Why a raffle allocates differently than a price

The same scarce good can be sold to the highest bidder or drawn at random among applicants. The two methods produce different recipients, and neither result is a mistake. Each mechanism records one kind of information about the applicants and discards the other, and the choice of mechanism is a choice about which kind counts.

What everyone sees

Random allocation is read as the fair option and price as the efficient one, with the decision framed as a trade-off between the two. That framing hides what is actually being selected: a lottery deliberately ignores intensity of need, while a price deliberately ignores everything except capacity to pay. Both are exclusions, and only one of them is usually described that way.

What is actually happening

Elster examined how institutions actually allocate scarce goods and found the rules in use — queues, lotteries, waiting lists, need assessments, prices — encode substantive judgements about which claims are legitimate, with different domains adopting different rules for reasons that are not primarily about efficiency. Roth’s work on repugnant transactions documents the complementary constraint: in some domains price allocation is refused outright regardless of efficiency, so the rule reflects what the community will accept as a valid basis for claiming. In both accounts the mechanism is not machinery attached to a decision. It is the decision, expressed as a procedure.

Why it stays hidden

The judgement hides because procedures look administrative. A raffle appears to be an absence of criteria rather than the criterion that all claims are equal, and a price appears to be a market fact rather than the criterion that willingness to pay measures entitlement. Once a mechanism is in place it is defended on operational grounds, and the substantive choice inside it stops being discussed.