The policy-ineffectiveness proposition (PIP) is a new classical theory proposed in 1975 by Thomas J. Sargent and Neil Wallace based upon the theory of rational expectations, which posits that monetary policy cannot systematically manage the levels of output and employment in the economy.
Prior to the work of Sargent and Wallace, macroeconomic models were largely based on the adaptive expectations assumption. Many economists found this unsatisfactory since it assumes that agents may repeatedly make systematic errors and can only revise their expectations in a backward-looking way. Under adaptive expectations, agents do not revise their expectations even if the government announces a policy that involves increasing money supply beyond its expected growth level. Revisions would only be made after the increase in the money supply has occurred, and even then agents would react only gradually. In each period that agents found their expectations of inflation to be wrong, a certain proportion of agents' forecasting error would be incorporated into their initial expectations. Therefore, equilibrium in the economy would only be converged upon and never reached. The government would be able to maintain employment above its natural level and easily manipulate the economy.
This behavior by agents is contrary to that which is assumed by much of economics. Economics has firm foundations in assumption of rationality, so the systematic errors made by agents in macroeconomic theory were considered unsatisfactory by Sargent and Wallace. More importantly, this behavior seemed inconsistent with the stagflation of the 1970s, when high inflation coincided with high unemployment, and attempts by policymakers to actively manage the economy in a Keynesian manner were largely counterproductive. When applying rational expectations within a macroeconomic framework, Sargent and Wallace produced the policy-ineffectiveness proposition, according to which the government could not successfully intervene in the economy if attempting to manipulate output. If the government employed monetary expansion in order to increase output, agents would foresee the effects, and wage and price expectations would be revised upwards accordingly. Real wages would remain constant and therefore so would output; no money illusion occurs. Only stochastic shocks to the economy can cause deviations in employment from its natural level.
Taken at face value, the theory appeared to be a major blow to a substantial proportion of macroeconomics, particularly Keynesian economics. However, criticisms of the theory were quick to follow its publication.
The Sargent and Wallace model has been criticised by a wide range of economists. Some, like Milton Friedman, have questioned the validity of the rational expectations assumption. Sanford Grossman and Joseph Stiglitz argued that even if agents had the cognitive ability to form rational expectations, they would be unable to profit from the resultant information since their actions would then reveal their information to others. Therefore, agents would not expend the effort or money required to become informed and government policy would remain effective.
The New Keynesian economists Stanley Fischer (1977) and Edmund Phelps and John B. Taylor (1977) assumed that workers sign nominal wage contracts that last for more than one period, making wages "sticky". With this assumption the model shows government policy is fully effective since, although workers rationally expect the outcome of a change in policy, they are unable to respond to it as they are locked into expectations formed when they signed their wage contract. Not only is it possible for government policy to be used effectively, but its use is also desirable. The government is able respond to stochastic shocks in the economy which agents are unable to react to, and so stabilise output and employment.
The Barro–Gordon model showed how the ability of government to manipulate output would lead to inflationary bias. The government would be able to cheat agents and force unemployment below its natural level but would not wish to do so. The role of government would therefore be limited to output stabilisation.
Since it was possible to incorporate the rational expectations hypothesis into macroeconomic models whilst avoiding the stark conclusions that Sargent and Wallace reached, the policy-ineffectiveness proposition has had less of a lasting impact on macroeconomic reality than first may have been expected. In fact, Sargent himself admitted that macroeconomic policy could have nontrivial effects, even under the rational expectations assumption, in the preface to the 1987 edition of his textbook Dynamic Macroeconomic Theory:
- 'The first edition appeared at a time when discussions of the 'policy-ineffectiveness proposition' occupied much of the attention of macroeconomists. As work of John B. Taylor has made clear, the methodological and computational implications of the hypothesis of rational expectations for the theory of optimal macroeconomic policy far transcend the question of whether we accept or reject particular models embodying particular neutrality propositions... The current edition contains many more examples of models in which a government faces a nontrivial policy choice than did the earlier edition.'
Despite the criticisms, Anatole Kaletsky has described Sargent and Wallace's proposition as a significant contributor to the displacement of Keynesianism from its role as the leading economic theory guiding the governments of advanced nations.
While the policy-ineffectiveness proposition has been debated, its validity can be defended on methodological grounds. To do so, one has to realize its conditional character. For new classicals, countercyclical stimulation of aggregate demand through monetary policy instruments is neither possible nor beneficial if the assumptions of the theory hold. If expectations are rational and if markets are characterized by completely flexible nominal quantities and if shocks are unforeseeable white noises, then macroeconomic systems can deviate from the equilibrium level only under contingencies (i.e. random shocks). However, no systematic countercyclical monetary policy can be built on these conditions, since even monetary policy makers cannot foresee these shocks hitting economies, so no planned response is possible. According to the common and traditional judgement, new classical macroeconomics brought the inefficiency of economic policy into the limelight. Moreover, these statements are always undermined by the fact that new classical assumptions are too far from life-world conditions to plausibly underlie the theorems. So, it has to be realized that the precise design of the assumptions underlying the policy-ineffectiveness proposition makes the most influential, though highly ignored and misunderstood, scientific development of new classical macroeconomics. New classicals did not assert simply that activist economic policy (in a narrow sense: monetary policy) is ineffective. Robert Lucas and his followers drew the attention to the conditions under which this inefficiency probably emerges.
- Barro, Robert J. (1977). "Unanticipated Money Growth and Unemployment in the United States". American Economic Review. 67 (2): 101–115. JSTOR 1807224.
- Barro, Robert J. (1978). "Unanticipated Money, Output, and the Price Level in the United States". Journal of Political Economy. 86 (4): 549–580. doi:10.1086/260699.
- Fischer, Stanley (1977). "Long-Term Contracts, Rational Expectations, and the Optimal Money Supply Rule". Journal of Political Economy. 85 (1): 191–205. doi:10.1086/260551.
- Glick, Reuven & Hutchison, Michael (1990). "New Results in Support of the Fiscal Policy Ineffectiveness Proposition". Journal of Money, Credit, and Banking. 22 (3): 288–304. doi:10.2307/1992561.
- Grossman, Sanford J.; Stiglitz, Joseph (1980). "On the Impossibility of Informationally Efficient Markets". American Economic Review. 70 (3): 393–408.
- Heijdra, Ben J. & van der Ploeg, F. (2002). Foundations of Modern Macroeconomics. Oxford: Oxford University Press. ISBN 0-19-877617-9.
- McCallum, Bennett T. (1979). "The Current State of the Policy-Ineffectiveness Debate". American Economic Review. 69 (2): 240–245.
- Phelps, Edmund S. & Taylor, John B. (1977). "Stabilizing Powers of Monetary Policy under Rational Expectations". Journal of Political Economy. 85 (1): 163–190. doi:10.1086/260550.
- Sargent, Thomas & Wallace, Neil (1975). "'Rational' Expectations, the Optimal Monetary Instrument, and the Optimal Money Supply Rule". Journal of Political Economy. 83 (2): 241–254. doi:10.1086/260321.
- Sargent, Thomas & Wallace, Neil (1976). "Rational Expectations and the Theory of Economic Policy". Journal of Monetary Economics. 2 (2): 169–183. doi:10.1016/0304-3932(76)90032-5.
- Sargent, Thomas (1987). Dynamic Macroeconomic Theory (2nd ed.). Academic Press. ISBN 0-12-619751-2.
After Keynesian Macroeconomics
The new classical macroeconomics is a school of economic thought that originated in the early 1970s in the work of economists centered at the Universities of Chicago and Minnesota—particularly, Robert Lucas (recipient of the Nobel Prize in 1995), Thomas Sargent, Neil Wallace, and Edward Prescott (corecipient of the Nobel Prize in 2004). The name draws on John Maynard Keynes’s evocative contrast between his own macroeconomics and that of his intellectual forebears. Keynes had knowingly stretched a point by lumping his contemporaries, a. c. pigou and Alfred Marshall, in with the older classical political economists, such as David Ricardo, and calling them all “classical.”
According to Keynes, the classics saw the price system in a free economy as efficiently guiding the mutual adjustment of supply and demand in all markets, including the labor market. Unemployment could arise only because of a market imperfection—the intervention of the government or the action of labor unions—and could be eliminated through removing the imperfection. In contrast, Keynes shifted the focus of his analysis away from individual markets to the whole economy. He argued that even without market imperfections, aggregate demand (equal, in a closed economy, to consumption plus investment plus government expenditure) might fall short of the aggregate productive capacity of its labor and capital (plant, equipment, raw material, and infrastructure). In such a situation, unemployment is largely involuntary—that is, workers may be unemployed even though they are willing to work at a wage lower than the wage the firms pay their current workers.
Later Keynesian economists achieved a measure of reconciliation with the classics. paul samuelson argued for a “neoclassical synthesis” in which classical economics was viewed as governing resource allocation when the economy was kept, through judicious government policy, at full employment. Other Keynesian economists sought to explain consumption, investment, the demand for money, and other key elements of the aggregate Keynesian model in a manner consistent with the assumption that individuals behave optimally. This was the program of “microfoundations for macroeconomics.”
Origins of the New Classical Macroeconomics
Although its name suggests a rejection of Keynesian economics and a revival of classical economics, the new classical macroeconomics began with Lucas’s and Leonard Rapping’s attempt to provide microfoundations for the Keynesian labor market. Lucas and Rapping applied the rule that equilibrium in a market occurs when quantity supplied equals quantity demanded. This turned out to be a radical step. Because involuntary unemployment is exactly the situation in which the amount of labor supplied exceeds the amount demanded, their analysis leaves no room at all for involuntary unemployment.
Keynes’s view was that recessions occur when aggregate demand falls—largely as the result of a fall in private investment—causing firms to produce below their capacity. Producing less, firms need fewer workers, and thus employment falls. Firms, for reasons that Keynesian economists continue to debate, fail to cut wages to as low a level as job seekers will accept, and so involuntary unemployment rises. The new classicals reject this step as irrational. Involuntary unemployment would present firms with an opportunity to raise profits by paying workers a lower wage. If firms failed to take the opportunity, then they would not be optimizing. Employed workers should not be able to resist such wage cuts effectively since the unemployed stand ready to take their places at the lower wage. Keynesian economics would appear, then, to rest either on market imperfections or on irrationality, both of which Keynes denied.
These criticisms of Keynesian economics illustrate the two fundamental tenets of the new classical macroeconomics. First, individuals are viewed as optimizers: given the prices, including wage rates, they face and the assets they hold, including their education and training (or “human capital”), they choose the best options available. Firms maximize profits; people maximize utility. Second, to a first approximation, prices adjust, changing the incentives to individuals, and thereby their choices, to align quantities supplied and demanded.
business cycles pose a special challenge for new classical economists: How are large fluctuations in output compatible with the two fundamental tenets of their doctrine? Here is how. The economy, they believe, is often buffeted by unexpected shocks. Shocks to aggregate demand are typically unanticipated changes in monetary or fiscal policy. Shocks to aggregate supply are typically changes in productivity that may result, for example, from transient changes to technology, prices of raw materials, or the organization of production. Ideally, firms would choose to produce more and to pay their workers more when the economy has been hit by favorable shocks and less when hit by unfavorable shocks. Similarly, workers would be willing to work more when productivity and wage rates are higher and to take more leisure when their rewards are lower. For both, the rule is “make hay while the sun shines.”
Employment, like output, would clearly rise with favorable shocks and fall with unfavorable shocks. But having rejected the very notion of involuntary unemployment, why do new classicals think that the unemployment rate would fall in the boom and rise in the slump? When a worker is laid off, he must seek a new job. He weighs the value of taking a lower-paid job that might be easily available (a machinist might become a day laborer) against the value of a better-paid, more suitable job that is harder to find. The new classicals do not argue that the unemployed job searcher is happy with his choice: being laid off was a bad draw, and, like everyone, he prefers good luck to bad. Rather, they argue that the worker chooses what he regards as the best available option, even when the options are poor. To remain unemployed (and to show up in the unemployment statistics) is something that he chooses based on his judgment that the benefits of the search outweigh the costs; this is not an exception to the rule that amount supplied equals amount demanded.
The fact that the economy experiences good and bad shocks is not enough to explain business cycles. An adequate theory must account for persistence—the fact that business cycles typically display long runs of good times followed by shorter, but still significant, runs of bad times. Those new classicals who regard demand shocks as dominant argue that the shocks are propagated slowly. It is always costly to adjust production levels quickly. Similarly, when higher production requires new capital, it takes time to build it up. And when lower production renders existing capital redundant, it takes time to wear it out or use it up. New classicals of the “real-business-cycle school” (led by edward prescott and finn kydland, corecipients of the 2004 Nobel Prize) regard changes in productivity as the driving force in business cycles. Because changes in technology may also come in waves, runs of favorable or unfavorable productivity (or technology) shocks may account for some of the persistence characteristic of business cycles.
Rational Expectations and Policy Ineffectiveness
Most economic decisions are forward looking. To know whether today is a day for work or for leisure, we need to decide whether tomorrow will be more or less productive than today; in short, we must have an expectation of the future. How should economists analyze expectations? The new classicals adopted John Muth’s “rational-expectations hypothesis” (see rational expectations). Muth argued that an economic model in which people’s expectations differ from the outcomes predicted by the model itself is poorly formulated. If the predictions of the model were correct—and therefore people’s expectations were wrong—then they could use the model to correct their own expectations. To fail to do so would result in economic losses and would be irrational. At one level, Muth’s hypothesis is just a technical consistency criterion for models. At another level, it appeals to the economic insight that people will not persist in easily correctable, systematic, and costly errors. The new classicals appeal implicitly (and sometimes explicitly) to Lincoln’s well-known adage: “You can fool some of the people all of the time, and all of the people some of the time, but you cannot fool all of the people all of the time.” They warn policymakers that a policy that depends on the assumption that the public systematically misunderstands its own interest is likely to fail.
Keynesian economists of the 1960s often appealed to the phillips curve, taking it to imply that monetary or fiscal policy that lowered the unemployment rate also caused a higher inflation rate. The interesting policy question was the trade-off: How much extra inflation was a one-point fall in the unemployment rate worth? The new classicals rejected the idea that there was any useful trade-off. They argued that an expansion of aggregate demand lowered unemployment only because the acceleration in prices was not anticipated. Firms that mistook higher market prices for higher real returns would be willing to produce more. Workers who mistook higher market wages for higher purchasing power would be willing, if unemployed, to take a job sooner. Increased output and lower unemployment would, however, be temporary because neither the returns to firms nor the purchasing power of workers was, corrected for inflation, really higher. As soon as they realized the mistake, firms and workers would return to old levels of production and labor supply.
What is more, having made the mistake once, they would not be easily fooled again by the same policy. The combination of rational expectations and the central tenet of new classical analysis that quantity supplied equals quantity demanded ensures that systematic, pure aggregate-demand policies do not have real effects on the economy. The Phillips curve trade-off can be observed in the data because some part of policy is always unanticipated. But policymakers cannot exploit it because the public will see through any systematic policy. Because it rejected the prevailing Keynesian view that monetary policy could offset a recession, this “policy-ineffectiveness proposition” became the most startling and controversial conclusion of the early new classical macroeconomics.
The policy-ineffectiveness proposition is frequently misunderstood. It is not a claim that no government policy affects the economy. Policies on government spending, for example, represent changes in the real claims the government makes on GDP and may affect output and employment. Rather, the proposition is limited to the effects of changes in government liabilities (the monetary base and the government debt) that may affect the rate of inflation. In this respect, the policy-ineffectiveness proposition is related to another new classical proposition: Ricardian equivalence (see government debt and deficits). Ricardian equivalence is the claim that whether a given path of government expenditure is financed through taxes or debt is unimportant: substituting debt for taxes appears to increase disposable income today. But since the debt must be repaid with interest, a rational taxpayer would save the entire windfall in order to afford the future tax bill, leaving his expenditure unchanged. Ricardian equivalence remains controversial because it depends on assumptions about the public’s foresight and grasp of the fiscal system closely related to the rational-expectations hypothesis and on debatable assumptions about the incidence of taxes and expenditure.
The Lucas Critique
Unanticipated policy has real effects, but, because it is unanticipated, it cannot be systematic—and therefore it cannot be used to direct the economy. The systematic element of policy can be viewed, implicitly at least, as a policy rule. Consider the Phillips curve again. How much does unemployment fall for a one-percentage-point increase in the price level? Lucas argued that the answer depends on the policy rule. If the rule had been one that held the inflation rate to zero (prices are constant), then the increase would be unanticipated and unemployment would fall. If the rule had been one that maintained a steady 1 percent rate of inflation (each year prices grow by 1 percent), then the increase would be just what systematic policy implied, the inflation would be perfectly anticipated, and unemployment would not change. And if the rule had been one that maintained a steady 2 percent rate of inflation, then a one-percentage-point increase would fall short of what had been anticipated, and unemployment would rise. These different trade-offs suggest that there is a different Phillips curve for each policy rule. A Phillips curve estimated under one policy regime would not predict accurately what would happen under a different regime.
Lucas argued that what is true of the Phillips curve in this example is true of the most important relationships in the econometric macroeconomic models used to evaluate economic policy. His analysis has come to be known as the “policy non-invariance” or “Lucas” critique.
The Legacy of the New Classical Macroeconomics
The new classicals profoundly changed the technical underpinnings of modern macroeconomics. Economists now widely accept the Lucas critique. To avoid it, economists sought ways to predict exactly how estimated relationships would change with the policy regime by integrating rational expectations and the public’s response to policy rules into their models. Because rational expectations depend on the structure of the whole economy, the program of microfoundations is no longer content to look at different markets separately, but concentrates on general equilibrium among them. Dynamic models have replaced static models: policy actions cannot be evaluated merely for what they do today, but for how they change people’s judgments about the future.
While few economists want to assume that the government can fool the public systematically, many remain skeptical of the rational-expectations hypothesis as a description of people’s actual expectations. Some, including founding new classicals such as Sargent, have explored models of learning that emphasize that overcoming expectational errors is a process that may take some time.
Most economists, even among the new classicals, no longer accept the policy-ineffectiveness proposition. It is widely agreed that wages and prices do not move quickly and smoothly to the values needed for long-run equilibrium between quantities supplied and demanded. Consequently, even if monetary policy is ineffective in the long run, it may be of considerable use in the short run.
While optimal search and voluntary changes in labor supply may explain a large portion of routine unemployment rates, many economists question whether they explain high unemployment in recessions. Was the 25 percent unemployment rate of the Great Depression the result of a mass decision to take a vacation? New Keynesian critics typically maintain that unemployment is, in fact, characterized by wages above the level needed to clear the labor market. But what, exactly, prevents firms from taking profitable advantage of the situation remains controversial.
About the Author
Kevin D. Hoover is professor of the departments of economics and philosophy at Duke University. He is past president of the History of Economics Society, past chairman of the International Network for Economic Method, and editor of the Journal of Economic Methodology.
Hartley, James, Kevin D. Hoover, and Kevin D. Salyer, eds. Real Business Cycles: A Reader. London: Routledge, 1998. Important articles for and against real business cycle models, including Finn E. Kydland and Edward C. Prescott’s seminal “Time to Build and Aggregate Fluctuations,” and the editors’ introduction that presents an accessible, critical account of the models.
Hoover, Kevin D. The New Classical Macroeconomics: A Sceptical Inquiry. Oxford: Blackwell, 1988. General account of the new classical economics.
Lucas, Robert E., Jr. Studies in Business Cycle Theory. Oxford: Blackwell, 1981. Some of Lucas’s own articles are collected here, including “Econometric Testing of the Natural Rate Hypothesis,” a formal but relatively accessible presentation of key new classical ideas, and “Understanding Business Cycles,” which lays out the basis for the new classical theory of the business cycle.
Lucas, Robert E., Jr., and Thomas J. Sargent, eds. Rational Expectations and Econometric Practice. London: Allen and Unwin, 1981. Many key articles covering the first decade of the new classical school, including Muth’s original article on rational expectations, Sargent and Wallace’s articles on the policy-ineffectiveness proposition, and Lucas and Sargent’s new classical manifesto, “After Keynesian Macroeconomics.”
Sheffrin, Steven M. Rational Expectations. 2d ed. Cambridge: Cambridge University Press, 1996. General account of the new classical economics.
Return to top