Crusty Logic Christianity & Liberty

Christianity, Statistics, and the misguided chuzpa of Bradley R. E. Wright

A recent USA Today article “Christians question divorce rates of faithful”, began:

It’s been proclaimed from pulpits and blogs for years — Christians divorce as much as everyone else in America.

But some scholars and family activists are questioning the oft-cited statistics, saying Christians who attend church regularly are more likely to remain wed.

“It’s a useful myth,” said Bradley Wright, a University of Connecticut sociologist who recently wrote “Christians Are Hate-Filled Hypocrites … and Other Lies You’ve Been Told.”

“Because if a pastor wants to preach about how Christians should take their marriages more seriously, he or she can trot out this statistic to get them to listen to him or her.”

Pastors aren’t exactly known for their due diligence when it comes to what they say from the pulpit and, sadly, trotting out inaccurate facts and figures is not uncommon.  And indeed, Christians who attend church regularly are more likely to stay together.  But there’s a lot more to it.

In his book ‘Christians Are Hate-Filled Hypocrites …And Other Lies You’ve Been Told’ Bradley Wright takes aim at many statistics that have made the rounds of U.S. Christian Pop Culture over the past several years.

He starts off questioning Barna’s research implying that evangelical Christians were second only to prostitutes in how little they are respected.  So far, so good.  While Barna’s statistics on this are accurate, the hyperbole was not.

Other parts of this book however, including the portions the USA Today article was based on, are surprisingly disingenuous and misleading.

In chapter 6 Wright takes on Barna’s research indicating that Christians and in particular, Evangelical Christians, divorce almost as often as non-Christians.  Referencing The General Social Survey (a great source of information btw), Wright says:

As for divorce, the survey reports how many respondents (who had ever been married) had been divorced or were currently separated from their spouse. Contrary to popular belief, Christians and members of other religions have lower divorce rates, about 42%, than do the religiously unaffiliated, about 50%. Among Christians, however, there was substantial variation. Catholics are the least likely to have divorced, at 35%, followed by Mainline Protestants (41%), Evangelicals (46%), and Black Protestants (54%).

But if we want to know whether or not the evangelical church’s teachings affect the actions of its members, perhaps an even more important question is whether cohabitation and divorce rates go down as church attendance goes up. As it turns out, they do, and the change is substantial. As shown in Figure 6.1, of the Evangelicals who rarely if ever attend church, 7% were cohabitating, compared to 5% of the monthly attendees and only 2% of the weekly attendees. Likewise, with divorce, 60% of the never-attendees had been divorced or were separated compared to only 38% of the weekly attendees[1].

The statistic’s Wright uses are spot-on and are actually in agreement with Barna.  It’s Wright’s analysis that doesn’t add up.  He chooses to turn cause and effect on its head, asserting that church attendance is the cause and divorce is the effect.  And this is accurate – occasionally.  More often though, divorce is the cause and church attendance is the effect.  If you look across the landscape of divorced Christians you will find a vast number who attended church regularly – until some point after their divorce.  It’s not so much that “divorce rates go down as church attendance goes up” as Wright asserts, but that church attendance goes down as divorce rates go up.

The most common scenario is a couple who are fairly strong evangelical Christians and who attend church regularly with their kids.  He’s caught getting some sex on the side, she divorces him, she continues to attend church, he eventually stops attending.  Perhaps the best news from this is that only a bit over half of such scenarios end in divorce.  Sadly though, in the majority of those that do, one or both spouses drop out of church as do some number of their children.  This is a hugely critical element of the divorce problem within U.S. Christianity that Wright sets aside.

Even if you choose, like Wright, to ignore those who drop out of Christianity after their divorce (a chunk of that 60% of never-attendees who’ve divorced that he references in his book) and only consider the current regular attenders worthy, you still have a huge problem – even Wright’s slimmed-down 38% divorce rate is a quite noticeable lump in history and is significantly higher than even non-Christians outside the U.S. today.  The argument Wright makes here is simple lunacy.

Divorce is one of the most harmful challenges impacting Christians and society, whether 38% or 50%.  Loss of a relationship with Christ is perhaps the only thing worse.  Wright dismisses both as inconsequential.

—–

Later in Chapter 6 Wright says:

The General Social Survey asks respondents if, when married, they have ever had sex with someone other than their husband or wife. As shown in Figure 6.3, 16% of Evangelicals reported that they had committed adultery at some time in their life. (It’s worth noting that the way the question is worded, we don’t know if the adultery happened before or after their initial involvement in evangelical Christianity). Mainline Protestants, Catholics, and Jews all reported similarly low levels of 14 to 16%. Black Protestants and the religiously unaffiliated reported the highest rates of extramarital sex, at about 25%. If we focus on the line labeled “Christians (all)” we see that, taken as a whole, Christians are committing adultery about one-third less than the unaffiliated. It appears that the commandment “Thou shalt not commit adultery” is, thankfully, still having an effect on the church.

… Just as we discovered with divorce rates, church attendance correlates well with sexual misconduct. As shown in Figure 6.4, Evangelicals who regularly attend church display far less sexual misconduct than those who attend less often. Twenty-two percent of Evangelicals who never attend church have committed adultery as compared to 13% of those who attend weekly.

As with divorce, Wright is telling us “nothing here folks, move along…”  But there is something here, it is the major cause of divorce among Christians, and we should not ignore it.

In order to convince us to ignore reality, Wright chose to ignore three critical elements in his analysis.

The first is correlated data.  Do the subject numbers jive with other numbers?  Studies of prostitution in the U.S. indicate that each year about 17% of adult men visit a prostitute[2].  That’s actually pretty close to Wright’s estimated range of 13% to 25% of men going out for sex while married.  However, is that all that ever will, or only those who have so far?

Next year it will be 17% as well, but a slightly different 17%.  A few guys will have stopped visiting prostitutes and a few will have begun doing so for the first time.  The vast majority of these visits are by guys in their forties so of all adult guys, we have a gob of guys in their twenties and thirties who have yet to visit a prostitute, but will, and a gob of guys in their fifties and sixties who once did but no longer do (not many old guys have the sexual prowess of Dominique Strauss-Kahn)[3].

Using conservative numbers we get about 62% of U.S. men visiting a prostitute at some point in their life with 37% (of all adult men) doing so while married.  More realistic numbers produce somewhat higher estimates.  And this doesn’t include the guys who get some on the side with affairs but never visited a prostitute.  Any way you look at it, this adds a fairly big question mark to Wrights’ numbers.  Either his numbers are extremely low or studies of prostitution are extremely high.

But one correlation does not a good analysis make, especially where it’s possible for both statistics to be off.  Another and perhaps more critical correlation to look at then is divorce rate vs. extra-marital sex.

Does Wright’s statement that 13% of weekly attending Evangelical Christians engage in extra-marital sex work with his own 38% divorce rate for this same group?  Let’s look at 200 people in 100 marriages.  Over the length of their marriages, using Wright’s numbers, 26 will go out for sex (13% of 200) resulting in 26 divorces which is about70% of his 38%.  That comes fairly close to estimates of divorces caused by extra-marital sex.  So far, so good.

There’s more to it though.  To get there we had to assume that every single person who engages in extra-marital sex gets caught and that every instance results in divorce[4].  Reality is closer to maybe 50% getting caught and 65% resulting in divorce.  So now, 26 go out and 13 are caught resulting in 8 divorces.  This accounts for only 20% of Wright’s 38 divorces.  Or put another way, according to Wright’s numbers about 4 in 5 divorces, or 80%, involve no extra-marital sex at all by either partner (as a cause of the divorce or not).  The lowest relatively reliable estimate that I’ve seen is that 25% of divorces involve extra-marital sex, the highest 80%, with most in the 50-65% range.  Pastors and marriage counselors I’ve talked to say a bit over 50%.

A more probable reality of our 200 weekly attending Evangelical Christians is that about 75 guys and 10 gals will, at some point in their marriage, visit a prostitute or have an affair.  Five of these overlap with both partners going out for sex, giving us 80 marriages impacted in some way.  In 40 of these marriages one or both will be caught resulting in 26 divorces accounting for 68% of Wright’s 38% estimate (or 60% of a more probable 44% divorce rate).  That’s closer and also jives with prostitution and other statistics.

So, correlated data casts suspicion on Wright’s analysis.  His numbers come from a reputable survey though, how can he be so wrong?

The second element Wright has chosen to ignore, similar to our 17% prostitution number, is time.  He applies a static number (13% having gone out for sex) to a continuum (marriage).  That doesn’t work.  And he knows it.

We know how many say that they have already gone out for sex, but what about those, specifically guys younger than about 42, who have not, but will in the future?  This is extremely important in this context because marriages extend over a period of time but going out for sex is limited to a relatively brief period of that time, usually about 6 to 10 years during their 40’s.  What we want to know is how many marriages will be impacted by this at some point in time during the marriage.

For example, consider menopause, which is something else that marriages experience for a relatively brief period of time .  If you ask these same 100 married couples how many have experienced menopause you’ll get a response of about 40.  In reality though, 100% of marriages will.  So, to say something like “40% of marriages are impacted by menopause”, implying that the rest don’t have to worry about it, is quite misleading.  (Similarly, you can ask how many are currently experiencing menopause and get a response of maybe 11% which is similar to our 17% of men currently visiting prostitutes each year.)

So, of our 100 marriages of weekly attending evangelical Christians that exist on Jun 12, 2012 (with ages ranging from 20’s to older than me), using Wright’s 13% number, we’ll have 26 having been impacted by extra-marital sex.  Over the next year, one or two more of these marriages will be impacted.  And one or two more the year after that.  In 10 years about 38 of these marriages will have been impacted.  In 20 years 50, and in 25 years, when the youngest have now reached their 50’s and are no longer likely to begin patronizing prostitutes, about 57.

And we haven’t yet accounted for the number of guys who were regular attenders in their twenties, thirties, and forties, got caught going out for sex in their forties, divorced, dropped out of being a regular attender and thus aren’t counted as a regular attender who went out for sex even though they were.  This will push our 57 impacted marriages up a bit, likely to about 70 or so.

The third thing Wright ignores is admission error.  Let’s start with his lowest number – 13% of Evangelicals who attend church weekly reported having engaged in extra-marital sex.  A more accurate statement would be 13% of Evangelicals admit to having engaged in extra-marital sex.  How many do it but don’t admit to it?  Even in a supposed confidential survey?

In almost any survey there is some level of admission error.  Obese folk often under admit to how much they eat, most of us over admit to how much we exercise, high school boys over admit to how often they’ve had sex and married guys under admit to how often they’ve gone out for sex.

The reasons for the latter vary.  Some married guys lie simply because they are concerned about getting caught.  They’re either not confident in the confidentiality of the survey or they’re concerned that someone is listening or looking over their shoulder when they’re responding or that in some other way they’ll be caught if they admit to it.  Another is that some respondents simply don’t think that a quickie with a hooker or a happy ending at a massage place, counts.  And then there is the issue of cognitive dissonance – innocence bias – “if I don’t admit to it then maybe I didn’t really do it and thus I am who I think I am and not who I am.”  It’s one thing to do something, another entirely to admit to it.

It’s rather critical to note here that the General Social Survey is conducted face to face.  “So Mr. Respondent, how often have you visited a prostitute?”  I think you get the point.

The impact of admission error is difficult to gauge.  If your only purpose here were to compare Christians to non-Christians then you might choose to ignore it since both will contain somewhat similar errors.  Some statisticians might note that Christians may be somewhat less likely to admit to it, but not calculate that in to the results.  If however, what you want to know is more of an absolute, like how many do in fact go out for sex versus how many do not, then you need to account for it.

We can assume that most of those who have not gone out for sex will not say that they have.  But how many who did go out for sex will not admit to it?  This is a commonly discussed phenomenon in general and in particular with socially undesirable attributes  For instance, the number of guys admitting in surveys to going out for sex is consistently well below the number needed just to support our known prostitution industry, not counting non-prostitution extra-marital affairs, and is consistently below any reliable estimates of the number of divorces caused by extra-marital sex.

Based on my experience I’d guesstimate that about 40% of weekly attenders who have gone out for sex will not admit to it in a survey (and slightly fewer, maybe 30%, for non-attenders).  This would put us in the mid 90’s for marriages impacted by extra-marital sex, a number that I guesstimate is high.

Note that critical word guesstimate.  We’re dealing with pretty ambiguous numbers here.  We can’t watch these folks 24 hours a day to know for certain how many go out for sex and how many do not.  In the end, taking in to account known statistics and correlated data, I’d guesstimate that anywhere from 65% to 90% of marriages of evangelical Christians who attend  church weekly are impacted by extra-marital sex, that in about half of these it becomes known to their spouse, and that maybe 65% of these result in divorce.

Getting below 65% of marriages of regular attending Evangelical Christians impacted by extra-marital sex requires accepting some rather unrealistic suppositions (like everyone gets caught, all of these result in divorce, everyone caught and divorced continues being a regular attender, and every one of these admits to it in surveys.)  Oh, and we’d have to have a lot of future admission as well “gosh, I’ve never done anything like that, but I expect that in ten years I’ll begin visiting prostitutes”.

The 90% upper bound is simply gut instinct.  I strongly believe that at least 10% of these marriages never have any extra-marital sex.

Most importantly though, these guesstimates correlate relatively closely to other data.

Conclusion

When it comes down to it, we don’t really know what the real numbers are.  We can, at best, make educated guesstimates.

We know that Wright’s 38% divorce rate is low and that reality is probably more like 44%, but it could be 40% or 48%.  In the end though, these differences don’t matter a wit, 38% or 44% is still an extremely major problem that impacts a ton of people, particularly children.  The difference in Barna’s and Wrights numbers simply don’t make any difference – we have a huge problem either way.

Underlying this, how many weekly attending Evangelical Christians have extra-marital sex?  Likely somewhere between about 34% and 47%.  A very long way from Wright’s 13%.  Most critical though is that we know that we have a high divorce rate, that these divorces are having an extremely detrimental impact on a large number of people, particularly children, and that extra-marital sex is at least a major cause of these divorces and likely the major cause.

Why then would Wright produce this?  Why would he try so hard to play down the problem of divorce?  He knows statistics, sociology, anthropology, and statistical analysis well enough to know better.  The report card he gives at the end of his book may provide a clue.  Some of his more interesting:

Gender Equality: C Christianity still a majority of women, except in leadership.

Crusty: Interesting that he says this here but doesn’t account for it in any of his analysis.

Racial Integration: B-  Church still predominantly White, but it’s getting better and more diversified in recent decades.

Crusty:  Still predominantly white?  What does he expect when the vast majority of our population is white?  Interestingly, Wright’s own chart indicates that U.S. evangelical Christianity better reflects the racial make-up of the U.S. than any other religion and in most cases it’s better by a huge margin. I’ve reproduced Wright’s chart below from the same data source with the only difference being sort order.  I’d give us a B+.

Divorce and living together: B  Relatively low rates, and less among frequent attendees, but increasing over time.

Crusty: Really?  Relatively low rates?  By Wright’s own estimate, evangelical Christians who attend church weekly have a 38% divorce rate.  And Wright’s number is low.  Easily an F in my book.  We have a major problem here that is having a huge negative impact on millions of people and on society.

Sex: A-  Relatively low rates of adultery, premarital sex, porn; these decrease with attendance.

Crusty: See Divorce and living together.

Loving Behaviors: C+  Could act more charitably to others, but this does increase with attendance.

Crusty: I don’t agree or disagree.  There is certainly room for improvement and likely always will be. 

Attitudes toward Blacks: D  Um, being Black is not a sin.  Gets worse with attendance, but improving over time.

Crusty: Again, there is certainly room for improvement. However, if this really was a D I doubt that evangelical Christianity would so well reflect the racial make-up of our overall society and be the most racially integrated religion in the world.

Attitudes towards gays: D  Not loving gays; gets worse with attendance, but improving over time.

Crusty: Agree.  Being gay is not a sin, acting on it is.  Even so, we should still love even those who most flagrantly act on it.  I know a lot of Christians who, I believe, act accordingly in this regard, but I also know several who have a very difficult time separating the sin from the sinner.

We may be getting somewhere. Wright does his best to check off the social justice agenda; Blacks, Gays, Gender Equality.  Then he tries to say that nearly half of Christian marriages ending in divorce, leaving millions of impacted children in its wake, isn’t much to be concerned with.

———

Notes:

Reproduction of Wright’s figure 4.3 using the same data and formatting.  I made two changes; 1) I sorted based on standard deviation from the norm of the Total U.S. Population, and 2) I used colors that more closely correlate with the groups we’re discussing.  The number in parenthesis indicates the percent of the total population accounted for by each group.

 

 


[1] Bradley R.E. Ph.D. Wright. Christians Are Hate-Filled Hypocrites…and Other Lies You’ve Been Told: A Sociologist Shatters Myths From the Secular and Christian Media (p. 133). Kindle Edition.

[2] Extremely quick numbers: Including erotic massage such as practiced by Jennifer Love Hewitt’s character in The Client List, there are about 132 million engagements per year by the 275,000 women working as prostitutes in the U.S. (or an average of 40 / mo / prostitute)  The average customer makes 7 visits per year (the majority average about once per month, but a significant number visit prostitutes once or twice per year, thus the average of 7 per year).  132 million engagements divided by 7 visits per customer gives us 18 million distinct customers which is 17% of the 111 million adult men in the U.S.

[3] This 40-something phenomenon comes from a number of sources including studies of prostitutes indicating that the majority of their customers are in their 40’s, polls on prostitution websites such as The Erotic Review that indicate most of the men involved are in their 40’s, audience statistics from Alexa and others indicating that those who visit prostitution sites are predominantly 38 – 47.

[4] We also assumed that each of these was a different marriage – E.G., that no marriage had both husband and wife having extra-marital sex.

  • Copyright ©2011 Crusty Logic. Best viewed in anything but Internet Explorer.