The outbreak of Coronavirus has swept across the world tragically and unexpectedly, jeopardizing our health and safety, stressing national health care systems and putting the global economy on hold, creating excruciating uncertainty for millions. Although the pandemic is likely to carry on running its destructive course for some time to come, we believe that it is essential to analyse the mechanisms of the economic recession that it will inevitably trigger in order to prepare for the recovery process. We begin our analysis by looking into the past occurrences of similar crises.

Should we classify the impending pandemic recession as genuinely unprecedented? Global outbreaks of infectious disease have devasted humanity historically, but so far in the past that many could claim this outbreak is indeed unparalleled. The last pandemic of comparable scale took place more than a hundred years ago when the Spanish flu outbreak disrupted Europe in the immediate aftermath of the First World War. Although it’s clear that the world was very different at that time, it is valuable to examine the implications of this pandemic recession. To do so, we summarize the research by Karlsson M. et al, (2014) which provides an excellent description of the context and a thorough analysis of the economic implications of the Spanish Flu outbreak in Sweden in the early twentieth century.

The historical Swedish story of pandemic recession is particularly interesting because it enables a quantification of the magnitude and nature of the economic shock caused by an outbreak more accurately. Firstly, Sweden retained a neutral position during the First World War. This limited its economic effect significantly compared to other, belligerent countries. Secondly, Sweden is renowned for its long tradition of keeping meticulously recorded statistical data. Founded in 1858, the Swedish Statistics Bureau has been publishing yearly data on nine different societal issues since 1911. Third, Sweden was a homogeneous society, stripped from the potential effects of cultural differences and significant asymmetry in demographic response, which might disrupt an analysis of a particular economic shock.

The researchers treat the Spanish flu pandemic as a labour supply shock, using it as a natural experiment allowing them to evaluate the predictions of economic growth models, deriving their hypotheses from the two-factor model of Lucas-Uzawa. Using seemingly random variation in the outbreak incidence rates throughout the Swedish counties, Karlsson M. et al, (2014) test their hypotheses using the difference-in-differences method on Swedish administrative data. They estimate the pandemic effects on personal earnings, physical capital returns and poverty rates. However, before looking into the hypotheses and corresponding results of the research, it is useful to outline the context of the influenza outbreak in Sweden in the early twentieth century.

At the time of the outbreak, Sweden was a capitalist state undergoing rapid industrialization. At the turn of the century, the country was still mostly agricultural, with 53% of the population making their living in the agricultural sector, while 29% worked in manufacturing. By 1930, the Swedish economy had undergone a dramatic structural shift, with only 39.4% of the population engaged in agriculture compared to 35.7% in manufacturing. (According to the World Bank, the modern Swedish service-sector orientated (80%), 2% are engaged in agriculture, while 18% are working in the industrial sector.)

The Spanish flu pandemic came in three waves that took the world by surprise. The first, very mild wave of influenza arrived in spring 1918 and would have remained unnoticed until the second, deadliest wave came in fall 1918. The second wave turned out to be 5-20 times more fatal than common influenza, which typically claimed the lives of 0.1% of those who contracted it. The last wave that took place in 1919 occurred mainly in Scandinavia and some islands of the South Atlantic. As a result of the pandemic, almost 1% of the Swedish population perished. Unlike COVID-19, the Spanish flu took the lives of healthy young people, with 50% of the death toll being made up of those aged 15-40. However, as noted above, there were dramatic differences in mortality within different counties, which have been outlined by the authors for research purposes.

As previously stated, Karlsson et al, (2014) assume that regional economies behaved according to the Lucas-Uzawa model, exhibiting a sharp decrease in labour supply caused directly by the pandemic. Therefore, within the first group of hypotheses, they expected the epidemic to lead to an immediate relative increase in earnings and physical capital returns in more affected compared to less-affected areas. This would manifest as a deepening of capital. Moreover, the researchers expected that profoundly affected districts would demonstrate a lower growth rate (in production, earnings and capital returns) during the transition phase after the pandemic until the proportions of capital and labour levelled out to long-term equilibrium levels.

The results of the study suggest that the pandemic had a strong negative impact on physical capital returns, which was presumed to have been caused by a combination of both short and medium-term effects. According to the researchers’ estimates, the highest affected quartile of counties (concerning influenza mortality) experienced a drop of 5 per cent during the pandemic and an additional 6 per cent afterwards. Moreover, the researchers did not find a significant effect on earnings either during or after the epidemic.

The second hypothesis concerns the effect of the pandemic on poverty rates. The researchers hypothesized that an immediate impact on poverty would be the net result of two opposing forces: an increase in household poverty owing to dependents losing the principal breadwinner and a decrease in the rate of the impoverished population due to rising wages and capital returns. In the medium term, researchers expected the differential regional poverty (from the outbreak) to close.

The outcome appears to reject the second group of hypotheses as well. The researchers identified a significant increase in poverty levels following the pandemic, which manifested once the last outbreak had receded by 1920. The top affected quartile of Swedish counties suffered an increase in poverty by 11 per cent compared to the bottom quartile.

There is no doubt that the world we live in is fundamentally different from the one a century ago. However, the COVID-19 outbreak has undoubtedly caused a negative shock for the global economy as its core manufacturing hub, China, closed down. As the pandemic unravels it produces many more local shocks, rendering thousands of businesses around the world untenable, leaving millions of people uncertain about their future. However, just like the Spanish flu pandemic, sooner rather than later, the COVID-19 outbreak will become a historical event. The impending pandemic recession of 2020 is not likely to be a result of a labour shock caused by increased mortality, but rather an aggregate supply/demand shock resulting from a conscious global policy decision to limit their economic activity to save as many human lives as possible. We believe, despite the hardship, it makes us extremely lucky.

References:
  1. Karlsson, M., Nilsson, T., & Pichler, S. (2014). The impact of the 1918 Spanish flu epidemic on economic performance in Sweden: An investigation into the consequences of an extraordinary mortality shock.   Journal of health economics36, 1-19.

  2. World Bank, World Bank Open Data. (2020) Employment in agriculture (% of total employment) (modeled ILO estimate). URL: https://data.worldbank.org/indicator/SL.AGR.EMPL.ZS

  3. World Bank, World Bank Open Data. (2020) Employment in industry (% of total employment) (modeled ILO estimate). URL: https://data.worldbank.org/indicator/SL.IND.EMPL.ZS

  4. World Bank, World Bank Open Data. (2020) Employment in services (% of total employment) (modeled ILO estimate). URL: https://data.worldbank.org/indicator/SL.SRV.EMPL.ZS