Tuesday, November 16, 2004

Thinking about Think Tanks

Geoff Pullum and Justin Busch (both linguists) debate about whether right-wing think tanks are always described as “well-funded.” (Thanks to Neal for the pointer.) I think Busch gets the better of the debate. While it’s not literally true that right-wing think tanks are “always” described as well-funded, it’s certainly true that they draw that particular description substantially more often than left-wing think tanks.

This seems like a good opportunity to reiterate a point I’ve made before: that it’s fallacious to conclude, from the fact that A funds B, that the funding of A leads to the positions of B. The causation may well run in the other direction: B holds certain positions already, and then seeks out funding from sources like A who are likely to agree with B’s position. The distinction is important, since the former explanation characterizes B as a gun-for-hire while the latter characterizes B as an honest advocate.

In reality, the explanation needs to draw on both supply and demand explanations. Causality probably runs in both directions, but the relative impact of each is less clear. A think tank that is highly committed to its ideological agenda, and which is free to seek other sources of support, need not depend on the contributions of any one supporter. In such a case, we would expect causality to run primarily from positions to funding. On the other hand, if there’s a think tank with less commitment to its agenda and fewer funding options, we would expect causality to run mainly from funding to positions (as the accusers of right-wing think tanks claim).

Aside from the post hoc ergo propter hoc fallacy, the veiled gun-for-hire accusation also implicitly commits the ad hominem fallacy: it asks us to dismiss arguments, policy positions, etc., based on the interests of those who state them. Certainly, the existence of vested interests ought to make us a skeptical, but skepticism cannot substitute for careful analysis and argument.


Ronald said...

Imagine a think tank with a set of positions on all possible policy issues, each formed independent of donors' wishes. Now, suppose corporate donors can earmark contributions to specific policy areas -- say, investment banks can earmark contributions for Social Security privatization, etc. How will this affect a think tank's behavior?

The answer of course is that it will lead to not all policy areas being pursued equally aggressively. The ones that raise the most earmarked funds will have staff and equipment dedicated to them, and those that don't raise money will be scaled back. It's simply a matter of opportunity cost.

It's in this sense that in practice think tanks are influenced by donors' wishes. It's not that their policy positions are corrupted by donor influence, but that the *relative importance* of policy positions are distorted by earmarked funds. For example, a libertarian think tank may believe philosophically that free speech is a more important issue than lawsuits against food manufacturers, but its de facto policy output will skew toward whatever issue raises the most funds rather than its true philosophical importance. All this is fairly common knowledge in nonprofit development departments.

Although it's not a damning critique of think tank scholarship altoghether, this is an important and often overlooked way that funding can influence think tank output.

Glen Whitman said...

Andrew -- yes, I agree. As you say, it's not a damning critique, since the think tanks still act as honest advocates on any given issue.

Also, it's worth noting that the fungibility of money mutes the effect. Suppose a think tank studies 10 issues and has a budget of $10 million, allocated equally over all 10 issues. A new funder donates another $1 million, stipulating that all $1 million must be spent on Issue #3. It does not follow that Issue #3 will now receive twice as much funding as any other issue. If the original $10 million is not earmarked, it can be reallocated so that each issue gets $1.1 million, with Issue #3 getting $1 million from the specific donor and $100,000 from the general fund. Of course, if every donor earmarked his funds, this would not be possible. So the think tank's degree of freedom depends on the proportion of its funds that come without strings.

Anonymous said...

Glen, the "gun-for-hire" causation definitely happens more on the commercial side. If we replace "think tank" with "analyst firm" (and probably "investment bank" as well) then you get a lot more of this "so-and-so just holds that position because they were paid to."

See here for a well-documented example... http://vowe.net/cgi-bin/wiki.cgi?RadicatiGroup

Anonymous said...

Regarding the fungibility qualification, Alexander's 1996 American Journal of Sociology article showed that art museum curators are so good at juggling grants that in the aggregate their behavior is free of the strings attached to specific pots of money.

As for the mercenary notion, ideologies are typically more consistent than either party positions or corporate interests. So one can test the mercenary hypothesis by finding instances where the patrons' interests conflict with the think tank's founding principles and see which triumphs. One useful example may be China since most businesses favors going easy on China but most schools of conservative (and liberal) thought favor taking a stronger human rights stance with China.

Gabriel Rossman