The upcoming 2020 Census Detailed Demographic and Housing Characteristics File A (Detailed DHC-A) greatly expands what we know about the racial and ethnic composition of the U.S. population.

While the 2020 Census has already revealed much about the overall Hispanic population and the major race groups (White, Black or African American, Asian and so on), the Detailed DHC-A provides data on detailed groups – such as German, Lebanese, Jamaican, Chinese, Native Hawaiian and Mexican – and on American Indian and Alaska Native (AIAN) tribes and villages like the Navajo Nation.

In total, we will release 2020 Census population counts for about 1,500 detailed race and ethnicity groups and AIAN tribes and villages. Depending on the size of the group, we’ll also provide sex-by-age statistics (the number of males and females within certain age categories). This is our most detailed racial and ethnic accounting of who we were as a nation in 2020.

Leading up to the 2020 Census, we conducted research and engaged with various communities and tribal nations to learn more about the ways in which people view and report their detailed racial, ethnic or tribal identities. This helped us improve the way that we collect, code and tabulate detailed data, providing a more vibrant portrait of the United States.

This blog describes what you should know about the Detailed DHC-A release on September 21.

Larger groups will receive tables with the age categories further split apart. For example, a county with 8,000 people who identified as Japanese alone would receive a table with nine age categories in a sex-by-age table and a county with 20,000 or more would receive a table with 23 age categories.

The margins of error (MOE) in table 1 also show how much disclosure avoidance-related uncertainty will be in the data about 95% of the time.

• For instance, a county with a Japanese alone population of 850 would have an MOE of ±11. This means we expect the published count of the total Japanese alone population in that county to be within 11 people of the enumerated count 95 percent of the time if this disclosure avoidance process was repeated many times.
• For groups with sex-by-age statistics, each age category published would have the MOE listed in table 1. (For example, if a group at the county-level receives four age categories, the count for each category would have an MOE of ±11.) The sex-by-age data then add up to the group’s total population count, and we use a formula to calculate its MOE. The result is the MOE for the total population count is larger for these larger groups. (For example, if each of the four age categories has an MOE of ±11, the total population count for the group would have an MOE of ±31.1).

Detailed information on MOEs for sex-by-age counts will be available in the “Data Accuracy and Margins of Error” section of the technical documentation (coming soon).

Using this type of adaptivity allows us to produce more data for detailed groups where the populations live and to provide accurate data while protecting respondents’ privacy and the confidentiality of the data.

For the 2020 Census, we also made two improvements that allowed us to provide population counts in more localities than ever before.

• We eliminated the minimum population counts required for a group to receive a population count at the national and state levels. This means values of 0, 1, 2, 3… 99 can appear in national and state totals. In contrast, for the 2010 Census, a group had to have at least 100 people to receive a population count, even at the national and state levels.
• We reduced the minimum population count required for a group to receive a population count at the county, place, census tract and AIANNH area levels from 100 in 2010 to 22 for detailed groups and 94 for regional groups (defined below) in 2020. 

Data will be available for the race alone and race alone or in any combination populations.

• The race alone population represents the minimum number of people who identified as that group. It includes those who reported only one response, such as only Fijian.
• The race alone or in any combination population represents the maximum number of people who identified as that group. It includes those who reported only one response, such as Fijian, and those who reported multiple responses, such as Fijian and Japanese or Fijian and Black or African American.

The concepts of alone and alone or in any combination apply only to responses to the race question. Following the U.S. Office of Management and Budget’s 1997 Standards for Maintaining, Collecting, and Presenting Federal Data on Race and Ethnicity, only one response for Hispanic or Latino origin was tabulated.

As discussed in a previous blog, we made several enhancements to our 2020 Census Hispanic Origin and Race Code List that allowed us to more accurately collect, process and tabulate data. We made updates based on feedback that stakeholders, advisors and tribal leaders provided to us and that fit within the OMB guidelines.

Below are some examples of the improvements we made to more accurately code and process data for specific detailed race and ethnicity groups and AIAN tribes and villages. Some of these improvements include differences in how groups were coded and tabulated in 2010 and 2020. 

• Based on research and consultation with experts and stakeholders, we added a new code for Afro-Latino and reclassified Garifuna from the American Indian and Alaska Native racial category to the Hispanic or Latino category. This is the first time these responses will be tabulated from the Hispanic origin question.

• In 2020, we expanded our code list to include additional detailed White (including Middle Eastern and North African) and Black or African American groups, as the race question elicited the collection of detailed White and Black or African American responses through dedicated write-in areas for the first time.
• Data will be available in the Detailed DHC-A for detailed White groups such as German, Lebanese and Cajun and detailed Black or African American groups such as Kenyan, Haitian and Jamaican.
• Data for detailed White and Black or African American groups traditionally have been tabulated from the question on ancestry in the American Community Survey, and this is the first time these data will be tabulated from the race question in the census. 

• Leading up to the 2020 Census, we conducted formal consultations with tribal leaders to better understand how their citizens identify so we could more accurately code tribal responses.
• Based on the consultations, data for individual American Indian and Alaska Native tribes are included in this product instead of tribal groupings.
• We updated the code list to account for any changes to how tribes are formally recognized. For example, in 2010, responses for “Assiniboine Sioux,” “Fort Peck Sioux,” “Fort Peck Assiniboine,” and “Fort Peck Assiniboine and Sioux Tribes of the Fort Peck Indian Reservation” were coded and tabulated as four separate tribes; in 2020 they were coded and tabulated together as “Assiniboine and Sioux Tribes of The Fort Peck Indian Reservation, Montana.”

• We reclassified several Central Asian groups (i.e., Afghan, Kazakh, Kyrgyz, Tajik, Turkmen and Uzbek) from the White racial category to the Asian racial category based on research and consultation with experts and stakeholders.
• After extensive engagement with the Sikh community, “Sikh” was included as a distinct detailed population group within the “Asian” racial category instead of being tabulated as “Asian Indian” like it was in the 2010 Census.
• New codes were added for Kazakh, Kyrgyz, Tajik, Turkmen, Uzbek, Sikh, Sindhi, Bruneian, Mien, Buryat, Kalmyk, Kuki, Lahu, Malay, Mizo, Pashtun, Tai Dam and Timorese, and data for these groups will be available for the first time. 

• New codes were added for New Caledonian, Nauruan, Cook Islander, Easter Islander, French Polynesian, Maori, Niuean, Rotuman, Tuvaluan, and Wallisian and Futunan, and data for these groups will be available for the first time.  
• Because we updated the checkbox for “Guamanian or Chamorro” to “Chamorro,” we’re now able to provide data for the Guamanian and the Chamorro populations. 

• Although we had codes for groups in the Some Other Race population in 2010, the 2020 Census will mark the first time we publish data for detailed groups, such as Brazilian, within this code range.

We’ve developed a crosswalk for data users interested in learning more about the differences between how we tabulated 2010 Census data and 2020 Census data for detailed race and ethnicity groups and AIAN tribes and villages. The Detailed Race and Ethnicity Crosswalk: 2010 to 2020 shows which codes were used to tabulate each group in 2010 and 2020.

Data comparisons between 2020 Census and 2010 Census detailed race data should be made with caution and take into account improvements we made to the question and the ways we code what people tell us. However, the detailed Hispanic origin data from the ethnicity question are comparable between the two censuses.

We’d like to note that we’ve previously estimated that we would publish data for approximately 370 detailed race and ethnicity groups and about 1,200 detailed AIAN tribes and villages. We estimated these were the maximum number of groups that could be eligible for data based on our code list. Now that we’ve finalized our iterations list used to tabulate the data, we know that the number of groups will be a little lower. This is because the code list includes every term that receives a unique code during processing, but not all unique codes receive their own tabulations in the Detailed DHC-A.

For example, Okinawan, Iwo Jiman, and Japanese all receive unique codes during our data processing. However, in the 2020 Census Hispanic Origin and Race Iterations List, Okinawan, Iwo Jiman and Japanese are all included in Japanese. Based on the final iterations list, 300 detailed race and ethnicity groups and 1,187 AIAN tribes and villages will be eligible for tabulation.

Leading up to the 2020 Census, we engaged with data users, stakeholders, researchers, tribal leaders and advisors who helped us make our code list one that more accurately represents our diverse nation and produces data fit for a variety of uses. The upcoming Detailed DHC-A will provide a rich source of information about myriad race and ethnicity groups and tribal nations.

More information about the upcoming release is available in the Detailed DHC-A press kit.