Mirrlees & Vickrey (1996): Designing Incentives When Information Is Hidden

The Information Problem at the Heart of Policy

Governments want to tax the rich more than the poor. But they can’t directly observe ability — only income. And income depends on choices: a talented person might work less if taxes are too high. Insurance companies want to charge risky drivers more. But they can’t observe riskiness — only claims history. Auctioneers want to sell to whoever values the item most. But they can’t read minds — only bids.

All these problems share a common structure: someone must design a system (tax code, insurance contract, auction) that works well even though the participants have private information they may not want to reveal. James Mirrlees and William Vickrey independently developed the theoretical tools to solve such problems — laying the foundations for what would become mechanism design theory.

Vickrey: The Auction That Makes You Tell the Truth

In 1961, William Vickrey asked a question that seems impossible: can you design an auction where every bidder’s best strategy is to bid exactly what the item is worth to them — no bluffing, no strategic underbidding?

The answer was the Vickrey auction (second-price sealed-bid auction):

• Each bidder submits a sealed bid
• The highest bidder wins
• But they pay the second-highest bid, not their own

Why does this work? If you bid less than your true value, you might lose to someone who values it less — a missed opportunity. If you bid more than your true value, you might win but pay more than it’s worth to you — a loss. Bidding your true value is always the safest strategy. Truth-telling is the dominant strategy.

This was far more than an auction trick. Vickrey established fundamental principles:

• Incentive compatibility: A well-designed mechanism can align individual incentives with social goals — making honesty the best policy
• Revenue equivalence: Under standard assumptions, all common auction formats (English, Dutch, sealed-bid first-price, sealed-bid second-price) generate the same expected revenue for the seller
• The revelation principle: Any outcome achievable by any mechanism can also be achieved by a direct mechanism where participants truthfully report their private information

These principles became the foundation of auction theory and mechanism design — with applications from spectrum auctions (worth billions) to online advertising (Google’s ad auction is essentially a Vickrey mechanism).

Vickrey’s Other Innovations

Vickrey was remarkably versatile. Beyond auction theory, he pioneered:

• Congestion pricing: Vickrey proposed in the 1950s that road tolls and transit fares should vary by time of day — higher during rush hour, lower off-peak — to reduce congestion. This idea, radical at the time, is now implemented in cities worldwide
• Marginal cost pricing for public utilities: Electricity, water, and transit should be priced at marginal cost, with peak-load pricing to manage demand
• Cumulative averaging for income tax: To smooth the tax burden over a lifetime rather than penalizing volatile incomes

Mirrlees: The Optimal Income Tax

James Mirrlees tackled what may be the most important question in public economics: how should a government design its income tax if it wants to redistribute income but can’t observe people’s abilities?

The dilemma: the government wants to tax high-ability people more. But it can only observe income, not ability. If it taxes high incomes heavily, high-ability people may choose to work less — reducing both their income and the tax revenue available for redistribution. The tax system must balance:

• Equity: Transferring resources from rich to poor
• Efficiency: Not discouraging productive work
• Incentive compatibility: Ensuring high-ability people don’t pretend to be low-ability by working less

Mirrlees’s 1971 paper solved this problem mathematically, deriving the optimal nonlinear income tax schedule — the tax structure that maximizes social welfare subject to incentive constraints.

Key results:

• The optimal marginal tax rate is not simply progressive: It doesn’t just rise with income. Under standard assumptions, the optimal marginal rate at the very top is zero — a counterintuitive result because taxing the highest earner’s last dollar more doesn’t change their ranking, only discourages their effort
• The tradeoff is quantifiable: Mirrlees showed exactly how much redistribution you sacrifice for each unit of efficiency, and vice versa — making the equity-efficiency tradeoff precise rather than vague
• Information constraints bind: The optimal tax is less redistributive than a first-best (full information) tax would be — because the government must respect the fact that it can’t observe ability directly

The Broader Legacy: Mechanism Design

Together, Mirrlees and Vickrey established the intellectual framework for mechanism design — the science of designing institutions and rules that achieve desired outcomes when participants have private information:

• Principal-agent theory: How should an employer design a contract when they can’t observe the worker’s effort? (Moral hazard)
• Screening and signaling: How can an uninformed party (insurer, employer) design a menu of options that induces informed parties (drivers, workers) to reveal their type?
• Public goods provision: How can a government determine the true demand for public goods when citizens have incentives to understate their willingness to pay?

Their 1996 Nobel Prize was awarded “for their fundamental contributions to the economic theory of incentives under asymmetric information.”

Tragically, William Vickrey died just three days after the Nobel announcement, at age 82.

Explain It to a Child

Imagine you’re selling your old toys at a yard sale, but you don’t know how much each buyer would really pay. In a normal auction, buyers try to pay as little as possible — they hide what they truly think. Vickrey invented a clever rule: the winner pays what the second-highest person bid, not what they bid themselves. This magic trick makes everyone bid honestly, because there’s no advantage to lying. Now imagine you’re a teacher who wants to give harder homework to smarter students — but you can’t read minds to know who’s smart. If you make the hard homework too painful, smart kids will pretend to be average. Mirrlees figured out the best way to assign homework (or taxes) so that everyone reveals who they really are — without making the system so harsh that people give up.