Yves here. This is an important survey on the impact of Covid-19 on different cohorts, but I wonder how much the results are undercut by being undertaken at different points in disease progress. And even though the general finding, that the young are least willing to curtail their activities (the bars! the beaches! the parties!), down here, I’ve seen more cases of mask defiance and improper use (exposed noses) among those in their 40s and 50s than any other group.
By Michèle Belot, Professor of Economics,Syngjoo Choi, Professor of Economics, Seoul National University, Egon Tripodi, PhD candidate in Economics, European University Institute, Eline van den Broek-Altenburg, Assistant Professor, University of Vermont, Julian C. Jamison, Professor of Economics, University of Exeter Business School, and Nicholas W. Papageorge, Broadus Mitchell Assistant Professor of Economics, Johns Hopkins University, Originally published at VoxEU
Almost all countries in the world have implemented drastic measures to contain the COVID-19 pandemic. This column documents the effects of the epidemic and containment measures using representative individual data on age and income from three Western and three Asian countries. Younger groups in all countries have been affected more, both economically and non-economically. Differences across income groups are less clear and less consistent across countries. The young are less compliant and supportive of the containment measures, no matter how hard they have been affected by them.
The COVID-19 pandemic has affected almost all countries in the world and has led to unprecedented measures being implemented to contain the virus. The adjustments required have had a dramatic impact on how we live, on our ability to work, and on our leisure activities. Countries have differed in their response to the epidemic. Some adopted stringent measures, such as shelter-in-place orders, while others implemented early and widespread testing and tracing procedures (Hale et al. 2020).
Within each country, society is structured in a way that not everyone would be equally affected by the measures. Younger individuals, for example, typically have more active social lives and more face-to-face interactions at work. Higher-income groups are more likely to be high-skilled and therefore to have jobs that can be performed from home (Milasi et al. 2020).
In a recent paper (Belot et al. 2020a), we document the effects of the epidemic and measures implemented in six different countries (three Western and three Asian) on different age and income groups. Data was collected in the third week of April 2020 on samples of around 6,000 individuals from the US, UK, and Italy in the West, and China, Japan, and South Korea in Asia (Belot et al. 2020b). The samples are nationally representative on three dimensions: age, gender, and income. The US data covers the four most populous states: California, Florida, New York, and Texas and the sample is also nationally representative along race.
At the time of data collection, the countries we examined were at different phases of the epidemic and had implemented different measures.
Our key questions are: How have the lives of individuals of different age and income groups been affected? And can these differences explain differences in protective behaviour and in public support for the measures implemented?
Income Dropped More Among the Young; Consumption and Leisure Dropped More Among the Rich
Across all countries, we observe a clear age gradient in economic impact: Younger groups are more likely to have experienced a drop in household income due to the pandemic, as well as a drop in consumption.
The pattern according to income is much less consistent across countries (Figure 1). In Italy and Korea, the richest households are less likely to have experienced a fall in income. The other countries (China, Japan, UK, and US) appear to have succeeded in the early months of the pandemic to shield lower-income groups from negative financial effects.
In the UK and the US, we do see an income gradient in fall in consumption: the richest are more likely to have reduced spending. This echoes evidence from credit-card transaction data reported in Chetty et al. (2020) for the US. The most likely explanation is the closure of shops and leisure-related facilities.
Figure 1 Age and income gradients on drop in household income and spending
a) Experienced loss in household income
b) Experienced drop in spending
Notes: Point estimates and 95% confidence intervals from a linear probability model of an indicator variable. This indicator denotes loss of household income during the pandemic in panel a and drop in consumption in panel b. Covariates include income quintile, age group, gender, and geographical controls. 18-25 and Income Q1 are baseline categories for age and income quintile groups, respectively.
In all countries, the younger groups (18-25 or 26-45 years of age) and higher-income groups are, in normal circumstances, more likely to engage in leisure activities with a social component. They are more likely to attend large social gatherings, visit large closed spaces (such as museums or shopping centres), go to large open spaces (such as public parks), and visit friends or family.
Since these activities were effectively discouraged or forbidden at the time of the survey, they also experienced a larger negative impact on their social life in most of the countries. However, we also see that the older groups have reduced their social interactions most and more than the younger groups.
Younger Groups Are More Affected Psychologically
Looking at negative non-financial effects, we find a similar pattern across all countries: younger groups are more likely to report negative non-financial effects (Figure 2). They are more likely to report experiencing negative psychological effects such as anxiety, boredom, and loneliness, strongly and robustly across all surveyed countries.
Differences across income groups are again less clear: All income groups report negative non-financial effects, and there is no clear gradient across countries. However, we do see a clear positive income gradient in positive effects. Higher-income groups are more likely to enjoy more free time, time with family, and reductions in noise and pollution.
Figure 2 Age and income gradients in psychological consequences
Notes: Point estimates and 95% confidence intervals from a linear regression model of number of negative non-financial effects due to the pandemic (which include: (i) boredom, (ii) loneliness, (iii) trouble sleeping, (iv) general anxiety and stress, and (v) increased conflicts with friends/family/neighbours) on income quintile, age group, gender, and geographical controls. 18-25 and Income Q1 are baseline categories for age and income quintile groups, respectively.
The Negative Experiences Do Not Explain the Lack of Compliance and Public Support for Measures
The picture emerging is one where younger people have been substantially more affected in their lives than the old. The heterogeneity across income groups is less clear and consistent across countries.
One could worry that those who have been affected more by the containment measures are more reluctant to follow health recommendations (such as wearing a mask in public spaces and social distancing). We may also be concerned that public support for restrictive measures may be lower among younger groups because of these different experiences.
We observe a clear age gradient in these behaviours and in public support for measures (younger people are less likely to wear a mask and are less supportive of the measures implemented by their governments) – especially in the UK and in the US. Surprisingly, this gradient persists even when we take into account the impact of the pandemic on people’s lives.
This implies that experiencing the negative consequences of the pandemic and public health measures play a relatively small role in explaining differences across age groups. It suggests, rather, that residual factors that correlate with age (e.g. objective health risks associated with the disease, political views) may be key drivers of attitudes and behaviours relevant for public health.
See original post for references