Figure 4


Download Data: Click Here

Source: World Values Survey 2006-2014, unless otherwise noted.

Note: In Central and South American countries where WVS data were not available for marriage, DHS data took priority over IPUMS. In IPUMS, it is likely that cohabiting and married people are combined, whereas they can be separated out in DHS data.

1: Source: Demographic and Health Surveys (DHS) 2001-2014.
2: Source: Integrated Public Use Microdata Series-International (IPUMS) 2000-2011. There are differences in definitions of marriage and cohabitation (see footnote 9 for details on cohabitation) in the IPUMS data source due to differences in wordings in individual censuses.
3: Source: Manpower survey of Israeli Central Bureau of Statistics (special analysis). Data on cohabiting not available.
4: Cohabitation was not a marital status response option in Qatar due to the cultural sensitivityof this item.
5: Based on men’s and women’s interviews which included living together as a distinct marital status; the household files included cohabitors with marrieds. Average of men’s and women’s proportions are weighted by the proportion in each gender at reproductive ages in the household population.
6: Estimate is based on women only.
7: Source: 2012 Chile Census.
8: Source: Organisation for Economic Co-operation and Development (OECD) 2013, Cohabitation rate and prevalence of other forms of partnership. Accessed on 11 June 2015). Data reflect all persons ages 20 and up for countries except New Zealand, where they reflect all persons ages 15 and up. For New Zealand, cohabitation data refer to the proportion of cohabiting people among those who have never married.
9: Source for cohabiting: IPUMS. In some censuses, cohabitors are identified only through relationship to household head. In these households, one person would be identified as an unmarried partner of the household head, and both the head and the partner could be assumed to be cohabiting. Limitations include 1) some censuses do not distinguish between spouses and partners, thus it is not possible to identify cohabitors; 2) unmarried partners yield an undercount of all cohabitors, since many cohabitors identify as a roommate instead of an unmarried partner; 3) cohabiting couples living in the household of a parent or other relative cannot be identified.