CH-18-19 Alternative Perspectives on Stabilization Policy – Macroeconomics : Chapter Notes

This Document Contains Chapters 18 to 19 CHAPTER 18 Alternative Perspectives on Stabilization Policy Notes to the Instructor Chapter Summary This chapter discusses issues in macroeconomic policy. Even though the AD–AS model suggests the desirability and simplicity of stabilization policy, in practice the benefits of stabilization policy are less clear and its complications are much more evident. The chapter addresses two key questions. First, should policymakers try to stabilize the economy or instead be content with doing no harm? And second, should policymakers commit to a rule or make policy decisions on a case-by-case basis? Accordingly, the chapter frames its discussion around debates over active versus passive policy, and over rules versus discretion. Comments The material in this chapter can be covered in two lectures. The taxonomy set up in this chapter makes discussion of policy issues very clear, and as a result students like this material. I emphasize that many economists believe that there is a case in principle for stabilizing the economy but that pragmatic difficulties may override this case. The material on the political business cycle is particularly striking to students and obviously could be given emphasis in election years. It might also be worth referring back to Ray Fair’s work discussed in Supplement 1-2. Use of the Web Site This is the natural time to discuss the presidential policy game in the Web-based software. Not surprisingly, students seem to like this game a lot. As well as being fun, it is intended to teach a number of lessons about the difficulty of stabilization and of how expansion can be beneficial in the short run but costly in the long run. Instructors might also choose (at some point) to explain to the students about the details of the model, as explained in the documentation. At the risk of turning fun into work, the game could be used as a basis for homework assignments—for example, asking students to write an essay on good and bad strategies. Instructors could also poll their classes to see who was most successful (in terms of score and/or length in office) and find out what strategy they used. It is also interesting to discuss what is left out of the game (open-economy considerations, forecasts of future shocks, elections, independent monetary and fiscal policymakers) and how inclusion of these features might change things. Use of the Dismal Scientist Web Site Go to the Dismal Scientist Web site and download data for the past three years on the rate of consumer price inflation for the Euro-zone, the United Kingdom, Canada, Australia, and New Zealand. These countries all have explicit inflation targets that their central banks use in setting monetary policy. Visit the central bank Web sites of these countries (or see Supplement 18-16) to learn the actual values (or ranges) of their inflation targets. Assess the extent to which recent inflation has been above or below the targets for these countries. In cases where the targets have not been achieved, discuss any unusual circumstances or policy reasons that may have led these outcomes. 425 Chapter Supplements This chapter includes the following supplements: 18-1 Menu Costs, Imperfect Competition, and the Welfare-Improving Effects of Policy 18-2 Profit Sharing as an Automatic Stabilizer 18-3 Leading Indicators in Action 18-4 The Pitfalls of Forecasting (Case Study) 18-5 Are Forecasters Rational? (Case Study) 18-6 Microfoundations and Aggregation 18-7 Spare a Thought for the Empirical Macroeconomist (Case Study) 18-8 The Response to Romer (Case Study) 18-9 Distrust of Policymakers 18-10 The Political Business Cycle 18-11 The Political Business Cycle at Its Worst 18-12 The Economy Under Democratic and Republican Presidents 18-13 Price Level Versus Inflation Targeting 18-14 Inflation Targeting (Case Study) 18-15 Central-Bank Independence and Growth (Case Study) 18-16 Measuring Central-Bank Independence (Case Study) 18-17 Additional Readings Lecture Notes Introduction Our interest in macroeconomic phenomena is motivated by a desire to understand not only the functioning of the economy but also how macroeconomic policies affect the economy, with a view to understanding, in turn, if there is a role for government policies directed at stabilization. Ultimately, we wish to answer two questions: Can the government improve the functioning of the economy by pursuing active stabilization policies? How should a well-designed economic policy be conducted? Both are considered in this chapter. To answer these questions, we first need answers to two prior questions: How would the economy perform in the absence of stabilization policies? How do government policies affect the overall functioning of the economy? Our analysis to this point provides answers. Given that flexible prices are important for the efficient allocation of resources in the economy, the Supplement 18-1, possibility of price stickiness in the short run suggests that recessions and booms may be a signal “Menu Costs, of inefficient misallocations of resources. To the extent that we believe that prices and wages Imperfect Competition, and may not be perfectly flexible in the short run, we have some reason to believe that a wellthe Economic designed macroeconomic policy could improve overall economic welfare. The aggregate Welfare-Improving demand–aggregate supply model, meanwhile, explains that government policies affect the level Effects of Policy” of output in the economy by altering aggregate demand. Given a short-run aggregate supply curve, shifts in aggregate demand affect output and employment. 18-1 Should Policy Be Active or Passive? The aggregate demand–aggregate supply model indicates that, in principle, policy-makers could manipulate aggregate demand in order to stabilize the economy. Despite such a conclusion, many economists do not advocate an active role for government policy. New classical economists believe that prices and wages are flexible and usually argue that stabilization policy is inappropriate. They believe that fluctuations arise either as an efficient response to shocks (the real-business-cycle view) or as a result of the information problems summarized in the imperfect-information theory of short-run aggregate supply. Since they believe in price and wage flexibility, their presumption is that markets serve to allocate goods efficiently. There will be room for policymakers to stabilize the economy only if they are better informed about shocks to the economy than are individuals in the economy. If individuals know as much about the shocks hitting the economy as policymakers do, then they can form as good a prediction of aggregate demand and hence of the price level. There is no way for policymakers to improve economic performance unless they have an information advantage. Lags in the Implementation and Effects of Policies If prices and wages are sticky, as Keynesians believe, then the theoretical case for intervention is much stronger. Nevertheless, some proponents of new Keynesian ideas are also skeptical about stabilization policy. In reality, the economy is not as easy to influence as our models suggest, and there are many practical difficulties associated with economic stabilization. Economists point in particular to lags in the policy process. First, it takes time to recognize and take actions to respond to shocks that hit the economy. As just noted, if policymakers cannot recognize shocks any more quickly than private individuals can, then they have no scope for stabilization. Even after a shock has been identified, it may take time to change the course of monetary and fiscal policies. This is known as the inside lag. It is relatively more important for fiscal policy than for monetary policy, since a change in monetary policy simply entails a decision by the Federal Reserve, while a change in fiscal policy requires agreement and action by the congressional and executive branches of government. Second, there is an outside lag, meaning the time it takes for policies to affect the economy. Fiscal policies affect spending and hence aggregate demand directly, and so do not have a significant outside lag. Monetary policy has a long outside lag, because monetary policies affect aggregate demand through the monetary transmission mechanism (whereby changes in the money supply affect interest rates and thus investment). The presence of inside and outside lags makes active use of fiscal and monetary policies imprecise. But these lags do not imply that policy should remain passive in the presence of a severe and lengthy economic contraction, such as the downturn that began in 2008. Some features of the economy help keep GDP close to the natural rate without direct Supplement 18-2, action by policymakers. When the economy is in a boom, tax revenues increase and transfer “Profit Sharing as payments, such as unemployment insurance and other welfare benefits, decrease. The opposite an Automatic occurs when the economy is in recession. Thus, taxes minus transfers (T) are procyclical: T Stabilizer” automatically increases in booms (decreasing disposable income) and decreases in recessions (increasing disposable income). These are known as automatic stabilizers. In terms of our model, they imply that a given shock to aggregate demand causes a smaller change in GDP than would be the case in their absence. The Difficult Job of Economic Forecasting A fundamental difficulty with good policymaking is that it depends upon the ability to forecast Supplement 18-3, future economic events. The large-scale macroeconometric models discussed in Chapter 12 are “The Labor of some use here, since they help predict the behavior of key economic variables, given Market in the assumptions about the future behavior of exogenous variables. But many events that have a 1990 Recession” significant effect on the economy, such as Iraq’s invasion of Kuwait in 1990 or the attacks of September 11, 2001, are inherently unpredictable. Supplement 18-4, The index of leading economic indicators, discussed in Chapter 10, provides some “A Different information on the future performance of the economy. This is a set of 10 data series that have Leading Indicator” historically been a good guide to future economic behavior. Case Study: Mistakes in Forecasting The difficulty of economic forecasting is well illustrated by the failure of forecasters to predict both the Great Depression and the severe recession of 2008-2009 with any accuracy. Almost all Figure 18-1 Supplement 18-5, forecasters predicted that the poor economic performance at the start of the Depression would be “The Pitfalls of short-lived and that recovery was imminent. And in the early stages of the economic downturn Forecasting” of 2008, forecasters were predicting a relatively mild slowdown, with unemployment projected Supplement 18-6, to increase only modestly. In fact, the unemployment rate rose sharply during 2008 and 2009. “Are Forecasters Forecasting is difficult even when we have a good economic model because of the uncertainty in Rational?” predicting the future behavior of exogenous variables. Ignorance, Expectations, and the Lucas Critique The economist Robert Lucas has criticized the use of traditional economic and econometric models for the evaluation of economic policy. Lucas’s point is that individuals’ behavior and actions—and particularly their expectations—in general depend upon the policies chosen by government. Therefore, Lucas argues, it is misleading to use standard models for policy evaluation unless we take account of the effects of policies on expectations. For example, if the economy has been experiencing moderate and fairly steady inflation, then people’s inflation expectations might be well approximated by adaptive expectations. But if the Fed then announced that it was going to start increasing the money supply by 100 percent per year, then individuals would certainly revise their inflation expectations in light of this announcement, and adaptive expectations would be a very unreliable guide to people’s beliefs. Supplement 18-7, Another implication of the Lucas critique is that we need to be careful to recognize that “Microfoundations some of the basic equations of our model might also change when policy changes. For this and Aggregation”” reason, many macroeconomists believe that it is important to pay attention to the microfoundations—that is, the microeconomic underpinnings—of our macroeconomic model. The textbook emphasizes such microfoundations at various times. For example, Chapter 16 shows that the effects of a tax cut on consumption are different depending upon whether individuals expect the tax cut to be temporary or permanent. The Historical Record One way to help decide if policy should be active or passive is to try to determine whether stabilization policies were successfully pursued in the past. Macroeconomists often look to historical data to answer this question. Casual examination of the data supports the idea that stabilization policies worked. In the period after World War II, policymakers used the insights of Keynesian economics as a basis for active demand management. The data on GDP and other variables indicate that this period was characterized by less severe fluctuations than the prewar period. This evidence is not conclusive, however. It may have been that the economy was simply not hit by such severe shocks. Alternatively, the changing composition of the economy may have lessened the importance of shocks. For example, suppose that the most important shocks to the prewar economy were weather shocks affecting agricultural output. Then as the agricultural sector became relatively less important in the U.S. economy, such shocks would be less relevant for aggregate fluctuations. Our previous discussion of the Great Depression illustrates why such analysis is often inconclusive. The spending hypothesis argues that the Depression was caused by a fall in consumption and investment demand, which could have been countered by expansionary monetary and fiscal policies. The money hypothesis argues that the Depression was caused by badly managed economic policymaking— in particular, a severe monetary contraction—and suggests that a passive policy of steady monetary expansion could have limited its severity. Case Study: Is the Stabilization of the Economy a Figment of the Supplement 18-8, “Spare a Thought Data? for the Empirical Macroeconomist” The economist and economic historian Christina Romer has suggested that it may not even be true that the economy has been more stable since World War II. She noted that the economy may Supplement 18-9, simply appear less volatile in recent years as a consequence of improved data collection. By “The Response to attempting to acquire better old data and by subjecting more modern data to the methods used in Romer”” constructing older data series, she demonstrated that at least part of the volatility of early data is an artifact of the methods used to construct those data. Some of her work even suggests that there is no evidence that the postwar economy is more stable than previously. Other economic historians disputed this claim. The current consensus appears to be that Romer is right to attribute part of the difference in volatility to data flaws, but that this does not explain all the difference. Case Study: How Does Policy Uncertainty Affect the Economy? Recent research investigates the effect of policy uncertainty on economic performance. The researchers developed an index of policy uncertainty that has three components. One component considers the number of articles in major newspapers during a given month which contain the words “uncertainty” or “uncertain”, “economic” or “economy, and at least one of the following: “congress,” “legislation,” “white house,” “regulation,” “federal reserve,” or “deficit.” A second component considers the number of temporary provisions in the tax code. These provisions are often extended at the last minute and so create uncertainty in financial planning for businesses and households. A third component considers the amount of disagreement among private forecasters concerning the future price level and the future amount of government spending. More disagreement among forecasters is taken as evidence of more uncertainty. The researchers then studied whether the index was associated with the level of economic activity. They found that higher uncertainty about economic policy tends to depress the economy. When policy uncertainty rises, investment, production, and employment are more likely to decline relative to their usual growth over the next year. A possible reason for this effect may be that uncertainty reduces aggregate demand for goods and services. An increase in policy uncertainty may lead firms and households to put off large purchases until the uncertainty has been reduced or eliminated. While some amount of policy uncertainty is unavoidable, some is under policymakers’ control. Reducing the policy uncertainty policymakers control may have payoffs in improved macroeconomic performance. 18-2 Should Policy Be Conducted by Rule or by Discretion? Even if economists agreed upon the desirability of active stabilization policies, they would probably still disagree about exactly how policy should be conducted. One possibility is discretionary policy, whereby policymakers respond to changing economic circumstances. Another possibility is a policy rule, whereby policymakers commit themselves in advance to follow a particular policy. At first sight, discretionary policy might seem the better option, for how could limitations on the conduct of policy improve economic outcomes? In fact, there are a number of reasons for favoring policy rules. Distrust of Policymakers and the Political Process In practice, the policies enacted by governments need not always be those that are best for the economy. The idea of a well-informed, benevolent policymaker choosing monetary and fiscal policy variables to stabilize the economy is a useful fiction for our analysis, but it hardly describes the realities and complexities of the political process. Policymakers may not be well informed, since good macroeconomic analysis is not easy. And policymakers may not be benevolent, but may instead be interested in manipulating the economy for their own political ends. The idea that governments use macroeconomic policy to help their reelection chances is Supplement 18-10, “Distrust of known as the political business cycle. Under these circumstances, simple fixed rules may do the Policymakers” least harm. Supplement 18-11, “The Political The Time Inconsistency of Discretionary Policy Business Cycle” Supplement 18-12, Time inconsistency provides a less pragmatic but perhaps more intriguing argument for favoring “The Political rules over discretion. The insight is that policymakers may want to change their minds about the Business Cycle at appropriate policies to pursue, giving them an incentive to renege on previous policy Its Worst” announcements. Individuals will understand this incentive and so will not believe the original announcements. Policymakers’ incentive to deviate from announced policies, in other words, causes those announcements to not be credible. A fixed-policy rule may be more credible, and so ultimately better. We discussed the importance of credibility when we considered the potential for painless disinflation in Chapter 14. Suppose that the economy is experiencing high inflation and Supplement 18-13, ”The Economy unemployment. The Federal Reserve, in order to reduce inflationary expectations, might under Democratic announce a tight money policy. If this announcement is believed, then the Phillips curve teaches and Republican us that inflation will fall. But the Fed then has an incentive to cheat—that is, to engineer an Presidents” unexpected increase in the money supply in order to bring unemployment down. If, as a consequence, individuals fail to believe the Fed’s announcement, then inflationary expectations and hence inflation will not fall after all. Case Study: Alexander Hamilton Versus Time Inconsistency In the process of achieving independence, the United States incurred a substantial public debt. Chapter 18 The Secretary of the Treasury, Alexander Hamilton, strongly opposed suggestions that the Appendix government default on its debts in order to avoid levying taxes to pay them off. Such a policy was tempting at the time, but it would have certainly meant that the government would have found it very difficult to borrow in the future. The decision to honor the government debt could be thought of as an investment in credibility. Rules for Monetary Policy Various rules for monetary policy have been proposed. Perhaps the simplest and most famous is Milton Friedman’s monetarist view that the growth rate of the money supply should be kept constant. The argument for such a rule is that changes in the money supply may be a cause of fluctuations (as, for example, in the money-hypothesis explanation of the Great Depression). If the velocity of money is constant, then stable money growth will avoid output fluctuations. If Supplement 18-14, velocity is unstable, though, such a policy will not lead to stable output. Economists who believe “Price Level Versus Inflation Targeting” that the velocity of money is prone to variation often advocate nominal GDP targeting instead. In this case, the Fed manipulates the money supply to try to keep nominal GDP on some announced path. A third possible policy rule is an inflation target. In this case, the Fed adjusts monetary policy in an attempt to maintain some announced target for inflation. All these policies are expressed in terms of nominal variables. While real variables ultimately are a better measure of economic welfare, there is a danger in directing monetary policy toward real variables. Suppose that the Fed is misinformed and believes the natural rate of unemployment to be lower than it actually is. Suppose also that the Fed uses monetary policy in an attempt to keep unemployment at this incorrect level. If unemployment is at the true natural rate, the Fed will increase the money supply. The Phillips curve tells us that this would lead to lower unemployment and higher (demand-pull) inflation in the short run. Over time, expected inflation rises and the economy returns to the natural rate. This, in turn, induces the Fed to inflate further. The consequence will be ever-increasing inflation. Targeting nominal variables avoids such problems. Case Study: Inflation Targeting: Rule or Constrained Discretion? Inflation targeting, which started to become a popular policy objective for central banks in the Supplement 18-15, “Inflation Targeting” late 1980s, is not a strict commitment to a policy rule. Inflation targets are usually a range (e.g., 1 to 3 percent), leaving central banks with a fair amount of discretion. In addition, in the face of temporary shocks, inflation is allowed to deviate from the target range without invoking action by the central bank. Even with such flexibility, inflation targeting increases the transparency of monetary policy and makes it easier for the public to judge the central bank’s performance. The Federal Reserve was a latecomer in adopting a policy of inflation targeting. Not until 2012 did the Fed set an inflation target of 2 percent. In explaining its reasoning for adopting a target, the Fed argued that a rate of 2 percent was consistent with its mandate for price stability and maximum employment. The Fed noted that a higher rate risked hindering the public’s ability to make accurate economic and financial decisions, and that a lower rate increased the chance of deflation. Figure 18-2 Case Study: Central-Bank Independence Supplement 18-16, Countries vary in the degree to which their central banks are independent from the government. “”Central Bank For example, in some countries the central bank is under the jurisdiction of the finance or Independence and treasury department while in other countries the central bank is separate from the government. Growth” Researchers have studied the relationship between the degree of independence of a Supplement 18-17, “Measuring Central country’s central bank and the performance of the macroeconomy. A key finding is that Bank countries with more-independent central banks have lower average inflation rates. Furthermore, Independence” this lower inflation rate does not come at the price of lower or more volatile growth. These findings have led some countries to pass legislation increasing the degree of independence of their central banks. 18-3 Conclusion: Making Policy in an Uncertain World The best and most careful economists are aware of the limitations of our understanding of the macroeconomy and are correspondingly cautious about giving advice. But an imperfect understanding is better than no understanding at all, and so economists cannot abdicate responsibility for the design of macroeconomic policy. If policy is not formulated on the basis of our best knowledge of the economy, then it will be designed instead on the basis of a lesser understanding. Appendix: Time Inconsistency and the Tradeoff Between Inflation and Unemployment This appendix contains an analytic treatment of the time-inconsistency problem. Suppose that the Fed’s preferences over inflation and unemployment can be represented by the loss function L(u, π) = u + γπ2. The Fed likes inflation to be as close to zero as possible and also likes unemployment to be as low as possible, so the Fed wants to minimize this function. For simplicity, we suppose that the Fed can simply choose the inflation rate. The behavior of the economy in the short run is described by a Phillips curve, u = un – α(π – Eπ). Given the public’s expectations of inflation (Eπ), the Fed chooses π so as to minimize the loss function. We can be show (by substituting for u in the loss function and using calculus to minimize the loss) that the Fed will always choose π = α/(2γ). But if the public has rational expectations, it will anticipate the Fed’s actions, so Eπ = π and u = un. The loss function thus takes on a higher value than it would if the Fed could credibly commit to setting π = 0 at all times. Figure 1 illustrates the problem in a diagram with inflation and unemployment on the axes. The Fed’s loss function can be illustrated using indifference curves familiar from microeconomics: all points on a given indifference curve are equivalent as far as the Fed’s preferences are concerned. Indifference curves further to the left are preferred. Given an expectation of inflation, we can draw the corresponding Phillips curve (passing through the point {un, Eπ}). The Phillips curve represents the constraint faced by the Fed, taking expected inflation as given, and so the Fed will choose the point on the Phillips curve where it is tangent to the loss function. For the simple loss function in the example, this point of tangency is always at π = α/(2γ), regardless of the value of expected inflation. To see the time-inconsistency problem, suppose that the Fed announced that it was going to set π = 0. If the public believed this announcement and if the Fed carried out its promise, then the economy would be at the optimal point, labeled O. This point is optimal in the sense that it is the best indifference curve that can be attained subject to the economy’s being at the natural rate. If the public expects zero inflation, however, the Fed faces a short-run constraint given by the Phillips curve passing through point O. It then has an incentive to renege on its promise and generate an unexpected inflation to put the economy on a better indifference curve, at point I (for inconsistent). Unfortunately, people recognize that the Fed is tempted to behave in this way, so its promise of zero inflation may not be credible. As a result, the economy ends up at point C, which is time consistent, but puts the economy on a worse indifference curve than at point O. ADVANCED TOPIC 18-1 Menu Costs, Imperfect Competition, and the Welfare-Improving Effects of Policy The presence of menu costs in the situation where firms set prices in imperfectly competitive markets provides scope for monetary policy to have large effects on economic welfare. This framework thus gives some support to an active role for policy in stabilizing the economy. The essential insight of menu costs is disarmingly simple—so much so that it is quite surprising that it was noted only very recently. Consider a price-setting firm that wants to maximize its profits. We can graph its profits against the price it charges as shown in Figure 1. The price that maximizes the firm’s profits is pm. But, as shown in Figure 1, there may be a wide range of prices, between p' and p'', that the firm could charge and earn almost as much profits as it would at pm. The point is sometimes summarized by the phrase: “A hill is flat at the top.” The firm, therefore, does not care very much whether it charges pm or a lower or higher price. Consumers, however, might care a great deal. More precisely, microeconomics teaches us that imperfect competition imposes social costs, and these can be significantly different depending on whether the firm charges a low or a high price. With the aid of a little bit of microeconomics, it is easy to show that small menu costs can have a large effect on welfare in the presence of imperfect competition. Consider the case of a monopolist who faces a downward-sloping demand curve and (for simplicity) constant marginal cost of production equal to k. The standard microeconomic diagram is shown in Figure 2. The firm sets its price (pm) where marginal revenue (MR) equals marginal cost (MC) and produces the quantity q. The area under the demand curve and above the marginal cost curve can be divided into three areas. The firm earns profits (∏) equal to pm – k on every unit. (These profits are sometimes referred to as producer surplus.) The area above the price and below the demand curve (labeled CS) is consumer surplus—it represents the difference between the amount consumers would be willing to pay and the amount they actually pay and so is a measure of the gain of consumers. 2 The final area is the deadweight loss (DW), which measures the cost to society of the firm’s monopoly behavior. Perhaps the easiest way to understand this loss is to note that, under perfect competition, the firm would set its price equal to marginal cost. The firm is then worse off by the area labeled ∏, but consumer surplus increases by II + DW, so the social loss in moving from perfect competition to monopoly equals DW. Now suppose that the firm must set its price before it knows for certain the demand curve that it will face. Specifically, we might think that the firm must set its dollar price before it knows what the aggregate price level will be. It can change its price when it learns the true state of demand, but at a cost. It wishes to set a real price equal to pm and so sets a nominal price (Pm) on the basis of its expectation of the price level (EP): Pm = pmEP. Its actual real price turns out to be Pm/P = pm(EP/P). Thus, if the actual price level is lower than the firm expected, then (EP/P) > 1, and the firm ends up with a higher real price; if the actual price level is higher than expected, the firm’s real price is lower. Let us therefore suppose that, perhaps as a result of an unexpectedly large increase in the money supply, the firm ends up setting a lower real price ( p') than it intended. It can change its price if it wishes to incur the menu cost. But for the reasons explained earlier, the gain in profits from moving to the profitmaximizing price from a price that is close might be small. If this gain in profits is less than the menu cost, the firm will not bother to change its price. This is shown in Figure 3. The firm’s loss in profits from charging the price p' rather than the optimal price pm is equal to (A – C) (in other words, if it changed its real price, it would lose an amount equal to the area C and gain an amount equal to the area A). As Figure 3 shows, this may be a small difference (this is because a hill is flat at the top). If A – C is smaller than the menu cost associated with changing prices, the firm will elect to stick with the price p'. Consumers gain substantially, however, by the firm’s charging p' rather than pm: consumer surplus increases by an amount equal to A + B. The total gain to society from the lower price is equal to the reduction in the deadweight loss (i.e., B + C). Thus, we find that an unanticipated increase in demand, the result perhaps of an unanticipated increase in the money supply, increases output and social welfare. The reason is that, because of menu costs, firms have sticky prices and so change quantities rather than prices in response to the demand shock. Because of imperfect competition, this increase in output is welfare improving from the perspective of society as a whole. Exactly the opposite conclusions hold for an unanticipated decrease in demand. ADDITIONAL CASE STUDY 18-2 Profit Sharing as an Automatic Stabilizer Economists often propose policies to improve the automatic-stabilizing powers of the economy. The economist Martin Weitzman has made one of the most intriguing suggestions: profit sharing. Today, most labor contracts specify a fixed wage. For example, General Motors might pay assembly-line workers $20 an hour. Weitzman recommends that the workers’ total pay should depend on their firm’s profits. A profitsharing contract for General Motors might pay workers $10 for each hour of work, but in addition the workers would divide among themselves a share of the firm’s profit. Weitzman argues that profit sharing would act as an automatic stabilizer. Under the current wage system, a fall in demand for a firm’s product causes the firm to lay off workers: it is no longer profitable to employ them at the old wage. The firm will rehire these workers only if the wage falls or if demand recovers. Under a profit-sharing system, Weitzman argues, firms would be more likely to maintain employment after a fall in demand. Under our hypothetical profit-sharing contract with General Motors, for example, an additional hour of work would cost the firm only $10; the rest of the compensation for additional workers would come from the workers’ share of profits. Because the marginal cost of labor would be so much lower under profit sharing, a fall in demand would not normally cause a firm to lay off workers. To provide evidence for the advantages of profit sharing, Weitzman points to Japan. Most Japanese workers receive a large fraction of their compensation in the form of year-end bonuses. Weitzman argues that, because of these bonuses, Japanese workers “think of themselves more as permanently employed partners than as hired hands.” And, as Weitzman’s theory predicts, employment in Japan is much more stable than in countries without any form of profit sharing. At least three arguments can be made for profit sharing: improved incentives, increased wage flexibility, and Martin Weitzman’s “excess-demand-for-labor” argument. Profit sharing may be desirable, first, because it improves workers’ motivation and thus increases their efficiency. Any individual worker’s increase in effort is likely to have a small effect on overall profits and hence a small effect on that worker’s pay. The presence of such free-rider effects leads economists to suggest that pay should ideally be linked to measures of individual effort rather than to overall profitability. Psychologists, however, look more favorably on group incentives than do economists: where economists see free-rider effects, psychologists see the potential for greater cooperation and improved motivation. A second argument for profit sharing is that it may make wages more flexible. Under profit sharing, wages will fall with profits in recessions and rise with profits in booms. Thus firms might be more willing to retain workers in recession, rather than laying them off. If wage rigidity is a cause of cyclical unemployment, profit sharing might therefore be desirable. Many economists are uncomfortable with this argument, however, because it simply takes wage rigidity as exogenously given. Without an explanation of why wages are rigid, we cannot be confident that the introduction of profit sharing would really make wages more flexible. For example, suppose that wages are rigid because of the power of insider workers. Such workers would be likely to oppose the introduction of profit sharing. Weitzman’s argument for profit sharing is more subtle than either of these and is often misunderstood. He contends that under profit sharing, firms are less likely to lay workers off in recessions because firms possess excess demand for labor. Under profit sharing, increases in employment reduce profits per worker and so reduce compensation. Weitzman thus argues that the marginal product of labor will exceed its marginal cost and firms will always hire all available workers. Contracts with profit sharing, according to Weitzman, will lead to less employment fluctuation than contracts that specify fixed wages, and so policies should be enacted to encourage profit sharing. This argument is open to the same objection noted previously: without knowing why firms and workers write fixed-wage contracts, we cannot be sure that profit sharing will really improve things.4 In fact, if wages are rigid for the sort of reasons explained in Chapter 6 of the textbook (such as efficiency wages), profit sharing could make matters worse. The reason is that profit sharing causes wages to fall in the face of adverse shocks. Firms might then actually lay more workers off in bad times in order to maintain the desired rigid wage. Moreover, profit sharing exposes workers to increased income risk and, as noted earlier, gives employed workers an incentive to try to influence hiring decisions. The New York Times dubbed Weitzman’s proposal “the best idea since Keynes.” Advocates of his theory want the government to provide tax incentives to encourage firms to adopt profit-sharing plans. Others, however, have expressed skepticism. They wonder why, if profit sharing is such a good idea, firms and workers don’t sign such contracts without prodding from the government. For these and other reasons, many economists doubt that widespread profit sharing would radically improve macroeconomic performance. It is probably fair to say that, whereas Weitzman’s proposal generated a lot of attention when it first emerged, interest and enthusiasm for the idea have since waned considerably. There may be good arguments for encouraging profit sharing, but they are not overwhelming. ADDITIONAL CASE STUDY 18-3 Leading Indicators in Action Figure 1 shows the percentage change in real GDP and the percentage change in the index of leading economic indicators for the period 1960–2014. Source: Department of Commerce, Bureau of Economic Analysis, and The Conference Board. The relationship between changes in the index of leading economic indicators and changes in real GDP is far from exact, but the index does frequently appear to “lead” movements in GDP. CASE STUDY EXTENSION 18-4 The Pitfalls of Forecasting The case study in Chapter 18 documents some of the difficulties of economic forecasting. Making economic forecasts is a good way to look foolish. For example, the following statements are from the eminent economist Irving Fisher. “Stock prices have reached what looks like a permanently high plateau.” (Fall, 1929) “There may be a recession in stock prices, but nothing in the nature of a crash.” (Sept. 1929) “I expect to see the stock market a good deal higher than it is today within a few months.” (Oct. 15, 1929) The stock market crashed on October 29, 1929. As John Kenneth Galbraith observes, “Irving Fisher was the most original of American economists. Happily there are better things . . . for which he is remembered.” On Monday, October 19, 1987 (the day the stock market fell over 500 points), meanwhile, The Wall Street Journal quoted one stock market expert on the previous week’s decline in the market: “It was a huge correction, but it was needed very very badly. Now I expect the market will be flat for a while.” CASE STUDY EXTENSION 18-5 Are Forecasters Rational? Although forecasters often make mistakes, they may still be doing the best that they can. The theory of rational expectations suggests a way of evaluating economists’ and others’ forecasts. Rational expectations suggest that forecasters should base their predictions on all available information. It should not then be possible systematically to improve upon a forecast using information that is publicly available at the time the forecast is made. If it were, then by definition, this could not be the best forecast. One common way to evaluate forecasts is thus to look at forecast errors (that is, the after-the-fact difference between the actual values of a variable and the forecast) and to see if those errors are correlated with available information. The evidence is mixed. Many studies have found that inflation forecasts are not rational in this sense. But a problem with many of these surveys is that they do not consider the forecasts of professionals who have a financial stake in the accuracy of their forecasts. Michael Keane and David Runkle describe this problem as follows : Suppose you’re an economics professor who’s busy writing a paper to be presented in three days. Someone phones you for your forecast of next quarter’s interest rate on three-month Treasury bills. Because you’re paid for writing papers, not making forecasts (and because you’re very busy), you quickly tell the caller the first number that pops into your head—8 percent. Later in that day, while reading The Wall Street Journal, you see that the forward rate on three-month T-bills is 9 percent. Since bond prices rise when interest rates fall, it seems logical that you’d rush out and buy bonds. But you don’t because, when you think about it, 9 percent seems reasonable. Thus, 8 percent is an erroneous measure of your true expectation because you don’t act in the market as if it really were your expectation. Keane and Runkle’s analysis of professional forecasts finds, contrary to other studies, that forecast errors were not predictable on the basis of available information. This evidence is thus supportive of rational expectations, at least as far as professional forecasters are concerned. LECTURE SUPPLEMENT 18-6 Microfoundations and Aggregation One area of some methodological dispute in macroeconomics involves the desirability of microfoundations—microeconomic underpinnings for macroeconomic relationships. Part V of the textbook considers microfoundations in detail. It shows how careful microeconomic analysis both provides broad support for, and suggests refinements of, some of the key behavioral relationships in our models, such as the consumption function, the investment function, and the money demand function. Some economists think that this does not go far enough. They argue that the use of behavioral equations at all is a bad idea because of the Lucas critique. For example, they point out that the marginal propensity to consume might take on a different value if government policies are changed. In this case, they point out, there is really no such thing as “the” consumption function; there is instead a different consumption function for any configuration of government policy. These economists advocate that macroeconomic models should always be based entirely on first principles of microeconomics. The theories they propose often take the form of Robinson Crusoe models, in which the economy is analyzed as a single individual making choices about consumption, investment, and labor supply, in order to maximize his or her utility (happiness). An advantage of such models is that the effects of policies can be clearly traced through their impact on individual behavior. Yet other economists are skeptical. They point out that such models simply assume that the aggregate behavior of the economy looks just like the behavior of a single individual. The Nobel-Prize-winning economist James Tobin expresses this criticism as follows2: It is possible to give macro relations the veneer of rigorous derivation from utility or profit maximization by assuming the aggregate of firms to behave as if they were one firm, and so on. Since the procedure involves a dubious assumption of aggregation, it is not clear that it is as much an improvement of the rigor and robustness of macro relations as a gratification of the trained consciences of the theorists. For some important issues of macroeconomics . . . the assumption that agents are all alike is clearly not appropriate and their differences in circumstances and behavior should somehow enter the model. Candid macro economists have never deceived themselves that the equations of their models were more than simple and approximate descriptions of the diverse responses of individual agents of ever-changing relative weights in the aggregates. The Robinson Crusoe models criticized by Tobin generally exclude by assumption any possibility of agents making inconsistent decisions: there is no way that Robinson the producer of goods will make choices inconsistent with the wishes of Robinson the consumer. In the real world, these decisions are made by different people and might not be consistent. But simple representative-agent models based on explicit microfoundations need not be Robinson Crusoe models. Coordination failure models (see Chapter 19), for example, typically feature just one type of agent but emphasize that the interdependencies of agents’ actions can lead to bad outcomes. Tobin undoubtedly makes an important point: aggregation is a difficult problem for macroeconomics, and representative-agent models simply sweep it under the rug. In other disciplines, moreover, it is not always more helpful to look for microfoundations: we can understand the collision of two billiard balls more easily using classical rather than quantum mechanics. Yet the Lucas critique is also very powerful and convinces many economists that microfoundations must have a central role in macroeconomic theories. CASE STUDY EXTENSION 18-7 Spare a Thought for the Empirical Macroeconomist One of the most important tests of a theory is its ability to explain the data. Yet over and over again in the textbook we find that the data do not provide a definitive answer to the questions that concern macroeconomists and do not allow us to distinguish with complete confidence between different theories. Two points should be made about this. First, it is true that disputes do exist in the field of macroeconomics. An important reason why macroeconomists disagree is precisely because the data are open to different interpretations. But the theories presented in the textbook represent current thinking in macroeconomics. Other theories proposed in the past have been rejected because of their inability to explain the data; they are no longer in the textbooks. Second, the case study in Chapter 18 drives home the point that we are not even really sure what the facts are. We have only about 70 years’ worth of reliable macroeconomic data for the United States. Whereas we do have some data going back to the nineteenth century, those data are incomplete and unreliable. As macroeconomists, we cannot carry out experiments to generate more data; the only way we can acquire more data is to wait. It will probably only be the gradual acquisition of more and better data that will ultimately allow us to decide among different macroeconomic theories. CASE STUDY EXTENSION 18-8 The Response to Romer Prior to Christina Romer’s work, economists accepted as a largely unquestioned stylized fact the view that the U.S. economy was much more stable after World War II than prior to it. Romer’s work showed that there could be no such presumption: differences in the quality of the data for the two periods mean that comparisons are difficult and perhaps unreliable. These same problems with early data, however, led some economists and economic historians to be skeptical of Romer’s conclusions, since they in turn questioned whether or not Romer had actually made appropriate corrections to the historical data. One problem concerns GDP data before 1919. Prior to that date, information on incomes was lacking and so estimates are based largely on data on commodity output. Unfortunately, this accounts for only about half of GDP, since it neglects transportation, distribution, and services. Measuring volatility on the basis of commodity output is misleading because it is more variable than total GDP; researchers make a correction for this on the basis of some period in which data exist on both series. The question then is what period should be used? Romer argued that the period of the Great Depression should be excluded, and that this correction makes early GDP series less volatile. The economic historian David Weir argued in turn that there was no good evidence that the relationship between GDP and commodity output was different during the Depression, but that there are problems with other time periods. After examining a number of possible time periods for this comparison, Weir concludes that “[s]tabilization of real output since World War II seems to be a robust finding of the historical evidence.” Weir also questions Romer’s conclusions about the relative stability of unemployment. One issue here concerns labor-force participation. Romer argued that traditional estimates of unemployment volatility before World War II assumed that participation rates were too smooth. A more strongly procyclical participation rate implies that employment is less volatile. Weir, however, argues that labor-force participation is very likely to have been less procyclical than in the postwar period because the “discouraged worker” phenomenon is probably stronger postwar and because women may make use of downturns to exit the labor force temporarily for purposes of childbearing. He concludes that “[i]mposing the postwar pattern of cyclical response cannot be justified as the best estimate of prewar labor-force fluctuations.”3 These criticisms of Romer’s work in no way deny the importance of her contributions to the debate on macroeconomic stabilization. As a direct consequence of her work, most economists now recognize and agree that the difference between the pre- and postwar economic fluctuations is not as great as we once thought. Simply by provoking the debate, Romer’s work led to other advances in our understanding of historical data. Weir concludes as follows: “No one who has worked to create historical data series . . . would deny the possibility of improvement. As economists and econometricians devise new questions to put to the data, new problems will emerge. To the simple question of whether cyclical fluctuations around trend in [GDP] and unemployment have become smaller since World War II the data are more than adequate to deliver a definitive answer: yes.”5 ADDITIONAL CASE STUDY 18-9 Distrust of Policymakers Some economists believe that there may be a good case in principle for government intervention in the macroeconomy but doubt the ability of policymakers to enact the right policy in practice. Certainly, there is evidence that U.S. presidents lack basic macroeconomic literacy. As one example, does former President Bush understand the distinction between real and nominal interest rates? The following is an excerpt from a reply made by then Vice President Bush to a question on civil rights, during the 1984 vice presidential candidates’ debate (held in Philadelphia on October 11, 1984)2: We think of civil rights as something like crime in your neighborhoods. And, for example, when crime figures are going in the right direction that’s good, that’s a civil right. Similarly, we think of it in terms of quality of life, and that means interest rates. You know it’s funny, Mr. Mondale talks about real interest rates. The real interest rate is what you pay when you go down and try to buy a TV set or buy a car, or do whatever it is. The interest rates when we left [sic] office were 21.5 percent. Inflation! Is it a civil right to have that going right off the chart so you’re busting every American family, those who can afford it the less? This is not a new problem. Going back a little further, Edward Tufte notes that presidents “have not been blessed with great economic sophistication, if anecdote can serve as evidence. President Kennedy said that the way he remembered the difference between fiscal and monetary policy was that monetary and Martin (William McChesney Martin, then chairman of the Federal Reserve) both began with M. President Nixon once pointedly expressed his distaste for extended briefings on foreign exchange problems, particularly those relating to the lira.” Perhaps our best hope is that, even if our leaders do not understand economics well, they still surround themselves with competent economic advisers. For example, President Kennedy’s Council of Economic Advisers and its accompanying staff/consultants in 1962 included three future winners of the Nobel Prize for economics, as well as many other illustrious economists. Two members of President Clinton’s Council of Economic Advisers, Alan Blinder and Joe Stiglitz, are authors of (other!) leading economics textbooks, and Stiglitz also was awarded the Nobel Prize. And a former chair of President Obama’s Council of Economic Advisers, Christina Romer, is the economist whose writings on the Great Depression and stability of the U.S. economy are discussed in Chapter 18. ADDITIONAL CASE STUDY 18-10 The Political Business Cycle That politicians may manipulate the economy to serve partisan aims has been documented by the political scientist Edward Tufte. As one example, Tufte considered transfer payments between 1961 and 1976. He noted that the amount of transfer payments was generally increasing through time, implying that we would expect the volume of transfer payments to be higher in December than in November, and in turn higher in November than in October. In all odd-numbered (that is, nonelection) years this was true, but in half of the election years, transfer payments instead reached their yearly peak in October or November. In those four years, moreover, the peak was in October when November transfer payments arrived after Election Day and in November when that month’s transfer payments arrived before Election Day. As another piece of evidence for the political business cycle, Tufte looked at the performance of 27 democracies between 1961 and 1972 and asked whether the growth rate of real income increased more often in election years than in nonelection years. In 19 of the 27 cases, the answer was yes.2 Tufte also discusses the 1972 election in some detail, providing evidence that the Nixon administration put pressure on the Federal Reserve (chaired at that time by Arthur Burns) to stimulate the economy. “The Fed accelerated the growth of money in 1971–1972. Whether it did so because of the election or for some other reason remains a matter of controversy. Critics of the Fed have pointed to Burns’s ties with Nixon for many years and the Burns-Nixon proposal to stimulate the economy before the 1960 election.” Fiscal policy was also expansionary. Whereas the federal government ran a $3 billion surplus in 1969 and a $3 billion deficit in 1970, it ran $23 billion deficits in 1971 and 1972. Tufte also documents substantial increases in almost all transfer payments in the last quarter of 1972. “It appears that in 1972 President Nixon was willing to sacrifice conservative anti-spending principles and absorb ideological defeats on beneficiary legislation in exchange for the increased flow of transfers immediately before the election.”5 ADDITIONAL CASE STUDY 18-11 The Political Business Cycle at Its Worst Theories of the political business cycle suggest that politicians may manipulate the economy to achieve political ends. For example, an incumbent president may try to generate a boom in an attempt to be reelected. There is a flip side to this. A leader presiding over an economy that is doing poorly may need other ways to increase his or her chances of reelection. Gregory Hess and Athanasios Orphanides investigated the possibility that U.S. presidents who face reelection in bad economic times may be more likely to take the country into a war. They define a war fairly broadly as “an international crisis in which the United States is involved in direct military activity that leads to violence.” For the period 1953–1988, they find that there is usually about a one-third chance of the country’s entering a war in a given year. But if the president is up for reelection and if the economy is doing poorly, the probability of war approximately doubles. (The exact findings vary depending upon the measure of economic performance that is used.) Incumbent presidents hope that the economy will be doing well when they are up for reelection. Hess and Orphanides’s findings suggest that the rest of us should hope so too. Supplement 3-8, “Wars and Interest Rates,” points out that wars often provide natural experiments for economic theories. Whereas fiscal policy may respond to the state of the economy, and so be endogenous, we would normally think of wars as exogenous events. One other minor implication of Hess and Orphanides’s results is to cast doubt on this presumption. Hess and Orphanides have found that even wars may be endogenous. ADDITIONAL CASE STUDY 18-12 The Economy Under Democratic and Republican Presidents As discussed in the Chapter 18 of the text and in the previous two supplements, one view of politicians is that they have a short time horizon that coincides with the election cycle. According to this view, policies are designed to ensure reelection rather than with the long-term health of the economy in mind. If this view is correct, then one would expect to find similar patterns of macroeconomic policies regardless of which political party is in office. But for the United States, data on real GDP growth suggest that the two political parties choose different macroeconomic policies. In other words, politicians are not simply opportunists trying to guarantee their reelection, but instead may be expressing the partisan preferences of their constituencies. Table 1 shows economic growth in each of the four years of presidential administrations since 1948. Average growth in Republican administrations is quite a bit lower than in Democratic administrations during the first two years of a term, but then is higher in the last two years. In addition, negative growth typically occurs in the second year of a Republican administration, whereas growth is usually booming during the second year of a Democratic administration. Table 1 Real GDP Growth During Democratic and Republican Administrations (percentage-change, 4th quarter over 4th quarter) Year of Term Democratic Administrations Truman -1.5 13.4 5.5 5.3 Kennedy/Johnson 6.4 4.3 5.2 5.1 Johnson 8.5 4.5 2.7 5.0 Carter 5.0 6.7 1.3 0.0 Clinton I 2.6 4.1 2.3 4.5 Clinton II 4.4 5.0 4.7 2.9 Obama I -0.2 2.7 1.7 1.6 Obama II 3.1 2.4 Average 3.5 5.4 3.3 3.5 Republican Administrations Eisenhower I 0.5 2.7 6.6 2.0 Eisenhower II 0.4 2.7 4.5 0.9 Nixon 2.1 -0.2 4.4 6.9 Nixon/Ford 4.0 -1.9 2.6 4.3 Reagan I 1.3 -1.4 7.8 5.6 Reagan II 4.3 2.9 4.4 3.8 Bush (senior) 2.8 0.6 1.2 4.3 G.W. Bush I 0.2 2.0 4.4 3.1 G.W. Bush II 3.0 2.4 1.9 -2.8 Average 2.1 1.1 4.2 3.1 Source: Bureau of Economic Analysis, U.S. Department of Commerce. A possible interpretation of this evidence is that the two parties have different preferences regarding inflation compared to unemployment. Republican administrations come into office seeking to bring down inflation and thus pursue restrictive policies, which may entail a recession. Democratic administrations enter office seeking to lower unemployment and thus pursue expansionary policies, even if these raise inflation. This interpretation does not necessarily argue for rules to constrain policymakers, since any rule would limit the ability of the electorate to express its will. LECTURE SUPPLEMENT 18-13 Price Level Versus Inflation Targeting Inflation targeting differs from price level targeting: with price level targeting the central bank must correct its past mistakes, which can increase the volatility of prices. To illustrate, suppose that in year 1 country A chooses to target its price level for the next three years along the path detailed in Table 1, while country B chooses to target its inflation rate for the next three years at 2 percent. At first glance it might seem that these are identical targets. The price level target is based on a 2 percent yearly rise in prices. However, operationally the targets are different. Suppose that both countries meet their target for year 2. In country A the price level is 102 and in country B the inflation rate is 2 percent. The following year both countries overshoot their targets. In country A the price level rises to 108.1 while in country B the inflation rate is 6 percent. Prices in both countries rose by 6 percent between year 2 and year 3. The central bank of country A must act to bring the price level back to its target. Because the actual price level in year 3 is above the year 4 target level, the central bank must deflate. To meet its target in year 4, prices must fall by 1.8 percent. In contrast, the central bank of country B merely needs to reduce the inflation rate, bringing it back to the 2 percent target. In country A prices must fall; in country B the rise in prices must be reduced. Table 1 Price Level Versus Inflation Targeting Year Price Level Target Inflation Target 1 100.0 2 102.0 2% 3 104.0 2% 4 106.1 2% CASE STUDY EXTENSION 18-14 Inflation Targeting In 1989 New Zealand passed legislation requiring the central bank (in consultation with the Ministry of Finance) to establish inflation targets. Over subsequent years, many other countries followed New Zealand’s example. And in 2012, the United States adopted a policy of targeting inflation, when the Federal Open Market Committee of the Federal Reserve set a target of 2 percent. Table 1 provides details on inflation targeting in seven industrial countries and the Euro-zone. Most of these countries set a range within which inflation is allowed to fluctuate and update the target every few years. The Federal Reserve seeks to “maintain an inflation rate of 2 percent over the medium term,” although actual inflation may differ from that rate in the short term. A key motivation for setting inflation targets is to enhance transparency in monetary policy. Most targets are specified over an intermediate window of time, usually about a year or so. Having a central bank committed to an inflation target helps households and businesses make better decisions by reducing uncertainty associated with policy. And, as the text notes in Chapter 18, even though most countries do not have explicit sanctions for failing to hit the inflation target, their presence does make central bankers more accountable and constrains central-bank discretion Table 1 Inflation Targeting in Industrial Countries Who Sets Country/Area Current Target Index Targeted Target? Australia 2–3% CPI Central Bank and Government Canada 1–3% CPI Central Bank and Government Euro-zone Below but close to 2% Harmonized Index of Consumer Prices Central Bank Israel 1–3% CPI Central Bank and Government New Zealand 1–3% CPI Central Bank and Government Sweden 2% (with small deviations permitted) CPI Central Bank United Kingdom 2% (with permissible fluctuations of ±1%) CPI Government United States 2% PCE Central Bank CASE STUDY EXTENSION 18-15 Central-Bank Independence and Growth Figure 1 shows the relationship between the degree of central-bank independence and economic growth in 16 countries from 1955 to 1987. There appears to be no connection between an independent central bank and higher or lower economic growth. Source: From A. Alesina and L. Summers, “Central Bank Independence and Macroeconomic Performance: Some Comparative Evidence,” Journal of Money, Credit, and Banking 25, no. 2 (May 1993): 151–62, reprinted by permission. Copyright 1993 by the Ohio State University Press. All rights reserved. CASE STUDY EXTENSION 18-16 Measuring Central-Bank Independence Studies of central-bank independence and the macroeconomy measure independence by first establishing a list of criteria thought to characterize independence and then examining central-bank charters and laws to determine which criteria are met by each central bank in the study. Legal measures of independence, however, may not be reflective of the true constraints affecting central banks. Thus a country believing that it can achieve low inflation simply by granting its central bank legal independence may be sorely disappointed. Some countries that recently passed legislation to increase the independence of their central banks are finding that operational independence is more difficult to achieve. For example, in June 1996 the Russian parliament passed a law ordering the Central Bank of Russia to transfer $1 billion to the government to finance President Yeltsin’s preelection spending promises. While the central bank protested that such a requirement jeopardized its independence, it complied with the measure. In January 1993 Venezuela increased the legal independence of its central bank. Sixteen months later the governor (head) of the central bank and several board members resigned, claiming that the government was trying to reduce the autonomy of the central bank. Most recently, in June 1996, the governor of Chile’s independent central bank resigned following a conflict with the government. Given that a conflict between the central bank and the government may end in the resignation of the head of the central bank, Alex Cukierman has suggested that the tenure of the head of the central bank may serve as a good proxy for the degree of independence. Cukierman studied the link between central-bank independence and inflation in a large sample of developed and developing countries. While he found that developed countries with legally independent central banks had lower average inflation than countries whose central banks are controlled by the government, this relationship did not hold for developing countries. However, in developing countries Cukierman found a negative relationship between the tenure of the head of the central bank and the rate of inflation. Another issue in measuring central-bank independence is the extent to which this should reflect goal independence versus instrument independence. In other words, should independence include freedom to select the ultimate goals of monetary policy or should it be limited to freedom in choosing procedures for achieving a given goal or set of goals. Guy Debelle and Stanley Fischer examine the link between the legal requirement that price stability be one of the goals of the central bank and inflation performance.5 They found that having a legally mandated goal of price stability results in lower inflation. They also found that the degree of independence for a central bank in choosing how it implements monetary policy—as measured by its ability to avoid financing the government budget and freedom to set interest rates—is negatively correlated with inflation. Thus, they argue that a central bank should not have goal independence but should have instrument independence. In their view, a central bank should be required to have price stability as its goal, but at the same time should be given operational freedom to achieve this goal. LECTURE SUPPLEMENT 18-17 Additional Readings Two very accessible and concise books on stabilization policy are Central Banking in Theory and Practice, The Lionel Robbins Lectures, by Alan S. Blinder (MIT Press, 1998) and Inflation, Unemployment, and Monetary Policy, The Alvin Hansen Symposium on Public Policy, by Robert M. Solow and John B. Taylor, edited by Benjamin Friedman (MIT Press, 1998). Students will find these books cover a wide range of issues related to macroeconomic policy. CHAPTER 19 Government Debt and Budget Deficits Notes to the Instructor Chapter Summary This chapter begins with an analysis of the size of the U.S. government debt and the outlook for the debt. It then discusses problems in accurately measuring the deficit and the debt. The last three sections concentrate on the effects of the government debt on the economy, looking at the effects through the traditional view and the Ricardian view, along with other perspectives. Comments As with the material on stabilization policy in Chapter 18, students find the study of government debt to be quite relevant for understanding the debates concerning economic policy that are frequently covered in the news media. The chapter includes a section that discusses balanced budgets and optimal fiscal policy—helping students understand why deficits might sometimes be beneficial. The case study about the fiscal future considers the reasons behind the troubling long-term budget outlook. The chapter also introduces the role of debt in recent financial crises, a topic covered more completely in Chapter 20. This chapter provides a particularly good time to review and bring together other material in the textbook (in particular, short-run and long-run crowding out and the implications for capital accumulation and growth). The material can be covered in one lecture. I stress the distinction between the debt and the deficit as an important piece of macroeconomic literacy and encourage students to listen for the all-too-common confusion on this topic. After covering the material in this chapter, I find it interesting to ask students what they think about Ricardian equivalence and which arguments (for or against) they find most compelling. Use of the Dismal Scientist Web Site Go to the Dismal Scientist Web site and download annual data over the past 20 years for the GDP price index and the federal debt outstanding (in the Flow of Funds data). Compute the inflation rate from the GDP price index. Use the inflation rate along with the debt outstanding to compute the component of interest on the debt that is due to inflation. Now download the federal government budget deficit over the past 20 years. Subtract the inflation component from the deficit to arrive at an adjusted deficit. Assess how this adjustment changes the pattern of the deficit over this period. Chapter Supplements This chapter includes the following supplements: 19-1 Debt and Deficits: The Data 19-2 How Important Is Crowding Out? (Case Study) 19-3 Structural and Cyclical Deficits 19-4 Generational Accounting 19-5 The Government Budget Constraint 19-6 Borrowing Constraints Using the Fisher Diagram 19-7 Social Security Benefits and Ricardian Equivalence 19-8 Is Everything Neutral? 19-9 Does Altruism Matter? (Case Study) 19-10 Unpleasant Monetarist Arithmetic 19-11 Inflation Indexed Bonds and Expected Inflation (Case Study) 19-12 Additional Readings 455 Lecture Notes Introduction One important controversy in macroeconomics concerns the role and significance of government debt. Government deficits and government debt moved to the forefront of U.S. economic and political debate in the 1980s, as the federal government increased its debt at a rate unprecedented in peacetime. More recently, during the 2000s, the reemergence of large deficits once again put this issue at the top of the policy agenda. 19-1 The Size of the Government Debt The net debt of the U.S. federal government was nearly 86 percent of GDP in 2014. This was Table 19-1 well below the 143 percent share of GDP for Japan’s debt but larger than the 3.5 percent share Figure 19-1 of GDP for Australia’s debt. Relative to other countries, the U.S. debt is roughly in the middle of the pack. Another way to view its size is to note that in 2014, the $12.9 trillion debt held by the public amounted to $41,000 for each of the 316 million people in the United States. Given that the average person’s lifetime earnings are about $2 million, a per capita debt of $41,000 does not appear overwhelming. Another important feature of the debt is that over time its size has fluctuated greatly. Debt has typically risen during wars and fallen during peacetime. At the end of World War II, for example, the debt held by the public reached the historic high of 106 percent of GDP and then fell sharply over the subsequent decades. A major exception to this pattern was the large increase in the debt during the 1980s and early 1990s. The debt declined as a share of GDP during the late 1990s as the budget swung from deficit into surplus due to tax increases, spending restraint, and strong economic growth. Recession, tax cuts, and two wars reversed this trend, and the debt began to rise again as a share of GDP during the early and mid-2000s. The debt then grew sharply from 2008–2011 as a result of the deep recession that began in late 2007 and the large fiscal stimulus plan that Congress adopted to prop up the economy. Case Study: The Troubling Outlook for Fiscal Policy The long-term budget outlook is distressing. According to the Congressional Budget Office Supplement 19-1, “Debt and Deficits: (CBO), spending on programs for the elderly over the next several decades is projected to The Data” increase more rapidly than the revenues supporting these programs, leading to rising deficits Supplement 19-2, and, thus, an accumulating federal debt. In particular, spending on Social Security, Medicare, “How Important Is and Medicaid has risen from less than 1 percent of GDP in 1950 to about 10 percent today and Crowding Out?” is projected to reach about 17.5 percent of GDP over the next half century. The projected increase in spending is due to the aging of the population as life expectancy has risen and fertility rates have declined. In addition, technological advances in health care that improve and extend people's lives also push upward the cost of health care programs. CBO forecasts show the federal debt as a share of GDP rising to historic highs by the latter part of this century. Long-term forecasts, like any forecast, involve much uncertainty. Small changes in assumptions about demographic trends or economic performance can have large changes in projected outcomes for deficits many decades in the future. Furthermore, the CBO’s long-term projection assumes that current policies remain unchanged—something not likely to occur in the face of intensifying pressure to address the problem of long-term fiscal imbalances. Of course, addressing the problem is easier if it is done sooner rather than later. 19-2 Problems in Measurement Some disagreements about fiscal policy arise because of difficulties in obtaining an accurate and economically meaningful measure of the deficit. There are many subtleties and arcane details of government financial statistics. Here we consider four particularly important measurement problems. Measurement Problem 1: Inflation The deficit as usually measured is not adjusted for inflation. Part of the deficit is interest payments on the government debt. By the Fisher equation, these interest payments equal iD = (r + π)D. The real interest payments are equal to rD, so the deficit is overstated by an amount equal to πD. When inflation is high, this overstatement can be large: In 1979, for example, inflation was 8.6 percent and the debt was $495 billion, implying that the deficit was overstated by about $43 billion. Corrected for inflation, the reported budget deficit of $28 billion turns into a budget surplus of $15 billion. Measurement Problem 2: Capital Assets The government’s budget deficit, as usually measured, accounts only for changes in the government’s liabilities and not for changes in the government’s assets. Thus, if the government were to sell a national park to developers and use the revenue to reduce its debt (liabilities), the budget deficit would be lower. In this case, however, the reduction in the deficit does not mean the government has shifted to a more responsible fiscal policy. Capital budgeting by contrast would measure the budget deficit as the change in liabilities minus the change in assets. Under capital budgeting, the sale of the national park would have no effect on the budget deficit. While superior in principle, capital budgeting is very difficult in practice because of difficulties in deciding exactly what sales and expenditures should be counted as changes in the government’s stock of assets. Measurement Problem 3: Uncounted Liabilities Certain liabilities of the government, such as government employee pensions and accumulated Social Security benefits, are excluded in calculation of the deficit. This is a particular problem in the case of contingent liabilities, such as federal deposit insurance, that are paid only if certain prespecified events (for example, a bank failure) occur. Measurement Problem 4: The Business Cycle Supplement 19-3, Automatic changes in the deficit occur due to the direction in which the economy is going. “Structural and During a recession, for example, the budget deficit rises due to depressed tax revenue and Cyclical Deficits” increased government spending. These changes make it more difficult to use the deficit to monitor fiscal policy because one needs to know if a rise or fall in the deficit has occurred because of a policy change or because of economic growth or decline. A cyclically adjusted budget deficit is based on estimates of what government spending and tax revenue would be if the economy were operating at its natural rate of output and employment. Summing Up These measurement problems make the task of assessing government fiscal policy difficult. The Supplement 19-4, “Generational only safe lesson is that any simple statistic, such as the government deficit, provides only one Accounting” imperfect piece of information about the state of the government’s finances. More generally, good economic analysis requires understanding that all data are imperfect. 19-3 The Traditional View of Government Debt We begin our theoretical analysis with a discussion of the standard IS–LM view of the deficit. An increase in the deficit means either lower taxes or else increases in government expenditures or transfer payments. All of these imply higher spending, either directly or through their effect on disposable income. Hence, increases in the deficit are expansionary and are associated with outward shifts of the IS curve. In the short run, therefore, increases in the deficit lead to higher interest rates and a higher GDP. The higher GDP is directly the result of the increase in spending. The higher interest rates arise because increased GDP increases the transactions demand for money, thus pushing up interest rates and reducing investment. This is the crowdingout effect. It is important when money demand is insensitive to interest rates and investment is 457 sensitive to interest rates. Also, higher interest rates lead to a capital inflow, causing an appreciation of the exchange rate and crowding out of net exports. This analysis implicitly makes an assumption that deficits are financed by borrowing. If the deficit were instead to be financed by increased M, we would see outward shifts of both the IS and LM curves. In the long run, after price adjustment returns the economy to full employment, we find further crowding out. Increases in the price level reduce the real money supply, pushing interest rates and the exchange rate up further and thus further decreasing investment and net exports. In the long run, we see complete crowding out: Private spending is reduced by an amount exactly equal to the increased government spending. FYI: Taxes and Incentives The textbook represents the tax system with a single variable, T, treating taxes as a simple lumpsum amount. But in reality, we need to consider how tax revenue is actually raised. The field of public finance spends a good deal of time addressing the benefits and costs of alternative types of taxes. An important conclusion is that taxes have effects on economic incentives. When people are taxed on income from working, they have less incentive to work, and when they are taxed on income from capital, they have less incentive to invest. Accordingly, changes in taxes lead to changes in incentives, and this may have effects on the overall economy. In particular, a reduction in taxes may encourage more work and investment, thereby increasing aggregate supply. Supply-siders believe that these effects are large enough to pay for themselves through increased tax revenue. Economists agree that taxes affect incentives and that incentives affect aggregate supply, but most do not believe that tax cuts, in general, pay for themselves. Discussions in recent years about tax reform, including the possibility of moving to a consumption tax, have centered on reducing the disincentive effects of taxes. 19-4 The Ricardian View of Government Debt There is not universal agreement on this reasoning. Some economists have argued that the logic behind the IS–LM model is substantially flawed and that we should not expect to see crowding out, nor should we expect to see deficits having any substantial impact on the economy. The main argument is known as Ricardian equivalence, so named because the basic idea was first noted by the nineteenth-century British economist David Ricardo. The current role of this theory in macroeconomic debate is principally the result of work by economist Robert Barro. The Basic Logic of Ricardian Equivalence Ricardian equivalence argues that the effects on the economy are the same whether government expenditures are financed by taxes or by borrowing (hence the term “equivalence”). Stated differently, Ricardian equivalence argues that a debt-financed tax cut should have no effect on the economy. This is in stark contrast to the standard view, in which the two types of financing are not equivalent at all. The essence of this argument is that consumers understand the government’s intertemporal budget constraint. If the government cuts taxes and issues more bonds today, it is committing itself to higher taxes in the future. Rational households will realize that higher taxes are ultimately necessary and so will save the tax cut today in anticipation of their future tax liability. Changes in private saving will exactly offset any changes in public saving. Ricardian equivalence does not imply that government expenditures have no effect on the economy. In general, changes in government spending, financed either by taxation or borrowing, will have significant real effects. Ricardian equivalence is a statement about two different ways of financing a given amount of spending. Consumers and Future Taxes Supplement 19-5, For Ricardian equivalence to hold, consumers must be forward-looking so that future tax “The Government liabilities have a large effect on current consumption. There are a number of reasons why this Budget Constraint” may not be so. Myopia One obvious possibility is that Ricardian equivalence assumes too much rationality; it may be that people do not take account of the government budget deficit when making their consumption/saving decisions. Borrowing Constraints Ricardian equivalence breaks down if individuals face borrowing Supplement 19-6, constraints. A liquidity-constrained individual consumes all her income and would consume “Borrowing more if she were able to borrow. If the government cuts her taxes, then she is able to increase Constraints Using her consumption and so will not save the tax cut, in contrast to the predictions of Ricardian the Fisher Diagram” equivalence. Supplement 19-7, “Social Security Case Study: George H.W. Bush’s Withholding Experiment Benefits and Ricardian In his 1992 State of the Union address, President George H.W. Bush reduced income tax Equivalence” withholding. The stated intention was to avoid overwithholding, thus ensuring that workers were paid more now but would receive lower refunds when they eventually submitted their income taxes. Such a policy can be viewed as a simple test of Ricardian equivalence: Workers knew that their current tax payments were lower but that their future payments (the next April 15) would be higher by an equal amount. A survey conducted after this policy was announced found that 43 percent of consumers planned to alter their consumption; that is, they planned to behave in a non-Ricardian way. Future Generations In reality, individuals’ and governments’ decisions are made over long time horizons. It is not the case that if the government cuts taxes this year, it will have to raise them next year. The government budget constraint merely tells us that if taxes are cut this year, the government will have to raise taxes at some unspecified future date. It could be next year, it could be 20 years from now, it could be 200 years from now. So, do we care if the U.S. government will have to raise taxes in the year 2212? We won't be around to worry about it. The economist Robert Barro has argued that we might indeed care. He suggests that if people care about their children’s welfare, then Ricardian equivalence could operate through bequests. If the government cuts taxes today, the current generation may not face higher taxes in the future, but future generations might. An individual’s tax cut today may mean a tax increase for her great-great-great-grandchildren. Now, if every parent cares about her children, then, by implication, every parent also cares indirectly about her grandchildren. (Note that you don’t have to care directly about your grandchildren for this argument to work.) In effect, Barro suggests that we think not in terms of individuals, but rather in terms of dynasties stretching into the future. The result is that, even though people die, they may behave as if they were going to live forever. The argument is elegant but hardly compelling. Many authors have noted problems with it. Supplement 19-8, Most of the objections are less elegant but nonetheless important. For example, some people do “Is Everything Neutral?” not have children so, for them, the tax cut is a straight gain. Moreover, even people who do have children may not wish to bequeath anything to them. Case Study: Why Do Parents Leave Bequests? Empirical evidence indicates that bequests are important for a large part of the population but Supplement 19-9, “Does Altruism that it is unclear how important altruism is as a reason for these bequests. Some economists have Matter?” suggested that parents use bequests strategically to control their children. Overall, the evidence on bequests is not particularly favorable to Ricardian equivalence. Making a Choice The debate over Ricardian equivalence is important because the two views of government debt have very different implications for economic policymaking. If Ricardian equivalence holds, then the government cannot use tax cuts to stimulate the economy, since consumers always view these as transitory. At the same time, tax cuts do not lead to crowding out of investment, since Ricardian equivalence implies that reduced government saving is offset by increased private saving. Traditional economists are worried about the effects of current deficits on our future standard of living, whereas economists who hold the Ricardian view regard concern about the deficit as misplaced. 459 FYI: Ricardo on Ricardian Equivalence In an 1820 article, “Essay on the Funding System,” David Ricardo poses the argument that bears his name: Ricardian equivalence. Ricardo, however, actually rejected the notion that people would consider debt finance to be as burdensome as current taxes. He today would disagree with what has become known as Ricardian equivalence. 19-5 Other Perspectives on Government Debt This section presents four additional perspectives on the effects of government debt that can modify either the traditional or Ricardian view. Balanced Budgets Versus Optimal Fiscal Policy Some commentators and politicians advocate a constitutional balanced-budget amendment, requiring the federal government to be in budget balance at all times. Such a rule has superficial appeal as a means to avoid large federal deficits. Most economists do not support this idea, however. One reason is that budget deficits and surpluses arise in part because of the automatic stabilizers discussed earlier. If the economy goes into a recession, tax revenues go down and transfers increase. Thus T falls, increasing the deficit. A balanced-budget amendment would eliminate such automatic stabilizers and so exacerbate economic fluctuations. Second, public finance teaches that the distortionary effects of taxation are minimized by smooth tax rates. A balanced-budget requirement would force the government to increase taxes when expenditures were high or revenues were low and so might increase distortions. Finally, deficits and surpluses are a means of intergenerational redistribution. If policymakers wish to pursue distributional goals across generations, then budget deficits and surpluses are a natural tool. Fiscal Effects on Monetary Policy According to this view, government debt affects inflation expectations because a high level of debt may encourage a government to resort to inflation to reduce the real value of the debt. Supplement 19-10, There is little evidence, however, of a link between inflation and the level of debt in developed “Unpleasant Monetarist countries. There are three explanations for this lack of evidence. First, most governments have Arithmetic” not had problems financing deficits by selling debt and hence have not needed to resort to inflationary finance. Second, central banks resist political pressures for expansionary monetary policy to reduce the real value of the debt. Third, even most politicians acknowledge that inflation is a poor solution for fiscal problems. Debt and the Political Process The economist Knut Wicksell in the nineteenth century was the first to articulate the idea that deficit finance hides the costs of the fiscal decisions made by politicians by pushing the cost on to future taxpayers. Most recently, James Buchanan and Richard Wagner have noted that deficit finance gives the illusion that one can get “something for nothing.” (Note that if Ricardian equivalence holds, individuals will reject this idea.) Martin Feldstein notes that only the requirement of a balanced budget forces government to determine if the benefits of spending justify the costs. Economists who support these ideas favor a balanced-budget amendment that would allow the government to run deficits only in times of war or depressions. Others argue that such a policy would unduly limit economic policy. International Dimensions High levels of government debt may increase the likelihood of default. Such a possibility can result in a sudden loss of investor confidence, leading to a withdrawal of capital and a collapse in the foreign exchange value of the currency in conjunction with a rise in interest rates. While the possibility of default may be insignificant for most advanced economies, the same is not true for the rest of the world. Several Latin American countries defaulted on their debts in the 1980s as did Russia in 1998. And in 2011, it appeared that Greece, an advanced economy and member of the eurozone, was likely to default on its debt. Another argument is that a nation’s political clout in world affairs is tied to its status as a creditor nation. To the extent that government deficits reduce national saving and result in borrowing from abroad, a country is less likely to be able to influence world affairs. While the theory may explain the emergence of the United States as a world power, it fails to explain the continued dominance of the United States despite its role as a debtor nation. Case Study: The Benefits of Indexed Bonds Supplement 19-11, In 1997 the U.S. Treasury began issuing bonds that were adjusted for inflation so that the real “Inflation Indexed value of the principal and returns was constant. While the primary reason for issuing such bonds Bonds and Expected is to eliminate the inflation risk borne by investors and hence increase the attractiveness of the Inflation” Supplement 5-11, bonds, there are other benefits. Indexed government bonds may encourage the private sector to “U.S. Treasury also issue similar indexed securities. These securities would be particularly attractive to those Issues Indexed looking for long-term saving instruments. Inflation indexed bonds also reduce the incentive for Bonds” the government to use surprise inflation, and they provide data that is useful to both government and private forecasters about expected inflation. 19-6 Conclusion The prominence of fiscal policy issues in the political arena ensures that economists and policymakers will continue to debate the short- and long-term effects of deficits and debt on the economy. ADDITIONAL CASE STUDY 19-1 Debt and Deficits: The Data The federal government receives revenue from taxes (principally income taxes, corporate taxes, and Social Security contributions) and spends on national defense and other purchases, transfer payments, and interest payments on the national debt. Table 1 provides revenues and expenditures for 2013 (in billions of dollars). Table 1 Federal Receipts and Expenditures Income taxes 1287 Social Security 1092 Other (including corporate) 734 Total receipts 3113 Total consumption expenditures 963 Transfer payments 2322 Interest payments 417 Other 60 Total current expenditures 3762 Deficit 649 Source: Department of Commerce, Bureau of Economic Analysis. Table 2 (see next page) provides data on the federal budget deficit and federal debt held by the public for fiscal years 1980–2013 (in billions of dollars).3 Table 2 Federal Budget Deficit and Debt Year Deficit (–)/Surplus (+) Debt Debt/GDP (%) 1980 –74 712 25 1981 –79 789 25 1982 –128 925 28 1983 –208 1,137 31 1984 –185 1,307 32 1985 –212 1,507 35 1986 –221 1,741 38 1987 –150 1,890 39 1988 –155 2,052 39 1989 –153 2,191 39 1990 –221 2,412 40 1991 –269 2,689 44 1992 –290 3,000 46 1993 –255 3,248 47 1994 –203 3,433 47 1995 –164 3,604 47 1996 –108 3,734 46 1997 –22 3,772 44 1998 69 3,721 41 1999 126 3,632 38 2000 236 3,410 33 2001 128 3,320 31 2002 –158 3,540 32 2003 –378 3,913 34 2004 –413 4,296 35 2005 –318 4,592 35 2006 –248 4,829 35 2007 –161 5,035 35 2008 –459 5,803 39 2009 –1,413 7,545 52 2010 –1,294 9,019 60 2011 –1,300 10,128 65 2012 -1,087 11,281 70 2013 -680 11,983 71 2014 -485 12,780 73 Source: Department of Commerce, Bureau of Economic Analysis. Note: Debt is end of fiscal year. CASE STUDY EXTENSION 19-2 How Important Is Crowding Out? Projections of future deficits should take into account the crowding-out effect of expansionary fiscal policies. Both the direct rise in interest rates and the dampening effect on output tend to increase the deficit. This occurs because higher interest rates increase the cost of financing existing debt, and less expansion in output (and incomes) means lower taxes than otherwise. As the IS–LM analysis indicates, expansionary government policy (running deficits) shifts the IS curve outward, increasing aggregate demand. However, because the increase in government spending raises interest rates, it “crowds out” some private investment and hence diminishes the effect on output. If government spending had no effect on interest rates (a flat LM curve), then crowding out would not occur. A Congressional Budget Office study examined the effect of an increase in government spending to support programs for the elderly under two scenarios: the assumption that crowding out occurs and the assumption that crowding out does not occur. Table 1 shows the projections for the federal deficit and debt when crowding out occurs, while Table 2 gives the projections if no crowding out occurs. In Table 1 the deficit is projected to increase to 23 percent of GDP by 2050, while in Table 2 the projected increase is only 7 percent of GDP. Similarly, while the debt increases to 206 percent of GDP if crowding out occurs (Table 1), the debt is 94 percent of GDP if no crowding out occurs (Table 2). Table 1 Projections of Federal Receipts and Expenditures with Economic Feedback (percentage of GDP) 2000 2010 2020 2030 2040 2050 Receipts 21 20 20 20 20 20 Expenditures 21 20 22 25 30 43 Consumption 5 4 4 4 4 4 Transfers, grants, and subsidies Social Security 4 5 6 6 7 7 Medicare 3 4 5 6 7 7 Other 6 6 6 6 7 7 Net interest 2 1 1 2 6 19 Deficit (–) or surplus 0 1 –1 –5 –10 –23 Debt held by the public 42 21 17 40 93 206 Source: Long-Term Budgetary Pressures and Policy Options, Congressional Budget Office, May 1998, Table 2-1. What explains these differences? Both scenarios show similar increases in spending on programs for the elderly as a share of GDP. If crowding out occurs, the increase in the deficit as a result of these increases in spending raises interest rates, driving up the cost of borrowing for both the government and private borrowers. This increase in borrowing costs is illustrated by the rise in net interest expenditures. If crowding out does not occur, interest costs rise from 2 percent of GDP in 2000 to 4 percent in 2050. If crowding out does occur, interest costs rise from 2 percent in 2000 to 19 percent in 2050. As interest costs raise the deficit, the government must borrow more, further raising interest rates and hence the cost of financing the deficit. This vicious circle is primarily responsible for the differences in the two forecasts. Table 2 Projections of Federal Receipts and Expenditures without Economic Feedback (percentage of GDP) 2000 2010 2020 2030 2040 2050 Receipts 21 20 20 20 20 20 Expenditures 21 19 21 24 26 27 Consumption 5 4 4 4 4 4 Transfers, grants, and subsidies Social Security 4 5 5 6 6 6 Medicare 3 4 5 6 7 7 Other 6 6 6 6 7 7 Net interest 2 1 1 1 2 4 Deficit (–) or surplus 0 1 –1 –3 –5 –7 Debt held by the public 42 21 15 32 61 94 Source: Long-Term Budgetary Pressures and Policy Options, Congressional Budget Office, May 1998, Table 2-4. ADDITIONAL CASE STUDY 19-3 Structural and Cyclical Deficits Deficits and surpluses arise naturally in an economy experiencing business-cycle fluctuations because of the presence of automatic stabilizers. This suggests a decomposition of the overall budget deficit (or surplus) into a cyclical deficit, because of automatic stabilizers, and a structural (or full-employment) deficit, which may better reflect the overall stance of fiscal policy. That is, the structural deficit is a measure of what the deficit would be if the economy were at the potential level of output, and the cyclical deficit is the difference between the actual and the structural. This distinction sheds particular light on the experience of the U.S. economy in the 1980s. The increase in the deficit in the early 1980s was primarily an increase in the cyclical deficit, caused by the 1982 recession. The 1982 tax cuts and spending increases, however, raised the structural deficit. After 1983 the cyclical deficit fell but the structural deficit continued to grow. By the end of the 1980s, output was close to the natural rate and the cyclical deficit was relatively small. The high deficits observed in the early 1990s were largely structural. The structural deficit declined sharply in the late 1990s and then moved into a small surplus by 2000 as a consequence of tax increases and spending cuts. With the economy operating above its potential level of output, the cyclical balance also moved into a moderate surplus. But a structural deficit emerged once again during the early and mid-2000s following tax cuts and increased spending associated with two wars. The onset of the financial crisis and steep recession led to a sharp increase in the structural deficit as a result of the stimulus program passed by Congress. Table 1 shows the structural balances of seven countries in 2009 and again in 2013. All of these countries had structural deficits in 2009 reflecting shifts toward expansionary fiscal policy in response to the Great Recession. Between 2009 and 2013, as fiscal policy became tighter, structural balances declined for all of these countries except for Japan, where the structural deficit widened, reflecting a move toward a more expansionary fiscal stance. Table 1 Structural Budget Balances, 2009 and 2013 Structural Balance as Percent of Potential GDP 2009 2013 United States –11.0 –4.4 Japan –7.3 –9.1 Germany –1.4 0.4 France –6.5 –3.4 Italy –3.7 0.2 Britain –10.1 –4.9 Canada –3.5 –2.6 Source: Organization for Economic Cooperation and Development, Economic Outlook, 2014. Note: Negative sign in table indicates deficit. LECTURE SUPPLEMENT 19-4 Generational Accounting Laurence Kotlikoff and others argue that the analysis of government budget deficits should be informed by the insights of the life-cycle model (discussed in Chapter 16). From the point of view of an individual, taxes and transfers in any given year are less important than the pattern of taxes and transfers over that individual’s entire lifetime. Kotlikoff’s system of generational accounting takes this approach. The aim is to calculate the net effect of lifetime taxes and transfers for a given cohort of people born at the same time. Generational accounting often comes to very different conclusions than the officially measured budget deficit about the effects of fiscal policy. For example, during the 1980s, many analysts viewed tax cuts and associated deficits as a burden on young workers who would some day have to pay off the accumulating debt. Kotlikoff agrees with this view but notes that other fiscal policy changes need to be accounted for before drawing overall conclusions. In particular, the 1980s also witnessed a significant long-term reform of the Social Security system, reducing benefits as well as raising taxes. When accounted for over people’s lifetimes, these changes actually redistributed income toward the younger generation. Despite its usefulness in providing insight over longer horizons, generational accounting is not likely to replace our standard measures of the budget deficit. One important shortcoming of the approach is the need to make assumptions about future policies over which people are likely to disagree. LECTURE SUPPLEMENT 19-5 The Government Budget Constraint To better understand the link between government debt and future taxes, it is useful to imagine that the economy lasts for only two periods. Period one represents the present, period two the future. In period one, the government collects taxes T1 and makes purchases G1; in period two, it collects taxes T2 and makes purchases G2. Because the government can run a budget deficit or a budget surplus, taxes and purchases in any single period need not be closely related. We want to see how the government’s tax receipts in the two periods are related to its purchases in the two periods. In the first period, the budget deficit equals government purchases minus taxes. That is, D = G1 – T1 , where D is the deficit. The government finances this deficit by selling an equal amount of government bonds. In the second period, the government must collect enough taxes to repay the debt, including the accumulated interest, and to pay for its second-period purchases. Thus, T2 = (1 + r) D + G2, where r is the interest rate. To derive the equation linking taxes and purchases, combine the two previous equations. Substitute the first equation for D into the second equation to obtain T2 = (1 + r) (G1 – T1) + G2. This equation relates purchases in the two periods to taxes in the two periods. To make the equation easier to interpret, we rearrange terms. After a little algebra, we obtain This equation is the government budget constraint. It states that the present value of government purchases must equal the present value of taxes. The government budget constraint shows how changes in fiscal policy today are linked to changes in fiscal policy in the future. If the government cuts first-period taxes without altering first-period purchases, then it enters the second period owing a debt to the holders of government bonds. This debt forces the government to choose between reducing purchases and raising taxes. Figure 1 uses the Fisher diagram (see Chapter 16 for more details) to show how a tax cut in period one affects the consumer under the assumption that the government does not alter its purchases in either period. In period one, the government cuts taxes by ∆T and finances this tax cut by borrowing. In period two, the government must raise taxes by (1 + r)∆T to repay its debt and accumulated interest. Thus, the change in fiscal policy raises the consumer’s income by ∆T in period one and reduces it by (1 + r)∆T in period two. The consumer’s lifetime income is the same as before the change in fiscal policy. Therefore, the consumer chooses the same level of consumption as she would have without the tax cut, which implies that private saving rises by the amount of the tax cut. Hence, by combining the government budget constraint and Irving Fisher’s model of intertemporal choice, we obtain the Ricardian result that a debtfinanced tax cut does not affect consumption. Figure 1 A Debt-Financed Tax Cut in the Fisher Diagram LECTURE SUPPLEMENT 19-6 Borrowing Constraints Using the Fisher Diagram Suppose a consumer faces two constraints: a budget constraint and a borrowing constraint (see Chapter 16 for more details). The budget constraint says that the present value of consumption must not exceed the present value of income. The borrowing constraint says that first-period consumption must not exceed first-period income. A debt-financed tax cut of ∆T raises first-period income by ∆T and reduces secondperiod income by (1 + r) ∆T, as shown in Figure 1. Because the present value of income is unchanged, the budget constraint is unchanged. Yet because first-period income is higher, the borrowing constraint allows a higher level of first-period consumption. The consumer now chooses point B rather than point A. Hence, Ricardian equivalence fails to hold. Figure 1 ADDITIONAL CASE STUDY 19-7 Social Security Benefits and Ricardian Equivalence David Wilcox’s work on consumers’ responses to changes in Social Security benefits (see Supplement 165) provides a test of Ricardian equivalence.1 Recall that Ricardian equivalence suggests that the timing of taxes is irrelevant to forward-looking, altruistic consumers. This result should hold equally for transfer payments: Consumers should view an increase in transfer payments today as signaling higher taxes in the future, so they would not alter their consumption behavior. (In the textbook, the variable T measures taxes minus transfer payments.) Wilcox found that consumption spending responds significantly to changes in Social Security benefits, suggesting that consumers do not behave in a Ricardian manner. 1 D. Wilcox, “Social Security Benefits, Consumption Expenditure, and the Life-Cycle Hypothesis,” Journal of Political Economy 97, no. 2 (April 1989): 288–304. Wilcox’s study is discussed in more detail in Supplement 16-5, “Do Consumers Anticipate Changes in Social Security Benefits?” ADVANCED TOPIC 19-8 Is Everything Neutral? One of the most clever arguments against Ricardian equivalence is a reductio ad absurdum put forth by Douglas Bernheim and Kyle Bagwell. They consider Robert Barro’s argument for intergenerational altruism and point out that, in fact, we would not only expect to see dynasties that are linked, but we should expect to see interconnected chains of individuals. The consequence is that everyone is ultimately related to everyone else, and so everything is neutral! The Barro argument contends that a redistribution of resources from members of one generation to another has no effects because individuals offset this transfer by means of bequests. The redistribution of resources among members of the same extended family (dynasty) is thus neutral. But in the Bernheim– Bagwell world, everyone is a member of the same family, and so all transfers turn out to be neutral. If the government takes $100 from Ms. A in Connecticut and gives it to Ms. B in California, then Ms. B will increase her bequest by $100 and Ms. A will decrease her bequest by $100, because they both know that they will have descendants in common. Bernheim and Bagwell conclude that this result is patently absurd and that the Barro argument of intergenerational altruism is correspondingly suspect. (Greg Mankiw, when discussing this paper at a conference, said words to the effect that of, “I would have been happy to have written this paper. But if its argument is right, I should be just as happy that they wrote it.”) CASE STUDY EXTENSION 19-9 Does Altruism Matter? Much of the debate over Ricardian equivalence considers the intergenerational altruism argument of Robert Barro. Even if current deficits will ultimately be paid for by taxes on future generations, the effects of these deficits may be limited by parents’ taking account of their children’s tax liabilities and leaving larger bequests. James Poterba and Lawrence Summers suggest that this aspect of the debate may be something of a red herring, at least for evaluating the effects of deficits on consumption and saving. Even though tax burdens may be shifted onto future generations so that the wealth of those currently alive is increased by deficits, life-cycle consumers have a small marginal propensity to consume out of wealth. Poterba and Summers calculated the size of this effect by considering the behavior of an economy peopled by life-cycle consumers who work for 45 years and are retired for 10. They then considered the effects of a $1 transfer to all living people, working or retired, financed by a government deficit (under an assumption that, after some period of time, the government would then tax workers to pay the interest on the debt). Poterba and Summers looked at the effects of this policy on consumption under different assumptions about population growth and the real interest rate. They found, for example, that a one-year deficit would increase consumption by only $0.06 per dollar, assuming a 3 percent real interest rate and population growth at 1 percent. A five-year deficit increases consumption by $0.22. Thus, the short-run effects on consumption and national saving are small. Poterba and Summers summarize their results as follows4: Although deficit policies like those the United States has historically followed shift substantial tax burdens to future generations, we find that this effect is not large enough to significantly alter national saving. The extent and nature of bequest motives are therefore of secondary importance for judging whether deficits have short-run crowding out effects…. The short-run saving effects are, however, sensitive to the presence of liquidity constraints, consumer myopia, or other factors that may cause consumption to respond to changes in disposable income by more than the lifecycle hypothesis would predict. LECTURE SUPPLEMENT 19-10 Unpleasant Monetarist Arithmetic Thomas Sargent and Neil Wallace emphasized the interdependence of monetary and fiscal policy that is implied by the government budget constraints: Dt = ∆Mt + ∆Bt; Bt = Bt–1 + ∆Bt. They note that there is a limit on the stock of (real, per-capita) debt that the private sector is willing to hold. As a consequence, fiscal policy and monetary policy will not be independent in the long run. Suppose that the fiscal authorities establish a policy setting out all current and future deficits. Since there is a limit to the amount of debt that the public will hold, it may eventually not be possible to finance deficits with new bonds. Instead, the monetary authorities will then be forced to finance the deficits by issuing new money. Now suppose that, in an environment such as this, the Fed decides to tighten money growth in the present. This then means that it will have to let money grow faster in the future, so tighter money now means more inflation later. Further, with rational expectations, the public will anticipate this higher growth rate in the future, and so this can feed back into the current rate of inflation.2 Sargent and Wallace then show that tighter money today can lead to higher, not lower, inflation today. Inflation becomes always and everywhere a fiscal phenomenon. This model gives us some instructive insight into the interdependence of monetary and fiscal policy but should probably not be taken too literally (although some economists argue that the Volcker deflation was hindered by these sorts of effects). In reality, the monetary and fiscal authorities play a complicated dynamic game, and it is hard to figure out who ends up having primacy and, hence, whether monetary or fiscal policy is set residually. CASE STUDY EXTENSION 19-11 Inflation Indexed Bonds and Expected Inflation One way to measure expected inflation, as the text points out, is to look at the difference in yields on nominal and real bonds. Figure 1 does this using the yields on ten-year nominal Treasury securities and ten-year Treasury Inflation Protection Securities (TIPS) over the period 2003-2014. According to these data, expected inflation fluctuated in the range of 2 percent to 2.5 percent from 2003 through mid-2008. With the intensification of the economic crisis in late 2008, expected inflation then plummeted to below 0.5 percent by December. In late 2009, the expected inflation rate had moved back up to about 2 percent. One should not, however, place too much emphasis on the inflation expectations number as measured by the difference in these yields on government securities. One factor may lead the difference in yields to overstate expected inflation and two other factors may lead it to understate expected inflation. First, because investors holding nominal bonds face the risk that actual inflation will turn out to be different than expected, the yield on nominal bonds incorporates a risk premium that compensates for this uncertainty. Thus, the yield on nominal bonds will exceed the yield on real bonds by the sum of expected inflation and this risk premium. The difference in yields will therefore tend to overstate expected inflation. Second, the market for TIPS is not as liquid as the market for standard Treasury securities. If investors are concerned about the ability to easily resell TIPS, they will demand a higher yield than if the market for TIPS were more liquid. The difference in yields will tend to understate expected inflation. Finally, the inflation adjustment in the principal of the TIPS is taxed in the year it accrues, rather than at maturity. For example, suppose you buy a $1,000 ten-year TIPS with an interest rate of 3.5 percent. In the first year you will earn $35 in interest. If inflation is 3 percent, the principal will increase by $30 to $1,030. You will owe tax on both the $35 interest payment and the $30 adjustment in principal, even though you continue to hold the bond and don’t receive the $30 until you sell it or it matures. Investors may demand a higher yield on TIPS to compensate for this accelerated taxation of principal, so the difference in yields will tend to understate expected inflation. Depending on the extent to which these factors matter, expected inflation measured as the difference in these yields might be overstated or understated. If we assume, however, that the relative importance of these factors does not shift much from year to year (i.e., inflation uncertainty is the same, the market for TIPS remains relatively small, and the tax treatment is unchanged), then we can at least use the difference in yields to illustrate the trend in expected inflation. In this case, the data, as shown in Figure 1, clearly indicate that expected inflation was steady over much of the 2000s before falling sharply in late 2008 and then returning during 2010 and 2011 to its earlier level. Source: Board of Governors of the Federal Reserve System. Note: Interest rates are market yields on Treasury securities adjusted to a ten-year constant maturity. LECTURE SUPPLEMENT 19-12 Additional Readings The Spring 1989 issue of the Journal of Economic Perspectives 3 contains a symposium on “Budget Deficits.” Two papers in particular address Ricardian equivalence: Robert Barro, “The Ricardian Approach to Budget Deficits,” pp. 37–55; and Douglas Bernheim, “A Neoclassical Perspective on Budget Deficits,” pp. 55–72,. Bernheim also discusses Ricardian equivalence in D. Bernheim, “Ricardian Equivalence: An Evaluation of Theory and Evidence,” in S. Fischer (ed.), NBER Macroeconomics Annual, 1987 (Cambridge, Mass.: MIT Press), pp. 263–316. A symposium on “Explaining Savings” in the Spring 1988 issue of the Journal of Economic Perspectives contains papers by Franco Modigliani and Lawrence Kotlikoff on saving and bequests. James Tobin explains his skepticism about Ricardian equivalence in J. Tobin, Asset Accumulation and Economic Activity (Oxford: Basil Blackwell, 1980), Chapter 3. The Economist has a “Schools Brief” on Ricardian equivalence in its issue of November 24, 1990: 77–8. Instructor Manual for Macroeconomics Gregory N. Mankiw 9781464182891, 9781319106058