Long-term declines in voter turnout ( United States Election Project, 2016 ) and discussions of disaffected voters ( Ignatius, 2016 ; DelReal, 2014 ; Jackson, 2015 ) are part of the current political landscape. As we approach the 2016 U.S. presidential elections, it is important to consider trends in American’s attitudes about voting. Do these trends support the generally pessimistic assessments about voting?
Voting behaviors and attitudes are important elements of citizenship, civil society, and the American Dream ( McElwee, 2015a, 2015b; 2016 ; Atkeson & Carrillo, 2007 ; Hanson & White, 2011 ). Although there is some diversity of opinion on the degree to which voting attitudes are related to voting turnout, many have argued that these attitudes affect voting behaviors ( Vecchione & Caprara, 2009 ; Blais et al., 2011 ). Additionally, there is agreement that attitudes about voting are critical in healthy democracies ( Blais, 2000 ; Barbour & Wright, 2012 ). Most models suggest three important correlates that are important in understanding voting related attitudes and behaviors―political involvement, political efficacy, and social connectedness ( Pinkleton & Austin, 2001 ; Timpone, 1998 ; Blais, 2000 ; Vechione & Caprara, 2009 ; Atkeson & Carrillo, 2007 ).
Can we make general conclusions about voting attitudes and correlates among Americans or do these attitudes vary by important social statuses? Increasing evidence suggests that simple statements about voting behaviors and attitudes are deceiving, and ignore considerable diversity across gender ( Atkeson & Carrillo, 2007 ; Verba et al., 1987 ; Shearer, 2016 ; Pratto et al., 1997 ; CAWP, 2015 ) and race/ethnic groups ( Bobo, 1990 ; Barreto, 2007 ; Shearer, 2016 ; Frey, 2013 ; Preston & Santos, 2012 ; Lopez & Gonzalez-Barrera, 2013 ).
Using data from the American National Election Studies (ANES), this research examines trends in voting related attitudes and in correlates of voting attitudes. We also address the question of whether trends in voting attitudes are related to the correlates. That is, do the voter attitude trends vary across groups of people that are high and low on correlates involving political involvement, political efficacy, and social connectedness? Finally, given the diversity in the
The data used here were retrieved from the Time Series Cumulative Data File (1948- 2012) available on the American National Election Studies website (http://www.electionstudies.org/). The ANES data are an invaluable resource for researchers interested in a rich variety of voting data for probability samples that allow comparisons across individuals, groups, and time. More specifically ANES involves a series of biennial National Election Studies where face to face surveys (sometimes supplemented by an internet sample) are conducted on probability samples of eligible voters in the U.S. The ANES data include measures on issues involving electoral participation, voting behavior, and public opinion as well as other demographic and attitudinal measures. The 2012 data are the 29th study in a series of election studies conducted during years of presidential elections since 1948. The years included in this trend analysis are: 1996, 2000, 2004, 2008, and 2012. These years were chosen in order to examine attitudes over a nearly 2-decade period. They were also chosen since the data for most years include repeated measures of our study variables. Although the ANES data is an excellent resource for answering our research questions about voting attitudes, correlates, and variation by gender and race/ethnicity, it is always desirable to have more extensive measures of the variables in the survey and greater representation of race/ethnic minority groups in the sample.
Our research questions focus on trends in voting attitudes and correlates as well as gender and race/ethnic variation in these trends. ANES includes variables that are indicators of the voting attitudes (see question in Table 1) and correlates involving political involvement, efficacy, and social connectedness (see questions in Tables 2-4). These measures are available for most variables in the years between 1996 and 2012.
Table 1. Intent for vote for president.
ANES: “Needless to say, the election for President of United States is a long time away in November. But I’d like to ask you for your best guess about who you will vote for in the election in November. Who do you think you will vote for?”
Table 2. Political involvement-interest in elections.
ANES: “Some people don’t pay much attention to political campaigns. How are you, would you say that you have been/were very much interested, somewhat interested, or not much interested in (1996: following) the political campaigns so far this year?”
Table 3. Political efficacy―government officials care.
ANES: “I don’t think public officials care much what people like me think”.
Table 4. Social connectedness―church attendance.
ANES: “Lots of things come up that keep people from attending religious services if they want to. Thinking about your life these days, do you ever attend religious services, apart from occasional weddings, baptisms, or funerals? (If Yes) “Do you go to religious services every week, almost every week, once or twice a month, a few times a year, or never?”
Trend analyses are an important part of understanding shifts in public opinion. These analyses of opinion over time are by nature descriptive. Thus our examination of voting related attitudes and correlates is not causal but provides essential base-line descriptive data on levels of support for key indicators of voting. Importantly, our research also describes variation in these indicators by two critical demographic characteristics― gender and race/ethnicity. Cross-tables are used to show trends in voting related attitudes and correlates. Line-graphs are used to show variation in these trends by gender and race/ethnicity.
Relationship between Voting Attitudes and Correlates of Voting Correlates. Before conducting the trend analysis of voting related attitudes, we sought to discover whether trends in voting attitudes and correlates (involvement, efficacy, and connectedness) are correlated in the ANES data. In analyses not shown here, we examined voting attitudes within groups of respondents who reported high and low political involvement, political efficacy, and social connectedness. More specifically we look at intent to vote within categories of correlates of voting attitudes. We found a strong connection between each of the correlates and voting attitudes―those with higher involvement, efficacy, and connectedness are the most likely to intend to vote.
Attitudes toward Voting. Data in Table 1 show trends in attitudes toward voting. We focus on the response category, “respondent does not intend to vote”. Data from 1996 through 2012 show that in each year through 2010, only 10% to 12% of Americans chose this response when asked about their intent to vote for President. However, in 2012 the percent that did not intend to vote fell to 3.6.
In Figure 1 and Figure 2 we examine trends for the intent to vote item by gender and race/ethnicity respectively. Figure 1 shows women were slightly more likely than men to say they did not intend to vote in 1996 (12.5 vs. 11.6), but less likely than men to report this in 2004 (9.7 vs. 12.11). By 2012 there was very little gender difference with 3.5 percent of men and 3.7 percent of women choosing the “does not intend to vote” response.
Intent to vote trends across race/ethnic groups are shown in Figure 2. In 1996 it was Whites who were the least likely to report that they did not intend to vote (11.4%) but by 2004 and beyond, Blacks were consistently the least likely to report that they would
Figure 1. Do not intend to vote by gender, ANES.
Figure 2. Do not intend to vote by race & ethnicity, ANES.
not vote. However, it should be noted that intent to not vote did decrease across race groups in 2012.
Correlates of Voting Attitudes―Political Involvement. Data in Table 2 provide trends on a measure of political involvement ?interest in political campaigns. There has been a considerable increase in those reporting they were “very much interested” in the political campaign this year. Only 27.5% of respondents noted this level of interest in 1996, but 44.5% were very much interested in 2008. The change in level of interest occurred most markedly between 2000 and 2004.
Figure 3 and Figure 4 show the trends in interest in this year’s political campaigns by gender and by race/ethnicity. In Figure 3 we see that men are more involved in politics than women in each survey year using this measure of interest in the political campaign. However, both men and women showed an increase in involvement during the survey period. The trend toward being “very much interested” in the political campaign this year was especially marked for women. The increase in percent of women choosing this option (24.2% in 1996 and 42.7% in 2008) was greater than the increase for men (31.7% in 1996 and 46.8% in 2008).
Trends in interest in this year’s political campaigns by race/ethnicity (Figure 4) show that Blacks were the most likely of the race/ethnic groups to report “very much interested” in 1996 (31.5%), but by 2008 they were the least likely to choose this option (30.5%). The percent reporting “very much interested” increased considerably for both Whites (27.6% to 45.8%) and for Hispanics (25.5% to 38.4%) over the same period. Note that we were unable to use the 2012 ANES data for this question since it used a set of response codes that were not consistent with earlier codes. On other measures of trends, we report a marked change in the public opinion of Blacks on political attitudes in 2012.
Correlates of Voting Attitudes―Political Efficacy. Data in Table 3 show overall stability for opinion on whether public officials care what people like the respondent think. Americans show low efficacy here. A majority of Americans in each year agreed with the statement that public officials do not care. The percent agreeing with this state- ment was stable at 60.9 in 1996 and 61.1 in 2012.
In Figure 5 and Figure 6 we examine the trends in the “public officials do not care” measure of external political efficacy by gender and by race/ethnicity. Trends in Figure 5 show that males are more likely than females to agree with the statement that public officials don’t care in 1996 and in 2012 (63.8% vs. 58.6% and 63.1% vs. 59.2%).
Trends in Figure 6 show that in 1996 and 2004 it was Blacks who had the lowest external political efficacy and were the most likely to agree with the negative statement
Figure 3. Very much interested in elections by gender, ANES.
Figure 4. Very much interested in elections by race & ethnicity, ANES.
Figure 5. Agree, government officials do not care by gender, ANES.
Figure 6. Agree, government officials do not care by race & ethnicity, ANES.
about public officials (e.g., in 1996 the figures on percent agree show 65.5% for Blacks vs. 49.5% for Hispanics and 59.9% for Whites). This trend reversed in 2012 with Blacks now having the highest efficacy and being the least likely among the race/ethnic groups to agree with the statement (54.6% vs. 59.7% for Hispanics and 63.5% for Whites). Thus, results on opinion about whether the respondent thinks public officials don’t care suggest that Whites had the highest external efficacy (relative to other race/ethnic groups) in 1996 and the lowest in 2012. Another interesting trend in Figure 6 involves the efficacy for Hispanics relative to the other race/ethnic groups. As the figures noted above suggest, Hispanics had the highest efficacy in 1996 on the item asking about public officials. In the most recent survey year (2012), they were higher than Whites but lower than Blacks on the measure of efficacy.
Correlates of Voting Attitudes―Social Connectedness. Table 4 shows trends in social connectedness as measured by attendance at religious services. Trends show a decrease in attendance over the 1996-2012 period. Those attending church more than once a week declined from 13.0% to 9.9%. However, the percent of respondents who reported attending “once a week” or “almost every week” remained stable over the period at approximately 11%. Finally, the percent of respondents who never attend increased from 30.3 to 40.0 during the survey period.
Figure 7 and Figure 8 show trends in the attendance at religious services measure for males and females and for race/ethnic groups. In every survey year, trends in Figure 7 show that women tend to be more likely than men to report attending religious services every week (either more than once or once). The gender difference remained somewhat stable over the survey years on the attending religious services more than once a week response. For example, 11.5% of males and 14.1% of females reported this frequent attendance in 1996 and 7.9% of males and 11.7 percent of females reported this attendance in 2012. Note the slight decline in percent attending more than once a week for both gender groups. However, an interesting gender trend occurs on the “once a week” religious attendance item. Men actually increased their response to this item between 1996 and 2012 resulting in virtually no gender difference in the percent reporting weekly attendance in 2012 (12.5 % for men and 12.7% for women).
Trends on attendance at religious services for different race/ethnic groups in Figure 8 show that across survey years, Blacks were the most likely to report high attendance (attending religious services more than once a week). The race/ethnic difference in attendance increases between 1996 and 2012 with Blacks becoming almost twice as likely as Whites to be in this category by 2012 (16.1% vs. 8.4%). In earlier years Hispanics
Figure 7. Church attendance at least once a week by gender, ANES.
Figure 8. Church attendance at least once a week by race & ethnicity, ANES.
were less likely than Whites to report this level of high attendance but by 2012 both Hispanics and Blacks were more likely than Whites to report attending religious services more than once a week (16.1% for Blacks, 9.1% for Hispanics, and 8.4% for Whites).
Discussion of Findings. Although other researchers have not used the trend data used here, our findings are consistent with others who have found a relationship between voting attitudes and the correlates involving involvement, efficacy, and connectedness ( Timpone, 1998 ; Atkeson & Carrillo, 2007 ). Our findings showing positive trends or stability on many of these attitudes and correlates are more optimistic than some research on voting turnout ( United States Election Project, 2016 ) and disaffected voters (e.g., Ignatius, 2016 ). Although there is limited research on gender and race/ ethnic diversity and voting attitudes, the encouraging trends shown here on some positive attitudes and correlates among women and among racial/ethnic minorities are supported by a growing body of research (e.g., Atkeson & Carrillo, 2007 ; Weiner, 2013 ; Preston & Santos, 2012 ).
Data from ANES were used to examine voting attitudes and correlates of voting attitudes. Results suggest some interesting trends contributing to optimistic conclusions about these attitudes and opinions that are important elements of democracy and civil society. U.S. respondents increasingly report that they plan to vote. Over the past 2 decades, trends show increased political involvement (interest in political campaigns). Trends show stability in social connectedness (weekly attendance at religious services) and efficacy (belief that public officials do not care). Findings suggest significant variation in trends on the voter attitudes and correlates across gender and race/ethnic groups. As we approach the 2016 presidential elections, conclusions from this analysis of trends suggest that Americans are overall positive in their attitudes about voting and on measures of political involvement and social connectedness. The results also suggest that these trends are not one size fits all―there is considerable variation across gender and race/ethnicity. Some of the optimism in the attitudes of American voters comes from the influence of women and race/ethnic minorities.
*Thanks to Janet Adesina for her research assistance.
 Vechione, M., & Caprara, G. V. (2009). Personality Determinants of Political Participation: The Contribution of Traits and Self-Efficacy Beliefs. Personality and Individual Differences, 46, 487-492.