Time to cut loose

The odds on the Nov. 3 election for the US presidency haven’t changed since last week, when the opinion poll aggregator Fivethirtyeight.com gave the Democratic Party’s Joe Biden an 88 percent chance, and the Republican Party’s Donald Trump only a 12 percent chance, of winning. Polls pick the winner 79 percent of the time, says 538, citing the average proportion in its pollster database, from 1998 onward, of correct calls of the winner in the final 21 days before elections.

Being correct on the winner depends, of course, on the size of the leader’s lead over the next candidate. For national leads of 20 percentage points or more in the last three weeks before the election, the polls in 538’s archive have been 99 percent correct. For leads of over 10 points, they have been 93 percent correct. For spreads of 6 to 10 points—which is Biden’s present lead over Trump—the polls have been 84 percent correct.

Another observer I respect is Claire Durand, past president of the World Association for Public Opinion Research and professor at the University of Montreal, who writes ahlessondages.blogspot.com (“survey” is sondage in French, encuesta in Spanish, and umfrage in German). In her latest blog “Polls of 2020—Where are we now?” (10/29/20), professor Durand points out that Biden now has around 55 percent of voting intentions, whereas Trump has 45 percent. Ever since the first debate, Biden’s score has been stable or, at worst, down by only 1 point.

This blog, intended for polling professionals, focuses on the effect of survey mode, which is whether the poll is done by live interviews, or by the web, or by IVR (interactive voice response, for example: “If you intend to vote for X, press 1; if you intend to vote for Y, press 2,” and so forth), or by a mix of these modes.

She says that web polls estimate the support for Trump 1.9 points lower than IVR polls, and 1.6 points lower than live interviewer polls; mixed-mode polls get an estimate 1.2 points higher. I read this as an assurance that the choice of mode does not affect a candidate’s score very much, and so does not change the expectation of the polling profession that Biden will win over Trump.

Meanwhile, what are world leaders’ positions regarding the US elections? The other day, I thought I heard presidential spokesman Harry Roque say on TV that President Duterte does not want to get involved in this matter, but is leaving the outcome up to the American electorate.

If so, then Mr. Duterte has apparently changed his view from last February, when his statement that Trump deserves to be reelected got him listed this week as one of a few world leaders who openly favor Donald Trump (Miriam Berger, “Ahead of election, a handful of world leaders voice support for Trump,” Washington Post, 10/28/20).

The other leaders are Prime Minister Viktor Orban of Hungary, President Jair Bolsonaro of Brazil, President Aleksandar Vucic of Serbia, and Prime Minister Janez Jansa of Slovenia (birthplace of first lady Melania Trump). Neither Mr. Duterte, nor any of these politicians, can influence the election anyway; why get on the wrong side of the next occupant of the White House?

As to the partisanship of Filipino Americans, my guess is that they lean more toward Democrats than Republicans. I think the business entrepreneurs among them (understandably pro-Trump) are a minority. The Democratic Party is generally the kinder one in terms of immigration policy toward Filipinos. Recently I saw a CNN clip of Biden extending profuse thanks to the Fil-Am health frontliners for their courageous work in the current pandemic. October 2020 happens to be Filipino American History Month; I think the people in its Facebook posts are mainly Democrats.

——————

Contact mahar.mangahas@sws.org.ph.

Read more...