Ideas How systems shape choices

Designing Rules People Will Not Game

Instead of hoping people will behave honestly, design the rules so that honesty is in their own interest, because people respond to incentives and keep private what they know.

The idea in plain words

Ask managers how much budget their teams need, and many will ask for more than they need, expecting to be cut. Ask bidders what an item is worth to them, and they will understate it to pay less. Rules that depend on people reporting honestly invite them not to.

Mechanism design turns the problem around. Instead of taking the rules as given and predicting behaviour, it asks what rules would produce the outcome you want when everyone acts in their own interest and keeps private what they know. The goal is incentive compatibility: rules under which telling the truth is the best strategy.

Where it comes from

In 1961 the economist William Vickrey studied sealed-bid auctions and found that if the highest bidder wins but pays the second-highest bid, every bidder’s best strategy is to bid their true value. Bidding more risks overpaying; bidding less risks losing an item they would have bought at the price. The design makes honesty pay.

Leonid Hurwicz generalised the approach in the 1960s and 1970s, defining incentive compatibility and showing how to analyse any set of rules as a mechanism for gathering and using private information. Roger Myerson and Eric Maskin extended it, Myerson in 1981 by showing how to design auctions that maximise a seller’s revenue. Vickrey received the Nobel Prize in 1996; Hurwicz, Maskin and Myerson shared it in 2007.

What it does to the mind

People are highly sensitive to the incentives in rules, often without consciously calculating. Teams learn quickly what a budgeting process rewards, sellers learn what a ranking rewards, and employees learn what a bonus formula rewards. Rules that make honesty costly teach people, gradually, that honesty is naive. Rules that make it pay can sustain cooperation among people who do not trust each other.

Where it shows up

The essay on web search describes the ad auction, a descendant of Vickrey’s, that weighted bids by how often users clicked and charged just above the next bidder, and the economists who later explained why it worked. The essay on payments describes interchange rules that shaped what cardholders and merchants did. The essay on marketplaces describes reputation systems designed to make honesty pay among strangers, and their inflation when they did not.

What it asks of someone deciding

Before adopting any process that depends on what people report, ask how it could be gamed and whether the rules reward the gaming.

Design for honesty: separate what people report from what they receive where possible, and use mechanisms like second-price rules, in which shading the truth gains nothing.

Remember who designs the mechanism. The party that writes and operates the rules can bend them to its own ends, which is why auctioneers and platforms attract scrutiny.

Argue with it

The strongest objection

Mechanism design assumes people are narrowly self-interested and fully strategic. Real people care about fairness and reputation, find rules confusing, and can be insulted into gaming systems that treat them as opponents.

Take it to your team

Questions to ask

  • If someone wanted to game this process, how would they do it, and would our rules reward them?
  • What private information do we need from people, and what do they gain by revealing it honestly?
  • Who designed the rules of this market, and whose interests do they serve?

Follow the thread

Connections

Read further

Bibliography

  1. William Vickrey, “Counterspeculation, Auctions, and Competitive Sealed Tenders,” Journal of Finance 16, no. 1 (1961), 8–37.
  2. Leonid Hurwicz, “The Design of Mechanisms for Resource Allocation,” American Economic Review 63, no. 2 (1973), 1–30.
  3. Roger B. Myerson, “Optimal Auction Design,” Mathematics of Operations Research 6, no. 1 (1981), 58–73.
  4. Benjamin Edelman, Michael Ostrovsky and Michael Schwarz, “Internet Advertising and the Generalized Second-Price Auction,” American Economic Review 97, no. 1 (2007), 242–259.

How to cite

Shrikant Kumar. “Designing Rules People Will Not Game.” Idea, Telos. https://shrikantk.com/ideas/mechanism-design/

↑ ↓ to move↵ to open⌘K to toggle