Why People Resist Marginal Cost Pricing

3 May 2010 at 10:50 am Peter G. Klein 4 comments

| Peter Klein |

Consumers everywhere seem to like all-you-can-eat/drink/exercise/drive pricing, even if their own consumption is modest and they are net losers in the process. This may simply be irrational. But (as cruise ship lines understand very well) it may also reflect the real psychological benefits of being able to imagine oneself in an environment of no scarcity, in which everything is “free” including calories. By the same token, salad-bar-style blanket pricing may relieve consumers of the very real costs of weighing costs against benefits at every turn of the menu.

One could interpret this in terms of bounded rationality, e.g., Akerlof and Yellen’s concept of near-rationality (1, 2): small deviations from “rational” behavior aren’t too costly, in such cases, so why be fully rational? Alternatively, call these decision-making costs some kind of transaction costs and deem the behavior fully rational, once all relevant costs are taken into consideration. (Some hand-waving required in both cases.) But some kind of cognitive explanation is probably best here.

Entry filed under: - Klein -, Ephemera.

Greenstein on Innovation Religion and Economic Development

4 Comments Add your own

1. Lawrence H. White | 3 May 2010 at 12:02 pm

All-you-can-eat restaurants are a small share of the restaurant market. So there must be diminishing returns to taking a holiday from evaluating food items at the margin (or, what many people like about buffets, having the opportunity to sample small portions of a wide variety of items).

Unlimited mileage rental cars predominate the market, though, so perhaps a different explanation is called for.
2. Recomendaciones « intelib | 3 May 2010 at 3:42 pm

[…] Why People Resist Marginal Cost Pricing, by Peter Klein […]
3. srp | 3 May 2010 at 8:20 pm

It’s probably a good idea to play Gary Becker and first ask what a simple optimizing reason for this behavior would be. In this case, the simple story is about the option value of a right to consume more than usual amounts at no additional price. A model like that would probably generate testable comparative statics.

If those predictions were falsified empirically, then the next move would be to introduce a cognitive cost of budgeting or economizing. Extra free parameters make testing this harder.

A third possibility is sophisticated time-inconsistency where the agent wants to precommit to a policy to overcome weakness of will. For example, buying a gym contract for a time period rather than paying for each visit helps overcome “myopic” resistance to working out. All-you-can-eat overcomes “myopic” tendencies to postpone pleasurable consumption (which has been claimed to happen by some behavioral economists).
4. Michael E. Marotta | 4 May 2010 at 7:59 pm

Barak-Kim-Shelton in their papers on the “CSI Effect” report that there is “no” CSI Effect among juror… that is the BIG HEADLINE. But when you read the papers, the truth is more complicated. One scary fact is that educated people require less proof while the uneducated expect television-style criminal science.

I mention that to cast some light on what “consumers” seem to “demand.” 73% of them? or 37% of whom?

Obviously, if you have to guess, it is better to buy a bit more than you need than a bit less, to charge a bit more than you think you are worth than to leave money on the table, to kiss good night twice rather than not at all. Most people would rather live an extra day than one less.

And yet… I seldom go to buffets or smorgasbords. I do not enjoy food, actually, but just have to eat it as a necessity. I have never in my entire life watched enough Super Bowl all together to make a single game’s worth. I have never watched “Dancing with the Stars” or “Survivor” or “Grey’s Anatomy.” We do have an NCIS disk to watch tonight. I am not a rightwing retreatist in a log cabin. But, as F. A. Hayek and Ludwig von Mises predicted, I am one of 6 or 7 billion individuals whose values are subjective.

I accept the statistical sampling and question the inference.