New proposed forms for the 2020 census would decrease breakouts of various subgroups within the Asian American, Native Hawaiian and other Pacific Islander communities, fears two national Asian American community groups (Photo by George Kelly).
Asian Americans Advancing Justice, the National Coalition for Asian Pacific Americans along with researchers Karthick Ramarkishnan of UC Riverside and Leiliani Nishime of the University of Washington have launched a letter writing campaign urging the census to revise its census forms.
In an open letter to the community the groups write:
“The Census is currently making critical decisions about how race will be collected in 2020. We’ve seen the proposed formats, results from experiments of different formats, and have some deep concerns about about the way in which data are being collected. Our primary concern is that some of the proposed formats decrease detailed reporting (ie specification of a particular ethnic or national origin group) among Asian Americans.”
The groups are asking academics, demographers, and national and local community advocates, and individuals to sign onto a letter to be sent Monday to Nicholas Jones, the Chief of Racial Statistics at the US Census.
You can add your name to the letter here.
In the letter to Jones, the groups write “While we appreciate the Census Bureau’s engagement with AANHOPI (Asian American and Native Hawaiian & Other Pacific Islander) advocates, we remain gravely concerned by the most recent testing proposals as it jeopardizes the accurate counting of the diverse AANHOPI community, and is a step backwards from progress made within the 2010 Census.
You can read the entire letter below:
Chief, Racial Statistics
U.S. Census Bureau
4600 Silver Hill Road
Suitland, MD 20746
Dear Mr. Jones:
We, the undersigned, write today to provide feedback on the Census Bureau’s efforts to assess opportunities to revise the way census surveys ask about race and ethnicity. On May 7, 2014, the U.S. Census Bureau met with researchers from academic institutions and community organizations that have long-standing experience and expertise with Asian American and Native Hawaiian & Other Pacific Islander (hereafter AANHOPI) populations. This meeting follows an earlier call in February between the Census Bureau and AANHOPI communities about concerns with the current testing of the race and ethnicity questions. While we appreciate the Census Bureau’s engagement with AANHOPI advocates, we remain gravely concerned by the most recent testing proposals as it jeopardizes the accurate counting of the diverse AANHOPI community, and is a step backwards from progress made within the 2010 Census. Representing academics, demographers, and national [and local] community advocates, the [#] groups and individuals signed onto this letter propose the following general principles for the 2020 census and more specific recommendations for the 2015 content test currently being planned.
Key Principles for 2020 Census
We offer the following overarching principles the Census Bureau should consider as it moves forward with its efforts on determining the race and ethnicity questions for the 2020 census. We also provide recommendations on how to ensure these principles are being met through this process.
In many of our conversations and in the design proposals for testing variations in the measure of race and ethnicity, the Census Bureau has noted “balance,” “equity,” and “symmetry” as central tenets of this effort. However, this approach overlooks two other important, and fundamental, tenets that are centrally important for Census data collection on the race and ethnicity of U.S. persons:
1) Ensuring that we do not move backwards from the detailed reporting achieved in the 2010 Census, which should serve as the baseline for effective practice and;
2) Ensuring the accuracy of the data collected.
Principle #1: 2010 Census as a Baseline for Effective Practice – We cannot go backwards
AANHOPIs represent our nation’s fastest growing and most diverse racial groups. Often viewed as homogenous, these communities include more than 50 detailed subgroups that can differ dramatically across key social and economic indicators. Through the decennial census, American Community Survey, and other national surveys, the Census Bureau stands as the single most important source of disaggregated data, currently providing data on the size and characteristics of 24 Asian American and 22 NHOPI detailed subgroups. It then becomes imperative than any data collected in future census meet the standard already achieved in 2010. This is a minimum or floor, with the hopes that future efforts can build on this foundation. We cannot go backwards in terms of the quality of detailed data collected on AANHOPI groups. In order to honor this principle, we recommend the following practices:
• A maximum number of check boxes should be included, with the number used during the 2010 Census serving as a minimum – Check boxes capturing detailed race groups improve detailed race reporting and should be utilized for all race groups, regardless of whether or not the race and Hispanic origin questions are combined.;
• A maximum number of examples should be included, with the number used during the 2010 Census serving as a minimum – Examples are critical in soliciting detailed race reporting from detailed race groups not represented by check boxes and should also be utilized;
• NHOPI response options should be clearly identified separate from Asian American response options – Increasing the visibility of NHOPI response options will help promote their detailed race reporting, which was particularly low during the 2010 Census.
Any further testing should:
• Keep (at a minimum) the practice of having 6 separate check boxes for Asian sub-categories and 3 separate check boxes for Pacific Islander sub-categories.
• “Other Asian” and ”Other Pacific Islander” categories should be included, with listings of the same number of examples used in the 2010 Census (at a minimum, 5 for Other Asian American and 2 for Other Pacific Islander).
• Improve upon this baseline, e.g. increasing the number of check boxes and listing of examples, not decreasing these options.
Principle #2: Accuracy
Without accurate data by detailed subgroup, the diversity in the AANHOPI community means some of the most disadvantaged in our communities are rendered invisible to policy makers, leaving their critical needs unmet. In reviewing proposed research panels, we observed that the proposed formats decreased both the number of separate checkboxes and listings of examples for Asian American and Pacific Islander race groups compared to Census 2010 – practices that have been demonstrated by research to be both ineffective and a threat to accuracy. The most recent AQE testing results confirm that:
• Removing check boxes used to capture detailed race groups also decreased the amount of detailed race reporting among Asian Americans. Indeed, these formats yielded the lowest detailed race reporting among Asian Americans of any format tested.
• Removing a subgroup or national origin from the list of examples for each broad race category reduced reporting for that group. Testing conducted as part of the 2005 National Census Test suggests that limiting or removing the list of examples has a negative impact on detailed reporting.
By testing and adopting these design practices, the Census Bureau is introducing new barriers for Asian American populations that did not exist before that will certainly harm the accuracy of the data being collected. Additionally, the manner in which data is collected for both large and small groups must take this accuracy principle into consideration and address the needs of both. Given the historical opportunity that the 2020 Census presents to accurately capture America’s changing racial and ethnic demographics – we cannot waste time or resources in re-testing bad designs and creating new barriers to accuracy in detailed reporting. In addition to the recommendations above regarding check boxes and examples, we recommend the following to ensure more accurate data on the AANHOPI community:
• Oversampling Asian Americans and NHOPIs by ethnic group in future testing protocols, ensuring an adequate sample of both large and small groups across all panels;
• Providing adequate Asian and Pacific Island language assistance to ensure meaningful response from limited-English proficient Asian Americans and NHOPIs and to ensure results are not biased by English-fluent respondents.
Finally, we appreciated the Census Bureau sharing its thoughts on possible future testing of the race and ethnicity panels at the May 7th meeting. The following represents specific, technical recommendations to the panels shared (which have been attached as an appendix for easy reference):
• Version 1 of the new possible designs for testing (see Page 1 of Appendix)
o The open text boxes can be reduced in width by 50%, and the space can be used to provide examples for write-in categories found in Census 2010
o Version 1 introduces language on the OMB definition of the group that is likely to generate confusion and opposition among many respondents to the questionnaire. As one of a handful of alternatives to be tested, resources would be better utilized by increasing sample size (see recommendation above)
• Version 2 of the new possible designs for testing (see Page 2 of the Appendix)
o The open text boxes can be reduced in width by 50%, and the space can be used to provide examples for write-in categories found in Census 2010.
• Version 3 of the new possible designs for testing (see Page 3 of the Appendix)
o The principles of balance and symmetry would be further enhanced by filling up two complete lines of examples. Currently, the write-in examples for Asian only take up 1.5 lines, with room to include additional categories that were provided in 2010. Critically, this change would help achieve the two central principles of accuracy in detailed reporting and using 2010 as a baseline for effective practice.
• Version 1 of the Internet-based instrument of the new possible designs for testing (see Page 4 of the Appendix)
o The principles of balance and symmetry would be further enhanced by filling up two complete lines of examples. Currently, the write-in examples for Asian only take up 1 line, with room to include additional categories that were provided in 2010. Critically, this change would help achieve the two central principles of accuracy in detailed reporting and using 2010 as a baseline for effective practice.
• Follow-up screen for Version 1 of the Internet-based instrument of the new possible designs for testing (see page 5 of the Appendix)
o This follows good practices of Internet-based survey research, by allowing for follow-up prompts based on particular choices that make for more efficient use of screen space, while not overwhelming the user with the full universe of detailed choices. Based on these design principles, and using the screen examples provided (see page 5 of the Appendix), adding choices to the second column (currently left blank) would be an optimal way to achieve the two central principles of accuracy in detailed reporting and using 2010 as a baseline for effective practice.
We would like to reiterate our commitment to working with the Census Bureau on the future formulation of the race and ethnicity questions with an eye towards addressing concerns and needs from both the Bureau and the AANHOPI community. We believe we share common goals around ensuring accurate data and improving data collection efforts on detailed groups. As we strive to find solutions that address all needs, it is imperative to t focus on outcomes and accuracy first and foremost. We should not view “equity” in a superficial manner (i.e., same number of check boxes, examples, etc.). Instead, the Census Bureau should commit to treating groups equally with respect to the quality of the data collected. We look forward to future conversations on this issue.
Cc: John H. Thompson, Director, United States Census Bureau
Paul Watanabe, Chair, U.S. Census Bureau National Advisory Committee on Racial, Ethnic, and other Populations
Congresswoman Judy Chu, Chairwoman of the Congressional Asian Pacific American Caucus (CAPAC)
Kiran Ahuja, Executive Director, White House Initiative on Asian Americans and Pacific Islanders