Random Events in Election Campaigns
Posted on: November 15, 2016 /
A persuasive post
asserts that random events were the most important events in the US election, far more significant than after-the-fact theories about brilliant or flawed political strategies: giving two examples, two key essentially random events relate to revelations from a sex tape and then from a pedophile texter.
Canadian election campaigns have such random events, with perhaps less salacious significance. Consider the event and the reaction to the fumbled football photo in the 1974 election campaign: unpredictable yet significant. Consider, more recently, the impact during the 2006 election campaign of the RCMP announcement of a criminal investigation into possible insider trading on income trusts prior to a government announcement on the subject. This latter was a pivotal determinant of the election’s outcome and, unless you assert egregious political interference by the RCMP, also to all intents and purposes random as to timing.
“Events, dear boy, events”
feared by Prime Ministers,
according to Harold MacMillan
Nonetheless, I believe it is too strong to assert that post hoc analyses of election outcomes are an arbitrary and unjustified imposition of a framework on essentially random events. Understanding the voter’s mind may be very speculative, but the structural underpinnings of an election are just as important: turn out matters; demographics matter. And I would say that character matters – perhaps in the wrong way.
With respect to turn out, 2106 voter participation fell in the US Election, compared to 2012. The causes are no doubt many, but it appears that Mr. Trump was elected President in 2016 with substantially fewer votes than gained by Mr. Romney in his loss to President Obama. The reasons for the fall in participation can be left for the political theorists (they could include those random events). However, it is indisputable that the Democratic vote dropped far more than the Republican (1).
With respect to demographics, there was a race-based split in voter preference. Whites favoured Mr. Trump by a strong margin (in the 60% range), Blacks and Latinos favoured Ms. Clinton by an enormous margin (in the 80% range). Women and lower income earners tended to favour Ms. Clinton; the reciprocal for Mr. Trump. I believe it is clear that the structure of the vote is more significant than random events in the campaign.
Assessments of a candidate’s character are far less objective than statistical descriptions of voters and their turn out. However, character assessments were a very strong theme in the campaign, and mostly negative about the other candidate. My assessment is that the charges against Mr. Trump were far more extensive (rapist, sexist, racist, bigot, bully, swindler, etc.) than those against Ms. Clinton (deceptive incompetent crook). My further assessment is that the various charges were essentially self-cancelling, that they brought both candidates down, but not one much more than the other.
However, in other cases, character matters. Here follows two Canadian Conservative Party examples, one illustrating good character, one not.
Robert Stanfield allowed the minority government of Lester Pearson to survive after it was defeated on a tax bill. Pearson enlisted the aid of the Governor of the Bank of Canada to persuade Stanfield that the fall of the government would have disastrous economic consequences. ‘Twas a dubious argument, but Stanfield accepted the word of the Prime Minister and the Governor – to his political cost. Pierre Trudeau became the new Liberal leader and the opportunity was lost.
Stephen Harper sought to prorogue Parliament, only six weeks after an election, rather than face a vote of non-confidence on a fiscal matter. Harper attacked the prospect of a vote as a political tactic by the opposition to gain unelected power, and sought to erase the whole non-confidence process by proroguing Parliament. Although Harper’s motives appeared entirely political, the Governor General allowed his request and the administration survived. There were no apparent electoral consequences for Mr. Harper, as the next election gave him a majority in Parliament.
Alas, in these examples, good character is not rewarded and poor character is. But character matters.
So the outcomes of elections depend on far more than random events. That said, theories that purport to understand the minds of the voters (until there is data on the matter) are indeed suspect. Nonetheless, there is enough data out there to see the workings of an election campaign.
(1) Although the Democratic vote was still numerically greater than the Republican.