The COVID-delayed results of the 2020 census are finally in, with totals for the 50 states and the District of Columbia at nearly one-third of a billion — 331,449,281 — and with surprises having to do with the short run and the long term.
The short-term news revolves around the function for which the framers of the Constitution mandated the world’s first regularly scheduled census: the reapportionment of seats of the House of Representatives among the states. That’s done according to a 1941 statutory formula that the Census Bureau conveniently applies.
The results were underwhelming. Only seven seats out of 435 were switched from one state to the other. Texas gained two, and Colorado, Florida, Montana, North Carolina and Oregon gained one each. Losing one each were California (for the first time in history), Illinois, Michigan, New York, Ohio, Pennsylvania and West Virginia.
Population and representation continue to flow from the Northeast and Midwest to the South and west and, generally, from Democratic states to Republican states. But they will also recognize the changes are small, nothing like the censuses in which one state gained eight seats (California in 1960) or another lost five (New York in 1980).
The partisan effects are likely to be small as well, expert forecasters agree. Sean Trende of Real Clear Politics predicts a Republican net gain of four seats. The Cook Political Report’s David Wasserman pegs it at 3.5 and Kyle Kondik of Sabato’s Crystal Ball at two. Not quite enough to overturn the 222-213 majority Democrats won in November 2020.
All three emphasize that the redistricting processes within the states could produce a wide range of results. According to Wasserman, Republicans control redistricting in states with 187 U.S. House districts, Democrats in states with 75 districts and theoretically bipartisan commissions in states with 121 districts. Control is split between parties in states with 46 districts, and six states have just one district each.
That’s less of an advantage than Republicans had in the 2010 cycle and about the same as they had in the 2000 cycle; it’s less than the advantages Democrats had in the 1960, 1970 and 1980 cycles. Democrats’ advantages then derived from their majorities in northern metro areas and near monopolies in the South. Republicans’ more recent advantages are due mainly to the clustering of Democratic voters in central cities, sympathetic suburbs and university towns, while Republican voters are more evenly spread around the country.
As for the long-term effects, the 2020 census shows less population change and less internal migration than government and private estimators expected, based on models from previous decades. Arizona grew 3.3% less than the census estimate and didn’t gain the seat widely forecast, and Texas and Florida each fell a seat short of expected gains.
On the other hand, population outflows were less than expected, especially in New York, New Jersey and Rhode Island. The latter two didn’t lose seats as expected, and New York was only 89 people short of not losing a seat for the first time since 1940.
Speaking of which, the picture the census paints of the 2010-20 decade closely resembles that of the long-past decade of 1930-40, when the nation’s population increased by only 7.3%. That’s eerily similar to the 7.4% in the decade that just ended, dominated by the sluggish Obama recovery of 2009-16 and the downscale-driven, pre-COVID Trump upturn of 2017-19.
These two stand out as the lowest growth intervals in American history. In every other 10-year period, the nation’s population has grown by double-digit (rounded off) percentages.