The Divorce Rate

by Russ Roberts on August 29, 2008

in Family

Charlie, in the comments to this post, refers to a post by Justin Wolfers (who was guest blogging at MR) on the supposed divorce myth. Wolfers argues that the divorce rate is falling, not rising. That may be (though it is hard to measure correctly). But even if it is falling, it may still be higher than it was in 1975. When I last looked at the data, I saw a surge in divorce in the beginning of the 1970s. It probably peaked at some point rather than continuing to rise steadily.

But the more important point is not divorce per se, but the increase in households headed by single moms. They could be single moms because they’re divorced. Could be they never married. But the increase in that category since the 1970s makes it hard to make comparisons across time about inequality.

Be Sociable, Share!



20 comments    Share Share    Print    Email


Charlie August 29, 2008 at 3:18 pm

Again the question: If I present data that the within group inequality of Americans changed over the period, will you change your beliefs/opinions about anything?

Or do you too believe inequality has risen over the period and you are just making a methodological point?


Russ Roberts August 29, 2008 at 3:32 pm


I'm willing to accept that inequality has risen in the last few decades.

What I find hard to accept is that the average person has seen no progress over the last 30 years.

Dave August 29, 2008 at 3:41 pm

What?! I get no credit for basically making the same point?!?!

Just kidding. Glad you agree.

Charlie August 29, 2008 at 3:53 pm

"What I find hard to accept is that the average person has seen no progress over the last 30 years."

I don't think any serious person thinks that. Even the Saez graph shows gains in household growth over the period. But surely showing some progress over 30 years isn't a very high burden to meet. That would be a huge depression under Ed Prescott's deviation from trend definition. If the economy is growing at 2% a year, then that is the relevent metric of comparison for what the median American is doing. Not everything is ok, if there is any progress.


Charlie August 29, 2008 at 4:02 pm

I got curious and looked up Real per Capita GDP growth from 1929 to 1939, it was 0.22%. So Russ you are saying is, "What I won't believe is that we experienced the equivalent of the great depression over the last thirty years."

Yes, I agree.

Russ Roberts August 29, 2008 at 4:34 pm


I guess it depends on what you mean by "serious person."

I'd mention Larry Summers, the people at the Center on Budget and Policy Priorities who did the chart (Chye-Ching Huang and Chad Stone) and Steven Pearlstein, who recently wrote that the typical male worker makes less than he did in 1973.

Hammer August 29, 2008 at 4:47 pm

One might further ask whether inequality is relevant at all to anything. In and of itself, some people doing better than other people doesn't mean anything.

Unit August 29, 2008 at 5:12 pm

Suppose A makes 1 dollar, B makes 2 dollars, and C makes 3 dollars. Then A finds gold and now makes 5 dollars. Inequality has risen.

In other words every time an average Joe strikes it rich, he or she contributes to rising inequality. OTOH every time a rich dude (or duderita) loses his (or her) shirt, that reduces inequality. So rising overall inequality, means that more people make it big compared to people going bankrupt.

Charlie August 29, 2008 at 5:41 pm


Please link to the Larry Summers quote that you are referring to. He is a very serious person, and I very much doubt that he said that. But you have an excellent opportunity to prove me wrong.

As I said earlier, even the chart shows the average person is better off (even assuming you were referring to the median, which I assume you were).

I don't know who the last person is.


Russ Roberts August 29, 2008 at 5:52 pm


I may blog on this on the front of the page at some point, but here's the Summers quote (I find part of it totally inexplicable, but focus on the next to last sentence):

"Indeed, in a recent paper on tax policy prepared for the Hamilton project, my collaborators and I concluded from Congressional Budget Office data that, since 1979, changes in income distribution had raised the pre-tax incomes of the top 1 per cent of the population by $664bn or $600,000 per family – an increase of 43 per cent.

"By definition what one group gains from changes in the distribution of income another group must lose. The lower 80 per cent of families are $664bn poorer than they would be with a static income distribution, which works out to $7,000 less in income per family or a 14 per cent loss. To put this in some perspective, the total gain in median family incomes adjusted for inflation between 1979 and 2004 was only 14 per cent. If middle income families had shared fully in the economy’s income growth over the past generation their incomes would have risen twice as rapidly!"

He's arguing that over a 25 year period, when GDP grew dramatically, a tiny portion went to the average person. Not zero, but 14% over 25 years is pitiful. The part about one person's gain being another's loss bewilders me.

But I'll plead guilty to exaggerating. He didn't say zero.

Steven Pearlstein is a business columnist for the Washington Post. He wrote last week:

"Hey, good news on the income front: The Census Bureau reported yesterday that median earnings for full-time male workers rose by $1,653 last year, to $45,113, after adjusting for inflation.

"Another year like that, and maybe the typical male worker will finally catch up to where he was in 1973.

"Truth is, despite the squishy nature of income data, things haven't been so great for the middle and working class for some time. Every now and again you get a good year like last year, when wages and household incomes increased. That's usually at the tail end of an economic expansion.

"But over the past 35 years, the typical American household has managed to eke out only a 15 percent increase in its pretax income."

So the first part argues negative growth. Then he has measure that is even more depressing than Summers's–15% over 35 years.

Unit August 29, 2008 at 6:11 pm

I was too quick. Here are better numbers.

Suppose A makes 1 dollar, B makes 2 dollars, and C makes 3 dollars. Then A finds gold and now makes 6 dollars. If you use the methodology that produced the graph in this post inequality has risen. That’s because the top third of the income distribution saw income rise from 3 to 6, a 100% increase; while the bottom two-thirds rose from 3 to 5, just a 66% increase.

Charlie August 29, 2008 at 9:46 pm


I think I understand what Summers is doing. He starts his essay with, "When I studied economics in graduate school a generation ago we were taught that it was a “stylised fact” that the US income distribution was very stable."

He's doing a thought experiment taking the income distribution from 1979 as given. Then he's applying GDP growth to that distribution. Then he's comparing it to the present data. He notes that the tail on the high earners is fatter, so by his assumption that fatness must comes from the other parts of the distribution. So it looks correct to me. He's sort of setting up a proof by contradiction by taking the statement above as given.

The 14% seems to match up reasonably to the Saez data as well. I think you have a good argument that income per household isn't the right measure, but I think that there is also a lot of evidence within group inequality has risen as well (though not as much).

Ken August 29, 2008 at 11:46 pm

A friend of mine claimed that the divorce rate for 1st time marriages is rather low- around 20%. He claimed that the high divorce rate that is quoted (I've heard as high as 60%) is due to habitual divorcers. I do believe that there is a higher probability of someone getting divorced given that they've all ready been divorced. However, I don't know if that's true and even if it was, could these serial divorcers account for the discrepancy?

Does anyone have any information on this?

Charlie August 31, 2008 at 3:24 am


Justin Wolfers has lots of data on marriage and divorce over the last 150 years in this paper.

Charlie August 31, 2008 at 3:25 am

different link

LowcountryJoe August 31, 2008 at 7:30 am

A question for my command & control, top down, central planning types: just how much inequality in incomes can be tolerated?

Martin Brock September 1, 2008 at 5:24 am

Divorce is only one factor contributing to declining household size. Children establish separate households earlier. Parents live with their children less frequently. Family size declines. People share quarters with fewer unrelated roommates.

Josh B. September 5, 2009 at 4:28 am

Just the other day, I was chatting with a new friend who is staying in the U.S. for a while on business. She said she read somewhere that the divorce rate in the U.S. has dropped because of the recession. So couples that want to divorce are staying in the marriage for financial reasons.

almost broken joi October 3, 2009 at 8:39 pm

The foremost reason why divorce cases are increasing is because the couples themselves accept the idea of divorce as the best solution to marriage problem.

Both of my parents cheated. My mom cheated on my dad when I was 10, and my dad cheated on my mom many times. I am 21 now, and my parents? They are still living together. As far as I can remember, when my dad wanted a divorce, the reason why he himself refused to do it was because of my brother and I. I was a daddy’s girl. And my brother was a mama’s boy.

If couples do think the same way as my parents did, divorce can never happen.

online divorce October 5, 2009 at 1:00 pm

Hi, I run an online divorce website, so am pretty upto date on this stuff – the divorce rate is rising – but not at the growth rate it was.
Additionally, the marriage rate is falling – which in turn means we should see a reduction in the divorce rate in the next decade. Damn!

Previous post:

Next post: