August 2, 2012 • 12:34 pm

“One in two couples marrying today can expect to split up.”

Daily Telegraph, 30 July 2012

It is a commonly and pessimistically held assumption that as we continue into the 21st century, more and more marriages are ending divorce. Commenting on the recent divorce of former TV presenter, Anthea Turner, the Telegraph put a statistic to these fears, claiming that 50 percent of couples signing a marriage license will also see their names on divorce papers.  

This isn’t the first time that Full Fact has lifted the veil on divorce claims, recently discovering that the rate of divorce is lower than the rate at which people switch their bank account providers. But do today’s newly-weds have only a 50/50 chance of sticking together, as the Telegraph claims?

Analysis

The paper’s figure can be traced to an Office for National Statistics (ONS) 2008 analysis on “The proportion of marriages ending in divorce”.

Using statistics collected prior to 2005, the article aimed to predict the future proportion of marriages ending in divorce from 2005 onwards. This is the latest study from the ONS to model this. 

Calculating this probability is not an easy matter. It is not as simple as considering the divorce rates in the chosen year (2005), as this is calculated from the number of divorces in that year and gives no indication of how long the marriages lasted before the divorce. 

To reach an estimate of the proportion of marriages ending in divorce, it is necessary to estimate the size of the married population of 2005.

In estimating this population, it is important to factor in the variants which could effect the number of married individuals at any one time. This report takes into account divorce, but also death and migration.

If one part of a married couple dies, both themselves and their partner (who is now “widowed”) are removed from the married population. Migrating couples can also cause fluctuation in marriage population, as they can leave England and Wales after marrying or marry abroad and move to England.

The report is the first study to take into account all factors affecting a married population when calculating its estimate, making it the most reliable calculation of the probaility of divorce thus far. 

Once these factors are controlled for, it is possible to estimate the age at which each individual in the population tied the knot according to the average duration of marriages. This will enable the proportion of marriages ending in divorce and the proportion ending in death to be calculated, as the likelihood of divorce is much higher for younger couples or a marriage which has existed for less than 10 years, whereas an older couple are more likely to be separated by death.

Having negotiated this maze of a methodology, the report calculates:

“Making the assumption that divorce rates and mortality rates remain unchanged from 2005, around 10 % of those marrying in 2005 will celebrate their diamond (60th) wedding anniversary, with 45% of marriages ending due to divorce and 45% of marriages ending in death.”

Clearly then, the statistics in the Telegraph article can be supported by these predictions (allowing for rounding).

But do such predictions made four years ago provide an adequate basis for judgements on today’s marriages?

Unfortunately, the predictions themselves haven’t been updated since they were relased seven years ago.

However, the ONS does collect figures on the proportion of marriages taking place each year that end in divorce, broken down for each year since the couple took their vows. The latest information covers marriages which took place in each calendar year up to 2010, meaning we can see how accurate the predictions have been for the past five years.

The table below shows the predicted proportion of marriages ending in divorce for those tying the knot in 2005, based upon the modelling undertaken that year:

 

Percentage divorced by anniversary (years)

Year of marriage

 

1

2

3

4

5

 

 

 

 

 

 

2005

1.2

3.6

6.6

9.6

12.4

Whereas, the next table shows what has really happened to the 247,805 marriages which occurred in 2005:

 

 

Percentage divorced by anniversary (years)

Year of marriage

Total number of marriages

 

 

1

2

3

4

5

 

 

 

 

 

 

 

2005

247,805

0

1

3

6

8

By comparison, we can see that the projections made in 2008 continuously overestimate the proportion of 2005 couples untying the knot. Furthermore, the data suggests that divorces are becoming rarer: whereas in 1987 11 per cent of married couples had separated by the fifth year of marriage, for the 2005 cohort only 8 per cent had done likewise. (The full dataset shows a downward trend in other years too.)

Unfortunately, we would have to wait another 60 or 70 years or so to fully test whether the prediction of a grand total of 45 per cent pans out in reality, but on the evidence we have so far, the forecast may be on the pessimistic side.

Conclusion

The Telegraph’s divorce statistics are sourced from a prediction made four years ago about the probability of divorce for married couples at the time, and they have been accurately relayed by the paper.

Whether or not these forecasts will be borne out in practice remains to be seen, but the most recent data we have suggests that they may not be doing justice to Britain’s romantics.

Share this article:
By • 12:34 pm
A non-profit company (no. 6975984) limited by guarantee and registered in England and Wales. © Copyright 2010 - 2014 Full Fact