After noting that “some of my best friends work for think tanks,” The Atlantic blogger Megan McArdle contends that these organizations cannot be trusted because the purported ideological homogeneity of their employees renders them intellectually blindered, lazy, and compromised. Since Cato was apparently the first think tank Ms. McArdle thought of in the context of school choice (and I’m not even one of her best friends!), I couldn’t resist the temptation to respond.

All human institutions are flawed (I used to work for one of the most successful corporations in the world, and it was no exception), but I haven’t noticed the crippling “groupthink” that McArdle warns of in my time at Cato. I have had stimulating debates with colleagues on a host of issues, including education policy. As it happens, I’m not alone. Cato made waves in the blogosphere not so long ago due to a very public disagreement among its staff over domestic surveillance law and policy. One of the bloggers who noted this lack of ideological groupthink at Cato was… Megan McArdle. Cato scholars have at times publicly disagreed on high profile foreign policy questions as well.

Getting to the heart of her argument, though, I’m puzzled as to why someone conversant with public choice theory would imagine that think tanks are categorically different from other sources of scholarship when it comes to ideological bias. McArdle contends that donors would balk if think tanks produced publications at odds with the donors’ preferences. Some do. Cato has in fact lost donors on a number of occasions for this reason. But this incentive system is not fundamentally different from what obtains in the academic world. In academia, career advancement is tied to journal publication, but journal editors and peer reviewers have their own biases (as do academic authors themselves), and these are felt in their decisions of what to publish. Hence, there is pressure on academics to conduct studies they expect to be more rather than less publishable.

The most notable difference between academic and think tank papers is that naive some readers grant academic papers an unwarranted presumption of impartiality. The same applies to government publications. For think tanks, the absence of that presumption means that our work is more stringently checked by the media when the ideological flavor of the medium is different from the presumed slant of the think tank. Since there are few libertarian media outlets, Cato output is subjected to more extensive fact checking by the media than is the typical academic paper, government pronouncement, or the media’s own reporting. For example, I recently wrote an op-ed noting that the press grossly understates per pupil spending in DC public schools. In making my case, I observed that local media had claimed “$8,322 is spent for each student” and then I showed that when the district’s total budget is divided by its enrollment the real figure is over $24,000 per pupil. The editor of the paper publishing my piece asked for my source for the inaccurate media figure, claiming that his own paper knew better. I pointed him to an article in his own paper that used it (among others).

This is not unusual. It is far more common that we at Cato point out inaccuracies in the mainstream media and in government claims and reports than the other way around.

In a world in which everyone who collects and analyzes data invariably is subject to complex economic pressures, there is only one reliable path to the truth: read their publications critically and assess them for yourself, on their intrinsic merits. In cases where you lack the expertise to critically evaluate a study yourself, the next best thing is probably to seek out a proxy reviewer — an expert in the field whose conclusions you have reason to trust. But simply dismissing an entire category of scholarly institutions due to a misplaced faith in the impartiality of the other categories is an epistemological error.