Monday, March 31, 2014

How Personality & Religious Values Interact to Influence Political Ideology

Recently, scholars have examined how religion interacts with other factors such as race, gender, income, region, and the like in predicting political attitudes. And there has been some consideration of the mediating influence of political information on the relationships between religion and political attitudes. But how do personality traits, the biological self-regulatory systems that help us evaluate our surroundings, work with learned religious values and beliefs in shaping political attitudes? No known research directly considers this question. 

It is clear from the literature that people have political attitudes that are consistent with their religious beliefs. Religious beliefs may also influence personality traits. Personality traits may influence religious beliefs. And given that these factors influence each other, it is also likely that personality and religious beliefs work together, interacting to generate evaluations of political objects. Recently, Caprara et al. (2009) look at how personality traits and values work together to predict politics. They find that individuals’ personality traits, Openness, Conscientiousness, and Agreeableness, explained variance in politically relevant values of security and universalism. In turn, they show that these politically relevant values explained voters’ political orientations. While not religious values, these findings show that it is possible for personality traits to interact with other socially constructed values in predicting political outcomes. In addition, Lewis and Bates (2011) use the personality system framework, which links three domains, the largely biologically driven tendencies of personality, characteristic adaptions defined by values, life-events, and cultural norms, and objective behaviors and attitudes, such as political attitudes. They argue that personality traits interact with (or are mediated by) characteristic adaptations (values) in generating political orientations. They find that personality traits influence political ideology indirectly through values, such as purity, fairness, authority, and the like, providing additional evidence that personality might work with religious values to explain political attitudes. In sum, given the centrality of religious beliefs in determining important values that are relevant to politics, this analysis posits that religious beliefs serve as characteristic adaptions that interact with one’s personality, their internal, biologically generated traits in facilitating evaluations of their political world.

But what specific personality traits interact with religious values in explaining political attitudes? It is likely that not all personality traits are politically relevant, especially in concert with religious values. I argue Agreeableness, which focuses on altruism, trust, and tender-mindedness, and Openness, which is most related to creativity and openness to new experiences – both traits which are typically associated with liberalism – are the most likely suspects to mediate the conservative (and Republican) pull of religious beliefs observed in American society today. Thus, the expectation is: individuals who are more agreeable and open should be less likely to be political conservative, Republican or take conservative stances on moral issues. And the effect of Agreeableness and Openness on these political variables should be greatest among the most religiously traditional members of society. In this way, personality is interacting with influence of religious values on political orientations. But if we observe that Agreeableness and Openness have the strongest influence on secular individual’s political attitudes, an additive, instead of interactive, relationship will be observed.

I test this expectation using the Baylor Religion Study (2007). The analysis finds that Agreeableness and religious beliefs interact when predicting Americans’ political ideology (Figure 3 below). In particular, the more agreeable a religiously secular individual is the less politically conservative they are. However, religious traditionalism serves to diminish the influence of Agreeableness on political ideology, suggesting that religious values serve as a force to diminish the influence of individuals' internal biologically driven systems on political evaluations. In addition, not shown here, there is also some very limited evidence that Openness and religious beliefs interact when predicting gay marriage attitudes. The more open a religiously secular individual is the less opposed to gay marriage they are. Overall, though there is limited evidence of an interactive relationship between personality traits and religious beliefs in predicting political attitudes, there are some unique and important findings presented here. As a result, this analysis is viewed as an addition to the literature that explores why individuals with similar religious beliefs do not always share the same political attitudes.



Sources: 
Caprara, Gianvittorio, Michele Vecchione, & Schwartz, Shalom H. (2009). Mediational role of values linking personality traits to political orientation. Asian Journal of Social Psychology 12, 82-94.
Lewis, Gary J. & Bates, Timothy C. (2011). From left to right: How the personality system allows basic traits to influence politics via characteristic moral adaptations. British Journal of Psychology 102, 546–558.
For a review of the religion and politics literature see: Smidt, Corwin E., Lyman A. Kellstedt, & Guth, James L. (2009). The Role of Religion in American Politics. Explanatory Theories and Associated Analytical and Measurement Issues. In Corwin E. Smidt, Lyman A. Kellstedt, & James L. Guth, James (ed.), The Oxford Handbook of Religion and American Politics (pp. 3-42). New York, NY: Oxford University Press.
  

Saturday, March 1, 2014

The Role of Positive & Negative Images of God in Generating Partisan Conflict

While examining the multitude of questions available in the Baylor Religion Study (Wave II, 2007), I noticed that there are several questions that ask respondents about how they view God, negatively or positively. From these questions, I generated two indices, one that summarized positive images of God and another that summarized negative images of God. Below, I list the variables included in these indices.

The Baylor Study asked two questions: 1) Based on your personal understanding, what do you think God is like? And, 2) How well do you feel that each of the following words describe God in your opinion?

Positive: Concerned with the well-being of the world (GODWORLD); Concerned with my personal well-being (GODPERS); Ever-present (GODPRES); Forgiving (GODFORGV); Friendly (GODFRND); Just (GODJUST); Kind (GODKIND); Loving (GODLOVNG)

Negative: Angered by human sin (GODANGER); Angered by my sins (GODMYSIN); Critical (GODCRIT); Distant (GODDIST); Punishing (GODPUNSH); Severe (GODSEVER); Wrathful (GODWRATH)
I was curious to see if the values and beliefs associated with different images of God would help us better understand how religion is connected to partisan politics today. The religio-political sorting over the past several decades has led to the expression of increasingly conservative, Republican politics among many of the most religiously traditional denominations, e.g., white evangelical Protestants and Mormons, and individuals who subscribe to highly orthodox and traditional religious beliefs (Biblical liberalism and high frequency of attendance at religious services). But these measures are far from universal predictors of religio-partisan coalitions. So what is it about these conservative religious groups that lead to conservative, Republican politics?

One explanations might be that these religious groups have become more focused on fear or negative aspects of religion. Political activists from the Republican Party have activated these fearful images by connecting the strength of American society of polarizing moral issues about the "goodness" or "badness;" "righteousness" or "evilness" of American culture, e.g., gay marriage. Thus, we should see that negative images of God have come to define the relationship between religious imagery and partisanship, with individuals who have more negative images of God being more likely to be Republicans than individuals who have less negative images of God. And, positive images of God should not be as important as negative images of God. This theoretical framework can certainly be, and should be, extended.

But, the simple model below does confirm this pattern. Table 1 shows that even after controlling for religious tradition, religiosity, and other demographic factors, negative images of God are related to partisanship. Whereas, positive images of God are insignificant. Individuals who view God more negatively were more likely to be Republicans than their counterparts who viewed God less negatively. But, to be fair the relationship is substantively weak. But given the other religious controls in the model, this finding may be a fruitful place to look for a mechanism behind the emergence of increasingly sorted religious groups into political coalitions. Specifically, it might help us explain why individuals do not always conform to their expected partisan positions based on their religious group. It might be because of the way they view their higher-power.

Table 1. Predicting PID with the Baylor Study of Religion Images of God Questions
PID (2007) Ordered-Logit
b
Negative Images of God
0.098***
(0.038)
Positive Images of God
0.003
(0.037)
Black Protestant
-1.106
(0.730)
Mainline Protestant
-0.650***
(0.184)
Catholic
-0.617***
(0.176)
Other Religious Traditions
-0.612**
(0.261)
Unaffiliated/Secular
-0.899***
(0.240)
Religiosity
0.299***
(0.058)
Income
0.219***
(0.045)
Education
-0.146***
(0.046)
Female
-0.467***
(0.131)
White
1.800***
(0.546)
Rural
0.075
(0.156)
East
-0.417**
(0.196)
Mid-West
-0.174
(0.182)
West
0.043
(0.183)
Notes: N=1,115; PRE = 11.77%. Estimates include a sampling weight. 
*** p < 0.01, ** p < 0.05, * p < 0.1

Monday, August 12, 2013

FORCING AN OPINION: CHRISTIAN-NATIONISTS’ JUSTIFICATION FOR BEING OUT OF STEP

Recently there has been ample debate about the “Christian-nation myth,” discussing it from a legal, historical, and cultural perspective (Billitteri 2010; Bolton 2004; Heclo 2007; Stone 2008; Hunter 1991; Wald and Calhoun-Brown 2011). Essentially the myth proclaims that Christianity contributed profoundly to the formation of the early republic, giving Christianity a privileged place in American society today (Wald and Calhoun-Brown 2011, p. 63). Hunter (1991) also provides a good discussion of the “Christian nation” myth, arguing that subscribers of the myth largely describe the status and global position of the U.S. is biblical terms, such as ordained, predestined, and the like (p. 108-113).

However, scholars have overwhelmingly discredited Christian-nationist claims as fundamentalist Christian propaganda. For example, in a seminal essay on the topic, Brian Bolton (2004) explains how the Founding Fathers were not of the same blend of Orthodox, fundamentalist Christian that exists today (they were deists, Unitarians, and seculars); shows evidence of the lack of Christian predominance in our foundational legal texts (our system is based on civil rights and civil liberties); and most importantly for this essay, argues that fundamentalists who make Christian-nation claims are out of step with the American public. He states:

"The national fundamentalist agenda opposes public schools, cultural pluralism, environmental protection, abortion, evolution, gays, women, gun control, and the right to die, and endorses public prayer, censorship, and capital punishment. In other words, fundamentalists reject America’s highest values: freedom of choice, the right to personal privacy, and tolerance of and respect for individual differences."

Bolton does not provide empirical evidence to support these provocative statement, leaving a gap in his argument. While Bolton is likely correct in his assessment, good social science provides empirical evidence. Thus, this essay uses public opinion data to assess the “Christian-nation myth,” focusing on the differences between Christian-nationists and average American adults.From Bolton’s statement two competing testable propositions can be generated. (1) If the U.S. is a Christian nation, individuals who make claims to this belief should have very similar opinions as the average American adult. Because Christian-nationists would (under this hypothetical “reality”) live in a “Christian nation,” they should be no different than other Americans. (2) On the other hand, if the U.S. is not a “Christian nation,” individuals who claim it is a Christian nation should have very different opinions than the average American adult.

Using data from the Baylor Religion Survey (2007), one of the most recent surveys asking the necessary questions about the U.S. religious orientation and public policy, Table 1 provides an empirical overview of the differences between Christian-nationists, non-Christian-nationists, and the adult Population in general, which includes both of these groups.[1] These data show a number of important trends, all of which have strong implications for the “Christian nation” myth today.

(1) Approximately a quarter of American adults agree or strongly agree that the federal government should be declared a Christian nation, far from a majority.[2] Therefore, it is clear that the U.S. is not a Christian-nation because Americans do not believe it should be a Christian-nation.

(2) In terms of beliefs on school pray, gay marriage, abortion, cultural pluralism (as indicated by opinions about making English an official language), and creationism, Christian-nationists are largely out of step with non-Christian-nationists, and more importantly, the American public in general. The first three issues listed above as well as views about creationism have an obvious connection to religion, making it clear that Christian-nationists have very different opinions on issues most related to religion than typical American adults. Thus, these data support the proposition that the U.S. is not a Christian-nation because individual who subscribe to such beliefs are very different from other American adults. As a result of these differences, the U.S. cannot be a Christian-nation. However, it is important to mention that in terms of views toward women in society, the death penalty, guns and the environment, differences between Christian-nationists and the rest of the American public are present, but they are not as large. However, enough evidence is presented in Table 1 to support the proposition that debunks the Christian-nation myth.

(3) For all of the cultural issues and political orientations examined here, the non-Christian-nationists group (approximately 67 percent of the American adult population) never has a majority of the group taking the conservative or fundamentalist viewpoint. Because being a non-Christian-nationist is the majority viewpoint, this group also includes the median American adult, also known as the median voter. Thus, by extension, the average perspective in the U.S. is very different from that of the Christian-nationists, providing more evidence that the U.S. is not a Christian-nation.

What are the implications of these findings? First, it is very clear that the U.S. is not a Christian-nation. Second, the Republican Party has a Christian-nationist problem. According to these data, nearly 40 percent of Republicans are Christian-nationists.[3] Having this much of your party base out of step with the American public is a losing strategy – at least nationally.[4] But this is not likely to change. Political activists within the Republican Party, who are also fundamentalist Christian, package misinformation about our religious origins and governance as a political rally cry. If they can get enough people to believe that the U.S. is a Christian-nation, they can also get these people, who are likely Christians, to be out of step with reality in terms of issue attitudes. The rationalization goes something like this: “if the U.S. is Christian, and I am Christian, than I must be right.”






NOTES: Data comes from the Baylor Religion Survey (2007). The data and the codebook for the Baylor Religion Survey are available at http://www.thearda.com/Archive/Files/Descriptions/BAYLORW2.asp.

* Indicates that there is a statistically significant difference between White-Christian-Nationists or White Non-Christian-Nationists p < 0.001 (two-tailed chi squared test).

The Christian-nation question is: “To what extent do you agree or disagree that the federal government should declare the United States a Christian nation?” Those who agreed or agreed strongly were classified as “Christian Nationists.” Those who disagree or strongly disagree were classified as “Non-Christian-Nationists.”

There are a number of important public policy issues and political orientations examined here. Because affect, or rather how strongly an individual holds an opinion, matters to public opinion, I classify only the strongest agreement or disagreement for the public policy issues. (1) Pro-School Prayer is a measure of those who strongly agree with federal government allowing school prayer. (2) Anti-Gay Marriage is a measure of those who strongly disagree that homosexuals should be allowed to marry. (3) Pro-Life Abortion is a measure of those who believe that an abortion is always morally wrong when the mother cannot afford the child. (4) Pro-Official-English is a measure of those who strongly agree that the federal government should declare English the official language. (5) Pro-Creationism is a measure of those who strongly agree that creationism should be taught in public schools. (6) Anti-Women is a measure of those who strongly agreed that men are more suited for politics than women. (7) Pro-Death Penalty is a measure of those who those who strongly disagree that the federal government should abolish the death penalty. (8) Pro-Guns is a measure of those who strongly disagree that the federal government should enact stricter gun laws. (9) Anti-Climate Change is a measure of those who strongly disagree with potential effect of climate change. (10) Conservative is a measure of those who are leaning conservative, conservative, or extremely conservative. (11) Republican includes respondents who were leaning Republican, Republican, or strongly Republican. Leaners are included because we know that there individuals tend to be closet ideologues and partisans.

Notes for text: 
[1] See the notes from Table 1 for a discussion of the methods.

[2] More recently a HuffPost/YouGov nationally representative survey found very similar patterns. 32 percent favor or strongly favor a “Constitutional amendment which would make Christianity the official religion of the United States?” (Swanson 2013).

[3] This is calculated by looking at how many Republican also say they are Christian-nationists.

[4] In light of these results, I find it particularly surprising that the recent report by the GOP about the future failed to mention this problem (GOP 2013).

WORKS CITED

Baylor University. (2007). The Baylor Religion Survey, Wave II. Waco, TX: Baylor Institute for Studies. Retrieved from: http://www.thearda.com/Archive/Files/Descriptions/BAYLORW2.asp

Billitteri, Thomas, J. (2010). Government and Religion: Was the United States founded as a “Christian nation?” CQ Researcher 20 (2), 27-45.

Bolton, Brian. (2004). America--Not a Christian Nation. Freethought Today, 21 (1). Retrieved from: http://ffrf.org/publications/freethought-today/item/13354-america-not-a-christian-nation

GOP. (2013). Growth and Opportunity Project. Retrived from: http://growthopp.gop.com/rnc_growth_opportunity_book_2013.pdf

Heclo, Hugh. (2007). Is America a Christian Nation? Political Science Quarterly 122 (1), 59-87.

Hunter, James D. (1991). Culture Wars: The Struggle to Define America. New York, NY: Basic Books.

Stone, Geoffrey R. (2008). The World of the Framers: A Christian Nation? UCLA Law Review 56 (1), 1-26. Swanson, Emily. (2013). Christianity As State Religion Supported By One-Third Of Americans, Poll Finds. Huffington Post, 4/6/2013. Retrived from: http://www.huffingtonpost.com/2013/04/06/christianity-state- religion_n_3022255.html

Wald, Kenneth D. & Allison Calhoun-Brown. (2011). Religion and Politics in the United States, 6th Ed. Lanham, MD: Rowman & Littlefield Publishers, Inc.


Friday, March 29, 2013

A Hypothetical World

Would election outcomes be different if the distribution of religious preferences in the U.S. were different? The graph below shows that the hypothetical predicted percent of two-party vote received by the Republican candidate for each election (1992-2008).  With different religious configurations in the U.S., the outcomes would be different, but the differences are not as large as one might expect. Even making all voters the max religiously traditional or secular does not lead to a scenario where Republicans gain 80 or 90 percent and 10-20 percent of the vote, respectively. That said, presidential races are decided by small margins; therefore, changes in the religious configuration of the electorate would certainly matter.  

Why does this matter? Both political parties have to attract individuals who are religious to their political party. But, the future does suggest that more individuals are becoming religiously tradition, but rather the younger generations are becoming more secular. It is yet to be seen if this pattern will remain. But if it does, in the future we might see the Republican Party trying to attract more secular voters. If they do not, assuming the same voting patterns remain, the future would look bad for the Republican Party. 

The GOP may slowly be starting to take notice with the recent switching of positions on the gay marriage issue.  

Data comes from the ANES cumulative data file and is limited to presidential election years from 1992 to 2008. Religion is measured as an interval level variable and was constructed from a series of religious indicators. Principal components analysis was used to construct the religion index. Hypothetical scenarios are estimates by changing the values of religion of the respondents, making them more or less religious. Then I counted the number that had a higher than .5 probability of voting Republican and those who had a less than .5 probability of voting Republican. The graph summarizes this data.


Tuesday, March 19, 2013

GOP, GROWTH, & THEIR MINORITY PROBLEM

The other day the GOP released its recommendations for growth. The central finding was that the Republican Party has an "old, rich, white man problem." The party lacks diversity. In response to this "epiphany," (we have all known this for a long time), this report suggests that the GOP should target racial minorities more regularly in the future, especially when mobilizing voters in national contests. They suggest that they are just fine when it comes to state politics.

How can the Republicans win minority votes at a national level? They suggest that they have more in common with minorities than one might think. Are they right? If we consider individual-level party identification, Black Protestants, Hispanic Catholics, Hispanic Mainline Protestants, and Hispanic Evangelicals are all on the Democratic side. However, when considering political ideology, all but Hispanic Mainline Protestants (which are moderate) are on the conservative side. Thus, the report's claim that they could highlight the important similarities between the GOP and these minorities groups may be true. What they are actually saying is that they should highlight their conservatism. The report even mentions promoting their message to minorities through religious communities.

However, the problem is that the GOP takes conservative stances on many of the economic policies that would benefit racial minorities groups and most voters are economic voters (as much as some scholars tell us otherwise). The overlap between the GOP and minorities is on social issues such as abortion and gay rights. My predication is that unless the GOP stops its recent hard-line stance against redistribution policies and education, they will likely not make much headway with racial minorities in the future.



Data comes from the PEW Religious Landscape Survey of 2007.

Monday, March 11, 2013

Race, Racial Composition of One's Place of Worship, and PID

It is clear that places of worship in the United States are segregated racially. Blacks, Whites, Hispanics, and Asians practice their beliefs largely with individuals of the same racial groups. While this is the tendency, there are a growing number of non-racially segregated places of worship in the United States. How might attending religious services with individuals that are unlike you racially influence your politics? Theoretically there are two ways in which the racial composition of the place of worship might matter. 1) Religious leaders are selling a product to their congregation and that product must be tailored to the median member to retain members. If there is more racial diversity, the median member is likely more Democratic or Independent leaning than Republican; therefore, messages will conform to this reality. 2). Places of worship are social networks with informal communication among attenders. This communication will likely impact individuals political preferences pulling them toward a Democratic or Independent leaning position.

Below we see that White respondents are not substantially different across the racial compositions of their places of worship. However, a large majority still attend primarily white places of worship, an interesting finding in and of itself.

However, among Blacks, there were slight differences in PID. Blacks who attend non-primarily-Black places of worship are slightly less likely to be Democrats and slightly more likely to be Independents than attenders of primarily Black places of worship.

With so called "modernization," the diminishing racial divides in America, and the like, we should see more individuals of different racial background interacting with each other in the future. The question posed here is only a glimpse into how race, religion and politics intersect.

The data comes from the National Politics Study, 2004.



Friday, March 8, 2013

Watching the News, Religious Attendance, and PID

Conventional wisdom suggests that the FOX News viewers and MSNBC viewers make up opposite ends of the political spectrum, with FOX viewers being Republicans and MSNBC viewers being Democrats. What happens if we control for religiosity (religious attendance) when examining the relationship between the network of news an individual views and their PID? A number of things stand out from the graphs below, which were constructed using data from PEW November 2008 Post-Election Survey. 

1. The number of respondents watching MSNBC and FOX does not dramatically change across types of religiosity (see the width of the bars for the Fraction by Network). 

2. The largest changes is PID across religiosity take place among viewers of local networks, other networks, and CNN. For both non-attenders and regular attenders, CNN viewers are more likely Democrats (roughly 75% for non-attenders and 60% for regular attenders).  

3. MSNBC viewers are overwhelming Democrats regardless of their religious preferences and FOX News viewers are overwhelming Republicans. This suggests perhaps that cable network news viewership leads to polarized political positions because networks specialize in a specific "favor" of politics (see Prior 2008). 

4. There are more independents among non-attenders (a point that will be returned to in future posts). 




Thursday, March 7, 2013

GOOGLE TRENDS

The other day I found new Google tool that might have wonderful applications for social science. Using the search term "Jesus," and limiting it to just the U.S., Google Trends provides an standardized output of searches on Google for "Jesus" or similar words in U.S. The spikes correspond with Christmas and Easter. Humans are apparently very consistent. Also, from the map, it is clear that the Bible-Belt is more active in searching for the term "Jesus." What does this mean? It confirms what we know about the distribution of Christians throughout the U.S, see the PEW map for secularism. There may be more to come using this tool.


Data from PEW Research Center's Measure of Secularism

Religious Beliefs, Traditions, and Party

Individual-level religious beliefs and an individual's association with a religious tradition are very different concepts in the United States. There is substantial variance in the religious beliefs of individuals within as well as across religious traditions. However, it is clear that American Christians are the most religiously "traditional" whereas minority religious groups and non-believers are religiously "secular." Being secular in this context does not necessarily mean atheist (although the atheists are the most "secular"). Considering the figures below, we see that Americans are heavily skewed to the traditional side of the spectrum. Additionally, as expected, atheists, agnostics, and nonbelievers in general are much more secular than all religious groups. Surprisingly, Jewish respondents were much more "secular" than expected. However, I imagine there would be variance among Jewish respondents if they were classified by type (e.g., Hasidic vs. modern). Most importantly for my purposes, the final graphic shows that the religious beliefs of Democrats and Republicans in America are not as different as the media would have us believe. Not all Democrats are purely "secular" and not all Republicans are purely "traditional." This is very much in line with the arguments presented by Morris Fiorina and his colleagues in the book Culture War? The lesson here is that 1) differences exist, but 2) they are not as large as pundits might have us believe.

Data comes from the PEW Religious Landscape Survey, 2007. The religious beliefs index in constructed using a factor analysis. The reliability of the index is high (alpha = .75).




Wednesday, March 6, 2013

Religious Populations, PID, and the Future!

America is remarkably diverse religiously, yet also dominated by its Christian heritage. While the U.S. is clearly dominated Christianity overall, there are a rather large number of other religious traditions present in the U.S. Below is a figure of the proportion of the U.S. population than falls in each religious tradition. This figure was generated from the PEW Religious Landscape Survey of 2007, one the most comprehensive examinations of religion in the U.S. to date.

But what the religious landscape mean politically?  The Democratic Party has long been viewed as the party of diversity (women, racial minorities, religious minorities, middle and lower class, unions etc.); whereas the GOP the party of white rich men. When examining the mean party identification of religious groups in the U.S., there are very few who lean Republican (White mainline and evangelical Protestants, which are not that different, and Mormons). There are a very large number of religious groups that make up the Democratic coalition.

Assuming these patterns remain stable, with all of the groups associated with the Democratic coalition (non-white, religious minorities) growing in numbers (see 2010 U.S. Census) and the White Protestant population becoming a smaller share, the future looks good for the Democratic Party. However, political parties are adaptive; therefore, we might see major shifts in the political targeting from the GOP in years to come.

FIGURE 1

FIGURE 2