Public Health Shocks and Product Pricing Behavior: Evidence from Life Insurance during the Influenza Pandemic

This blog is based on the authors presentation to the Economic History Society’s Annual Conference, Academic Session ASIB

by Gustavo S. Cortes (University of Florida) and Gertjan Verdickt (KU Leuven)

An extensive literature focuses on the impact of financial shocks on a firm’s activities. Despite the recent COVID–19 outbreak drawing renewed attention to the importance of disruptions for firm decisions, there is little research on how firms change their product pricing behavior after a systemic  public health shock.  Studying how health shocks propagate through firms and the general economy requires detailed, granular data, on product prices and firm fundamentals.  One potential explanation for the limited number of studies in this area  is  severe data limitation.  Moreover, to the extent that health crises are relatively rare in history, data scarcity becomes even more of a limitation.

In this paper we resort to history and explore one of the most severe health shocks in history: the Influenza Pandemic of 1918–19.  We introduce hand-collected balance sheet data on the universe of US life insurance companies between 1911 and 1922, to study both the short and long-term effects of public health crises on a crucial part of the financial system (Figure 1).  Analyzing life insurance companies is valuable and informative for at least two reasons.  First, life insurers offer a relatively homogeneous product, allowing us to draw lessons for similar industries.  Second, health shocks affect insurers through the increased liabilities they incur, which reduces the value of their investment portfolio. This multifaceted exposure to public health shocks underscores our contribution to the literature on financial frictions, which is  often limited to the asset side of insurers’ balance sheets.

The influenza pandemic is an interesting historical episode within our framework because of the nature of the disease: a large number of younger adults died. Since this cohort is generally of low risk for life insurers, this can significantly affect actuarial models, leading to a change in premiums. Moreover, unlike the COVID–19 crisis, the influenza outbreak was a health shock that did not coincide with a recession. This conveniently deals with the confounding effects of the recession present in the COVID–19 crisis studies, and it permits a ‘cleaner’ identification of the impact of the influenza pandemic.

We find that insurers more affected by the abnormal mortality shock charged higher premiums vis-à-vis less-affected insurers. This suggests that health shocks are transmitted to consumers. During the pandemic, young adults experienced an increase in excess mortality, hence, impacting this cohort’s longevity. Life insurers priced in this increase in mortality risk in newly-issued policies. This finding is in contrast to the current evidence from financial shocks. For instance, life insurers with low asset growth and high leverage decrease prices significantly more during the financial crisis. Similarly, a recent study documents that, to lure in new policyholders, insurers reduced prices on those policies that immediately translated into income. More generally, an influential paper finds that affected firms reduced prices more significantly following the Lehman Brothers’ bankruptcy. We, in fact, find that premiums increased in more affected companies (Figure 2).

Our evidence is robust to other, important market frictions. For instance, we document that financial constraints, estimated as ex-ante leverage, loss ratio, or an insurer’s capacity, did not play a significant role in explaining product pricing differences. We also find that the confounding effects arising from World War I, or any unobserved characteristics of retired and new life insurers, do not affect our findings. More importantly, narrative evidence suggests that the increase in premiums was the result of a change in firm behavior, rather than changed consumer preferences.

Critically, we uncover the real consequences of this extraordinary pricing behaviour. First, we show that the competitiveness between insurers significantly increased due to the pandemic. Interestingly, this did not come at the expense of affected insurers. Nevertheless, affected insurers were more likely to exit an affected state. Hence, public health shocks impacted entry and exit decisions made by insurers. Second, we document that insurers in financial distress generally charged higher prices. This implies that the observed price differentials increased the sector’s fragility. Similar to COVID-19, the Influenza Pandemic had a strong impact on a firm’s financial quality.

To contact the authors:

Gustavo S. Cortes, Gustavo.cortes@warrington.ufl.edu

Gertjan Verdickt, Gertjan.verdickt@kuleuven.be