ch. 14). If so, it becomes a matter of more than academic interest to understand better the forces that shape the political outlook of young adults. It is imperative to renew the long-interrupted tradition of research into political socialization. But this time around, unlike a generation ago, researchers cannot afford to overlook the impact of formal civic education and related school-based experiences on the formation of the civic outlook of young adults. Visit the Annual Reviews home page at www.AnnualReviews.org LITERATURE CITED Astin AW, Sax LJ. 1998. How undergraduates are affected by service participation. J. Coll. Stud. Dev. 39:251–63 Astin AW, Sax LJ, Avalos J. 1999. Long-term effects of volunteerism during the undergraduate years. Rev. High. Educ. 22:187–202 Astin AW, Vogelgesang LJ, Ikeda EK, Yee JA. 2000. Executive Summary: How Service Learning Affects Students. Los Angeles: High. Educ. Res. Inst., Univ. Calif. Barber BR, Higgins RR, Smith JK, Ballou J, Dedrick J. n.d. Civic Skills and Service Learning: Research in Democratic Theory and Practice. New Brunswick, NJ: Walt Whitman Cent. Cult. Polit. Democr., Rutgers Univ. Barber BR, Smith JK, Ballou J, Higgins RR, Dedrick J, Downing K. 1997. Measuring Citizenship Project Final Report. New Brunswick, NJ: Walt Whitman Cent. Cult. Polit. Democr., Rutgers Univ. Battistoni RM. 2000. Service learning in political science: an introduction. Polit. Sci. Polit. 33:615–16 Bennett SE. 1997. Why young Americans hate politics, and what we should do about it. Polit. Sci. Polit. 30:47–53 Billig SH. 2000. Research on K-12 schoolbased service-learning: the evidence builds. Phi Delta Kappan 81:658–64 Blyth DA, Saito R, Beikas T. 1997. A quantitative study of the impact of service-larning programs. See Waterman 1997, pp. 39–56 Campbell DE. 2000. Making Democratic Education Work: Schools, Social Capital, and Civic Education. Presented at Conf. Charter Sch., Vouchers, Public Educ., Progr. Educ. Policy Governance, Harvard Univ. Cent. Civic Educ. 1994. National Standards for Civics and Government. Calabasas, CA: Cent. Civic Educ. Conover PJ, Searing DD. 2000. A political socialization perspective. See McDonnell et al 2000, pp. 91–124 Delli Carpini MX. 1997. Review of NH Nie, J Junn, and K Stehlik-Barry, Education and Democratic Citizenship in America. Am. Polit. Sci. Rev. 91:971–72 Delli Carpini MX, Keeter S. 1996. What Americans Know About Politics and Why It Matters. New Haven, CT: Yale Univ. Press. 397 pp. Elkin SK, Soltan KE, eds. 1999. Citizen P1: FXY April 17, 2001 9:32 Annual Reviews AR130-09 Annu. Rev. Polit. Sci. 2001.4:217-234. Downloaded from www.annualreviews.org by University of Michigan - Ann Arbor on 09/20/10. For personal use only. CIVIC EDUCATION Competence and Democratic Institutions. University Park: Penn. State Univ. Press. 424 pp. Frazer E. 2000. Review of RG Niemi and J Junn, Civic Education. Gov. Oppos. 35:122– 25 Galston WA. 1991. Liberal Purposes: Goods, Virtues, and Diversity in the Liberal State. Cambridge, UK: Cambridge Univ. Press. 343 pp. Giles Jr DE, Eyler J. 1998. A service learning research agenda for the next five years. In Academic Service-Learning: A Pedagogy of Action and Reflection, ed. RA Rhoads, JPF Howard, pp. 65–72. San Francisco: JosseyBass Gray MJ, Ondaatje EH, Fricker R, Geschwind S, Goldman CA, et al. 1999. Combining Service and Learning in Higher Education: Evaluation of the Learn and Service America, Higher Education Program. Santa Monica, CA: RAND Educ. Greene JP. 2000. Review of RG Niemi and J Junn, Civic Education. Soc. Sci. Q. 81:696– 97 Hart-Teeter 1997. Public Service and Government Effectiveness: The View of Young Americans. Presented to Partnership for Trust in Government, Washington, DC Hepburn MA, Niemi RG, Chapman C. 2000. Service learning in college political science: queries and commentary. Polit. Sci. Polit. 33:617–22 Hildreth RW. 2000. Theorizing citizenship and evaluating public achievement. Polit. Sci. Polit. 33:627–32 Hoffman T. 1998. Rationality reconceived: the mass electorate and democratic theory. Crit. Rev. 12:459–80 Krosnick JA. 1998. Review of MX Delli Carpini and S Keeter, What Americans Know about Politics and Why It Matters. Ann. Am. Acad. Polit. Soc. Sci. 559:188–89 Langton K, Jennings MK. 1968. Political socialization and the high school civics curriculum in the United States. Am. Polit. Sci. Rev. 62:862–67 233 Leming RS. 1996. We the People... The Citizen and the Constitution. ERIC Digest EDO-SO96–4. Bloomington, IN: ERIC Digest Lutkus AD, Weiss AR, Campbell JR, Mazzeo J, Lazer S, et al. 1999. NAEP 1999: Civics Report Card for the Nation, NCES 2000–457. Washington, DC: Natl. Cent. Educ. Stat. March JG, Olsen JP. 2000. Democracy and schooling: an institutional perspective. See McDonnell et al 2000, pp. 148–73 Markus GB, Howard JPF, King DC. 1993. Integrating community service and classroom instruction enhances learning: results from an experiment. Educ. Eval. Policy Anal. 15:410–19 Mason JL, Nelson M. 2000. Selling students on the elections of 2000. Chron. High. Educ. Sept. 22, p. B16 McDonnell LM, Timpane PM, Benjamin R, eds. 2000. Rediscovering the Democratic Purposes of Education. Lawrence: Univ. Press Kansas. 280 pp. Medill News Serv. 2000. National Survey, yvote 2000. Washington, DC: Youth Vote 2000 Melchior A, Frees J, LaCava L, Kingsley C, Nahas J, et al. 1999. Summary Report: National Evaluation of Learn and Serve America. Waltham, MA: Cent. Hum. Resourc., Brandeis Univ. Natl. Assess. Gov. Board. 1996. Civics Framework for the 1998 National Assessment of Educational Progress. NAEP Civics Consensus Project. Washington, DC: Natl. Assess. Gov. Board Natl. Assoc. Secr. State. 1999. New Millenium Project—Phase I: A Nationwide Study of 15– 24 Year Old Youth. Washington, DC: Natl. Assoc. Secr. State Nie NH, Junn J, Stehlik-Barry K. 1996. Education and Democratic Citizenship in America. Chicago: Univ. Chicago Press. 268 pp. Niemi RG, Junn J. 1998. Civic Education: What Makes Students Learn. New Haven, CT: Yale Univ. Press. 204 pp. Owen D. 2000. Service learning and political socialization. Polit. Sci. Polit. 33:639–40 Popkin SL, Dimock MA. 1999. Political P1: FXY April 17, 2001 Annu. Rev. Polit. Sci. 2001.4:217-234. Downloaded from www.annualreviews.org by University of Michigan - Ann Arbor on 09/20/10. For personal use only. 234 9:32 Annual Reviews AR130-09 GALSTON knowledge and citizen competence. See Elkin & Soltan 1999, pp. 117–46 Popkin SL, Dimock MA. 2000. Knowledge, trust, and international reasoning. In Elements of Reason: Cognition, Choice, and the Bounds of Rationality, ed. A Lupia, M McCubbins, SL Popkin, pp. 214–38. New York: Cambridge Univ. Press. 330 pp. Putnam RD. 2000. Bowling Alone: The Collapse and Revival of American Community. New York: Simon & Schuster. 541 pp. Rahn WM. 1999. Americans’ engagement with and commitment to the political system: a generational portrait. Presented at Congressional Briefing, Washington, DC Sax LJ, Astin AW, Korn WS, Mahoney KM. 1998. The American Freshman: National Norms for Fall 1998. Los Angeles, CA: High. Educ. Res. Inst., UCLA Grad. Sch. Educ. Inform. Stud. Sax LJ, Astin AW, Korn WS, Mahoney KM. 1999. The American Freshman: National Norms for Fall 1999. Los Angeles, CA: High. Educ. Res. Inst., UCLA Grad. Sch. Educ. Inform. Stud. Shapiro RY. 1998. Public opinion, elites, and democracy. Crit. Rev. 12:501–28 Skinner R, Chapman C. 1999. Service-Learning and Community Service in K-12 Public Schools, NCES 1999–043. Washington, DC: Natl. Cent. Educ. Stat. Somin I. 1998. Voter ignorance and democracy. Crit. Rev. 12:413–58 Torney-Purta JV. 1997. Review essay: links and missing links between education, political knowledge, and citizenship. Am. J. Educ. 105:446–57 Torney-Purta JV. 1999. Review of RG Niemi and J Junn, Civic Education. Am. J. Educ. 107:256–60 Wade RC, Saxe DW. 1996. Community servicelearning in the social studies: historical roots, empirical evidence, critical issues. Theory Res. Soc. Educ. 24:331–59 Waterman AS, ed. 1997. Service-Learning: Applications from the Research. Mahwah, NJ: Erlbaum. 184 pp. Weiler D, Lagoy A, Crane E, Rovner A. 1998. An Evaluation of K-12 Service-Learning in California: Phase II Final Report. RPP Int. Sacramento, CA: CalServe Off., Calif. Dep. Educ. Weiss AR, Lutkus AD, Grigg WS, Niemi RG. 2000. The Next Generation of Citizens: NAEP Trends in Civics—1988 and 1998. NCES 2000–494. Washington, DC: Natl. Cent. Educ. Stat. Wolf PJ, Greene JP, Kleitz B, Thalhammer K. 2000. Private schooling and political tolerance: evidence from college students in Texas. Presented at Conf. Chart. Sch., Vouchers, Public Educ., Progr. Educ. Policy Gov., Harvard Univ. Youniss J, McLellan JA, Yates M. 1997. What we know about engendering civic identity. Am. Behav. Sci. 5:620–31 Zaller JR. 1992. The Nature and Origins of Mass Opinion. Cambridge, UK: Cambridge Univ. Press. 367 pp. Annual Review of Political Science Volume 4, 2001 Annu. Rev. Polit. Sci. 2001.4:217-234. Downloaded from www.annualreviews.org by University of Michigan - Ann Arbor on 09/20/10. For personal use only. CONTENTS TRANSNATIONAL POLITICS: Contention and Institutions in International Politics, Sidney Tarrow MECHANISMS IN POLITICAL PROCESSES, Charles Tilly DEMOCRATIZATION AND ECONOMIC REFORM, V. Bunce PSYCHOLOGY AND INTERNATIONAL RELATIONS THEORY, J. M. Goldgeier, P. E. Tetlock POLITICAL TRADITIONS AND POLITICAL CHANGE: The Significance of Postwar Japanese Politics for Political Science, Bradley Richardson, Dennis Patterson RELIGION AND COMPARATIVE POLITICS, Anthony Gill TOWARD A FOURTH GENERATION OF REVOLUTIONARY THEORY, Jack A. Goldstone POLITICAL CONSEQUENCES OF MINORITY GROUP FORMATION, M. Hechter, D. Okamoto POLITICAL KNOWLEDGE, POLITICAL ENGAGEMENT, AND CIVIC EDUCATION, William A. Galston THEORIES OF DELEGATION, J. Bendor, A. Glazer, T. Hammond TIME-SERIES–CROSS-SECTION DATA: What Have We Learned in the Past Few Years?, Nathaniel Beck THE ORIGINS AND WAR PRONENESS OF INTERSTATE RIVALRIES, John Vasquez, Christopher S. Leskiw THE POLITICAL ECONOMY OF INTERNATIONAL MONETARY RELATIONS, J. Lawrence Broz, Jeffry A. Frieden BIOLOGY AND POLITICS: Linking Nature and Nurture, R.D. Masters WOMEN'S MOVEMENTS AT CENTURY'S END: Excavation and Advances in Political Science, Karen Beckwith TAKING STOCK: The Constructivist Research Program in International Relations and Comparative Politics, Martha Finnemore, Kathryn Sikkink VIACRATIC AMERICA: Plessy on Foot v. Brown on Wheels, Douglas W. Rae 1 21 43 67 93 117 139 189 217 235 271 295 317 345 371 391 417