I never said this election indicated a disaster for Democrats. I merely pointed out there was no 6 point swing. I also stated the voter enthusiasm was about the same for both parties.. I stated some reasons why this might be so. There may be other reasons.
For now the election in Texas 27 stands alone. It is a anomaly. To be a harbinger it must be confirmed.
I am late posting this, as I went to church.
Those are fair statements. As you said, it's just one data point among others. There have been Congressional specials this cycle with some small swings, some large ones (all toward the Democrats). There's naturally going to be some variation in results among the entire set of such elections. The point is that the final margin in November is likely to be closer to the average of these results, rather than to an extreme on either end.
This is the list of results compiled by Harry Enten (who, incidentally, defines swing as the change from the district's partisan lean):
Source: https://twitter.com/ForecasterEnten/status/1013418919363858433
(No, I don't think the final margin in November will be D+16. As I've said elsewhere, I think this is a reasonable upper bound on the likely Democratic margin, with the GCB average as a reasonable lower bound. )
Before 2006 the special election swing average was like D+18, and resulted in a D+7 house vote. The reason that special election swing is correlated to final swing but not particularly predictive of the final margin is that it is a great measure of enthusiasm, but neglects two key factors. First, that there is an incumbent running in most districts (which in and of itself is worth about five points) and second that more people show up to vote in a midterm than just the motivated portion of your base, thus diluting the influence of an enthusiasm gap (though obviously not erasing it in any way).
Based on the special election performance I think a prediction of D+16 is very silly, even as a ‘ceiling’. There’s no reason to think Democrats will even come close to that unless they have a great campaign cycle and the generic ballot lead opens up a lot wider. Unfortunately such a swing would be impossible to see in special election results because there is only one more prior to November.
