The Census Rainbow

As the census machine chugs to a stop this year -- and all of us in the genealogical community celebrate turning ours in and looking forward to the release of the 1940 census in April of 2012 -- I learned something this week that gave me pause. Our beloved census, already fraught with accidental inaccuracies or those borne of untrusting ancestors, has also been guilty of recording intentional ones.

I don't mean intentional ones from our ancestors. We all know that this happened. We might curse our paranoid great-grandfather and end up taking another route to find the information he didn't feel like the government should know.

No, I mean intentional changes that the Bureau chose to do itself, like altering the sex of its citizens.

Yes, this has happened. It happened in 1990.

No doubt some of you are asking questions like, What? Why? How?

According to the Urban Institute : "Prior to 2000, there was no designation that enabled same-sex couples to accurately describe their relationship, and when a same-sex partner checked off the designation of "husband or wife," the Census Bureau altered the gender of one of the two people, assuming they had accidentally misidentified their gender."

In the 2000 census these standards changed slightly. That census allowed for an "unmarried partners" category, whereas two people of the same sex could categorize themselves without fear of arbitrary gender reassignment by the Bureau. However, if two people of the same sex described themselves in a spousal relationship, the Bureau recategorized them as same-sex unmarried partners. At the time, it was illegal for two people of the same sex to be married.

This census year, for the first time, the number of same-sex couples who self-identify as married — license or no license — will be tabulated. Thank goodness the Bureau won't go around reassigning genders again! Gah!

But this self-identification presents its own set of problems. On the one hand, this glimpse into the lives of people and their thoughts about their relationships gives us genealogists valuable pieces of life stories. That's wonderful. But then there's the other side of me that's perhaps just a stickler. That side prefers my census data to be lawfully accurate and able to point me in concrete directions. That side of me sees in the census that a couple was married and goes scurrying to find the records that prove it to be true.

If I was a researcher 70-odd years from now doing work in the 2010 census and had a gay couple in my tree that I was trying to trace and they indicated they were married, I would likely start looking for that license. I doubt I'd be alone. I anticipate there will be many such hiccups in research when these censuses finally come up for public release. It will be the responsibility of genealogists researching those censuses to learn the laws that were in effect and adjust their research accordingly. A smart genealogist does that already, anyway.

What I can't help but wonder is this: with the advent of the computer, the censuses primarily processed through it are likely the ones most vulnerable to digital altering. The random reassignment of sex is just one glaring example of how such a broad-based change could be applied to an entire nation's tabulation. While the censuses taken before 1980 or so and not originally processed via computer likely have the data relatively intact as it was written -- how can we be sure? As they are digitally processed, will they, too, be changed?

And lastly, when you run across that random 'mistake' of gender in your early censuses -- especially between the two primary adults in the household -- perhaps this might give you cause to stop and pause a moment. And wonder.

Things are never cut and dried. Even then.