POLITICAL EFFICACY, CIVIC EDUCATION, AND SOCIAL INEQUALITIES

Although few of us describe what we do as “teaching civics,” a significant part of what we do serves that function. This is most pronounced in undergraduate American government courses, in which significant attention familiarizes students with the theory and practice of America’s political system and which is one of the most common general-education electives taken by undergraduates. American political science has a history of involvement in civic education, going back to John Dewey (Reference Dewey1916). Concern with “civic” values has been a driving concern of scholars including Almond and Verba (Reference Almond and Verba1963), Barber (Reference Barber1984), Dahl (Reference Dahl1998), Putnam (Reference Putnam2001), and Young (Reference Young2002). An “institutional” phase began with the formation of the APSA Task Force on Civic Education (Ostrom Reference Ostrom1996) and the subsequent formation of the APSA Committee on Civic Education and Engagement.

We chose political efficacy rather than other values (e.g., tolerance and egalitarianism) because we believe it is a fundamental component of democracy and because it seems to be the most “content-neutral” component. Political efficacy is “the feeling that individual political action does have, or can have, an impact upon the political process…the feeling that political and social change is possible and that the individual citizen can play a part in bringing about this change” (Campbell, Gurin, and Miller Reference Campbell, Gurin and Miller1954, 187). Scholars also have long distinguished between internal and external dimensions of political efficacy (Balch Reference Balch1974; Converse Reference Converse, Campbell and Converse1972). Whereas internal political efficacy reflects an individual’s belief that he or she can understand politics, external political efficacy reflects that individual’s belief that he or she can influence political decisions.

We believe political efficacy is an important component of civic education. The most recent report by the APSA Committee on Civic Education and Engagement (McCartney, Bennion, and Simpson Reference McCartney, Bennion and Simpson2013) includes several chapters that discuss political efficacy and its relation to civic education or civic learning (see, especially, Van Vechten and Chadha Reference Van Vechten, Chadha, McCartney, Bennion and Simpson2013). However, most of the edited volume focused on deliberate pedagogical strategies, approaches, or other types of “interventions” (e.g., service-learning projects) and assessing their impact on “civic education” benchmarks or outcomes. Moreover, no chapter focused explicitly on the ways in which civic learning might vary by race or gender. The notable exception is the chapter by Owen (Reference Owen, McCartney, Bennion and Simpson2013), which included both race and gender as demographic controls in her study of whether junior and high school civic courses with “active-learning elements” improved students’ engagement with the 2008 presidential campaign.

Many elements of civic education are closely tied to political efficacy. Discussions of civic education often focus on “knowledge acquisition” (Galston Reference Galston2001; Niemi and Junn Reference Niemi and Junn2005), and there is a long-established link between education and political efficacy (Almond and Verba Reference Almond and Verba1963; Campbell, Gurin, and Miller Reference Campbell, Gurin and Miller1954). In the tradition of Dewey (Reference Dewey1916), civic-education proponents argue that increasing “political literacy” improves individuals’ political efficacy, which improves the health of our democracy (Feith Reference Feith2011; Gutmann Reference Gutmann1999; Westeimer and Kahne Reference Westheimer and Kahne2004). Traditionally, civic education includes a broad appreciation of how the political system works and how individuals can participate effectively. Studies consistently find a relationship between education and political efficacy, although that relationship is complex. Studies also have shown that deliberate “civics” instruction improves knowledge acquisition (Niemi and Junn Reference Niemi and Junn2005), as does the discussion of controversial topics (Hess Reference Hess2009). Research on political socialization has long explored the relationship between childhood and early-adult education and individuals’ political attitudes—including political efficacy (Easton and Dennis Reference Easton and Dennis1967; Meyer Reference Meyer1977). Many scholars even argue that developing or improving individuals’ political efficacy should be an explicit goal of civic education (Kahne and Westheimer Reference Kahne and Westheimer2006; Pasek et al. Reference Pasek, Feldman, Romer and Jamieson2008).

However, there is reason to be skeptical about the effects of civic education—particularly regarding deeply entrenched social inequalities such as those related to gender and race. If civic education is partly a means to impart political efficacy—the belief that one can affect political outcomes—then it must confront the realities of institutional sexism and racism embedded in American society. We know that levels of political efficacy tend to be lower among women (Verba, Burns, and Schlozman Reference Verba, Burns and Schlozman1997) and African Americans (Abramson Reference Abramson1972; Rodgers Reference Rodgers1974) and that these gender differences may be a product of socialization (Bennett and Bennett Reference Bennett and Bennett1989). Studies by the National Center for Education Statistics found that students who are poor, whose parents have less education, or are African American or Hispanic do less well on civic–political knowledge tests (National Center for Education Statistics 1999; 2007). Programs that are self-consciously oriented toward civic education must address these inequalities.

There is growing interest in assessing the effectiveness of civic-education programs and pedagogies. McCartney, Bennion, and Simpson (Reference McCartney, Bennion and Simpson2013) largely focused on this endeavor. However, the chapters in that volume—other than Owen (Reference Owen, McCartney, Bennion and Simpson2013) and Van Vechten and Chadha (Reference Van Vechten, Chadha, McCartney, Bennion and Simpson2013)—paid little attention to racial, gender, or other inequalities in civic education. Studies that focus on key social differences such as race and gender offer reasons for skepticism. For example, although service-learning and extracurricular programs comprise an increasing component of civic education (Baldi et al. Reference Baldi, Piere, Skidmore, Greenberg, Hahn and Nelson2001; Kahne and Sporte Reference Kahne and Sporte2008), they may not raise levels of political efficacy (Kahne and Westheimer Reference Kahne and Westheimer2006). Even if they did, students with lower socioeconomic status and ethnic minorities have fewer opportunities to participate in such programs (Conover and Searing Reference Conover, Searing, McDonnell, Michael Timpane and Benjamin2000; Kahne and Middaugh Reference Kahne and Middaugh2008). If civic education is an important component of political science curriculum in the twenty-first century, we must grapple with ways in which socioeconomic inequalities affect how students acquire the benefits of such an education, as well as the possibility that efforts aimed at civic education may have limited impact—or even negative, unintended consequences.

THE SURVEY: DATA AND METHODS

Our goal was to assess the civic-education impact of undergraduate introductory courses at our institution. Rather than develop and evaluate specific civic-education programs, we developed a panel survey of the more than 1,000 students enrolled in three introductory courses offered by the department (i.e., Introduction to AP, Introduction to CP, and Introduction to IR). We distributed our survey in two waves, during the first and last weeks of the Fall 2011 semester at the University of Mississippi, a public flagship university in the South. Although the faculty teaching the relevant courses were aware of our survey’s scope, we neither expected nor encouraged them to alter their syllabus in any way. In addition to asking questions about political efficacy, we asked about students’ background characteristics, political attitudes, and behavior. We also asked about their news-media consumption and social-media usage. Because our survey was not anonymous, we were able to match individual responses across both survey waves. This also meant that we could use students’ individual end-of-semester grades as a control variable. We collected 613 survey responses during the first wave and 444 during the second wave, which represents a 50.3% response rate at start-of-semester and a 39.9% response rate at end-of-semester. We are confident that our sample was representative relative to the total population enrolled in the three courses during that semester and the student population at large (see the appendix for more details).^{Footnote 2}

FINDINGS

Our findings suggest that students did not have different self-reported political efficacy across background characteristics such as race and gender at start-of-semester but that a racial “efficacy gap” developed by end-of-semester. We also found little evidence that completing an introductory political science course significantly impacted students’ political efficacy—even when controlling for individual students’ grades and teacher quality.

Multivariate Analysis of Changes in Self-Reported Political Efficacy

To account for potential interactions between our variables and political efficacy, we used “jackknife” multivariate ordered logit regression.^{Footnote 8} Overall, multivariate analysis confirmed many of our earlier findings but provided additional nuance; however, we note that the models were not very robust.

Table 1 presents regression estimates for start-of-semester internal and external political efficacy. The models for start-of-semester internal political efficacy, which were the most robust, confirmed that it was lower for students enrolled in AP and higher for social science majors, even when controlling for other factors. We found no evidence of a start-of-semester racial efficacy gap, but we found evidence that women had lower start-of-semester internal political efficacy.

Table 1

Jackknife Ordered Logit Estimates of Self-Reported Political Efficacy at Start-of-Semester

Notes: *p<0.05, **p<0.01, ***p< 0.001. Jackknife standard-error estimates in parentheses.

Table 2 presents regression estimates for individual-level changes in political efficacy by end-of-semester. The models confirmed the emergence of an end-of-semester racial efficacy gap, even when controlling for factors such as parents’ level of education, students’ grades, and teacher quality. We found evidence that students reported higher internal political efficacy after completing an AP course. However, these models were not robust, suggesting that other unobserved factors affect how students gain or lose political efficacy during a semester.

Table 2

Jackknife Regression Estimates for Change in Self-Reported Political Efficacy

Notes: *p<0.05, **p<0.01, ***p< 0.001. Jackknife standard-error estimates in parentheses.

IMPLICATIONS FOR TEACHING INTRODUCTORY POLITICAL SCIENCE

Our findings raise concerns about political science and civic education. We found evidence that completing an AP course improved students’ internal political efficacy, regardless of their grades. Students who completed an AP course believed that they understood (American) politics better than they had at start-of-semester. However, we also found evidence that race is a significant factor in how external political efficacy was mediated through civic education. If external political efficacy is a measure of an individual’s belief that he or she can affect social or political change, then a racial efficacy gap is a serious concern. Moreover, we are troubled by the possibility that African American students may be leaving introductory courses with less external political efficacy than when they entered. Are black students (inadvertently or not) being taught that they have even less political power than they thought they had?

In thinking through this issue—and discussing it with colleagues at our institution and beyond—we considered two possible explanations. Some suggested that “institutional racism” is felt on campus (and in the classroom) in ways that negatively affect black students’ political efficacy. This may be a particular issue at our institution, a flagship university in the South, with all the historical baggage that that entails. It also is possible that biases (implicit or otherwise) affect black students’ reactions to faculty. We are not fully convinced by the institutional racism explanation, although we encourage more research on the subject. First, we know the department’s faculty are generally committed to racial equality and focus significant attention (particularly in AP) on discussing civil rights and liberties and social-justice movements.

Second, it seems unlikely that black students first encounter institutional racism on our campus and not previously. Recall that we observed no significant difference in start-of-semester external political efficacy between white and black students. We also were unable to fully test this explanation empirically with our data. That semester, all faculty teaching AP were white males; there were two female instructors (both in CP) and two Hispanic/Latino faculty (also both in CP).^{Footnote 9} Thus, we could not use instructor’s race or gender as a control variable. We also cannot identify individual students’ teaching-evaluation surveys to determine whether students’ race or gender affected instructor evaluations. We found a racial “grade gap” of 0.57 GPA points between black and white students (t=3.94, p<0.001). Again, however, black students’ external political efficacy declined even when controlling for end-of-semester grades.

The second possibility is intriguing but more difficult to test. The majority of faculty of color with whom we discussed our results were unsurprised, suggesting that a declining external political efficacy after completing an introductory course fit their own experience. They suggested that many black students arrive on campus with unrealistically heightened external political efficacy (several suggested that this reflects communities’ “empowerment” efforts) and were confronted with a more pessimistic reality in the classroom—particularly if the university was their first experience in a predominantly white institution. Those colleagues suggested that this was a typical response to completing an introductory course and that later courses reversed it, increasing external political efficacy by focusing on practical strategies rather than basic “civic” knowledge. We cannot test this claim because our data encompass only introductory courses. However, this explanation also concerns us because we know that most students who take introductory courses are not political science majors. If introductory courses “break down” the external political efficacy of black students, what happens to accounting, chemistry, and theater majors? Moreover, this explanation does little to explain the significant increase in external political efficacy of white students.

Ultimately, we have no clear, satisfactory explanation for why we observed a significant efficacy gap between white and black students who completed an introductory political science course at our institution. However, if the racial efficacy gap we observed exists, then it has profound consequences for how we think about civic education in twenty-first-century America. We encourage others to carefully consider these issues and to include race, gender, and other background characteristics in future studies of civic education pedagogies.

SUPPLEMENTARY MATERIAL

To view supplementary material for this article, please visit https://doi.org/10.1017/S1049096518000380