Opinion, Berkeley Blogs

Data! We want data! (so sayeth the social scientists)

By Martha Olney

Data are the life blood of the social sciences.  Imagine you want to answer this question:  What is the impact of the availability of legal marriage on gay individuals?  You could refine the question to look at child rearing, or saving for retirement, or patterns of spending, or labor mobility, or any of a host of things.

You might want to design your study to compare gay and straight people in states with and without legal marriage.  How did the introduction of marriage in a state change behavior of gay individuals?

You might then want to expand your study to see whether the behavioral change is due to the availability of marriage in one's home state, or the possibility of marrying even in a neighboring state.  Is it just making the commitment or also having that commitment recognized in my home community that matters?  And then you might want to expand your study further to see if the behavioral change is due to the availability of marriage in any state in the country.  If a couple from North Carolina travels to San Francisco and marries in October 2008 (Hi Chris & Darlene!), does being married alter their behavior?

You'd want to compare married individuals with those who are registered domestic partners, to see whether marriage per se matters.  And you'd want to look at the difference between marriage & domestic partnerships for gay individuals and then for straight individuals, to see if there's a different marriage effect for one group than for the other.

You'd be doing the sort of work that is done by The Williams Institute at UCLA's School of Law:  "advancing critical thought in the field of sexual orientation law and public policy."  You'd be doing the sort of work captured in Gary Gates' research summary, "Same-Sex Couples in the 2008 American Community Survey" -- but with data for every state and for every couple, not just a sample.  (The ACS is great for some things, but how many gay married couples living in North Carolina are picked up by the ACS?  Probably not enough to analyze.)

Historian Margo Anderson of University of Wisconsin Milwaukee published a book twenty years ago entitled The American Census: A Social History (Yale UP, 1988).  You can find it in the Main or Moffitt library at HA37.U55 A53 1988.  The census questions evolve and change as our society evolves and changes.  Some of us are married now.  We'll mark "married" when the census form comes.   Whether the census reports the data won't change how married we are.  But how married we are might ultimately change what the census reports.