I learned some wonderful news the other day.  One does not have to be a graduate sudent to have access to World Values Survey data.  It’s online!!  This is a lot of rich information concerning about 99 countries.  It is a sociologist’s dream.  Check it out.  I just did a research project using this as my data source. 

Religion and Politics: will you be Filing Jointly or Separately?

By Believer 1

Abstract

This study looks at how people respond to four key statements that explore the relationship between religion and politics.  I use the 2005 World Values Survey to try to answer the questions: What types of people support a link between politics and religion, and what types of people support a separation of politics and religion?  Although there are differences in how people responded to each statement, there are also some similarities. The variables for religious person, and highest educational attainment play an important role in explaining people’s responses to all four statements.  The variable for voted in the most recent elections does not explain people’s responses to any of the statements.     

Introduction and Literature Review

            This study tries to answer the questions: What types of people support a link between politics and religion, and what types of people support a separation of politics and religion? American scalars rarely talk about the relationship between politics and religion without mentioning the first amendment of the United States constitution (Van Alstyne, 1963; Tamney, 1974;  Esbeck, 1985; Rohrer, 1987; Stephen, 2002).   Van Alstyne starts his article by saying, “Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof….” (Van Alstyne , 1963, P. 863).

For the United States in particular, the debate about the proper relationship between religion and politics really heats up with the start of the conflict over Sunday mail in 1810.  Prior to this, there is a connection between the government, especially state government, and religion that goes relatively unchallenged.  At this time, many states have what they call “moral laws” which among others include not working on Sunday (Rohrer, 1987).

            The Sunday mail debate consists of two opposing camps of people the Sabbatarians and the Anti-Sabbatarians.  The Sabbatarians believe in a covenant theology.  Put simply, this means the United States has a covenant with God that says that if Americans obey God, the country will be blessed by God.  However, the Sabbatarians do not use their covenant theology to argue against Sunday mail.  They use the constitution instead.  They argue that the first amendment prohibits the government from keeping people from practicing their religion, and that since it is against many people’s religion to work on Sunday, Sunday mail is unconstitutional.  They also argue that Sunday mail violates the moral laws of several states, and that the federal government should not go against state government (Rohrer, 1987). 

At first, the Anti-Sabbatarians, consisting of less dominate religious groups such as Unitarians, Universalists, and Baptists, simply argue that not having Sunday mail hurts businesses.  Later, they express a fear of one government endorsed religion.  This fear is brought a head,

“when, in 1827, Presbyterian minister Ezra Stiles Ely issued an influential call for the creation of a ‘Christian party in politics’.  Ely proclaimed that a moral reformation of America could be accomplished only if Christians selected leaders ‘orthodox in their faith’.  The Presbyterians alone, he argued, ‘could bring half a million electors into the field,’ while the five largest protestant denominations ‘could govern every public election in our country’” (Rohrer, 1987, P. 64-65).       

Stephen Carter, a law professor at Yale, says this about the separation of church and state argument in America’s early days, “It is vital that we in our legalist ahistoricism not forget that the Protestant separatists believed in dividing church from state, not God from state. The purpose of the separation was not to protect the state from religious believers but to protect the church…” (Stephen, 2002).  The Sabbatarians try to distance themselves from Ely claiming that he acted alone, but their efforts are in vain.  By 1830, the Sabbatarians loose the Sunday mail debate paving the way for a pro-separation of church and state viewpoint to prevail (Rohrer, 1987).  

            Carl Esbeck writes about five viewpoints concerning the relationship between politics and religion.  The first two are the strict separatists, and the pluralistic separatists viewpoints.  Like the Anti-Sabbatarians in Rohrer’s article about Sunday mail, people who subscribe to these viewpoints fear that a close link between religion and politics results in a loss of freedom, especially religious freedom, for those in less dominate groups (Esbeck, 1985; Rohrer, 1987). 

            As one might imagine, based on the title strict separatists, people with this viewpoint want a strict and complete separation of politics and religion.  Pluralistic separatists want a separation between politics and religion, but when moral issues such as those involving social welfare and peace are involved, these people have no trouble inserting their religious views into their political participation.          It is important to note that both of these groups may contain religious, as well as, nonreligious people.  The third point of view concerning religion and politics is referred to as the institutional separationists viewpoint.  People in this group want a stronger connection between politics and religion, but are not in favor of a theocracy.  They believe that both the religious realm, and the political ream are ordained by God.  As a result, there should, and will be some interplay between the realms.  At the same time however, each real has its own purpose and destiny (Esbeck, 1985).

            People who subscribe to the forth viewpoint are referred to as nonpreferentialists.  Like institutional separationists and other separatists, these people are against the government supporting a particular religion.  Nonpreferentialists attack this issue from a different angle than the separatists.  They argue the government should support all religious organizations, as opposed to not supporting any religious organizations.  American nonpreferentialists may not be alone in their approach.  Joseph Tamney argues that people in Indonesia believe that their government should support every religion.  Nonpreferentialists argue that supporting religious organizations reduces government costs, because these organizations provide services to communities at a lower cost, and in a more personal manner than the government.  As one might accept, political conservatives are often nonpreferentialists (Tamney, 1974;  Esbeck, 1985).

            Lastly, “restorationists believe that the United States is a Christian nation or was originally intended as one, and they often argue for the restoration of the nation’s high view of Christianity as it existed in the founding period.  Not only is the public theology explicitly Christian in its creed, but much of restorationism has a decidedly Puritan or at least a ‘chosen people’ cast to it” (Esbeck, 1985, P. 371).  For these people, religion and politics cannot be separated.  The government must protect the church; while at the same time avoid interfering with the church (Esbeck, 1985). 

            Based on the literature, I hypothesis that the variable for religious person will explain more of the variance in the dependent variables, than will the variable for religious denomination.  I argue that the dominance of particular religions are different depending on the society considered.  Put another way, one country’s dominate religion, may be another country’s least dominate religion, and vice versa.   Therefore, I argue that for a world sample, such as the sample for this study, religiosity is more important that religious denomination.  I hypotheses that the variable for voted in most recent elections will play an important role in explaining the variance in the dependent variables.  It makes sense that people who participate politically have probably thought about how government interacts with other social institutions, such as religion.  Lastly, I hypothesis that several demographic variables will play an important role in explaining the variance in the dependent variables.     

Methodology

            I use the 2005 World Values Survey for my study.   The World Values Survey consists of face-to-face interviews of a randomly selected, representative sample of people living in 99 countries.  There are 67,268 respondents.  The sample is made up of individuals with very low income all the way up to individuals with very high income, with each income level fairly represented.  Highest educational level attained includes people with no formal education all the way up to people with college degrees.  Age is presented as an open-ended question.  There are more than thirty categories for religious denomination, and the sample is 48% male and 52% female.  Nearly three quarters of respondents said they voted in recent parliament elections, and a majority of the respondents said they are religious. 

My dependent variables are four key statements that explore the relationship between religion and politics.  Statement number one is: Politicians who do not believe in God are unfit for public office.  People who agree with this statement support a link between religion and politics, and people who disagree with this statement favor a separation between religion and politics.  Statement number two is: Religious leaders should not influence how people vote in elections.  People who disagree with this statement support a link between religion and politics, and people who agree with this statement favor a separation between religion and politics.  Statement number three is: It would be better for (insert Country) if more people with strong religious beliefs held public office.  People who agree with this statement support a link between religion and politics, and people who disagree with this statement favor a separation between religion and politics.  Statement number four is: Religious leaders should not influence government decisions.  People who disagree with this statement support a link between religion and politics, and people who agree with this statement favor a separation between religion and politics.

My independent variables are country/region, religious denominations, religious person, voted in most resent parliament elections, sex, age, highest educational attainment, and income.  The answer options for religious person are religious, nonreligious, and committed atheist.  Voted in most resent parliament elections is a yes or no question.  The answer options for sex are male and female.  Income data is coded in a scale of income.  I used linear regression to analysis the data for my study.        

Results

            Table 1 explains some of the variance in the dependent variable: Politicians who do not believe in God are unfit for public office.  The variable for religious person, which explains the most, explains 9.5% of the variance in the way people respond to the statement.  Table 1 shows religious people are significantly more likely, than nonreligious people or atheists to agree with the statement.  The variable for religious denomination explains .5% of the variance in the way people respond to the statement.  When explaining this dependent variable, Hypothesis 1 is correct.  The variable for voted dose not explain any of the variance in the way people respond to the statement.  Therefore, hypothesis 2 is incorrect. 

            The variable for highest educational attainment, which explains the second largest amount, explains 1.9% of the variance in the way people respond to the statement.  Table 1 shows that those with less education are significantly more likely, than those with more education to agree with the statement.  Because only one demographic variable plays an important role in explaining the variance in the dependent variable, hypothesis 3 is incorrect.  Model 9 shows that considering all of the independent variables explains 7.5% of the variance in the way people respond to the statement.          

 

 

 

Table 1

Regression Results for Dependent Variable:Politicians who do not believe in God are unfit for public office.
Model 1	ConstantCoefficient	CountryCoefficient	RSquared as %
	2.893***	.002	.5%
Model 2	ConstantCoefficient	Religious DenominationCoefficient	RSquared as %
	3.447***	-.006***	.5%
Model 3	ConstantCoefficient	Religious PersonCoefficient	RSquared as %
	3.980***	-.734***	9.5%
Model 4	ConstantCoefficient	VotedCoefficient	RSquared as %
	3.055***	-.050***	0%
Model 5	ConstantCoefficient	SexCoefficient	RSquared as %
	2.880***	.080***	.1%
Model 6	ConstantCoefficient	AgeCoefficient	RSquared as %
	3.139***	-.003***	.2%
Model 7	ConstantCoefficient	Highest EdCoefficient	RSquared as %
	3.402***	-.076***	1.9%
Model 8	ConstantCoefficient	IncomeCoefficient	RSquared as %
	3.229***	-.050***	.8%
Model 9	ConstantCoefficient	All inCoefficients	RSquared as %
CountryReligious Denomination Religious Person Voted Sex Age Highest Ed Income	4.936***	.002***-.007*** -.640*** -.095*** -.009 -.005*** -.046*** -.027***	7.5%
P Value = * < .05, ** < .01, *** < .001

            Table 2 explains very little of the variance in the dependent variable: Religious leaders should not influence how people vote in elections.  The variable for religious person, which explains the most, explains 1.1% of the variance in the way people respond to the statement.  Table 2 shows religious people are significantly less likely, than nonreligious people or atheists to agree with the statement.  The variable for religious denomination explains does not explain any of the variance in the way people respond to the statement.  When explaining this dependent variable, Hypothesis 1 is correct.  The variable for voted only explains .1% of the variance in the way people respond to the statement.  This demonstrates further that hypothesis 2 is incorrect.  Nevertheless, table 2 shows that those who voted are significantly more likely, than those who did not to agree with the statement. 

            The variable for highest educational attainment, which explains the second largest amount, explains 1% of the variance in the way people respond to the statement.  Table 2 shows that those with more education are significantly more likely, than those with less education to agree with the statement.  Because no demographic variables play an important role in explaining the variance in the dependent variable, there is still no support for hypothesis 3.  Model 18 shows that considering all of the independent variables only explains 1.2% of the variance in the way people respond to the statement.

Table 3 explains some of the variance in the dependent variable: It would be better for (insert Country) if more people with strong religious beliefs held public office.  The variable for religious person, which when considered by itself, explains 13% of the variance in the way people respond to the statement.  Table 1 shows religious people are significantly more likely, than nonreligious people or atheists to agree with the statement.  The variable for religious denomination explains .1% of the variance in the way people respond to the statement.  When explaining this dependent variable, Hypothesis 1 is correct.  The variable for voted dose not explain any of the variance in the way people respond to the statement.  This demonstrates further that hypothesis 2 is incorrect.

 

 

Table 2

Regression Results for Dependent Variable: Religious leaders should notInfluence how people vote in elections.
Model 10	ConstantCoefficient	CountryCoefficient	RSquared as %
	3.810***	.000*	0%
Model 11	ConstantCoefficient	Religious DenominationCoefficient	RSquared as %
	3.722***	.001*	0%
Model 12	ConstantCoefficient	Religious PersonCoefficient	RSquared as %
	3.527***	.204***	1.1%
Model 13	ConstantCoefficient	VotedCoefficient	RSquared as %
	3.886***	-.062***	.1%
Model 14	ConstantCoefficient	SexCoefficient	RSquared as %
	3.843***	-.029**	0%
Model 15	ConstantCoefficient	AgeCoefficient	RSquared as %
	3.727***	.002***	.1%
Model 16	ConstantCoefficient	Highest EdCoefficient	RSquared as %
	3.559***	.045***	1%
Model 17	ConstantConfident	IncomeCoefficient	RSquared as %
	3.718***	.017***	.1%
Model 18	ConstantCoefficient	All inCoefficients	RSquared as %
CountryReligious Denomination Religious Person Voted Sex Age Highest Ed Income	3.264***	.000.002*** .120*** -.042** -.002 .002*** .039*** .002	1.2%
P Value = * < .05, ** < .01, *** < .001

Table 3

Regression Results for Dependent Variable: It would be better for (insert Country)If more people with strong religious beliefs held public office.
Model 19	ConstantCoefficient	CountryCoefficient	RSquared as %
	2.973***	.002***	1.1%
Model 20	ConstantCoefficient	Religious DenominationCoefficient	RSquared as %
	3.388***	-.002***	.1%
Model 21	ConstantCoefficient	Religious PersonCoefficient	RSquared as %
	4.198***	-.802***	13%
Model 22	ConstantCoefficient	VotedCoefficient	RSquared as %
	3.142***	-.028*	0%
Model 23	ConstantCoefficient	SexCoefficient	RSquared as %
	2.951***	.115***	.2%
Model 24	ConstantCoefficient	AgeCoefficient	RSquared as %
	3.283***	-.004***	.3%
Model 25	ConstantCoefficient	Highest EdCoefficient	RSquared as %
	3.616***	-.093***	3.3%
Model 26	ConstantCoefficient	IncomeCoefficient	RSquared as %
	3.373***	-.054	1%
Model 27	ConstantCoefficient	All inCoefficients	RSquared as %
CountryReligious Denomination Religious Person Voted Sex Age Highest Ed Income	4.960***	.002***-.004*** -.704*** -.091*** .015 -.006*** -.059*** -.021***	10.7%
P Value = * < .05, ** < .01, *** < .001

The variable for highest educational attainment, which explains the second largest amount, explains 3.3% of the variance in the way people respond to the statement.  Table 3 shows that those with less education are significantly more likely, than those with more education to agree with the statement.  The fact that only one demographic variable plays an important role in explaining the variance in the dependent variable, further shows that hypothesis3 is incorrect.  Model 27 shows that considering all of the independent variables explains 10.7% of the variance in the way people respond to the statement.

            Table 4 explains very little of the variance in the dependent variable: Religious leaders should not influence government decisions.  The variable for religious person, which explains the most, explains 1.6% of the variance in the way people respond to the statement.  Table 4 shows that religious people are significantly less likely, than nonreligious people, or atheists to agree with the statement.  The variable for religious denomination explains none of the variance in the way people respond to the statement.  When explaining this dependent variable, Hypothesis 1 is correct.  The variable for voted dose not explain any of the variance in the way people respond to the statement.  This demonstrates further that hypothesis 2 is incorrect.     

            The variable for highest educational attainment, which explains the second largest amount, explains .2% of the variance in the way people respond to the statement.  Table 4 shows that those with more education are significantly more likely, than those with less education to agree with the statement.  The fact that no demographic variables play an important role in explaining the variance in the dependent variable, further shows that hypothesis 3 is incorrect.  Model 36 shows that considering all of the independent variables explains only 1% of the variance in the way people respond to the statement.

Table 4

Regression Results for Dependent Variable: Religious leaders should notInfluence government decisions.
Model 28		ConstantCoefficient	CountryCoefficient	RSquared as %
		3.665***	.000***	0%
Model 29		ConstantCoefficient	Religious DenominationCoefficient	RSquared as %
		3.658***	.000	0%
Model 30		ConstantCoefficient	Religious PersonCoefficient	RSquared as %
		3.350***	.256***	1.6%
Model 31		ConstantCoefficient	VotedCoefficient	RSquared as %
		3.736***	-.039**	0%
Model 32		ConstantCoefficient	SexCoefficient	RSquared as %
		3.788***	-.064	.1%
Model 33		ConstantCoefficient	AgeCoefficient	RSquared as %
		3.628***	.002***	0%
Model 34		ConstantCoefficient	Highest EdCoefficient	RSquared as %
		3.574***	.022***	.2%
Model 35		ConstantConfident	IncomeCoefficient	RSquared as %
		3.604***	.018***	.1%
Model 36		ConstantCoefficient	All inCoefficients	RSquared as %
CountryReligious Denomination Religious Person Voted Sex Age Highest Ed Income		3.197***	.001***.001* .199*** -.029* -.030* .002*** .011*** .013***	1%
	P Value = * < .05, ** < .01, *** < .001

Dissection

This study tries to answer the questions: What types of people support a link between politics and religion, and what types of people support a separation of politics and religion?  The first hypothesis is that the variable for religious person will explain more of the variance in the dependent variables, than will the variable for religious denomination.  This hypothesis is supported by several scalars mentioned previously (Tamney , 1974; Esbeck, 1985; Rohrer, 1987).  First Rohrer states that at first, the Anti-Sabbatarians, consisting of less dominate religious groups such as Unitarians, Universalists, and Baptists, simply argue that not having Sunday mail hurts businesses.  Later, they express a fear of one government endorsed religion.  He says that both groups in the Sunday mail controversy are made up religious people.   The Anti-Sabbatarians do not fear religion: The fear dominance (Rohrer, 1987).

In his article discussing different viewpoint of the relationship between religion and politics, Esbeck says that nonpreferentialists argue the government should support all religious organizations, as opposed to not supporting any religious organizations.  American nonpreferentialists may not be alone in their approach.  Joseph Tamney argues that people in Indonesia believe that their government should support every religion (Tamney , 1974; Esbck, 1985).  Given all of this information, one might wonder how my first hypothesis is supported.  I argue that the dominance of particular religions are different depending on the society considered.  Put another way, one country’s dominate religion, may be another country’s least dominate religion, and vice versa.   Therefore, I argue that for a world sample, such as the sample for this study, religiosity is more important that religious denomination.  The first hypothesis is correct.  When determining what kind of relationship people support beteen religion and politics, it is more important to know whether a person is religious or not, than it is to know their religious denomination.  This importance of this finding stretches beyond the church-state issue.  The finding shows commonality between religions. 

My second hypothesis is that the variable for voted in most recent elections will play an important role in explaining the variance in the dependent variables.  When starting my study, It made sense that people who participate politically have probably thought about how government interacts with other social institutions, such as religion.  My second hypothesis is incorrect.  My error in thinking may have been the result of my extensive training as a sociologist.  Maybe political participation dose not result in thinking about how government interacts with other social institutions, such as religion for the average person.  Future studies could explore this in more depth by looking at other forms of political participation. 

My third hypothesis is that several demographic variables will play an important role in explaining the variance in the dependent variables.  This hypothesis is incorrect.  Highest educational attainment is the only demographic variable that plays an important role in explaining differences in the way people respond to the statements. 

In conclusion, this study makes a modest yet important contribution to answering the questions: What types of people support a link between politics and religion, and what types of people support a separation of politics and religion?  It is more important to know whether a person is religious or not, than it is to know their religious denomination.  Religious people tend to support a link between politics and religion. Nonreligious people and atheists tend to support a separation of politics and religion.  People with higher levels of education tend to support a separation of politics and religion.  People with lower levels of education tend to support a separation of politics and religion. 

Bibliography

Carter, S. L. . (2002). The J. Byron Mccormick Lecture: Reflections on the Separation of Church and State. Arizona Law Review, 44(293), Retrieved from http://sb6nw2tx4e.scholar.serialssolutions.com/?sid=google&auinit=SL&aulast=Carter&atitle=Reflections+on+the+Separation+of+Church+and+State&title=Arizona+law+review&volume=44&date=2002&spage=293&issn=0004-153X

 Esbeck, C. H. (1985). Five Views of Church-State Relations in Contemporary American Thought. Brigham Young University Law Review, Retrieved from http://lawreview.byu.edu/archives/1986/2/esb.pdf

Rohrer, J. R. (1987). Sunday Mails and the Church-State Theme in Jacksonian America . Journal of the Early Republic, 7(1), Retrieved from http://www.jstor.org/stable/3123428?seq=1

 Tamney , J. B. (1974). Church-State Relations in Christianity and Islam. Review of Religious Research, 16(1), Retrieved from http://www.jstor.org/stable/3510193?cookieSet=1

Van Alstyne, W. W. (1963). Constitutional Separation of Church and State: The Quest for a Coherent Position. The American Political Science Review, 57(4), Retrieved from http://library.csus.edu/guides/amatab/History/jstorex.pdf