The 2016 Nobel Prize: Incentives, Property Rights, and Ownership
The 2016 Nobel Prize in economics was awarded today to Oliver Hart and Bengt Holmström, two of the most eminent economists specializing in the analysis of contracting and organizational structure. Hart, a British economist who teaches at Harvard, and Holmström, originally from Finland and now teaching at MIT, are leading practitioners of the formal, mathematical analysis of firms, contracts, and organizations. Hart is best known for his contributions to the “incomplete-contracting” or “property-rights” approach to the firm, while Holmström is considered the founder of modern principal-agent theory. Both have long been considered frontrunners for the prize (Paul Krugman tweeted that the two are so obviously deserving his first thought was, “didn’t they have it already?” — something said about Krugman’s own Nobel by No One Ever).
As Nicolai Foss pointed out in an insightful 1994 article, and I have also discussed in my own work, the Austrian economists anticipated many of the points in the modern economics of contracting and organization. Take principal-agent theory. While Mises’s 1920 article on socialist economic calculation offers the definitive statement on why a rational allocation of resources is impossible under common ownership of the means of production, it is sometimes forgotten that Mises 1922 book Socialism, which includes the calculation argument, also includes a lengthy discussion of incentives under socialism, and Hayek discusses the problem in detail in his contribution to Collectivist Economic Planning. Mises also addresses the separation of ownership and control in Human Action.
The Principal-agent Problem
Holmström is best-known for a specific interpretation of the principal-agent problem (articulated in his most-cited article, “Moral Hazard and Observability”). Holmström examines a situation in which one party (the “principal”) has some task to be performed, but must hire another party (the “agent”) to perform the task. Performing the task is costly to the agent, so the principal must provide some incentives to get the agent to do what the principal wants. If the principal cannot observe or understand the agent’s actions directly — which is plausible, otherwise the principal could simply perform the task — then an incentive contract based on some observable, but noisy signal of output is problematic, because it exposes the agent to risks related to the noisiness of the signal. (For example, incentivizing salaried managers by giving them stock options helps to align their incentives with those of shareholders, but exposes them to the risk of market fluctuations that are caused by the Fed, by the actions of other firms, or other forces beyond their control.) Hence principals face a specific tradeoff between providing incentives for agents (by using performance-based pay) and insuring agents against risks beyond their control (by using fixed salaries). Exactly how this tradeoff should be managed depends on the particulars of the situation.
While the Austrians did not explore the specific aspects of the agency problem highlighted by Holmström — this tradeoff between incentives and insurance — they did recognize the importance of the agency problem, and Mises in particular offered a nuanced discussion of various means by which principals mitigate agency problems, anticipating a well-known argument of Henry Manne (see my discussion here). Austrian ideas about entrepreneurship, heterogeneous capital, and tacit knowledge can also improve our understanding of agency problems (Foss et al. 2007). (For further discussion of Austrian economics and agency problems see Padilla 2003 and Hülsmann 2008).
Property Rights and Ownership
While entrepreneurship does not get a mention in the Nobel citation, Hart’s insights on property rights and ownership have greatly influenced my own understanding of the entrepreneurial function. Following Frank Knight, I have argued that entrepreneurs establish firms because entrepreneurial judgment is non-contractible. In other words, an entrepreneur’s idiosyncratic understanding of the future — what Knight and Mises call judgment — can only be expressed through ownership of resources. Ownership provides the ultimate control of productive resources, meaning that owners get the final say in how resources will be used. A resource owner can contract for advice (e.g., by hiring a consultant), but only the owner can make the final decisions about deploying resources (the owner chooses whether or not to hire a consultant, which consultant to hire, and whether to act upon the consultant’s advice). Judgment is the act of exercising this form of control over resources. In other words, there are markets for assets, but not markets for judgment, because judgment implies the right to control an asset. To exercise judgment, the entrepreneur most own assets, i.e., must establish or maintain a firm. (This is a somewhat different perspective on entrepreneurship than Israel Kirzner’s well-known approach.)
Hart’s work with Sanford Grossman and with John Moore (hence, the “Grossman-Hart-Moore” theory of the firm) is based on a particular concept of asset ownership. Ownership, in this approach, is defined as residual rights of control — i.e., the right to decide how an asset will be used in situations not covered by prior agreement. In a world of perfect certainty (akin to Mises’s evenly rotating economy), individuals could write very complex and detailed contracts about how various resources will be used under particular circumstances. With contracts like these, “ownership” is vague and indeterminate; it doesn’t matter who has formal title to an asset because the asset will be used in exactly the same way regardless.
In the real world of uncertainty, however, such contracts are impossible, because we cannot anticipate every potential future event and agree in advance on what we would do. In other words, all feasible contracts are “incomplete,” meaning that they contain some omissions or gaps. Property rights are thus necessary to fill in the gaps. If party A is the owner of the asset, then even though party B may have certain contractual rights to use the asset in particular ways, if an unforeseen contingency arises, the decision is made by party A. This is a useful and parsimonious way to think about ownership.
In Hart’s work (his 1995 book is the most accessible treatment), these ownership incentives are important because they affect what kinds of investments parties are willing to make in creating and maintaining specialized assets. I find their theory of the firm insightful, though there are many potential problems with the details. More generally, though, their understanding of ownership helps clarify some ambiguities about property rights. For example, people often confuse residual control rights – the right to determine how a resource will be used when unanticipated circumstances arise – with residual income rights. But residual income rights can be partly delegated to non-owners; I can own a restaurant collect a fixed fee, while letting the manager keep the net profit. That does not make the manager the owner of the firm, because he cannot change this arrangement, or sell the restaurant, or do a host of other things without my permission.
Mises works with a similar definition: “Ownership means full control of the services that can be derived from a good” (Human Action, p. 678). He goes on in that passage to distinguish this “catallactic” definition with the legal definition; i.e., ownership for economics means de facto control, not de jure control. (Under interventionism, as Mises points out, the control of legal owners is severely circumscribed by government policy.) Mises also contrasts the ultimate control afforded by ownership with the proximate, day-to-day control rights that can be delegated to non-owners, in the context of the principal-agent problem in corporations: “The general direction of a corporation's conduct of business is exercised by the stockholders and their elected mandataries, the directors. The directors appoint and discharge the managers. ... A successful corporation is ultimately never controlled by hired managers (Human Action, p. 304).” As I have emphasized in my own work, Mises always associated entrepreneurship exercised under uncertainty with resource ownership. Outside the imaginary construction of the evenly rotating economy, “A capitalist is always also virtually an entrepreneur and speculator” (Human Action, p. 254).
My overall reaction to this year’s Nobel is positive. As with Oliver Williamson’s work, there is much in the writings of Holmström and Hart for Austrian economists to admire. Unlike Williamson, they are less explicitly influenced by thinkers such as Hayek, and they write in the style of modern mathematical economists. But their work is worth careful study and has interesting implications for Austrians — something I would definitely not say about all the recent Laureates.