Monday, April 6, 2020

What shape recovery?

Will the recovery be V shaped, quickly roaring back to the previous level? It does that every January 2 after the long Halloween-Thanksgiving-Christmas-New Years slowdown, and it did in 1984. Or will it be an agonizingly slow U or L shape, as the recovery from 2008 turned out to be?

I had early hope for a V, but a fear that shuttered businesses and permanently fired people would turn it into an L. Those take much more time to reorganize. Hence, lots of blog posts advocating a more nuanced policy than a blanket lockdown.

But now I think it's clear the virus will not end with a sudden all-clear, like January 2 or an air raid. If, as we all hope, the current unbelievably costly lockdown does its job, we will in a month or two emerge with the curve bent, a stable or declining number of cases. But the vast majority of the population will still not have been exposed. We will not have "herd immunity" -- and a good thing too as 1% of the herd will not have died to get there. And cases both home and abroad will not be zero.

Nothing short of a cheap, effective, incredibly safe vaccine given to just about everyone on the planet will change that.

That means the virus is ready to reemerge promptly. All it takes is one person to travel to a town, go to a restaurant or club meeting, wait two weeks, and you have an outbreak all over again. We will have hotspots and flare-ups needing intense testing, contact tracing, local lockdowns, travel restrictions, and so forth -- if our bureaucracies are finally up to the task of doing anything competently.

On the individual and public health level this means almost all of us -- who have not gotten it or don't trust that you can't get it again -- will be practicing some sort of social distance for a long time. And, getting to the point, this all suggests a period of very slow economic activity. I don't want to call it "recession" as that word implies simple lack of aggregate demand, the Keynesian uni-causal story. No amount of printed or borrowed money will get the social distance economy going again.

Ross Douthat nicely sketched a picture of the coming economy in the Sunday NYT:
Life at half capacity: Right now our institutions must survive while essentially closed — with few or no customers, moviegoers, travelers. But soon they will have to figure out how to reopen while maintaining the social distancing that semi-normalcy requires.
... fewer people will come out, and because there will be rules governing how many people can come in.
...the scenes at some grocery stores right now, the line of people six feet apart waiting to come inside and shop, may become a permanent feature of the semi-normal landscape. Churches will hold services with every other pew occupied. Restaurants will seat every other table. Planes could fly without a single middle seat occupied. Sports may resume without spectators, relying on TV revenue alone.
Ross doesn't fully draw the economic conclusions of this vision. Such grocery stores can only serve a fraction of the number of customers, yet need more employees and still have to pay the rent. Such restaurants make half as much money yet still must pay the rent. Such airlines still pay the pilots, flight attendants, fuel, and larger cleaning and disinfecting crews. This is not a sustainable economy -- at today's prices.

Ross turns to the government
And since the flow of money and custom and attendance won’t come close to what existed just a month ago, any government response will have to be calibrated to a half-capacity world — where institutions are technically open for business, but they still need help to stay alive.
I have bad news. The government is also a limited resource. We cannot go on for months on end with the government paying half the bill of everything. Just who is buying all those government bonds? With what income? A trillion dollars a month adds up.

The answer is, this is an enormous negative supply shock, together with a big shift in demand. If only  half the seats can be filled, running an airline just got twice as expensive. If only half the tables are filled, running a restaurant just got twice as expensive. Those prices have to double. Which in turn, will drive customers away, towards driving (RV sales should go up), cooking at home, fancy takeout, and so forth.

There is likely also to be a shift towards precautionary savings. I think lots of people and businesses have figured out that keeping some cash or money market investments around is a good idea, and overall appetite for  risk is going to be lower. I diagnosed markets as suffering from a panicked demand for cash last month. But the standard business cycle mechanism of lower "risk appetite" makes a lot of sense. The shift towards a desire for safe investments may also keep markets low for a long time. This is the standard business cycle mechanism. (Don't think about saving vs. investment. Think about desire for risky vs. safe investments. Business cycles are about risk premiums.)

Torsten Slok writes by email (summarizing gated DB research) suggesting
Increase in precautionary savings for households... More space between seats at restaurants, cinemas, sports events concerts, conferences, trains, buses and airplanes. Fewer people traveling on vacation and going out... Older generations staying at home, less willing to put parents in retirement homes. Limits on the numbers of people un supermarkets, more online shopping, more online doctor visits. Fewer people going to fitness centers, doing group sports. More people driving their own car to avoid public transportation. 
Aside. This could be the kiss of death for public transport in the form of busses and trains. If there is anything that cannot stand a doubling of its cost, and attendant decline in demand, that's it.
Less business travel.. more video conferencing... fewer buybacks, lower dividend payouts  [more equity less debt] 
and, I am not alone worrying
more supply of government bonds, increasing risk of a debt crisis.  
Policy will face the usual cruel tradeoff: The more help you give the unfortunate, the more disincentives, and the slower the recovery. Noah Williams writes perceptively of unemployment expansion (which, whatever its faults, is probably the best of the government's responses -- better than cash payments to everyone which will arrive late summer, better than bailouts for airline stock and bondholders, and municipal bond holders)
the program is poorly designed. It provides incentives for employers to lay off workers. In the future—assuming the pandemic restrictions are lifted before August—it will discourage people from returning to work.
The current federal relief package extends unemployment benefits to 39 weeks, plus additional payments....Under the new expansion, the average replacement rate across states would increase to roughly 116 percent 
You're running a business. You have some cash around, and could keep people on, at least at reduced hours and health insurance. If you fire them, though, they can get 116% of their salary from the government, and Obamacare. It's a no brainer.

There is good news in this however. It means that much of what looks like unemployment may really be furlough. The people and employer know where each other is and can snap back more quickly.

39 weeks of 116% of salary though gives people little incentive to answer that phone call. Especially while schools are closed, day care is closed, and gardeners aren't allowed to come around.

Moreover, there is already a shift in demand -- to cleaning crews, online services, and so on. Paying people to sit at home makes sense in the lockdown. But much less in life at half capacity -- and rapidly changing -- economy.

This seems heartless, but it is brainless to ignore that there is always a tradeoff between help and incentives. The last recession and half-hearted recovery was a chaos of bad incentives. Noah:
In general, unemployment-benefit programs try to balance insurance with incentives, seeking to provide relief when needed while also offering motivation to look for work. States typically require recipients to search for a job. Setting the replacement rate well below 100 percent is usually a strong encouragement for them to do so.
Covid-19 presents unusual circumstances because unemployment has been enforced by government decree. Much of this joblessness will likely be temporary, with workers rejoining their employers once the pandemic subsides and restrictions are removed. Further, while some employers (Amazon, Walmart, grocery stores) are adding jobs, most companies are, at best, putting a freeze on hiring. The disincentive effect of unemployment benefits in the current crisis is minimal, while relief needs are large; thus, Washington has increased benefits, and many states are waiving job-search requirements.
Policymakers should be wary, though, of implementing relief provisions that will delay economic recovery, as occurred during the Great Depression and the 2008–2009 Great Recession. The Federal Pandemic Unemployment Compensation program is time-limited, but if the shutdown ends within the next four months, the aggressive unemployment-benefit replacement rates well in excess of 100 percent would hamper the labor market’s recovery.
In short, great generosity makes sense in the lockdown, but must be much more carefully calibrated if we do not want an L shaped recovery.

And emerging chaos at unemployment offices and small business administration suggests the help may come just as it is no longer needed.


  1. If this does turn out to be the kiss of death for public transport, will it also be the kiss of death for big cities as we know them? Can a city like New York function while maintaining social distancing on public transport and elevators? That would also be a major change from the current trend, with major economic implications.

  2. “We will not have "herd immunity" -- and a good thing too as 1% of the herd will not have died to get there.“

    I think we need to be really careful with this conclusion. The issue of the Covid-19 infection fatality rate (IFR) has to be thought of it terms of competing risks. It’s the excess mortality that matters. A lot of the Covid-19 patients who die, would have died from a competing risk (e.g., the flu, etc.) on a similar time scale. If, as a recent report in The Lancet shows, the IFR is really about 0.7% then the excess mortality might be barely measurable. A non-trivial fraction of the population susceptible to death from the flu and other viral infections end up dying of Covid-19 instead. Competing risks analysis drives our understanding of excess mortality, which is what ultimately matters.

    In fact, due to the lock-downs, social distancing, and increased awareness on hand-washing, the mortality rate in Europe is lower right now than what we would expect at this time of year precisely because these efforts remove competing risks too. Even now, Covid-19 is the background relative to the flu in terms of numbers of deaths this year.

    1. Please give support for your assertion that overall mortality rate in Europe is lower than expected. In fact, WSJ reported the opposite, with regions of Italy seeing 3x-4x the normal amount of deaths for this time of year. LBS report from Apr 5 also see a similar trend where mortality across Italy, Portugal and France are 2x-4x what is normally expected for this time of year. ( WSJ link: )

    2. Excellent points.

      I disagree with John Cochrane on this one. I don't want to live this way, especially by government ukase.

      If people wish to stay home, that is their business.

      We cannot destroy an economy in order to save it.

      600,000 people a year die from cancer in the United States, often triggered by artificial carcinogens. That is every year, not a one-time event. We do not shut down the economy in order to remove carcinogens in the environment.

      Sometimes you have to take the least-bad option. The least-bad option now is to go back to work.

    3. gsb05,

      The statement that European death rates have dropped below expectation for this time of year and that Covid-19 deaths might be unreported can both be true. The lock-downs mitigate a lot of the competing risks and taken together, the competing risks are larger than Covid-19. It's excess mortality that we need to pay attention to.

      Here is the reference (European Monitoring of Excess Mortality):

  3. You describe a world in which demand is running at a rate less than half the rate pre-COVI-19. With demand for goods and services at one-half the pre-COVID1-9 level, employment will be one-half or less (depending on technology) and incomes at a similar depressed level. What would you anticipate the general price level to be in those circumstances? Should it not be one-half the price level pre-COVID-19? Prices reflect income levels, as always.

    You mention rents for retail space and office space. Rents and lease rates will reflect income levels--the landlord has no choice in the matter, he either reduces the rental rate or the tenant moves to another landlord's premises offering the reduced rental rate. Vacancies for all but residential living space will rise--employment at one-half the prior "normal economy" level will see to that. Capital will be costlier because it is more scarce, in real terms.

    Airline travel at one-half the previous "normal economy" level?--then only one-half the airplanes will be flying and producing one-half the CO2 emissions (bonanza!). Oil production?--one half (another bonanza for the environment!)

    Globalization?--busted. This epidemic has demonstrated the fallacy of relying on others to produce the goods that the economy requires.

    The actions of the U.S. government in procuring N95 medical grade face masks demonstrated for all foreign governments that relying on U.S. suppliers wherever their production plants are located is a fool's gambit.

    Berlin has learned this lesson; Canada has too. Result?--return to the branch plant economy with local country control through a local board of directors such as the Swiss require. NAFTA or USMCA?--the point of either is what now?

    Free movement across borders is all but ended; has ended in eastern Canada between Canadian provinces, and between Canada and the U.S., and between the EU and the U.S. This is the new normal. The traveller?--no longer wanted--he poses an unacceptable risk to the indigenous, the locals--so see him off, ward him off, with the police if necessary.

    Rebound?--unlikely. Bon chance!

  4. The Back Death lasted from 1347 to 1351. As history has shown, in addition to wiping 30% of Europe's population, it created massive economic disruption that broke the back of medieval feudalism.

    I think it's far too early to discuss outcomes. There is no vaccine. Attempts to develop an HIV vaccine have been unsuccessful for decades. Hydroxychloroquine seems to work in vitro but it has also been shown to work in vitro for viruses like Zika and dengue but fails in clinical use. Results thus far for hydroxychloroquine have been mixed and after reviewing what's been published I would say it works well in patients at low risk of mortality and not for patients with more severe disease.

    Anthony Fauci has publicly stated that he has looked at all the models and found them all deficient.

    In summary, nobody knows how this will play out. Everything is speculation. Don't count on a vaccine or treatment (nothing looks very encouraging and a vaccine is at least a year away - maybe 30 years away like the HIV vaccine), don't count on monetary or fiscal intervention (nothing will work when there are massive reductions in services, production or consumtion) and dont count on herd immunity, which will require culling of the herd.

    IMHO barring a miracle this is going to get worse in terms of deaths, social disruption, and economic ruin, and it could last years.

    Hope is not a plan.

    1. This is a extremely pesimistic view. The spanish flu of 1918 is a case much closer to Covid19 and lasted just from january 1918 to december 1920. Being mostly harmless since october 1918 (despite being much more devastating during the initial outbreak).

      There is no clear explanation for that, but virus mutation is probably behind this extreme change in the mortality rate of the spanish flu.

      Hope is not an strategy but when you add up the probabilities of:

      1.- Virus mutation (virus tend to become less lethal ... incentives work pretty well also for them!)

      2.- Effective vaccine development

      3.- Increased herd inmunity

      4.- Improvement in treatment. The HIV that you mencione is a good example. Even if we have not been able to develop a vaccine the mortality rates have gone, in USA, from 16.3 per 100,000 in 1995 to 3.7 in 2007.

      When you add up all this, the chances of having the pandemic completely forgotten in less than a year (as in fact happened with the spanish flu in 1918) are pretty large.

      Actually the HIV is, maybe, a good example since there were business that did take a hit from the virus but were back in pretty good shape in a sort period of time. And the kind of "social distancing" that the virus sparked are mostly gone leaving a change in behaviour that we are probably also going to see with Covid19.

    2. This is just a recap of wishful thinking. AIDS had far less economic impact. What businesses were impacted? Bath houses?

      The AIDS fatality rate dropped due to improved antiviral medication. There is still no vaccine after 40 years of trying.

      I doubt anyone is going to forget this in a year.

      What herd immunity? Where is your evidence that surviving C19 infection provides immunity? At least 8 strains have been identified. You might survive type 1 but you could still be vulnerable to 2 through 8.

      I dont know where you got the impression that viruses tend to mutate into weaker forms. Look up the terms genetic drift and genetic shift for viruses.

      The tendency to evolve new strains is one reason flu vaccines dont work very well. We've experienced some colossal epidemics just recently because they predicted the wrong strains. Its like having the weatherman pick the strains for the next flu season.

      And how come if you catch a cold you can still catch a cold over and over? Everyone has had a cold. Where's our herd immunity?

      And Spanish Flu "lasted just from january 1918 to december 1920." JUST? Thats 3 years.

  5. Well, we're already seeing line mechanisms in place to accommodate social distancing. How long until we return to mostly a price mechanism? Hard to say - unless! We do, wait for it: more testing that's fast and efficient.

  6. I am afraid the scenario will be similar but worse than the 80s no growth with inflation. The reorganization of consumption and production. The related technological and organizatinal/institutional innovation may provide some relief. But I'm afraid much pain and perhaps social and political unrest are waiting round the corner.

  7. Just like saving businesses through the epidemic, it's good to save people. No one disagrees. What may make the recovery harder is making unemployment pay more. 'Twould in fact be better to give the unemployed unconditional cash than to tie the funds to unemployment compensation.

    As for the future, whatever behavioral changes may come, the free price system will make the best of it, middle seats filled or not.

  8. One thing you are missing is the perspective of the now-dominant economic force in the country. This is no longer the baby boomer cohort that you belong to and are familiar with, but millennials. Millennials will seamlessly head back to work, back to restaurants and bars, back to parks, etc. They aren't nearly as afraid of re-engaging with the economy as you fear – and they are rapidly (and finallly) taking over the economic reigns from the boomers. Note: I'm in neither cohort, but I can see this transition coming like a freight train, and we'll need it over the next 12 months.

  9. While the economic fate of workers in the food and hospitality industries is horrible, keep in mind that millions of workers in banks, insurers, academia and high tech are still getting very nice salaries.

    Yet they have very little to spend this money on. When and if the virus subsides, there will be a huge pent-up demand for new cars, new houses, et al.

    From my reading, that is how we got an economic boom in 1946. The relative lack of consumer goods during the war, plus higher salaries to domestic workers plus good savings habits -- all worked together for prosperity. It could happen again.

  10. The economic dislocation from COVID-19 need not be as serious as it appears at first sight. To see why, consider a simplified perspective, Case A.

    Suppose COVID-19 requires shutting down non-essential firms, e.g., theatres, sports events, restaurants, cruise ships. These firms and their employees lose their income. However, those who would have purchased their goods and services have money they otherwise would have spent that exactly equals the lost income of the closed firms and their employees. In principle, the federal government could tax this money away from those who would have spent it and pay it to those they would have spent it on. This would leave all firms and their employees with the same income as before with all essential services being provided exactly as before.

    Once the COVID-19 crisis ended, all that would be necessary is to cancel the tax, at which time everyone could return to their previous production and spending patterns.

    Case A requires no borrowing, no budget deficit, and no serious economic hardship on anyone, and a quick return to the previous pattern of economic activity once the crisis is over.

    There are several issues with Case A. For example, Case A fails to take advantage of the closed firms’ laid off employees by having them do something useful to the extent possible. This suggests that implementing something like Case A can be done in a way to yield a better result than Case A, since at least some of these newly unemployed people would do something useful (deliver food from supermarkets).

    Another issue with Case A is that it is not practical to determine who would have spent how much on the closed firms’ goods and services. This makes impractical an accurate assessment of who should pay how much tax. On the other hand, there is no good reason why only the people who would have bought the closed firms’ goods and services should pay the whole tax – that would mean that the closed firms and their employees, who lose nothing under Case A, gain at the expense of their customers. It seems better if everyone shares in paying the tax. This would make things easy – simply charge a surtax on everyone’s current tax bill (remember, the closed firms and their employees retain their gross income in Case A). Call this Case B, which is better than Case A.

    Things are not quite so simple. The COVID-19 crisis requires large additional expenditures for health care, although not as large as it appears at first sight, because only incremental costs must be counted.

    Incremental medical goods and services come at the expense of leisure time and giving up production and consumption of other items to provide the time and materials to produce medical goods and services for COVID-19 use. All this is a rearrangement of the use of person-hours in real time and represents only a re-allocation of resources, not magically creating them. Consequently, just as above, increased taxes are a feasible mechanism to pay for it all with no borrowing and no budget deficit, and as few adverse consequences as possible.

    Thinking in real terms sometimes makes things clear when dollar considerations are confusing.

    1. You can always call Ricardo to the rescue and eliminate the constraint of the tax and the "money sending" of Case B having to take place simultaneously.

      You just send the money now to the former employees of all those "non essential" activities and wait for the Covid-19 crisis to fade away before charging a surtax in everybody tax bill (including, as it is fair, the none essential activities employees which would be, at that point, back to work).

      Business as usual ... not working as intendend as usual.

  11. John: one thing I do not understand about the long-term social distancing you envision... isn't it far more likely we'd just start wearing high quality PPE, and otherwise go about our business as normal? Sure bars & restaurants* will suffer -- can't eat through a mask -- but why would anything else need to change? You'd think high-quality, comfortable PPE would become widely available pretty quickly in the scenario you envision.

    The above is not a rhetorical question: I see enough "permanent social distancing" talk from smart people who I respect to appreciate there's probably something I'm missing here.

  12. I think permanent social distancing can make people more hygienic than before!


Comments are welcome. Keep it short, polite, and on topic.

Thanks to a few abusers I am now moderating comments. I welcome thoughtful disagreement. I will block comments with insulting or abusive language. I'm also blocking totally inane comments. Try to make some sense. I am much more likely to allow critical comments if you have the honesty and courage to use your real name.