But the Left has its own nonsense. How many liberals still believe that George W. Bush “stole” the 2000 presidential election? Jonathan Chait of New York Magazine, hardly a denizen of the fever swamps, was declaring the 2000 recount stolen as recently as last month. And if you want fever swamps, consider a 2006 Scripps Howard poll found that half more than half of registered Democrats believed George W. Bush was complicit in the September 11 terrorist attacks, with respondents split about evenly between calling Bush’s involvement “very likely” and “somewhat likely.”

There’s a connection between the two. As “elite” media figures know, stories — true and false — trickle down, implanting themselves in the minds of hundreds of thousands or millions of citizens too busy or too lazy to do their own research. When Vox writes, “The election probably wasn’t hacked. But Clinton should request recounts just in case,” it’s legitimizing a seed of doubt.

It’s no surprise, then, that Paul Krugman — Princeton economist, New York Times columnist, Nobel laureate — spent Tuesday night on Twitter calling for an “independent investigation” of election results, based on a New York Magazine report that a handful of “prominent computer scientists and election lawyers” think results in Wisconsin, Michigan, and Pennsylvania “may have been manipulated or hacked.” There’s no meaningful evidence to support that charge, as the Times’ Nate Cohn immediately pointed out, but it’s now an active point of discussion on cable news.

Where are the lines dividing “fake” news from real? Why was voter fraud a rightwing “conspiracy theory” when conservatives push it, but an urgent matter of electoral transparency now that it’s coming from liberals? Why are right-wingers fabulist nuts, but left-wingers devotees of triumphant Reason? And when Ezra Klein neglects the context that effectively invalidates his thesis, is it a mistake or a lie — or “fake” news?