Religiosity and Voting Behavior

Religiosity and Voting Behavior

Aim�e K. Gibbs

Abstract

Throughout time there has been a strong emphasis in the understanding of voting behavior. Religion is a strong indicator of an individual�s voting behavior. Religion has an important impact in how the voter looks at specific issues, such as abortion, homosexuality, the environment, and economics. The voter then uses his or her response to these issues as a guideline to help determine which presidential candidate to vote for. It is therefore reasonable to understand that religion has a significant influence on an individual�s voting behavior in presidential elections. This study looks at the influence of religion on various issues regarding abortion, homosexuality, the environment, and economic issues, and the vote for president in the 2000 and 2004 U.S. presidential elections.

Introduction

Voting behavior has been studied for quite some time and many conclusions are made, but what factor influences voting behavior the most? Some say that race is a factor, others that gender is a factor. Many studies, however, look at the influence religion, or lack thereof, has on the voting behavior of citizens, thus making religion one of the most important factors in determining voting behavior. Different religious groups also have differing opinions on issues regarding abortion, homosexuality, the environment and economics, thus leading to a variety of conclusions when looking at influences on voting behavior. Secular groups also have varying opinions on issues adding more confusion to voting behavior studies. In studying this wide spectrum of opinions on the issues of abortion, homosexuality, environment and economics, it is difficult to understand what drives individual voting behaviors.

Looking at the varying opinions of religious and non-religious groups is important in the current society to understand voter preferences and to understand the cultural differences that appear to be dividing the country. As it seems as though the country is dividing upon these cultural differences our government will reach a point of standstill when it is unable to meet the wants and needs of its citizens. It is also beneficial for those who are campaigning or running a campaign to know who will vote on certain issues and how they will vote. It is also important for those in office who are attempting to please their constituency as they will now be better equipped to understand the voting habits of their constituency. As the society changes rapidly with an influx of new ideas and issues, a study which looks at how issues are ranked in importance to different groups allows politicians and others to understand the influence of religion or lack of religion in voter behavior in order to adapt to a �modern� society.

Hypothesis

It is my hypothesis that religiosity will be a leading factor regarding voting behavior on issues regarding abortion or homosexual rights more so than on issues dealing with the environment, or economic issues due to the moral values connected with the former issues. I also hypothesize that Protestants are more inclined to vote on these issues than the two other main religious groups: Roman Catholic and Jewish, due to stronger religious connections to the issues and to the morality associated with such issues in Protestant denominations. It is possible to divide the Protestant group into Mainline Protestants and Evangelical Protestants, but for some of the issues being discussed I believe the viewpoints are similar within the specific religious groups under the larger category or Protestant and it is therefore unnecessary to divide the group.

My hypothesis stems from two theoretical traditions. The first of which is a social-psychological theoretical tradition. This tradition centers on demographics to determine which factors have the greatest influence on individuals. My hypothesis looks at the religious group membership to see if religion has a stronger influence on individuals than the individual�s membership in other demographic groups such as race, gender, class, etc., especially regarding the individual�s voting behavior. The second theoretical tradition my hypothesis stems from is the prospective voting theory. This theory centers on policies rather than on demographics. My hypothesis claims that social/moral policy issues are becoming more important than economic or environmental policy issues. While these two theories are quite contrasting they are connected in that there is a probable correlation between the increase in the importance of religion as a demographic characteristic and an increased focus on social/moral policy issues.

Literature Review

When looking at the voting behavior of American citizens many factors can influence the voter�s decisions. Studies often look at the influence of race, gender, and ideology. Many studies also emphasize the influence of religion on voting behavior. Though studies have looked at many variables in order to determine influences on voting behavior, I argue that religion has an overriding influence, especially on the issues of abortion and homosexuality that tends to permeate many of these other factors.

Racial influence on voting behavior is studied more often in recent years than previously. Branton (2004) looked at the effect of racial and ethnic diversity in initiative elections. She argues that race and ethnicity have a strong influence on many issues because of the origin of the issues in a white-dominated society thereby excluding some of the racial minorities and their opinions. Branton (2004) also looks at how white voters in areas dominated by minorities vote on these issues. Minorities and whites vote differently on the issues because they interpret the effects of the issues in a different manner.

There are many issues that are directly affected by race and ethnicity, for example initiatives dealing with affirmative action. Areas that are more racially diverse, such a major cities, are going to have more racially tolerant voters than areas, such as rural areas, that are not vastly diverse (Branton, 2004). Branton states that the voters in homogeneous areas �tend to be less tolerant, which can affect their attitudes toward racially relevant policies.� In this study the hypothesis is that racially relevant initiatives will show how voters will either support or oppose policies affecting racial/ethnic minorities based upon their own race or ethnicity (Branton, 2004).

Branton (2004) also takes into consideration initiatives that are not racially relevant, much like abortion. In these situations economics and benefits are the common driving factors that voters consider when voting on these initiatives. Branton claims that these are part of voter self-interest and that voters will vote on what they think will be best for their group. She claims �voters are even less likely to make the link between the vote and benefits for their own racial group, further diluting any impact of race or racial diversity on their vote� (Branton, 2004). Branton�s hypothesis in these regards is that race has little impact on initiatives that are non-racially relevant. This can also be similar with religion�s impact on issues in that both race and religion will be more effective regarding some issues more than others.

Branton (2004) found that when dealing with racially relevant initiatives, the race and ethnicity of an individual does have some significance. Branton states �the relationship between individual-level voting and diversity is consistently negative, which is in line with the expectations regarding the relationship between diversity and white voting on racially tinged ballot initiatives.� More importantly are Branton�s findings regarding non-racially relevant initiatives. After looking at ten different initiatives (�gambling, medicinal marijuana, parental rights, partial-birth abortion, jurisdiction over juveniles, and sentencing for controlled substance possession) Branton found that �racial and ethnic diversity is statistically only relevant in one of the 10 models�� thus supporting her hypothesis that there is no correlation between race and the voting behavior of these initiatives.

Johnson, Stein, and Wrinkle (2003) focused on the voting of Latino Americans. The Johnson et al. study focused on Mexican Americans in Texas and their voting tendency. Johnson et al. (2003) found that �the establishment of ties to an ethnic group in a majority-minority context over time mitigates the negative relationship between the use of Spanish as a primary language and voting.� Johnson et al. state that �voters who choose not to consume information or communicate in English find themselves at a disadvantage and are less likely to be mobilized or informed about the means of political participation� thus making them less opinionated in matters both racially centered and non-racially centered issues. However, Johnson et al. (2003) also claim that �individuals with strong cultural ties and � strong community ties [�] are particularly likely to participate in politics.� In conclusion, Johnson et al. (2003) found that there is a limited connection between the Latino race and voting behavior.

Gender has also been an important when studying the voting behavior of citizens regarding a variety of issues. Sanbonmatsu (2002) looks at the different stereotypes associated with men and women regarding �candidate beliefs, issue competency, and traits, and voter gender.� In regards to gender stereotypes Sanbonmatsu claims that the general population tends to view women as more capable to understand issues dealing with women�s rights, whereas men are more capable to deal with issues such as foreign policy and defense. Sanbonmatsu reports that their gender stereotypes influence a preference for certain genders in office. Sanbonmatsu (2002) found that in regards to issues that voters �who think women are more likely to take their position on abortion � a stereotypically female strength � are more likely to prefer the female candidate.� The voters, especially female voters, who are pro-choice and support women�s rights believe that a female representative will be more likely to vote pro-choice than a male representative. Ultimately, Sanbonmatsu (2002) found that �voter gender is only part of the explanation for voters� baseline preference; not only voter gender, but stereotypes about traits, beliefs, and issue competency explain the baseline gender preference.� The stereotypes about a issues such as abortion, or foreign affairs show a correlation with gender, thus affecting how people vote.

Kaufmann (2002) also looks at the effects of gender on voting behavior and how the gender gap between the parties has widened. Kaufmann (2002) states that �For women, the issues themselves � reproductive rights, female equality, and legal protection for homosexuals � have become increasingly important determinants of party identification. For men, the influence of cultural conflict on partisanship is argued to be equally pervasive.� Kaufmannn (2002) looks at the cultural factors that have pushed women to the Democratic Party while men tend to be a part of the Republican Party. Kaufmann (2002)looks at the fact that newer social policies regarding women�s rights and homosexual rights are threatening to a society based upon tradition, thus leading to the split between men and women between the parties.

In looking at the division between the genders and the political parties, Kaufmann (2002) looks at the opinions and cultural values on different issues. Kaufmann (2002) states that �women are generally more liberal than men across most of the issue dimensions and in particular on cultural and social welfare issues.� Kaufmann (2002) sees how cultural attitudes toward different issues such as women�s rights, homosexual rights, and abortion rights are different for men and women. Kaufmann (2002) claims that

There is strong evidence that the culture wars � in particular partisan polarization over abortion and homosexual rights � may have recruited additional women to the ranks of the Democrats. Even though men and women share similar attitudes on abortion and women�s rights, the greater emphasis that women place on these views tend to exacerbate the partisan differences between them. This can also be true for the influence of religions in that particular religious groups, such as Protestant groups, in that their overwhelming voting influence is what further separates the voters in the different parties. Ultimately, Kaufmann�s (2002) findings show that women are more interested in the particular issues such as abortion rights or homosexual�s rights leading them to be more liberal with partisanship whereas men are more conservative regarding the issues because their focus is on policies regarding social welfare. Kaufmann (2002) strongly argues that there are many gender implications when looking at culture issues in the United States.

Another key factor when looking at voting behavior is the ideology of the individual who is voting on a particular issue. One�s ideology reflects what is important to that individual whether it is religious influence, race/ethnicity, gender or another factor. Chressanthis, Gilbert, and Grimes (1991) look at the ideology and voting record of United States Senators on the issues of abortion which is a way to show how ideology has an important influence on voting behavior. Chressanthis et al. (1991) found that senators were often swayed based upon their own beliefs upon the issue. Chressanthis et al. (1991) argue �that ideology plays a prominent role in matters of social policy issues like abortion and school prayer.� It is important to look at the Congressional voting because it is a representation of ideologies throughout the country.

Chressanthis et al. (1991) support two different sources of ideological voting: promoting an idea may allow the individual to feel a sense of satisfaction if they feel they have improved the life of others, and that even if there is no effect on others, individuals may still be content in knowing they have �done the right thing.� Chressanthis et al. (1991) also found that a �lack of election competitiveness and voter control increases the likelihood that senators will pursue pure ideologically motivated actions�.� Also, by using ideology, the senator will not have to do as much research on the issue or the constituency being represented (Chressanthis et al., 1991). Using religion as one of the ideologies shows that some religions will support and others reject laws regarding abortion depending on the nature of the law (Chressanthis et al., 1991). Another ideology that was researched by Chressanthis et al. (1991) was the party affiliation each member of the Senate was associated with; Republicans are associated with more pro-life amendments whereas Democrats are associated with more pro-choice amendments (Chressanthis et al., 1991). Chressanthis et al. (1991) also look at geographic region as a factor of ideology.

Chressanthis et al. (1991) found that �The significance of the senator-specific religious affiliation does suggest that to some extent senators are motivated by individual-specific influences which are not derived from or affected by constituent interests.� Chressanthis et al. (1991) concluded that the ideology of a senator will have a larger impact than constituency opinion regarding their voting behavior on issues such as abortion. Ultimately it was determined that �the results clearly indicate that ideology is the most significant factor in explaining senator voting behavior� (Chressanthis et al., 1991).

Fastnow, Grant, and Rudolph (1999) looked at the effect of religion in the House of Representatives to see how religion affected vote choice. It was found that �members� votes represent both their [the representatives�] own religious affiliation, and the religious groups within their districts� (Fastnow et al., 1999). The goal of Fastnow et al. (1999) was to show how both specific issues and broad voting patterns are related to religion.

Fastnow et al. (1999) used the represented denominations to group together those denominations with similar beliefs and traditions to make the study more sufficient. In this study the authors used the highly religious topic of abortion to understand the impact of religion on a specific issue (Fastnow et al., 1999). Fastnow et al. (1999) also �expect religious tradition to affect the general ideological orientation of members.� When looking at the correlation between abortion legislation and religious affiliation, Fastnow et al. (1999) found that different religions affect how a representative will vote. For instance, Mainline denominations and Jewish groups will generally take pro-choice stances whereas Catholics, Evangelical groups, and Mormons will generally take a pro-life stance (Fastnow et al., 1999). Fastnow et al. (1999) also notes the effects of religion on abortion over time: �religious tradition is an important determinant of abortion voting, but that this direct effect has lessened as party has become more important.� Religion�s influence on broad voting is also notable as we tend to find that generally Evangelical Protestants and Mormons will vote conservatively while Catholics, Jews, and some Black Protestant groups vote liberally (Fastnow et al., 1999).

Fastnow et al. (1999) came to the conclusion that religion not only plays a role in the voting behavior of representatives, but also in the decision-making process of the legislature. Fastnow et al. (1999) determined that �Religion stands up to the competition of a variety of other theoretically powerful predictors of voting behavior, including party, constituency preferences, and important demographic indicators.� Religion also reflects the party affiliations, voting behavior, and attitudes of groups throughout the country, thus affecting the legislation brought forth from Congress (Fastnow et al., 1999).

Overwhelming amounts of literature have been devoted to the importance of religion in regards to voting behavior. It is apparent that there is a connection between how voters vote on issues and their religiosity, especially regarding Evangelical Protestant religious groups. Smidt and Kellstedt (1992) look at the influence of the Evangelical religious movement. Kellstedt and Smidt (1992) noted how there had been in influx of evangelicals in the political arena. At this point, many evangelicals associated themselves with the Republican Party whereas non-evangelicals tended to associate themselves with the Democratic Party (Kellstedt and Smidt, 1992).

Another factor at this point is the regional differences across the country (Kellstedt and Smidt, 1992). No matter the religion, southerners were less Republican than the rest of the country, and evangelicals throughout the country were more Republican than others (Kellstedt and Smidt, 1992). Kellstedt and Smidt (1992) noted that the South and the North continue to become more Republican over time, thus affecting their influence in voting behavior regarding certain issues. During Presidential elections Kellstedt and Smidt (1992) found that white evangelicals and nonevangelicals consistently voted for Republicans, but the influence of the region also affected the percentages of each group voting for the Republican candidate.

Ultimately, Kellstedt and Smidt (1992) found four conclusions regarding their study. The first conclusion was that there has been a shift in partisanship of white evangelical voters to the Republican Party (Kellstedt and Smidt, 1992). Secondly, Kellstedt and Smidt (1992) concluded �that white evangelicals have become more politicized during the past decade.� The third conclusion made by Kellstedt and Smidt (1992) was that �there is a growing divergence between the voting patterns of white evangelicals and white nonevangelicals, with evangelicals being more likely to cast their ballots for the GOP presidential candidate.� Finally, Kellstedt and Smidt (1992) concluded that the South may be one region where there is a growing split between the voting behavior of evangelicals and nonevangelicals.

Fox and Richardson (2001) look, once again, at the influence of religion on voting regarding the topic of abortion. Fox and Richardson (2001) used four variables to look at the voting behavior associated with abortion: �religious affiliation, party, age, and rural-urban constituency.� They found that religious affiliation was a better predictor of voting behavior than any of the other three variables (Fox and Richardson, 2001). Fox and Richardson (2001) also establish some of the key beliefs regarding abortion. The Catholic Church opposes laws that support abortions, Protestants generally do not support laws supporting abortion, though it frequently depends on the law, and Mormons are not likely to support abortion laws due to their notions regarding life (Fox and Richardson, 2001).

Fox and Richardson (2001) found �that the only variable that consistently allowed prediction of voting behavior is religious affiliation.� When other variables were included in this study, there was little significance in voting behavior (Fox and Richardson, 2001). Fox and Richardson (2001) concluded that �religious affiliation is always the best predictor, even though in some cases its effect is small.� Though some religious groups have a stronger stance than others on abortion issues, those who have a religious affiliation are more apt to vote against abortion than those without a religious affiliation (Fox and Richardson, 2001).

Brooks and Manza (1997) study the relationship that religion has with political behavior. Brooks and Manza (1997) emphasize the fact that religion has a larger effect on voting than any other factor such as class. Looking at post- New Deal elections, Brooks and Manza (1997) note that the Democratic Party has been greatly associated with the Jewish and Catholic religions, whereas the Republican Party has been associated with mainline Protestant religions (Brooks and Manza, 1997). Of their many theses, Brooks and Manza (1997) look at �the political behavior of specific religious groups in the electorate.�

Brooks and Manza (1997) explore theses such as the Christian Right Thesis, the Catholic Dealignment Thesis, and the Liberal/Mainline Protestant Dealignment Thesis. In the Christian Right Thesis Brooks and Manza (1997) discuss the realignment of conservative voters of a Protestant denomination. The Catholic Dealignment Thesis looks at the prospect of Catholic voters to a neutral position from a formerly Democratic position (Brooks and Manza, 1997). Brooks and Manza (1997) explain that in the Liberal/Mainline Protestant Dealignment Thesis the extremely Republican religious voters are often overrepresented. Brooks and Manza (1997) ultimately found that the religious affiliation with politics is greater than any correlation between race, gender, or class and politics. It was also concluded that �it is [�] a rare event for a specific religious group to move decisively toward or away from support for a major party�s presidential candidates.� Religion has a profound influence on the party affiliation and voting tendency of individuals, though change has been occurring slightly over time (Brooks and Manza, 1997).

Several other articles have focused on the influence of religion on voting behavior. One such article is �A Catholic Vote?� by Reverend Andrew M. Greeley (2004). Greeley (2004) looks at abortion issues and claims �Most Americans [�] are neither consistently pro-life nor consistently pro-choice. Rather, they are ideologically inconsistent, approving legalized abortion when there is a serious threat to the mother�s health and rejecting it when it is merely abortion on demand.� Greeley (2004) also notes that within different religious groups, especially the Catholic religion, there is an inconsistency in voting for abortion laws. Greeley (2004) finally states �I consider data inviolate in the face of ideological prejudice� showing how data and ideology can coincide with politics.

It is understandable to see a relationship between religion and voting behavior; however it is limited in the most recent research. The most recent election, the 2004 election, has not been figured into these studies. These studies have also not looked at how religion affects voters regarding a variety of issues � which issues are more important to religious voters and to secular voters. This can also be further broken down into the various religious groups and their emphasis on particular issues. My hypothesis regarding the influence of religion on voters and their choice to vote on certain issues more than others will address the incomplete information mentioned above. It will support the idea that religion does affect voting behavior in the fact that some issues are more important to religious voters than secular voters, and that certain religious groups will be more inclined to vote on certain issues than other religious or non-religious groups.

Research Design

This study will look at the National Election Studies put forth by the University of Michigan. The 2000 National Election Study and the 2004 National Election Study will be used. These studies will be looked at in order to understand the voting behavior of all citizens on the issues of abortion, homosexual rights, environmental issues, and economic issues. I will be looking at specific questions regarding the implementation of laws or policies to understand the voting habits of the respondents. I expect to find that religiosity will be more influential in predicting voting behavior on issues regarding abortion and homosexuality than issues regarding the environment and economic issues. This also implies that people who do not find religion important will be less compelled to vote on those issues. I also expect to find that the people within Protestant religious groups will be more apt to participate in voting on the issues of abortion, and homosexual rights than the other religious denominations. This voting behavior prediction is due to the close religious ties and influence of religion on daily life in the Protestant religions with the issues of abortion and homosexuality. The data found in these studies will reflect the voting behaviors of a wide variety of citizens across the United States and will include a variety of issues on which the voters could vote. They will also include information regarding other factors that can/do influence voting behavior. Key variables that will be studied include the different religious affiliations, how important individuals view religion in their life, which presidential candidate the individual voted for, as well as views on abortion, homosexual, economic and environmental issues.

I first looked at the 2004 National Election Study in order to understand the voting habits of the respondents and then looked at the 2000 National Election Study to see if this was true in prior elections. To look at the different issues I looked at specific questions regarding the issues. For abortion in the 2004 National Election Study I looked at the following questions: the importance of the abortion issue to the respondent, does the respondent favor or oppose government funds be used to pay for abortions, and does the respondent favor or oppose a ban on late-term/partial-birth abortion; for issues regarding homosexuality: respondent�s position on gay marriage, should laws protect homosexuals against job discrimination, should homosexuals be allowed to serve in the U.S. armed forces, and should homosexual couples be allowed to adopt children; for issues regarding the environment: respondent�s self-placement on environment versus jobs tradeoff scale, and the importance of environment/jobs issues to the respondent; and for economic issues: the respondent�s self-placement on private or government insurance, federal spending on welfare programs, and federal spending on social security. To understand the religious factors the questions used were �is religion an important part in the respondent�s life� and �if the respondent attends religious services which major religion do they classify their services.� The dependent variable used in both studies was �if respondent voted in the election, who did respondent vote for� (in this instance President George W. Bush or Senator John Kerry). This variable allows us to show the correlation between how the respondent voted for the presidential candidate and their opinion on the issues of abortion, homosexuality, environment and economics. It also allows us to determine how strongly religiosity influences issue-based voting. Similar, if not exact, questions were used when studying the 2000 National Election Study.

To determine how strong of a correlation there is between the factors looked at in the 2004 election, a look at a previous election dealing with similar issues was needed, therefore I looked at how influential these variables are in the 2000 National Election Study. Again, I looked at specific questions regarding each of the issues in order to best gauge the influence of religion and other variables on the larger issues of abortion, homosexuality, environment and economy. Abortion questions include the following: the importance of the abortion issue to the respondent, does the respondent favor or oppose a law in his/her state requiring girls under the age of eighteen to have parental permission in order to have an abortion, and does the respondent favor or oppose a law banning late-term/partial-birth abortions; for homosexuality issues: should laws protect homosexuals against job discrimination, should homosexuals be allowed to serve in U.S. armed forces, and should homosexual couples be allowed to adopt children; for environmental issues: respondent�s self-placement on environment versus jobs scale and does the respondent favor or oppose tougher government regulations on businesses to protect the environment; and for economic issues: the respondent�s self-placement on private or government insurance, should federal spending on welfare programs increase or decrease, and respondent�s opinion of federal spending on Social Security. Again, to understand the religious factors, the questions used were �is religion an important part in the respondent�s life� and �if the respondent attends religious services which major religion do they classify their services.� Again I incorporated the question regarding who the respondent voted for in the election � Vice-President Al Gore or Governor George W. Bush � in order to show the correlation between the issues and how the respondents voted for their presidential candidate.

Results of Analysis

My first hypothesis is that people who feel that religion is important in their life will be more apt to vote on issues regarding abortion and homosexuality than on issues regarding the environment or the economy. Here I split the data set in order to look at the voting behavior of the respondents claiming that religion was an important aspect of his or her life (see Appendix A) with the voting behavior of the respondents claiming that religion was not an important aspect of his or her life (see Appendix A). When looking at the correlations between how respondents who claim that religion is important in their lives voted for President and their opinions on the twelve issues regarding abortion, homosexuality, environment and economics, five of the twelve issues held a strong significance. Of the five issues holding strong significance, three of the issues dealt with abortion, one dealt with homosexuality, and one dealt with the environment. Abortion appears to be the strongest factor regarding how people who find religion important vote. Abortion was also the strongest factor regarding how people who do not find religion important voted, but it was the only issue with a strong significance of the twelve issues. As my hypothesis suggested, those who find religion an important part of their lives are more apt to vote on these issues, especially abortion and homosexuality. For example, not only do people who are religious and pro-life vote for George W. Bush, but those individuals who are religious and pro-choice will vote for John Kerry. As only one of the twelve issues tested showed a significance with those who do not find religion important shows that there is not as strong correlation between the lack of religiosity and how one votes.

I took my first hypothesis one step further in saying that people who associate themselves with a Protestant religion are more apt to vote on issues regarding abortion and homosexuality than people associating themselves with other religious faiths. I also looked at the issues regarding the environment and economics issues to see if my hypothesis remained true for those issues as well. For this information I split the data to study those who claim that they are Protestant with those who associate themselves with other religious groups (see Appendix A and A). For those who claim themselves as Protestants had strong significance with four of the twelve issues; three of these were the three issues regarding abortion and the fourth regarded homosexuality. For those who associate themselves with religious groups other than Protestant had only one issue with strong significance and it regards abortion. As there are more issues strongly connected with how Protestants vote than how other religious groups vote my hypothesis is once again supported in that Protestants will be more inclined to vote on the issues of abortion and homosexuality than those individuals belonging to other religious groups.

In order to see if my hypotheses held true in a previous election I also tested them in the 2000 National Election study. My first hypothesis was to determine if there was a correlation between how important the respondent believes religion is in his/her life and the issues of abortion, homosexuality, environment and economics. Again, I first divided the data set in order to study those who claim that religion is an important aspect of their lives with the issues of abortion, homosexuality, environment and economics (see Appendix B) with those who do not consider religion to be an important aspect of their lives (see Appendix B). For those who claim that religion is important in their lives six of the eleven issues have a strong significance. The issue with the strongest significance and beta value is one regarding economics, but the others dealt either with abortions or homosexuality. For those who claim that religion is not important only two of the eleven issues have strong significance, however both of them are issues regarding homosexuality. Though my hypothesis is upheld, one does have to consider that homosexuality was important to voters who held religion as an important part of their lives and those who do not consider religion to be important in their lives.

Taking this hypothesis one step further, I looked at the issues in comparison with the major religions; keeping in mind my second hypothesis that Protestants will have a stronger relation with the issues and will be more likely to vote on them than other religious groups. Once again, I divided the data set into those who claim to be Protestant and those who associate themselves with other religious groups in order to best distinguish the differences in voting behavior between these two groups of voters (see Appendix B and B). For both groups of voters there are four issues of the eleven that hold a strong significance. Also for both groups the same three issues are significant with strong beta values � one regarding economics, one regarding abortion, and one regarding homosexuality. The fourth issue of significance for Protestant voters regards the environment while the fourth issue for other religious groups regards abortion. In this instance my hypothesis is not upheld because other religious groups have the same or more significance with issues of abortion and homosexuality.

Conclusions

Throughout the study it becomes apparent that religious influence is noted in many aspects of voting behavior. The results of my data analyses show that in fifty-seven of ninety-two total tests, my hypothesis is supported in that regarding issues of abortion and homosexuality religion is one of the strongest factors in determining voting behavior. The results of my analyses show that religion has a strong influence on which policies citizens use to determine their vote. My results also supported my second hypothesis in the 2004 election, but not the 2000 election, in that especially on issues of abortion and homosexuality, Protestant religious groups frequently have a stronger correlation with the issues than other religious groups. As the studies showed, religion has been an important factor when voting on the issues regarding abortion and homosexuality in both the 2000 and 2004 elections, thus showing that there has been a correlation for more than just a few years. Though my hypotheses are generally upheld, there seem to be some exceptions, for example in the 2000 National Election Study other religious groups held a stronger significance with abortion issues than Protestant groups. There is also the oddity that in the 2000 National Election Study, when religion of any denomination was a factor there held a strong correlation with federal spending on Social Security.

My study is still incomplete, however. I was limited in that I was not able to look at other demographics such as class, gender, or race in order to support the fact that religion is becoming an increasingly more important demographic category. My paper more easily supports my hypothesis in that I was able to compare policy issues to find that there was a stronger influence of religion on the moral issues than on the issues regarding economics and the environment (twenty-two significant findings compared to five significant findings). Since I was unable to incorporate other demographics, I am not able to compare the importance of religion as a demographic with other demographic categories to strengthen my hypothesis that religion is one of the strongest factors in determining voting behavior. In order to gain more understanding about these exceptions a more detailed study about the specific variables should be done, as well as studying these same variables in more election studies. A study that includes other demographic variables should also be explored. An analysis of religious influence on the voting behavior regarding issues of abortion, homosexuality, environment and economy over time in previous elections will also create enough data to create a more concrete theory about the connections between religion and the voting behavior of citizens on these issues and how it affects their presidential voting behavior. As religion seems to be a significant factor regarding abortion and homosexuality it seems as though religious beliefs should be considered in some way when creating legislation or rulings on issues regarding abortion or homosexuality. My hypotheses were generally upheld revealing that religion is still an important influence in the way that people vote on certain issues, namely abortion and homosexuality issues, and as such should be considered when analyzing the public opinion on such issues.

References

Branton, Regina P. (2004, Fall). Voting in initiative elections: Does the context of racial and ethnic diversity matter? State Politics and Policy Quarterly. 4(3), 294-317.

Brooks, Clem and Jeff Manza. (1997, July). The religious factor in U.S. Presidential election 1960-1992. American Journal of Sociology, 103(1), 38-81.

Chressanthis, George A., Kathie S. Gilbert, and Paul W. Grimes. (1991, September). Ideology, constituent interests, and senatorial voting: The case of abortion. Social Science Quarterly. 72(3), 588-600.

Fastnow, Chris, J. Tobin Grant, and Thomas J. Rudolph. (1999, December). Holy roll calls: Religious tradition and voting behavior in the U.S. House. Social Science Quarterly. 80(4), 687- 701.

Fox, Sandie Wightman and James T. Richardson. (2001). Religion and voting on abortion reform: A follow-up study. Journal for the Scientific Study of Religion, 14(2), 159-164.

Greeley, Andrew M. (2004, December). A Catholic vote. America, 191(18), 6.

Johnson, Martin, Robert M. Stein, and Robert Winkle. (2003). Language choice, residential stability, and voting among Latino Americans. Social Science Quarterly. 84(2), 412-424.

Kaufmann, Karen M. (2002, September). Culture wars, secular realignment, and the gender gap in party identification. Political Behavior. 24(3), 283-307.

Kellstedt, Paul and Corwin Smidt. (1992). Evangelicals in the post-Reagan era: An analysis of Evangelical voters in the 1988 Presidential election. Journal for the Scientific Study of Religion. 31(3), 330-338.

Sanbonmatsu, Kira. (2002, January). Gender stereotypes and vote choice. American Journal of Political Science. 46(1), 20-34.

Appendix A

Appendix B