Pandemics Initially Spread Among People of Higher (Not Lower) Social Status: Evidence From COVID-19 and the Spanish Flu

Pandemics Initially Spread Among People of Higher (Not Lower) Social Status: Evidence From COVID-19 and the Spanish Flu. Jana B. Berkessel et al. Social Psychological and Personality Science, September 7, 2021. https://doi.org/10.1177/19485506211039990

Abstract: According to a staple in the social sciences, pandemics particularly spread among people of lower social status. Challenging this staple, we hypothesize that it holds true in later phases of pandemics only. In the initial phases, by contrast, people of higher social status should be at the center of the spread. We tested our phase-sensitive hypothesis in two studies. In Study 1, we analyzed region-level COVID-19 infection data from 3,132 U.S. regions, 299 English regions, and 400 German regions. In Study 2, we analyzed historical data from 1,159,920 U.S. residents who witnessed the 1918/1919 Spanish Flu pandemic. For both pandemics, we found that the virus initially spread more rapidly among people of higher social status. In later phases, that effect reversed; people of lower social status were most exposed. Our results provide novel insights into the center of the spread during the critical initial phases of pandemics.

Keywords: social status, pandemic spread, COVID-19, Spanish flu

Discussion

According to a staple in the social sciences, people of lower social status are particularly exposed to pandemics (O’Sullivan & Bourgoin, 2010; von Braun et al., 2020). Acknowledging that pandemics evolve in phases (World Health Organization, 2017; Zhang et al., 2020), we provided a more nuanced account. Specifically, we proposed that during the critical initial phases of pandemics, people of higher (not lower) social status are at the center of the spread. Our research builds on established social status theory in social psychology: People of higher social status exhibit more independent behavior, are more mobile, and possess more diverse social networks; people of lower social status exhibit more interdependent behavior, are less mobile, and possess more homogenous social networks (Carey & Markus, 2017; Kraus et al., 2012; Thomson et al., 2018). Further, our research helps reconcile findings from the ongoing COVID-19 pandemic that appeared contradictory (Drefahl et al., 2020; Mogi et al., 2020). Finally, our research makes much-needed progress toward a distinctively psychological theory of pandemics (Betsch, 2020; Van Bavel et al., 2020).

Across three nations, two pandemics, two levels of analysis, and different data sources, we found full support for our hypothesis. Study 1 used region-level infection data and showed that in the initial phases of the COVID-19 pandemic, the virus spread primarily in higher income regions. In the later phases of the pandemic, however, the virus spread primarily in lower income regions. Study 2 showed that the 1918/1919 Spanish Flu pandemic in the United States initially spread more strongly among people of higher social status. Later on, by contrast, the pandemic spread most strongly among people of lower social status. Study 2 relied on a unique historic data set of millions of memorials, which we web-harvested using the latest social data science techniques. As such, the study highlights the potential of adding a historical perspective to derive insights into pandemic processes (Hatchett et al., 2007) and showcases the potential of social data science to scrutinize such historical data at large scale (Adjerid & Kelley, 2018; Kosinski & Behrend, 2017). By combining large-scale contemporary data and historical data, our work is one of the few psychological studies that examines historic events (Arnett, 2008; Muthukrishna et al., 2021). A limitation of our work is its focus on three Western, developed nations (Henrich et al., 2010a, 2010b). Accordingly, it will be an important task for future research to probe whether our findings extend to non-Western, less developed nations.

The COVID-19 pandemic hit the world underprepared (Remuzzi & Remuzzi, 2020; Walker et al., 2020). Hence, far-reaching societal decisions had (and still have) to be made under great uncertainty. Our study aims at reducing these uncertainties and ultimately at improving allocation of resources. We provide a more nuanced understanding of the initially most exposed societal strata during pandemics, informing both, scientific theory and real-world decision making. In fact, our findings have actable implications for containment efforts: First, people of higher social status are at the center of the initial spread. As such, higher income regions should probably be a focus of outbreak monitoring whenever a new pandemic seems to rise. Second, the longer pandemics last, the more they will spread among people of lower social status. Importantly, our results suggest that previous research might even have underestimated the risk people of lower social status face at later pandemic phases. This is so because previous estimates did not differentiate between earlier and later phases and, thus, lumped together their opposed effects.

As we write this article, we are amid a second COVID-19 wave. It, therefore, remains an important task for future research to examine whether our results generalize to later waves. Recall that our theoretical reasoning rests on the preconditions that (a) the virus is not yet widespread and that (b) social behaviors are unrestricted. These preconditions are necessarily fulfilled during the initial phases of a pandemic’s first wave. Regarding later pandemic waves, however, it is less clear whether those preconditions are fulfilled. On the one hand, in many nations, the virus was almost entirely contained, and life has largely gone back to normal in-between COVID-19 waves (e.g., reopened restaurants, hotels, and shopping centers). From this perspective, first and later waves might show similar spreading patterns (i.e., initial spread among higher social status people in both waves). On the other hand, social behaviors were never entirely reset to their prepandemic state (e.g., mask-wearing, avoidance of large gatherings, and restricted flying). Furthermore, the virus survived at low levels in-between the waves in all societal strata. If anything, it probably survived better among people of lower social status. From this perspective, first and later waves might show different spreading patterns (i.e., initial spread among higher social status people in the first wave, but initial spread among lower social status people in later waves). Considering these competing perspectives, there might be no universal answer to the question of whether first and later waves initially spread similarly across societal strata. Rather, the spreading patterns of later waves more likely depend on a nation’s specific pandemic trajectory (e.g., consider the United States that never contained the virus vs. Germany that almost entirely contained the virus in-between waves).

Taken together, our findings hint at a most relevant (and tragic) dynamic: While people of higher social status may import novel viruses and cause their initial spread, people of lower social status carry the major burden once the pandemic unfolds. Importantly, this spread among people of lower social status during later pandemic phases is particularly consequential. Specifically, people of lower social status not only suffer from reduced access to health resources (which leads to a higher death toll) but will also need longer to economically recover from a pandemic crisis (which perpetuates existing social inequalities; Walker et al., 2020). Importantly, our findings highlight that these adverse effects for people of lower social status may be preventable. Specifically, if containment measures are implemented early and rigorously, it might be possible to stop pandemics before they reach the most vulnerable societal strata.