Objectifs

Part 1: Hakimov

Market Design

The class aims to introduce students to one of the most applied areas of microeconomics - market design. This class is about how to design mechanisms to allocate scarce resources and how to create successful marketplaces and platforms. Different marketplaces have different tasks to accomplish. Practical market design that aims to design marketplaces that will be adopted, implemented, and maintained can be thought of as a kind of economic engineering.

A market is an institution where goods and services are exchanged or traded, and its role is to determine who gets what. Some markets, like commodity markets, typically function without any effort of the designer, while demand and supply are the primary designers for them. Thus, in these markets prices decide who gets what. This is not the case however for a variety of other cases. Sometimes money cannot be used for one reason or another (for instance in case of allocating seats in public elementary schools), or sometimes the market is not thick enough to easily determine the price and ensure the transaction (for instance, if you want to sell a painting of Picasso). We will study these markets. Our primary tool will be economic theory. In most classes, we will start with a class of problems and develop a theoretical model that helps us identify what a “good” solution to the problem is, what are the incentive problems that might arise and whether some mechanisms or designs might lead to a good solution. Then we will consider applications where we think about a particular setting and compare what is done in practice to the theory. Often, we will look at empirical evidence to see whether the theoretical solutions are supported by the outcomes in the real applications or controlled experimental settings. At the end of each class, we will have a practice exercise session to get more familiarity with the studied solutions.

The class divides roughly into two main parts: auctions and matching. The first part of the class is on auctions. Asymmetric information is an essential part of auctions, as potential buyers might have limited knowledge about the value of the asset or rights for sale. Here we start with the simplest case of a seller who wants to allocate a single item (a house, a company, a painting, or the rights to a natural resource such as oil or timber), and look at how different types of auctions work in theory and practice. Then we will consider complex allocation problems where a seller wants to allocate goods and bidders may want to buy more than one thing. We will look at the auctions used to allocate multiple items, auctions that are used to allocate rights to mineral resources, online ads, and others. William Vickrey received Bank of Sweden Prize in Economic Sciences in Memory of Alfred Nobel in 1996 for, among the others, his seminal contribution to the auction theory. Also, Paul Milgrom and Robert Wilson received the prize in 2020 for improvements to auction theory and inventions of new auction formats.

In the second part, we are going to study matching markets, i.e., markets where there’s no money and no prices. Instead, the designer asks people what they want and try to satisfy their preferences subject to the scarcity of resources. Examples include problems such as assigning students to schools in large cities where families can ask to be placed outside their neighborhood school, placing doctors in residency positions, matching couples on dating websites, centralized college admissions or assigning donated kidneys to transplant patients. We will study a set of algorithms that have desirable theoretical properties and are often used in practice to solve these problems. Matching is one of the most recent and fast-developing field of economics, which provides the solution to the allocation problems across the globe, and the contribution of economists to this area was awarded Bank of Sweden Prize in Economic Sciences in Memory of Alfred Nobel to Alvin Roth and Lloyd Shapley in 2012.

Finally, in the last lecture, we will consider high-frequency trading markets and the adverse effects of the continuous allocation systems of the possibility of the black market development. In these markets, some undesirable effects can arise due to some players having the speed advantage over the other players. Independent of the presence of prices in these markets, the insights from matching market design solutions can be used to solve these problems.

An important goal in this class is to introduce the areas of economic theory such as game theory and mechanism design and connect this theory to practical applications. The hope is that once students get familiar with the ideas, they will be able to read academic papers or browse the internet and see lots of examples where this theory is relevant and useful.

Part 2: Santos-Pinto

The Economics of Asymmetric Information

The goal of this part of the course is to introduce students to the economics of asymmetric information. At the end of this part of the course students should know that informational asymmetries can lead to market failures and that there exist decentralized solutions (e.g., signaling and screening) and centralized solutions (e.g., mandatory insurance) that can reduce or prevent these market failures. Students should also know that informational asymmetries can cause welfare losses in contractual relationships not mediated by a market and that contract theory shows us how to reduce or prevent these welfare losses.

One of the implicit assumptions of the two fundamental theorems of welfare economics is that the characteristics of all commodities are observable to all market participants. In reality, many markets are characterized by asymmetric information, that is, actors on one side of the market have much better information than those on the other. For example, in financial markets, borrowers know more than lenders about their repayment prospects and, in insurance markets, prospective clients know more than insurance companies about their accident of risk.

An individual who possesses better information than others has in effect monopoly power over his/her own piece of private information. Hence, to understand the impact of asymmetric information on market outcomes, we must forsake general equilibrium models and resort to game-theoretical tools.

A number of questions immediately arise about markets with asymmetric information: How do we characterize market equilibria in the presence of asymmetric information? What are the properties of these equilibria? Are there possibilities for welfare-improving market intervention? During the 70's George Akerlof, Michael Spence, and Joseph Stiglitz laid the foundation for a general theory of markets with asymmetric information. Applications have been abundant, ranging from agricultural to financial markets.

Akerlof (1970) studied markets where sellers have more information than buyers about product quality. He showed that low-quality sellers may squeeze out high-quality sellers in such markets. As the average quality of the product falls, the price that buyers are willing to pay decreases and this further drives out high-quality sellers from the market. This process can, in extreme cases, lead to market unravelling or market failure. Generally, adverse selection is said to occur when an informed individual’s trading decision depends on her hidden (or unobserved) characteristics in a manner that adversely affects the uninformed agents in the market. In financial markets, adverse selection can lead to market freezes and liquidity hoarding, reflecting buyers’ beliefs that most securities offered for sale are of low quality. Government intervention can mitigate problems of adverse selection in financial markets. The effectiveness of policy responses depends on the cause of a market freeze.

Spence (1973) asked how better informed individuals on a market can credibly transmit, signal, their information to less informed individuals, to avoid the problems associated with adverse selection. He showed how the high ability agents may improve the market outcome by taking costly actions to signal information to poorly informed recipients. An important example is education as a costly signal of high individual productivity in the labor market. Applications of the costly signaling model of Spence to issues in corporate finance include Leland and Pyle (1977), Ross (1977), and Bhattacharya (1979, 1980) in the context of entrepreneurs seeking financing for projects that only they know the value of, managers of firms signaling their firm’s better prospects with debt financing, and managers signaling firm cash flows by precommitting to higher levels of dividends, respectively.

Rothschild and Stiglitz (1976) asked how the poorly informed individuals on a market can extract information from the better informed, to avoid the problems associated with adverse selection. They showed that an insurance company (the uninformed party) can give its clients (the informed party) effective incentives to “reveal” information on their risk situation through screening. Rothschild and Stiglitz (1976) makes a strong case for government intervention in insurance markets: because of asymmetric information, mandatory insurance may improve all agents’ welfare.

Asymmetric information also plays a very big role in many contractual relationships not mediated by a market. For example, after an owner of a firm hires a manager, the owner may be unable to observe how much effort the manager puts into the job. Similarly, the manager will often end up having better information than the owner about the profit opportunities available to the firm.

The existence of informational asymmetries in contractual relationships not mediate by a market can lead to welfare losses compared to situations were information is either complete or symmetric. Anticipating the development of such informational asymmetries, the contracting parties seek to mitigate the difficulties they cause. In the late 70’s and the mid 80's James Mirrlees, William Vickrey, Oliver Hart, and Bengt Holmström laid the foundation for a theory of contracts with asymmetric information (in short, contract theory).

If the owner of a firm is unable to observe how much effort the manager puts into the job, effort is costly, and the owner of the firm pays the manager a fixed wage, then the manager has an incentive to exert the lowest possible effort level, which, in turn, lowers the profit of the owner. Generally, moral hazard is said to occur when (i) two parties are involved in a contractual relationship, (ii) the hidden (or unobserved) actions of one party affect the payoff (utility) of both parties, (iii) the parties have opposite interests, and (iv) the informed party chooses “hazardous” actions to the uninformed party.

Mirrlees (1975), Holmström (1979), Grossman and Hart (1983), and Holmström and Milgrom (1987) analyzed how a principal, for example, the owner of a firm, should formulate a contract for an agent, for example, the firm's CEO, whose actions the principal cannot fully monitor. The owner of the firm wants to choose terms of compensation, which gives the firm’s CEO incentives to act in accordance with the owner’s interests, for example, by maximizing the owner’s expected profits. In designing such an incentive scheme, the firm’s owner has to take into account the costs of giving the firm’s CEO incentives to act in accordance with the owner’s interests. The higher the CEO’s sensitivity to punishment and the larger the amount of information about the CEO’s actions contained in the outcome, the lower these costs. The contract allows the firm’s owner to address the moral hazard problem by making the firm’s CEO bear part of the cost of undesirable outcomes or receive part of the profits from favorable outcomes. Contract theory has been significant for, among other things, governance of companies and the design of laws and institutions.

For their seminal contributions, Mirrlees and Vickrey won the 1996, Akerlof, Spence, and Stiglitz won the 2001, and Hart and Holmström won the 2016 Bank of Sweden Prize in Economic Sciences in Memory of Alfred Nobel.

Contenus

Part 1: Hakimov

1. Introduction to market design and simple auctions

1.1 General description of the course

1.2 Introduction to market design

1.3 Simple auctions - theory

1.4 Empirical evidence on simple auctions

1.5 Application: eBay

2. Other auctions

2.1 VCG auction

• Case study: Keywords auctions

2.2 Common value auctions

• Case study: US treasury bills auctions and IPO

2.3 Exercise for simple auctions

3. Two-sided matching: Marriage market

3.1 Two-sided matching: Marriage market model

3.2 Application: Medical match

3.3 Exercises for other auctions

4. Assignment problem and kidney exchange. School choice

4.1 One-sided matching: Assignment problem

4.2 Application 1: On-campus housing

4.3 Application 2: Kidney exchange

4.4 One-sided many-to-one assignment model: School choice

model

4.5 Application 1: School choice programs in US

4.6 Exercises for the marriage market

5. High-frequency trading and black markets in first-come-first-served systems

5.1 Trading: current system, problem, and solutions

5.2 Booking appointment systems: model, experimental test, and

recommendations

5.3 Exercises for assignment problems,kidney exchange, and school choice

5.4 Exercise for current lecture

5.5 Overview of exam structure

6. Midterm 1

Part 2: Santos-Pinto

1. Introduction

1.1 The Economics of Asymmetric Information

1.2 Five Main Lessons from Course

1.3 Asymmetric Information (Private Information or Hidden Information)

1.4 Two Kinds of Hidden Information: Hidden Characteristics and Hidden Actions

1.5 Hidden Characteristics: Definition and Examples

1.6 Hidden Actions: Definition and Examples

1.7 Two Kinds of Problems Caused by Asymmetric Information: Adverse Selection and Moral Hazard

2. Hidden Characteristics and Adverse Selection

2.1 Gresham’s Law

2.2 Adverse Selection in the Used Car Market

2.3 Akerlof's (1970) "Market for Lemons": The Binary Type Case

2.4 Akerlof's (1970) "Market for Lemons": The Continuous Type Case

3. Basic Game Theory Concepts

3.1 Incomplete Information

3.2 The Notion of Type and Strategy

3.3 Bayesian Nash Equilibrium

3.4 Dynamic Games of Complete but Imperfect Information

3.5 Beliefs and Sequential Rationality: Introduction to PBE

3.6 Beliefs and Sequential Rationality: PBE

4. Perfect Bayesian Equilibrium (PBE)

4.1 PBE in Dynamic Games of Incomplete Information

4.2 Bayesian Updating

4.3 Separating and Pooling Equilibria

4.4 The Beer-Quiche Game

4.5 Multi-Stage Games with Observed Actions and Incomplete Information

4.6 The One-Shot-Deviation Principle

5. Decentralized Solutions to the Adverse Selection Problem: Signaling

5.1 Spence's (1973) "Job Market Signaling"

5.2 Leland and Pyle's (1977) "Equity Signaling"

6. Decentralized Solutions to the Adverse Selection Problem: Screening

6.1 Rostchild and Stiglitz's (1976) "Screening in Insurance Markets"

7. Centralized Solutions to the Adverse Selection Problem: Mandatory Insurance

8. Hidden Actions and Moral Hazard

8.1 What is Moral Hazard?

8.2 How do Insurance Companies Protect themselves against Moral Hazard?

8.3 Moral Hazard in Principal-Agent Relationships

9. Solutions to the Moral Hazard Problem in Principal-Agent Relationships

9.1 The Principal-Agent Model

9.2 Bonuses and Other Forms of Incentives, Monitoring, and Auditing

9.3 Definition of contract, participation constraint, and incentive constraints

9.4 Holmström’s (1979) “Moral Hazard and Observability”

9.5 The “Informativeness Principle”

9.6 Grossman and Hart’s (1983) “Principal-Agent Problem”

9.7 Holmström and Milgrom (1987)

10. Midterm 2

Références

Part 1: Hakimov

Textbooks:

1. Two-Sided Matching: A Study in Game Theoretic Modeling and Analysis by Alvin Roth and Marilda Sotomayor published by Econometric Society. The textbook is optional and supplemental to the slides covered in class.

2. Market Design by Guillaume Haeringer. The textbook is optional and supplemental to the slides covered in class.

Part 2: Santos-Pinto

Textbooks:

1. The Economics of Asymmetric Information by Brian Hillier

2. The Economics of Contracts: A Primer by Bernard Salanié

3. The Theory of Incentives: The Principal-Agent Model, by Laffont and Martimort.

Pré-requis

Microeconomic Theory (Analyse économique: Microéconomie)

Evaluation

1ère tentative

Examen:

Sans examen (cf. modalités)  

Evaluation:

Part 1: Hakimov

The grade of this part of the course is 100% from midterm 1. Midterm 1 is closed-book. Midterm 1 will take place on week 7 of the course. Students are expected to be able to solve the allocation task, know the properties of the different allocation mechanisms, and also be aware of the main theoretical and empirical results on each of the topics. Students are expected to know the proofs of the main results if given in the class.

Part 2: Santos-Pinto

The grade in this part of the course is 100% from a midterm 2. Midterm 2 will take place in the last week of the course. Midterm 2 is written, closed-book, and lasts 3 hours.

Final Grade

Your final grade at the course is 50% of your grade in Hakimov and 50% of your grade in Santos-Pinto.

Rattrapage

Examen:

Sans examen (cf. modalités)  

Evaluation:

Each midterm has to be retaken (50% for Hakimov and 50% for Santos-Pinto). The two retakes are closed-book and back-to-back.