Information presented in Minnesota Compass's Cultural Community Profiles comes from Minnesota Compass analysis of the Integrated Public Use Microdata Series (IPUMS):
Steven Ruggles, Sarah Flood, Ronald Goeken, Josiah Grover, Erin Meyer, Jose Pacas, and Matthew Sobek. IPUMS USA: Version 9.0 [dataset]. Minneapolis, MN: IPUMS, 2019.
IPUMS collects, preserves, and harmonizes data from the U.S. Census Bureau's American Community Survey (ACS), which includes questions about race, ethnicity, ancestry, and place of birth. Minnesota Compass uses responses to these questions to categorize respondents into one or more cultural communities.
The ACS estimates we assemble are based on data collected over five years and therefore describe average characteristics for that five-year time period. Compass created profiles for cultural communities with at least 175 unweighted records available to analyze.
DEFINING CULTURAL COMMUNITIES
Minnesota Compass takes responses from a combination of questions and categorizes respondents into cultural communities. We count a respondent as a member of a cultural community if that person meets any of the criteria listed below. Individuals are categorized into all cultural communities to which they may belong. People who identify with multiple cultural groups—Vietnamese and Chinese, for example, or Mexican and German—are included in both data profiles.
The ACS asks respondents to identify their race: "What is Person 1's race? Mark (X) one or more boxes." Response categories include:
- Black or African American
- American Indian or Alaska Native → Print name of enrolled or principal tribe.
- Asian Indian
- Other Asian → Print race, for example, Hmong, Laotian, Thai, Pakistani, Cambodian, and so on.
- Native Hawaiian
- Guamanian or Chamorro
- Other Pacific Islander → Print race, for example, Fijian, Tongan, and so on.
- Some other race → Print race.
Based on these responses, we identified cultural communities for three types of respondents: White (non-Hispanic), African American, and Asians.
Asian respondents have the opportunity to select from a handful of checkbox categories—Chinese, Asian Indian, Vietnamese, among others—or to write in an Asian race that is not listed. We use the responses to these questions as an Asian respondent's community.
See below for more detail on the white (non-Hispanic) and African American profiles.
The ACS asks respondents if they are Hispanic, Latino, or of Spanish origin: "Is Person 1 of Hispanic,
Latino, or Spanish origin?" Response categories include:
- No, not of Hispanic, Latino, or Spanish origin
- Yes, Mexican, Mexican American, or Chicano
- Yes, Puerto Rican
- Yes, Cuban
- Yes, another Hispanic, Latino, or Spanish origin → Print origin, for example, Argentinean, Colombian, Dominican, Nicaraguan, Salvadoran, Spaniard, and so on.
Any person answering "yes" and listing a cultural group under the Hispanic origin question is included in that cultural community's profile.
The ACS asks about respondents' ancestry or ethnic origins: "What is this person's ancestry or ethnic origin" Survey respondents can list up to two responses.
Any person who lists a cultural group in this question is included in that cultural community's profile. This is the first opportunity that some people have to indicate their background. For example, a Somali respondent has no opportunity under the ACS question about race to indicate that they are Somali (they can select Black/African American or another race, but cannot indicate Somali ancestry). Compass would include that respondent in the Somali profile based on this question. For Asian respondents, their response to the ancestry question may or may not match the detailed race that they listed, but they will be
included in a profile if they list that cultural community in either location.1
The ACS asks respondents to indicate their birthplace: "Where was this person born?" Response categories include:
- In the United States → Print name of state.
- Outside of the United States → Print name of foreign country, or Puerto Rico, Guam, etc.
Those who select "Outside of the United States" are asked to identify the country of their birth. If respondents are born in another country, we count them in that country's profile unless they indicate some other background in the race or ancestry questions. For example, respondents born in Ethiopia who identify their ancestry as Somali are counted as Somali and not as Ethiopian.
We count minor children living with their parents as belonging to any cultural communities that their parents belong to, unless the parent indicates a different, conflicting cultural group for the child in the race or ancestry questions.
We construct a white (non-Hispanic) cultural community profile by counting people who select white for race. We exclude respondents who indicate that they are also another race. We further exclude respondents who indicate that they are of Hispanic, Latino, or Spanish origins.
We chose not to disaggregate white (non-Hispanic) by ancestral origins (e.g., Norwegian, German) because of the similarity of the data, except for population counts, which we provide for the largest groups.
We construct a Native American cultural community profile by counting people who select American Indian or Alaska Native (AI/AN) alone or with another race. We also count individuals who list AI/AN or a related identity (such as "Indian") as their ancestry. Additionally, we include the minor children of parents who list AI/AN race or ancestry. We cannot disaggregate AI/AN by tribe because of data suppression.
When respondents indicate that they are AI/AN and another race, the U.S. Census Bureau suppresses their tribal identification.2
We construct an African American cultural community profile by counting people who select Black or African American for race and are born in the Unites States. We include respondents who indicate that they are of Hispanic, Latino, or Spanish origins. We also include respondents who select other races in addition to Black or African American. We exclude respondents whose parents are foreign born if the respondent lives in the same household as their parent. We further exclude respondents who indicate Somali or Ethiopian ancestry.
OVERCOUNT AND UNDERCOUNTS
Estimates in the profiles are based on information collected from a sample of the total population. Relying on a sample to describe the population (i.e., an entire cultural community) introduces error. Our estimates would likely vary if the same survey were conducted with a different sample of the population.
Furthermore, the profiles rely on self-reporting in the ACS. Respondents to the survey may not indicate their cultural community for any number of reasons, including fear of persecution. This may result in an undercount. Respondents to the ACS may also indicate a cultural community that the Census Bureau does not recognize. In such cases, the Census Bureau recodes the unrecognized cultural communities to a different cultural community, one with which the respondent may not associate. This could result in an overestimation of the recoded cultural community.
All profiles include margins of error (+/-), which give us a measure of statistical uncertainty. Adding and subtracting the margin of error from an estimate gives us a range, and we have a certain level of confidence that the true population value falls within that range. In the profiles, margins of error are provided for a 90% confidence level. We suppress data when the margins of error equal 70% or more of the estimate.
DIFFERENCES WITH OTHER EFFORTS TO CREATE CULTURAL PROFILES
The Minnesota State Demographic Center (MDC) released a report, The Economic Status of Minnesotans 2018, that presents information for 17 cultural groups in Minnesota. There are a handful of differences between how the MDC approached its report and how Minnesota Compass created cultural profiles.
- MDC presents margins of error for a 95% confidence interval for its estimates; Minnesota Compass uses a 90% confidence interval.
- In developing group estimates, MDC includes cultural groups where there are 300 or more individual pooled responses to the ACS between 2012 and 2016; Minnesota Compass developed profiles with a threshold of 175 individual pooled responses between 2013 and 2017.
- MDC developed estimates for Dakota and Ojibwe tribal members. We do not disaggregate AI/AN by tribe because of data suppression (See above "Native American cultural community profile" note and corresponding footnote for more information on data suppression).
- The MDC estimates for African American exclude people who indicated a Liberian ancestral heritage; the Minnesota Compass African American profile includes Liberian.
- The groups in the MDC report account for 94% of Minnesotans; the Minnesota Compass profiles account for 97% of Minnesotans.
1 For both the race and ancestry questions, the Census Bureau recodes some open-ended responses into its pre-existing categories. Respondents indicating that they are Karen or Oromo, for example, are recoded by the Census into Burmese and Ethiopian, respectively, with nothing to indicate their original response. In those cases, the individuals are included in the cultural groups corresponding to their Census Bureau reclassification.
2 In the American Community Survey, and in tables that the Census Bureau publishes based on the ACS, tribe is reported for all respondents who indicate that they are American Indian or Alaskan Native, regardless of whether they also identify with another race. However, in the microdata files on which IPUMS is based, tribe is suppressed for respondents who are American Indian and another race, which results in a substantial undercount of Minnesotans identifying with specific tribes. In 2013-2017, Compass estimates that 33% of Ojibwe people and 45% of Dakota people in Minnesota cannot be properly identified with their tribe in the IPUMS microdata.