RELIGIOUS INFLUENCE ON VOTING BEHAVIOR OF MILLENNIALS By JULIAN OLIVIER BRAVO B.A., Metropolitan State University of Denver, 2012 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Arts Political Science Program 2015
2015 JULIAN OLIVIER BRAVO ALL RIGHTS RESERVED
ii This thesis for the Master of Arts degree by Julian Olivier Bravo has been approved for the Political Science Program by Michael Berry, Chair Tony Robinson Michael S. Cummings 04/24 /2015
iii Bravo, Julian Olivier (M.A., Political Science) Religious Influence on Voting Behavior of Millennials Thesis directed by Associate Professor Michael Berry ABSTRACT Millennials have different voting behavior than generations before them. In the 2008 and 2012 presidential elections they voted for the Democratic candidate in greater numbers than the Republican candidate. The purpose of this study is to figure out why Mi llennials were more likely to vote for the Democratic candidate. The factors examined include party identification, religion, race and income. It is found that Millennials have different voting behavior than other generations and it could be a result of va rious factors. Party affiliation has seen a shift toward Millennials remaining independent but preferring to register as a Democrat than a Republican. Millennials have fewer religious individuals than other generations and the amount of religious individua ls continues to decrease. Millennials are as likely to vote based off of racial identification as generations before them. Millennials are as likely as other generations to vote for a Republican candidate if their income is higher. The findings from this s tudy show that Millennials are different from the generations before them and it has influenced their voting behavior. The form and content of this abstract are approved. I recommend its publication. Approved: Michael Berry
iv ACKNOWLEDGEMENTS I would fir st like to thank my committee chair and advisor Michael Berry for his effort and professionalism throughout the process. Without Mike, this project never would have been e. have asked for a better mentor. stages of this project. She provide d me with the motivation I needed to work on the project from the start and made sure that I stuck to my task. When the project reached its most difficult stages, she helped me work through them and reach the end goal. She was a true friend from the start. Thanks Alejandrita for your part. project. We lost a family member during this stage and I was able to keep moving forward through her support and assistance. It is safe to say that I would not be near this stage without everything she has done. My efforts are a reflection of everything she has done as a person. Thanks for everything mom, you are the best.
v TABLE OF CONTENTS CHAPTER I. INTRODUCTION ................................ ................................ ................................ ........................ 1 II. RESEARCH TOPIC AND QUESTION ................................ ................................ ..................... 3 III. LITERATURE REVIEW ................................ ................................ ................................ .......... 6 Generational Differences at the Ballot Box ................................ ................................ ................. 6 Generational Lifetime Voting ................................ ................................ ................................ .... 10 Uniqueness of Millennials ................................ ................................ ................................ .......... 12 Millennials and the Media ................................ ................................ ................................ .......... 14 Individual Influences on Voting Behavior ................................ ................................ ................. 1 8 Religious Influence on Individuals ................................ ................................ ............................ 22 Racial Influence on Individuals ................................ ................................ ................................ .. 26 Social and Economic Influence on Individuals ................................ ................................ .......... 28 IV. EXPLANATORY FRAMEWORK AND HYPOTHESIS ................................ ...................... 31 V. DATA SAMPLE AND MEASUREMENT ................................ ................................ .............. 33 VI. METHODOLOGY ................................ ................................ ................................ .................. 39 VII. 2012 PRESIDENTIAL ELECTION ................................ ................................ ...................... 4 1 Party Registration in 2012 ................................ ................................ ................................ .......... 4 1 Presidential Vote in 2012 ................................ ................................ ................................ ........... 4 3 Religion in 2012 ................................ ................................ ................................ ......................... 4 6
vi Income in 2012 ................................ ................................ ................................ ........................... 5 1 Race in 2012 ................................ ................................ ................................ ............................... 5 3 VI I I. 2008 ELECTION ................................ ................................ ................................ .................. 5 8 Party Registration in 2008 ................................ ................................ ................................ .......... 5 8 Presidential Vote in 2008 ................................ ................................ ................................ ........... 60 Religion in 2008 ................................ ................................ ................................ ......................... 6 1 Income in 2008 ................................ ................................ ................................ ........................... 6 6 Race in 2008 ................................ ................................ ................................ ............................... 6 8 IX. DISCUSSION ................................ ................................ ................................ .......................... 7 3 X. CONCLUSION ................................ ................................ ................................ ......................... 7 9 REFERENCES ................................ ................................ ................................ .............................. 8 3
vii LIST OF TABLES TABLE 1.1 2012 Party Registration Vote (ANES) ................................ ................................ ..................... 4 1 1.2 2012 Millennial Party Registration and Vote Choice (ANES) ................................ .............. 4 2 2.1 2012 Presidential Vote (ANES) ................................ ................................ ............................ 4 4 2.2 Logit Regression Model of 2012 Presidential Vote (ANES) ................................ ................. 4 5 2.3 Logit Regression Model of 2012 Presidential Vote (ANES) ................................ ................ 4 5 3.1 Evangelical Vote (ANES) ................................ ................................ ................................ ..... 4 6 3.2 Evangelical Vote (CNN Exit Polls) ................................ ................................ ....................... 4 7 3.3 Logit Regression of Presidential Vote (ANES) ................................ ................................ ..... 4 7 3.4 Religion and the 2012 Presidential Vote (ANES) ................................ ................................ 4 8 4.1 2012 Racial Vote (ANES) ................................ ................................ ................................ ..... 5 4 4.2 Regression Model of 2012 Obama Vote Choice ................................ ................................ ... 5 7 5.1 Party Identification and Vote Choice (ANES) ................................ ................................ ...... 5 8 6.1 2008 Presidential Vote (Pew) ................................ ................................ ................................ 60 7.1 Religion and 2008 Presidential Vote (ANES) ................................ ................................ ....... 6 2 7.2 Evangelicals and Presidential Vote (CNN Exit Poll) ................................ ............................ 6 3 8.1 2008 Racial Vote (ANES) ................................ ................................ ................................ ..... 70
viii LIST OF FIGURES FIGURE 1.1 2012 Presidential Vote Percent by Protestantism (CNN) ................................ ...................... 4 9 2.1 2012 Presidential Vote Percent by Income (CNN) ................................ ................................ 5 2 3.1 2012 Presidential Vote Percent by Race (CNN) ................................ ................................ ... 5 5 4.1 2008 Presidential Vote Percent by Protestantism (CNN) ................................ ...................... 6 5 5.1 2008 Presidential Vote Percent by Income (CNN) ................................ ............................... 6 6 6.1 2008 Presidential Vote Percent by Race (CNN) ................................ ................................ .... 71
1 CHAPTER I INTRODUCTION This study seeks to examine factors that affected voting behavior in the 2008 and 2012 elections with a focus on generational cohort effects. Of particular interest is the Millennial generation, which is defined as includin g all individuals born between 1981 through 1994 (Pew Research 2012). 1 While generational differences in voting behavior are expected, this study seeks to explain why the voting behavior of Millennials would differ from other generations. Possible reasons for that include religion, race, income, ideology and party affiliation. Religiosity is the primary focus of this aspect of the study since the new generation of voters is less religious than earlier generations (Pew Research 2012). In short, the analysis presented in this study seeks to test for differences in generational voting behavior and provide explanations for patterns that exist. The study will begin with a breakdown of the 2008 and 2012 presidential elections using data from the American National Election Study (ANES) as well as exit poll data from CNN and Pew. A number of individual level factors are examined as potential influences on voting behavior. This will look at what defines each generation and how generational characteristics might have a ffected both elections. The study will then focus on the new generation, voting behavior and religious influence on voters. By breaking down these categories, the study will present a framework for possible explanations to existing patterns found in the st udy and give a hypothesis for expected results. The study will then analyze data from various sources to see if any variable can be determined to have a significant impact on voters. 1 The 2008 e lection will only represent Millennials born through 1990 since Millennials born after were not old enough to vote. The 2012 election will represent the full Millennial demographic.
2 This subject is worth studying because the uniqueness of the Millennial g eneration could change the landscape of politics in the future. A growing voter base that is secular and socially liberal could benefit more socially liberal candidates. Focusing on the Millennial generation adds to the study of voting behavior as a whole because it might also lead to a trend in how the next generation might vote. This study will contribute to the political science literature in the area by examining voting behavior of the Millennial generation in the two most recent presidential cycles. T he expected results should be in line with earlier research showing that Millennials are more likely to vote for Democratic presidential candidates (Pew Research 2012). The purpose of this study is to test whether the primary reason for that is because Mil lennials tend to be less religious than other generations and that impacts their decision to vote for the Democratic presidential candidate. The implications of the study include finding out how people are likely to vote given their religion and age. If re ligion does play a significant role in voting behavior then it would explain how candidates might try appealing to voters in the future. With religious identification on a decline, the study examines the possibility that candidates might become more libera l in social issues to appeal to the Millennial generation. If the study concludes that there is an increasing amount of Millennials voting for the Democratic candidate then the Democratic Party might do more to continue winning support from Millennials.
3 CHAPTER II RESEARCH TOPIC AND QUESTION Voting behavior is a major topic within the field of American politics that offers an in depth view of how and why people choose their elected officials. This is important because it helps scholars better understand how people vote in presidential elections and how they might vote in future elections. Studies in this area have explored the mindset of the voters and what issues are important when it comes to their decision making process. A wealth of existing research has focused on people and how they choose to vote based on important issues at the time (Bartles 2000 Brooks and Manza 1997, Coleman 2004 ). This study continues in that tradition by focusing on the 2008 and 2012 elections. The research question asked in this behavior differ from earlier generations? By many accounts, Millennials appear to be more socially liberal compared to the generations before them (Pew Research 2012). If that is true then we would expe ct that members of the Millennial cohort would be more likely to vote for Democratic presidential candidates. Beyond ideology, the religiosity of a person is expected to strongly influence how one votes. Previous studies have shown that the Millennial gene ration is the least religious generation in the history of the United States and religious affiliation is often connected to social ideals (Pew Research 2012). Millennials are the least religiously involved generation and there could be a possible connecti on to the lack of religion and the decision to vote for the Democratic candidate. The questions that need to be addressed are, if a lack of religion does lead to more liberally based social ideas, and, if that leads to a preference for the Democratic candi date.
4 It is possible, however, that these assumptions of each generation might be false and that decisions to vote for either candidate. Millennials could pr efer more economically liberal policies as well and that could influence their vote as much as their preference to socially liberal policies. The same could be true for non Millennials who might favor socially conservative policies that influence their beh avior in a more significant way than economic issues. This study looks to first breakdown the generations and the uniqueness of the Millennial generation when compared to the generations before it. If it can be determined that Millennials are significantly different than the generations before them, then the study finds a basis for why they might have different voting behavior than other generations. One argument this study makes is that Millennials are substantially different than generations before them. The uniqueness of their ideals are expected to play a major role in how they chose to vote during the 2008 and 2012 elections when they had more influence on the outcome. The next part of the study focuses on generational voting behavior. Traditionally gen erations vote in different manners and breaking down how they voted could serve to find how each generation was different (Pew Research 2012). By determining the difference in voting behavior, the study could focus on what issues are most important to each generation. This could also find just how different the Millennial generation is compared to the earlier generations. Because the Millennial generation is less religious than other generations, the analysis will then look at how much of a role religion pl ays in shaping voting behavior. The assumption is that religion is among the greatest determining factors on candidate selection. It is expected that members of religions that align more with the Republican agenda are more likely to vote for the Republican candidate. It is also expected that nonreligious people will be significantly more
5 likely to vote for the Democratic candidate because they are less likely to share the conservative social values of the Republican candidate.
6 CHAPTER III LITERATURE REVIEW As referenced in the introduction, this study seeks to assess generational differences in voting behavior and provide an explanation for any differences that are found. Thus, an important first step is to simply determine whether Millennials differ in thei r voting behavior relative to other generations. Of particular interest is the effect of religiosity as a casual factor that may help explain differences that exist. In order to provide such an analysis, generational cohorts and groups must be examined to determine their voting priorities. Before doing so, however, it is important to first access the state of existing research on this topic. The literature review proceeds in three parts. First discussed are studies that have tested for generational differe nces in voting behavior. This section is followed by one that examines the uniqueness of the Millennial generation. While many differences exist, not all are expected to affect the political actions of Millennials. The final section surveys the primary con clusions from individual level studies of voting behavior. Among the most important factors from this are race, income, ideology and party affiliation. Generational Differences at the Ballot Box There are many differences between Millennials and other generations in how information is obtained and how that impacts voting behavior. A study by Poole and Palfrey shows that a more informed generation is more likely to vote, but Millennials are not voting at the same rate as other generations (1987). Consid ine, Horton and, Moorman found that Millennials were obtaining more political information than generations before them, but still did not have a strong voter turnout (2009).
7 While studies have shown that people with more information are more likely to vote this has not been the case with Millennials. Voter predictability seeks to find why Millennials are not voting in higher numbers if they do get more information from the Internet than other generations. Millennials are expected to have predictable voting behavior. Millennials are more likely to vote for a Democratic presidential candidate which might mean that they are more informed if this voting pattern is predictable. The other assumption is that Millennials are less informed than other generations and therefore, are less likely to take part in elections. These are two assumptions that can be made about Millennials in connection to the Palfrey and Poole study (1987). While Millennials use the Internet as their primary source, it does not mean that they are more informed than other generations. The voting patterns of other generations are less predictable than that of Millennials, but they also show higher voter turnout. It would be worth noting that Millennials contain the highest number of unaffiliated voters out of any generation (Pew Research 2012). Unaffiliated voters tended to vote more for the Democratic candidate in the 2008 presidential election (Pew Research 2012). Another connection to the Palfrey and Poole study is that Millennials are not wel l informed because they align themselves with a party could show uncertainty. The perception that Millennials remain apathetic could be true. Johnson and Kaye found that they are participating in other areas of politics, but that might not translate to more informed voters (2003). Wells and Dudash ask how youth participation affects presidential elections (2007) It is unknown how much influence younger generati ons can have on election outcomes because their voting pattern is somewhat unpredictable. Riker and Ordeshook (1968) discussed how the importance of elections might influence voters to take part if they feel the outcome is in doubt
8 and that might have impa cted the results of the 2008 election, but still does not explain why turnout decreased for the 2012. The predictability of Millennial voting behavior is difficult because they are obtaining information at greater levels than other generations but still ch oosing not to vote. Con sidine, Horton and Moorman (2009) find that the amount of information obtained through the Internet from Millennials is significantly greater than other generations, but Wells and Dudash (2007) question the reliability of that inform ation. Palfrey and Poole found that more knowledgeable voters were more likely to vote, but this does not explain why it was not the case for the 2004 election (1987) it could lead to them opting not to vote rather regardless of how much information they have on a candidate. Wells and Dudash (2007) come to the same conclusion from Considine, Horton and Moorman that the Internet is the primary source of information for the younger generation (2009) While this is important, Wells and Dudash also find that they will discuss information with their frie nds and family which also helps shape their beliefs (2007) The study also finds that the much of the younger generation discusses political issues on a daily basis (Wells and Dudash 2007). A problem that came up was the reliability of the information that they received. The younger generation acknowledged that the information it receives from the media and Internet is a good starting point, but it needs to be further explored to uncover more credible information (Wells and Dudash 2007). Previous generation s had different means of receiving their information about politics and elections, but the Millennial generation focuses on the Internet while also commonly
9 receiving information from different media including television, advertising and film. This diverse media landscape has a significant influence on the ideology of Millennials by providing an abundance of information from differing political perspectives (Considine, Horton and Moorman 2009). There has been an increase in the amount of people that are soc ially liberal and it is a result of a more open and accepting media portrayal of formerly controversial issues like gay rights (Considine, Horton and Moorman 2009). The findings from Wells and Dudash were in line with Palfrey and Poole in the sense that th e younger generation feels that they are more likely to take part in elections if they are better informed (2007). Riker and Ordeshook discussed the concept of civic duty which seems less prevalen t in the younger generation (196 8). The study by Wells and D udash did find a few cases in which learning more lead to a decrease in participation (Wells and Dudash 2007). The more the younger generation found out about candidates, the less likely they were to vote for them because they felt disillusioned by the can didate. The participants in the study believe that they either know a significant amount about politics or want to know more about politics. The Galatas and Pressley study did question their results when they found that there were respondents with a low se nse of civic duty, and felt that the outcome was not close, but decided to vote anyway (2010). There were also those with a high sense of civic duty, and felt that the outcome was close, but decided not to vote. These go counter to the conclusions of the s tudy, but they are seen as isolated circumstances (Riker and Ordeshook 1968). The Millennial generation had a lower turnout rate than the generations before it and this could be connected to the assumption that it is apathetic, b ut it could also be because Millennials felt that the election was not in doubt. The 2008 and 2012 elections both had potential for closeness, but that did not
10 Millennials generally lack politi closeness will lead to an increase in voter turnout. Millennials are perceived to have the same qualities as those that participated in the Wells and Dudash study. Low voter turnout in the 201 2 election could be explained by disillusioned rather than lack of knowledge. Millennials assume that they have a significant amount of are more likely to take pa rt in other areas of politics. Millennials are actually highly engaged in discussing politics and gathering information, but that does not lead to an increase in voter turnout. Millennials are different from other generations because they are better infor med but still turnout in smaller numbers. Part of this could be their unpredictability or because of disillusionment. The next part will examine the voting behavior of other generations. Generational Lifetime Voting Millennials have a low voter turnout bu t it could be a result of a few reasons. One reason could be that each generation is different and there is a general decline in party affiliation (Abramson 1976). The other reason might be that generations like Generation X and Baby Boomers, start off sl owly but participate more as they age (Lyons and Alexander 2003). A study conducted by Abramson looks at the connection between age and partisanship voting from 1952 to 1974. One of the major findings was that weak party loyalty from young adults was not a result of their youth but rather that the generation as a whole was less likely to identify with a party (Abramson 1976). The study determined that there was not an increase in party loyalty as each generation got older but rather that the percentage sta yed the relatively similar throughout the 22 year cycle of the study. This connects to Millennials because it shows
11 that they are less likely to identify with either party compared to generations before them. This is not a result from this generation being younger, but rather it is because, as a whole, it is just less likely to identify with either party. A study by Lyons and Alexander finds evidence that both supports and contradicts the Abramson study (2003). A significant finding was that there has been a decrease in party loyalty within each generation beginning in 1932 (Lyons and Alexander 2003). A different finding from the study was that generational cohorts are slower to enter the electorate, but they eventually catch up by the age of 50. This contra dicts the Abramson study because it shows that generational cohorts eventually show greater party loyalty as they age. Since Millennials show weak party loyalty, Lyons and Alexander argue that it is because of their youth and not their generation. A 2011 election study conducted by Pew looks at the differences in voting behavior over typically voted for the Democrat throughout their entire lifespan showing that they did not become more conservative. One reason for this could be the political realignment that saw the Democrats eventually shift to being more liberal and Republicans being more conservative as a result of the New Deal Coalition. 83 in 2011) was initially more likely to vote Republican but saw a shift halfway through their life. They eventually shifted back to being more likely to vote Republican and had the most Republican voters of any gene ration in recent elections (Pew Research 2011). 68 in 2011) had varying results based on their age. The younger Boomers (49 54) voted Republican more than the middle
12 (55 60) or older Boomers (60 6 9) in every presidential election. The young and middle Boomers both started out voting more Republican, but the middle Boomers started voting Democrat more often as they aged. The older Boomers began voting more Democrat but eventually became more likely to vote Republican as they aged. This shows differences within the Boomer generation. 48 in 2011) was more likely to vote Democrat in their first elections but there were also differences within the gener ation. Older Gen Xers (37 48) have voted more Republican than other generations in every election since 1996. The younger Gen Xers (29 36) voted more Democrat in every election since 2004. Within the generation, there are significant differences. Older Gen Xers are more likely to vote Republican than younger Gen Xers. Millennials have always been more likely to vote Democrat. In every election since they have entered the electorate, they have been one of the highest Democrat voting generations. The results are mixed as Boomers and Gen Xers conflict within the group but, as a whole, most of the generations became more likely to vote Republican as they aged. The older generations have been more likely to vote Republican in the last few elections but that is a more aged (Pew Research 2011) Next, the study will look at how Millennials are different through their expectations, work environment, focus on technology and civic engagement to help give a basic understanding of Millennials before examining their voting behavior. Uniqueness of Millennials Many studies have examined ways in which Millennials are unique. These areas have spanned a broad number of topics including economics (Ng, Schweitzer and Lyons 2012) and
13 media consumption (Considine, Horton and Moorman 2009). In this study, it is expected that Millennials will have radically different voting behavior from other generations because they grew up in a completely different environment. The important differences in Millennials include their expectations, work environment, focus on technology an d civic engagement. The study finds that Millennials are significantly different in these areas compared to other generations. The important parts of understanding Millennials are their expectations, work environment, focus on technology and their civic en gagement. With different expectations of the economy and job market than other generations, Millennials have different expectations for their job success. With that they also expect things to be more comfortable for them as a sacrifice for more performance based jobs that lack that security. This differs from other generations, but this does not lead to apathy among the generation. Millennials can see things as bleak at times, but also use technological innovation as their avenue for political participation They are not less likely to take part than other generations, but engage in other manners online. These differences could be an explanation for the uniqueness of Millennials to other generations in political attitude. The Millennial generation has differ ent attitudes toward the country, in terms of career expectations, than generations that preceded it (Ng, Schweitzer and Lyons 2012). One of the most notable areas of focus is on the assessment of the job market. Previous generations expected that there wo uld be a career available for them upon completion of college but Millennials see career opportunities as limited. Economic factors play significantly in voting behavior and the attitude of each generation could play a significant impact by determining how beneficial it is to vote.
14 Schweitzer and Lyons 2012). They seek jobs that offer them a chance for an increase in pay and advancement in their employment. Part of t hat is the understanding that they might not be able to have success upon graduation. Millennials have the realistic expectation that finding ideal jobs is not attainable early on in their career (Ng, Schweitzer and Lyons 2012). Such economic consideration s may filter into the voting booth because the economic situation plays a role in determining how they will vote (Ng, Schweitzer and Lyons 2012). If the economy is struggling and opportunities are limited then Millennials are more likely to vote for someon e that they feel would be better suited in creating better opportunities. Other things Millennials seek include working in a comfortable environment. The Millennial generation is attached to the idea of being nurtured and they want a comfortable and nurtu ring job environment. Ng, Schweitzer and Lyons also focus on whether work expectations and priorities differ in the Millennial generation (2012) Their results support the notion that Millennials have different expectations or priorities among different de mographics. A few secondary findings included their preference to base pay instead of work incentives and flexible spending instead of long term financial investment (Ng, Schweitzer and Lyons 2012). This may also connect to voting behavior because Millenni als expect guaranteed security over future investment. Millennials tend to see things in a more short term model instead of long term solutions. That could play a role in the manner in which they choose to vote. Millennials and the Media A misconception a bout Millennials is that they are less likely to be politically engaged thirds claim that their involvement in politics has increased or greatly increased since they firs t became
15 would be more able to engage politically than during the time this study was conducted. Having grown up in an age of technological innovation, Millennia ls are predisposed to utilizing the Internet as the primary source of information when compared to earlier generations (Considine, Horton and Moorman 2009). This has led to an increase in interest in politics. This has also led to an increase in voting, po litical interest and campaign interest. According to Johnson and Kaye, the Internet is significantly influencing election s in new ways (2003). People were more heavily reliant on the Internet to gather political information during the 1996 and 2000 electio ns. It was found that people that people are growing increasingly more reliant on the Internet as a means of gathering political information and engaging in political activity (Johnson and Kaye 2003). There has been a significant increase in the amount of people that get political information and engage politically online. This increase obtaining information (Considine, Horton and Moorman 2009). The connection thi s has to Millennials is the attitude they have for voting. If media does have as much influence as Ng, Schweitzer and Lyons (2010) and Considine, Horton and Moorman (2009) noted, then Millennials might be even less likely to vote in general. By gathering i nformation from the media that projects a one sided election, Millennials may feel that their vote does not have a significant impact and they might choose not to vote. Galatas and Pressley discussed the significance of engaging the Millennial generation i n civic participation and political education. A study by Urbatsch shows that Millennials tend to lack political interest them more likely to vote in general.
16 one of the lowest voter turnout among all generations. The abundance of information available on the Internet is seen as a positive in advancing the political interests of Millennials. Other outlets can be costlier or filtered and that limits the access to those sources. Scholars believe that this abundance of information leads to a more level play ing field as more vantage points are represented while limiting their censorship (Johnson and Kaye 2003). The hypothesis that Millennials are significantly more liberal than other generations could be connected to their access to the Internet. Considine, H orton and Moorman discusses how Millennials use the Internet to reinforce their beliefs about the state of the country and that could lead to different perceptions of the shape of the country. If a Millennial feels that the country is in poor shape then th ey might utilize the Internet to justify their belief. Their increased opportunity to engage civically could also help reinforce their beliefs if they are practiced more often. If there is increased opportunity to engage civically, the religious beliefs of Millennials could be important in determining how they will choose to engage. Like all generations, Millennials do consider economic factors to be important in their decision making process. The difference with Millennials and other generations throughou t these elections was that economic issues were not see n as important as social issues (Considine, Horton and Moorman 2009) A main focus of this study is how different the Millennial generation is compared to other generations. In order to appeal to this generation, there needs to be a shift in the way news is delivered. Politicians spend a significant amount of money on media and it has the ability to sway voters (Considine, Horton and Moorman 2009). The Considine, Horton and Moorman study has a connectio n to the Ng, Schweitzer and Lyons study in the sense that the information
17 Millennials gathered from the media, about the job market, could influence their perception of the job market. Millennials have been told that the job market is difficult at the mome nt and they have allowed that to influence their hopes for the future. Palfrey and Poole discuss the relationship between information, ideology and voting behavior. The study finds that people that are more informed in elections are more extreme in their political alignment and are more likely to vote in elections than those that are not as well informed (Palfrey and Poole 1987). Palfrey and Poole also find that more informed voters are more predictable in their voting patterns. Much of this can be connect ed to Millennials through how informed their generation is compared to other generations. Considine, Horton and Moorman focused on how much Millennials use the Internet as their source of gathering information and use it as a means of civic participation. If Millennials are more informed than the average person then they would be more likely to vote. The issue with this is how many Millennials are seeking out information to become more informed with elections. An important part of the Millennial generation is its civic participation. As Considine Horton and Moorman mention, there are efforts to find different methods of teaching Millennials and shifting their attitude toward civic participation. Galatas and Pressley discuss the role of civic engagement with the Millennial generation. According to the study there has been a on academia to encourage it. Millennials seem disinterested in civic participation that lacks complexity, but are more willing to engage if they are engaged in problem solving situations (Galatas and Pressley 2010). The Galatas and Pressley study supports the conclusions found in the Considine, Horton and Moorman study by focusing on media influen ce. The idea of disenchantment with politics is one of the reasons there is a declining amount of civic
18 participation in younger generations and it can be connected to the negativity developed through the media. Millennials are unique to other generations through their means of obtaining political information and the media (Considine, Horton and Moorman 2009). Individual variables also influences voting behavior. Individual Influences on Voting Behavior Conformity can influence party affiliation (Coleman 2004). The voting behavior of Millennials seems to shift based off of trends. Generations as a whole experience political waves that influence how they vote in each election, but Millennials are one of th e more unpredictable generations. They are not registering with either party in greater numbers than other generations and the independent vote seems to shift based off of trends. Voting conformity and partisan voting are both areas that might explain the changes trends among Millennials and other generations. Election results have shifted in the last few presidential election cycles with the Democratic Party winning the White House twice. In order to find how much religion affects Millennial voting, voting the intellectual aspect of voting. The major conclusion Urbatch has is that people with higher cognitive capacities are more likely to vote on personal interest when compared to others. Personal interest may include issues like gay rights or education funding, singular issues that have a direct impact on someone that might be homosexual or someone attending school. The other part is that people disinterested in politics do not see their vote as an expression of themselves. People that are interested in politics tend to turn out and vote regardless of whether they feel their vote might have a significant impact on the outcome. This is different from the
19 rationality of people with higher co gnitive capacities because they should understand that the importance would be diminished if it will not have any implications (Urbatsch 2012). (Coleman 2004). A study b y Coleman focuses on the idea of voting in conformity and that people will vote in elections if they feel that everyone else is voting. The study also discusses how conformist behavior can also influence how the individual votes. The connection to the 2008 and 2012 elections is that there was significant support for Barack Obama and it is possible that conformist behavior from the people led to increased support for Obama. Voter turnout in 2008 increased among the voting eligible population compared to 2004 and is possibly explained by the conformity of the support for Obama as well as the expectation to vote in that election. The 2012 election saw a significant decrease in voter turnout among the eligible voting population that might be explained by less co nformist expectations and support for either candidate. might be connected to the idea of conformity and less to the idea of religious affiliation. s conducted before the 2008 and 2012 elections and it determined that there was a possible connection to voting conformity and election outcomes, but he also acknowledged that more research needed to be conducted to see the significance of voting conformit y (Coleman could be a result of conformity as well. By having people become more accepting of socially liberal values through conformity, then it could potent ially lead people to more likely vote for the Democratic candidate.
20 A connection between conformity and partisanship voting might be explained through the studies from Coleman (2004) and Bartles (2000) Coleman found that voters shift their voting behavior based off of trends. It is possible that a shifting social trend could lead to an increase in voters registering with one party over another. Layman found that there were changes in the role of religion in the parties that led to a shift in party registra tion (1997) Bartles also found that partisan voting has increased significantly in the last few election cycles meaning people are more likely to vote for someone in the same party than ever before (2000) Millennials are more likely to register as Democr ats, something that could explain the difference in voting behavior between them and non Millennials. A further avenue of study may test whether there is conformity within the belief system of the individual. Socially liberal issues have gained support wit hin the country as they become more socially acceptable. There is a potential connection that it is a result of conformity and that the lack of religion among Millennials might also be a result of conformity. The expectations of the 2016 election could sho w how much impact conformity has on the election. This study believes that the Democratic candidate should garner even greater support because there will be more Millennials eligible to vote and they will be less religious. If there is a swing in favor of the landscape plays a bigger factor than religious identification. One possible issue within this study is that the connection might not be with the person but with thei r political affiliation. A person might have chosen to vote for Obama in 2008 because there was a greater turnout meaning more Democrats came out to vote for him. The 2012 election saw a decrease in voter turnout when fewer Democrats turned out to vote for
21 more supporters for Democrats in a particular presidential election leading to a greater margin of victory for the Democratic candidate. The result would be a n egative correlation between party entropy and turnout, and partisanship would be the explanation, not c onformity 80). One aspect of voting behavior is the idea of partisanship voting within the eligible voting population. Voters are becomin g more likely to vote for the presidential candidate that is of the same party (Bartles 2000). During Bartles examination he noticed that these levels have greatly a in which parties in government seem increasingly consequential, the public may increasingly come to develop and apply partisan predispositions of exactly the sort described by the authors of, The American Voter predispositions that lead to conclusions about what each party represents and that is one explanation for why they choose to align with a party and vote for them in presidential elections. Millennials have found themselves more likely to align with the Dem ocratic Party, but registration within either party is not high. Millennials often choose to stay independent. Bartles broke down voters into weak, strong and leaning partisan categories (2000). Of course, strong partisans are much more likely to vote for where Millennials stand could be an important indicator of how they vote. With more Millennials registering as a Democrat, it needs to be determined how loyal they are to their party. Most Millennials stay independent and it is also important to find which party they lean toward. Millennial generation are significantly more likely to vote for the Democratic candidate in each p residential election and that there are significantly more Democratic leaning independents
22 within the generation. Republican registered Millennials are just as likely to remain loyal to their party, but it is expected that there would be significantly less Republican leaning Millennials than those that lean Democrat. Conformity does influence political identification (Coleman 2004). Religion is another variable that can influence party identification. Religious Influence on Individuals Little existing rese arch directly connects Millennials and religion as influences on voting behavior. Finding such disparity between religious affiliation in Millennials and other generations is the primary reason for this study to see whether it is influencing the changes in voting behavior. The important components found in this section include the religious affiliation religion of a candidate could potentially impact voters. A Pew st udy found that there was a trend between Millennials becoming less politically affiliated while seeing an increase in the number of religiously unaffiliated people (Drake 2014). This demonstrates that religion might play a role in different voting patterns for Millennials compared to other generations. By examining the overall impact of religion influence on voting, the study will seek to explain how much the decrease in religiously affiliated Millennials might influence their voting behavior. A study by L ayman examines the influence of religion on political behavior from 1980 to 1994. The study demonstrates how voting behavior changes throughout time by examining how different religions have changed their voting patterns in different election cycles. Tradi tionally conservative religions, like Protestants, were more likely to vote Republican throughout this study while traditionally liberal religions, like Jews, were more like to vote for a Democrat. What
23 the study also examines is the distinction between do ctrinal conservatives and doctrinal liberals. Doctrinal conservatives are more likely to align themselves with the Republican Party and vote for the Republican candidate while doctrinal liberals are not more likely to align themselves with the Democratic P arty or vote for the Democratic candidate (Layman 1997). A study by Brooks and Manza shows that just 1% of voters identified as having no religion in 1960 while that number increased to 12% of the voting population in 1992. The shift came from liberal Prot estants (25% to 14%) and moderate Protestants (24% to 17%) (1997). Depending on how the moderate Protestants voted, it is possible that they became more likely to vote for a Democrat due to their secular shift. Assuming this trend continued, there would be even fewer liberal and moderate Protestants today while the amount of nonreligious cleavages grew. This would be significant if the hypothesis that nonreligious people are more likely to vote for the Democratic candidate is true. A 2006 study of religious influences on the 2004 presidential election discusses how it helped secure a Republican victory. The general consensus is that religion plays a significant role in election outcomes, but Guth believes that religion had a greater role in determining the 2 004 day Saints, and traditionalists from all major (and most minor) religious traditions, with the bulk of their remaining support coming from centrists in the three while the Democrats were more likely to rely on religious minorities and secular citizens. Considering the increasing amount of religious minorities within the Millennial generation, it is possible to explain why the Democrats saw an increase in votes during the elections that followed.
24 A study by Layman found that religion does play a role in voting trends (1997). People that are disillusioned by religious ideals could spark social trends that influence partisanship. If a trend does become big enough, a potential for a bandwagon effect might causes people to vote with the popular norm. Examples are gay rights and abortion in wh ich the social trend might cause people to vote against a religious ideal. Social issues often influence trends and can influence conformity among voters. Another conclusion from this study is how devoutness of a religious sect also influences an individua compared to less doctrinal sects of Protestantism. Democrats saw an increase in their an 1997). Millennials are the least doctrinal generation and that could explain the decrease in alignment with the Republican Party. Millennials are more likely to stay secularist or religiously liberal making them more likely to align with Democrats. Laym may have indirect effects on partisanship and the vote, but their infl uence may be exerted loyalty to traditional conservatism or traditional liberalism might lead to voting for a third party. These questions influence the result of t his study because the Millennial generation has attached also be influenced by outside variables that are connected to their religious affiliation. Mitt Ro
25 focus is on Evangelical candidates and how the public views them. The important factor is how the public perceives them and the average voter often sees them as more ideologically extreme. Evangelical candidates will attempt to appeal to the average voter by mo ving closer to the center with a more moderate platform, but the stereotypes leave voters with a preconceived notion of The preconceived notion of Evangelical candidates being more conservative can be positive an d negative for the candidate. Generally it leads to a disadvantage for these candidates among average voters. Millennials tend to consider themselves to be less spiritual, and more socially liberal (McDermott 2009). Evangelical candidates typically hold st rong conservative beliefs, something that does not align with the average Millennial voter. Millennials feel less connected to Evangelical candidates regardless of how they try to appeal to the Millennial demographic (McDermott 2009). This study demonstrat es that the media and technological stereotype of Evangelicals that are not always in line with their political ideology, but it can be a significant way that an a (McDermott 2009). Republicans have been successful in convincing Evangelical voters to align with their party, but have been unsuccessful with racial minorities. African American and Latino Evangelicals tend to hold the same social beliefs as white Evangelicals, but they ha ve not followed the same political appeal of the Republican Party (McDaniel and Ellison 2008). According to McDaniel and Ellison, there has been limited movement from racial minority Evangelicals in the shift to conservative social policies. The study show s that they were more
26 likely to join the Democratic Party or stay unpartisan. The Republican Party supports socially conservative policies that are often held by ethnic minorities, but it seems that other issues are influencing their voting behavior (McDan iel and Ellison 2008). important factor in the way people vote. Traditionally conservatives tend to vote earlier and are less likely to place more emphasis on a candida connection to McDermott is that it goes against the idea that voters can see candidates, primarily challengers, as l ess relatable. There is still evidence that they take personality into consideration, but it is not the primary factor in their decision making. vote. Evangelical can didates are seen as less moderate while being stereotyped with more conservative social ideals. This alienates groups of voters of different races and moderates because they are unable to relate to them. A candidate can appeal to voters that share similar religious beliefs because they feel that their positions on social issues are similar. The demographic of Millennials is different than other generations meaning that there is a more diverse demographic of voters. This could be an explanation for the uniqu e voting behavior of Millennials. R acial Influence on Individuals There is a significant counter argument on how much religion actually influences an might influe nce such as political ideology. Patrikios discusses other political variables that influence voting behavior including church attendance, party identification, and ideology.
27 Patrikios agrees with McDaniel that conservatives have been successful in appealin g to Evangelicals, but things like church attendance also have a significant role in voting behavior (2008). Patrikios notes that normal church attendance leads to practicing greater conservatism. This makes normal church attendees more likely to fall in l ine with social beliefs of the Republican Party (Patrikios 2008). Many white Evangelicals come from the same poor economic situation as ethnic Evangelicals, but they still primarily choose to vote more conservative. This is because they relate more to the Republican Party on social issues. The opposite effect is seen in racial minorities as they are more attracted to liberal policies that tend to favor their economic situation (McDaniel and Ellison 2008). Ethnic Evangelicals tend to have the same preference s on social issues as white Evangelicals, but they weigh the economic factors as being more significant (McDaniel and Ellison 2008). Evangelical candidates are having issues appealing to a range of demographics. McDermott noted the difficulty of appealing to the average voter and McDaniel notes how they have difficulty connecting to the ethnic voters. While religion seems to be an important indicator in voting behavior, race also has a significant influence. Patrikios further argues that politics and relig ion could be linked through reciprocal causation. Religious voters could lean more conservative because they are more likely to represent ideas taught within the church (Patrikios 2008). This begs the question of how much religion affects voting behavior. McDaniel believes that the stereotype of Evangelicals voters being more socially conservative is enough to influence voting behavior, but McDermott and Patrikios argue that there are other factors at play. Race and church attendance are two leading contrib utors in how people choose to vote in elections. Since Millennials are less religious and
28 less likely to attend church as a whole, these might not be as influential as they had been for earlier generations. The religious influence on voting behavior seems to center around race and religion according to the studies conducted by McDaniel and Ellison (2008), Guth (2006) and Patrikos religion of the candidate can have an i mpact on voting behavior. Millennials are perceived to be more likely to register Democrat than Republican, but are more likely to stay independent. This could be explained by either disenfranchisement with the Republican candidate or a decrease in the amo unt of Millennials that are Protestant compared to non Millennials. Social and Economic Influence on Individuals Another topic considered in the Guth study is that religious organizations differ on foreign policy and economic issues. Social issues are a ma jor determining factor in how religious groups vote, but the difference in foreign policy and economic issues also play a role in which candidate they prefer. An expectation was that Republicans would continue trying to appeal to their traditional voter ba se, but also attempt to appeal more to religious minorities (Guth 2006). It was also expected that Democrats would continue expanding their appeal to religious groups. Depending on the results of this study, it can be determined how successful each party w as in appealing to religious minorities and their traditional base. Social issues play a significant role in trends, but economic issues are the other important aspect people consider when voting. The economic state of the country is important in determini ng how people might vote and if it has the same influence as social trends on Millennials.
29 A study by Markus discusses the importance of economics on voting behavior. The study focuses primarily on the economic state of the United States and ho w that might influence the outcome of presidential elections. Through the study Markus identified party identification and race as two of the most influential indicators of how an individual will vote. Economic status of the country does have an impact on support for an incumbent running for reelection (Markus 1988). If there was a Republican incumbent that presided over a state of declining or stagnant economic productivity, they would expect to be defeated at the polls. President Jimmy Carter presided ove r a period of economic decline and was defeated in his run for reelection. A lot of this is attributed to the lack of economic growth during his presidency. One issue with this study is how the economic growth of the country is perceived. As no ted earlier, Millennials are more likely to get their information from the Internet which provides positions from various points of view. Markus highlights a study by Tufte that focuses than making assumptions about the economy as a whole, an individual might make up their minds about the economy based off of their own financial situation. It is found that people do not generally feel the financial situation is significantly worse in the ir election cycle (Markus 1988). Ng, Schweitzer and Lyons focused on how Millennials consider economic factors to be an important aspect of how they choose to vote (2012). Much like the conclusions seen from Markus (1988) Ng, Schweitzer and Lyons found that Millennials are as likely to vote for change campaigns are not irrelevant when it comes to elections because they can play a major r ole in influencing the 3 percent of the un affiliated vote ( 1988). When it comes to Millennials, their attitude toward candidates does not differ significantly from other generations if they are not
30 racially or party motivated. If an individual is strugglin g financially, Millennials are just as likely to look to new methods of correcting their situation (Ng, Schweitzer and Lyons 2012). A study by Nadeau and Lewis Beck looks at how large the impact of the economy is on the outcome of presidential elections. One perception of the government is that the President is in charge of the economic status of the country and that, even if there is a split government, the blame of a struggling economy should rest on the President (Nadeau and Lewis Beck 2001). The economic impact does largely rest on whether either candidate is an incumbent in the upcoming Lewis Beck 2001). Voters are more likely to praise and blame the incumbent running for reelection than they would a different candidate from the same party. The Nadeau and Lewis Beck study finds that only certain people see the economic state of the country as an important factor in determining who they will vote for. This connects to the hypothesis of this study that most Millennials do not see economic issues as the most important factor in voting behavior. The Nadeau and Lewis Beck study is similar to the Markus study because they both conclude that the most important determining factors on economic voting are whether the candidate is an incumbent and how people perceive the economic success of the country (Markus 1988). Ng, Schweitzer and Ly ons believe that Millennials follow this trend, but account for economic factors in their vote, but not to the same extent as social issues. Religion does influe nce the voting behavior of individuals, but the research found that there were other factors that people considered. In order to better understand Millennials and voting behavior as a whole, voting behavior needs to be addressed.
31 CHAPTER IV EXPLANATORY FRAMEWORK AND HYPOTHESIS In order to examine voting behavior across multiple presidential elections, this study uses the individual as the unit of analysis. The purpose is to compare how Millennials voted during the last few election cycles when compared to previo us generations. The primary independent 2012 election. The first hypothesis is that, if an individual is a Millennial, then they are more likely to vote for the Democratic presidential candidate. Conversely, non Millennials are expected to be more likely to vote for a Republican presidential candidate. It is a lso expected that the conservative candidate. In order for this analysis to be made, the Millennial generation must be compared to the non Millennial generations. Thes e generations are more likely to have higher levels of religiosity compared to the Millennial generation. The hypothesis for party is that Millennials will be more likely to register Democrat but both parties will vote for their candidate by a significant majority. This means that it likely will not matter whether the individual is a Millennial or a non Millennial because it is expected that they will likely vote for the candidate of the party they are registered. The hypothesis for income is that the wealt hier an individual is, the more likely they are to vote for the Republican candidate. It is expected that there should be less wealthy Millennials compared to non Millennials possibly providing a reason for why they are more likely to vote for
32 the Democrat ic candidate. The expectation is that wealthy Millennials and wealthy non Millennials will still be more likely to vote for the Republican candidate than less wealthy individuals. The hypothesis for race is that Caucasians will be more likely to vote for t he Republican candidate and racial minorities will be more likely to vote for the Democratic candidate. One expectation is that there will be a more diverse voting population leading to a higher concentration of minorities in the Millennial generation. It is expected that Millennials of all races will be more likely to vote for the Democratic candidate than their non Millennial counterparts. A Pew Research Center study recently examined how many Millennials consider themselves to be religious or attend chur ch. The study shows that 32 percent of adults aged 18 29 do not identify with any particular religion which greatly contrasts that of other age groups (Pew Research 2012). The same study also shows how that has changed from five years ago and provides a b asis for how many people consider themselves to be religious. Accordingly, another hypothesis of this study is that, if an individual is a non Millennial and an Evangelical Christian, they will be more likely to vote for the Republican candidate.
33 CHAPTER V DATA SAMPLE AND MEASUREMENT The primary data source for this research is from the American National Elections Studies (ANES). ANES data are appropriate for this study because the data source is the most common source of U.S. electoral survey data. ANES s urvey data provide information related to will be a cross sectional study examining the 2008 and 2012 presidential elections. One limitation to using thes e data is the fact that they represent only a sample of the American electorate. Each study had limited amounts people that participated, and that does not truly reflect the entire voting population of the country during the time of these elections. Becaus e of the skew in some of the variables in the ANES data, exit poll results from Pew and CNN are also presented and analyzed. The ANES sample sizes consist of 2,323 cases for the 2008 election and 4,188 cases for the 2012 election. Once again, the main vari able of focus is Millennial generation membership, which is designated as those individuals born between 1982 and 1998. By differentiating survey respondents through generational cohort, it is possible to examine how each generation voted. This study will measure the dependent variable of presidential vote choice using a binary level variable. For the purposes of the analysis, only those that voted for one of the two major party candidates are included in the sample. In 2008, 1,466 individuals reported vot ing for Obama or McCain, and 4,188 reported voting for Obama or Romney in 2012. There is a skew that occurs within the information from ANES given that the total electoral percentage voting for the Democratic candidate was higher in
34 both elections. In 2008 shows 66.6% of people voting for Obama and just 33.4% voting for McCain. ANES shows a higher percentage of voters that voted for the Democratic candidate than the amount that actually vote d for him. The primary concern for the validity of this variable is how accurate is the information from ANES given the disproportionate number of individuals that reported voting for Obama in both elections. The survey asked people to give the ir own responses and there is no way to determine if the answers given by the participants were truthful. For the most part, there should not be a significant amount of wrongfully provided information to introduce bias. Traugott and Katosh discuss the pit falls in a study with finding accurate information to represent the voting population. One issue that comes up is over reporting when respondents did not actually participate in elections (1979). Researchers have also shown evidence of a bandwagon effect w here survey respondents are more likely to report having voted for the winning candidate. Though, misreporting and candidate favoritism after the election was conducted are a concern, these survey pitfalls are not unique to the ANES. The ANES study appears to suffer from bias introduced through the bandwagon effect or sample selection. The overall findings show that more participants in the study claimed to have voted for Obama then the actual outcome of the election. This is a validity concern, yet it is n ot sufficient to abandon the use of ANES for this study. The collection of the ANES data is valuable because the data do provide greater detail than the exit polls and it allows for each voter to be analyzed based off of the responses to hundreds of survey questions. However, any sort of misreporting can potentially throw off the findings. Anyone that did not vote, but participated in the survey and said they voted, will also
35 skew the results. The study seeks to eliminate non s to ensure that their preferences and demographics are not accounted for in the final results for the ANES study. Findings from the ANES data set should be looked at with a Democratic bias given the deviation between the final voting percentage in the ele ction and the findings from the study. There are further concerns that the study might not accurately represent either population of the Millennials or non Millennials. The type of people willing to respond to a questionnaire might not accurate ly represent the voting population from the election. The study might be more accurate if there was a more reliable measure of what the religious percentage of both the Millennial and non Millennial generations is in reality. By taking the sample of respon dents from both groups provided by ANES, there is a possibility that we might get more members of one religion being represented by either generation then there actually are within the generation. This would not impact the hypothesis of how each generation votes based off of religion, but it could impact the final results of how big the difference was in the voting behavior of Millennials. One of the questions in this study asks how old the respondent was during the time of the election. Based on the descri ption of the Millennial generation, the study is able to separate those Millennials from non Millennials based on the age of the respondent. This allows for a breakdown of the demographic. By separating these groups, the study will determine differences in the voting patterns in these groups of respondents. The best way to remedy this issue is reworking the study to account for the demographic of Millennials and non Millennials. In order to do this, the percentage of each religious group must be broken down between the generations and respondents that fit those stipulations must be sought in order to receive a more accurate representation of the population. The concern with
36 changing the study in this manner is that the data might not be available for certain religious sects of the population as a whole, but it would provide a more valid study. A second independent variable is religious identification. The study conducted by ANES breaks down religion into multiple religious categories and then fur ther breaks those down into sects of those religions. The primary focus will be on Evangelical voters because they represent the sect of Christianity most likely to vote for the Republican (McDermott 2009). The study will also examine other sects of Christ ianity and other religions comparatively. Those include Judaism, Baptist, Muslim, and nonreligious, among others. The religious part of the study will be focused around Evangelicals and nonreligious voters. The expectation is that those two groups will in cur the greatest deviation in voting behavior with Evangelicals being the largest group of voters that voted for the Republican candidate and the nonreligious being the largest group of voters that voted for the Democratic candidate. Evangelicals will be c ompared to all other religions to see whether being an Evangelical is the most significant religious indicator of voting behavior while nonreligious voters will be compared to all other religions to see if religiosity affects voting behavior. The study see ks to compare religion and Evangelicals to nonreligious individuals and non Evangelicals in both a Millennial and non Millennial capacity. Since ANES has a low religious demographic of Millennials, these data will be supplemented with the exit poll data fr om CNN and Pew. The measurement of the independent variable is through the varying religious options provided in the ANES study including Judaism and its sects, Catholicism and its sects, Protestantism and its sects, nonreligious affiliation and other reli gions. The study will employ the binary variable s through a comparison of each religious identification. An example is whether
37 being Evangelical compared to those that are not Evangelical and the same conducted for other religions and sects. The variables used in the analysis are Protestants and non Protestants for 2008 and Evangelicals and non Evangelicals for 2012 because 2008 did not have a high enough number of Evangelicals in the study. One part of the study is the individual and their reli gious affiliation. The question specifically asks which religion they identify with and goes into further detail including which sect or that religion they identify with. This analysis of the individual will be incorporated into both parts of the study by examining any possible relationship between more religious voters and their propensity to vote Republican compared to less religious voters and their propensity to vote Democratic. This will also be beneficial in comparing the results of people that identi fy themselves as religious compared to the Pew Study to see if there are similarities in both studies on how many people identify as religious. Also included will be a comparison between religiosity and non religious Millennials as an independent variable. The study will also include the individual and their party identification. The question seeing whether Democrats or Republicans voted for their candidate in greater numbers than their counterparts. It will also be beneficial to see where independents voted in both elections. mark the answer that includes the i ncome of all members of your family living here in 2011 will benefit the study to see whether income is a more important variable than others. It is also beneficia l to see where, if anywhere, voting behavior changes based off of income. The 2012 ANES study breaks down the income of the individual from under $5,000 to over $250,000.
38 The last variable included will be the race of an individual. The question asks the vote Democratic than Caucasians. This benefits the study to see if different races have difference preferences in voting for a candidate. The Pew study provides a base for calculating an accurate representation of religious Millennials while ANES provides a base for examining the voting patterns within those groups of religious individuals. With the Pew study showing a significant difference between gener ational religious identification and the ANES study showing the likelihood of nonreligious Millennials voting liberal, there is sufficient information to show that religion and generation plays an important role in voting behavior. The study will determine how much of an impact that difference played in the 2008 and 2012 elections. The Pew data set is from an exit poll conducted after the election that looked at all the variables included in this study. It broke down the election by people that voted Democ rat, Republican and Independent as well as their age, religion, party identification, race and income. The CNN exit poll was similar but was conducted separately from Pew. It also identified which candidate an individual voted for, as well as their age, re ligion, party identification, race and income.
39 CHAPTER VI METHODOLOGY In order to properly test the hypothesis, all of these statistical indicators are needed to see if there is a significant connection between these explanatory variables and presidential vote choice. As an initial way to examine relationships between variab les, a series of crosstabs are presented. The crosstab analysis is supplemented with bivariate and multivariate regression, which allow for estimates of the direction and strength of each variable on voting behavior. The result expected fro m the analysis is that the data will show a significant disparity in voting patterns between several groups analyzed in the study. With a focus on religion, the expectation is that religion will be the most significant indicator of how the individual voted during either election. Another expected result is that there will be a significant difference between the amount of religious and nonreligious voters between generations. The Millennial generation is expected to have one of the lowest amounts of religiou s voters compared to the generations before them that are expected to have higher levels of religious voters. This difference is expected to show the Millennial generation being much more likely to vote for the liberal candidate compared to the other gener ations. there are significantly more people that identify themselves as nonreligious in the Millennial generation than the generations before it. In the ANES and Pew studies there is a higher percentage of people that identified themselves as religious as their age increased. Also, in the studies, there is a correlation between the amount of people that either voted Democratic or
40 claimed they would vote fo r the Democratic candidate based off of their age. This shows that there is a generational difference in the voting patterns between the different generations.
41 CHAPTER VII 2012 PRESIDENTIAL ELECTION Party Registration in 2012 The data from the 2012 ANES survey shows that there were 1,203 (59.44%) individuals registered as Democrat, 146 (7.21%) as pure independent and 675 (33.35%) as Republican. Barack Obama got the most votes from the Democratic partisans and independents with different levels of sup port from each group while Mitt Romney got the most votes from Republicans. CNN exit polls shows that around 36% of the voting demographic identified as Democrats, 32% identified as Republican and 29% identified as independent. This difference is attributa ble to the fact that CNN characterizes independent leaning partisans as independent. Of the 1,203 that identified as Democrats, 1,063 voted for Obama. This was roughly 88% of the registered Democrats that voted along party lines. This shows dis parity from the CNN exit polls which shows that Democrats voted for Obama at a 92% rate while voting for Romney at a 7% rate. The ANES study does have a few notable skews in favor of the Democratic candidate so the findings would expect an inflated number in terms of support for Obama. Table 1.1 2012 Party Registration Vote (ANES) Barack Obama Mitt Romney Total Democrat 1,063 (88.4%) 140 (11.6%) 1,203 Republican 138 (20.4%) 537 (79.6%) 675 Independent 102 (69.9%) 44 (30.1%) 146 Total 1,267 (62.6%) 757 (37.4%) 2,024 Table 1.1 reports that 675 respondents that identified as Republicans. 537 of the 657 voted for Romney. This was roughly 80% of the registered Republicans that voted along party
42 lines. The CNN exit polls shows that Republicans voted for Romney at a 93% rat e and Obama at a 6% rate. This is an even greater disparity and does not even come to the same conclusions as the ANES study with both findings having different results for how strongly the Republican Party supported Romney. The ANES study does have the De mocratic skew that does favor Obama which could explain why the exit poll data shows greater support for Romney within the Republican Party. Of the 146 independent participants, 102 voted for Obama. This was roughly 70% of independent voters t hat favored the Democratic candidate in the 2012 el ection. The CNN exit polls show 50% of independents voting for Romney and 45% voting for Obama. The amount of independent voters that participated in the study was significantly lower than what the exit polls projected for the total independent voting demographic. This was because it wa s measured between people that were registered for a party and not those that leaned toward either party. The findings from this variable show there is a conclusive relationship between party identification and voting behavior. The CNN exit polls show that there is a significant relationship between party affiliation and voting behavior in addition to the ANES study that shows a significant relationship with Republican registered participants voting for Romney, albeit in smaller numbers. Table 1.2 2012 Mi llennial Party Registration and Vote Choice (ANES) Non Millennial Millennial Total Democrat 1267 (87.8%) 176 (20.2%) 1,443 Republican 716 (89.1%) 88 (10.9%) 804 Independent 156 (72.9%) 58 (27.1%) 214 Total 2,139 (86.9) 322 (13.1%) 2,461 Table 1.2 shows that, of the 749 respondents that identified as Democrats, 168 were Millennials. Of the 358 that identified as Republicans, 81 were Millennials. This shows that the
43 two groups make up roughly 22% of their respective parties and the amount of Millennials that register for either party is not disproportionate to the other. This shows that 249 out of the 1,107 respondents were Millennials. With Democrats having more participants as a whole, it still shows that Millennials, like the other generations, are just as likely to identify as a Democrat. If Millennials have roughly the same party identification as other generations then it would not explain why they were significantly more likely to vote for Obama since the findings were clos er than examined. The study has already identified that Millennials are more likely to vote for the Democratic candidate and they are more likely to register Democrat. Even though there is a relationship between affiliation and voting behavior through the ANES study, Millennials are not registering as Democrats in greater proportions than non Millennials. Presidential Vote in 2012 The ANES data set was broken down into people that reported voting in the 2012 presidential election for either Bara ck Obama or Mitt Romney. The amount of people eligible in that group was 4,188 and of those 2,496 (59.6%) voted for Obama while 1,692 (40.4%) of those voted for Romney. The issue with this finding is that the actual outcome of the election was roughly 51.1 % for Obama and 47.2% for Romney. This shows that the statistics accumulated by ANES are skewed in favor of Obama and does not represent the Romney base as strongly as it should. ANES have shown a bias in favor of the Democratic candidate in all their prev ious studies. The first part of the study is testing whether Millennials do vote different than other generations so the ANES study is broken down by age groups. In order to correctly single out the Millennials as a separate variable, the age group of 29 or younger was taken out. Out of the
44 4,188 voters it was found that 230 fell into the age group of Millennials. Among those 177 (76.96%) voted for Obama compared to just 53 (23.04%) for Romney. This is in line with the hypothesis that Millennia ls are more likely to vote for the Democratic candidate and it also shows that the difference is substantial. The ANES study is skewed toward Obama, but there is a major difference among the non Millennials as there was over an 18% difference in voting per centage between the Millennials and non Millennials. Another sampling issue with the ANES data set is that it only shows Millennials as representing 5.5% of the 2012 electorate while exit polls typically show that Millennials represented around 19% of the population for that election. The likelihood of the skew is that Millennials would have still voted for Obama, but not to the same extreme levels. The outcomes from exit polls show that about 60% of Millennials voted for Obama in the 2012 election while Ro mney averaged 37%. Exit polls also show greater Romney support among older demographics. It also shows that people born right before Millennials were also more likely to vote for Obama in the 2012 election. Table 2.1 2012 Presidential Vote (ANES) Barack Obama Mitt Romney Total Non Millennials 2319 (58.6%) 1639 (41.4%) 3958 Millennials 177 (77%) 53 (23%) 230 Total 2496 (59.6%) 1692 (40.4%) 4188 Table 2.1 shows that there is a significant relationship between voting behavior of Millennials and other generations in the 2012 election. While the ANES study shows that they are participating less than other generations, the exit polls show that Millenn ials made up a significant representation of the voting population. The only other age group that voted in favor of Obama was the 30 39 age group while the rest of the groups were more likely to vote for
45 Romney. Support from the Millennial and Generation X generations helped bolster President Table 2.2 Logit Regression Model of 2012 Presidential Vote (ANES) Number of Observations= 4,149 R squared= .0053 Vote Coefficient Standard Error Millennial .109* .023 Constant .582 .008 Note: mean s significance at the 0.1 level Table 2.2 reports the results of a logistic regression model where the dependent variable indicates whether an individual voted for Obama or Romney with Obama votes coded as 1. The results show that Millennials are significantly more likely to vote for Obama. This shows that Millennials did favor Obama in 2012 election more than other generations. Table 2.3 presents a logit regression model testing whether individual Millennial subgroups are each more likely to vote Democratic. Table 2.3 Logit Regression Model of 2012 Presidential Vote (ANES) Number of Observation= 4,188 R s quared=.0046 Vote Coefficient Standard Error Millennial (17 20) .572* .146 Millennial (21 24) .222* .161 Millennial (25 29) .719* .244 Constant .335* .033 Note: mean s significance at the 0.1 level One issue within the Millennial generation is determining whether the generation has a significantly different voting behavior within the generation. According to the regression model
46 of Millennials and age, (see Table 2.3) Millennials of all ages were sig nificantly more likely to vote for the Democratic candidate in 2012 relative to non Millennials. In addition to this finding, the regression model also found that the older age range of Millennials (between the age of 25 and 29) were the most likely to vot e for Obama while the middle age range of Millennials (between 21 24) were the least likely to vote for Obama. The importance of these findings is that the Millennial generation as a whole was more likely to vote for Obama while the older age range of Mill ennials were most similar to the younger group of Millennial voters. Religion in 2012 The results for the Evangelical vote show that, among the 4,188 voters, 943 considered themselves to be Evangelical (see table 3.1). This was the highest representation o f any religion in the entire study. The results from the crosstab presented in table 3.1 do not provide support for the religion hypothesis because Obama received 57.58% of the Evangelical vote compared to Romney receiving 42.42%. Among non Evangelicals th e v ote was not too different with Obama receiving 60.18% of the non Evangelical vote while Romney received 39.82%. Table 3.1 Evangelical Vote (ANES) Barack Obama Mitt Romney Non Evangelical 2224 (60.18%) 1471 (39.82%) Evangelical 543 (57.58%) 400 (42.42%) These results conflict with the Pew Research data that show radically different results. Pew found that 79% of Evangelicals and Born again Christians voted for Romney while just 20% voted for Obama. The hypothesis of this study was that the tren d would be that Evangelicals would be more likely to vote for the Republican candidate. The conflict with the ANES study could be seen as an issue with gathering the information. All the previous studies
47 have seen that, regardless of age, Evangelicals will be more likely to vot e for the Republican candidate. Table 3.2 shows the exit polls taken also show that Evangelicals voted for Romney over Obama by a significant margin. The CNN 2012 exit poll show that 78% of Evangelical and Born again Christians voted for Romney while 21% voted for Obama. This is compared to 60% of non Evangelicals voting for Obama while 37% of non Evangelicals voted for Romney. Table 3.2 Evangelical Vote (CNN Exit Polls) Barack Obama Mitt Romney Evangelical 21% 78% Non Evangelical 60% 37% There are many reasons the data from the ANES study might not accurately represent the preferences of the Evangelical voter base. The most obvious reason is that the results are over high the p ercentage of voters in any category would be. A couple other issues with the ANES study include Evangelicals only representing 22.5% of the population in the study while a more accurate representation would be around 26%. Given that the study does focus mo re on non Millennials, it would be expected that there would be a skew toward having more religious participants. Table 3.3 Logit Regression of Presidential Vote (ANES) Number of Observations= 4,111 R squared= .0053 Vote Coefficient Standard Error Millennial .107* .024 Religiosity .019 .182 Constant .586* .009
48 Table 3.3 reports the results of a logistic regression mod el where the dependent variable indicates whether an individual voted for Obama or Romney with Obama votes coded as 1. The results show that Millennials are significantly more likely to vote for Obama but religious individuals are not. This shows that Millennials and religious individua ls have different voting behavior. The finding fro m the analysis of the ANES data presented in table 3.1 is that there is not a significant relationship between Evangelicals and voting behavior compared to non Evangelicals. There are major validity concern s with the findings from the ANES study as they go against all other information obtained about the subject and it might be beneficial to gather a larger and more diverse sample of information about Evangelical voters to find out if the data gathered was i naccurate. The contradictory nature of the study means that it is unknown whether being Evangelical was more likely to vote for a Republican candidate in the 2012 election. The other hypothesis about religion was that nonreligious voters were m ore likely to vote for the Democratic candidate. In the ANES data, 779 of the 4,118 participants were nonreligious which accounts for around 18.9% of the people in the study. This is higher than the representation from exit polls which estimated that aroun d 12% of the voting population was nonreligious. The Pew study estimated that about 24% of the voting population in 2012 had no religious identification meaning that there are questions as to how many voters were nonreligious. Table 3.4 Religion and the 2 012 Presidential Vote (ANES) Barack Obama Mitt Romney Religious 1,891 (56.6%) 1,448 (43.4%) Nonreligious 559 (71.8%) 220 (28.2%)
49 Note: mean s significance at the 0.1 level Table 3.4 shows that among ANES respondents, 71.8% of nonreligious voters voted for Obama while 28.2% of nonreligious voters voted for Romney. Religious voters voted for Obama at a 56.6% rate while voting for Romney at a 43.4% rate. The qualification for the religious There is a significant relationship within this finding that shows that less religious people are more likely to vote for Obama. The D emocratic skew in the ANES data is still present and it would be expected that a more accurate representation of the data would show a higher percentage of religious voters voting for Romney. The Democratic skew could also account for a higher proportion of Democratic voters in the nonreligious findings as well. Figure 1.1 2012 Presidenti al Vote Percent by Protestantism (CNN)
50 In the Pew data, nonreligious voters voted for Obama at a 70% rate and nonreligious voters voted for Romney at a 26% rate. These proportions mirror the results of the ANES study. The religious voters show differing trends between religions with African American Protestants being the most likely to vote for Obama at 95% while Born again Evangelicals were more likely to vote for Romney at 79%. Overall, the total of all religious groups together was still more likely to vote for the R epublican candidate in 2012 than the nonreligious groups. The CNN exit poll also found that about 70% of nonreligious voters voted for Obama while 26% voted for Romney. The CNN exit poll broke down the religious voting between, Protestant, Cat holic, other religion and non religion and saw the greatest amount of Obama voters coming from the other religion category with 74% of them voting for Obama and just 23% of them voting for Romney. Figure 1.1 shows that Protestants of all denominations vote d for Romney at a 57% rate while voting for Obama at a 42% rate. This shows another significant difference in voting behavior between nonreligious voters and religious voters in the 2012 election. All three studies found a significant relationshi p between voting behavior of nonreligious voters compared to religious voters (see table 3.3). One avenue worth exploring is how much these other religions play a role in voting behavior. Despite not finding a significant relationship between Evangelicals in the ANES study, it is still possible that this is the most significant religion when it comes to voting for the Republican candidate. It was also found that African American Protestants and other religions voted for Obama in even higher percentages than nonreligious voters. This difference possibly shows that there were other factors that played a role in the voting behavior of religious voters.
51 The overall findings on the religious influence of voting behavior in the 2012 presidential electi on is that being nonreligious was a major factor in voting for the Democratic candidate. The inconclusive results regarding the Evangelical remains in question due to the findings of the ANES study, but it is still likely that being an Evangelical plays a role in whether someone will vote for the Republican candidate based off of the findings from other studies. Income in 2012 The income increments are larger at the higher end of the scale because there are fewer people in the higher income gaps. There is t he largest number of under $5,000 income gap people that voted for Obama with 240 of the 338 participants that fell in that income range. The most voters Romney had were in the $80,000 $89,000 income range. The $80,000 $89,000 income range had 101 Romney v otes, but that was only 50% of people in that income range. This shows that there was no significant relationship in that income range as the votes was split evenly. Figure 2.1 shows the 2012 CNN exit polls by income. The polls show different numbers as i t does not break it down into small categories, but instead into increments of $50,000. One exit poll shows the comparison between people making less than $50,000 which also shows those voters being likely to vote for Obama in significant numbers. Accordin g to the exit polls, 60% of people making under $50,000 voted for Obama and 38% voted for Romney. The other income ranges show different results from the ANES study as voters were more likely to vote for Romney if they made between $50,000 $100,000 or over $100,000. As a whole, voters making over $50,000 were more likely to vote for Romney by a 53% to 45% margin. When that number is increased to over $100,000, there is not a significant increase as the margin increases to 54% to 44%.
52 Figure 2.1 2012 Presidential Vote Percent by Income (CNN) The similarity in both studies is that poorer voters were significantly more likely to vote for the Democratic candidate in the 2012 election. The results show different results for how the middle and upper class voted in the 2012 election as, broken down into smaller gaps, it was difficult to distinguish any pattern with voting behavior based on the in crease of income. The ANES data does have the Democratic skew, but it seems as though the results from the voters would not have differed much if that was not present. The increase in income did not mean that the highest incomes were any more likely to vote for the Republican candidate than the Democratic candidate. It does seem that the likelihood of an individual vo ting for the Democratic candidate is greater if the income of the individual is around $44,999 or less.
53 Income of Millennials varies in range, but Millennials feel that they are not earning as much as they expect (Pew Research 2012). Compared to other ge nerations, they feel that they are not earning enough, but remain hopeful that they will earn more in the future. A majority of Millennials, however, is optimistic about their earning potential in the future which is 23% more than Generation X. Determinin g how this connects to income is difficult because Millennials have the sense that they are not making enough, but many still fall above the $44,999 income level upon graduation. According to the ANES study, Millennials do not differ significantly in incom e disparity when compared to other generations. This makes it difficult to determine if income plays a role in why Millennials are more likely to vote Democrat. One avenue of further exploration is determining whether the economic expectation of Millennial s for each candidate might affect their voting behavior. If Millennials feel that the Democratic candidate has a better policy to help them earn more in the future, then they might be more inclined to vote for them. Given that the biggest difference betw een Millennials and other generations is their dissatisfaction with their current pay; their vote might be influenced by economic policy. As far as current income goes, there is no significant conclusion from the ANES and CNN exit polls that shows a connec tion between Millennials being more likely to vote for the Democratic candidate over the Republican candidate because of their income in the 2012 election. Race in 2012 In the 2012 ANES study, there were two ethnicities that had a significant amount of respondents, white and African American. Of those two groups, they were the only ones with over 32 respondents that voted for either Obama or Romney in the 2012 election. O f the 806
54 white respondents, 426 voted for Obama while 380 voted for Romney. Of the 424 African American respondents, 413 voted for Obama and 11 voted for Romney (see table 4.1). This distribution shows a major disparity in voting behavior between white an d African American voters with both being more likely to vote for Obama, but African American voters much more likely to vote for him than white participants. Table 4.1 2012 Racial Vote (ANES) Barack Obama Mitt Romney Total White 426 (52.9%) 380 (47.1% 806 African American 413 (97.4%) 11 (2.6%) 424 Total 839 (68.2%) 391 (31.8%) 1230 The 2012 CNN exit polls show different numbers and also have a Latino group included. Of the white demographic, 59% voted for Romney while 39% voted for Obama. Of the African American demographic, 93% voted for Obama and 6% voted for Romney. The Latino demographic also heavily favored Obama by a 71% to 27% margin. The CNN exit polls representation of Latinos could explain why there was a higher percentage of Obama voter s in the ANES study since Latinos were more likely to be included in the white demographic since there was no Latino category available. The voting demographic differed in the ANES study and CNN exit polls as the ANES study shows a higher representation o f African American voters than the CNN exit polls. The representation given in the CNN exit polls is around 72% white, 13% African American and 10% Latino with other minorities representing 3% or less. The high number of African American participants in th e ANES study could explain why there is the significant Democratic skew. The Pew study from the 2012 election focuses on the combination of race and religion in voting behavior. One major finding was that African American Evangelicals had the highest
55 perc entage of Obama voters with 95%. This is in contrast to white Evangelical voters that had the highest number of voters for Obama with 79%. Of the Evangelicals in the ANES study, 256 were African American while 213 were white. This explains the high number of Obama voters in the Evangelical category since there were more African American Evangelicals than white Evangelicals. Given their likelihood to vote for the Democratic candidate, this explains why there was a significant Democratic skew in the findings of the ANES study. Figu re 3.1 2012 Presidential Vote Percent by Race (CNN) The Pew study also shows that African Americans were the highest voting racial group by percentage in the 2012 election. Latinos and Asians were the only two major ethnic groups that voted beneath 50%. One outside factor that could be considered is that Ob ama is partially African American and that helped influence African American voters to vote in higher numbers.
56 African Americans only represent about 13% of the voting population, but their propensity to vote Democratic can play a significant role in the o utcome of an election. Figure 3.1 shows that whites voted for Romney at a 59% rate while voting for Obama at a 39% rate. This shows another significant difference in voting behavior between white and non white voters. It also shows that non whites voted for Romney at a 16% rate while voting for Obama at a n 82% rate. White Millennials represented a total of 11% of the voting population and they were more likely to vote for Romney by a 51% to 44% margin. Latino Millennials represented a total of 4% of the voting population and they favored Obama by a 74% to 23% margin. African American Millennials represented 3% of the voting population, and voted for Obama by a 91% to 8% margin. The Millennial breakdown shows that white Millennials represent around 61% of the voting demographic while African American Millennials represent around 17%. Latino Millennials represented around 22% of the voting population. This shows that the Millennial generation had more Latino and African American voters than other generations Since both racial groups were more likely to vote for the Democratic candidate, there is a potential link between the increased likelihood of Millennials voting for the Democratic candidate and race. The greater racial diversity of the Millennial genera tion could be one possible explanation for why they were more likely to vote for the Democratic candidate. Data from ANES, CNN, and Pew show that both African Americans and Latinos are significantly more likely to vote for the Democratic candidate than whi tes. Millennials have mirrored the voting behavior of other generations in terms of voting based off of race, but the increasing prevalence of other races could explain why Millennials were more likely to vote for Obama in the 2012 election.
57 The findings from this study also explain why there was a high Democratic skew in the ANES study. The high number of African Americans that participated in this study could explain why the Evangelical group shows a higher likelihood of voting for Obama. The white categ ory was also mixed with the Latino category in the ANES study which meant that the white totals could have been skewed since Latinos are significantly more likely to vote for the Democratic candidate. Table 4.2 Regression Model of 2012 Obama Vote Choice Number of Observations= 3,673 R squared= .5045 Vote Coefficient Standard Error Coefficient Standard Error Millennial .411* .177 Millennial (17 20) .410* .389 Millennial (20 24) .415* .300 Millennial (25 29) .408* .245 Evangelical .671* .141 .671* .141 Democrat 2.467* .126 2.467* .126 Republican 2.645* .143 2.645* .143 Income .013* .007 .014* .007 Racial Minority 1.786* .158 1.79* .158 Constant .148 .128 .148 .128 Note: mean s significance at the 0.1 level The findings in the table 4.2 regression model show that there was significant relationship between all the variables included. Millennials, Democrats and racial minorities were more likely to vote for Obama while Republicans and wealthier and evangelical indi viduals were less likely to vote for Obama. One finding within the Mi llennial generation was that age did not play a significant role in whether a Millennial was more or less likely to vote for Obama. The 17 20 Millennials voted for Obama in nearly the sam e amount as the 25 29 Millennials.
58 CHAPTER VIII 2008 PRESIDENTIAL ELECTION Party Registration in 2008 In 2008, 734 of the ANES respondents considered themselves to be Democrats, 339 were Republican and 393 were independent. Of the 734 Democrats, 692 (94.3%) voted for Obama. Of the 339 Republicans, 310 (91.4%) voted for McCain. Of the 393 independents, 255 (64.9%) voted for Obama (see table 5.1). This shows high partisanship among voters in the 2008 election which clashes with the findings from the 2012 electi on which show differences in party affiliation. Table 5.1 Party Identification and Vote Choice (ANES) Barack Obama John McCain Total Democrat 692 (94.3%) 42 (5.7%) 734 Republican 29 (8.6%) 310 (91.4%) 339 Independent 255 (64.9%) 138 (35.1%) 393 Total 976 (66.6%) 490 (33.4%) 1,466 The 2008 CNN exit polls show that about 39% of the voting electorate identified as Democrats, 32% identified as Republicans and 29% identified as independent. This differs from the ANES study in that their study included significantly more Democrats and a higher percentage of independents than the exit polls. The disparity in the amount of people that label themselves as Democrats and Republican in both studies shows that there is a Democratic skew in the amount of partici pants in the ANES study because ANES only counts pure independents. This is insignificant because the quantity of party affiliated voters does not influence the overall partisanship of their voting behavior. According to the 2008 ANES study, of the 39% th at labeled themselves as a Democrat,
59 89% voted for Obama and 10% voted for McCain. Of the 32% that labeled themselves as Republican, 90% voted for McCain and 9% voted for Obama. Of the 29% that identified as independent, 52% voted for Obama and 44% voted f or McCain. These numbers are in line with the ANES study that also shows high partisanship within voting behavior in the 2008 election. The amount of independent voters that voted for Obama in the ANES study is higher, but that can be accounted for by the Democratic skew within the study. The Pew study reveals that the amount of Democrats was around 39%, Republicans around 32% and independents around 29% which was identical to the CNN exit polls. What the Pew study also shows is how the electorate shifted from the 2004 election when roughly 37% of voters were Republican and 37% were Democratic meaning that the major shift came with less people considering themselves to be a Republican and more being a Democrat. The affiliation of the electorate did not chan ge much in 2012 with 38% labeling themselves as a Democrat while 32% labeled themselves as a Republican. From 2004 to 2008 there was a major shift in the amount of registered Democrats and Republicans that might have been an important factor in influencing the outcome of the 2008 and 2012 elections. The findings from the ANES, CNN exit polls and Pew study is that party affiliation did matter in the 2008 election with a high amount of partisan voting as well as an increase in the overall amount of Democrats that voted in the election. Millennials being more likely to be Democrats could be one of the reasons why there has been an increase in the amount of registered Democrats, but it would not explain why there was not another major increase in the amount of Democrats during the 2012 election. The outcome does show that there is a significant relationship between party affiliation and voting behavior. Thus, one reason Millennials are more
60 likely to vote for the Democratic candidate is because more of them are registering as Democrats. Presidential Vote in 2008 The presidential vote in the 2008 ANES time series is broken down into 1539 cases in which a participant either voted for Obama or McCain. Of those participants, 1025 (66.6%) reported voting for Obama whi le 514 (33.4%) voted for McCain. The actual vote count was roughly 52.9% for Obama and 45.7% for McCain. This shows that the accumulated data from ANES is even more skewed in the 2008 survey than it was for 2012. This means that the Democratic skew will be even more present in the results of this study so the actual amount of voters in each demographic will favor Obama. The first analysis presented in this section compares the distribution of the Millennial vote to non Millennials. A challenge to do this fo r the 2008 election is the fact that ANES breaks down age groups by the age group of 18 34 meaning that Millennials are mixed in with non Millennials in this study. The other issue is that only 15 participants in the data set fall into the 18 34 age category. Eight of these individuals voted for McCain while seven of which voted for Obama. The sample size is too small to determine whether there is a significant relationship between voting behavior of Millennials compared to no n Millennials. Many of the participants in this study did not disclose how old they were and the study did not go into greater detail on their specific age. Even if that information were disclosed, the amount of participants is still not high enough to com e up with any reliable conclusions. Table 6.1 2008 Presidential Vote (Pew) Barack Obama John McCain Non Millennials 48% 50% Millennials 66% 31%
61 According to the Pew study, the 18 29 age range represented around 19% of the total voting demographic. This is the same percentage as that of 2012, but it also includes four years of non Millennials that would have ranged in age from 26 30 at the time of the study. Like the 2012 results, the likelihood that someone was going to vote for McCain increased as their age increased, but Obama won all age groups besides the 65 and older group. A major difference is that 66% of the 20 29 age group voted for Obama which was 14% more than the next closest age group. The exit polls broke down the age into the age group of 18 24 with little difference compared to the 25 29 age group. The findings actually show that both groups voted for Obama at a 66% rate, but the 25 29 group voted for McCain 1% less than the 18 24 age group (see table 6.1). There are quite a few differences from the findings in 2008 and 2012, the most prominent being that the overall vote for Obama was much greater in 2008 than it was in 2012. What wa s notable was the age groups within the Millennials generation were even more likely to vote for Obama in 2008 than they were in 2012. This finding shows that there is a significant relationship between Millennials being more likely to vote for the Democra tic candidate than non Millennials. Like the 2012 study, reasons for why Millennials are more likely to vote for the Democratic candidate should be examined. The hypothesis remains that the decreasing amount of religiously affiliated voters in the Millenni al generation is the main reason why they are more likely to vote for the Democratic candidate. Religion in 2008 The ANES study broke religious affiliation into just four religious categories, Protestant, Catholic, Jewish and other. The issue with this is that it does not separate nonreligious group from the other category meaning that there is no way to determine which of the participants
62 identified with no religion. The ANES study will not be a determining factor in the final results of wheth er nonreligious people are more likely to vote for the Democratic candidate. The first category examined is the Protestant category because it is a sect of Christianity. The results of the ANES study shows that 536 of the participants were cons idered Protestant and, of those, 5 of them were part of the Evangelical sect offered as an option in the study. The biggest response was for Baptists which had 466 participants. The Evangelical sect is also discounted from the 2008 results because there is not a significant amount of Evangelical participants in this study. The ANES study shows that as a whole, Protestants were more likely to vote for McCain in the 2008 election than Romney in the 2012 election. Both sets of data were skewed in f avor of favor of the Republican candidate. Of the 428 Protestant participants, 219 voted for McCain and 209 voted for Obama. While this is a small difference, i t is significant because of the Democratic skew as well as the other religions voting heavily in favor of Obama. Table 7.1 Religion and 2008 Presidential Vote (ANES) Barack Obama John McCain Total Protestant 209 (48.8%) 219 (51.2%) 428 Catholic 232 (66.7%) 116 (33.3%) 348 Other 404 (78.1%) 113 (21.9%) 517 Total 845 (65.4%) 448 (34.6%) 1293 Table 7.1 shows that, of the 348 Catholics, 232 voted for Obama and 116 voted for McCain. This shows a significant relationship between Catholics being more likely to vote for
63 McCain received 113 votes. Both of these indica te that religions other than Protestantism heavily supported Obama while Protestantism was the only religion that had close results. The 2008 CNN exit polls show the same results with Protestant voters being more likely to vote for McCain at a 54% to 45% rate. The other religions also show that non Protestants were significantly more likely to vote for Obama with all four of the other major categories in the study (Catholic, Jewish, other and none) all voting in favor of Obama. Of those categori es, 75% respectively. The disparity between Protestant and non Protestant voters in both studies shows that Protestants are more likely than other non Protestant grou ps to vote for the Republican candidate. Table 7.2 Evangelicals and Presidential Vote (CNN Exit Poll) Barack Obama John McCain Evangelical 24% 74% Non Evangelical 62% 36% The 2008 CNN exit polls do also include a white Evangelical/born again question. Those that identified as Evangelical/born again voted for McCain at a 74% level while non Evangelicals voted for Obama at a 62% level (see table 7.2). To further breakdown the comparisons, the exit polls also show Evangelic al/born again compared to non Evangelical Protestants. Those that labeled themselves as Protestant, but non Evangelical, voted for McCain at a 55% rate. The disparity shows that Protestants as a whole seem more likely to vote for the Republican candidate, but Evangelicals are even more likely to vote for the Republican candidate than Non Evangelical Protestants. The Pew study found that all religions were more likely to vote for Obama in 2008 election than in 2012. In 2004 Bush dominated the Evangelical vo te with 79% but that decreased
64 to 73% in 2008 for McCain. This was only the third biggest change as unaffiliated religious voters saw in 8% increase in voting for Kerry in 2004 to voting for Obama in 2008. This was the largest shift in religious voting beh category that did not see an increase for Democratic candidate voters in 2008, but instead saw a 1% increase for the Republican candidate. This could possibly show that this religious group of voters remains consistent in their voting behavior. It was still heavily in favor of the Democratic candidate in 2004 and 2008. other faith and Jewish religious groups. Protestant groups and white Catholics. The Evangelical sect of the Protestant group is where McCain received the largest margin of victory. The Pew study determined that there was an overall increase in the amount of Evangelical voters from 2004 to 2008 by 3%. Unaffiliated voters also saw an increase of 2% overall to around 12% of the total electorate. Despite the increase in Evangelicals, the Protestant group remained at around 54% of the electorate whi ch has not changed from 2000. The 2% increase in unaffiliated voters is significant though because of their likelihood to vote for the Democratic candidate. The religious variable holds merit in the 2008 election as much as it did in the 2012 election. Th e ANES study was in line with exit polls and the Pew study, all of which show that being Protestant meant you were more likely to vote for the Republican candidate while not being Protestant meant you were more likely to vote for the Democratic candidate. A further breakdown of that revealed that Evangelicals were the highest percentage of all religiously affiliated voters that voted for the Republican candidate while nonreligious voters were among the highest percentage of voters that voted for the Democra tic candidate.
65 Figure 4.1 shows that Protestants of all denominations voted for McCain at a 54% rate while voting for Obama at a 45% rate. It also shows that non Protestants voted for McCain at a 35% rate and Obama at a 53% rate. This shows another significant difference in voting behavior between nonreligious voters and religious voters in the 2008 election. Figure 4.1 2008 Presidentia l Vote Percent by Protestantism (CNN) The connection to Millennials is that it is seen that Millennials are the least religiously affiliated generation in years with increasing percentages of them being nonreligious. There is a strong case that one of the reasons Millennials vote in favor of the Democratic candidate is because they are religiously unaffiliated.
66 Inco me in 2008 The 2008 ANES study broke down the income ranges in categories that ranged from under $2,999 to over $150,000. The most votes for Obama and McCain both came in the $60,000 $74,999 income range as Obama received 88 of the 142 voters that fell in to that category while McCain received the other 54 votes. Figure 5.1 shows that people making under $50,000 voted for McCain at a 38% rate while voting for Obama at a 60% rate. It also shows that people making over $50,000 voted for McCain and Obama at a n equal rate of 49%. This shows another significant difference in income of voters in the 2008 election. Figu re 5.1 2008 Presidential Vote Percent by Income (CNN)
67 Votes by these lower income ranges mirrored those from the 2012 ANES study as both found that the decrease in income led to a higher likelihood that the voter would vote for the Democratic candidate. The difference was that this study shows that an increas e in income led to a higher likelihood that the voter would vote for the Republican candidate as the over $150,000 income range was significantly in favor of McCain. Despite this difference, the income ranges below $150,000 show no significant difference b etween Obama and McCain. The 2008 CNN exit polls broke down income in under $15,000 and over $200,000 as their highest measures with ranges as low as $15,000 and as high as $50,000. The highest percentage of Obama voters came from the under $15,000 range w hich show that voters were more like to vote for Obama than McCain by a 73% to 25% margin. The highest percentage of McCain voters came in the $100,000 $150,000 range with a 51% to 48% margin. As the income increased from the range, voters were more likely to vote for Obama according to the exit polls. The exit polls also break down the income ranges by over and under $50,000 and over and under $100,000 and it shows that both income ranges were split in the over categories. Both over categories were 49% in favor of Obama and 49% in favor of McCain. This shows that an increase in income did not mean that voters were more likely to vote for the Republican candidate in the 2008 election. The decrease in income did show a significant relationship in likelihood that the voter would vote for the Democratic candidate in the 2008 election. The ANES study and the CNN exit polls both show that the lower incomes were significantly more likely to vote for the Democratic candidate. The studies conflicted on the findings on the higher income individuals as the ANES study found that people earning above $150,000 were more likely to vote for the Republican candidate, but the CNN exit polls found that they were no more likely than individuals earning $50,000. These findings lead to
68 inconclusive results as to whether an increase in income means that an individual is more likely to vote for the Republican candidate. Like the 2012 election, Millennials have varying incomes. This means that it is difficult to determine if there is a relationship between their voting behavior compared to other generations for the 2008 election. Since Millennials do not greatly differ in income disparity compared to other generations, it is inconclusive whether that income might influence the votin g behavior of Millennials. Like the 2012 election, one income variable that could explain this behavior is perception of the market and its future. With Millennials optimistic about their earning potential, they could feel that the Democratic candidate off ers them greater earning potential than the Republican candidate. The overall finding with income is that there is no significant relationship between the income of an individual and their voting behavior. That means that Millennials are no different tha n the other generations when it came to their voting behavior based off of income. The only significant connection found was that the decrease in income means an individual would be more likely to vote for the Democratic candidate, but Millennials have not seen an abundance of individuals with lower incomes. Race in 2008 The 2008 ANES study shows that there were 954 white participants that voted for either Obama or McCain in the 2008 election. There were also 418 African American participants and 51% of the total vote with 486 voting for him. The African American group voted for Obama with a 99.5% to .5% margin with 416 out of the 418 participants claiming to h ave voted for
69 for McCain (see table 8.1). This shows that African Americans were significantly more likely to ere also more likely to vote for Obama.
70 Table 8.1 2008 Racial Vote (ANES) Barack Obama John McCain Total White 468 (49.1%) 486 (50.9%) 954 African American 416 (99.5%) 2 (.5%) 418 Total 884 (64.4%) 488 (35.6%) 1372 The 2008 CNN exit polls show that white voters voted for McCain by a 55% to 43% margin while all other races favored Obama, the highest of which was the African American population which voted for Obama by a 95% to 4% margin. The Latino group, which represented 9% of the total voting demographic, voted for Obama by a 67% to 31% margin. There was a significant difference in voting behavior from Millennials and other generations in terms of race. Millennial whites favored Obama b ay 54% to 44% margin which was different from non Millenni al whites. African American Millennials were no different from the overall African American demographic as they voted for Obama by a 95% to 4% margin. Latino Millennials did have a significant difference in their voting behavior as they voted for Obama by a 76% to 19% margin. Figure 6.1 shows that whites voted for Romney at a 55% rate while voting for Obama at a 43% rate. It also shows that non whites voted for McCain at an 18% rate and Obama at an 80% rate. This shows another significant difference in voti ng behavior between white and non white voters in the 2008 election. The Pew study found that the African American voting population had its highest turnout ever in the 2008 election. That was later bested by the turnout in 2012. There was over a 5% incre ase in the amount of African Americans that voted. Whites represented the highest voter turnout percentage, but it was less than 1% more than the turnout of African Americans. The other ethnic groups remained below 50% for voter turnout.
71 There was a dispa rity between voting patterns of Millennials in 2008 and 2012 as the 2008 Millennials were significantly more likely to vote for Obama. White Millennials represented 11% of the total voting demographic in 2008, but they voted for Obama by a 54% to 44% margi n. This was 10% higher than they voted for Obama in 2008 while Romney received 7% more voted than McCain. African American Millennials were also significantly more likely to vote for Obama in 2008 than 2012. Figure 6.1 2008 Presidential Vote % by Race (CNN) T hey represented 3% of the total voting demographic in both years, but they voted for Obama by a 95% to 4% margin. This decreased in 2012 by 4% while Romney received 4% more of the African American Millennial vote in 2012. Latino Millennials also saw a decr ease in in likelihood of voting for Obama from 2008 to 2012. In 2008 they represented 3% of the total
72 voting population, which increased to 4% in 2012, and they voted for Obama by a 76% to 19% d to 23%. The comparisons between the 2008 and 2012 studies show different trends in voting behavior of Millennials. In 2008, they were significantly more likely to vote for Obama based off of age but also based off of race. Romney was able to increase th e amount of votes he received from all ethnic demographics by significant margins, but they were not enough to win the election. There are trends with how race might influence the voting behavior of individuals, but, given the shift from 2008 to 2012, it m ight not be the most influential factor in determining how an individual will vote. The expectation was that Millennials would continue the trend of being more likely to vote for Obama, but that was not shown based off of the findings from both elections. The findings of race as a whole still suggest that it does play a role in how an individual votes. African American and Latino voters were still more likely to vote for the Democratic candidate in both elections by a significant margin while white voters w ere more likely to vote for the Republican candidate in both elections. It is also shown that the Millennial generation of each respective race was significantly more likely to vote for the Democratic candidate than the ethnic demographic as a whole. This could possibly mean that something different influences why Millennials vote differently than other generations as a whole.
73 CHAPTER IX DISCUSSION Based off of the findings from the 2008 and 2012 elections, it appears that several of the variables play a role in influencing the voting behavior of Millennials. Comparatively speaking, it appears that they are significantly more likely to vote for Democratic presidential candidates. Religious influence seems to yield the most significant results in terms of voter predictability. Evangelical whites remained consistently the most likely religion to vote for the Republican candidate while Evangelical African Americans remained the most likely to vote for the Democratic candidate. Nonreligious voters were significantl y more likely to vote for the Democratic candidate while Protestants, as a whole, were more likely to vote for the Democratic candidate. Millennials could become more religious as they grow older, but that would have to be something looked at in the futur e (Pew Research 2012). The trends from Millennials are that they are the least religious generation ever and that trend might continue to an even less religious next generation. Looking strictly at the numbers between 2007 and 2012, there is a noticeable t rend in the decrease of major religion as a whole (Pew Research 2012). Seventy eight percent of the population said that they were Protestant in 2007 and that number has since decreased to 73% as recently as 2012. The other side of that trend has seen the number of unaffiliated religious groups increase by 4.3%. Most of that increase in unaffiliated religious groups is from people that label themselves is ar ound a 2% change, which would be roughly 6 million Americans. The trend is the fastest
74 growing religious shift in the history of the United States as unaffiliated Millennials are 32% unaffiliated while the next closest generation, Generation X, is around 2 1% unaffiliated. The growth before Generation X in religious change was around 6%, but this shows nearly double that change over one generational cohort. The breakdown within the Millennial generation even shows disparity between religiosity of the differ ent age groups within the generation. The 32% unaffiliated demographic is split between the older Millennials (1982 1989) and younger Millennials (1990 1994) in which the older Millennials are around 30% unaffiliated while the younger Millennials are aroun d 34% unaffiliated. If this trend continues then the next generation will see a larger spike in the overall number of unaffiliated individuals. Party affiliation of Millennials shows that they are not particularly aligned with either party with significan t support. Millennials are more likely to register as Democrats instead of Republicans, but the overall majority of Millennials still chose to remain independent. The results from the 2008 study show significant partisanship from voters that did identify w ith either party, but neither party have seen a huge spike in registration among Democrats. Determining why Millennials are more likely to remain independent is an area worth studying. The assumption with Millennials is that they are more progressive and identify more closely with Democrats on social issues, but also identify better with Republicans on economic issues. This split could explain why Millennials feel that neither party truly represents their value. As a whole Millennials are a lot less disenf ranchised by the government compared to other generations (Pew Research 2012). Millennials trust the government more than other generations as well meaning that they feel that the role of larger government might not be a bad thing.
75 Millennials are the gen eration most likely to view the Democratic Party in a positive light while they are also the most likely to view the Republican Party in a negative light. Abortion is an issue where Democrats are seen as more favorable by Millennials. Only 32% of Millennia ls see the Republican position on abortion as favorable while 52% see the Democratic position as favorable. Millennials also feel that Republicans do a poor job on the healthcare issue as 36% see the Republican position as favorable while 52% see the Democ ratic position as favorable. The only issue that Millennials see Republicans as the favorable party is on the budget and deficit. Millennials favor Republicans on the budget issue by a 46% to 39% margin. Compared to other generations, this is still relativ ely lower (Pew Research 2012). These Pew studies show that Millennials put a lot of weight on social issues since they are more likely to support the Republican Party on financial issues. If it is connected to religiosity, the argument could be made that the reason Millennials are more likely to favor Democrats is because they are less religious than other generations. Abortion was the biggest social issue that causes the Republican Party to be viewed negatively by Millennials while other generations saw i t as less of a significant issue. The controversy could be a result of increasingly nonreligious individuals. Even though Millennials are registering in lower numbers than other generations, voting along party lines is still common among those Millennials that are registered. Since Millennials are not registering in high numbers, they are not giving the Democratic candidate an advantage in that regard. The majority of independent voters seem to favor the Democratic candidate in both elections meaning that t here is likely another reason that Millennials vote for the Democratic candidate in greater numbers.
76 The same issue occurs with the belief that income could be the reason Millennials are more likely to vote for the Democratic candidate. The results from th is study show that it was difficult to find a connection to higher income and individuals voting for the Republican candidate. The only significant finding within this study was that lower income individuals were significantly more likely to vote for the D emocratic candidate in the 2008 and 2012 elections. Millennials are no more likely to be in economically dire situations compared to other generations so it is difficult to find a connection with Millennial income and an increase their propensity to vote f or the Democratic candidate. The economy as a whole might be a better indicator of Millennial voting behavior. Millennials tend to have a different perception of the economy and the expectations for it in the future. Millennials do not feel that their personal finances are better off than the Boomer or Silent generations, but they feel that it is better than that of Generation X. As a whole, Millennials are quite content with their financial situations compared to the perceptions of other generations du ring the same time in their lives. The biggest support for Republicans from Millennials comes on the budget issues, but other economic issues still find Millennials supporting Democrats. Millennials are significantly more likely to trust the Democratic Par ty with fixing the job situation and solving social security. Other generations do feel the Democratic Party is better suited for fixing social security in even greater numbers than Millennials showing that this is one of the few issues where Millennials a re more in line with Republicans than the Boomer generation. The connection between Millennials and income does not yield enough significant results to explain the increased likelihood of Millennials voting for the Democratic candidate. The
77 economic issue s connected with perception of income might play a marginal role, but it does not seem to be the primary reasoning behind voting behavior of Millennials. Race seems to be one of the biggest factors in determining why Millennials are more likely to vote for the Democratic candidate. Millennial whites had different voting patterns compared to whites from other generations. Finding a reason behind that could be another important indicator of why they are more likely than other generations to vote for the Democ ratic candidate. The other part is that the makeup of Millennials is significantly different than other generations. The Silent and Greatest generations both were comprised of about 80% whites while Millennials are the first generation to go below 60% of i ts total demographic being white. The biggest growth has come from the Latino population going from around 6% during the Greatest generation to around 20% in the Millennial generation. Aside from the Asian demographic, which has fluctuated between 4% and 8 %, the only generation that has seen a significant decrease in the demographic of its population is whites. African Americans have also seen an increase in their overall population from the Greatest generation to the Millennial generation with a 6% increas e. White Millennials are still more likely to vote for the Republican candidate compared to other ethnicities. In 2008, white Millennials did favor the Democratic candidate, but that changed to Republican candidate in 2012. Despite the increases in the Lat ino and African American population, that has not translated into higher voter turnout from those ethnic groups. The African American population saw a major increase in overall voter turnout from their demographic, but Millennial African Americans have not been part of that increase. Latinos have not seen a high voter turnout and Latino Millennials have not increased their turnout either.
78 This study has demonstrates that, of the 19% of the voting population that Millennials contribute to election, 11% of t he total vote still comes from white Millennials. African American Millennials contribute around 4% of the total vote, but that is above what is expected with the size of their population. The 2008 and 2012 election can be seen as special cases since the D emocratic candidate was a African American candidate. Latino Millennials and white Millennials were both more likely to vote for the Democratic candidate than other generations. It was only the African American Millennials that were more likely to vote for the Republican candidate than the overall African American population. It appears that there is a connection with race and voting behavior for Millennials, but the connection could still be questioned since ethnic minorities are not voting in higher numbe rs. It must also be examined why white Millennials are increasingly more likely to vote for the Democratic candidate than non Millennial whites.
79 CHAPTER X CONCLUSION Analysis from this study shows that Millennials do, in fact, have different voting tendencies than other generations and are more inclined to vote for Obama in the 2008 and 2012 elections. It was determined that as age increased, the individual was more likely to vote for the Republican candidate. The variables tested show only two results that cou ld explain why Millennials were more likely to vote for the Democratic candidate. Income and party affiliation demonstrates significant connections to Millennials and their voting behavior. The issue with income was that there was not a greater representa tion of rich or poor Millennials compared to other generations. Economic issues were important to Millennials as well, but they favored the Republican candidate on the budget deficit. Millennials do possess a much different perception of economics than oth er generations with a more positive outlook on the future. The other variable of party affiliation shows that the highest numbers of Millennials were independent voters. There were significantly more Democratic registered Millennials than Republican regi than other generations. One finding was that Millennials did see themselves identifying with Democrats more than Republicans, but still did not register with them in signif icant numbers. The shift with white Millennials favoring Obama in 2008 and Romney in 2012 shows that they are more open to aligning with either party depending on the circumstances. Religion was the first variable that shows significant results. The ANES, CNN, and Pew data show that Protestants as a whole were more likely to vote for the Republican candidate.
80 White Evangelicals had the highest percentage of Republican voters out of all religions while nonreligious voters had one of the highest percentages of Democratic voters. This was significant because the Millennials demographic has seen a decline in the total number of white Evangelical Millennials compared to other generations while seeing an increase in the amount of non religious Millennials. The o ther finding was that race does play a role in voting behavior of Millennials because the composition of Millennial ethnicity greatly differs from other generations. With significantly more ethnic minorities than other generations, Millennials have more in dividuals that are more likely to vote for the Democratic candidate. Race did play a role in determining how an individual will vote with whites being more likely to vote for the Republican candidate while Latinos and African Americans were more likely to vote for the Democratic candidate. One possible reason behind Millennial whites being more likely to vote for the Democratic candidate is because of religion. With an increase in the total number of nonreligious whites in the Millennial generation, it is possible that they are not swayed by socially conservative ideals. The study found that Millennials heavily favored the Democratic candidate on social issues. Millennials also saw Republicans as being extreme compared to the Democrats when it came to most social issues and that could disenfranchise them as voters. The ability to sway white Millennials could influence how Millennials vote in the future since there are so many independent voters in the Millennial generation. As for Latino and African American Millennials, it seems that they will likely continue the trend of voting for the Democratic candidate. The regression model from the ANES study shows that there was statistical significance for Millennials, religion and race while there was no statistica l significance for income. What
81 this shows was that during the 2012 elections, Millennials, non Evangelicals and non Whites were significantly more likely to vote for Obama than they were for Romney. The insufficient findings for income show that the total income did not have an impact on the voting behavior of individuals. With the increased number of ethnic minorities and nonreligious individuals in the Millennial generation, it appears that these are two possible reasons why the Millennial generation i s more likely to vote for the Democratic candidate than other generations. If the trend continues then it is likely that the next generation will be even more likely to vote for the Democratic candidate than Millennials. The main strengths of this study include information provided from the ANES study that provides accurate information that is more reliable than exit polls in determining the age, religion and vote. By breaking down the categories needed, the study examines the vote on a personal level and not having to rely solely on exit poll data. With the separation of categories done by ANES, the study is better able to find which religions have the most significant impact on voting behavior. The weakness of the design is determining whether religion is a significant indicator of election outcomes. There is the potential of no causal relationship between religion and a vote for either candidate; but that does not show a direct connection to either religion or generation. The other weakness of the study is the limitation of data available. While the ANES study does a good job of providing a strong and accurate representation o f the population, it is difficult to decide if its research represents the voting population. Without universal voting information from every person that voted in both elections, it is impossible to give perfect
82 measurement results. There are also reliabil ity issues within the research conducted from the ANES study because people could have misreported information. It is also possible that the study has a bias in the data because of which people were given the opportunity to take part in the study. There w as a huge movement within the Millennial generation that voted for the Democratic candidate that could explain why Barack Obama won both elections. The last implications aim to show that, since there has been a decrease in religious identification among Mi llennials, Barack Obama won the 2008 and 2012 elections because Millennials are more likely to vote for the Democratic nominee than the generations before them. Future avenues that could be conducted from this study include examining the voting behavior of the generation that follows the Millennials and whether they continued the voting trends of the Millennial generation or if they mirrored that of the generations before Millennials. There could also be research conducted in relation to the generations bef ore Millennials and examining whether these results were part of a trend that occurred in those generations. Breaking down voting behavior through generations could explain why a candidate was successful in a given election while also explaining how candid ates might change their policies in order to be more successful in their campaigns. Another avenue that might come from this study is comparing the importance of religion in voting behavior to other variable like race. One strong variable that shaped this study was how race has a significant impact in influencing voting behavior over religious ideology and generation.
83 REFERENCES America: 1952 1974 American Political Science Review : 70, 469 478. "ANES 2008 Time Series Study." ANES Data Center N.p., n.d. Web. 11 Apr. 2014. ANES > Data Center > Study Pages > ANES 2008 Time Series Study "ANES 2012 Time Series Study." ANES Data Center N.p., n.d. Web. 11 Apr. 2014. http://www.electionstudies.org/studypages/anes_timeseries_2012/anes_timeseries_2012. htm American Journal of Political Science : 44(1), 35 50 American Sociological Review : 62(6), 937 946. ing the Political Psychology : 33(5), 619 63 0 Political Analysis : 12(1), 76 96 Considine, David, Horton, Julie and Moorma Journal of Adolescent & Adult Literacy : 52(6), 471 481 http://www.pewresearch.org/fact tank/2014/03/07/6 new findings about millennials/ (June 14, 2014)
8 4 erica's Fiscal Future in the PS: Political Science and Politics : 43(2), 323 326. Presidential Studies Quarterly : 36, 223 242 Harvard Internat ional Journal of Press /Politics : 8(3), 9 34. Katosh, John and Traugott, Michael. 1981. "The consequences of validated and self reported voting measures." Political Opinion Quarterly : 45(4), 519 545. Political Opinion Quarterly : 61(2), 288 316 wo Electorates: Generational Journal of Politics : 62, 1014 1034. Presidential Vote: A Pooled Cross American Journal of Political Science : 32(1), 137 154 Political Research Quarterly : 61(2), 180 191. Political Research Quarterly : 62(2), 340 354
85 Nadeau, Richard and Lewis The Journal of Politics : 63(1), 159 181. Journal of Business and Psychology : 25(2), 281 292 Palfrey, Thoma American Journal of Political Science : 31(3), 511 530 between R Political Behavior : 30(3), 367 389 Pew Research Center for the People and the Press. 2011. "The Generation Gap and the 2012 Election." November 3. http://www.people press.org/2011/11/03/the generation gap and the 2012 election 3/ (June 14, 2014) Pew Research Centers Religion Public Life Project RSS. 2012. "How the Faithful Voted." November 7. http://www.pewforum.org/2008/11/05/how the faithful voted/ (June 18, 2014) Pew Research Centers Religion Public Life Project RSS 9. http://www.pewforum.org/2012/10/09/nones on the rise/ (June 14, 2014) Pew Research Centers Social Demographic Trends Project RSS. 2014. "Millennials in Adulthood." March 7. http://www.pewsocialtrends.org/2014/03/07/millennials in adulthood/ (June 18, 2014) The American Political Science Review : 62(1), 25 42.
86 Taylor, Paul and Lopez, Mark H. 2011. "Six take aways from the Census Bureau's voting report." Pew Research Center (March) http://www.pewresearch.org/fact tank/2013/05/08/six take aways from the census bureaus voting report/ (June 14, 2014) Traugott, Mi chae l W. and Katosh, John. 1979. Public Opinion Quarterly : 43(3), 359 377 Public Choice : 150(3/4), 511 524 Wells, Scott D. and Dudash, Elizabeth A. 2007. "Wha'd'ya Know?: Examining Young Voters' Political Information and Efficacy in the 2004 Election." American Behavioral Scientist 50, 1280 1 289.