Tyler Cowen excerpts this NYT piece by Barry Schwartz on whether we have too many choices. I found titibits like this pretty damn obvious:
Sheena Iyengar and Mark Lepper, psychologists at Columbia and Stanford respectively, have shown that as the number of flavors of jam or varieties of chocolate available to shoppers is increased, the likelihood that they will leave the store without buying either jam or chocolate goes up. According to their 2000 study, Ms. Iyengar and Mr. Lepper found that shoppers are 10 times more likely to buy jam when six varieties are on display as when 24 are on the shelf.
If you're optimizing with respect to jam, then an increase in your number of choices increases your search costs. If it looks like the cost of sorting through all the jam is going to be fairly high, and your desire for jam isn't urgent, you're likely to just walk out jamless. You don't have to be optimizing, etither. More likely you'll be happy with anything that passes some threshold. But thresholds like these tend to be context sensitive (the worst jam at Whole Foods might be better than the best jam at Giant, but in both cases, you may aim for the middle of what's on offer), so you won't be sure where it is until you get a sense of your options. This, too, costs.
Tyler presents these cases as “brickbats” for libertarians and economists. Well, OK. To me, this points to the economic importance of “editors”. If people get turned off when the choice set gets too big, but people will buy something in the set if its smaller, then the money is in packaging smaller choice sets and knowing who to present them to (like the shoe salesman Tyler mentions). To some extent, this is precisely the difference between a boutique and a department store. Part of what you're paying for in a boutique is the editorial skill of the buyer & salesperson. The trim they choice set so you don't have to.
Methodological digression: Schwartz's results point to an fascinating area of research for experimental economists. The establishment science fiction economics isn't happy to recognize the scarcity of computational resources, and so just assumes that everybody is able to costlessly and immediately represent the entire choice set and come up with some preference-ordering over all those choices. Of course, we don't do this. We represent a tiny fraction of the potential choice set, and the fraction that we do represent seems to be primed by context together with our belief systems (and other stuff). Somebody, please please tell us: HOW DOES THIS WORK?!