Echelon Insights has an estimate--to be precise, 157,408,480! That's quite a high level, working out to be about 65.8 percent of the Voting Eligible Population (VEP), which is a higher turnout rate than 2008 and an increase of about 18.6 million voters over 2016.
Is this a plausible estimate? I've looked at their methodology and it's quite respectable. What we're looking for here is not them hitting turnout on the nose--the precise estimate is offered a bit tongue in cheek I think--but being in the right neighborhood. In that sense, I think it is a plausible estimate.
So, potentially a good result for democracy. But what about for Joe Biden? This is harder to suss out. One might assume that higher turnout is generally good for the Democrats though of course that really depends on the distribution of voters who do show in larger numbers.
The Echelon Insights turnout model does provide some broad bush answers on distribution. It indicates that nonwhites should be 2 points higher as a share of voters, presumably good news for Biden. But, in Patrick Ruffini's article on Medium, he points out some good news for Trump: an estimated 42.9 percent of voters should be white noncollege, Trump's best demographic group. He notes that this figure is higher than the figure many pollsters are assuming and is also higher than estimates generated by the States of Change project. (Ruffini carefully explains the reasons for this divergence in his article.)
But, based on figures provided to me by Ruffini, there is also bad news for Trump about his good news. Compared with their analogous 2016 figures, their 43 percent estimate for white noncollege in 2020 would represent a 3 point drop in vote share from the previous election. So that's not so good for Team Red.
So, stay tuned. A lot of moving parts here!