Historical data on marriage and divorce in the US
Dalrock | August 7, 2011 | by Dalrock
-------------------------
I found the paper Marriage and Divorce: Changes and
their Driving Forces.  This excerpt from their analysis of the chart
in Figure 1
[http://bpp.wharton.upenn.edu/betseys/papers/JEP_Marriage_and_Divorce.pdf#page=3]surprised
me :

> Yet when viewed over a longer time period, we see that while the
> 1970s had exceptionally high divorce rates, the low divorce rates in
> previous decades were also somewhat exceptional. Fitting a simple
> trend line to the divorce rate between 1860 and 1945 (thereby
> excluding the post–World War II surge in divorce) as shown in
> Figure 1, suggests that some of the run-up in divorce in the latter
> third of the twentieth century reflects the divorce rate simply
> reverting to levels consistent with earlier trends, following
> unusually low divorce in the 1950s and early 1960s. Indeed, based on
> extrapolation, family scholars as early as the turn of the last
> century had predicted future divorce rates like those actually
> witnessed in the 1980s (Coontz, 2005). While the 1970s overshot the
> trend, the subsequent fall in divorce has put the divorce rate back
> on the trend line, and by 2005, the annual divorce rate projected by
> the pre-1946 trend is quite close to actual divorce rates.

I think they are putting too much weight on the extrapolated trend
line.  I don’t believe our current divorce rate is pre ordained,
and I especially don’t believe that divorce rates are something
which should naturally go up for well over 100 years without reaching
a natural limit
[http://socialpathology.blogspot.com/2011/05/law-of-parabolic-returns.html]
which drives them back down.  I’m also not very impressed by the
“unusually low” divorce rates in the 50s and early 60s or
convinced that divorce in the 70s and 80s was a sort of reversion to
the natural trend.  The divorce rates of the 50s and early 60s
aren’t really that striking compared to the trend, and if anything
appear to be a hangover effect from the very impressive spike in
divorce just following WWII.

The post WWII spike is very interesting, and it almost appears as if
there was a one time shuffling of spouses after the war.  If I
didn’t know better I would guess someone had proclaimed a one time
divorce/remarry amnesty.  Unlike the bulge which peaked in the 80s,
the post WWII spike in divorce corresponds with a spike in marriage. 
The two lines move so closely together after WWII that they are nearly
impossible to distinguish during the peak of the spike.  On the other
hand, with the bulge that peaked in the 80s marriage rates began to
fall around the same time divorce was on the steep incline, and have
continued this steady fall ever since.

They have another chart which I found very interesting. I can only
imagine that those who married in the 1960s assumed they were signing
up for the same deal as the people who married 10 years prior.
Obviously this was not the case. Also note how little the much vaunted
reduced divorce rates have improved the outcomes of those who married
in the 1990s in Figure 2
[http://bpp.wharton.upenn.edu/betseys/papers/JEP_Marriage_and_Divorce.pdf#page=5].

The length of the line for the 1990-1999 cohort appears short at
first, but given that the data source is from 2001 it actually strikes
me as being too long.  While the divorce rate for the first few years
includes all or nearly all of the marriages which occurred in the
decade, the latter part of the line has to be made up only of those
marriages which occurred in the very beginning of the decade. 
Assuming things are improving, this would overstate the longer term
divorce rate for that cohort.  Either way, the authors put what I
consider to be a happy face on some very troubling data:

> Figure 2 analyzes data from marital histories to assess the fate of
> first marriages, grouping them by the decade in which the wedding
> occurred. For those marriages that occurred in the 1950s through the
> 1970s, we know a lot about their eventual outcomes, and the figure
> clearly shows that the probability of divorce before each
> anniversary rose for each successive marriage cohort until the
> 1970s. For marriages that occurred in the 1970s, 48 percent had
> dissolved within 25 years, roughly confirming—for this specific
> cohort—the popular claim that “half of all marriages end in
> divorce.”1 Yet for first marriages that occurred in the 1980s, the
> proportion that had dissolved by each anniversary was consistently
> lower, and it is lower again for marriages that occurred in the
> 1990s. While it will take several more decades for the long-term
> fate of recent marriages to be realized, it appears likely that
> fewer than half of these recent marriages will dissolve.
 
-------------------------
Archived from https://theredarchive.com/blog/Dalrock/historical-data-on-marriage-and-divorce-in-theus.12240