In his article "PolitiFact is All Wrong About Clooney-Clinton Fundraiser," Miller gives two examples of the bias harming Sanders.
First, Miller questions PolitiFact's revised "Half True" rating (down from its original "Mostly True" rating) of George Clooney's claim that a fundraiser he attended would mostly help down-ballot Democrats:
I suppose we could really use a logical approach to understanding the rating system. What parts of Clooney’s statement is true?Miller goes on to conclude that both main elements of Clooney's statement were wrong, and that the "Half True" rating remains way too generous to Clooney.
Clearly not “overwhelming amount”, since only 1% went to the states. That qualifier on the word “money” was the point of his statement, though; it isn’t even correct if you change it from “overwhelming amount” to “most of”; or even “a lot of”. It’s only accurate if you change it to “a little bit”. His whole statement only has persuasive power if the vast majority of the money is going to downballot races, and it’s clear that that is not true.
Next, Miller hits PolitiFact for its rating of Sanders' claims that national polls show him performing better than Clinton does against Donald Trump.
PolitiFact rated Sanders' claim "Mostly True" after finding Sanders representation of the poll results was right.
Miller (bold emphasis Miller's):
Apart from the disingenuousness of saying “yes it’s true but we don’t think that’s what matters”, everything in their “additional context” itself requires a fact-check. Politifact didn’t bother to cite or quote their “polling experts”, or say what they mean by “well before”. Sanders does a lot better than Clinton against Trump right now, and that is relevant to November. Their justification is loaded with opinion and presumption, and is far from the simple checking of the numbers that it could and should be.Both of Miller's criticisms point out real problems with PolitiFact. But there's a problem. Miller apparently doesn't realize that he's criticizing PolitiFact's standard operating procedure. PolitiFact's history of ratings is littered with many examples of false-but-true and true-but-false ratings. PolitiFact's justifications are often "loaded with opinion and presumption."
And does it really pass the sniff test that PolitiFact could be biased against Sanders to Clinton's benefit but not show a similar bias against Clinton's GOP opponent in the general election?
Miller makes the point that PolitiFact's parent paper, the Tampa Bay Times, endorsed Clinton.
Politifact’s parent paper, the Tampa Bay Times, endorsed Clinton. Others have shown their unfairness toward Sanders. It’s laborious to make each case, since their ratings system must involve some subjectivity, and you have to do a lot of work to fact check anything (hence the value in having a neutral fact-checker, like Politifact aspires to be, and most of the time is).It's true the Times endorsed Clinton. But the Times also endorsed Jeb Bush in the Republican primary. Is that a good supporting argument for the idea that PolitiFact's coverage is weighted in favor of Bush against the other Republican candidates?
What would Miller say if he knew that the Times has never in its entire history endorsed a Republican presidential candidate in the general election, going back over 100 years?
Most of the time a neutral fact-checker?