DOC PREVIEW
Berkeley ECON 231 - Optimal Paternalism

This preview shows page 1-2-22-23 out of 23 pages.

Save
View full document
View full document
Premium Document
Do you want full access? Go Premium and unlock all 23 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 23 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 23 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 23 pages.
Access to all documents
Download any document
Ad free experience
Premium Document
Do you want full access? Go Premium and unlock all 23 pages.
Access to all documents
Download any document
Ad free experience

Unformatted text preview:

Optimal Paternalism∗Manuel Amador, MIT Iv´an Werning, MIT and NBERPreliminary and IncompleteAbstractIf one accepts some paternalistic objectives, how should this shapepolicy? This paper addresses this issue by considering the trade-offthat results when agents ha ve superior information regarding theirown situation and tastes but a paternalistic principal views agents’tastes as as being biased. Applications include: sa vings behavior byhyperbolic discoun ting consumers, schooling choice by teenagers andmany situations with externalities. We show that under certain condi-tions the optimal mechanism takes a simple threshold form. Particularattention is given to the savings case and the implication for forcedminimum savings policies.1IntroductionIf people suffer from self-control problems, what should be done to helpthem? Most analysis lead to a simple conclusion: the optimality of takingover individual’s choices and running their lives for them. For example, inmodels with quasi-hyperbolic agents it is generally desirable to impose aparticular savings plan on individuals. To obtain less extreme prescriptionswe require some meaningful trade-off to the benefits of commitment. In thispaper we model a trade-off between commitment and flexibility.It is con venient to reformulate the same issue using a “paternalistic” in-terpretation. Thus, consider the maximizing a “paternalist’s” utility function∗We’d like to thank Andy Atkeson and George-Marios Angeletos for helpful commentsand discussions.1over two goods consumed by an “agent”, subject to a standard budget con-straint. We assume the paternalist’s utility function differs from the agent’sand, in particular, that the agent’s preferences are biased — relative to thepaternalist’s preferences — towards one particular good. The question we askis: What is the optimal mechanism in this situation?If this were the whole story then clearly the optimal thing to do wouldbe to force the agent to consume a certain allocation: that which maximizesthe paternalist’s utility function. However, it is likely that the individual’shas superior and relevant information about their own situation. If the pa-ternalist finds this information relevant then imposing a particular allocation— thus disregarding the individual’s information completely — is likely to beundesirable.Our main example modifies the in tertemporal taste-shock preference spec-ification introduced by Atkeson and Lucas (1995). Eac h period agen ts receivean i.i.d. taste shock that affects their desire for current consumption. To thispreference specification we incorporate quasi-hyperbolic discounting [Laibson(1997)]. Quasi-hyperolic discounting implies that preferences over consump-tion plans change over time and generate a desire for commitment. In thissetup we characterize the optimal allocation from the point of view of thetime 0-self. One can interpret this solution as providing the optimal com-mitment device, constrained only by the incen tive problems generated by theasymmetric information.Although we focus on the h yperbolic discounting model as our main ap-plication, the crucial feature of our analysis is a disagreement in preferencesbet ween a ‘paternalist’ and an ‘agent’. Hyperbolic models provide one ra-tional for disagreement in preferences bet ween the different time selves. Wediscuss some other rationales for disagreement about preferences.The rest of the paper is organized as follows. After a brief discussion ofrelated literature, Section 2 lays out the basic intertemporal model. Section3thenbriefly detours to discuss some alternative interpretations. Section4 begins the analysis of the intertemporal model by solving the case withtwo taste shocks. Section 5 shows the kind of problems encountered withmore than two shocks. Sections 6 and 7 analyze the case with a continuumof shocks — these sections contain the main results of the paper. Section 8extends the analysis to arbitrary time horizons. Section 9 concludes. Anappendix collects some proofs.21.1 Related LiteratureSeveral strands of literature are related in different ways to this paper andwe review briefly those most related.O’Donahue and Rabin (2003) advocate studying paternalism normativ elyby modeling the errors or biases agents make explicitly and applying stan-dard public finance analysis to these environments. We subscribe to thissuggestion and take a further step by studying the mechanism design prob-lem associated with our environment instead of optimizing over a particularset of policies1.Economists have long been interested in the implications of and justifi-cations for discounting the future at a lower rate than individuals. Phelan(2002) studies the implications for long-run inequality from insurance in ataste shock framework where the planner does not discount the future whileagents discount the future at a positive rates. Caplin and Leah y (2001) dis-cuss another justification for a welfare functional that discounts the futureat a lower rate than individuals. Note that in both papers the planner andagents discoun t the future exponentially.There is a large literature on social securit y policies that attempts to takeinto account the possible “undersaving” b y individuals. Diamond (1977) dis-cussed the case where agen ts may undersave due to mistakes. Feldstein(1985) examines the case where agents discount the future at a higher ratethan the planner to study the optimal pay-as-you-go in an OLG model. Us-ing h yperbolic-discounting preferences Imrohoroglu, Imrohoroglu and Joines(2000) perform a quan titative exercise measuring the welfare effects of pay-as-you-go systems. Diamond and Koszegi (2002) use a model with hyper-bolic discounting agents to study the policy effects of endogenous retirementchoices.Athey, Atkeson and Kehoe (2002) study a problem of optimal monetarypolicy with a trade-off between comittment and flexibility. In this regardtheir paper is closely related to ours. In their model the cen tral banker isbenevolent but suffers from a time-consistency problem that could lead tohigh average inflation. Each period the central banker receives a shock thataffects its objective function. They study the optimal policy design problemwhen this shock is private information.1DellaVigna and Mermandier (2003) study a mechanism design problem that resultsfrom the maximization of profits for a firm facing hyperbolic, and possibly naive,


View Full Document

Berkeley ECON 231 - Optimal Paternalism

Download Optimal Paternalism
Our administrator received your request to download this document. We will send you the file to your email shortly.
Loading Unlocking...
Login

Join to view Optimal Paternalism and access 3M+ class-specific study document.

or
We will never post anything without your permission.
Don't have an account?
Sign Up

Join to view Optimal Paternalism 2 2 and access 3M+ class-specific study document.

or

By creating an account you agree to our Privacy Policy and Terms Of Use

Already a member?