I’ve underestimated Trump since January (I was more bullish on him than most until that point), and I think it would be a mistake to write him off completely for the general election. While I think the media will turn on him, and his persona will ultimately not wear well, he’s facing an opponent who is not, to put it mildly, a particularly deft politician. Moreover, the political science is reasonably compelling that elections are mostly driven by things like the economy and incumbent job approval; my Senate model suggests that the more controversial Senate candidates of the past few years, like Ken Buck, Christine O’Donnell, or Richard Mourdock, probably only cost their party two or three points (four or six, net). That’s not nothing, but it isn’t enough to guarantee a loss.
If there’s cause for concern, whether from a National Front-type Republican Party or a Syriza-style Democratic one (and to be clear, given the nature of our political system, both parties will likely remain more moderate than their European counterparts), it probably comes in 2020. Given the length of the business cycle, the probability of a recession during the next four years is extremely high, and many people have not recovered from the last one. The results of that could be catastrophic, depending on the severity of the downturn, and could convince more voters to try something radically different. A Supreme Court with a swing vote justice who is no longer a culturally cosmopolitan Republican who is reluctant, but willing, to utilize his power to push social change could continue the alienation of traditionalists from mainstream dialogue.