Does the Spanish flu offer lessons in how to tackle a pandemic?

The influenza pandemic of 1918-19 is the subject of renewed research interest as scholars look back to history for insights into Covid-19. Influenza-18 or the Spanish flu, as it is commonly known, infected perhaps a third of the world’s population and had a death toll of between 50 and 100 million, with 2.64 million deaths in Europe – 1.1% of the continent’s population.

Significant variation in mortality outcomes of this virus across different populations has inspired epidemiologists, historians and social scientists to explore the efficacy of public policy in disease control. But there is still precious little work on the economic impact of the Spanish flu.

What does evidence from epidemiologists, historians and economists tell us?

• The 1918-19 influenza pandemic occurred before the modern epidemiological transition had fully completed, when background mortality was still high and infectious disease outbreaks were more commonplace.
• The pandemic originated from an unknown source and arrived in countries through major trading ports, carried, among others, by military personnel returning from the battlefields of the First World War.
• Influenza-18 diffused through populations in a process of spatial contagion in three waves, the second being the deadliest.
• Public transport networks were a significant source of disease transmission.
• The disease proved particularly deadly in developing countries with inadequate public health systems. India is likely to have suffered the highest death toll in the world.
• There was significant variation in the flu’s health impact within countries, an outcome driven by local differences in demography, density, economy, environment and policy.
• Demographic factors explain much of the flu’s health outcomes. Men tended to be affected more than women, and the disease proved particularly fatal to infants, the 20-40 age group and the elderly.
• The reason why adults aged over 40 were less affected is not fully understood. One hypothesis is that they enjoyed some inbuilt immunity thanks to exposure to pandemic influenza in the late nineteenth century. Another is that younger adults were simply most exposed to the disease, given the nature of the economy.
• The use of non-pharmaceutical public health interventions (lockdown measures) varied significantly across the world, and within countries, but none were as long or extensive as those implemented today.
• Where lockdown measures were used, they helped to reduce disease transmission and bring down peak death rates.
• The decentralised funding of the UK’s health system in the early twentieth century is likely to have hampered policy-makers’ ability to combat the disease.
• While there is insufficient work on the short-term economic impact of the pandemic to draw any firm conclusions, it is likely that those affected by Influenze-18 while in utero suffered long-run health consequences into adulthood.

How reliable is the evidence?

The evidence on the Spanish flu is published in academic publications in different disciplines, including epidemiology, mathematics, health economics, demography, medical history and historical geography. Each of these disciplines has different norms on what constitutes reliable and useful knowledge. Their different approaches to knowledge creation present us with challenges if we want to collate this knowledge and learn lessons for policy-making today.

Contrasting works of epidemiology and health economics highlight the nature of the problem. Both disciplines make use of large datasets and conduct rigorous statistical analysis, but ‘rigour’ in both disciplines constitutes something different. It is noticeable that the statistical techniques of epidemiology tend to be more straightforward than those used in modern econometrics; they base their insights on descriptive statistics and multivariate correlations and do not think about causality in the same way that economists do.

But ‘simpler’ is not necessarily worse; there have recently been some very fragile studies of the Spanish flu by economists that rely too much on technique and are not reliable when scrutinised further (see, for example, the very high-profile pre-print paper by Correia et al, 2020, and the forensic reply by Lilley et al, 2020).

Predictably, epidemiologists focus on environmental and demographic factors, while economists tend to focus on more social and economic variables in their analysis of the Spanish flu. Both disciplines are interested in the role of public policy, but perhaps have different outcome variables in mind. Epidemiologists are naturally less interested in the economic implications of public health policy interventions – that is, the potential trade-offs between public policy interventions and a country’s economic performance.

An influential epidemiological study on public policy during Influenza-18 shows cities in the United States that attempted to reduce infectious contact between people early in the 1918 pandemic had significantly lower peak death rates than cities that were late to adopting disease containment policies (Hatchett et al, 2007). A useful contribution from economics demonstrates the strains that the Spanish flu put on Sweden’s social security system (Karlsson et al, 2014).

Of the few economic studies that examine the impact of Influenza-18, it is worth highlighting one on the pandemic's short-run impact on wage growth (Garrett, 2009), and two that measure its long-run impact by tracking what happened to the generation affected by the disease while in utero (Almond, 2006; Parman, 2015). These studies use US data; no such study has been carried out with UK data.

While economists and epidemiologists tend to be interested in general patterns, historians look at specific case studies and focus much of their attention on outliers. Historians make use of archival evidence that is not easily quantifiable but is nevertheless necessary to understand the specific institutional context of the historical period. In particular, the detail they provide on public health infrastructure is required to interpret any of the quantitative findings of epidemiologists and economists.

Some of the statistical analysis presented in the writings of historians of the Spanish flu is not very reliable; unfortunately, most historians working in academia today do not use statistics. Their strengths lie elsewhere and their contribution to knowledge of this pandemic remains invaluable. There is a useful review of much of the recent historical scholarship of the Spanish flu (Phillips, 2014).

What else do we need to know?

Influenza-18 has naturally piqued the interest of economists who see this historical pandemic as a way to develop a better understanding of Covid-19. Certainly, the advantage of such historical scholarship is that economists can study the full course of the pandemic disease, from its initial emergence through to its long-run economic consequences.

But economists should be cautious and acknowledge that their chosen subject of study is inherently interdisciplinary. The toolkit of an econometrician is insufficient on its own to take full account of all the relevant confounding factors. The variables on which epidemiologists and demographers focus in their studies of the flu, such as age and sex, must also feature in our work (Colvin and McLaughlin, 2020). The limitations of historical statistics should be understood. And the context and nuance provided by historians is crucial to drawing useful applied lessons for policy-making today.

The job of interdisciplinary scholars, such as economic historians, is to piece together the work of these disparate disciplines into a coherent story which can then be used as a basis for learning. But given the scale of the mortality shock resulting from the pandemic, it is surprising how little attention economic historians have spent examining the Spanish flu compared with other events in the early twentieth century. A rare example of recent work on the Spanish flu published in an economic history journal suggests a direct link between air pollution and influenza mortality (Clay et al, 2018).

The lack of attention of economic historians is now likely to change in light of Covid-19. Indeed, a host of new studies have already been distributed as pre-prints. Much of this new work is being distributed in Covid Economics, a new publication from the Centre for Economic Policy Research. For example, one study looks at economic recovery from Influenza-18 in Denmark (Dahl et al, 2020), and another looks at the case of Italy (Carillo and Jappelli, 2020).

But we need to remain very careful when linking these studies with specific policy proposals. Not only is the knowledge not yet peer-reviewed, but the historical context of the Spanish flu is not necessarily directly applicable to the situations we face today. As L. P. Hartley’s 1953 book The Go-Between begins: ‘the past is a foreign country, they do things differently there’.

Where can I find out more?

Coronavirus and Spanish flu: economic lessons to learn from the last truly global pandemic: Chris Colvin and Eoin McLaughlin explore lessons from this historical pandemic for today’s policy-makers.

What the 1918 flu pandemic tells us about whether social distancing works: Nancy Bristow on lessons for today.

The 1918 influenza did not kill the US economy: Efraim Benmelech and Carola Frydman look for a possible explanation for the limited adverse effects of the flu despite similar social distancing requirements.

Re-evaluating the benefits of non-pharmaceutical interventions during the 1918 pandemic: Guillaume Chapelle looks at the debate on the efficacy of lockdown measures during the Spanish flu pandemic.

Pandemic 1918: BBC Radio 4 series on the origin, spread and reaction to the pandemic that devastated much of the planet just over 100 years ago.

We’re stuck fighting a 21st century public health crisis with 100-year-old medical techniques: Caitjan Gainty and Agnes Arnold-Forster look at how public health policy developed during the 1918 pandemic is still being used today

Who are UK experts on this question?

• Chris Colvin, Senior Lecturer in Economics, Queen’s University Belfast, is an economic historian of the interwar era, working on the demographic impact of the Spanish flu in the case of Ireland.
• Joan Roses, Professor of Economic History, London School of Economics, conducts research into historical economic geography and is an expert on the Spanish flu in the case of Spain.
• Caitjan Gainty, Lecturer in History of Science, Technology and Medicine, King’s College London, conducts research into the ways medicine’s history can usefully inform healthcare policy today.
• Mark Honigsbaum, Lecturer in Journalism, City University London, is the author of several articles and monographs on the history of the Spanish flu pandemic in the UK and elsewhere.
• Matthew Smallman-Raynor, Professor of Analytical Geography, University of Nottingham, is an expert on the spread of the Spanish flu throughout the UK and elsewhere.

Author: Chris Colvin, Queen’s University Belfast