Our response to the ONS consultation on marriage statistics

Oct 19, 2017

The Office of National Statistics has opened a consultation with data users on whether it should continue to provide separate estimates for the national population of opposite-sex and same-sex married couples.

For those interested in responding, the deadline is October 24th and the consultation documents can be found on the portal here: https://consultations.ons.gov.uk/data-analysis-unit/user-need-for-separate-estimates-of-the-population/ 

Below we have reproduced our response to the exercise should any supporters be interested:

Question one 

How do you require estimates of the population who are married? 


  1. Presented separately for same-sex couples and for opposite-sex couples 

Question two 

Please explain why you need separate estimates.  

Please include how you would use these estimates, and what your analysis would help to inform. 


Accurate and uncontested estimates of the stock of same-sex and opposite-sex marriages are essential in order to support evidence based policy making and to ensure that the Government may be held accountable for claims made at the time of the passage of the Marriage (Same-Sex Couples Act) 2013.  

Same-sex marriage represented a major change to established law in an important area of social policy. It is therefore a matter of public interest that the impact of this policy be made open to proper scrutiny through the analysis of accurate statistical data. 

In the guidance issued by the Government Equality Office in April 2014, it was asserted that: 

“Marriage is a hugely important institution in this country. The principles of long-term commitment and responsibility which underpin it bind society together and make it stronger. The Government believes that we should not prevent couples from marrying unless there are very good reasons – and loving someone of the same sex is not one of them.” 

The then Minister for Women and Equalities, Maria Miller, stated in the Second Reading debate in Parliament on 5th February 2013 that: 

“What marriage offers us all is a lifelong partner to share our journey, a loving stable relationship to strengthen us and mutual support throughout our lives. I believe that that should be embraced by more couples. The depth of feeling, love and commitment between same-sex couples is no different from that depth of feeling between opposite-sex couplesAll couples who enter a lifelong commitment together should be able to call it marriage. 

In combination these assertions provide an important and testable basis on which the Government legislated to legalise same-sex marriage. The “lifelong” or alternatively “long-term” commitment to a monogamous union with one another shown by same-sex couples was held to be identical with that shown by opposite-sex couples.  

It will be possible to test and verify this assertion over time using separated samples for the stock of marriages for both sets in conjunction with the separately published data for marriages and dissolutions.  

It is necessary to have a separate estimate of stock in order to calculate the divorce rate. Without a stock estimate, the population of married couples in either group may be misstated because of the lack of an estimate for deaths. It is likely that the two groups will have different demographic profiles given the recent legalisation of same-sex marriage, meaning that a combined figure will likely overstate the number of marriages ending by death of same-sex couples and therefore give an artificially high divorce rate if we attempt to calculate changes of stock ourselves from combined data. 

On a broader point, successive governments have asserted the importance of marriage to the individuals concerned and the nation, as per the quotations above. The availability and existence of core statistics is important to ensure that emerging trends are fully discernible for both opposite and same-sex unions and that government policy making is able to react to this. 

Question three 

What would the impact be of not getting separate estimates? 


As stated in the previous answer, there would be two main impacts: 

Firstly, the quality of data available to test the evidential assumption on which same-sex marriage was legalised in Britain would decline significantly. This would both harm the accountability of policy-making in this country and remove an important data source for other nations considering similar legal changes. 

Secondly, while it would be possible to produce a crude reconstruction of the stock of both populations using other available data on the annual number of marriages and dissolutions in combination with a death rate for the combined population, this is likely to understate the population of same-sex marriages as the demographic is likely to be younger given recent legal changes. Again, this would hamper evidence based policy-making and analysis.