Census clarifies same-sex couple data

The U.S. Census Bureau released revised same-sex couple data Tuesday, saying errors by census workers in recording responses to the 2010 Census "artificially inflated" the tally. The bureau also released data on same-sex couples who said they were married, the first time it has recorded that information.

According to the bureau's revised totals, called the 2010 "preferred estimate," there were 48,932 same-sex couples in New York, 10,125 of whom were same-sex spouses, down from the 2010 Census count of 65,303 same-sex couples, of which 23,978 were recorded as spouses.

Nationally, there were 646,464 same-sex couples, according to the 2010 preferred estimate, down from 901,997 in the 2010 Census. And according to the preferred estimate, there were 131,729 same-sex spouses in the nation, down from the 2010 Census count of 349,377. Same-sex couple households represent less than 1 percent of all U.S. households, the bureau said.

The problem in the tabulation of same-sex couples largely came as enumerators went door-to-door to follow up with people who did not mail back the census form. The enumerators used a form that was designed differently from the one that was mailed to households, Martin O'Connell, chief of the bureau's fertility and family statistics branch, said at a news teleconference. The enumerators checked the wrong box for gender during interviews of heterosexual couples, making it seem as if there were more same-sex couples than actually existed, he said. Small errors in a small population, such as same-sex couples, "can produce very large error rates," O'Connell said.

He said the bureau would not be releasing same-sex couple information for geographic areas below the state level, citing the unreliability of the data in smaller geographies and confidentiality concerns.

David Kilmnick, executive director of the Long Island GLBT Services Network, said though the Census was the "best attempt" at counting the nation's same-sex households, it was flawed. "If we want to see real data on how many same-sex households we have, then we need to ask the question straight out, no pun intended."