Paper Genocide: The Erasure of Native People in Census Counts

Native people were excluded from the first 70 years under the Constitution, which regarded “Indians not taxed” as not countable.

[Photo: A watercolor painting of a native person. In the background is a night sky.]
Inaccurate counts of blood quantum have had an enormous effect on Native individuals and nations. Marcy Angeles for Rewire News Group

This is the first article in a two-part series. Read the second article in the series here.

The Trump administration’s fear-mongering tactics against immigrant communities surrounding the 2020 census may have been shocking to some, but it’s not the first time the government has used the census as a tool of oppression. Native people, in particular, are the most undercounted ethnic group in the census’ history. Native people were excluded from the first 70 years under the U.S. Constitution, which explicitly regarded “Indians not taxed,” or those living on reservations or unsettled territories, as not countable. In more recent years, the U.S. Census Bureau’s own data has shown significant undercounting. In the 1990 census, 12.2 percent of Native people on reservations were undercounted, according to the Census Bureau’s findings. A decade later, the census seemed to improve, with the bureau not reporting a statistically significant undercount. But then in 2010, it jumped back up to 4.9 percent.

This is particularly devastating for Indigenous people because of how census data has been used to help determine many aspects of tribal sovereignty, such as tribal recognition and enrollment.

Judy Shapiro, an attorney practicing Indian law for 34 years, told me that the federal government uses census data as part of the mechanism for “gatekeeping” for federal recognition. “Through the federal recognition process, they determine who is Native, who continues to exist, and who they are responsible for [maintaining trusts and treaties].”

The erasure of Native people from the U.S. census and tribal rolls amounts to paper genocide, a systematic destruction of Native identity by reclassifying people into non-Native racial groups on government records. By erasing our existence via the census, our treaty rights are further ignored and funding streams for our tribal nations are gutted. What’s worse, the details surrounding the 2020 census suggest there is no real change in sight.

A brief history of Native erasure in the U.S. census

The census is written into the U.S. Constitution and must occur every ten years. The first census was taken in 1790, after the inauguration of President Washington and before the second session of the first Congress ended.

Today, the Census Bureau typically collects data by sending self-reporting forms to households across the country. Any household that doesn’t return a form is then visited or called by a census enumerator. This process has changed over the years, and many Native households fall through the cracks and don’t get counted. “American Indian and Alaska Natives” are designated by the Census Bureau as a hard-to-count population due to issues including non-traditional addresses, high rates of renters and houselessness, and difficulties accessing more rural lands.

Nonetheless, census data is used in a variety of ways. It determines everything from how many congressional representatives communities have, to the apportioning of federal funds for community needs, to the enforcement of the Voting Rights Act of 1965, and much more. For Native people, the census is crucial to their state and federal recognition, the enforcement of treaties, and the economic health and well-being of tribal governments and its citizens. And yet, since the census began, Native people have not been counted accurately.

Native people didn’t appear in the census until 1860. Before that, they were not identified in the 1790-1850 censuses as a result of discrimination in the U.S. Constitution. One exception was the 1850 census of the New Mexico Territory, which listed Pueblo people as “Copper.” (They were designated as “Indian” in the 1860 and 1870 censuses of the territory.) Enumerators began counting Native people living in the general population, meaning those living among white and free people of color, for the first time in 1860, when, according to the National Archives, they received instructions to count the “families of Indians who have renounced tribal rule, and who under state or territory laws exercise the rights of citizens.”

From 1790 to 1870, the U.S. Marshals were responsible for collecting census data. They were replaced in 1880 by specially appointed census supervisors, who were required by an act of Congress to hire enumerators to take the census count.

Up to the 1950 census, how a person’s race was classified was often based entirely upon the perception of the enumerator. “If a different census taker comes to the door and they leave their paper bag at home, then maybe they’re Indian that year,” Shapiro said, referring to the idea that a person’s race could be determined based on whether their skin color was darker than a paper bag.

According to Shapiro, Native people often went underground to evade persecution. They would sometimes hide when census takers would come, Shapiro explained, changing their appearance, making strategic marriages, and employing other strategies to hide their indigeneity in the hopes they wouldn’t be exiled. This affected the accuracy and quality of the counts.

Rose Buchanan of the National Archives told me that the 1890 census marked the first time Native people living both among the general population and in tribal communities were counted. However, nearly all of the 1890 census schedules were destroyed in a fire at the Department of Commerce in 1921. As a result, the 1900 census is typically referred to as the first inclusive count of Natives in the United States.

Tribal rolls, also known as the Indian census rolls, were the primary means for counting Native people until 1940. However, the tribal rolls counted only groups of Native people, whereas the U.S. census collected data of all U.S. citizens. Between 1885 and 1940, officials took 692 rolls on Native nations. The Act of July 4, 1884 required every Bureau of Indian Affairs (BIA) agent to submit an annual “census on the Indians at his agency or upon the reservation under his personal charge.” There were no instructions, though, on how to count Native people, so it stands to reason that some of this data is inaccurate.

Whether the data was accurate seems beside the point for the U.S. government. As the Leadership Conference’s “Race and Ethnicity in the 2020 Census” report explains, the U.S. government began collecting such data at the end of the 19th century in order to “carry out policies designed to maintain the white majority’s influence and power in the political, social, and economic arenas.” Indeed, many Native people weren’t citizens of the United States until the 1924 Indian Citizenship Act and therefore had no right to vote in U.S. elections. Some Native people, though, were granted citizenship under the Allotment Act of 1887, also known as the Dawes Act.

Measuring “Indian blood”

Beyond the paper bag test, government officials have used blood quantum for racial identification since the turn of the 20th century. And just as the government used the one drop rule to discriminate against African people and keep them enslaved, it used blood quantum to harm Indigenous communities.

In theory, blood quantum measures the amount of “Indian blood” a Native person possesses, which is then captured on a Certificate of Degree of Indian Blood issued by the BIA. Officials use the following federal government records to measure blood quantum: census rolls between 1885 and 1940, the 1900 special Indian census, the Dawes Rolls, Durant Rolls, and land conveyances involving Native people. During this period, sexual violence became a common form of genocide against Native people, which some elders have attributed to an effort to lower the blood quantum of future generations.

There are only three types of living beings in the United States that have to register their blood quantum with the U.S. government: dogs, horses, and Native people.

The census of 1880 was the first to include “a complex rubric for racial identification” for Native people, according to a paper by University of Colorado librarian Margaret Jobe. “The enumerators needed to distinguish between full-blood tribal members and individuals of mixed racial or tribal origin.”

According to Jobe’s paper, the superintendent of the 1870 census, Francis A. Walker, considered three options for counting mixed Natives:

to assign race based on the “condition” of the father or the mother; to assign race based on the “superior or inferior blood”; or to assign race by “the habits, tastes, and associations of the half-breed”. Walker believed that the latter was the most logical and least cumbersome choice …. The first alternative—the “condition” of the father or the mother was particularly problematic. During the era of slavery, if the condition of the mother was “slave,” then the child was considered a slave.

Jobe’s paper further explains, “With the passage in 1887 of the General Allotment (Dawes) Act, the United States government institutionalized the distinction between full- and mixed-blood Indians. To receive an allotment, Indians had to become enrolled members of their respective tribes. To enroll in a tribe, an individual needed to prove a certain degree (purity) of Indian blood.”

The instructions for the 1900 U.S. census directed enumerators to count blood quantum by writing “0” if the Indian has no white blood. If they had white blood, they were assigned a fraction that was nearest to their blood quantum, such as 1/2 or 1/4. The 1930 Indian census rolls only allowed three categories of blood quantum, including “full blood, 1/4 for one-fourth or more Indian blood, and -1/4 for less than one-fourth.”

Shapiro noted that the enumerators for the 1930 U.S. census were instructed to count people of mixed white and Indian blood as Indian unless their blood quantum was “very small” or if they were  “regarded as a white person in the community where he lives.” Instructions were also given to count “a person of mixed Indian and Negro blood” as Negro unless the “Indian blood predominates and the status of an Indian is generally accepted in the community.”

By institutionalizing blood quantum, the federal government put indigeneity, and that of future generations, in the hands of white agents of the state who were determined to exterminate Native people. As I’ve written previously, “It is not our blood quantum or even our tribal citizenry that makes us who we are; it is our connection to our ancestors, people, and the seven generations to come.”

Inaccurate counts of blood quantum have had an enormous effect on Native individuals and nations. For example, one of my family members is listed as full blood in one roll and one-half in another. If children were members of more than one tribe, officials would count them on the roll of only one tribe as opposed to every roll taken for all of their tribal nations.

My blood quantum is registered with the BIA as one-eighth. While the Cherokee Nation doesn’t use blood quantum to determine tribal citizenship, some tribal nations do. This incorrect data has a direct impact on the ability of future generations to gain tribal citizenship and be entitled to their treaty rights.

The passage of the Indian Citizenship Act in 1924 created new issues for counting Indigenous people. As Jobe’s paper explains, “The Census Bureau was concerned that Mexican laborers might attempt to pass themselves as Indians in the states that share a border with Mexico. To get an accurate count of the Indian population, the bureau instructed enumerators to take special care to differentiate between the two groups in the states of California, Arizona, and New Mexico.” To this day, Indigenous people from what is now known as Mexico and Central and South America aren’t counted as Indigenous to those lands. They can identify on the census as American Indian or Alaskan Native, but are often counted as Hispanic or Latino.

The census did not begin using self-reporting forms until 1960, allowing Native people to finally record their race themselves. However, it wasn’t until the census of 2000 that multiracial individuals could self-identify with more than one race. Many individuals of mixed heritage, who might have identified as white, Black, or Asian in earlier years, were then able to identify themselves as Indian as well.

The impact of undercounting

Since the first iterations of the census and tribal rolls, the U.S. government has used this data to cheat Native peoples out of allotments of land, distribution of goods and money, and treaties it has failed to honor.

The eligibility for treaty-related services, such as Indian Health Services care, is equated with tribal enrollment status even when accurate historical tribal enrollment data doesn’t exist. Whether or not a Native person was eligible for goods and services was tied to these allotment lists, tribal rolls, and prior censuses under U.S. Indian law. If the Census Bureau hadn’t counted a person on any of these lists, then the government did not consider the person as Native. This was especially troublesome when, as in 1930, the Commissioner of Indian Affairs sent directives to Indian agents to lose the “names of Indians whose whereabouts have been unknown for a considerable number of years.” Their whereabouts could have been unknown for any number of reasons: they were hiding from government officials, were away hunting or fishing when the counts occurred, or they were incarcerated or had been murdered.

The U.S. census and tribal rolls were also a means of assimilating Native people to European or Western traditions. As Jobe’s paper notes, “The Census of 1880 introduced a special enumeration schedule for the Indian Division that could be used to measure the degree to which an Indian had adopted a European way of life. For example, it asks if a person was a chief or war chief, wore citizen’s dress, was supported by civilized industries in whole or part, or was supported by hunting, fishing, or gathering.” This effectively measured how “civilized” and assimilated Native people had become to the white man’s ways as opposed to living their traditional ways of life.

Racist enumerators and agents also forced Indigenous people to assume Christian names, or inaccurately spelled Native names. A directive given in 1902 to agents taking Indian census rolls suggested how to translate “Indian names to English.” Indian names were instructed to be recorded as such unless they were “too difficult to pronounce or remember.” Animal-related names could be translated to English, though agents should avoid “foolish, cumbersome, or uncouth translations which would handicap a self-respecting person.” My last name, Deerinwater, is spelled in multiple ways over several rolls of my people. I’ve been left to wonder what my ancestor’s name originally was before U.S. government officials got their hands on it.

And these counting systems counted men as the heads of households. Under the 1902 directive, officials assigned women and children the surname of their husbands and fathers even though this was not the way many nations and clans traditionally assigned names.

Census data today

The inaccurate and often racist and colonizing U.S. census, as well as historical tribal rolls, are still used to this day to determine important matters, such as federal recognition. Federal recognition is essentially the U.S. government making a decision on who is a legitimate Native nation and as a result, who is entitled to treaty rights, trust land, and many other goods and services. Shapiro, who has worked with tribal governments on this, told me that this is a “terrible, difficult process.”

For example, census data has played a key role in the recognition process for the Mashpee Wampanoag of Massachusetts. In the 1970s, the Mashpee filed a land claim that took them to trial. The standard imposed in the case required the Mashpee to show they were a “unitary Native people,” Shapiro said, but the census data was inconsistent. They were then challenged as to whether they were culturally Native based on standards of distant tribes. The jury ultimately decided that the Mashpee weren’t really Native.

In 2007 the Mashpee finally gained federal recognition. In such recognition cases, the courts had to consider how long the tribe has been in a “continual community and political existence,” said Shapiro. This could become an issue for tribes that were displaced due to land loss.

This wasn’t the end of the Mashpee’s problems. In 2015, they gained land in trust from the U.S. government. Having trust land meant that the Mashpee could build a casino, generating a crucial source of revenue for the tribe. But as the Mashpee began developing their casino project, a group of plaintiffs partially funded by a white developer who wanted to build a competing casino challenged the trust land decision in federal district court.  In 2018, the U.S. Department of Interior determined the Mashpee do not qualify for land in federal trust, putting the Mashpee’s reservation, historic homeland, and casino project at risk. The tribal government is currently appealing this decision, and a bill to restore the Mashpee land’s federal trust status passed the U.S. House of Representatives in May.

Gaining trust land can be incredibly powerful for tribal nations, and not only because it means that the land goes off the tax rolls. It establishes tribal jurisdiction and limits the power of state governments to control Native people’s economic development. When a tribe has land in trust, they can do economic development projects, including gaming, which help tribal nations prosper and gain economic and political power. “ A number of tribes have done this, but a small number of them have succeeded to the point that they wield powerful voices in the capital and state houses,” Shapiro said.

“In a lot of places people only like tribes when they’re poor, invisible, or quaint. Assembling a land base, particularly for a tribe that’s been hiding for years, is significant,” Shapiro added.

As such, the United States often works against the well-being of tribes.

Take U.S. Supreme Court Justice Clarence Thomas’ opinion in the 2009 Carcieri v. Salazar decision holding that if tribes weren’t “under federal jurisdiction” in 1934, when the Indian Reorganization Act was passed, then they can’t hold land in trust. This affects tribes that were not federally recognized before 1934, often because the government used the existence of intermarriage and assimilation to deny their status as Indian nations. This history is now being used against them, particularly for tribes mixed with Black people.

Despite Native erasure through the U.S. census and tribal rolls, these forms of data collection can still provide a historical account of Native people. Shapiro has poured through old censuses and says “it’s an amazing resource” in terms of uncovering history. “It shows who was stubborn enough to stay. You can see people going away, but coming back. They could be off whaling one year, but they were back the next. You can see a sweep of stubborn remaining history.”

A correct recording of our people is a form of historical preservation and is necessary for the well-being of future generations of Native people. Just as we shouldn’t repeat the errors and atrocities of the past, the 2020 census must not repeat its genocidal mistakes. But as I explore in part two of this series, the federal government hasn’t yet learned from its history.

Author’s note: For the purposes of this article “Native” is used to mean American Indian and Alaskan Natives, and Indigenous denotes the Indigenous people of the land who do not fall under the categories of American Indian and Alaskan Native, such as Native Hawaiians.