Martin N. Ségur-Cabanac

The United States of America today are home to one of the most religious societies in the world; it certainly is the most religious western democracy. Despite increasing secularisation and ever more abandoned pews in modern Europe, it defies the sociologist theory that modernisation and secularisation necessarily go hand in hand. And for the past five years it has been led by the most religiously outspoken president in recent history. This, together with what some perceive as a rise of the “religious right”, has fuelled both media speculation and academic research into the area. Around the world, George W. Bush’s black-and-white doctrine of “you’re either with us or against us” in his “fight against terror” has given dangerous credence to the idea of a global war between (radical) Christianity and (radical) Islam. This, in conjunction with the controversy surrounding the 2000 U.S. presidential election, has probably made its successor the most globally media-covered election in history.

Increasingly, the question of religious faith and especially the frequency with which a person attends sermons has become the most precise indicator (and therefore most cherished by sociologists and political strategists alike) of voting behaviour in the U.S. Since the more devout churchgoers traditionally tend to vote Republican, the “religious right” has tried to hijack George W. Bush’s re-election, as Richard Land, a prominent Southern Baptist, claims that “white evangelicals were the driving engine” of Bush’s success. And he does have a point, since ever more politically organised religious groups took the election to the pulpit, effectively educating their congregation on who to vote for, thereby risking the violation of U.S. tax law and blurring the constitutional line of separation between church and state. And an exit poll conducted by Edison Media Research and Mitofsky International for the National Election Pool, a consortium of major U.S. television channels and the Associated Press, purports that the question of “moral values” was the most important factor in decision making in the election.

While it cannot be denied that religion did play an important role in the 2004 U.S. presidential election, it certainly cannot be regarded as the decisive issue. It will be shown that religion in the U.S. is active and important in the political arena, but that its growing influence in elections is a result of interpretations of statistical findings and subsequent categorisation, and the jumping to political action of the conservative and religiously devout because of the implied power given to them. In addition, it will be demonstrated that the National Election Pool exit poll “moral values” question is unscientific and consequently has to be disregarded as an indicator of voting behaviour. Finally, the much more precise and revealing statistic presented by John C. Green of the Bliss Institute at the University of Akron will be used to make an attempt at indicating the actual role religion played in the election.

Religion in America

The French aristocrat Alexis de Tocqueville commented in his two-volume study of the American people and their political institutions, Democracy in America, that “Religion in America takes no direct part in the government of society, but it must be regarded as the first of their political institutions.” This is still an apt description of the role faith plays in the U.S. today, more than oneand-a-half centuries after it was put forth by Tocqueville, even though the constitutional separation of church and state as it is put into practice might suggest a secular society. For, upon examination of the basis of this separation, it reveals itself to be not quite as rigid as one might expect in a country where no religious artefact may adorn a public school classroom. It is written in article six, clause three, of the United States Constitution that “no religious Test shall ever be required as a Qualification of any Office or public Trust under the United States.” And the First Amendment to it says that “Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof.” These two statements have become the pillars upon which the separation of church and state in the U.S. rests. Considering the increasing influence religion seems to be having in the political arena, it is interesting to note that the wordings do not include a vice versa clause. Therefore, it can be deducted that the original purpose of these statements was to remove the influence of the government on religion and thereby guarantee the freedom of religion in society (as opposed to state churches in the Old World), but not necessarily to imply that religion shall not exercise any influence on the government.

Indeed, there can never be a complete separation between church and state because both by implication have to deal with the same fundamental questions of morality and virtue, or at the very least are motivated by them. For, on the surface, it does not matter so much if one deems theft or murder to be wrong because it is stated in the Ten Commandments or because careful philosophical reasoning and causal deduction tells one so, as long as one comes to the conclusion that they are indeed wrong. A society based on morality clearly is preferable to one void of fundamental principles of human behaviour and peaceful co-existence, as has been shown repeatedly throughout human history. However, such values can become dangerous when grounded not on reason (and therefore debatable and introspective) but on blind belief in scripture and dogma (and therefore above and beyond all but basic analysis and interpretation), as they then become non-negotiable in the face of differing or outright different values. George M. Marsden, professor of history at the University of Notre Dame, Indiana, argues that the irony is that,

such a sense of the importance of virtue, which the Puritan heritage helped provide, can lead to an arrogant moral superiority that transgresses the very rights of others that the moral system claims to protect.

Such a perceived American arrogance has been one of the root causes for the growing gap between the U.S. and western Europe, as well as many other regions around the world, and has prompted the director of the Foundation for Strategic Research in Paris, François Heisbourg, to doubt if the United States and Europe can still be regarded as part of the same civilisation.

This religious heritage of morality is important in trying to understand the influence religion has on presidential elections in the U.S., as it is the growing distancing of society and politics from this very heritage since the social revolution of the 1960s that has prompted the re-emergence of what some call a politically active “religious right”. Public scandals like Nixon’s “Watergate” in the 1970s, Reagan’s “Iran-Contra” affair in the 1980s, and, more recently, Clinton’s impeachment, the Enron and WorldCom bankruptcies, and (even though not strictly a U.S. issue) the United Nation’s “Oil for Food” scandal have, it seems, persuaded the religiously conservative to enact countermeasures to bring morality back into politics. Indeed, this objective was one of president Bush’s strategies to win the election against Al Gore in 2000, as he proclaimed that he would “return the highest standards of hono[u]r to the highest office in the land.” The conservative protestant “Christian Right”, as Manfred Brocker, professor for politics and religion at the Catholic University of Ingolstadt, Germany, identifies it, wants to bring America back to the 1950s of president Eisenhower, and has taken up massive political lobbying in Washington of quantitatively and qualitatively hitherto unseen dimensions. Incidentally, it is president Eisenhower who in 1954 added the phrase “one Nation under God” to the Pledge of Allegiance, a line that has prompted Californian lawyer and founder of the “First Amendment Church of True Science (FACTS)” Michael Newdow to go to court claiming it violates the first amendment. In November 2005 he also filed a lawsuit against the words “In God We Trust” printed on U.S. currency. As of the writing of this text, both lawsuits have yet to yield a final verdict.

It is important to understand this climate of a perceived lack of moral clarity in both the political and business arenas that the 2004 U.S. presidential elections took place in. Outside of the U.S., it was mostly foreign policy – especially the war in Iraq – which people following the election on 2 November 2004 cared about. But on a national level, other topics seemed to encourage both voter turnout and behaviour, as the following exit poll suggests.

The National Election Pool exit poll

On election day, 2 November 2004, the pollsters Edison Media Research in partnership with Mitofsky International conducted an exit poll in all of the 50 U.S. states. A representative group of 13,660 voters were asked for personal information, including race, gender, income, education, religion, what they thought of the candidates and their election campaigns, and who they voted for and why.

One of the questions asked in the poll was what the “most important issue” was for determining who respondents gave their vote to. In order of magnitude, 22 percent of voters cited “moral values” as being most important, 20 percent the “economy” and “jobs”, 19 percent “terrorism”, 15 percent “Iraq”, 8 percent “health care”, 5 percent “taxes”, and 4 percent “education”. On the surface, this means that more people – a fifth of all respondents – gave their candidate their vote because of “moral values” than because of any other of the given reasons. As “moral values” can broadly be defined to incorporate issues such as abortion, gay rights, stem cell research, euthanasia, and human cloning – all of which the religiously devout tend to have strong opinions on – this does seem to make a firm case for a significant impact of the “religious vote” on the election. As Kevin J. McMahon, David M. Rankin, Donald W. Beachler, and John Kenneth White, all university professors, argue, if the “moral values” voters would have been excluded from the election, “Kerry would have won in a landslide.” And indeed, a re-calculation of the election result state by state shows that Kerry would have amassed 436 of the 538 electoral votes, as well as 54.7 percent of the popular vote. However, this reasoning rests on an erroneous assumption: that the “moral values” issue represents a scientifically valid statistical question. As Gary Langer, Director of Polling for ABC News, points out, the questions of “morals” and “values” represent “personal characteristics”, whereas the other issues (economy/jobs, terrorism, Iraq, health care, taxes, and education) are “discreet political issue[s].” Mixing these together distorts the exit poll data, rendering the “most important issue” findings pseudoscientific and therefore useless. Furthermore, the issue of the Iraq war certainly is not free of moral implications, and neither is the war on terror or the subject of education. And if a respondent was unable to give an answer to the question, e.g. because he or she voted a given way because his or her whole household did, he or she was arguably much more likely to choose the “moral values” option than any other, precisely because of its vagueness. Even if one is prepared to accept the “moral values” question as valid, the other issues then need to be more broadly categorised in order to create a more level statistical playing field. Combining “Iraq” with “terrorism” into the issue of “national security” (or, alternatively, “foreign policy”), and the “economy/jobs” issue with “taxes” into a broader “economy” issue, the picture is different yet again. Now, in the order of magnitude, the most important issues are “national security” with 34 percent, followed by “economy” with 25 percent, and “moral values” with 22 percent at third place. While this may still mean that more than a fifth of voters were motivated by “moral values”, it takes away its overly prominent (and erroneous) first place.

Another question asked by the pollsters further helps to refute the picture drawn by the “moral values” question: what the “most important quality” in a president is. Here, the personal quality of “religious faith” comes second to last, with only 8 percent of respondents deeming it the most important quality. Most significant is that a president “will bring change” with 24 percent, followed by his “clear stand on issue[s]” and his being a “strong leader” with 17 percent each, his being “honest and trustworthy” with 11 percent, and that he “cares about people” with 9 percent. Astonishingly, only 7 percent of respondents cited “intelligen[ce]” as the most important quality in a president. If the purported one-fifth of “moral values” voters does represent an uprising of a “religious right”, surely more respondents would have picked “religious faith” as the prime quality in a presidential candidate.

Two further questions asked in the exit poll could have been included in the “most important issue” question, since they represent discrete political issues while arguably still falling under the category of “moral values”: the questions of “abortion” and “same-sex couples”. According to pollsters, 21 percent of voters thought that abortion should “always [be] legal”, while 34 percent thought it ought to be only mostly legal”. Another 26 percent wanted abortion to be “mostly illegal” and 16 percent meant it should “always [be] illegal”. As the percentage gap between the two camps (“always” and “mostly” legal versus “always” and “mostly” illegal) is not that big, with those “pro-choice” representing a slight majority, abortion might be considered a critical because closely contested issue in the election. Under the assumption that those believing abortion should be illegal under any circumstances do so out of religious beliefs, the data suggests that a little over one-sixth of voters were religiously motivated. The question of a “policy towards same-sex couples” suggests an ever higher number. 25 percent of respondents thought that gay couples should be allowed to “legally marry”, while 35 percent supported at least “civil unions”. 37 percent wanted “no legal recognition” for same-sex couples. Assuming again that the more religiously devout would surely oppose any legal standing for homosexual partnerships, over one-third of voters with that opinion might be considered religiously motivated. This represents a significantly higher number than the “moral values” question suggests. Under the stated assumptions, religiously motivated voters prefer Bush by an overwhelming margin: 77 percent of those wanting abortion to be “always illegal” and 70 percent of those opposing legal recognition for same-sex couples voted for the incumbent president.

The National Election Pool exit poll also queried the religious affiliation of voters. Nationally, 54 percent of respondents identified themselves as Protestant, of which 59 percent voted for the incumbent president. Even though Senator John F. Kerry is Catholic, the 27 percent of Catholic respondents gave 52 percent of their votes to George W. Bush. Only among the bottom 20 percent of respondents, of which 3 percent were Jewish, 7 percent from other religions, and 10 percent not religious at all, did Kerry garner majorities. However, those groups were much more clearly divided, with 74 percent of the Jewish and “other” votes and 67 percent of the non-religious votes going to Kerry. Even though Bush managed to win more than half of the Protestant and Catholic votes, the margins are by no means large enough to justify any sentiment of an uprising of a “religious right” to take back their country.

Much more telling is the data combining voting choice with church attendance rather than religious affiliation. Of the 16 percent of respondents claiming to go to church “more than weekly”, 64 percent voted for Bush. The 26 percent that go to church “weekly” also favoured the president by 58 percent over Kerry. While those 14 percent attending sermon “monthly” were evenly divided, the 28 percent who go to church “a few times a year” and the 15 percent who “never” attend religious ceremonies favoured Kerry with 54 percent and 62 percent respectively. Accordingly, it seems that, no matter what voters’ religious affiliation was, the more often and more regularly they went to church, the more likely they were to vote for George W. Bush. This, combined with the data presented earlier, refutes the argument put forth by Marsden that the “religious dimensions” of the United States can be viewed as “contest between those who can be considered as broadly in the Puritan-evangelical tradition and those who are not.” Therefore, a different kind of poll is needed to describe more accurately the causal relationship between religion and voting choice in the 2004 presidential election.

The Fourth National Survey of Religion and Politics

John C. Green, a political scientist at the University of Akron, Ohio, and director of the university’s Ray C. Bliss Institute of Applied Politics, conducted the Fourth National Survey of Religion and Politics in the spring of 2004 with 4,000 interviews, as well as a post-election survey in November and December of 2004 with 2,730 re-interviews. The purpose of this study was to “provide a baseline for analy[s]ing the underlying impact of religion during and after the 2004 campaign.” As he has been carrying out these surveys for every presidential election since 1992, he can be considered an expert on the matter.

Based on religious affiliation and race, Green identifies twelve voter groups: Evangelical Protestants, Mainline Protestants, Latino Protestants, Black Protestants, Catholics, Latino Catholics, Other Christians, Other Faiths, Jewish, Unaffiliated Believers (no religious affiliation but nonetheless a high level of belief), Seculars (no affiliation but modest beliefs), and Atheists/Agnostics (non-theistic). He furthermore defines three subdivisions for Evangelical Protestants, Mainline Protestants, and Catholics: Traditionalists, Centrists, and Modernists. This serves to incorporate the concept of the importance of church attendance and adherence to religious doctrine on voting behaviour discussed earlier. Green sees Traditionalists as those who display a high level of orthodox beliefs, attend church regularly, and have “a desire to preserve such traditional beliefs and practices”, whereas Modernists follow a much more heterodox approach and attend church less frequently. The group of Centrists represents those who fit in neither of the two other groups and have mixed feelings about their religious beliefs.

Of the 26.3 percent of respondents identified as Evangelical Protestants, 12.6 percent are Traditionalists, making them the single largest religious group in the U.S. today. Green suggests that this is the group now commonly referred to as the “religious right”. Of those Traditionalist Evangelical Protestants, an overwhelming 88 percent voted for George W. Bush, making this group his single largest religious constituency. The second-largest Republican voter group was Other Christians with 80 percent, followed by Traditionalist Catholics with 72 percent and Traditionalist Mainline Protestants with 68 percent. The combined votes of those four groups giving more of two-thirds of their votes to Bush amount to almost one-fifth of the incumbent president’s total vote. For the Democratic candidate, the most important religious constituency was Black Protestants with 83 percent, followed by Atheists/Agnostics with 82 percent, Modernist Mainline Protestants with 78 percent, Other Faiths with 77 percent, Jews with 73 percent, and Seculars with 70 percent. The combined vote of those top groups for Kerry amounted to more than one-fifth of his total vote. To correctly analyse the importance of religion for these voter groups, the “most important issue” question needs to be raised again.

Here, Green relies on roughly the same type of conflated poll suggested earlier. Respondents were queried on three issues (or, rather, issue categories): “social issues” (e.g. abortion and gay rights), “foreign policy” (e.g. the Iraq war and terrorism), and “economic issues” (e.g. taxes and jobs). Firstly, they were asked to define the relative importance of each of these issues for their voting choice by rating them as “very important”, “somewhat important”, and “not very important”. Secondly, interviewees were asked to choose the single issue that was “most important” in their voting choice. Since Green chooses to use the label “social issues” rather than “moral issues”, he prevents the ambiguous interpretation of a “moral values” vote that was a result of the National Election Pool exit poll. The four top religious groups for the incumbent candidate also turned out to be the ones with the highest percentage deeming “social issues” to be “very important”, with Traditionalist Evangelical Protestants at 78 percent, Traditionalist Catholics at 68 percent, Other Christians at 67 percent, and Traditionalist Mainline Protestants at 55 percent. Even though the numbers for “social issues” when it comes to choosing the most important issue are significantly lower, they are still at the top of the statistic with those groups. 47 percent of Traditionalist Evangelical Protestants thought “social issues” was the most important topic, followed by Other Christians with 38 percent, Traditionalist Catholics with 39 percent, and Traditionalist Mainline Protestants with 30 percent (here, Latino Protestants show a slightly higher percentage than Traditionalist Mainline Protestants, with 33 percent). Overall, the most important issue for the respondents was “foreign policy” with 35 percent, followed by “economic issues” with 33 percent, and “social issues” in third place with 24 percent. Even though, as mentioned, Green does not face the “moral values” question, categorising the topics his way still does not reveal what influence a person’s religious beliefs had on their decision in combination with the issue he or she deemed to be the most deciding factor in the 2004 election. For, it is not so much the fact that people cast their vote because of “social issues” that reveals a religiously grounded voting behaviour, but the reasoning and motivation behind the choice. People might, after all, deem any of those issues most important for a number of reasons other than their personal faith.

What Green does offer, however, is a general analysis of the question whether a respondent’s personal faith was “more important”, “about as important”, “less important”, or “not important” as other factors in the voting decision. Even though, unfortunately, Green does not present the data in conjunction with the political issues, this still is the most important and precise indication of the influence religion had on the 2004 U.S. presidential election available to date. Again, Traditionalist Evangelical Protestants are at the top of the list, with 56 percent claiming that faith was “more important” to their voting choice than any other factor. This is the only group with a majority in this regard. (If this section of the Traditionalist Evangelical group can be regarded as the most orthodox and “fundamentalist” group – i.e., the “religious right” – it would encompass about 7 percent of the voting U.S. population, or roughly 8.5 million citizens.) Second in line are the Latino Protestants, with 40 percent deeming faith to be the most important factor, followed by Other Christians with 39 percent, and Traditionalist Mainline Protestants with 32 percent. These are contrasted by much higher numbers of voters regarding faith as not important to their voting choice, with, unsurprisingly, Atheists and Agnostics at the top with 76 percent, followed by Seculars with 69 percent, Unaffiliated Believers with 56 percent, Modernist Catholics with 52 percent, and Modernist Mainline Protestants with 51 percent. Overall, 38 percent of respondents deemed their faith to be not important for their voting choice, and 15 percent felt faith was less important than other factors. On the other side, 26 percent thought faith was about as important as other factors, and 21 percent saw it as the top factor. Unfortunately, Green does not combine this data with the voting choice of the respondents. Therefore, it is only possible to look at the religious constituencies most important for the two candidates to get a picture of where they stand. Unsurprisingly, the top religious groups deeming faith to be more important than other factors gave a majority of their votes to George W. Bush, while the top groups for whom faith is not important preferred John Kerry.

Lastly, the Fourth National Survey of Religion and Politics does also give the data of voters’ frequency of worship attendance, be it however again without direct relation to voting choice. Out of the entire sample, 43 percent said they attend worship “regular[ly]“, 32 percent “often”, and 25 percent “rarely” (unfortunately, respondents were not given the option to state if they never attend worship). Split into religious groups, Traditionalist Evangelical Protestants together with Traditionalist Catholics top the list with 87 percent of “regular” worship attendance each, followed by Latino Protestants with 63 percent, and Traditionalist Mainline Protestants with 59 percent. On the other side, 83 percent of Atheists/Agnostics attend worship “rarely”, followed by Seculars with 79 percent, and Unaffiliated Believers with 58 percent. This corresponds with respondents’ voting choice, as the top groups for “regular” attendance gave majorities to president Bush while the top groups attending worship “rarely” favoured Senator Kerry. These findings corroborate those of the National Election Pool exit poll as well as the general consensus mentioned in the introduction that the frequency of church attendance seems to be a precise indicator of voting behaviour.

Conclusion

In order to paint a concise picture of the role religion played in the 2004 U.S. presidential election, statistics – as imperfect as they are – are the most scientific vehicles to arrive at a conclusion most representing the truth. For elections themselves in their processes are nothing but surveys on a massive scale and with a direct effect. From the numbers given by the two statistics presented here, it is vivid that religion indeed did have its impact on the outcome of the election. There seems to be a split not so much between religious denominations, but between the more conservatively devout and those practising their faith in a more modernist and liberal way. Both Democrats and Republicans recognised this and consequently tried to mobilise their respective constituencies while at the same time winning swing-voters from others. Senator Kerry hired a “director of religious outreach” (an evangelical no less), while Bush’s campaign apparatus staged a “Catholic Outreach Tour” and consulted, among others, the Rev. Richard Land. This feat took less effort from the Republicans, as their religious coalition was far more homogeneous and therefore easier to mobilise, as John Green suggests. Furthermore, the traditionalist groups have been far more politically active then their modernist counterparts. On the other hand, secularists may be on the rise, as the U.S. General Social Survey reported a steady growth in their numbers during the 1990s.

The findings, however, cannot be considered on their own as proof of a “religious vote”. As mentioned at the outset, the connection between church attendance and voter behaviour has been known and discussed for quite some time now. This public awareness in itself is likely to create a feedback loop that in return fuels the next survey. It conditions those who cannot pinpoint the exact reason why they voted for their candidate (which is entirely legitimate in a democracy) to choose, when asked, a vague option like “moral values” or their religious affiliation and personal beliefs as their “official” reason. Therefore, as James W. Ceaser and Andrew E. Busch, who have analysed U.S. presidential elections since 1996, suggest, the “actual line of causality” in surveys may run against pollsters’ expectations, with respondents “embrac[ing] some policy explanation to justify” their voting choice. The categorisation into “religious” and “secular” voters gave political activists on both sides the fuel they need to make their voices and opinions heard. Only the secular faction has so far been more silent than their counterpart. It remains to be seen how strong their efforts in the next presidential election in 2008 will be.

PDF-Version including Bibliography: Religion and the 2004 U.S. Presidential Election