There is a curious disjuncture in this claim that you can draw strong inferences about national candidate quality from state and local election results:
By this measure Bernie and Beto look pretty good, Harris/Warren look bad, and Klobuchar/Brown/Baldwin look like champs winning big in *the* key swing region.
— Matthew Yglesias (@mattyglesias) December 29, 2018
You see the problem here? Where one might expect to see actual evidence for the claim that previous election results translate into national candidate quality, there’s just…a bunch of confident ex ante speculation. I would note as well that even on its own terms these inferences are highly contestable. How much more impressive is Bernie winning by 40 points in a historic wave year than Pat Leahy winning by 28 points in a slightly pro-Republican year, really? And what does political success in a tiny, virtually all-white atypically liberal rural state tell us about appeal on a national stage anyway? One can say something similar about Beto losing a state by two points that Clinton lost by nine; once you boil off the much more favorable structural circumstances of 2018 it’s pretty weak tea. [ETA] The argument about Harris being weak seems to be based on a single low-profile statewide election in a Republican wave year.Why should Warren’s underperformance of presidential tickets matter more than Mitt Romney’s much greater overperformance of them? If strong political performance in a midwestern purple state is evidence of a strong presidential candidacy, say hi to Presidents Jeb! Bush and Scott Walker.
Anyway, to be meaningful this claim would have to be supported by historical evidence, and I assume it isn’t because it can’t be. Obama, for example, lost the only remotely tough election he ever fought, and I don’t see how his win in a deep blue state against a farcical candidate parachuted in at the last minute tells us more than Clinton beating Lazio in what was at the time considered a competitive race. (Those contemporaneous assessments of Lazio might have been wrong, but this underscores the fact that this is all just an endless rabbit hole of tautologies that can produce any conclusion you want.) The 2016 Republican primaries, chock full of politicians who had had apparently high candidate WARs, were won by someone who had never been elected to anything, and the runner-up was someone who arguably had the least impressive relative electoral record of the serious candidates.
Basically, trying to project “electability” in a national election is a fool’s game, and to the extent that there are tangible measures (like fundraising) these will be tested by the primary process itself. Support the candidate you think would be the best president.
…EliHawk in comments:
I mean, in theory there are ways it could be relevant, but it’s amazing how few, if any, recent Presidents had any kind of “overperform relative to the fundamentals” success, like, ever.
Obama: Only statewide race was a blowout after the GOP candidate quit
W: Knocked off an incumbent Governor (good) by 8.5% (good) in a sturdy GOP state (weak) who had only won 49.5% of the vote in 1990 (weak) while the GOP Senator was winning by 22% (weak).
Clinton: Arkansas is kind of small enough for gubernatorial elections to be sui generis but did manage to lose a House race in the Watergate midterm and lose the Governor’s race in the bad Dem year of ’80 and win it back in the good Dem year of ’82
Bush: Only won two House races; lost Senate races in both bad GOP years (’64) and mediocre ones (’70)
Reagan: Managed to win CA-Gov in a good GOP midterm against an incumbent facing ‘third term’ fatigue in a lean-GOP state
Carter: Pre-74 or so Southern Politics is a totally different beast, but his most difficult political accomplishment involved winning the 1970 Democratic Primary by nuking popular ex-Gov. Carl Sanders with race baiting demagoguery, which is, um, not an ‘electability’ message anyone was pitching at National Dems in 1975