Census here and there

Tuesday night was census night in Australia, so Australians this week have been filling out their forms to provide a snapshot of our population as of that time – a task made easier by the fact that much of the country is in a Covid-induced lockdown. (Here’s where to go in case you missed it.)

But as we noted last year, the census in Australia is of limited political importance. The allocation to states of seats in the House of Representatives is done every three years on the basis of population estimates, not actual census figures (although the latter provide an important check on the accuracy of the estimates), and the drawing of boundaries, both state and federal, is based on enrolment figures, not population.

So while the census results, when they are gradually released over the next twelve months, will tell us interesting things about the effect the pandemic has had on the composition of Australia’s population, they have no immediate impact on the fortunes of our political parties.

Things are different in the United States, where the constitution requires that apportionment of congressional seats take place on the basis of population every ten years. The census is always held in years ending in zero, and its results are used to apportion seats and then redraw boundaries in time for the election two years later – in this case, next year’s mid-term election.

The processing of last year’s census was delayed due to the pandemic, so state population figures were not released until last April (here’s my report at the time). Thirteen states (an unusually small number) saw a change in their entitlement to seats in the House of Representatives. But all states, except for the six that are only entitled to one seat, will revise the boundaries of their districts to keep pace with shifts in the distribution of population.

Yesterday the census bureau released the data on which that redistricting will be based. It’s all in pretty raw form; the “full redistricting data toolkit” is not due for release until 30 September. But the press release summarises some of the interesting features, including the fact that “the U.S. population is much more multiracial and more diverse than what we measured in the past” – due, the bureau says, to better measurement as well as actual demographic change.

Increasing urbanisation and increasing ethnic diversity both spell long-term political trouble for Republicans. But the immediate effect of this round of redistricting is expected to work in their favor, for the simple political reason that Republicans control the process in more states than Democrats do. Nonpartisan redistricting (something that most of the world takes for granted) has made gains in the US since the previous census, but in far too many states the drawing of boundaries is still completely politicised.

Despite Democrat majorities in both houses of Congress, their chance of passing any comprehensive electoral reform that might address this problem is essentially nil. And even a more limited proposal that would specifically target partisan gerrymandering (as many commentators have urged) seems most unlikely to be passed in time, if at all.

The late release of data already means that many states are going to have trouble meeting their various practical or legislative deadlines. It’s not clear, however, whether that will make them more ready to compromise or simply provide an excuse for rushing maps through with less consultation. (And compromise does not always favor democracy; often it just means the two parties agree to protect each other’s incumbents.)

FiveThirtyEight.com has a new feature where you can track the progress of redistricting across the country. Derek Muller provides a helpful list of other sites covering the issue. It should be an interesting one to watch over the next few months.