Introduction
Recent studies of political attitudes and voting patterns in the United States presidential elections demonstrate that voter loyalties and voting trends have undergone a major shift since the original large-scale voting studies were initiated in the 1940s and 1950s (Guterbock, 1980).
In the American context, the presidential elections are often a hotly contested affair between the traditionally business-oriented Republicans and the labor-oriented Democrats (Gelman, Kenworthy & Su, 2010). However, as suggested by Gonzalez (2012), U.S. presidential elections are won on a number of platforms other than the philosophical leanings of the candidates or party interests.
In this regard, it is important to evaluate the determinants of voting participation in the U.S. presidential elections not only to assist political parties to polish their campaigns during the electioneering period, but also to inform policy decisions at the party level.
While it is a well known fact that most Americans consider themselves as Republican or Democrat, a strand of existing political literature (e.g., Newman, 2001; Pasek et al., 2009; Weisberg, 2007) demonstrates that presidential election results are paradoxically predicated upon a multiplicity of factors other than partisan voting.
While these studies exhibit rekindled interest in the existence of factors outside party identification that are intrinsically involved in determining the results of the presidential election (Holbrook & McClurg, 2005), less attention has been devoted to examining the mechanisms by which these factors activate or deactivate partisanship and mobilize core supporters toward voting for a particular presidential candidate Wildavsky, 2011).
Interestingly, many of these studies tend to deal with individual factors to understand voting patterns in U.S. presidential elections though common practice demonstrates that a number of factors are involved in efforts to shape the voting discourse and trajectory.
The present study aims to fill this research gap by analyzing four possible determinants of voting participation in U.S. presidential elections: media, religion, social economic status and level of education. A comprehensive review of these factors is presented in the subsequent sections.
Available literature demonstrates that the media (print, electronic, Internet) has an agenda-setting capacity or “the ability to influence not what people think, but what they think about” (Newton & Brynin, 2001 p. 225).
This view is reinforced by Gonzalez (2012), who acknowledges that most modern media platforms have the capacity to prime or frame issues in a manner that leads the audience or users to think about them in one way rather than another.
Extending and supporting the work in this nascent area of research, Newman (2001) claims that victory in the U.S. presidential election often goes to the candidate who wages the best marketing campaign using available media platforms not only to make an emotional connection with the people, but also to project an image of honesty, compassion and toughness in the minds and hearts of the American people.
Other scholars note that “both candidates and voters have increased their use of the Internet for political campaigns” (Robertson, Vatrapu & Medina, 2010 p. 11).
Presidential candidates, according to these authors, have adopted many Internet-based tools to communicate with voters, collect contributions, foster community and organize political campaigns, whereas voters have adopted Internet tools to relate to the presidential candidates, engage in political discourse, follow activist causes, and share information.
The relationship between religion and the U.S. presidential elections has been investigated by a number of scholars. In his seminal work on religious identity and the U.S. presidency, Gonzalez (2012) found that “the relationship between religion and the presidency impacts both the viability of candidates and the manner in which decisions are made in the voting booth” (p. 568).
In the 2012 presidential elections, for instance, Republican candidate Mitt Romney suffered considerable stigma from the American voters due to his close association with Mormon religious doctrines (Gonzalez, 2012).
Manza and Brooks (1997) are of the opinion that in the U.S political landscape, religious-oriented cleavages may have been a more fundamental factor for understanding the social roots of voter alignments than the class cleavage owing to the fact that Americans normally claim higher levels of church membership and attendance at religious gatherings and hence are more likely to believe in God and claim that religion is of substantial importance in their lives.
Social economic status has been shown as a possible determinant of voting participation in the U.S. presidential elections. A study by Southwell (2004) shows that unemployed and economically-disadvantaged people are less likely to take part in voting, whereas their employed and rich counterparts derive much satisfaction from participating in presidential elections.
This author further explains that persons experiencing financial difficulties are “less likely to participate in elections because the stressful nature of economic adversity forces a preoccupation with personal economic problems and makes the individual withdraw from political or community matters as a result” (p. 237-238).
Guterbock (1980) used ecological data from the Midwestern city of Middletown to demonstrate that although there is a perceived weakening of the relationship between socioeconomic status (SES) and electoral choices, a considerable number of eligible Americans continue to vote along class and racial lines.
According to the researcher, wealthy people and those in white-collar occupations continue to vote for Republican candidates to maintain the status quo, while middle class (working class) voters and immigrants vote for Democrat candidates because the policies projected by the Democratic Party are perceived as more responsive to their interests.
Lastly, a number of research studies have investigated the relationship between a voter’s level of education and his or her voting participation in the U.S. presidential elections.
Although the results are not conclusive, Coley and Sum (2012) “reveal a startling stratification at the nation’s polling stations, from a voting rate of 3.5 percent for voting-age high school dropouts to 80.5 percent for well-off, advanced-degree holders between the ages of 55 and 64” (p. 2).
These authors found a significant association between the level of education and civic engagement (e.g., participating in elections), leading to the conclusion that the nation’s less-educated, lower-income eligible voters have willingly disenfranchised themselves form the voting process.
One Canadian study analyzing the last federal election found that “the voting rate among people with a university degree was 78% compared with rates of 60% or lower among those with a high school education or less” (Uppal & LaRochelle-Cote, 2012 para. 12). These figures demonstrate that education may have a ‘positive effect’ on voting patterns not only in the U.S. but also internationally.
The present study is interested in testing the following hypotheses. The first hypothesis is that voters with high media exposure (TV) are more likely to participate in U.S. presidential elections than voters with low media exposure. The second hypothesis is that voters with a solid religious orientation are more likely to determine the outcomes of the U.S. presidential election than voters with a secular orientation.
he third hypothesis is that social economic status is a strong predictor in determining the probability of voting in the U.S. presidential elections. The last hypothesis is that voters with low levels of education are less likely to take part in presidential elections than voters with high levels of education.
Methods
The data for this study were extracted from a larger database known as the General Social Survey (GSS), which is basically a nationwide survey intended to capture the demographic, behavioral and attitudinal views of Americans on a wide range of issues. The GSS is a probability sampling national survey completed through personal questionnaires targeted at non-institutionalized individuals over the age of 18 years.
Information from the GSS official website shows that “the 1972-2012 GSS has 5,545 variables, time-trends for 2,072 variables, and 268 trends having 20+ data points” (General Social Survey, 2013 para. 2). The 1996 data set was used in this study, and the sample size drawn for analysis consisted of 1,419 Americans. Data relevant to the dependent and the independent variables were used to test the hypotheses.
Questions were posed to the participants and the responses entered into the corresponding categories in line with a quantitative approach. However, some items were not operationalized and required the respondents to give their responses in an open-ended manner. These responses were later operationalized by the researcher around underlying themes and then analyzed quantitatively using the IBM SPSS Statistics program.
While the ordinal level of measurement was mostly used when values for the responses represented categories with some intrinsic ranking, the nominal level of measurement was used when values for the responses showed no form of intrinsic ranking, whereas the scale level of measurement was employed when values for the responses represented ordered categories demonstrating a meaningful metric (Balnaves & Caputi, 2001).
For this study, the independent variables include respondent’s highest level of education (measured using ordinal level by ranking participants against the intrinsic categories for educational achievement), TV hours and Internet hours (measured using scale level by stating the number of hours respondents use per day watching TV), respondents income for the last year (measured using ordinal level by entering the respondent’s income for the last year into predetermined intrinsic categories), and belief in life after death (measured using nominal level as a “YES/NO” response).
The dependent variable is whether the respondents voted in the 1996 U.S. presidential election. As already mentioned, the sample size for this study is 1,419.
Quantitative techniques were employed to analyze the data with the view to testing the stated hypotheses. Descriptive statistics (frequency distributions and cross tabulations) were used to demonstrate the frequency of occurrence and the relationships between the dependent and independent variables.
Chi-square tests were also done for purposes of identifying which frequencies and relationships could be considered statistically significant. The results are presented in the following section.
Results
Table 1 shows the cross-tabulation of the number of number of hours per day respondents spent watching TV and if they voted in the 1996 presidential elections. It is imperative to note that over two-thirds (67.9%) of respondents who spent a minimum of two hours per day watching the TV voted against only 118 (25.1%) who spent the same number of hours but did not vote.
Similarly, 205 (67.9%) of respondents who spent 3-5 hours watching the news voted in the 1996 general elections against 23 (27.6%) who spent similar number of hours but did not vote. The Pearson Chi-Square test showed a df of 45 and two-sided significance of 0.290, while the Spearman Correlation Coefficient showed an approximate significance of 0.082.
Consequently, the level of occurrences and relationship is significant enough to prove that voters with high media exposure are more likely to participate in voting than voters with low media exposure.
Table 1: Hours spent watching TV and Participation in 1996 General Elections
Table 2 shows the cross-tabulation of perceptions of belief in life after death (to demonstrate religiosity) and if respondents voted in the 1996 presidential elections. From the cross-tabulation, it is clear that out of 521 valid cases of respondents who voted in the 1996 presidential elections, 449 (86.2%) believed in life after death (religiosity) while only 72 (13.8%) of those who voted said there was no life after death.
The Pearson Chi square and linear-by-linear association showed a weak relationship between the independent and dependent variable (0.003 and 0.004 respectively at 0.05 significance level) but the high occurrence of those who voted in the 1996 election and demonstrated a faith or belief in life after death proves that voters with a solid religious orientation are more likely to determine the outcomes of the U.S. presidential election than voters with a secular orientation.
Table 2: Belief in Life after Death and Participation in the 1996 Presidential Elections
Table 3 demonstrates the cross-tabulation between the social economic status (measured by income for last year) of respondents and if they voted in the 1996 presidential elections.
From the cross-tabulation, it is evident that of the 573 respondents who participated in the 1996 presidential election, 105 (18.3 %) earned a salary of up to $12,999 per year, 341(59.5%) respondents earned between $12,500 and $49,000, and a further 127 (22.2%) earned $50,000 or more per year.
The Pearson Chi-square and linear-by-linear association (both at 0.000 at 0.05 significance level) demonstrated no significant association that could have given credence to the hypothesis in symmetric measures. However, the cross-tabulation analysis proves that social economic status (as measured by respondents’ income for the previous year) is a strong predictor for participation during presidential elections.
Table 3 Respondents Income for Last Year and Participation in 1996 Presidential Election
Table 4 shows the cross-tabulation analysis of the respondents’ highest level of education and participation in the 1996 presidential election.
The Pearson chi-square analysis and linear-by-linear association both demonstrated that there was a significant relationship between level of education and participation in presidential voting (Pearson Chi-square = 103.702, df-6, p =.004; linear-by-linear association = 93.526, df = 1, p = 0.000). The Lambda measure of association revealed a strong association between variables (0.774).
Table 4: Educational Level and Participation in 1996 Presidential Election
The cross-tabulation above demonstrates that out of the 883 respondents who participated in the 1996 presidential voting, 778 (89.2%) had a high school degree and above. This analysis together with the measures of association and significance proves that voters with low levels of education are less likely to take part in presidential elections than voters with high levels of education.
Discussion
The findings of this study demonstrate that media exposure, religion, social economic status and level of education are important determinants of voter participation in the U.S. presidential elections. Consequently, the present study reinforces findings of other studies that have evaluated individuals attributes (variables) and found them to have a significant influence on voting behaviors and patterns.
In this study, media exposure has been found to be positively associated with a high likelihood of participating in presidential elections.
This can be explained in terms of the capacity of media platforms to set the agenda of political campaigns with the view to influencing what the voters think about (Newton & Brynin, 2001), and also in terms media’s capacity to prime or frame political issues in a manner that will lead the audience to see the need for casting their vote on the election day (Gonzalez, 2012).
Through priming and framing of issues in Television channels, candidates are able to not only project an appealing image to the audience, but also create an emotional bond with viewers (Robertson et al., 2010, hence sustaining the audience’ desire to participate in elections.
This study has also demonstrated how religious orientation is critical to informing voter decision to participate in presidential elections.
However, as insinuated by Manza and Brooks (1997) that religious-oriented cleavages may have been a more fundamental factor for understanding the social roots of voter alignments than the class cleavage, the present study found both variables to be equally important in influencing voter participation in U.S. presidential election.
In social class, this study reinforces the findings of other previous studies (e.g., Guterbock, 1980; Southwell, 2004) that economically disadvantaged Americans are less likely to vote than their well-off counterparts.
The level of education has also being shown as a strong predictor to voting participation during the presidential elections, with findings demonstrating that eligible voters who have been unable to graduate from high school are less likely to vote than high school graduates and diploma/degree holders.
Although the voting pattern (3.5% for eligible high school dropouts to 80.5% for well-off, advanced degree holders) demonstrated by Cole and Sum (2012) has not been replicated in this study, the view that uneducated voters are less likely to participate in elections than more educated voters has been well reinforced.
There exist some limitations to the present study. First, the use of secondary data has brought difficulties in operationalizing some variables such as religion. The researcher had to rely on evaluating if respondents believed in life after death to determine their religious orientation.
However, common knowledge demonstrates that not all people who believe in life after death are religious and not all religious people believe in life after death. The case of missing data values also presented a challenge during data analysis. Additionally, it can be said that some of the variables used are limited in scope and therefore could not be relied upon in a rigorous scientific research.
Because presidential elections are a closely contested affair in the United States, it is imperative for policy makers and political players to know the factors that determine the participation of voters in the election.
Knowledge of such determinants (media, social economic status, religion, and level of education) will not only help in prioritizing campaign needs for political parties, but also in ensuring that effective strategies are employed to woo voters to participate in elections.
It should be remembered that presidential candidate Mitt Romney lost considerable number of votes due to poor understanding of religious orientation as an important underpinning in U.S. elections. It is therefore suggested that more research needs to be done to analyze the dynamics of these determinants and how they are played out in party politics.
References
Balnaves, M., & Caputi, P. (2001). Introduction to quantitative research methods: An investigative approach. Thousand Oaks, CA: Sage Publications.
Coley, R. J., & Sum, A. (2012). Fault lines in our democracy: Civic knowledge, voting behavior, and civic engagement in the United States. Web.
Gelman, A., Kenworthy, L., & Su, Y. S. (2010). Income inequality and partisan voting in the United States. Social Science Quarterly, 91(5), 1203-1219. Web.
General Social Survey. (2013). Web.
Gonzalez, M. A. (2012). Religion and the US presidency: Politics, the media, and religious identity. Political Theology, 13(5), 565-585.
Guterbock, T. M. (1980). Social class and voting choices in Middletown. Social Forces, 58(4), 1044-1056.
Holbrook, T. M., & McClurg, S. D. (2005). The mobilization of core supporters: Campaigns, turnout, and electoral composition in United States elections. American Journal of Political Science, 49(4), 689-703.
Manza, J., & Brooks, C. (1997). The religious factor in U.S. presidential elections, 1960-1992. AJS, 103(1), 38-81. Web.
Newman, B. I. (2001). An assessment of the 2000 US presidential election: A set of political marketing guidelines. Journal of Public Affairs, 1(3), 210-216.
Newton, K., & Brynin, M. (2001). The national press and party voting in the UK. Political Studies, 49(2), 265-285.
Pasek, J., Tahk, A., Lelkes, Y., Krosnick, J. A., Payne, B. K., Akhter, O., & Tompson, T. (2009). Determinants of turnout and candidate choice in the 2008 U.S. presidential election: Illuminating the impact of racial prejudice and other considerations. Public Opinion Quarterly, 73(5), 943-994. Web.
Robertson, S. P., Vatrapu, R. K., & Medina, R. (2010). Off the wall political discourse: Facebook use in the 2008 U.S, presidential election. Information Polity: The International Journal of Government & Democracy in the Information Age, 15(1/2), 11-31.
Southwell, P. (2004). Economic voting in volatile times. Journal of Political & Military Sociology, 32(2), 237-247.
Uppal, S., & LaRochelle-Cote, S. (2012). Factors associated with voting. Web.
Weisberg, H. F. (2007). The one thing you need to know about voting in American presidential elections. Web.
Wildavsky, A. (2011). Presidential elections: Strategies and structures of American politics. New York, NY: Rowman & Littlefield Publishers.