Would Trump Have Won So Bigly Under Alternative Voting Systems?

Hillary Clinton won over 2.8 million more votes than Donald Trump. But, Trump is President-elect because he won 57% of Electoral College delegates. 
http://www.cnn.com/election/results

Delegates are awarded to the candidate who wins the most votes in each state. Trump did a better job targeting his campaign on large states with close contests. Trump’s 46% of the national vote was distributed state-by-state more effectively than Clinton’s 48%.

Meanwhile, back in my native Canada the House of Commons Special Committee on Electoral Reform released a report on December 1^st.

What does Trump’s triumph tell us about the voting reform options under consideration in Canada?

The Canadian and American systems are not carbon copies, but have much in common. In American elections, all Electoral College delegates go to the candidate who wins the most votes in a state, no matter how small the margin. In Canada, our first-past-the-post (FPP) method awards a House of Commons seat to the candidate who wins the most votes in a constituency.

The winner-take-all aspect of the American and Canadian systems has similar effects. In the US, Trump turned 46% of the national vote into 57% of Electoral College delegates. In Canada, Stephen Harper in 2011 and Justin Trudeau in 2015 both won control of 54% of House seats even though their parties had less than 40% of total votes. (See footnote 1 at end of text.)  

How could voting reform reduce the chances in Canada of a party with minority support winning a majority government? The solution is a voting system that measures popular support better than FPP.

The House Committee on Electoral Reform consulted about two alternatives.

The first is proportional representation (PR) used in most of Europe. Under PR, a party’s proportion of legislative seats is close to its percentage of total votes. If the Conservatives, Liberals and New Democratic Party (NDP) won 40%, 35% and 25% respectively of votes in a future election, the Conservatives would get 40% of the seats, the Liberals 35% and NDP 25% under PR. (See footnote 2.) After the election, parties would negotiate with each other searching for enough in common to form a government with majority support.

The other option is the preferential voting (PV) system used in Australia. Under PV, a voter can rank local candidates. If no one wins 50% of first-choice votes, the candidate with the fewest first-choices is dropped. Ballots are recounted based on second-choices until a candidate reaches 50%.

What does the American election tell us about these alternatives? Trump would have won even under PV.

Exit polls showed a slight advantage for Clinton when the 6% who voted for minor candidates were asked to choose between Trump and Clinton. Michigan is the only state that might have switched to Clinton under PV. Trump would have won an electoral college majority even under Australian rules. (See footnote 3.)

What about PR? Trump and Clinton would have tied with 267 delegates each -- 3 short of the 270 needed to win. 

Under PR, Trump and Clinton would have had to make deals with minor-party leaders. Libertarian Gary Johnson shares Trump’s isolationist approach to foreign affairs and would likely have instructed his 2 delegates to support the Donald.

If Green candidate Jill Stein and independent Evan McMullin had each sent 1 delegate to Clinton, a 269-269 electoral college tie would have sent the election to the House of Representatives. The Republican House would elect Trump. 

Candidate	Number of Electoral College Delegates
	Actual Result: First Past the Post (FPP)	Simulated Result: Preferential Voting (PV)	Simulated Result: Proportional Representation (PR)
Donald Trump	306	290	267
Hillary Clinton	232	248	267
All Others	0	0	4

Note: The above simulations are not intended as precise predictions of what the results would have been under PV and PR. Politicians campaign differently when second choices matter under PV or when all votes in all states count under PR. The direction of change is what is important in the above table. It is very likely that Trump would have won significantly fewer delegates under PR. His share of total delegates would have been much closer to his share of the national vote.

Why would Trump still be tied in delegates with Clinton under PR even with 2.8 million fewer votes than Clinton? The electoral college is tilted to favour states with small populations, most of which are Republican. This tilt in the electoral college cannot be fixed easily, as it was part of the bargain when the former British colonies formed the United States.

Yes, Trump would likely have won even under PR. But, the message for Canadians is that a PR election would have been more difficult for Trump. (See footnote 4.) 

One American election cannot furnish statistically significant proof that PR would be better than FPP for Canadians. But, we will never have the controlled experiment conditions needed for conclusive evidence about which voting system is best. Trump’s victory in the United States adds to what we already know from our experience in Canada about the shortcomings of winner-take-all FPP. Let’s act now on voting reform in Canada before it’s too late.

Footnotes:

1. The risk is even greater in Canada that a government can be elected without majority support. The top two Canadian parties, Liberals and Conservatives, combined to win 71% of the total vote in our 2015 election. By contrast, the top two American parties, Democrats and Republicans, just took 95% – a bit below the recent average. With more parties as serious players with significant support, the potential is even greater in Canada for the distribution of House of Commons seats to deviate from party shares of the national vote.

2. In proportional representation (PR) examples with more parties and more complicated rules such as minimum threshold requirements to win seats based on either national or regional vote shares, the end result would still be party seats broadly in line with vote shares.

3. Relative to first-past-the-post (FPP), preferential voting (PV) would not always benefit left-wing parties at the expense of right-wing parties. British Professor John Curtice’s simulations show that Margaret Thatcher’s Conservative Party would have won majority mandates in 1983 and 1987 even under PV. At that time, the British Labour Party would have lost seats under PV. http://news.bbc.co.uk/nol/shared/spl/hi/uk_politics/10/alternative_vote/alternative_vote_june_09_notes.pdf In Australia, the Liberal Party has been the natural party of government winning most national elections since PV was introduced in 1919. The Australian Liberal Party traces its roots to 19^th century, free trade liberals and is that country’s right-wing mainstream party.

4. One objection to PR is that the Nazis rose to be the largest party in Germany in PR elections. Hitler was then invited to take over as Chancellor. The Nazi example demonstrates the institutional weakness of the German Weimar Republic, not the failure of PR. Hitler became Chancellor in 1933 with one-third of the seats in the Reichstag. Hitler then moved immediately to dictatorship and never had to build the coalition necessary under PR. Neither courts, police nor military acted to defend democracy.