The COVID-19 Recession: Unprecedented Collapse and the Need for Macro Policy

In the past week, it has quickly become clear that the social distancing required to mitigate the spread of Covid-19 will create a severe economic fallout. This short essay begins by exploring the severity of the economic challenges the U.S. economy faces. I then analyze how policies can be best structured to contain the damage to households and businesses and to facilitate a quick recovery.

It is now very likely there will be a large federal fiscal response to the crisis. Votes in Congress on the package are imminent, but as of this writing (late afternoon of March 25, 2020) details remain sketchy. Policies will undoubtedly evolve.

Unprecedented Macroeconomic Conditions

Large parts of the United States economy are almost entirely shutting down. I made estimates of the direct effects on several key industries on March 18 was shocked to learn real GDP would likely plummet by at least 5% to 10% from mid-March to mid-April. This estimate is based just on a few service industries; it does not include any multiplier effects or spillovers to manufacturing, business and residential investment, or international trade. (See appendix 1 at the end of this most for details.)

While enormous uncertainty remains, these estimates provide sobering perspective. An output collapse of this magnitude over a month or two is unprecedented in U.S. economic history going back to 1950. Most economists consider the “Great Recession” of 2007-09 the worst contraction since the 1930s, but in that crisis output fell “only” 4% , and that decline took place much more slowly, over about 18 months.

Because of the way quarterly data are annualized in the official GDP figures, the headline quarterly growth number for the second quarter of 2020 could easily be as bad as -25%.[1] The annualized decline during worst quarter of the Great Recession (2008 Q4) was 8.4%. Even if we are lucky and the decline is half as large as my figures imply, the second quarter of 2020 seems as if it will look like the worst quarter in modern economic data going back to the late 1940s.

Other factors could make things even worse. To name just a few, these numbers do not include “multiplier” effects that arise when workers lose jobs and cut back on spending across all parts of the economy. Reduced business capital investment will magnify the size of the contraction. Manufacturing will also decline. The drop in production of goods could be mitigated relative to lower sales if firms expand their inventories. But higher inventories in the next few months will likely slow any recovery of manufacturing output.[2]

When I first did these calculations, I was stunned, wondering if I had made some kind of basic error to project such a severe collapse. But in the past days, other forecasters have put out similarly bleak figures.

The data look scary. But a silver lining to the economic cloud could be a quick recovery. We have reason to hope the health threats of the virus will pass in a few months. Data from China on new infections look encouraging. Even if the virus infections linger or a second wave of infections emerges in the fall, it is unlikely that severe social distancing will continue at the same level as late March and April. At some point, probably after the greatest strain on the health care system abates, people and their governments will likely conclude life for the vast majority with mild or no symptoms should return to more normal conditions and the virus will need to run its course, hopefully creating herd immunity in the global population.

When things calm down, Americans will want to return to normal life and demand for social activity will return. Restaurant visits and entertainment activities will resume. Pent up demand for durable goods purchases and vacations postponed during the crisis could surge. We can hope the recovery proceeds quickly (although it is still likely to be somewhat slower than the unprecedented collapse in March and April).

But this optimistic outcome is not guaranteed. There is a chance the economic effects of Covid-19 will trigger further contraction that will persist even as the health crisis recedes. A recession of this magnitude can severely compromise household and business finances. The result could be that households do not have the means to return to normal economic activity and businesses may have failed and so cannot fulfill rising demand.

Effective and quick federal policy response is critical to create conditions for a quick recovery. As of this writing, government leaders are coming close to passing a large package of measures to combat economic damage. Let us consider three key objectives this policy package needs to achieve.

Address Needs of the Medical Infrastructure, And Do it Now!

First and foremost, the federal government must move aggressively to expand health care resources necessary to treat Covid-19. This includes rapidly ramping up production of protective gear for health care workers, surge hospital capacity, and many new ventilators to treat severe respiratory distress, among other things identified by public health experts. This is not a time for “wait-and-see” responses. Health care infrastructure must rise as fast as possible all over the country, regardless of whether some locations have yet to experience “hot spot” strains on existing resources.[3] Government put its head in the sand about known public health risks prior to the crisis. It must now do everything it can to catch up. In addition, the federal government must assure costs for treatment of any respiratory illness during this crisis are covered for uninsured patients.

The legislation now working now working its way toward passage provides extra federal funding for the health care system. Hopefully, it will be adequate.

Critical Support for Unemployed Workers

Second, economic multiplier effects must be limited. To prevent the collapse in further spending by workers who lose their jobs due to the direct effects of social distancing, the most important need is to shore up unemployment insurance benefits. This policy both assists those who suffer the worst consequences of the crisis and cuts off multiplier effects that magnify economic contraction.

Standard state-level unemployment insurance programs are, in most cases, woefully inadequate to this task. In my state of Missouri, unemployment benefits replace just over 50% of wages and benefits are capped so that there is no replacement at all for monthly wages above $2,667 ($32,000 at an annual rate).[4] At replacement rates like these, workers unemployed, obviously through no fault of their own, will be severely challenged to pay rent, service debt, and purchase basic needs; they certainly will have no budget for discretionary expenditure. This kind of financial stress will magnify output losses during the crisis and greatly hinder recovery in its aftermath as households suffering unemployment struggle to regain their financial footing.

Enhanced unemployment benefits should cover 100% of lost wages up to a modest level of income. In addition, workers who earned more than the cutoff for 100% benefits should receive a substantial share of wages up to a reasonably high upper cap. For illustrative purposes, appendix 2 provides a rough estimate of the monthly cost of a program like this with wages fully replaced up to $2,500 per month and a 50% replacement rate for the amount of wages from $2,500 per month to about $6,700 per month.[5] Suppose the monthly unemployment rate hits at a very high level of 20%. In that case, I estimate the cost of this program at about $108 billion per month. We may need a program like this for several months, but most likely the unemployment rate will not be as high as 20% over much of that time. This cost is easily accommodated within the magnitudes of federal fiscal responses being discussed by Congress and the White House.[6]

As of this writing, it appears Congress will vote on expanding unemployment benefits for up to $600 per week above state levels, a significant step. But in Missouri this maximum amount would support the proposal in the previous paragraph for just an estimated 60% of the affected population. We do not yet know how this benefit will be allocated, especially how it will replace wages for the lowest part of the income distribution where the case for a 100% replacement rate, or close to it, is compelling. Furthermore, it is not clear how the legislation will handle the massive increase in funding necessary from the states just to provide the often meager standard benefits to the explosion of workers who need support.

Extra cash in American pockets from the proposed stimulus checks sent to every household will also help maintain overall spending, especially when social distancing requirements ease. I support payments to every household, in part to create broad political support for the full policy package. But these payments must not be a substitute for much enhanced unemployment benefits. For households that keep their labor income during the crisis, extra income may have little effect immediately because there are few channels for discretionary spending available. Payments to all households may help speed recovery after social distancing eases. Although the temporary structure of these payments will reduce their effect on demand. The need for greatly enhanced unemployment coverage is both more socially just and more effective at limiting economic damage. Based on what we know about the plan coming out from the government, it appears unemployment support is low compared to unfocused payments to households. If the overall cost of the package imposes a limit on what can be done, the large expenditure on checks to all households at the cost of less support for unemployment benefits is not an effective response.

Keep American Businesses Viable with Working Capital Loans

Third, businesses need significant, but well designed, support. The financial damage in this crisis is much more concentrated than in a typical recession. Without cash flow for two months or more, many businesses will not survive to participate in the hoped-for quick recovery. Americans cannot go out to restaurants in big numbers this coming summer and fall if a big share of the restaurants are permanently shut down! The problem is sizeable and requires a sizeable response from the federal government. Business needs are, in one sense, parallel to the problem of labor unemployment. Indeed, supporters of financial support for business have called it a response to the “unemployment of capital.” But the solution is more complicated. Machines, assembly lines, delivery trucks, and office buildings do not have to be fed, clothed, and provided with medical care. That said, there are rent, maintenance, and debt service costs. If firms cannot cover these fixed costs, they may fail and be unable to restart.

One group that deserves special attention is small business owners of modest means. Many small business owners have middle-class incomes. “Gig” workers who might be considered self-employed likely have lower incomes on average. The kind of enhanced unemployment benefits described previously must be available to self-employed people who run their own businesses. There should also be appropriately adjusted benefits for individuals who credibly demonstrate they have lost part, but not all, of their income.

To remain viable through the worst months of the crisis, businesses need access to working capital to take care of short-term expenses that do not go away as production shuts down. This kind of funding is best handled with bridge loans with easy terms and a reasonably long repayment schedule. Perhaps there should be some interest rate charged on these loans to discourage businesses who do not really need the funds from taking “free money” and clogging up the system. It would be reasonable to impose some underwriting requirements to assure that the borrower was a viable business before the crisis. But access to bridge loans should be easy and efficient. There does not seem to be any need to decide a priori what kinds of businesses qualify for loans or how much different kinds of firms or industries can borrow. While administrative practices might differ by firm size, in principle this kind of program could provide from a few thousand dollars to a restaurant to pay its rent up to billions of dollars to an airline to keep its grounded planes in condition to start flying again when travel restrictions ease.

The policy package now under discussion provides loans, and possibly some grants, for businesses small and large. It remains unclear at this time how this facility will operate. There is some suggestion that there will be specific allocations of funds to particular industries (even particular firms). No doubt, there is unevenness in the effect of the crisis across industries. But it seems impossible to decide at any level of government how much different sectors need or which businesses are more “deserving” than others. A decentralized system of easy credit access for all who need it seems like a better solution. In particular, authorizing the Federal Reserve to implement such a program through the commercial banking system and funded by increases in the Fed’s balance sheet (that is, effectively “printing money”) could be a good approach. Remember, these are loans, not grants. There will be some default, but there will also be some positive interest charges. Any effect on the long-term monetary base will likely be small.

Although I am fully behind the need to support workers whose incomes are decimated by the crisis, I am not enthusiastic about policies that force businesses to keep employees on the job in order for them to borrow from the bridge loan program.[7] It does not make much sense for businesses to keep people working if the productivity of those workers does not justify their employment. This is a central reason that greatly enhanced unemployment benefits are so critically important.

There is one part of labor costs to firms, however, that should be retained even if the workers are temporarily laid off. Because the U.S. connects so much of health insurance to employment, it makes sense to subsidize health insurance costs for laid off employees rather than deal with the disruption that might arise if all unemployed workers must receive health coverage through a program like Medicaid.

Providing unemployment grants to workers, but loans with interest to business, may seem asymmetric. Should we support profit income as well as wage income? Surely a risk to the economy is the massive financial instability and wealth loss evident in equity markets over the past month. For large firms, I believe loans make more sense than grants. The loans should be on easy terms, but they should be paid back. This kind of policy means profits will take a large hit for some time. But as a wise financial manager I talk with regularly points out, owners of equity (stockholders) know the cash flows of the businesses they own are risky. Accepting profit disruption in recessions comes with the territory. Over time, stockholders have been compensated by the market for the extra risk they bear with much higher average financial returns than those who seek the safety of bonds or certificates of deposit.

In addition, if we support the economy well over the next few months the effects on stock prices should be much less severe than the panic conditions of recent weeks suggest. I simulated a 50 percent decline in profits for a representative firm for the full year of 2020, assuming profits recover in 2021. This is a big negative shock. Recognize that profits for the first two and a half months of the year were fine, and recovery will hopefully be underway by, at least ,some time in the fourth quarter of 2020 (assuming a good policy response to the crisis). The effect of this profit shock on stock prices depends on assumptions about long-term growth and discount rates one chooses. But most reasonable assumptions imply a decline in the “fundamental value” of firms today between 2.5% and 4.0%. This kind of modest wealth decline will have very limited macroeconomic effects. The much greater drop in stock prices and the associated massive volatility of recent days are more likely due to uncertainty and fear and about concerns of inadequate policy responses than they are about the long-term profitability of corporate America.

The objective is not to protect short-term profitability but to assure the continued viability of businesses affected by the short-term social distancing Covid-19 requires. If the crisis extends longer than expected or other conditions make the burden of working capital loans too high, especially for small business, some loan forgiveness or restructuring could be implemented down the road. For now, easy-to-obtain, low-interest loans seems the way to go to keep businesses viable.

A Time for Effective Federal Leadership

In normal times, the states bear most of the cost of the unemployment insurance system. States also fund much of the Medicaid program. But the states cannot be expected to shoulder the massive financial burden associated with the Covid-19 crisis. States usually must balance their budgets. Some states have “rainy day” funds for emergencies, but these are totally inadequate to meet the needs created by this crisis. In contrast, the federal government can borrow as needed and can coordinate with the Federal Reserve to fund various activities. This financial privilege is limited to national governments who control a sovereign currency. It is not available to states within a unified federal financial system.

Without massive federal support, states will be forced to cut back on basic non-medical services and they will not come anywhere close to meeting economic needs. Contraction of state governments in a recession creates another channel for multiplier effects and magnifies the size of the problem. The vast majority of the economic response needed must come from Washington.

Economic data in the coming weeks will almost certainly look very bad. It is too late for macroeconomic policy to prevent a severe contraction. The focus of economic policy now needs, first, to support people most severely affected by the crisis and, second, to prevent a hangover of economic weakness once the medical crisis passes, creating conditions for a quick recovery.

Appendix 1: Macroeconomic Effects of the Crisis

Any numerical projection in this rapidly changing environment is speculative, but here are some stunning numbers to help understand the size of the possible economic contraction:

The first column provides data from the U.S. Bureau of Economic Analysis on the share of the economy for industries most obviously affected by social distancing. The second column is a rough estimate about how much of these industries’ output will be lost at the worst point of the current crisis. Multiplying the first two columns shows the possible effect of each industry on GDP (gross domestic product) and adding up the rows gives a ballpark estimate of how much the U.S. economy will contract.

There may be some offsetting positive effects in industries such as health care and grocery stores. But there are further negative effect caused by multipliers, investment, and international trade not accounted for. For these reasons, I bracket the total figure from the table and project a decline in GDP of 5% to 10%.

Because the headline statistics on quarterly economic growth are annualized, the reported figures will be approximately four times as large, depending on how the overall collapse is allocated across calendar quarters.

Appendix 2: Cost Estimates of Expanded Unemployment Benefits

The estimated cost of expanded unemployment benefits assumes the distribution of wages among unemployed workers equals the distribution of personal income from 2018 Census data. This assumption could overstate the actual cost for two reasons. First, personal income includes other sources of income besides labor income. Second, the distribution of unemployed workers, especially in this crisis, is likely more skewed toward the lower end of the income distribution than in the population as a whole.

I assume a replacement rate of 100% of pre-tax wages up to $2,500 monthly ($30,000 at an annual rate). The replacement rate for higher income workers is $2,500 plus 50% of the amount of their monthly pre-crisis income, capped at $6,667 ($80,000 at an annual rate). For example, an unemployed person who earned $6,667 per month before the crisis would receive the maximum benefit of $2,500 + 0.5 *($6,667 - $2,500) = $4,580. Individuals earning more than $6,667 per month have the benefit capped at $4,580 per month.

I compute the weighted average of replacement benefits for all income groups (defined in $2,500 increments) with the weights equal to the share of the population in each income group. This average is $3,261 monthly for each unemployed worker. The U.S. workforce in February of 2020 consisted of 165,000,000 individuals. An unemployment rate of 20% in a month implies 33,000,000 people out of work. At an average cost of $3,261, the total benefit paid equals $108 billion.

[1] Suppose the economy declines 8% from its peak to trough and does not begin to recover until after the end of the second quarter. Also assume that 1% of this decline takes place in the first quarter and the economy drops 7% in the second quarter. Annualization reports this figure as if the economy declined 7% for a full year. To calculate the annualized growth rate use the formula (GDP in Q2 / GDP in Q1) ^ 4 – 1 = 0.93 ^ 4 – 1 = -0.252 or -25.2%

[2] The Purchasing Managers Index (PMI) is an early signal of the direction of US economic activity. Any number below 50 signals contraction. The March 24 PMI release for manufacturing is already down to 49.2 (from a modest expansion figure of 50.7 the previous month). The services PMI dropped to a dismal value of 39.1 on March 24, an all-time low.

[3] This need is magnified by the pressure for social distancing to end prior to the elimination of the virus threat and the possibility of more than one wave of Covid-19 outbreak.

[4] Benefits are also limited to 26 weeks, but this restriction is not critical in the current crisis, assuming recovery is reasonably quick.

[5] Median monthly income for working individuals in 2018 from the Current Population Survey was about $3,250. One can make a good argument for replacing a higher share of wages above $2,500 per month to limit multiplier problems. But considerations of social equity suggest those who are not working should receive some reduction in compensation relative to workers that continue to be employed, beyond income levels necessary for meeting the most basic needs. Concerns about households needing more funds during the crisis could be mitigated with a government-provided credit line for households.

[6] Because unemployment insurance is administered by the states, the federal cost may be reduced. It is important, however, that the enhanced benefits are not whittled away by stingy state-level discounts.

[7] Another proposed restriction would limit share buybacks, and possibly dividends, for any firm that takes government funding through a crisis-relief program. The intuition for this restriction is obvious: why should the government subsidize businesses that simply pass the money on to their shareholders? The need to pay back working capital loans with interest should curtail, if not eliminate, this kind of activity.