Friday, November 11, 2016

Fact-checking fact-checkers

I've seen some gloating Trumpkins say the mere fact that their hero won "delegitimates" his critics. For instance, that pundits like Michael Medved and Ben Shapiro shot their "credibility" by opposing Trump. 

i) There are political prognosticators by trade. When they make the wrong call, that injures their credibility because that's how they make their living. It's an argument from authority, like expert witnesses. So, for instance, Nate Silver's star has waned. He was wrong about Trump in the primaries, and wrong about Trump in the general election. 

ii) However, even when people like Michael Barone are wrong, they can still be useful for post-election analysis. Although they may be unreliable in advance of the fact, they can explain after the fact why a candidate won or lost. There's more to Barone than prognosticator. He has immense knowledge of American political history and demographics. 

iii) Why do they get it wrong? I notice that political prognosticators often rely on historical precedent. The assumption that the future resembles the past. 

Sometimes this amounts to little more than numerological superstition. That if something hasn't happened in 100 years, it's unlikely to happen this time. The fallacy is to generate cumulative probabilities from causally independent events. 

iii) Then you have partisan operatives who mask themselves as nonpartisan "fact-checkers". It's funny how many people are taken in by that ruse. Just call yourself a "fact-checker," and people let down their guard. 

If, however, it becomes apparent that they are rooting for one team, then they lose credibility since their ostensible raison d'être was to play the role of impartial referees who aren't betting for one team or another. An argument from authority that collapses if the authority is suspect.

iv) By the same token, gov't agencies like the DOJ and FBI lose credibility when they become apparatchiks. Likewise, debate moderators who give one side advance knowledge of the questions. 

v) People who endorse candidates can lose credibility if they put their reputation on the line to vouch for the candidate's bona fides. That's another argument from authority. A high-profile figure offers himself as a character witness for the candidate. But that's risky. If the candidate proves to be untrustworthy, then the person who endorsed him looks like a dupe. That does delegitimate them. 

vi) Finally, credibility is important in a candidate. If you can't trust him to try to keep most of his campaign promises, why vote for him? 

Credibility is important in that context because it's about the future. How he will act in the future. The voter doesn't know the future. He doesn't have direct access to the candidate's intent. He can't independently verify or falsify the campaign promise–unless it's obviously unworkable. 

v) By contrast, "credibility" is the wrong yardstick to measure pundits like Medved and Shapiro. They aren't asking the audience to take their word for it. And electoral predictions aren't their stock and trade. 

Rather, they reason for their positions. They offer evidence for their positions. It's not an argument from authority. 

No comments:

Post a Comment