I agree to a very large extent; I was just commenting on a response to an article about increasing voter turnout. The suggested approach (mail-in-only voting),

I agree to a very large extent; I was just commenting on a response to an article about increasing voter turnout. The suggested approach (mail-in-only voting), while nice in many ways, does not (empirically) have that particular effect. I think voter turnout has become the axis on which all modern (presidential) elections have hinged. Trump didn't beat Hillary in the rust belt by convincing hoards of Obama supporters to bot Republican. What the Obama voters did, though, was stay home. The numbers bear this out (although we don't have individual tallies) -- Trump didn't win any more votes in those states than Romney did in the last election. The same voters who pulled that lever came out and did the same thing they did in the last cycle. That's also why the polls were so off -- they rely on "likely voter" models, not realizing that what they should be measuring is not who likely voters will vote for, but whether or not people will show up at the polls. Instead, they used "likely voter" criteria that they learned in the Obama election and gauged sentiment in that group.