Bill Sweigert: U.S. census conceals “Warren Effect” on Indian population

By Bill Sweigert

The demographic diversity of U.S. workers is emerging in new data from the U.S. Census Bureau. These tables become the primary source for sociological analysis about jobs and who has them. They are used for employer analysis of their affirmative action programs and for assessing progress from the prior census. The bureau always changes something each census, making line-to-line comparison tough between decades.

Last time, the bureau adopted a new “rounding” rule, making numbers between one and seven equal to four. The rationale was to protect the privacy of individuals in jobs of low occupancy. In other words, if there were one Black doctor in Wayback County, Tennessee, the census would say there were four, just so nobody would learn the race of the Black doctor from the government. If that is the best they can do, policy-makers might do their arithmetic with a big grain of salt.

Another thing the bureau did was to split out “NHOPI” (Native Hawaiian and Pacific Islander, except Japan) from “Asian” at the request of Sen. Daniel Inouye. Most journalists didn’t get the memo and conflate them, talking about the trends in bias against a group they blend together as “Asian American/Pacific Islander.”

In the latest census, the bureau chose not to report Americans who checked two or more ethnicities. Last time, we knew the numbers around “Black and white,” “Asian and white,” “American Indian/Alaska Native and white,” or “American Indian/Alaska Native and Black” and “Other two or more” heritage. So the number of Americans at least part Black is unknown, meaning the percent of “Black or part Black” will appear lower than it was last census.

People who checked one of these combination categories are now lumped together with “race unknown” and what I call the “protest” or “nonsense” responses, like “human, Vulcan, Hobbit, Klingon.” This is a shame because we won’t be able to see how the percent of protest answers to the race inquiry has changed since last time. Colleges report something around 15% of applicants protest the race inquiry.

Another loss is we won’t be able to see the “Warren Effect” in action. That’s what I call the jump in the percentage of American Indians when you also count white folks who claim some Indian heritage. In the 2010 census reports, it jumped from roughly 0.6% of all workers to 1.1%. No other group count has anything near this large an effect, close to doubling. Now we will only know those who claim to be from a single heritage. As we become more and more diverse, government data will reflect it less.

Bill Sweigert, Sweigert Management Associates LLC, Spokane