PPS231S.01 Law, Economics, and Organization

Please download to get full document.

View again

of 34
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Similar Documents
Information Report
Category:

Documents

Published:

Views: 68 | Pages: 34

Extension: PDF | Download: 2

Share
Description
PPS231S.01 Law, Economics, and Organization. Spring 2012 I.3. Incentives. Incentives. Principals and Agents In this module, we truly begin our study of the interaction between law, economics, and the study of organizations. We study “incentives,” which is an awfully broad topic.
Transcript
PPS231S.01Law, Economics, and OrganizationSpring 2012I.3. IncentivesIncentivesPrincipals and AgentsIn this module, we truly begin our study of the interaction between law, economics, and the study of organizations.We study “incentives,” which is an awfully broad topic. Likewise, asking “Do incentives matter?” is asking abroad question – One to which we could answer: “Yes, we know that the Law of Demand (generally) holds.”IncentivesWhen economists discuss “incentive theory,” what they refer to most often is agency theory, or contract theory, i.e., the study of principal-agent relationships.What is a contract? The simplest answer is “Anything that is not a spot transaction.”But then, what is a spot transaction? It is a transaction that occurs spontaneously – buyer and seller meet on the market; buyer pays seller; seller gives up good or provides service to seller immediately.IncentivesContracts involve a certain length of time, even though they are often modeled as static games with a sub-game perfect Nash equilibrium.Examples:Ordering a book from Amazon.comRenting an apartmentFinancing a real estate purchase with a mortgageLeasing out a plot of land to a tenant… There are thousands upon thousands of examples!Incentives“All models are wrong, but some are useful.” – George Box.To study agency relationships, we strip down the contractual environment to its bare essentials.Typically, a contract involves two parties: (i) the principal; and (ii) the agent. And typically, the principal hires the agent to perform a certain task, in exchange for compensation.IncentivesIncentivesTwo Essential Features of Agency Relationships1. The agent’s effort and/or type is unobservable, or only partially observable at prohibitive cost.Example: Agricultural uncertainty in land tenancy; Riskiness of drivers in automobile insurance.2. The incentives are not perfectly aligned between the principal and the agent.The principal would really like the agent to exert the highest effort possible, but the agent really only wants to do enough to get paid.IncentivesRecall this scene from “Office Space” (1999). Peter works for Initech, which has hired two consultants to downsize the company. The consultants interview each employee.http://movieclips.com/2pyJo-office-space-movie-motivation-problems/What is the incentive problem (i.e., misaligned incentives) here? How does Bob offer to fix it? What’s the causal mechanism?IncentivesThus, the study of incentive problems is only interesting when the two conditions above hold, i.e., agent effort is unobservable and the incentives of the principal and the agent are not aligned. Can you see why?What if agent effort is observable?What if the incentives of the principal and the agent are aligned?IncentivesTwo Information AsymmetriesThere are two types of information asymmetries between the principal and the agent, and they are not mutually exclusive:1. Agent effort is unobservable; and/or2. Agent type is unobservable.(Note: Both problems mean that the First Fundamental Theorem of Welfare Economics no longer holds.)IncentivesTwo Information Asymmetries1. Moral HazardMoral hazard occurs when the principal cannot observe the behavior of the agent ex post, i.e., after the contract is signed.Milgrom and Roberts (1992): moral hazard is “the form of post-contractual opportunism that arises because actions that have efficiency consequences are not freely observable and so the person taking them may choose to pursue his or her private interests at others’ expense.” IncentivesIncentivesTwo Information AsymmetriesIn the neoclassical model of section I.1, moral hazard was not a problem. Consumers could perfectly observe the quality of the goods they purchased from firms, firms could perfectly observe the quality of the inputs they purchased from consumers, consumers who supplied their labor to firms did not shirk on the job, and the firms acted in the interest of their stockholders. IncentivesIncentivesTwo Information Asymmetries2. Adverse SelectionAdverse selection occurs when the principal cannot observe the type of the agent ex ante, i.e., before the contract is signed.Thus, adverse selection is a form of pre-contractual opportunism. The presence of heterogeneous types (e.g., differences in quality, riskiness, etc.) on the market causes inefficiencies, and the First Fundamental Theorem of Welfare Economics no longer holds.IncentivesTwo Information AsymmetriesTo understand adverse selection, let’s consider a simple yet classic example: Akerlof’s (1970) market for lemons.Two types of cars on the market: good cars (G) and bad cars (B) – bad cars are referred to often as “lemons”.Buyers’ WTP for a good car is PG and PB for a bad car, and of course PB < PG.Assume for simplicity that sellers’ WTA is symmetric.The proportion of good cars on the market is α, so that the proportion of bad cars on the market is 1 – α. IncentivesTwo Information AsymmetriesWhen presented with the opportunity to buy a car, a buyer’s WTP will simply be his expected WTP considering the quality uncertainty of the car she is presented with. That is,E(P) = αPG + (1 – α)PBObviously, PB < E(P) < PG.IncentivesTwo Information AsymmetriesBut then, since every buyer is only willing to pay E(P), no seller of a good car will be willing to sell at that price, and all sellers of bad cars will be willing to sell at that price, i.e., the market completely unravels and becomes a market for lemons.In reality, there often exists a continuum of types (or close to one), so that the market will not completely unravel. An example of complete market unraveling would be a voluntary market for automobile insurance. How so?IncentivesTwo Information AsymmetriesHere’s a personal example: The Universitéde Montréal’s student government’s vision and dental insurance plan during the 1997-1998 academic year.The student government asked students whether they would like the option to join a (voluntary) vision and dental insurance plan – 58 percent voted “Yes.” In January 1998, the insurance plan came into effect.A year later, the insurance company wanted to break the contract with the student government. Can you see why?IncentivesThe Identification Problem of Contract TheoryThe following is a true story.In 1990, enumerators were sent to several of the busiest intersections in Washington, DC to gather data on the cars that run red lights, all on the same day, for the same length of time.→ Of all the cars running reds, it turns out that Volvos were significantly overrepresented!IncentivesThe Identification Problem of Contract TheoryThe $64,000 Question:Is this an example of adverse selection or moral hazard?IncentivesThe Identification Problem of Contract TheoryThe $64,000 Question:Is this an example of adverse selection or moral hazard?We don’t know!IncentivesThe Identification Problem of Contract TheoryVolvos have a reputation for being the safest cars money can buy. So eitherVery bad drivers choose to buy Volvos as a form of partial insurance or commitment device → Adverse selectionOtherwise indistinguishable drivers who buy Volvos feel “too safe” in them and are less careful → Moral hazard.IncentivesThe Identification Problem of Contract TheoryThe key point is that without anything more (e.g., random assignment of drivers to cars), it is impossible to determine which of the two information asymmetries causes a departure from efficiency.The term “identification problem” comes from the statistical/econometrics literature: it is impossible to identify the causeof the inefficiency.IncentivesPrendergast (1999)The title of the paper is “The Provision of Incentives in Firms,” but Prendergast’ssurvey concerns more than just firms.Examples: Piece rates, options, discretionary bonuses, promotions, profit-sharing, efficiency wages, deferred compensation, etc.Two things: (i) underlying assumption that agents respond to incentives; and (ii) do firms write contracts with that in mind?, i.e., do contracts correspond to theoretical predictions.IncentivesPrendergast (1999)Note that these two questions correspond to the two major lines of inquiry in applied contract theory:1. Do incentives matter?2. Do we observe in the data the contracts that the theory says we should be observing?IncentivesPrendergast (1999)Main theme in the literature: contracts trade off risk sharing and incentives (we will be discussing this at length in a few weeks.)Incentives = Pay-for-performanceRisk = Depends on contextIncentivesPrendergast (1999)The upper bound on pay-for-performance in any contract is the amount of risk borne by the agent.Conversely, the constraint on the amount of insurance (the inverse of the amount of risk borne by the agent) provided in any contract is its incentive nature.IncentivesPrendergast (1999)1. Do incentives matter? By and large, they do. Agents respond to incentives. A more interesting question is whether incentive power increases as the performance measure becomes less noisy. Here, the evidence is not as clear cut.IncentivesPrendergast (1999)Lazear (1996) on wages vs. piece rates and autoworkers installing windshields (productivity increased by 35 percent, one-third attributed to selection).Paarsch and Shearer (1996) on Canadian tree planters. Climatic and soil conditions determine either piece rate or salary. Raw productivity difference is again 35 percent; more than two-thirds attributed to selection).Also: British jockeys; weight reduction of agricultural workers in the Philippines; etc.IncentivesPrendergast (1999)2. Do contracts reflect agency concerns?This is a much harder question. The bulk of this literature has used CEO compensation data, which we will look at explicitly this semester. In this case, the dependent variable is contract choice (e.g., wage; piece rate; mixture of both.)IncentivesPrendergast (1999)In that case, there is some evidence that contracts are signed to trade off risk-sharing and incentives, but it’s not completely convincing.One of the main problems is that risk aversion, while a simple concept, is extremely difficult to measure, and most proxies are useless (Bellemare and Brown, 2010).IncentivesPrendergast (1999)Miscellaneous Points:1. Multitasking2. Tournament theory (e.g., Millennium Challenge Corporation)3. Efficiency Wages (Shapiro and Stiglitz, 1984)4. Deferred compensation and relational contracts (example of baseball player on a wage, but with contract renewal pending.)SummaryWhat are the key points from this section?What’s a contract? ExamplesTwo essential features: effort observability, alignment of incentivesMoral hazard and adverse selectionIdentification problem of contract theoryPrendergast on the strands of research in applied contracts: incentives and predicted contractsPrendergast on the empirical evidence
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks
SAVE OUR EARTH

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!

x