Same-sex couples in the census

Several European Union countries have legalised same-sex marriage or created new forms of civil union open to gays and lesbians. But counting same-sex couples is still a complicated undertaking, note Clara Cortina (CSIC, Spain) and Patrick Festy (INED, France), in a working paper for the European Families And Societies research project. As they see it, no surveys or censuses have yet found the optimal means of identifying same-sex couples.

 

What makes it difficult to count same-sex couples?

  • Their low numbers: Same-sex couples are a relatively small group, both in absolute terms and relative to the population at large. Only extremely wide surveys such as population censuses can collect sufficient data.
  • The way census questions are formulated: On some census forms, those questions are ambiguous; they do not always allow for clearly identifying same-sex couples or distinguishing between different marital or union statuses. Coding errors during data collection or processing are common for the same-sex couple category precisely because of how few such couples there are compared to different-sex couples.
  • Underreporting of homosexuality: Some respondents may be reluctant to report living with a same-sex partner, and this may lead to underestimating the number of same-sex couples.

The situation in France:

Same-sex marriage was legalised in France in May 2013. But in 1999 the country had already instated a type of civil union, the PACS, that is open to gay and lesbian couples. According to INSEE, in 2011 France had 99,000 same-sex couples (0.6% of the total number), approximately 43,000 of them in civil unions and 56,000 cohabiting.

France is a particular case in that the afore-cited estimate comes from the 2011 edition of the Famille et Logements Survey (Family and housing), conducted about once a decade since 1954 and associated with the country’s population census.

  • The Famille et Logements survey: The 2011 questionnaire was the first to include a question on partner’s sex, making it possible to identify same-sex couples directly rather than through a question about the PACS. Later, the PACS was included among answers to the marital status question. The survey also takes account of couples whose members do not live under the same roof. Since this survey is associated with the population census, the information can be crosschecked to limit the risk of error. The collected data is then corrected.
  • The census: Census questionnaires are less explicit than those in the Famille et Logements survey; respondents cannot directly report belonging to a same-sex couple. On the general census form, respondents fill out a table of all the inhabitants of the given household; each member is asked to specify his or her kinship tie or relationship to the head-of-household. In the individual census forms, each inhabitant specifies sex, whether or not they are a member of a couple and if so their marital status. This provides INSEE with information that must then be tabulated and crosschecked.

Source: Clara Cortina and Patrick Festy, 2014, Identification of same-sex couples and families in censuses, registers and surveys, FamiliesAndSocieties, Working paper series 8 (2014)

Contact: Clara Cortina and Patrick Festy

Online: May 2014