Tim Lambert had a go a while back at Glenn Reynolds over Liberal v. Conservative books being prominently displayed in bookstores. Now it may be a little foolhardy of me to leap in here given Tim’s recent success at ripping me a new one over statistics but here goes.
The basic problem as enumerated by Reynolds (leaving aside sock puppets and the like) is that Liberal books appear to get better and larger placings in bookstores than Conservative books. Several commentators at Lambert’s site above point out certain tricks of the trade, such as slotting, where a publisher pays the bookstore for such space. Other things, such as Liberal books appearing to come from big publishers while Conservative ones come from small ones, are also mentioned. No one would be too surprised that a large publisher will be able to mount a larger publicity campaign, or that a bookstore would take more notice of their reps. There’s even (a view which I suspect actually has some truth to it) the idea floated that since academia is left slanted, and academics read more, thus Liberal books will sell better.
All of the above could be and quite probably are true, and none of them in any way indicate market failure. However, there’s one other argument floated, categorised as an economic one, which needs a little bit of elucidation. This is the idea that as bookstores are in business to make profits, they will therefore always do things to maximise such profits. Most certainly their shareholders would hope that they do. However, it is well known that this is not actually the case. It isn’t a perfect world. One way of describing this is agency theory. Yes, the shareholders want to maximise profit, but they have to hire other people to actually go and make it for them. Making sure that the interests of employees are precisely aligned with the interests of the shareholders is very difficult….one example would be higher wages v. higher profits. It isn’t actually possible to align everyone’s interests so that they are all pursuing the same goal. The best one can do is get people to grumpily aim in vaguely the right direction.
Now, let us set up a hypothesis. Imagine that those who actually work in book stores were, for some reason or another, more left wing or politically correct than the average in the population. We could have fun thinking up reasons why this might be so (Republicans are all off doing MBAs? bookstores tend to hire Liberal Arts Graduates? bookstores don’t pay much and poverty is not a known driver of Conservative political views? bookstore employees must be literate and this is not a markedly obvious attribute of Conservatives? your choice.) but at the moment we’re just going to assume it. Now, given that, could we expect to see what Reynolds describes? Most certainly.
The shareholders are looking for profit maximisation. The employees are not. There is a more complex calculation going on with them, sure, interest in the money income but also in propagation of their views (not that this is exclusive to the left. All humans are motivated by more than just money. Corporations less so.). We could go further perhaps (note, perhaps,) and say that the biases of the ground level employees are biasing the ordering process higher up the management tree…..under-reporting of folks who’ve come in and asked for the Conservative books would, of course, reduce the number ordered.
One doesn’t have to be a conspiracy theorist to believe in this. If our starting assumption was that bookstore staff are more right-wing than the general population we would expect to see the same problem in reverse. Actions by the workforce to the perceived benefit of that workforce, by that workforce, to the detriment of the profit-maximising desires of the shareholders.
One of the joys of the modern world is that a number of statistical techniques have been devised to enable one to test such assumptions. As above, it quite obviously ain’t gonna’ be me who does so. But two ways come to mind to test this, rather, each method could test part of the theory.
1) Take a bookshop with known to be Conservative leaning staff. An Evangelical Christian shop might serve as a model for this (maybe not. Perhaps someone could suggest an alternative?). Do they choose to display different books from the mainstream? Do they do so to the detriment of the profits of the shareholders but to the benefit of the staff’s perceived world view? Voila, we would appear to have shown that there are indeed agency problems at work.
2) Look at sales of Liberal v. Conservative books as reported by those stores which we think are indeed staffed by leftists and compare them to neutral forums such as Amazon. If we find that books sell in equal proportions via the two distribution methods then we know that everyone is indeed acting economically rationally. The publishers buying the slot space, publishing the books themselves, the employees (apparently) pushing the Liberal ones and so forth. There are no agency problems and no market failures. If, however (as I suspect but certainly would not want to try and have to prove), we find that online sales of Conservative books do better, proportionately, than they do in bookstores then we have, no, not proven, for there are other possible answers (Conservatives have greater access to the net?), at least indicated that there are two economic problems. One, the agency problem, plus one of a market failure, in that no chain has as yet stepped in to fill the gap. (Note, the old joke about Chicago economists refusing to believe that there is a $10 bill on the sidewalk because someone would already have picked it up is indeed a joke, one at the expense of said School of thought which has at times been accused of an overly, rigorous shall we say, view of how close the real world gets to perfect information.)

If you, as I suspect many of Tim Lambert’s readers will, find this description of agency problems a little difficult to swallow, allow me to recast it slightly. Do you think that Michael Eisner’s untold squillions in pay and options are really in the best interests of Disney shareholders? Yes? then their interests are aligned. No? Then this is the agency problem again, perhaps in an anecdote you find more believable.