When you sat down to fill out the 2010 Census form, what category did you choose for your relationship to the household head? Did you choose “husband or wife”? Or maybe “stepson or stepdaughter”? “Roommate”? Did it strike you as odd that you couldn’t choose “concubine,” or “polygamous wife”? Or, better yet, did you wonder why the form even requested your relationship status in the first place?
When the idea of a census was written into the U.S. Constitution it was a mechanism to ensure a representative democracy:
Representatives and direct Taxes shall be appointed among the several States which may be included within this Union, according to their respective Numbers. (Article 1, Section 2)
In those early years, the Census was primarily concerned with three things – age, gender, and slave vs. free status. From this count of individuals – “whole Number for free Persons… three fifths of all other Persons” – the members of the House of Representatives were divided among the states.* The government does not appear to have been be concerned with the relationship status of these early Americans
Over the years, the Census expanded beyond its original purpose to include all manner of questions including disability status, value of property owned, and occupation status, to name a few. It wasn’t until 1880 that questions about the “relationship to head of family” were added to the forms. Some of the relationships feel familiar by today’s standards – husband and wife, parent and child, brother and sister. But others provide an intriguing snapshot into households across the country:
- Live-in domestic employees like nurses, maids, and cooks, and their families.
- Boarders or lodgers
- Polygamous wives
- Concubines and their children
While it is debatable how common some of these living arrangements might have been, they were, nonetheless, available choices.
After 1930, all of these codes were scrapped and family relationships focused primarily on connections through blood and marriage. One could argue that these codes were eliminated because they became obsolete (How many live-in concubines have you met recently?), or that they have become incorporated into other codes (boarders become “roommates”). But the change in these codes also points to ways that the Census classifies people into categories which are socially acceptable. Polygamous relationships, while not common, are still practiced today (at least on “Sister Wives”) even if the category for it no longer exists.
These are not the only ways that the census form constrains our ability to accurately understand household relationships. Through the 1980 Census, romantic couples who lived together without marrying either had to classify their relationship as “roommates” or claim to be spouses. In 1990, the Census Bureau added “unmarried partners” as a relationship choice for opposite sex couples, providing a more accurate picture of domestic life.
For same-sex couples, clear recognition of their relationships has taken longer. In 1990, for same-sex couples that listed their relationship as “spouses”, the gender of one of the partners was recoded so all spouse relationships were male-female. It wasn’t until 2000 that same-sex couples were coded as “unmarried partners,” ten years after opposite-sex couples gained this designation. Furthermore, the 2010 Census was the first time the Bureau used Census data to report on same-sex couples separate from opposite-sex couples, in spite of the shared relationship category.
While classifying same-sex partners as “unmarried partners” is an improvement over previous decisions, it still fails to differentiate between those who choose not to marry and those who would marry if the option were available. In effect, it treats all same-sex relationships as if they were the same as opposite-sex cohabiting relationships. This portrays same-sex relationships as temporary and hides the fact that many same-sex relationships have marriage-like stability despite their lack of legal marriage status. In preparation for the 2020 Census, the Bureau is testing questions for same-sex partners to capture these differences.
While it may no longer make cultural sense to count all the concubines, it is important to remember that the Census, and other surveys like it, does more than simply count. It also sorts us into categories which reflect our cultural values. In addition, these proscribed categories constrain the ways we think about classifying our own relationships. In light of that, we should be mindful of who is represented in these categories and those who become overlooked.
* The 14th Amendment, ratified in 1868, allocates Representatives based on the “whole number of persons in each State,” thus revoking the three-fifths clause.