Two articles on Vox caught my attention:
The first follows up on the notion floated by Ed Rendell recently about the Democratic strategy that targets Republican voters should Trump gain the nomination, that, in face of Trump nomination, many regular Republicans will stay home or even vote for Clinton, even while Trump gains working class whites who have been largely missing from the polling place. As far as the presidential elections are concerned, this may be a reasonable bet, as per the significantly limited willingness of Republicans to back Trump as I pointed to earlier. Still, I think it is a long shot: the educated, upper income partisans are the most loyal of the lot, and their defection from the party ranks is not something I could count on under any circumstance. As for the prospect of their voting Democratic down the ballot, I think it is a ludicrous fantasy. It is worth remembering that many Democratic voters who voted for Reagan in 1980s loyally returned Democratic congressmen to the House. Educated and affluent suburbanites vote with very high probability: they may defect to HRC with some probability, but it is absurd to think that they will vote for Democrats down the ballot just because they dislike Trump any more than the love of Reagan caused the Reagan Democrats to ditch the Democratic House members three decades ago. There may be a few Democratic pickups here and there, but the demographics of the presidential year makes it somewhat likely, considering the problem Republicans have in mobilizing their voters relative to Democrats. Some unexpected suburban districts may switch parties, but probably not numerous enough for anything like the Democratic takeover.
The second point relies on dismissing the presently available data as meaningless. Given the timing, that is not unreasonable: the reason Sanders leads in head-to-head polls is because he leads among the indepedents. In the recent Quinnipiac polls, Sanders outperforms Clinton by about 10% against every Republican candidate among the “independents” and that accounts for why Sanders does better in head to head match ups. The independent voters are, of course, the least informed and least likely to have made up their mind at this stage. But we have had more data than just the polls: the primaries have been showing the lay of the land much more precisely than any polls have, and they confirm the polls: practically in every state so far, Sanders does better (or is at least competitive, in the states he lot badly) among the independents compared to Clinton. There is no mistaking that Clinton wins on the basis of Democratic votes, while Sanders does on independents so far in the primaries. Will the pattern continue on the election day? There is no reason to believe it will be otherwise. Sanders would probably outperform Clinton significantly among the independents on the election day, were he the nominee, and in competitive states like Virginia, Ohio, or Michigan, the inability to appeal successfully to independents will cost Democrats the state. Indeed, the demonstrated lack of ability on the part of Clinton to appeal to independents is a serious weakness that has been exposed in the primaries so far, even in the states that she won handily. This points to a significant electability argument in favor of Sanders.
Just my two cents.