Bloomington’s 2020 census count: Plot (of dots) thickens for IU enrollments, dormitories

Ten days ago, when the news broke that Bloomington’s 2020 Census count had dropped, compared to the 2010 numbers, many residents reacted with disbelief.

The leading theory for the 1.5-percent drop, from 80,405 to 79,168, is based on the fact that Bloomington is home to Indiana University’s flagship campus.

The city experienced a pandemic-related mass exodus of students right around the time the census count was taking place, in late March and April. The university delivered remote-only instruction for the rest of the spring semester.

Thousands of students were undercounted in the 2020 Census, goes the theory, which would explain why the vintage 2019 estimates by the US Census Bureau put the population of Bloomington at 86,630, or about 7,000 more than the actual count done in 2020.

A theory based on student undercount is consistent with The B Square’s coarse-grained look at the precinct-by-precinct geographic distribution of population losses.

If students were undercounted, then that should be reflected in losses in on-campus and near-campus neighborhoods where many students live. The geographic distribution does show diminished numbers in areas on and near campus.

A different theory of that geographic distribution does not rely on the idea of a student undercount: If there were fewer students actually living in those campus areas, fewer students would be counted there.

That’s a theory that looks like it could have some support from two possible perspectives.

Continue reading “Bloomington’s 2020 census count: Plot (of dots) thickens for IU enrollments, dormitories”

Census 2020 analysis: Lower numbers in Bloomington’s student neighborhoods point to possible undercounting

The 1.5-percent decrease in population reported for Bloomington by the US Census last week has generated local skepticism about the accuracy of the count.

The most recent estimates from the US Census had pegged the city’s population at around 86,000. But the 2020 numbers came in under 80,000, less even than the actual count in 2010.

The result was a big enough surprise that the city’s mayor, John Hamilton, issued a statement last Friday, the day after the results were released. The statement raises the possibility that the reported numbers are not accurate, because of an undercount of Indiana University students.

Hamilton’s statement points out: “[M]uch of the census data collection began the very week thousands of university students were directed to leave for the semester due to the COVID pandemic.”

If an undercount of university students contributed to Bloomington’s lower numbers, the undercount would likely be detectable in the geographic distribution of population losses in Bloomington, as counted by the US Census.

Based on a B Square plot of precinct population counts in 2020 compared to 2010, the idea that college students were undercounted has some statistical support.

It was predominantly areas on and near the Indiana University campus that showed lower counts in 2020 compared to 2010.

Continue reading “Census 2020 analysis: Lower numbers in Bloomington’s student neighborhoods point to possible undercounting”