New numbers for America

I mentioned a couple of weeks ago that two Australian states, Victoria and Western Australia, were currently undergoing redistributions of federal electoral boundaries, to provide for changes in the number of MPs that states are entitled to due to population shifts. This week a similar, but different, process comes to the United States.

In Australia, state and territory populations are reviewed for this purpose at least every three years; they do not have to be actual census figures, just “statistics … that the Australian Statistician has … compiled and published in a regular series.” But in the US, only the census will do, and it is only held every ten years. Last year was the year.

In theory, the population was counted as of 1 April last year, and normally the figures would have been ready for release by 31 December. But for a variety of reasons, some connected with Covid-19 and some with the Trump administration’s efforts to politicise the process, the release has been delayed until this week.

But now the numbers are out. The United States (including the District of Columbia but excluding the other territories) has 331,449,281 people, an increase of 7.4% since 2010. All but three states (Illinois, Mississippi and West Virginia) increased in population; the fastest growth was 18.5% in Utah, while the biggest increase in absolute numbers was in Texas, which grew by 16.1%, or a little over four million.

That will change the allocation of seats in the House of Representatives. The total number is fixed at 435, but as of the next election, in November next year, seven states will each lose a seat: California, Illinois, Michigan, New York, Ohio, Pennsylvania and West Virginia. Five states – Colorado, Florida, Montana, North Carolina and Oregon – will gain a seat each, and Texas will gain two.

That’s an unusually small change: last time around, 18 states saw some change to their representation, with Texas then gaining four seats. But the general direction of change is the same as it has been for some time. Population is shifting, at least in relative terms, from the north and mid-west to the south and west, from older (and colder) industrial areas to sunnier and more dynamic ones. California’s loss of a seat was the exception – in fact it is the first time its representation has ever declined.

The political effects, as far as Congress is concerned, will depend on exactly how the new boundaries are drawn, a process that in the US is heavily politicised. The states that are gaining population tend to lean Republican and those that are losing it lean Democrat, which benefits the Republicans in two ways: first because there will be more Republican votes to go around, and secondly because Republicans are more likely to control the redistribution process (“redistricting”, as the Americans call it) in growth states.

But it’s not all bad news for the Democrats. Although the sunbelt states still mostly vote Republican, the areas of them where the growth is, principally the suburbs of big cities, are much more Democrat-leaning. So it may require considerable creativity in some of them to avoid giving the Democrats an extra seat. Conversely, the states losing seats tend to be under Democrat control, so they may be able to ensure that it is the Republicans who lose out.

And in one of the declining states, West Virginia, even though the process is controlled by the Republicans we know that it will be a Republican seat that disappears, because there are no Democrat-held seats there to start with!

The other thing to note is the way the changes flow through to the electoral college, in which each state chooses a number of electors equal to its total Congressional delegation, representatives plus senators. Last November, Joe Biden carried five of the seven states that are losing a seat, and only two of the states that are gaining one. So if the same result was repeated with the new entitlement, his margin in the college would come down from 306-232 to 303-235.

That’s not a big difference, although in a close election – and the US seems destined to have close elections for a while – it could happen to be decisive. It’s also not certain that it would work to the Republicans’ advantage. The growth states currently vote more Republican, but that advantage is declining; the relative movement tends to be towards the Democrats. Last year the six growth states recorded a median swing to Biden of 2.0%, 0.8% better than the national average. The seven states whose representation will go down had a median swing of just 0.1%.

In other words, if the electoral map continues shifting the same way, the reweighting of the electoral college could end up benefiting the Democrats – if not in 2024, then quite plausibly in 2028. If the Republicans by then have picked up Michigan and Pennsylvania but lost Florida, North Carolina and Texas, population change will no longer look like their friend.