There have been two ‘wrong winners’ in UK general elections over the last 70 years. In 1951, Labour got more votes than the Conservatives but the Tories won more seats. In February 1974, the position was reversed, with the Conservatives winning more votes but Labour gaining a higher number of seats. This is always a possibility with Westminster’s First Past the Post (FPTP) electoral system, where the distribution of votes across different types of seats can sometimes skew the overall result.
In the US, we have seen the same number of Presidential ‘wrong winner’ elections since the turn of the century. The Democratic Party candidate for President has won the popular vote in four of the last five elections but has won the Presidency on only two of those occasions.
In 2000, Al Gore won around half a million more votes than George W. Bush, across the whole of the United States but Bush won the Presidency. In 2016, Hillary Clinton’s popular vote victory over Donald Trump was even larger, at nearly three million votes, yet it was Trump who emerged as President.
The reason for these ‘wrong winner’ elections is the way that the votes of US citizens are turned into votes at the Electoral College. We’ve previously explored the history of the Electoral College. For our purposes, we simply need to note that each state (and the District of Columbia) is allocated a number of Electoral College votes, roughly in proportion to their population. The number of Electoral College votes varies from 55 for California to three each for Alaska, Delaware, District of Columbia, Montana, North Dakota, South Dakota, Vermont and Wyoming.
The vast majority of states apply a winner-takes-all system, whereby the candidate who wins the most votes in a state are assigned all of that state’s Electoral College votes. A closer look at the 2016 Presidential election result highlights how, because of this, Trump was able to beat Hillary Clinton, despite losing by nearly three million in the popular vote. The key states in 2016 turned out to be Wisconsin, Michigan and Pennsylvania. Three states in the upper mid-west of the United States, which had voted for the Democratic candidate in every Presidential election since 1992 (Michigan and Pennsylvania) or 1988 (Wisconsin). Trump was able to outperform his popular vote performance in these three states and flip them in the Republican’s favour, winning all of them by wafer-thin margins of less than one percentage point.
Overall, Clinton beat Trump by two percentage points (48.1% to 46.0%). However, in these crucial states, Trump slightly outperformed his overall vote share and Clinton slightly underperformed hers. Trump won Pennsylvania (48.2% to 47.5%), Michigan (47.5% to 47.3%) and Wisconsin (47.2% to 46.5%). Taken together, these three states account for 46 Electoral College Votes – Pennsylvania (20), Michigan (16), Wisconsin (10). If Clinton had won these three states, she would have won the Electoral College and become President. Like the UK’s FPTP system, this highlights how under the winner-takes-all Electoral College system, the distribution of votes across particular geographic areas can be crucial.
Given that a ‘wrong winner’ election has happened twice in the last twenty years, it is worth exploring the likelihood of it happening again this year. On the face of it, it would appear unlikely. As of 30 October, the FiveThirtyEight website forecasts that the Democratic Party candidate, Joe Biden, will receive 53.3% of votes and Donald Trump 45.4%. This would be a far higher overall victory margin of 8 points, compared to Hillary Clinton’s 2-point margin in 2016. FiveThirtyEight are also forecasting a Biden victory in the Electoral College, by 346 votes to 192. If this did turn out to be the outcome, it would represent a comfortable victory, along the lines of Barack Obama’s 2008 win over John McCain, both in terms of the popular vote and the Electoral College.
However, a closer look at FiveThirtyEight’s forecasts reveals how there is still a potential mismatch between the popular vote and the Electoral College. While Trump’s chances of winning the popular vote are put at a mere 3%, as of 30 October, his chances of winning the Electoral College are put at 10%. Pennsylvania again looks like being a key state, with FiveThirtyEight forecasting it to be the ‘tipping point’ state, the point at which Biden would have enough Electoral College votes for victory. As of 30 October, Biden is forecast to have a 5-point lead in Pennsylvania (52.2% to 47.1%), slightly but notably lower than his forecast 8 point overall lead. When you take into account that FiveThirtyEight’s final forecast for Pennsylvania in 2016 had Clinton 3 points ahead (48.9% to 45.2%), you can see why it would be foolish to say another ‘wrong winner’ election in 2020 is impossible.
Although a ‘wrong winner’ election this year looks unlikely, though not impossible, if the trends that have led to two ‘wrong winners’ in the last 20 years continue, it may well be the case that calls for reform of the Electoral College system will be increasingly heard.
Image Donald Trump av Gage Skidmore. CC BY SA 2.0
Find out more about the campaign for reform of the electoral college