Content-Type: text/html Media, efficacy, and participation Media influences on normative expectations of citizen efficacy and effects on political participation Michael R. McCluskey Doctoral Student Sameer Deshpande Doctoral Candidate Hye Lim Yoo Master's Student Dhavan V. Shah Assistant Professor Doug M. McLeod Professor Mass Communication Research Center School of Journalism and Mass Communication University of Wisconsin-Madison 821 University Avenue 5115 Vilas Hall Madison, WI, 53706 Phone: 608 262-3691 Fax: 608 262 1361 Email: [log in to unmask] The authors also thank participants in the Mass Communication Research Center project, including Prof. Jack M. McLeod, Prof. Zhongdang Pan, Jessica Zubric, Jaeho Cho, Heejo Keum, Cory Armstrong, Michael Boyle, Mark Heather, Hernando Rojas, Michael Schmierbach, Mark Shevy, Seungahn Nah, and So Hyang Yoon, and students of the 2001 Communication and Public Opinion class (J614) for their contributions. Revised version of paper submitted to the Communication Theory and Methodology Division of the Association for Education in Journalism and Mass Communication Annual Conference Miami, August 7 -10, 2002. Abstract Media influences on normative expectations of citizen efficacy and effects on political participation The mass media may encourage a sense that citizens both can and should be able to influence politics and public policy and engage in self-governance, yet this normative value of political efficacy (the "should" evaluation) has been subsumed by a research emphasis on people's evaluation of their actual effectiveness (the "can" evaluation). Using data from an RDD community survey, this research demonstrates (1) that normative efficacy is conceptually different from evaluative efficacy; (2) that mass media use, community integration and demographics significantly predict both types of efficacy; and (3) that each type of efficacy predicts different types of political participation. Implications for future research are discussed. Media influences on normative expectations of citizen efficacy and effects on political participation Democratic societies have long carried the normative expectation that citizens participate in the process of self-governance. Indeed, some democratic theorists assert that an enlightened, engaged citizenry is a cornerstone of democracy. Whether citizens live up to participatory expectations is in part a function of their sense of efficacy, the feeling that they might make a difference and that their efforts have some effect. Normatively speaking, the media should play an active role in encouraging participation and cultivating a sense of efficacy above and beyond their role of informing the public (Gurevitch & Blumler, 1990). But more than communicating a sense that members of the public "can" make a difference, the media may help the public realize that they "should" make a difference, and that the system should be structured so as to permit their input. This paper examines the role of media use and other factors related to citizens' evaluations of how much efficacy they do have (evaluative efficacy), and their expectations about how much influence they should have (normative efficacy). Citizens with large gaps between their assessments of how much influence they should have and how much influence they actually do have are likely to become disenchanted with the existing system, thus reducing the likelihood that they will participate. It is expected that patterns of media use influence individual assessments of both dimensions of political efficacy and indirectly shape political participation through their effects on efficacy. To test these possibilities, we conducted a random digit dialing community survey and then tested the direct and indirect effects of public affairs media use and evaluative and normative efficacy alongside demographic and structural variables in our models predicting various modes of civic participation. Literature Review Efficacy, participation and democracy Notions of participation and citizen efficacy have long been part of democratic philosophy. Citizens in ancient Greek city-states were expected to participate in political decision-making and serve in leadership roles. The philosophers of the day were instrumental in bringing citizen efficacy into the debate over democracy. While Plato was skeptical of the ability of common citizens to govern, Aristotle believed in the combined judgment of groups of individuals (Oates, 1941). For Aristotle, citizens were expected to participate in governance if they wanted to enjoy the rights of citizenship (Lipset, 2001). The philosophy behind this early democracy included the recognition that a key to the success of a democratic system was a sense of citizen efficacy. More recent democratic philosophers adopted the essential element of citizen self rule, placing an emphasis on high levels of citizen competence. As modern democratic systems have evolved, they have retained the ethic of the importance of citizen effectiveness that was developed in the democratic philosophies of John Locke (Cecil, 1984). Writing in the 17th and 18th centuries, he argued for the popular sovereignty of majority decision-making (Cecil, 1984; Locke, 1997; Lipset, 2001; ). An active citizenry was also evident in the writings of Jean-Jacques Rousseau and John Stuart Mill. Rousseau, in the 18th century, proposed conditions for self-rule, in which individuals would be free from control by others and subject only to their own choices and decisions (Lipset, 2001; Rousseau, 1994; Street, 1996), and, further, would play an active rather than a passive role in governance (Rousseau, 1994). Likewise, Mill, in the 19th century, believed in the intelligence of the majority to mak e good decisions (Lipset, 2001). Mill found a high degree of popular involvement in public life as crucial to safeguarding liberty, suggesting an inevitable, almost compulsory involvement in the activities of government (Biagini, 1996). The framers of the American Constitution drew heavily on Locke's theories of law and government (Cecil, 1984), including rights of citizens to have a voice in government to redress grievances and to vote. Thus, our modern democratic system rests on the normative assumption that competent, self-governing citizens are capable and responsible for exercising their will in democratic arenas (Rowley, 1997, p. 4). Political efficacy in social science research Thus, the history and political philosophy of democracy encompass an expectation that citizens will participate and be effective in self-governance (although there is certainly disagreement about how much influence citizens should ideally have), so it makes sense that efficacy has played an important role in political science research. This is true up to a point. Though efficacy has been studied empirically, the focus has been almost exclusively on evaluating how influential citizens believe they are - their self-assessments of evaluative efficacy - leaving out the normative element of how influential citizens believe they should be. Evaluative political efficacy has appeared on National Election Study surveys since the 1950s. Building on the work of Lane (1959), scholars often have separated political efficacy into two dimensions, internal and external. Internal efficacy refers to belief about one's own competence to understand and participate effectively in politics, while external efficacy refers to beliefs about the responsiveness of governmental authorities and institutions to citizen demands (Niemi et. al., 1991). Researchers have typically found that evaluative efficacy matters in voting (Abramson & Aldrich, 1982; Acock et. al., 1985; Finkel, 1985; Stewart et. al. 1992), and in other types of political participation, including contacting officials (Hirlinger, 1992; Sharp, 1982; Verba & Nie, 1972; Vedlitz & Veblem, 1980). While internal and external efficacy have proven to be useful for analysis, the dimensions do not consider how much influence individuals believe they ideally should have to shape political decisions. Considering that education through schools and the mass media include an expectation that citizens in a democracy should have an influence and a role in self-governance, it seems logical for these citizens to form a normative expectation of how much influence they should have. Yet this normative efficacy may differ from individuals' assessment of how efficacious they are. Differences in one's normative expectation of efficacy and evaluative assessment of influence - call it an efficacy gap -- could hold important explanatory power for variance in political participation. It seems likely that normative and evaluative efficacy have different causes and effects. Demographics, community integration and mass media usage all have been shown to influence evaluative efficacy; we might expect these variables to affect normative efficacy as well, but in different ways. Further, evaluative and normative efficacy can be expected to have different influences depending on the type of political participation and level of effort it takes to participate. It seems logical to assume that citizens have higher mean levels of normative efficacy than evaluative efficacy -- that is, individuals have a higher expectation of how influential they should be than assessments of how influential they actually are. The opposite seems unlikely, that citizens would assess they are more efficacious than they should be. These expectations lead to the first hypothesis. H1: Respondents will have higher levels of normative efficacy than evaluative efficacy. This research will emphasize evaluative and normative efficacy in the local community, where citizen input is a more realistic goal for the common person. Participation in social science research Literature on political participation has traditionally focused on voting (e.g., Rosenstone & Hanson, 1993; Teixeira, 1987; Wolfinger & Rosenstone, 1980), but recent communication research has examined other forms of political participation, such as attending town-hall meetings or deliberating over issues (Fishkin, 1995; Gastil & Dillard, 1999; McLeod et. al., 1999a and 1999b). Accordingly, political participation has been described as a continuum. Milbrath (1965) organized political acts from least to most difficult. Individuals performing more difficult acts also tended to perform those that were less difficult. American citizens have been categorized based on the frequency and intensity with which they participated, with Verba and Nie (1972) suggesting categories of "voting specialists," "parochial activists" (citizens who are active only with respect to problems directly affecting their lives), "communalists," and "campaigners." Verba et al. (1995) recategorized those groups into four dimensions: voting, campaign, contact, and community. Neuman (1986) divided political participation into three groups: the highly sophisticated activist group (5% of the population); the moderately sophisticated mass public (75%); and the largely apolitical remaining 20% of the population. Schlozman, Verba and Brady (1995) found that political activity was motivated not by self-interest, but by the desire to affect policy or civic gratification, such as satisfying a sense of duty. Thus, scholars have differentiated political participation by motives and sophistication of the individuals, and types of political participation, separating voters from other types of activists. Political participation may also be classified by the effort it takes. This research suggests that typical predictor variables may have differential influences on various types of participation. Evaluative efficacy has been closely tied to political activity, often shown as strongly associated with increased participation (Chavis & Wandersman, 1990; Milbrath, 1965; Peterson, 1990; Zimmerman & Rapport, 1988). But Verba, Scholzman and Brady (1995) describe this relationship as bidirectional. That is, evaluative efficacy leads to political participation, but such behavior may also enhance participant(s efficacy. Similar circularity of effects has been discussed by Dahl (1961), Eldersveld (1956) and Milbrath (1965). Verba et. al. (1995) showed that evaluative efficacy has some effect on time-based acts of political participation, but small effect on voting and no effect on making political contributions and political discussion. In a similar vein, Milbrath (1965) showed a stronger relationship between evaluative efficacy and spectator activities (voting, proselytizing, and wearing a button) than with gladiator activities (giving money, attending meeting, joining a club, and working in a campaign). Giving money and attending meetings - which are transitional activities and have a higher percentage of persons doing them - showed a higher correlation with efficacy than working in a campaign or joining a political club. These expectations lead to the next hypotheses. H2: Efficacy will have a positive effect on all types of participation. H3a: Evaluative efficacy will play a more significant role in the collective forms of participation. H3b: Normative efficacy will play a more significant role in the individual forms of civic engagement. Predictors of efficacy and participation Both efficacy and participation are expected to be predicted by similar factors - demographics, community integration and media use. Demographics. Much research suggests that evaluative efficacy is positively associated with socioeconomic status (Milbrath, 1965; Verba, & Nie, 1972; Peterson, 1990), education (Milbrath, 1965), and being male (Milbrath, 1965; Peterson, 1990), as highly educated individuals and males have typically been afforded greater access to decision-making power. Similar antecedents may foster normative efficacy, as individuals sharing these characteristics are likely to be socialized to expect more influence. H4: Demographics, including socioeconomic status, education, gender and age, are positively associated with both evaluative and normative efficacy. Similarly, education, income, age, and gender are strong predictors of traditional political activities (McLeod et al., 2000; Milbrath, 1965; Verba & Nie, 1972; Verba, Nie, & Kim, 1978; Verba, Schlozman, & Brady, 1995). In addition, more educated people are more likely to have higher income, and higher income is positively related to political participation (Verba et al., 1995; Milbrath, 1965). Young people and women are less likely to participate in political activities (Verba & Nie, 1972; Milbrath, 1965). The higher the socioeconomic status, with more income and more education, the more likely a person is to develop a sense of evaluative efficacy and consequently to participate in more political activities (Verba & Nie, 1972; Peterson, 1990). H5: Demographics, including socioeconomic status, education, gender and age, are positively associated with civic participation. Community integration. McLeod et al. (2000) suggest that community integration, enhanced by local media and interpersonal discussion, can increase feelings of personal and group evaluative efficacy, ultimately promoting citizen participation. Similarly, Chavis and Wandersman (1990) found a "sense of community" could act as a catalyst, heightening one's perceived personal and group efficacy to influence local action. Maton and Rapport (1984) also found that the development of a sense of community in a religious community was positively associated with psychological empowerment. Since this research focuses on local community issues, community integration would seem to be an important factor in a sense of efficacy. Considering that ties to the community give individuals a greater chance to interact with others, it should also allow them to assess how efficacious they can be. H6: Community integration will significantly predict both evaluative and normative efficacy, and will play a stronger role in evaluative efficacy than in normative efficacy. Community integration, according to McLeod et al. (1999a), can be a necessary condition or prerequisite to local political participation. They suggest that community integration provides the structure and mass media or interpersonal communication provide the knowledge or incentives to participate. Friedland (2001) asserts that community is necessary for democracy to work. Moreover, communication is the basis of the constitution of community. The mass media, especially local newspapers, have long been seen as playing a crucial role in community integration (e.g. Dewey, 1927; Park, 1940/1967). Newspapers connect people through associations, allowing for collective action (Tocqueville, 1835/1969, pp. 517-518), the conduct of group affairs (Lynd & Lynd, 1929), and information acquisition that could lead to collective action (Bryce, 1889; Lemert, 1981). Each of these scholars reflected either primarily or exclusively on local mass media, especially newspapers. The distinctive influence of local media on political participation may be due to their important roles in defining community institutions and facilitating community integration (McLeod et al., 2001; Friedland & McLeod, 1999; McLeod et al.,1996a). Friedland & McLeod (1999) asserted that community integration emerged as a concept that connects local news media with their potential for enlarging democracy at the community level, and that local media can be seen as creating a particular conception of community and interpreting its boundaries and norms. H7: Community integration will significantly predict all types of participation and will have its strongest impact on participation that is closely related to the local community. Media use. McLeod et. al. (1996b) found both newspapers and television to be positively related to internal efficacy, but not significantly related to external efficacy, both types of evaluative efficacy. They also found evaluative efficacy to be an intervening variable between knowledge and participation. Delli Carpini and Keeter (1996) contend that political knowledge promotes civic attitudes and behaviors (see also Verba & Nie, 1972), such as political interest and efficacy, which in turn promote an understanding of why politics is relevant and increases the likelihood of political participation. In particular, local media may be important for normative efficacy by providing information about government affairs that citizens believe that they should be able to shape. This is not to disregard the fact that news media contain mobilizing content that may be critical for evaluative efficacy (Lemert, 1981). However, certain news sources, particularly those that emphasize interpreta tive or critical coverage, may encourage expectations of citizen involvement that outpace their self-assessments of actual political power. H8: Media use will significantly predict both evaluative and normative efficacy. Research on the relationship between mass media use and political participation tends to revolve around citizens' electoral and campaign-related behaviors (e.g., Glaser, 1965; Kennamer, 1987; McLeod et. al. 1979; McLeod et al., 1996b; McLeod & McDonald, 1985; Shaffer, 1981). Research has found that people who read newspapers are more likely to be engaged in civic life than others. Newspaper reading predicts the likelihood of voting (McLeod et al., 1996b; Viswanath et. al., 1990), community political involvement outside of voting (Viswanath et. al., 1990), and participating in public forums (McLeod et. al., 1999a). By contrast, television usage has been linked to political apathy (Bennett, 1986) and a decline in voting turnout (Shaffer, 1981), while watching television news was unrelated to participation in public forums (McLeod et. al., 1999a). Use of media may spur participation by providing information that raises awareness and allows people to learn about the subject. Delli Car pini and Keeter (1996) found that those with the highest levels of political information were the most engaged in campaign activities beyond voting. McLeod et al. (2000) found that local media (especially the local newspaper) had both direct and indirect effects on political activities. They explained that local media use had a strong positive influence on political participation, mediated by interpersonal discussion. It also was positively associated with citizen efficacy. McLeod et al. (2001) note that media effects on civic participation are strong but mainly indirect through knowledge, cognitive complexity, and citizen efficacy. The distinctive influence of local media on political participation can be due to their important roles in defining community institutions and facilitating community integration (McLeod et al., 2001; Friedland & McLeod, 1999; McLeod et al., 1996a). H9: Media use will significantly predict all types of participation, but will play a stronger role in more demanding types of participation. Further, local newspapers will be especially strong predictors of political participation. Methods The Survey This study is an analysis of 657 telephone interviews administered by students of an advanced research methodology class under faculty supervision. The sampling method followed conventional probability sample procedures. A combination of systematic sampling and a variant of random digit dialing were employed to ensure the inclusion of unlisted phone numbers in the sample. The respondents, who are adult residents of Madison, Wisconsin, and its surrounding areas, were randomly selected from each household. The interviews were conducted between October 18 and November 7, 2001, with a response rate of 49.8%. The Variables Two different constructs of efficacy were employed, evaluative efficacy and normative efficacy. Both types of efficacy measures were additive indices of three items each. Evaluative efficacy was derived from three items, amount of influence perceived by the respondent that three kinds of people (the average person, themselves, and people working together) can have (mean inter-item correlation = .61). Similarly, normative efficacy index was based on three items: amount of influence perceived by the respondent that three kinds of people (the average person, themselves, and people working together) should ideally have (mean inter-item correlation = .60). Since we discussed two type of efficacy, normative and evaluative, we felt the difference between the two (labeled as relative efficacy) may be important to consider. We conducted tests on both normative and evaluative as well as relative efficacy (mean inter-item correlation = .51). Civic participation, a construct that serves as the dependent variable in our main analysis, was measured by an additive index of seven behavioral items that asked whether the respondent had participated in the last two years: voted in a local election; worked for a political campaign locally; attended a civic forum or meeting where citizens spoke about local issues; worked for a political campaign; worked on behalf of a social group, cause or organization; signed a petition for a local candidate or issue; contacted a local public official; contributed money to local organizations; donated blood, food, or clothing. All items composing the participation in local politics or community life were measured dichotomously (i.e., yes or no). For complete question wording, see the appendix 1. Factor analysis was conducted on past participation items, revealing two factors that were easily interpretable. Factor 1 included four items: 'worked for a political campaign,' 'participated in a civic forum,' 'contacted local public official,' and 'worked on behalf of social group.' These items share a focus on more active and collective engagement in society, and hence this factor was labeled as 'collective participation.' Factor 2 included following three items: 'contributed money to local organization,' 'signed a petition for local candidate or issue,' and 'voted in a local election.' These items reflect less active and more individualistic participation and hence the second factor was labeled as 'individual participation.' Reliability analysis was conducted for each factor and the mean inter-item correlation was moderate, with values equal to r = 0.31, and r = 0.24 for the two indices, respectively.[1] Established demographic control variables were included in our model ( a total of five variables that indicate the respondent's gender, age, work status, level of education, and household income. Another group of controls concerned the respondent's integration to the community: ownership of a residence, number of years lived in a town, and psychological attachment to the community. Community attachment was measured by an additive index of three items, "I'd be content to live in the Madison area for the rest of my life," "I feel at home in the community I live in," and "I feel close to many people in my neighborhood" (mean inter-item correlation =.41). The third group of controls included media use variables. As for the media variables, the current study employed 'information and surveillance' uses (as opposed to entertainment uses) across three distinct types of media. The measurement of the respondent's level of exposure and attention to newspaper hard news consisted of ten items. Newspaper use was measured by items asking about reading newspaper articles on international affairs, national government and politics, local government and politics, community issues, and urban growth in Dane county (mean inter-item correlation =.61). In the regression model, we utilized one additive variable for newspaper use, which combined all newspaper exposure and attention items. Similarly, attention to television hard news was measured and combined into a single variable. Five items were asked on a ten-point Likert scale to attain the respondent's level of attention to news contents of international affairs, national government and politics, local government and politics, community issues, and urban growth in Dane county (mean inter-item correlation =.58). Internet use was constructed with four categories of on-line activities: e-mail to communicate with people the respondent knows, following news developments over the Internet, searching for information over the Internet, and expressing the respondent's opinion about issues and politics (mean inter-item correlation =.69). The frequency of engaging in each of these activities was measured on a ten-point scale, and then pooled into a single variable. Results Antecedents of evaluative, normative and relative efficacy H1 predicted that normative efficacy would be significantly greater than evaluative efficacy on community issues, and was supported by the results. Normative efficacy (M=21.50) had a higher mean value than evaluative efficacy (M=19.56) that was significantly different when tested by a paired-samples t-test (t -9.4, p<.01). H4, which hypothesized that demographic variables would significantly predict both evaluative and normative efficacy, was not supported (see Table 1). Demographics were not significant predictors of evaluative efficacy (0.7% of variance) or normative efficacy (0.4%). Among the specific variables, just age was a significant predictor, and only for evaluative efficacy. Interestingly, age was not a significant predictor when controlling for basic demographics, but achieved significance once community integration and media variables were entered (Beta= -.03 at initial entry, Beta = -.13 after other predictors entered). The effect of age on evaluative efficacy was negative, indicating that older people felt less efficacious than younger people, contrary to expectations. No demographic variables predicted normative efficacy. Figure 1, a plot of age and the two types of efficacy, indicates that the gap between evaluative and normative efficacy grows with age. These data do not allow us to disentangle life cycle and generational effects, though both merit consideration. [ INSERT TABLE 1 ABOUT HERE ] [ INSERT FIGURE 1 ABOUT HERE ] H6, which predicted that community integration would significantly predict both evaluative efficacy and normative efficacy, was partially supported. Community integration was a significant contributor to evaluative efficacy as expected, explaining 5.4% of the variance, but not to normative efficacy (1.0% of the variance). Also as hypothesized, community integration played a stronger role in evaluative than for normative efficacy. Community attachment was the lone significant variable (Beta = .25 for evaluative and Beta = .09 for normative at initial entry). Neither years in the community nor home ownership were significant predictors. H8, which predicted a relationship between media use and both types of efficacy, was supported by the results. Media use explained 2.2% of the variance in evaluative efficacy and 2.3% in normative efficacy. Within media variables, newspaper (Beta = .18) and television (Beta = .18) were significantly related to evaluative efficacy when basic demographics were controlled, but online communication was not significant. In the final regression equation, television news use (Beta = .10) remained significant and newspaper reading fell to marginal significant (Beta = .09, p<.10). For normative efficacy, all three media variables had significant before-entry betas (newspaper = .14, television = .10, online communication = -.10), but only newspaper (Beta = .12) and online communication (Beta = -.11) remained significant after community integration was entered. When the effect of media use on type of efficacy was plotted, the differences in evaluative and normative efficacy becomes clearer (see Figures 2, 3 and 4). The gap between the evaluative and normative efficacy remains similar across levels of newspaper use, as both steadily increase with greater newspaper use. Conversely, the gap reduces with greater use of television and online communication. With greater television use, evaluative efficacy increases, while normative efficacy remains steady. On the other hand, normative efficacy goes down with greater use of online communication, while evaluative efficacy remains steady. [ INSERT FIGURES 2, 3 AND 4 ABOUT HERE ] Additional analysis found that relative efficacy - the gap between evaluative and normative efficacy - was significantly predicted by community integration (explaining 3.3% of the variance), but not by demographics (1.0%) or media use (0.5%). Gaps between actual and ideal levels of efficacy were shaped by a few specific variables - age (Beta = .14) and community attachment (Beta = -.17) were significant at final entry, and television news use (Beta = .09) at just initial entry. Antecedents of past participation Results supported H2, which predicted that efficacy would have positive effects on both collective and individual types of past participation (see Table 2). Results showed that efficacy significantly explained variance on both collective (R2 = 2.4%) and individual participation (R2 = 0.9%). In support of H3a, evaluative efficacy beta levels (.20 at initial entry, .14 at final entry) were higher than normative efficacy beta levels (.13 at initial entry, not significant at final entry) for collective participation. H3b was also supported, as normative efficacy beta levels (.12 at initial entry and .10 at final entry) were higher than evaluative efficacy beta levels (.11 at initial entry, not significant at final entry) for individual participation. [ INSERT TABLE 2 ABOUT HERE ] H5, testing the effect of demographic variables on collective and individual participation, was generally supported. Demographics were significant for all forms of participation, explaining 15.8% of the variance for collective participation and 19.5% for individual participation. Among specific variables, education was a strong predictor of both collective (Beta = .19 at final entry) and individual participation (Beta = .22 at final entry), and household income was also significant (Beta = .08 for collective and Beta = .10 for individual participation) at final entry. Age was a significant predictor initially for both forms of participation, but fell below significance when community integration, media and efficacy variables were introduced. Gender was a significant predictor of individual participation. H7, dealing with effect of community integration on past participation, was partially supported. Community integration explained 6.2% of the variance for individual participation and 3.0% for collective participation, both highly significant. Community attachment (Beta = .12 and .19), years in town (Beta = .11 on both) and home ownership (Beta = .18 and .24) were significant for both collective and individual participation at initial entry. After media use and efficacy were entered, home ownership remained significant for both types of participation (Beta = .13 collective and .20 individual), community attachment remained significant for individual participation (Beta = .13) and lost significance for collective participation, and years lived in town lost significance on both types of participation. H9, which predicted the effect of media use on past participation, was strongly supported. Media use had significant effects on both types of participation, as predicted. In addition, as hypothesized, media use explained more variance in the more demanding collective participation (5.5%) than individual participation (2.0%). This pattern was the opposite from the contribution of demographics and community integration, which explained more variance for individual participation. Expectations that newspaper use would be an especially strong predictor of participation was also supported, as newspaper use was significant for both collective (Beta = .28 and .24) and individual participation (Beta = .21 and .16) at both points of entry. Television was significant on collective participation just at initial entry (Beta = .11). In an additional analysis, the gap between evaluative and normative efficacy - relative efficacy -failed to account for a significant amount of variance (0.3% on collective and 0.2% on individual participation). Relative efficacy was significant at initial entry (Beta = -.09) for collective participation, suggesting that individuals who see a greater gap between their actual and ideal level of political efficacy are less likely to participate in forms of participation that demand cooperation and collaboration. Discussion Analysis suggests that efficacy is conceptually different when separated by normative and evaluative perceptions, that these types of efficacy affect political participation, and that the media play a distinct role in both efficacy and participation. Predictor variables explained 8.4% of the variance of evaluative efficacy, mostly from community integration (R2 = 5.4%) and media use (R2 = 2.2%). Closer analysis reveals that community attachment and use of newspapers and television were crucial in establishing whether people felt they could affect community affairs, consistent with previous literature. By contrast, predictor variables accounted for just 3.7% of the variance for normative efficacy, with just media use (2.3%) making significant contributions. This shows that community integration makes significant contributions to one's sense of evaluative efficacy and little difference in normative efficacy, while media use contributes nearly equally to both types of efficacy. This suggests that normative efficacy is less rooted in one's community ties and perhaps more oriented in broader socialization processes. Further, the finding that older individuals have lower levels of evaluative efficacy raise the questions of whether aging makes one more realistic about influence - that is, more aware of the complexities of political action and thus less idealistic -- or aging makes one less able or interested in influencing politics. The media measures formed intriguing patterns to explain efficacy. While newspaper use was significant for both evaluative and normative efficacy on both loadings (evaluative efficacy at final loading was marginally significant), different patterns emerged for television and online communication. For newspaper users, this suggests that print's effectiveness in general learning may be important building blocks for forming both normative and evaluative efficacy. Television was significant on final loading for just evaluative efficacy, while online communication was significant (a negative relationship) on just normative efficacy. Television may be a significant contributor to evaluative efficacy by providing the context individuals need to assess how efficacious they are. The negative relationship of online communication to normative efficacy is intriguing, especially since it remained after demographics were controlled. Perhaps those more likely to use online communication are cynical toward traditional political values like how efficacious one should be. On the participation measures, the predictor variables all made significant contributions, but intriguing differences were found between the two types of participation. Media use and efficacy contributed more variance on collective participation - more difficult and demanding types of activities - than on individual participation. Perhaps one needs a greater sense of efficacy and must pay attention to the media to engage in the least common types of participation. By contrast, demographics and community integration accounted for more variance in individual participation than in collective participation. The significance of the demographic measures suggest that individuals' personal characteristics - such as age, income and education - place those individuals in physical or psychological states that make individual participation more accessible or attractive. The same arguments may apply to the significance of the community integration measures, suggesting that a sense of attachme nt to the community may be a necessary condition in many cases to prompt individual participation. Among the media use measures, newspaper use was significant for both types of participation at both points of entry, which supports previous research linking newspaper readership to participation. However, television and online communication showed interesting patterns. Television was significant only at initial entry for collective participation and significance disappeared at later entry, while online communication emerged as marginally significant only at later entry. This suggests that television was significant only by working through other variables, while these variables masked the influence of online communication. Television and online communication were not significant for individual participation. Although not hypothesized, the gap between normative and evaluative efficacy deserves further attention. Analysis shows that this gap - or relative efficacy - is predicted by specific demographic, community integration and media use variables. In addition, relative efficacy was significant at initial entry for collective participation - activities that require a higher level of effort and cooperation than individual participation activities. This gap between one's sense of normative and evaluative efficacy may be linked to political cynicism or decisions not to participate in politics, raising some interesting questions for future research. This analysis indicates that evaluative and normative efficacy measure distinctly different types of a citizen's perceived influence, suggesting that research into political behavior should recognize these differences. Predictor variables, including media use, appear to play distinct roles in one's sense of normative and evaluative efficacy, another intriguing line of research for communication scholars. Additionally, future research should examine whether these differences between normative and evaluative efficacy, found at the local level, also exist at the national level. Further, what effect does this have on participation? Do media (and other predictor variables) play significant roles on normative and evaluative efficacy at a national level? The answers to these questions and the role of media in fostering higher assessments of evaluative and normative efficacy are fruitful areas for future study. References Abramson, P.R. (1983). Political attitudes in America: Formation and Change. San Francisco: W. H. Freeman. Abramson, P.R. & Aldrich, J.H. (1982). The decline of electoral participation in America. The American Political Science, Review, 76, 502-521. Acock, A., Clarke, H.D. & Stewart, M.C. (1985). A new model for old measures: A covariance structure analysis of political efficacy. The Journal of Politics, 47, 1062-1084. Bennett, S. E. (1986). Apathy in America 1960-1984: Causes & consequences of citizen political indifference. Dobbs Ferry, NY: Transnational Publishers. Biagini, E.F. (1996). Liberalism and direct democracy: John Stuart Mill and the model of ancient Athens. In Biagini, E.F. (Ed.) Citizenship and Community: Liberals, Radicals, and Collective Identities in the British Isles, 1865-1931, pp. 21-44. New York: Cambridge University Press. Bryce, J. (1889). The American commonwealth, Vol. II. London and New York: MacMillan and Co. Carmines, E.G. & Zeller, R.A. (1979). Reliability and Validity Assessment. Newbury Park, CA: Sage. Cecil, A.R. (1984). Introduction. In Stevenson, A.E.; Cecil, A.R.; Neuhaus, R.J.; Payton, R.L.; Stanlis, P.J. & Young, R. (Eds.) The Citizen and His Government. Dallas: The University of Texas Press. Chavis, D.M., & Wandersman, A.(1990). Sense of community in the urban environment: a catalyst for participation and community development. American Journal of Community Psychology, 18, 55- 81. Dahl, R.A. (1961). Who governs? Democracy and power in an American city. New Haven: Yale University Press. Delli Carpini, M. X., & Keeter, S. (1996). What Americans know about politics and why it matters. New Haven: Yale University Press. Dewey, J. (1927). The Public and its problems. New York: Holt-Rhinehart & Winston. Eldersveld, S.J. (1956). Experimental propaganda techniques and voting behavior. The American Political Science Review, 50, 154-165. Finkel, S.E. (1985). Reciprocal effects of participation and political efficacy: A panel analysis. American Journal of Political Science, 29, 891-913. Fishkin, J. S. (1995). The voice of the people: Public opinion and democracy. New Haven, CT: Yale University Press. Friedland, L.A. & McLeod, J.M. (1999). Community integration and mass media: A reconsideration. In D.P. Demers & K. Viswanath (Eds.) Mass media, social control, and social change. Ames, IA: Iowa State University Press. Gastil, J., & Dillard, J. P. (1999). Increasing political sophistication through public deliberation. Political Communication, 16, 3-23. Glaser, W. A. (1965). Television and voting turnout. Public Opinion Quarterly, 29, 71-86. Gurevitch, M., & Blumler, J. G. (1990). Political communication systems and democratic values. In J. Lichtenberg (Ed.), Democracy and the mass media (pp. 269-289). Cambridge: Cambridge University Press. Harper Jr., G.M. (1941). Democracy at Athens. In The Greek Political Experience, Studies in Honor of William Kelley Prentice, pp. 36-49. London: Oxford University Press. Hirlinger, M.W. (1992). Citizen-initiated contacting of local government officials: A multivariate explanation. The Journal of Politics, 54, 553-564. Hollis, M. (1996). Machiavelli, Milton and Hobbes on liberty. In Bellamy, R. & Ross, A. (Eds.) A Textual Introduction to Social and Political Theory, pp. 63-90. Manchester, New York: Manchester University Press. Janowitz, M. (1991). The social dimensions of the local community. In J. Burk (Ed.) Morris Janowitz: On social organization and social control. Chicago: University of Chicago Press. Kebbel, G. (1985). Strength of political activity in predicting newspaper use. Newspaper Research Journal, 6, 1-7. Kennamer, J. D. (1987). How media use during campaign affects the intent to vote. Journalism Quarterly, 64, 291-300. Lane, R.E. (1959). Political life: Why and how people get involved in politics. New York: Free Press. Lemert, J. B. (1981). Does mass communication change public opinion after all? A new approach to effects analysis. Chicago: Nelson-Hall. Lipset, S.M. (Ed.) (2001). Political Philosophy: Theories, Thinkers, Concepts. Washington, D.C.: CQ Press. Locke, J. (1997). Political Essays. (M. Goldie, editor.) Cambridge: Cambridge University Press. Lynd, R. & Lynd, H.M. (1929). Middleton. New York: Harcourt, Brace, Jovanovich. Maton, K. I., & Rapport, J. (1984). Empowerment in a religious setting: A multivariate investigation. Prevention in Human Services, 3, 37-72. McLeod, J. M., Bybee, C. R., & Durall, J. A. (1979). Equivalence of informed political participation: The 1976 presidential debates as a source of influence. Communication Research, 6, 463-487. McLeod, J. M., & McDonald, D. G. (1985). Beyond simple exposure: Media orientations and their impact on political processes. Communication Research, 12, 3-33. McLeod, J.M., Daily, C., Guo, Z., Eveland, W.P., Bayer, J., Yang, S., & Wang, H. (1996a). Community integration, local media use and democratic processes. Communication Research, 23, 179-209. McLeod, J. M., Guo, Z., Daily, K., Steele, C. A., Huang, H., Horowitz, E., & Chen, H. (1996b). The impact of traditional and nontraditional forms of mass media in the 1992 presidential election. Journalism & Mass Communication Quarterly, 73, 401-416. McLeod, J. M., Scheufele, D. A., & Moy, P. (1999a). Community, communication, and participation: The role of mass media and interpersonal discussion in local political participation. Political Communication, 16, 315-336. McLeod J.M., Scheufele, D.A., Moy, P., Horowitz, E.M., Holbert, R.L., Zhangn, W., Zubric, S., & Zubric, J. (1999b). Understanding deliberation: The effects of discussion networks on participation in a public forum. Communication Research, 26, 743- 774. McLeod, J. M., Zubric, J., Kwak, N., & Powell, M. (2000). Communication and citizen activity in community problem solving: Media and interpersonal influences on perceptions of institutional and citizen efficacy and voluntary participation. Paper presented at the joint conference of the World Association for Public Opinion Research and the American Association for Public Opinion Research, Portland, Oregon. McLeod, J.M., Zubric, J., Keum, H., Deshpande, S., Cho, J., Satein, S. E., & Heather, M. (2001). Reflection and connecting: Testing a communication mediation model of civic participation. Paper presented to the Communication Theory and Methodology Division at the annual convention of the Association for Education in Journalism and Mass Communication, Washington, D.C. Milbrath, L. W. (1965). Political participation: How and why do people get involved in politics? Chicago, Il: Rand McNally & Company. Neuman, W. R. (1986). The paradox of mass politics: Knowledge and opinion in the American electorate. Cambridge, MA: Harvard University Press. Niemi, R.G., Craig, S.C. & Mattei, F. (1991). Measuring internal political efficacy in the 1988 National Election Study. The American Political Science Review, 85, 1407-1413. Oates, W.J. (1941). The ideal states of Plato and Aristotle. In The Greek Political Experience, Studies in Honor of William Kelley Prentice, pp.187-213. London: Oxford University Press. Ordeshook, P.C. (1997). Lessons for citizens of a new democracy. In Rowley, C.K. (Ed.) Classical Liberalism and Civil Society, Section 10, pp. 1-135. Cheltenham, U.K., Northampton, MA: Edward Elgar. Park, R. (1940/1967). News as a Form of Knowledge. In Robert E. Park: On social control and collective behavior. Chicago: University of Chicago Press. Peterson, S. A. (1990). Political behavior: Patterns in everyday life. Newbury Park, CA: Sage. Rosenstone, S. J., & Hansen, J. M. (1993). Mobilization, participation, and democracy in America. New York: Macmillan. Rousseau, J.J. (1994/1762). The Social Contract (C. Betts, translation). Oxford, New York: Oxford University Press. Rowley, C.K. (1997). On the nature of civil society. In Rowley, C.K. (Ed.) Classical Liberalism and Civil Society, pp. 1-24. Cheltenham, U.K., Northampton, MA: Edward Elgar. Schlozman, K. Lehman, Verba, S. & Brady, H.E. (1995). Participation's not a paradox: The view from American activists. British Journal of Political Science, 25, 1-36. Shaffer, S. D. (1981). A multivariate explanation of decreasing turnout in presidential elections, 1960-1976. American Journal of Political Science, 25, 68-95. Sharp (1982). Citizen demand making in the urban context. American Journal of Political Science, 28, 654-670. Stamm, K.R., & Fortini-Campbell, L. (1983). The relationship of community ties to newspaper use. Journalism Monographs, no.84., 1-27. Street, J. (1996). Rousseau and James Mill on democracy. In Bellamy, R. & Ross, A. (Eds.) A Textual Introduction to Social and Political Theory, pp. 205-228. Manchester, New York: Manchester University Press. Stewart, M.C., Kornberg, A., Clarke, H.D. & Acock, A. (1992). Arenas and attitudes: A note on political efficacy in a federal system. The Journal of Politics, 54, 179-196. Teixeira, R. A. (1987). Why Americans don't vote: Turnout decline in the United States, 1960-1984. Westport, CT: Greenwood Press. Tocqueville, A. (1835/1969). Democracy in America. New York: Harper and Row. Vedlitz & Veblem (1980). Voting and contacting: Two forms of political participation in a suburban community. Urban Affairs Quarterly, 16, 31-48. Verba, S., Burns, N. & Schlozman, K. Lehman (1997). Knowing and caring about politics: Gender and political engagement. The Journal of Politics, 59, 1051-1072. Verba, S., & Nie, N. H. (1972). Participation in America: Political democracy and social equality. New York, NY: Harper & Row. Verba, S., Nie, N.H., & Kim, J. (1978). Participation and political equality : A seven-nation comparison. Cambridge, England: Cambridge University Press. Verba, S., Schlozman, K. L., & Brady, H. E. (1995). Voice and equality: Civic voluntarism in American politics. Cambridge, MA: Harvard University Press. Viswanath, K., Finnegan Jr., J.R., Rooney, B., & Potter, J. (1990). Community ties in a rural Midwest community and use of newspapers and cable TV. Journalism Quarterly, 67, 899- 911. Wolfinger, R. E., & Rosenstone, S. J. (1980). Who votes? New Haven, CT: Yale University Press. Zimmerman, M. A., & Rapport, J. (1988). Citizen participation, perceived control, and psychological empowerment. American Journal of Community Psychology, 16, 725-750. Figure 1: Influence of Age on Efficacy Figure 2: Influence of newspaper use on efficacy Figure 3: Influence of television use on efficacy Figure 4: Influence of online communication use on efficacy Table 1 Antecedents of Evaluative and Normative Efficacy: Hierarchical Regression Analysis of Demographic, Community Integration and Media Use Influences Evaluative Normative Efficacy Efficacy Before Final Before Final N 581 581 Demographics Gender (Female) -.03 -.02 -.03 -.02 Income .03 .01 -.05 -.04 Work Status .02 .03 .03 .03 Age -.03 -.13* .04 -.01 Education .05 .06 .02 .02 Incremental R2 (%) 0.7 0.4 Community Integration Community Attachment .25** .21** .09* .08 Years lived in a town .05 -.03 -.04 -.08 Own or rent house .05 -.00 .01 -.01 Incremental R2 (%) 5.4** 1.0 Media Use Newspaper .18** .09# .14** .12* Television .18** .10* .10* .04 Online communication -.04 -.04 -.10* -.11* Incremental R2 (%) 2.2* 2.3** Final R2 (%) 8.4** 3.7* Notes: (1) # = p< .10 * = p< .05 ** = p< .01 (2) Values are betas from OLS regression. Betas in the left column are those immediately before entry of the block into the equation, controlling for demographic variables block. Betas in the right column are coefficients from the final equation with all variables controlled. (3) Evaluative efficacy = amount of influence average person, self, and people working together) can have. (4) Normative efficacy = amount of influence average person, self, and people working together) should have. Table 2 Antecedents of collective and individual political participation: Hierarchical Regression Analysis of Demographics, Community Integration, Media Use and Efficacy Influences Collective Individual Participation Participation Before Final Before Final N 571 573 Demographics Gender (Female) -.03 -.03 -.08* -.06# Income .19** .08# .21** .10* Work Status .02 .00 .06 .06 Age .17** .05 .23** .05 Education .25** .19** .24** .22** Incremental R2 (%) 15.8** 19.5** Community Integration Community Attachment .12** .00 .19** .13** Years lived in a town .11* .06 .11* .03 Own or rent house .18** .13** .24** .20** Incremental R2 (%) 3.0** 6.2** Media Use Newspaper .28** .24** .21** .16** Television .11** -.02 .07# -.04 Online communication .07 .08# .01 .02 Incremental R2 (%) 5.5** 2.0** Efficacy Evaluative Efficacy .20** .14** .11** -.00 Normative Efficacy .13** .03 .12** .10* Incremental R2 (%) 2.4** 0.9* Final R2 (%) 26.7** 28.6** Notes: (1) # = p< .10 * = p< .05 ** = p< .01 (2) Values are betas from OLS regression. Betas in the left column are those immediately before entry of the block into the equation, controlling for demographic variables block. Betas in the right column are coefficients from the final equation with all variables controlled. (3) Collective participation = 'worked for a political campaign,' 'participated in a civic forum,' 'contacted local public official,' and 'worked on behalf of social group.' (4) Individual participation = 'contributed money to local organization,' 'signed a petition for local candidate or issue,' and 'voted in a local election.' Appendix 1. Survey Question Wording Demographics Gender: Coded by interviewer Income: I would like an estimate of your total 2000 household income. Please estimate the combined income for all household members from all sources. Work status: Could you tell me if you work outside the home or if you are a student? Age: What was your age on your last birthday? Education: How many years of school have you completed (Including elementary school)? Community integration Community ties: I'd be content to live in the Madison area for the rest of my life. I feel "at home" in the community I live in. I feel close to many people in my neighborhood. Years in town: How many years have you been living in the Madison area? Rent/Own: Do you rent or own? Media use Newspaper Use (1) Exposure and (2) attention to a. News about international affairs b. News about national government and politics c. News about local politics d. News about community issues e. News about urban growth Television Use Attention to a. News about international affairs b. News about national government and politics c. News about local government and politics d. News about local politics e. News about community issues f. News about urban growth Online Communication Use: How often are you involved in each of the following Internet activities? a. Use e-mail to communicate with people you know b. Follow news developments over the Internet c. Search for information over the Internet Efficacy Evaluative: Some people think that public and private groups can influence community issues. Where one means LITTLE INFLUENCE and ten means STRONG INFLUENCE, how much influence do you think a. the average person can have? b. what about you, yourself c. people working together Normative: Now, apart from how much influence those people can have, how much influence should they have? Where one means LITTLE INFLUENCE and ten means STRONG INFLUENCE. Ideally, how much influence should: a. the average person can have? b. what about you, yourself c. people working together Participation: In the past two years have you ... a. Voted in a local election? b. Worked for a political campaign locally? c. Attended a civic forum or meeting where citizens spoke about local issues? d. Worked on behalf of a social group, cause or organization? e. Signed a petition for a local candidate or issue? f. Contacted a local public official? g. Contributed money to local organizations? Appendix 2 Descriptive Statistics of Predictor and Criterion Variables VARIABLE Mean S.D. Inter-item correlation DEMOGRAPHICS Gender (female-high) 49.5% female Income Median = $30,000-50,000 Work Status 79.7% work Age 43.41 16.50 Education 15.52 2.75 COMMUNITY INTEGRATION Community Attachment 21.37 6.03 .41 Years in a town 21.54 18.08 House Rent or own? 64.2% own MEDIA USE Newspaper 61.94 23.74 .61 TV 30.91 11.39 .58 Online communication 15.38 10.48 .69 EFFICACY Evaluative 19.56 5.48 .61 Normative 21.50 5.09 .60 Relative 1.92 5.14 .51 PARTICIPATION Collective participation 1.78 1.55 .31 Individual participation 2.36 0.83 .24 [1] Since our factors have four or less items, we report only inter-item correlation to indicate internal consistency of scale (Carmines and Zeller, 1979).