It’s by no means an original thought to say that returning the United States political system to some sort of normality will require most of all a change in the Republican Party. As numerous analysts have pointed out, the existence of a sane, mainstream centre-right party, committed to constitutional government, is of fundamental importance in a democracy.
But for most of the last four years such change has seemed a distant prospect. It’s even been difficult to see the fault lines along which a division between the Trumpist Republicans and their hypothetically sane counterparts could occur: the party’s unanimity in support of the president has been remarkable. A number of its elders and intellectuals supported the anti-Trump Lincoln Project, but it barely registered with the rank and file.
Now, however, as Donald Trump’s term comes to a close, things are changing. The attempted insurrection in Washington earlier this month – and the question of Trump’s responsibility for it – have split the Republican Party down the middle. It’s now possible to say something about which people are on which side.
So, prompted by this map posted on Twitter by Grant Gregory,* I had a look at the roll call of the vote in the House of Representatives on the bogus Trump-inspired challenges to the vote in the electoral college. On the question of Arizona’s votes, 121 Republicans voted to uphold the challenge, with 83 voting against and seven not voting. (I’m using Arizona because it was the more even split; on Pennsylvania the Republicans split 138 to 64.)
I segmented the Republican vote in two ways; firstly according to how safe the member’s district is, measured by whether or not the Republican vote at the most recent Congressional election reached 62% (a figure chosen because it divides the list neatly in half). Secondly, according to geography: I marked out a Trumpian heartland, consisting of the deep South and Appalachia (Alabama, Arkansas, Georgia, Kentucky, Louisiana, Mississippi, Oklahoma, South Carolina, Tennessee and West Virginia), a borderland area, consisting of the states surrounding that block (Florida, Indiana, Iowa, Kansas, Missouri, Nebraska, North Carolina, Ohio, Pennsylvania, Texas and Virginia), and a remainder, consisting of all the rest of the country.
The geographic effect is very clear. Of the 51 Republican members in the heartland, 37 (72.5%) voted to uphold the Arizona challenge. Of the 94 in the borderland, it was 55, or 58.5%. But in the remainder of the country, only a minority of Republican members – 29 out of 66 (43.9%) – voted the same way.
But safety of district matters as well. Of the 106 members with marginal seats, a narrow majority (55, or 51.9%) voted to uphold the challenge. The Republicans with safe seats, however, backed it by a larger margin: 66 to 39, or 62.9%.
It’s easy to identify a causal mechanism for this latter factor. Republicans with safe seats don’t have to worry much about the voters at large; the main threat to them comes from challengers in the Republican primary, so pleasing committed Republicans is more important to them. Conversely, those in marginal districts are more concerned about swinging voters and less about Republican activists, which could lead them to take more mainstream positions.
And because Republican seats tend to be safer in the heartland than elsewhere, that could help to account for the geographic pattern. The question is, is that all that is going on, or is geography predictive of a member’s position independently of how safe their seat is?
The answer is yes, it is. Heartland Republicans were more likely to support the challenge, regardless of how safe their seats were – in fact, those in marginal seats supported it at a slightly greater rate (75.0% to 71.8%). Those in the remainder area voted against in both cases, again with the safety of the district having the opposite effect to what we predicted (47.9% and 33.3%). Only in the borderland did a district’s margin seem to have a big influence, with 50.0% of those in marginal seats backing the challenge but 66.7% of those in safe seats.
These results should be taken with some caution; for one thing, there aren’t a lot of marginal heartland districts or safe remainder districts (12 and 18 respectively), so we’re dealing with small sample sizes. But they suggest that the parts of the country where safe Republican districts predominate really are electing a different sort of Republican: those members are not just behaving differently due to the political pressure on them.
More encouragingly, they also suggest that the constitutionalist minority in the party has its own power base: it does not just consist of members for swing districts who are ignoring their party activists. But the picture is complicated, and there are plenty of cases that run the other way.
Geographical polarisation is always a bad sign, and with a few unhinged Republicans already throwing around talk of civil war it’s far from comforting to find that there might be some territorial basis for it. In fact I think there’s less to that threat than meets the eye, but that will be a topic for another day.
* Thanks to Stephen Davies for drawing my attention to it. Some of this discussion started out as comments on his Facebook wall.