The Pandemic Requires a Different Macroeconomic Policy Response Jan Kregel

Let me start by noting that I have been on shelter in place in New York since the beginning of March so that my comments will be coloured by conditions here. I would like to start by setting out an alternative approach to the economic problems raised by the pandemic in difference from both the mainstream approach and some Keynesian inspired policy responses.

My knowledge of the mainstream analysis comes from the widely ventilated interviews of economists on the major news programmes. Not surprisingly these were usually couched in terms of the impact on supply and demand. Since the initial outbreak was first announced in China, and met with a rigid local quarantine, the initial focus was on the reduction in production due to disruption of global supply chains that put pressure on supplies and prices. When contagion started to be visible in Europe, and then in the United States, leading to suspension of non-essential production activities the discussion quickly shifted to the collapse of demand problem as workers were furloughed or dismissed. So, the discussion shifted quickly from an inflationary spiral to a recession or depression problem or both.  The dominant policy discussion was around the most appropriate remedies for the supply bottlenecks and failing demand. With the recent financial crisis in mind, most recommended the financial support measures similar to those that had been employed after 2008 – higher government deficit spending and low or zero interest rates to offset the declines in output and to support demand. The Federal Reserve moved quickly on interest rates and expansion of its balance sheet while Congress produced multiple “stimulus” packages to offset the impact of the pandemic. Keynesian democrats joined Cheney Republicans to argue that there was nothing to fear from deficit spending and expanding national debt. I think this was the wrong place to start, and led to the wrong policy response, because it was a mischaracterisation of the problems facing the economy.

Was there an alternative point of departure and method of analysis? I am not a virologist or epidemiologist and thus totally unqualified to say anything at all on the impact of the Coronavirus pandemic – and have attempted to keep silent. On the other hand, there is an alternative approach that is possibly relevant and that I am presumed to know something about – uncertainty and expectations. Quite early on, developments in China made it clear that the analysis of the corona virus was a case of absolute, complete and total uncertainty as to its nature and impact. Keynes had already pointed out that the natural response to a case like this would be to presume that tomorrow would be much like today. Thus, the first response was to back to something that you know. And this is what happened – the references were to the experience of SARS and MERS, to the death rates, contagion rates, which were contained at what were thought to be manageable values. The official response was “do not panic,” we know that SARS and MERs were eventually controlled, so we really don’t have to be too concerned. Further, as was initially thought with MERS, there was no confirmed person-to-person contagion. With no idea what the virus was, or what the impact was going to be, the natural response was that it would be like the past, so the first order of business was to prevent panic; which outside of China was quite easy as there was no real evidence of the presence of the virus before February.

But as evidence in China accumulated, the transmission rate turned out to be much higher than it should have been, and mortality was much higher than expected, reinforced by the rapid spread and mortality in parts of Northern Italy. As the physical evidence of overcapacity in hospitals and morgues emerged from Wuhan and then Lombardy, it became necessary to form new expectations. Forecasters started to model the prospective evolution of infection and mortality based on parameters to reflect this new reality; and the numerous model projections of the spread of the disease quickly produced estimates that exceeded capacity of potential treatment facilities. The panic was as much due to the fear of insufficient hospital beds and ventilators as to the potential explosion in the death rate.

But the basic parameters of these models, such as the now famous R₀ (or reproduction rate) and the mortality rates, were just best-guesses and were initially thought to be much higher than historical norms. And more importantly, reporting from China raised the suspicion that unlike the earlier episodes, there was clear evidence of asymptomatic viral shedding and person-to-person contagion (paradoxically both had been characteristic of the Spanish Flu, but there was no historical memory) that called into question the estimates of R nought. If you don’t know how many people are infected and spreading the disease before showing symptoms you cannot know the spread rate. The same was true for the death rate, since asymptotic spread means that the real death rate would be much lower than recorded because the denominator is much higher than reported when as many as 25-40% of the infections are asymptomatic. And this miscalculation was aggravated by the initial response of “don’t panic”, which allowed the spread to ramp up largely unseen. As we now know, the virus has been identified in November and was already global and extensive in December in China and Europe and in January in the US.

This meant that the parameters that were required for the model projections were not only uncertain, they were unknowable until full scale testing determined asymptomatic infection and transmission rates. This problem was further aggravated in the US by the fact that test kits were in short supply, were defective, and only offered to those demonstrating clear symptoms. So, two months after the official Wuhan announcements, uncertainty was complete and total and it appeared that health services would be wholly insufficient to provide even minimal treatment.

In these conditions, the only available medical response was to take physical measures to stop transmission – full scale lockdown appeared to be the only way to avoid pandemic. The policy problem was no longer how to manage the impact on the economy, to avoid recession, it was to influence R₀ to stop contagion and the breakdown of the health system. The economic problem became the management of expectations in pursuit of one objective: move R₀ > 1. The recession was thus the collateral result of active policy decisions to achieve that objective. Non-essential employment had to be suspended. It was not the result of cyclical movement in demand, it was imposed on the economy. The problem was to ensure that the lockdown could be maintained until the objective was achieved. And this presented another unknown. In Wuhan, the lockdown started the end of January and the relaxation in April, which was around three months. This suggested that the cost of fighting the virus would be something less than a quarter of GDP in lost income and production, depending on how rapidly the subsequent resumption of activity could take place.

Many economists responded with traditional policies to fight recession, not realising that the economic problem was not to provide stimulus to offset the “recession”, but how to ensure that the economy could sustain the likely GDP and employment loss under lockdown. There was no possible “stimulation” response – stimulus doesn’t kill the virus, money is useless if you are sheltering in place and can’t go out to shop except at Amazon. Job guarantee programmes are not useful if the point is not to work, if the employment loss is imposed. Guaranteed basic income programmes do not work if you have to provision in public.

What had to be done was to make sure that everyone managed to survive the pandemic, which meant staying quarantined at home in order to avoid public contagion and overcrowded hospitals. And, in the aftermath of the mistakes made with stimulus in response to the Great Recession, it was important that this should be done in an equitable manner, to ensure it did not make income and wealth inequality worse.

The response thus called for a different page from the New Deal experience – building confidence that the threat could be overcome. It is now recognised that there was no formal blueprint for a New Deal when Roosevelt took office, and indeed many of the policies implemented were borrowed from the Hoover Administration. But Roosevelt recognised the need to provide a sense of confidence and security, to free decisions from the fear of fear itself. Thus, he moved to restore confidence in the banking system via rapid legislation; and he moved to restore confidence by providing income through the myriad employment programmes and finally a blueprint for recovery. Ira Katznelson notes that what “observers and commentators” of the dilemma facing the incoming administration shared was “an understanding that theirs was a time when uncommon uncertainty at a depth that generates fear had overtaken the degree of common risk that cannot be avoided. … When deep uncertainty looms, the ability to choose is transformed. …  Measurable risk generates worry. Unmeasurable risk about the duration and magnitude of uncertainty spawns fear… Under conditions of fear… people develop a heightened mindfulness and self-awareness about the constraints on free action, and take, as a central goal, the desire to restore a higher degree of coherence and certainty; that is, they try to reduce deep uncertainty to ordinary risk.” (Katznelson 2013, 33) And this is precisely what was required in order to generate support for the only rational response to the virus – loss of income and employment via a strict lockdown.  But it was necessary for this policy to recreate a sense of safety and security, of a clear path to exit from the threat of disease and unemployment, rather than fear and uncertainty of employment loss.

How could it have been done? It should seem quite clear that neither guaranteeing income nor providing government employment would work. Lockdown means not working, not working means no demand for labour and staying home means you don’t have the possibility to spend much anyway. What was required was a guarantee not of income but of survival, of what we may call “social provisioning”. So, the first thing is to make sure that if everybody stays home, and even if they lose their jobs, they will have enough to eat and survive without fear and the constraint of loss of income.

The US has a minimum system of state provided unemployment benefit and of food support (known by the acronym SNAP), which provides provide food security for low income families. One possibility would have been to support everybody’s income by expanding these programmes to everyone classified as non-essential and subject to lockdown. These are cash transfer programmes that require the ability to use income for provisioning in the public market. But the essence of the lockdown means suspension of the market mechanism, suspension of the labour market, of the consumption market, of public activity. Providing income transfers requires provisioning through participation in the market; if there is no market, then it cannot work. Provisioning must be provided by central organisation, by government.

Here’s where the New Deal comes in again. A food distribution organization—some sort of Civilian Conservation Corps—would have been required to make sure that everyone was provided with enough to eat. In most large cities, restaurants donate their excess supplies to food banks and other non-profit private or religious welfare agencies. It is obvious that in the shutdown restaurants cannot operate; the supplies that they would have bought would become redundant. Here was excess supply that could provide food security, if only it could be organised and distributed. In the event it took press reports that the excess food production from the countryside’s producers was being trashed before a system was set up to channel it to the unemployed through non-profit beneficial agencies and soup kitchens.  The first order of business should have been the reorganisation of the food service sector into a food distribution sector; social provisioning to insure a minimum subsistence and survival to everyone in lockdown. Hebert Hoover did this for Belgium in the First World War, so it clearly could have been done instead of sending everyone $1,200 to chase down non-existent goods on empty store shelves, and benefiting speculators who profited from creating shortages and creating incentives to break quarantine in search of supplies.

The next measure would have been to look at other types of expenditure. If you are a good Keynesian economist or stock flow modeller, you know that every income is an expense and if you know your national accounting, you know that production creates income and expenditure to take off that production. The “stimulus” response was a focus on incomes only and forgot about costs and earnings. A more sensible approach would have looked at the need for a balance between costs and incomes in the lockdown. If the firm is not producing, it is not earning; variable costs go down; it does not need a government loan to support employment if workers have to stay home. If households receive government distributed food subsistence, they don’t need wages from employment to survive.

But in the shutdown, there are also fixed costs that have to be met, for both the family unit and the firms—rents, mortgages, leases etc.—which create insecurity if they are not met. But it is not necessary to provide income transfers to ensure that these payments are met. Fair burden sharing of the costs of the crisis implies that the rational approach is to suspend all fixed cost payments.  A successful lockdown requires every economic unit to have minimum costs and minimum income, but to be assured to survive the lockdown period in condition to recommence normal activity. Thus, suspending financial operations and avoiding unnecessarily increased debt that will be carried over to the recovery and creates difficulty in repayment is the order of the day. to support and maintain social provisioning.  Government support and finance and organisation of the production and distribution of subsistence are required, not the provision of incomes to maintain capital values.

There is no reason why rental income or capital income should have preference over labour income. If everyone qualifies for social provisioning, the landlord gets food security just like the furloughed production worker. Here is the key to the idea of an equitable distribution of the burden of fighting the pandemic. Capital incomes and capital values have to be treated the same way as human capital and labour incomes. Household non-essential workers are losing, say, three months of income; so should all other wage or capital income recipients. Everyone should shoulder the capital loss created by the reduced flow of income, both human and financial.

The basic principle is that to stop all flow costs, since everybody’s cost is somebody else’s income, we don’t have to give anybody money to cover those costs or offset the lost incomes. Capital incomes have to share in the costs of fighting the virus. This means not only that we suspend trading in capital assets, but that the incomes of all corporate administrative and management personnel are eliminated. They also now go down to subsistence provisioning. That’s the way we try and share equitably the cost of the shutdown. This would also mean that the government would not have to engage in massive deficit spending aside from the support of social provisioning. The government will have to engage in a certain amount of organization; and we’re now back to the Roosevelt administration. It took about a month to set up the Civilian Conservation Corps, under the Employment Act. But we are not really interested in the impact on jobs and income, but the decisive action to set up the organization and management in record time. What was needed in this case was not government money: government did not need to spend a great deal in order to solve this problem. What was needed was to get the government to provide an equitable means for sharing the costs of quarantine to fight the virus, and to make sure that particular sectors and categories didn’t pay a higher price than others.

In the 2007-2008 crisis the government bailed out the banks and the management of private corporations, but let households lose their homes. Instead of an equitable distribution of capital loss, there was a capital transfer. This time was supposed to be different. But it looks very much the same in the US. There was stimulus support for banks, for firms, and this was supposed to maintain jobs—for workers the companies did not need. Should one be surprised that the job maintenance of these programmes was extremely low? The major impact of the stimulus was to expand government debt and produce fee income for the financial institutions managing them. And what of the income transfers and unemployment benefits? It is telling that household savings hit record levels during this period.

And as a result of the attempt to stimulate during the lockdown there is an increase in government debt. Is this a problem? We all know that it really doesn’t make that much difference. And there are any number of Keynesians who support the government debt finance stimulus because they argue deficits don’t matter. And they don’t, but if they are unnecessary then there is a very big political cost, because every time the government deficit increases the debt, Congress attempts to cut essential services; which are precisely the things that are needed in order to respond to the crisis at the federal level.

However, in this regard the real problem is on the state level where governments have been responsible for most of the health care costs. Since state governments run on a legally imposed balance budget principle, the eventual response is going to be cuts in state government expenditures that are primarily the income and employment of those essential workers in the forefront of the fight against the virus. Medicare and education, the basics that we need, are going to be cut as a result of the mistaken emphasis on stimulus.

But there is a much simpler argument against the use of stimulus to offset a policy induced recession. Consider Keynes on the appropriate role of government expenditure: “… if our central controls succeed in establishing an aggregate volume of output corresponding to full employment as nearly as is practicable, the classical theory comes into its own again from this point onwards. If we suppose the volume of output to be given, i.e. to be determined by forces outside the classical scheme of thought, then there is no objection to be raised against the classical analysis of the manner in which private self-interest will determine what in particular is produced, in what proportions the factors of production will be combined to produce it, and how the value of the final product will be distributed between them. …  Thus, apart from the necessity of central controls to bring about an adjustment between the propensity to consume and the inducement to invest, there is no more reason to socialise economic life than there was before.”

Those who justify additional deficit financed stimulus expenditures with arguments about the unimportance of government debt rely on the first part of the argument, but overlooks the fact that the market system which is presumed to function to allocate these expenditures to produce output and employment no longer functions, since the very response to the pandemic has been to shut down the labour market and suspend private decision making. The appropriate policy should be the “necessity of central controls” over federal expenditure, in order to offset rising unemployment that has been created by central controls suspending the market mechanism. If we take Keynes at his word, the response to the pandemic should have been not only central controls on aggregate output, but central controls on the determination of production and distribution. Without the latter the former is likely to be ineffective as response to the pandemic, as the equivalent of Keynes’ well-known references to burying bank notes in the ground: better than nothing, but not efficient in fighting the virus. Better food distribution and building hospitals and providing loans to business operating in a system without a market mechanism.

If the objective is to eliminate the virus contagion and it requires shutting down the economy, there is no need to support full employment income through government expenditure. The problem is rather to provide the appropriate redirection of production and an equitable burden sharing in which social provisioning is assured.  Debt, deficits, transfer payments are not part of this. Central control of the market should be the goal. Relying on the crippled market mechanism cannot do it. If you fight a war, this is what has been required in the past.

As a final coda, there is the problem of the essential workers and those who manage to remain employment while in quarantine. This is part of the problem of equity. Since it is difficult to suspend these incomes, the appropriate solution would be a progressive tax on the incomes of home workers who can telecommute without loss of income to be used as a supplement to incomes of essential workers in the sanitary support sector, to offset the increased contribution and risks incurred in fighting the spread of the virus.  As a metaphor, the idea is for the economy (like Snow White) to fall asleep, while the trusty essential workers manage to the fight against the coronavirus, and once R₀ is sufficiently reduced the economy can resume without unacceptable disparity in the distribution of costs of the lockdown.

Why Stimulus cannot solve the Pandemic Depression
Jan Kregel (SSER IDEAs Online Lecture Series)
May 11, 2020.