The Political Economy of Moral Hazard

This article originally appeared as “The Political Economy of Moral Hazard” in the Czech journal Politická ekonomie (February 2006): 35–47.

A central occupation of economists is to analyze the nature, causes, and effects of incentives — the circumstances that are held to motivate human action. Economists agree on the positive role that “good” incentives play to increase production. They also agree that “perverse” incentives have an opposite impact. One of these perverse incentives is called moral hazard, the subject of our present essay.

Moral hazard is the incentive of a person A to use more resources than he otherwise would have used, because he knows, or believes he knows, that someone else B will provide some or all of these resources. The important point is that this occurs against B’s will and that B is unable to sanction this expropriation immediately. The mere incentive to rely on resources provided by others is not per se problematic. For example, the announcement of a future inheritance might prompt the prospective heir to spend more in the present than he would otherwise have spent. In such cases we would not speak of moral hazard. A genuine moral-hazard problem appears however if A has the possibility to use B’s resources against B’s will and if he knows this. Laymen would call A’s incentives a “temptation to steal” or a “temptation to act irresponsibly.” Economists, ever weary of moralizing, have espoused the technocratic expression “moral hazard.”

Thus the essential feature of moral hazard is that it incites some people A to expropriate other people B. The B-people in turn, if they realize the presence of such a moral hazard, have an incentive to react against this possible expropriation. They make other choices than those that they would consider to be best if there were no moral hazard.

Many economists have therefore concluded that moral hazard entails market failures; it brings about a different allocation of resources than the one that would exist in the absence of moral hazard. Conventional economic theory explains moral hazard as a consequence of the fact that market participants are unequally well informed about economic reality. In other words, moral hazard results from “asymmetries of information” and the theory of moral hazard is therefore considered to be a part of the economics of information.

In the present paper we will criticize this conventional approach and propose an alternative. We will argue that information asymmetries are just one among several causes of moral hazard. Most importantly, they entail disequilibria and the expropriation of third parties only accidentally and ephemerally, because these third parties can avoid them by better judgment (by anticipation). By contrast, moral hazard also results from government interventionism; and in this case it creates disequilibria and expropriation of a sort that cannot be avoided, not even in the absence of informational asymmetries.

We will first present the conventional “informational” theory of the causes and consequences of moral hazard. Then we will study moral hazard on the free market and under government interventionism. Finally we will outline how our findings fit the important case of government interventionism in money and financial markets.

The Conventional Theory of Moral Hazard

The fact that people act on the basis of different knowledge about the real world will hardly be contested. The baker knows other things than the astronaut, the opera singer other things than the teacher of mathematics. Neither can it be doubted that people are unequally well informed about the real world. Some bakers know more about cakes, cake making, and the cake market than others, and so on. In short, information asymmetries are a universal aspect of human life as we know it. They are both a cause and a result of the division of labor. There is no reason to suppose that they are a priori harmful or a sign of imperfection. Conventional theory therefore stresses an additional condition to explain the emergence of moral hazard, namely, the separation of ownership and control. Two main cases can be distinguished: co-ownership and agency contracts.

In the case of co-ownership, any one owner has control over a given piece of property, but not exclusive control. Informational asymmetries can then produce moral hazard in conjunction with this separation of ownership and control. Whenever one co-owner of a swimming pool cannot effectively monitor the activities of his fellow-owners, the latter have an incentive to swim without cleaning up, repairing the fences and so on, thus increasing their own (monetary and psychic) income at his expense.

Similarly, in the case of an agency contract, moral hazard can arise when an economic good is not effectively controlled by its owner (the “principal”) but by a different person called the “agent,” for example, by an employee. Again, informational asymmetries produce moral hazard in conjunction with this separation of ownership and control. The agent, who is fully informed about his own activities, has an incentive to act in his own material interest against the material interests of his less informed principal. Whenever the principal cannot effectively monitor the activities of his agent, therefore, the latter has an incentive to increase his own (monetary and psychic) income at the expense of the former.

The standard case of moral hazard in an agency setting is an insurance contract. Here the insurance company is the less informed principal and the insured person is the agent. Automobile insurance, for example, creates a moral hazard for drivers; it creates an additional incentive for risky driving because other people (other clients of the insurance) will pay a part of the costs of the agent’s accidents. Similarly, in the presence of unemployment insurance, an unemployed person has an additional incentive to stay unemployed because other people will pay at least a part of his living expenses. Or, in the presence of health insurance, insured people will have an additional incentive to engage in risky activities or lifestyles because others will pay at least a part of the treatment in case of illness or accidents.

However, moral hazard is in no way a particular problem of the insurance industry. It can arise in almost any other field of human activity where there is a separation of ownership and control. Employees can be subject to moral hazard to the extent that they can reduce their efforts without fearing reduced pay. Debtors may be subject to moral hazard if they believe they can squander the money without negative consequences when they turn out to be incapable of paying back. Certain auditing firms have been subject to moral hazard when they sold consulting services to the very companies they were supposed to audit (for example, in the Enron case). A central bank can produce moral hazard in the banking community if the commercial bankers perceive the central bank as a lender of last resort. The IMF can produce moral hazard among debtor governments. Taxpayers are said to be subject to moral hazard if they can evade high-tax regions, and so on. Similarly, in the literature on public choice and constitutional political economy, governments and parliaments are often portrayed as agents prone to moral hazard, whereas the voters are the less informed principals.

It is a rather significant fact that, in the light of the conventional theory, information asymmetries produce moral hazard only in conjunction with the separation of ownership and control. We will argue that the latter is indeed the decisive element, whereas information asymmetries are but a sideshow. More to the point, we will argue that moral hazard systematically entails expropriation only when ownership and control of a resource are separated in a special way, namely, without the consent of the owner.

To demonstrate our contention, we will first analyze the voluntary separation of ownership and control, as it occurs on a free market and then turn to the most important case of involuntary separation, namely, the case of interventionism.

Moral Hazard on the Free Market

When we speak of a free market, we mean that each property owner has the unhampered right to use his property as he sees fit. This includes the right to let other people use his property under conditions he approves, as well as the right to co-own property with other people. In short, it includes purely voluntary separations of ownership and control.

The question we have to answer now is whether such a voluntary separation of ownership and control is per se likely to lead to any sort of expropriation due to information asymmetries. In other words, we have to examine whether the conjunction of these two conditions is sufficient for one man to be able to enrich himself at the expense of another man, or whether, to the contrary, even in the presence of the combined two conditions, such expropriation of man by man is a contingent consequence. We will argue that moral hazard does not necessarily entail expropriation whenever information asymmetries combine with a separation of ownership and control; that, whenever moral hazard results from them accidentally, there are strong forces at work to eliminate expropriation; and that moral-hazard-induced expropriation is therefore not only accidental, but also ephemeral on the free market.

The decisive fact on which we can base our argument is that the interplay between information asymmetries on the one hand, and the separation of ownership and control on the other hand, finds on the free market an effective antidote in expectations. Consider the problem of principal-agent relationships. Suppose the owner of a grocery shop hires a clerk for a salary of €1,000 per month. Suppose further that that the owner has various other shops to run and that he can therefore not effectively supervise his new clerk all the time. According to the conventional theory, the clerk is likely to behave differently than he would in the permanent presence of his employer. He can have longer naps over the lunch break, be sloppier in dealing with customers, and so on.

All of this, we can admit right away. But the crucial question is: does the sloppy clerk necessarily enrich himself at the expense of his employer? Does the principal necessarily bear costs entailed by the unfaithful agent? Only if this were the case could we speak of expropriation. But it is not the case. The clerk does not necessarily enrich himself at the expense of the owner because the latter can anticipate the behavior of the former. To the extent that he does anticipate it, the payment of €1,000 would properly reflect the discounted marginal value product of sloppy labor. And this is of course precisely what prudent employers do, on a free market, whenever they hire new people. They discount the marginal value product of good or normal work by a factor that represents the risk of bad work due to lack of supervision.

The role of expectations can hardly be overstated in the theory of moral hazard. To the extent that the expectations of the principal are correct, moral hazard on the part of the agent cannot lead to a situation in which the latter expropriates the former.

Moreover, principals acting on a free market do not merely have to rely on their foresight of the future actions of their agents. They are also free to design contractual relationships in ways that minimize (a) the danger of moral hazard arising in the first place and (b) the danger of moral hazard, once there, affecting them negatively. The insurance industry offers well-known illustrations. In health insurance, for example, exclusions, deductibles, and copayments are tools intended to reduce moral hazard. Similarly, moral hazard in road traffic has been effectively reduced with the help of radar controls and black boxes in vehicles.

Most importantly, on a free market, principals have the liberty to sever contractual relations with agents at any time. The threat of being fired is probably the most powerful incentive to deliver diligent work rather than succumb to moral hazard. Reputation effects and blacklists operate in the same direction. Analogous considerations apply to all forms of co-ownership. Here the various co-owners, being aware that the utilization of the common resource must be restricted to preserve its value, have an incentive to set up and accept rules for its use. If the number of co-owners is high, individual capitalist-entrepreneurs might step in and provide the co-property and the rules, as for example in urban development.

The institutional settings designed to neutralize moral hazard are usually considered to be “second-best” solutions, as distinct from the “first-best” solutions that would exist in a world inhabited with human beings that enjoy perfect information. Based on this distinction, some writers infer that competitive markets are inefficient in the presence of moral hazard. In other words, moral hazard would be a systematic cause of market failures even on a free market.

However, the distinction between first-best and second-best solutions is a figment of the imagination. It is based on the premise that human beings could choose to inhabit a world in which they would be endowed with perfect foresight and in which they would not need any institutions designed to deal with moral hazard. Only in comparison to such an alternative world could the world that we know — the world with information asymmetries and moral hazard — be called inefficient. But the alternative does not exist. The only meaningful standards to evaluate the efficiency, say, of deductibles and copayments are other contractual provisions or similar devices that could be used to attain the same end.

Moral Hazard and the Definition of Property Rights

Before turning to the analysis of moral hazard under government interventionism, it might be useful to first consider the hybrid or intermediate case in which property rights are not clearly defined. The case is not only interesting from a systematic point of view, but also because, in practice, for a great number of economic goods there are no clearly defined property rights. Most notably we might think here of air and oceans and their inhabitants; birds and fish are considered to be the commonwealth of humanity. Human beings appropriate these resources with only feeble attempts at creating something like property rights. Thus we have a system of co-ownership without rules. Not surprisingly, the result is universal moral hazard; everybody has an incentive to use the resource in question as much as possible. This is a sure recipe for quick depletion, as has been known at least from the times of Aristotle. In more recent years, the phenomenon has been called “tragedy of the commons.”

Now, in this practically highly important case, information asymmetries play no role whatever. Moral hazard would be at work here even if each co-owner were perfectly aware of the activities of his fellow owners. We could even say that moral hazard would be here at work especially in the case in which every co-owner was perfectly aware of the activities of his fellow owners.

Interventionism and Moral Hazard

Important though moral hazard stemming from a deficient definition of property rights might be in practice, it is no match for moral hazard that results from a forced separation of ownership and control. To this type of moral hazard we turn now.

By a “forced” separation of ownership and control we mean a separation brought about against the will of its owner. Although owners might be forced both by governments and by private parties, government interventionism is far more important in practice. This is so not only because of the greater quantitative impact, but also because, in our western societies at least, interventionism is usually enshrined in the law and thus can be anticipated.

Let us first stress that government interventionism must not be confused with a mixed economy. In the latter the government is one of several owners and it controls only its own property. By contrast, an interventionist government commands other property owners to use their resources in a different way than these owners themselves would have used them. In so doing, the interventionist government makes some person or group A (for example itself) the uninvited co-owner of other agent B’s property. The essence of interventionism is precisely this: institutionalized uninvited co-ownership.

Government makes itself the uninvited and unwanted co-owner whenever it taxes, regulates, and prohibits. The specific forms of taxation, regulation, and prohibition are myriad. The important fact is that any form of government interventionism, by its very nature, entails a forced separation of ownership and effective control.

Taxation means that the government proclaims itself the owner of (a certain share of) resources belonging to its subjects; and that it forces them to eventually hand over these resources, which the latter would not have yielded voluntarily (otherwise one would not speak of taxation, but of donations to the government). Today taxation does not concern concrete physical items, but their monetary equivalent. It follows that, until the tax is paid, the government imposes itself as the co-owner of virtually all physical assets of the taxpayers. However, until the tax is paid, the resources in question are typically controlled by the citizen.

Regulation means that the government proscribes a certain use of certain resources. This use is typically not the one that the citizens would have chosen (otherwise the regulation would be pointless). Again, the government thereby proclaims itself the co-owner of these resources. Consider the case of price controls. If the government fixes a minimum wage rate, it effectively proclaims itself the co-owner of workers, because it does not allow them to work under conditions they see fit. And it also proclaims itself the co-owner of the capitalists or, more precisely, of the money that the latter plan to spend on labor. However, the government does not permanently control the actions of the workers, and it does not interfere with other uses of the capitalists’ money.

Prohibition means that the government outlaws a certain use of certain resources altogether. Again, it thereby proclaims itself the co-owner of all resources that could be put to the prohibited use. For example, if it prohibits the production and sale of alcoholic beverages it effectively imposes itself as a co-owner of all resources that could be used for the production and sale of such drinks. However, it does not interfere when those resources are used in other employments.

Interventionism does not abolish private property. The citizens still have ownership and control of their property, even though they have to share both ownership and control with the government and its agents. It is true that this forced co-ownership is usually a matter of degree. Increased interventionism increases the share of government control of resources, though without outlawing other people’s simultaneous control of these same resources. But the forced nature of the co-ownership itself is not a matter of degree. It is a categorical and essential feature of any intervention, be it ever so small.

Government interventionism always and everywhere entails a forced separation of ownership and control. It always and everywhere creates unwanted “partnerships” between the citizens and their government. It follows that, by its very nature, it creates a moral hazard both for the citizens and for the government. Most importantly, it creates a situation in which each of the parties involved (the citizens on the one hand and the government on the other hand) desires to expropriate the resources subject to interventionism at the expense of the other parties.

First consider the reaction of the citizens to taxation, regulation, and prohibition. From the very fact that intervention entails forced co-ownership, it follows that the citizens have an incentive to evade the intervention. They can to some extent evade it because they have some control of their property. To avoid taxation, for example, they can choose to invest capital in a country with low taxes rather than in a country with high taxes; they can choose to emigrate to low-tax countries rather than stay in high-tax places; they can choose a profession that is less taxed than other professions; or they can choose to make fraudulent declarations of their income and capital. To evade regulations, they can choose not to buy or sell commodities subject to price controls, or they can choose to buy and sell them on the black market. To evade prohibitions, they can buy and sell prohibited items on the black market. However, operating on the black market is risky and thus very costly, and evasion to other countries is costly too. Thus it follows that there is an incentive for the citizens to use a greater part of their property for personal consumption rather than invest it. Hence, the general tendency of interventionism on the citizens is to entail excessive consumption and to make production more costly because of the necessity to evade the intervention.

But moral hazard also comes into play on the side of the government itself. Governments rely on the resource use that comes through taxation and regulation. They will therefore tend to tax more and regulate more in order to neutralize the ways in which the citizens evaded its previous intervention. It will seek to “close the loopholes.” We have here a basic mechanism of the dynamics of government interventionism. Interventionist governments have an incentive to extend taxation to all branches of economic life; to regulate industries that have so far escaped regulation; and to beat into submission the countries that serve as tax havens. The ultimate result is to reinforce the tendencies that we characterized above: excessive consumption and insufficient production; in short, a general impoverishment of society.

It can hardly be overemphasized that it is not co-ownership per se that causes these excesses. They are caused by the uninvited and unwanted co-ownership that springs from interventionism.

Notice that our foregoing analysis in no way depends on the existence of asymmetrical information. We can assume for the sake of argument that all citizens are perfectly aware of the government’s activities, and that the government is also perfectly well informed about the activities of all citizens. All of this would not alter the picture. Government interventionism entails moral hazard both on the side of the government and on the side of the citizens. And this moral hazard cannot be neutralized by choosing appropriate contractual devices, because it has no contractual basis at all; it is imposed. It cannot be sidestepped by choosing to avoid the moral-hazard-prone situation altogether, because the situation itself is imposed. The very meaning of interventionism is, as we have said, to overrule the choices of property owners.

And similarly, the workings of moral hazard cannot be eliminated or diminished by correct expectations, as in the case of moral hazard on the free market. The case is exactly the reverse. It is precisely when the citizens correctly anticipate how high the next tax will be, and when it will hit them, that a moral hazard will start bearing on them and incite them to evade the tax.

The Case of Fiat Money

Let us now turn to the discussion of an important case in which government interventionism produces moral hazard on a large scale: monetary interventionism.

The fundamental intervention, on which all other interventions in this field are built, is the imposition of a legal tender. The latter is a means of payment that the market participants are obliged to accept, even if they made contracts that stipulated payments in terms of other media of exchange. This intervention creates a moral hazard for the market participants to hoard or export the media of exchange that in their eyes are better than the legal tender, but which they are legally bound to use at par with the legal tender. The paradoxical result is that only the legal tender — the medium of exchange that everybody seeks to avoid — remains in circulation. Economists call this phenomenon “Gresham’s Law.”

All contemporary monetary systems are based on such legal tender privileges. Paper money — as well as electronic money (”central-bank liquidity”) — does not compete with other monetary products on the free market, but is imposed by special privilege. It is therefore called fiat paper money. This institution is of great interest from the point of view of the theory of moral hazard. In fact it entails moral hazard on the greatest imaginable scale, again, both on the side of its producer and on the side of its users.

Fiat paper money creates moral hazard for the producer because he has the possibility of creating ex nihilo virtually any amount of money and, thus, to buy virtually any amount of goods and services for sale. The only limit to this capacity is the hyperinflation that invariably results in the case of a great inflation of the money supply.

But fiat paper money also creates moral hazard on the side of the money users — the citizens, the banks, and the governments — because they sooner or later come to realize that the masters of the printing press have the power to bail out virtually any bankrupt firm or government. Thus they engage in more or less reckless financial planning, expecting that the monetary authorities will not allow a great mass of reckless planners to go bankrupt. This speculation has been borne out by the last thirty years. Public and private debts are at record heights all over the world.

Monetary theorists were aware of this danger early on, even though they did not use the word “moral hazard” in this context. Here we should mention in the first place all defenders of sound money — that is — of competitive money, such as Ludwig von Mises, F.A. Hayek, Murray Rothbard, and many other economists of the Austrian School. But neither has the problem escaped the attention of the champions of fiat money. Nineteenth-century economist Walter Bagehot, who coined the term “lender of last resort,” warned that the central bank, while freely lending in times of financial duress, would have to avoid at any cost the impression that it would bail out the market participants. Some sixty years later, Herbert Simon stressed the need for rules rather than discretion in economic policy, essentially for the same reasons. And his disciple Milton Friedman reiterated this point even more famously in his proposal for a constant growth rate of the money supply.

The idea that rules could prevent large-scale moral hazard in a fiat paper money system defies human logic. After all, the only possible use of a printing press is to produce more money than would have been produced on a free market. What sense does it make to assert that this power could be handled responsibly? In analogy to a felicitous phrase from Anthony de Jasay, we might say that praising the virtues of monetary stabilization while maintaining legal tender laws for paper money is like using a chastity belt for Miss World while letting her keep the key. Let us add, though, that, to make the analogy with our monetary constitution perfect, we would have to advertise day in day out in the New York Times that Miss World definitely owns the key to her chastity belt.

Financial bubbles are the unavoidable result of such a state of affairs. If more or less every major participant to the financial market is subject to moral hazard, then in due time even the smaller traders realize that the bigger fish play the moral hazard card, and thus they too venture to set out on the same path. This means that the market participants sooner or later come to base their plans on the availability of a far greater quantity of goods and services than is really available in the economy. In short, paper money by virtue of its mere existence produces massive error on a large scale, until the bubble bursts in a crisis.

Again, as in the other cases of interventionism-induced moral hazard, these effects cannot be neutralized, avoided, or diminished through anticipations. And this means that they cannot be managed through the management of expectations. Mainstream economists have in the past thirty years labored to bring expectations into the picture of monetary policy. The next thirty years will presumably be needed to take account of institutionalized moral hazard. The problem is that our monetary system will not survive that long, if we can extrapolate the speed of the events of the past thirty years.