What 2012 Population Estimates Could Mean in 2020

What 2012 Population Estimates Could Mean in 2020

By Sean Trende - December 21, 2012

While Washington bickers over the fiscal cliff, and normal people frantically finish their Christmas shopping, numbers geeks have been eagerly awaiting the 2012 census population estimates. These are the annual revisions that the U.S. Census Bureau does to the decennial count, and they give us some insight into how the population is changing.

The numbers are now out, and the shifts again confirm what we already mostly knew about population trends: The Northeast and Midwest grew very slowly, the West grew at a modest pace, and the South grew about 3.5 times as fast as the Northeast and Midwest combined. The pace of immigration picked up slightly, but the overall pattern remains the same: relatively slow population growth, accompanying slow economic growth.

But one of the more interesting things that we can do with the census estimates is compare them to the 2010 numbers, and extrapolate those changes in numbers out to 2020. With those estimates for the next decennial count, we can calculate what the apportionment of congressional seats the next time around would look like. I did this in 2011, and came up with the following results:

Winners: California (+1), Colorado (+1), Florida (+1), North Carolina (+1), Oregon (+1), Texas (+3).

Losers: Illinois (-1), Michigan (-1), Minnesota (-1), New York (-1), Ohio (-1), Pennsylvania (-1), Rhode Island (-1), West Virginia (-1).

If we take the latest estimates and do a simple linear extrapolation (take the difference between the 2010 and 2012 numbers, multiply by five, add that to the 2010 figures), we get the following updated results:

Winners: California (+1), Colorado (+1), Florida (+1), North Carolina (+1), Texas (+2), Virginia (+1).

Losers: Illinois (-1), Michigan (-1), Minnesota (-1), Ohio (-1), Pennsylvania (-1), Rhode Island (-1), West Virginia (-1).

Some of these changes have some significance. California resumes its upward march in seats. New York fails to lose a seat for the first time since the 1940 census. Michigan falls to 13 seats, its fewest since the 1910 census, while Ohio falls to 15 districts, the fewest it has had since the 1820s.

Many of these seats are tentative. Only a few thousand people separate seat number 435 (California’s 54th district) from “just missed” seat number 436 (Texas’s 39th district). Seats number 431 through 434 are, in order, Virginia’s 12th, Colorado’s 8th, Alabama’s 7th, and New York’s 27th. Those just missing seats (437 through 440) are West Virginia’s 3rd, Oregon’s 6th, Minnesota’s 8th, and, interestingly, Montana’s 2nd.

But perhaps we should weight recent population shifts more heavily. Instead of just looking at the difference between 2012 and 2010, let’s extrapolate from an average of the shift from 2010 to 2011 and 2011 to 2012, weighting the latter shift twice. Under that scenario, we get the following results:

Winners: Arizona (+1), California (+1), Colorado (+1), Florida (+1), North Carolina (+1), Texas (+4), Virginia (+1).

Losers: Alabama (-1), Illinois (-1), Michigan (-1), Minnesota (-1), New York (-1), Ohio (-1), Pennsylvania (-2), Rhode Island (-1), West Virginia (-1).

Under this scenario, the last five seats go to California, Wisconsin, Arizona, Texas, and Illinois. The five closest misses are from Pennsylvania, Florida, Oregon, California, and Alabama.

It’s worth noting that the overall trend continues to be one from blue states to red states. Under the first scenario, Romney would have picked up two electoral votes over his 2012 totals. Under the second, he would have picked up four electoral votes over his 2012 totals, and the “sweep the South, plus Ohio, plus one other state” strategy would have dropped the “some other state” part. Some argue that Texas and Arizona will be purple by 2024 (the first presidential year under these numbers), but I am skeptical. This is an argument for another article, but it is worth noting that, despite the demographic shifts, the Republican vote shares here haven’t budged in the past eight years.

In the bigger picture, this is more consequential. Under the second scenario, Bush would have won 287 electoral votes in 2000 and 294 in 2004; he could have lost Ohio and either Iowa or New Mexico (the two closest states) but still would have won.

Again, these will get clearer as we draw closer to 2020. By 2007, we had a very good look at what 2010 would look like, but even the 2003 estimates got us into the ballpark. 

Sean Trende is senior elections analyst for RealClearPolitics. He is a co-author of the 2014 Almanac of American Politics and author of The Lost Majority. He can be reached at Follow him on Twitter @SeanTrende.

Sean Trende

Author Archive

Follow Real Clear Politics

Latest On Twitter