Of the numerous reasons why most observers on both sides expected Hillary Clinton win the 2016 election heading into Election Day, perhaps the most glaring was the fact that she had seemingly already won the state of Florida based on early-voting alone. It’s why it came as such a shock when Donald Trump somehow pulled off the upset in the state by one percent of the vote, a difference which would have swung the entire election. But in yet another piece of evidence that the voting tallies may have been rigged, a closer examination of the early-voting numbers suggests that Trump’s victory in Florida wasn’t just unlikely – it was mathematically insurmountable.
Here’s what we know about Florida for sure: of the 9.1 million people who voted in the state, roughly seventy percent of them – or 6.6 million people – voted early. The early voters consisted of only slightly more registered Democrats than registered Republicans. However, the most prominent exit poll of Florida early voters showed that 28% of registered Republicans had voted for Hillary Clinton, with just 6% of registered Democrats having voted for Trump. Republican strategist Mike Murphy expressed skepticism that the number could have been that high, but even he pegged it as being closer to the mid-teens.
When you do the math, you see that Hillary Clinton had such a massive early voting lead in her pocket in Florida that – depending on which end of the crossover vote totals you want to lean toward – it turns out Donald Trump would have needed to get somewhere between 59% and 71% of the Election Day voting in the state in order to catch up. Here’s the breakdown, based on numbers available from the state of Florida’s official early vote tallies:
If the early voting exit poll I linked to two paragraphs above is correct, it means Hillary Clinton got 94% of the registered Democrat early vote, or 2472223 out of 2630025 votes. It means she also got 28% of the registered Republican vote, or 709418 out of 2533635 votes. The exit polling didn’t reveal how early voters from neither party broke down. But let’s assume the 1443922 independent early votes were split down the middle between Clinton and Trump. In this we’re giving Hillary 721961 out of the 1443922 independent votes. I can make a demographic argument for why registered independents tend to skew younger and therefore may have been more likely to vote for Hillary, but I’m going to stick with an even split.
That gives Hillary an early vote total of 3903602 votes, with Trump getting the remaining 2703890 early votes (his early vote tally being based on 72% of registered Republicans, 6% of registered Democrats, and 50% of registered independents). That means Hillary received 59% of the total early vote in Florida.
This was across roughly seventy percent of all votes cast in the state, meaning that based on early voting along, she already had (59% of 70%) a whopping 41.3% of all the votes she needed to win the state – and she only needed to get another 8.7% of the total on election day to reach the 50% mark. In turn that means that she would have needed just 29% of the election day voting, which represented a total of 30% of the vote, in order to get the remaining 8.7% she needed. In other words, in this model, Trump would have needed to get 71% of the election day voting in order to catch up and get back to even.
If we’re to instead use Mike Murphy’s on-air estimate on MSNBC that a mid-teen number of republicans crossed over to vote for Hillary Clinton in Florida, which we’ll formalize as being 15%, the above numbers shift accordingly. Since he didn’t provide any estimate of Democratic crossover for Trump, we’ll once again go with 6% for the sake of consistency, and we’ll stick with each candidate getting 50% of the independent vote. These numbers work out to Hillary getting 54% of the early voting, or based on seventy percent of all votes having been cast early, 37.8 of all the total votes cast. That means on election day she would have needed another 12.2% of the total vote, or 41% of the election day vote. In other words, even in this model, Trump would have needed 59% of the election day vote in order to catch up and get back to even.
So these two models, one based on Florida exit polling and the other based on the expert opinion of a fair minded Republican political operative, say that Donald Trump would have needed somewhere between 59% and 71% of the election day vote in Florida in order to overcome Hillary Clinton’s early vote lead and tie things up in the state. Democrats are somewhat more likely to early vote and Republicans are somewhat more likely to wait until the day of voting. And both of my above models are based on Trump simply coming back to tie things up in Florida. For him to have won the state by one percent of the vote, as claimed, he would have needed around 62% to 74% of the election day vote, an even steeper hill to climb. But while that’s mathematically possible in the theoretical sense, it’s mathematically insurmountable in any real world scenario.
So what the hell happened? That’s a good question. Despite the exit poll documentation showing just how many registered Republican Cuban-Americans were crossing over to vote for Hillary Clinton due to Donald Trump’s extreme anti-Hispanic policies, the official vote tallies in Hispanic dominated Miami-Dade County insist that Hillary only got around the same percentage of the vote in 2016 that Barack Obama got in 2012 despite the fact that his opponent Mitt Romney was not seen as being overtly racist. Trump’s impossible win in Florida was just one of numerous mathematical unlikelihoods and impossibilities around the nation that I’ve documented thus far.
Bill Palmer is the publisher of the political news outlet Palmer Report