The start of the COVID-19 pandemic sent global economies into crisis. COVID-19 cases were increasing each day, and the unemployment rate was skyrocketing. Thankfully, unemployment rates and COVID-19 cases have lowered now relative to the pandemic's onset in 2020. In the summer of 2020, I thought it would be important to get a clearer picture of how each country’s job market would be affected. The result of the data gathering and consequent analysis was an Employment Resilience Index. I constructed this index to estimate how each country’s job market would recover in a post-pandemic world. A higher score on the index meant a higher likelihood of a quick recovery for a country's labor markets, whereas a lower score meant that the country's job market would likely recover at a slower pace. Of course, quick and slow here are relative to the recovery rates of other countries and do not necessarily provide an estimate with a time range.
The Data
School and workplace closures stemming from the COVID-19 pandemic led to companies downsizing, job loss and consequently a spike in unemployment rates. Throughout the pandemic and amid these closures, multilateral organizations such as the World Health Organization and Center for Disease Control urged governments globally to establish social distancing measures. Effective social distancing would mean keeping COVID-19's rate of spread low, which would likely translate to establishments re-opening sooner rather than later. Countries that practiced effective social distancing could re-open their workplaces sooner, and therefore begin their recovery process earlier, gaining a head start on other economies on the road towards recovery.
Given this information, the first question I asked was to what level were governments establishing social distancing measures in their respective countries? I sourced this information by looking at the Stringency Index by Our World In Data. This index looks at how governments responded to the pandemic through efforts such as school and business closures and assigns them a score. Particularly, the stringency index uses nine metrics to calculate the intensity of each government's response in enforcing social distance measures. These include workplace closures, cancellation of public events, restrictions on public gatherings, closures of public transport, stay-at-home requirements, public information campaigns, restrictions on internal movements, and international travel controls such as border closures.
Beyond policy, I also asked: were people in the countries actually social distancing? Governments could write the rules and enforce them at an institutional level. At the end of the day, however, it was up to each person to follow through on the decision to stay home. To better understand this I sourced mobility data from Google’s Community Mobility Reports. These reports provide data on people's movement across Geographical areas using geolocation data categorized into key areas of interest. These include residential areas, workplaces, parks, retail and recreation areas, transit stations as well as groceries and pharmacies. The reports are created while adhering to the strictest privacy protocols. For my data analysis, I chose to focus residential and workplace mobility: how were people moving in relation to residential areas and workplaces? Were they actually staying home or were they visiting each other often? What proportion of people in each country were still routinely going to their offices? Analyzing the mobility data helped answer these questions.
As the pandemic continued, more and more people started remote work. I also collected estimates on the percentage of jobs that could be done from home in each country. This data was collected at the University of Chicago. They are referred to as teleworkable jobs. The proportion of teleworkable jobs in each country depends on its industry composition in three main employment sectors: agriculture, industry and services. I categorized each of the occupations listed in the University of Chicago data as being either in the agriculture, industry or services sector. I then used data from the International Labor Organization to find the proportion of teleworkable jobs for each country according to the sector.
The type of jobs countries create for their people also largely depends on their GDP. I estimated each country’s level of income before the pandemic with 2018 GDP per capita data as it was recent at the time. I sourced this data from the World Bank.
For each country, I also collected data on the number of people connected to the internet. At the time, tracking healthy populations through Covid-19 cases and death rates was essential. I also included these statistics in the dataset.
I made a principal component to create the index through Principal Component Analysis (PCA). The results can be seen in the map below. You can also view the entire data story on Tableau.
