Within Europe, Italy has been at the forefront in the war against COVID-19. The first country to introduce social distancing, Italy is now struggling to devise a safe way out. Various schemes have been proposed for the transition strategy, each choosing a different segment of the population as a candidate for returning to workplaces first, among them: the tested, the immune, the healthier, and the residents of the less affected South. Ichino et al. (2020) advocate for starting with the young.
The idea of sending women first has also been proposed. Worldwide, as in Italy, men are dying at unquestionably higher rates than women. However, when making a decision about sending someone back to work, the likelihood of their contracting the virus is as important as their likelihood of dying from it, due in part to the consequences for the further spread of the contagion.
At first glance, Italian women appear to be less susceptible than men to COVID-19, as they represent 47.8% of the diagnosed cases (as of 9 April 2020). However, sex-and-age disaggregated data reveal that the contrary is true for Italian women aged 20-49 (Table 1). For instance, in the 20-29 age group, 56.5% of the diagnosed cases are female; only after age 50 are women outnumbered by men (Riccardo et al. 2020, Task force COVID-19 2020). A comparison between the sexes is therefore meaningful only if age is also taken into consideration.
Table 1 Diagnosed COVID-19 cases by sex and age. Italy, 9 April 2020
Note: The table does not include cases of unknown sex.
Source: Task force COVID-19 (2020).
Sex, age, and work
Virologists and epidemiologists have proposed several explanations for the apparent advantage that women hold in the face of COVID-19, ranging from biological and genetic factors to epidemiological and behavioural ones (Wenham et al. 2020). None of these explanations has accounted for age or for age reversal.
Without denying the concurrence of the aforementioned correlates, socioeconomic factors should also be evaluated. My hypothesis is that the occupation profile of Italian women can explain both the gender difference in susceptibility and its age reversal (Bertocchi 2020). National statistics by ISTAT reveal that the age groups in which women are more susceptible than men are those in which the female employment rate is closer to the male rate. Despite the fact that women are less likely than men to be employed at any age, in 2019 the employment gender gap is lowest for the 25-34 age group; remains relatively contained until age 54; and increases markedly at 55-65, which corresponds to the age at which working-age women acquire the largest advantage over men in terms of risk of contagion. Only more detailed individual data will allow for a formal test of this hypothesis, but it is legitimate to conjecture that the higher susceptibility of working-age women relative to working-age men is associated with exposure to contacts through the workplace.
But why is it that younger women (aged 20-49) are more susceptible than men, even though their employment rate remains lower? Setting aside, once again, non-socioeconomic factors, one explanation rests on the fact that women are much more represented in jobs that expose them to a higher risk of contagion. Barbieri et al. (2020) show that the two sectors that are most exposed to infections and diseases – namely, health and education – employ a disproportionate fraction of female workers. In the health sector, where women represent 70% of the workforce and 61% of them are below age 50, the disease exposure index is 54 (against 8.8 for the Italian economy as a whole). In education, the index is 15.5; women are 75% of the workforce and 47% of them are below age 50. Barring hotels and restaurants (where young women are also overrepresented), education and health score highest, even in terms of a physical proximity index – another aggravating factor. Consistently, women’s representation is close to men’s in sectors not subject to lockdown, health among them (Casarico and Lattanzio 2020). While education workers have been sheltered since the earliest phase of the lockdown, they will be on the front line when schools will open again.
The above data on the demographics of diagnosed cases and sectoral employment do not account for the feminisation of the overall population that starts to be noticeable after age 40 and becomes crucial past retirement age. Taking feminisation into account would make the female advantage reported in Table 1 at age 50-79 even more noticeable (women outnumber men past age 80 because women are overrepresented in that age group). But what is driving the stark reversal of the risk of infection when gender differences in employment become negligible? It must depend on the same factors that determine women’s higher life expectancy in general, and men’s higher fatality rates specifically for COVID-19. The medical literature points to a smaller incidence for women of co-morbidities such as chronic heart and lung diseases. Other relevant risk factors that are in turn associated with higher risk of co-morbidities in men, such as smoking and drinking, are as ingrained in cultural norms as attitudes toward women’s work (Baranov et al. 2020, Purdie et al. 2020).
A cross-country comparison
Though data by sex and age are scant, those available have been assembled by Global Health 50/50 (last accessed on 10 April), a research initiative housed in the UCL Centre for Gender and Global Health. Preliminary evidence for the 28 countries that collect data by gender is mixed. Women represent the minority of diagnosed cases in only ten countries (36%, Italy among them), while the opposite is true in twelve countries (43%), and it is a tie in the remaining six (21%). Only eight countries, six of them in Europe, provide information stratified by both sex and age (the other two countries are Australia and Peru).
Table 2 European countries providing sex-and-age disaggregated data. 6-7 April 2020
*Per 100,000 women.
Sources: Global Health 50/50 (accessed on 10 April 2020) for confirmed cases, % women, and age groups; OECD for female employment rates.
To mitigate differences in demographic, labour market, and health system structure, Table 2 summarises information on confirmed cases for the six European countries. Strikingly, Italy stands alone with a share of female cases below 50%. In the other European countries, women outnumber men (with a tie in Germany). The table also reports which age group in each country shows a prevalence of women among confirmed cases (within working age groups, over 100,000). Once again, Italy stands out with a prevalence of women for the age ranges up to age 49. In other words, in the 20-29, 30-39, and 40-49 age groups, Italy reports more cases for women than for men, per 100,000 women and men respectively. By contrast, in the other five countries, women remain in a more vulnerable position up to age 59. Though this evidence is only suggestive, it can be read in light of labour market factors since (as shown in the last column), Italy exposes a huge cross-country gap in the female employment rate, as low as 50.1%; the next lowest is 57.9% for Spain. The likely reason that women are more vulnerable in other countries is that they work more, and up to an older age.
Conclusion
It would not be the first time that women replaced men in the workplace. It happened during WWI and WWII, when men were fighting at the front. Nevertheless, at this stage, there is no evidence that working women find themselves in a better position than men in the fight against COVID-19, at least in terms of susceptibility. In fact, the opposite is true. It follows that a policy counting on women as a replacement for men, as the lockdown is lifted, might well make the problem worse rather than solving it.
Several questions still need answers. First, a proper causal analysis is prevented by data availability. We cannot even rule out the possibility that selection may be present in the current data, since the potentially high fraction of asymptomatic individuals implies that those diagnosed may not reflect the actual distribution of the infected population by sex and age. Further, definitions of recorded cases vary widely by country, and the extent to which social distancing has been imposed is far from uniform.
To shed light on the gender differences in susceptibility to COVID-19 and their interaction with socioeconomic factors, the relative feasibility of working from home also needs to be accounted for. Moreover, understanding the determinants of sex-and-age specific transmissibility requires information on co-residence patterns. In principle, given the high contagiousness of the infection and given that working age roughly coincides with the age at which individuals tend to live with a partner, a gender gap in the infection rate should not occur. The fact that we observe one may depend both on COVID-19 specific characteristics, such as a potential gender gap in the proportion of asymptomatics, and on underlying factors, such as the cross-country variation in co-residence and age homogamy patterns.
While the focus here has been on susceptibility to the disease, the impact of the COVID-19 outbreak on gender equality extends to much broader realms, including the differential outcomes in terms of job losses caused by the economic downturn, as well as the implications for the division of labour within households, in both the short and long run (Alon et al. 2020a 2020b).
References
Alon A, M Doepke, J Olmstead-Rumsey, M Tertilt (2020), “The impact of COVID-19 on gender equality”, COVID Economics: Vetted and Real-Time Papers, Issue 4 London: CEPR Press.
Alon A, M Doepke, J Olmstead-Rumsey and M Tertilt (2020), “The impact of COVID-19 on gender equality”, VoxEU.org, 19 April.
Baranov V, R De Haas and P Grosjean (2020), “Men”, VoxEU.org, 2 April.
Barbieri T, G Basso and S Scicchitano (2020), “Italian workers at risk during the COVID-19 epidemic”, INAPP Working Paper no. 46/2020, 12 April.
Bertocchi, G (2020) “Ma davvero le donne sono più resistenti al COVID-19?”, lavoce.info, 8 April.
Casarico A and S Lattanzio (2020), “La demografia del lockdown”, lavoce.info, 7 April.
Global Health 50/50 (2020), “Sex, Gender and COVID-19: COVID-19 Sex-Disaggregated Data Tracker”, Centre for Gender and Global Health at University College London, accessed 10 April.
Ichino A, G Calzolari, A Mattozzi, A Rustichini, G Zanella, M Anelli (2020), “Transition steps to stop COVID-19 without killing the world economy”, VoxEU.org, 25 March.
Purdie A, S Hawkes, K Buse, K Onarheim, W Aftab, N Low and S Tanaka (2020), “Sex, gender and COVID-19: Disaggregated data and health disparities”, BMJ GH Blogs, 24 March.
Riccardo F, M Ajelli, XD Andrianou, A Bella, M Del Manso, M Fabiani, S Bellino, S Boros, A Mateo Urdiales, V Marziano, MC Rota, A Filia, FP D'Ancona, A Siddu, O Punzo, F Trentini, G Guzzetta, P Poletti, P Stefanelli, MR Castrucci, A Ciervo, C Di Benedetto, M Tallon, A Piccioli, S Brusaferro, G Rezza, S Merler and P Pezzotti, for the COVID-19 working group (2020), “Epidemiological characteristics of COVID-19 cases in Italy and estimates of the reproductive numbers one month into the epidemic”, medrxiv, 11 April.
Task force COVID-19 del Dipartimento Malattie Infettive e Servizio di Informatica, Istituto Superiore di Sanità (2020), Epidemia COVID-19, Aggiornamento nazionale: 9 April.
Wenham C, J Smith and R Morgan, on behalf of the Gender and COVID-19 Working Group (2020), “COVID-19: The gendered impacts of the outbreak”, The Lancet 395 (10227): 846-848.
