In any field of scientific interest, a mark of a champion is the ability to show impossibilities. In this sense, Kurt Gödel is a champion in the field of logic for establishing that it is impossible for any logical system—by which we mean a set of axioms—to be consistent and complete at the same time. Albert Einstein showed that it is impossible for an object to travel faster than the speed of light. Kenneth Arrow argued that it is impossible to aggregate social preferences to achieve a suitable representation without it being dictatorial.
Oliver Hart and Bengt Holmström carried on that lineage—they showed that it is impossible to design a ‘complete contract’. They are the recipients of the Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel, in 2016.
Contract theory is a field of study within law and economics that focuses on establishing agreements between two or more parties given asymmetric and imperfect information. Imagine walking into a restaurant to reward yourself with a much-needed break after meeting a tight deadline. One beer on top of another and you quickly lose count of the number of bottles. Much to your pleasant surprise, so does the person sitting at the cash counter. He is too lazy to think back and count how many beer bottles you all ordered and just asks you to make a guess and pay accordingly. You somehow arrive at a ‘lower estimate’ for the actual number and pay for a much shorter list of items consumed. As you step out of the restaurant, you are likely to ask—how can anyone hire this fellow? He is getting paid for nothing.
As it turns out, an economist will ask precisely the same questions.
Before we go further, let us emphasise that this is actually a true story that one of us lived through, in one of the many convivial bars of Sao Paulo. The experience was a reminder of Hart’s talk at the Indian Statistical Institute, New Delhi, on informal agreements and renegotiation. An interesting finding of his research was that contractual expectations can be managed by informal communication between interested parties to (partially) overcome rigidities implicit in written agreements. Our cash-minister would have probably met his periodic sales target before that episode and couldn’t possibly care any more than his fixed salary.
Think of situations where a shop-owner (our ‘principal’—in the jargon of economics) hires labour (‘agent’) or where bank-owners hire managers for daily operations or where a hospital hires a surgeon. In all these situations, a principal would want to pay the agent according to the effort the latter puts in. The central problem lies in the fact that the principal doesn’t get to observe the agent’s effort, and only gets to make inferences on the basis of some imperfectly observed outcomes.
The hospital wants to know how much to pay the surgeon: First, besides luck, a doctor’s success depends a lot on the collective effort from the team. As a result, incentive provided to a doctor should be located within the larger incentive structure for the entire team. Second, the doctor may be trying to manage a number of conflicting objectives, such as increasing patient comfort and lowering costs for the hospital—a doctor’s incentive should then depend on the outcomes of conflicting tasks, some of which are fairly hard to measure. Third, contracts for doctors should adjust payments that other doctors are receiving—they may contain information about the changes in training and technology that are hard to otherwise know. One of Holmström’s seminal paper shows that contracts should include any information that reduces uncertainty about agent’s behaviour—this automatically decreases the risk premium needed for an agent to accept the contract. Surprisingly, contrary to Holmström’s prediction, pay of senior executives in large firms are seldom linked to industry performance benchmark—one possible index to capture the industry-wide uncertainty.
Consequently, a CEO of an IT exporting firm may end up receiving a fat bonus on a boom year, despite delivering a weak performance and vice-versa.
Hart’s contribution lay more on the contracts and their relations to theory of firms. In one of his important early papers, Hart along with his co-author argues why firms, as owners of assets, are in a relatively better bargaining position if, either due to mutual interest or due to an unforeseen state of the world, contracts need to be renegotiated.
Further, he analysed why firms often deliberately leave contracts incomplete. He found firms would needlessly focus on easily measurable outcomes at the expense of not-so-easily measurable ones, if an extensive set of expectations are written down in a contract.
Hart and Holmström belong to a generation of economists who used complex mathematical arguments to model and predict outcomes. Not only did they have on offer rich insights about mechanics of contractual agreements, but also provided the discipline with some fundamental tools of analyses, which form one of the methodological bedrocks in economics today. Theirs was an era where empirical support was constrained by both data and computing power. Now with these constraints relaxed, many of their predictions have been validated by data. That, in itself, is a remarkable testimony of the prescience of their work.
Ritwik Banerjee is assistant professor of Economics at IIM Bangalore, India.
Subrato Banerjee is economist at Queensland University of Technology, Australia