Today Crooked Timber points to various reasons not to trust The Economist. CT's been down this road before; see this 2006 post, which highlights an instance of The Economist's astonishing cluelessness about American politics:
This piece on the demise of Mark Warner’s and George Felix Allen’s respective president hopes is a case in point. Most of the article is pretty unexceptionable. The peculiar bit is this summation of the current state of the race for the Democratic presidential nomination.
But whatever the reason, [Warner’s] retreat has created a vacuum. He had positioned himself as the centrist alternative to Hillary Clinton, the early front-runner for the Democratic nomination and the darling of the party’s liberal activists. Southerners, Westerners and moderates are now shopping for a new candidate, perhaps Governor Bill Richardson of New Mexico or Governor Tom Vilsack of Iowa, Senator Evan Bayh of Indiana or former Senator John Edwards of North Carolina, the vice-presidential nominee in 2004.So Hillary Clinton is apparently the "darling of the party’s liberal activists." ... [deletia] ... Equally bizarre is the suggestion that centrists might want to gravitate towards John Edwards. This could just be the result of sloppy thinking that telescopes “Southerners, Westerners and moderates” into a unified category, but to the extent that Edwards might appeal to Southerners and Westerners, it’s not because he’s a moderate. It’s because he’s running the most economically populist campaign that a serious candidate for the Democratic nomination has run in recent history. These claims don’t seem biased to me so much as clueless.
Now, if this magazine, whose primary readership is the educated upper middle class in the US and UK, cannot understand basic facts about politics in the US — an English-speaking nation with an open, internationally distributed press, and where 47% of its readers live (cite) — how can you believe anything they say about, say, Kyrgyzstan or Qatar or China?
Brad DeLong rounds up some reactions to the above CT post. A few months earlier, of course, he had called new Economist editor John Micklethwait possibly the stupidest magazine editor alive.
Then there's this James Fallows piece in The Atlantic from 1991.
So basically The Economist is unreliable. It is frequently both biased and clueless, and if you read it, then it is very hard to tell when they are being one or the other or neither or both. (Note that merely biased, but not clueless, would be far better, because then you might be able to mentally correct for that bias.)
When a publication reaches a certain threshold of unreliability, reading it becomes less like improving your image of the world than like adding white noise to the image. Individual points may be perturbed closer to reality, but overall you'll just end up with a fuzzier image. Has The Economist has passed that threshold? I honestly don't know enough to say for certain, but given how annoying I find its tone and its politics, I'm not terribly motivated to give them the benefit of the doubt.