2
was not legally possible in their country.
Likewise O’Connell and Lofquist (2009) reported that of the
same‐sex couples identified in the 2000 Census, 43 percent reported as spouse (prior to an edit that
reclassified them as unmarried partners). A more recent study conducted a few months after the 2010
Census by Gates (2010) attempted to understand how same‐sex couples reported in the 2010 Census.
Using an internet panel with pre‐identified same‐sex couple households, Gates presented respondents
with a facsimile of the 2010 Census relationship question, then asked which category they had selected.
He found that over 90 percent of same‐sex couples either were legally married, lived in a state that
recognized the marriage, and reported choosing “husband/wife” or lived in state that did not recognize
same‐sex marriage and reported choosing “unmarried partner.” Additionally, he reported that 84
percent of those in civil unions or registered domestic partnerships used “unmarried partner” to
describe their relationship, while 97 percent without any legal relationship status reported selecting this
category.
Far fewer studies have systematically examined different approaches to collecting data on same‐sex
couples (and the change in estimates that result from these differences). A review of demographic
surveys conducted by the Census Bureau revealed a surprising variety of categories and questions for
both relationship and marital status. Some surveys do not contain a relationship category to easily
classify unmarried same‐ (or opposite‐) sex couples (e.g., the National Crime Victimization Survey ‐
NCVS, the Fishing Hunting and Wildlife‐Associated Recreation Survey – FHWAR, and the American
Housing Survey ‐ AHS), while others do present an unmarried partner option (the 2010 Decennial
Census, the Current Population Survey ‐ CPS, the National Health Interview Survey ‐ NHIS, and the
Consumer Expenditure Survey ‐ CE). Still others have introduced new/expanded categories with less
traditional offerings, for example:
2
Mortensen and Orten (2009) report that the category “married” tends to be overreported in Eurostat statistics,
but do not disaggregate between same‐sex and opposite‐sex relationships.
4