I take some of those points. I don’t know the state or trend of polarization in other countries (only countries with two main parties would be of interest for this model, multi-party elections yield a different system). Though I maintain that one country is still an observation to be explained, requiring a hypothesis. I could also believe that donations or activism don’t drive election performance.
But fundamentally, I think you’re addressing an issue with my suboptimal framing of the puzzle, rather than the puzzle itself. I’m not really interested in the time trend (increase?) of the polarization of politicians. From the perspective of MVT, it’s surprising that there are any differences between the two main candidates’ positions, irrespective of the amount of polarization in the electorate. Certainly it’s surprising that when elections are close, politicians don’t move further towards the center in order to win—something must be restraining them. What is it?
I agree individual politicians could be irrational (in the sense of not optimizing for getting elected, though the precise mechanism still interests me). For example, maybe they do not change their true ideological positions to appeal to voters even when getting elected is at stake (is this likely?). But you’d still need to explain why other politicians, whose views are genuinely more moderate, don’t emerge and win the elections. In the same way that individual producers and customers could be biased/selfish, but the market could still end up efficient & useful, the political system could have generated the most electable politician, who is a moderate. Why doesn’t it? I can’t think of any hypothesis not listed or raised so far.
This is a very good point.