Bryan Caplan  

Are Centers a Mistake?

Substitute goods and reasoning... Google and the EU foreign poli...
Suppose you want to convert the world to belief X.  You decide that professors are the best spokesmen for X.  There are ten universities, and you have enough money to fund ten professors.  There are two possible ways to spend the money:

1. Give ALL the money to one school to create a Center for the Study of X - an academic cluster where all ten pro-X professors work together.

2. Give EACH school enough money to hire ONE professor, so every university has a lone proponent of X.

Question: Which option will provide the best return on your charitable investment?  Definitionally, #1 is better if there are economies of scale, and #2 is better if there are diseconomies of scale.  But which description best fits the real world? 

My main thoughts: #1 is probably much more fun for the faculty.  Being part of a center of like-minded folks beats being a lone voice in the wilderness.  #1 also plausibly attracts more media attention.  A Center of ten professors is more visible than ten isolated professors.  However, #2 also has a big advantage: It avoids redundancy.  When a lone pro-X professor converts a student, it's hard to say, "It would have happened even if this professor hadn't been hired."  But if one out of ten pro-X professors converts a student, "It would have happened even if this professor hadn't been hired" is quite plausible.

Bonus question: Suppose #1 is definitely better than #2.  If each school already has one pro-X professor, would it be worthwhile for a donor to give one school enough money to "poach" all the pro-X professors from the other schools?  Could clustering really be that valuable?

Please show your work.

P.S. Thanks to all the great people I met in Ohio.  I'll be back!

Comments and Sharing

COMMENTS (22 to date)
Scott Wentland writes:

For having impact on students, it seems like the second option is likely the most efficient one for reasons you have already stated. However, is a center's greatest impact per dollar ("bang for the buck") on students or the general public? Professors can get a lot of bang for their buck through research (i.e. "converting" their colleagues among the profession) and writing for broader audiences (e.g. policy-makers and the general public). I think option #1 is far more effective at the latter, which is probably the greatest overall impact. Showing my work...the centers at GMU do a great job with research and impact more broadly.

Steve Waldman writes:

There's a very serious problem with #2 that #1 avoids.

Under #2, the professors you buy may get converted by their colleagues! #1 ensures useful reinforcement of correct beliefs.

dangerman writes:

It strikes me that converting enrolled students is one of the *least* effective ways either scenario makes the world more pro-X. Because enrolled students are so limited in number (unnecessarily so, of course).

Although I realize that it's one of the activities the professors spend most of their time actually doing.

ilya writes:

The first option, because it incentivizes other people to donate as well.

Bonus answer: yes, but still second best to creating new pro-X professors.

Max M writes:

Suppose #2 is definitely better than #1. Does this mean a donor interested in promoting position X should fund a big center for position anti-X to gather all the anti-X advocates in one relatively-less-effective concentration?

Glen Smith writes:

Can you pull your money if the center goes off agenda? In any case, it would probably be easier for you to make sure the profs in a center don't 'expand your idea' in a way that destroys its core.

Keith K. writes:

It depends somewhat on what the center does.

If the center is good at turning out high quality grad students/professors in your intellectual tradition then it is probably going to have a longer lasting influence. The biggest problem I can see from spreading out the professors is simply that (given academic politics), it is more likely that "the one fringe guy" in each location will simply be marginalized by the other folks who control the department via superior numbers.

August writes:

I think it would make far more sense to spend the money putting the message out directly, perhaps starting a conference or two with like minded people who either already agree or respond well to your message. If you look at the paleo-ancestral health movement, you will see people did this more or less naturally.
Universities do not buy you much these days.

1. More fun for faculty should translate into lower cost for option #1.
2. Economies of scale seem plausible, but perhaps at a smaller number of faculty than 10--option #1.5 with 5 faculty at two centers might yield a higher ROI than either #1 or #2.

Ricardo writes:

Bryan, how much would I have to pay you to give up your lunches with fellow GMU academics? My guess is a lot. Michael Giberson is right, option 1 costs a lot less than option 2.

Chris writes:

Option 3.

3 centers with 3 profs each and the extra leftover money used for a yearly award for the prof with the best published work.

J Storrs Hall writes:

Google Population Viability Analysis.

Tom West writes:

Suppose you want to convert the world to belief X. You decide that professors are the best spokesmen for X.

Conversion goes from the top down. A center garners more media attention and more important, attention from elites (who use media attention as a gauge of importance).

Even with regard to proselytizing, the real gain is from seeding graduate students into professorial positions, and that to is mostly based on status of the graduating institution. A center has status, making it's graduate students more likely to get hired.

Lastly, the idea that fun means you can pay less is at odds with every place that I've worked. The more I've been paid, the more fun the work and the better I've been treated.

In my experience, economics only rarely trumps the human instinct to value people by how much they're paid (unless it reaches a truly egregious disparity).

Jonathan writes:

#1 is the better option.

It gains certain benefits that #2 does not get (more donation, more focused research, better shared research).

It also attains the benefits of #2. #1 will produce more students willing to be professors that think similarly. Those new students likely won't work at the original university, so you get the sprinkling of professors at other universities.

Thomas Sewell writes:

It's more complicated than just one strategy over the other. You need #1 up to a certain critical mass, then after a certain point of diminishing returns, you'll get more of a return from switching additional professors to #2.

As an example you may not have considered, Mormons began by proselyting and recommending gathering into one location, then only once that location had a sizable self-reinforcing community of the like-minded, switched to proselyting and encouraging folks to stay where they were converted and start a new LDS community there.

John writes:

Depends also on what the center is called. If it's The Center for X then everyone will ignore what the professors there say - oh, she would say that, she's employed at The Center for X, etc

blink writes:

I see two overlooked reasons favoring centers: (1) substitution and (2) monitoring.

(1) Presumably the goal is to increase the number of professors supporting X, but earmarks are fungible. This is just the food stamps problem. Substitution is also related to poaching.

(2) Hiring professors who best fit the desired category (or even closely) requires interpretation. Centers, with many like-minded professors, are more likely to self-monitor and prevent wolf-in-sheep's-clothing hires.

ThomasH writes:

Economies of scale ought to play a role. If the idea needs "big data" to be promoted, then the Center strategy is probably better.

Although I'm happy the Scott Sumner will be coming to GMU, his advocacy of NGDP targeting did not seem to suffer from being "alone" at Bentley. Advocating a carbon tax however probably requires dealing with climate models, international trade models, and political economy models that people scattered around different faculties might not be able to address as well.

Joshua Katz writes:

Option 1 causes option 2: A center does a good job producing future faculty. I don't mean so much persuading current graduate students, but rather teaching those already inclined in that direction. They can then go out and infect other institutions - except that the most capable (who are frequently not the best spokespeople, I might add) will end up working at the center.

AS writes:

#2 seems better. Then graduate students at each of the ten universities have the choice to work in X and continue the chain of research. If all professors in X gather at a single school, that deprives students of the other nine universities from working in X.

X may also be complementary with non-X. Key insights are often achieved through combining ideas from different fields in novel ways. If a university specializes in X, it misses out on that opportunity and its research may eventually become repetitive and stale.

Duncan Frissell writes:

I think a Center gives more flexibility because in addition to proselytizing you can work together easily and also perform tricks that a lone prof can't.

Par example, Hillsdale is a "Center" and it can teach a classics curriculum 50% of the time. Thomas Aquinas or St. Johns can teach a classics curriculum 100% of the time. A lone prof can't.

Centers can also afford more extensive defensive works than lone profs. Hence the castle.

If centers choose to adopt institutional planting techniques based on the Pentecostal Church model, they can also spread and take over the world more easily.

William M writes:

#2 is clearly the better option. If your goal is to convert people to belief X, then you must be able to interact, understand, argue with, etc. people who are anti-X. You also want to influence people (students/professors) who haven't really thought about the issue or fully formed their own opinion with regard to X yet. The cocoon of a center for X doesn't really lend itself to either of these scenarios.

The benefits of a center for X can mostly be replicated without requiring the physical presence of all 10 pro-X professors at a single site. The journals, articles, book publishing, conferences, media publicity can all be handled administratively by non-professors.

Also, most (or at least the superstar) professors don't mind being the lone voice in the wilderness. They know they are correct on policy X and don't need the 9 other professors sitting down the hall from them to assist them in their proselytizing.

This example is one way that the marketplace for ideas is different than the marketplace for goods.

Comments for this entry have been closed
Return to top