Harry Benson, November 2023 1 www.marriagefoundation.org.uk
What is the Divorce Rate?
Harry Benson, Marriage Foundation
‘What is the divorce rate’ is a much harder question to answer than most people realise. Most
commentary refers to divorces in a particular year the ‘year of divorce’ method. But what we
really want to know is what happens to marriages over time – the ‘year of marriage’ method.
For years I have been keeping record of divorces by duration of marriage, by commissioning
data specially from the Office for National Statistics. This allows me to map marriages that
ended after say 1, 2, or 3 years with the number of marriages that began 1, 2, or 3 years
ago. This means I can track what happens to couples who married in any particular year, which
gives a far better indication of the real trends in divorce.
Since the 1960s, divorce rates have followed a highly predictable pattern, rising through the
earliest years of marriage, peaking in years 3-7, then falling gradually thereafter. Divorce rates
therefore vary only during the first decade of marriage. After ten years, there is little to no
variation, regardless of which year couples married from 1963 to the present day.
Divorce rates overall rose for couples marrying in the 1960s, 70s and 80s and then fell for
couples marrying in the 1990s, 2000s and 2010s. Divorce rates are now back to 1970 levels,
down 21% overall from their peak for 1986 marriages, down 24% for couples in their first
decade of marriage and down 48% for couples in their first 5 years of marriage.
Most striking of all is that husband-granted divorce rates have barely changed since the 1970s.
Almost all the change has involved wife-granted divorces in their first decade of marriage. This
tells us that long term changes in divorce rate are not about women’s economic independence
(which should mean rates rise) or changes in age or economics (which should affect husband-
and wife-granted divorces equally). Instead, the key driver is men’s commitment. Rising rates of
cohabitation and continuing social pressure to get married in the 1960s and 1970s meant
increasing numbers of women marrying less-committed men. As marriage has become more
optional since the 1990s, men who marry today are increasingly the more committed men who
really want to do so. Hence fewer divorces initiated by women in their first decade of marriage.
The overall lifetime risk of divorce is now 35%, the same as for couples marrying in 1970, down
from 44% for couples marrying in 1986 who were the cohort with the most divorces ever.
Harry Benson, November 2023 2 www.marriagefoundation.org.uk
How do we calculate divorce rates?
Almost all reporting on ‘divorce rates’ involves
some comparison of the number of divorces in
the most recent year divided by the number of
marriages in the most recent year, or the number
of adults in the population, or the number of
adults who are married. This is called the ‘year of
divorce’ method.
The problem with this method is that we’re not
really comparing like with like. Imagine a simple
scenario where there are 100,000 divorces in two
consecutive years. But in the first year, its mostly
old marriages ending and in the second year, its
mostly younger marriages ending. The headline
‘divorce rate’ would show no change. Yet a great
deal is happening behind the scenes.
What we really want to know is what has already
happened to couples who married in the same
year, whether 1970 or 1980 or 1990. That tells us
the real trend. We also want to know what will
happen to couples who get married this year, on
present trends. This is called the ‘year of
marriage’ method.
The big advantage of the ‘year of marriage’
method is that it should give a really clear
indication of trends in divorce. How are today’s
couples likely to fare, compared to couples who
married in the 1980s or 1990s. The big
disadvantage of this is that it’s much harder to
calculate.
I use much the same method used by the Office
for National Statistics on the rare occasion when
they report ‘year of marriage’ divorce rates. Their
most recent paper on this is from 2012 when
they calculated the lifetime risk of divorce at
42%. This is the figure most often cited in the
media (ONS 2012a).
Because ONS do not routinely publish the data in
sufficient detail, Marriage Foundation has
specially commissioned data on four separate
occasions, including this year (ONS 2012b, 2015,
2019, 2023)
The method used is straightforward enough. I
take the number of divorces involving marriages
under one year and divide by the total number of
marriages that took place last year. I then do the
same for divorces that took place after one to
two years of marriage and divide by the total
number of marriages two years ago.
The method isn’t perfect because of overlap – for
example not all divorces within a year were
marriages that took place last year. But so long as
the changes aren’t too dramatic, the method is
accurate and gives an excellent guide to trends.
Adjusting for overseas weddings
I also make one very important adjustment to the
number of new marriages recorded in any given
year, as the table above shows. The Annual
Passenger Survey produces an estimate of the
number of people returning to the UK after
getting married. Because many or most of these
people will in fact be one half of a returning
couple, I make a conservative adjustment to my
wedding numbers by adding 50% of this estimate
to the figure for England & Wales weddings.
Between 2002 and 2015, some 30-70,000 British
nationals married abroad in each year (adjusted
for England & Wales population). Reducing these
numbers by half adds an average of 12%, ranging
between 6% and 17%, to the total number of
weddings in any of these years (ONS 2008, 2017).
Because these figures are based on relatively
small survey samples, ONS do not adjust their
divorce rate estimates for overseas weddings.
ONS divorce rates will therefore always overstate
the true divorce rate because they underestimate
the number of weddings. I on the other hand
don’t have to be quite so precise. I am therefore
confident that the addition of a conservative
estimate for overseas weddings significantly
improves the quality of all divorce rate estimates.
The consistent pattern of divorce
Using this ‘year of marriage’ method, it is
possible to construct an entire spreadsheet
showing the trends in divorce rates for couples
who marries in any particular year.
The rather messy chart below shows the trend in
annual divorce rates for every cohort of couples
marrying in the years 1985 and 2020.
The purpose of showing this is to illustrate that
there is a very clear trend in divorces as
marriages progress over time. What is striking is
Harry Benson, November 2023 3 www.marriagefoundation.org.uk
the consistency of pattern. No matter which year
we are looking at, divorce risk tends to rise in the
first few years of married life, peak after between
3 and 7 years of marriage, before declining in a
highly predictable trend from about 10 years
onwards. After about 30 years, annual divorce
risk is very low.
Trends in divorce – back to 1970
Most people know that divorce rates have risen
since the 1960s. Few people also know that
divorce rates are now back at 1970 levels.
The chart below illustrates this with a comparison
of what happened to couples who married in
1970, 1986, and 2010.
The blue line shows the actual annual divorce
rate of couples who married in 1970. The green
line shows the actual annual divorce rate of
couples who married in 1986 – the year when
more couples than in any other year went on to
divorce. Finally, the red line shows divorce rates
for couples who married in 2010 and have
therefore had ten years of divorce data.
Although we can’t be completely sure what will
happen to the remaining 2010 couples, the
pattern of previous years means we can be
confident that their annual divorce rate will trend
slowly down in the same way as for every
previous year group. Their final divorce rate will
therefore be similar to the 1970 couples and well
below that of the 1986 couples.
The first decade of marriage
The two previous charts strongly suggest that the
big changes in divorce rates take place in the first
few years of marriage. So let’s have a look at
what happened to couples in their first, second
and third decades of marriage.
Because the latest divorce figures were reported
in 2021, we have data on the first decade of
marriage for couples who married any time
before 2011. We have data on the second decade
for couples who married any time before 2001
and third decade for couples who married any
time before 1991.
The chart below shows these trends. What is very
clear from the trends shown in the chart below is
how the overall divorce rate for any given year of
marriage depends almost entirely on what
happens to couples in their first decade of
marriage.
The chart shows how the blue line, the divorce
rate during the first decade of marriage, rose
steadily from 8% among couples who married in
the mid-1960s to 23% among couples who
married any time between the mid-1980s and
mid-1990s. Thereafter divorce rates have fallen
back to 1970s levels and currently stand at 17%.
In sharp contrast, divorce rates among couples in
their second or third decade of married life have
remained remarkably consistent and predictable
at 13% during the second decade and 6% during
the third decade.
Of note is the recent upwards blip in early
divorce, about which more later.
The gender gap in divorce rates
Even less known is that up until April 2022 when
the law changed, almost all divorces were initially
petitioned by and finally granted to either
husband or wife.
Harry Benson, November 2023 4 www.marriagefoundation.org.uk
We have produced several papers over the past
decade highlighting the gender gap in divorces
rates accruing to either husband or wife.
The two charts below show the trends in this gap
over time during the first and second decades of
marriage. The first chart shows the rise and fall of
divorce but also highlights that these changes are
driven almost entirely by changes in divorces
granted to wives during that first decade.
Whereas the divorce rate accruing to wives
marrying at any time after 1975 fluctuated within
a 6.7% range from 10.2% to 16.9%, the divorce
rate accruing to husbands fluctuated in a much
lower and narrower 1.7% range between 5.1%
and 6.9%.
During the second decade, fluctuations remained
within a narrow 1.7% range for divorces granted
to wives and 1.3% range for divorces granted to
husbands.
What about the ‘blip in divorce rates?
The most recent ‘year of divorce’ data show a
9.3% increase in divorces from 102,438 to
111,934 in 2021. It is tempting to assume this is a
post-lockdown phenomenon, where couples who
grimly survived the domestic pressures of 2020
lockdown finally gave up on their marriages.
However, the evidence strongly suggests this is a
system issue rather than a divorce rate issue. The
increase is not due to more marriages giving up
but more likely to a clearing of previous backlogs
and/or faster processing of previous delays.
First, Ministry of Justice figures for ‘divorce cases
started’ were falling during 2020 and showed no
indication that there would be a 9% increase in
divorces completed during 2021 (Ministry of
Justice 2023) see chart below. Second, our own
analysis of survey data showed a reduction, not
increase, in ‘thoughts about divorce’ during
lockdown itself (Benson & McKay 2021). Thirdly,
the ‘blip’ in divorce is evenly spread across all
durations of marriage. The number of divorces by
duration of marriage show a consistent increase
of between 200 and 600 extra divorces during
2021 for every duration of marriage between 2
and 25 years. Had there been a genuine increase
in divorce rates, we should have seen this
concentrated among those vulnerable years 3-7.
Cumulative changes in divorce rates
The next chart shows the cumulative progression
of actual divorce rates over time among different
cohorts. After 30 years of marriage, 35% of
couples who married in 1970 had divorced.
Among the 1980 cohort, 41% had divorced.
Among the 1990 cohort 42% had divorced.
Although the 2000 cohort have only been
married for 20 years, the addition of the highly
predictable divorce rates in years 20-30 mean
that 40% of this cohort are likely to have divorced
within 30 years.
Harry Benson, November 2023 5 www.marriagefoundation.org.uk
Using the same technique, the 2010 cohort
already have lower divorce rates in their first ten
years that makes them look more like the 1970
cohort. A long-range projection for them, albeit
with a slightly higher margin of error, is that 35%
will have divorced within 30 years of marriage,
i.e. by 2040.
Lifetime divorce rates
Continuing this projection over 50 years for all
cohorts who have married since 1963 produces
the following chart.
Overall lifetime divorce rates have risen from
28% among couples marrying in 1963 to peak at
44% among couples marrying in 1986 the
cohort with the highest ever rates of subsequent
divorce before falling to the current best
estimate of 35%. The chart also shows the
relative contributions made by divorce granted
either to wives or husbands.
Note that these predictions included a
conservative adjustment estimate for overseas
weddings that are not included in official figures
for weddings but nonetheless contribute to the
real annual number of weddings.
Leaving out this important adjustment, divorce
rates would be proportionately higher, rising
from 28% among 1963 couples to 46% among
1986 couples and falling to the current estimate
of 38%.
References
Benson, H. (2022) Married poor more stable than
unmarried rich. Romford: Marriage Foundation.
Benson, H. & McKay, S. (2021) No ‘divorce boom’ in
sight. Romford: Marriage Foundation.
Ministry of Justice (2023) Family Court Statistics
Quarterly, April to June 2023.
https://www.gov.uk/government/statistics/family-
court-statistics-quarterly-april-to-june-2023
Office for National Statistics (2008) Marriages abroad
2002-2007. Population Trends, 133, 65-70.
Office for National Statistics (2012a) Divorces in
England and Wales: 2011
Office for National Statistics (2012b) Number of
Divorces in England and Wales by duration of marriage
by party to whom granted (husband and wife) and by
previous marital status 1963 to 2010, commissioned
data for Marriage Foundation.
Office for National Statistics (2015) Number of
Divorces in England and Wales by duration of marriage
by party to whom granted (husband and wife) and by
previous marital status 2011, 2012 and 2013
Office for National Statistics (2017) British citizens
married abroad. FOI request
Office for National Statistics (2019) Divorces in
England and Wales by duration of marriage by party to
whom granted (husband and wife) and by previous
marital status, 2010 to 2017
Office for National Statistics (2023) Opposite-sex
divorces in England and Wales by duration of marriage
by party to whom granted (men and women) and by
marital status prior to marriage 2018 to 2021
Tables
Cumulative divorce rates, Overall
by year of marriage & duration of marriage
Actual Actual Predicted
1st 5 2nd 5 1st 10 2nd 10 3rd 10 Lifetime
1965 2% 8% 10% 12% 6% 31%
1970 5% 10% 15% 12% 6% 35%
1975 7% 11% 18% 12% 6% 38%
1980 8% 12% 20% 13% 6% 41%
1985 10% 12% 22% 13% 6% 43%
1990 10% 12% 22% 13% 5% 42%
1995 10% 13% 23% 12% 43%
2000 9% 11% 20% 12% 40%
2005 7% 11% 18% 39%
2010 6% 10% 16% 36%
2015 5% 35%
Latest 5% 11% 18% 13% 6% 35%
1986 10% 13% 23% 13% 6% 44%
vs 1986 -48% -12% -24% -5% -1% -21%
Cumulative divorce rates, by party granted
by year of marriage & duration of marriage
Wife-granted
Husband-granted
Predicted lifetime
1st 10 2nd 10 3rd 10 1st 10 2nd 10 3rd 10 Wife Husband
1965 7% 9% 4% 3% 4% 2% 21% 10%
1970 10% 8% 4% 4% 3% 2% 24% 10%
1975 13% 8% 4% 5% 3% 2% 26% 11%
1980 15% 9% 4% 5% 4% 2% 29% 12%
1985 16% 9% 4% 6% 4% 2% 31% 12%
1990 16% 9% 3% 6% 4% 2% 29% 13%
1995 16% 8% 7% 4% 29% 14%
2000 14% 8% 6% 4% 26% 14%
2005 12% 6% 25% 14%
2010 10% 6% 23% 14%
Latest 11% 8% 3% 7% 5% 2% 22% 13%
1986 17% 9% 4% 6% 4% 2% 31% 13%
vs 1986 -36% -14% -5% +12% +18% +7% -31% +3%