How does a data set like the US census deal with people who defy its expectations? How can a data set encompass those its designers never imagined? These questions can’t be answered with the published numbers, the final facts. They can only be answered by seeking out stories deep within the data, stories like this one: Buy This Book At: If you buy something using links in our stories, we may earn a commission. This helps support our journalism.
Learn more . Margaret Scattergood had been one of the few women in the room where the census questions had been debated. A year and two months later, Scattergood faced those questions herself.
She had to fit herself into the frame her colleagues had constructed. Scattergood had been an outlier in the Commerce Department auditorium; she would be an outlier on the census sheets too. The 1940 manuscript census schedules record the results of conversations that took place across millions of doorsteps as over 120,000 census takers spanned the nation and started asking questions.
They tell us, for instance, that an enumerator named Richard Gray visited Scattergood’s house in Fairfax County, Virginia, on May 25, 1940. He was late, reaching this country home nearly a month after rural spaces were supposed to be enumerated—it seems likely that he had attempted to enumerate the household earlier, but could not find the residents at home. On this visit, he appraised the house’s value at an impressive $50,000.
It was (and is), by all accounts, beautiful, bucolic, and grand. Gray then listed three residents: 57-year-old Florence Thorne, white, single, with four years of college education, the “assistant editor” for a “labor union”; 45-year-old Margaret Scattergood, white, single, college educated, and a “researcher” for a “labor union”; and 50-year-old May Stotts Allen, divorced and (apparently) entirely unschooled—she was listed as “W” for white, which was then scratched over with a darker “Ng” for “Negro. ” Gray was supposed to mark which of the people in the household he spoke to directly, but he did not in this case, and so it’s impossible to say if Scattergood encountered the questions herself.
We do know how Gray made sense (in census terms) of these three middle-aged women living together. He made the eldest, Florence Thorne, the “Head” of the household, writing her name first. Allen he listed last, as the “Maid,” related by her servile status.
Scattergood, in the middle, became a “Partner. ” “Partner” is a curious label, a term that can have a jumble of meanings. Partners might run businesses or law firms.
Some of us have partners in crime. These days, partner mostly means lover or companion. Used by queer and straight alike, married or not, partner now often indicates a long-term intimate connection.
That usage isn’t even new; in his 1667 masterpiece Paradise Lost , John Milton made the parents of humanity into partners, placing in the ur-lover’s mouth, the mouth of Adam, speaking of Eve, this lament: “I stand Before my Judge … to accuse My other self, the partner of my life. ” Is that what the census taker had in mind when he labeled Margaret Scattergood partner? There were just over 200,000 other people in the continental United States labeled partners in 1940. Were they all romantically involved? For most of the US census’s first century, it wasn’t possible to be labeled a partner, as Scattergood was.
The “Relation” column, which asked all individuals in a household to explain their position vis-à-vis the family’s “head,” didn’t even come into being until 1880. This column reminds us of the census’s hidden-in-plain-sight secret: the survey may list individuals to make the data easier to manage, but individuals are not its fundamental unit. To be counted by the census, a person must be placed within something resembling a family.
Before the 1850 census, only heads of household (male by default) were named in the survey, and the rest of the household became simply a set of tally marks or numbers attached to that name. The 1850 shift to naming each individual looked revolutionary, but it was mostly a data-management strategy, a way to better keep track of answers to a lengthening list of questions. It left the essential nature of the census unchanged—the basic organizing unit of the count was what its architects originally called the “family.
” The statisticians devising the census instructed enumerators to name male heads of household first and then list the wife, children, and then servants or lodgers or other members of the “family” who were not related by blood (such as an apprentice). In coming decades, families were rebranded as “households,” but the logic remained unchanged. Even in 2020, any American who filled out a census form had to place themselves within a household (though without any explicit “head”), even if just in a household of one.
So why did it become necessary to put labels on household “relations”? Why make that shift in 1880? It might have had something to do with another big change that happened at the same time: The statisticians were also given the authority, for the first time, to organize the enumeration and oversee the recruitment of enumerators (with the approval of each ruling-party congressman, of course). As Washington, DC, won more control over the question-askers in the field, it simultaneously might have worried about the agents it couldn’t control. The “Relation” column might have been a tool for keeping an eye on the people asking the questions as much as on those who were answering them.
If that sounds strange or like wild speculation, it isn’t. In fact, it’s the precise justification given by key Census Bureau professionals for keeping the “Relation” column in later years. On the one hand, Census Bureau officials faced the enduring problem of trying to squeeze too many questions into too little space.
That consideration led to a vote in 1917 to get rid of the “Relation” question, freeing up a column for other inquiries. On the other hand, officials faced an enduring problem that would ultimately rescue the column from removal: Enumerators sometimes made up people, a fraudulent practice common enough to have its own census-specific name: “padding. ” (There was actually a whole census-specific vocabulary for making up people, including the lovely verb curbstoning that alluded to the enumerator who filled out the census sheet with imagined data while sitting at the curb, unencumbered by actual knowledge of actual people in actual homes.
) The 1910 census had been plagued by padding fiascos—most spectacularly in Tacoma, Washington. Boosters intent on inflating their city’s status and influence conspired with local enumerators to make up more than 30,000 people. The enumerators themselves, who were paid for each head they “counted,” proved perfectly happy to help.
They invented about a quarter of the city’s reported population! Census Bureau officials wondered what they could do to prevent another epidemic of fraudulent entries. Unsurprisingly, they decided to ask Congress to change the census law to make it easier to prosecute rogue enumerators. Their other big move was more surprising, and also quite telling.
Detailed relation data served as a valuable tool for sniffing out padded entries, so officials decided to overturn their earlier space-saving decision: The “Relation” column had to stay. The relationship category—as a category, as part of the frame of the data—points to a bureaucratic system concerned about disciplining possibly corrupt workers. It also tells us something about the nation and the institutions that required it.
It gives evidence that the data’s designers believed that there was such a thing as a normal family and that they could distinguish normal from abnormal households, and that real people would belong to normal households. “The fiction of the census is that everyone is in it,” wrote the theorist of nationalism, Benedict Anderson, “and that everyone has one—and only one—extremely clear place. ” But governments exert power by deciding what each possible “clear place” is, and people with power have an easier time getting themselves placed as they’d like to be placed.
“Partners” became “partners” because no “extremely clear place” for them could be found. They defied norms. Census enumerators resorted to the partner label just over 50,000 times in their very first effort to label every American’s spot in a “family” in 1880.
That translated to about 0. 1 percent of everyone enumerated, 0. 1 percent of the population with no other “clear place.
” Some of those partners look like holdovers from the earlier era of family-centered enterprises. The dry goods merchants Henry and George Combs might have been brothers, but an enumerator in 1880 labeled George a “partner” to Henry. There are, in fact, quite a few dry goods merchants or dealers who show up in that census as partners.
The term proved useful for fitting merchant capitalism into the form. It also made it possible to aggregate unattached, usually male workers who were at that moment driving the development of American industry in the West. In Nevada an enumerator might (and did) fill a sheet with partners, all Italian immigrants working as “coalburners.
” In California, enumerators used the “partner” label to describe large groups of Chinese immigrant men living together, at a moment when a quarter of the California workforce was Chinese. The Chinese, like the Italian miners, were busy building California. The census translated their impromptu labor camps into households of partners.
Such partnerships scared many white Americans, who cast them as direct competitors to the nuclear family. Sinophobes, especially in California, did not see laborers powering industrial development. They saw an unassimilable people whose failure to live in traditional family arrangements and whose willingness to work for lower wages threatened the ability of white working men to act as proper heads of their households.
One enumerator in San Francisco came upon a household of Chinese workers living together, working in a laundry. He didn’t bother to classify them as partners. Instead, he turned their occupation into a familial relation.
They became laundrymen in both the relation and occupation columns, creating another category for those who didn’t fit the census’s expectations. “Laundryman” wasn’t an official label in 1880, but neither (as far as I can tell) was “partner. ” In the instructions to enumerators that year, the Census Bureau felt the need to define a family but not to specify all the acceptable family relation labels.
The bureau cautioned enumerators that families for statistical purposes might not look like, well, families. “The word family , for the purposes of the census, includes persons living alone,” read the instructions, “equally with families in the ordinary sense of that term, and also all larger aggregations of people having only the tie of a common roof and table. ” Note that “table” mention.
Tables mattered. Tables defined families—urban families, at least: “In the case … of tenement houses and of the so-called ‘flats’ of the great cities, as many families are to be recorded as there are separate tables. ” Over the next 60 years, through six censuses, “partner” slid into an officially sanctioned status.
It became a clear place for placing the otherwise unplaceable. The 1940 instructions told enumerators: “If two or more persons who are not related by blood or marriage share a common dwelling unit as partners, write head for one and partner for the other or others. ” It was the very last line in the section on relations, almost an afterthought, a literal last resort.
The “partner” puzzle first presented itself to me some months after I began researching this book. I had been reading through 1940 census sheets for over a year and had never noticed a “partner” label. Then I received a message from a colleague at Columbia University’s Teacher’s College, Ansley Erickson, asking what it meant when two women were living together and one was marked “partner” to the other.
Erickson had found such a record. I did not have an answer for her then, but I knew I needed to find one. I also had a personal reason for wanting to learn more.
“Partner” was a descriptor we had started to use more often in our family, a term that fit better as our marriage got queerer and queerer. But I had no idea where the term in the census had really come from, so I began looking around. Beyond my personal reason, I was drawn to studying the “partner” label by a hunch.
I believed that chasing after “partners” would help me think about how census data dealt with queer and other marginalized people generally. I believed that chasing after “partners” might also provide an opportunity to offer some token of historical recognition to people who lived on the margins of their society, the sort of people who always have less of a shot at making it into the history books. Chasing after “partners” did not disappoint on either count.
Indeed, it led me to a series of doorstep encounters I could never have imagined from examining the final census numbers alone. On April 2, 1940, Lillian Rita Davis knocked on her very first door, an apartment in a small building on New York City’s West Eleventh Street. It was a marginal neighborhood—or maybe better to call it edgy—at the edge of respectability, the edge of wealth and legitimacy.
New York’s powerful labeled the area a 3 or C, colored yellow for Caution. It was “a very old district” made up mostly of “tenements” according to the accompanying report. About 20 percent of the buildings were “Rooming houses, etc.
Miscellaneous. ” Any reference to rooming houses augured a poor grade, because rooming houses filled up with people unbound from conventional nuclear families. (The exceptions were a couple of the poshest neighborhoods, near Central Park, where the rooming houses were luxury residential hotels.
) Emily H. Brand answered the door. She told Davis she worked as a secretary for a theater, getting by with not much money.
She was college educated, 29, white, and single. She told Davis too that she lived with Katrina S. Grant, who was also college educated, 27 years old, white, and single.
Grant earned quite a bit more money as a social worker. We don’t know if Brand called Grant her partner, or her roommate, or her friend, or her companion, or her lover. All we know is that Davis saw the pair—a pair seemingly plucked from the training materials—as a head (Brand) and a partner (Grant).
Davis went on to record 12 more “partners” in her district, many more than the three that pure chance would predict. Partners had a way of coming in clumps, clumps clustered at the margins, in places like Greenwich Village. Davis’ partners usually paired to people designated with the same sex.
(The census, then and now, reduces the actual variety of the world to only two options for sex or gender: man or woman, excluding or misgendering nonbinary and transgender people in the process. ) Among those partnerships, five partners were women living with another woman, the pairs always within a few years in age. They included a pair of doctors, two travel agents (who may have been business partners too), an editor and a secretary, and a secretary and a stenographer who both worked for the YMCA.
Another stenographer was aptly named Tessie Finger and partnered to a secretary whose name was the delightfully alliterative, pleasure-proclaiming Lee Lustgarden. (At least those were the names the enumerator wrote down. ) Two of the partners were male-male, including a police detective who lived with a patrolman.
There is no way of knowing if these living arrangements enabled intimate relationships, if they were romantic or sexual, or if they were simply the result of people finding roommates to help pay the rent. Whatever the reasons for the pairings, these partnerships troubled the premise that every household had to have a “head,” and for that reason, we can count them queer. They pushed the boundaries of the data’s assumptions and transgressed the Question Men’s ideas about what a normal household looked like.
Uptown, in Harlem, Vera Maude Smith knocked on her first door in Enumeration District (E. D. ) 31-1723B on April 16, 1940.
The neighborhood Smith enumerated had been shaded red, scored with a D or 4. The accompanying description explained: “Formerly a good residential area largely one-family dwellings. Now almost entirely negro–tenements and converted dwellings into rooming houses.
Rentals are fair due to crowding. ” It’s worth pausing on that last line, to let it sink in. In the eyes of the banker or real estate man, the crowding of Harlem driven by the strict racial segregation enforced by the realtors drove up rental rates and so made rentals the one thing that were worth investing in.
The rest of the neighborhood suffered from both withheld credit and artificially cramped quarters. If in Greenwich Village partners looked as though they had walked right out of the sample sheets used to train enumerators, the Harlem partners reveal the limits to such illustrations. Looking to Harlem, we see evidence that what a “partner” meant and how it manifested varied from place to place, from community to community.
Partners came in clumps, and each clump had a distinct character, variations that we can read and recognize in the manuscript data. The first time the enumerator Vera Maude Smith applied the partner label, it was in a household that was at once typical of the other 14 partner pairs in Smith’s district and in some respects a significant outlier. What made it typical was what made it most different from the partner households in Greenwich Village: It was a partnership featuring at least one married person—this was the norm for partnerships in Smith’s Harlem.
Smith’s first partner was 42-year-old Elizabeth Hickson, who was listed as divorced. Smith marked Hickson’s race as “Ind” for Indian—as in Native or Indigenous—but that appears to have been an error on Smith’s part, because in a supplementary question, Hickson revealed that her mother was born in India (and her father in Indiana), which would have made Hickson’s race “Hindu” according to the census. Hickson became “partner” to the head of her household, William Roth, a 45-year-old handyman by trade and one of the few white men living in Smith’s district.
Smith marked Roth with an “M” for married. But he was not married to Hickson. The census was silent on the identity of his spouse.
The census was not finished in its marking of the Roth record, though—just as it would not be finished with others who were separated from spouses. Some weeks or months later, after the enumerator Smith’s portfolio of sheets had traveled to Washington, a Census Bureau editor would take a pen to Smith’s sheets (and every other census sheet), preparing the entries Smith made so that they could be punched into paper cards and eventually tabulated. Whenever that editor found a married person living apart from their spouse, that editor placed a line through the letter “M” and replaced it with a number: 7.
Ordinarily, every “M” on a sheet was supposed to be translated onto a punch card as a 2, in the second row. (Row 1 was for “S,” or single people. ) But married people living apart from their spouses did not get to be ordinary M2s—they instead were reclassified as a distinct group, M7s, at the bottom of the punch card, an explicitly marginal category.
The Census Bureau treated married people living apart from their spouses as anomalous and also as a group subject to statistical scrutiny. As Smith made her rounds enumerating, she ran into more pairs that she judged to include a partner, and at least one person in most of those pairs was married—just not married to the person they lived with. So Smith kept writing down Ms that an editor in Washington would change to a 7, rendering the record remarkable or possibly suspect.
There was Olivia Parker, a 31-year-old Black woman, an M7 pastry chef with two children, who Smith listed as a partner to Ollie Simmons, a 25-year-old single Black man, a dancer. There was Herbert Hill, a 25-year-old Black man, a porter, partnered to James Parker, a 22-year-old Black man who worked in a garage as an auto mechanic. There was Inez Reid, a single 25-year-old Black woman with two years of college education, working as a servant, partnered to Frannie Dozier, a 27-year-old Black woman, an M7 from South Carolina who also worked as a servant.
A couple doors down from them lived Arline Brooks and Marie Wescott, both M7s, both in their twenties and from South Carolina, working in a dress factory, and so Smith made Marie partner to Arline. The Great Migration of African Americans from the rural South to cities like New York and neighborhoods like Harlem enabled or caused or forced many spouses to separate, for a short while or for good. Hearts were broken; others were freed.
The census sheets cannot tell us which were which. Each of these partner clusters turned up in neighborhoods that money and capital had relegated to the margins. Each was also, not coincidentally, a queer neighborhood.
According to George Chauncey in his book Gay New York , Lillian Rita Davis’ Greenwich Village was considered New York City’s “most infamous gay neighborhood by outsiders. ” And Harlem played host to the “riotous Black girls, troublesome women, and queer radicals” who historian Saidiya Hartman celebrates in her book Wayward Lives, Beautiful Experiments . In each neighborhood, the existence of rooming houses and the possibility of renting a room—often a furnished room—as a single person or part of an unmarried pair created the conditions of possibility for queer communities to form, and the conditions of possibility for partners to cluster in the census.
If every normal household is the same, every peculiar one is peculiar in its own peculiar way. Partner labels made it possible to place every household by absorbing those peculiar households that otherwise wouldn’t fit. The partner label made it possible for a household census to count at the margins of society.
It created a big tent filled with an enormous variety of arrangements for living together. In early 2020, Washington Post reporters investigated Margaret Scattergood’s partnership with Florence Thorne. The reporters weren’t interested in the census; they were interested in the Central Intelligence Agency.
Thorne and Scattergood had been in bed with the spies, in a manner of speaking. The pair sold their house to the CIA in 1948 on the condition that they could go on living in it until they died. (Thorne died in 1973; Scattergood died in 1986.
) The house has since been turned into a conference center for the CIA’s Langley headquarters (the Scattergood-Thorne Conference Center). The reporters dug up the 1940 census record and pondered its possible meanings. They wrote: “If that partnership included romance, it was a well-kept secret.
” According to their investigation, “Scattergood’s family members, who have diligently kept records of her life, say the pair had separate bedrooms and never acknowledged a relationship beyond friendship. ” The problem is that the “partner” designation cannot really tell us that much about the people it was applied to. It wasn’t meant to reveal intimate details.
After Richard Gray enumerated Margaret Scattergood, one of the few women among the Question Men, her record would have made its way back to a Virginia district office, then on to Washington, DC. There, a punch card operator would have translated Gray’s handwritten observations into holes in a tabulating card. The column on the card for relation had a limited number of options, because the card had only so much space.
The options were: Hd, Wif, Chi, Par, GrC, OtR, Lod, Ser, Ot, and Inm. Hd for “head” is clear enough, and so is the second option: “Wif” for “wife”—not spouse, but wife. (Even the punch cards enforced the norm of the patriarchal family.
) The other abbreviations stood for child, parent, grandchild, other relative, lodger, servant, other, and inmate. That “Par” is tantalizing, but it isn’t for “partner. ” Scattergood, as a partner, would have been punched in and tabulated as “Lod,” a “lodger.
” When it came to making statistics, all the partners of this chapter simply disappeared. That is why their stories cannot be found in the published numbers. It had never really been about the partners; the point had never been to single out partnered households.
The philosopher Ian Hacking has explained how “enumeration demands kinds of things or people to count. ” He continued: “Counting is hungry for categories. ” If suitable categories don’t yet exist, the designers of data must invent them.
The language of class that defined the political economy of Karl Marx: That began as the language of worried bureaucrats counting people. The idea of poor revolutionaries climbing the barricades (possibly singing)—call them “Les Misérables”—that too began as a “standard set of pages in statistical reports” produced by nervous officials in early-19th-century Europe. Hacking has shown that the categories created to keep track of a worrisome group, the categories printed in blue-ribbon reports and official statistics—those categories can take on a life of their own.
That didn’t happen with partners in the census. Concerns about adults living together outside of marriage did not inspire or require the partner label. In fact, the government cared so little about this group that it never tabulated or printed any data about partners.
With no partners on the punch cards, a tabulation wasn’t even possible. So, then, why bother labeling Margaret Scattergood as a partner in the first place? Partner labels preserved the social structure and the data structure. The “relation” column persisted on into 1940 by bureaucratic fiat: Officials believed it helped to tamp down fraud in the enumeration process.
They banked on being able to sniff out abnormal household patterns, or just thought requiring a relation label would help deter false enumerations. But not everyone who enumerators encountered at the nation’s doorsteps measured up to the official ideal of a normal household—one with a clear head, usually male. The partner option provided a release valve for enumerators encountering queer couples, colonial laborers, rent-sharing roommates, or recent migrants from other counties or other countries—recognizing that in many cases, “partner” labeled people who might fit more than one of those descriptions.
The “partner” was an “other” option, but one that might have been easier for enumerators to use in doorstep negotiations. The way the label appears in clumps also hints that some census takers and some communities took to it on their own, perhaps because the term already made some sense to them. Though we have seen the tremendous variety of people and circumstances encompassed by partnerships, the job of making statistics rests on smoothing out variations and abstracting away difference.
Back in DC, the bureau decided to subsume the variety of partners and the specificities of its use within the category of “lodger. ” They charged its editors and card punchers with that translation. Such are the sorts of choices that statisticians must make—though there seems to have been little chance for public discussion of this choice.
Enumerators at over 200,000 doorsteps judged people to be “partners. ” Those judgments made those people countable and allowed them to show up in final statistics, though not as partners. Reading beyond the numbers, we can recover a series of lost histories, of communities of people who did not otherwise seem to fit, from Harlem to Hawaii.
Margaret Scattergood became a partner, then dissolved into the mass of the nation. Reading the data deeply now, we can see her and acknowledge her, and her choice to live outside the bounds of a patriarchal household, and the unseen negotiations that made her, maybe only briefly, a “partner. ” When Americans completed the census in 2020, they found two partner options: “opposite-sex” and “same-sex.
” These partnerships would not be ignored, thanks in large part to decades of queer advocacy . The Census Bureau had included a partner option in 1990, to capture unmarried heterosexual households, but LGBTQ activists organized their communities to adopt the label as their own. Those same-sex partnerships weren’t allowed into the final tallies by bureau officials, who declared that they couldn’t tell actual same-sex partners from those who had accidentally checked the wrong box.
Continued lobbying reversed that policy, allowing the same-sex partners of 2000 to live on in statistical annals, and then finally in 2020, the bureau introduced separate checkboxes for queer and straight partners and spouses. More advocacy remains needed. (M/F sex binary, I’m looking at you.
) At the end of March in 2020, I noticed a Reddit discussion of strategies employed by nonbinary folks to make themselves fit. One Redditor, an agender nonbinary person married to the same, wanted to be counted. So, they compromised, choosing “Female” as the sex for both and then checking the box for a same-sex couple.
“So I guess what I’m saying is now we’re political lesbians,” they wrote. “Census lesbians. Censusbians.
” In the end, the diversity of desires will always overwhelm the allowances of bureaucracies. We need a more capacious census (and right now, as the bureau seeks public comment on the 2030 census, we have a chance for change), but even as we do better at stripping away straight norms, creativity and compromise will remain paramount in order for everyone to count. Excerpted from Democracy’s Data: The Hidden Stories in the US Census and How to Read Them, by Dan Bouk.
Published by MCD an imprint of Farrar Straus and Giroux. Copyright © 2022 by Dan Bouk. All rights reserved.
.
From: wired
URL: https://www.wired.com/story/us-census-queerness-data/
