[An updated version of this article can be found at Information in the 2nd edition.]
In the past two decades, an important strand of economic research, sometimes referred to as information economics, has explored the extent to which markets and other institutions process and convey information. Many of the problems of markets and other institutions result from costly information, and many of their features are responses to costly information.
Many of the central theories and principles in economics are based on assumptions about perfect information. Among these, three stand out: efficiency, full employment of resources, and uniform prices.
At least since Adam Smith, most economists have believed that competitive markets are efficient, and that firms, in pursuing their own interests, enhance the public good "as if by an invisible hand." A major achievement of economic science during the first half of the twentieth century was finding the precise sense in which that result is true. This result, known as the Fundamental Theorem of Welfare Economics, provides a rigorous analytic basis for the presumption that competitive markets allocate resources efficiently. In the eighties economists made clear the hidden information assumptions underlying that theorem. They showed that in a wide variety of situations where information is costly (indeed, almost always), government interventions could make everyone better off if government officials had the right incentives. At the very least these results have undermined the long-standing presumption that markets are necessarily efficient.
Full Employment of Resources
A central result (or assumption) of standard economic theory is that resources are fully employed. The economy has a variety of mechanisms (savings and inventories provide buffers; price adjustments act as shock absorbers) that are supposed to dampen the effects of any shocks that the economy experiences. In fact, for the past two hundred years economies have experienced large fluctuations, and there has been massive unemployment in the slumps. Though the Great Depression of the thirties was the most recent prolonged and massive episode, the American economy suffered major recessions from 1979 to 1982, and many European economies experienced prolonged high unemployment rates during the eighties. Information economics has provided explanations for why unemployment may persist and for why fluctuations are so large.
The failure of wages to fall so that unemployed workers can find jobs has been explained by efficiency wage theories, which argue that the productivity of workers increases with higher wages (both because they work harder and because employers can recruit a higher-quality labor force). If information about their workers' output were costless, employers would not pay such high wages because they could costlessly monitor output and pay accordingly. But because monitoring is costly, employers pay higher wages to give workers an incentive not to shirk.
While efficiency wage theory helps explain why unemployment may persist, other theories that focus on the implications of imperfect information in the capital markets can help explain economic volatility.
One strand of this theory focuses on the fact that many of the market's mechanisms for distributing risk, which are critical to an economy's ability to adjust to economic shocks, are imperfect because of costly information. Most notable in this respect is the failure of equity markets. In recent years less than 10 percent of new capital has been raised via equity markets. Information economics explains why. First, issuers of equity generally know more about the value of the shares than buyers do, and are more inclined to sell when they think buyers are overvaluing their shares. But most potential buyers know that this incentive exists and, therefore, are wary of buying. Second, shareholders have only limited control over managers. Information about what management is doing, or should be doing, to maximize shareholder value is costly. Thus, shareholders often limit the amount of "free cash" that managers have to play with. They do so by imposing sufficient debt burdens to put managers' "backs to the wall" so that managers must exert strong efforts to meet those debt obligations, and so that lenders will carefully scrutinize firms' behavior.
The fact that firms cannot (or choose not to) raise capital via equity markets means that if firms wish to invest more than their cash flow allows—or if they wish to produce more than they can finance out of their current working capital—they must turn to credit markets, and to banks in particular. From the firm's perspective borrowing has one major disadvantage: it imposes a fixed obligation on the firm. If it fails to meet that obligation, the firm can go bankrupt. (By contrast, an all-equity firm cannot go bankrupt.) Firms normally take actions to reduce the likelihood of bankruptcy by acting in a risk-averse manner.
Risk-averse behavior, in turn, has two important consequences. First, it means that a firm's behavior is affected by its net-worth position. When its financial position is adversely affected, it cuts back on all its activities (since there is some risk associated with virtually all activities); activities that are particularly risky—such as long-term investments—are cut the most.
Second, it means that if a firm perceives an increase in the risk associated with production or investment—such as when the economy appears to be going into a recession—it cuts back on those activities. Since risk perceptions are notoriously volatile, this too helps explain the economy's volatility.
Similarly, costly information explains why banks ration credit. Why ration rather than simply charge higher interest rates to higher-risk borrowers? Because often the only borrowers who will borrow at high rates are those who are the highest risk, and on whom, therefore, the lenders are most likely to lose. Also, higher interest rates may even induce borrowers to undertake greater risks.
Banks, in turn, can be viewed as highly leveraged firms that borrow from depositors. Their "production" activity is making loans (screening loan applicants, monitoring loans, etc.). When their net worth is reduced, or when they perceive that the risk from lending has increased, they (like any other risk-averse firm) cut back on their activities: they make fewer loans. But this in turn has strongly adverse effects on producing firms, particularly as the economy goes into a recession. Firms' cash flows are reduced. To maintain their production and investment levels, given their reluctance to issue equity, they turn to banks for credit. And it is precisely when they need the credit the most that banks may be cutting back their credit rather than expanding it. Thus, the recessionary pressures are exacerbated. As one might expect, these effects are particularly important for small and medium-size firms, for which the issuing of commercial paper is not a viable alternative.
Thus, the characteristics of credit and equity markets—characteristics that can be explained by imperfect, costly, and asymmetric information—help us understand the volatility of the economy. Information economics helps explain economic volatility in another important way. In standard theory, changes in economic circumstances lead to changes in wages, prices, and interest rates. Adjustments in these variables act as "shock absorbers." In fact, Keynes noted that prices, wages, and interest rates are not so flexible, and a major strand of Keynesian research has placed these rigidities at the center of macroeconomic fluctuations.
The explanation of such rigidities remains controversial, however. Perhaps the most convincing explanation is that firms are uncertain of the consequences of their actions, and the larger the change in any action, the more uncertain they feel. The greater their perceived uncertainty, the more conservative their actions. They change prices and wages only slowly because the consequences of changing them are so uncertain.
A third major principle of economics (besides the efficiency of market economies and the fact that resources, including labor, are fully utilized) is referred to as the Law of the Single Price. Under this law, there is a uniform price in the market, and price differences are quickly eliminated by arbitrage. In fact, many markets are marked by noticeable differences in prices. The differences in observed prices and wages are far larger than can be accounted for simply by differences in attributes of, say, location, differences in quality, and nonpecuniary characteristics of jobs. As George Stigler pointed out in a seminal article in 1962, costly information provides a ready explanation: arbitrage is costly. It is costly for consumers to search for the lowest price or the highest-paying jobs.
But the consequences of imperfect information are even more fundamental. Firms recognize that consumers and workers face costly search. In some special cases this may lead each firm, less concerned about losing customers or workers to rivals, to raise its price or lower its wage.
In some cases it has been shown that even though there are many firms, prices might be raised to the monopoly level, even when search costs are very small. To see why, consider a case where all firms charged the same price. If any firm were to raise its price just a little—by an amount less than the cost to customers of switching to another firm—that firm would lose no customers. Thus, so long as the price is below the monopoly price, it pays that firm to raise its price by a little. But it pays each firm to do so. They thus all raise their prices—and the process continues until the monopoly price is reached.
In other cases it has been shown that markets create their own "noise," so that an equilibrium in which all firms charge the same price cannot exist. If all firms were charging the same price (and there was, accordingly, no need to search for the store with the lowest price), it would pay some firm to raise its price to exploit those who are particularly price insensitive because their search costs are high.
Many market institutions, practices, and structures can be viewed as the economy's responses to these informational problems. We have already noted three of these: the prevalence of the use of credit rather than equity as a source of finance for new investment; the widespread occurrence of credit rationing; and the fact that firms pay wages higher than strictly necessary in order to obtain workers, both to enable them to acquire a higher-quality labor force and to induce workers to work harder.
Three other market responses to costly information are particularly important. First, firms need to have reputations so that customers know they won't be cheated, and workers need reputations so firms know that they won't shirk (see Brand Names). This means that firms and workers must have an incentive to maintain their reputations. Usually, the most severe punishment that a customer can impose on a firm that has sold her a shoddy product is to stop dealing with the firm and to tell one's friends and associates. The most severe punishment a firm can impose on a worker who has shirked in performing duties is to fire the worker. But in traditional economic theory neither of these acts would make much difference: firms make zero profits at the margin, and workers are paid their opportunity cost (that is, the amount they could earn elsewhere). Therefore, there is no difference between the wage paid by the firm and what they could obtain elsewhere.
Thus, for reputation mechanisms to work, firms must at the margin receive some profits, and workers must receive wages in excess of their opportunity costs. The presence and importance of these higher-than-normal profits and wages, though long recognized, had not been previously explained.
A second response to imperfect information is advertising. Because information is costly, both suppliers and demanders must spend resources to acquire and disseminate information. Just as customers search for the lowest price and workers search for the highest wage, stores advertise to provide information to potential customers concerning the location, price, availability, and qualities of their products (see Advertising).
Middlemen are a third example of a market response to costly information. Much popular literature vilifies the role of the middleman. Press reports point out the huge difference between the prices received by farmers and the prices paid by customers, suggesting that evil middlemen are engaged in robbing farmers and consumers. But middlemen provide a vital function in ensuring that goods are delivered to where they are wanted. They are in the business of ensuring the efficient allocation of the economy's scarce resources. For the most part competition in this sector is keen. The fact that so much is paid for these services reflects their value in allocating resources efficiently. The fact that there are often high profits simply reflects that some individuals are able to perform those services much better than others.
The standard theorems that underlie the presumption that markets are efficient are no longer valid once we take into account the fact that information is costly and imperfect. To some, this has suggested a switch to the Austrian approach, most forcefully developed during the forties and later by Friedrich Hayek and his followers. They have not attempted to "defend" markets by the use of theorems. Instead, they see markets as institutions that have evolved to solve information problems. According to Hayek, neoclassical economics got itself into trouble by assuming perfect information to begin with. A much better approach, wrote Hayek, is to assume the world we have, one in which everyone has only a little information. The great virtue of free markets, he wrote, is that they allow each person to efficiently use his own information, and do not require that anyone have all the information. According to Hayek, government planning requires the impossible—that a small body of officials have all this information.
The new information economics substantiates Hayek's contention that central planning faces problems because it requires an impossible agglomeration of information. It agrees with Hayek that the virtue of markets is that they make use of the dispersed information held by different participants in the market. But information economics does not agree with Hayek's assertion that markets act efficiently.
The fact that markets with imperfect information do not work perfectly provides a rationale for potential government actions. The older theory said that no government, no matter how well organized, could do better than markets. If that was true, then we had little need to inquire into the nature of government. The modern theory says that government might improve upon matters, but to ascertain whether or not this is the case requires a closer examination of how governments actually behave, or might behave under various rules.
The modern study of political economy has uncovered many inefficiencies associated with government behavior, just as the modern study of firms has uncovered many inefficiencies associated with market behavior. An important line of research has focused on identifying how government differs intrinsically from other organizations in the economy (their powers and constraints, including the limitations on information that they face and their powers and incentives to acquire information) and, based on these distinctive features, on determining the appropriate economic roles of governments and markets.
Joseph E. Stiglitz is an economics professor at Columbia University. He was previously chief economist at the World Bank and chairman of President Clinton's Council of Economic Advisers. In 1979 he received the American Economic Association's John Bates Clark Award, given every two years to the economist under age forty who is judged to have made the most significant contribution to economics. He is a founding editor of the AEA's Journal of Economic Perspectives. He shared the Nobel Prize in economic science in 2001.
Hayek, Friedrich A. "The Use of Knowledge in Society." American Economic Review 35 (1945): 519-53.
Stiglitz, Joseph E. "The Causes and Consequences of the Dependence of Quality on Prices." Journal of Economic Literature 25 (March 1987): 1-48.
Stiglitz, Joseph E. "Information and Economic Analysis." In Current Economic Problems, edited by Parkin and Nobay. 1975.
Stiglitz, Joseph E. "Information and Economic Analysis: A Perspective." The Economic Journal, Vol. 95, Supplement: Conference Papers. (1985), pp. 21-41.
Stiglitz, Joseph E. "On the Economic Role of the State." In The Economic Role of the State, edited by A. Heertze. 1989.