Black Candidates and Black Political Empowerment: An



Empowerment isn’t just for Elected Officials: An Analysis of Obama’s Candidacy on Change in Black Efficacy over the 2008 Election.

Christopher T. Stout

University of California, Irvine

Ph.D. Candidate

Department of Political Science

University of California, Irvine

3151 Social Science Plaza

Irvine, CA 92697

Cstout@uci.edu

Abstract

Very few studies have assessed the role that black candidates have on black political behavior. Fewer still have studied the relationship between black candidates and black political empowerment. To address this shortcoming in the literature, I analyze Barack Obama’s candidacy for president on blacks and whites feelings of internal and external political efficacy over the course of the 2008 election. Using latent growth curve models and the 2008 ANES panel survey; I show that blacks experienced significant growth in both forms of efficacy throughout the 2008 election. Furthermore, blacks’ level of political efficacy grew at a significantly higher rate than whites. The results indicate that black candidates, like black elected officials, can play an integral role in the political incorporation of African Americans.

The 2008 election was a significant moment in African American history. A group that no less than fifty years ago faced widespread social and political disenfranchisement witnessed the election of this country’s first black president. On one hand, the success of Barack Obama represents the vast progress that blacks have made in America since the passage of the 1965 Voting Rights Act. On the other, blacks still lag behind whites in many facets of American politics. For example, blacks are less likely than whites to be interested in politics, feel efficacious, and trust government (Tate, 1993, 2003; Verba et al., 1995; Wolfinger and Rosenstone, 1980). Early research attributed these lower levels of political empowerment with black’s low social standing and lack of economic opportunities (Rosenstone and Hansen, 1993; Wolfinger and Rosenstone, 1980).

During the mid 1980’s scholars began to argue that the observed political disparities between blacks and whites could be attributed to more than socio-demographic differences (Abney and Hutchinson, 1981; Banducci et al, 2004; Bobo and Gilliam, 1990; Gay, 2002; Tate, 2003). These authors find that blacks who are descriptively represented are more likely know their representatives, have positive attitudes about the political system, and are more participatory than their counterparts who are represented by white elected officials. This research shows that black elected officials play an integral role in black political incorporation.

While these studies expand our understanding of the effects of descriptive representation, they have several shortcomings that I address in this article. First, previous studies ignore the role that prominent black leaders and candidates – who don’t hold elected office – play in altering black political attitudes and behaviors. It is not inconceivable that high profile black candidates may have a similar effect on the black electorate as black elected officials. Second, due to lack of data, previous studies have only examined descriptive representation’s effect on political attitudes at one or two points in time. While this offers some understanding of the impact that descriptive representation has on political attitudes, these studies provide virtually no insight into the trajectory of change for this relationship. Finally, even though there is an extensive literature linking descriptive representation to changes in political attitudes, few studies have examined its impact on political efficacy. This is a surprising omission given the importance of efficacy in predicting several political attitudes and behaviors (Abramson and Aldrich, 1982; Dawson, 1994; Tate 1993;Wattenberg, 2002).

To address these shortcomings in the literature, I measure the effect that Obama’s candidacy had on blacks and whites feeling of internal and external efficacy over the course of the 2008 Presidential Election. In order to accomplish this goal, I first define internal and external efficacy and discuss several shortcomings in the current explanations of its development. Second, along the theoretical lines of the descriptive representation literature, I argue that high profile black candidates – who are not incumbents – have the same potential to increase black political efficacy as black elected officials. Furthermore, I argue that during the course of the election blacks will become significantly more efficacious (both internally and externally) than whites.

To test my hypotheses, I use the advance release of the 2008-2009 American National Election Study Panel Survey. This panel data set has a large number of black and white respondents, which allows for the measurement of individual change in levels of political efficacy for both racial groups. Moreover, this data collected individual’s political attitudes at multiple points during the 2008 election which allows researchers to examine how Obama’s bid for president influenced individual change in internal and external efficacy over time. As a result, this data set provides a more complete understanding of descriptive representation’s effect on individuals in ways previous data could not.

Using this data, I estimate and test several latent growth curve models for both black and white respondents. The results indicate that black’s and white’s feelings of efficacy grew throughout the campaign. The growth in black internal and external efficacy, however, was significantly greater than the change in white efficacy. To ensure that this growth in efficacy is a black candidate effect rather than an “Obama Effect”, I also analyze the 1996 Black National Election Study. The results from this analysis indicate that even in states with US Senate and Gubernatorial black candidates who failed in their bid for elected office, blacks were more efficacious than blacks in states without these black candidates. Overall, the results suggest that high profile black candidates can play a vital role in the political empowerment of black voters.

Political Efficacy and Empowerment

Political scientists have identified two distinct types of political efficacy. First, there is internal efficacy which is an individual’s belief that they are competent enough to understand and affect government (Balch 1974; Converse 1972). Those with higher levels of internal efficacy have more confidence in their abilities to influence politics and policies. Second, external efficacy captures individual’s perceptions about government responsiveness to its constituents (Balch 1974; Converse 1972). Individuals with higher levels of external efficacy are more likely to believe that government will be receptive to the demands of the public.

The earliest research on political efficacy focused on the individual. This research concluded that those with more resources in terms of income and education, were the most efficacious (Clarke and Acock, 1989; Hayes and Bean, 1993; Hougland and Christenson, 1983; Soss, 1999).Other studies have hypothesized that political efficacy is built through participation (Finkel 1985, 1987; Pattie and Johnston, 1988). Based on these findings, most of the initial studies blamed blacks’ lower social standing as the culprit for differences in feelings of efficacy (Abramson, 1983; Foster, 1978; Howell and Fagan, 1988).

Following these studies, a wave of research has shown that lack of descriptive representation could also explain several political disparities between blacks and whites (Abney and Hutchinson, 1981; Bobo and Gilliam, 1990; Gay, 2002; Howell, 2000; Lublin, 1997; Tate, 2003). Furthermore, these studies show that descriptive representation is a stronger predictor of political differences between blacks and whites than socio-economic resources (Abney and Hutchinson, 1981; Bobo and Gilliam, 1990; Howell, 2000, Shingles 1981). While these studies give us some understanding of the role that descriptive representation has on black’s feelings of efficacy, there are three major shortcomings in the literature that I address in this study.

First, by only studying black elected officials’ effect on the electorate, descriptive representation scholars ignore the possibility that black political empowerment occurs before the black candidate is elected to office and the sustainability of the effects of descriptive representation after the black elected official leaves office. In other words, imagine that there are three hypothetical pieces to the descriptive representation puzzle; the effects of the black candidacy, the effects of the early and middle stages of the terms of the black elected officials, and the effects after the elected official has left office. Many studies have examined the role that descriptive representation has on political attitudes and behaviors at the early and middle stages of a black elected official’s term (Abney and Hutchinson, 1981; Bobo and Gilliam, 1990; Gay, 2002; Howell, 2000; Lublin, 1997; Tate, 2003). Gilliam and Kaufman (1998) partially addressed the last part of this puzzle, finding that black elected officials over a longer period of time could actually lead to lower levels of black empowerment.

Yet very little research explores the relationship between black candidates and changes in political attitudes. This is a surprising omission for several reasons. First, it is plausible that changes in black attitudes occur during the course of a viable black candidate’s campaign for elected office and remains constant after the candidate is elected. If this is true, previous research which concludes that black elected officials are the impetus for change in political attitudes may just be measuring the residual effect from the African American’s candidacy for elected office. Second, while most black candidates will fail in their bid for high profile elected office, they may still have a positive effect for black politics that should be explored.

Second, the methods used to assess the influence that descriptive representation has on political attitudes are problematic. To make comparisons, most of these studies use cross-sectional methods comparing blacks in areas with black elected officials to blacks who are not descriptively represented (Bobo and Gilliam; 1990, Howell, 2000; Lublin, 1997; Tate, 2003). Other studies only examine elected officials effect on the electorate at two time points; one point before the election and another during the black elected officials term (Abney and Hutchinson, 1981; Gilliam and Kaufman, 1998).

Both methodologies miss the nuances of the effect descriptive representation has on changing black political attitudes over time. Thus, they provide an incomplete understanding of the trajectory of the hypothesized relationship. For example, do the changes in political attitudes grow gradually over time or do they have a punctuated effect early in the election and just remain stable for the remaining years? Although I do not examine change in political opinions over a long period, the use of latent growth curve models and the unique opportunity to examine change in individuals over the course of the 2008 election will allow me to understand the nature of growth that black candidates have on black and white political attitudes.

A third major weakness in this literature is the lack of research on the influence descriptive representation has on political efficacy. This is surprising, given that lower levels of efficacy between blacks and whites could explain many other disparities between the two groups (Abramson and Aldrich, 1982; Dawson, 1994; Tate, 1993, 2003; Wattenberg 2002). Only Tate (2003) examines the role that black descriptive representation has on the electorate using multi-variate statistics and she finds that there is no significant relationship between descriptive representation at the congressional level and higher levels of black political efficacy.

Contrary to Tate’s (2003) findings, I expect that black office seekers can affect black’s feelings of efficacy. By only comparing differences in levels of efficacy between groups, Tate could be missing the growth that descriptive representation has on efficacy for blacks. Therefore, her results may be misleading. Blacks in areas that are descriptively represented may not have significantly higher levels of efficacy than blacks in other areas, but this does not preclude them from experiencing significant amounts of growth in political efficacy. As a result, the presence of a black political leader may have a positive affect on efficacy that is not being properly captured by previous methods.

Hypothesis

All of the same factors that explain why black elected officials influence black political attitudes can also be applicable to viable high profile black candidates. Mansbridge (1999) argues that descriptive representation allows minorities to feel a sense of inclusion into the political system. Furthermore, descriptive representation signals to the black electorate that their representative cares about their interests and will work to enact policies that are beneficial to their community (Abney and Hutchinson, 1981; Gay, 2002).

Black candidates should have a similar impact on black voters, especially viable black candidates who campaign for a high profile elected position[1]. Campbell and Wolbrecht (2008) find that viable women candidates have potential to engage women who are outside of the political arena through inspiring an increased feeling of belonging to the political system. I expect that black candidates will have a similar affect on black voters. Second, while black candidates can’t directly influence policy, studies have shown that high profile black candidates can alter the platforms of other elected officials, candidates, and parties (Dawson 1994; Tate, 1991, 1993). For these reasons, I expect that black voters may feel a heightened sense of efficacy if there is a black candidate on the ballot.

Research also shows that the mediating factor between descriptive representation and changes in political behavior and attitudes can be found in the visibility of the politician (Atkeson 2003). Atkeson’s (2003) findings can partially explain why black mayoral candidates have a larger impact on political behavior than the lesser identifiable black House Representatives. Due to the fact that viable black candidates rarely get nominated for high profile elected office, viable black candidates are very visible to the public. Therefore, I would expect that these black candidates will have a similar positive impact on black attitudes as black elected officials.

I also anticipate that black candidates will have a significant effect on efficacy because of their potential to influence policy. Gilliam and Kaufman (1998) find that the influence of descriptive representation on political attitudes wanes after long periods of times. They argue that blacks are initially engaged when a black politician is elected Mayor, but blacks become disillusioned over time when black politicians fail to make broad reforms that alter the social opportunities for blacks. Gilliam and Kaufman’s (1998) study suggests that the black officeholder’s potential to change the political system is one of the factors that leads descriptive representation to affect the electorate. Viable black candidate should inspire the same excitement as black elected office holders in the early stages of their term because of their potential to legislate change in the political system.

While black representation has positive effects for blacks, most previous descriptive representation research finds that whites living in areas with black elected officials have lower levels of political empowerment (Bobo and Gilliam, 1990; Gay, 2002; Tate, 2003). Most of the studies attribute lower levels of empowerment in white voters to their lack of confidence in black elected officials to represent their interests (Bobo and Gilliam, 1990; Gay, 2001, 2002). I expect that black candidates who are frontrunners will have a similar effect on white voters. While whites may have more of an opportunity to keep blacks from getting elected at this stage, white voters who believe that the election of the black candidate is inevitable may also feel that they have less power to influence the political system. The empirical findings in this section lead to the following two hypotheses.

H1: Barack Obama’s candidacy will lead to a significant increase in African Americans feelings of internal and external efficacy

H2: Black voters will experience significantly more growth in internal and external efficacy than whites over the course of the 2008 election.

Data:

To measure Obama’s effect on blacks and white’s political attitudes, I use the 2008-2009 American National Election Study Panel Survey (ANESPS)[2]. The ANESPS collected a nationally representative sample of blacks and whites and tracked their political opinions monthly from January 2008 to September 2009. On January 31st, 2009, the ANES released an advanced version of this data set which includes data from the primaries (January and February 2008) and the general election (September-November 2008). The data set has a large sample of black and white respondents[3], a low attrition rate, and it captured political attitudes over multiple points in time. These factors allow for the examination of individual change in blacks and whites feeling of efficacy over the course of the 2008 election.

My dependent variables are measures for internal and external efficacy. To measure internal political efficacy I use a question which asks “How much can people like you affect what the government does?” For external efficacy, I use the following question “How much do government officials care what people like you think?” For both questions, the respondents were given a five point scale that ranged from not interested at all (1) to extremely interested (5).

I use several independent variables to isolate the effect that Obama’s candidacy has on black’s and white’s feelings of efficacy over the course of the 2008 election. I control for socio-demographic factors such as age, gender, marital status, and employment status. For marital status, I compared respondents who were married to those who were not. For employment status, I considered whether the respondent was employed. I also include socio-economic variables such as income and education. It is well established in the political behavior and attitudinal literature that that certain socio-demographic and socio-economic variables influence individual’s levels of efficacy (Verba, Schlozman, and Brady 1995, 1997). Generally, white married males with more income and education are more likely to feel efficacious (Verba, Schlozman, and Brady 1995, 1997).

In addition to the more traditional controls, I also account for the ideology of the respondent[4]. I expect that liberal voters will feel more efficacious in the 2008 election because Democrats were predicted to perform well at all levels of government. By controlling for this, I can ensure that the results of my analysis are not driven by ideology. This takes on added significance because blacks tend to be overwhelmingly liberal democrats (Barker et al. 1999, Tate 1993). I also control for several participation variables, including whether the respondent was registered to vote in the 2008 election and if they had voted in the 2004 election. Researchers have argued that political participation is associated with higher levels of efficacy (Finkel 1985).

Finally, I control for the candidate choice of the respondent. I create dichotomous dummy variables for respondents who supported Clinton in the primaries and a separate variable for respondents who supported Obama in the primaries[5]. I also create a third candidate dummy variable which encompasses respondents who supported McCain in the general election. One could imagine that Clinton and McCain voters may have felt deflated because of poor poll performance by their preferred candidate and this could lead to lower levels of both internal and external efficacy.

Methodology

In order to measure change in political attitudes over the course of the 2008 election, I use latent growth curve models. Latent Growth Curve Models (LGCM) are forms of structural equation models and allow for the estimation of both the intercepts and slopes of individual’s political efficacy over the course of the 2008 election. Therefore this method is suitable for this project, in which I am interested in estimating growth in political efficacy over time. Furthermore, this method accounts for the possibility that individuals have different initial levels of political efficacy and LGCM measures whether growth is associated with an individual’s starting position. Using this method, we gain a more nuanced understanding of the variables that influence growth in political efficacy.

For both the internal and external efficacy models, I constrain the intercepts to be 1 at each time point and across cases because this is the starting point in our data set. Second, for the internal efficacy model, I constrain the slopes so that the estimates grow linearly over time. These estimates correspond with the month in which the levels of political efficacy were recorded. For external efficacy, the best fitting trajectory was a latent growth curve[6]. As a result, I allow the estimates in between January 2008 and November 2009 to be estimated. For both models, I also constrain the means of the dependent variable to be 0 and I assume that all error terms in my model are equal. Both are commonly used techniques for LGCM. The model is presented graphically in figure 1.

[Insert Figure 1 about Here]

To measure internal and external efficacy over time, I run two separate sets of LGC Models. For each set, I estimate a linear LCGM predicting internal efficacy, and a latent curve LGCM predicting external political efficacy. In the first set, I estimate change in internal and external political efficacy for blacks and whites without any controls. This will allow me to test whether either racial group experienced a significant growth in political efficacy over the course of the campaign. In the second set of models, I test whether the trends (slopes) for black and white voters are significantly different controlling for several variables that may be associated with higher levels of political efficacy (Socio-Demographic, SES, political orientation, and candidate choice variables).

Results

[Insert Table 1 about Here]

Table 1 displays the change in black’s and white’s feelings of internal and external efficacy over five periods of time (Jan. 2008, Feb. 2008, Sep. 2008, Oct. 2008, Nov. 2008). In January of 2008, blacks have higher levels of internal political efficacy than whites. Both racial groups experienced growth in internal efficacy over the course of the election, but this growth seems to decrease during certain months. Both blacks and white decreased in internal efficacy during February and October of 2009. Despite the similarities in the shapes of their growth patterns, Blacks internal efficacy grew at a much higher rate than whites. As shown in the difference column of table 1, the gap between blacks and whites feelings of internal efficacy grew linearly over the course of the election.

At the beginning of the 2008 election cycle, blacks and whites have similar levels of external political efficacy, but both of these groups diverge over the course of the election season. Unlike internal political efficacy, growth in external political efficacy follows a more stable upward trajectory for both racial groups. According to the difference column in Table 1, blacks appear to become increasingly more efficacious as the election year progresses. In November, when Obama was elected President, the difference between black and white voters was over third of a point based off of a five point scale for political efficacy. This cursory analysis suggests that Obama’s candidacy did have a positive impact on black’s and white’s levels of political efficacy. Blacks, however, become much more efficacious than whites over the course of the election.

[Insert Table 2 about Here]

Table 2 displays two unconditional latent growth curve models. As previously discussed, the LGCM estimates the slope and intercept for each individual’s level of political efficacy. The means estimate for the slope (Y1) presented in table 2, shows the average amount of growth for each group over time. A significant slope (Y1) estimate indicates that on average individuals within this group experienced a significant amount of growth in political efficacy.

The results presented in Table 2 suggest that as the election year a progress, external efficacy for both blacks and whites and internal efficacy for blacks grows at significant rates. For internal efficacy, black start at a higher rate than whites and grow at a rate of .24 points for every month in the data set. Whites on the other hand, experience no significant growth in internal efficacy. Blacks and whites have similar initial levels of external efficacy, but blacks grow at more than twice the rate of whites over the course of election. Overall, both fit statistics (CFI and RMSEA) show that the parameters in these models fit the observed data relatively well. The data provides support for my first hypothesis: Blacks became more efficacious over the course of the election.

[Insert Table 3 about Here]

Table 3 presents two models, the first is a linear growth curve model predicting internal political efficacy and the second is a latent growth curve model which predicts external political efficacy. LGCM provides two estimates for each model. The first is the direct effect on I which estimates the variables’ effect on the intercept. A significant intercept coefficient indicates that this variable has a strong effect on an individual’s initial level of political efficacy. The second estimate is the direct effect on S which estimates how each variable influences change in political efficacy over time. A significant slope coefficient indicates that this variable influenced the level of growth in political efficacy over the course of the campaign. Even though the LGCM calculates estimates for the intercept and the slope, I am primarily interested in the effects that Obama’s candidacy has on the rate of change in political efficacy. Therefore, the discussion will generally focus on the slope estimates for each group.

[Insert Table 4 about Here]

The results presented in table 4 provide additional support for my hypothesis. Not only did blacks have significantly higher levels of internal and external efficacy at the beginning of the election cycle, holding several variables constant, but black’s level of both types of political efficacy grew at a significantly greater rate than whites. According to the model, black respondents experienced about a fifth of a point (on a five point scale) increase in both forms of political efficacy each month in the data set. These results reiterate the previous findings presented in Tables 1 and 2 and also confirm my second hypothesis that Obama’s candidacy lead blacks to experience significantly more growth in political efficacy than whites.

Surprisingly, few other variables influenced change in internal and external political efficacy over time. Conservative ideology had a negative influence on change for both forms of efficacy. Respondents who identified as being conservative grew significantly less confident in their abilities to influence government and became less certain that government would be responsive to their needs as the election progressed. Other than ideology no other variables affected the slope for internal efficacy.

Income and whether the respondent supported Obama or Clinton in the primaries influenced the amount of change in individual’s levels of political efficacy over the 2008 election cycle. Respondents with less income became more efficacious over the course of the campaign. While this may indicate that Obama was able to energize voters with less resources, this result could also be due to the fact that those with more income had less room to grow in terms of efficacy. Finally, respondents who supported one of the leading Democrats in the primaries experienced significantly more growth in external efficacy than those who supported other candidates.

For both models, a respondent’s initial level of political efficacy significantly affected their growth in both forms of political efficacy. Those who start with less internal and external efficacy were more likely to become more efficacious over the course of the 2008 election. The estimated models in table 3 fit the observed data very well. Both RMSEA scores are well below the cutoff point of .07 for both models and both model’s CFI scores exceed .9. Overall, the results suggest that black candidates can have a positive impact on black political attitudes and more specifically black respondent’s feelings of political efficacy. Unfortunately, based solely on the 2008 election it is difficult to distinguish whether the change in black efficacy is due to Obama as an individual or if all high profile black candidate’s can inspire the similar change in black political efficacy. Therefore, further analysis is necessary.

[Insert Table 5 about Here]

To test whether the findings from the 2008 election are generalizable beyond Obama’s candidacy, I analyze prominent statewide black candidates’ effect on efficacy using the 1996 Black National Election Study (BNES). The 1996 BNES collected a large sample of black respondents and asked them questions related to politics and the 1996 election. This survey also contains questions about internal and external efficacy. For internal efficacy, the survey asks respondents on a 1 (Strongly Agree) to a 5 (Strongly Disagree) point scale how much they agree with the following statement “ R Doesn’t have a say about What Government Wants”. For external efficacy, the question uses the same scale and asks whether “R believes that Public officials don't care what people think”.

Using these questions, I run two OLS regressions estimating the effect that high profile blacks candidates have on political efficacy. In order to do this, I compare black’s levels of political efficacy in states with major party black US Senate or Gubernatorial Candidates, who ran within five years of the collection of the 1996 BNES, to those in states without these black candidates. To test this hypothesis, I create a dummy variable for states with black candidate who ran in recent elections[7]. In order to isolate the effect of the candidate, I control for individual level characteristics such as age, income, education, gender, employment status, partisanship, political trust, and political participation. In a separate model, I combine these individual factors, but I also include context variables such as the percent of Democrats and blacks in the state legislature in 1996[8]. Also, I controlled for the proportion of the black population in each state and created a dummy variable for whether the state was in the south.

Table five shows the OLS regression results for internal and external efficacy controlling first for individual characteristics and second for political and social context. The results show that high profile black candidates can have a significant effect on external efficacy for black voters. Holding several variables constant, blacks in states which had high profile black candidates had higher levels of external efficacy than blacks in states without these black candidates. The link between black candidates and internal efficacy does not appear to be as robust. When controlling for only individual level factors, black respondents in state with black candidates have higher levels of internal efficacy than other blacks. However, when you also control for the political and social context of the state this relationship becomes insignificant. Overall, the data suggests that black candidates can have a strong impact on external efficacy and in some circumstances impact blacks feelings of internal efficacy.

Conclusion

While the idea that descriptive representation (through elected officials) could erase black and white differences in political behaviors and attitudes is hardly novel, previous studies have ignored the role that viable black candidates have on the political empowerment of black voters. The results in this paper show that black candidates were associated with positively changing black’s feelings of efficacy. Both blacks and whites felt more efficacious as the 2008 election progresses, but blacks’ efficacy grew at a faster rate than whites. Furthermore, the results presented in table 4 indicate that states with blacks running for high profile statewide elected office can also have a positive impact on black political attitudes.

The results from this study imply that the relationship between descriptive representation and the change in black political attitudes begins at an earlier stage than when the black elected official is in office. Moreover, even when a black candidate is unsuccessful (as we saw with the state level black candidates) these candidates can still have a profound impact on black political attitudes As a result, this finding may create potential problems for studies that only examine black elected officials’ effect on the electorate. By not accounting for the effect of the candidate, previous studies may be mistakenly attributing changes in black political attitudes to black elected officials. This claim becomes especially problematic if a majority of the growth occurs when the elected official was a candidate. As a result, more research should examine the effects of descriptive representation over time. Future studies then should make comparisons between black candidates and black elected officials to determine if black elected officials influence change in political attitudes above and beyond the effects from their candidacies.

Many authors argue that low levels of political efficacy lead to an apathetic and frustrated electorate. Low levels of efficacy could lead to less participation, and some have even linked the absence of efficacy to aggressive forms of non-traditional political participation (Finkel 1987). Many studies hypothesize that disparities between black and whites in terms of socio-economic status lead black voters to have lower levels of both internal and external efficacy and this leads to lower levels of political engagement with the political system (Abramson 1972). While blacks lag behind whites in efficacy in most cases, the results in this paper show that prominent black candidates can help increase black confidence in their own political abilities and in the government to be responsive to their needs. As a result, high profile black candidates may help erase disparities in efficacy between blacks and whites. The results suggest that black candidates can play a very important role in the political incorporation of the black electorate.

References

Abney, F. G. and J. D. Hutcheson (1981). "Race, representation, and trust: Changes in attitudes after the election of a black mayor." Public Opinion Quarterly 45(1): 91-101.

Abramson, P. R. (1983). Political attitudes in America, W. H. Freeman San Francisco.

Abramson, P. R. and J. H. Aldrich (1982). "The decline of electoral participation in America." The American Political Science Review: 502-521.

Atkeson, L. R. (2003). "Not all cues are created equal: The conditional impact of female candidates on political engagement." The Journal of Politics 65(04): 1040-1061.

Balch, G. I. (1974). "Multiple indicators in survey research: The concept ‘sense of political efficacy.’" Political Methodology 1(2): 1-43.

Banducci, S. A., T. Donovan, et al. (2006). "Effects of minority representation on political attitudes and participation." Diversity in Democracy: Minority Representation in the United States: 193.

Barker, L. J., M. Jones, K, Tate. (1999) African Americans and American Political System, New York, Prentice Hall.

Bobo, L. and F. D. Gilliam Jr (1990). "Race, sociopolitical participation, and black empowerment." The American Political Science Review: 377-393.

Campbell, D. E. and C. Wolbrecht (2006). "See Jane run: Women politicians as role models for adolescents." The Journal of Politics 68(02): 233-247.

Clarke, H. D. and A. C. Acock (1989). "National elections and political attitudes: The case of political efficacy." British Journal of Political Science: 551-562.

Converse, P. E. (1972). "Change in the American electorate." The human meaning of social change: 263-337.

Dawson, M. C. (1994). Behind the mule: Race and class in African-American politics, Princeton University Press.

Finkel, S. E. (1985). "Reciprocal effects of participation and political efficacy: A panel analysis." American Journal of Political Science 29(4): 891-913.

Finkel, S. E. (1987). "The effects of participation on political efficacy and political support: Evidence from a West German panel." The Journal of Politics: 441-464.

Foster, L. S. (1978). "Black Perceptions of the Mayor An Empirical Test." Urban Affairs Review 14(2): 245.

Gay, C. (2001). The Effect of Minority Districts and Minority Representation on Political Participation in California, Public Policy Institute of CA.

Gay, C. (2002). "Spirals of trust? The effect of descriptive representation on the relationship between citizens and their government." American Journal of Political Science: 717-732.

Gilliam Jr, F. D. and K. M. Kaufmann (1998). "Is there an Empowerment Life Cycle?: Long-Term Black Empowerment and its Influence on Voter Participation." Urban Affairs Review 33(6): 741.

Hayes, B. C. and C. S. Bean (1993). "Political efficacy: A comparative study of the United States, West Germany, Great Britain and Australia." European Journal of Political Research 23(3): 261-280.

Hougland, J. G. J. and J. A. Christenson (1983). "Religion and politics: The relationship of religious participation to political efficacy and involvement." Sociology and Social Research Los Angeles, Cal. 67(4): 405-420.

Howell, S. E. (2000). "Racial Polarization, Reaction to Urban Conditions and the Approval of Black Mayors." Black and Multiracial Politics in America: 60–83.

Howell, S. E. and D. Fagan (1988). "Race and trust in government: Testing the political reality model." Public Opinion Quarterly 52(3): 343-350.

Lublin, D. (1997). The paradox of representation: racial gerrymandering and minority interests in Congress, Princeton Univ Pr.

Mansbridge, J. (1999). "Should Blacks Represent Blacks and Women Represent Women? A Contingent" Yes"." Journal of Politics: 628-657.

Rosenstone, S. J. and J. M. Hansen (1993). Mobilization, participation, and democracy in America, Macmillan Pub Co.

Shingles, R. D. (1981). "Black consciousness and political participation: The missing link." The American Political Science Review: 76-91.

Soss, J. (1999). "Lessons of welfare: Policy design, political learning, and political action." American Political Science Review: 363-380.

Steiger, J. H. and J. C. Lind (1980). Statistically based tests for the number of common factors.

Tate, K. (1991). "Black political participation in the 1984 and 1988 presidential elections." The American Political Science Review: 1159-1176.

Tate, K. (1994). From protest to politics: The new black voters in American elections, Harvard University Press.

Tate, K. (2003). Black faces in the mirror: African Americans and their representatives in the US Congress, Princeton University Press.

Verba, S., K. L. Schlozman, H Brady. (1995). Voice and equality: Civic voluntarism in American politics, Harvard University Press.

Wattenberg, M. P. (2002). Where have all the voters gone?, Harvard University Press.

Wolfinger, R. E. and S. J. Rosenstone (1980). Who votes?, Yale University Press.

Table 1: Average Change of Political Efficacy and Interest for Black and White Respondents throughout the 2008 Election.

Internal Efficacy External Efficacy

| |Whites |Blacks |Difference |Whites |Blacks |Difference |

|Wave 1 |1.633 |2.05 |-0.417 |1.39 |1.39 |0 |

|Wave 2 |1.584 |2.01 |-0.426 |1.37 |1.46 |-0.09 |

|Wave 9 |1.629 |2.13 |-0.501 |1.43 |1.53 |-0.10 |

|Wave 10 |1.518 |2.09 |-0.572 |1.4 |1.6 |-0.2 |

|Wave 11 |1.657 |2.43 |-0.773 |1.65 |2.01 |-0.36 |

|Average Growth |0.006 |0.095 |-0.5378 |0.06 |0.15 |-0.09 |

| | | | | | | |

Difference=White Efficacy-Black Efficacy

Table 2: Unconditional Growth Models for Internal and External Political Efficacy

Internal Efficacy External Efficacy

|Means |Black Linear Growth |Whites Linear Growth |Black Unconditional Latent |White Unconditional |

| | | |Growth |Latent Growth |

|Y0 |2.00*(.093) |1.60 (.024) |1.32* (.086) |1.33* (.025) |

|Y1 |.024* (.008) |-0.001(0.002) |.063* (.008) |.028* (.002) |

|Variance | | | | |

|Y0 |.961 (.180) |670* (.037) |.57* (.095) |.465* (.024) |

|Y1 |.000(.001) |.002* (.000) |.001 (.001) |.001* (.000) |

|I with S |-0.011(0.012) |-.015*(.003) |-.01 (.009) |-0.005 *(0.002) |

|Chi Square(DF) |30.424 (14) |92.654(14) |13.77 (11) |68.93 (11) |

|CFI |0.941 |0.977 |0.989 |0.982 |

|RMSEA(CI) |0.066 (.034-.099) |.048 (.039-.057) |.031 (.000-.075) |046 (.036-.057) |

* Significant at .05

Table 3: Latent Growth Curve Models Predicting the Intercept and Slope of Internal and External Efficacy

| |Direct Effect on I (Internal|Direct Effect on S |Direct Effect On I (External|Direct Effect On S (External|

| |Efficacy) b(SE) |(Internal Efficacy) b(SE) |Efficacy) b (SE) |Efficacy) b (SE) |

|Direct Effects | | | | |

|Black |0.441 *(.088) |0.185* (.085) |0.100+(0.060) |0.175*(0.066) |

|Education |0.006 (.011) |0.009 (.011) |0.036*(0.008) |0.002(0.008) |

|Age |-0.005*(0.002) |0.002 (.001) |-0.001(0.001) |0.001(0.001) |

|Income |0.006 (.005) |0.002 (.005) |0.021*(0.004) |-0.010*(0.004) |

|Male |-0.081+(.043) |-0.027(0.042) |-0.075*(0.030) |-0.044(0.032) |

|Employed |-0.072 (.052) |-0.023 (.05) |-0.072*(0.035) |-0.004(0.038) |

|Married |-0.033 (.051) |-0.038 (.049) |-0.045(0.034) |0.049(0.037) |

|Political Orientation | | | | |

|Political Interest |0.141*(.022) |-0.003 (.021) |0.059*(0.015) |0.016(0.016) |

|Conservative ID |0.016 (.017) |-0.025+ (.015) |0.010(0.013) |-0.049*(0.014) |

|Registered |0.077 (.088) |-0.071 (.076) |0.028(0.067) |0.007(0.072) |

|Voted in 04 |0.28* (.087) |0.001 (.079) |0.132*(0.061) |0.204*(0.063) |

|Candidate Support | | | | |

|Prim Clinton Sup |.005 (.087) |0.058 (.078) |0.046(0.056) |0.134*(0.061) |

|Prim Obama Sup |.05 (.087) |0.072 (.083) |0.114+ (.061) |0.125+(0.066) |

|General McCain Sup |0.073* (.032) |-0.047 (.029) |0.072*(0.022) |-0.034(0.023) |

|Cov: I with S | |-.109*(.022) | |-0.032*(0.014) |

|Latent Variables | | | | |

|Y0 (Mean-Inter.) | |1.003* (.206) | |0.395*(0.145) |

|Y1 (Mean-slope) | |-.051(.196) | |0.148(0.160) |

|Y0(Var-Inter) | |.621 *(.030) | |0.419*(0.016) |

|Y1(Var-Slope) | |.130*(.025) | |0.016(0.027) |

|Chi Square |264.957 (59) | |149.28 (56) | |

|CFI |0.955 | |0.977 | |

|RMSEA |.034(.030-.038) | |.023 (0.019 0.028) | |

|N |3049 | |3049 | |

|Rsquare Range |.58-.61 | |.550-.554 | |

| | | | | |

+ Significant at .10 *Significant at .05 Dependent Variable are Internal Efficacy (5 point Scale from 0 (strongly disagree) to 4 (strongly agree) to the following statement “How much can people like you affect what the government does?”) and external efficacy (point Scale from 0 (strongly disagree) to 4 (strongly agree) to the following statement “How much do government officials care what people like you think?”). Whites, females, employed respondents, single, and Republicans are the omitted/comparison categories.

Table 4: OLS Regression predicting Internal and External Efficacy using the 1996 Black National Election Study

| |External Efficacy |Internal Efficacy |External Efficacy |Internal Efficacy |

| |Baseline |Baseline |Context |Context |

|Black Candidate |0.23+ (.12) |0.27+ (.14) |0.28+(.14) |0.24 (.17) |

|Education |-0.02*(0.014) |-0.03(.02) |-0.03*(0.01) |-0.03*(0.01) |

|Income |0.01* (.00) |0.01+ (.00) |0.01* (.02) |0.01 (.00) |

|Age |-0.01* (.00) |0.00 (.00) |-0.01*(0) |0.00 (.00) |

|Male |0.01(.10) |-0.03 (.12) |0.02 (.10) |-0.04*(0.12) |

|Unemployed |0.10 (.19) |0.17 (.23) |0.12 (.20) |0.12 (.23) |

|Democrat |-0.05(.11) |-0.01 (.14) |-0.04(0.12) |-0.02*(0.14) |

|Republican |0.24 (.29) |0.60+ (.34) |0.30 (.29) |0.58 (.34) |

|Voted In 96 |.15 (.12) |0.30 (.14) |0.14 (12) |0.27 (.14) |

|South | | |-0.05 (.11) |0.17 (.13) |

|% Blacks In State Legislature | | |2.32 (2.13) |-0.02 (.87) |

|% Black Population | | |-1.34(1.43) |-0.17*(1.68) |

|% Democrats in State Legislature | | |0.43 (.74) |2.62 (2.51) |

|Constant |3.02* (.29) |2.99* (.35) |2.73(.45) |2.65 (.53) |

|Adj. R Square |0.03 |0.02 |0.03 |0.02 |

|Number |751 |782 |751 |782 |

+ Significant at .10 *Significant at .05 Ommitted/ Comparison Groups are Black Respondents in states w/o a black candidate, females, independents, and non southern states.

Figure 1: Conditional Linear Growth Curve Model Predicting Internal and External Efficacy over the course of the 2008 Election Cycle.

[pic]

Appendix

Question Wording and Variable Coding for the 2008-2009 ANESPS

Internal Efficacy: Question- “How much can people like you affect what the government does?” Internal Efficacy is coded as follows: Not at All=0, A little=1, A Moderate Amount=2, A Lot=3, A Great Deal=4

External Efficacy: Question-“How much does government officials care what people like you think?” External Efficacy is coded as follows: Not at All=0, A little=1, A Moderate Amount=2, A Lot=3, A Great Deal=4

Note: All Independent Variables are measured using each respondent’s initial response.

Education: Question- “What is the highest degree or level of school you have completed?” Education was measured as years in school: No Schooling=0, 4th grade=4, 5th or 6th Grade=5.5, 7th or 8th grade=7.5, 9th grade=9, 10th grade=10, 11th grade=11, High School Graduate=12, Some College=14, Associate’s degree=14, Bachelor’s degree=16, Master’s degree=18, Professional or Doctorate=20.

Married: Question- “Are you now married, widowed, divorced, separated, or never married?” Married is coded as follows: Married=1, Widowed=0, Divorced=0, Separated=0, Never Married=0

Employed: Question-“Which statement best describes your current employment status?” Employed is coded as follows: Working - as a paid employed/Working - self-employed=1, Not working-on temporary layoff from a job/looking for work/retired/disabled/other=0

Income: Question- “Was your household income in the past 12 months….” Income was coded as follows: less than $5,000=2.5, $5,000 to $7,499=6.25, $7,500 to $9,999=8.75, $10,000 to $12,499=11.25, $12,500 to $14,999=13.75, $15,000 to $19,999=17.5, $20,000 to $24,999=22.5, $25,000 to $29,999=27.5, $30,000 to $34,999=32.5,$35,000 to $39,999=37.5, $40,000 to $49,999=45, $50,000 to $59,999=55, $60,000 to $74,999=67.5, $75,000 to $84,999=80, $85,000 to $99,999=92.5,$100,000 to $124,999=112.5, $125,000 to $149,999=137.5, $150,000 to $174,999=162.5, $175,000 or more=175

Conservative: Question- “When it comes to politics, would you describe yourself as liberal, conservative, or neither liberal nor conservative?” Conservative is coded as follows: Liberal=0, Conservative=1, Moderate=0

Registered: Question- “Are you registered to vote, or not?” Registered was coded as follows. Yes=1 No=0

Vote in 04: Question- “Which one of the following best describes what you did in that election in 2004?” Vote 04 is coded as follows: Definitely did not vote=0,Definitely voted in person at a polling place on election day=1, Definitely voted in person at a polling place before election day=1, Definitely voted by mailing a ballot to elections officials before election day=1, Definitely voted in some other way=1.

Or Question- “If you had to guess, would you say that you probably did vote in the election held in November 2004, or probably did not vote in that election?” Again Vote 04 is coded as follows: 1=Probably Voted, 0=probably did not vote.

Political Interest: Question- “How interested are you in information about what’s going on in government andpolitics?” Interest was coded as follows: 0=Not Interested at all 1=Slightly Interested 2=Moderately Interested 3=Very Interested 4=Extremely Interested.

Obama Primary Supporter: Question- “Which presidential candidate did you vote for?” Obama Primary Supporter is Coded as Follows: Barack Obama=1, Hillary Clinton=0, John Edwards=0, Bill Richardson=0, Joe Biden=0, Chris Dodd=0, Dennis Kucinich=0, Mike Gravel=0,someone else=0

Clinton Primary Supporter: Question- “Which presidential candidate did you vote for?” Clinton Primary Supporter is Coded as Follows: Barack Obama=0, Hillary Clinton=1, John Edwards=0, Bill Richardson=0, Joe Biden=0, Chris Dodd=0, Dennis Kucinich=0, Mike Gravel=0, someone else=0.

McCain General Election Supporter: Question- “Who did you vote for in the election for President?” McCain General Election Supporter is Coded as Follows: John McCain=1, Barack Obama=0, Someone Else=0

Question Wording and Variable Coding for the 1996 BNES

Internal Efficacy: Question- Please tel1 me how much you agree or disagree with these statements. Do you strongly agree, somewhat agree, neither agree nor disagree, somewhat disagree or strongly disagree? Public officials don't care much what people like me think? Internal Efficacy was coded as follows: 1=Disagree Strongly, 2=Disagree somewhat, 3= Neither agree nor disagree, 4= Agree Somewhat, 5=Agree Strongly

External Efficacy: Question- Please tel1 me how much you agree or disagree with these statements. Do you strongly agree, somewhat agree, neither agree nor disagree, somewhat disagree or strongly disagree? People like me don't have any say about what the government does. External Efficacy was coded as follows: 1=Disagree Strongly, 2=Disagree somewhat, 3= Neither agree nor disagree, 4= Agree Somewhat, 5=Agree Strongly

Education: Question- “What is the highest degree or level of school you have completed?” Education was measured as years in school: Some Grade School= 5, Some High School=10, High School Graduate=12, Some College=14, Associate’s degree=14, Bachelor’s degree=16, Some Graduate/Master’s degree=18, Professional or Doctorate=20. DK/REFUSED were dropped

Income: Question- “Was your Family income in the past 12 months….” Income was coded as follows:

10=Up to $10,000, 12.5=$10,000-$15,00, 17.5=$15,000-$20,000, 22.5=$20,000-$25,000, 27.5= $25,000-$30,000,35= $30,000-$40,000, 45= $40,000-$50,000,62.5=$50,000-$75,000, 82.5= $75,000-$90,000, 97.5= $90,000-$105,000, 105= $105,000 AND MORE. DK/REFUSED were dropped

Unemployed: Question: Employment status. Employment status is coded as follows: WORKING NOW ONLY/WORKING NOW/RETIRED--WORKING 20+/WORKING NOW/PERMANENTLY DISABLED/WORKING NOW/HOMEMAKER, WORKING /WORKING NOW/STUDENT WORKING 2O+=1 TEMPORARILY LAID OFF/UNEMPLOYED/RETIRED--NO OTHER OCCUPATION/RETIRED AND WORKING LESS THAN 2/PERMANENTLY DISABLED--NOT WORKING/PERMANENTLY DISABLED AND WORKING NOW, WO/HOMEMAKER--NO OTHER/HOMEMAKER AND WORKING NOW, WORK/STUDENT--NO OTHER OCCUPATION=0

Vote in 1996: Question: In talking to people about elections, we often find that a lot of people were not able to vote because they weren’t registered, they were sick, or they just didn’t have time. How about you--did you

vote in the elections this November? Vote in 1996 is coded as follows:1=Yes, voted, 2=No, didn’t vote. DK/REFUSED were dropped

South: The South dummy variable contains respondents from the following states: Alabama, Arkansas, Florida, Georgia, Kentucky, Louisiana, Maryland, Mississippi, North Carolina, South Carolina, Tennessee, Texas, Virginia.

% Black State Legislature: This variable includes the proportion of blacks in both the upper and lower houses in each state. The data was obtained from the Joint Center for Political and Economic Studies.

% Black in each State: Data was collected from the US Census

% Democrats in State Legislature: This variable includes the proportion of democrats in both the upper and lower houses in each state.

-----------------------

1 I consider candidates who run for statewide elected office and president as high profile candidates.

[1] The current data set has not gone through normal ANES quality control procedures. As a result, ANES warns those who use the data set to recognize that some changes may occur when the data set is fully released in September. Possible changes include revisions to the weights, changes to the missing value codes, addition or deletion of cases, and re-computation of summary variables.

[2] To maximize the N for both groups, MPLUS imputed values for missing data using maximum likelihood estimation. See for more details.

[3] I also considered including party identification into the model. Unfortunately, the party identification variable was coded incorrectly in the advanced release of the Data Set. However, we do not expect this to significantly influence our results because we use candidate choice and ideology as proxies for partisanship.

[4] I did not control for Obama support in the general election because this variable was highly collinear with Obama and Clinton support in the primaries.

[5] See Supplemental Index for an analysis of best model fit for external and internal efficacy.

[6] States that had black candidates run for Governor or US Senator from 1991 to 1996 are as follows: North Carolina 1996, Louisiana 1995, Missouri 1994, Washington 1994, and Maryland 1992. While Carole Moseley Braun also ran for US Senator in 1992, Illinois was excluded from my black candidates dummy variable. During the collection of the 1996 BNES, Moseley Braun was serving as a US Senator and to avoid a potential confound by including a black elected official, I omitted Illinois from this variable. A separate model not reported in this study was run with Illinois included in the black candidates dummy variable and this made states with black candidates more significant than the variable reported in this analysis.

[7] Data was obtained from the Joint Center of for Political and Economic Studies 1996 BEO Roster.

-----------------------

Internal/

External

Efficacy

(Intercept)

Internal/

External

Efficacy

(Slope)

Political Efficacy (Jan)

Political Efficacy (Feb)

Political Efficacy

(Sep)

Political Efficacy (Oct)

Political Efficacy (Nov)

1

1

1

1

1

0

1

8

9

10

E1

E2

E3

E4

E5

Covariates: Race, Gender, SES, Socio-Demographic, Pol Ideology, and Candidate Choice

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download