Consumer perspective canonical runs · main experiment · no intervention

Who pays, who gains, who loses

Personalized pricing is not uniformly bad for consumers: it brings in people the uniform price excluded and it squeezes the ones it can identify as willing to pay. The question is the distribution.

Does the burden fall on the poor or the rich?

Consumer surplus per period by income quintile, for each information regime.

Does the burden fall on high-valuation consumers?

Consumer surplus per period by willingness-to-pay decile.

How close does each algorithm get to charging your valuation?

Average offered price against valuation (20 bins). The 45° line is perfect extraction.

How concentrated is the remaining surplus?

Lorenz curves of consumer surplus (all consumers; non-buyers at zero).

Which behaviours protect you?

Surplus per period by consumer type under the individualized boosting seller vs uniform.

Are vulnerable consumers exposed?

Bottom income tercile × top urgency tercile: price, surplus and price-to-income ratio by regime.

Winners and losers

Share of consumers whose steady-state surplus is higher / lower than under uniform pricing, matched consumer-by-consumer within seed.

Reading the vulnerability metric. Income does not enter valuations directly in the model; it is a correlate. "Vulnerable" is a definition we impose (low income and high urgency) so that the price-to-income ratio is a measure of burden, not of demand. Alternative definitions are one line in spo/population.py.