Re: the efficiency of fair-pricing. I have long considered consumer surplus more important than producer surplus, but I've never bothered to actually construct the argument:
1. Most people are both consumers and producers, so on the face of it you might be indifferent to gains in C vs S.
2. But waged employees - i.e. most economic participants in their "producer" role - don't capture much of producer surplus
3. Standard U = log W social welfare move: we should weight surplus by the marginal utility of income of whoever receives it
(That argument isn't really about consumer surplus but boring distributional efficiency, except insofar as consumer surplus happens to fall on more people with higher marginal utilities.)
Could easily be overwhelmed by entrepreneurial incentives and investment multiplier effects on U though.
great point, you're right that 50/50 is probably not the end of the story.
two things I'd say in defense of the simplification: (1) the symmetric Nash split is already enough to get you the interesting structure (deadweight loss elimination, the α-dependence, the FDT enforcement mechanism) without needing to parameterize the competitive landscape. and (2) once you add competition, the Shapley value in the multi-player game shifts the split toward buyers anyway (their outside options are stronger), which is directionally what your Ramsey argument also says.
so I think you're pointing at the right correction: log utility all the way through gives you an asymmetric, buyer-favoring split, and competition pushes the same direction, but not all the way to uniform pricing.
It seems to me that you are treating the seller as a monopoly, which is what gives them an incentive to charge above marginal cost. This allows for the price discrimination to unlock more deals increasing the total surplus. Where as in a perfectly competitive market, all worthwhile trades already happen so there is no way to increase both consumer and producer surplus.
Probably worth making that assumption more explicit.
Re: the efficiency of fair-pricing. I have long considered consumer surplus more important than producer surplus, but I've never bothered to actually construct the argument:
1. Most people are both consumers and producers, so on the face of it you might be indifferent to gains in C vs S.
2. But waged employees - i.e. most economic participants in their "producer" role - don't capture much of producer surplus
3. Standard U = log W social welfare move: we should weight surplus by the marginal utility of income of whoever receives it
4. Pricing is then the [distributional](https://www.jstor.org/stable/1821470) [many-person Ramsey rule](https://eml.berkeley.edu/~saez/course/Diamond_JPubE(1975).pdf) rather than fair-price
(That argument isn't really about consumer surplus but boring distributional efficiency, except insofar as consumer surplus happens to fall on more people with higher marginal utilities.)
Could easily be overwhelmed by entrepreneurial incentives and investment multiplier effects on U though.
great point, you're right that 50/50 is probably not the end of the story.
two things I'd say in defense of the simplification: (1) the symmetric Nash split is already enough to get you the interesting structure (deadweight loss elimination, the α-dependence, the FDT enforcement mechanism) without needing to parameterize the competitive landscape. and (2) once you add competition, the Shapley value in the multi-player game shifts the split toward buyers anyway (their outside options are stronger), which is directionally what your Ramsey argument also says.
so I think you're pointing at the right correction: log utility all the way through gives you an asymmetric, buyer-favoring split, and competition pushes the same direction, but not all the way to uniform pricing.
It seems to me that you are treating the seller as a monopoly, which is what gives them an incentive to charge above marginal cost. This allows for the price discrimination to unlock more deals increasing the total surplus. Where as in a perfectly competitive market, all worthwhile trades already happen so there is no way to increase both consumer and producer surplus.
Probably worth making that assumption more explicit.
That is a very good point! I have some discussion of that in footnote 2, but maybe I should have pulled it into the main body
Great piece to feed to AI-for-collective-bargaining-of-consumers before launching; that's still some years away, but thanks in advance!