For misleading headlines, it’s hard to beat this one from the AP yesterday: “Exclusive: 4 in 5 in US Face Near-Poverty, No Work.” Sounds like a cultural and economic Armageddon in the making. But the Wall Street Journal’s James Taranto demolishes the whole thing in his inimitable style.
It turns out that the headline left out a key phrase: “in their lifetimes.” If you have ever, for even a very short period of time, been out of work or on some sort of assistance program, then you’re in the 80 percent who face near-poverty and no work. “Near-poverty” is defined as having less than 150 percent of a poverty-level income. As Taranto points out, that’s like saying that a man who is 8 feet 9 inches tall is of near average height.
In other words, this headline is on a par with one that reads, “Half of Americans Have Below-Average IQ.”
But Fox News’s normally sensible Special Report with Brett Baier led with this story last evening and never questioned its statistics.
What scientists call “confirmation bias”—the tendency to unquestioningly accept results that confirm one’s hypothesis—lead us to not question statistics that support our political philosophy. Politicians (and far too many editorial and op-ed writers) fall victim to confirmation bias. Others know that such statistics are rarely challenged and use them to lie to the public.
So I have a suggestion. Every serious newsroom in the country should have on staff someone whose job is to vet statistics for intellectual honesty and statistical rigor. Do they compare apples with oranges? Do they use a misleading baseline? Do they manipulate the shape of a graph to make things look good or bad? Are the definitions valid? Was the poll sample properly chosen?
Such automatic second guessing would greatly improve the level of public discourse. But I won’t hold my breath waiting for it to become standard journalistic procedure. Misleading headlines can sell a lot of newspapers.