TV pundits and op-ed writers of every major newspaper epitomize how the Democratic establishment has already reached a consensus: the 2020 nominee must be a centrist, a Joe Biden, Cory Booker or Kamala Harris–type, preferably. They say that Joe Biden should "run because [his] populist image fits the Democrats’ most successful political strategy of the past generation" (David Leonhardt, New York Times), and though Biden "would be far from an ideal president," he "looks most like the person who could beat Trump" (David Ignatius, Washington Post). Likewise, the same elite pundit class is working overtime to torpedo left-Democratic candidates like Sanders.

For someone who was not acquainted with Piketty's paper, the argument for a centrist Democrat might sound compelling. If the country has tilted to the right, should we elect a candidate closer to the middle than the fringe? If the electorate resembles a left-to-right line, and each voter has a bracketed range of acceptability in which they vote, this would make perfect sense. The only problem is that it doesn't work like that, as Piketty shows.

The reason is that nominating centrist Democrats who don't speak to class issues will result in a great swathe of voters simply not voting. Conversely, right-wing candidates who speak to class issues, but who do so by harnessing a false consciousness — i.e. blaming immigrants and minorities for capitalism's ills, rather than capitalists — will win those same voters who would have voted for a more class-conscious left candidate. Piketty calls this a "bifurcated" voting situation, meaning many voters will connect either with far-right xenophobic nationalists or left-egalitarian internationalists, but perhaps nothing in-between.

From: There is hard data that shows that a centrist Democrat would be a losing candidate | Salon.com