Josh Marshall: “We are now waiting to see the upshot or impact, if any, of the two successive national party conventions. The general consensus was that the Democrats did very well. Then last week – at least as I was hearing it – many seemed to think that the Republican convention was more effective than Democrats anticipated. Suddenly the tide seemed to shift. I felt some of this myself. I have no idea which of these is true. But I can offer one observation that I’m pretty certain is accurate. It’s born of years of experience watching elections. Regardless of the objective realities, Democrats will consistently anticipate loss or worry about loss while Republicans will consistently be confident of victory.”

“This may be obscured by the genuine shock and horror Democrats experienced on election night four years ago. Democrats were pretty confident and all their worst fears were realized. But a closer look shows the general pattern was actually in effect through much of the 2016 cycle. Indeed we saw a particular example of it during the 2018 midterm election. The fall of 2018 was chock full of theories and predictions about how two years of ‘resistance’ activism were coming up short. It was the ‘caravan’. It was Trump’s 12 dimensional chess. It was low turnout among young voters. The pervasive were Democrats’ latent fears of coming up short that they actually persisted well into election night and even the first couple days after the election – until late returns, results of close call races and just the actual numbers made clear Democrats had won a decisive victory. One place you can see this tendency is in those polls which look not at what candidates people want to win but those which they believe will win. Republican candidates routinely over-perform in these polls. The reason is straightforward. Democrats are generally more uncertain of victory while Republicans are confident. The net effect is polls that show most people anticipating Republican victories – or at least anticipating it more confidently than more objective measures merit.