As COVID-19 cases surge across the country, U.S. policymakers are struggling to pass a pandemic relief package before emergency stimulus programs expire by the end of the year.

With negotiations ongoing, signs have emerged on Capitol Hill of a thawing of tensions, but the two sides remain far apart. All eyes are now on Dec. 18, the new likely date when Congress needs to pass another continuing resolution--where policymakers have signaled aid provisions may be included--to keep the U.S. government open.

The overall impact of pandemic relief on the economic recovery boils down to current economic conditions and projected conditions through the end of this year and next. The headline question is the size of the package. But how it's spent will ultimately dictate its effectiveness in guiding the U.S. economy back to health. Thus, the most important question for Congress may be: How much economic activity does one fiscal dollar generate, or, to put it another way, what is the bang for the buck?

In this analysis, employing short-term demand multipliers from the Congressional Budget Office (CBO), modified by S&P Global Economics' U.S. social distancing assumptions, we came up with some initial answers to this question. In our most recent U.S. economic forecast report, we forecast the economy would gradually reach fourth-quarter 2019 GDP levels by third-quarter 2021, assuming a $1 trillion stimulus package is in place at the start of next year--a faster recovery than in our September forecast, which reached pre-pandemic levels by fourth-quarter 2021, assuming a smaller $500 billion package.

Not surprisingly, we then found that an even bigger, largely demand-driven stimulus package of $1.5 trillion in place by the start of next year would provide even more juice to the recovery, with the U.S. reaching pre-pandemic levels by second-quarter 2021. Meanwhile, should Congress fail to produce a package altogether, the U.S. economy would continue to suffer over the short run, not reaching pre-pandemic levels until first-quarter 2022, almost a year later than in our $1.5 trillion scenario.

The temporary--and primarily demand-preserving--fiscal stimulus measures currently discussed in Congress include tax rebates (in the form of one-time stimulus checks to households), extended unemployment insurance (with a booster), support to small businesses, and transfers to state and local governments (to avoid drastic budgetary spending cuts). The direct effects of such temporary fiscal measures (shocks) are generally limited to the duration of the measures themselves, with the persistence of the fiscal multipliers decaying in subsequent periods(1).

Under normal conditions, for every dollar spent, the economic activity generated would be under a dollar to varying degrees, with some policies closer to break-even. It depends on where the U.S. economy is in the cycle. Fiscal multipliers are generally higher in a weak economy with interest rates low. The table below shows the range of fiscal multipliers used in our calculations for each policy we analyzed in this report. We directly rely on the CBO fiscal multipliers for aid to small businesses. The others are in line with the CBO's estimates in September 2020 (see the appendix for more on our methodologies) (2).

The Keynesian approach being discussed in Congress may be just the short-term relief needed to provide the bridge to recovery. But we emphasize that not all government spending is created equal and that these short-term policy fixes are by design temporary and don't pay for themselves. They can't bring the economy to a higher sustainable growth path. They will accelerate "filling the demand hole" but will not change the longer-run growth rate of the economy, which is a function of growth in productivity and the labor force.

Other fiscal policies, if chosen wisely, may be a better solution. Investments in physical infrastructure and human capital (public health and education), for example, are such long-term investments that they pay for themselves--and some more--and are urgently needed after decades-long underinvestment.

Where Is The Recovery Headed?

The U.S. economy has found bottom and is on the mend. However, we expect the climb out of the hole to be slow and painful, with risks along the way. We expect U.S. economic growth to decline by 2.3% (quarter over quarter annualized) in the fourth quarter. On its own, a one-quarter decline (driven by an assumption of a pullback in December consumer spending due to coronavirus restrictions) does not signal recession. But it increases chances that the U.S. will see another downturn soon (see "Economic Research: Staying Home For The Holidays," published Dec. 2, 2020).

S&P Global Economics expects real GDP to contract 3.9% for this year and grow a modest 4.2% in 2021. All of this assumes passage of a $1 trillion stimulus package before year-end. Even with that boost, by the fourth quarter of 2023, real GDP would still be $96 billion (or 1.9%) smaller than what we expected in our December 2019 forecast. Adding to the pain is that the U.S. unemployment rate is unlikely to reach its pre-pandemic low until after 2023. This will give the Fed good reason to keep interest rates near zero until mid-2024, as we expect.

We continue to see the risk of recession over the next 12 months as near the top of the 25%-30% range. The U.S. economy, which has recouped two-thirds of the economic losses from the COVID-19-induced recession, is now showing signs of weakness this holiday season amid climbing COVID-19 infection rates.

As the recovery unfolds, the big risks have increased: No coronavirus vaccine is widely available in the winter flu season as people head indoors, and new fiscal stimulus is lacking with more pandemic-related stimulus expiring at year-end. With no agreement in the making, the worry now is that there will be no stimulus until next year, if at all. Moreover, indecision among policymakers will influence private-sector expectations and spending plans. Households, for example, may delay spending on worries that the modest recovery may collapse. Increasing trade tensions with China only make matters worse.

The Coronavirus Aid, Relief, and Economic Security (CARES) Act has helped support struggling families through large fiscal transfers to households, such as stimulus payments, enhanced unemployment insurance, and moratoriums on mortgages. But the doors are closing. Already extended boosters to jobless benefits have also expired, and with the majority of out-of-work Americans unemployed for at least 27 weeks, their 13-week extensions to traditional benefits will also expire soon. Moreover, further job cuts from state and local governments, whose budgets have been crushed by COVID-19, will add pressure if federal stimulus isn't available. They would need to tighten their belts further, increasing chances of more jobs lost in December.

It's not a far-fetched possibility the U.S. economy could see no more fiscal stimulus and a COVID-19 resurgence that cripples growth in the fourth quarter. In our downside scenario, GDP would drop by 4.4% in 2020 and rise by only 0.8% in 2021.

S&P Global Ratings believes there remains a high degree of uncertainty about the evolution of the coronavirus pandemic. While the early approval of a number of vaccines is a positive development, countries' approval of vaccines is merely the first step toward a return to social and economic normality; equally critical is the widespread availability of effective immunization, which could come by mid-2021. We use this assumption in assessing the economic and credit implications associated with the pandemic (see our research here: www.spglobal.com/ratings). As the situation evolves, we will update our assumptions and estimates accordingly.

Short-Term Fiscal Stimulus: How Long A Bridge To The Other Side?

Our analysis considers several possible scenarios, incorporating the size of a hypothetical short-term stimulus and the size of the fiscal multiplier. We focus on three possible stimulus scenarios: a $1 trillion package, $1.5 trillion package, and no stimulus.

We assumed short-term stimulus would be divided equally across programs that are currently being discussed on Capitol Hill: tax rebates, extended unemployment insurance, support to small businesses, and transfers to state and local governments.

Chart 1

Our $1 trillion scenario sees the biggest bang in first-quarter 2021

A $1 trillion stimulus package (incorporated in our December baseline forecast) would bring the economy back to the pre-COVID-19 level in third-quarter 2021. We assume equal fiscal spending of $250 billion on rebates, unemployment insurance, state and local governments, and small businesses. First-quarter 2021 would show the biggest bang for the buck, as we assume households would receive their tax rebate checks in this quarter, and literature suggests they tend to spend nearly 40% of the money received in this quarter, assuming increased social distancing(3). Unemployment insurance payments would provide more persistent aid to the economy, since we assume they would be gradually distributed in the next two years while the unemployment rate stays high.

We assume households, businesses, and state and local governments would receive help, with about three-fifths of the $1 trillion coming in 2021 and two-fifths over 2022 and 2023. Assuming a multiplier effect in line with Brookings, Hamilton, and CBO estimates, the weighted average impact would be $770 billion over our forecast horizon of 2021-2023, adding about 2.5 percentage points to growth over the three-year period(4).

Even with a $1 trillion package, the new pace of growth has shifted lower, well below the pre-pandemic path of expansion that we expected in December 2019 (dotted black line in chart 1), highlighting the permanent loss from COVID-19.

If a 2.5 percentage point boost to economic activity through deficit-financed government transfers were to occur, it would still not be enough to fill the estimated 3.6% negative output gap (as of third-quarter 2020). Perhaps that is why the Fed is not likely to raise rates until 2024. Based on our analysis, the output gap will finally get to its historic average of negative 0.2% sometime in 2024.

In our analysis, we then used the highest possible fiscal multiplier estimates from the CBO. Fiscal multipliers are the largest when the economy is weak and the Fed keeps the policy rate at the zero lower bound (as has been the case for the U.S. in recent years with unemployment high, inflation low, and interest rates near zero). Here, the economic impact from a $1 trillion package is a bit stronger (pink line in chart 1), with the U.S. economy closing in on its pre-pandemic GDP level by second-quarter 2021, one quarter sooner than when we assume average fiscal multipliers.

However, the demand-driven stimulus boost would be short-lived, while the economy would recover faster than without stimulus; with no boost to productivity, the pace of growth would decelerate back to GDP levels nearer to its potential growth rate by fourth-quarter 2023 in both stimulus packages.

Our $1.5 trillion scenario leads to higher growth for longer

In another scenario, lawmakers agree to an even larger stimulus package, to the tune of $1.5 trillion (compared with $1 trillion in our baseline). This helps sustain a higher growth path for a longer time compared with our baseline.

In this hypothetical $1.5 trillion stimulus scenario, assuming average fiscal multipliers, the U.S. economy rebounds at a quicker pace of around 5% in 2021 (black solid line in chart 1). GDP remains higher than in our $1 trillion stimulus scenario. But as the stimulus wears off, the difference between the two scenarios shrinks, with GDP levels for the $1.5 trillion scenario moderately higher in fourth-quarter 2023 than for the $1 trillion scenario.

The $1.5 trillion stimulus scenario creates a bigger boost in economic activity than the other scenarios, holding near its pre-pandemic expansion path through 2022. However, the new pace of growth for the $1.5 trillion package also eventually shifts lower, toward the economy's lower potential growth (trend) rate and well below the pre-pandemic path of expansion that we expected in December 2019 (dotted black line in chart 1).

We also used the highest possible fiscal multiplier estimates from the CBO. Here, the economic impact from a $1.5 trillion package is even stronger, with the U.S. economy reaching pre-pandemic economic activity earlier in second-quarter 2021. Moreover, economic activity in dollar terms for the $1.5 trillion stimulus scenario with high fiscal multipliers catches up to pre-pandemic GDP levels by fourth-quarter 2021 and remains near that level for one year.

As with the other hypothetical scenarios, economic activity in the $1.5 trillion scenario with high fiscal multipliers still drifts down closer to its potential GDP growth rate as the temporary, demand-driven stimulus unwinds.

The Multiplier Impact By Policy

The impact of the various stimulus policies depends on how large the federal aid but also how quickly the money is spent (the marginal propensity to consume, or MPC). For our illustrative purposes, we keep the model simple by using the CBO's recent MPC estimates for individual stimulus channels, but with smaller adjustments for social distancing (since we do not envision the country going back to the strictness experienced in April). The two charts below show the boost from different stimulus policies on the recovery.

Lower-income households have larger MPCs than higher-income households, and thus, policies that benefit the lower-income group--unemployment benefits, in this case--show the highest MPCs. They are also longer lasting by design, given the expectations of labor market recovery.

On the other hand, one-time rebate checks have slightly smaller MPCs, and the bulk of the effect is front-loaded due to their one-time nature. Earlier in the year, tax rebates went out to households at a relatively generous income threshold level. For example, married couples filing jointly at an income threshold below $150,000 received the full amount. After that threshold, the stimulus checks became lower. Past the $198,000 threshold, a household would not get a paycheck--which meant MPCs were lower on average compared with the pool receiving unemployment benefits.

State and local governments were modeled according to historical experience, given both the delivery of the aid and the speed with which it is spent are slower than for the other stimulus channels. Small businesses have the smallest impact because the stimulus is likely to be used in balance-sheet repair to avoid shutdowns.

Regardless of a short-term bump in the economy, in both cases, the economy decelerates to trend as the demand-driven boost to economic activity is temporary, returning to the potential growth rate by 2023.

Chart 2

Chart 3

Bridging The Gap: An Investment And A Solution

The emergency pandemic stimulus is meant to extend the bridge so that the U.S. economy can get to the other side of the recovery. The stimulus would help stabilize the health of the American people and the American economy.

But more may be needed to bring the economy back to full health.

In just two quarters in 2020, U.S. economic activity shrank by 10.2%, losing $1.95 trillion of GDP in real terms--two times the loss of the Great Recession, but in one-third the time. The U.S. GDP will be smaller, in dollar terms, on average in 2020 than in 2019. We now expect U.S. labor productivity gains to eke out just 0.5% on average in 2021-2023 in the aftermath of the sudden-stop recession. That's half the already meager 1.0% labor productivity gain that we had in 2019.

Chart 4

This does not have to be the case--and S&P Global Economics calculated in an earlier analysis that a revival of public infrastructure spending relative to GDP, to levels seen in the mid-20th century, could provide the salve that this injured economy so desperately needs. We see infrastructure investment as one way to get the U.S. back on track once COVID-19 makes its exit, and it may even help stave off the virus' next attack.

In our earlier report, we found that if the U.S. invested $2.1 trillion into public infrastructure spending over a 10-year horizon, it could create 2.3 million jobs by 2024 as the work is being completed, helping those millions of unemployed workers displaced by COVID-19 (see "Infrastructure: What Once Was Lost Can Now Be Found--The Productivity Boost," published May 6, 2020).

The infrastructure investment would, over the long run, provide the productivity boost needed to help get the expansion back on track, essentially steepening the slope of the curve, with the "infrastructure" scenario catching up to the pre-virus expansion at a faster rate than the "no infrastructure" scenario.

Moreover, the productivity boost from the infrastructure investment could add as much as $5.7 trillion to the U.S. over the next decade (see chart 5). The additional 0.3% boost to productivity per year would generate a net 713,000 jobs on the books by 2029. Wallets would also be fatter, with per capita personal income $2,400 larger than in the "no infrastructure" scenario.

Chart 5

There would be an initial lift-off over the first five years, with infrastructure spending as a share of GDP reaching levels seen during the mid-20th century. In our scenario, spending would then slow to a higher steady state of about 1.7%, more than the historically low share of 1.3% in 2018 (half of what it was 60 years ago).

This scenario is based on the 2019 agreement between the White House and Democratic lawmakers on a $2 trillion infrastructure project that never came to fruition. President-elect Joe Biden has proposed a similar plan, also for $2 trillion. However, Biden is open to infrastructure projects that include climate change and public health, in addition to the usual (and needed) bridges and roads.

Public infrastructure investment could also "crowd in" or encourage private investment, with every dollar spent having a multiplier effect(5). The argument is that public capital enhances the productivity of private capital, raising its rate of return and encouraging more investment. If a public infrastructure project was wisely devised, the multiplier from the investment would be larger than the money spent.

For evidence of this, look no further than former President Dwight D. Eisenhower's Interstate Highway System. Costing about $500 billion in today's dollars, it has clearly paid for itself, given all the products and people that travel on its 48,000 miles of roads on any given day. Ike's project is estimated to have a multiplier equal to 6--or, for every dollar spent, the U.S. got $6 back.

COVID-19 has created an urgency to invest in much-needed public health infrastructure. But the story doesn't end there. Six months from now, we may also hear news about a collapsed bridge that had been in disrepair for years, or we could see another Superstorm Sandy, which called attention to the need for investing in infrastructure to prevent damage from climate change. Let's hope the U.S. will respond.

By prioritizing infrastructure once again, we could again take pride in investing not only in the physical health of our nation, but also its economic health, creating more jobs today and, with the boost to productivity, more economic activity and jobs in the future.

We know a possible infrastructure package is in the concept stages--shovels won't hit the ground anytime soon. But the post-pandemic recovery is also expected to be slow. Moreover, many planned projects have stalled with shovels left in the ground. Given that we currently expect the unemployment rate to be above full-employment levels until fourth-quarter 2023, those extra jobs coming sooner would likely be a welcome relief.

With the U.S. only slowly climbing out of a hole much worse than the Great Recession, the returns on investment from the infrastructure boost would be likely higher over the near term, given costs are lower in the now-soft jobs market. And assuming the infrastructure spending was prudent, the productivity gains later in the decade would generate a bigger boost to economic activity, with the multiplier averaging 2.7 over the 10-year period, helping to give the post-pandemic expansion more fuel as it nears 2030. This translates to $2.70 back for every dollar spent, much larger than average multiplier estimates of a still-high 1.3-1.8 range for infrastructure spending, or the 0-0.4 multiplier for the corporate tax provision(6).

A 2012 San Francisco Federal Reserve report studied the effect of unexpected highway grants on state GDPs (GSPs) back to 1990 and found that, on average, each dollar of infrastructure spending increases the GSP by at least $2, a multiplier of 2.0(7). Moreover, the report found that spending in 2009 and 2010 was roughly four times more than average, indicating that it can be highly effective during periods of "very high economic slack, particularly when spending is structured to reduce the usual implementation lags."

Appendix

We followed the CBO's approach to analyze the impact of fiscal stimulus on the economy. In this appendix, we detail assumptions we made in the calculations.

We calculated the direct and indirect effects of the potential stimulus package on output as measured by real GDP (change in "Y"). This output multiplier measures the change in total economic output generated by each dollar of budgetary cost of a change in fiscal policy (change in "d"). The direct effects refer to how much and when the stimulus recipients want to spend the money received. The indirect effects--or demand multiplier--capture how other participants in the economy enhance or offset the direct effect.

The combined effect of a change in fiscal policy is simply the product of the direct and indirect effect and is referred to as the output multiplier or fiscal multiplier on output. According to the CBO, when the economy is weak, if stimulus recipients spend one dollar, it could affect the economy in the following eight quarters, with a cumulative fiscal multiplier ranging from 0.5 to 2.5(8).

An adjustment factor for social distancing is also applied on both direct and indirect effects. The first adjustment factor, denoted by the alpha on the marginal propensity to consume (MPC) and M (the demand multiplier under normal conditions), reduces the direct and indirect effects on demand from a policy change at time "t". When the two adjustment factors are incorporated into the definition in this equation, the output multiplier under social distancing can be expressed as:

The power of fiscal stimulus depends on how much and when stimulus recipients want to spend the money received. When analyzing the behavior of households when they receive tax rebates and unemployment insurance payments, we assume households tend to spend 73% of the money received. This assumption is based on evidence from 2008 tax rebates (9) and the latest analyses of households' spending behavior after they received unemployment insurance payments in 2020(10).

In addition, we account for the impact of social distancing on households' spending patterns. We assume social distancing would no longer be an issue in first-quarter 2022(11). From first-quarter 2021 to first-quarter 2022, the degree of social distancing would gradually abate. Due to the inability to go out in 2021, households' near-term spending would be dialed down. We assume they would gradually spend 60% of what would have been consumed in the first half of 2021 after the economy fully reopens(12).

We assume state and local governments would gradually spend 72% of the federal aid over the next three years. Relevant studies show that what state and local governments have received is likely to surpass their expected revenue shortfalls(13). However, they may face additional budget pressures in the following two years as the economy remains weak. Therefore, it is likely state and local governments would still make use of the aid in the following years.

A similar study by Brookings assumes state and local governments would gradually spend all of the aid from the federal government from 2021 to 2023(14). We made similar assumptions in this report but assume a more conservative MPC, at 73%, which is the midpoint estimate made by the CBO(15).

The impact of support to small businesses is harder to analyze, since business owners don't directly spend the money received. In addition, the aid to small businesses is aimed at increasing the survival probability of those businesses, working on the supply side of the economy rather than the demand side, as is the case with households and state and local governments. Therefore, in our analysis of the stimulus' impact on small businesses, we used the CBO's estimate that the cumulative effect on GDP from one-dollar spending on the Paycheck Protection Program is 0.36 dollar from 2020 to 2023(16).We assume three-fourths of the cumulative effect would take place from 2021 to 2023.

Notes

1) It is generally held that forward-looking agents are not affected by temporary changes in their disposable income, while credit-constrained ones are only affected over the duration of the shock. The persistence of multiplier effects (indirect effects) also declines steadily past the duration of the shock. By contrast, the effect of permanent fiscal shock is more persistent.

2) "The Effects of Pandemic-Related Legislation on Output." CBO.

3) Consumers tend to spend 50% of the money received from tax rebates. Due to social distancing, we adjusted it down to 37.5%. Parker, Souleles, Johnson, McClelland. "Consumer Spending and the Economic Stimulus Payments of 2008." American Economic Review 103(6): 2530-2553.

4) See "The Effects of Pandemic-Related Legislation on Output," CBO, Seliski, Betz, et. al, "Key Methods That CBO Used to Estimate the Effects of Pandemic-Related Legislation on Output." CBO. "What could additional fiscal policy do for the economy in the next three years?" The Brookings Institution, Oct. 9, 2020.

5) An example of "crowding in" can be seen in the Isaksson, Anders, Working paper 15/2009, "Public Capital, Infrastructure and Industrial Development." Research and Statistics Branch Programme Coordination and Field Operations Division, UNIDO.

6) Studies on multipliers include: Bivens, Josh, EPI Briefing Paper #374, "The Short- And Long-Term Impact of Infrastructure Investments on Employment and Economic Activity in the U.S. Economy," July 2014; CBO, "Estimated Impact of the American Recovery and Reinvestment Act on Employment and Economic Output from April 2011 Through June 2011," 2011; Council of Economic Advisers (CEA), "The Economic Impact of the American Recovery and Reinvestment Act, 2009 Seventh Quarterly Report. Washington, D.C.: Executive Office of the President"; CBO, "Estimated Impact of the American Recovery and Reinvestment Act on Employment and Economic Output from October 2011 Through December 2011," February 2012.

7) Leduc, Sylvain, Wilson, Daniel. "Highway Grants: Roads to Prosperity." FRBSF Economic Letter 2012-35.

8) Whalen, Reichling. Working Paper 2015-02, "The Fiscal Multiplier and Economic Policy Analysis in the United States." CBO.

9) Parker, Souleles, Johnson, McClelland. "Consumer Spending and the Economic Stimulus Payments of 2008." American Economic Review 103(6): 2530-2553.

10) Farrell, Ganong, Greig, Liebeskind, Noel, Vavra. "Consumption Effects of Unemployment Insurance during the COVID-19 Pandemic." JPMorgan Chase Institute.

11) "The Effects of Pandemic-Related Legislation on Output." CBO.

12) Seliski, Betz, et.al "Key Methods That CBO Used to Estimate the Effects of Pandemic-Related Legislation on Output." CBO.

13) "How Much Would Each State Receive Through the Coronavirus State Fiscal Relief Fund in the Heroes Act?" Center on Budget and Policy Priorities.

14) Edelberg, Sheiner. "What could additional fiscal policy do for the economy in the next three years?" Brookings.

15) Seliski, Betz, et. al. "Key Methods That CBO Used to Estimate the Effects of Pandemic-Related Legislation on Output." CBO.

16) "The Effects of Pandemic-Related Legislation on Output." CBO.