APDU Conference Panel Notes: Diversity, Equity, and Inclusion in Public Data


Arturo Vargas, CEO of the National Association of Latino Elected and Appointed Officials (NALEO) Educational Fund and Terry Ao Minnis, Senior Director of Census and Voting Programs for Asian Americans Advancing Justice (AAJC), discussed diversity, equity, and inclusion (DEI) in public data at the APDU (Virtual) Annual Conference on Tuesday, July 27. Moderated by Hansi Lo Wang, National Correspondent for NPR, the panelists explored changes to census data collection practices, the relationship between the quality of data and equity, and how NALEO and AAJC can use upcoming Census data to achieve their organization’s goals.

In large part because of the Latino population, Texas gained two congressional seats in the last Census. Quality data fosters accurate representation and helps distribute resources equitably. The panelists plan to keep an eye out for many components of change, even those that appear less notable.  Protecting the civil rights of Latin Americans and Asian Americans, NALEO and AAJC use public data to influence program design, implementation, and even fundraising.  Since detailed analysis is yet to come out, the impact of the most recent Census is uncertain. While upcoming data will not change reapportionment, it can change funding formulas or influence a policymaker’s decisions.

Asian American and Latino populations have grown over the years. By state down to the metro level, Minnis notes that it is important to keep track of changes in populations across the board because “without the data we cannot show elected officials who makes up their community and who needs their issues addressed.” However, the Census Bureau has struggled to collect complete and accurate demographic data. Advocates like Vargas and Minniswant to change that.

Confusing race and ethnicity questions often are a source of incomplete data. The panelists posit that more disaggregated, descriptive self-identifier questions would rectify this issue. Disaggregated questions mean better data and better data means more effectively allocating funds based on need. Disruptions from COVID-19, hesitancy from the Trump Administration, and other pending Census changes delayed alterations in the 2020 Census. Transitioning to the Biden Administration, proposed changes to race and ethnicity questions are on the table again. If the Office of Management and Budget (OMB) approves the proposed changes, Hansi notes that “it would allow the Census Bureau to ask about race and ethnicity in a radically different way.”

While the Census Bureau is working to improve inclusivity in data collection, privacy concerns in the publication of data are also at the forefront. The Census Bureau is adopting differential privacy plans and a disclosure avoidance system (DAS) at the block level. To scramble data, the Census Bureau will distance race and ethnicity from “as enumerated” data, so that smaller populations are not as susceptible to compromised personal information.

Vargas believes that this change is toeing the line between “what level a certain amount of noise is acceptable and at what level the noise deteriorates the data so that it may undermine civil rights protections.” Minnisfurthers Vargas’ point by noting that the Census Bureau admits that, because block level data is what is used for redistricting, some districts on the margins could be lost. The extent to which confidentiality will affect equity in redistricting is still uncertain, but in time we will know better if it has protected, corrupted, or had no impact on the data. The panel acknowledged the challenges both of utilizing current data and improving practices for future collection. The Biden Administration’s recent efforts to identify gaps in data is promising, but more research is necessary. To learn more or revisit the presentation, a recording of the session is available to conference registrants through Whova.