What can we learn from past pandemics?

In this week’s article we explore the macroeconomic consequences of previous pandemics. We outline how pandemics have typically influenced key economic variables such as GDP, wages and interest rates. We look at the factors that have influenced the magnitude of those economic effects, including the scale and severity of the pandemic itself and explore how economic policymakers have responded to previous pandemics. And finally outline the key lessons that are likely to apply to the Covid-19 pandemic.

Setting the scene

Appendix 1 shows the major pandemics in the past 1000 years. The Black Death of 1347-1353 and the Spanish Flu of 1918-1920 have received the most attention in analyses of previous pandemics, in part because they caused by far the largest loss of life. But other pandemics, including various outbreaks of cholera in 19th Century Europe and flu in the 19th and 20th Centuries have also had a devasting impact on lives and livelihoods.

More recently, while the 2003 SARS epidemic in China and its near neighbours was less deadly, it is still of interest since it occurred at a time when economic structures and institutions were very similar to today. Compared to previous pandemics, at least, we also have reasonable economic data on the countries affected. As a broad rule of thumb, the further we go back the worse the availability and quality of the data and the greater the differences in the economic and political systems compared to today.

But we have basic data covering the key macroeconomic variables for most of these pandemics. And where data deficiencies prevent us from coming to quantitative judgements, we have used other sources to form qualitative ones.

A conceptual framework

Before we start it is worth establishing a theoretical framework for thinking about the potential economic consequences of pandemics. One obvious consideration is the effect on output. The loss of life, and restrictions on activity to stem the spread of the virus, both voluntary and mandatory, will tend to depress output. In addition to the scale of the output fall, the time taken for it to return to pre-virus levels is also an important determinant of the economic impact. At the same time, pandemics could in theory affect how the proceeds of growth are distributed. The loss of life (and subsequent contraction of the labour force) will, all other things being equal, increase the amount of capital per worker and therefore lower investment demand. At the same time, precautionary saving by households may rise. Both factors would act to push down real interest rates, thus meaning the income return to capital falls. The flip side of this would be a rise in the share of GDP flowing to labour.

The UK provides the richest source of data. Indeed, it is the only country for which annual output data go as far back as the Black Death in the 14th Century. Since then the UK has experienced eight pandemics with death tolls of over 100,000. These led to an average peak-to-trough fall in output of 5.5%. What is more, it took on average four years for GDP to return to its pre-virus level.

Likewise, there is some evidence in contemporary accounts to support the idea that pandemics produce a shift in how income is distributed. For example, according to one account, in the year following the end of the Black Death, having previously been flat, the pay of textile workers in northern France increased three times. Likewise, in the decade after the Great Plague of London in 1665, the daily wages of skilled building workers rose by 45%. This is also evident in historical data on real wages produced by the Bank of England (dating back to 1311!). These show that real earnings have increased by an average of 60% in the decade following the onset of the same eight pandemics. Equally, there is some evidence that the real rate of interest has tended to fall in the aftermath of pandemics.

At first sight, then, it appears that pandemics tend to cause large falls in output, from which it takes several years to recover, and also tend to shift the proceeds of growth away from capital and towards labour. There is no “normal” pandemic However, there have been a wide range of outcomes around the average. It becomes clear that the “average” effect of a pandemic is a somewhat meaningless concept.

For example, while the peak-to-trough fall in GDP has averaged 5.5%, in several cases (including the Great Plague of London in 1665) there has been no discernible effect on GDP. On the other hand, during the Black Death, there was a peak-to-trough fall in GDP of 23.5% and the same is true of the effects on real wages and interest rates. In the past, the extent of economic damage has been influenced by the scale of human suffering. The deadliest pandemics, including the Black Death and the Spanish Flu, left a big mark on GDP. But less deadly pandemics, including the Hong Kong Flu of 1967-69 and the 2003 SARS outbreak, had smaller effects.

This reflects the fact that the economic disruption caused by previous pandemics was due to either the death or incapacity of workers. This directly reduced the supply of labour and stopped the rest of the population from going about their usual business due to fear and uncertainty. What is more, both increased with loss of life. In contrast, the economic disruption caused by Covid-19 has been caused in large part by measures to stop the spread of the virus itself. Previous pandemics have seen people voluntarily adopt some aspects of the Covid-19 lockdowns. In his diaries, Samuel Pepys notes how London’s streets were “mighty thin of people” at the height of the Great Plague of 1665. Pepys himself decamps to his country residence in Greenwich and references to work in his office at the Navy Board plunge as the Plague intensifies.

While there were city-wide quarantines as far back as the Black Death, Covid-19 is the first pandemic in which restrictions imposed on economic activity by governments to curtail the virus have been so widespread. In this respect, during the coronavirus, the economic damage has been caused not by the loss of life, but by measures to preserve it. Likewise, a variety of institutional and structural constraints had prevented monetary and fiscal policy from being loosened to cushion the blow to demand during previous pandemics. On the monetary side, these restrictions were largely the result of the straight-jacket imposed by past systems tied to gold. On the fiscal side, not only was the state far smaller during past pandemics, but most of these outbreaks occurred in the pre-Keynesian era where fiscal expenditure would not have changed in response to an exogenous shock.

Economic structures, institutions and behaviour

While the economic impact of previous pandemics varied widely, perhaps the more important historical lesson relates to their impact on consumer behaviours, economic structures, and institutions. The coronavirus has clearly affected almost everyone’s way of life, with fewer visits to public places, workplaces and a decline in tourist activity. But these changes are not unique. Even as far back as the Black Death there is evidence that outbreaks of disease not only had a significant impact on consumer behaviours but also often preceded significant institutional change. In the following section, we break down the various effects into three broad categories. The first are those that relate directly to the shock from the virus. The second are those that are indirectly related to the virus. And the third are those changes that were already underway, but that were accelerated by the virus.

Changes directly related to the virus

The first, and probably most obvious, changes are those that are the direct result of the virus. In the coronavirus pandemic, these include changes in consumer behaviour such as an increase in mask wearing, working from home, reduced travel and fewer social gatherings. But they also include institutional change. And while some of these changes are specific to the coronavirus, some are not– even as far back as the Black Death, there is evidence of similar effects.

One key similarity between the impact of the coronavirus and previous pandemics is the desire to avoid infection. In some cases, it has been because of restrictions imposed by authorities – the word “quarantine” originated during outbreaks of the plague. Indeed, in Italy, all ships were required to isolate for forty (quaranta) days before they could dock. But in other instances, it has been because individuals have adapted their behaviour. During the Black Death, lockdowns were imposed sporadically in the UK. In the areas where quarantine measures were not imposed, there is plenty of evidence that individuals chose to spend more time at home, chose not to work, or if they work, they only did so for considerably more money.

“Will the coronavirus permanently change behaviours?”, Samuel Pepys’ diary provides some insight into behavioural changes during the Great Plague of London in 1665. Indeed, the number of references to certain places and activities in his diaries provides a guide to what parts of his life Pepys put on hold during the plague and for how long. For instance, as well as not visiting the office in the second half of 1665 and retiring to Greenwich, the number of visits to pubs and shops also dropped significantly.

The travel and tourism sector has often been hit very hard during disease outbreaks. Even during the Black Death in the 14th Century, when travel was very limited, there were travel restrictions in certain cities in Italy. Milan denied access to foreign visitors. Perhaps more relevant, though, was the sharp drop in tourism during the SARS outbreak in 2003. At the height of the SARS outbreak in May 2003, the hotel occupancy rate in Hong Kong fell to just 20% of “normal” levels. All these changes were fairly short-lived. After all, pubs did not die off in the 1300s! Instead, people tended to return to their daily lives fairly quickly once the threat of disease dissipated. Following the first outbreak of the Black Death, for instance, there is evidence that people returned to work once the immediate threat had eased and as wage growth accelerated. There is even evidence that, having lost their spouse to the plague, the financial security that marriage offered at the time led people to get remarried quite quickly too.

More recently, during the SARS outbreak, evidence suggests that tourism rebounded to “normal” levels within a year of the crisis. Hotel occupancy in Hong Kong, for instance, returned to its pre-virus levels after about six months. Admittedly, the SARS outbreak was shorter, more contained, and less deadly than today’s coronavirus outbreak. And we expect it to take longer for the travel and tourism industry to recover this time. Other shocks to the sector, such as terrorist attacks, have weighed on passenger numbers for a far longer period. Following the 9/11 attacks, it took two years for US flight numbers to rebound back to the levels before that date.

But the weight of the historical evidence suggests that there is unlikely to be a permanent change in tourist activity and demand will recover eventually. There are some changes in individual behaviour that have been more permanent, though. During the Spanish Flu outbreak at the start of the 20th Century, makeshift masks became a popular method of protection in Japan. Further outbreaks of the flu in the middle of the century coupled with some natural disasters solidified their widespread use. In 2019, for instance, Japanese consumers spent $230 million on surgical masks. And following the SARS outbreak, the custom of wearing masks became more prevalent in other Asian countries, including China and Taiwan.

Indeed, the historical evidence suggests that masks and other PPE products are likely to become a common feature of life in Europe and the US. A desire to maintain stocks of these products therefore is likely to be more permanent. As well as the impact of the virus at the individual level, disease outbreaks have preceded institutional change. The most obvious is that the size of the state has often increased as a result of disease. This virus has been no exception: fiscal support announced as a result of the pandemic is now equivalent to 7% of GDP in the US, 8.7% in the UK and as high as 12.5% in Germany. Not only has government spending increased, but governments have intervened directly in individuals’ lives. Examples of this are the restrictions placed on people’s movements and the forced, temporary closure of certain businesses.

Pandemics have also sparked major structural changes in how health care is delivered and who provides it. The first international health boards were developed during the first wave of the plague among a number of cities in 14th Century Italy. More recently, the Spanish Flu gave a significant boost to the concept of socialised medicine and healthcare. Indeed, following the disorganised approach to the Spanish Flu in 1919, the Canadian government set up its first federal healthcare system. And in Russia, the government funded the first public healthcare system – the premise of which was later adopted in many other European countries. So, although direct fiscal support is likely to be withdrawn as the virus fades, there are longer-term spending implications insofar as the public will demand greater public health spending.

Pandemics accelerate changes already underway

So far, we have discussed changes that relate directly to the virus, but pandemics have had indirect effects on behaviour too. These have tended not to come out of the blue, but instead relate to an acceleration of trends that were already underway. One example is an increase in distrust – both in authority and amongst different groups and individuals. For instance, the Black Death heightened suspicion towards the Church and preceded a decline in the number of religious observants in the 14th and 15th Centuries. This extended further, in that it led to an overall distrust in authority. Some economic historians suggest that the plague was one factor that drove peasants in the revolts in the UK at the end of the 14th Century. More recently, there is evidence that the Spanish Flu increased the racial divide in South Africa at the beginning of the 20th Century. Indeed, scapegoating by white South Africans began the legislative process that would eventually lead to Apartheid.

In today’s world, fault lines run between China and the US – we will publish more on this over the coming weeks. But on a domestic level, the evidence suggests that the coronavirus may exacerbate distrust in experts and the establishment, and potentially lead to a further fracturing of society. When it comes to changes in consumer behaviour, the most obvious in the past have related to where people chose to live. For instance, the Black Death sped up the desertion of the countryside. Survivors abandoned inferior land and moved to the city, attracted by the improved infrastructure and the increased housing supply. Admittedly, the poor sanitation conditions in cities at the start of the 14th Century meant that urbanisation was one of the reasons the plague spread as widely and as quickly as it did. But the flipside of that was that cities were quick to improve sanitation standards. This time, there is plenty of evidence that some different consumer trends are already gathering pace. While leisure travel will probably return as fears over the virus subside, technological progress means that there are   more acceptable substitutes for business travel. Admittedly, new business relationships that require trust are probably far easier to form face-to-face rather than screen-to-screen. But existing relationships can be adequately maintained via video conferencing. Similarly, while the shift towards online spending and the use of electronic payments has been underway for a decade now, the coronavirus has fast-tracked these trends and brought them forward. Indeed, in the UK, the US and the euro-zone, non-store retail sales, the vast majority of which is online sales, have risen by around 50%, 40%, and 25% since the start of the year, respectively. And the shift towards working from home will probably be more permanent given that both employees and employers have been moving towards doing so for a long time. One broader key macroeconomic shift that has started to play out over recent years is the retreat of globalisation. Pandemics and globalisation are more closely linked than one might imagine. The Black Death, for instance, spread from east to west along trade routes linking East Asia and Europe. As we discuss in a forthcoming Focus, we believe that the current pandemic has already threatened globalisation. Since January, world trade volumes have declined by over 15%, passenger numbers on flights dropped to almost zero at the peak of the lockdowns and the coronavirus has exacerbated the already taut relationship between the US and China, threatening a further hit to world trade.

Conclusions

Since the economic implications of past pandemics have been so broad, our analysis tells us that there is no such thing as a “normal” pandemic. And given that the macroeconomic consequences have previously been determined by the number of people who died, Covid-19 is likely to have a very different impact on the economy. The effects of this virus will also be distinguished by the fact that the policy response by governments and central banks has been particularly rapid and particularly large. The most important lesson from history that is likely to apply to today’s crisis is that seismic shocks tend to accelerate changes to economic structures, institutions and behaviours that were already underway. This implies that the current pandemic could leave a legacy that touches on everything from globalisation to the future of work.

The opinions expressed are those of Irongate using the independent research of Capital Economics and are subject to market or economic changes. This material is not a recommendation or intended to be relied upon as a forecast. You may get back less than the amount invested. The value of an investment with St. James's Place will be directly linked to the performance of the funds selected and may fall as well as rise.