A few salient facts are known about the Americans whose lives might be changed by a Supreme Court ruling on same-sex marriage expected this summer.
About one in five gay and lesbian couples is raising children younger than 18. One in 10 men with a male partner or spouse is a military veteran. As many as 6 million Americans, roughly 2 percent of the population, have a parent who is lesbian, gay, bisexual or transgender (LGBT).
These nuggets of demographic insight into same-sex couples were contained in an amicus brief filed in connection with cases before the Supreme Court on the constitutionality of California's gay marriage ban and the Defense of Marriage Act.
A decade ago, such precise statistics were impossible to come by. Even now, many of the numbers commonly used to shape government policies are, for gays and lesbians, nonexistent.
But as gays become more visible in politics, demographic research into lesbians and gays is emerging from the shadows. Some gay advocates say it's time for surveys to ask people their sexual orientation point-blank.
"As a political and cultural issue, it's very important for us to understand how big and visible this population is," said Gary Gates, a prominent demographer of gay statistics who wrote the amicus brief.
However the Supreme Court rules, demographic knowledge about gays and lesbians is poised to expand further.
The National Health Interview Survey of 35,000 Americans has started asking respondents their sexual orientation, aiming to identify health-care needs. Last year, the federal government began putting a sexual orientation question in the annual workforce survey.
A Gallup poll last month found that 3.5 percent of American adults identify themselves as LGBT.
Gay activists say more research is needed, just to make the case that they exist.
"When our legislative affairs director goes into congressmen's offices, they're often told, 'I have no gay people in my district,' " said Fred Sainz, a spokesman for the Human Rights Campaign, a civil rights group urging protections for gays. "That's why this demographic information is so incredibly important."
Yet gay activists have been some of the most vocal critics of Gates, who lives with his husband in Seattle and is affiliated with the Williams Institute, a UCLA School of Law think tank that researches sexual orientation and gender identity. They say his work showing that 3.8 percent of Americans are LGBT underestimates their numbers and marginalizes their concerns.
Some of the controversy is rooted in a 1940s-era study by sexologist Alfred Kinsey, who estimated 10 percent of men had had same-sex experiences. That figure has been cited often by gay activists, according to Gates, to make the case they could not be ignored.
Some opponents of gay marriage say more demographic research can correct misperceptions of the size of the gay community. Gallup has found that Americans believe, on average, that 25 percent of the population is gay, several times higher than any research estimate.
"I think serious data on this is important to inform the debate," said Peter Sprigg, a senior fellow at the Family Research Council, which has argued that same-sex marriage is harmful to society. "People have an exaggerated view of how many homosexuals there are in the population. They don't realize what a relatively small population it is."
Yet some activists and academics say they suspect the percentage is higher than demographic research shows.
"When people look at demographics, they're often about white gays living in the 'gayborhoods,' " said Jaime Grant, a feminist researcher at Kalamazoo College who heads the Arcus Center for Social Justice Leadership. "But in communities of color, there's already a resistance to filling out census forms, and it's even more so for gays of color."
It was long considered career suicide to specialize in research into lesbians and gays. That began changing in 1992, when election-day exit polls asked voters their sexual identity.
"I had some interest, being gay," said Murray Edelman, who ran exit polling for the networks. "If you weren't on the exit poll, then you didn't exist. . . . It was another way of being counted."
Since then, more lesbians and gays have been willing to identify themselves and demand for gay statistics has grown from public-health agencies, the military and marketers.
Some of the biggest hurdles remain. Gathering data about gay people is much more complex than asking about race or income. Should people be asked about their sexual attractions, behavior or identity? Which of the many synonyms for gays should be used? How do researchers count gay individuals who are not living with a mate?
"We have guesses we're missing over half the gay and lesbian population if we just focus on partnered individuals," said Amanda Baumle, a University of Houston sociologist and demographer.
The decennial census now provides hard numbers on the growing number of same-sex couples willing to identify themselves as such. The 2010 census counted 646,000 same-sex couples, including 130,000 who were married.
The National Gay and Lesbian Task Force, which waged a Queer the Census campaign urging same-sex couples to identify themselves, wants more government surveys to ask about sexual identity and behavior.
"It makes us visible and helps us identify ways our community needs support," said Darlene Nipper, deputy executive director of the group.
Gates, 51, a former software engineer and seminarian, sits on advisory committees that help formulate questions on government surveys. He said demographics can only bolster the case for marriage equality.
"If the arguments wind up being anecdotal, the arguments of both sides end up being given equal weight," he said. "The challenge is, science doesn't always find everything that fits a particular political agenda."