How to Combat Recession; Stimulus without Debt - Laurence Seidman - 2018 by ElectricStorm

HOW TO COMBAT RECES SION

How to Combat Recession Stimulus without Debt Laurence Seidman

1 Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and certain other countries. Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America. © Oxford University Press 2018 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by license, or under terms agreed with the appropriate reproduction rights organization. Inquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above. You must not circulate this work in any other form and you must impose this same condition on any acquirer. Library of Congress Cataloging-in-Publication Data Names: Seidman, Laurence S., author. Title: How to combat recession : stimulus without debt / Laurence Seidman. Description: New York, NY : Oxford University Press, [2018] Identifiers: LCCN 2017053675| ISBN 9780190462178 (hardcover : alk. paper) | ISBN 9780190462192 (epub) Subjects: LCSH: Recessions—United States. | Monetary policy—United States. Classification: LCC HB3743 .S45 2018 | DDC 339.5/20973—dc23 LC record available at https://lccn.loc.gov/2017053675 1 3 5 7 9 8 6 4 2 Printed by Sheridan Books, Inc., United States of America

CONTENTS

1. Introduction | 1 2. How Would a Benevolent Ruler Combat a Recession? | 7 3. What Is Stimulus without Debt? | 17 4. Do Tax Rebates Work in a Recession? | 58 5. What about Other Kinds of Fiscal Stimulus? | 97 6. Would Stimulus without Debt Be Inflationary? | 127 7. Would Stimulus without Debt Weaken the Fed’s Balance Sheet? | 137 8. Would Stimulus without Debt Undermine the Fed’s Independence? | 150 9. Can’t Monetary Stimulus Overcome a Severe Recession? | 157 10. Can Stimulus without Debt Be Used by Other Countries? | 173 11. What Have Others Written about Stimulus without Debt? | 179 v

Contents

227

221

HOW TO COMBAT RECES SION

Chapter 1

Introduction

Are we ready to combat the next severe recession? We can be, but we’re not. A severe recession always involves a plunge in aggregate demand for goods and services that compels producers to sharply cut back production and employment. To recover from the recession, aggregate demand must be boosted all the way back up to normal. Economic analysis and experience shows that waiting for the free market to reverse the plunge in aggregate demand takes much too long—usually half a decade to a full decade. Monetary stimulus—cutting interest rates to zero—is much too weak to induce a huge boost in demand because experience shows that sensible consumers and business managers aren’t willing to go deeper in debt by borrowing to spend in a severe recession. Fortunately, a huge boost in demand can be achieved by a large fiscal stimulus—a temporary large increase in tax rebates for households, federal grants to state and local governments, tax credits to partly reimburse firms for purchases of capital goods, and government spending on infrastructure maintenance projects.

H ow to C ombat R ecession

So why do I say we aren’t ready? Because a large fiscal stimulus has always in the past required large borrowing by the Treasury and therefore a large increase in government debt. Experience shows that any policy that requires a large increase in government debt is strongly opposed by many policymakers and citizens. This is especially true in a severe recession, because the recession itself causes a plunge in tax revenue that forces the Treasury into huge borrowing to avoid large cuts in government spending, thereby sharply increasing government debt, which alarms policymakers and citizens. It is hardly surprising, then, that in a severe recession a proposal for a large fiscal stimulus that requires even more borrowing would be met with intense opposition. That, of course, is exactly what happened during the Great Recession in the United States, when the fiscal stimulus proposed in early 2009 met stiff resistance. Yes, despite opposition, a two-year fiscal stimulus was enacted in early 2009 and was large by historical standards. But simple calculations at the time showed that the fiscal stimulus needed to be at least twice as large in 2009 and 2010 to overcome this severe recession, and later calculations showed that it needed to be three times as large. Yet even most advocates of fiscal stimulus didn’t dare propose a larger fiscal stimulus because even they worried about making the increase in government debt even larger. Despite the continuing weak recovery, after 2010 fiscal stimulus was made much smaller, not larger, because of worry about government debt. So it took until 2016 for the unemployment rate to return to normal. The lesson, therefore, is sobering: As long as a large fiscal stimulus requires a large increase in government debt, Congress won’t make it large enough to successfully combat a severe recession. The one policy—large fiscal stimulus—that has the capacity to overcome a severe recession won’t be used to its full potential strength. So we are indeed not ready to combat the next severe

Introduction

recession as long as it assumed that fiscal stimulus must increase government debt. The stimulus-without-debt proposal, however, is not simply a tactic for getting a large fiscal stimulus enacted in a severe recession. It is certainly better for stimulus to be implemented without a large increase in government debt. Large government debt— that is, government debt that is a large percentage of GDP—may generate negative economic consequences and risks in the future. If government debt becomes a high percentage of GDP, the government may incur a heavy interest burden if interest rates rise, forcing cuts in worthwhile government programs or tax increases. Moreover, as I will explain later, there is a possibility of an anxious reaction by financial investors around the world to US government debt that is high and rising as a percentage of GDP, which may lead to a US recession or a financial crisis. Thus, I have two reasons for proposing stimulus without debt. The first is political: a large fiscal stimulus is unlikely to be enacted by Congress if it causes a large increase in government debt as a percentage of GDP. The second is economic: government debt that is large and rising as a percentage of GDP may have negative economic consequences and risks. Fortunately, the assumption that fiscal stimulus requires an increase in government debt is false. In fact, it is astonishingly easy to implement even a very large fiscal stimulus without any increase in government debt. All it takes is this: When Congress enacts fiscal stimulus, the Federal Reserve can decide to make a transfer (not loan) to the Treasury roughly equal to the fiscal stimulus so the Treasury doesn’t have to borrow. That’s it. Moreover, the large stimulus would be phased out as the economy approaches full employment, so it would not be inflationary. But isn’t the paper money injected by the Fed also government debt? The answer is no. Paper money is not counted in official government debt. Nor should it be. Government paper money

H ow to C ombat R ecession

was government debt in the era when the government promised gold to any holder of government paper money who requested it. But when the government nearly a century ago removed its promise to provide gold or anything else to the holders of paper money, the paper money ceased to be government debt. By contrast, government bonds are government debt because the government promises to pay government paper money—principal plus interest—on schedule to all holders of government bonds. Another question arises: Does fiscal stimulus really work? Most empirical studies find that the answer is yes. Moreover, just think about it using common sense. Suppose Congress decides, as it should, that the main component of the fiscal stimulus package will be tax rebates to each household. Suppose that the US Treasury mails out two rebate checks to each household— one in June, one in December—each check for $6,000. Is there anyone who seriously believes that households, in a severe recession when most employees haven’t received a raise and some have been laid off, would save the $12,000? In a recession, doesn’t it seem more likely that hard-pressed households would spend a substantial portion within six months? The best empirical studies support common sense: in a recession households do indeed spend about two thirds of their tax rebates within six months. Similarly, does anyone seriously think that in a severe recession, when state and local tax revenues plunge, that cash grants from the federal government would be saved instead of used to maintain normal state and local government expenditures? If you managed these governments during a recession, would you really save the grants, and then do lots of borrowing or slashing of expenditures? Most empirical studies of state and local government behavior in a recession support common sense: federal grants are mainly used to keep state and local governments from

Introduction

cutting spending or raising taxes, so federal grants prevent a fall in spending by state and local governments and a fall in spending by consumers who would not spend as much if their state and local taxes were raised. So the grants prevent a fall in aggregate demand for goods and services. In this book I explain how a temporary large fiscal stimulus can be implemented without any increase in government debt or inflation. I also present analysis and evidence that fiscal stimulus works in a recession—it increases aggregate demand for goods and services, which in turn leads to an increase in production and employment. Stimulus without debt isn’t the only thing that must be done when a severe recession hits. The Fed, Treasury, and FDIC must perform financial rescues of key firms and inject funds into financial firms to keep credit from freezing up. These essential interventions are not addressed in this book. Some analysts believe that these interventions are all that’s needed in a severe recession. I strongly disagree. A severe recession always involves a plunge in aggregate demand for goods and services, and once that plunge occurs, it will not be reversed simply by rescuing key firms and restoring the flow of credit. Several things are needed to make us truly ready for the next severe recession. First, a lot of economists, policymakers, members of Congress, financial market participants, and others must learn that it is possible to implement a large fiscal stimulus without any increase in government debt. Second, they have to be persuaded that fiscal stimulus—particularly, a tax rebate to every household—works. Third, they have to be convinced that a large transfer from the Federal Reserve to the Treasury during recession won’t be inflationary. Fourth, Congress must enact an amendment to the Federal Reserve Act empowering the Federal Reserve’s Open Market Committee to decide whether to make a

H ow to C ombat R ecession

large transfer (not a loan) to the Treasury to finance a fiscal stimulus enacted by Congress. If these things happen, we will be truly ready to combat the next severe recession. The purpose of this book is to help make all these things happen.

Chapter 2

How Would a Benevolent Ruler Combat a Recession?

In this book I will propose “stimulus without debt,” a policy to combat recession that is designed for our actual institutions: Congress, the Treasury, and the Federal Reserve. These institutions impose important constraints on the design of a practical policy. But before I turn to stimulus without debt for our actual institutions, I want to set the stage by considering how a benevolent ruler with complete power, who takes the place of the Federal Reserve, Congress, and the Treasury, could combat a recession without increasing government debt. The policy that is implemented by a benevolent ruler will serve as useful guide for a stimulus without debt policy that is implemented by the Federal Reserve, Congress, and the Treasury.

Government Money Held by the Public Is Not Government Debt Before I turn to the benevolent ruler, I need to make a fundamental point: government paper money held by the public is 7

H ow to C ombat R ecession

not government debt. When government paper money was introduced during the past three centuries, it was usually “backed” by gold (or silver)—that is, if any holder of paper money wanted the government to exchange it for gold, the government promised to make the exchange. Thus, the government owed gold to any holder of paper money, and the government had to be ready to provide gold to any holder of paper money who requested it. Because of this promise, government paper money held by the public was viewed as government debt and included in the liability column of the government’s balance sheet. This promise to pay gold was probably necessary to win the acceptance of government paper money by the public. But in the last century, the public in most economically advanced countries gradually gained confidence in government paper money. The governments of these countries gradually withdrew their promise to provide gold to any holder of paper money who requested it. Despite the withdrawal of this promise, most of the public continued to be willing to hold paper money and use it in transactions. One reason for this public willingness was that the government guaranteed that the public could use its paper money to pay taxes; another was that the government stated that its paper money could be used by the public to pay off private debts. But it is possible that even without these guarantees by the government, the public would have been willing to hold and use government paper money because of confidence gained over decades of use for transactions. Thus, it is now the case that in most economically advanced countries, the government does not owe anything to holders of its paper money. Paper money held by the public is therefore no longer government debt. By contrast, each government bond held by the public is government debt because the government promises to pay its paper money to holders of bonds according to the schedule of interest and principal on the bonds;

How Would a Benevolent Ruler Combat a Recession?

the government therefore owes paper money to the holders of government bonds. Nevertheless, current central-bank accounting has ignored the fundamental change that occurred when the government withdrew its promise to pay gold (or silver) to any holder of its paper money. Paper money held by the public continues to be listed in the liability column of the central bank’s balance sheet. Moreover, when the accountants produce the consolidated balance sheet of the government and the central bank, paper money held by the public continues to be included in the liability column of the consolidated balance sheet. This inclusion causes the liabilities of the central bank, and of the consolidated government, to be greatly overstated. By contrast, official government debt correctly includes government bonds held by the public and correctly excludes paper money held by the public. The official government debt correctly focuses attention on the government’s obligation to pay money to bondholders and ignores paper money held by the public because the government has no obligation to pay anything to the holders of paper money. Some economists, however, not just central-bank accountants, continue to call government paper money held by the public “government debt,” usually without giving any justification for using the term “debt.” Why do they do this? One reason may be inertia: government paper money was indeed government debt when the government promised to pay gold to anyone holding the paper money who wanted gold. Another reason may be an intuition that there can’t be a “free lunch.” It seems like a free lunch when the government writes checks to members of the public and prints enough paper money to pay check recipients who request paper money. Surely it must be true, some think, that the government is incurring a debt when it prints pieces of paper to give to the public; it can’t really be that easy for the government

H ow to C ombat R ecession

to create an asset, paper money, without also creating a debt. But in a paper money system, it really is that easy. Although paper money held by the public is not government debt, injecting too much paper money into an economy that is already at full employment of resources will make aggregate demand for goods and services exceed potential output and therefore generate rising prices—inflation. Injection of money into the economy should not cause concern about government debt, but should cause concern about inflation if the economy is already at full employment. The policy I will call “stimulus without debt” prescribes injecting money only in a recession when employment is below its potential, so that, as I’ll explain later, an increase in demand for goods and services will cause an increase in output, not an increase in prices.

Stimulus without debt in the Benevolent Ruler’s Economy In the benevolent ruler’s economy, money consists of official paper notes, and all transactions in this economy occur in official government paper notes. These paper notes were once backed by gold, but are no longer backed by gold or anything else, so they are not government debt. Assume the benevolent ruler’s economy is initially at full employment. Now suppose a recession occurs because of a fall in aggregate demand for goods and services. There are several possible causes of a fall in aggregate demand. Consider a fall caused by a dramatic and sustained plunge in the stock market. In response, anxious consumers with less stock market wealth cut their spending, so consumer demand for goods and services falls. Producers of these goods and services respond by cutting production and employment. Managers in firms making consumer goods or providing consumer services react by cutting

How Would a Benevolent Ruler Combat a Recession?

their demand for equipment to produce more consumer goods or services, so producers of equipment—investment goods— cut production and employment. Thus, in response to the fall in consumption and investment demand, most firms cut back production and employment, so the economy falls into recession. To combat a recession caused by a fall in aggregate demand for goods and services, a policy must be implemented that will increase aggregate demand. To increase aggregate demand for goods and services, the ruler deposits a specific amount of paper notes in the bank account of each household. The deposit is a transfer from the government to the household, not a loan that the household must repay. The ruler calls the transfer to each household a “tax rebate” because it gives back some of the tax that the household paid in the previous year. The tax rebates are called “fiscal stimulus” because they are a government expenditure (“fiscal”) that increases (“stimulates”) consumer demand for goods and services. Households spend a portion of their tax rebates and save the rest, and the portion they spend causes producers of consumer goods and services to increase their production and employment. As managers in firms making consumer goods or services observe the revival of consumer demand, they spend more to increase their equipment, so producers of these investment goods raise their production and employment. But how does the benevolent ruler obtain the paper notes needed to give tax rebates to households? Assume that at the beginning of the year the ruler has no notes on hand. The ruler’s adviser points out that when other governments have faced a similar situation, some of these governments have borrowed from the public by selling government bonds to the public to obtain the paper notes. But the ruler replies that there is no need to sell bonds and thereby incur government debt. Instead, the ruler

H ow to C ombat R ecession

simply orders the printing of the amount of paper notes required to give the tax rebates. As explained in the previous section, paper money issued by the government and held by the public is not government debt. Thus, to combat the recession, the benevolent ruler implements “stimulus without debt.” The ruler keeps paying out tax rebates with new paper money until the economy approaches full-employment output. As this happens, consumer and business confidence gradually rises, enabling the ruler to gradually phase out the tax rebates. The ruler’s adviser concedes that the ruler’s policy (stimulus without debt) has worked, but points out that when other governments have faced a similar situation, some have printed money and used the money not for tax rebates but to buy government bonds from the public in “the open market.” These are bonds the public previously bought from the government but which some members of the public now want to sell. These governments have called this bond buying “open-market operations” and asked their central bank to carry it out. The ruler replies that using new money to buy government bonds in the “open market” is a much less efficient to way to increase aggregate demand for goods and services than using new money to give tax rebates to households. The reason, says the ruler, is that in a recession households will very likely spend a larger share of the tax rebates they receive than bond sellers would spend. Why? When a household receives a tax rebate, the household knows it can spend some of it, save some of it, and use some of it to pay down debt; the household instinctively realizes that the tax rebate has increased its wealth, enabling it to do more of all three. By contrast, when someone sells a government bond, the seller instinctively realizes that his wealth hasn’t changed: the seller knows he now has more cash, but he also knows he no longer has the government bond. Why did he sell the government bond? Although his wealth is the same, he

How Would a Benevolent Ruler Combat a Recession?

may have wanted to replenish his declining checking or savings account, or buy a corporate bond, or a corporate stock, or pay taxes, or pay down debt, or buy goods or services, and he was willing to give up the bond to do it. The typical bond seller has much higher wealth and income than the typical tax-rebate recipient. It seems likely that money used to send tax rebates to households will increase aggregate demand for goods and services much more than if that same money were used to buy government bonds from bondholders willing to sell some bonds. After full-employment output is achieved by the stimulus- without-debt policy, confidence has returned to normal, and the tax rebates to households have been phased out, there will be more money in the economy than before due to the tax rebates. If the ruler thinks this extra money might cause too much spending and therefore inflation, the ruler can remove it from the economy by temporarily cutting government spending so it is less than tax revenue, or by temporarily raising tax revenue so it is greater than government spending. Either action results in more money coming into the government than the government spends. The government can remove this surplus money from the economy until money in the economy is back to normal. Then government spending can be set equal to tax revenue.

Whose Writing Guided the Benevolent Ruler? When the benevolent ruler was asked whose writing was most influential, the ruler replied that the greatest influence came from two economists: John Maynard Keynes and Abba Lerner. The ruler said Keynes (1936) taught the crucial importance of aggregate demand for goods and services: if aggregate demand falls, it causes the economy to go into recession, so demand must be raised. Keynes warned that in a recession monetary

H ow to C ombat R ecession

stimulus—lowering interest rates—would prove too weak to raise demand up to normal. He therefore recommended that the government increase its spending. The ruler said that Lerner specifically recommended printing money to pay for fiscal stimulus (an increase in government transfers to households, an increase in government purchases of goods and services, and/or a decrease in taxes on households) to combat a recession. Lerner said that government should practice “functional finance,” not “sound finance,” and explained why in his “functional finance” chapter in each of his two books (Lerner 1944, 1951). The title of his 1944 book is The Economics of Control, and the title of his 1951 book is The Economics of Employment. The ruler also cited chapter 1 of his 1951 book, entitled “The Economic Steering Wheel,” as particularly clever and insightful. Lerner wrote that if the unemployment rate is above normal, the government should decrease taxes so households spend more, or increase its own spending and pay for the excess of spending over taxes by printing money instead of borrowing. But won’t printing money be inflationary? Lerner said it would indeed be inflationary if it were done when there is already full employment. Why? Because with full employment each firm can attract employed workers away from other firms only by raising wages, which increases costs and compels firms to raise prices. But if it were done when unemployment is high, firms would be able to attract unemployed workers without raising wages, so there would be no cost increases, and no need for firms to raise prices.

Lessons for Combating Recession with Actual Institutions It should be possible to implement stimulus without debt under our actual institutions because the benevolent ruler was able to

How Would a Benevolent Ruler Combat a Recession?

do it. Congress and the president should control government spending, taxes, and fiscal stimulus, with the Treasury as their administrative agent. The Federal Reserve should control the printing of money and its injection into or withdrawal from the economy. With this institutional separation of powers, how would stimulus without debt be implemented in a recession? The Federal Reserve would decide whether to give a transfer (not loan) to the Treasury to be used for fiscal stimulus, and if so, how much. The Fed would make its decision by estimating the depth of the recession and the magnitude of the fiscal stimulus needed to combat it. If all transactions were conducted using Federal Reserve notes (not by writing checks or crediting bank accounts), the Fed would print new Federal Reserve notes in the amount it wanted to transfer to the Treasury. In practice, the Fed would either write a check to the Treasury or credit the Treasury’s checking account at the Fed, and print the amount of new Fed notes needed to meet requests for Fed notes from banks and the public. The Fed would continue to use its standard instruments of monetary policy such as sales or purchases of government bonds in the open market, and decide how to adjust these sales and purchases in light of the Fed’s transfer to the Treasury for fiscal stimulus. Congress and the president would decide how much fiscal stimulus to enact in light of the magnitude of the transfer the Fed was willing to give to the Treasury for fiscal stimulus. They could enact a larger fiscal stimulus than the Fed’s transfer to the Treasury, but the Treasury would have to borrow the difference. If they enacted a smaller fiscal stimulus than the Fed’s transfer to the Treasury for fiscal stimulus, some of the Fed’s transfer would go unused and be returned to the Fed. Congress and the president would decide the composition as well as the size of the fiscal stimulus. As I will explain shortly, I recommend that a large portion of the fiscal stimulus be tax rebates to households.

H ow to C ombat R ecession

Conclusion The most important point is this: fiscal stimulus does not require an increase in government debt. To get high unemployment down to normal, the government should implement fiscal stimulus: increase its spending (mainly tax rebates to households, but also some purchases of goods and services and other expenditures) and/or cut taxes. When the government does this, it doesn’t need to borrow. It can get the money it needs from its printing press. As long as it does this only when unemployment is high, it will not be inflationary. Thus, when unemployment is high, fiscal stimulus can be implemented without debt and without causing inflation.

Chapter 3

What Is Stimulus without Debt?

“Stimulus without debt” is a policy that would increase aggregate demand for goods and services in a recession without increasing government debt. Stimulus without debt consists of a transfer (not loan) from the central bank to the nation’s treasury so that the treasury does not have to borrow to finance fiscal stimulus enacted by the legislature. In most of this book I illustrate stimulus without debt with reference to the United States, but stimulus without debt can be implemented in other countries and in the eurozone.

The Strategy behind Stimulus without Debt Most recessions, including the US Great Recession of 2008, involve a fall in demand for goods and services. When the US housing bubble burst in 2007, followed by the plunge in the stock market and the failure of firms like Lehman Brothers in 2008, consumer wealth and confidence fell sharply. Anxious consumers cut back their spending, so consumer demand for goods and 17

H ow to C ombat R ecession

services fell. Producers of consumer goods and services had no choice but to cut back production and lay off workers. Business firms reacted to this fall in consumer demand by cutting their demand for investment goods—why expand plant and equipment to produce more when consumers won’t buy more? In response to the fall in consumption and investment demand, firms cut back production and employment. To combat a severe recession involving a sharp fall in aggregate demand for goods and services, fiscal stimulus (an increase in government transfers to households, an increase in government purchases or goods and services, and/or a decrease in taxes on households) must be enacted to increase aggregate demand for goods and services because monetary stimulus—lowering interest rates—alone is too weak to combat a severe recession (Seidman 2001, 2003, 2011, 2012a, 2012b). Under stimulus without debt, Congress would enact a fiscal stimulus package that consists mainly of cash transfers (tax rebates) to households but also other temporary expenditures and temporary tax cuts; the fiscal stimulus would raise aggregate demand. The Federal Reserve would use new money to give a large transfer (not loan) to the Treasury equal to the fiscal stimulus package so that the Treasury would not have to borrow to pay for the package. Hence, there would be no increase in government debt (Seidman 2013). The Fed’s transfer to the Treasury injects an amount of money into the economy that is equal to the Fed’s transfer. By contrast, fiscal stimulus alone would require the Treasury to borrow an amount equal to the fiscal stimulus by selling new government bonds, thereby significantly increasing government debt. Stimulus without debt differs crucially from a standard fiscal- monetary stimulus. Under a standard fiscal-monetary stimulus, the Treasury borrows to pay for the fiscal stimulus by selling new government bonds to the public, and the Federal Reserve

W hat Is Stimulus without Debt?

enters the “open market” and buys an equal amount of government bonds from the public. Government debt increases by the amount of the fiscal stimulus, and the Fed increases its holding of Treasury bonds by the same amount. A crucial point is that the Fed’s action does not prevent the increase in Treasury debt outstanding: official Treasury debt increases by an amount equal to the fiscal stimulus. Money injected into the economy is the same under stimulus without debt and standard fiscal-monetary stimulus, but government debt is greater under standard fiscal- monetary stimulus. Official government debt includes all Treasury bonds outstanding whether held by the public or by the Fed. But should it include Treasury bonds held by the Fed? My answer is yes. It is true that the Fed returns to the Treasury most of the interest it receives from the Treasury. But at any moment, the Fed can decide to sell Treasury bonds to the public, and the public will expect principal and interest to be paid on schedule. Thus, the Treasury must be ready to pay principal and interest on schedule on all Treasury bonds outstanding, including bonds held by the Fed. Congress, the Treasury, and the citizenry are therefore correct to focus on the official government debt figure that includes bonds held by the Fed; and they would be correct to oppose excluding bonds held by the Fed from official government debt. Thus, the difference between stimulus without debt and standard fiscal- monetary stimulus is genuine both in theory and in practice. Government paper money held by the public is not government debt. Government paper money was government debt a century ago when the government promised gold to any holder of government paper money who requested it. But when the government removed its promise to provide gold or anything else to the holders of government paper moneypaper money ceased to be government debt because the government owes nothing to holders of its paper money. By contrast, government

H ow to C ombat R ecession

bonds are government debt because the government promises to pay government paper money—principal plus interest—on schedule to all holders of government bonds; the government owes government paper money to holders of government bonds. Fiscal stimulus without debt, therefore, means fiscal stimulus without an increase in government bonds. I have two important reasons for proposing fiscal stimulus without debt instead of the standard fiscal stimulus with debt. The first reason is political and the second reason is economic. My political reason comes from facing the fact that many policymakers, financial investors, citizens, and economists believe that large government debt will generate negative economic consequences and risks in the future, so they oppose enactment of a large fiscal stimulus in a severe recession if it causes a large increase in government debt. Without the stimulus-without- debt plan proposed in this book, a fiscal stimulus large enough to combat a severe recession would cause a large increase in government debt—and it is therefore unlikely to be enacted by Congress. By contrast, with the stimulus plan proposed here, even a large fiscal stimulus would cause no increase in government debt. The stimulus-without-debt plan would therefore remove one key obstacle to obtaining the support of policymakers and the general public for a large fiscal stimulus in a severe recession. My economic reason is that large government debt—that is, government debt that is a large percentage of GDP—may generate negative economic consequences and risks in the future. Although large government debt may not lead to this negative scenario, it is a risk that is worth avoiding: (1) If interest rates rise in the future, large debt will result in large interest payments; (2) large interest payments will force Congress to raise taxes or cut spending or borrow more; (3) more government borrowing and still larger government debt may at some point make financial investors anxious and cause them to sell corporate stocks,

W hat Is Stimulus without Debt?

resulting in a fall in the stock market; (4) a fall in the stock market would reduce wealth and confidence and cause consumers and businesses to cut spending, precipitating a recession; (5) the fall in the stock market and the recession would generate still more anxiety among financial investors and managers, generating a financial crisis, worsening the recession. Thus, the public and Congress are right to be concerned about the possible negative economic consequences of letting government debt become a large percentage of GDP. The phrase “stimulus without debt” also means “without private debt.” Standard monetary stimulus reduces interest rates in order to induce households and/or businesses to borrow more in order to spend more on goods and services, so standard monetary stimulus works by inducing households and firms to incur more private debt in order to spend more on goods and services. By contrast, fiscal stimulus in the form of tax rebates (cash transfers) to households enables recipients to spend more without increasing their debt; empirical studies that I will review later show that households spend a significant portion of their tax rebate and use the remaining portion to pay down debt and save. Under the stimulus-without-debt policy, the Federal Reserve would transfer (not lend) X dollars to the Treasury. In turn, the Treasury, after authorization by Congress, would mail out a large portion of the X dollars as “tax rebate” checks (cash transfers) to households (a small portion of the X dollars would be spent on other kinds of fiscal stimulus such as temporary government purchases of goods and services or temporary tax cuts). Thus, the Treasury would not have to borrow to finance X dollars of fiscal stimulus because of the Fed’s transfer (not loan) of X dollars to the Treasury. Moreover, households would spend some or most of their tax rebate without doing any borrowing. Thus, this policy would increase aggregate demand for goods and services without increasing government debt or private debt.

H ow to C ombat R ecession

The stimulus-without-debt plan is designed to maintain a separation of powers and checks and balances: both Congress and the Federal Reserve play crucial independent roles. Congress sets the size and composition of the fiscal stimulus package. The Federal Reserve decides the size of the transfer (not loan) it will make to the Treasury; the Fed therefore sets the amount of fiscal stimulus that can be implemented without increasing government debt, but Congress is free to enact a fiscal stimulus that is larger or smaller than the Fed’s transfer to the Treasury. The plan for stimulus without debt in a recession has five elements:

1. If the Federal Reserve judges that the GDP of the economy is significantly below the Fed’s estimate of potential GDP, the Fed would decide whether to give a transfer (not a loan) to the Treasury, and if so, how much, on the condition that it be used only for fiscal stimulus. Authority for the Fed to make this transfer might require an amendment to the Federal Reserve Act. The Fed could implement its transfer by writing a check to the Treasury or crediting the Treasury’s checking account at the Fed. 2. Congress would decide whether to enact fiscal stimulus, and if so, how much; its main component would be tax rebates to households, though other components should also be included in the fiscal stimulus package. The Treasury would mail tax rebate checks to households in amounts specified by Congress. 3. If the amount Congress enacts for fiscal stimulus is no greater than the Fed’s transfer, then the Treasury would not have to borrow to finance the fiscal stimulus; if the fiscal stimulus is greater than the Fed’s transfer, the Treasury would have to borrow the difference.

W hat Is Stimulus without Debt?

4. The Fed would decide how much to adjust its bond purchases or sales to try to keep employment high and inflation low. It is very likely that the Fed, having injected money into the economy through its transfer to the Treasury, would decide either to inject less money into the economy through bond purchases or to withdraw money from the economy through bond sales. 5. The Fed would order (from the Treasury’s Bureau of Engraving and Printing) an amount of new Federal Reserve notes equal to its transfer to the Treasury, and would store these notes in the Fed’s vault; this Fed vault cash would be an asset on the Fed’s balance sheet, and as a consequence of this new vault cash the Fed’s transfer to the Treasury would not reduce the Fed’s capital (net worth) on its balance sheet.

Would Stimulus without Debt Be Temporary or Permanent? In response to a recession caused by a shock like a plunge in housing prices or stock prices, stimulus without debt would be a temporary policy. As the economy recovers toward full employment and consumer and business confidence returns, stimulus without debt would be gradually phased out: Congress would slow the growth of government spending and/or raise taxes, thereby reducing fiscal stimulus; and the Fed would reduce the growth of money in the economy, thereby reducing monetary stimulus; the Fed would do this by buying a smaller amount of bonds from the public than it otherwise would have bought. When the economy plunges into recession, surveys confirm the obvious: consumer and business confidence and expectations for the economy fall sharply. The fiscal stimulus in stimulus

H ow to C ombat R ecession

without debt, primarily tax rebates to households, would initially be set large enough so that, even with low consumer and business confidence and expectations, spending on goods and services would increase enough to launch a strong recovery. As a solid recovery takes hold, surveys again confirm the obvious: consumer and business confidence and expectations start rising. This rise in confidence and expectations raises demand for goods and services. As confidence and expectations return to normal, stimulus can be gradually phased out. Some macroeconomists have ignored the role of confidence and expectations in determining demand for goods and services, and the role of the state of the economy in determining confidence and expectations. But many economists and laymen rightly regard the connection, confirmed by surveys and anecdotes, as obvious. As the economy moves from recession to a strong recovery generated by stimulus, confidence and expectations will rise in step with the economy, thereby enabling the gradual phasing out of the stimulus. There is one caveat. Suppose the economy hasn’t plunged into recession due to a shock like a plunge in housing prices or stock prices, but has instead experienced a gradual decline in demand relative to potential output—a “demand-induced secular stagnation.” If demand-induced secular stagnation is really a problem, could stimulus without debt treat it? Should it? With secular stagnation, would the treatment have to be permanent rather than temporary? I will discuss this at the end of this book.

Tax Rebates to Households The purpose of having the US Treasury mail a tax rebate check (a rebate of a portion of the income, payroll, and sales taxes paid by the household in the previous year) to every household

W hat Is Stimulus without Debt?

in a recession is to increase consumption spending and thereby stimulate production of goods and services. Strong evidence for a significant impact of tax rebates on consumption spending will be presented in depth in the next chapter. A tax rebate to households is one type of fiscal stimulus (tax cuts and/or government spending). A tax rebate is a giving back to each taxpayer a portion of the taxes (income, payroll, and sales) that the taxpayer paid in the previous year. Congress does this in a recession to help citizens cope with the recession and to boost their spending on goods and services, which will stimulate production and employment. Why focus on tax rebates to households rather than other kinds of fiscal stimulus? There are at least eight reasons. First, and most important, as I will document in chapter 4, tax rebates work: a significant portion of tax rebates are spent within half a year of being received. This should not be surprising. Both surveys of recipients and econometric analysis of spending data following the payment of tax rebates indicate that households spend a significant portion of it: one-third within three months and two-thirds within six months. Note: any recession almost always involves a fall in consumption below its trend growth path. Second, tax rebates clearly increase household spending on all goods and services rather just a subset of goods and services, so all businesses would recognize that rebates boost customer demand for their goods or services. Third, with tax rebates, there would be no “shovel-ready” problem that may occur with infrastructure projects; there is no limit to how much households can spend promptly on goods and services. Fourth, with tax rebates there would be no temporary or permanent increase in the size of government: rebates simply give spending power to millions of individual consumers, whose spending stimulates the private sector. Fifth, tax rebates to combat a recession are clearly temporary and require a new vote by Congress to be continued. Sixth, every

H ow to C ombat R ecession

household would receive a rebate check in the mail from the US Treasury, so every voter would actually see a concrete personal benefit from this kind of fiscal stimulus. Seventh, the inclusion of every household would cause most voters to regard tax rebates as a fair way to implement fiscal stimulus. Eighth, rebates have been enacted with bipartisan support three times—1975, 2001, and 2008. Thus, a tax rebate has many important advantages as an instrument for fiscal stimulus in a recession. There is a good reason why tax rebates were able to pass with bipartisan support three times. Conservatives and liberals, Republicans and Democrats, have differing long-term agendas for government spending and taxation. For example, conservatives generally want permanent tax cuts, while liberals want permanent increases in social insurance and education programs. Neither side wants an antirecession stimulus that would advance the agenda of the other side if it became permanent. Tax rebates are obviously and inherently temporary. They do not favor the long-term agenda of either side. This is undoubtedly one reason that tax rebates were enacted three times with bipartisan support, whereas most other proposals, such as permanent tax cuts, or permanent increases in social insurance or education programs, have generated partisan opposition. The tax rebate does not favor one side’s long-term agenda over the other. It is neutral toward long-term agendas. The details of the design of a tax rebate in recession will be discussed in the last section of chapter 4.

A Transfer from the Fed to the Treasury in a Recession The purpose of the Fed’s transfer to the Treasury in a recession is to enable the Treasury to pay tax rebates to households and undertake other fiscal stimulus without borrowing. The Fed’s

W hat Is Stimulus without Debt?

transfer to the Treasury is not a loan; the Fed would not receive Treasury bonds in return for its funds. If the Federal Open Market Committee (FOMC) judges that the total amount of tax rebates chosen by Congress is appropriate to the severity of the recession, the FOMC would decide to make its transfer roughly equal to the total amount of the tax rebates so that the rebates don’t increase the debt of the federal government. Later in this chapter, I will explain why the Fed should have transferred $450 billion to the Treasury every six months from June 2008 through December 2010, six transfers summing to $2,700 billion, and Congress should have authorized $450 billion of fiscal stimulus every six months from June 2008 through December 2010, so each year there would have been $900 billion of fiscal stimulus—6% of GDP ($15,000 billion). Most of the fiscal stimulus should have been tax rebates to households. The fiscal stimulus would not have required any new borrowing by the Treasury—it would not have required any sale of new Treasury bonds. Thanks to the Fed’s transfer to the Treasury, the fiscal stimulus would not have increased federal debt.

What the Federal Reserve Would Do under Stimulus without Debt It is important to distinguish between two roles for the Federal Reserve when the economy suffers a plunge in aggregate demand resulting in a severe recession: (1) acting aggressively as a lender of last resort to perform financial rescues and unfreeze credit markets; (2) cutting interest rates to try to induce households and firms to maintain borrowing and spending. The Fed should do both aggressively. The first role can succeed; the second role can help, but cutting interest rates is not nearly powerful enough to overcome a severe recession.

H ow to C ombat R ecession

The recession began in late 2007 in response to the plunge in housing prices but proceeded moderately through the first half of 2008 with below-normal growth in aggregate demand. In September 2008, however, in part due to the failure of Lehman Brothers, a financial crisis was triggered as fear swept over managers of financial institutions and investors that they might follow Lehman Brothers to a similar fate. Institutions and investors sought liquidity and sharply cut their lending. In a financial crisis, the Federal Reserve has a vital role to play as a lender of last resort, providing emergency loans to financial institutions and other firms in order to prevent the freezing up of credit markets and to keep funds flowing to business firms so that production can continue. The Fed and the Treasury aggressively engaged in financial rescues in the fall of 2008 and the winter of 2009. In my judgment, these actions were essential to preventing a full-scale great depression. The Fed must also make sure that interest rates are cut sharply to zero in order to give households and firms as much inducement as possible to keep borrowing and spending. The Fed also performed this task well in 2008. But a key question then becomes whether the Fed’s sharp cut in interest rates is enough to stimulate a sufficiently large increase in aggregate demand for goods and services in a severe recession. When the Federal Reserve wants to stimulate demand for goods and services, its standard method is to lower interest rates in the hope that consumers and businesses will respond by borrowing more in order to spend more on goods and services. The Fed buys US Treasury bonds from financial firms and other institutions. When the Fed writes checks to buy the bonds, it is injecting money into the economy. When the sellers of these bonds receive checks from the Fed, they deposit the checks in their banks. The banks try to induce households and firms to borrow by cutting interest rates. But in a severe recession, it is

W hat Is Stimulus without Debt?

doubtful that cutting interest rates can induce enough borrowing and spending to fully reverse a deep plunge in aggregate demand because in a deep recession people and businesses are pessimistic and reluctant to take on more debt. Under the stimulus-without-debt policy, the Fed injects money into the economy a new way. When the Fed transfers X dollars to the Treasury, it injects X dollars of high-powered money into the economy because the Treasury, after authorization by Congress, mails out most of the transferred X dollars as tax rebates to households and spends the rest. The Fed would decide how much to adjust its bond purchases or sales to try to keep employment high and inflation low; it is likely that the Fed, having injected money into the economy through its transfer to the Treasury, would inject less money into the economy through bond purchases or would withdraw money from the economy through bond sales. Under the stimulus-without-debt policy, the Fed would order X dollars of new Federal Reserve notes and store these notes in its vault. This vault cash would be an asset on the Fed’s balance sheet. As a consequence of this action, the stimulus without debt would have no effect on the Fed’s net worth (capital) on its balance sheet.

Money for Tax Rebates versus Money to Buy Government Bonds Now let me compare money for tax rebates versus money to buy government bonds. Under stimulus without debt, the Fed creates money and uses it to make a large transfer to the Treasury, which, at Congress’s request, uses the money to pay tax rebates to households. By contrast, under standard monetary stimulus, the Fed creates money and uses it to buy government bonds “in

H ow to C ombat R ecession

the open market” from members of the public who want to sell government bonds that they previously bought. Which of these two uses of new money created by the Fed is likely to cause a larger increase in aggregate demand for goods and services? A person who receives a tax rebate realizes she can spend more, save more, and pay down more debt. The rebate increases her wealth, enabling her to do all three. But when someone sells a government bond, the seller realizes her wealth is the same because the cash replaces her bond. The seller may have sold the bond to replenish her declining checking or savings account, or buy a corporate bond, or a corporate stock, or pay taxes, or pay down debt, or buy goods or services, and was willing to give up the bond to do it. The typical bond seller has much higher wealth and income than the typical rebate recipient. It seems likely that money used for tax rebates will increase aggregate demand for goods and services more than if that same money were used to buy government bonds. Thus, to increase aggregate demand, paying rebates is likely to be a more efficient use of money by the government than buying bonds.

Debt Worry Prevents Sufficient Stimulus in a Big Recession Influential economists have warned against enacting a very large fiscal stimulus to combat a severe recession because they worry about the rise in government debt that it will generate (Seidman 2011, 2012a, 2012b, 2013). I believe some economists have exaggerated the risks of the resulting rise in government debt. But I will now cite a few economists to show that a very large fiscal stimulus in a severe recession will be opposed by influential economists as long as it generates a very large increase in debt.

W hat Is Stimulus without Debt?

Olivier Blanchard, for many years a professor of economics at MIT and then chief economist of the International Monetary Fund during part of the Great Recession, supported some fiscal stimulus during the Great Recession but opposed a fiscal stimulus that would have been large enough to overcome that recession. He explained why in a section of his intermediate macroeconomics textbook entitled “High Debt, Default Risk, and Vicious Cycles” (2017). If financial investors start worrying about whether the government can fully pay principal plus interest on schedule for all its bonds held by the public (domestic and foreign), they will refuse to buy new government bonds unless the bonds offer high interest rates to compensate for default risk; but these high interest rates on government bonds will make it even harder for the government to fully pay principal plus interest on schedule. To avoid this scenario, Blanchard argues that each government must limit its fiscal stimulus in a recession so that its debt doesn’t get high enough to worry financial investors. Carmen Reinhart and Kenneth Rogoff, the authors of This Time Is Different: Eight Centuries of Financial Folly (2009), presented a paper (2010) at the American Economic Association’s annual conference in which they asserted that their empirical study of data from many countries over the past two centuries found that government debt greater than 90% of GDP on average reduces a country’s annual economic growth rate by one percentage point—for example, from 3% to 2%—in this example, a 33% reduction in the growth rate. Their paper received a lot of attention and publicity. They said the sharp rise in government debt as a percentage of GDP during the Great Recession was worrisome because it might slow future economic growth. They implied that fiscal stimulus to combat a recession should therefore be limited. They said that when government debt as a percentage of GDP

H ow to C ombat R ecession

rises to a high level, it is likely that financial markets will at some point generate a sharp jump in market interest rates, resulting in slower economic growth. The Reinhart-Rogoff assertion that debt over 90% of GDP it likely to result in slower economic growth has been persuasively challenged by Herndon, Ash, and Pollin (2014) and others. But the Reinhart and Rogoff paper, especially their 90% threshold, has received much more publicity and attention than the replies of critics; I suspect their claim of a 90% threshold has persuaded some economists and policymakers to oppose fiscal stimulus in a recession.

My 2013 Article Entitled “Stimulus without Debt” In 2013 I proposed and explained “stimulus without debt” in an article with that title in the economic policy journal Challenge (Seidman 2013). My article began by noting that a grim lesson from the Great Recession is that widespread worry about government debt generates strong political resistance to enacting a fiscal stimulus large enough to overcome a severe recession. But fortunately there is a way to implement fiscal stimulus in a recession without increasing government debt. I stated that the purpose of my article was to explain and defend a “stimulus without debt” plan under which fiscal stimulus enacted by Congress would be accompanied by a transfer from the Federal Reserve to the US Treasury of the same magnitude so that the Treasury would not have to borrow to finance the fiscal stimulus. I then contrasted stimulus without debt with standard fiscal- monetary stimulus. Under standard stimulus, Congress cuts taxes or raises government spending (transfers or purchases) in order to raise aggregate demand for goods and services, and

W hat Is Stimulus without Debt?

the Treasury borrows an amount equal to the fiscal stimulus by selling US Treasury bonds to the public, thereby increasing government (Treasury) debt held by the public. The Fed then buys an equal amount of Treasury bonds from the public in the open market so that the Fed, not the public, ends up holding the increase in Treasury debt. A crucial point is that the Fed’s action does not reverse the increase in Treasury debt: official Treasury debt increases by an amount equal to the fiscal stimulus, whether or not the Fed buys Treasury bonds from the public. Standard stimulus involves monetizing the debt; it does not prevent debt from being issued by the Treasury I set out the two key components of stimulus without debt. The first component of the stimulus-without-debt plan is fiscal stimulus (an increase in government spending and/or tax cut) enacted by Congress. My recommendation for fiscal stimulus is a tax rebate for households supplemented by other kinds of fiscal stimulus. A tax rebate was implemented during the US recessions in 1975, 2001, and 2008. The second component of the stimulus- without-debt plan is a transfer from the Federal Reserve to the Treasury. In a severe recession the Federal Open Market Committee would give a transfer to the Treasury in an amount decided by the FOMC that, in its judgment, would promote the Federal Reserve’s dual legislative mandate—enacted years ago by Congress—of promoting both high employment and low inflation. It must be emphasized that the Federal Reserve would not be buying bonds from the Treasury; the Treasury would not be incurring a debt—it would be receiving a transfer. I emphasized that a crucial feature of the stimulus-without- debt plan is that it preserves the separation of powers and checks and balances in the implementation of fiscal and monetary policy. The first component—fiscal stimulus—is under the control of Congress (and the president, whose signature is required unless Congress can obtain a two-thirds majority to

H ow to C ombat R ecession

override the president’s veto) but not the Federal Reserve. The second component—the transfer from the Federal Reserve to the Treasury—is under the control of the Federal Reserve but not Congress.

Stimulus without Debt versus Alternative Stimulus Plans I will now contrast stimulus without debt with the following alternatives: (1) helicopter money; (2) monetizing the debt; (3) quantitative easing by the Federal Reserve; (4) transfers from the Federal Reserve to households; (5) money creation by the Treasury as authorized by Congress; (6) not counting bonds held by the Federal Reserve as official government debt; (7) having the Fed burn or shred Treasury bonds equal to the fiscal stimulus. Each alternative will be considered in turn.

Helicopter Money Stimulus without debt is similar to “helicopter money” in certain ways but very different in others. The concept of helicopter money comes from a parable written by Milton Friedman in 1969 in which money is dropped on the population by a helicopter. Stimulus without debt is a policy in which the Federal Reserve gives a transfer to the Treasury, and the Treasury, after authorization by Congress, gives transfers (“tax rebates”) to the population. Hence, the similarity is that under both the parable and the policy, new money is given (not loaned) to the population. In contrast to helicopter money, stimulus without debt specifies how to achieve a separation of powers by assigning particular roles to particular institutions in its practical implementation. In Friedman’s parable, the money in the economy

W hat Is Stimulus without Debt?

increases by the amount dropped from the helicopter. By contrast, under stimulus without debt, when the Fed injects money into the economy by giving a transfer to the Treasury, the Fed is likely to decide to inject less money buying Treasury bonds, so the increase in money in the economy is likely to be much less than the Fed’s transfer to the Treasury. Finally, and most crucially, in Friedman’s parable the helicopter drops money on an economy that is at full employment, whereas the stimulus-without-debt policy should only to be used when the actual output of the economy is well below potential output. With the economy at full employment, giving money to the population would lead to spending that raises prices, not output. Friedman ignored what would happen if the economy were in recession. In that important case, giving money to the population would lead to spending that raises output, not prices. In chapter 11 I look at recent writing by other economists on helicopter money.

Monetizing the Debt The stimulus-without-debt plan proposed in this book does not involve “monetizing the debt” because it creates no debt to monetize: the Treasury sells no bonds, and no additional Treasury bonds are held by either the public or the Federal Reserve; the official federal debt stays constant. By contrast, standard fiscal- monetary stimulus involves “monetizing the debt”: the Treasury sells bonds to the public, the Fed buys Treasury bonds from the public, and official Treasury debt increases.

Quantitative Easing by the Federal Reserve Under quantitative easing by the Fed, the Fed buys bonds in the open market and pays bond sellers with checks that the sellers

H ow to C ombat R ecession

deposit in their banks, thereby increasing bank reserves, which is expected to lead to a reduction in the interest rates that banks offer borrowers, thereby raising borrowing and spending by households and business firms, resulting in more production and employment. To work, quantitative easing must therefore induce households and businesses to incur more debt. But in a severe recession households and business firms are usually reluctant to run up more debt. A huge run-up of the ratio of household debt to household income was one cause of the Great Recession of 2008 and the Great Depression of 1929 (Mian and Sufi 2014). Just prior to each collapse, heavily indebted households became unwilling to borrow more in order to continue their spending, and spending collapsed. Standard monetary stimulus—lowering interest rates—was unable to stimulate a significant increase in consumer spending because deeply indebted households were trying to pay down some of the debt they incurred during the run-up. Under stimulus without debt the Treasury mails tax rebate checks to households in order to raise households’ ability to spend more without incurring more household debt. With tax rebates, households use some of their rebate to pay down debt, some for saving, and some to increase their spending.

Transfers from the Federal Reserve to Households Under Federal Reserve transfers to households, in a recession the Fed would give each household a transfer—for example, $2,000 per adult plus $1,000 per child. To implement this transfer, the Fed would have to obtain the addresses of millions of households—presumably from the Internal Revenue Service—or request that the IRS do the mailing for the Fed; the IRS would need the approval of Congress either to provide the Fed with taxpayers’ mailing addresses, or to do the mailing on

W hat Is Stimulus without Debt?

behalf of the Fed. Under this plan the Fed, not Congress, would specify the dollar amount that would be sent to each household. Many citizens, as well as members of Congress, might question the appropriateness of the central bank, rather than Congress, deciding the specific amounts to go to each household. By contrast, the stimulus-without-debt plan preserves a separation of powers in which Congress retains control over the amount sent to each household and the Federal Reserve’s decision is limited to the total amount of the transfer it gives to the Treasury. When Congress enacted tax rebates for households in 1975, 2001, and 2008, no one questioned the appropriateness of Congress deciding the specific amounts that would go to each household. By contrast, many might question the appropriateness of the Federal Reserve deciding the specific amounts. Moreover, as I will argue in c hapter 5, there are good arguments for Congress to include several other kinds of fiscal stimulus in its fiscal stimulus package: specifically, temporary grants (transfers) from the federal government to state governments, temporary tax incentives for business investment, and federal spending or grants to states for infrastructure maintenance projects. Most citizens would agree that Congress, not the Federal Reserve, should decide whether to include these kinds of fiscal stimulus.

Money Creation by the Treasury as Authorized by Congress If there were no independent central bank, and instead Congress directly controlled not only government spending and taxes but also money creation, then in a recession Congress could set taxes below government spending and authorize the Treasury to create money, rather than sell bonds, to cover the difference. Under this plan, Congress would enact tax rebates for households—for example, $450 billion—and then authorize the Treasury to create

H ow to C ombat R ecession

$450 billion of new money to pay for the rebates. Money creation by the Treasury would eliminate the need for the Treasury to borrow $450 billion by selling bonds, so there would be no increase in government debt. Nor would there be any participation by the Federal Reserve under this plan. Congress would enact the fiscal stimulus and then authorize the Treasury to create the money to pay for it rather than borrow. But if Congress authorizes money creation by the Treasury, then there would be a breakdown of the current separation of powers and checks and balances. Congress would directly control money creation as well as spending and taxation. This would enable Congress to set government spending well above taxes in a normal economy when no stimulus is warranted, and create money to cover the difference, thereby unilaterally injecting a combined fiscal-monetary stimulus that overheats the economy and generates inflation. By contrast, under the stimulus-without- debt plan, Congress, not the Federal Reserve, would authorize the tax rebates to households, and the Federal Reserve, not Congress, would decide the specific amount of money to transfer to the Treasury.

Not Counting Treasury Bonds Held by the Fed as Government Debt Under this plan, the US Treasury and the Congressional Budget Office (CBO) would be instructed by Congress to exclude Treasury bonds held by the Federal Reserve from the calculation of official US government debt. I once argued (Seidman 2001, 2003) that Treasury bonds held by the Fed should not be officially counted as government debt because the Fed is a special lenient creditor. In contrast to other holders of Treasury bonds, the Fed returns most of the interest earned on its Treasury bonds to the Treasury.

W hat Is Stimulus without Debt?

But I now realize that not counting Treasury bonds held by the Fed has a fatal flaw. The Fed may at any time sell these Treasury bonds to the domestic or foreign public, which will, in contrast to the Fed, expect and demand full interest and principal payments from the Treasury on schedule. Thus, the Treasury must be ready to pay the principal and interest for these bonds on schedule in case they are sold by the Fed to the public. Hence, not counting Treasury bonds as part of government debt just because they are currently held by the Fed would understate how much the Treasury might be obligated to pay in the future. Thus, policymakers, members of Congress, financial investors, and the public would be fully justified in insisting that Treasury bonds held by the Fed should continue to be counted as part of official government debt.

Having the Fed Burn or Shred Treasury Bonds Under the “bond-burning” plan, the Treasury would finance the fiscal stimulus by selling new Treasury bonds to the public, the Fed would buy an equal amount of Treasury bonds from the public, and then the Fed would burn these bonds. I concede that the bond-burning plan should, in theory, have the same effect as stimulus without debt because both plans have the same fiscal stimulus, the same amount of money injected into the economy, the same constancy of government debt, and the same effect on the Fed’s balance sheet. The bond-burning plan can be made less incendiary by having the Fed “shred” rather than burn the bonds. The Treasury may prefer that the Fed give the bonds, without compensation, to the Treasury so that the Treasury can shred them and therefore be absolutely certain that they have been shredded. So are there any reasons for preferring stimulus without debt to the bond-shredding plan? I think there are three. First, if the

H ow to C ombat R ecession

bonds are going to be created only to be burned or shredded, why create them in the first place? My transfer from the Fed to the Treasury saves two steps: bond creation by the Treasury and bond shredding by the Fed or Treasury. Second, I think stimulus without debt makes it much clearer to the public and Congress that there is no increase in government debt because the Treasury does not create or sell any new bonds—it does not borrow—to finance the fiscal stimulus. By contrast, under the bond-shredding plan, the Treasury creates and sells new bonds—it borrows—so the Treasury’s action in itself increases government debt. It is true that the shredding of bonds by the Fed or Treasury can keep government debt constant despite the Treasury’s creating and selling new bonds. But why make the public and Congress worry about whether the Treasury’s new borrowing and bond sales will be fully offset by later bond shredding? Third, I suspect that a proposal for bond burning or shredding by the Fed or Treasury would be received more negatively by the public than stimulus without debt. Bond burning or shredding may seem like a trick that lets Congress off the hook for its borrowing; burning or shredding may seem immoral or bizarre. In contrast to burning or shredding bonds, the Fed writes checks every day in order to buy Treasury bonds in the open market or make loans to banks. Stimulus without debt simply proposes that the Fed write several checks to the Treasury (for transfers, not loans) so that Congress can combat a recession without requiring the Treasury to borrow.

Worry about Inflation Would the stimulus-without-debt plan raise inflation? Under the plan, the magnitude of the stimulus would be set with the aim of

W hat Is Stimulus without Debt?

raising aggregate demand for goods and services just enough to achieve full employment. As long as demand for goods and services is not raised significantly above this level for a significant period of time, there would be little or no inflation. As consumer and business confidence rises in response to the economic recovery generated by stimulus without debt, the stimulus would be gradually phased out to avoid an excessive rise in the demand for goods and services once the economy has reached full employment. Chapter 6 explains in greater depth why stimulus without debt would not be inflationary.

Worry about the Fed’s Balance Sheet How would the stimulus-without-debt plan affect the Fed’s balance sheet? If the Fed buys a Treasury bond in the open market, it obtains an asset, but if the Fed gives the Treasury a transfer, it obtains no asset. On the Fed’s balance sheet, the Fed’s “capital” (“net worth”), defined as assets minus liabilities, would therefore be lower if the Fed gave the Treasury a transfer instead of buying Treasury bonds. So at first glance, it might seem that stimulus without debt would cause the Fed to have a balance- sheet problem. But the Fed can avoid this balance-sheet problem. Under the stimulus-without-debt policy, just before the Fed makes its transfer to the Treasury, the Fed would order the printing of an amount of new Federal Reserve notes equal to its transfer to the Treasury and the Fed would then store these notes in the Fed’s vault. Cash in the Fed’s vault is a Fed asset on its balance sheet, so the Fed’s total assets would increase by an amount equal to the Fed’s transfer to the Treasury—the same increase in total assets that would have occurred if the Fed had instead spent the same amount of money buying Treasury bonds. With this

H ow to C ombat R ecession

storing of new Federal Reserve notes in the Fed’s vault, stimulus without debt would not alter the Fed’s capital (net worth), just as the Fed’s purchase of Treasury bonds does not alter the Fed’s capital. Chapter 7 explains further why stimulus without debt would not weaken the Fed’s balance sheet.

Worry about Fed Independence Worry: “Under the stimulus-without-debt plan, Congress will set the magnitude of its fiscal stimulus, and will then pressure the Federal Reserve to provide a transfer to the Treasury equal to the amount of the fiscal stimulus. Therefore, under this plan the Fed will sacrifice its independence. It will be compelled to provide the money Congress wants.” But under the stimulus-without-debt plan, the Fed provides a transfer to the Treasury only if the Fed judges that this transfer would advance the Fed’s mandate, prescribed by Congress many years ago, of promoting high employment and low inflation. Under the plan the Fed is called upon to set the magnitude of its transfer to the Treasury to promote this mandate according to the Fed’s judgment. Of course, some in Congress may try to pressure the Fed to provide the magnitude of transfer to the Treasury that they would like. But the same is true today of interest rates and Fed open-market operations: there are some in Congress who try to pressure the Fed to adjust interest rates or open-market operations to their liking. Nevertheless, despite such pressure from some members of Congress, the Fed has generally maintained its independence about adjusting interest rates and open-market operations. It seems equally likely that the Fed would maintain its independence in deciding whether to give a transfer to the Treasury, and if so, how much the transfer should be. Chapter 8

W hat Is Stimulus without Debt?

further discusses why stimulus without debt would not undermine the Fed’s current degree of independence from Congress and from the president.

Would an Amendment to the Federal Reserve Act Be Needed? I have not been able to find a clear answer to the question of whether the Fed is prohibited under current law (the Federal Reserve Act or any other law) from (1) giving a large transfer to the Treasury; or (2) ordering a large amount of new Federal Reserve notes (equal to its transfer to the Treasury) to keep in its own vault and record as an asset on its balance sheet. Legal scholars may have written about it, but I have been unable to find such writings. Ideally, before the next recession advocates of stimulus without debt will find a member of Congress who is willing to introduce a bill that would amend the Federal Reserve Act to explicitly permit the Fed to take the two actions listed above. Committee hearings on the bill would generate a debate on the stimulus-without-debt policy. If the next recession hits before Congress has considered the amendment, advocates should intensify their effort to have the bill introduced. In the meantime, I recommend that the Fed consult its own legal department and external legal scholars, and if the Fed concludes that there is no clear prohibition under current law, the Fed should go ahead and take actions (1) and (2) so that Congress can enact fiscal stimulus equal to the Fed’s transfer without borrowing by the Treasury and without weakening the Fed’s balance sheet. If most legal scholars say the Fed isn’t prohibited from taking these two actions in a recession, the Fed faces a choice. It can wait for the next recession and then take action without seeking

H ow to C ombat R ecession

approval from Congress. Or it can ask Congress now to amend the Federal Reserve Act to explicitly give the Fed permission to take these two actions in the next recession. On the other hand, if most legal scholars say that under current law the Fed is prohibited from so acting, then the Fed has no choice but to ask Congress to amend the Federal Reserve Act now if it wants the option of playing its key role in implementing stimulus without debt in the next recession. Suppose the issue is ignored until the next recession. If the Fed chair asks the Federal Reserve’s Open Market Committee to authorize a large transfer to the Treasury for fiscal stimulus to combat the recession, it is likely that at least a minority of the FOMC will contend that the Fed is prohibited under current law. If a majority of the FOMC voted to do it and therefore a large transfer for fiscal stimulus was made (either by writing a check to the Treasury or by crediting the Treasury’s account at the Fed), it would probably be challenged in court. I don’t know whether a court can issue an injunction for the Treasury to return the transfer to the Fed or to hold the transfer rather than use it. Opponents of the Fed transfer might propose legislation that would order the Treasury to return the transfer to the Fed and prohibit the Fed from making a transfer in the future. Supporters of the Fed transfer in a recession might propose an amendment to the Federal Reserve Act permitting a transfer. In the meantime, Congress wouldn’t know whether any fiscal stimulus it considered enacting would be financed by the Fed’s transfer or would require borrowing by the Treasury. A legislator might vote for fiscal stimulus if it would not require borrowing by the Treasury, but vote against fiscal stimulus if it would require borrowing. Clearly it would be better if Congress took action now to clarify the Fed’s power before the next recession. Stimulus-without-debt

W hat Is Stimulus without Debt?

advocates should work to find a member of Congress to introduce a bill now and get hearings held in the House or Senate committee that has jurisdiction. Committee hearings on the amendment would generate a debate on the stimulus-without- debt policy. The bill would amend Section 14 of the Federal Reserve Act, currently entitled “Open-Market Operations.” The title of Section 14 would be changed to “Open-Market Operations and Transfers of Funds to the Treasury.” Section 14 currently has two subsections, (1) and (2). The amendment would add two new subsections, (3) and (4). Subsection (3) would give the Federal Open Market Committee the power to transfer funds to the Treasury to be used only for fiscal stimulus enacted by Congress in a recession or slow recovery or stagnant economy. Subsection (4) would require the Federal Reserve to order new Federal Reserve notes from the Bureau of Engraving and Printing equal to the amount of the Fed’s transfer to the Treasury, and require the Fed to keep these new notes in a Federal Reserve vault; these new notes would therefore be recorded as an asset on the Fed’s balance sheet. As explained earlier in my discussion of “worry about the Fed’s balance sheet,” subsection (4) would ensure that stimulus without debt leaves the Fed’s net worth (capital) unchanged on its balance sheet. The political fate of the proposed amendment to the Federal Reserve Act would depend on the composition of the fiscal stimulus package that legislators expected. Discussions between members of the two parties should begin now over whether they can come to a tentative agreement over the composition of a fiscal stimulus package for the next recession. I urge both parties go back to the 2008 package that passed with bipartisan support and use it as a starting point to negotiate a tentative fiscal stimulus package for the next recession. In 2008, the

H ow to C ombat R ecession

president was Republican, and Democrats were a majority in the House. With bipartisan support in Congress, a fiscal stimulus package was enacted that contained just two components: tax rebates to households (two-thirds of the package) and temporary tax incentives for business investment (one-third). In chapter 4 I’ll argue for making tax rebates the main component in a fiscal stimulus package, and in chapter 5 I’ll recommend also including in the package temporary tax incentives for business investment, temporary aid to state and local governments, and temporary federal spending on infrastructure maintenance projects, but excluding from the package components that favor the long- term agenda of one political party but not the other. By contrast, in 2009 and 2010 Democrats had a majority in the House and Senate and a Democrat in the White House, and shaped the composition of the fiscal stimulus package to largely favor their priorities. Democrats were able to the pass the fiscal stimulus bill, but Republicans voted unanimously against it. Ideally, agreement on the composition of a fiscal stimulus package for the next recession and enactment of the proposed amendment to the Federal Reserve Act would both be achieved before the next recession. Unfortunately, Congress may not take action until the pressure of an actual recession is upon us. If some preliminary work has been done, however, it might be possible to move quickly as soon as the next recession hits. Under the pressure of recession, it might be possible for a bipartisan fiscal stimulus package to be negotiated. With both parties interested in avoiding higher federal debt, Congress might be willing to pass the amendment to the Federal Reserve Act. Finally, the Federal Reserve’s Open Market Committee might be willing, after learning about the contents of the negotiated fiscal stimulus package, to make the transfer to the Treasury that would enable the fiscal stimulus package to be implemented without any new borrowing by the Treasury.

W hat Is Stimulus without Debt?

Can’t Monetary Stimulus Overcome a Severe Recession? Can’t monetary stimulus overcome a severe recession? As I emphasized earlier in this chapter, the Fed can and should act as a lender of last resort by aggressively performing financial rescues and unfreezing credit markets, and should cut interest rates to zero in order to help maintain some borrowing and spending by households and firms. But it is crucial to understand why cutting interest rates to zero cannot reverse the large plunge in aggregate demand that caused the severe recession. In the first edition to his influential college economics textbook, with the experience of the Great Depression clearly in mind, future Nobel laureate Paul Samuelson (1948) gave the explanation shared by early Keynesians concerning the impotence of low interest rates to overcome a severe recession In a section entitled “The Inadequacies of Monetary Control of the Business Cycle,” Samuelson said that few economists (in 1948) thought the Federal Reserve could overcome a severe recession. Yes, the Fed can increase the supply of money and bank deposits, but the Fed can’t compel the money to be used to demand goods and services, thereby stimulating production and employment. In the middle of a depression, banks hesitate to make loans to business firms or households, so though the Fed can buy government bonds in the open market and thereby increase bank reserves, the banks will simply hold the reserves and there won’t be more borrowing and spending. Even if banks cut interest rates, questionnaire studies of businessmen’s behavior show that a fall in interest rates does not induce borrowing and spending on new machinery a in a severe recession when customer demand is weak, so business investment is likely to be unresponsive to the interest rate. The same is true about people’s decisions on how much to spend on consumption; a cut in interest rates won’t

H ow to C ombat R ecession

induce nervous consumers to take on more debt in a deep recession. Samuelson therefore concluded that monetary policy is too weak to overcome a severe recession. But doesn’t the “quantity theory of money” show that injecting more money into the economy must result in more spending on output, thereby stimulating production? Samuelson, like Keynes, rejected this “quantity theory.” He said that in the language of the quantity theory of money, the “velocity” of circulation of money does not remain constant. He invoked the analogy that you can lead a horse to water, but you can’t make it drink. The Fed can inject money into banks, but in a recession it can’t make pessimistic businessmen and consumers borrow and spend that money. The quantity of money in the economy will go up, but in a recession the velocity of circulation of money will go down, so spending and output will stay constant. Samuelson’s analysis, which I believe is still correct today, expresses the view of the Keynesian economists who dominated the economics profession in the 1940s, 1950s, and 1960s, such as Keynes and future Nobel Prize winners Tinbergen, Samuelson, Hicks, Klein, Tobin, Modigliani, and Solow; and current Keynesians such as Blinder and Nobel Prize winners Krugman, Stiglitz, Akerlof, and Shiller. To explain most recessions, traditional Keynesians focus on aggregate demand for goods and services: if aggregate demand falls, output and employment will fall. Aggregate demand can fall for a variety of reasons, including the bursting of a stock market or housing bubble, or a retrenchment in spending by consumers or business managers in reaction to their own excessive spending and the running up of excessive debt, or a psychological shift of consumers or business managers from optimism to pessimism, and from confidence to anxiety. Standard monetary stimulus—a Fed injection of high-powered money into the banking system— can lower interest rates, but in a deep recession won’t induce most potential spenders to risk borrowing and spending because

W hat Is Stimulus without Debt?

loans must be paid back. Low interest rates can’t overcome a severe recession. New Classical economists and even many New Keynesian economists differ from traditional Keynesians in key respects. New Classical economists contend that the economy will automatically recover quickly from recession as long as the central bank keeps the money supply from contracting; and that fiscal stimulus doesn’t work but does generate harmful government debt. New Keynesian economists are “Keynesian” in their view that keeping the money supply from contracting is not enough to generate a quick recovery from a deep recession. But prior to the 2008 recession, many New Keynesians believed that active standard monetary stimulus—cutting interest rates—would be sufficient to generate an adequate recovery from a deep recession and that fiscal stimulus was either unnecessary or ineffective. The experience of the 2008 recession caused some New Keynesians to recognize a basic tenet of traditional Keynesian economics as they came to realize that fiscal stimulus is essential for overcoming a deep recession (David Romer 2012). The recent experience of the Great Recession provides strong evidence that when the Fed injects money into the economy by buying bonds and making loans, reserves sit idly in banks because anxious consumers and businesses are afraid to borrow. In their book on the 2008 recession, Mian and Sufi (2014) take issue with “the bank-lending view,” which holds that the key to recovery from recession is for the Fed to inject lots of money into banks in the hope that the banks will lend lots of money to consumers and businesses at low interest rates. They ask whether households and firms really want to borrow when the economy is depressed. They cite evidence from surveys of small businesses by the National Federation of Independent Businesses (NFIB), which asks business managers to indicate

H ow to C ombat R ecession

their most important concern and offers choices such as poor sales, regulation and taxes, and financing and interest rates. The percentage of managers who put financing and interest rates as their main concern never rose above 5% from 2007 to 2009. This contradicts the view that small businesses were held back because they were unable to get financing. By contrast, from 2007 to 2009 the percentage of businesses citing poor sales as their top concern jumped from 10% to almost 35%. Businesses held back because of a sharp cut in customer demand. Finally, even if monetary stimulus alone could overcome a severe recession, it would be better to rely mainly on tax rebates to households. There are at least two reasons for preferring tax rebates to cutting interest rates. First, tax rebates to households spread the increase in demand across all goods and services, roughly replicating the fall in demand across all goods and services during a recession. To respond to this demand, most firms would seek to rehire the same workers they had laid off. By contrast, monetary stimulus concentrates its direct effect on interest-sensitive sectors, a much narrower portion of the economy. Firms in the rest of the economy would not seek to rehire workers they had laid off. Thus, stimulus from tax rebates roughly restores the prerecession mix of goods, services, and jobs, whereas stimulus from cutting interest rates does not. Second, relying on monetary stimulus alone means interest rates must be cut to zero and kept there for a long time. This puts a large and unfair burden on retirees who built up savings in their bank account during their work years, hoping that in retirement their bank would pay a normal interest rate. With stimulus from tax rebates restoring demand across all goods and services, interest rates would stay normal instead of being cut to zero. Chapter 9 further explains why it would be a mistake to rely on monetary stimulus to overcome a severe recession.

W hat Is Stimulus without Debt?

What Should Have Been Done in the Great Recession? By the beginning of 2008 it was clear that the economy was falling into a recession. At that time, however, the severity of the recession was not widely anticipated. But by the middle of 2008, it was evident that the recession would be substantially worse than had been expected at the beginning of 2008. By the fall of 2008, it was clear that the economy was plunging into a severe recession. What fiscal/monetary stimulus was applied in 2008, 2009, and 2010? What should have been applied in 2008, 2009, and 2010? Throughout this section I will assume, to simplify the calculation, that potential GDP—the GDP that would have occurred annually had there been no recession in these years—would have been $15,000 billion per year. In early 2008, a fiscal stimulus bill of about $150 billion passed Congress with bipartisan support and was signed by President Bush. It consisted primarily of a tax rebate for households ($100 billion), but also contained an investment bonus for business firms ($50 billion). With annual potential GDP equal to $15,000 billion, the fiscal stimulus was about 1% of potential GDP. In the remainder of this section, I will drop the word potential and simply write “1% of GDP.” The empirical studies that I examine in c hapter 4 find that a plausible estimate of a tax-rebate multiplier in recession is 1.0; a tax- rebate multiplier is less than a government-purchases multiplier because consumers save a portion of the rebate. Thus, this 1% of GDP tax rebate stimulus would be estimated to raise real GDP about 1%. With an output-to-unemployment estimate of two to one, this 1.0% increase in GDP would be estimated to reduce the unemployment rate 0.5%—for example, from 7.0% to 6.5%. Although enacted in early 2008, the tax rebate checks were not sent to households by the US Treasury until the summer of 2008.

H ow to C ombat R ecession

Because the 2008 fiscal stimulus was 1% of GDP, it made the federal deficit 1% of GDP larger than it would have been without the fiscal stimulus—for example, if the deficit would have been 2% of GDP without the 1% of GDP fiscal stimulus, it would have been 3% with the fiscal stimulus. By the summer of 2008, and certainly by the fall, it was clear that another, much larger fiscal stimulus would be needed in 2008. By summer, fiscal stimulus advocates (including me) were writing op-ed articles calling for the prompt enactment of a much larger stimulus. But in the second half of 2008 the president did not propose, nor did Congress enact, another fiscal stimulus. This was a policy failure of enormous magnitude. After the shocking bankruptcy of Lehman Brothers in September, attention was rightly focused on immediate rescues of financial firms to prevent a financial panic. What was unjustified, however, was the failure to even consider the enactment of a large fiscal stimulus in the second half of 2008. The election of November 2008 meant that a pro-fiscal- stimulus president would take office in mid-January, and pro- fiscal-stimulus legislators would have a majority in both the House and the Senate in the new Congress. With remarkable speed, in early January the soon-to-be-president and the new Congress began developing a two-year fiscal stimulus, which was enacted into law in mid-February. The two-year stimulus would inject about $750 billion—$375 per year (2.5% of a potential GDP of $15,000 billion) into the economy in 2009 and another $375 billion (2.5% of GDP) in 2010. Over the three years 2008, 2009, and 2010, the fiscal stimulus was $900 billion ($150 + $375 + $375), 6% of GDP over three years, or an average of 2% of GDP per year. With a multiplier of 1.0, the stimulus would have raised GDP by 2% per year. With an output-to-unemployment estimate of two to one, the stimulus would have made the unemployment rate one percentage point lower than it otherwise

W hat Is Stimulus without Debt?

would have been—for example, at the end of 2010, 10% (its actual value) instead of 11%. Thus, the fiscal stimulus implemented from 2008 through 2010, though large by historical standards, was too small to fully overcome the Great Recession. At the end of 2010, aggregate demand for goods and services was still well below potential output, the unemployment rate was still 9%, and consumer and business confidence was still well below normal. Yet the one-year fiscal stimulus legislation enacted in February 2008 was not renewed; nor was the two-year fiscal stimulus legislation enacted in February 2009 (the American Recovery and Reinvestment Act). Some additional fiscal stimulus was enacted and implemented in 2011 and 2012, but it was significantly smaller than the ARRA. The $150 billion fiscal stimulus enacted in February 2008 consisted of tax rebate checks to households made in the summer of 2008 and bonus depreciation allowances for business firms on the 2008 tax return; there was no provision for anything beyond 2008. The $750 billion fiscal stimulus ($375 billion for 2009 and $375 for 2010) enacted in February 2009, the American Recovery and Reinvestment Act (ARRA), consisted of grants to state governments, tax cuts, and government purchases made primarily in two years, 2009 and 2010; there was no provision for any renewal of its components. The lesson from the actual fiscal stimulus in the Great Recession is not that that fiscal stimulus doesn’t work, but rather that it was too small and phased out too quickly to overcome the severe recession. Republicans, who overwhelmingly opposed fiscal stimulus, won a majority in the House of Representatives in November 2010 and blocked renewal of ARRA components starting in January 2011. The Obama administration was able to get some new fiscal stimulus through Congress in 2011 and 2012—such as a temporary payroll tax cut—but the stimulus was too small to turn the recovery from sluggish to robust.

H ow to C ombat R ecession

Suppose Congress and the president had enacted and implemented a fiscal stimulus consisting mainly of tax rebates that was three times as great—6% of potential GDP each year from 2008 to 2010. With a multiplier estimate of 1.0, the 6% of GDP fiscal stimulus would have raised GDP in each year 6% above what it otherwise would have been, and with an output- to-unemployment estimate of two to one, reduced the unemployment rate three percentage points below what it otherwise would have been each year—for example, at the end of 2010 the unemployment rate would have been 8% instead of 11%. Suppose the entire fiscal stimulus had taken the form of tax rebates. With 300 million people, $450 billion (3% of GDP) each June and December (6% of GDP for the year) would have supported a rebate of $1,500 per person (for example, $2,000 per adult, $1,000 per child) each June and December, so twice a year a family of two adults and two children would have received a $6,000 rebate check. Why wasn’t a 6% of GDP fiscal stimulus enacted in each year from 2008 through 2010? In those years no one considered proposing a transfer from the Fed to the Treasury—a key element of stimulus without debt. It was therefore assumed that a 6% of GDP fiscal stimulus each year would raise government debt by 6% of GDP each year. With debt already rising sharply as a percentage of GDP due to the recession itself, many opponents of fiscal stimulus voiced alarm about making government debt larger, and even many supporters of fiscal stimulus worried about larger debt. Suppose, however, that prior to 2008, stimulus without debt had been proposed by economists and that Congress had amended the Federal Reserve Act to permit the Federal Reserve to give a transfer to the Treasury for fiscal stimulus in a recession. As the economy plunged into severe recession in 2008, Congress might have been willing to enact fiscal stimulus equal to 6% of

W hat Is Stimulus without Debt?

GDP each year for three years provided the Fed was willing to provide a transfer to the Treasury equal to 6% of GDP each year for three years. If the Fed decided to make the 6% of GDP transfer each year from 2008 to 2010, and Congress enacted the 6% fiscal stimulus each year from 2008 to 2010, the unemployment rate at the end of 2010 might have been 8%, three points below what it would have been without any fiscal stimulus (11%), and two points below its actual value (10%) which resulted from the actual 2% of GDP fiscal stimulus (the average of 1% in 2008, 2.5% in 2009, and 2.5% in 2010). By summer of 2010 it was clear that despite the 2% of GDP fiscal stimulus, the recovery was still not strong enough. Pro- stimulus legislators still had majorities in House and Senate. But these pro-stimulus legislators were afraid of making government debt still larger, so they didn’t go forward. If stimulus without debt had been available, these pro-stimulus legislators might have passed another two-year fiscal stimulus for 2011 and 2012. So with stimulus without debt, we might have had a much larger fiscal stimulus for another two years. The stimulus would have strengthened the recovery and we might have reached 5% unemployment two years earlier than we did (2013 instead of 2015). Thus, stimulus without debt is a means of getting a larger and/ or longer fiscal stimulus passed by Congress in a recession or a slow recovery.

Would Stimulus without Debt Increase Government Deficits? Under the stimulus-without-debt policy, the government deficit would not increase because the government’s Treasury would receive revenue from the Federal Reserve equal to the fiscal stimulus. For example, suppose the fiscal stimulus consists of $900

H ow to C ombat R ecession

billion of spending on tax rebates and other programs; then the Treasury would receive $900 billion in revenue from the Fed, so the increase in the Treasury’s total revenue ($900 billion) would equal the increase in its spending. This is the same budget treatment given by state governments to receipt of transfers from the federal government; the state government adds federal transfers received to state tax revenue to obtain total revenue, which is then compared to total state government spending (outlays) to determine the state government’s deficit. But isn’t fiscal stimulus the same thing as an increase in the deficit? No, it isn’t. Fiscal stimulus is an increase in government spending and/or a cut in taxes because both of these raise aggregate demand for goods and services. A simple definition of fiscal stimulus is the difference between government spending and tax revenue. If government spending (at a given level of GDP) increases, aggregate demand for goods and services (at that level of GDP) increases; if tax revenue (at a given level of GDP) decreases, aggregate demand (at that level of GDP) increases. This definition is too simple because an X-dollar increase in government purchases of goods and services increases aggregate demand more than an X-dollar increase in government transfers because recipients save a portion of transfers that they receive; for the same reason, an X-dollar decrease in tax revenue causes a smaller increase in aggregate demand than an X-dollar increase in purchases. These complications need to be kept in mind when using my simple definition. The government deficit is the difference between government spending and total revenue, not just tax revenue. In particular, if the government receives transfers (grants) from any other entity, the transfers received by the government increase the government’s total revenue. For example, suppose a state government spends $250 billion, collects $200 billion in taxes, and receives $50 billion in grants (transfers) from the federal

W hat Is Stimulus without Debt?

government. Then its deficit is $0 even though its spending is $50 billion greater than its tax revenue. This state government does not have to borrow; its deficit is $0 and there is no increase in its debt. If it is under a balanced budget requirement by law, it has met the requirement. Suppose initially the federal government spends $3,000 billion, collects $3,000 billion in tax revenue, and receives no transfer from the Federal Reserve. Under my simple definition of fiscal stimulus, its fiscal stimulus is $0. Its deficit is also $0. Now suppose its spending is raised to $3,900 billion while its tax revenue remains $3,000 billion, so its fiscal stimulus is $900 billion. If it still receives no transfer from the Fed, its deficit is also $900 billion, but if it receives a transfer of $900 billion from the Fed, its deficit is $0. Thus, fiscal stimulus can be increased without increasing the deficit if the stimulus is accompanied by an equal increase in transfers received from the Federal Reserve.

Chapter 4

Do Tax Rebates Work in a Recession?

The most recent uses of tax rebates to households in the United States to combat a recession were in the summer of 2001 and the summer of 2008. Several studies published in the year following each rebate claimed that the rebates did not work. But in my view the study using the best methodology and the best data to analyze the impact of the 2001 tax rebate was coauthored by Johnson, Parker, and Souleles (2006) and published in a top- ranked, peer-refereed journal, the American Economic Review, in its December 2006 issue, five years after the 2001 rebate occurred; it estimated that consumers spent about one-third of the rebate within three months and about two-thirds of the rebate within six months. The study using the best methodology and the best data to analyze the impact of the 2008 tax rebate, was published in the American Economic Review in its October 2013 issue about five years after the 2008 rebate occurred and was, coauthored by Parker, Souleles, Johnson, and McClelland (2013); it estimated a slightly larger and faster response. Another study coauthored by Broda and Parker (2014) and published in the Journal of Monetary Economics in its fall 2014 issue estimated a response 58

Do Tax Rebates Work in a Recession?

similar to the two other studies. The conclusion from these three studies is that the tax rebates worked: the average household recipient spent about one-third of the rebate within three months and about two-thirds within six months. This chapter will review the major studies of the 2001 and 2008 tax rebates, explain and contrast the methodologies and data in the studies, and draw policy conclusions.

Do Tax Rebates to Households in a Recession Actually Work? At first glance it seems commonsensical that the consumption spending of a typical household varies directly with its current disposable income, even if its spending also depends on other factors. Because the consumption spending of a typical household is a high percentage of its current disposable income, at first glance it also seems commonsensical that if a typical household receives a tax rebate in a recession, within a year it will spend a significant percentage of its rebate. Some economists, however, are skeptical that tax rebates to households in a recession will be spent during the recession. These economists contend that because households know that the rebates will stop when the recession ends, they will save the rebates during the recession. Other economists, including me, however, find it plausible that the percentage of a recession tax rebate that the typical household spends within a year is substantially less than 100% but also substantially greater than 0%. Most economists agree that empirical research based on study of the actual response of actual consumers to actual tax rebates in a recession is needed to determine how effective tax rebates are in stimulating consumer spending during a recession. The US Treasury mailed out one round of tax rebate checks to households in the 2001 recession

H ow to C ombat R ecession

and one round in the 2008 recession. This chapter will examine the major empirical studies of these two episodes. It will be useful at the outset to provide a big-picture summary. Details will be supplied in the rest of this chapter. The effectiveness of the summer 2001 tax rebate was quickly thrown into doubt by a consumer survey, conducted and interpreted by two respected professors of economics at the University of Michigan, Mathew Shapiro and Joel Slemrod, who asked a sample of consumers what they planned to do with their rebate. In fall 2001 the coauthors of the survey reported that “only” 22 percent of consumers said they planned to “mostly spend” their rebate, and these authors communicated to the media that their survey showed that the tax rebates didn’t work. Besides generating media attention and publicity in the fall of 2001, they published an article with their survey results and interpretation in the American Economic Review (Shapiro and Slemrod 2003a) and another article with additional analysis of their survey in a volume on tax policy and the economy (Shapiro and Slemrod 2003b). Their survey, study, and interpretation influenced many to conclude that tax rebates didn’t work. It took five years after the Shapiro-Slemrod fall 2001 consumer survey before another study, coauthored by three other respected economists, David Johnson, Jonathan Parker, and Nicholas Souleles (2006), was published in the American Economic Review in late 2006, using data on actual expenditures reported by individual consumers following the 2001 rebate rather than a survey of consumer intentions, and concluded that consumers had in fact spent about one-third of their rebate dollars within three months of receiving it and about two-thirds of their rebate dollars within six months—in other words, the tax rebates of 2001 had in fact been very effective in boosting consumer spending above what it otherwise would have been in the half- year following their payment to households.

Do Tax Rebates Work in a Recession?

The initial claim that the summer 2008 tax rebates were ineffective in combating the 2008 recession came from two influential professors of economics, Martin Feldstein and John Taylor, who separately and independently published op-ed articles in the Wall Street Journal concluding that the 2008 tax rebates had been ineffective (Feldstein 2008; Taylor 2008), and presented their analysis at the January 2009 annual conference of the American Economic Association (AEA), which received substantial media coverage during the period when Congress and the new president were deciding the composition of a fiscal stimulus package to combat the Great Recession. Their articles based on their presentations were published in the May 2009 Papers and Proceedings of the AEA conference (Feldstein 2009; Taylor 2009). Their op-ed articles and presentations had a significant impact on the view of policymakers and the public. It took three years before two other economics professors, my colleague Kenneth Lewis and I (Lewis and Seidman 2012), published a detailed analysis and rebuttal of the articles by Feldstein and Taylor in an economics journal that has a much smaller readership and gets much less media coverage. At the same time, the two professors of economics who conducted the consumer survey in fall 2001, Shapiro and Slemrod, conducted a consumer survey in fall 2008 and presented their analysis at the same January 2009 conference of the AEA. Their article (Shapiro and Slemrod 2009) based on their presentation was also published in May 2009 AEA Papers and Proceedings. They also published a second article (Sahm, Shapiro, and Slemrod 2010) on the 2008 rebate in an annual volume on tax policy and the economy in 2010, joined by coauthor Claudia Sahm. The featured headline from both of their articles was that their survey showed that “only about a fifth” of consumers said they planned to “mostly spend” their rebate, and these authors communicated to the media that once again, as in 2001, their survey showed that tax rebates didn’t work.

H ow to C ombat R ecession

Once again, it was five years after the 2008 rebates were mailed to households before another study, conducted and interpreted by the same professors who had examined actual expenditures reported by consumers following the 2001 rebate, Parker, Souleles, Johnson, with a new coauthor McClelland, was published in the American Economic Review in October 2013. It estimated that the response of consumers to the 2008 rebate was slightly larger and faster than they had estimated for the 2001 rebate. They concluded that tax rebates were very effective in promptly raising consumer spending. It took six years for a study by Broda and Parker using different data to be published in the Journal of Monetary Economics in fall 2014; it estimated a similar magnitude for the impact of the 2008 rebate on consumer spending. Although the recovery from the Great Recession of 2008 was very slow, no second tax rebate was issued after 2008. There were, however, two other policies that were used from 2009 through 2012 that at first glance seem similar to tax rebates and were viewed by some policymakers as replacements for the 2008 tax rebate. In 2009 and 2010, Congress enacted a “Making Work Pay” tax credit for employees ($400 for singles and $800 for married couples) that was implemented by requiring employers to reduce federal income tax withholding according to a schedule set by Congress (Seidman 2009). In 2011 and 2012, Congress replaced the Making Work Pay credit with a temporary payroll tax cut for employees from 6.2% to 4.2%, which was implemented by requiring employers to reduce federal payroll tax withholding according to schedule set by Congress (the Social Security Trust Fund would be reimbursed by general revenue so that future Social Security benefits of these employees would be unaffected). At the end of this chapter, I will compare the method of cutting withholding (either for federal income tax or federal payroll tax) to mailing out tax rebate checks to households by the US Treasury.

Do Tax Rebates Work in a Recession?

The 2001 Tax Rebate Prior to the onset of recession in late 2000, Republican presidential candidate George W. Bush in early 2000 proposed a permanent across-the-board cut in income tax rates for the same reason that candidate Ronald Reagan had done in 1980: conserv ative philosophy and supply-side economics. Like Reagan, Bush believed in lower taxes and less government domestic spending. Like Reagan, Bush accepted supply-side economics, which held that high marginal tax rates were discouraging people from supplying labor and saving, thereby holding back the pace of economic growth and of the economy. During the 2000 campaign, Democrats opposed Bush’s permanent tax cut proposal. But in early 2001 there was a new consideration: a recession, precipitated by the bursting of the stock market bubble in 2000. Democrats continued to reject the reasons Bush gave for his permanent across-the-board cut in income tax rates. But some Democrats instead supported a temporary tax rebate with equal dollar benefits for households of all incomes to help them cope with the recession and boost demand for goods and services to combat the recession. On January 28, 2001, my op-ed article was published in the Philadelphia Inquirer with this title: “In a Recession, Think Tax Rebates, Not Across-the-Board Cuts in Taxes.” My op-ed began by reminding readers that a tax rebate for households had been enacted with bipartisan support during the 1975 recession to stimulate consumer spending. President Ford signed the Tax Reduction Act of 1975 in January. Its key element was a 10% rebate of each household’s 1974 personal income tax up to a maximum of $200 per household. I urged the enactment of a similar tax rebate in 2001. In fact, the 2001 Bush tax-rate cut proposal, while originally not containing a tax rebate, contained one element that could

H ow to C ombat R ecession

be implemented through a tax rebate. Under the income tax schedule then in effect, the tax rate that applied to the first dollars of taxable income was 15%. Although the Bush plan emphasized the cut in the top marginal tax rates, it also contained a proposal to create a new 10% bracket that would apply to the first $12,000 of taxable income for a couple filing jointly, and the first $6,000 for a single person; thereafter, the household would enter the 15% tax bracket. This cut of five percentage points (15% to 10%) on the first $12,000 of taxable income for a couple meant a $600 tax cut; on the first $6,000 for a single person, a $300 tax cut. Taxpayers of all incomes pay the first-bracket rate on their first dollars of taxable income, so all couples with taxable income above $12,000 would receive an equal dollar tax cut—$600 under the Bush plan (and all singles with taxable income above $6,000 would receive a $300 tax cut). This one element of the Bush plan appealed to many Democrats. The final tax act, passed by Congress near the end of May, reflected the fact that the Republicans controlled both Congress and the White House. The Democrats proposal to enact an additional one-time check to people who paid payroll tax as well as income tax was omitted. However, the president’s five-percentage- point cut on the first $12,000 (for a couple), instead of being implemented through small changes in withholding rates and a final adjustment on the 1040 tax return due April 15, 2002, would be implemented more promptly, just as the Democrats proposed: tax rebate checks would be mailed out by the US Treasury quickly (between July and September 2001), $600 per couple ($300 for an individual). On June 7, 2001, President Bush signed the tax bill into law. In early July, millions of taxpayers received a “Dear Taxpayer” letter from the US Treasury saying that they would receive a check for $600 (couple) or $300 (single) due to action by Congress and the president. The Treasury did not mail out all the checks on the same day. Instead, a fraction

Do Tax Rebates Work in a Recession?

of the checks were mailed out each week for three months. Each recipient’s Social Security number determined the week the recipient received the check. (The tax rebate of 2001 is more fully described in Seidman 2003.) In the summer of 2001, real GDP growth stayed below normal, the stock market continued to fall, and the unemployment rate continued to rise, so some of us came to the conclusion that preparation should begin immediately for a second round of tax rebates in the fall. For example, my letter to the editor appeared in the New York Times on September 1, 2001, with the title “The Rebate Helps. Do It Again.” However, momentum for repeating the rebate was stopped in its tracks by the announcement by two respected economists at the University of Michigan, Mathew Shapiro and Joel Slemrod, that results from their survey of consumers indicated that the rebates were ineffective because only 22% of recipients indicated the rebate dollars would be mostly spent.

Shapiro and Slemrod’s 2001 Consumer Survey In the fall of 2001, Shapiro and Slemrod released results to the media of their survey of roughly 1,500 consumers who had received the 2001 tax rebate that summer. They subsequently published two full-length articles on their survey of the 2001 rebate, the first in a top-ranked economics journal (Shapiro and Slemrod 2003a), the second in a respected annual volume on tax policy and the economy (Shapiro and Slemrod 2003b). Their results received substantial top media coverage beginning in the fall of 2001 that definitely influenced the debate in Congress against enacting a second rebate in late 2001 or in 2002. They asked survey questions of households in August and September

H ow to C ombat R ecession

about their plans for the tax rebate mailed by the US Treasury in July, August, and September, and reported that only 22% said they “would mostly increase spending”; most respondents said they would to use the rebate either to pay down debt or to save. “Only” was the word the authors emphasized in their report, and “only” was the word the media focused on. The report and its coverage conveyed the message to the public that the rebate didn’t work. They soon wrote up their results in a National Bureau of Economic Research (NBER) working paper, and that paper, “Consumer Response to Tax Rebates,” was subsequently published in the American Economic Review (Shapiro and Slemrod 2003a). In their survey they asked this question: “Thinking about your family’s financial situation this year, will the tax rebate lead you mostly to increase spending, mostly to increase saving, or mostly to pay off debt?” Note the phrase “Thinking about your (family’s) financial situation.” Why did Shapiro and Slemrod insert such a phrase? Why tell the respondent what he or she should be thinking about in order to decide what to do with the rebate? If you’re told to think about your financial situation, you’re likely to think, “I really want to buy some badly needed goods and services, but thinking about my financial situation makes me realize I better instead save or pay off debt.” The introductory phrase may well have influenced some respondents to answer either “mostly to increase saving” or “mostly to increase debt.” Suppose instead the phrase had been “Thinking about the goods and services your family badly needs . . . ” Surely that phrase would have influenced some respondents to answer “mostly increase spending.” The best way to be neutral would have been to have no phrase at all. The second problem with their wording is the time frame. The respondent is asked to think about the family’s financial situation “this year,” but the sentence does not end with a specific

Do Tax Rebates Work in a Recession?

time frame such as “over the next month” or “over the next two months” or “by the end of this year” or “over the next twelve months.” The respondent is given no explicit guidance about the time frame. It seems likely that many respondents would read the phrase “this year,” and assume the question meant, “What will you do by the end of this calendar year?” The US Treasury mailed the rebates out every week from late July to late September, so the average recipient received the rebate in late August. After receiving the rebate, the recipient had four months left until the end of the calendar year. Thus, it seems likely that the average recipient would interpret the question as, “What will you mostly do with the rebate in the next four months?” The third problem with their question is that it does not ask, “What percentage or fraction of the rebate do you plan to spend this year?” In response to the release by Shapiro and Slemrod of their survey results in the fall of 2001, Seidman and Lewis (2002) pointed out that the survey did not measure the percentage of rebate dollars spent on consumption (the marginal propensity to consume rebate dollars). They noted that Shapiro and Slemrod did not ask the household what percentage or fraction of its additional tax-home pay it intended to spend. If they had asked about the percentage or fraction, they would have received an answer about the respondent’s intended marginal propensity to consume (MPC). The answer to their question did not give the respondents’ intended MPC. Consider a hypothetical example. Suppose every household intended to spend 40%, save 10%, and use 50% to repay debt. Then 0% would answer that they intended to spend most of it. Yet in this example, the percentage of additional take- home pay households intended to spend would be 40%, not 0%; the MPC would be 40%, not 0%. Because the percentage was not asked, the study did not obtain information on the percentage of rebate dollars that households intended to spend (the intended MPC). I made this point in Seidman (2003).

H ow to C ombat R ecession

In their American Economic Review article, Shapiro and Slemrod (2003a) gave readers the impression that they had measured the MPC. They said that the 22% figure was well below what most other research would suggest for the propensity to consume for increments to income. In their AER article, Shapiro and Slemrod did not alert readers that their 22% did not mean that households intended to spend 22% of the rebate. They did not warn readers that their survey had not obtained the MPC. Recall that the Seidman and Lewis article published in 2002 emphasized that Shapiro and Slemrod’s 22%—the percentage who said they would mostly spend the rebate—was not the MPC out of the rebate. In their second article on the 2001 rebate, Shapiro and Slemrod (2003b) acknowledged this point and conceded that that their 22% was not the MPC. They included a section with the title “Converting Survey Responses into an Aggregate MPC.” They conceded that the aggregate MPC for their sample was not obtained directly from their survey. Instead, their survey provided self-reported estimates of the fraction of people who would either mostly spend the rebate or mostly save it, either by adding it to assets or repaying debt. They said they decided not to ask survey participants for their MPC because that might be “too complicated.” As a consequence of that decision, their survey was unable to provide an answer to the important question: What fraction of the total rebate dollars was spent? Having made that decision, they tried to persuade their readers that it was still possible to determine the aggregate MPC by making deductions based on “plausible assumptions.” They wrote that they could infer the MPC by making a few assumptions about what range of individual MPCs corresponds to mostly spending or mostly saving and the distribution of those individual MPCs. They mentioned the Seidman-Lewis (2002) numerical example above but called it extreme and misleading. Yet with their assumptions, they computed an aggregate MPC of 36%, just a few points below the MPC of 40% in the Seidman-Lewis example. Their statement

Do Tax Rebates Work in a Recession?

that 36% of the rebate was spent is therefore not a direct finding from their survey, but a deduction based on assumptions and a complex translation method that they called “plausible.” Let me now summarize my three objections to the wording of their question and what can be learned from their survey. First, they discouraged respondents from answering “mostly spend” by introducing the phrase “Thinking about your (family’s) financial situation,” so it is likely that the percentage who reported that they would “mostly spend” was less than the percentage who really intended to “mostly spend.” Second, they did not specify the time frame; because they used the phrase “this year,” it seems likely that with checks received in the summer, the typical respondent assumed the time frame was the rest of the calendar year—roughly four months (because rebates were mailed by the Treasury from late July through late September). Third, they did not ask what percentage or fraction of the rebate dollars the respondent intended to spend, so they had to make assumptions that would enable them to translate the answers to their questions into the MPC out of the rebate—the percentage or fraction of the rebate dollars respondents intended to spend over the next four months. The translation assumptions they made led them to an estimate of 36% for the four-month MPC, but other translation assumptions would have led to other estimates. Given these important shortcomings, my interpretation is that their survey results imply at least 40% of the rebate dollars would be spent within four months.

Johnson, Parker, and Souleles’ Study of the 2001 Rebate It took over five years from the summer of 2001 when tax rebates were mailed to households until the Johnson, Parker, and Souleles (JPS) article was published in the December 2006

H ow to C ombat R ecession

issue of the American Economic Review with the title “Household Expenditure and the Income Tax Rebates of 2001.” In contrast to Shapiro and Slemrod, who asked roughly 1,500 individuals what they planned to do with their rebates, JPS used the Consumer Expenditure Survey conducted by the US Bureau of Labor Statistics (BLS), in which about 13,000 individuals reported estimates of what they had spent on various goods and services in the preceding three months. JPS began their article by explaining that a fraction of the rebate checks were mailed out each week over three months, which enabled them to estimate the change in household expenditures due to rebate receipt by comparing the expenditures of households that received rebates in different weeks. Their study took advantage of the fact that the timing of the mailing of each rebate was based on the second-to-last digit of the Social Security number of the tax filer who received it, a digit that is effectively randomly assigned. JPS explained that the regular consumer expenditure (CE) data did not have enough information for a full analysis of the impact of tax rebates on spending because the CE survey did not record the timing of taxes and transfers within the year, nor did it record the Social Security numbers of household tax files. But immediately after the passage of the 2001 Tax Act, university economists Parker and Souleles teamed up with BLS economist Johnson and with his cooperation and participation as well as other BLS and government staff, a special module of questions about the tax rebates was added to the CE survey, thereby making the study possible. This module asked households about the timing and amount of each rebate check they received. The module was included in the survey from shortly after the rebate mailing began until the end of 2001. The JPS paper was the first to use the new tax rebate module and exploit the randomized timing of the rebates in the CE survey.

Do Tax Rebates Work in a Recession?

It took five years for Parker, Souleles, and Johnson to embark on their study in 2001, develop their module, obtain data from CE households, analyze the data, write their paper, submit it to the American Economic Review, wait for referees to review it, respond to reviewers, have their paper accepted, and then finally see it appear in print in the December 2006 issue. From the fall of 2001 to December 2006, the Shapiro and Slemrod study had left economists and policy makers with the impression that the 2001 rebate had been ineffective. Finally, in December 2006, the Johnson, Parker, and Souleles study showed economists and policymakers that, on the contrary, it was very likely that the 2001 rebate had in fact been very effective. JPS reported that the average household in their sample spent about one-third of its rebate within three months of receiving it, and about two-thirds within six months. They also reported that households with low levels of liquid assets or low income spent significantly more of the rebate than the average household. They ran regressions with the individual household’s consumption as the dependent variable and the amount of that household’s rebate as the key independent variable. They found a statistically significant short-run impact of the rebate on consumer spending. Moreover, they said that the estimated effects were economically significant, implying a substantial increase in aggregate consumption spending. In their conclusion, they said their study found significant evidence that households spent much of their rebates shortly after they arrived and that the rebates generated a substantial stimulus to the national economy in 2001, helping to end the recession because the rebates directly (ignoring multiplier effects) increased total personal consumption expenditure by nearly 1.0% in the third quarter of 2001 and 0.5% in the fourth quarter. The Johnson, Parker, Souleles study, finally published at the end of 2006, reversed the conventional wisdom that the Shapiro

H ow to C ombat R ecession

and Slemrod study had established five years earlier in the fall of 2001, and the reversal came just in time because an economic storm was brewing in 2007 in the US housing market that would soon plunge the United States into its worst recession since the 1930s.

The 2008 Tax Rebate After the bursting of the housing market bubble in 2007, the US economy began to fall into recession. In early 2008, surprisingly, tax rebates were put on the front burner in Congress. It’s not clear why. As we’ve seen, soon after the 2001 rebate checks had been sent out, Shapiro and Slemrod obtained impressive media coverage and publicity for their claim that the tax rebate failed. By contrast, Johnson, Parker, and Souleles obtained little publicity for their study concluding that the 2001 rebate had worked. Maybe some members of Congress remembered that they had passed a tax rebate in 2001 and that common sense told them that writing checks to consumers would raise consumer spending. In February 2008, with the bipartisan support of a Republican president (George W. Bush) and a Democratic Speaker of the House (Nancy Pelosi), a large bipartisan majority in Congress voted overwhelmingly to enact the Economic Stimulus Act of 2008, which devoted two-thirds of its funds to a tax rebate for households. The estimated total expenditure under the act was roughly $150 billion, and the estimated expenditure for tax rebates was roughly $100 billion. Under the tax rebate, single individuals received between $300 and $600, and married couples received between $600 and $1,200. Those with dependent children received an additional $300 per child. A broad group of households—those with an income tax liability or those with at least $3,000

Do Tax Rebates Work in a Recession?

of qualifying income—were eligible for the 2008 rebate. Qualifying income included forms of income that are not fully taxable such as Social Security benefits. The rebate was phased out at higher income levels by reducing the payment by 5% of any income above $75,000 for single filers and $150,000 for joint filers. The Treasury distributed the rebates electronically (direct deposit) if it had taxpayer personal bank account information. Otherwise it mailed a paper check. Conditional on the form of the payment, the exact timing of the payment depended on the recipient’s Social Security number. The payments, in particular the electronic fund transfers (EFTs), were distributed quickly. For tax returns processed by the IRS by April 15 for which the IRS had a private electronic routing number (not the number of the tax preparation firm) rebates were distributed via direct deposit by May 16 (i.e., over a three-week period). The distribution of rebates via paper check began in the final week of the direct deposit distribution and was completed by July 11 (a nine- week period). Before the payments, all eligible households were sent a letter from the US Treasury stating the amount of their forthcoming rebate. The total amount of rebate income paid out by the Treasury in 2008 was April, $2 billion; May, $48 billion; June, $28 billion; July, $14 billion; August–December, $4 billion, for a total of $96 billion. Technically, the rebate was a credit for tax year 2008, but to expedite the payments, the rebate amount was estimated (in terms of eligibility and for the purpose of calculating the high-income phaseout) using 2007 taxable income. If actual income in a 2008 tax return implied a higher rebate value, then the difference was paid out to the individual (or applied to any tax liabilities from the 2008 return). Individuals who were “overpaid” on the basis of their 2007 income did not have to repay the excess amount of their rebate.

H ow to C ombat R ecession

The 2008 rebates differed from the 2001 rebates in several ways. In 2001, the tax rebates were “advanced payments” of $300 for singles and $600 for couples that corresponded to the benefit from a new 10% income bracket. Hence, the 2001 rebates were linked to tax liabilities and for the most part went only to individuals with tax liabilities. The 2001 rebates went to a smaller fraction of the population, were smaller in aggregate relative to the size of the economy, and were not phased out for high-income individuals. The 2001 rebates were distributed by check over a ten-week period, so their disbursement was more spread out over time than in 2008. As in 2008, the paper checks in 2001 were mailed out on the basis of the recipient’s Social Security number. These 2008 stimulus payments were large and were disbursed quite quickly. They amounted to 8.7% of 2008 federal personal tax payments. Most of the payments were disbursed in May and June.

Taylor and Feldstein’s Wall Street Journal Op-Ed Articles During the Great Recession, the professional economist who probably did more than any other economist to try to discredit fiscal stimulus in general, and tax rebates in particular, was John Taylor of Stanford. Taylor wrote widely read op-ed articles in the Wall Street Journal, published articles in economics and policy journals, gave quotes to the media, appeared on TV, and spoke at conferences. Taylor certainly had help from other economists, including Martin Feldstein of Harvard and Shapiro and Slemrod of the University of Michigan. But in my judgment Taylor probably had the greatest impact in 2008–2009. The rebates were mailed out in the summer of 2008, and by the fall Feldstein and Taylor had separately and independently

Do Tax Rebates Work in a Recession?

published Wall Street Journal op-eds asserting that the summer and fall aggregate data for the US economy showed that the rebates had failed to stimulate consumer spending. Each asserted that aggregate time-series data from the US Bureau of Economic Analysis’s National Income and Product Accounts before and after the rebate proved that the rebate failed. At the January 2009 annual conference of the AEA, Feldstein and Taylor each presented lengthier papers claiming that the 2008 rebates had failed and argued that it would be a serious mistake to enact another tax rebate to try to combat the deepening 2009 recession. Their presentations received substantial media coverage. They published articles in May 2009, based on their January 2009 presentation, in the annual Papers and Proceedings of the AEA. Their articles and presentations were one reason that, despite the deep recession and slow recovery from 2009 through 2012, Congress did not authorize the US Treasury to mail out any more tax rebates to households. Before discussing Feldstein and Taylor, I should note how a rebate is supposed to work. After receiving a rebate check from the US Treasury, a household deposits the check and the household’s savings initially increase by the amount of the rebate. Gradually, the household spends more than it otherwise would have. Thus, immediately after a household receives a rebate check, there is a spike in saving, not a spike in spending, relative to what it would have been the case without the rebate. The key issue is the time path of consumption spending following receipt of the rebate compared to what spending would have been—in particular, the spending differential during the year following the receipt of the rebate. It is important to note that advocates of tax rebates expect recipients to use the rebate partly to pay down debt, partly to save, and partly to spend. A crucial feature of the rebate is that it lets households do some additional spending while also reducing

H ow to C ombat R ecession

their debt. This is in sharp contrast to standard monetary stimulus, which lowers interest rates to induce households to spend more by borrowing—by increasing their debt. It is hard to get households that have taken on excessive debt in the preceding boom to borrow more. Thus, a crucial difference between tax rebates and standard monetary policy is that rebates enable households to spend more while reducing their debt, whereas standard monetary policy depends on debt-burdened households being willing to take on still more debt in order to spend more. To assess the impact of any policy, a comparison is required between what actually happened after the policy was implemented, and what would have happened if the policy had not been implemented. What would have happened can only be estimated—it cannot be known with certainty. Yet Feldstein and Taylor claimed that before-and-after data prove conclusively that the rebate failed. In his WSJ article, published August 6, 2008, Feldstein asserted that the increase in consumption due to the rebate was only 15% of the rebate, a conclusion he reached simply by examining data before and after the rebate. He said the tax rebates of $78 billion arrived in the second quarter of the year, but government data showed that the level of consumer outlays only rose by an “extra” $12 billion, or 15% of the rebates, so the rest went into savings, including the paying down of debt. Notice the misleading word “extra” preceding “$12 billion.” What did Feldstein mean by “extra?” He meant that consumer outlays were $12 billion more in the second quarter than they were in the first quarter. He implicitly made the entirely unjustified assumption that, if there had been no rebate, consumer outlays in the second quarter would have been the same as in the first quarter. But his assumption (not “fact”) that outlays would have been the same in the two quarters (had there been no rebate) was implausible because, in the second quarter, house prices, stock market price indices, and the University of

Do Tax Rebates Work in a Recession?

Michigan’s index of consumer sentiment all plunged, and Bear Stearns nearly failed in March, rattling financial markets. These dramatic second-quarter negative economic events might well have caused personal outlays to shrink (had there been no rebate). If so, then Feldstein’s $12 billion is a substantial underestimate of how much the rebate raised consumption above what it otherwise would have been, given these negative events. For example, if consumption outlays would have been $13 billion lower in the second quarter than the first quarter due to these negative events (had there been no rebate), then the rebate caused consumption outlays to be $25 ($13 + $12) billion higher than what they would have been in the second quarter (had there be no rebate). In his WSJ article, November 25, 2008, Taylor addressed the impact of the mid-2008 tax rebate to households by presenting a chart that he claimed showed the tax rebate didn’t work. The horizontal axis of the chart was the months from January 2007 to October 2008. For each month two points were plotted: the higher point showed household disposable income and the lower point showed household consumption. From January 2007 to April 2008 the disposable income curve and the consumption curve grew slowly in parallel, but in May–July 2008 the disposable income curve jumped up, reflecting the payment of tax rebates in these three months, but the consumption curve did not jump up. Taylor asked readers to observe that consumption showed no noticeable increase at the time of the rebate. Taylor concluded from his chart that the rebate did little or nothing to stimulate consumption. Taylor therefore claimed that actual data from May through July proved the rebate didn’t work. But what Taylor’s chart of actual data didn’t show is what would have happened to consumption after April 2008 had there been no rebate. In mid-2008, several other influences already mentioned—the plunge in house

H ow to C ombat R ecession

and stock prices, the unprecedented high level of consumer debt, and the fall in the consumer confidence index—would likely have reduced consumption after April. Yet consumption didn’t fall until September. It is therefore plausible that the rebate kept consumption from falling from May until September when it otherwise would have fallen. Christina Romer, chair of President Obama’s Council of Economic Advisers in 2009 and 2010, provided a critique of Taylor’s WSJ article on the 2008 tax rebate in a 2011 public lecture. She pointed out that Taylor’s analysis failed to consider what besides the rebate was going on at the time. The tax rebate was enacted because Congress realized the economy was heading downward and consumption was likely to fall. The subprime mortgage crisis had taken hold. House prices were tumbling. Mortgage lenders were in deep trouble. Economists worried that consumption was about to plummet. For most families, their home is their main asset. When house prices fall, people are poorer, and so tend to cut back on their spending. Contrary to the claims by Feldstein and Taylor, Romer contended that the fact that consumption held steady around the time of the tax rebate shows just how well it was working: it kept consumption up for a while despite the strong downdraft of falling house prices. You can’t deduce the effect of a tax rebate or some other policy just by looking at outcomes. You have to think hard about what else was going on, and where the economy was heading in the absence of policy. Taylor and Feldstein didn’t.

Lewis and Seidman’s Critique of Taylor and Feldstein After asserting in their Wall Street Journal op-eds that the 2008 rebate failed, Taylor and Feldstein turned to the standard

Do Tax Rebates Work in a Recession?

regression method and report results in their 2009 AER articles. Both Taylor and Feldstein used monthly data from the Bureau of Economic Analysis (BEA) of the US Department of Commerce. In a detailed critique of these regressions, Lewis and Seidman (2012) emphasized that the rebates in 2001 and 2008 were paid out in only six months, three months in each year (a tiny amount of rebate was paid out in the month before and the month after the three months in 2001 and 2008). Hence, even though Taylor’s sample runs 106 months (from January 2000 to October 2008) and Feldstein’s sample runs a much larger number of months (from January 1980 to November 2008), they each had only six months in their sample with a tax rebate. Taylor and Feldstein both said that the rebate variable was “statistically insignificant” and that their regressions therefore proved that the rebate had no effect. Lewis and Seidman point out that “insignificant” meant only that, given the large standard error due to the small number of rebate months, the hypothesis that the true rebate coefficient was zero couldn’t be rejected; it didn’t mean the hypothesis that rebate coefficient was zero should be accepted. Taylor and Feldstein certainly knew this basic distinction between rejection and acceptance, taught in every introductory econometrics course and textbook, but they misled readers by asserting that their regressions proved that a rebate had no effect on consumption. Lewis and Seidman examined the variables in Taylor’s equation. In Taylor’s regression, the dependent variable was personal consumption expenditure (PCE), and there were only three right-hand variables: lagged PCE, rebate payments (which occurred only in three months in 2001 and three months in 2008), and disposable income excluding any rebate payments (DPY). Taylor stated that the rebate was statistically insignificant and much smaller than the statistically significant impact of disposable personal income excluding the rebate. But Lewis and

H ow to C ombat R ecession

Seidman replied that this wasn’t surprising because there were 106 months of data on disposable income but only six months of data on a tax rebate—three months in the summer of 2001, and three months in the summer of 2008—too little tax rebate data to precisely estimate its impact. Neither Feldstein nor Taylor alerted readers to the implications of the fact that they had only six months of tax rebate data. They gave the impression they were as certain about the impact of tax rebates as they were about the impact of regular disposable income, without warning readers that they a lot more data on disposable income than they had on tax rebates. Lewis and Seidman criticized Taylor for omitting several variables that were surely bringing down consumption in mid-2008 when the rebates were paid out. Housing prices were falling, home foreclosures were rising, Bear Stearns had been barely rescued in March, the index of consumer sentiment was collapsing, and the stock market was plunging. As rebate checks were being received in June, the University of Michigan’s consumer sentiment index fell to a low point of 56.4 (in contrast to its January value of 78.4) and the Dow Jones average plunged 1,288 points. All of these downward currents together certainly pulled down personal consumption expenditures. Yet Taylor apparently did not try to control for these downward currents. Lewis and Seidman found that inclusion of the consumer sentiment index nearly doubled the estimated coefficient and the t-statistic of the rebate. Inclusion of the Dow Jones average (instead of the consumer sentiment index) had an effect similar to the inclusion of the consumer sentiment index; it raised the estimated rebate coefficient and t-statistic. Finally, the inclusion of both the consumer sentiment index and the change in the Dow Jones average had a stronger effect on the rebate coefficient than either one alone. The estimated rebate coefficient became

Do Tax Rebates Work in a Recession?

slightly greater than half of the estimated disposable income coefficient. Lewis and Seidman said that the hypothesis that a rebate had roughly half as much positive impact on consumption as regular disposable income definitely could not be rejected and actually seemed roughly accurate. This hypothesis was proposed years ago by Alan Blinder (1981) based on his econometric study of the effect of temporary tax changes and transfers on consumption using aggregate times-series data. Blinder concluded that a temporary tax change had half the impact of a permanent change of equal magnitude. Next Lewis and Seidman addressed Feldstein’s regressions. In contrast to Taylor, who presented his regression results with details in a table, Feldstein (2009) provided only a paragraph. He said that he and Stephen Miran estimated a consumer expenditure equation using monthly data from January 1980 through November 2008. They estimated that the marginal propensity to consume out of real per capita disposable income was 0.70 but the estimated MPC from the corresponding rebate variable was only 0.13 (standard error 0.05). The other variables in the equation were the unemployment rate and the 10-year interest rate. In response to Feldstein, Lewis and Seidman noted that Feldstein did not include the consumer sentiment index or the change in the Dow Jones average to capture consumer anxiety. The unemployment rate variable was not a satisfactory substitute because it lagged behind the economy. For example, in May 2008 the unemployment rate was still only 5.4%. By contrast, the consumer sentiment index had already plunged. Finally, Lewis and Seidman made one other important point: neither Taylor nor Feldstein reported regressions with quarterly data, which are commonly used in macroeconometric models. Moreover, quarterly data may be preferable for testing the impact of a rebate on consumption because when a

H ow to C ombat R ecession

household receives a rebate check, it usually deposits the check, initially raising its saving, and only gradually raises its spending over the next year, so one month might not be enough time to detect the impact of the rebate on spending. Lewis and Seidman ran the same regressions with quarterly data and found that the performance of the rebate variable was stronger than with monthly data. The two-quarter MPC was fairly stable at roughly 0.28 and the four-quarter MPC was fairly stable at roughly 0.44. In summary, I reach these conclusions. Although many economists and policy makers have heard that Taylor of Stanford and Feldstein of Harvard showed that tax rebates don’t work, both of their studies were flawed. Neither Feldstein’s arithmetic in his Wall Street Journal op-ed nor Taylor’s chart in his Wall Street Journal op-ed showed that the rebate failed, contrary to their assertions. Their regressions using monthly data, reported in their articles, don’t yield a precise robust coefficient estimate for the tax rebate variable because there were very few months with a tax rebate—rebates were paid in only six months (three months in 2001 and three months in 2008) out of the 106 months in Taylor’s sample (January 2000 to October 2008). Their rebate coefficient estimates and standard errors were very sensitive to the inclusion of other plausible variables that they excluded, like the stock price index and the consumer confidence index, which both plunged during the period when the rebates were received. Lewis and I showed that the hypothesis that the true rebate coefficient is half the true ordinary disposable income coefficient could not be rejected and in fact seemed likely to be roughly accurate. Taylor and Feldstein misinterpreted their regression results and claimed erroneously that their regressions showed that tax rebates don’t work. Economists and policymakers should be informed that that Taylor’s and Feldstein’s studies of tax rebates are deeply flawed and their conclusions about tax rebates should therefore be discounted.

Do Tax Rebates Work in a Recession?

Sahm, Shapiro, and Slemrod’s Study of the 2008 Tax Rebate The claim by Feldstein and Taylor that the 2008 rebate failed was soon echoed by Shapiro and Slemrod, who in a paper presented at the January 2009 annual conference of the AEA said their new fall 2008 consumer survey showed that “only one-fifth of consumers mostly spent” the rebate. Their paper was published in May 2009 in the annual Papers and Proceedings of the AEA. With a new coauthor, they published a second article on the 2008 rebate in a 2010 volume (Sahm, Shapiro, and Slemrod 2010) on tax policy. They admitted their survey did not ask how much of the rebate was spent (the MPC), but they claimed they could estimate it was “about one-third,” using assumptions that I described earlier when they tried to deduce the MPC out of the 2001 rebate from their consumer survey questions that did not ask for the respondent’s MPC. Their MPC estimate of one-third applied to the four months remaining in 2008 after the rebate was received in the summer. But with the bias they injected by asking recipients to think about their “financial situation” instead of about goods and services they badly needed, my interpretation is that the results of their 2008 rebate consumer survey imply that over 40% of the rebate was spent within four months. Note this is the same as my interpretation of their 2001 rebate study (Seidman 2003).

Parker, Souleles, Johnson, and McClelland’s Study of the 2008 Tax Rebate With Feldstein; Taylor; and Sahm, Shapiro, and Slemrod all asserting in 2008 and 2009 that the 2008 tax rebate was largely ineffective, policymakers did not enact another tax rebate despite

H ow to C ombat R ecession

the deepest and most prolonged recession since the 1930s. With a Democratic president and Democratic majorities in the House and Senate in 2009 and 2010, significant fiscal stimulus was enacted despite near-unanimous opposition from Republicans in Congress. One of the elements of the fiscal package, the Making Work Pay tax credit, was at first glance similar to a tax rebate in trying to get more cash into the pockets of consumers, but was implemented very differently: by a cut in income tax withholding with each paycheck instead of a rebate check mailed out from the US Treasury. And in 2011 and 2012, the Republicans were willing to support a cut in the employee payroll tax withholding as a replacement for the cut in income tax withholding under the Making Work Pay tax credit. After 2012, the cut in payroll tax withholding was terminated, so that for the first time since 2008 there was no longer any tax method—either large tax rebates or cuts in withholding—being used to get more cash in the pockets of consumers during the slow recovery of 2013 and 2014. Unfortunately, it took until the October 2013 issue of the American Economic Review before Johnson, Parker, and Souleles, now joined by McClelland, published their study of the 2008 rebate using essentially the same methodology and data source that they had used to study the 2001 rebate. They explained that the rebates were paid out to households over a nine-week period and that the final two digits of a household’s Social Security number determined which week the household received its rebate. They closely followed the methodology of Johnson, Parker, and Souleles, which analyzed the 2001 tax rebates. With the cooperation of the staff of the BLS, they added additional questions to the Consumer Expenditure Survey. These supplemental questions asked CE households to report the amount and month of receipt of each stimulus payment they received. Their main finding is the opposite of the early claims of Feldstein; Taylor; and Sahm, Shapiro, and Slemrod. They

Do Tax Rebates Work in a Recession?

estimated that spending was slightly larger and faster out of the 2008 rebate than had been estimated in the study of the 2001 rebate. They estimated that within three months of receiving the 2008 rebate the average household spent about one-half of its rebate (instead of about one-third). More than half the spending was on durable goods and related services, so less than half was on nondurable goods and services. They reported that their results were economically and statistically significant across a variety of specifications. Because their estimate of spending out of the 2008 rebate was only slightly larger and faster than the estimate of spending out of the 2001 rebate, I summarize the estimates from these two studies this way: about one-third of a rebate was spent within three months, and about two-thirds was spent within six months. Johnson et al. provided some useful details about the 2008 tax rebate. To speed the payments to households, the Treasury used its data from tax year 2007 returns, so only those filing 2007 returns received the payments. If subsequently a household’s tax year 2008 data implied it was entitled to a larger payment, the household could claim the difference on its 2008 return filed in 2009. But if the 2008 data implied a smaller payment, the household did not have to return the difference. Households without tax liability received basic payments of $300 ($600 per couple) so long as they had at least $3,000 of qualifying income (which includes earned income and Social Security benefits). The total stimulus payment phased out with income, being reduced by 5% of the amount by which adjusted gross income exceeded $75,000 ($150,000 for couples). In aggregate the stimulus payments in 2008 were historically large, amounting to about $100 billion, which in real terms was about double the size of the 2001 rebate program. The stimulus payments constituted about two-thirds of the total ESA package.

H ow to C ombat R ecession

Johnson et al. noted that they generally maintained consistency with the methodology they used in their study of the 2001 rebate. Their main estimating equation had the dependent variable equal to the increase in the household’s consumption from t to t+ 1, and independent variables that included the amount of the ESP and control variables. Lags in ESP were used to allow for gradual impact over time on spending. They presented ordinary least squares (OLS) results with the dollar change in the individual household’s consumption expenditures as the dependent variable and the contemporaneous amount of the payment (ESP) to that individual household as the key independent variable. The resulting estimate of the coefficient measured the average fraction of the payment spent within the three-month reference period in which the payment was received. Johnson et al. looked carefully at subsets of consumer durables and found a key role for vehicles. Auto purchases, although weakening during the recession, would have been even weaker in the absence of payments. They noted that their finding of a large spending response on new cars implied an important role for liquidity constraints: the tax rebate may have helped provide down payments for debt- financed purchases of cars.

Broda and Parker’s Study of the 2008 Tax Rebate In contrast to the studies that used data from the Consumer Expenditure Survey of the Bureau of Labor Statistics, this study by Broda and Parker used different data, the Nielsen Consumer Panel (NCP). To measure the spending effects of the economic stimulus payments (ESPs), they conducted a survey of roughly 60,000 households in the NCP during the spring and summer of 2008. The NCP contained annual information

Do Tax Rebates Work in a Recession?

on household demographics and income, and weekly information on spending on a set of household goods. Participating households were given barcode scanners to report spending on households goods. Broda and Parker prepared a supplemental survey, designed in conjunction with Nielsen, that collected information on the date of arrival of the first ESP received by each household, as well as its amount. They identified the change in spending caused by the receipt of an ESP at the household level, using the fact that the particular week in which the funds were disbursed depended on the second-to-last digit of the taxpayer’s Social Security number. Broda and Parker estimated that the propensity to consume from a tax rebate was 30% to 45% of the rebate amount during the quarter of disbursement and 20% to 30% during the following quarter; hence, roughly one-third was spent within three months and two-thirds within six months. Households in the bottom third of the 2007 income distribution had larger propensities to spend out of their EPSs in the month of arrival than households in the top third. Households in the bottom 40% of the distribution of liquid wealth spent at roughly triple the rate of the rest of the households during the month of receipt, and at roughly double the rate during the three months starting with receipt, so that households with low liquid wealth accounted for the majority of the estimated spending response.

Parker’s Summary of Tax Rebate Research Parker (2014) summarized his research (with coauthors) on tax rebates and his conclusions for policy. He presented evidence that rebate-type payments policies generated substantial increases in demand for goods and services. In particular, a large portion of tax rebates were spent rapidly.

H ow to C ombat R ecession

Parker said that a decade ago there was general agreement that central banks should combat recessions by cutting interest rates; fiscal stimulus was unnecessary. Parker was unfortunately correct that almost no one among the leading economists at elite universities argued for fiscal stimulus in the 1990 recession or the 2001 recession—either “New Keynesians” or “New Classicals.” Unfortunately, Parker did not mention that the minority of economists who were traditional (not “new”) Keynesians in fact did call for fiscal stimulus in both of those recessions (see Seidman 2001 and 2003). Parker noted that since the plunge into severe recession in 2008, there had been some revival in interest in fiscal stimulus. Several empirical studies confirmed that transfer payments to households can provide timely and temporary economic stimulus. In 2001 and in 2008, the United States distributed more than 2% of quarterly GDP in tax rebates to the majority of US tax filers within a few months of the start of each recession. Distributing tax rebates to stimulate demand is at odds with the textbook understanding of economic behavior according to which one-time payments are not spent immediately but instead used to raise spending a small amount over many years, a lifetime, or forever. Further, many economists believed that households would assume that rebates in the present implied higher taxes in the future and would therefore save their rebates to pay the future taxes. Let me interrupt Parker to point out that traditional Keynesians have always agreed that consumers respond more to a $1,000 raise from their employer (permanent income, Friedman) than a $1,000 tax rebate in recession, and that when consumers receive a one-time tax rebate in recession they will spread some of it (life cycle planning, Modigliani). But traditional Keynesians have never accepted the claim by Barro (1974) that consumers will save the entire $1,000 tax rebate because

Do Tax Rebates Work in a Recession?

they think they better get ready to pay future taxes. Barro has never shown a shred of empirical evidence that ordinary people think this way, yet his hypothesis, called Ricardian equivalence, has been accepted by many economists and has been assumed in models presented in advanced textbooks and journal articles. Parker cited his own research with coauthors. They found that the arrival of a tax rebate caused a significant increase in spending over the next few weeks. The effect declined over time, but the cumulative effect was large. Households spent three- quarters of the rebate within three months on durables and nondurable goods and services (one-quarter on nondurables). The majority of spending was done by households with insufficient liquid assets to cover two months of expenditures (about 40% of households). These households spent at a rate six times that of households with sufficient liquid wealth.

Rebates versus Cuts in Withholding The fiscal stimulus package called the American Recovery and Reinvestment Act (ARRA) authorized federal spending of roughly $800 billion over 2009 and 2010 (hence, roughly $400 billion per year). It was enacted into law in February 2009 with virtually unanimous support from congressional Democrats and virtually unanimous opposition from congressional Republicans. Instead of enacting a 2009 tax rebate as a follow-up to the 2008 rebate, the Democrats inserted in the fiscal package a Making Work Pay tax credit recommended by the president that would give a single individual a $400 income tax credit and a married couple an $800 credit. The Making Work Pay tax credit could have been implemented as a tax rebate—the US Treasury could easily have mailed $400 or $800 checks to households in 2009 just as it mailed rebate checks in 2008.

H ow to C ombat R ecession

Instead, a reason was given for cuts in withholding from each paycheck instead of a tax rebate mailed to the household. It was argued that if small increments in funds were integrated into regular paychecks through small cuts in withholding, then employees would spend the small increments in funds the same way they spent most of their paycheck. But surely an important reason for making the switch was the incorrect claim by economists Feldstein, Taylor, and Sahm et al. that empirical studies showed that little of the 2008 rebate had been spent. I published an article in Tax Notes (Seidman 2009) that explained in numerical and diagrammatic detail how the $400 and $800 Making Work Pay tax credit would be implemented by changing specific numerical amounts withheld in each paycheck, and discussed various aspects and options for the withholding method of delivering cash to households, such as whether the amount of the withholding cuts should be changed every quarter in light of the path the economy was taking. It soon became apparent, however, that one great practical difference between rebates and withholding cuts had not been sufficiently appreciated. Tax rebate checks sent by regular mail from the US Treasury to individual households seldom escaped the notice of most recipients. Publicity occurred for weeks in 2001 and 2008 from the president and members of Congress, who wanted to make sure voters knew they had enacted the tax rebates. Also, the media told people to watch their mail for their rebate check from the US Treasury. No doubt a few people may have tossed out the envelope from the US Treasury with other junk mail. But most people in 2001 and 2008 watched their mail for the US Treasury envelope they had heard so much about, opened the envelope eagerly when it came, and deposited it quickly. Depositing an $1,800 rebate check (for two adults and two children in the household) makes the adult depositor aware that the federal tax rebate program has made the household

Do Tax Rebates Work in a Recession?

$1,800 richer—that the household has the ability to spend, pay down debt, and save a total of $1,800 more than it otherwise would have, thanks to the rebate check. By contrast, surveys showed that many individuals were completely unaware of the withholding cut due to the Making Work Pay program. This isn’t surprising. The change in the dollar amount in each paycheck was small and was often deposited directly by the employer into the employee’s bank account. Even if an individual noticed a small change in withholding or take- home pay, the reason for the small change in withholding might be unclear. The amount withheld out of a paycheck often varies by a small amount month to month for a variety of reasons. There was no easy way for even an observant individual to know why these small changes had occurred in her paycheck and whether it would be reversed in her next paycheck. Thus, even the observant individual had no way to know whether the small withholding change meant she had really become richer for the rest of year, and therefore would be able to afford to buy more goods and services. Many individuals didn’t notice the withholding change because the change did not require the individual to take any action. Either your paycheck is deposited in your bank by your employer, or you make your monthly (or biweekly) deposit of your regular paycheck that you receive once or twice every month in the mail or from your employer at work, but nothing alerts you to look at the numbers to see if they’ve changed, and if you do look at the numbers, there is no action for you to take. By contrast, if a rebate check is sent to you by regular mail, in an envelope labeled US Treasury, you must take an unusual action. You have to open an envelope you usually don’t receive from the US Treasury and then deposit this special check. It is true that some individuals did not have to take action because their 2008 rebate check was electronically deposited in their

HOW TO COMBAT RECESSION

bank account. In the future, to make sure everyone is aware of receiving a tax rebate, it would be better to have all rebate checks sent through regular mail. It is easy to alert the public to watch for a large rebate check in the mail, to explain in a short cover note that rebates are being sent only because of the recession or weak recovery, and that they won’t be repeated unless the economy becomes even weaker. That’s what happened when tax rebates were mailed out to the majority in 2001 and 2008. The Republicans, who won back a majority of the House in the November 2010 elections, were particularly eager to get rid of the Making Work Pay credit when it expired at the end of the 2010 because it had been recommended by President Obama. With a tax rebate still widely thought to be ineffective, what else might be done? Some Republicans feared being accused of levying a tax increase if they simply ended the Making Work Pay cut in withholding, thereby causing a small increase in withholding and a small cut in take-home pay. Even though many employees probably wouldn’t notice the change, Republicans worried that some would notice. So at the end of 2010, Democrats and Republicans came to an agreement to get rid of the Making Work Pay cut in withholding and replace it with a cut in employee payroll tax withholding from 6.2% to 4.2%. It was quickly agreed by both Republicans and Democrats that the loss in payroll tax revenue to the Social Security Trust Fund would be immediately made up by an equal infusion in general revenue, so there would be no change in the funds in the trust fund and no change in anyone’s Social Security benefits. One serious problem with this payroll tax cut was that it began to generate nervousness among elderly and older workers, and therefore across the political spectrum, that somewhere down the road the general revenue reimbursements might fall short of fully compensating for the lost payroll tax revenue, and that the Social Security Trust Fund and retiree benefits would eventually

Do Tax Rebates Work in a Recession?

be weakened if this “temporary” payroll tax cut continued year after year. So after being extended from 2011 to 2012, a majority in both parties agreed to end the temporary cut in the payroll tax from 6.2% to 4.2% and restore the full 6.2% starting in 2013.

The Design of a Future Tax Rebate A tax rebate is a return to the taxpayer of a portion of income, payroll, and sales taxes paid in the previous tax year. In the most recent tax rebate, 2008, the rebate was the same dollar amount per adult ($600), and per child ($300), for most households. Based on the income tax return of the previous year (2007), the rebate was phased down and out for high-income households, phased down for low-income households that had paid less income tax than the standard rebate amount (the rebate could not exceed the amount of income tax the household had paid), and there was no rebate for households with very low income (less than $3,000). In my view there is an economic and political advantage in giving every household a rebate, though not necessarily the same dollar amount. Almost every household has borne a burden from some tax—income tax, payroll tax, and/or sales tax. A valid way to make more low-income households eligible for a tax rebate is to recognize the burdens from payroll and sales taxes. It is impractical and unnecessary for the IRS to try to determine each household’s exact payroll tax or sales tax burden in the preceding year. Using the wage income reported on the household’s 1040 return, the IRS can take 12.4% of the household’s wage income up to last year’s Social Security ceiling as the IRS’s estimate of the household’s Social Security payroll tax burden; and take 2.9% of the household’s wage income as the IRS’s estimate of the household’s Medicare payroll tax burden.

H ow to C ombat R ecession

The IRS should use 12.4% rather than 6.2%, and 2.9% rather than 1.45%, because economic analysis implies that workers bear nearly the entire burden of the payroll tax because employers reduce wages in order to pay the employer payroll tax. The IRS can use 5% of total income reported on the 1040 as its estimate of the household’s sales tax burden; even in states without a sales tax, prices may be bid up as border shoppers choose to shop across the border so that shoppers bear some burden from a sales tax in adjacent states. Although recognizing payroll and sales tax burdens will make most low-income households eligible for a substantial tax rebate, some may still have zero tax burden in the preceding year. In my view it would be worth setting a minimum tax rebate for all households even if that minimum amount exceeds the estimated total tax burden of the household. I also recommend, in contrast to the tax rebate of 2008, phasing down but not out the tax rebate for high (and even very high) income households. It is true that high-income households usually have a lower propensity to consume than low-income households. But there is evidence that most high-income households do have a positive marginal propensity to consume. Some even have a propensity as large as the typical low-income household (Kaplan, Violante, and Weidner 2014). Sending tax rebates to the affluent, the poor, and everyone in between during a recession should broaden the political support for tax rebates. To make sure that every household recognizes that it has received a tax rebate, I recommend that all tax rebates be sent by regular US mail rather than deposited electronically in bank accounts. A check in the mail requires the recipient take several concrete actions: opening the envelope, recognizing that a check is enclosed, looking at the short letter explaining why the check has been sent, and then depositing the check in a bank account or cashing the check. There is bound to be media

Do Tax Rebates Work in a Recession?

coverage of individuals opening envelopes from the US Treasury containing their rebate check, and the media will communicate the fact that every household will be receiving a rebate check in the mail. This widespread awareness of tax rebates should boost support for the tax rebate program.

Conclusion The impact of tax rebates in the 2001 recession and in the 2008 recession has been subject to a set of studies by highly respected economists. The three studies that in my view used the best methodology and data to estimate the impact of tax rebates on consumer spending in a recession (Johnson, Parker, and Souleles 2006; Parker et al. 2013; Broda and Parker 2014) estimated that tax rebates were very effective in stimulating consumer spending: about one-third was spent within three months and about two-thirds was spent within six months. By contrast, the studies that concluded that these rebates had little effect had serious flaws, as explained in detail in this chapter. Whenever the US economy is hit with a recession, tax rebates should be the mailed out with a size that fits the severity of the recession. Suppose a recession reduces real GDP by 6% below normal, which is estimated to raise the unemployment rate by three percentage points—for example, from 5% to 8%. To simplify the calculation, assume that the fiscal stimulus consists entirely of tax rebates (in the next chapter I will make the case for including several other programs in the fiscal stimulus package). To raise GDP by 6% would require total tax rebates equal to 6% of GDP, assuming a tax rebate multiplier of 1.0 in a recession. If the rebate is to be paid out twice a year—for example, in June and December, then each payment would be 3% of GDP. Suppose in 2018 potential GDP is $20,000 billion and a recession hits.

H ow to C ombat R ecession

This would require total tax rebates by the US Treasury in June, and then in December, of 3% of $20,000 billion, or $600 billion in June, and then $600 billion in December. With a US population of roughly 300 million, the average rebate in June and in December would be $2,000 per person. Congress might specify that the rebate to each household would be $2,667 per adult and $1,333 per child in June and in December, so a household with two adults and two children would receive an $8,000 tax rebate check in June and in December. Under stimulus without debt, the Fed would make a transfer to the Treasury of $600 billion in June and $600 billion in December, so that the Treasury does not have to borrow to finance the tax rebates.

Chapter 5

What about Other Kinds of Fiscal Stimulus?

In this book I argue that tax rebates to households should be the main component of a fiscal stimulus package in a severe recession for the reasons and evidence given in c hapters 3 and 4. In this chapter, I consider other fiscal stimulus programs that might be included with tax rebates in a fiscal stimulus package. The chapter begins with several fiscal stimulus programs that I recommend for inclusion. Then I examine several other fiscal stimulus programs and explain why I don’t recommend them. Next I review and comment on components of the fiscal stimulus enacted in early 2009 and implemented in 2009 and 2010—the American Recovery and Reinvestment Act (ARRA). I then explain a common mistake made by many economists that leads them to underestimate the “multiplier effect” of fiscal stimulus programs. Finally, I report on and evaluate several empirical studies of fiscal stimulus programs utilized in the Great Recession.

H ow to C ombat R ecession

Recommended Supplements to Tax Rebates In this section I recommend several fiscal stimulus supplements to tax rebates for inclusion in a fiscal stimulus package: a temporary increase in federal aid to state governments, temporary tax incentives for business investment, and infrastructure repairs and maintenance.

A Temporary Increase in Federal Aid to State Governments I recommend a temporary increase in federal aid (grants) to state governments. A formula should be used for distributing the aid across states. In a recession, most of the aid will be spent, not saved, thereby helping to combat the recession. A recession automatically reduces tax revenue at all levels of government— federal, state, and local. If the governmental unit responds by cutting its spending or raising its tax rates in an effort to keep its budget balanced, it will further reduce aggregate demand for goods and services, thereby making the recession worse. Just as tax rebates help sustain consumer spending when the recession reduces household income, so temporary federal grants to state governments, which in turn provide temporary grants to local governments, should help sustain state and local government spending, and also help sustain consumer spending by reducing state and local tax increases needed to balance governments’ budgets. Moreover, federal aid to state governments should be (as it usually is) implemented according to an objective numerical formula based, for example, on the state’s population, income per capita, unemployment rate, and so on. Such a formula spreads the federal aid across all states. If the elements included in the formula are generally perceived as fair, federal grants have the

W hat about Other Kinds of Fiscal Stimulus?

potential to win the support of a majority of voters, thereby generating support for passage of the fiscal stimulus package by Congress. Use of a formula minimizes favoritism and pork- barrel politics. Politics will certainly play a role in choosing the elements that enter the formula and the numerical weights given to each element. But once the formula is set, the funds going to each state are not affected by political dealing and lobbying. Another simple method would be to raise the federal matching rate for Medicaid (the “Federal Medical Assistance Percentage,” FMAP). This was done under the fiscal stimulus bill enacted in February 2009, which will be described below. Medicaid is a matching federal/state program, and Medicaid constitutes a large percentage of state government spending in most states. During a recession, more households qualify for Medicaid. Raising the federal matching rate is a simple and quick way to inject federal funds into state governments. Even these simple methods of injecting federal funds into state governments are subject to objections by critics. Some critics believe that just as a recession forces households to “tighten their belts” and cut spending, so all levels of government should have to “tighten their belts” and cut spending. Of course, tax rebates attempt to ease the belt tightening of households. Some critics of government “waste” believe a recession is a good time to squeeze some of the waste out of government. Some conservative critics hold that recession is a good time to reduce the size of government at all levels—federal, state, and local. If the federal grants are specifically targeted to a large list of particular programs and projects, then there is criticism of favoritism and pork-barrel politics; many opponents of the 2009 fiscal stimulus package made this criticism. Some opponents have argued that federal grants will be almost entirely saved by state governments, just as some opponents of tax rebates have argued that rebates are almost entirely saved. But one of the empirical studies reviewed

100

H ow to C ombat R ecession

at the end of this chapter provides an estimate of how much is saved, and this study estimates that a substantial percentage of the grant is not saved. As opponents point out, there is some risk that the federal grants that are intended to be temporary will not be ended when the economy recovers from the severe recession. In both the 2001 and 2008 recessions, however, temporary federal grants to states were in fact ended once recovery began. The grants were enacted for a particular year, and once recovery occurred, Congress did not renew them for the next year. Ideally, it might be better to enact the grants with a clause that automatically winds down and terminates the grant according to an indicator such as the unemployment rate. But past behavior by Congress shows that such a “terminator clause” may not be necessary.

Temporary Tax Incentives for Business Investment Tax rebates to households, by raising consumer spending, should indirectly cause an increase in business investment because firms producing consumer goods and services will need to expand their capacity to meet the rise in customer demand. It might be argued that tax rebates for households provide sufficient stimulus for business investment. Nevertheless, there are several good reasons why a tax incentive for firms to invest in plant and equipment should be included in a fiscal stimulus package. Over the past half-century, two tax incentives encouraging business firms to buy plant and equipment have been enacted in the United States: (1) an investment tax credit; (2) bonus depreciation. Under the first, a firm receives a tax credit on its income tax equal to a certain percentage of its expenditure on an investment good—in effect, a firm is reimbursed by the federal for that percentage (for example, 20%) of its expenditure. Under the second incentive, a firm is permitted to depreciate a certain

W hat about Other Kinds of Fiscal Stimulus?

101

percentage (for example, 50%) of its expenditure on an investment good on its income tax in the year of purchase in addition to its regular depreciation, thereby reducing its taxable income and its tax that year. If Congress uses the word “temporary” in the title of the legislation for business tax incentives, it puts pressure on business firms to act during the recession to take advantage of the temporary cut in the price of investment goods while it lasts. There are two reasons for including temporary tax incentives for business investment in a stimulus package. First, most economists agree that a temporary cut in the price of investment goods should, in theory, induce some increase in purchases. Second, inclusion of a program that directly sends funds to business firms that invest may improve the political feasibility of the fiscal stimulus package. For example, the 2008 fiscal stimulus package contained two components: a tax rebate that directly sent funds to households (two-thirds of the package), and a bonus depreciation tax incentive that directly sent funds to business firms (one-third of the package). The 2008 fiscal stimulus package passed with bipartisan support. There is a risk, however, that advocates of permanently lower tax rates on business firms will press to continue the tax incentive even after a strong recovery from the recession has been achieved. The effectiveness of the tax incentive to stimulate investment spending during the recession will be weakened if business managers believe it will be permanent, not temporary, just as a 50% price cut during a sale at a retail store will stimulate less consumer spending if consumers believe the price will be permanent, not temporary. It might be particularly important to include language indicating that the incentive is temporary, intended only to combat the recession, or even include a terminator clause that phases it out as the unemployment rate declines to normal.

102

H ow to C ombat R ecession

Infrastructure Repairs and Maintenance There is strong bipartisan support in all sections of the country for infrastructure spending to repair and maintain current roads and bridges. For federal infrastructure this would involve direct federal spending; for state and local infrastructure, this would involve federal grants to state governments earmarked for this purpose. This program would promptly increase spending in every state, but this spending directly increases demand in only certain sectors of the economy—construction firms and workers and the materials they use. By contrast, tax rebates spread their increase in aggregate demand across most goods and services in the economy, and federal aid to states stimulates all state and local government spending, not just infrastructure maintenance. Thus, infrastructure spending for repairs and maintenance is a useful small complement to, but not a substitute for, a large tax rebate or federal aid to states. A fiscal stimulus bill in a severe recession is not the place to put new, costly, controversial long-term infrastructure projects. In contrast to repairs and maintenance, new infrastructure projects involve significant delays in starting construction. Each of those projects should be subjected to cost-benefit analysis, should not be judged as an antirecession measure, and should be voted on separately from the fiscal stimulus package.

Not Recommended In this section I explain why I do not recommend the following in a fiscal stimulus package: a temporary cut in income tax withholding, a temporary cut in payroll taxes, a cut in income taxes, or an increase in spending on long-term programs. Two different criteria guide the decision not to recommend the following policies for inclusion in a fiscal stimulus package in

W hat about Other Kinds of Fiscal Stimulus?

103

a severe recession. First, the tax rebate to households dominates the other ways listed below to get cash to households to increase their consumption spending. Second, the fiscal stimulus package devised to combat a severe recession should not be used to achieve and implement a long-term agenda by either liberals (Democrats) or conservatives (Republicans). Instead, the aim should be to construct a fiscal stimulus package that is effective against the severe recession and neutral concerning the long- term agenda of each side, so that the package can be enacted quickly with bipartisan support and renewed if the recession is more severe than expected. The place and time for partisan debate over long-term agenda policies should be outside the fiscal stimulus package and after it is enacted.

A Temporary Cut in Income Tax Withholding Mailing out tax rebate checks to households is better than a cut in income tax withholding. A tax rebate was enacted only in the first year of the Great Recession, 2008. In 2009, the new president was a Democrat and Democrats had a majority of both the House and Senate. Together the Democrats fashioned and passed the American Recovery and Reinvestment Act, which contained many fiscal stimulus programs and will be discussed more fully later in this chapter. The bill was passed over the virtually unanimous opposition of Republicans in the House and Senate. Instead of including a repeat of the 2008 tax rebate to households as one component of ARRA, the Democrats replaced it with a program advocated by candidate Obama during his campaign in 2008 called “The Making Work Pay Tax Credit.” Like a tax rebate, the Making Work Pay Tax Credit provided more cash in the pockets in a majority of the nation’s households in the hope that they would use a significant portion of the cash to buy goods and services, thereby stimulating production of these goods and

104

H ow to C ombat R ecession

services. Most households received a credit of $800. Like the 2008 tax rebate, the MWP tax credit excluded the affluent. There was, however, one crucial difference between the Making Work Pay Tax Credit of 2009 and the tax rebate of 2008. The tax rebate of 2008, like the tax rebate of 2001 and 1975, was a single large check sent out in an envelope by regular mail to each household by the US Treasury (though a minority of households did receive their rebate through an electronic deposit from the Treasury to the individual’s bank account). By contrast, the Making Work Pay Tax Credit was paid out in small amounts in each worker’s paycheck by cutting the income tax withheld by the employer. Most households that received the tax rebate in 2008 were aware of receiving the large rebate check in the mail and depositing it in their bank, so most knew how much they had received from the rebate. They received a cover letter telling them this money came from the president and Congress. They knew that if their rebate check was $1,200, they were $1,200 richer and could afford to increase their spending up to $1,200 without borrowing. By contrast, in 2009 most households were not aware of receiving a small cut in income tax withholding due to the Making Work Pay Tax Credit, and therefore a small increase in take-home pay. Small changes in withholding, and in take-home pay, occur for many reasons, and surveys showed that many workers did not even notice the change, or if they did notice, did not realize it was due to the MWP tax credit. Most important, most did not feel richer, and therefore did not think they could afford to do $800 more in spending, saving, and paying down debt.

A Temporary Cut in Payroll Taxes Mailing out tax rebate checks to households is better than a temporary cut in the employee Social Security payroll tax rate. When

W hat about Other Kinds of Fiscal Stimulus?

105

the Making Work Pay Tax Credit expired at the end of 2010, the Republicans had regained a majority in the House and were opposed to renewing the MWP tax credit. Instead of turning back to tax rebates, the two parties agreed on a two-percentage-point cut in the Social Security payroll tax for employees, from 6.2% to 4.2% in 2011. Over a year’s time, an employee earning $40,000 would receive $800 more in take-home pay. Like the MWP tax credit, the cut in the employee payroll tax rate was implemented by a small reduction in withholding in each worker’s paycheck and thus a small increase in take-home pay in each paycheck. As in the case of the cut in income tax withholding, surveys showed that many workers were unaware of the change in take- home pay, or if they noticed the change, did not know why it had changed. Many employees did not feel $800 richer after a year had passed than they would have felt without the reduced tax, and they may not have felt they could afford to do $800 more in spending, saving, and paying down debt. For policymakers and for citizens who were aware of the cut in payroll taxes, moreover, the cut raised concern about whether it would weaken the finances of Social Security and the benefits of current or future retirees. The legislation enacting the two- percentage-point cut in the payroll tax stipulated that general revenue would annually replace the reduction in payroll tax revenue coming into the Social Security Trust Fund so that the finances of the fund would remain unchanged and benefits to all Social Security recipients current and future would be unaffected. But some were concerned that eventually political pressure to use general revenue to fund other programs might reduce the funds flowing into the Social Security Trust Fund. Although the payroll tax cut was renewed in 2012, it was allowed to expire at the end of 2012, so that in 2013 the payroll tax rate reverted from 4.2% to 6.2%. This relieved the fears of those who worried, during 20111 and 2012, about the financing of Social Security.

106

H ow to C ombat R ecession

A Cut in Income Taxes Most conservatives and Republicans support a permanent cut in income tax rates as a central element in their long-term agenda of reducing tax revenue and government domestic spending as a percentage of GDP and giving permanent incentives for people to supply more labor and capital to the economy because they keep more of what they earn. Many liberals and Democrats oppose a permanent cut in income tax rates because it would prevent a central element in their long-term agenda of strengthening social insurance and education programs. To combat a severe recession, mailing out tax rebate checks is better than a cut in income tax rates, which should be debated between conservatives (Republicans) and liberals (Democrats) as a matter of long-term policy.

An Increase in Spending on Long-Term Programs Many liberals and Democrats support a permanent increase in spending on social insurance, education programs, and environmental projects as a central element in their long-term agenda of providing the general public with more protection and opportunity, but many conservatives and Republicans oppose this role for government and the higher taxes it requires. These permanent spending increases should be debated as long-term policies, not as elements of a temporary fiscal stimulus package to combat a severe recession.

Rebates and Bonuses in 2008, ARRA in 2009–2010, and Fiscal Stimulus in 2011–2012 In February 2008 a $150 billion fiscal stimulus was enacted— 1% of the GDP of $15,000 billion—consisting of $100 billion of tax rebates to households mailed out in the summer, and $50

W hat about Other Kinds of Fiscal Stimulus?

107

billion of bonus depreciation for business firms to encourage investment spending. It was enacted with bipartisan support and signed by President Bush. In the fall of 2008, as the economy plunged into deep recession following the failure of Lehman Brothers, President Bush failed to propose, and Congress failed to enact, another fiscal stimulus, despite urgent newspaper columns from some economists, including me. In my judgment, this was a major policy failure that has not received the attention it deserves. The American Recovery and Reinvestment Act (ARRA) was enacted in February 2009 just one month after a new Democratic president, Barack Obama, took office, with Democrats having a majority in the House and nearly (but not quite) sixty votes (out of one hundred) in the Senate. Democrats therefore had the votes to pass a stimulus bill containing components that were consistent with their long-term objectives and therefore almost unanimously opposed by Republicans. ARRA included several government spending programs favored by Democrats as permanent policies. In the 2001 recession, when a Republican occupied the White House and Republicans had a majority in Congress, the stimulus bill promoted the Republicans’ long-term agenda by enacting long-term (10-year) cuts in all personal income tax rates, with larger tax cuts, as measured by the total dollar amount, going to high-income households. In defense of the Democrats’ stimulus bill, it must be emphasized that the Republicans showed no interest in a fiscal stimulus package in the fall of 2008, especially one containing any government spending. The Democrats recognized that it was up to them to shape a fiscal stimulus bill and pass it without Republican votes. Some Democrats also wanted to use the fiscal stimulus to advance their long-run agenda. Here is my rough estimate of the impact of fiscal stimulus (tax rebates and bonus depreciation) of 1% of GDP in 2008 and ARRA in 2009 and 2010. ARRA was enacted in February 2009

108

H ow to C ombat R ecession

and began entering the economy in April 2009. The US unemployment rate had already shot up from 5.0% in April 2008 to 9.0% in April 2009. ARRA was a roughly $750 billion fiscal stimulus package implemented mainly over two years (2009 and 2010)—roughly $375 billion per year, or 2.5% of GDP ($15,000 billion) per year. Thus, the average fiscal stimulus per year from 2008 through 2010 was 2% of GDP (the average of 1%, 2.5%, and 2.5%). Conservatively assuming a fiscal stimulus multiplier of 1.0, GDP was raised an average of 2% per year above what it otherwise would have been. Assuming a ratio between the increase in GDP and the reduction in the unemployment rate of 2, fiscal stimulus reduced the unemployment rate one percentage point below what it otherwise would have been; with fiscal stimulus the unemployment rate reached a peak of 10% in October 2009, but without fiscal stimulus, the unemployment rate would have peaked at 11%. The message from this calculation is that the fiscal stimulus certainly helped, but was simply too small for the Great Recession. At the time, some economists, including me, wrote columns calling for fiscal stimulus at least twice as large. I now think the stimulus should have been three times as large. If fiscal stimulus had been 6% of GDP in 2008, 2009, and 2010 instead of averaging 2% per year, the unemployment rate would have been reduced 3%, not 1%, from 11% to 8%, not from 11% to 10%. If further strong stimulus had been continued in 2011 and 2012, instead of being blocked by the House Republican majority that was elected in November 2010, a strong recovery would have been generated that would have brought the unemployment down to 5% by 2013 instead of 2015, as was actually the case. In its account five years after ARRA’s enactment, the Obama administration’s Council of Economic Advisers (CEA) presented its analysis of the impact of ARRA (chapter 3 of The Economic

W hat about Other Kinds of Fiscal Stimulus?

109

Report of the President 2014). The half of professional economists (including me) who believe that fiscal stimulus works in a recession find the CEA analysis of the ARRA’s impact on the economy largely accurate, but the other half of professional economists who believe that fiscal stimulus does not work in a recession do not. Most economists, however, would find the information about the ARRA provided in the CEA’s chapter to be very useful. The chapter begins with an important fact: President Obama signed the ARRA on February 17, less than a month after taking office. I want to emphasize that the speed with which the new Democratic majority in the House and Senate enacted the ARRA stands as a refutation of the conventional view that fiscal stimulus always takes many months to enact. There is a common impression that the two-year ARRA, implemented in 2009 and 2010, was the entire fiscal stimulus implemented during the Great Recession. But the CEA pointed out that this impression is incorrect. In the four years following passage of the ARRA, the president signed into law over a dozen fiscal measures aiming to speed job creation. Critics have claimed that the fiscal stimulus (the ARRA and subsequent fiscal stimulus measures) enacted during the Great Recession was wasteful government spending. But the CEA replied that nearly half of the job measures in the ARRA and subsequent legislation were tax cuts—with most of them directed at families. The other half was for rebuilding bridges and roads, supporting teacher jobs, or providing temporary help for the unemployed. The CEA provided additional information about the ARRA. The initial cost projections showed the law would be fairly evenly distributed across tax cuts ($212 billion billion), expansions to mandatory programs such as Medicaid and unemployment benefits ($296 billion), and discretionary spending ($279 billion) in areas ranging from direct assistance to individuals to

110

H ow to C ombat R ecession

investments in infrastructure, education, job training, energy, and health information technology. Although over a fourth of ARRA stimulus was tax cuts to individuals, most people were unaware of this component—the Making Work Pay Tax Credit— because it was implemented through small cuts in withholding in each paycheck. The CEA provided this additional information about the subsequent fiscal stimulus measures. There is a common impression that fiscal stimulus ended after 2010 because the Republicans, who had criticized fiscal stimulus, regained a majority in the House in the November 2010 election. But the CEA report showed that this was not so. The Obama administration succeeded in getting Republicans in the House to maintain roughly 80% as much fiscal stimulus as the ARRA (2% of GDP per year instead of the ARRA’s 2.5%) in 2011 and 2012. While the ARRA was the first and largest fiscal action undertaken after the financial crisis to create jobs and strengthen the economy, many subsequent actions extended, expanded, and built on the ARRA. Parts of the act were extended to address the continuing needs of the economy, including Emergency Unemployment Compensation, accelerated depreciation of business investment for tax purposes (that is, “bonus depreciation”), measures for teacher jobs, and aid to states for Medicaid. In other cases, new measures expanded on elements in the ARRA, such as the temporary payroll tax cut in 2011 and 2012, which was nearly 50% larger than the Making Work Pay credit it replaced, and an even greater allowance for businesses to write off the cost of investments when computing their tax liability (that is, “expensing”). There is also a common impression that the Obama administration never asked for an enlargement of fiscal stimulus after ARRA. But this is untrue. The president proposed further measures for the economy, most notably the American Jobs Act, which was proposed in a special speech to a joint session of

W hat about Other Kinds of Fiscal Stimulus?

111

Congress in September 2011 and would have provided additional investments in everything from infrastructure to teacher jobs to a robust tax credit for small business hiring. But the House was unwilling to pass it. The CEA said that most studies have found that fiscal stimulus (the ARRA and subsequent measures) helped stop the plunge in the economy and helped generate the economic recovery (several empirical studies of the impact of the ARRA will be examined later in this chapter). Most studies found that the ARRA substantially boosted employment and output. The CEA estimated that, by itself, the act saved or created an average of 1.6 million jobs a year for four years through the end of 2012. In addition, the ARRA alone raised the level of GDP between 2% and 2.5% from late 2009 through mid-2011. In a future recession, in contrast to the ARRA, it would be better to enact a fiscal stimulus package that is neutral with respect to the long-term agenda of conservatives or liberals, because such a package can be enacted even when there is divided government, with no single party (Democrat or Republican) having the presidency and a majority in the House and sixty (enough to prevent a filibuster) out of a hundred senators. It would be better to establish a tradition that a fiscal stimulus package to combat a severe recession is not the time or place to push for measures that favor only one party’s long-term agenda. It is, of course, politically tempting for a party that controls the White House and Congress to push measures that advance its long-term agenda. But this guarantees that the other party will attack the stimulus package and try to discredit it with the public. It is certainly debatable whether in early 2009 Republicans and Democrats would have been willing to put aside long- term priorities and join in supporting the neutral fiscal stimulus package advocated here. Whose fault was it that the fiscal stimulus bill that was actually enacted—the ARRA—favored

112

H ow to C ombat R ecession

the long-term priorities of Democrats? Some Republicans insist they would have supported a politically neutral package but Democrats preferred a partisan bill. Other Republicans admit they would have opposed any fiscal stimulus bill because of their worry about government deficits and debt. In the second half of 2008, as the recession deepened, it is a sobering fact that neither the Republican president nor Republicans in Congress proposed a large fiscal stimulus package. After the November 2008 election, Democrats felt it was urgent to move quickly to enact a strong fiscal stimulus package, and that time would be wasted trying to negotiate a bipartisan, strong, neutral package with Republicans. Although the ARRA was not ideal in its composition, in my view the evidence is strong that it was effective: without it the recession would have been deeper and lasted longer. However, it is also likely that the stimulus package would have been larger, and would have been sustained for four years instead of two, and therefore have been much more effective, if it had been neutral toward long-term priorities and if both parties had been willing to put aside long-term priorities and concerns about government deficits and debt to pass a neutral package.

A Common Mistake in Estimating Fiscal Multipliers Some economists have contended that the CEA assumed fiscal stimulus multipliers larger than have been found in pre-2008 empirical studies of the quantitative relationship between fiscal stimulus and the resulting increase in output. Unfortunately, many (not all) of these studies have made a fundamental mistake in estimating multipliers—a mistake that results in a significant underestimate of the effectiveness of fiscal stimulus in a severe

W hat about Other Kinds of Fiscal Stimulus?

113

recession when the “output gap” is large—that is, when actual output is far below potential output (the output that would be produced if labor were fully employed and the real capital stock were used to full capacity). A fiscal stimulus multiplier is the ratio of the increase in output to the increase in government spending or tax cut that generates it. Assume the economy is at full employment and full capacity utilization when the government increases spending or cuts taxes. With hardly any unemployed labor or capital available, real output hardly increases in response to an increase in government spending or a tax cut, so the multiplier is near zero. By contrast, suppose the economy is in a severe recession with high unemployment and low capacity utilization. Then the increase in spending or tax cut would cause employers to hire unemployed workers and utilize idle machines, thereby increasing real output; hence, the multiplier would be positive. Moreover, the newly employed would enjoy an increase in income, enabling them to raise their consumption spending, inducing producers of consumer goods to hire unemployed workers, utilize idle machines, and raise real output, thereby making the multiplier larger. What matters for fiscal stimulus to combat a recession is the size of the multiplier in a recession when unemployment is high and capacity utilization low, not the size of the multiplier in a fully employed economy. Consider the textbook aggregate-demand/aggregate-supply diagram shown in Figure 5.1, in which the supply curve is initially flat but curves upward to become steep at the full-employment level of real output Y0 (Seidman 2012a, 2012b). When the economy is in severe recession at a low value of real output, Y1, with high unemployment and low capacity utilization, a shift right of aggregate demand (D) can raise real output with only a slight bidding up of wages, costs, and prices, so the aggregate supply (S) curve is relatively flat. But when the economy is at full

114

H ow to C ombat R ecession S

D' D D' D Y1

FIGUR E 5.1 The Stimulus Multiplier in Recession vs at

Full-Employment

employment output Y0, a shift right of aggregate demand curve mainly bids up wages, costs, and prices, with hardly any increase in real output Y, so the aggregate supply curve is steep. Thus, when the economy is in a severe recession, a shift right of the D curve by magnitude ∆D causes a relatively large increase in real output ∆Y—from Y1 to Y2—hence, the multiplier is large—but when the economy is at full-employment output Y0, a shift right of the D curve by the same magnitude ∆D causes a relatively small increase in real output ∆Y—hence, the multiplier is small. Thus, it is a fundamental error to estimate the value of the multiplier in a fully employed economy and then assume this value holds when the economy is in or still recovering from a severe recession. Yet Robert Barro made exactly this error. He tried to estimate a value for the multiplier using data from a fully employed economy and then asserted that this multiplier value would hold when the economy was in a severe recession. Barro summarized

W hat about Other Kinds of Fiscal Stimulus?

115

his research in a truly revealing article in Economists’ Voice entitled “Voodoo Multipliers” (2009). He argued that the best evidence of the size of the multiplier came from the massive expansion of US defense expenditures during World War II. There was a dramatic reduction in the US unemployment rate from 1939 through 1942 driven partly by the sharp rise in military and related spending in preparation for a possible entry into a world war from 1939– 1941 and actual entry in 1942. An analysis confined to 1939 to 1942 might therefore have been useful. Instead, Barro focused on 1943–1944 when the US economy was at full employment. He estimated that World War II raised US real defense expenditures by $540 billion (1996 dollars) per year at the peak in 1943–1944, amounting to 44% of trend GDP. He also estimated that the war raised real GDP above trend by $430 billion per year in 1943– 1944. Thus, the multiplier was 0.8 (430/540). Barro is not the only economist who estimated the magnitude of the fiscal stimulus multiplier in a recession by using data generated in an economy that was not in recession. There is a large empirical literature that presents regressions of the change in real output against the change in government spending or taxes for all quarters in the sample. But only a small fraction of the quarters in the sample were recession quarters, so these studies mainly estimate the value of the multiplier when the economy is not in recession. What matters for countercyclical fiscal policy, however, is the value of the multiplier when the economy is in or still recovering from recession so there is still substantial slack in the economy. One empirical study (Auerbach and Gorodnichenko 2012) focused specifically on distinguishing between the numerical value of the government purchases multiplier in a recession versus its value in an expansion. Using a regime-switching model, they found large differences in the size of spending multipliers in recessions and expansions, with fiscal policy being much more

116

H ow to C ombat R ecession

effective in recessions. Their finding was consistent with the traditional Keynesian explanation: when the economy has slack, expansionary government spending shocks are unlikely to crowd out private consumption or investment because idle labor and capital can be brought into production. They criticized the majority of empirical studies of fiscal multipliers that don’t provide separate estimates for multipliers in recession, noting that much empirical research in this area is based on models that assume that multiplier values don’t depend on the state of the economy. They described exactly my AD-AS diagram in Figure 5.1. The AS curve is flat at low levels of output (Y)—recession—but becomes steep at high levels of output (Y)—full employment of labor and utilization of capital. When AD shifts right in recession, it generates a large increase in Y; but when AD shifts right at full employment, it generates a large increase in P but only a small increase in Y. They found that the size of the size of the multiplier varied considerably over the business cycle. Typically, the multiplier is between 0.0 and 0.5 in expansions and between 1.0 and 1.5 in recessions. A dollar increase in government spending in 2009 during the Great Recession would raise output by about $1.75. Auerbach and Gorodnichenko (2013 extended their analysis of the government purchases multiplier to other OECD countries, and found that multipliers of government purchases were larger in a recession. They concluded that their results were consistent with the traditional Keynesian model, but not with New Keynesian or New Classical models. They noted that the New Keynesian model doesn’t have spare capacity or slack, a key feature of the traditional Keynesian model. They observed that during the Great Recession, many countries adopted expansionary fiscal policies aimed at counteracting the large negative shocks to their economies, despite of skepticism among many economists about the potential of fiscal policy to stimulate

W hat about Other Kinds of Fiscal Stimulus?

117

economic activity. Their two studies found that fiscal stimulus worked in recession—it raised mainly real output, not inflation, and multipliers were large. Their empirical results contradict the assumption of New Keynesian models that crowding out is large and the fiscal multiplier is small. Instead, their results are consistent with the traditional Keynesian model, which assumes a recession is characterized by “slack”—labor that is involuntarily unemployed and machines that are idle.

Empirical Studies of the ARRA Fiscal Stimulus (2009–2010) I will now review several empirical studies of fiscal stimulus using data from 2009 and 2010. Recall that the American Recovery and Reinvestment Act was enacted in February 2009 and its funds were injected into the economy from spring 2009 through the end of 2010 (with some spending continued beyond 2010).

Fiscal Spending Jobs Multipliers: Evidence from the 2009 ARRA This subsection heading reproduces the title of an article by Daniel Wilson (2012). Under the ARRA, federal grants to states varied according to a formula. Wilson applied regression analysis to a cross section of states to determine to what extent jobs in a state were affected by ARRA funds received in that state from the federal government. His strategy was to exploit the cross- sectional geographic variation in ARRA spending and the factors that determined that variation. Wilson explained that, in contrast with studies of the economic effects of fiscal stimulus that rely on time-series variation, the use of cross-sectional variation greatly reduces the risk

118

H ow to C ombat R ecession

of confounding fiscal stimulus effects with effects from other macroeconomic factors, such as monetary policy, that are independent of the geographic distribution of stimulus funds. His study made use of the fact that most of the ARRA funds were allocated according to statutory formulas based on exogenous factors such as the number of miles of highway in a state and the youth share of its population. Wilson noted that his study measured local multipliers and did not include reverberations outside the region. Wilson explained how his methodology differed from two other methodologies that have been used to measure the impact of ARRA on jobs. The first methodology employs a large-scale macroeconometric model to obtain a no-stimulus forecast and compares that to a simulated forecast where federal government spending includes the ARRA. This is the methodology used in reports by the Congressional Budget Office, the White House’s Council of Economic Advisers, and private forecasters such as Macroeconomic Advisers, HIS Global Insight, and Moody’s Economy.com, as well as a number of academic studies. The second methodology is an attempt to count the jobs created or saved by requiring first-round recipients of certain types of ARRA funds to report the number of jobs they were able to add or save as a direct result of projects funded by ARRA. These counts are aggregated across all reporting recipients by the Recovery Accountability and Transparency Board (RATB), which was established by ARRA. But the number of jobs created or saved reflects only first-round jobs tied to ARRA spending, such as hiring by contractors and their immediate subcontractors working on ARRA-funded projects, and excludes both second-round jobs created by lower-level subcontractors and jobs created indirectly due to spillovers such as consumer spending made possible by the wages associated with these jobs. By contrast, Wilson’s methodology used employment totals as reported by the Bureau of

W hat about Other Kinds of Fiscal Stimulus?

119

Labor Statistics so that all direct and indirect jobs created by ARRA would be reflected in the results. Wilson estimated that ARRA spending created or saved about 2.1 million jobs, or 1.6% of pre-ARRA total nonfarm employment, in its first year. Jobs created or saved grew to 3.4 million by March 2011. He noted that his results were in line with estimates of the ARRA’s impact generated by studies using the macroeconometric modeling approach.

Does State Fiscal Relief during Recessions Increase Employment? This subsection heading reproduces the title of the study by Chodorow-Reich, Feiveson, Liscow, and Woolston (2012). Like Wilson, they applied regression analysis to a cross section of states. But unlike Wilson, they focused on Medicaid spending. They tried to determine how jobs in a state were affected by ARRA Medicaid funds received by that state from the federal government. Most state and local governments have balanced-budget requirements that limit their borrowing even in a recession, so when recession reduces tax revenue, these governments are under pressure to lay off employees, cut spending and transfer programs, and raise tax rates. The designers of the ARRA sought to give fiscal relief to state and local governments by directing roughly a third of the ARRA’s funds to these governments. An important share of ARRA funds was used to increase the federal match of state Medicaid expenditures. Chodorow-Reich et al. noted that few studies focused on the response of state and local governments to federal grants aimed at combating a recession. Federal grants might have a small or zero immediate impact on economic outcomes if states simply used them to bolster their rainy-day funds, saving instead of spending the federal funds. On the other hand, states might use

120

H ow to C ombat R ecession

the money to reduce tax increases or avert spending, allowing money to enter the economy more quickly than direct federal purchases that require project selection and approval. They decided to focus on the federal Medicaid match for several reasons. First, the amount of money distributed through this program was large enough to plausibly generate a detectable effect on employment. Out of a total of $88 billion dedicated to an increase in Medicaid matching funds, states had received $61.2 billion by June 30, 2010, the end of the Chodorow-Reich et al. period of study. Second, increasing the federal match effectively transfers money into state budgets that states can use for any purpose—the money is fungible. Many states reported that the increase in the Medicaid match enabled them to allocate money quickly to areas that otherwise would have undergone deeper budget cuts. Third, the level of additional money received by states as of June 2010 per person age 16 or older varied greatly across states, from a low of about $100 to a high of about $500. This variation made possible a cross-sectional econometric strategy. Chodorow-Reich et al. focused their analysis on the effect on jobs because the effectiveness of the ARRA centered largely on its impact on jobs. Moreover, high-quality monthly state-level employment data made it possible to obtain more precise estimates of fiscal multipliers than were possible with the existing state-level income data. The ARRA increased the percentage of Medicaid expenditures that the federal government paid for all states by 6.2 percentage points and increased the match rate by more for states that experienced especially large increases in unemployment. Like Wilson, Chodorow-Reich et al. noted that their study measured local effects and ignored reverberations outside the recipient state. They found that ARRA transfers to states for Medicaid had an economically large and statistically robust positive effect on jobs. They estimated that $100,000 in Medicaid transfers resulted in

W hat about Other Kinds of Fiscal Stimulus?

121

nearly four jobs that lasted one year; they estimated that three of the four jobs were outside the government, health, and education sectors, implying that the states must have used the funds to avoid tax increases or spending cuts. Moreover, the federal transfers did not appear to have increased the states’ end-of-year balances.

Macro Fiscal Policy in Economic Unions: States as Agents This subsection heading reproduces the title of article by Carlino and Inman (2016). They noted that a striking feature of the ARRA was its reliance upon intergovernmental transfers to state and local governments for implementing central government macroeconomic fiscal policy. Carlino and Inman used time-series data over the last several decades, in contrast to the studies by Wilson and by Chodorow-Reich et al. reviewed above, who used data only from 2009 and 2010 following the enactment of the ARRA during the Great Recession. A major concern about the Carlino and Inman study is that they did not isolate the state response to federal aid during recessions from the state response when the economy was not in recession. The Carlino and Inman estimates were based on data generated every year over the past several decades, so the results they reported were a weighted average of state response to federal aid in a normal economy and state response in recession, with a much larger weight for the response in a normal economy. In a normal year when state or local tax revenue is sufficient to avoid spending cuts, it is conceivable that a substantial share of federal aid might be saved in a rainy-day fund. But in a recession, when tax revenue has fallen and spending cuts are looming, it is more likely that a large portion of federal aid will be used to avoid unpopular spending cuts that would cause hardship, rather than

122

H ow to C ombat R ecession

to build a rainy-day fund. Hence, Carlino and Inman’s estimates for state response to federal aid in an average year in their sample are likely to substantially underestimate the state response to federal aid in a recession. For an average year in their sample (not a recession year), their estimate of the GDP multiplier for federal project aid was modest—between 0.0 and 1.0—and not statistically significant, but the multiplier for federal welfare aid was large—between 1.6 and 2.3—and statistically significant. They estimated that states saved about half of federal project aid and spent all of matching welfare aid on lower-income assistance and tax relief for general taxpayers. I conjecture that multipliers for both project aid and welfare aid in a recession year would be much higher than their estimates for an average year.

An Empirical Analysis of the Revival of Fiscal Stimulus in the 2000s This subsection heading reproduces the title of an article by John Taylor (2011). Taylor has probably done more to oppose fiscal stimulus in a recession than any other American economist. In this Journal of Economic Literature article, a journal read by many academic economists, Taylor wrote about other components of a fiscal stimulus package as well as tax rebates. In the abstract for his article, Taylor stated that none of the components of the fiscal stimulus packages enacted in the United States since 2000 actually worked in recession. He claimed that households saved tax rebates, and state and local governments saved their federal aid. In chapter 4, I critiqued Taylor’s articles in opposition to tax rebates. Here I will critique his analysis of federal aid to state and local governments. In a recession, state and local tax revenue falls, but state and local governments are required to balance their budgets as

W hat about Other Kinds of Fiscal Stimulus?

123

soon as possible. These governments may delay balancing their budgets by emergency borrowing, but before long they will be forced to choose one of two very unpopular alternatives: cutting spending or raising taxes. Suppose emergency federal aid is now given to these governments. Taylor claims these governments will save the emergency aid and cut spending programs and raise taxes just as much as if there were no additional federal aid. How many state or local officials who must run for re-election would support saving the federal aid instead of using the aid during a recession to avoid either unpopular spending cuts or unpopular tax increases? Here is a simple numerical example I constructed that casts doubt on Taylor’s claim that federal aid in a severe recession had no effect on the expenditures of a state or a local government. Table 5.1 shows possible impacts of a recession and federal aid on the budget of a state or a local government. In any row of the table, Tax Revenue + Federal Aid + Borrowing = Expenditures + Gross Saving. The left side gives the sources of funds for the government, and the right side gives the uses of funds. Note that TABLE 5.1 The Impact of Federal Aid on State and Local

Expenditures in Recession Year

Tax revenue

Federal aid

Borrowing

Expenditures

Gross saving

0N 1N 1R 1RB 1RA

$ 950 $1,000 $ 800 $ 800 $ 800

$190 $200 $200 $200 $340

$0 $0 $0 $ 140 $0

$1,140 $1,200 $1,000 $1,140 $1,140

$0 $0 $0 $0 $0

1RAS

$ 800

$340

$1,000

$ 140

Note: In each row, tax revenue + federal aid + borrowing = expenditures + gross saving.

124

H ow to C ombat R ecession

Total Revenue = Tax Revenue + Federal Aid. Also note that net saving is defined as gross saving minus borrowing, so moving borrowing to the right of the equation would change the gross saving term to net saving. In Table 5.1, borrowing is $0 in every row except one (row 1RB), so in all other rows gross saving and net saving are equal. If the economy were normal (N)—not in recession—in years 0 and 1, it would experience normal growth from year 0 to year 1 and would move from row 0N to row 1N. I assume there would be normal federal aid of $190 in year 0 and $200 in year 1, and that the budget would be balanced each year, with borrowing and gross saving each equal to $0. To simplify the numbers, I assume each number in row 0N is 95% of the corresponding number in row 1N—equivalently, that there would be 5.3% growth in each number from year 0 to year 1 if the economy stayed normal. In row 1N, total revenue, which equals tax revenue ($1,000) plus federal aid ($200), equals expenditures ($1,200). Now suppose the economy falls into recession in year 1, automatically reducing tax revenue to $800. Consider the possibilities for year 1 with the recession. Rows 1R, 1RB, 1RA, and 1RAS give four possible outcomes. In row 1R, I assume normal federal aid (the federal aid that would have occurred in a normal economy in year 1) is $200 in row 1N, and that the government is required to keep its budget balanced. In this row I assume that borrowing and saving are each $0. With tax revenue $800 and federal aid $200, expenditures would equal $1,000, which is $200 less than in row 1N ($1,200)— what it would have been in year 1 without recession—and $140 less than last year in row 0N ($1,140). There are two ways to prevent expenditures from falling to $1,000 in year 1. The first is to permit the government to borrow as shown in row 1RB. The second is to increase federal aid as shown in row 1RA.

W hat about Other Kinds of Fiscal Stimulus?

125

In row 1RB, the government is permitted to borrow, and it borrows $140, just enough to prevent a cut in spending below last year’s $1,140 in row 0N, but still below $1,200 in row 1N. With expenditures $1,140 and total revenue (tax revenue plus federal aid) equal to $1,000, the government deficit is $140. Note that net saving, defined as gross saving minus borrowing, is −$140 in this row. In row 1RA, the government must balance its budget. I assume borrowing and gross saving are each $0. But federal aid increases $140 from $200 to $340, so expenditures are $1,140. Thus, when the S&L government is unable to borrow, $140 of additional federal aid can prevent it from having to reduce expenditures below $1,140. Finally, row 1RAS shows the outcome that Taylor claims will occur despite a $140 increase in federal aid (from $200 to $340). In row 1RAS, expenditures are still only $1,000, just what would have happened in row 1R without an increase in federal aid. This happens because the government chooses to respond to the $140 increase in federal aid by increasing its gross saving by $140 and cutting its expenditures to $1,000, which is $140 below last year’s expenditures ($1,140 in row 0N). Taylor claims the government would choose this outcome even though such an unpopular cut below last year’s expenditures ($1,140) could have been prevented just by keeping gross saving at $0 ($1,140 in 1RA). It is very unlikely that the government would choose Taylor’s row 1RAS in a recession. With popular and political pressure to avoid severe expenditure cuts that cause hardship, it seems much more likely that a state or local government would give top priority to avoiding severe expenditure cuts rather than to increasing gross saving, so that the government is much more likely to choose row 1RA rather that Taylor’s row 1RAS. Thus, as shown in row 1RA, the increase in federal aid would prevent a severe cut in expenditures rather than cause an increase in gross saving.

126

H ow to C ombat R ecession

Conclusions In chapters 3 and 4, I made the case for having tax rebates to households be the largest component of any fiscal stimulus package to combat a recession. In this chapter, I recommended certain fiscal stimulus programs for inclusion in a fiscal stimulus package (along with tax rebates to households), but recommended against inclusion of certain other fiscal stimulus programs. For each fiscal stimulus program I explained why I recommended for or against inclusion. I recommended these programs for inclusion: (1) a temporary increase in federal aid (grants) to state governments; (2) temporary tax incentives for business investment (for example, an investment tax credit, or bonus depreciation); (3) a temporary increase in federal spending for infrastructure repairs and maintenance (either direct spending by the federal government or grants to the states for this purpose). I recommended against these programs: (1) a temporary cut in income tax withholding; (2) a temporary cut in payroll taxes; (3) a cut in income taxes; (4) an increase in spending on long-term programs. I described and commented on components of the fiscal stimulus enacted in early 2009 and implemented in 2009 and 2010—the American Recovery and Reinvestment Act (ARRA). I then discussed a mistake made by many economists that leads them to underestimate the “multiplier effect” of fiscal stimulus programs. Finally, I reviewed and evaluated several empirical studies of fiscal stimulus programs utilized in the Great Recession.

Chapter 6

Would Stimulus without Debt Be Inflationary?

Would stimulus without debt be inflationary? This question should be divided into two parts. First, will stimulus without debt generate inflation before the economy fully recovers from the recession? I will explain why my answer is no. Second, can stimulus without debt be phased out by the time the economy reaches full employment so that it doesn’t generate inflation? I will explain why my answer is yes.

Stimulus without Debt Is Not Inflationary in Recession Inflation is defined as rising prices. Prices and quantities of goods and services are determined by demand and supply for goods and services. When demand increases for goods and services, how much of the increase in demand will go into raising quantities of goods and services, and how much will go into raising prices? It depends on whether the economy is in recession or at full employment. 127

128

H ow to C ombat R ecession

Suppose the economy is in a recession with high unemployment and low capacity utilization, so that unemployed workers would be glad to work even without any increase in wages and idle machines are available for these workers. Then the increase in demand would cause employers to hire unemployed workers and utilize idle machines, thereby increasing real output with hardly any increase in costs or prices, so output would expand with hardly any inflation. By contrast, suppose the economy is at full employment and full capacity utilization when demand increases. With little or no unemployed labor or machinery available, real output would hardly increase in response to the increase in demand. Instead, the increase in demand would bid up wages and prices, generating inflation. Figure 6.1, a textbook aggregate-demand/aggregate-supply diagram, illustrates these points. In the figure, real output Y is plotted horizontally and the price level P is plotted vertically. Both Y and P are determined by the intersection of the demand S

D' D D' D Y1

FIGUR E 6.1 The Impact of Stimulus on Inflation in Recession vs at

Full Employment

Would Stimulus without Debt Be Inf lationar y?

129

curve and supply curve. In a recession, the demand curve intersects the supply curve at Y1, well below Y0. To combat the recession requires policy that increases aggregate demand for goods and services and therefore shifts the aggregate demand curve to the right. Figure 6.1 illustrates what happens in recession versus full employment when the demand for goods and services increases and the aggregate demand curve shifts to the right. When the economy is in severe recession at a low value of real output, Y1, unemployment is high and capacity utilization is low, so a shift right of aggregate demand (D) can raise real output (from Y1 to Y2) without bidding up wages, costs, and prices, so the aggregate supply (S) curve is nearly flat in this range. But when the economy is at full-employment output Y0, a shift to the right of aggregate demand curve mainly bids up wages, costs, and prices, without much increase in real output Y, so the aggregate supply curve is steep in this range. Thus, when the economy is in a severe recession, a shift right of the D curve causes a large increase in real output (from Y1 to Y2) but hardly any inflation (rise in P). When the economy is at full-employment output Y0, a shift right of the D curve by the same magnitude causes little increase in real output (rise in Y) but instead causes a lot of inflation (rise in P). The stimulus-without-debt policy should only be used in a recession, not when the economy is at full employment. Figure 6.1 implies that as long as the stimulus-without-debt policy is phased out before the economy reaches full-employment output Y0, it will cause little or no inflation. Here is another way that economists reach a similar conclusion. If high demand for goods and services pushes the unemployment rate below un, the rate that corresponds to full employment, employers find it hard to retain their own workers, who have good job options, and employers are therefore compelled to raise wage increases, which raises cost increases and price increases,

130

H ow to C ombat R ecession

so inflation rises. Symmetrically, when the unemployment rate is pushed above un by low demand for goods and services, employers find it easy to retain their own workers, who have bad job options, and employers are therefore able to reduce wage increases, which reduces cost increases and price increases, so inflation falls. Thus, high demand for goods and services and the resulting low unemployment rate cause inflation to rise, and low demand for goods and services and the resulting high unemployment rate cause inflation to fall. The stimulus-without-debt policy would only be used in a recession, not when the economy is at full employment. The model just described implies that as long as the stimulus-without-debt policy is phased out as the economy reaches un, it will cause little or no rise in inflation. What’s the difference between the first explanation shown in Figure 6.1 and the second explanation? The first explanation assumes an economy where inflation is initially zero, while the second explanation assumes an economy where inflation is initially constant at a steady low rate. But both explanations lead to a similar conclusion: inflation won’t start or worsen due to an increase in aggregate demand for goods and services as long as the economy has not yet reached full employment. The experience of the Great Recession is fully consistent with this analysis. In 2009 and 2010, when GDP was far below normal (potential) and unemployment was far above normal (10% instead of 5%), a large fiscal stimulus and monetary stimulus was implemented for two years. In 2009 some critics warned that this combined stimulus would be highly inflationary, and predicted a sharp rise in inflation over the next few years. Inflation, however, stayed constant at a low rate of 2%, not just in the depths of the recession (2009 and 2010), but as the economy the economy gradually recovered from the Great Recession to achieve full employment (2011–2015). As this book goes to press in 2018,

Would Stimulus without Debt Be Inf lationar y?

131

after two years of full employment (2016 and 2017), inflation is still steady at about 2%. So contrary to the warnings of some, the monetary and fiscal stimulus implemented during the Great Recession in 2009 and 2010 has not raised inflation above 2%. But what would have happened to inflation under stimulus without debt with a fiscal stimulus three times as large as the stimulus implemented in 2009? Recall that at the end of chapter 3, I contended that a fiscal stimulus three times as large was needed to overcome the severity of the Great Recession. The analysis of this section implies that inflation would have stayed at 2%—its rate prior to the recession—because this stimulus would not have reduced the unemployment rate below 5%. The larger stimulus would have generated a faster recovery so that full employment would have been achieved several years sooner than it did. But the phaseout of the stimulus would have prevented the economy from overheating and generating inflation.

Can Money Injected since 2008 Be Withdrawn? As of mid-2017, the Fed had only just started to withdraw a significant amount of money from the economy because the gradual recovery from the Great Recession took a full decade. The unemployment rate has at last fallen below 5% (from its peak of 10% in 2010); monthly job creation has been adequate, but not strong enough to generate a full-fledged economic boom. Moreover, inflation has remained low, though some critics continue to predict that the Fed’s reversal will fail and inflation will rise sharply. Starting in 2008, the Fed injected a large amount of “high- powered money” (also called “monetary base”) into the economy by writing checks to buy bonds or make loans. When the Fed writes checks, the recipients deposit them in their bank accounts

132

H ow to C ombat R ecession

so banks receive reserves from the Fed. In normal economic times, banks lend out a large fraction of their reserves, and the borrowers increase their demand for goods and services. But in a severe recession, the banks make few loans partly because anxious consumers and businesses don’t want to borrow, and partly because banks fear their loans won’t be repaid. Critics say the large injections of high-powered money into banks starting in 2008 will eventually cause substantial inflation. If the economic recovery becomes strong, banks will rapidly draw down their huge reserves to make a large volume of loans and give borrowers corresponding checking accounts that they will use to demand goods and services; given the huge stockpile of reserves, critics say this process will rapidly generate excessive demand for goods and services, fueling a significant rise in inflation. They say that the fact that there wasn’t any rise in inflation from 2008 through 2017 does not refute their contention, because it took eight years (from 2008 to 2016) for the economy to finally return to full employment; they say the inflation will finally take off when the recovery grows stronger. But it would be straightforward for the Fed to withdraw the high-powered money it injected into the economy since 2008 for one simple reason: whenever the Fed injected high- powered money, it acquired assets. As the economic recovery gets stronger, the Fed can and should gradually sell most or all of those assets, thereby withdrawing most or all of the money it injected. The withdrawal of money from banking system would reduce the funds that banks have available for making loans, and this reduction in the supply of loanable funds would cause banks to raise the interest rates they charge household and business borrowers. This rise in interest rates would reduce borrowing and spending by households and businesses, thereby preventing inflationary pressure by keeping aggregate demand for goods and services from rising above full-employment output.

Would Stimulus without Debt Be Inf lationar y?

133

Can Money Be Restored to Normal after Stimulus without Debt? Can money be restored to normal after the injection of money under stimulus without debt? This question is not easy to answer for one simple reason: when the Fed gives a transfer to the Treasury instead of buying bonds in the open market, the Fed doesn’t acquire assets that can later be sold to reverse the process and withdraw money from the economy. Thus, the answer is no longer obviously yes. But I will now show that if stimulus without debt had been implemented during the Great Recession with a fiscal stimulus three times larger (my recommendation in chapter 3) than was actually implemented, the Fed would have been able to restore money to normal in time to avoid generating inflation as the economy reached full employment. It must be emphasized that the Fed wouldn’t need to reduce money in the economy to its amount prior to the recession. In a growing economy, the Fed steadily injects high-powered money into the economy by buying bonds so that the amount of money in the economy grows each year in line with the growth of GDP. When stimulus without debt is implemented, money temporarily grows faster than normal, so the amount of money in the economy rises above its normal growth path. The Fed would only need to reduce money to its normal growth path, not to its prerecession amount. To bring money down to its normal growth path, the Fed needs to make money grow more slowly than normal until it reaches its normal growth path. To make money grow more slowly than normal, the Fed would need to buy fewer bonds each year than it otherwise would have bought. Under the stimulus-without-debt plan that I proposed in chapter 3 for the Great Recession, there would have been a fiscal stimulus equal to 6% (instead of the actual 2%) of GDP each year

134

H ow to C ombat R ecession

from 2008 through 2010, so the Fed’s transfers to the Treasury would have injected an amount of money each year equal to 6% of GDP—$900 billion (GDP was $15,000 billion)—for a total of $2,700 billion over three years. Thus, to bring money down to its normal growth path, the Fed would need to buy $2,700 billion less of bonds than it otherwise would have. In other words, the Fed would need to inject less money through bond buying to offset its injection of money due to its transfer to the Treasury. From 2008 to 2011 the Fed actually bought $1,200 billion of Treasury securities (its holdings of Treasury securities increased from $500 billion in 2008 to $1,700 billion in 2011) and $800 billion mortgage-backed securities (its holdings of mortgage- backed securities increased from $0 in 2008 to $800 billion in 2011)—a total for these two bonds of $2,000 billion (Federal Reserve Board 2017). Under stimulus without debt, I would have recommended that the Fed purchase only $200 billion of these two bonds instead of the $2,000 billion it actually purchased from 2008 to 2011 ($1,800 billion less), so the Fed’s money injection into the economy from 2008 to 2011, instead of being $2,700 billion above its normal growth path, would have been only $900 billion above its normal growth path. From 2011 to 2013 the Fed bought $500 billion of Treasury securities and $700 billion of mortgage-backed securities—a total of $1,200 billion (Federal Reserve Board 2017). Under stimulus without debt, I would have recommended that the Fed purchase only $300 billion of these two bonds instead of the $1,200 billion it actually purchased from 2011 to 2013 ($900 billion less), so the Fed’s money injection into the economy from 2011 to 2013, instead of being $900 billion above its normal growth path, would have been $0 billion above its normal growth path. Thus, by 2013 money in the economy would have been back to its normal growth path.

Would Stimulus without Debt Be Inf lationar y?

135

Inflation will only threaten if aggregate demand for goods and services rises above full-employment output. Raising interest rates sufficiently can prevent aggregate demand from rising above the potential output of the economy. The Fed can raise interest rates by reducing the injection of money so that money rises more slowly than GDP; the Fed can do this by reducing its purchases of bonds. This is the standard way that the Fed has raised interest rates in the past. There are two other policies that the Fed can also use to raise interest rates. Under the first, recommended by Bernanke (2009), the Fed can raise the interest rate it pays to banks on bank deposits at the Fed. Banks won’t make loans to households and businesses at an interest rate less than the interest rate the Fed pays banks for keeping their funds at the Fed, so by raising its rate on bank deposits, the Fed can cause a rise in interest rates facing households and businesses, thereby reducing their borrowing and spending. Under the second policy, recommended by Siegel (2013), the Fed can raise the reserves each bank is required to hold to back its deposits. Raising the bank’s required reserve ratio would reduce the funds banks can supply for loans, and would result in banks charging higher interest rates to borrowers, thereby reducing borrowing and spending by households and businesses. Finally, fiscal policy could in theory reduce aggregate demand by raising taxes and/or cutting government spending whenever aggregate demand is excessive and threatens to generate inflation. But in practice Congress is very reluctant to raise taxes or cut spending because these actions are unpopular; by contrast, Congress is much less reluctant to enact fiscal stimulus (cut taxes or raise spending) because each component is popular. In practice, then, it is prudent to rely on monetary policy rather than fiscal policy to prevent inflation.

136

H ow to C ombat R ecession

Conclusion If stimulus without debt had been implemented in the Great Recession, the large fiscal stimulus and the large injection of money from the Fed transfer to the Treasury would have caused little or no inflation. The large stimulus would have raised aggregate demand for goods and services enough to achieve full employment. If demand rose further, the Fed would have been able to eliminate this excess demand by raising interest rates. Although stimulus without debt would have temporarily raised money in the economy above its normal growth path, the Fed would have reduced money to its normal growth path before the economy reached full employment. Thus, stimulus without debt would have brought the economy out of the Great Recession without generating inflation.

Chapter 7

Would Stimulus without Debt Weaken the Fed’s Balance Sheet?

At first glance it might seem that stimulus without debt would cause the Fed to have a balance-sheet problem. Why? When the Fed buys a Treasury bond in the open market, it obtains an asset, but if the Fed gives the Treasury a transfer, it obtains no asset. On the Fed’s balance sheet, the Fed’s “capital” (“net worth”), defined as assets minus liabilities, would therefore be reduced by the amount of the transfer if the Fed gives a transfer instead of buying Treasury bonds. But the Fed can avoid a balance-sheet problem by taking the following action. Just before the Fed writes its transfer check to the Treasury, the Fed should order an amount of new Federal Reserve notes (from the usual place, the Treasury’s Bureau of Engraving and Printing) equal to its transfer to the Treasury and then store these notes in the Fed’s vault. Cash in the Fed’s vault is an asset on the Fed’s balance sheet, so the Fed’s total assets would increase by an amount equal to the Fed’s transfer to the Treasury—the same increase in total assets that would have occurred if the Fed had instead spent the same amount of money buying Treasury bonds. Moreover, just as Treasury bonds 137

138

H ow to C ombat R ecession

held by the Fed are not used to meet requests by banks or the public for Fed notes, neither should these new notes in the Fed’s vault; requests for Fed notes should be met as usual by ordering additional new Fed notes, just as if the vault contained Treasury bonds instead of Fed notes. With this ordering and storing of Fed notes in the Fed’s vault, stimulus without debt would not reduce the Fed’s capital, just as the Fed’s purchase of Treasury bonds does not reduce the Fed’s capital. For example, suppose the Fed plans to transfer $450 billion to the Treasury for fiscal stimulus in a severe recession and Congress enacts a $450 billion fiscal stimulus. Just before making the transfer, the Fed would order $450 billion in new Fed notes from the Treasury’s Bureau of Engraving and Printing, and upon receiving these notes the Fed would store them in its vault. To simplify, suppose the $450 billion fiscal stimulus consists entirely of tax rebate checks to households and that the households deposit these checks at their banks and request $450 billion of Fed notes. Under current balance-sheet accounting for the Fed, the $450 billion of Fed notes held by the public would be listed on the Fed’s balance sheet as a liability, but this Fed liability would be matched by $450 billion of new Fed assets—new Fed notes in the Fed’s vault—so there would be no change in the Fed’s capital. If the banks request only $50 billion of Fed notes (instead of $450 billion) and increase their deposits at the Fed by $400 billion, once again Fed liabilities increase $450 billion ($50 billion plus $400 billion), so once again there would be no change in the Fed’s capital. By contrast, if a firm or household writes a transfer check to someone, its net worth decreases by the amount of the transfer check, and the firm or household is not able to order and store new Federal Reserve notes to prevent a decrease in its net worth. But the Fed can. That’s exactly the point. Congress has given the Fed, the central bank of the United States, the power to order

Would Stimulus without Debt Weaken Fed’s Balance Sheet? 139

new Federal Reserve notes, while Congress prohibits firms or households from doing the same. Moreover, the Fed does not have to “back” Federal Reserve notes with gold or anything else. My recommendation illuminates the fundamental point that the Fed has the power to raise its total assets promptly by as much as it deems necessary. Thus, the Fed can always make sure its assets exceed its liabilities; it can always make sure its net worth is positive. This chapter could end here because the ordering and storing of new Fed notes solves the balance-sheet problem without any change in the current balance-sheet accounting used by the Fed. But it is worth looking further into whether the current balance- sheet accounting used by the Fed should be reformed, and if so, to make sure that stimulus without debt still would not reduce the capital (net worth) of the central bank.

Should the Fed’s Accounting Be Reformed? There are two fundamental reasons why the current balance- sheet used by the Fed should be reformed: (1) its overstatement of the Fed’s total liabilities; (2) it failure to recognize that the Fed has the power to promptly raise its total assets. After explaining these two problems, I recommend a reform of the Fed’s current balance-sheet accounting.

Overstatement of the Fed’s Total Liabilities On the current balance sheet of the Federal Reserve, the item “Federal Reserve notes held by the public” (including banks) is listed as a liability. This is consistent with the view held by some that government paper money held by the public is government debt, just like Treasury bonds held by the public.

140

H ow to C ombat R ecession

Recall, however, that in c hapters 2 and 3 I explained why paper money held by the public is not government debt. Listing paper money as a liability on the Fed’s balance sheet made sense when the Fed pledged to pay gold to anyone presenting Fed notes and requesting gold. The Fed had to worry about whether it could always obtain enough gold to fulfill its promise if a panic should sweep holders of Fed notes, who might create a “run” on the Federal Reserve by demanding gold in exchange for their Fed notes. But for many decades, the Fed has not pledged to pay gold or anything else to anyone presenting Fed notes. Hence, Fed notes held by the public are no longer a Fed liability because the Fed doesn’t owe anything to holders of the notes. It follows that Fed notes held by the public should not be listed as a liability on the Fed’s balance sheet. Some background is useful to appreciate why Fed notes in circulation should not be counted as a liability of the Fed on its balance sheet and to grasp the significance of the fiat money system most countries have today. In their macroeconomics textbook, Sachs and Larrain (1993) give an insightful explanation of the central bank’s balance sheet under fiat money that supports not counting Fed notes held by the public as a liability of the Fed. They begin by pointing out that official government paper money is “legal tender,” which means that a creditor must accept government paper money from a debtor, and that taxes are paid with government paper money (or a bank check backed by government paper money). They then explain how fiat money evolved. When government paper money was first introduced in the past two centuries, the government pledged to provide ounces of gold or silver to any holder of paper money who requested it. But once the public became confident about the paper money after many years of use, the government withdrew its pledge. Yet even without a government pledge to provide gold or silver, the paper money continued

Would Stimulus without Debt Weaken Fed’s Balance Sheet? 141

to be used. Paper money unbacked by gold or silver is called fiat money. Fiat money has been widely used only in the twentieth century. Today virtually all countries use fiat money. In a system dependent on gold or silver, the economy was buffeted by either new discoveries or the absence of new discoveries of gold and silver. New discoveries enabled governments to expand paper money, and the economy experienced an inflationary boom. Absence of new discoveries prevented governments from increasing paper money as fast as potential real output and the economy experienced a deflationary bust. An economy that depends on the vagaries of gold or silver discoveries is inherently unstable. During the 20th century most advanced economies converted to a fiat money system by taking two steps. The first step was government paper money backed by gold or silver. This gold or silver backing was necessary to get the public willing to use paper money; as they used it, the public gained confidence in it. The second step was the removal of the gold and silver backing. It turned out that despite this removal, the public continued to use paper money with confidence. These two steps remind me of the three steps used to get children to ride a two-wheel bicycle. The first step is to provide training wheels—without training wheels, many children would be afraid to try for fear of falling. The second step is to remove the training wheels but have an adult hold the seat of the bicycle to keep the child from falling. The third step is to have the adult let go of the seat. Sachs and Larrain make a key point about the current treatment of Federal Reserve notes held by the public as a liability on the Fed’s balance sheet. They point out that the outstanding stock of Fed notes held by the public is usually the largest liability of the Fed. But they ask this question: In what sense is this really a liability—something that the Fed owes? Under the gold standard, holders of government paper money

142

H ow to C ombat R ecession

had the legal right to convert money into gold at the Fed at a fixed price. This made the paper money a clear liability because the Fed had to supply gold in return for paper money to anybody that demanded gold. But once that government pledge was removed, there is no longer an automatic right to convert paper money into anything else. Sachs and Larrain provide an anecdote that makes the point. In 1961 Senator Paul Douglas, then chairman of the Joint Economic Committee, met with Douglas Dillon, the US secretary of the treasury. Senator Douglas gave Secretary Dillon a $20 bill, urging him to honor his liability. Dillon didn’t hesitate: he accepted Douglas’s $20 bill and gave him two $10 bills. Similarly, Karl Whelan (2015), a professor of economics at University College Dublin, wrote that a central bank’s “liabilities” differ from the liabilities of a private bank or business. He pointed out that under the gold standard, a central bank owed gold to anyone holding paper money who wanted to exchange paper for gold, but in a modern fiat currency system, the central bank doesn’t owe anything to a holder of paper money. In a fiat money system, the conventional listing of central bank notes held by the public as a “liability” on the central bank’s balance sheet should be viewed as “notional.” Later in his article Whelan stated that there is no good basis for the view that central banks assets need to be greater than their “notional” liabilities in order for the central bank to function successfully. I wish Whelan had simply written that central bank notes held by the public should not be listed at all under central bank liabilities. Listing government paper notes held by the public in the liabilities column, even with a footnote that this entry is “notional,” would continue to mislead many who glance at or read the central bank’s balance sheet. In my view, the proper solution is to delete central bank paper money held by the public from the list of central bank liabilities, rather than list it as a “notional”

Would Stimulus without Debt Weaken Fed’s Balance Sheet? 143

liability and then try to persuade the public not to worry about the central bank’s apparent negative capital (net worth). In an article with the title “Accounting for Sovereign Money: Why State-Issued Money Is Not ‘Debt,’ ” Dyson and Hodgson (2016) contend that government paper money held by the public is not government debt in a fiat money system. They begin with an examination of two sources that accountants use: The International Accounting Standards (IAS) and International Financial Reporting Standards (IFRS). They report that a “financial liability” is defined as “a contractual obligation to deliver cash or another financial asset to another entity.” Debt obliges the debtor to deliver something of value to the creditor. Usually the debt contract specifies when it must be settled and what must be delivered. When the government is the supposed debtor and the holder of government paper money is the supposed creditor, the government has no obligation to settle or deliver anything to the holder of paper money. Thus, government paper money is not government debt. By contrast, a government bond is government debt because the government is obliged to pay specific amounts of money—principal plus interest—to the holders of government bonds. Dyson and Hodgson then observe that paper money held by the public is still listed as a central bank liability even though the central bank has no obligation to pay anything to the holders of paper money.

The Fed Can Create Assets The Fed can order new Fed notes and store them in its vault. These vault notes are Fed assets. The volume of notes the Fed can order and store can be very large if necessary. Of course, the Fed’s goal should be to inject the right amount of money into the economy—enough to avoid recession, but not enough to generate inflation. If paying off Fed liabilities would inject too

144

H ow to C ombat R ecession

much money into the economy and generate inflation, the Fed can and should sell Treasury bonds and other assets to withdraw the excess money from the economy. Thus, the Fed can make its total assets as large as necessary by ordering and storing new Fed notes in its vault without generating inflation. As long as the Fed exercises its power to order and store notes, it can always make its total assets exceed its total liabilities. Virtually all experts concede that the Fed can do this. One prominent expert is former Fed chairman Alan Greenspan. In a speech when he was chairman of the Fed, Greenspan (1997) stated clearly that under the current fiat money system the Fed could create assets by ordering new Federal Reserve notes. Greenspan said that a nation’s credit rating depends on the soundness of its fiscal, monetary, and regulatory policies. If there is “confidence in the integrity” of the government, in a fiat money system the central bank can issue a large amount of paper money that will be accepted by the public. The government “cannot become insolvent.” He warned, however, that issuing too much money would be inflationary and harmful. By contrast, when the Fed pledged to pay gold to holders of Fed notes, the Fed could not issue a large amount of paper money without risking becoming insolvent— unable to meet its gold pledge to holders of paper money.

Comments by Specialists in Central Bank Accounting Several specialists on central bank accounting have recognized that current accounting is misleading. Unfortunately, they do not advocate the accounting reform I propose, but instead accept current central-bank accounting and simply point out where it is misleading. The problem with that approach is that the public and financial market participants will continue to be influenced by the current official balance sheet of a central bank. Nevertheless, here I document their concerns.

Would Stimulus without Debt Weaken Fed’s Balance Sheet? 145

In a foreword to a paper entitled “Central Bank Finances,” Caruana (2013) wrote that, in contrast to commercial banks, central banks do not seek profits; nor do they face the same financial constraints. In fact, most central banks could keep losing money until their equity turned negative, yet still function completely successfully. The problem is that many people don’t realize this and become alarmed when they learn that a central bank’s accounting equity has gone negative. Markets may well respond poorly due to the mistaken belief of financial investors that balance-sheet losses imply an inability to conduct effective policy. Moreover, many politicians may jump to the conclusion that negative equity means that poor decisions have been made and that the central bank now needs a government bailout with taxpayer money. If politicians and financial investors knew what central bank specialists know, they would relax on learning a central bank has negative equity, and this relaxation would enable central banks to function effectively despite their negative equity. But they may not relax, and this may cause serious problems. Losses or negative capital may raise “erroneous” doubts about the central bank’s ability to deliver on policy targets, and generate political pressure on the central bank to take actions to turn its equity positive. In the paper itself, “Central Bank Finances,” Archer and Moser-Boehm wrote that for more than a century the mission of central banks has been public policy, not private profit. Governments are the effective owners of central banks. Shares in most central banks are not for sale, so financial investors don’t need to compare central bank assets and liabilities in order to decide whether to buy shares. Central banks are protected from insolvency proceedings and can legally to operate with negative equity. Although central banks are really monetary authorities, not banks, they have retained the language and balance sheets of banks. This has misled many into thinking that a central bank,

146

H ow to C ombat R ecession

like a commercial bank, cannot operate effectively for very long if its conventional balance sheet shows assets that are less than liabilities.

Balance-Sheet Reform and Stimulus without Debt Recall my recommendation at the beginning of this chapter for how stimulus without debt should be handled on the Fed’s balance-sheet. Just before the Fed writes its transfer check to the Treasury, the Fed should order an amount of new Federal Reserve notes (from the usual place, the Treasury’s Bureau of Engraving and Printing) equal to its transfer to the Treasury and then store these notes in the Fed’s vault. Cash in the Fed’s vault is an asset on the Fed’s balance sheet, so the Fed’s total assets would increase by an amount equal to the Fed’s transfer to the Treasury—the same increase in total assets that would have occurred if the Fed had instead spent the same amount of money buying Treasury bonds. Moreover, just as Treasury bonds held by the Fed are not used to meet requests by banks or the public for Fed notes, neither should these new notes in the Fed’s vault; these requests for Fed notes should be met as usual by ordering additional new Fed notes, just as if the vault contained Treasury bonds instead of Fed notes. With this ordering and storing of Fed notes in the Fed’s vault, stimulus without debt would not reduce the Fed’s capital, just as the Fed’s purchase of Treasury bonds does not reduce the Fed’s capital. But if Fed notes held by the public are deleted, as they should be, from the Fed’s list of liabilities on its balance sheet, then my recommendation in the previous paragraph would require one change: the Fed should still order and store in its vault Fed notes equal to the Fed’s transfer to the Treasury, but now the Fed’s vault cash should be used to meet requests from the public

Would Stimulus without Debt Weaken Fed’s Balance Sheet? 147

(including banks) for Fed notes. Here is a numerical example that shows that, with this treatment, stimulus without debt would not change the Fed’s capital (net worth) on a reformed balanced sheet that correctly does not list Fed notes held by the public as a Fed liability. Suppose in a recession Congress enacts a $450 billion fiscal stimulus package. In anticipation of writing a check for $450 billion to the Treasury, the Fed orders $450 billion of new Federal Reserve notes that the Fed stores in its vault. The Fed’s increase in vault cash of $450 billion is recorded as an increase of $450 billion in its assets on its balance sheet. So far, prior to the writing of its check to the Treasury and the implementation of the fiscal stimulus, the Fed’s capital has increased by $450 billion due to the $450 billion increase in Fed notes in its vault. After the Fed writes a $450 billion check to the Treasury, and the $450 billion of fiscal stimulus is implemented, recipients of the stimulus deposit $450 billion into their banks. Now consider three scenarios:

1. Suppose the banks keep $450 billion as deposits at the Fed. These bank deposits at the Fed really are a Fed liability because the Fed must be ready to provide Federal Reserve notes to the banks holding these deposits at the Fed. Thus, this is recorded on the Fed’s balance sheet as an increase of $450 billion in liabilities, matching the increase of $450 billion in Fed assets due to the increase in $450 billion of Fed notes in the Fed’s vault. Thus, from beginning to end, there has been no change in the Fed’s capital because Fed assets have increased $450 billion and Fed liabilities have increased $450 billion. 2. Suppose the banks keep no deposits at the Fed, but instead ask for $450 billion of Federal Reserve notes. The

148

H ow to C ombat R ecession

Fed provides them with $450 billion of Fed notes from its vault, so the Fed records a reduction in its assets of $450 billion, and Fed assets are the same as before the fiscal stimulus. But the $450 billion of Fed notes held by banks are not listed as a liability on the Fed’s balance sheet. Thus, from beginning to end, there has been no change in the Fed’s net worth (capital) because Fed assets are unchanged and Fed liabilities are unchanged. 3. Suppose the banks keep $400 billion as deposits at the Fed, and ask the Fed to provide them with $50 billion in Fed notes. The Fed provides them with $50 billion of Fed notes from its vault, so $400 billion is left in the Fed’s vault and the Fed records a reduction in its assets of $50 billion. The $400 billion of deposits at the Fed is recorded as an increase of $400 billion in liabilities on the Fed’s balance sheet. Thus, from beginning to end, there has been no change in the Fed’s net worth (capital) because Fed assets have increased $400 billion and Fed liabilities have increased $400 billion.

Thus, if the Fed balance sheet is reformed so that Fed notes held by the public (including banks) are no longer listed as Fed liabilities on its balance sheet, and if the Fed’s new vault cash is used to meet requests from the public (including banks) for Fed notes, then stimulus without debt would not change the Fed’s capital on its reformed balance sheet.

Summary The stimulus-without-debt policy would not weaken the Fed’s balance sheet. The policy would not reduce the capital (net worth) of the Federal Reserve on its balance sheet as long as the Fed

Would Stimulus without Debt Weaken Fed’s Balance Sheet? 149

orders and stores in its vault new Federal Reserve notes equal to the Fed’s transfer to the Treasury under the policy. The chapter also considers a reform of the Fed’s balance-sheet accounting and shows how stimulus without debt would not reduce the Fed’s capital on the Fed’s reformed balance sheet.

Chapter 8

Would Stimulus without Debt Undermine the Fed’s Independence?

Would stimulus without debt undermine the current degree of independence of the Federal Reserve from Congress and from the president? I contend that the answer is no for two reasons: (1) The Fed would decide when and whether to transfer funds to the Treasury for fiscal stimulus, and if so, how much; (2) the Fed would decide whether and how much to adjust its monetary policy instruments, taking into account its transfer (if any) to the Treasury, the fiscal stimulus enacted by Congress, its mandate from Congress to pursue high employment and low inflation, and everything else (for example, whether there is a financial crisis). Thus, stimulus without debt does not require the Fed to make the transfer to the Treasury for fiscal stimulus in a severe recession, or specify how large the transfer must be; nor does it tell the Fed how much to adjust its monetary policy instruments. Stimulus without debt would therefore preserve the current degree of independence of the Federal Reserve. In this chapter I don’t address the important question: What is the socially optimal degree of independence of the Fed from Congress and from the president? This is a book about stimulus 150

Would Stimulus without Debt Undermine Fed’s Independence? 151

without debt, not Fed independence. Consequently, in this chapter I focus on whether stimulus without debt would change the current degree of independence of the Fed, not whether the current degree of independence is socially optimal.

The Current Degree of Fed Independence The US Constitution assigns to Congress the power to create money, but Congress has delegated this task to the Federal Reserve, which Congress created. The Fed is governed by congressional legislation—the Federal Reserve Act. Consequently the Fed is certainly not independent of Congress. Also, the president nominates the members of the Board of Governors of the Federal Reserve, including the chairman, and the Senate must approve the nominations. So the Fed is not independent of the president either. But once appointed, members cannot be removed because of their policies—just as Supreme Court justices can’t be removed for their written opinions. In contrast to other government agencies, the Fed has its own source of funding: the interest earnings from its holding of securities (bonds) and from its loans to banks. After using a small portion to cover its expenses, the Fed returns most of its earnings to the Treasury. As a consequence, the Fed does not have to seek appropriations from Congress. Stimulus without debt would not affect this source of Fed independence. The Fed is governed by congressional legislation: the Federal Reserve Act as amended over the years. In a 1977 amendment to the Federal Reserve Act, Congress assigned to the Fed a dual mandate: it specified that the Fed must pursue both high employment and low inflation. In 2008 the Fed was therefore expected to lower interest rates to try to induce borrowing and spending on goods and services to combat the deepening recession and

152

H ow to C ombat R ecession

rising unemployment. In fact, in 2008 the Fed did lower interest rates for this purpose. By contrast, the European Central Bank has been given a single mandate: to pursue low inflation. If Congress had given the Fed the same single mandate prior to 2008, the Fed would have been unable to lower interest rates in 2008 to try to combat the fall in employment. Stimulus without debt would permit, not require, the Fed to make a transfer to the Treasury for fiscal stimulus. This permission would not affect the Fed’s current degree of independence because the Fed would be free not to take this action. But wouldn’t Congress exert great pressure on the Fed to make a large transfer for fiscal stimulus to the Treasury so that Congress’s fiscal stimulus didn’t raise the deficit or debt? Congress might. Wouldn’t Fed decision-makers feel they had to make a transfer as large as Congress wants? No, they would not. Speeches given by current and past members of the board of governors and current and past Fed regional bank presidents (who are also on the Federal Open Market Committee) make clear that most Fed decision-makers have always believed that the Fed should counter excessive fiscal stimulus in order to prevent inflation.

A Biased Definition of Fed Independence In 1977 Congress enacted as an amendment to the Federal Reserve Act that gave the Fed a dual mandate: to pursue both high employment and low inflation. Thus, in 2008 the Fed explained its reduction of interest rates as an attempt to prevent a fall in employment. Some critics believe the Fed should pursue only low inflation, not high employment; they believe pursuing high employment is likely to lead to high inflation. These critics usually view Congress as likely to pursue high employment more

Would Stimulus without Debt Undermine Fed’s Independence? 153

than low inflation, and that the Fed must be independent so that it can counter Congress by pursing only low inflation. These critics contend that the Fed is not really independent unless it pursues only low inflation. One such prominent Fed critic was the late Allan Meltzer, a professor of economics at Carnegie-Mellon and author of a detailed history of the Federal Reserve. In an article entitled “Federal Reserve Independence,” Melzer (2014) said that the Fed’s quadrupling of its balance sheet without oversight during the Great Recession in less than five years at first glance seems to indicate its independence. He contended, however, that much of the balance-sheet expansion was undertaken to finance the outsized budget deficits during the recession, asserting that independent central banks do not finance budget deficits. Thus, Meltzer asserted that “independent” means avoiding a specific action that Meltzer opposes: Fed financing of “outsized” budget deficits. The severe recession of 2008 caused an automatic drop in federal tax revenue well below federal spending, causing the Treasury to borrow the large difference, and in 2009 Congress enacted and the president signed a large fiscal stimulus to combat the recession that further widened the federal deficit, causing the Treasury to borrow more from the public. There was no direct transfer of funds from the Fed to the Treasury. Meltzer clearly disapproves of the budget deficits, the fiscal stimulus, and the Fed’s purchase of Treasury bonds in the open market during the severe recession. But his disapproval does not mean the Fed sacrificed its independence. The Fed bought a large amount of Treasury bonds in the open market because, in the Fed’s judgment (expressed in speeches, op-eds, and congressional testimony by Fed chairman Bernanke), such action would pursue high employment without sacrificing low inflation. As the author of a history of the Fed, Meltzer contended that the original Federal Reserve Act in 1913 did not permit any

154

H ow to C ombat R ecession

Federal Reserve support of the Treasury and that the founders wanted the Fed to follow the gold standard and adhere to a rule that prohibited financing the Treasury and the budget. Those two rules, Meltzer argued, supported an independent Federal Reserve during the 1920s. Whenever the Fed followed a rule or took actions that Meltzer approved of—for example, monetary tightening that reduces inflation while ignoring the short-term negative impact on employment—Meltzer called the Fed “independent.” Consider Meltzer’s comments on Fed chairman (1979–1987) Volcker. Meltzer said Volcker was a relatively independent chairman because he was committed to a policy of reducing inflation. Volcker shared Meltzer’s view that the Fed should reduce money growth to lower inflation. Meltzer, a monetarist, called Volcker a “pragmatic monetarist” who shared Meltzer’s belief in what Meltzer called an “anti-Phillips curve,” which implied that the way to reduce the unemployment rate was to lower expected inflation and actual inflation. Volcker reduced inflation in the early 1980s by cutting Fed money injections (via Treasury bond purchases) into the economy, which caused a sharp rise in interest rates and fall in consumer and business borrowing and spending, which threw the economy into a severe recession with high unemployment; it was this high unemployment and weak product demand that brought down wage increases, cost increases, and price increases. Now consider Meltzer’s comments on Fed chairman (1987– 2006) Greenspan’s independence. Meltzer said Greenspan, who succeeded Volcker, further reduced inflation and was a “relatively independent chairman” who resisted “open criticism” from the George H. W. Bush administration of Greenspan’s anti-inflation policy during the 1992 election year. In my view, the criticism from the Bush administration and others in 1992 was warranted because the unemployment rate rose from 5.6% in 1990 to 7.5%

Would Stimulus without Debt Undermine Fed’s Independence? 155

in 1992, and yet the Fed did not cut the federal funds rate below 3.0% due to its primary focus on preventing inflation. Meltzer said Greenspan rightly followed a Taylor rule instead of erratic discretion, moving interest rates in response largely to inflation with only minor responses to unemployment, and Greenspan’s implementation of the Taylor rule produced a long period of growth, short and mild recessions, accompanied by low inflation, a period now called “the great moderation.” Greenspan was able to maintain Federal Reserve independence because his policy maintained popular support. Following the Taylor rule preserved Fed independence. Endorsing the Taylor rule is quite a concession for Meltzer because the Taylor rule, with the specific numerical weights advocated by Taylor, calls for at least some response to high unemployment. The Taylor rule response, however, is much milder than the response implemented by the Fed under Bernanke in the 2008 recession. Finally, Meltzer turned to the 2008 recession. He gave credit to the Fed for preventing a financial collapse. But then his view of the Bernanke Fed’s actions turned negative. He pointed out that the Federal Reserve pursued the most expansive policy in its history and that idle excess reserves in banks rose from less than $800 billion to more than $2.5 trillion. Meltzer emphasized that this policy financed massive government budget deficits at very low interest rates, which is “the very opposite of what an independent central bank does.” He said he did not know of any example, anywhere, in which base money creation to finance large budget deficits avoided higher inflation. He was disappointed that the Fed had not revealed a credible policy to prevent inflation when the idle reserves were no longer idle, and worried about a possible “tsunami of idle reserves” that would spill over the domestic and international economy. Meltzer made no mention of fiscal stimulus. In my view, fiscal stimulus is much more effective than standard monetary

156

H ow to C ombat R ecession

stimulus in a severe recession; the Fed would have been much more effective if it had made a large transfer to the Treasury to finance a large fiscal stimulus instead of buying bonds. Thus, in a severe recession the Fed should finance a large fiscal stimulus. Meltzer refers to “massive budget deficits” without acknowledging that the recession itself caused large budget deficits. I agree with Meltzer, however, that the Fed must be prepared to raise interest rates once the economy enters a strong recovery and excessive demand threatens to generate inflation; I discussed how the Fed could do this in chapter 6. Finally, Meltzer defines Fed “independence” as the Fed not financing fiscal stimulus and focusing mainly on inflation. But Fed independence is the Fed having the choice of whether to finance fiscal stimulus. Under stimulus without debt, the Fed is permitted, not required, to finance fiscal stimulus. The Fed’s degree of independence is the same whether Congress has prescribed a single or a dual mandate. Thus, Meltzer’s definition of “independence” is biased toward the monetary policy he prefers.

Summary Stimulus without debt would not undermine the current degree of independence of the Federal Reserve from Congress and from the president because the Fed would decide whether to transfer funds to the Treasury and how much, and whether and how much to adjust its monetary policy instruments. Stimulus without debt does not require the Fed to make the transfer to the Treasury in a severe recession, or specify how large the transfer must be; nor does it tell the Fed how much to adjust its monetary policy instruments.

Chapter 9

Can’t Monetary Stimulus Overcome a Severe Recession?

In this chapter I explain why monetary stimulus is not enough in a severe recession. To avoid misunderstanding, I want to underline the important role of the Federal Reserve in helping to combat a severe recession. As I emphasized in c hapter 3, in a severe recession the Fed must act as a lender of last resort by aggressively performing financial rescues and unfreezing credit markets, and must cut interest rates to zero in order to help maintain borrowing and spending by households and firms. These actions are essential, but they are not sufficient to overcome a severe recession because cutting interest rates to zero cannot fully reverse the plunge in aggregate demand that caused a severe recession. To raise aggregate demand all the way to normal, the Fed needs substantial help from fiscal stimulus. There is another channel by which standard monetary stimulus—the buying of Treasury bonds in the open market by the Fed—can work: the portfolio rebalancing effect. If the Fed injects money into the economy by buying Treasury bonds, it will raise their price and reduce their yield. In response, financial investors will try to shift their portfolios toward other 157

158

H ow to C ombat R ecession

assets—stocks and bonds—thereby bidding up their prices. The rise in the prices of stocks and bonds should have a positive effect on consumer and business confidence and on actual consumption and/or investment spending. But in a severe recession, the increase in demand from this effect will not be enough. In this chapter I will evaluate the strength of both channels of standard monetary stimulus—lower interest rates, and higher prices for stocks and bonds—in a severe recession. Although the Federal Reserve, along with many others, underestimated how much the bursting of the housing bubble in 2007 would jolt the entire economy, in the fall of 2008 the Federal Reserve finally recognized the acuteness of the financial and credit crisis and the magnitude of the plunge of the economy. The Fed then acted aggressively to pump reserves into the banking system with two objectives: first, to provide liquidity to banks to withstand the financial crisis and the freezing of credit markets, and to assist the Treasury in rescues of failing firms; second, to lower interest rates by injecting reserves into the banking system in the hope of stimulating substantial borrowing and spending to combat the recession. The provision of liquidity and financial rescues certainly played a key role in overcoming the financial and credit crisis and thereby preventing a full-scale Great Depression. But the Fed was unable to achieve its goal of stimulating substantial borrowing and spending on goods and services despite pumping a huge amount of reserves into the banking system. The injection of reserves failed to stimulate enough borrowing and spending to generate a strong recovery from the deep recession. Thus, after helping prevent a Great Depression in late 2008 by cutting interest rates to zero and assisting financial rescues, the Fed still faced an economy in 2009 and 2010 that showed no signs of recovery, as the unemployment rate continued to rise gradually towards a peak of 10% in late 2009 and then stayed well

Can’t Monetar y Stimulus O vercome a Severe Recession?

159

above 9% through the end of 2010. Why didn’t the cut in interest rates to zero in 2008, 2009, and 2010 induce enough borrowing and spending by households to generate a solid recovery by the end of 2010? The answer is that many households had run up excessive debt between 2000 and 2007 and were now (2008– 2010) trying to avoid taking on more debt, even when offered low interest rates to borrow more. When most households are determined to avoid taking on more debt, cutting interest rates to zero induces only a small increase in borrowing and spending.

Why Cutting Interest Rates Is Not Enough In a severe recession, aggregate demand for goods and services is much too low, and in response to low customer demand, producers have little choice but to cut production and employment. To combat a severe recession, there must be a large increase in customer demand for goods and services. When the Fed buys Treasury bonds from private bondholders through its standard open-market operations, the private bond sellers deposit their Fed checks in their banks. The bond sellers have money instead of bonds, so their wealth is largely unchanged and there is no reason to expect them to significantly increase their demand for goods and services. Their banks have excess reserves that they would like to lend provided borrowers will be able to repay them. But in a severe recession, banks may doubt that borrowers will be able to repay, and even if banks offer to lend at a very low interest rate, consumers and businesses will be reluctant to borrow and spend because they don’t believe they will be able to repay the principal plus interest on the loan. So the Fed’s standard open-market operations will be unable to lift aggregate demand for goods and services nearly enough to propel a strong recovery from the severe recession. The same is true if the Fed

160

H ow to C ombat R ecession

makes loans to banks. Even if the banks offer to lend at a very low rate, consumers and businesses will be reluctant to borrow. Thus, the Fed’s standard open-market operations will inject reserves into banks, but will only modestly increase demand for goods in services in the economy. Even if the Fed offered to make loans directly to consumers and businesses, few consumers and businesses would be willing to borrow. Recall that in chapter 3 I paraphrased what Nobel laureate Paul Samuelson (1948) wrote in the first edition of his enormously influential college textbook (which went through many editions for decades). Samuelson explained that monetary stimulus alone can’t overcome a severe recession because “You can lead a horse to water, but you can’t make him drink.” The central bank can buy bonds with new money, but it can’t make the recipients of new money spend that money in a deep recession. The central bank can get banks to lower the interest rates they charge households and firms by injecting money into the banks, but it can’t make households and firms borrow and spend in the deep recession. Samuelson expressed the view of traditional Keynesian economists. Many economists today, however, reject this view. New Classical economists and even many New Keynesian economists reject the basic tenets of traditional Keynesian economics. New Classical economists argue that the economy will automatically recover promptly from recession as long as the central bank keeps the money supply from falling; fiscal stimulus is unnecessary, ineffective, and harmful because it generates large government deficits and debt. New Keynesian economists are Keynesian in their view that keeping the money supply from contracting is not enough to generate a quick recovery from a deep recession; but until the Great Recession, many New Keynesians believed that active standard monetary stimulus—cutting interest rates—would be sufficient to generate a satisfactory

Can’t Monetar y Stimulus O vercome a Severe Recession?

161

adequate recovery from a severe recession, and that fiscal stimulus was unnecessary, or ineffective, or harmful, or undermined by politics. The Great Recession caused some New Keynesians to reconsider one crucial tenet of traditional Keynesian economics as they realized that fiscal stimulus is essential for overcoming a deep recession (Romer 2011). Even if monetary stimulus alone could overcome a severe recession, I contend it would be better to rely mainly on tax rebates to households than on cutting interest rates. With tax rebates, households can spend more on goods and services without going deeper into debt by borrowing. By contrast, interest rate cuts will only boost spending if households are willing to take on more debt during a deep recession. With tax rebates, household demand for goods and services is spread across all goods and services. With interest rate cuts, the increase in demand is concentrated on interest-sensitive sectors, a much smaller portion of the economy. Moreover, relying on monetary stimulus alone means interest rates must be cut to zero and kept there for a long time. This causes an unfair burden on retirees who built up savings in their bank account during their work years, hoping that in retirement their bank would pay a normal interest rate. With tax rebates restoring demand for goods and services, interest rates would stay normal instead of being cut to zero.

Why the Portfolio Rebalancing Effect Is Not Enough In a speech to the annual Jackson Hole conference on August 31, 2012, Chairman Bernanke (2012) set out his view of how Fed purchases of long-term government securities stimulated spending on goods and services. Bernanke said that these Fed purchases lowered long-term interest rates and also induced

162

H ow to C ombat R ecession

the private sector to rebalance its portfolio by purchasing corporate stocks, thereby causing a rise in stock prices and private sector wealth, which in turn induced more consumption and investment spending. After referring to the Fed’s purchases of long-term government securities, Bernanke said these purchases would stimulate demand for goods and services through a portfolio balance channel noted by Nobel laureates Tobin, Friedman, and Modigliani. For example, Fed purchases of mortgage-backed securities (MBSs) should raise the prices of those securities; as investors rebalance their portfolios by replacing the MBSs sold to the Federal Reserve with other assets, the prices of the assets they buy should rise. These rising asset prices should raise wealth, resulting in more spending on goods and services. Bernanke contended that Fed bond purchases had led to rising stock prices, wealth, and spending on goods and services. Thus, even if households don’t borrow more just because interest rates have been cut to zero, a rise in the value of their corporate stocks, which raises their wealth, may increase their consumption. Even if business managers don’t borrow more just because interest rates have fallen to zero, a rise in the value of their corporation’s stock may increase their investment. Bernanke then cited simulations with the Federal Reserve’s FRB/US model of the economy that found that the first two rounds of asset purchases by the Fed during the Great Recession raised the level of output by almost 3% and increased private payroll employment by more than two million jobs relative to what otherwise would have occurred. However, in my view it is likely that the Fed’s model overestimated the impact of the Fed’s stimulus in the Great Recession because its estimates use data from all past quarters, capturing the response to a cut in interest rates on stock prices and spending in a normal economy, when household and business confidence is normal; the response in a severe recession when confidence has plunged would surely

Can’t Monetar y Stimulus O vercome a Severe Recession?

163

be smaller. Bernanke conceded this point when he noted that the impact in a deep recession might be less than a model with coefficients estimated mainly from nonrecession US data. But even if the Fed’s estimate of an increase in employment of 2 million was accurate, 2 million was not nearly enough to overcome the severe recession, because from the first quarter of 2008 to the fourth quarter of 2009 employment fell 8 million, from 138 million to 130 million (and the unemployment rate still rose 4.9% from 5.0% to 9.9%). Without the Fed’s bond purchases employment would have fallen 2 million more to 128 million, but to overcome the severe recession the portfolio rebalancing effect would have had to increase employment by 10 million, not 2 million.

Household Debt and Consumption from 2000 through 2008 In their book House of Debt (2014), Atif Mian (professor of economics at Princeton) and Amir Sufi (professor of finance at the University of Chicago) explain why the cut in interest rates to zero in 2008, 2009, and 2010 failed to generate a solid recovery by the end of 2010. They make the case that excessive consumer borrowing from 2000 to 2007 in the United States led to an unsustainable rise in the ratio of household debt to household income. When overindebted households finally cut sharply their consumption in 2008 in an attempt to pay down some of their debt, the sharp fall in aggregate demand for goods and services generated a severe recession, and a cut in interest rates to zero in 2008, 2009, and 2010 was unable to induce overindebted households to resume borrowing and spending again. They begin their book with the fact that the United States saw an exceptional rise in household debt between 2000 and

164

H ow to C ombat R ecession

2007—total debt doubled from $7 trillion to $14 trillion, and the household debt-to-income ratio shot up from 1.4 to 2.1. From 1970 to 2000 this ratio rose very gradually, but then accelerated dramatically from 2000 to 2007. Mian and Sufi cite studies by economists who present evidence that a similar huge run- up of household debt relative to income occurred in the 1920s prior to the Great Depression that began in 1929. From 1920 to 1929, there was an explosion in mortgage debt and installment debt for cars and furniture. Mortgages tripled. The spread of an innovation, installment financing, revolutionized how people bought washing machines, cars, and furniture. According to Martha Olney, an economic historian of consumer credit, in her 1999 article entitled “Avoiding Default: The Role of Credit in the Consumption Collapse of 1930,” the 1920s were the turning point in the history of consumer credit. For the first time, many consumers were going into debt to buy durable goods, made possible by the availability of credit. As a consequence, consumer spending in the 1920s rose faster than income, and consumer debt as a percentage of household income more than doubled. One scholar wrote in 1930 that the prosperity of the 1920s was propelled by a huge expansion in credit by retailers and debt by households. Like the Great Recession, the Great Depression began with a large drop in household consumption. Mian and Sufi (2014) cite the economic historian Temin, who wrote that the consumption decline at the start of the Great Depression was too big to be explained just by falling income and prices; Mian and Sufi believe the consumption decline was mainly due to a decade of debt accumulation by households. They point out that other countries in other periods have followed the same sequence, and severe recessions abroad have often been preceded by a large run-up of household debt followed by a sharp drop in household spending. Moreover, there is evidence that the bigger the increase in debt, the sharper the

Can’t Monetar y Stimulus O vercome a Severe Recession?

165

fall in spending. Mian and Sufi cite a 2010 study of the Great Recession in the 16 OECD countries showing that countries with the largest increase in household debt from 1997 to 2007 were exactly the ones that suffered the largest decline in household spending from 2008 to 2009. They dispute the view that the failure of Lehman Brothers in September 2008 was the main cause of the US recession. Instead, the main cause was the overindebtedness of households that peaked in 2007. They say that the decline in spending was in full force before the fall of 2008. The National Bureau of Economic Research dates the beginning of the recession in the fourth quarter of 2007, three quarters before the failure of Lehman Brothers. The collapse in residential investment and durable consumption was dramatic well before the events of the fall of 2008. What happened in the fall of 2008 intensified a recession that was already well underway. Mian and Sufi emphasize that the plunge in residential investment began in 2006. Starting in the second quarter of 2006, residential investment declined substantially. Spending on durable goods started declining before September 2008. They compare consumer spending in January through August 2008 to January through August 2007 and say a clear pattern emerges. In 2008, auto spending was down 9%, furniture spending was down 8%, and home-improvement expenditures were down 5%. These declines all occurred before the failure of Lehman Brothers. Hence, the sharp cut in household spending on durable goods had to have been triggered by something other than the financial crisis in September 2008. The timing suggests that the decline in household consumption spending, not the financial crisis, was the main cause of the recession. Job losses rose because households stopped buying. The decline in business investment was a reaction to the massive decline in household spending. When businesses see no demand for their products, they cut back on investment.

166

H ow to C ombat R ecession

With households finally determined to reduce their debt in 2008, monetary policy had little power to induce them to spend more by borrowing more. Cutting interest rates can induce more borrowing and spending when consumers are willing and able to take on more debt, but not when they are determined to reduce their excessive debt. Nor can cutting interest rates induce firms to invest more when consumers are cutting their spending. Recall that in chapter 3 I cited the survey of small-business managers by the National Federal of Independent Business (NFIB) reported by Mian and Sufi; it found that the most important concern of managers during the Great Recession was low customer demand, not access to credit or lower interest rates.

The Weakness of Monetary Stimulus in a Balance-Sheet Recession The phrase “balance-sheet recession” is used by Richard Koo in three books (2003, 2007, 2015). In his most recent book, Koo (2015) asserts that the bursting of a stock market or housing market bubble results in a balance-sheet recession. In such a recession, the private sector is determined to avoid taking on more debt because debt incurred during the bubble remains, while the value of assets bought with borrowed funds collapses when the bubble bursts, leaving balance sheets underwater. With most households and firms trying to pay down debt, even zero interest rates won’t induce much borrowing and spending. Koo says that Japan’s recession that began in 1990 sheds light on the Great Recession in the United States and Western Europe in 2008. He points out that Japan’s recession that began in 1990 with the bursting of a massive debt-financed bubble. Koo explains why Japanese firms in the 1990s, following the 1990 plunge in the prices of real estate, sharply reduced borrowing

Can’t Monetar y Stimulus O vercome a Severe Recession?

167

and spending despite zero interest rates. After the bubble burst, commercial real estate prices plunged 87% from their peak. Stock prices also plunged. Koo says the loss of national wealth amounting to three full years of GDP. The money borrowed by households and businesses to acquire real estate and corporate stocks assets remained intact, sending many households and firms underwater—their debt exceeded the new value of their assets. Koo points out that the bursting of Japan’s real-estate and stock-market bubble in 1990 should have caused a sharp fall in GDP. But it didn’t. Why, then, did Japan’s GDP never fall below its bubble-era peak? Koo says the answer is that the government decided to borrow and spend. Government spending kept GDP above the bubble-era peak in spite of a dramatic cutback in spending by households and firms. The boost in government spending despite low tax revenue required substantial government borrowing year after year and left Japan with a huge public debt. But if the government had not stimulated the economy in this way, Koo argues, GDP would probably have fallen to half or less than half its peak. Consider that the crash in US asset prices during the Great Depression destroyed wealth equivalent to a year of 1929 GDP, and as a consequence output plunged 46%. Japan lost wealth equal to more than three years of 1989 GDP, so the resulting hit to the economy would almost certainly have been substantially greater. But this disastrous outcome was averted only because the government administered fiscal stimulus early on and continued to do so over an extended period of time. Japan’s fiscal stimulus, Koo concludes, was therefore extremely successful. Koo contends that when the central bank buys bonds in the open market, and sellers of bonds deposit central bank checks in their bank accounts, the banks are forced to hold the funds because there are few willing borrowers. The problem is not that

168

H ow to C ombat R ecession

the banks are unwilling to lend, but that potential borrowers are unwilling to borrow because of the debt already weighing down their balance sheets. When the private sector stops borrowing money even at an interest rate of zero, any funds supplied to financial institutions by the central bank remain within the financial system because there are no borrowers. That is why growth in private credit and the money supply has been so sluggish post- Lehman despite dramatic expansion of the monetary base by central banks. The key implication here is that the effectiveness of monetary policy diminishes dramatically as the private sector tries to reduce its debt. Koo hastens to add that monetary stimulus can play an important role in a financial crisis. He emphasizes that after a bubble bursts, monetary policy has a vital role to play. When a bubble collapses, two problems arise that must be carefully distinguished. The first is a balance-sheet recession, which is a borrower-side problem. The second is a financial crisis, which is a lender-side problem. A financial crisis occurs when the value of assets owned or held as collateral plunges after a bubble bursts, leaving many financial institutions holding loans unlikely to be repaid. These institutions take defensive actions by hoarding cash and avoiding any new loans or investments. As fear spreads among these institutions, the financial system freezes up. During such a panic, the financial system can collapse unless the central bank steps in as a lender of last resort and provides institutions with the funds they need for settlement. Providing liquidity at such times is an essential function of a central bank. After the Lehman failure, the Federal Reserve rightly injected funds to rescue the US financial system and succeeded in preventing its collapse. These Fed rescues deserve great praise. Once the financial crisis was over, however, Koo believes, the Federal Reserve should have recognized that it was unable to solve the first problem: the balance-sheet recession. Instead

Can’t Monetar y Stimulus O vercome a Severe Recession?

169

of insisting it could cure a balance-sheet recession by making huge bond purchases and lowering interest rates, the Fed instead should have repeatedly asked Congress to extend the two-year fiscal stimulus that had been enacted in 2009.

Blinder and Zandi’s Estimates of the Impact of Policies Blinder and Zandi (2015) use a macroeconometric model developed by Zandi (chief economist of Moody’s Analytics) to estimate the impact of monetary stimulus and fiscal stimulus on the economy using quarterly data from the past several decades. Blinder and Zandi agree that their study does not attempt to estimate stimulus impacts by studying data generated during the Great Recession. Instead, given their estimates of stimulus impacts from past data, and the assumption that these past stimulus impacts continued to hold during the Great Recession, they infer how much worse the economy would have been during the Great Recession if the stimulus had not been implemented. I agree with them, and most other users of macroeconometric models, that estimating policy impacts using past data, and then assuming these policy impacts continued during the period of interest (for example, the Great Recession years beginning with 2008), is one reasonable way to make inferences about the impact of stimulus policies during the period of interest. Zandi’s model is a traditional Keynesian macroeconometric model. In the 1960s, traditional Keynesian demand-oriented models were regarded by most economists as the best macroeconometric models. In the 1970s these demand-oriented models were initially unable to handle supply shocks like the oil price increase of the 1970s, but were soon modified to include supply shocks while retaining Keynesian demand. Since

170

H ow to C ombat R ecession

the mid-1970s many macroeconomists have criticized these models for lacking an explicit microeconomic foundation and have tried to develop models built on such a foundation. But many of these microfoundation models have done a poor job of explaining and tracking the Great Recession—for example, some of these models assumed that any involuntary unemployment would be quickly eliminated by a fall in wages that would induce employers to hire everyone willing to work. By contrast, traditional Keynesian models did a good job explaining and tracking the Great Recession. Having studied the structure and specific equations in the Zandi model, I believe it is well suited to explain and track the Great Recession. Blinder and Zandi estimate that if there had been no financial rescues, no monetary stimulus other than the usual cutting of short-term interest rates in a recession (no Fed purchase of long-term securities, “quantitative easing,” in an attempt to cut long-term interest rates), and no fiscal stimulus other than the usual automatic stabilizers, then the peak quarterly unemployment rate would have been 15.8% instead of its actual 9.9%, an increase of 5.9%, and in 2010 the unemployment rate would have been 15.0% instead of its actual 9.6%, an increase of 5.4%; 17 million jobs would have been lost, 9 million more than the actual job loss of 8 million; the peak-to-trough decline in real GDP would have been 14% instead of the actual decline of 4%; the economy would have contracted for three years instead of the actual contraction of a year and a half; and the budget deficit would have peaked at 20% of GDP instead of its actual peak of 10%. They then isolate the impact of each policy. Removing rescues and monetary stimulus would have increased the 2010 unemployment rate 2.7 percentage points (from 9.6% to 12.3%); removing fiscal stimulus would have increased the 2010 unemployment rate 1.2 percentage points (from 9.6% to 10.8%); because of positive interaction effects, removing all policies would

Can’t Monetar y Stimulus O vercome a Severe Recession?

171

have increased the 2010 unemployment rate 5.4 percentage points (from 9.6% to 15.0%). Dividing the policies into three categories—rescues, Fed purchases of long-term securities, and fiscal stimulus—each contributed roughly one-third of the mitigation of the severity of the recession. But even with the significant improvement from these policies, the unemployment rate in 2010 was a still very high 9.6%. Which of these policies could have been significantly increased from 2008 to 2010 to achieve a significantly lower unemployment rate in 2010? Rescues in the fall of 2008 and winter of 2009 were sufficient to end the financial panic and credit freeze, so more rescues were not needed and would not have further improved the economy. Purchase of long-term securities under “quantitative easing” in 2009 and 2010 reduced long-term interest rates, so it is unlikely that more purchases of long-term securities would have induced much more borrowing and spending. By contrast, large tax rebates (triple the size of the 2008 rebates) could have been sent out in 2008, 2009, and 2009, and this would have generated a significant increase in consumer spending even if a significant portion of the rebates were saved. There is no shovel-ready problem with tax rebates—the larger the rebate households receive, the more they will spend on goods and services. The only thing holding back the enactment of large tax rebates is concern about government deficits and debt. The key rationale for having the Federal Reserve give a transfer (not a loan) roughly equal to the federal expenditure on rebates—the stimulus-without-debt policy proposed in this book—is to make sure the rebates (or other fiscal stimulus) do not increase government deficits and debt. Thus, the conclusion I draw from the Blinder and Zandi study is that rescues, monetary stimulus (through the purchase of both short-term and long-term securities), and fiscal stimulus

172

H ow to C ombat R ecession

all kept the Great Recession from being significantly worse. But the Great Recession was still bad enough despite these policies, and more monetary stimulus would not have been able to make much more improvement. By contrast, fiscal stimulus (primarily tax rebates to households) could have been made much larger— tripled, for example—and could therefore have made a much larger improvement in the economy without increasing federal deficits or debt if the Fed gave a transfer (not a loan) to the Treasury that was roughly equal to the fiscal stimulus.

Conclusion In a normal economy, when households have modest debt, monetary policy can stimulate substantial borrowing and spending by reducing interest rates. But monetary policy can’t stimulate substantial borrowing and spending in a severe recession when households have excessive debt. Monetary stimulus provided some help during the Great Recession, but it is clear that monetary stimulus alone is not nearly enough to overcome a severe recession. The Great Recession taught once again the lesson of the Great Depression: lowering interest rates to zero is simply no match for a severe recession.

Chapter 10

Can Stimulus without Debt Be Used by Other Countries?

Thus far I have focused exclusively on the United States. What about other economically advanced countries?

An Economically Advanced Country with Its Own Central Bank Stimulus without debt for any economically advanced country with its own central bank is the same as for the United States. In a recession, the central bank makes a transfer to the national government’s treasury, and the national government’s legislature enacts fiscal stimulus roughly equal to the transfer from the central bank. The largest portion of the fiscal stimulus should be tax rebates to households, but the legislature should also include other kinds of fiscal stimulus in the package. The central bank therefore sets the amount of fiscal stimulus that can be implemented without increasing government debt. The national legislature sets the composition of the fiscal stimulus package and the national executive implements through the 173

174

H ow to C ombat R ecession

treasury the enacted fiscal stimulus package. The national legislature can enact a fiscal stimulus that exceeds the central bank’s transfer to the national treasury, but then the treasury must borrow an amount equal to the difference. If a transfer from the central bank to the treasury in recession to finance fiscal stimulus is illegal under current law, the legislature should enact a law that makes such a transfer legal.

Stimulus without Debt in the Eurozone In the eurozone there is one central bank, the European Central Bank (ECB), but many member countries, each with its own national government and legislature for setting government spending, tax rates, and fiscal stimulus. I propose that under a stimulus-without-debt plan for the eurozone, the ECB would use a formula to determine the transfer it would give to the treasury of each national government in a recession, just as the US Congress has used a formula to determine the federal transfer it gives to the treasury of each state government in a recession and also for nonrecession programs. Under the stimulus-without-debt plan, both the ECB and the legislatures of each member government would play crucial roles. The ECB would make transfers (not loans) to each treasury, and each national legislature would enact a fiscal stimulus package roughly equal to the transfer it receives from the ECB. The largest portion of the fiscal stimulus should be tax rebates to households, but the national legislature should also include other kinds of fiscal stimulus in the package. The ECB’s transfer therefore sets the amount of fiscal stimulus that can be implemented by each national government without increasing its deficits and debt. Each legislature sets the composition of the fiscal stimulus package, and each national government implements through

Can Stimulus without Debt Be Used by Other Countries?

175

its treasury the enacted fiscal stimulus package. A legislature can enact a fiscal stimulus package that is larger than the ECB’s transfer to its treasury, but its treasury must then borrow the difference between its fiscal stimulus package and the transfer it receives from the ECB. The ECB needs to use a formula that determines the magnitude of the transfer that it would give to each national treasury for the purpose of financing fiscal stimulus in a recession. Suppose the ECB decides to give each country the same transfer per person (R*). Then each country’s treasury would receive a total transfer from the ECB equal to R*Ni, where Ni is the country’s total population. Of course, the ECB might want a formula that doesn’t give the same transfer per person to each country. For example, in a recession the rise in each country’s unemployment rate differs. If the rise in a country’s unemployment rate is above average, its transfer per person could be set above R*, and if a country’s rise is below average, its transfer per person could be set below R*. An online column by John Muellbauer (2014), professor of economics at Oxford University, appeared with the title “Combating Eurozone Deflation: QE for the People.” He said that printing large sums of money and distributing it to the public would increase the demand for goods and services in a recession. But the ECB would have to find a way to ensure fair distribution. He noted that a simple solution would be to distribute the funds to governments and let the governments decide how best to spend them in their countries. But he contended that the EZ’s rule against using the ECB to finance government spending would prevent this simple solution. Instead of calling for a change in the EZ’s rule, Muellbauer proposed that the European Central Bank give €500 to each person in the eurozone. He recommended providing all workers and pensioners who have social security numbers (or the local equivalent) with a payment from the ECB, which governments

176

H ow to C ombat R ecession

would help distribute. He noted that another option would be to use the electoral register, a public database that the ECB could use independently of governments. Roughly 90% of the 275 million adults in the eurozone are on the electoral register. He contended that there is nothing in EZ law that forbids the ECB from undertaking such an independent action. Muellbauer said there is an advantage to having the ECB implement this policy by itself: a central bank might be more responsible than politicians worried about winning elections. Elected politicians who control national governments might be tempted to boost spending on popular programs or cut taxes as an election approaches, thereby generating inflation and running up government debt, whereas the ECB wouldn’t be subject to such temptation and would be trusted by the public to set the magnitude of the transfer to hit its low inflation target. I’m not persuaded by this argument. In my judgment, it is better to have a division of labor and checks and balances between the unelected central bank and the elected national legislatures. Have the central bank decide whether to give a large transfer to each national treasury in a recession, and if so, how much. Have each national legislature decide how much fiscal stimulus to enact in a recession, and what the composition of the fiscal stimulus package should be—the composition of spending, tax rebates, and tax cuts. The legislature can enact a fiscal stimulus larger than its transfer from the central bank, but then it must borrow the difference by selling new government bonds, thereby increasing its national debt. If the ECB is prohibited under current regulations from making transfers (not loans) to member national governments, then I recommend, as I do for the United States, that a change be made in the regulations to permit such transfers in a recession. In an online column entitled “Legal Helicopter Drops in the Eurozone” (February 24, 2016) Eric Lonergan (2016) asserts

Can Stimulus without Debt Be Used by Other Countries?

177

that helicopter drops in the eurozone may be a legal obligation of the ECB. He says the rules governing the ECB are clear. The objective of the ECB is price stability, which the ECB has defined as a 2% increase per year. The ECB must act independently. Policies aimed at financing the budgetary policies of governments are prohibited. Finally, the ECB can use whatever monetary operations are consistent with these constraints. Like Muellbauer, Lonergan recommends that the ECB implement helicopter money without the help of national governments. But in contrast to Muellbauer, who recommends having the ECB give transfers to households, Lonergan recommends having the ECB make loans. Why? Lonergan is worried about the ECB’s balance sheet. If the ECB makes a loan, its net worth is unchanged, whereas its net worth falls (under conventional accounting) when it gives a transfer. But if households must pay back a loan they didn’t seek, they won’t spend it. To work—to get households to increase their spending in a recession—households must receive a transfer, not a loan. If they don’t have to pay the “loan” back, then it is really a transfer, and ECB accountants won’t count the “loan” as an asset on the ECB’s balance sheet. So Lonergan’s solution to the balance-sheet problem won’t work. I discussed this central bank balance-sheet problem briefly in chapter 3 and more fully in chapter 7. Recall that my solution to the balance-sheet problem is to have the central bank order an amount of new central bank notes equal to its transfers to the national treasury or treasuries and to store these notes in the central bank’s vault; the central bank vault cash would clearly be an asset on the central bank’s balance sheet, and as a consequence of this new vault cash, the central bank’s transfers would not reduce its net worth. My solution to the balance-sheet problem applies whether the central bank gives transfers to national treasuries, as I propose, or to households, as Muellbauer and others propose.

178

H ow to C ombat R ecession

Conclusion Stimulus without debt can be implemented in any economically advanced country with its own central bank that it able to print its own government paper money. Stimulus without debt can also be implemented in the eurozone; it just requires agreement on a formula to determine the transfer to each national treasury in the zone. The US Congress has often enacted such a formula to determine the federal transfer to each state government, and the eurozone should be able to do the same.

Chapter 11

What Have Others Written about Stimulus without Debt?

In the past few years there have been many blogs, short articles, and comments through various media about stimulus without debt and helicopter money. In this chapter I will review some of these writings and react to them. But I want to begin with an article written decades ago by Milton Friedman.

Milton Friedman’s Helicopter Parable Friedman (1969) asked readers to imagine, as a thought experiment, a helicopter drop of money on a fully employed economy with national output (income) $10,000 per capita, and money $1,000 per capita held by the population. Each dollar number is an average; some individuals have higher-than-average income or money, others lower. Friedman analyzed the effect of a helicopter drop of $1,000 per capita—a doubling of the money held by the population to $2,000 per capita. Initially Friedman assumes each person picks up an amount of money equal to the amount he held before. Each now has more wealth than before, 179

180

H ow to C ombat R ecession

so each person would decide to spend more. But with resources fully employed so that real output (Q) cannot increase, the result of this greater spending is an increase in prices (P)—temporary inflation. Friedman says that prices would double, real output and income per capita would stay constant, and nominal income per capita would double (from $10,000 to $20,000). Recall that in my brief discussion of helicopter money in chapter 3, I said that stimulus-without-debt is similar to helicopter money in some ways but differs in others. It’s similar because in both the helicopter money parable and the stimulus- without-debt plan the population receives newly created money as a transfer that can be kept, not a loan that must be repaid. Stimulus without debt differs from helicopter money because stimulus without debt creates checks and balances by assigning specific roles to specific institutions in its implementation. In the parable, the helicopter drops the new money on the population— it’s a thought experiment, obviously not intended as a practical proposal for a modern economy. In stimulus without debt, the Federal Reserve creates the new money and gives it as a transfer (not a loan) to the Treasury, and after authorization by Congress, the Treasury mails tax rebate checks to the population. Stimulus without debt is intended as a practical proposal for a modern economy. In Friedman’s helicopter parable the economy is at full employment when the helicopter drop occurs, whereas the stimulus-without-debt policy is used only when the economy is in recession, with the actual output of the economy well below potential output. Hence, in Friedman’s parable, helicopter money causes a rise in prices to a new level—a temporary inflation—whereas stimulus without debt in a recession causes an increase in real output and employment, not inflation. The sole purpose of Friedman’s parable is to show that in a fully employed economy an injection of money will raise the price level— it will cause a temporary inflation. Friedman made no comment

W hat Have Others Written about Stimulus without Debt? 181

on what would happen if the helicopter dropped money on an economy in recession rather than at full employment. By contrast, my analysis in chapter 6 explained why stimulus without debt would move the economy from recession to full employment with little or no rise in prices. In Friedman’s helicopter parable the money in the economy increases by the amount dropped from the helicopter, whereas under stimulus without debt, the Fed, in order to hit its interest rate target, will offset its money injection from its transfer to the Treasury by cutting its purchase of Treasury bonds in the open market. Hence, under stimulus without debt the net increase in money in the economy is likely to be much less than the Fed’s transfer to the Treasury. Two key elements of stimulus without debt are (1) fiscal stimulus; (2) a transfer (not loan) from the Fed to the Treasury for fiscal stimulus so the Treasury doesn’t have to borrow (sell new bonds) to finance the fiscal stimulus. In his helicopter parable Friedman never used the phrase “fiscal stimulus” or “fiscal policy” and gives the false impression that his parable is solely about monetary stimulus or monetary policy. But this is incorrect. In our actual society, in contrast to Friedman’s imaginary experiment, cash transfers (such as tax rebates) to households are delivered by checks mailed out by the US Treasury under congressional legislation authorizing the transfers—by fiscal policy, not monetary policy. Friedman’s helicopter delivers a combined fiscal-monetary expansion.

A Central Bank Transfer to the Treasury in Recession A key component of the stimulus-without-debt proposal is a large transfer (not loan) from the central bank to the Treasury

182

H ow to C ombat R ecession

so that the Treasury does not have to borrow (sell new Treasury bonds) in order to pay for fiscal stimulus enacted by Congress in a recession. When I published my stimulus-without-debt article in 2013, I was unaware of a short online column by Ricardo Caballero (2010). As far as I can tell, he has not written anything else on his proposal. Caballero, a professor of economics at MIT, noted that insufficient demand for goods and services kept output and employment far below potential in 2010, and what was needed was a boost in aggregate demand. He pointed out that, unfortunately, the Federal Reserve has the resources—money that it can create—to increase aggregate demand, but not the instruments of fiscal policy (like tax rebates) to get that money to people who will spend it. It is Congress, not the Fed, that has the fiscal policy instruments (like tax rebates), but not the resources (the ability to create money). It stands to reason, he wrote, that what is needed is a transfer from the Fed (the money creator) to Congress (which controls the fiscal instruments) so the Treasury, the agent of Congress, can implement its fiscal stimulus (such as mailing tax rebate checks to households). In chapter 3 I gave a concise exposition of the main points from my 2013 article entitled “Stimulus without Debt,” which made the case for the policy in this book. Most but not all of the aspects treated in depth in this book were at least mentioned in that article. To my knowledge, my 2013 article was the first to give a lengthy exposition of stimulus without debt. A direct transfer from the Fed to the Treasury in order to achieve stimulus without debt was proposed in a Wall Street Journal op-ed by Daniel Arbess (2013), an investor and policy analyst. Arbess noted that the Fed printed about $2.5 trillion of new money during the Great Recession, but much of it sat in banks instead of getting into the economy through household spending on goods and services. He said there was a more direct

W hat Have Others Written about Stimulus without Debt? 183

way to channel Fed money into the economy. He recommended that the Fed bypass the credit channel: instead of buying bonds in the open market, which merely led to a buildup of funds at banks, the Fed should send cash straight to the Treasury, where it would be deployed as directed by Congress. Adair Turner (2013), a former chairman of Britain’s Financial Services, recommended “overt monetary finance” of fiscal stimulus in a recession, but his recommendation was not stimulus without debt because he called for the treasury to sell new bonds to the public to pay for the fiscal stimulus and the central bank to buy treasury bonds from the public with new money. His proposal was a standard combined fiscal-monetary expansion with an increase in government debt equal to the fiscal stimulus. But two years later Turner (2015) proposed stimulus without debt. Under his 2015 proposal, in a recession fiscal stimulus—a tax cut or increase in government spending enacted by the national legislature—would be financed by having the central bank deposit newly created money into the treasury’s bank account in an amount equal to the fiscal stimulus. There would be no sale of new government bonds by the treasury and no increase in government debt.

Bernanke’s 2016 Blog on Stimulus without Debt Before turning to Ben Bernanke’s significant 2016 blog urging serious consideration of stimulus without debt for the next severe recession, I want to go back to 2000 to describe the evolution of Bernanke’s thinking prior to 2006 when he became Chair of the Federal Reserve (Seidman 2006). Before he became a member of the Federal Reserve Board, Bernanke (2000) wrote an article on how Japan could stimulate a strong recovery from

184

H ow to C ombat R ecession

its lingering recession. Bernanke referred to Friedman’s helicopter drop but noted that the Bank of Japan (BOJ) doesn’t have the authority to rain money on the Japanese population. He recommended that Japan make “money-financed tax cuts” to domestic households: Japan’s treasury would sell government bonds to the public to obtain the funds needed to maintain government spending despite the tax cut, and the BOJ would buy the bonds from the public with new money. He stated that the policy he proposed—a money-financed tax cut—would be a combination of fiscal and monetary policies and said that some cooperation would be required between the BOJ and Japan’s legislature. It was not this 2000 article, however, that led some to call him “Helicopter Ben.” Instead, it was a single sentence from a speech (Bernanke 2002) he gave after becoming a member of the Fed board of governors in which he expressed support for a money-financed tax cut, and said that a money-financed tax cut is “essentially equivalent” to Milton Friedman’s famous helicopter drop of money. Bernanke, however, in his 2000 article and his 2002 speech did not propose a transfer from the BOJ to Japan’s treasury. He proposed instead that Japan’s treasury borrow to pay for the fiscal stimulus by selling new bonds to the public, and that the BOJ buy the bonds from the public with new money. Under his proposal, Japan’s government debt would increase by the amount of the fiscal stimulus. Thus, Bernanke’s proposal was not stimulus without debt. It was a standard combined fiscal- monetary stimulus with the increase in government debt held by the central bank rather than the public. In his article Bernanke (2000) made an important point concerning the independence of Japan’s central bank. He noted that his proposal required some cooperation between the BOJ and Japan’s treasury, and that this cooperation would not undermine BOJ independence. The BOJ would be taking a voluntary action

W hat Have Others Written about Stimulus without Debt? 185

that promoted its goal. Bernanke’s point about central independ ence applies equally to stimulus without debt in the United States. The central bank would be taking voluntary action— giving a transfer to the Treasury—in pursuit of its mandate set by Congress to pursue high employment and low inflation. As Fed chairman during the Great Recession, Bernanke took essential actions to combat the financial crisis. He also called on Congress to provide fiscal stimulus, implying that he recognized that monetary stimulus alone was not enough to overcome a severe recession. However, he never publicly raised the possibility of the Fed giving a transfer to the Treasury so that Congress could enact fiscal stimulus without increasing government debt. It was only after leaving the Fed and completing his book on his tumultuous term at the Fed that he publicly considered the possibility of stimulus without debt. In his April 2016 blog post, Bernanke took the significant step of addressing stimulus without debt, which he called a “Money- Financed Fiscal Program” (MFFP). Bernanke said that an MFFP should be seriously considered in the next severe recession. He gave this illustration. Congress would approve a $100 billion one-time fiscal program—a $50 billion increase in public works spending and a $50 billion one-time tax rebate. Bernanke said this would increase the deficit $100 billion. Instead of paying for the $100 billion fiscal stimulus by issuing new government debt to the public, the Fed would credit the Treasury with $100 billion in the Treasury’s checking account at the Fed, and the Treasury would use those funds to pay for the new spending and the tax rebate. He then added that, “alternatively and equivalently,” the Treasury could issue $100 billion in debt that the Fed agreed to purchase and hold indefinitely, rebating any interest received to the Treasury. I have two comments on Bernanke’s recommendation. First, Bernanke is correct that if the Fed credits the Treasury with

186

H ow to C ombat R ecession

$100 billion in the Treasury’s checking account at the Fed, then clearly the Treasury would not issue any debt to finance the fiscal stimulus. This really would be stimulus without debt. But I disagree with his phrase “alternatively and equivalently.” Under his alternative, the Treasury would issue $100 billion in debt, and the official federal debt would increase $100 billion. That alternative, therefore, would not be stimulus without debt. I hope that Bernanke will, on reflection, delete that alternative. Second, if the Fed gives a $100 billion transfer billion to the Treasury—whether by writing a $100 billion check to the Treasury, or by crediting the Treasury’s checking account at the Fed with $100 billion, there is no increase in the deficit because the Fed’s $100 billion transfer increases Treasury revenue $100 billion, which matches the fiscal stimulus of $100 billion. Bernanke’s language seems to imply, incorrectly, that the deficit increases $100 billion. Bernanke then turned to practical implementation. In his view, the most difficult practical issues surrounding MFFPs would involve their governance—who decides what? He said an MFFP would require the cooperation of Congress and the Fed, and noted that critics might claim that such cooperation sacrificed Fed independence. Another concern is whether an MFFP would be a slippery slope for members of Congress, who might try to get the Fed to finance spending or tax cuts in a fully employed economy. To implement an MFFP, Bernanke proposed the creation by Congress of a special Treasury account at the Fed. He offered a possible arrangement, set up in advance, that would work this way. Congress would create, by statute, a special Treasury account at the Fed, and give the Federal Reserve’s Open Market Committee the sole authority to fill the account up to some prespecified limit. The account would be empty most of the time. The Fed would use its authority to add funds to the account only when the FOMC assessed that an MFFP of specified size was needed to achieve the

W hat Have Others Written about Stimulus without Debt? 187

Fed’s goals. If the FOMC filled the special account, Congress and the administration—through the usual, but possibly expedited, legislative process—would decide how to spend the funds—for example, how much on public works, how much on tax rebates. Congress and the administration would have the option of leaving the funds unspent. If the funds were not used within a specified time, the Fed would be empowered to withdraw them. Bernanke emphasized the separation of powers in his proposal. It would leave to the Fed the responsibility of conducting the technical analysis of whether an MFFP was needed to achieve the Fed’s mandated goals, and of determining the corresponding amount of money to be created. Those conditions would help preserve the Fed’s policy independence and limit the ability of Congress to use monetary financing when the Fed judged such financing to be inappropriate for the economy. Congress and the administration would retain their constitutional authority to determine whether public funds would be spent, and if so how. Like my stimulus without debt, Bernanke’s MFFP focuses on achieving a division of labor and a separation of powers: The Fed decides the amount of fiscal stimulus that Congress can enact without increasing government debt, and Congress decides the amount and composition of fiscal stimulus. In his April 2016 blog, Bernanke said his MFFP should be seriously considered in the next severe recession. Hopefully, when the next severe recession occurs, Bernanke will forcefully advocate the implementation of his MFFP—stimulus without debt.

Central Bank Transfers to Households in a Recession My stimulus-without-debt proposal assigns important roles to both the central bank and the national legislature: the central bank would give a large transfer of new money to the treasury,

188

H ow to C ombat R ecession

and if the elected national legislature enacts fiscal stimulus, the treasury would use the new money (instead of borrowing by selling new bonds to the public) to pay for the fiscal stimulus. An alternative proposal would omit any role for the elected national legislature. Under this alternative, the central bank itself would give transfers of new money to households. Like stimulus without debt, this proposal would provide stimulus without any increase in government debt. As I explained in c hapter 3, I strongly prefer my proposal because it provides a separation of powers and checks and balances between the central bank and the national legislature. In 1999 Gregory Mankiw, a professor of economics at Harvard who a few years later served as chairman of the Council of Economic Advisers for President Bush, wrote a short article in Fortune magazine giving advice to Japan on how to fully recover from its recession. The title of his article was “Memo to Tokyo: Cut Taxes, Print Money.” He recommended “printing some new 100,000 yen [$1,000] notes and sticking one in the pocket” of every citizen. But in his short article, Mankiw didn’t clarify how he wanted the newly printed money to reach the pockets of citizens. There are three different ways to make this happen. Under the first, the central bank would mail transfer checks to citizens; there would be no increase in debt but no role for the national legislature. Under the second, the central bank would write a large transfer check to the treasury and the treasury, if authorized by the national legislature, would mail transfer (tax rebate) checks to citizens; this is the proposal I set out in detail in my 2013 article. Under the third, as Bernanke recommended in 2000 and 2002, the treasury would borrow by selling bonds to the public, the central bank would buy the bonds from the public with new money, and the treasury would mail transfer (tax rebate) checks to citizens. Under the first and second, there would be no increase in

W hat Have Others Written about Stimulus without Debt? 189

government debt, but under the third, government debt would increase. In his short article, Mankiw did not say which approach he recommended. In his online column entitled “How about Quantitative Easing for the People?” Anatole Kaletsky (2012a) called for transfers from central banks to people. Kaletsky lamented the fact that the huge amounts of new money created by the Federal Reserve and the Bank of England to combat the Great Recession had all gone to buy government bonds. He proposed that instead of giving newly created money “to bond traders,” the central bank should distribute new money directly to the public. He noted there would be no increase in government debt. He said that giving large amounts of money to people might seem “wildly irresponsible,” but he pointed out that during the Great Recession central banks had used large amounts new money to buy bonds and make loans to banks, and that redirecting the new money to the public would be fairer and more effective because people would use a portion of their transfers for consumption. A week later Kaletsky (2012b) followed up with a second column addressing objections he received in response to his first column. He said the most powerful objection was that it seemed too good to be true that new money could give given to people to stimulate spending, production, and employment without increasing taxes, government debt, or inflation. Economists often say that there’s no such thing as a free lunch, and this proposal seemed to violate that assertion. Kaletsky replied that in a recession with idle workers and idle machines, output was below potential, so that a free lunch could indeed be obtained by a policy that employed idle workers and utilized idle machines. In an article entitled “Print Less but Transfer More: Why Central Banks Should Give Money Directly to the People,” Blyth and Lonergan (2014) wrote that central banks should “hand consumers cash directly.” They said the central bank might limit

190

H ow to C ombat R ecession

eligibility to taxpaying households, might distribute cash equally to all households, or might just distribute it to the bottom 80% of households. But raising these options reveals the difficulty of excluding the national legislature from the process. Which is the more appropriate institution to make these decisions: an unelected central bank or an elected national legislature? Elected national legislatures make these decisions for many tax and spending programs. Legislatures decide how much a household’s tax will vary with a household’s income, or whether benefits paid to households should be the same for all households or vary with household income. Under my stimulus-without-debt proposal, the national legislature—in the United States, Congress—would make these decisions for the components of the fiscal stimulus package.

Other Writings In an online article, Simon Wren-Lewis, professor of economic policy at Oxford, gave a primer on helicopter money (Wren-Lewis 2014) that also applies to stimulus without debt. He agreed with my point in c hapter 3 that financing fiscal stimulus with new money from the central bank (helicopter money) would not increase government debt, in contrast to a standard combined fiscal-monetary stimulus. Recall that under the standard combined stimulus, the treasury borrows by selling new government bonds to the public, and the central bank buys an equal amount of government bonds from the public. Wren-Lewis contended, as I did in chapter 3, that government bonds held by the central bank should be included in government debt because at any moment the central bank might sell bonds to the public, which would expect to be paid principal and interest on schedule. He also agreed that an important share of tax rebates would be spent, increasing aggregate demand for goods and services.

W hat Have Others Written about Stimulus without Debt? 191

Wren-Lewis revisited helicopter money in another online column a few months later (Wren-Lewis 2015) to focus on whether helicopter money can be reversed in time to prevent inflation in a strong recovery. When a central bank buys government bonds in the open market, it can later sell bonds to withdraw money from the economy. But if a central bank gives a transfer to the treasury, it doesn’t obtain an asset it can later sell to withdraw money. I gave my answer to this question in chapter 6. I showed that in an economy where output, income, and money are all growing, the central bank can prevent inflation by temporarily reducing the amount of government bonds it buys and therefore temporarily reducing the growth rate of money; the central bank doesn’t need to sell bonds; it just needs to buy fewer bonds, thereby slowing the growth of money, to avoid inflation. An advocate of monetizing the debt in a severe recession is Michael Woodford (2013), a professor of economics at Columbia, who stated in an interview that the fiscal authority would make the transfers to households, sell bonds to pay for them, and later tax people to service the debt; the monetary authority would conduct open-market operations in the amounts needed to keep nominal GDP on the target path. In his interview Woodford did not call for the central bank to give a transfer to the Treasury so that the Treasury could finance a fiscal stimulus without borrowing. Australian economist Richard Wood (2012), in an article entitled “The Economic Crisis: How to Stimulate Economies without Increasing Public Debt,” proposed that the treasury or ministry of finance of a nation, not its central bank, create the money to pay for the fiscal stimulus enacted by its legislature. The problem in the United States is that if Congress authorizes money creation by the Treasury, there would be a breakdown of current checks and balances because Congress would directly

192

HOW TO COMBAT RECESSION

control money creation as well as spending and taxation. This raises the possibility that Congress might raise government spending well above taxes in a normal economy when no stimulus is warranted, and create money to cover the difference, thereby unilaterally injecting a combined fiscal-monetary stimulus that overheats the economy and generates inflation. By contrast, under the stimulus-without-debt plan, Congress, not the Federal Reserve, would authorize the tax rebates to households, and the Federal Reserve, not Congress, would decide the specific amount of money to transfer to the Treasury. In an article entitled “The Simple Analytics of Helicopter Money: Why It Works—Always,” Willem Buiter (2014) gave a mathematical analysis of the impact of helicopter money. He emphasized that “fiat base money” is “irredeemable”— viewed as an asset by the holder but not as a liability by the issuer. I agree: government paper money is not government debt; only government bonds are government debt, so that my policy, under which fiscal stimulus is financed by new government money rather than government bonds, really is stimulus without debt. Buiter also said that helicopter money raises aggregate demand because it is net wealth to the private sector. Once again I agree: tax rebates to households that are financed by new money raise the wealth of households, which therefore spend some of it.

Articles in The Economist The Economist weekly newspaper (magazine) is a good place to see how seriously a new economic policy proposal is being taken. In 2016 several editorials (“leaders”), articles, and blogs in The Economist called for serious consideration of helicopter money and/or stimulus without debt. An editorial (Economist 2016a)

W hat Have Others Written about Stimulus without Debt? 193

entitled “Out of Ammo?” called for legislators to work together with their central banker to stimulate their economy. It called on legislators to enact transfers or tax cuts for households and the central banker to finance them with newly printed money, so new money would go from the central bank to the legislature to households, thereby avoiding banks and financial markets. A “Free Exchange” one-page article (Economist 2016b) entitled “Money from Heaven: To Get out of a Slump, the World’s Central Bankers Consider Handing out Cash” begins by saying that helicopter money is a daring approach to monetary policy: printing money to fund government spending or to give people cash. This daring approach is now being taken seriously. Mario Draghi, the president of the European Central Bank, called helicopter money a very interesting concept. Though daring, helicopter money is a less radical departure than it sounds. Milton Friedman’s helicopter money parable first got serious attention in the early 2000s when economists tried to figure out how Japan might escape from its decade-long stagnation and mild deflation. Ben Bernanke, as I have mentioned, made reference in an article and a speech to Friedman’s parable, and soon was being called “Helicopter Ben.” Today, many of the world’s advanced economies are plagued with stagnation in the aftermath of the financial crisis. The stagnation has not yielded to either conventional or even unconventional monetary stimulus. A new strategy is surely needed. The “Money from Heaven” article says that most advocates of helicopter money argue for fiscal stimulus (government spending, tax cuts, or direct payments to citizens) enacted by the national legislature, financed with newly printed money provided by the central bank to the treasury rather than through borrowing by the treasury. Some advocates even call for the central bank to act without the national legislature, directly distributing new money to households; Jeremy Corbyn, leader

194

H ow to C ombat R ecession

of Britain’s Labour Party, has proposed this, calling it a “people’s QE.” In an economy with idle workers and machines, helicopter money would mainly raise output and employment, not prices. Modern central banks in advanced economies are wary of inflation and therefore likely to stop printing money once the economy is nearing full employment. The real problem with helicopter money is that there is likely to be resistance to using it. European law generally forbids the central bank from financing fiscal stimulus enacted by national legislatures, and many central bankers appear to be support this prohibition. Draghi barely managed to get support for the European Central Bank to buy national government bonds. Many national legislatures seem more concerned about cutting deficits than about stimulating their economies. The article ends by saying that what’s missing is the political will to use helicopter money.

Conclusion These articles and blogs about stimulus without debt and helicopter money indicate that the proposal has made headway, receiving attention and some support in a period when most advanced economies have been in recovery, not recession. The next time dark recession clouds loom on the horizon, it seems likely this proposal will receive the attention it warrants.

Chapter 12

Can Stimulus without Debt Combat Secular Stagnation?

Two possible sources of “secular stagnation” must be distinguished. The first source is chronically insufficient aggregate demand for goods and services; insufficient demand is demand that is less than the potential output of the economy. The second source is slow growth in the potential output of the economy. The US economy has only recently recovered from the Great Recession, so it is too soon to know whether it will suffer from secular stagnation. If secular stagnation occurs due to chronically insufficient demand, a stimulus-without-debt policy that is applied as long as demand would otherwise be insufficient can keep actual output equal to potential output and therefore can achieve and maintain full employment (because potential output is defined as the output that would be produced in a given year if labor were fully employed in that year); every year, the level of output and employment would be equal to potential output instead of being below potential output. Also, as I will explain later in this chapter, keeping actual output equal to potential output each year can slightly increase the growth rate of potential output; 195

196

H ow to C ombat R ecession

a stimulus-without-debt policy is unlikely, however, to significantly raise the growth rate of potential output.

Chronically Insufficient Aggregate Demand For several years Lawrence Summers has contended that the United States and other economically advanced countries are likely experiencing chronically insufficient aggregate demand. In an article Summers (2016) provided an exposition of his case. He began by documenting the weak recovery from the Great Recession. The recession was not as deep as the Great Depression, and recovery began sooner because aggressive fiscal stimulus was applied sooner (Summers helped shape and enact the fiscal stimulus as a top economic adviser to President Obama in 2009); but the recovery has been slower than usual in a recession, so that, like the Great Depression, the Great Recession has generated a “lost decade.” Summers noted that the term “secular stagnation” was introduced in 1938 by Harvard’s Alvin Hansen, a recent convert to Keynes’s perspective, who contended that changing demography and an exhaustion of investment opportunities would generate a chronic insufficiency of aggregate demand. In Summers’s view, Hansen was wrong about the next few decades after 1938 due to military spending during World War II, the release of consumer demand after the war that had been pent up during the war, and demand due to massive suburbanization (for example, for highways, automobiles, and housing). Summers said, however, that Hansen’s thesis appears to be correct for the first two decades of the 21st century, because despite very low interest rates that should stimulate borrowing and spending, actual GDP has continued to remain below potential GDP and actual inflation has remained below the target of 2%. Consumer demand has been too low because increasing inequality

Can Stimulus without Debt Combat Secular Stagnation?

197

has raised the share of income of the rich, who have a lower propensity to consume than the nonrich. Investment demand has been too low because slow growth of population and labor force has reduced demand for housing and plant and equipment. The economy temporarily had enough demand from 2003 to 2007 only because of unsustainable borrowing by consumers. If slow growth were due to supply, not demand, interest rates and inflation would be high; instead, both are low, supporting the hypothesis that it is demand, not supply, that has caused the slow growth in output. Summers warned that when the next recession occurs, monetary policy will be unable to cut interest rates because interest rates are already near zero. Hence, there should be a radical rethinking of the role of fiscal policy in economic stabilization. With monetary policy able to do much less, fiscal policy will need to do much more. Summers assumed that fiscal stimulus must involve deficits and debt, but argued that it is worth incurring these deficits and debt. With interest rates chronically low, even large debt burdens can be safely incurred, and should be incurred in order to get the economy to grow. I agree with Summers that even if the Treasury had to borrow to finance fiscal stimulus in a period of chronically insufficient demand, Congress should enact sufficient fiscal stimulus every year and instruct the Treasury to borrow to finance it. I also agree with Summers’s point that with interest rates very low and construction workers still suffering high unemployment, it makes sense to borrow to fund repairs to school buildings and airports. The problem with Summers’s proposal to borrow to pay for fiscal stimulus is that many will continue to oppose fiscal stimulus if it involves deficits and debt. One purpose of this book is to show that fiscal stimulus does not have to involve deficits and debt. I applaud Summers for making the case for ongoing fiscal stimulus if the US economy ever suffers from chronically insufficient aggregate demand. But he should recognize that ongoing fiscal stimulus can be provided without deficits and debt.

198

H ow to C ombat R ecession

Slow Growth in Potential Output Indirectly, a stimulus-without-debt policy, by keeping actual output equal to potential output, should generate a slightly higher growth rate of potential output. With actual output kept equal to potential output, each year the level of investment, including research and development, would be slightly higher than it otherwise would have been. A slightly higher level of investment and R & D each year should result in a slightly higher rate of technological progress—a slightly higher rate of introducing new production processes and new goods and services—and thus, a slightly higher growth rate of potential output. A stimulus-without-debt policy, however, is unlikely to generate a large increase the rate of technological progress and therefore is unlikely to significantly increase the growth rate of potential output. There is a vast literature on the causes of the rate of technological progress and policies that may be able to significantly raise that rate. These policies are supply-side, not demand-side, policies. For example, incentives that result from public policy for patents and property rights in new inventions, or from the tax treatment of R & D expenditures, may have a significant impact. Some periods of history, however, may simply be more conducive to generating increases in the rate of technological progress, and other periods less conducive.

Conclusion Stimulus without debt can combat secular stagnation if the cause of the stagnation is insufficient demand. Stimulus without debt—tax rebates to households and transfers from the Federal Reserve to the Treasury—can and should be implemented for as long as the demand-induced stagnation lasts. Without the

Can Stimulus without Debt Combat Secular Stagnation?

199

transfers from the Fed to the Treasury, the necessary fiscal stimulus would cause debt to rise to a very high percentage of GDP, rightly causing concern and resistance to the fiscal stimulus. Stimulus without debt would solve the insufficient demand problem without causing a serious government debt problem. If the stagnation is caused by slow growth of potential output, stimulus without debt can make only a modest improvement. By raising aggregate demand, stimulus without debt can keep the level of output and employment each year as high as possible without generating a rise in inflation; with investment slightly higher than it otherwise would have been each year, the level of the capital stock and technology should be slightly higher and the growth rate of potential output is likely to be slightly higher than it otherwise would have been. Thus, if secular stagnation is due to a slow rate of technological progress, a stimulus-without-debt policy can provide a little help. If secular stagnation is due to chronically insufficient demand, however, stimulus without debt can provide a lot of help.

Chapter 13

Would Stimulus without Debt Work in a Plausible Model?

I and my colleague Kenneth Lewis studied the impact of stimulus without debt in a plausible macroeconomic model (Seidman and Lewis 2015) developed by Laurence Ball, a professor of economics at Johns Hopkins University and an associate of the National Bureau of Economic Research (Lewis and I made several modifications to Ball’s model). Our paper, “Stimulus without Debt in a Severe Recession,” was published in the Journal of Policy Modeling (2015). In this chapter I will explain the model so that readers can decide whether they agree with Ball, Lewis, and me that the model is plausible—that it is likely to accurately describe what would actually happen in the economy in response to stimulus without debt in a severe recession. I will then answer the question: Would stimulus without debt work in this model?

A Description of the Equations of the Model

200

− Y* t ) /Y* t = λ ( Yt − 1 − Y* t − 1 ) /Y* t − 1  − β ( rt − 1 − r* ) + δ (Gt − 1 /Y* t − 1 ) + st .

(1)

Would Stimulus without Debt Work in a Plausible Model?

201

In equation (1) Yt is real output this year (year t) and Y*t is potential real output this year; Y* t is the largest output that can be produced this year without causing the inflation rate to increase. Suppose potential output ( Y* t ) is $1,000 and actual output ( Yt ) is $925; then actual output is $75 below potential output and the “output gap,” a percentage, is −7.5%. The left side of equation (1), the output gap (Yt − Y* t )/Y* t, would equal ($925 − $1,000) / $1,000 = −7.5%. In the United States in 2009, the output gap was roughly −7.5%. Suppose the last three terms on the right side of equation (1) are zero and λ = 0.8; if the output gap last year (t − 1) was −7.5%, then consumer and business confidence and demand will be low this year, so this year’s output gap would be −6.0%; but if last year’s output gap was 0%, then consumer and business confidence and demand would be normal this year, so this year’s output gap would be 0%. A recession last year causes low confidence, demand, and output this year, whereas a normal economy last year causes normal confidence, demand, and output this year. The last three terms on the right side of equation (1) that affect this year’s output gap are last year’s interest rate ( rt-1 ) , last year’s tax rebates to households (Gt-1 ) , and this year’s shock to aggregate demand ( st ) . π t = π e t + α ( Yt − 1 − Y* t − 1 ) /Y* t − 1 

where π e t = π t − 1 , and α > 0.

Pt = (1 + π t )Pt − 1 .

(2) (3)

Equation (2) says that the inflation rate this year (πt) varies directly with the expected inflation rate this year, which is assumed to equal the actual inflation rate last year (or 0% if last year’s inflation rate was negative). But πt also varies with last year’s output gap: if the output gap is negative (the economy is

202

H ow to C ombat R ecession

in recession), then this year’s inflation rate will be less than last year’s inflation rate, but if the output gap is positive, then this year’s inflation rate will be greater than last year’s inflation rate. Equation (3) simply says that this year’s price level exceeds last year’s price level due to this year’s the inflation rate.

iT t = r* + π t + a ( π t − π * ) + b ( Yt − Y* t ) /Y* t ), a > 0, b > 0, (4)

ln (MD t /Pt Yt ) = k − γit , so MD t /Pt Yt = exp[ k − γ ( it )], where γ > 0 and it ≥ 0.

(5)

it = [k − ln(Mt /Pt Yt )]/γ , if it ≥ 0,

(6)

Mt /Pt Yt = exp[k − γ (it )], if it ≥ 0.

(7)

Mt = Mt − 1 + Zt + R t

(8)

Equation (4) says that the Federal Reserve adjusts its target (T) for this year’s interest rate (iTt) according to how much this year’s inflation rate differs from the Fed’s target inflation rate (2%) and how much this year’s output gap differs from the Fed’s target output gap (0%). We assume that magnitude of the Fed’s adjustment is just what John Taylor prescribes; hence, we assume the Fed follows the Taylor rule with Taylor’s numerical coefficients (a= 1, b= 1). In equations (5)–(7), we assume that, given its target interest rate for this year, the Fed adjusts the supply of high-powered money so that the interaction of the supply and demand, shown in Figure 13.1, makes the actual interest rate equal the Fed’s target interest rate. Equation (5) shows how money demand varies with income and the interest rate, and equation (6) gives the actual interest rate in the economy resulting from a particular money supply. Equation (7) rearranges equation (6) to give

Would Stimulus without Debt Work in a Plausible Model? i

MD/(PY)

203

M/(PY)

8% 4% 0%

10.1%

15%

22.4%

M/(PY)

FIGUR E 13.1 Money Demand and Money Supply

the money supply that makes the actual interest rate equal the Fed’s target interest rate. Equation (8) says begin with money in the economy at the end of last year: if the Fed buys bonds this year, money in the economy increases; if the Fed sells bonds, money decreases; and if the Fed gives a transfer to the Treasury, money increases.

Bt = Bt − 1 + Ft .

Ft = fsPt Yt + θ(Pt Yt * − Pt Yt ) + Pt Gt − R t .

(10)

Ft /Pt Yt = fs + θ[(Yt */Yt ) − 1] + (Pt Gt − R t )/Pt Yt .

(11)

B t = Bc t + B p t .

(12)

(9)

Equation (9) says begin with Treasury debt at the end of last year: if there is a budget deficit (Ft > 0) this year, Treasury debt increases; if there is a budget surplus, Treasury debt decreases. Equation (10) says the budget deficit equals the structural deficit (the first term) plus the cyclical deficit (the second term) plus fiscal stimulus spending (the third term) minus the transfer received from the Fed (the fourth term). The structural deficit is assumed to be a fixed percentage (fs) of nominal income in the economy. The cyclical deficit is zero if nominal income equals

204

H ow to C ombat R ecession

potential nominal potential, and rises as nominal income falls below potential nominal income. Dividing equation (10) by nominal income (PtYt) yields equation (11). Equation (12) says that Treasury debt is held either by the central bank (the Fed) or the public. We used these parameter values: λ = 0.8, β = 0.33, δ = 1.20, r* = 2%, Y* grows at rate g* = 2%, α = 0.25, a = 0.5, b = 0.5, π* = 2%, γ = 10, k = −1.49712, θ = 0.33, fs = 3%.

The Impact of Alternative Policies to Combat Recession Initially real output equals potential real output, government debt is 75% of nominal income, and the government deficit is 3% of nominal income, so the deficit is 4% of the debt, which implies that the debt grows 4% per year. The nominal interest rate is 4%, and the inflation rate is 2%, so the real interest rate is 2%. Money is 15% of nominal income, money injections through Fed bond purchases are 0.6% of nominal income, and the Fed transfer to the Treasury is zero, so money grows 4% per year. Initially the economy is in a positive-growth steady state. Potential real output grows 2% per year, real output grows 2% per year, and inflation is 2% per year, so potential nominal income grows 4% per year. Each of the following grows 4% per year: Treasury debt, the deficit, money, and money injections through Fed bond purchases. The economy is in its initial steady-state in year 1. We compare four paths. Under the first (SS), the economy stays in its positive-growth steady state. Under the three other paths, a recession shock (st in equation (1)) hits in year 2, and only in year 2, that reduces aggregate demand and output 7.5% below potential in equation (1) in year 2 so the output gap in is −7.5% in year 2; monetary policy responds by following the

Would Stimulus without Debt Work in a Plausible Model?

205

Taylor rule. Under the second path (“shock”) there is no fiscal stimulus. Under the third path (G), there is fiscal stimulus (G) in year 2 only; the Treasury borrows to finance the fiscal stimulus. Under the fourth path (GR), there is the same fiscal stimulus (G) in year 2 only; the Treasury does not have to borrow to finance the fiscal stimulus because it receives a transfer (R) from the Federal Reserve equal to the fiscal stimulus. The eight charts in the Figure 13.2 show what happens to eight key variables for 16 years under each of the four paths, but we ran our simulations for many years beyond 16. Let me state a crucial result of our simulations: Along all three paths (shock, G, and GR), all variables in the model converged to their steady-state values. In particular, although this cannot be seen in the inflation rate (π) chart, which shows only up to year 16, in all three simulations the inflation rate (π) never exceeded 2% along its path and eventually converged to 2%. Thus, stimulus without debt did not raise the inflation rate in the short run or long run. Although stimulus without debt (the GR path) did not raise the deficit as a percentage of nominal income (F/PY) or the debt as a percentage of nominal income (B/ PY) in the long run—on all three paths, F/PY converged to 3% and B/PY converged to 75%—it achieved a lower F/PY and B/ PY in the short run compared to stimulus with debt (the G path) The YGAP (output gap) chart shows that when there is no fiscal stimulus, so that the economy is on the “shock” path, it takes over seven years for the output (Y) gap to return to 0% despite monetary stimulus that follows the Taylor rule. With sufficient fiscal stimulus (4.3% of nominal income PY) applied in year 2, whether financed by Treasury borrowing (so the economy is on the G path) or a Fed transfer (so the economy is on the GR path), the output gap is restored to 0% in year 3. What difference does it make how the fiscal stimulus is financed? The F/PY chart shows the deficit (F) as a percent of

2% 1% 0% –1%

YGAP

12%

G, GR 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

–2% –3%

–6%

–7% –8%

4.0%

Shock

3.5% G

85%

70%

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

4.5% B/PY

90%

75%

G GR Shock SS

–5%

80%

F/PY

10% 8%

Shock

–4%

95%

3.0%

G, GR

2.5% 2.0%

1.5% SS

1.0%

Shock

0.5% 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

0.0%

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

2.5% π

2.0%

1.5%

1.0%

G, GR

Shock

0.5%

–0.5%

0.5% 0.0%

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

2.0% 1.5%

1.0%

0.0%

2.5%

–0.5%

G, GR Shock 1 2 3 4 5 6 7 8 9 10111213141516

–1.0%

–1.5%

–2.0% 7% 6%

24%

M/PY

22% 20%

G, GR Shock

18%

12%

G GR Shock SS

3% 2% 1%

16% 14%

Z/PY

0% –1%

1 2 3 4 5 6 7 8 9 10111213141516

FIGUR E 13.2 Simulation Results

–2%

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

208

HOW TO COMBAT RECESSION

nominal income (PY). The shock itself causes an automatic drop in tax revenue and a rise in F/PY in year 2 along the “shock” path from its steady state value of 3% to 5.7%. If the fiscal stimulus in year 2 is financed by borrowing, F/PY rises to 10.0% along the G path. By contrast, if the fiscal stimulus is financed by a Fed transfer to the Treasury, this revenue to the Treasury matches the fiscal stimulus, so F/PY is the same 5.7% on the GR path as it is on the shock path. Thus, the Fed transfer to the Treasury prevents an additional rise of F/PY of 4.3% (from 5.7% to 10.0%). This difference is also evident in the B/PY chart that shows government debt (B) as a percentage of nominal income (PY). The shock itself raises B/PY from its steady-state value of 75% to 83.4% along the “shock” path. If the fiscal stimulus is financed by borrowing, B/PY rises to 87.5% along the G path. By contrast, if the fiscal stimulus is financed by a Fed transfer, B/PY is the same 83.4% on the GR path as it is on the shock path. Thus, the financing of the fiscal stimulus by borrowing would further increase the deficit and debt (above the increase caused by the recession shock itself), but the financing of the fiscal stimulus by a transfer to the treasury prevents this further increase in the deficit and debt. The Fed transfer to the Treasury injects money into the economy, but in order to follow the Taylor rule, the Fed cuts its purchase of bonds by the same amount, so that money injected into the economy is the same whether fiscal stimulus is financed by borrowing or by a Fed transfer. To see this, look at three charts: i, M/PY, and Z/PY. The i chart shows the nominal interest rate; the steady-state value of i is 4%. When the recession shock hits in year 2, the Taylor rule calls for a sharp cut in i to nearly 0%. To achieve this cut, high-powered money in the economy must be increased. In the M/PY chart, M/PY must be raised in year 2 from its steady-state value of 15% to 21.8%. To raise M in equation (8) to the level required to achieve the i called for by

Would Stimulus without Debt Work in a Plausible Model?

209

the Taylor rule, the required increase in Z (the purchase of bonds by the Fed) depends on whether R (the transfer from the Fed to the Treasury) is 0 or positive; if R is zero, the required Z is larger than if R is positive. In the Z/PY chart, the steady-state Z/PY is 0.6%. When the shock hits in year 2, the Taylor rule calls for Z/ PY to jump to 6.5% if the fiscal stimulus is financed by borrowing (so R is 0), but only to 2.0% if the fiscal stimulus is financed by a Fed transfer to the Treasury (so R is positive); this cut in Z/PY from 6.5% to 2.0% offsets the injection of money due to the Fed transfer to the Treasury, so that the increase in money stays the same—from 15% to 21.8%. Thus, in the M/PY chart, the path for M/PY is the same on the G path and the GR path, so in the i chart, i is the same. Note that i is higher along these two paths than along the shock path because when fiscal stimulus reduces the output gap to 0%, the Taylor rule prescribes less monetary stimulus—a higher i. In the inflation rate (π) chart, the recession shock in year 2 causes an output gap of −7.5% in year 2, which reduces π from 2% to 0% in year 3 and −1% in year 4. Fiscal stimulus enacted in year 2 affects the output gap in year 3 and π in year 4. In year 4, π doesn’t fall as much with fiscal stimulus as without it. As the economy recovers, π rises toward its steady-state value of 2.0%. Thus, the large fiscal stimulus in year 2 that eliminates the output gap in year 3 is not inflationary: it reduces the fall in inflation, but doesn’t raise inflation above 2%. Although the chart only shows π through year 16, I can report the following results from our simulations, which ran many more years than 16. In all three simulations (shock, G, and GR), the inflation rate (π) never exceeded 2% along its path and eventually converged to 2%. Stimulus without debt (the GR path) did not raise the inflation rate in the short run or long run compared

210

H ow to C ombat R ecession

to stimulus with debt (the G path) or the shock path, which has monetary stimulus prescribed by the Taylor rule.

Stimulus without Debt Works in This Plausible Model Thus, in this macroeconomic model, stimulus without debt definitely works in a severe recession. The large fiscal stimulus eliminates a large output gap in one year. Because the large fiscal stimulus is financed by a large transfer (not loan) from the Federal Reserve to the Treasury, so that the Treasury doesn’t have to borrow to finance it, the large fiscal stimulus doesn’t increase the deficit or debt beyond the increase caused by the recession shock itself. By contrast, without the Fed transfer to the Treasury, the same large fiscal stimulus would cause the deficit or debt to rise significantly more than with the recession shock alone. Moreover, stimulus without debt is not inflationary: the inflation rate never rises above its low target value. In the short run, the stimulus-without-debt path is much better than relying solely on monetary stimulus according to the Taylor rule to combat the severe recession, and is also better than the fiscal stimulus path, when the Treasury must borrow instead of a receiving a large transfer from the Fed.

Chapter 14

Are We Ready for the Next Severe Recession?

Prior to the Great Recession, I shared an optimistic assumption made by most economists who have advocated fiscal stimulus in past recessions. Most of us assumed that in the anxious climate generated by a severe recession, a majority in both houses of Congress and the president could be counted on to support fiscal stimulus that would be large enough to overcome the recession, even if it entailed a large temporary increase in federal deficits and debt. Given this assumption, there was no urgent reason for any of us to consider whether fiscal stimulus could be enacted without deficits and debt. But in the Great Recession that began in 2008, we learned that our assumption was naive. In early 2009, a majority in both houses of Congress and the president favored fiscal stimulus, but they were also worried about rising federal deficits and debt caused by the recession itself, and were therefore reluctant to propose a fiscal stimulus large enough to overcome the recession. The minority in Congress that opposed fiscal stimulus warned that it would cause a dangerous increase in federal deficits and debt. As a consequence, the fiscal 211

212

H ow to C ombat R ecession

stimulus that was enacted and implemented was, according to my calculations in early 2009, only a half of what was needed, and according to my later calculations, a third of what was needed. This sobering experience taught me that we are still not ready to combat a severe recession. So for the first time I asked whether it would be possible to implement a large fiscal stimulus without any increase in government deficits and debt. Once I asked the question, it wasn’t long before I saw that there was an astonishingly simple way to do it. I soon discovered that Keynesian economist Abba Lerner had seen clearly how to do it in the 1940s, and more recently, a few others have also. The way to do it is simply have the Federal Reserve use new money to finance the fiscal stimulus. There is no need for the government to borrow to pay for fiscal stimulus. Federal Reserve paper money is not government debt because the government no longer promises to pay gold or anything else to any holder of paper money. This wouldn’t be inflationary as long as the stimulus is implemented only during the recession and is phased out as the economy approaches full employment and consumer and business confidence returns to normal. But how would this be implemented in our actual economy? Simply by having the Federal Reserve give a large transfer (not loan) to the Treasury equal to the fiscal stimulus so the Treasury does not have to borrow to pay for the fiscal stimulus enacted by Congress. But as of today, to my knowledge, there are few economists, or economic policymakers, or members of Congress, or members of the Federal Reserve’s Open Market Committee who have even thought about whether it is possible to implement fiscal stimulus without government debt. Unless this changes, whenever the next severe recession hits, fiscal stimulus with debt will be proposed, triggering public and political opposition because of the large deficits and debt it will generate. As in 2009, the supporters of fiscal stimulus will either have to scale down their

Are We Ready for the Next Severe Recession?

213

fiscal stimulus to make its debt acceptable, thereby making it too small, or if they refuse to scale it down, their stimulus proposal will likely be politically defeated. The stimulus-without-debt proposal, however, is not simply a tactic for getting a large fiscal stimulus enacted in a severe recession. It is certainly better for stimulus to be implemented without a large increase in government debt. Large government debt—that is, government debt that is a large percentage of GDP—may generate negative economic consequences and risks in the future. If interest rates rise in the future, large government debt will require large interest payment by the government to bondholders, which will force Congress to raise taxes or cut spending or borrow more, so in the future citizens will have to pay higher taxes or receive lower government benefits. More government borrowing and still larger government debt may at some point make financial investors anxious and cause them to sell corporate stocks, resulting in a fall in the stock market, which in turn would reduce wealth and confidence and may cause consumers and businesses to cut spending, precipitating a recession. The fall in the stock market and the recession may generate still more anxiety among financial investors and managers, thereby generating a financial crisis, worsening the recession. So it is certainly better to do a large stimulus without debt than with a large debt. My stimulus-without-debt plan has five elements:

1. If the Federal Reserve judges that the GDP of the economy is significantly below the Fed’s estimate of potential GDP, the Fed decides whether to give a transfer (not a loan) to the Treasury, and if so, how much, on the condition that it be used only for fiscal stimulus; authority for the Fed to make this transfer might require an amendment to the Federal Reserve Act; the Fed could implement its transfer

214

H ow to C ombat R ecession

by writing a check to the Treasury or by crediting the Treasury’s checking account at the Fed. 2. Congress decides whether to enact fiscal stimulus, and if so, how much; its main component would be tax rebates to households, though other components should also be included in the fiscal stimulus package; the Treasury would mail tax rebate checks to households in amounts specified by Congress. 3. If the amount Congress enacts for fiscal stimulus is no greater than the Fed’s transfer, then the Treasury does not have to borrow to finance the fiscal stimulus; if the fiscal stimulus is greater than the Fed’s transfer, the Treasury has to borrow the difference. 4. The Fed decides how much to adjust its bond purchases or sales to try to keep employment high and inflation low; it is very likely that the Fed, having injected money into the economy through its transfer to the Treasury, would decide to inject less money into the economy through bond purchases or withdraw money from the economy through bond sales. 5. The Fed orders (from the Treasury’s Bureau of Engraving and Printing) an amount of new Federal Reserve notes equal to its transfer to the Treasury, and stores these notes in the Fed’s vault; this Fed vault cash would be an asset on the Fed’s balance sheet, and as a consequence of this new vault cash the Fed’s transfer to the Treasury would not reduce the Fed’s capital (net worth) on its balance sheet.

My proposal would preserve the separation of powers and checks and balances between Congress and the Federal Reserve. It would be up to Congress to decide whether to enact fiscal stimulus and to set its size and composition. It would be up to the Fed

Are We Ready for the Next Severe Recession?

215

to decide whether to give a transfer (not loan) to the Treasury and to set its size. But why choose tax rebates (cash transfers) to households as the main element of the fiscal stimulus? There are at least eight reasons.

1. Most important, tax rebates work: a significant portion of tax rebates are spent within a year of being received. This should not be surprising. Both surveys of recipients and econometric analysis of spending data following the payment of tax rebates indicate that households spend a significant portion of it: a quarter within three months and a half within a year. 2. Tax rebates clearly increase household spending on all goods and services rather just a subset of goods and services, so all businesses would recognize that rebates boost customer demand for their goods or services. 3. With tax rebates, there would be no shovel-ready problem; there is no limit to how much households can spend promptly on goods and services. 4. With tax rebates there would be no temporary or permanent increase in the size of government: rebates simply give spending power to millions of individual consumers whose spending stimulates the private sector. 5. Tax rebates to combat a recession are clearly temporary and require a new vote by Congress to be continued. 6. Every household would receive a rebate check in the mail from the US Treasury, so every voter would actually see a concrete personal benefit from this kind of fiscal stimulus. 7. The inclusion of every household would cause most voters to regard tax rebates as a fair way to implement fiscal stimulus.

216

H ow to C ombat R ecession

8. Rebates have been enacted with bipartisan support three times—1975, 2001, and 2008. Thus, a tax rebate has many important advantages as an instrument for fiscal stimulus in a recession.

There is a good reason why tax rebates were able to pass with bipartisan support three times. Conservatives and liberals, Republicans and Democrats, have differing long-term agendas for government spending and taxation. For example, conservatives generally want permanent tax cuts, while liberals want permanent increases in social insurance and education programs. Neither side wants an antirecession stimulus that would advance the agenda of the other side if it became permanent. Tax rebates are obviously and inherently temporary. They do not favor the long-term agenda of either side. This is undoubtedly one reason that tax rebates were enacted three times with bipartisan support, whereas most other proposals, such as permanent tax cuts, or permanent increases in social insurance or education programs, have generated partisan support and opposition. The tax rebate does not favor one side’s long-term agenda over the other. It is neutral toward long-term agendas. Should a fiscal stimulus package consist solely of tax rebates? No. I made the case for several fiscal stimulus supplements to tax rebates for inclusion in a fiscal stimulus package: a temporary increase in federal aid to state governments, temporary tax incentives for business investment, and infrastructure repairs and maintenance. But I also made the case against including any of the following in a fiscal stimulus package: a temporary cut in income tax withholding, a temporary cut in payroll taxes, a cut in income taxes, and an increase in spending on long-term programs. I considered the main objections to my stimulus-without- debt plan and gave my responses. Would stimulus without debt

Are We Ready for the Next Severe Recession?

217

be inflationary? No, because stimulus would be phased out as the economy returns to full employment and consumer and business confidence returns to normal. Would stimulus without debt weaken the Fed’s balance sheet? No, because the Fed would print an amount of new paper money equal to the fiscal stimulus and keep the new paper money in its vault as an asset. Would stimulus without debt undermine Federal Reserve independence? No, because the Fed would decide whether to make a transfer to the Treasury, and how much the transfer should be. Can’t monetary stimulus overcome a severe recession? No, because in a severe recession, cutting interest rates to zero doesn’t induce much borrowing and spending because of consumer and business pessimism. It would be straightforward to implement stimulus without debt in any country that, like the United States, has a central bank with the power to print the money it uses. But it would also be straightforward to implement stimulus without debt in the eurozone, where only the European Central Bank (ECB) can print euros. The ECB would transfer euros to each national treasury, using a formula that specifies how many euros should go to each particular treasury, in the same way that the Treasury of the US federal government implements revenue sharing by transferring dollars to the treasury of each state government, using a formula enacted by Congress that specifies how many dollars should go to each treasury. Although stimulus without debt was conceived as a tool to combat recession—especially a severe recession like the Great Recession that began in 2008—it could also combat chronic insufficient aggregate demand. If a nation ever faces secular stagnation due to chronic insufficient aggregate demand for goods and services, applying stimulus without debt each year would keep demand at the level needed to maintain full employment without generating inflation. Each year the Fed would give a

218

H ow to C ombat R ecession

transfer to the Treasury to finance fiscal stimulus (primarily tax rebates to households) enacted by Congress. As long as the magnitude of the stimulus is set so that it increases demand to, but not above, potential output, it won’t generate inflation. Thus, stimulus without debt is a solution to demand-induced secular stagnation. In conclusion, we are not yet ready for the next severe recession (or demand-induced secular stagnation) for two reasons. First, most economists, policymakers, members of the Federal Reserve’s Open Market Committee, and members of Congress assume that a large fiscal stimulus must generate a large increase in government debt, and many would oppose any policy that generates such an increase in debt. A major purpose of this book is to try to communicate to them that it is indeed possible to implement a large fiscal stimulus without any increase in government debt, and to persuade them to seriously consider stimulus without debt in the next severe recession. Second, Congress must amend the Federal Reserve Act to empower the Federal Reserve’s Open Market Committee to authorize, if it chooses to do so, a large transfer (not loan) to the Treasury to finance a large fiscal stimulus during a recession (or demand-induced secular stagnation). Without such an amendment, the FOMC may be unwilling to make the transfer, and even if it does, its action would be challenged in court. Ideally, it would obviously be better for Congress to hold hearings on and debate such an amendment prior to the next severe recession. In practice, it may take the reality of a looming severe recession to get Congress to act. So advocates of stimulus without debt must work hard to communicate that it is possible to implement large fiscal stimulus in a severe recession without any increase in debt and to persuade people—economists, policymakers, and the public— that stimulus without debt should implemented in the next severe recession. Prior to the next recession, advocates should try

Are We Ready for the Next Severe Recession?

219

to get Congress to hold hearings on and enact an amendment to the Federal Reserve Act that empowers the Federal Reserve’s Open Market Committee to authorize a large transfer (not loan) from the Fed to the Treasury to finance fiscal stimulus enacted by Congress in a recession. If this cannot be achieved prior to the next severe recession, advocates should be ready to offer the amendment to Congress as soon as a severe recession hits. Stimulus without debt is not the only action that must be taken the next time a severe recession occurs. As in the Great Depression and the Great Recession, it will be necessary for the Federal Reserve and/or the Treasury to perform financial rescues of key firms and inject funds into financial firms to keep credit from freezing up. These essential interventions are not addressed in this book. There are many analysts, however, who believe that these interventions are all that is necessary—that once key firms are rescued and the flow of credit is restored, then nothing more needs to be done. This book strongly rejects that view. A severe recession always involves a plunge in aggregate demand for goods and services, and once that plunge occurs, it will not be reversed simply by rescuing key firms and restoring the flow of credit. This book explains how a large fiscal stimulus can be implemented without any increase in government debt or inflation. It simply requires a large transfer (not loan) from the Federal Reserve to the Treasury to finance the fiscal stimulus enacted by Congress. If the argument of this book is widely accepted, and if Congress amends the Federal Reserve Act to empower the Fed, if it chooses, to make a large transfer to the Treasury for fiscal stimulus, then we really will be ready for the next severe recession.

REFERENCES

Arbess, Daniel (2013). “Bring on the ‘Helicopter Money.’” Wall Street Journal op-ed, May 14. Archer, David, and Paul Moser-Boehm (2013). “Central Bank Finances.” Bank for International Settlements Papers No. 71, April. Auerbach, Alan, and Yuriy Gorodnichenko (2012). “Measuring the Output Responses to Fiscal Policy.” American Economic Journal: Economic Policy 4(2): 1–27. Auerbach, Alan, and Yuriy Gorodnichenko (2013). “Fiscal Multipliers in Recession and Expansion.” In Fiscal Policy after the Financial Crisis, edited by Alberto Alesina and Francesco Giavazzi, 63–98. Chicago: University of Chicago Press. Barro, Robert (1974). “Are Government Bonds Net Wealth?” Journal of Political Economy 82(6): 1095–1117. Barro, Robert (2009). “Voodoo Multipliers.” Economists’ Voice, February www.bepress.com/ev. Bernanke, Ben (2000). “Japanese Monetary Policy: A Case of Self- Induced Paralysis?” In Japan’s Financial Crisis and Its Parallels to U.S. Experience, edited by Ryoichi Mikitani and Adam Posen, 149– 66. Washington, DC: Institute for International Economics. Bernanke, Ben (2002). “Deflation: Making Sure ‘It’ Doesn’t Happen Here.” Remarks to National Economists’ Club, Washington DC, November 21. Bernanke, Ben (2009). “The Fed’s Exit Strategy.” Wall Street Journal op- ed, July 21. Bernanke, Ben (2012). “Monetary Policy since the Onset of the Crisis.” Speech at Jackson Hole Symposium, Kansas City Federal Reserve 221

222

References

Bank, August 31. www.federalreserve.gov/newsevents/speech/ bernanke20120831a.htm. Bernanke, Ben (2016). “What Tools Does the Fed Have Left? Part 3: Helicopter Money.” Blog post, Brookings Institution, April 11. https://www.brookings.edu/blog/ben-bernanke/2016/04/11/ what-tools-does-the-fed-have-left-part-3-helicopter-money. Blanchard, Olivier (2017). Macroeconomics. 7th ed. Harlow, UK: Pearson. Blinder, Alan (1981). “Temporary Income Taxes and Consumer Spending.” Journal of Political Economy 89(1): 26–53. Blinder, Alan, and Mark Zandi (2015). “The Financial Crisis: Lessons for the Next One.” Center on Budget and Policy Priorities, October 15. http://www.cbpp.org/research/e conomy/the-f inancial-c risislessons-for-the-next-one. Blyth, Mark, and Eric Lonergan (2014). “Print Less but Transfer More: Why Central Banks Should Give Money Directly to the People.” Foreign Affairs, September, 98–109. Broda, Christian, and Jonathan Parker (2014). “The Economic Stimulus Payments of 2008 and the Aggregate Demand for Consumption.” Journal of Monetary Economics 68 (Fall): S20–36. Buiter, Willem (2014). “The Simple Analytics of Helicopter Money: Why It Works—Always.” Economics: The Open-Access, Open-A ssessment E-Journal 8: 2014–28. http://www.economics-ejournal.org/economics/journalarticles/2014-28. Caballero, Ricardo (2010). “A Helicopter Drop for the Treasury.” Vox, August 30. http://www.voxeu.org/article/ helicopter-drop-us-treasury. Carlino, Gerald, and Robert Inman (2016). “Fiscal Stimulus in Economic Unions: What Role for States?” Tax Policy and the Economy 30(1): 1–50. Caruana, Jaime (2013). Foreword to “Central Bank Finances,” by David Archer and Paul Moser-Boehm. Bank for International Settlements Papers No. 71, April. Chodorow-Reich, Gabriel, Laura Feiveson, Zachary Liscow, and William Gui Woolston (2012). “Does State Fiscal Relief during Recessions Increase Employment? Evidence from the American Recovery and Reinvestment Act.” American Economic Journal: Economic Policy 4(3): 118–45. Dyson and Hodgson (2016). “Accounting for Sovereign Money: Why State-Issued Money Is Not Debt.” https://www.google.com/#q=D yson+and+Hodgson+(2016)%2C+%E2%80%9C.

References

223

Economist (2016a). “Out of Ammo?” Editorial, February 20. Economist (2016b). “Money from Heaven: To Get out of a Slump, the World’s Central Bankers Consider Handing out Cash.” Free- Exchange Article, April 23. Federal Reserve Board (2017). “Credit and Liquidity Programs and the Balance Sheet.” https://www.federalreserve.gov/monetarypolicy/ bst_fedsbalancesheet.htm. Feldstein, Martin (2008). “The Tax Rebate Was a Flop. Obama’s Stimulus Plan Won’t Work Either.” Wall Street Journal op-ed, August 6. Feldstein, Martin (2009). “Rethinking the Role of Fiscal Policy.” American Economic Review Papers and Proceedings 99(2): 556–59. Friedman, Milton (1969). “The Optimum Quantity of Money.” In The Optimum Quantity of Money and Other Essays. Chicago: Aldine. Greenspan, Alan (1997). “Remarks by Chairman Alan Greenspan at the Catholic University Leuven, Leuven, Belgium.” January 14. https:// www.federalreserve.gov/boarddocs/speeches/1997/19970114.htm Herndon, Thomas, Michael Ash, and Robert Pollin (2014). “Does High Public Debt Consistently Stifle Economic Growth? A Critique of Reinhart and Rogoff.” Cambridge Journal of Economics 38(2): 257–79. Johnson, David S., Jonathan A. Parker, and Nicholas S. Souleles (2006). “Household Expenditure and the Income Tax Rebates of 2001.” American Economic Review 96(5): 1589–610. Kaletsky, Anatole (2012a). “How about Quantitative Easing for the People?” August 1. http://blogs.reuters.com/anatole-kaletsky/ 2012/08/01/how-about-quantitative-easing-for-the-people. Kaletsky, Anatole (2012b). “Suddenly, Quantitative Easing for the People Seems Possible.” http://blogs.reuters.com/anatole-kaletsky/ 2012/08/09/suddenly-quantitative-easing-for-the-people. Kaplan, Greg, Giovanni Violante, and Justin Weidner (2014). “The Wealthy Hand-to-Mouth.” Brookings Papers on Economic Activity, Spring, 77–138. Keynes, John Maynard (1936). The General Theory of Employment, Interest, and Money. New York: Harcourt, Brace and World. Koo, Richard (2015). The Escape from Balance Sheet Recession and the QE Trap. Singapore: John Wiley. Lerner, Abba (1944). The Economics of Control. New York: Macmillan. Lerner, Abba (1951). Economics of Employment. New York McGraw-Hill. Lewis, Kenneth, and Laurence Seidman (2008). “Overcoming the Zero Interest Rate Bound: A Quantitative Prescription.” Journal of Policy Modeling 30(5): 751–60.

224

References

Lewis, Kenneth, and Laurence Seidman (2012). “Did the 2008 Rebate Fail? A Response to Taylor and Feldstein.” Journal of Post Keynesian Economics 34(2): 183–204. Lonergan, Eric (2016). “Legal Helicopter Drops in the Eurozone.” February 24. http://www.philosophyofmoney.net/legal-helicopter- drops-in-the-eurozone/. Mankiw, N. Gregory (1999). “Memo to Tokyo: Cut Taxes, Print Money.” Fortune, January 11. Meltzer, Allan (2014). “Federal Reserve Independence.” Journal of Economic Dynamics and Control 49: 160–63. Mian, Atif, and Amir Sufi (2014). House of Debt. Chicago: University of Chicago Press. Muellbauer, John (2014). “Combatting Eurozone Deflation: QE for the People.” Vox, December 23. http://voxeu.org/article/ combatting-eurozone-deflation-qe-people. Olney, Martha L. (1999). “Avoiding Default: The Role of Credit in the Consumption Collapse of 1930.” Quarterly Journal of Economics 114(1): 319–35. Parker, Jonathan A. (2014). “The Effectiveness of Tax Rebates as Countercyclical Fiscal Policy.” Vox, June 17. http://www.voxeu. org/article/macro-stabilisation-and-tax-rebates. Parker, Jonathan A., Nicholas S. Souleles, David S. Johnson, and Brendan McClelland (2013). “Consumer Spending and the Economic Stimulus Payments of 2008.” American Economic Review 103(6): 2530–53. Reinhart, Carmen, and Kenneth Rogoff (2009). This Time Is Different: Eight Centuries of Financial Folly. Princeton NJ: Princeton University Press. Reinhart, Carmen, and Kenneth Rogoff (2010). “Growth in a Time of Debt.” American Economic Review Papers and Proceedings 100(2): 573–78. Romer, Christina D. (2011). “What Do We Know about the Effects of Fiscal Policy? Separating Evidence from Ideology.” Lecture, Hamilton College, November 7. Romer, David (2012). “What Have We Learned about Fiscal Policy from the Crisis?” In In the Wake of the Crisis, edited by Olivier Blanchard, David Romer, Michael Spence, and Joseph Stiglitz, 57– 66. Cambridge, MA: MIT Press. Sachs, Jeffrey, and Felipe Larrain (1993). Macroeconomics in the Global Economy. Englewood, NJ: Prentice Hall.

References

225

Sahm, Claudia R., Mathew D. Shapiro, and Joel Slemrod (2010). “Household Response to the 2008 Tax Rebate: Survey Evidence and Aggregate Implications.” In Tax Policy and the Economy, vol. 24, edited by Jeffrey R. Brown, 83–109. Chicago: University of Chicago Press. Samuelson, Paul (1948). Economics. New York: McGraw-Hill. Seidman, Laurence (2001). “Reviving Fiscal Policy.” Challenge 44(3): 17–42. Seidman, Laurence (2003). Automatic Fiscal Policies to Combat Recessions. Armonk, NY: M.E. Sharpe. Seidman, Laurence (2006). “Learning about Bernanke.” Challenge 49(5): 19–32. Seidman, Laurence (2009). “A Temporary Cut in Income Tax Withholding to Deliver Stimulus.” Tax Notes, September 28. Seidman, Laurence (2011). “Great Depression II: Why It Could Happen and How to Prevent It.” Challenge 54(1): 32–53. Seidman, Laurence (2012a). “Keynesian Fiscal Stimulus: What Have We Learned from the Great Recession?” Business Economics 47(4): 273–84. Seidman, Laurence (2012b). “Keynesian Stimulus versus Classical Austerity.” Review of Keynesian Economics, Autumn, 77–92. Seidman, Laurence (2013). “Stimulus without Debt.” Challenge 56(6): 38–59. Seidman, Laurence, and Kenneth A. Lewis (2002). “A New Design for Automatic Fiscal Policy.” International Finance 5(2): 251–84. Seidman, Laurence, and Kenneth A. Lewis (2015). “Stimulus without Debt in a Severe Recession.” Journal of Policy Modeling 37: 945–60 Shapiro, Matthew D., and Joel Slemrod (2003a). “Consumer Response to Tax Rebates.” American Economic Review 93(1): 381–96. Shapiro, Mathew D., and Joel Slemrod (2003b). “Did the 2001 Tax Rebate Stimulate Spending? Evidence from Taxpayer Surveys.” In Tax Policy and the Economy, vol. 17, edited by James M. Poterba, 83–109. Chicago: University of Chicago Press. Shapiro, Mathew D., and Joel Slemrod (2009). “Did the 2008 Tax Rebates Stimulate Spending?” American Economic Review Papers and Proceedings 99(2): 374–79. Siegel, Jeremy (2013). “Fed Must Revert to Tested Reserves Policy.” Financial Times op-ed, March 27. Summers, Lawrence (2016). “Secular Stagnation and Monetary Policy.” Federal Reserve Bank of St. Louis Review 98(2): 93–110. http:// dx.doi.org/10.20955/r.2016.93-110.

226

References

Taylor, John (2008). “Why Permanent Tax Cuts Are the Best Stimulus.” Wall Street Journal op-ed, November 25. Taylor, John (2009). “The Lack of an Empirical Rationale for a Revival of Discretionary Fiscal Policy.” American Economic Review Papers and Proceedings 99(2): 550–55. Taylor, John (2011). “An Empirical Analysis of the Revival of Fiscal Activism in the 2000s.” Journal of Economic Literature 49(3): 686–702. Turner, Adair (2013). “Debt, Money, and Mephistopheles: How Do We Get out of This Mess?” Group of 30, Washington, DC, May. Turner, Adair (2015). “The Case for Monetary Finance: An Essentially a Political Issue.” IMF Jacques Polak Research Conference, November 5, International Monetary Fund. Whelan, Karl (2015). “On Central Bank Balance Sheets.” Economic Policy 29(77): 79–137. Wilson, Daniel (2012). “Fiscal Spending Jobs Multipliers: Evidence from the 2009 American Recovery and Reinvestment Act.” American Economic Journal: Economic Policy 4(3): 251–82. Wood, Richard (2012). “The Economic Crisis: How to Stimulate Economies without Increasing Public Debt.” Center for Economic Policy Research Policy Insight No. 62, August. https://www. google.com/#q=Wood%2C+Richard+(2012).+%E2%80%9CThe+E conomic+Crisis. Woodford, Michael (2013). “Helicopter Money as a Policy Option.” Interview, May 20. Vox. http://www.voxeu.org/article/helicopter- money-policy-option. Wren-Lewis, Simon (2014). https://mainlymacro.blogspot.com/2014/ 10/helicopter-money.html Oct 22. Wren-Lewis, Simon (2015). https://mainlymacro.blogspot.com/2015/ 02/helicopter-money-and-government-of-html Feb 22.

INDEX

“Accounting for Sovereign Money” (Dyson and Hodgson), 143 Akerlof, George, 48 American Jobs Act proposal (2011), 110–11 American Recovery and Reinvestment Act (ARRA, 2009) accelerated depreciation of business investments and, 110 Council of Economic Advisers’ analysis of, 108–12 Democratic Party support for, 89, 103, 107, 109, 111–12 emergency unemployment compensation and, 109–10 empirical studies of, 111, 117–22 estimated impact of, 107–8, 111–12 federal deficit and, 153 government spending provisions in, 53, 89, 109 inflation and, 130–31 infrastructure provisions and, 109 Making Work Pay credit and, 62, 84, 89–92, 103–5, 110 Medicaid funding and, 99, 109–10, 119–21

need for additional stimulus following, 53, 55, 108, 212 Republican Party opposition to, 53, 84, 89, 103, 107, 112 size of stimulus in, 52–53, 108 state balanced budget requirements and, 119 statutory funding formulas in, 117–18 tax cuts and, 53, 109 teaching jobs prioritized in, 109–10 transfers to state and local governments under, 53, 99–100, 117–21 unemployment and job creation addressed in, 52–53, 108, 111, 118–21 Arbess, Daniel, 182–83 Archer, David, 145 Ash, Michael, 32 Auerbach, Alan, 115–16 “Avoiding Default” (Olney), 164 balance-sheet recessions, 166–69 Ball, Laurence, 200 Bank of England, 189

227

228

Index

Bank of Japan (BOJ), 184–85 Barro, Robert, 88–89, 114–15 Bear Stearns, 77, 80 Bernanke, Ben on Federal Reserve interest rates for bank deposits, 135 Great Recession recovery efforts and, 153, 155, 185 helicopter money and, 184, 193 on Japan, 183–85 on the portfolio rebalancing effect, 161–63 stimulus without debt proposal and, 185–88 Blanchard, Olivier, 31 Blinder, Alan, 48, 81, 169–71 Blyth, Mark, 189 Broda, Christian, 58–59, 62, 86–87, 95 Buiter, Willem, 192 Bush, George H. W., 154 Bush, George W., 51, 63–64, 72, 107 Caballero, Ricardo, 182 Carlino, Gerald, 121–22 Caruana, Jaime, 145 central banks. See also European Central Bank; Federal Reserve accounting equity and, 145 balance-sheet problems and, 177 commercial banks compared to, 142, 145–46 government bond purchases by, 167–68, 191 in Japan, 184–85 monetary stimulus and, 88 stimulus without debt proposal and, 173–74, 177–78, 191 transfers to households during a recession and, 187–90, 193–94 transfers to Treasuries during recessions and, 177, 181–83, 191

Chodorow-Reich, Gabriel, 119–21 “Combating Eurozone Deflation” (Muellbauer), 175 Congress elections (2010) and, 110 Federal Reserve and, 5, 42–46, 54, 138–39, 150–53, 156, 186–87, 218–19 Great Recession recovery and, 52, 54–55 proposal to allow unilateral money creation by, 37–38, 191–92 standard fiscal stimulus and, 2–4, 20, 185, 197, 211 stimulus without debt proposal and, 5–6, 15, 18, 21–23, 27, 29, 32–34, 37–38, 40, 44–45, 147, 150, 186–87, 190, 192, 214–15, 218–19 tax rebates and, 4, 18, 22, 25, 29, 37–38, 46, 63–64, 90, 96, 182–83 temporary tax incentives and, 38, 101 consumer demand consumer confidence and, 23–24 Great Recession and, 17–18, 53, 78, 80, 82, 167 helicopter money and, 192, 194 household debt and, 164–65 housing bubbles and, 48 income inequality and, 196–97 inflation and, 10, 127–30, 135–36 interest rates and, 28, 36, 47–48 Keynesianism and, 13–14, 48 marginal propensity to consume (MPC) and, 67–69, 81–83, 94 portfolio rebalancing effect and, 157–58 production output linked to, 10–11, 18 recessions and, 1, 5, 10–11, 13–14, 25, 38, 159, 164, 219

Index

secular stagnation and, 195–99, 217 slow population growth and, 197 small business survey data regarding, 50 standard fiscal stimulus and, 5, 18 stimulus without debt proposal and, 17, 23–24 stock market declines and, 10, 21, 48 tax rebates and, 11–12, 21, 24–25, 30, 50, 58–62, 65–72, 75–91, 95, 100, 171, 192, 215 Corbyn, Jeremy, 193–94

229

need for additional stimulus following, 52–53, 107–8, 212 size of, 51, 53, 72, 107–8 tax rebates and (see tax rebate of 2008) European Central Bank (ECB) balance sheet of, 177 inflation targets and, 152 Longeran’s proposal regarding helicopter money and, 176–77 Muellbauer’s proposal regarding quantitative easing and, 175–77 prohibitions against financing fiscal stimulus enacted by national legislatures via, 194 stimulus without debt proposal and, 174–77, 217 Eurozone countries (EZ), 17, 174–78, 217

demand-induced secular stagnation. See secular stagnation Democratic Party American Recovery and Reinvestment Act supported by, Federal Deposit Insurance 89, 103, 107, 109, 111–12 Corporation (FDIC), 5 economic stimulus debates in federal grants to state governments Congress (2009–10) and, 46 American Recovery and long-term tax cuts opposed by, 106 Reinvestment Act and, 53, payroll tax cut (2011–13) and, 92 99–100, 117–21 spending on long-term programs conservative criticisms of, 99 supported by, 106 Medicaid matching rates and, 99 Dillon, Douglas, 142 objective numerical formulas Douglas, Paul, 142 for, 98–99 Draghi, Mario, 193–94 phasing out of, 100 Dyson, Ben, 143 political pressure to avoid severe spending cuts and, 123, 125 Economics of Control (Lerner), 14 recession of 2001 and, 100 Economics of Employment (Lerner), 14 state balanced budget requirements Economic Stimulus Act of 2008 and, 57, 119, 122–23 bipartisan support for, 45–46, state government borrowing 72, 107 contrasted with, 123–25 bonus business depreciation states’ accounting rules provision in, 46, 51, 53, regarding, 56–57 101, 106–7 stimulative effect of, 4–5, 98–99, economic impact of, 52, 107–8 122–26, 216

230

Index

Federal Reserve. See also central banks accounting rules regarding government bonds held by, 19–20, 33, 38–39, 41, 143, 190 accounting rules regarding paper money held by, 3–4, 137–43, 146–49, 212, 217 asset creation by, 143–44 balance sheet of, 23, 41–42, 45, 137–49, 153, 217 bank deposit rates at, 135 Board of Governors of, 151–52 “bond-burning plan” and, 39–40 Congress and, 5, 42–46, 54, 138–39, 150–53, 156, 186–87, 218–19 credit liquidity and, 5, 27–28, 47, 157–58, 168, 219 Federal Open Market Committee (FOMC) and, 5, 27, 33, 44–46, 152, 186, 212, 218–19 Federal Reserve Act amendment of 1977 and, 151–52 Federal Reserve Act of 1913 and, 5, 22, 43–46, 54, 151, 153–54, 213, 218–19 fiat money system and, 140, 144 funding sources for, 151 government bond purchases by, 15, 19, 23, 28, 35–36, 133–35, 137–38, 146, 153, 156–57, 159, 161–63, 169–70, 183, 214 Great Recession recovery policies and, 153, 155, 158, 168–69, 185 housing bubble of 2007 and, 158 independence of, 42–43, 150–56, 186–87, 217 inflation targets and, 23, 29, 33, 42, 151–55 interest rates set by, 27–28, 42, 47–48, 135, 151–52 lender of last resort role of, 27–28, 47, 157

mortgage-backed securities purchased by, 134, 162 paper money held by, 15, 23, 29, 41–43, 45, 137–40, 143–44, 146–49, 214, 217 president of United States and, 151 rescue of financial firms and, 28, 158, 168, 219 stimulus without debt proposal and, 27–28 transfers to Treasury under stimulus without debt plan and, 3, 5–6, 11–12, 15, 17–18, 21–22, 26–27, 29, 32–35, 37, 40–41, 43–46, 55–56, 96, 133, 136–38, 146–47, 150, 152, 156, 171, 180–82, 185, 192, 198–99, 203–5, 208–10, 212–15, 217–19 transfer to households proposal and, 36–37 unemployment targets and, 23, 29, 33, 42, 151–52 Feiveson, Laura, 119–21 Feldstein, Martin American Economic Association presentation (2009) by, 75 standard regression analysis by, 78–79, 81–82 tax rebate of 2008 analyzed by, 61, 74–84, 90 Wall Street Journal op-ed (2008) by, 74–77, 82 fiat money. See under paper money fiscal stimulus. See standard fiscal stimulus; stimulus without debt proposal Ford, Gerald, 63 Friedman, Milton helicopter money parable of, 34–35, 179–81, 184, 193 portfolio rebalancing effect and, 162 on raises versus tax rebates, 88 functional finance, 14

Index

231

Gorodnichenko, Yuriy, 115–16 consumer demand and consumer government bonds. See also confidence during, 17–18, 53, government debt 78, 80, 82, 167 accounting rules regarding Federal credit liquidity and, 28, 168 Reserve holding of, 19–20, 33, federal deficit and, 153, 155–56 38–39, 41, 143, 190 Federal Reserve’s role in “bond-burning plan” and, 39–40 recovery from, 153, 155, 158, portfolio balancing effect linked to 168–69, 185 Federal Reserve purchases of, household debt levels preceding, 157–58, 161–63 36, 159, 163–66 sellers’ motivations and, 12–13, 30 housing bubble and, 17, 28, 48, 78, tax rebates compared to Federal 80, 158, 165 Reserve purchasing of, 29–30 interest rate cuts during, 28, government debt 158, 172 anxieties regarding standard fiscal Keynesian models and, 170 stimulus and, 2–3, 20, 30–32, Lehman Brothers collapse (2008) 54–55, 171, 211, 218 and, 17, 28, 52, 107, 165 Blanchard’s warnings regarding, 31 Meltzer on, 153–56 Federal Reserve-held Treasury monetary stimulus and, 51, 130, bonds and, 19–20, 33, 38–39, 41, 155–56, 158–59, 170–72, 185 143, 190 recommendations regarding interest rates and, 3, 20, 213 fiscal stimulus following, investor confidence and, 3, 54–55, 133–34 20–21, 31, 213 rescue of financial firms and, 28, negative economic consequences 52, 158, 168 of, 3, 20–21, 31–32, 213 secular stagnation and the slow paper money contrasted with, recovery from, 196 7–8, 19–20 standard fiscal stimulus and, 2, 31, Reinhart and Rogoff’s warnings 49, 51–53, 88, 116–17, 170–71, regarding, 31–32 185, 196 standard fiscal stimulus and, 2–3, stimulus without debt proposal 18–19, 29–30, 33 and, 27, 32, 54–55, 133–34, Great Depression 136, 156, 171–72 asset price decline during, 167 stock market declines and, household debt levels preceding, 78, 80, 82 36, 164 subprime mortgage crisis interest rate cuts and, 47, 172 and, 78 Great Recession (2008–10). See tax rebate proposals also American Recovery and and, 54, 171 Reinvestment Act unemployment and, 2, 52–55, Blinder and Zandi’s analysis 108, 130–31, 158–59, of, 169–71 162–63, 170–71 Congress and, 52, 54–55 Greenspan, Alan, 144, 154–55

232

Index

Hansen, Alvin, 196 Federal Reserve’s mandate helicopter money regarding, 23, 29, 33, Bernanke and, 184, 193 42, 151–55 Buiter on irredeemable nature of full employment economies and, fiat money and, 192 10, 14–15, 127–29, 135 consumer demand and, 192, 194 helicopter money and, 35, 180, 191 Draghi and, 193 interest rates and, 132, 135, Friedman’s parable of, 34–35, 154, 156 179–81, 184, 193 paper money and, 10, 14, full-employment economies and, 141, 143–44 35, 179–81 in recessionary economies, 127–32 inflation and, 35, 180, 191 stimulus without debt proposal Lonergan on the European Central and, 5, 13, 40–41, 127–36, 180, Bank and, 176–77 191, 209–10, 212, 216–17 monetary stimulus and, 181 infrastructure repairs, 102, 126, 216 stimulus without debt compared Inman, Robert, 121–22 to, 34–35, 180–81, 190 interest rates Wren-Lewis on, 190–91 bank reserve requirements Herndon, Thomas, 32 and, 135 Hicks, John, 48 consumer demand and, 28, “High Debt, Default Risk, and Vicious 36, 47–48 Cycles” (Blanchard), 31 Federal Reserve’s authority to set, Hodgson, Graham, 143 27–28, 42, 47–48, 135, 151–52 “Household Expenditure and government debt and, 3, 20, 213 the Income Tax Rebates of Great Depression and, 47, 172 2001” (Johnson, Parker, and Great Recession and, 28, 158, 172 Souleles), 70–72 inflation and, 132, 135, 154, 156 House of Debt (Mian and Sufi), 163 monetary stimulus through “How about Quantitative Easing for the lowering of, 1, 13–14, 18, 21, People?” (Kaletsky), 189 27–29, 31, 36, 47–50, 76, 151–52, 158–63, 166–72, “The Inadequacies of Monetary 197, 217 Control of the Business Cycle” quantitative easing and, 36 (Samuelson), 47 retirees and, 161 “In a Recession, Think Tax Rebates, small business survey data Not Across-the-Board Cuts in regarding, 50 Taxes” (Seidman), 63 Taylor rule and, 155 inflation International Accounting Standards American Recovery and (IAS), 143 Reinvestment Act and, 130–31 International Financial Reporting consumer demand and, 10, Standards (IFRS), 143 127–30, 135–36 International Revenue Service (IRS), 36, definition of, 127 73, 93–94

Index

Japan central bank in, 184–85 government spending during recession in, 167 Mankiw on, 188 public debt in, 167 real estate market decline (1990s) in, 166–67 recession (1990s) in, 166–67, 183– 84, 188, 193 Johnson, David tax rebate of 2001 analyzed by, 58, 60, 62, 69–72, 85, 95 tax rebate of 2008 analyzed by, 58, 62, 84–86

233

size of, 62, 89, 104 taxpayers’ awareness of, 91, 110 tax withholding reduction mechanism for, 62, 84, 89–92, 104, 110 Mankiw, N. Gregory, 188–89 marginal propensity to consume (MPC), 67–69, 81–83, 94 McClelland, Brendan, 58, 62, 84–86 Medicaid, 99, 109–10, 119–21 Medicare, 93 Meltzer, Alan, 153–56 “Memo to Tokyo: Cut Taxes, Print Money” (Mankiw), 188 Mian, Atif, 49–50, 163–66 Miran, Stephen, 81 Kaletsky, Anatole, 189 Modigliani, Franco, 48, 88, 162 Keynes, John Maynard, 13–14, monetary stimulus 48, 196 balance-sheet recessions Klein, Lawrence, 48 and, 166–69 Koo, Richard, 166–68 Friedman’s helicopter parable Krugman, Paul, 48 and, 181 government debt and, 19, 190 Larrain, Felipe, 140–42 Great Recession (2008–10) and, “Legal Helicopter Drops in the 51, 130, 155–56, 158–59, Eurozone” (Lonergan), 176 170–72, 185 Lehman Brothers collapse (2008), 17, inflation and, 192 28, 52, 107, 165 interest rate cuts and, 1, 13–14, Lerner, Abba, 13–14, 212 18, 21, 27–29, 31, 36, 47–50, Lewis, Kenneth 76, 151–52, 158–63, 166–72, plausible economic model testing 197, 217 of stimulus without debt in Japan (1990s), 184 and, 200 Keynesianism and, 48 tax rebate of 2001 analyzed limits of, 1, 13–14, 18, 28–29, 31, by, 67, 79 36, 47–50, 76, 157, 159–61, tax rebate of 2008 analyzed by, 163, 166–69, 172, 185, 193, 61, 79–82 197, 217 Liscow, Zachary, 119–21 New Classical economists and, Lonergan, Eric, 176–77, 189–90 49, 160 New Keynesianism and, 49, 160 Making Work Pay credits (2009–11) phasing out of, 23 Republican Party’s push to private debt and, 21 eliminate, 92, 105 quantitative easing and, 36

234

Index

monetary stimulus (cont.) quantity theory of money and, 48 recessions addressed through, 1, 51, 130, 155–72, 185 tax rebates compared to, 50, 76, 161 Taylor rule and, 155, 202, 205, 208–10 monetization of debt, 33, 35, 191 Money-Financed Fiscal Program (MFFP, Bernanke), 185–87 “Money from Heaven” (Economist article), 193 mortgage-backed securities (MBSs), 134, 162 Moser-Boehm, Paul, 145 Muellbauer, John, 175–77 multiplier effect of stimulus programs definition of, 113 Keynesianism and, 116–17 mistakes in estimating, 97, 112–17 recession economies versus full-employment economies and, 113–16 tax rebates and, 51 unemployment levels and, 115 World War II and, 115

temporary payroll tax cut (2010–13) and, 53 Olney, Martha, 164 Open Market Committee. See under Federal Reserve “Out of Ammo?” (Economist article), 193

paper money Federal Reserve’s holding of, 15, 23, 29, 41–43, 45, 137–40, 143–44, 146–49, 214, 217 as “fiat money,” 140–44, 192 government accounting rules regarding, 3–4, 8–10, 12, 19, 137–43, 146–48, 192, 212, 217 government debt contrasted with, 7–8, 19–20 inflation and, 10, 14, 141, 143–44 lack of guarantees for, 4, 8–10, 19, 139–42, 192, 212 previous gold guarantee for, 4, 8–9, 19, 140–42, 144 tax rebates as means of providing, 11–12 Parker, Jonathan tax rebate of 2001 analyzed by, 58, 62, 69–72, 85, 95 tax rebate of 2008 analyzed by, 58– 59, 62, 84–87 National Federation of Independent tax rebate research of, 87–89 Businesses (NFIB), 49–50, 166 payroll taxes New Classical economists, 49, 88, employer payroll taxes and, 94 116, 160 Medicare payroll tax and, 93 New Keynesian economists, 49, 88, recommendations against 116–17, 160–61 temporary cuts to, 104–5, 126, 216 Obama, Barack regressive nature of, 93 American Jobs Act proposal (2011) Social Security Trust Fund and, 62, and, 110–11 92, 105 American Recovery and temporary cut (2011–12) to, 62, Reinvestment Act and, 84, 92–93, 105, 110 107, 109–10 Pelosi, Nancy, 72 Making Work Pay Credit and, Pollin, Robert, 32 92, 103

Index

portfolio rebalancing effect, 157–58, 161–63 “Print Less but Transfer More” (Blyth and Lonergan), 189–90 quantitative easing (QE), 35–36, 170–71, 189, 194. See also Federal Reserve: government bond purchases by quantity theory of money, 48 Reagan, Ronald, 63 recession of 1990, 88 recession of 2001, 63, 100, 107 recessions. See also Great Recession (2008–10) balance-sheet recessions and, 166–69 benevolent rulers’ approach to combating, 7, 10–14, 16 consumer confidence and, 23–24 consumer demand decline and, 1, 5, 10–11, 13–14, 25, 38, 159, 164, 219 credit liquidity and, 5, 27 financial crises and, 168 household debt and, 164 housing market declines and, 23, 166 in Japan (1990s), 166–67, 183–84, 188, 193 monetary stimulus and, 1, 51, 130, 155–72, 185 stock market declines and, 21, 23, 166 tax rebates as means of addressing, 4, 25–26, 33, 59, 95, 97, 103 tax revenue declines during, 2, 98, 119, 153 Reinhart, Carmen, 31–32 Republican Party American Recovery and Reinvestment Act opposed by, 53, 84, 89, 103, 107, 112

235

economic stimulus debates in Congress (2009–10) and, 46 election of 2010 and, 110 long-term tax cuts supported by, 106 Making Work Pay Credit and, 92, 105 payroll tax cut (2011–13) and, 92 spending on long-term programs opposed by, 106 Ricardian equivalence, 89 Rogoff, Kenneth, 31–32 Romer, Christina, 78 Sachs, Jeffrey, 140–42 Sahm, Claudia, 61, 83–84, 90 sales taxes, 93–94 Samuelson, Paul, 47–48, 160 secular stagnation chronically insufficient demand as a cause of, 195–99, 217 introduction of the term, 196 slow growth in potential output as cause of, 195, 198–99 stimulus without debt proposal and, 24, 195–99, 217–18 Seidman, Laurence on accounting rules regarding Federal Reserve-held government bonds, 38–39 Great Recession stimulus recommendation by, 108 Making Work Pay credit analyzed by, 90 plausible economic model testing of stimulus without debt and, 200 “Stimulus without Debt” article (2013) by, 32–33, 182 tax rebate of 2001 analyzed by, 67–68, 79 tax rebate of 2008 analyzed by, 61, 79–82 tax rebate proposal for 2001 by, 63

236

Index

Shapiro, Matthew American Economic Conference presentation (2009) by, 83 tax rebate of 2001 analyzed by, 60, 65–72 tax rebate of 2008 analyzed by, 61, 74, 83–84 Shiller, Robert J., 48 Siegel, Jeremy, 135 “The Simple Analytics of Helicopter Money” (Buiter), 192 Slemrod, Joel American Economic Conference presentation (2009) by, 83 tax rebate of 2001 analyzed by, 60, 65–72 tax rebate of 2008 analyzed by, 61, 74, 83–84 Solow, Robert, 48 Souleles, Nicholas tax rebate of 2001 analyzed by, 58, 60, 62, 69–72, 85, 95 tax rebate of 2008 analyzed by, 58, 62, 84–86 standard fiscal stimulus assumptions regarding bank lending and, 49–50 Congress and, 2–4, 20, 185, 197, 211 consumer demand and, 5, 18 effectiveness of, 1, 4–5, 169–72 federal deficit and, 197, 199, 210–11 government bond purchases and government debt increases in, 2–3, 15, 18–19, 29–30, 33 government spending and, 1, 18, 32, 56 grants to state governments and, 1, 4, 56–57 Great Recession and, 2, 31, 49, 51–53, 88, 116–17, 170–71, 185, 196

inflation and, 192 in Japan, 183–84 Keynesian perspectives on, 88 monetization of debt and, 33, 35, 191 New Classical economists perspectives on, 88, 160 New Keynesian perspectives on, 88, 160–61 stimulus without debt compared to, 18–20, 29–30, 32–33, 35 tax cuts and, 1, 56 tax rebates and, 1, 4–5, 11, 18, 21, 25, 32, 181 Taylor’s criticism of, 122–23, 125 Stiglitz, Joseph, 48 “Stimulus without Debt” (2013 journal article by Seidman), 32–33, 182 “Stimulus without Debt in a Severe Recession” (Lewis and Seidman), 200 stimulus without debt proposal Bernanke’s proposal regarding, 185–88 “bond burning” proposal contrasted with, 39–40 Congress and, 5–6, 15, 18, 21–23, 27, 29, 32–34, 37–38, 40, 44–45, 147, 150, 186–87, 190, 192, 214–15, 218–19 consumer confidence and, 12–13, 23–24, 41 consumer demand and, 17, 23–24 in countries other than the United States, 173–78, 217 economy below full employment as a stipulation for, 10, 12–16, 35, 41, 129–30, 180 Eurozone countries and, 174–78, 217 federal grants to state governments and, 38, 46

Index Federal Reserve balance sheets and, 23, 41–42, 137–39, 143–44, 146–49, 217 Federal Reserve independence and, 42–43, 150–56, 186–87, 217 Federal Reserve transfers to households compared to, 36–37, 187–90 Federal Reserve transfers to the Treasury under, 3, 5–6, 11–12, 15, 17–18, 21–22, 26– 27, 29, 32–35, 37, 40–41, 43–46, 55–56, 96, 133, 136–38, 146–47, 150, 152, 156, 171, 180–82, 185, 192, 198–99, 203– 5, 208–10, 212–15, 217–19 government deficits and, 55–57, 185–86, 199 government spending and, 13–14, 21, 23, 33 Great Recession (2008–10) and, 27, 32, 54–55, 133–34, 136, 156, 171–72 growth rate of potential output and, 51 helicopter money compared to, 34– 35, 180–81, 190 inflation and, 5, 13, 40–41, 127–36, 180, 191, 209–10, 212, 216–17 monetization of debt compared to, 35 phase-out period and, 3, 13, 23– 24, 41, 127, 129–31, 133–36, 212, 216–17 plausible macroeconomic model testing of, 200–210 potential court challenges regarding, 44 president’s role in, 15, 33–34 private debt and, 21 quantitative easing compared to, 35–36

237

secular stagnation and, 24, 195–99, 217–18 separation of powers in, 15, 22, 33–35, 37–38, 176, 180, 187– 88, 192, 214–15 standard fiscal stimulus compared to, 18–20, 29–30, 32–33, 35 tax increases following, 13, 23 tax rebates and, 5–6, 12, 16, 21– 24, 26–27, 29, 33–34, 36, 46, 138, 188, 198, 214–15 temporary nature of, 23–24 temporary tax incentives for business investment and, 21, 38, 46 unemployment and, 14, 16, 181 Sufi, Amir, 49–50, 163–66 Summers, Lawrence, 196–97 supply-side economics, 63 tax cuts of 2001, 107 tax rebate of 1975, 38, 63 tax rebate of 2001 amounts dispersed to households in, 64, 74 Congress and, 38, 63–64 consumer spending and, 58–60, 62, 65–72, 85 Democratic Party opposition to elements of, 63 disbursement schedule for, 59, 64–65, 67, 69–70, 74, 79, 88 economy’s performance following, 65 Johnson on, 58, 60, 62, 69–72, 85, 95 Lewis on, 67, 79 paper check format of, 59–60, 64–65, 70, 74, 90, 92, 104 Parker on, 58, 62, 69–72, 85, 95 presidential election of 2000 and, 63 recession of 2001 and, 63

238

Index

tax rebate of (cont.) cutting withholding of taxes Seidman on, 67–68, 79 compared to, 89–93 Shapiro on, 60, 65–72 effectiveness of, 4–6, 25, Slemrod on, 60, 65–72 58–96, 215 Souleles on, 58, 60, 62, Federal Reserve purchase of 69–72, 85, 95 government bonds compared tax rate cuts accompanying, 64 to, 29–30 tax rebate of 2008 government spending bipartisan support for, 45–46 following, 13, 23 Broda on, 58–59, 62, 86–87, 95 long-term spending increases Congress and, 37 compared to, 25–26, 103 consumer spending and, 58–59, long-term tax cuts compared to, 61–62, 75–87, 90 25–26, 103, 106 disbursement schedule for, paper money provided 59–60, 73–74, 79, 85, 88 through, 11–12 electronic fund transfers (EFTs) raises from employers and, 73, 91–92, 104 compared to, 88 Feldstein on, 61, 74–84, 90 recommendations regarding Great Recession and, 33, 72 size of, 54 housing bubble and, 72, 78, 80 recommended future designs Johnson on, 58, 62, 84–86 for, 93–96 Lewis on, 61, 79–82 standard fiscal stimulus and, 1, 4– McClelland on, 58, 62, 84–86 5, 11, 18, 21, 25, 32, 181 paper checks and, 73–74, standard monetary stimulus 89–90, 92, 104 compared to, 50, 76, 161 Parker on, 58–59, 62, 84–87 stimulus without debt proposal Sahm on, 61, 83–84, 90 and, 5–6, 12, 16, 21–24, Seidman on, 61, 79–82 26–27, 29, 33–34, 36, 46, 138, Shapiro on, 61, 74, 83–84 188, 198, 214–15 size of, 72–73, 85, 93, 106 tax increases following, 13, 23 Slemrod on, 61, 74, 83–84 temporary nature of, 25–26, 215–16 Souleles on, 58, 62, 84–86 Treasury and, 4 subprime mortgage crisis and, 78 Tax Reduction Act of 1975, 63 Taylor on, 61, 74–84, 90 tax withholding tax rebates. See also tax rebate of Making Work Pay credits via 2001; tax rebate of 2008 reductions in, 62, 84, 89–92, bipartisan support for, 26, 216 104, 110 Congress and, 4, 18, 22, 25, 29, 37– recommendations against 38, 46, 63–64, 90, 96, 182–83 temporary cuts to, 103–4, consumer spending and consumer 126, 216 demand impacted by, 11–12, 21, taxpayers’ awareness of small 24–25, 30, 50, 58–62, 65–72, changes to, 91 75–91, 95, 100, 171, 192, 215 tax rebates compared to, 89–93

Index

Taylor, John American Economic Association presentation (2009) by, 75 standard fiscal stimulus during recessions critiqued by, 122–23, 125 standard regression analysis by, 78–79, 81–82 tax rebate of 2008 analyzed by, 61, 74–84, 90 Taylor rule and, 155, 202, 205, 208–10 Wall Street Journal op-ed (2008) by, 74–75, 77–78, 82 Temin, Peter, 164 temporary tax incentives for business investment bonus depreciation provisions and, 100–101 business pressure to establish permanence of, 101 Economic Stimulus Act of 2008 and, 101 investment tax credits and, 100 stimulative effect of, 100–101, 126, 216 stimulus without debt proposal and, 21, 38, 46 Tinbergen, Jan, 48 Tobin, James, 48, 162 Treasury of the United States “bond-burning” plan and, 39–40 bond payment obligations and, 39 credit liquidity and, 5 rescue of financial firms and, 28, 158, 219 standard fiscal stimulus and, 18

239

tax rebates and, 4 transfers from Federal Reserve under stimulus without debt plan and, 3, 5–6, 11–12, 15, 17–18, 21–22, 26–27, 29, 32– 35, 37, 40–41, 43–46, 55–56, 96, 133, 136–38, 146–47, 150, 152, 156, 171, 180–82, 185, 192, 198–99, 203–5, 208–10, 212–15, 217–19 Turner, Adair, 183 unemployment American Recovery and Reinvestment Act (2009) and, 52–53, 108, 111, 118–21 Federal Reserve’s mandate regarding, 23, 29, 33, 42, 151–52 Great Recession and, 2, 52–55, 108, 130–31, 158–59, 162–63, 170–71 stimulus without debt proposal and, 14, 16, 181 Volcker, Paul, 154 “Voodoo Multipliers” (Barro), 115 Whelan, Karl, 142 Wilson, Daniel, 117–21 Wood, Richard, 191 Woodford, Michael, 191 Woolston, William Gui, 119–21 Wren-Lewis, Simon, 190–91 Zandi, Mark, 169–71

How to Combat Recession; Stimulus without Debt - Laurence Seidman - 2018

Articles inside

REFERENCES

INDEX

14. Are We Ready for the Next Severe Recession?

Stagnation?

Debt?

Countries?

Would Stimulus without Debt Work

Recession?

Independence?

Sheet?

Can Stimulus without Debt Combat Secular

What Have Others Written about Stimulus without

Can’t Monetary Stimulus Overcome a Severe

Can Stimulus without Debt Be Used by Other

Introduction | 1

What Is Stimulus without Debt? | 17

Would Stimulus without Debt Undermine the Fed’s

What about Other Kinds of Fiscal Stimulus? | 97

Do Tax Rebates Work in a Recession? | 58

Would Stimulus without Debt Weaken the Fed’s Balance

Would Stimulus without Debt Be Inflationary? | 127

How Would a Benevolent Ruler Combat a Recession? | 7