Wednesday, March 11, 2020

Some Coronavirus Economics

Back in the mid-1980s, when I worked for a few years at the San Jose Mercury News as an editorial writer, my boss would sometimes remind us (channeling Murray Kempton): "An editorial writer is someone who comes down from the hills after the battle is over and shoots the wounded." Similarly, authors of books about important events have the luxury of time and distance before they commit themselves to print. But Richard Baldwin and and Beatrice Weder di Mauro, much to their credit,  decided to step into the arena of arguments about an appropriate response to the novel coronavirus while the disputes are ongoing by editing an e-book: Economics in the Time of COVID-19 (March 2020, free with registration from VoxEU.com). The very readable book was literally produces over a long weekend: it includes an "Introduction" and 14 short essays, many of them summarizing and drawing on longer work. Here, I'll draw up on some comments from the book as well as my own thoughts. 

1) The hard question is how bad the novel coronavirus will get, and the short answer is that nobody really knows. 

It is already clear that COVID-19 is worse than the SARS outbreak in 2002-3. Worldwide, that ended up being slightly more than 8,000 total cases and slightly less than 800 deaths. The Johns Hopkins School of Medicine maintains a continually updated page on confirmed cases of coronavirus around the world, as well as deaths and recoveries. As I write, it already has more than 120,000 cases and more than 4,000 deaths. 

For some context, the Centers for Disease Control estimates each year the cases and deaths from flu in the US. In the last decade or so, 2011-12 was a low mark for flu-related deaths, with "only" 12,000. Conversely, 2014-15 and  2017-18 were especially bad flu seasons in the US, with 51,000 and 61,000 deaths respectively. The 2009 Avian flu (N1H1) ended up causing between between 151,700 and 575,400 people deaths worldwide (according to Centers for Disease Control estimates), most of them in the US and Mexico. 

Predicting the path of an epidemic is difficult. Baldwin and Weder di Mauro offer a useful diagram, showing that in the early stages, a straight-line prediction will dramatically understate the harms, while in the middle stages, a straight-line prediction will dramatically overstate the harms. They offer a comment from Michael Leavitt, a former head of the US department of Health and Human Services: “Everything we do before a pandemic will seem alarmist. Everything we do
after will seem inadequate.” The challenge is to predict the length and peak of the curve --which depends not only on the epidemiology of the disease but also on what public health steps are taken. 
In addition, there is no guarantee that the coronavirus will ever disappear. AsBaldwin and Weder di Mauro note: "[T]he virus might become endemic – that is to say, a disease that reappears
periodically – in which case COVID-19 could become one of humanity’s constant
companions, like the seasonal flu and common cold."

2) What are some common estimates of potential economic losses from the coronavirus? In their chapter, Laurence Boone, David Haugh, Nigel Pain and Veronique Salins of the OECD  estimate a base scenario and a downside scenario. 
In a first best-case scenario, the epidemic stays contained mostly in China with limited
clusters elsewhere. ... In this best-case scenario, overall, the level of world GDP is reduced by up to 0.75% at the peak of the shock, with the full year impact on global GDP growth in 2020 being around half a percentage point. Most of this decline stems from the effects of the initial reduction in demand in China. Global trade is significantly affected, declining by 1.4% in the first half of 2020 and by 0.9% in the year as a whole. The impact on the rest of the world depends on the strength of cross-border linkages with China. ...

In the downside scenario, the outbreak of the virus in China is assumed to spread much
more intensively than at present through the wider Asia-Pacific region and the major
advanced economies in the northern hemisphere in 2020. ...  Together, the countries affected in this scenario represent over 70% of global GDP ... Overall, the level of world GDP is reduced by up to 1.75% (relative to baseline) at the peak of the shock in the latter half of 2020, with the full year impact on global GDP growth in 2020 being close to 1.5%.
Warwick McKibbin and Roshen Fernando simulate seven economic scenarios--three where the disease stays mainly in China, three where a pandemic spreads worldwide, and one in which a mild pandemic recurs each year into the future. For a sense of the range, their low pandemic scenario (S04) estimated 15 million deaths globally, with 236,000 in the US. Their most aggressive pandemic scenario (S06) is based on 68 million deaths worldwide, more than 1 million of them in the US. In this scenario, US GDP falls 8.4 percent in 2020, and the world economy falls by a similar amount.  To get a sense of what this scenario means, it is roughly equivalent to half the world's population being infected by the coronavirus, with a mortality rate of 2% for those infected.

3) How will the coronavirus affect the world trading system? Weber di Mauro writes: 
Supply chain disruptions may also turn out to be larger and more extended than is currently evident. Maersk, one of the world’s largest shipping companies, has had
to cancel dozens of container ships and estimates that Chinese factories have been
operating at 50-60% of capacity. Shipping goods to Europe from Asia via sea takes
about five weeks, so at the moment goods are still arriving from pre-virus times. The
International Chamber of Shipping estimates that the virus is costing the industry
$350m a week in lost revenues. More than 350 000 containers have been removed
and there have been 49% fewer sailings by container ships from China between mid
January and mid February. ... China has become a major source of demand in the world economy and many core European industries are highly dependent on the Chinese market. Sales in China account for up to 40% of the German car industry’s revenues, for example, and they have collapsed over the last weeks.
Richard Baldwin and Eiichi Tomiura write:
There is a danger of permanent damage to the trade system driven by policy and firms’ reactions. The combination of the US’ ongoing trade war against all of its trading partners (but especially China) and the supply-chain disruptions that are likely to be caused by COVID-19 could lead to a push to repatriate supply chains. Since they supply chains were internationalised to improve productivity, their undoing would do the opposite. We think this would be a misthinking of the lessons. Exclusively depending on suppliers from any one nation does not reduce risk –  it increases it. ...  We should not misinterpret pandemic as a justification for anti-globalism. Redundant dual sourcing from multiple countries alleviates the problem of excess dependence on China, though with additional costs. Japanese multinationals have already begun diversifying the destinations of foreign direct investment away from China in recent years, not foreseeing COVID-19 but prompted by Chinese wage hikes. We hope more intensive use of ICT enables firms to more effectively coordinate global sourcing.
4) Perhaps there will be a separation of global trade, which isn't likely to transmit pandemics, and free movement of people, which is more likely to do so. Joachim Voth raises this question clearly:

Fortunately, many – but not all – of the benefits of globalisation can be achieved without enormous health risks. The free exchange of goods and capital does not have to be restricted; only very few diseases are transmitted by contaminated goods. The free movement of people itself also contributes to the advantages of globalisation, but it is far less important for production. It is not obvious that running the risk of coronavirus outbreaks every few years – or worse – is a price worth paying for multiple annual vacation trips to Paris and Bangkok, say. Severe restrictions may well be desirable and justifiable, bringing to an end a half-century of ever-increasing individual mobility. In addition, specific restrictions could be brought in. For countries where, for example, wild animals are regularly sold and eaten (such as China, until recently), the certification for travel could be withheld without restrictions; anyone who comes or returns from there must undergo a medical examination and possibly spend a few weeks in quarantine. This would not only build a virtual plague wall against the next major outbreak, it would also put pressure on health authorities around the world to restrict dangerous practices that allow pathogens to jump from one species to the next. Even if airlines, hoteliers and tour operators would suffer from such rules in the short term and would complain, the lesson from Wuhan should be that we need a broad discussion within and outside of academia about how much mobility is actually desirable.
Voth also reminds us of some grim historical episodes:
The ship, Grand Saint Antoine, had already come to the attention of the port authority of Livorno. A cargo ship from Lebanon loaded with expensive textiles, it reached the port of Marseille in 1720. The Health Commission had its doubts – the plague was widespread in the eastern Mediterranean. Like all ships from affected regions, the Grand Saint Antoine was placed in quarantine. Normally, the crew and the property would have had to stay on board for 40 days to rule out the possibility of an infectious disease. But a textile fair near Marseille, where the importing merchants hoped for rich business, would soon begin. Under pressure from the rich traders, the health agency changed its mind. The ship could be unloaded, the crew went to town. 
After only a few days it was clear that changing the initial decision had been a mistake. The ship had carried the plague. Now the disease spread like a forest fire in the dry bush. The city authorities in Marseille could not cope with the number of deaths, with corpses piling up in the streets. ... At the behest of the French king and the pope, a plague wall (Mur de Peste) was built in Provence. Tourists can still see parts of it today. The wall was over two meters high and the watchtowers were manned by soldiers. Those who wanted to climb over it were prevented from doing so by force. Although some individuals managed to escape, the last major outbreak of black death in Europe was largely confined to Marseille. While probably 100,000 people – about a third of the population – died in Marseille, the rest of Europe was spared the repeated catastrophe of 1350 when millions of people lost their lives. 


5) Should the economic policies in response to the coronavirus be general or targeted? 

By general policies, I mean policies that refer to cuts in interest rates by central banks, or plans for government to send out checks to everyone (or in a US context, to cut Social Security payroll tax rates). By specific policies, I mean economic policies where the government focuses on specific issues like sick pay for workers not covered by employers, medical bills, support for small/medium firms with cash-flow problems, making sure banks have funds to lend and are not pushing firms into bankruptcy right now, and support for specific hard-hit industries like airlines and tourism.

John Cochrane put it this way:
We need a detailed pandemic response financial plan, sort of like an earthquake, flood, fire, or hurricane plan that (I hope!) local governments and FEMA routinely make and practice. Is there any such thing? Not that I know of, but I would be interested to hear from knowledgeable people if I am simply ignorant of the plan and it’s really sitting there under “Break glass in emergency” down in a basement of the Treasury or Fed. Without a pre-plan, can our political system successfully make this one up on the fly, as they made up the bank bailouts of 2008?
Then we have to figure out how to prevent the atrocious moral hazard that such interventions produce. Pandemics are going to be a regular thing. Ex-post bailout reduces further the incentive for ex-ante precautionary saving. Too good a fire department, and people store gasoline in the basement.
This starts down the same bailout and regulate road that suffocates our debt-based banking system. I welcome better ideas.
6) Will manufacturing or services be hit harder? 

Richard Baldwin and Eiichi Tomiura emphasize the problem for manufacturing:

An important point is that manufacturing is special. Manufactured goods are – on the whole – ‘postpone-able’ purchases. As we saw in the Great Trade Collapse of 2009, the wait-and-see demand shock impacts durable goods more than non-durable goods. In short, the manufacturing sector is likely to get a triple hit.
  1. Direct supply disruptions hindering production since the disease is focused on the world’s manufacturing heartland (East Asia), and spreading fast in the other industrial giants – the US and Germany.
  2. Supply-chain contagion will amplify the direct supply shocks as manufacturing sectors in less-affected nations find it harder and/or more expensive to acquire the necessary imported industrial inputs from the hard-hit nations, and subsequently from each other.
  3. Demand disruptions due to (1) macroeconomic drops in aggregate demand, i.e. recessions, and (2) precautionary or wait-and-see purchase delays by consumers, and investment delays by firms.
However, Catherine Mann points out that while manufacturing may be hit more in the short-term, it is also more likely to recoup its losses: 
Manufacturing will show a ‘V’ or ‘U’ shape. Manufacturing spillovers from factory closures loom large in the near term, but production will rebound to restock inventories once quarantines end and factories reopen. However, the duration of closures, as well as spillovers through supply chains and through virus cases and closures worldwide, will generate a set of Vs that should take on a U-shape in the global data. Importantly, the loss to global growth momentum will drag on both in individual country data and global rebound economic data, particularly trade and industrial production. Services, on the other hand, will experience an ‘L’ shape. The shock to tourism, transportation services, and domestic activities generally will not be recovered, and the projected slowing of global growth will further weigh on the L-shape evolution of demand for these non-storable tradeable services. Domestic services also will bear the brunt of the outbreak, depending in part on the responses of authorities, business, and consumers.