Every columnist, I suspect, has a hidden longing for social and political chaos — turmoil of a kind he might never actually advocate. I am no exception. The brute economic incentives of our trade require that we have frightening phenomena to interpret and to reassure about. But it is not just a question of incentive. Daydreaming about the world going to heck seems to be a fundamental, or at least very common, human impulse.
Look how much of the stuff in newspapers and magazines involves calamity scenarios whose failure to materialize — over years, generations or centuries — would seem to have shamed their vendors. People keep warning of naval war between China and some neighbour, or a sudden brutal collapse in the Canadian housing market, or violent global conflicts over drinking water, or the instantaneous, permanent uselessness of antibiotics. Why don’t these cries of “wolf!” eventually induce widespread reader fatigue and annoyance? Maybe it’s because they are not forecasts in the sense that the weather report is. They are a form of speculative fiction.
I mention this because I keep hearing that the U.S. presidential election scheduled for Tuesday is a close race — a “dead heat,” in the words of one of our own recent headlines. When I hear this, I turn back to my own methods of understanding a U.S. election, and over and over I come up with the same answer: Republican presidential candidate Donald Trump’s chances are probably better than Mike Dukakis’s in 1988, but not much better. The stream of “dead heat” headlines is being fuelled by unabashed cherry-picking of individual polls, sometimes not even national ones.
Thursday morning, to take the latest example that made me frown, The Economist made a big deal out of a poll showing that Democratic presidential candidate Hillary Clinton and her rival, Trump, are only 0.1 percentage points apart in Florida, implying that the state tends to decide presidential elections by itself, and therefore … the nauseating ellipsis hangs threateningly. In fact, Florida is one of a small group of big states that typically decide elections, but they do it as a group. When I watch U.S. election coverage, I tend to think of a mini-tournament called “FlOhPa” — Florida-Ohio-Pennsylvania. And in my head, Michigan is perched a little way downstage — the Ringo of the quartet.
On the current map, Trump merely being level with Clinton in Florida would appear to be pretty bad news for him. Pennsylvania, where he ought to be competitive if he is to have a real chance, has tumbled well out of reach: Clinton looks to be up three to five points there and one recent state poll has her 11 points up, if you enjoy cherry-picking. Both major-party campaigns are said to be targeting Michigan in upcoming days, and Michigan is a volatile, socially divided state where pollsters’ assumptions about turnout will count for a lot. (In plain English: it’s Canada, but with one huge city that is 80 per cent black.) Thing is, pretty much all the pollsters say Clinton is about six or seven points ahead in Michigan. It looks as though they are, as a group, in unusually close agreement on that.
In an ordinary election, Trump visiting Michigan would betoken the professionally chosen strategy of a desperate long shot. We have just got used to not regarding Trump’s campaign as in any way professional, even though he has locked down core Republican states and made the overall electoral map look a lot more like the ones Mitt Romney and John McCain faced. In the sense that it won’t be a Reagan-Mondale blowout going against the Republicans, the election certainly is now “close.” Clinton probably does not need Florida to win the Electoral College — but, as it happens, she is ahead there by a point or two anyway, if you look at the median of the recent polls.
I could invent reasons to be skeptical of my own evolved, common-sense approach to anticipating a U.S. presidential election. But that is the point: I would have to invent them. For awhile there, it seemed as though Gary Johnson’s Libertarian candidacy was going to start influencing swing states, making the overall Electoral College count less certain. That moment, alas, evaporated: the political scientists’ Sixth Party System prevailed.
And it looks like the Federal Bureau of Investigation’s late reopening of the investigation into Clinton’s State Department emails is not going to have an observable national effect on the polls. It is a question mark hovering in the void, to be addressed after a super-polarized electorate expresses deep, sturdy emotional preferences that have not really changed in six months. Some Americans will certainly cast a defensive or a feminist vote for Clinton, while harbouring the specific hope that she is ousted soon.
There are a lot of complicated things happening all at once in the news environment. The increasing prominence of poll aggregators has probably led to some deliberate gaming of their approach by political operatives, which is probably not difficult. Automated aggregating systems are definitely being fed a barrage of mystery-meat data from the marketing company SurveyMonkey, and since the systems are generally predicated on a Bayesian philosophy of regarding even bad data as valuable, it is unclear whether they are designed to respond appropriately. But it is not really the aggregators that are doing the most to fabricate the perception of a close election. It’s just the good, old media.
Related
- Michelle Hauser: The American electoral map is turning a certain shade of pink
- Dick Morris: James Comey’s bombshell, and what it means for Hillary Clinton’s electoral chances
National Post
ccosh@postmedia.com
Twitter.com/colbycosh
No comments:
Post a Comment