In my 2008 and 2012 election posts, I correctly predicted Obama’s victories right here on Daily Kos. Four years ago, I called 49/50 states and had Obama winning with 51% of the popular vote — just 0.1% off from his actual share of 50.9%! Tonight I’ll be giving my predictions for this year’s election.
Several years ago, I developed my own statistical model to forecast election results. Why should you trust it? Because it has a good track record: it correctly called the US presidential election in 2012, the Brazilian presidential election in 2014, and the general elections in Canada and Britain last year. It slightly missed Brexit, giving the Leave side only a 30% chance. But Brexit is the only ‘election’ (actually that was a referendum) that the model has called incorrectly, racking up over a dozen correct predictions over the last four years.
According to my calculations, Hillary’s win probability is 93%. I predict that Hillary Clinton will win the presidency with 323 electoral votes to Donald Trump’s 215. I also predict that she will get 50.7% of the popular vote to Donald Trump’s 45.2%. The rest of the vote will go to third-party candidates. Of the main battleground states, I’m picking Hillary to win Florida, North Carolina, and Nevada. I’m giving Iowa, Ohio, and Arizona to Trump. However, it shouldn’t surprise anyone if Hillary ends up winning one of these three states — or all three of them assuming she has a great night. My full electoral map can be found here.
I have not had time to run simulations of the Senate, so I’m just relying on the betting markets and other forecasters for the Congressional predictions. I think the Democrats will pick up the Senate, just barely, by winning in Nevada, Missouri, New Hampshire, Wisconsin, Pennsylvania, and North Carolina. I think the Republicans will retain the House, but the Democrats will gain about 15 seats — give or take two seats.
What do you all think will happen tomorrow?