Start with Alvin Roth's Who gets what--and why. It is the trade book by the economist who built the kidney exchange and the medical residency match, and it explains what market design is — clearing a market when prices alone cannot do it — without a line of algebra. Every formal chapter later in this path is a model of something described in it.
The honest warning belongs at the top, because it determines how far down the list you should go. Stage one requires no mathematics. Stage two is advanced-undergraduate and graduate game theory. Stages three onward are graduate economics: they assume calculus, probability, and comfort with static and dynamic games, and they use Bayesian Nash equilibrium without stopping to explain it. A reader who wants to understand what the field does, rather than do it, can stop after stage one and lose nothing.
What market design is, before any mathematics
Who gets what--and why first. Then Reinventing the Bazaar, John McMillan's wider framing of what any market needs in order to function at all — property rights, information, trust, competition — surveyed from Dutch flower auctions to post-Soviet transition. It is there to stop market design being read as a purely technical exercise. Discovering prices is Paul Milgrom's short lecture-derived account of the FCC incentive auction, the largest market ever deliberately designed. It closes the stage because it is the field's showpiece and the most concrete answer to what a market designer actually does, and it is still readable before any formal auction theory.
The game theory the field assumes
Game Theory by Steven Tadelis is the gentlest of the three and the right start if your last formal economics was intermediate microeconomics: static and dynamic games and incomplete information, taught through worked examples rather than theorem-proof density. An Introduction to Game Theory by Martin Osborne is a step up in rigour and the standard advanced-undergraduate text; read it second for the cleaner treatment of extensive-form and Bayesian games that everything later takes for granted. Game Theory by Roger Myerson is the graduate reference, by the economist who proved the revelation principle — the single result auction theory leans on hardest. Do not read it cover to cover. Use it to settle mechanism-design questions raised by the later stages. Two different books here are catalogued as simply Game Theory, so check the author when buying.
Auction theory proper
Auctions is Paul Klemperer's compact survey plus his essays on what actually went wrong in the European 3G spectrum auctions. Read it before the textbook, because it tells you which theorems matter in practice. Auction Theory by Vijay Krishna is the standard graduate textbook and the mathematical spine of the path: private-value, common-value and multi-object auctions in one framework. It assumes stage two outright. A solutions manual circulates separately under a similar name, so make sure you are buying the textbook.
Putting Auction Theory to Work is where the theory meets a design brief, written by the economist who designed the FCC auctions. Read it after Krishna: it assumes the equilibrium results and spends its energy on package bidding, activity rules, and the practical failures the theory does not predict.
Matching, and designing a whole market
Two-Sided Matching by Roth and Marilda Sotomayor remains the definitive statement of matching theory — Gale-Shapley deferred acceptance, the lattice structure of stable matchings, and the impossibility results that constrain every real match. It is the formal counterpart to the kidney and residency stories in Roth's trade book. Market Design by Guillaume Haeringer is the most recent textbook to treat auctions and matching as one subject rather than two literatures, which is how practitioners now think about it, and it consolidates the two preceding stages.
Computation and large-scale implementation
Combinatorial Auctions is the edited volume, assembled by Peter Cramton with Yoav Shoham and Richard Steinberg, that made package bidding a practical tool: winner-determination algorithms, bidding languages, and the empirical record. Read it after Milgrom, whose designs it formalises. Algorithmic game theory — the Nisan, Roughgarden, Tardos and Vazirani volume — closes the path by reframing everything above as a computational question: not what exists in equilibrium, but what can actually be computed and implemented at scale.
Follow the full ordered path here: The Best Books on Market Design and Auction Theory, in Order.
