Same-Sex Marriage and the 2010 United States Census

The Constitution commands that an enumeration of the population be made by the federal government every ten years, for the immediate purpose of allocating seats by population in the House of Representatives and, by extension, electoral votes for presidential elections.  Of course, this decennial census is important for many reasons apart from those.  Ever since the 1960s, when the Supreme Court adopted the "one person, one vote" rule for all legislatures except the US Senate, the Census has also become vitally important for drawing up district lines for state, county and municipal legislatures.  And, as federal programs that premise rights or benefits have grown, population data has become essential for the dispersal of federal funding for programs.  States and localities also rely on Census data for their budgeting and allocation of resources for programs.

The Defense of Marriage Act (DOMA) says in Section 3 that the federal government will not recognize same-sex marriages for any purpose of federal law.  Taking that as a literal command, the Bush Administration took the position that the 2010 Census would not count as married any same-sex couples, regardless whether they were legally married under the laws of a foreign country or one of the states that authorizes same-sex marriages.  When the last Census was held, in 2000, the question of counting same-sex marriages was irrelevant, because same-sex couples were not allowed to get married anywhere in the world!  It is amazing to think that in less than 10 years same-sex marriage has become available in several other countries and that by the time the Census is taken next spring, it may be legal in more than half a dozen states.  (The picture in the U.S. is complex.  As of now, same-sex couples can marry in Massachusetts, Connecticut, and Iowa, with new laws extending the right to marry to go into effect in the coming months in Vermont, New Hampshire, and Maine.  But the Maine law may not go into effect if sufficient signatures are submitted to the state government to force a "people's veto" on the ballot.  The NY State Assembly has passed a marriage bill that is pending before the State Senate.  The NJ Legislature may be poised to enact a marriage bill after the November election.  It is possible that pending litigation will revive the right to marry in California, but certainly not in time for the 2010 Census; on the other hand, there are about 18,000 couples who married in California prior to the passage of Proposition 8 and whose marriages have been validated by the California Supreme Court, and there is still an open question about the marital status of California residents who married in other jurisdictions prior to the passage of Proposition 8.  I said this was complex....)

So, what should the U.S. Census do in 2010 when U.S. residents who live as legally married same-sex couples mark their Census form as "married"?  The software in use from 2000 would note the genders of the individual respondents and reject the category "married" for such couples, instead classifying them as unmarried partners.  (The 2000 Census was the first one to begin providing anything approaching a reasonable statistical picture of same-sex couples in the U.S., by including a category of unmarried partners instead of just counting every unmarried person in a household as totally unattached to the other unmarried people in a household.)  Last winter some gay scholars and journalists took up the cry that the Bush Administration's interpretation of DOMA as precluding an accurate count of same-sex married couples in the U.S. was preposterous and homophobic.  With the election of a new administration that had announced its dedication to achieving equal rights for gay people, this was an important challenge.  And some unhappiness was expressed as the new administration went into its fifth month with no announcement that this issue would be addressed.

Then late last month there was a big blow-up in the gay blogosphere and media as more and more people became informed about a legal brief filed by the Justice Department (DOJ) in a pending lawsuit in California that challenges the constitutionality of the Defense of Marriage Act.  DOJ defended DOMA's constitutionality, arguing that the issue was so clear-cut that the trial court should dismiss the case outright, using some arguments and examples that commentators found offensive and contradictory to President Obama's election promises.  In the wake of the outcry, there came a rush of announcements of various administration initiatives on gay issues.  One of them involves the Census.

By the end of June, word had come out that administration lawyers had determined that DOMA does not require failing to count and enumerate and report same-sex marriages, although of course they can't be recognized legally for any purpose of federal law until DOMA is repealed.  (The president has pledged to work for repeal of DOMA, but not right now.)  Now the Census Bureau faces logistical challenges of quickly gearing up to implement the administration's decision that same-sex marriages should be counted and reported as marriages.

An article in USA Today that appeared in several US newspapers on July 6 sets out some of the problems they are confronting.  They must make a change in the software that was used to read the machine-readable census forms in 2000, since that software will immediately reject the classification of "married" if the individuals involved are of the same sex.  Writing and testing new software in time for the spring administration of the Census is difficult, as anybody who has been involved in a project to custom-write and install new software can appreciate.  You don't want lots of glitches on something as important as the Census.  At my school, the introduction of new software for the class registration system took years and went through several trial runs before we actually were able to conduct on-line course registration.  Now project those difficulties onto a project to enumerate and classify more than 300 million people... Doing this is not easy.

In addition, there are important questions about how the data should be presented.  For one thing, if DOMA has not been repealed by the time the Census data is ready to be reported, it seems likely that there will have to be distinctions made in the reported data between same-sex and opposite-sex marriages since so many different programs of federal, state and local governments use census data for their budgeting purposes, and appropriations for programs are also directed on the basis of Census data.  If these marriages can't be recognized for programmatic purposes, then they will have to be presented as a separate item line.  And, in addition to federal DOMA, more than 40 states are governed by constitutional amendments or statutes that similarly prohibit them from recognizing same-sex marriages, so the complications mount.

The article also points out the phenomenon of inaccurate self-reporting.  Many same-sex couples who have gone through commitment ceremonies, or holy union ceremonies through their religious denominations, will consider themselves "married" and may check that box, even though they have not taken the step of legal marriage.  There are now enough jurisdictions offering same-sex marriage without a residency requirement that anybody who wishes to marry and has the resources to travel can do so, but that's a substantial trip in many parts of the country.  If what the Census hopes to do is generate reasonably accurate data on legal same-sex marriages, that may prove a hopeless task without some method of cross-validating the answers against state marriage registries.  I'm not aware that they make any attempt to do cross-validating on different-sex marriages, and I bet there are plenty of inaccuracies there as well.  (I was told the story of a major non-profit agency that adopted domestic partnership benefits for same-sex partners and decided, as a matter of equity, that they couldn't require documentation of a partnership without requiring documentation from different-sex partners that they were actually married.  The result was to discover a significant proportion of their employees receiving spousal benefits who were not legally married....  so I suspect that Census data on marriage is already inflated by heterosexuals who self-report as married but have not taken the legal steps.) 

The Census Bureau people have many issues to ponder... which makes it even more ridiculous that Republicans in the Senate have been holding up confirmation of President Obama's highly-qualified appointee to head the agency, Professor Robert M. Groves, because of their articulated fears that he is an advocate of "sampling" in order to correct for undercounts of hard-to-count populations.  While the Republicans fight about a policy that they fear will lead to a more accurate (or, in their view, possibly inflated) count of transient and minority voters, who are more likely to vote Democratic than Republican, the Census Bureau remains without its designated leadership as the days tick away and the actual enumeration required by the Constitution looms.  In addition, the Bush Administration starved the Census Bureau in its appropriations, so this vital job has been underfunded and there is lots of catch-up going on to get ready to conduct the enumeration. This is one government operation that can't be put off.  Redistricting to reflect shifting population must be done in time for the 2012 federal election.  State legislatures need to get down to work in time to assure that the next federal elections can be conducted in a timely manner, and that the allocation of electoral votes for the 2010 presidential election properly reflects the county's population.  Let's get moving, guys!!

Comments

[Ken Sanders]

Remember that "jumping over the broom", or self-proclamation of marriage ("common-law" marriage) was a valid form. For an era in which religiosity has altered form and importance, the validity of religious marriage has been subsumed into the civil. Perhaps we should again recognize mutually proclaimed marriage as being valid, whether implicitly or explicitly declared to the world. The difficulties should not be insurmountable. Perhaps we could use the reasonable spouse standard, i.e. if a reasonable spouse would believe a couple were married, then the couple is married. Further, why should we penalize any individual partner of a pair when the other partner decides it is quits? Commitment is commitment and unlike contracts, it only takes one partner to make a marriage (at least if that one is deluded by the other). Of course, it only takes one partner to make a divorce. I say: live together under specified conditions ... and I leave it to legal philosophs to determine those conditions (although I favor the reasonable spouse standard) ... and satisfying those conditions one should be married with certain shared obligations flowing from each partner toward the other. In a "divorce" in this form of marriage, it is necessary to go through a formal divorce. Of course, it must be hard to get a jury of reasonable spouses, so perhaps the reasonable person standard would be better.

[Art Leonard]

I think one of the important policy considerations underlying the decision in all but a small handful of states of abolishing common-law marriage was that so many rights and responsibilities now turn on marriage - hundreds in state law, more than a thousand in federal law - that we need a bright-line test, we can no longer rely on the kind of fact-specific, case-by-case determination that was possible in a simpler time. So we need the official state ceremony. I'm not sure how those common-law states get along without it....