Fiscal Policy for the Coronavirus Maelstrom
A cacophony of economists is now calling for “whatever it takes” on the fiscal because, notwithstanding the many steps to be taken on the monetary and financial sides, the core of the macroeconomic policy response to the pandemic has to be fiscal.
But exactly how much, when, how, how long, by whom? are not just “details, dear boy, details”. Absent specifics, such calls are mere grandstanding. It is time to get specific.
The challenge is to tackle these specifics given little knowledge of the depth or duration of the pandemic or its economic fallout—beyond that until such as a vaccine or universal testing to allow identification and isolation of all asymptomatic carriers brings the pandemic to an end, it is bad.
And at best, all fiscal policy can do at a macroeconomic level is to minimize, not prevent, downturn. As consumers flee products the purchase or consumption of which requires close physical proximity to strangers, or production is hit by stay/work-at-home or lockdowns, output will fall. The macroeconomic task is to limit the drop to that.
All that is precisely the context in which fiscal automatic stabilizers work best because they dispense with need for policymakers to know in advance exactly how much intervention is required and exactly when; when well-designed, they deliver both automatically.
But the automatic stabilizers in fiscal systems in the advanced world now are poorly designed to handle the specific nature of macro-economic fallout from Coronavirus.
In particular, firms—facing pandemic-collapsed demand for their products as consumers socially distance (eg: airlines and high-street retailers), or which are hit by extended lockdowns or work/stay-at-home constraints, including indirectly through supply chains—are liable to lay off and/or stop paying workers.
Why? Because with demand collapsed, such workers are idle. So even if their employers have access to credit, paying them is loss-making—and more-so the longer the pandemic goes or is anticipated to go on. Such firms need to avert losses and, ultimately, bankruptcy.
That is the core shortcoming of the guarantee and loan programs announced by the UK (15 percent of GDP) and Germany (unlimited) and of the bold but generic French assertion that “no firm will go bankrupt”. Despite their big numbers, these programs deliver only a modest punch because credit does not avert losses.
And such unpaid or released workers should not simply shift “elsewhere in the economy”, aided by existing automatic stabilizers, because the shock is temporary: the pandemic will, eventually, pass and consumption and production patterns will, broadly, return to normal.
So absent action to prevent such layoffs at source, the upshot with automatic stabilizers “as is”, even with the announced credit initiatives, is significant amplification of what will inevitably be a major downturn, globally, as pandemic strikes all countries near simultaneously. As income and prospects collapse for such unpaid/laid off workers, multipliers kick in.
The core of the fiscal response to this prospect should be to establish automatic stabilizers now to pre-empt those layoffs or to minimize their multipliers, and to let those new stabilizers operate for as much and for as long as it turns out is necessary.
One way is to transform unemployment insurance: raise replacement ratios to 100 percent or so (so that the unemployed receive in benefit the same amounts that they were being paid in wages and salaries) and abolish work search requirements and any entitlement lags until the pandemic subsides. This minimizes multipliers as the income of affected workers is maintained.
But there are key drawbacks. Workers would actually be laid off to claim benefit and they would have to breach shelter-in-place and lockdown orders en masse in order to register for it. And benefit administrations may struggle to cope with the surge (with many administrators working from home, applying new rules, and some ill themselves). But affected workers would eventually suffer little loss of income and so maintain spending, minimizing multipliers. And those unemployed anyway reap a windfall unless these benefit changes are grandfathered.
A better approach is to avert these layoffs/pay cuts in the first place, by intervening directly via firms by establishing a temporary tax credit, a concrete “buyer of last resort” approach.
Thus, firms experiencing falls in their value added (due to sales drops or lockdowns or work-from-home) of, for example, 70 percent compared to the same month in 2019, should earn a tax credit for that month. The credit would be an equivalent percentage of their wage bill for that month (70 percent of the wage bill that month in the example). But the credit would be conditional on the firm paying out the full wage bill to its workforce for that month without releasing any of them beyond retirements or voluntary quits.
So if a firm has a fall in its value-added of 70 percent in April 2020 compared to April 2019, then it earns a tax credit of 70 percent of its payroll for April 2020, so long as it releases no workers and pays them all full base pay (excluding overtime and bonuses) in April 2020. If matters improve in May 2020 and the fall in its value-added is only 50 percent of the level in May 2019, then the tax credit for May 2020 is just 50 percent of that month’s wage bill, if all pre-crisis workers are paid and retained in May. And so on.
The tax credit could be set against all taxes firms collect and pay—including VAT/sales tax, and personal income and social security taxes, not just corporate taxes. The setup could be monthly as in the example, or quarterly, as convenient. The firm can decide where and how to allocate its workers in the meantime, through work-sharing, paid leave, or training.
Thus, the tax credit is highly adapted to different circumstances facing different firms in different sectors in both cross-section and time-series senses. And it works even if firms take a different view from governments about the likely duration and depth of their individual or the aggregate demand shortfalls. A “de minimus” criterion should be included—so the fall in value-added should be at least say 10 percent before the credit kicks in. Employees, part-timers, and contractors should be included in the definition of “workers”.
Thus, workers are not released and do not need to break travel embargoes to register with unemployment administrations, de minimus means the fiscal effort is targeted on more exposed firms, the administrative arrangements of paying out the tax credits can be incorporated into tax administration over time, and affected workers receive completely uninterrupted pay from their employers in the meantime without having to do anything.
Basing eligibility for the credit on firm value-added rather than gross final sales addresses the highly varied integration of affected firms in supply chains. For countries with VATs, this “tax credit base” can be derived from VAT returns; for those without, it can be built from P&L tax returns. And this approach avoids bespoke packages for, say, airlines, which always gives rise to risk that equally severely affected sectors—say the events industry or restaurants—will be overlooked for lack of a similarly high public and political profile.
And far better than discretionary fiscal injections—public investment projects and the like—this stabilizer dispenses with need for policymakers to possess a list of “shovel ready” projects in their top drawers, or to know either the array of firm-specific details or overall pandemic duration in order to get the calibration of fiscal policy broadly right. The latter features give these proposals key advantages over related Irish or Danish schemes.
The tax credit requires firms to secure finance across the pandemic, with the tax credit constituting both the collateral for the borrowing and the shield for their P&L accounts. Thus, actions taken by central banks and regulators to ensure the broader functioning of credit markets for corporates—even with more to be done on that front—will thereby be put to good purpose.
To prevent firms which were unprofitable even before the pandemic from simply stopping all operations—thereby deliberately decimating their monthly value added to secure eligibility for the tax credit—all so as to reduce their monthly losses, this tax credit should be defined as a debt of any company that goes into bankruptcy within some 3 years of the end of the pandemic, and as the most senior of such debts. Alongside, a cap on increases in eligible monthly wage bills can check outlier abuse of the tax credit by that means.
Unlike, say, calls to issue everyone with a $1000 cheque (too little for those in need to zero multipliers; too much for those not in need) or to cut payroll taxes across the board (ditto), this tax credit curbs amplification of the underlying macroeconomic shocks at source, thus constituting a highly targeted automatic stabilizer and job guarantee.
That setup diminishes exposure of such as gig workers to a minimum—as aggregate demand falls are minimized—and their remaining income shortfalls may be addressed directly by transfers through their platforms. And the tax credit preempts cascading chaotic payments arrears of rent, mortgages, utilities, trade credits, taxes etc.
How much will it cost? A lot—that part of the wage bill of firms which they cannot pay as a result of the pandemic, whatever that part of the wage bill turns out to be. To give an order of magnitude: if 70 percent of pre-crisis GDP is private, 25 percent of which is hit by bigger-than-de-minimus demand falls in a given quarter, which average 60 percent, with an all-inclusive wage bill of 70 percent of its pre-crisis value-added, then the tax credit sums to some 7½ percent of pre-crisis GDP at an annual rate.
That—even with other numbers—is a very big fiscal action.
But it is calibrated to what turns out to be actual macroeconomic need and no more, and its duration may be less than a year—if a vaccine or universal testing and isolation of asymptomatic carriers brings the pandemic to an end soon enough. It costs less than the parametrically equivalent “transformation of unemployment benefit schemes” and it secures greater benefit, both because it takes effect faster and is administratively simpler. The de-minimus criterion minimizes deadweight losses—if some firms hit by pandemic demand falls would have retained their full workforce anyway. And it would replace—rather than be additional to—part of the credit initiatives policymakers have already announced.
And with interest rates low and a big macroeconomic hit coming even if the tax credit is fully implemented immediately, there is a compelling case to err, if at all, on the side of doing too much.
Prompt credible policy announcements and implementation are essential because firms are facing these challenges right now. That is a key advantage of the tax credit over the unemployment insurance route. And as the economic effects of the pandemic ease, so the tax credit stabilizer automatically scales back until the scheme can be formally withdrawn.
By thus securing employment and minimizing the macroeconomic fallout from the virus, other fiscal policy actions—including full fiscal provision for health providers (whatever they need for emergency care, vaccine research, and securing universal testing), extension of universal paid sick leave benefits to encourage sick-at-home behavior, and interventions for such as gig workers—can proceed in the best of feasible circumstances.
Though introduction of an automatic stabilizer does not immediately lend itself to the sort of fanfare international fiscal announcement of the sort proclaimed at the London G-20 in 2009, impact would be greatly enhanced if carried out by all advanced countries together in a blaze of publicity. Thus, Euro Area fiscal rules should adjust to accommodate this for highly indebted members like Italy—with the tax credit design assuring all that no more is done there than is necessary. But in any