Making Everyone Count: An Examination of the Global Census Undercount Issues of Indigenous Populations
Written by Maggie Smith and Alyson Marks
2020 was supposed to be a critical year for census enumeration in many countries around the world. Then came the COVID-19 pandemic, which has resulted in disruptions to censuses in more than 60 countries worldwide and prompted many nations, including Brazil, Ecuador, and the Seychelles to postpone their census to 2021. However, in 2021, conducting a safe and accurate census is likely to still pose a significant challenge for many countries. This issue is particularly concerning for marginalized groups, including indigenous populations, who are already systemically undercounted.
From dictating the apportionment of federal funding for essential health, education, and other basic services to determining representation in government, having an accurate census is vital for governments to create effective and sustainable policies that benefit all citizens. As we look to this New Year, and as and when censuses do take place, countries with indigenous populations need to pay special attention to ensuring that these groups are fully counted in the census and how their census counts vary along ethnic lines.
Undercounts and Why They Matter
Census undercounts refer to the number of individuals within a particular group (often reported in percentages) that were omitted in the data, and indigenous populations are often the most significantly impacted. For example, in 2016, the Australian Bureau of Statistics found that 137,750 Aboriginal or Torres Strait Islander people had not been counted in their census – nearly 18% of the total population. In the 2010 U.S. Census, indigenous people living on reservations were undercounted by 4.9%, the largest undercount of any ethnic group in the country. It’s estimated that one in seven Native Americans living on reservations were left out of the census. Additionally, a study found that the indigenous population in Toronto, Canada is four times higher than officially recorded in the census.
These undercounts have direct repercussions on indigenous communities. Over one trillion dollars of federal funding in the U.S. is distributed based on census data, with 5.6 billion allocated to Native American tribes for services, such as healthcare and improving reservation infrastructure. In New Mexico, a state with one of the highest indigenous populations, it’s estimated that $5,000 USD is lost each year for every person not counted. And a majority of American tribes surveyed by the National Congress of American Indians (NCAI) also use census data for internal matters, such as grant reporting and establishing tribal priorities, which compounds these inaccuracies. Furthermore, census undercounts can impact a country’s other data sources, such as administrative data. In Canada, many of its national surveys, such as the Canadian Community Health Survey and the National Longitudinal Child and Youth Survey, use census sampling frameworks.
How Undercounts of Indigenous Populations Occur
Some of the current issues with the measurement of indigenous populations have to do with the census survey itself, particularly in the U.S. In the U.S., indigenous respondents have found that their names don’t fit in the allotted box, or that their reservation homes don’t have the traditional type of address that is accepted on the census. Additionally, the only Native American language the U.S. census is translated to is Navajo, despite the existence of many other indigenous languages spoken throughout the country. Census respondents are also asked to self-identify as indigenous as part of the question pertaining to race, which has been largely criticized by indigenous data activists because it disregards the sovereign political status of Native Americans. Also, if an individual selects that they are an American Indian or Alaska Native, they are then asked to add the name of their “principal tribe.” However, the census doesn’t provide a definition of principal tribe, and many Native Americans may live outside of tribal lands, which can result in inaccurate reporting. Similar issues also occurred in Australia, where nearly one-third of respondents didn’t answer the census question pertaining to indigenous identity. And in Kenya, for decades, a majority of indigenous communities were grouped with major ethnic groups on the census, depriving smaller tribes of resources and government recognition.
Census accessibility issues can also cause undercounts in indigenous populations. For instance, nearly 79% and 65% of Native Americans living in New Mexico and Alaska respectively reside in what are deemed “hard to count areas.” The COVID-19 pandemic has further exacerbated these issues. In the U.S., reservations closed their borders to non-residents, and many census bureaus and national statistics offices, have paused in-person enumeration efforts, which are often the only way to reach indigenous populations who often lack internet and mobile phone access.
Lastly, a lack of trust in the government is also a contributing factor. Many indigenous communities are often skeptical of revealing personal information to the federal government after years of policies that have caused them serious harm. For example, according to an NCAI study, 81% of U.S. tribes were concerned about how the federal government uses census data. This lack of trust likely results in poor census response rates. There are also issues of discrimination, as in Latin America, where the undercount is believed to be due to fear of discrimination for identifying as indigenous.
Warning Signs and the Road Ahead
Despite increased efforts by the U.S. government to remedy indigenous undercounts, we are already seeing some alarming 2020 U.S. census quality issues, many of which are likely a result of COVID-19. While the U.S. self-response rate was 67%, no tribal lands had a response rate higher than 56%. Moreover, the response rate of the largest reservation in the country, Navajo Nation, was a mere 21%. With high rates of poverty on many reservations and a need for increased government funding for basic services, these census issues are particularly concerning.
However, there is evidence that some countries are paying more attention to these issues and taking steps to remedy them. This year, Kenya added questions to their census to account for and record a wider range of tribal minorities, an important step forward to improve the livelihoods of smaller tribes that had been previously ignored or grouped with others in the census. Additionally, to remedy their poor indigenous response rates and trust issues, the Australian Bureau of Statistics has committed to hiring enumerators from indigenous communities and released targeted messaging for indigenous populations to encourage them to participate in the census. Activists in New Zealand have also formed an organization called Te Mana Raraunga, or the Maori Data Sovereignty Network, that has pushed for increased indigenous governance on data collected about the Maori population.
These examples all point to things that need to change to garner increased census participation from indigenous communities, including:
Greater collaboration between indigenous communities and census bureaus and national statistical offices. This will help to foster greater trust and knowledge sharing on indigenous peoples' needs to better design census surveys and enumeration efforts.
Increased efforts to hire indigenous community members to administer the census.
Additional funding for census outreach efforts in indigenous communities. This is especially needed in U.S. states with high indigenous populations, such as Oklahoma, North Dakota, and South Dakota, which lack a budget for this.
All countries should undertake post-census validation work to more accurately measure potential undercounts by ethnic group.
Countries should consider alternative and/or complementary ways to survey indigenous communities to accompany the census and improve the accuracy of the data. For instance, to help estimate the indigenous Sami population in Norway, the country uses voters’ lists from the Sami parliament.
Having accurate census data is essential for a myriad of critical government functions, but it is impossible to achieve without placing a special focus on the groups most likely to be undercounted. To ensure that everyone is counted as we move into 2021, we must remedy these acute gaps in census data collection.