Thursday, August 03, 2006

Divorce: Part Two of Travels in Wingnuttia



This post is about divorce in the Bible Belt and more generally in Wingnuttia, the imaginary country which extreme right-wingers inhabit. My earlier post on divorce tried to answer the question whether divorce is always a bad thing, and another one started the travels-in-Wingnuttia series. In this post, I will simply assume that there are people who regard the rising divorce rate as deplorable, and at least some of these people want social engineering by the government to reduce the rates of divorce. Or this would be called social engineering if liberals were carrying it out. When it's the conservatives, we call it something else: Defence of Marriage, Covenant Marriages (laws which make getting divorce harder) or schemes to get poor mothers off welfare programs by making them marry more.

The same people who support such schemes also tend to view high rates of divorce as part and parcel of the mythical liberal culture of permissiveness and vice. Sodom and Gomorrh. Not much evidence is ever presented to explain exactly how liberals go around getting people unhitched, but that doesn't matter when the need is to look for one simple scapegoat. And that scapegoat is the liberal culture, supported by Hollywood.

It comes then as quite a shock to find out that Massachusetts, that sinful home of lasciviousness and ribaldry, is actually the state with the lowest divorce rate, and that almost all of the reddest states have high rates of divorce. Indeed the Bible Belt divorce rates are fifty percent higher than the national average.

The 2003 figures are instructive. In that year the Massachusetts divorce rate was 5.7 divorces per 1,000 married people. Comparable figures for Kentucky, Mississippi and Arkansas were 10.8, 11.1 and 12.7 respectively.

In other words, if it is the permissive liberal culture that causes high divorce rates it somehow doesn't work in those areas where such a culture should be at its strongest. Other reasons for the high divorce rates should be sought.

The common candidates for such reasons are income, education and the average age at marriage. People with higher incomes and education levels are less likely to divorce, perhaps, because the life shocks caused by financial difficulties are less likely to push those with a financial cushion to a point of marital dissolution. People who marry very young may not have the experience to choose their spouses well.

Other variables affecting divorce rates are urbanization, the general amount of population movement into and out of an area and the rate of women's labor market participation. Divorce rates are higher in urban than in rural areas, possibly because divorce is more socially acceptable and/or easier in big cities. Areas with a lot of change tend to cause instability which may register in higher rates of marital dissolution. The higher the percentage of women who work the more likely it is that they are able to leave poor marriages.

These variables can explain part of the geographic pattern of divorce rates. The Bible Belt states have more poorer and uneducated people and the average age at marriage is lower there. On the other hand, urbanization rates, for example, are higher on the East Coast.

And what is the role of religion in all this? I found two conventional answers to this question, and one rather surprising one. The first conventional answer is that religious attachment makes divorce less likely. Most religions frown on divorce, and belonging to a religious group may raise the difficulty of divorce by the "shunning" or judgment a divorcing individual. Religion may also improve the quality of the marital relationship to those who believe in a particular faith, especially if the spouses share these beliefs.

This answer would account for the low divorce rates in Massachusetts by its large Catholic population. The Catholic Church is traditionally opposed to divorce. Ironically, this has been proposed as the explanation by some wingnuts who'd rather blame the liberal corrupt culture. Which suggests that Massachusetts citizens are more religious than those in the Bible Belt.

The second conventional answer consists of plugging some religion-related variable into a model of divorce which also controls for the income, education and age-at-marriage variables, to see if religion exerts any additional influence on the likelihood of divorce. The usual result from doing this is to find that religious attachment reduces divorce probabilities, over and above the impact of the other included explanatory variables.

I'm dissatisfied with this approach. It fails to distinguish between different religious affiliations or between, say, those who are Easter-and-Christmas Christians from those who attend some kind of service at least weekly. It's also a modeling error, because controlling for the age-at-marriage variable in this way doesn't allow us to study whether religious affiliation itself might affect the age at first marriage. I can see a fairly credible argument for such an impact to exist, but a model which holds the age-at-marriage variable constant in this way would fail to measure it empirically..

And what is this "credible argument", you might ask. Well, consider the abstinence view of many fundamentalists. If sex outside marriage is sinful, what do you do to your restless teenagers? You might suggest that they get married young, to avoid the temptations to sin by having premarital sex. Thus, the overall effect of religiosity on divorce might work through several channels, some making divorce less likely, others making it more likely. Holding age-at-marriage constant disguises one factor in the last-mentioned group.

All this technical speech is needed to explain why I searched for studies on this topic for such a long time and why I don't really like any one of those I found. It's not that these studies wouldn't be good for the purposes they were designed to answer, but they don't answer the question I have, which is to find out the impact of religious wingnuttery on divorce.

I did find out that it is the church-going conservative Christians who are most likely to worry about divorce and other signs of "cultural collapse". What makes this both clearer and less clear is the (and yes, I'm finally coming to the surprising answer I mentioned far, far up in this post) fact that the conservative Christian fundamentalists suffer from a pretty steep divorce rate themselves, higher than the rate for any other religious group. Put that into your pipe and smoke it, Pat Robertson and Jerry Falwell.

Take the group of born-again Christians. This group is not identical with right-wing fundamentalism but it's a pretty good approximation. According to Barna Group, which specializes in studies of religion, the born-again are as likely to get divorced as atheists and agnostics, and more likely than other religious respondents:

Although many Christian churches attempt to dissuade congregants from getting a divorce, the research confirmed a finding identified by Barna a decade ago (and further confirmed through tracking studies conducted each year since): born again Christians have the same likelihood of divorce as do non-Christians.

Among married born again Christians, 35% have experienced a divorce. That figure is identical to the outcome among married adults who are not born again: 35%.

George Barna noted that one reason why the divorce statistic among non-Born again adults is not higher is that a larger proportion of that group cohabits, effectively side-stepping marriage – and divorce – altogether. "Among born again adults, 80% have been married, compared to just 69% among the non-born again segment. If the non-born again population were to marry at the same rate as the born again group, it is likely that their divorce statistic would be roughly 38% - marginally higher than that among the born again group, but still surprisingly similar in magnitude."

Barna also noted that he analyzed the data according to the ages at which survey respondents were divorced and the age at which those who were Christian accepted Jesus Christ as their savior. "The data suggest that relatively few divorced Christians experienced their divorce before accepting Christ as their savior," he explained. "If we eliminate those who became Christians after their divorce, the divorce figure among born again adults drops to 34% - statistically identical to the figure among non-Christians." The researcher also indicated that a surprising number of Christians experienced divorces both before and after their conversion.

Multiple divorces are also unexpectedly common among born again Christians. Barna's figures show that nearly one-quarter of the married born agains (23%) get divorced two or more times.

The survey showed that divorce varied somewhat by a person's denominational affiliation. Catholics were substantially less likely than Protestants to get divorced (25% versus 39%, respectively). Among the largest Protestant groups, those most likely to get divorced were Pentecostals (44%) while Presbyterians had the fewest divorces (28%).

...

Barna stated that there is no end in sight regarding divorce. "You can understand why atheists and agnostics might have a high rate of divorce, since they are less likely to believe in concepts such as sin, absolute moral truth and judgment. Yet the survey found that the percentage of atheists and agnostics who have been married and divorced is 37% - very similar to the numbers for the born again population. Given the current growth in the number of atheists and agnostics, and that the younger two generations are predisposed to divorce, we do not anticipate a reversal of the present pattern within the next decade."

So what have we here? Let's do a summary: First, it is the fundamentalist right-wing Christians who are most worried about divorce, and it is this group which attributes the divorce culture to a liberal influence. But, second, it is also the fundamentalists who appear to be as likely to get divorced as atheists and agnostics, and more likely to get divorced than, say, the believers in more liberal theologies. The third point concerns the avenues of influence which might cause the relatively high divorce rates among the born again (fundamentalist) Christianity. One suspect for this is the impact of a fundamentalist religion on a person's age at first marriage. Others are possible lower income or education levels and the difficulty of following the very rigid sex roles prescribed by some literal readings of the Bible, especially the required subjugation of wives to their husbands.

Note that the "corrupt liberal culture" isn't a very good candidate for inclusion among the explaining variables of any model of divorce. How could this culture affect the born agains but not the rest of the religious people, for example? Indeed, one would expect that the born agains would be especially vigilant in warding off the deleterious effects of evil liberal influences.

All this smacks of projection. It seems that it is the Christian right-wing marriages which are in trouble, but somehow the fault for that is put at the feet of the liberal culture, even when liberal marriages seem to fare better. But perhaps this is not too astonishing, given the personal histories of many famous wingnuts:

Even a cursory look at the leading members of the forces of social conservatism in America reveals the same pattern. Rush Limbaugh, the top conservative talk-radio host, has had three divorces and an addiction to painkillers. Bill O'Reilly, the most popular conservative television personality, just settled a sex harassment suit that indicated a highly active adulterous sex life. Bill Bennett, guru of the social right, was for many years a gambling addict. Bob Barr, the conservative Georgian congressman who wrote the Defense of Marriage Act, has had three wives. The states that register the highest ratings for Desperate Housewives, the hot new television show, are Bush states.

----
Some of my sources are files some of my readers kindly sent me. The files themselves may not be available for consultation without a payment. E-mail me if you want to know the source of a particular argument that doesn't have a link attached to it.