Diffuse Support

As a descriptive term for a body of scholarly work, political socialization seems to have first appeared in 1959 when Herbert Hyman published his seminal book, Political Socialization. Drawing on a substantial number of studies from a variety of disciplines and on a variety of subjects, Hyman provided a theoretical framework uniting diverse works. Hyman's theoretical focus, characteristic of sociologists, was on the process by which social institutions instill political values on the young.

Many political scientists followed Hyman's lead and treated political socialization as one component of social structure. Other scholars, however, turned to various psychological theories. Impressed by the application of Freudian analysis to political life developed by Harold Lasswell (1930, 1948) and Robert Lane (1959, 1962), several scholars adopted this framework to theorize about political socialization. Thus, much of the early research focused on Lasswell's concern with authoritarian and democratic personality structures. Fred Greenstein's concentration on children's attitudes towards political authority typified this Freudian approach to political socialization studies (1960, 1965a, 1965b). Still other researchers incorporated early versions of associative conditioning models (Campbell et al. 1960) and later cognitive theories (Campbell 1962). Whether the theoretical approach was Freudian, neo-Freudian, associative conditioning, or cognitive modeling, political socialization research focused largely on the attitudes and perceptions of children. Childhood was viewed as the key to understanding the adult world. As Easton and Hess (1962) put it, “what is learned early in life is hard to displace.”

Researchers proceeded over the next decade to scrutinize children's perceptions of political abstractions and the agents that formed these perceptions. Studies of children's perceptions of presidents and to a lesser extent policeman (e.g., Easton and Dennis 1965; Greenstein 1965a, 1975; Hess and Torney 1967; Jaros 1967, 1968), their orientations toward political parties (e.g., Greenstein 1965; Merelman 1971), their sense of political efficacy (Easton and Dennis 1967; Hess and Torney 1967) and their sense of nationality (Davies 1968) proliferated in the 1960s and 1970s. So prodigious were students of political socialization that by the end of the 1960s Greenstein declared, “Political socialization is a growth stock” (1970, 969).

Political socialization research may have been a growth stock in 1970 but by 1985 Timothy Cook (1985) was pointing out that the bull market had turned decidedly bearish. The number of publications in political socialization between 1977 and 1982 markedly declined and more importantly the number of articles dealing with those younger than high school seniors all but disappeared from political science journals. As Cook (1985, 1080) noted, “childhood… disappeared in political science.” This rapid decline is often attributed to methodological quarrels centering on the use of survey instruments. Many researchers began to doubt the validity of administering survey instruments to children. Indeed, to some, the responses gleamed from surveys were simply nonattitudes, susceptible to instability and produced by the children's willingness to provide socially correct answers, further contaminated by response set bias (Kolson and Green 1970; Connel 1972; Vaillancourt 1973). But methodological qualms alone cannot explain the fall from favor that political socialization research experienced. Although survey research was the dominant methodology, inventive researchers found other ways to address the questions. Greenstein (1975), for instance, used semi-projective tests and Merelman (1971) used hypothetical dilemmas. For Cook, the decline of political socialization research in the late 1970s is attributable to the failure to take into account psychological models of development. The field's focus on the attitudes of children lead researchers to ignore, sometimes unintentionally, other times quite intentionally (see Greenstein 1970), the fundamental questions of cognition.

Reviving and Revising Interest in Political Socialization

During the 1990s, a renewed interest in political socialization and more generally the roots of civic engagement has lead researchers to examine the cognition of politically relevant facts. At least since the publication of the American Voter (Campbell et al. 1960), students of voting behavior have noted Americans' apparent lack of politically relevant information. In recent years, however, there has been considerable attention directed at measuring Americans' political knowledge. Indeed, it has often taken the character of a race to discover the most appalling lack of knowledge. A 1987 survey conducted by the National Constitution Center, for instance, concluded that 62% of the respondents could not name all three branches of government. (This is perhaps particularly poignant, given that a separate poll two years earlier showed that 59% of Americans could name all three Stooges.) One consistent theme emerging from studies of citizens' political knowledge is that young people are the least cognizant. In 1993, for example, a study of undergraduates at elite colleges demonstrated that the students were often unable to recall the names of U.S. Senators, the line of presidential succession, and the name of the Prime Minister of Great Britain (General Knowledge 1993).

As provocative as these studies sometimes have been, most have suffered from a lack of systematic assessment. Pronouncements of a widespread knowledge deficit were generally based on a small number of questions, often as few as six, of unsubstantiated importance. However, in the latter half of the 1990s, several studies taking a more systematic approach were published. The work of Delli Caprini and Keeter (1996) is particularly important because they draw on over 50 years of survey results and over 2,000 factual questions. Their findings present a mixed verdict regarding political knowledge. Generally, they find political knowledge approximates a normal distribution. A plurality of citizens demonstrates a base of political knowledge with smaller “knowledge rich” and “knowledge poor” groups. The authors conclude, “The American public, while not as politically informed as one might hope, is also not as uniformed as some characterizations have suggested” (1996, 69). Interestingly, their work demonstrates that people exhibit greater knowledge of the institutions and processes of politics than of political leaders, domestic politics, or foreign affairs (1996, 68–69). Of course, information about institutions and processes is more stable than the other categories and requires less consistent monitoring, but it is also the case that institutions and processes are the subject of civic courses in schools. Nevertheless, in none of the four subject area categories did the median response exceed 50% correct.

That the current generation is largely uninformed is an argument made through analysis of the 1998 National Assessment of Educational Process (NAEP). According to the NAEP, three levels of achievement—basic, proficient, andadvanced—identify the expectations of student performance in civics. The 1998 results reported that 39% of twelfth graders scored at the basic level, 22% at the proficient level, and only 4% reached the advanced level. Put another way, 35% of twelfth graders were below even the basic level of civics performance.

The Importance of Political Knowledge

One of the strengths of the Delli Caprini and Keeter (1996) book is its effort to go beyond cataloging what Americans know about politics. Central to their work is the important question: What difference does knowledge make? To many this may seem like an easy question to answer; obviously, knowledge is central to democratic citizenship. As is so often the case, however, this is a contested question.

Rational choice theorists, for instance, explain the political world by comparing elections to economic competition (e.g., Downs 1957; Riker and Ordeshook 1973). Political parties compete like businesses for profits/votes. Citizens are in both cases consumers with capital to expend. The voters weigh their likely benefits from the victory of a particular party or candidate against the costs of voting—including information costs. For many citizens, the costs of acting will outweigh the benefits that they expect to receive from the political process. Indeed, the assumption is that most citizens will find that the political world has little to offer and that their time and effort is better spent outside the political sphere. Of course, as Delli Caprini and Keeter (1996, 52) point out, even this model requires some baseline of political knowledge. Without at least some information, citizens cannot make the rational calculation as to whether participation is or is not in their interests.

Other political scientists have reacted to evidence that citizens possess little political information by suggesting a variety of low information models or what Popkin (1991) refers to as “low information rationality.” Citizens may not have much political information, but they do not need much. Under this view, citizens are information “misers” who use heuristics or informational shortcuts that allow them to make rational decisions with little information (see Page and Shapiro 1992; Sniderman, Brody, and Tetlock 1991; Zaller 1992). Perhaps these theories are reasonable descriptions of political reality, although Delli Caprini and Keeter's finding that citizens are generalists when it comes to political knowledge suggests otherwise. But even low information rationality assumes that citizens have some minimal level of information. Using cues—political parties, candidates, or interest groups—requires some information about the positions that these individuals and organizations take.

On the other hand, political scientists have long concluded that education is strongly related to propensity to vote (e.g., Wolfinger and Rosenstone 1980) and to engage in other forms of participation (e.g., Verba and Nie 1972; Verba, Schlozman, and Brady 1995). More specifically, Delli Caprini and Keeter have demonstrated that it is not just years of education but the amount of political knowledge possessed that predict political participation. Those most knowledgeable politically are most likely to participate in politics. Moreover, diffuse support is clearly linked to political knowledge. As Niemi and Junn point out, “Those who fail to understand the significance of democratic norms often fail to believe in them” (1998, 10). Similarly, as Morin (1996, 6) has shown, the more people know about government and politics, the more faith they express in the American system. Whether one looks at it from the micro perspective, what is best for individual citizens, or the macro perspective, what is best for the political system, political knowledge is central.

The questions then are: How much knowledge is necessary? And, as important: What type of political knowledge is necessary? Unfortunately, there is no easy answer to these questions. In part, the questions are complicated by the obscurity of the meaning of “citizenship.” If we mean only the status of legal citizenship, then apparently the knowledge base is very low and highly focused on reinforcing diffuse support of the regime. The citizenship exam administered by the Immigration and Naturalization Service remains the definitive standard by which we judge an individual's attainment of citizenship in the United States. You pass the exam; you are qualified to participate fully as a U.S. citizen. More than anything else, the exam is a gatekeeper mechanism for justifying the awarding of citizenship; it does not necessarily aim to instill the citizenship goals of civic engagement but focuses more on the objective of imparting some minimal level of political literacy and instituting and reinforcing regime support.

For most, citizenship is a birthright. As Niemi and Junn (1998, 1) have pointed out, there is no “cognitive tax.” There is no political knowledge test that need be passed. Of course, as we've just discussed, immigrants seeking U.S. citizenship do experience a cognitive tax, but even a cursory examination demonstrates that the knowledge required is very basic. If, on the other hand, the vision of citizenship requires only a voting public exercising its franchise rights, the knowledge demands are obviously somewhat higher. Nevertheless, the knowledge requirements of citizen- as-voter are most likely somewhat less than that expected of citizens in what Barber (1984) calls a “strong democracy” or in Dagger's (1997) term the “citizen as republican liberal.” Although there are differences over what Barber and Dagger's versions of citizenship mean, generally it requires active participation and engagement in politics. This kind of political engagement most surely requires a stronger knowledge base than simply the citizen-as-voter.

At this point, perhaps all that we can say about requisite political knowledge is the more the better. As Delli Caprini and Keeter point out, all information is relative and thus, “all other things being equal, more information is better than less information” (1996, 14–15). But the authors also note that information is situational (1996, 14). Knowledge of the intricacies of Congress, for instance, does not necessarily transfer to effective knowledge of local government affairs. What citizens need is not simply more information, but information in context. This creates a problem: political knowledge is a necessary precondition to civic engagement, but information per se is unlikely to be a sufficient precondition to civic engagement.

Is Education the Answer?

If political knowledge is a necessary precondition to civic engagement it follows that, as political thinkers from Jefferson to Dewey have assumed, more and better education is the solution. Furthermore, the long-standing empirical observation that years of formal education are highly correlated with political knowledge seems to support this solution (e.g., Hyman, Wright, and Reed 1975; Delli Caprini and Keeter 1996; Nie, Junn, and Stehlik-Barry 1996). Once again, however, the empirical evidence generates more questions than answers. Most importantly, we do not know the mechanism by which education affects political knowledge. At least until recently, the role of education has constituted a theoretical black box.

Indeed, some scholars have challenged the traditional view by arguing that education is largely a spurious correlate to political knowledge. Robert Luskin (1987, 1990) has been particularly critical of the linkage of education and political knowledge. His analysis argues that after controlling for intelligence, interests, and other individual attributes education has no significant effect on the level of political sophistication regarding political candidates or parties. Eric Smith (1989) makes a similar point when he argues that the selec tion bias inherent in pursuing higher levels of education inflates the correlation between education and political knowledge. According to Smith, those who went on to pursue higher levels of education had more preexisting knowledge of which reports of higher levels of political knowledge are simply a reflection.

Accepting the futility of education may not be necessary, however. Recent research demonstrates that civics instruction can have an impact on political knowledge after all. For instance, early evidence indicates that both the “We the People… Project Citizen” (Atherton 2000) and the “Kids Voting USA” endeavors may defy the conventional wisdom (Chaffee 2000). Most importantly, however, is the recent work of Niemi and Junn (1998). Their analysis of the 1988 NAEP civic assessment data shows that recent civics course work alone raises political knowledge by 4%. When this recent course work is combined with curriculum that covers a wide range of topics and a pedagogical approach that utilizes discussions of current events, the increase in political knowledge reaches 11%, after controlling for gender, ethnicity, home environment, and interest in government. Although the effects of classroom experiences on measures of trust in government are smaller than exhibited for political knowledge, the authors do find a positive effect. Contrary to over 30 years of research on these questions, Niemi and Junn conclude that, “the civics curriculum has an impact of a size and resilience that makes it a significant part of political learning” (1998, 145).

Challenges Ahead and a Note of Caution

All scholarly enterprises suffer from discontinuities. Indeed, the development of knowledge seldom runs in a continuous line. Political socialization is certainly no exception. After a promising start, interest seemingly dissipated only to be revitalized and redirected. While the past four decades have been marked by unevenness in the quantity and quality of political socialization research, both theoretical and empirical, this essay recognizes a number of studies that have contributed significantly to our understanding of political literacy and civic engagement.

Evident in this collective research is a lack of a single window of maturation, a bounded point in time, when individuals “learn about politics” or perceive a link between democratic citizenship and “acceptable” civic engagement. We think that investigating and understanding the developmental links between early childhood and adolescence and the ongoing adult process of political socialization are fundamental to our understanding of political literacy and participation in a nation's population (see, for example, Jennings and Stoker 2001, 18, and Jennings and Niemi 1974). An important key to this understanding is greater awareness of and attention paid by political scientists to developmental psychology and theories of the learning processes. Flanagan and Sherrod remind us that “[d]evelopmental research during the past two decades has emphasized lifelong plasticity and the importance of the sociohistorical contexts in which children grow up” (1998, 447). They argue, and we concur, that it is time to renew research on the political development of young people as they mature into adulthood.

It is also clear to us that for political socialization research to thrive and address the important questions of civic education and engagement for young people, future research needs to integrate the effect of the full range of agents. Political scientists have long been interested in the impact of the family, the education system, the media, and political campaigns on the socialization process. Scholars need to undertake future work regarding the implicit connections among these various agents in the development or lack of development of political awareness.

Finally, we offer a note of caution. Programs that seek to teach and encourage citizenship education and engagement often engender different and sometimes contradictory beliefs regarding what “good” citizenship constitutes and what comprises “acceptable” civic education and civic engagement. Kahn and Westheimer maintain that:

Kahn and Westheimer remind us that the study of youth, political literacy, and civic activism is both a complex undertaking and normatively loaded. For this very reason, the research demands not only rigor in design but the exploration of new venues for a better understanding of those agents, activities, and interactions that shape young peoples' visions of the political world and their choice to participate or not participate in it. In truth, while we can point to a number of excellent studies on civic education and civic engagement over the past four decades, we still know relatively little about what knowledge, both qualitatively and quantitatively, is necessary and desirable for an informed and active citizenry. While one may forcefully argue that the nation continues to flourish as a democratic institution under the present conditions of political literacy and engagement, the limited empirical knowledge that we have of the quality of civic life in our country suggests the need for a clearer, developmental understanding of what we know about politics and how, if at all, we link that knowledge to civic engagement. Only then can we make informed (and normative) decisions about appropriate standards of civic education and civic engagement for our youth and throughout the life-cycle.