Trevor Chow Blog Archive Newsletter About The Blog

Back-of-the-envelope low-quality musings: #takes

Longform stuff I've put time and thought into is at #effortposts.

To hear from academics and policymakers, #interviews.

Compiled scribbles on what I read can be found in #notes.

Topic-related tags include #economics, #coronavirus and #miscellaneous.

A Conversation with Nobel Laureate Oliver Hart on Contract Theory

14 May 2020

Professor Oliver Hart is the Andrew E. Furer Professor of Economics Harvard University. He specialises in contract theory and the theory of the firm, having served as the President of the American Law and Economics Association and as vice president of the American Economic Association. His work on contracts and ownership in particular won him the 2016 Nobel Prize in Economics, alongside Bengt Holmström. It was an honour and delight to speak to him about his work and its applications in various contractual relationships. (The interview is edited for clarity and brevity.)

Trevor Chow: Professor Hart, thank you for agreeing to come and speak about your work. I want to begin with your Nobel Prize winning work on contracts - in particular, why are long-term contracts inherently incomplete and what are some of the main problems that arise when not all outcomes can be specified?

The basic issue is that in a long-term relationship, it is very difficult to foresee all the things that can happen during the course of the relationship and include them in the contract as contingencies that the contract deals with. One example of this is the situation right now, with the coronavirus upending various business relationships. Many people who write contracts may find that they cannot fulfill the terms of their contract, or find it unreasonable to do so, due to the nature of this current crisis. Within the contract, they did not say how the relationship would be adjusted if a pandemic of this scale hits within the contract. If I am a small business renting my premises from you and we had foreseen this, it might have been in the contract that I don’t pay rent for three months, or maybe I pay less now and pay more later when things pick back up.

This is an example of why all the contingencies you could imagine being part of a contract are not all included. Economists and many others think of an ideal contract as state contingent - it would say, you will supply this much to me at this price if this event happens, and if that event happens, you’ll supply a different amount of a different good at a different price. That is, it would be full of contingent clauses - but this is too difficult to do in real life, causing the incompleteness and messiness when unanticipated events occur. Normally, the change in circumstances is much less drastic, and is a local shift that calls for a different business relationship that you didn’t prepare for in the initial contact. All these problems arise because there are so many things possible you cannot put them all down.

TC: What does the existence of these incomplete contracts incentivise in terms of distorting business behaviour and producing poor outcomes?

A lot of the thinking about incomplete contracts is not yet in its final state - we are still considering how parties respond to contractual incompleteness. The natural way to respond is to renegotiate - for example, if there is a coal mine supplying a type of coal to a power plant and circumstances change such that a different type of coal is required, the power plant and the coal mine could reach a new agreement. There are circumstances where this works rather well, and the fact that we didn’t put the contingent clauses in the original contract doesn’t really matter that much, since we reached an efficient outcome via the bargaining in the renegotiation. However, one line of the literature has argued that this will not occur that easily, as there might be lots of arguments about how much the price should change. That kind of argument can be quite costly and time consuming, and may lead to some bad feeling - these ex post renegotiation costs may be quite considerable in practice. Because one side may be dependent on the other, one side can raise their prices significantly above cost, holding up their business partner.

If businesses anticipate this, they may act to make themselves less dependent on the other party, such as by moving closer to another coal mine to preserve their options, which may be inefficient. Yet another reaction to this could be the purchase of the coal mine. This is a form of vertical integration, and the literature on this starts with Ronald Coase. Coase’s The Nature of the Firm (1937) first raised the question of why we have firms at all, if markets are so good at allocating efficient - one reason I have examined is the desire to avoid this dependency by integrating vertically. Once the mine is purchased, the firm may be less vulnerable to hold up, as the only mechanism of hold up is the coal mine manager threatening to deprive the firm of their managerial skill. They can be replaced by another worker, and thus the hold up is less problematic, especially if the value of the human capital in the production process is limited.

TC: A solution to the problems arising from incomplete contracts is the idea of a formal relational contract, similar to the keiretsu concept in Japan. In many ways, it appears to be like an employment contract - with a greater focus on the relationship rather than the specific exact details. Could you explain what these are and where these contracts are most useful?

There is a large quasi-legal literature on relational contracts, pioneered by the likes of Macneil and Macaulay. As people realize they are in long-term relationships, they often do not act as opportunistic short-term maximisers. This is because it may hurt them in the longer-term if hold ups are reciprocated, and we don’t want to get into this tit-for-tat behaviour. The game theory literature on these sorts of repeated games is extensive, with the prisoner’s dilemma being a prominent example of how a one-off game may encourage defection, while an infinitely repeated game would result in cooperative behaviour. The work I’ve done involves formal relational contracts - this is different in that it does not rely on infinitely repeated play, as well as being much more designed. The idea is that when we enter into a business relationship, we sit down and write the contract, but we recognise from the beginning that trying to think of all the things that can go wrong is a fool’s errand. We are not going to manage to do it, and it might be counterproductive to try and put as many detailed things as possible into the contract. Instead, a better approach might be to have a simpler contract that lays out some of the essential elements, but then supplement that with principles that we are going to use to fill out the contract when it is incomplete - when there are occurrences that the contract doesn’t govern, these are the guiding principles we use to fill in the gap.

This work came about by working alongside David Frydlinger, a Swedish lawyer, and Kate Vitasek, who works in supply management in the University of Tennessee. I met Frydlinger when I went to get the Nobel Prize in 2016, as his law firm was one of the places I was invited to speak at - he had heard that I spent the last decade working on behavioural contract theory, which resonated with the ways he was advising his clients on writing better contracts. David and Kate’s track record of implementing these contracts with guiding principles accompanied the theory I was looking at. There are six norms within their process, of equity, loyalty, honesty, integrity, autonomy and reciprocity. We don’t just write these norms down, but we talk about what these values mean to each party, such that when something unanticipated happens later on, we have agreed in advance how we might renegotiate while not being selfish. Although they sound informal, they are enshrined within the contract, making it a formal relational contract where both parties will try and understand each other’s perspective. This makes renegotiation easier, but the theory I have been developing also shows that it is less likely either of us feels aggrieved. If there isn’t this feeling of fairness, I may find some way to get back at the other party, such as by cutting my costs at the expense of quality, which hurts you. We avoid this by ensuring contractually that the negotiations will be mutually acceptable.

With respect to where these formal relational contracts apply, my colleagues seem to think that these contracts can apply anywhere. There is no need to have this complex machinery for a simple transaction, where a normal contract works fine because it is a short-term relationship, or because the good is fairly standard. The cost of doing the whole process is quite burdensome in that case, since we have to communicate at the beginning and throughout the relationship. However, if it is a more complex relationship, it can work across various businesses, as long as both parties sign onto it. What David and Kate find with clients is that some have a mindset that is not fully into this model, and much like the fact that not every friendship or marriage works because not everyone hits it off, both parties have to be willing to think in these ways, which may not be possible due to internal organizational structures or some other issues.

TC: Moving beyond business-to-business contracts, another contractual relationship is between companies and shareholders. You’ve spoken about an alternative take on shareholder primacy - where the duty is to maximize shareholder welfare, rather than shareholder value. By what mechanism can boards and companies take into account the broader interests of shareholders?

Let me summarize what I did with my co-author Luigi Zingales. We wrote a paper where we took issue with the dominant way of thinking about a company’s duties, which is epitomised by Milton Friedman famously arguing that the only purpose of a firm is to make money for its shareholders, with social goals being left to other parties. However, since shareholders themselves have non-monetary interests, some may prefer a reduced profit with a more socially beneficial outcome. To be clear, our argument is not to be confused with stakeholder arguments, which argue that companies should act on behalf of workers, consumers or other stakeholders. We don’t go down that line of argument, because most companies seem to be set up to act on behalf of shareholders, because they have the votes. Indeed, you can set up a company that isn’t just for the shareholders, and cooperatives do exist which act on behalf of other stakeholders such as workers. However, it isn’t a great idea for governments to compel companies that are not set up in such a way to do so, as it is an intrusion into private contractual relationships.

As such, what we advocate is that the fiduciary duty of a company’s board is not only to maximise profits, because shareholders have other objectives. Boards could ask shareholders what they want via a vote, and not doing so might in fact be disloyal. We support increased shareholder engagement, and if you are a shareholder, divesting from what you see as socially disfavourable company may be counterproductive, since the shareholders left will be advocating for disfavourable actions. One thing you could instead have is mutual funds and index funds that are designed to push companies they have shares in towards more pro-social directions. Socially inclined investors could invest in these funds if they were interested in the aims the fund was pushing for, such as reducing carbon footprint or not trading with corrupt regimes.

TC: Contract theory is perhaps less well known to the general public compared to some other economic subfields. What is the frontier of the current research that you find particular exciting?

I don’t think people are aware of it at all - if I said I worked on contracts, they would think I’m a lawyer, and even within the profession, it isn’t that huge compared to some other subfields such as game theory. Perhaps that’s because there’s been a movie about game theory, and we haven’t had one yet on contract theory. In terms of what’s interesting, I would divide contract theory into two parts. The first is what my co-laureate Bengt Holmström has worked on, which is principal-agent theory. The second is incomplete contracts, which is a much messier area. Just taking our discussion on formal relational contracts, this is still very new and has not caught on with most of the profession yet, let alone the majority of firms writing contracts. That is certainly an area for development, but more generally, incomplete contracts are everywhere, ranging from treaties between governments, constitutions or marriages. The theory could be fleshed out across many of these avenues, because there are so many situations where what people agree on in advance does’t always work out, and if you can find ways to ensure that things get adjusted in a good way, that would a huge improvement.

TC: I’d like to thank you for giving up your time to talk about contract theory. It has been very fascinating to find out more about the ways in which we can understand various contractual relationships and some mechanisms to design contracts and firms in such a way as to make them more beneficial.

Thank you.

[ economics  interviews  ]