2. Trump's unpopularity is very inelastic.
To me, this is his biggest problem. I think that the good economy is already "baked into" his approval ratings and is unlikely to save him even if we don't fall into a recession. There's a strong (but far from inevitable) chance that in fifteen months many on this board will be saying that Trump's defeat was always inevitable due to his consistently low approval ratings.
The idea that 2018 was a leading indicator and thus is more predictive of the next presidential election is an interesting thought, but there's also a strong chance that either the realignment won't turn out like we expect (in terms of the composition of the new Democratic coalition) or that the realignment won't happen as fast we expect. Though Pericles does make a good point about how the high turnout in 2018 could mean that it's more reflective of the 2020 election.
