Much ink has been spilled trying to explain the surprise election of President-elect Donald Trump. This post isn’t meant to make any comment or provide analysis regarding the pros or cons of the election but rather to empirically clarify what actually happened since much of the existing analysis has oversimplified a rather complicated election result.

As of the time of writing (Nov. 15), Secretary Hillary Clinton (hereafter simply referred to as Clinton) won the popular vote with 61,324,576 votes versus Mr. Trump’s 60,526,852. Too much has been made of the popular vote since presidents are not elected by it in the U.S. Trump handily won the Electoral College 290 to 232, which actually decides the presidential election.

This is the third straight decline in the number of votes received by the Democratic presidential nominee. In 2008, then-Senator Barack Obama received 64,498,516 votes, defeating his GOP rival Senator John McCain. President Obama was re-elected in 2012 but with 2.9 per cent less votes (1.9 million). Clinton’s 61.3 million votes is a further decline from President Obama’s 2012 performance of 1.3 million votes or 2.1 per cent.

Of the 50 states, Clinton’s vote total was higher than President Obama’s 2012 performance in 11 though one could argue that Arizona, South Carolina and Virginia were essentially the same (Clinton had 0.3 per cent, 0.4 per cent, and 0.6 per cent more votes, respectively). Clinton under-performed President Obama’s vote totals in the remaining 39 states.

Some of Clinton’s state-level improvements mattered. For instance, Clinton made Florida more competitive by improving on President Obama’s 2012 performance. However, some of the run-up in votes didn’t really matter because it occurred in states such as New York and Massachusetts, which are both Democratic strongholds or in states such as South Carolina or Utah that are equally as strong Republican states.

Trump essentially won the election by winning the states Mitt Romney won in 2012 and flipping another six states that in 2012 supported President Obama: Florida, Iowa, Michigan, Ohio, Pennsylvania and Wisconsin. An interesting question regarding these six states is the degree to which Trump actually won them versus Clinton losing them. The table below shows the presidential vote count for both the Democratic and GOP candidates for both 2012 and 2016 for these six states.

Florida is an example where Trump clearly won. He recorded a 10.7 per cent increase over Romney’s total in 2012. Secretary Clinton’s 5.9 per cent improvement over President Obama’s 2012 performance simply wasn’t not enough to overcome Mr. Trump’s vote surge in the state, though it made it more competitive.

Alternatively, Clinton clearly lost Wisconsin. Trump’s vote (1,409,467) was almost identical to Romney’s vote (1,408,746) in the state four years prior. The state went for Trump because Clinton’s vote collapsed. Her vote count of 1,382,210 was 14.4 per cent lower than President Obama in 2012.

The question of who won and who lost the remaining four swing states is more complicated. First and foremost, Clinton underperformed in delivering Democratic votes in all four states. Specifically, the vote for Clinton relative to President’s Obama’s 2012 vote declined by 20.3 per cent in Iowa, 11.5 per cent in Michigan, 14.1 per cent in Ohio, and 2.2 per cent in Pennsylvania. To some extent, then, it’s Clinton who lost the election by not delivering voters in these key states who had previously supported the Democratic nominee.

Second, Trump improved on the GOP’s performance compared to 2012 in all four states. Specifically his vote count compared to 2012 improved by 9.8 per cent in Iowa, 7.9 per cent in Michigan, 6.9 per cent in Ohio, and a stunning 11.2 per cent in Pennsylvania. It was, therefore, a combination of both Clinton under-performing 2012 and Trump over-performing compared to 2012 that explains the swing in these four key states.

Perhaps more illustrative, though, of Trump’s improvement is comparing his vote performance (2016) in these six swing states against President Obama’s performance in 2012. President Obama’s performance in 2012 compared to Trump’s 2016 performance would have resulted in three states that Trump won being retained by the Obama Democrats. Specifically, Mr. Obama’s vote totals would have retained Iowa, Michigan and Wisconsin. That means, however, that Trump would still have won Florida, Ohio and Pennsylvania, which would have resulted in a Trump presidency in 2012, though narrowly with 273 Electoral College votes versus President Obama’s 265.

If the same exercise is replicated in these six states with President Obama’s initial 2008 vote count he wins the presidency by claiming victory in Ohio and Pennsylvania as well as the three states from the 2012 analysis. Put simply, part of the explanation for the Trump victory in 2016 and hypothetically in 2012 is the reduction in the voting for the Democratic nominee rather than a surge in voting for the Republican nominee.

As citizens and particularly those active in the political marketplace digest the 2016 results and begin looking forward to 2018 and 2020, it’s key to understand what actually explains the Trump victory as well as the Clinton defeat.