The chapter looks at three recent episodes where, for better or for worse, fiscal policy played or is playing a major role.
Published onMar 30, 2022
Chapter 6| Fiscal policy in action
The chapter looks at three recent episodes where, for better or for worse, fiscal policy played or is playing a major role. The purpose is not to review them in full, which would take another book, but to show and discuss fiscal policy choices in the light of the analysis so far.
To caricature just a bit, the three episodes can be thought of “too little’’, “just right?’’, “too much?’’
Too little? The first section looks at the period of “fiscal austerity’’ that took place in the wake of the Global Financial Crisis. After the large initial increase in debt due to the crisis, the focus quickly turned to debt reduction. This was particularly true in the European Union, which embarked on a strong fiscal consolidation. Today, there is fairly wide agreement that, at least in Europe, the fiscal consolidation was too strong, too much based on the traditional view of debt, both by markets and by policy makers, and came at a substantial output cost.
Just right? The second section looks at the Japanese economy over the last three decades. Japan experienced the ELB constraint starting in the mid-1990s, earlier than either the United States or Europe, and has remained close to it ever since. Japanese macroeconomic policy is often characterized as a failure, with the central bank unable to achieve its inflation target, a low growth rate, and debt ratios steadily rising to reach more than 170% for net debt and 250% for gross debt. I think it should be seen instead as a qualified success, with the use of aggressive fiscal and monetary policies to compensate for very weak private demand: Output has remained close to potential. Growth is low, but mostly because of demographics, not because of debt. Inflation is low, lower than the target, but this is not a major failure. Looking forward however, there are reasons to worry: The debt ratios are very high. So far, investors do not mind, and 10-year nominal rates are close to zero. But can the build-up of debt continue? What happens if interest rates increase? Are there alternatives?
Too much? The third section looks at the effects of the American Rescue Plan, the stimulus program put in place by the Biden administration in early 2021. In 2020, the focus of fiscal policy had been protection, of both households and firms. In early 2021, the goal partially shifted from protection to sustaining the recovery. The size of the program was extremely large relative to apparent output gap. The strategy (intentional or not) was in effect twofold. For the Treasury, to strongly increase aggregate demand, and thus increase the neutral rate so as to relax the ELB constraint. And for the Fed, to delay adjustment of the policy rate to the neutral rate, allow for some overheating and generate slightly higher inflation in the process. To a number of observers, the size of the program appeared too large, leading to worries about overheating and excessive inflation. The section takes stock of where things are at the time of writing.
6.1 Fiscal austerity in the wake of the GFC
Go back to the Global financial crisis, which started in earnest in the fall of 2008.
For the remainder of 2008 and much of 2009, the reaction of governments was to spend whatever appeared needed, without worrying much about debt. The response was strongest in the United States, where the “American Recovery and Reinvestment Act” (ARRA), passed in February 2009, committed to additional spending and tax measures adding to 830 billion dollars, or 5.9% of GDP. The 2009 primary balance was equal to -11.2% of GDP, up from -4.6% in 2008. In Europe, the “European Economic Recovery Plan” (EERP), adopted in November 2008, recommended that governments take measures adding to 200 billion euros for the European Union as a whole, or 1.5% of EU GDP. The Euro area 2009 primary balance was -3.8%, up from 0.4% in 2008. Japan passed a number of programs as the crisis got worse, for a total of 29 trillion Yen in additional measures, or 5.8% of GDP. Japan’s primary deficit for 2009 was -9.3%, up from -3.8% in 2008.1
As a result, at the end of 2009 the (net) debt ratio stood at 63% for the United States, up 11% compared to the end of 2008; it stood at 62% for the euro area, up 8%, and at 96% for Japan, up 11.3%. In addition to the increase in official debt, there was a large increase in contingent liabilities, as governments embarked on off-balance sheet operations, such as the “Troubled Asset Relief program” (TARP), passed in the United States at the end of 2008, authorizing the purchase or the insurance of assets up to 700 billion dollars.
In parallel, central banks decreased rates down to or close to the zero lower bound. In the United States, the federal funds rate decreased from 2.0% in August 2008 to 0.1% by December. In Europe, the Euro discount rate decreased from 5.25% in September of 2008 to 1.75% by March of 2009. In Japan, which was already de facto at the effective lower bound before the crisis, the policy rate was reduced from 0.75% in September to 0.30% by the end of the year.
From the start, governments had insisted on the need for “timely, targeted, and temporary” fiscal measures. By the end of 2009, as economies appeared to slowly recover, the focus turned to debt consolidation. To get a sense of the shift in emphasis, I examined the G20 communiqués and the executive summaries of the IMF Fiscal Monitor. I used the following grading from +2 to -2: A focus on output stabilization rather than debt reduction was graded +2; a focus on debt reduction rather than output stabilization was graded -2. More balanced statements were graded between +2 and -2. The exercise is obviously subjective (although not obviously less reliable than an AI approach to the same issue) but turns out to be highly suggestive.2
Source: G20 communiqués and IMF Fiscal Monitor Executive summary
The results are presented in Figure 6.1. After expressing strong support for output stabilization until early 2010, the G20 communiqués took a sharp turn and focused nearly exclusively from then on on the need to decrease debt, despite the fact that central banks were still at or very close to their effective lower bound. The IMF was more dovish, also shifting in 2010 to an emphasis on debt reduction, but, as the recovery sputtered, putting more weight on output stabilization from 2011 on.
The shift to a focus on debt reduction was particularly strong in the European Union.
Pre-crisis, in June 2008, the assessment of the European Union Commission Public Finance reports had been that most countries satisfied the Maastricht criteria.
In 2009, the Commission expressed support for the EERP, but insisted on the “temporary” aspect of the stimulus.
In 2010, the Commission decided that there would be no suspension of the EU fiscal rules, and shifted the focus to an “exit strategy.” Given the increase in debt, it put most countries in the so called “corrective arm” of the Stability and Growth Pact, and started the Excessive Deficit Procedure (EDP), requiring countries to return to the medium term objective for debt within a horizon of 2 to 3 years. It invoked increasing spreads and thus the need to reassure markets, but the message was more general, applying to all member countries, including those whose spread had not increased. The message was unambiguous:
In view of the challenges, the planned pace of the fiscal consolidation should be ambitious, and will have to go well beyond the benchmark of 0.5% of GDP per annum in structural terms in most Member States.
In 2011, despite the slowing recovery, and the fact that the ECB was still close to the effective lower bound, the Commission doubled down, insisting on the need not just to stabilize debt ratios but to decrease them:
The issue of sustainability has emerged as the key concern in the immediate post-crisis years. Soaring deficits and off-balance-sheet operations to support the financial sector have led to a large increase in debt for nearly all European Union countries. Despite the fact that a return of GDP growth, a gradual withdrawal of the temporary support measures and the start of consolidation is starting to reduce deficits, debt is still expected to continue increasing for the next year or so in most cases. Once it has reached its peak, the issue is not over. It will not be sufficient to stem the increase; rather, additional consolidation measures will be required to reduce it from its new level, not least because population ageing is due to have an increasingly negative effect on the public finances and put pressure on their sustainability in coming decades.
It proposed changes in the rules, which in effect made the rules tougher. It added expenditure benchmarks, introduced numerical criteria for the minimal speed of debt ratio adjustment, namely 1/20th of the difference between the debt ratio and 60%.
In 2012, the Commission tripled down, despite negative growth, -0.8% (year on year) for the euro area.
The significant increase in debt ratios seen since the start of the crisis alongside the still sizeable deficits mean that there is little scope for many Member States to ease off the fiscal tightening, despite the extra pressure that this might put on already faltering growth. Amid the debate about how best to continue to respond to the crisis, concerns have been raised that further fiscal consolidation amid weak growth prospects may have self defeating effects on debt ratios. Part III presents a detailed analysis that highlights how such effects may arise but concludes that such cases are rather theoretical and anyhow short-lived under reasonable economic assumptions.
What is interesting in that quote is that the concern is not about growth per se, but about whether lower growth will make debt reduction self defeating.
In 2013, growth was still negative, -0.2%. “Whatever it takes”, announced by Mario Draghi in July 2012, had considerably reduced the spreads on sovereign bonds, especially so for Portugal, Italy, and Ireland, but there still was no change in the message: Admittedly, “deleveraging of the public sector” was tough and affected output, but it was necessary to continue fiscal consolidation:
Improvements in the financial conditions have had limited impact on the real economy so far. Economic activity disappointed in the second half of 2012, and turned out weaker than expected in the first quarter of 2013. This was due to two interrelated set of factors. First, because of persistent weaknesses in the banking sector, the improvement in the financial markets’ situation has not yet fed in the credit growth. [...] Second, the process of deleveraging of the private and the public sector is still on-going in many economies, and this weighs on aggregate demand. In particular, domestic demand remains muted due to high unemployment and as a result of persistent uncertainty amongst households and enterprises regarding the future economic outlook and the development of the debt crisis. At the same time, given the remaining fiscal sustainability concerns, governments in many Member States have to continue the necessary fiscal retrenchment.
How costly, in terms of output, was fiscal austerity? Counterfactuals are not available, but what can be done is to look at the cross-country evidence. Simple bivariate graphs have obvious limits and show correlation rather than causality, but they are suggestive. Figure 6.2 plots the change in the output gap between 2014 and 2008 against the change in the cyclically adjusted primary balance during the same period, both as ratios to GDP, for the 27 advanced economies for which the data is available. There is a clear significant and negative relation between the two. The regression coefficient, -0.48, implies that a 1% higher cyclically adjusted primary surplus was associated with a worsening of the output gap of about 0.5%, both as ratios to GDP. The results are similar when only European Union members are included (the .regression coefficient is -0.32)34
Source: World Economic Outlook for output gap, Fiscal monitor for cyclically adjusted primary balance.
In summary: In the wake of the global financial crisis, and while central banks were still at or close to the ELB, too much weight was put by policy makers on debt reduction relative to output stabilization. Put another way, it was a case of too little fiscal support. The costs of high debt were perceived to be very high, higher than they truly were, and the multipliers were underestimated, leading to an underestimate of the output costs of fiscal consolidation.567
6.2 The Japanese experience. Success or failure?
The economic experience of Japan since the early 1990s, and the underlying monetary and fiscal policies, are seen by some as a major failure.8 I shall make the argument that it is in fact a qualified success.9 Nevertheless, implications and the risks of the high levels of debt must be carefully considered. These are the themes of this section.
The case for failure appears straightforward: Poor output growth, a consistently missed inflation target, large budget deficits, and a legacy of very high public debt: Japan’s growth since 1992 has averaged 0.8% relative to an OECD average of 2.0%.10 Fiscal policy has been characterized by large deficits and a steady increase in debt ratios, with net public debt equal to 171% of GDP in 2020, gross debt equal to 250%. CPI inflation has run at an average of 0.3%, much lower than the target of 2%.
The case is however weaker that it seems.
Lower Japanese growth reflects mostly lower population growth, 0% since 1995, and by implication, lower labor force growth, 0.1%.11 Productivity growth measured as output per worker has been 0.6%, similar for example to 0.5% for the EU19 (the 19 advanced economy members of the European Union before the last extension), although lower than the 1.6% for the United States. Productivity growth measured as output per hour has been 1.3%, higher than the EU19 1.0%, but again lower than the US 1.7%.
The unemployment rate has remained low, only slightly exceeding 5% twice, first in 2001 after a steady increase in response to the crash of the asset bubble (which happened 10 years earlier), then during the Global Financial Crisis in 2009, but returning over time to 2.8% in 2021, close to its value of 2.1% in 1990.
Inflation has remained substantially lower than the target of 2%, alternating periods of small inflation and small deflation around an average of 0.3% for CPI inflation since 1992. How should this be interpreted? The Phillips curve relation implies that, if expectations are stable, an inflation rate lower than the expected inflation rate indicates that unemployment is above the natural rate, that there is a negative output gap. Evidence on inflation expectations suggests however that inflation expectations have remained lower than the target, around 1% on average, but still not quite as low as actual inflation.12 This suggests that, on average, output has remained below, but close to, potential.
It is true however that this was accompanied by a steady string of large deficits and a large increase in public debt, to which I now turn.
There are two main reasons why governments run budget deficits. The first is unintentional: Faced with adverse shocks or adverse trends, governments are unable to raise sufficient revenues to cover public spending. The other is intentional: Faced with weak private demand, governments run deficits to sustain demand and output (automatic stabilizers are a mix of the two). The question is which one of the two best characterizes Japanese fiscal policy since 1990.
The evolution of primary deficits, as ratios to GDP, in Japan since 1990, is shown in Figure 6.3.
It makes clear that, in the 1990s, the first reason played a central role. The sharp shift from a primary surplus of 3% in 1990 to a primary deficit close to 10% in 1998 was essentially due to the dramatic decrease in growth (from 4.3% in the 1980s to 1.3% in the 1990s) following the bursting of the asset bubble. But aging has also had an important and adverse role on public finances through most of the period: The proportion of people 65 years and older has doubled, reaching 30% in 2021, leading to steadily higher health and retirement spending pressure. Successive governments have found it hard to sufficiently increase taxes in parallel, leading to unintentional deficits and attempts by governments to reduce them.
At the same time, governments have been aware of the need to sustain demand and output in the face of weak private demand, and of the macroeconomic risks in reducing deficits.
This tension between the two goals has led to somewhat schizophrenic but not unusual outcomes. In the 1990s, ambivalence about whether to limit the increase in debt or sustain demand led to stop-and-go policies: A large expansion in 1995 was followed by contraction in 1996 and 1997.13 From 2000 on, as shown in the figure, governments started to set 10-year out zero primary balance targets, implying a steady reduction in deficits, of the order of 0.6% to 0.8% per year over the following 10 years, independent of the need for demand support. These announced target paths were actually more than words: Figure 6.3 shows that, leaving aside the adverse shocks from the Global Financial Crisis and the Covid crisis, which each led to a large increase in deficits, successive governments were actually staying close to their announced paths. But those two crises led to much larger deficits and a reset of the adjustment from large initial deficit ratios.
The following statement by Prime Minister Abe in 2017 is a good example of how recent governments perceive the balance between the desirability of reducing debt and the need for macro stabilization:14
If we were to cut the budget for the next fiscal year in half, the primary balance would turn into surplus. The moment it turns into surplus, the Japanese economy will be as if it is dead, and from the following year onwards, disastrous things will happen.
The current government has made clear that there may be a need to run deficits to sustain demand if needed, together with the fact that with (r−g)≤0, this allows for government to run some deficits while allowing the debt ratio to stabilize or even to decrease. The following quote from the current vice minister of finance states it as follows:15
In order to achieve fiscal soundness, the single year deficits must be reduced more than sufficiently (or, to be precise, at least to within“the surplus of the growth rate minus the interest rate”) during the interest rate bonus period. [note: the period during which g exceeds r is called the bonus period] If this is done, further deterioration of public finances can be avoided somehow. This is the bottom line (minimum goal) that Japan should aim for.
The tension between sustaining demand and decreasing debt is still there. Whether the statement in the quote above is consistent with the promise to go back to primary balance by 2025, starting from a primary deficit of 8.4% in 2020, and with the Bank of Japan (BOJ) still at the effective lower bound, is doubtful. Indeed, IMF forecasts of the Japanese primary balance in 2025 (which however includes social security funds) is -2.0%.
Overall, one can see the combination of fiscal and monetary policy, that is large (partly unintentional, partly intentional) deficits on the one hand and monetary policy at the ELB on the other as having maintained output in Japan close to potential in the face of chronically low private demand.16 This is the sense in which Japan can be considered a qualified success (again, it is hard to know what the counterfactual would have been).
Still, deficits have been so large that they have much more than compensated the favorable debt dynamics coming from (r−g)<0; gross debt ratios have increased from 63% in 1990 to a forecast of 250% by the end of 2021, net debt ratios have increased from 19% in 1990 to a forecast 171% by the end of 2021. This raises three sets of questions:
What deficits can Japan run to sustain output while preventing the debt ratio from increasing further?
To answer this question, a rough computation, based on the basic equation for debt dynamics, is useful. 5-year nominal rates are equal to -0.1%; 5-year forecasts of inflation are 1.0% annually; 5-year forecasts of growth, leaving out the high growth rate from the recovery from Covid in 2022, are 0.8% annually.17 The net debt ratio is 171%. This implies that (r−g)b is equal to (-0.1% - 1.0% - 0.8%) *171%, so -3.2%. This implies that the Japanese government can, in expected value, run a primary deficit of 3.2% and stabilize the debt ratio.
Is it enough to get and keep output at potential? For this, one needs a macro model, and assumptions about the strength of private demand post Covid, but the answer would appear to be yes: IMF forecasts for 2025 are that, with primary deficit ratios of 2%, the unemployment rate will be 2.3%, so close to the natural rate. Under those assumptions, the IMF forecasts a slight decline in the net debt ratio down to 169% from 171% in 2021.
This suggests that, under current forecasts, and despite very high debt, Japan does not face a debt sustainability problem. A theme of Chapter 4 on debt sustainability was however that a full assessment required to explore outcomes not just under the point forecasts, but taking into account the uncertainty and thus the distributions of the various variables involved. So, this triggers the next question, what would happen if interest rates increased substantially?
What would happen if interest rates increased substantially?
All the variables affecting debt dynamics are uncertain, from the primary balance needed if Covid turns out to be persistent, to the underlying growth rate of the economy, etc. Clearly the main worry (from the point of view of debt sustainability) is that real interest rates might increase substantially and make it difficult to avoid a debt explosion. To discuss the issue, we can again draw on the discussion of debt sustainability in Chapter 4.18
How much one should worry about interest rate increases depends first on the average maturity of the public debt. The longer the maturity of the government debt, the less the government budget constraint is affected by a temporary increase in interest rates, and the more time the government has to adjust to a permanent increase. The average maturity of public debt is 8.2 years.19 As a result of massive QE however, 45% of the debt is now held by the Bank of Japan, which in turn has issued zero-maturity central bank reserves for the corresponding amount. Thus, the average maturity of the debt of the consolidated Japanese government (Treasury and Bank of Japan) is roughly half of the average maturity of government debt. This is substantially shorter, and exposes the Japanese government to a substantial amount of interest risk.
The next question is thus: Why might interest rates increase and what does it imply?
The first possibility is a sunspot/sudden stop increase: We have just seen that, absent an increase in interest rates, Japanese debt appears sustainable. But, as we discussed in Chapter 4, if investors become worried and require a spread to hold Japan government bonds (JGBs as they are called), then their worries can be self fulfilling. For example, an increase in the real rate from -1.1% to say 2% would require the Japanese government to increase its primary balance by 5.3% of GDP. It might not be able to do so, both because of the catastrophic output effects, and political constraints on how much taxes can be increased or spending can be cut. Can the BOJ eliminate that risk? The answer is probably yes. First, JBGs have a very stable investor base: only 13% of government debt is held by foreign investors. Traditionally, Japanese investors have been more stable. Second, the BOJ is now the major holder of JBGs, and would play the role of a stable investor, unwilling to sell with other investors, and indeed being willing to buy when others sell. Indeed, it may not have to buy on a very large scale, so a commitment by the BOJ to maintain low spreads is credible.
The second possibility is an increase in Japan’s private demand, leading to an increase in r∗ given fiscal policy. Were fiscal policy to remain unchanged, the result would indeed be an increase in r as well, if the BOJ sets the policy rate equal to the neutral rate. But the problem in this case has a natural solution, which follows the conclusions reached in Chapter 5: As aggregate demand increases, the government can reduce its deficit, thus offsetting at least in part the increase in private demand, and thus limiting the increase in r∗ and, in turn the required increase in r.20
Let me again go through a blunt computation. Start from a stable debt ratio, and the assumption that r=r∗=0%. Suppose, that with unchanged fiscal policy, private demand increases by 2% of GDP and, assuming a unit elasticity of private demand to the real interest rate, r∗ increases from 0% to 2%. Suppose now that, in response to private demand, the government reduces the primary deficit by 2%, and the multiplier is equal to 1. The result is no change in r∗ nor in r, and a lower primary deficit. The debt ratio now decreases, other things equal, by 2% each year. One can think of variations on this scenario. Suppose that the government reduces the primary deficit by only 1%, and the BOJ keeps the nominal rate unchanged. Assume that the resulting overheating leads inflation to be higher by 1%. Assuming debt is nominal, the real value of the debt decreases by 1%; the combination of the decrease in the real interest rate by 1% and the decrease in the primary deficit by 1% combine to decrease debt further over time. If the initial debt ratio is 100%, then debt decreases initially by 2% a year.
The third possibility is an increase in aggregate demand in the world, leading to an increase in foreign r∗ and r. If we assume that Japanese financial markets are largely integrated, and if the BOJ does not match the increase in foreign r∗, this is likely to lead to a yen depreciation, and some induced inflation. Again, the likely expansionary effects of the depreciation (and given that Japanese public debt is not in foreign currency) allow for a decrease in deficits while keeping output at potential. And inflation reduces the real value of the debt, and decreases r, leading to more favorable debt dynamics.
The point of this discussion is not to argue that even with high debt levels there is no danger from higher rates. Indeed, reducing deficits, even if they do not lead to lower output, is difficult to achieve politically, and may not happen, leading to questions about debt sustainability.21 But the danger may be less than is usually argued.
What if private demand remains very weak, forcing large deficits and further increase debt ratios?
What may be more worrying than strong private demand and pressure on rates is actually the opposite outcome, namely continued deep private demand weakness, forcing the BOJ to stay at the ELB and requiring deficits so large as to lead to a further steady increase in debt ratios. As the computation above suggests, it would take large deficits to lead to further steady increases in debt ratios, but it cannot be excluded. This forces one to think of other ways to maintain aggregate demand. This in turn takes us to the earlier discussion of what factors might be behind weak private demand and low neutral rates, and whether any of them can be affected by policy.
Focusing first on the saving side: I argued in the previous chapter that increases in social insurance, besides being desirable on their own, can decrease the need for precautionary saving, and thus increase consumption demand. Although social insurance is already high in Japan, there may be room to do more. Prime Minister Kishida has proposed to expand retirement insurance for non-regular workers. This may not boost saving very much but goes in the right direction
Focusing on the investment side: Past public investment has a mixed reputation in Japan, with remembrances of bridges to nowhere. But public investment can take other forms. In particular, as discussed in the last chapter, there may also be a coincidence between the desirability of green investment to fight global warming, and its desirability on macroeconomic grounds. To the extent that public green investment has substantial spillovers to private investment, it may lead to an increase in aggregate demand without the need to run large deficits. And to the extent that investment increases growth in the future (which green investment may not, but other types of public spending, such as better child care and other measures to increase the fertility rate and reverse the population decline might), it also can improve debt dynamics by increasing g relative to r.
Given its high debt level, thinking about other ways than deficits to sustain demand should be a priority for Japan. But it is an important issue for other countries to start exploring as well.
6.3 The Biden bet: r,r∗, and g
Given the delay between the time a manuscript is finished and the book comes out, it is dangerous to discuss current policies and what they may imply for the future. In this last section, I shall take the risk, and discuss what I call the Biden bet, i.e. the very large fiscal package put in place by the Biden administration in early 2021. I shall do so because it shows how to apply the approach developed in the previous chapter, as well as the complexities involved in designing fiscal policy in the current environment.
When, in early March 2020, the dangers from Covid became clear, the Trump administration reacted quickly. It implemented lockdowns and stay-at-home measures, which came however at a high output cost (GDP went down by 31% at an annual rate in the second quarter of 2020). In parallel, it put in place several large fiscal programs throughout the year: The Coronavirus Preparedness and Response Supplemental Appropriations Act in March 2020, for $192 billion; the Paycheck Protection Program and Health Care Enhancement Act in April for $483 billion; the Coronavirus Aid, Relief, and Economic Security Act (CARES) in June 2020, for $2.3 trillion (which included about $1.0 trillion in loans, half of them potentially forgivable). The Fed decreased the policy rate from 1.5% in February to 0.05% by March 2020. The approach to both monetary and fiscal policy was: “Whatever it takes.” It focused on protection of households and firms as the goal rather than on sustaining demand. It led to a large increase in the net debt ratio, from 83% at the end of 2019 to 90% the end of 2020.
During 2020, the situation improved, with fewer lockdowns and confinements. Medical progress on fighting Covid was rapid. Genetic sequencing was achieved early in the year. By March, human evaluations were started. By December, a number of vaccines were approved and the general assumption became that Covid would be largely under control by mid-2021. The question became how to move from protection to recovery, how best to support demand and go back quickly to potential output. This is what I want to focus on here.
Two major programs were put in place. The first was the Impact Aid Coronavirus Relief Act, put in place by the Trump administration in December 2020 for $870 billion, and then the American Rescue Plan (ARP) put in place in March 2021 by the Biden administration for $1,800 billion.
This last program, coming on top of the earlier program by the Trump administration was a major fiscal expansion. Consider the following computation:
Consider first the potential size of the output gap to be filled, as it could be assessed at the time. In January 2020, the unemployment rate had been 3.5 percent, the lowest since 1953; it could reasonably be taken as being close to the natural unemployment rate. Put another way, in January 2020, output was probably very close to potential. The Congressional Budget Office (CBO) estimated potential real output growth for the earlier years to have been around 1.7%. Assuming that potential output continued to grow at the same rate during 2020, and given that actual real GDP in 2020 Q4 was 2.5 percent below its level a year earlier, this CBO estimate implied an output gap in 2020 Q4 of 1.7% + 2.5% = 4.2%, or, in nominal terms, about $900 billion.
Given the supply restrictions due directly or indirectly to COVID-19, it was clear even then that $900 billion was undoubtedly an overestimate of the gap that could be filled by an increase in demand. The pandemic had severely lowered potential output and was likely to continue to do so for at least a good part of at 2021. Supposing, conservatively, that potential output was still down by 1% in 2021 relative to where it would have been absent COVID-19, the output gap that needed to be filled in 2021 by an increase in demand was only $680 billion.
On the demand side, the issue was how much the program would increase aggregate demand. To do so required assumptions about multipliers. Decomposing the ARP program between its different components and using the set of mean values of multipliers given in the 2014 Report of the Council of Economic Advisers (, Chapter 3, Table 3-5) and using them to assess the likely effect of the program on aggregate demand, implied an overall multiplier (the ratio of the increase in aggregate demand to the size of the package) equal to 1.2, and thus an increase in spending of $2.1 trillion, so 3 to 4 times the estimated size of the output gap.22 But the degree of uncertainty was very large: The overall multiplier, under the low multiplier estimates, was 0.4; using the high multiplier, it was nearly 2.0. To the effects of the ARP had to be added the effects of the Trump administration program passed in December for $870 billion, with similar uncertainty about the relevant multipliers. And yet another source of uncertainty came from the fact that, in 2020, as a result of the various fiscal programs, households had saved an estimated $1.6 trillion more than usual; how much of it they would spend was very difficult to assess. In any case, even with conservative assumptions about multipliers and about consumption out of accumulated savings, it appeared that the increase in aggregate demand would vastly exceed the estimate of the output gap, leading to overheating.
On monetary policy, the Fed indicated that they would keep rates “low for long,” waiting for output to be back at potential and inflation to exceed 2% before increasing the policy rate and, meanwhile, they would continue to buy government bonds and mortgage based securities. To cite from the FOMC statement of July 2021:23
The Committee seeks to achieve maximum employment and inflation at the rate of 2 percent over the longer run. With inflation running persistently below this longer-run goal, the Committee will aim to achieve inflation moderately above 2 percent for some time so that inflation averages 2 percent over time and longer-term inflation expectations remain well anchored at 2 percent. The Committee expects to maintain an accommodative stance of monetary policy until these outcomes are achieved. The Committee decided to keep the target range for the federal funds rate at 0 to 1/4 percent and expects it will be appropriate to maintain this target range until labor market conditions have reached levels consistent with the Committee’s assessments of maximum employment and inflation has risen to 2 percent and is on track to moderately exceed 2 percent for some time.
Did this fiscal and monetary policy strategy make sense? (It is not clear that it was a thought-out coordinated “strategy”. It is not clear that the Biden administration intended to achieve substantial overheating, nor that the program was designed in coordination with the Federal Reserve. But we can look at it this way.) In terms of the discussion in Chapter 5, one could think of three options opened to US policy makers at the start of 2020:
A minimal approach: Have a fiscal expansion just sufficient to increase the neutral rate r∗ back to rmin, the ELB rate; equivalently to have a fiscal expansion sufficient to bring output back to potential while keeping the policy rate at the effective lower bound. Anything smaller would be undesirable, leaving a negative output gap. Given the substantial uncertainty, ideally make it partly contingent on the realization of private demand over time to avoid doing too little or too much.
A more ambitious approach: Have a larger fiscal expansion, thus an increase in the neutral rate r∗ above rmin, and have the Fed increase the policy rate r in line with the increase in r∗ to keep output at potential. We saw in Chapter 5 the arguments of such a strategy, namely giving some room for monetary policy to react to adverse shocks were they to happen later. Again, design the program to be partly contingent, so as to adjust it over time if needed.
An even more ambitious approach: Have the same fiscal expansion as in the second option, so an increase in r∗ above rmin, but have the Fed delay the adjustment of the policy rate r to r∗, thus leading to overheating for some time, and a temporary increase in inflation. The potential advantage over the second option, in addition to an implied decrease in the real value of the debt and the effect of a lower real rate for some time on debt dynamics, is to generate an increase in the inflation rate to compensate for an inflation rate below the target in the past, and thus achieve an average inflation rate equal to the target. Again, design the program to be partly contingent.
I believe that the third option is a fair description of the Biden administration and Fed strategy. Proponents argued for the program on four grounds:
First, that the overheating would be limited, either because people would not spend much, or because potential output was higher than the estimate above. Second, that the fiscal expansion following the Global Financial Crisis should have been larger, so the same mistake should not be made again. Third, that given the difficulty of achieving the inflation target in the past, more inflation for some time was desirable. Fourth, that even if there were overheating, the Phillips curve relation between inflation and unemployment was very flat, and thus there would be little inflation.
Others, including me, pointed to risks, agreeing about the direction of fiscal policy, but worried about its magnitude.24 The worry was that private demand could be quite strong on its own, in particular because of the accumulation of excess savings during the pandemic. Another worry was that the program was heavily front-loaded, and could not be reduced in size if demand turned out to be too strong. The risk therefore was that there would be substantial overheating, and that the increase in inflation would be much larger than implied by the historical Phillips curve relation, forcing the Fed to react and increase interest rates more than was intended, at least for some time.
The story is still happening. Nine months later, at the time this is written, where do things stand?
Potential output has turned out to be lower than forecast. Many workers have not returned to work, and the participation rate, at 61.9% in December 2021, up from 60.2% in April 2020, but down from 63.4% in January 2020. Demand has been strong, with an increase in the demand for goods relative to services. The unemployment rate has come down dramatically, from 14.8% in April 2020 to 3.9% in December 2021. A high ratio of vacancies to unemployment, and a high quit rate, both indicate that the labor market is tight. Strong world demand, partly due to U.S. demand, has led to a large increase in commodity prices, increasing costs of production. Supply chain disruptions, partly due to Covid, but mostly due to high demand, have led to shortages and higher prices.
As a result, inflation has increased substantially more than proponents of the program forecast. As shown in Figure 6.4, CPI inflation (measured as quarter over quarter of the previous year) exceeded 5% in 2021 Q3; much of it is due to the increase in commodity prices; in 2021 Q3, a global commodity price index for the United States was more than 60% higher than a year earlier. And the tight labor market has led to an increase in wage inflation, although so far the increase remains moderate.
What happens next? One can think of two scenarios (assuming that there are no major new disruptions due to Covid or other factors):
The first is that participation rates return at least to their pre-Covid levels, that supply chain disruptions disappear, that commodity prices return to lower levels, that private demand weakens, decreasing pressure in the labor market; that productivity growth turns out to be high, reflecting changes in the organization of firms during the crisis, and limiting the effects of wage increases back on prices. If so, inflation may naturally return closer to 2%, without the Federal Reserve having to increase the policy rate substantially. Real rates remain low.25
The second is that these conditions are not met. Participation rates remain lower, because for example, workers who retired early do not go back to the labor market; supply chain disruptions disappear only slowly; private demand remains strong, as households spend their accumulated excess savings; wages continue to rise faster, due to the tight labor market strengthening the bargaining power of workers, together with their desire to compensate for the increase in price inflation. In this case, inflation remains high, the Federal Reserve has to increase the policy rate more than it intended, and real rates may be substantially higher for some period of time. This is what I referred to as a potential temporary increase in rates due to the Biden stimulus in the conclusion to Chapter 3: Even if fundamentals suggest that real rates will remain low over the medium run, they may be much higher for a while, due to fiscal policy. I believe this scenario to be more likely than the first.
My purpose in this section was not to give forecasts, but to think about the effects of the stimulus program in light of the discussion of fiscal policy in the book, namely the combination of a fiscal expansion to increase demand and by implication increase r∗, so as to increase the room for monetary policy, together with a delayed adjustment of r to r∗ on the part of the Federal Reserve to allow for some overheating and a temporary increase in inflation. My conclusion is that, while the intent of the strategy (if indeed it was a strategy) was the right one, the size of the program was too large, leading potentially to a difficult adjustment.26
This discussion omits the possibility for r* to increase due to adverse supply shocks, which would imply different trade-offs. If there are reasons for this omission, I would suggest exposing them.
“boost” should be “reduce”
“Prime Minister Kishida has proposed to expand retirement insurance for non-regular workers. This may not boost saving very much but goes in the right direction“boost” should be reduce.
“the 46 billion allocated in the ARRA package passed in the United States in 2009 led to more than 150 billion in private and non federal private investment” should be “the 46 billion allocated in the ARRA package passed in the United States in 2009 led to more than 150 billion in private and non federal public investment
I guess these should be in line. I will try.
Masaya Sakuragawa : Your evaluation for Japan fiscal reform looks a bit optimistic, I guess. In practice in 2010s, risk-free rate is lower than growth rate by about 1-2 percent in Japan, but this arose at the chance of QE since 2013. Before that, interest rate was higher about 1-2 percent., meaning that questionable is if r<g of Japan is structural. Keeping a target for fiscal surpluses is a safe policy, I guess.
But inflation needs to be more like 4 per cent, you point out above.
There seems to be a large unused capacity to issue long-dated or even perpetual bonds. Discussion of this would be useful