The Employment Impact of COVID-19 on Vulnerable Populations
Analyzing country-specific vulnerability to the employment crisis caused by the COVID-19 pandemic.
Introduction
In response to Covid19 governments all over the world scrambled to enact policies to protect the health of their citizens. These policies were primarily directed towards mitigating the health crisis, but they used a rapid rise in unemployment almost everywhere in the world. Governments that did not pre-empt this crisis in employment were forced to enact additional policies to mitigate the employment crisis.
In this project by Omdena, we used publicly available data to look at countries around the world and analyze their vulnerability to the employment crisis caused by the COVID-19 pandemic.
Context: Lockdown and economic policies
In order to analyze the situation in a particular country, and be able to compare policies between them, the University of Oxford released a dataset comprising the government responses by country, updated on a daily basis, and showing the changes in three main groups of policies: Containment and closure, economic response or fiscal measures, and health-related policies. [1]
Among the Confinement policies, there are three specific ones that, according to the ILO [2], have had the most impact on the world of work. These three policies are:
● C2 (Workplace closedown)
● C5 (Closedown of public transportation)
● C7 (Restriction in the internal movement of citizens)
Each of these have different stringency values, ranging from 0 (no enactment of restriction) to 2 (or 3 in the case of C2) being the highest level a mandatory closedown with few exceptions. With these levels in mind, countries are categorized according to the level of lockdown they have.
The categories are:
Full lockdown: All three policies are in mandatory closing.
Partial lockdown: At least one of the policies is mandatory (the rest are working either just on a recommend closing level or are not enacted).
Weak lockdown: None of the policies is mandatory.
The Stringency of the policies applied by a country determines a stringency index, which is calculated by the creators of the policy dataset [1]. This index goes from 0, which is no policies being taken, to 100, meaning the country is at its maximum stringency and lockdown level. This index is updated on a daily basis, creating what we refer to as the Stringency curve.
Variables to quantify vulnerability
Since this article focuses on those populations that have a higher vulnerability regarding unemployment during the COVID-19 crisis, some indicators need to be introduced that help represent the vulnerability of the population within a country. The indicators considered here are the following.
● High Impacted Sector Exposure. The 14 economic sectors considered by the ILO, have been aggregated into 5 categories according to the level of impact that the current economic crisis has brought to them in particular (measured from real-time and financial data for each sector from the ILO). (Source: ILO Monitor. COVID-19 and the world of work. Second Edition [3]). From this, high-impact sectors are those that lie on the most negatively impacted side of this scale (these are Accommodation and food services, Real estate, Manufacturing, and Wholesale and retail trade). It is worth mentioning that other sectors, like Education and Agriculture, are also impacted by the crisis, but in general to a lower extent than those mentioned above.
● Inequality-adjusted Human Development Index (IHDI). The IHDI is an indicator available currently for 150 countries, that takes into account the country’s average achievements in health, education, and income. Furthermore, it weights these three dimensions with their distribution across the population (their level of inequality). It can, therefore, be considered as a general measure of the resiliency of each country to adverse effects on health, education, and income. (Definition is taken from the United Nations Development Program [4]).
● Informality Rate. The informality rate is the percent of people participating in the informal economy, out of the total labor force of that country. There are several criteria defined by the ILO to consider an employee as being an informal worker, such as; No contributions to social security and not an entitlement of a worker to paid annual leave and or sick leave. (Source: ILO. Women and Men in the informal economy: A statistical picture [5]).
● Stringency Index. This index measures the severity of implementation of the affecting policies, as mentioned before.
In order to be able to jointly use these variables to compare across countries, we define two indicators:
Impact Weighted Informality Rate: % share of high impact sectors * Informality rate
This indicator aims to define vulnerability in terms of the exposure of a country to the most adversely affected economic sectors and the share of workers in the informal economy that are likely affected by those adverse effects. The higher the value of this indicator, the higher the vulnerability.
Impact Weighted IHDI: % share of high impact sectors *(1-IHDI)
This indicator aims to define vulnerability in terms of the exposure of a country to the most adversely affected economic sectors and the resilience of that country to those adverse effects. The higher the value of this indicator, the higher the vulnerability.
A global perspective
Considering the indicators of the vulnerability described above, we dive into how these look on a global and then on a regional basis.
In Figure 2, countries that appear colored are those that, as of May 1, 2020, are reported to be under full lockdown measures by the Oxford policy database, based on the definition provided above. Furthermore, these countries have little or no income support from their governments (meaning less than 50% or no loss of wage is being compensated by government financial support). As of May 1, 2020, most governments globally have attained a plateau in their stringency curve, which means most of the strict lockdown measurements are already enacted and have been for at least a month in most countries. In figure 2, the color of the country represents its IHDI index, and the intensity of it represents the share of the population working in highly impacted sectors. The more intense (darker) the color, the higher the people that work in sectors like Accommodation, Manufacturing, Real Estate, or Retail, which are those most severely affected by the crisis.
The scatterplot in Figure 3 shows the same countries shown on the map, describing on the x-axis the share of the population in highly impacted sectors, and their stringency index on the y-axis. The sizes of the circles in the scatterplot represent the joint indicator Impact Weighted Informality Rate defined above. The following table represents the top 15 countries ranked by both this indicator and the Impact Weighted IHDI, in descending order of value (vulnerability).
One point to highlight is that some of the countries appear almost at the top of both lists (Nigeria, Guatemala, Honduras), which means that they are vulnerable considering all of the factors we consider in our vulnerability definitions, i.e. these countries are in full lockdown, have a high informality rate (for these countries it goes beyond 80% of the labor force), a low or medium human development, and a relatively high share in highly impacted sectors.
As can be seen, Nigeria is an African country among those in full lockdown and no government income support, that has the biggest share of the population working in highly impacted sectors, and also an overall informality rate of over 80% of its population (as can be seen later in Figure 4). Furthermore, it is classified as having a low human development index, based on the IHDI, which makes it less resilient as to overcome a potential economic crisis and massive amounts of its workforce either losing wages or their main sources of income altogether. This is exacerbated by the fact that neither formal nor informal workers have significant economic support in that country while being subjected nevertheless to strict lockdowns, which drastically prevents workers (and in particular informal workers) from going to work. Even though these people might still need to go out to get an income, the country’s conditions make it very difficult.
Other countries on the list above, as seen in the scatterplot (figure 3), have a similar circle size, which means that they are relatively close in terms of both informality (higher than 70% of the workforce in all these countries) and share of workers in highly impacted sectors (between 30–45%). Since the lockdown measures of Nigeria closely resemble those of the other countries on the list above (full lockdown in all cases), they all lie on the highest side of the vulnerability spectrum, being Nigeria nonetheless the only one among them on the Low human development category. Most African countries have a lower share of people working in high impacted sectors, even if they have some of the highest informality rates globally (which is still a very significant factor to be considered while addressing policies to them).
As can be seen in Figure 3, many European countries have a high share of their working population in high impacted sectors. However, many of them are not even seen on the scatterplot, because, as is the case for the USA, Canada, and Australia, they have had either milder lockdowns (like Sweden) or higher income support to workers.
This brings us to the conclusion that it is not either one of our vulnerability factors alone, but the combination of them, that makes us consider a country highly vulnerable or more vulnerable with respect to others. In an economy that has a high IHDI (more resiliency, the capacity of bigger fiscal investments, and income support to workers) and where the rate of informality is smaller, formal unemployment may still be increasing, but several tax relaxations and income support might be helping most people. In the opposite case (those shown in our list of most vulnerable), governments are not capable or willing to compensate for most of the working population that is being forced to stay home, and a much higher share of the population depend on a daily or short-term income to survive.
Countries where the informality rate is high, IHDI is low-medium, and many people work in a sector with high economic impact, should, therefore, be considered much more deeply the adverse consequences of their strict lockdowns in the economic sectors that are being more negatively impacted, and policies should be addressed accordingly, to prevent a bigger shock than what is inevitable.
Finally, even though we treat the sector impact in percentages relative to the workforce of each country, the population factor can play an important role here, particularly in Southern Asia, where India is located. The high population (as well as population density) and high informality (which yields a small percentage of the working population contributing to taxes and social security, therefore reducing public resources to provide income support to workers in the informal economy) adds up to the vulnerability of this population.
In the following sections, an overview of the situation in some specific regions of the world with the highest informality rate is presented.
A Regional Overview
Africa
Figure 4 shows the African countries that, as of May 1, 2020, are under full lockdown measures, as well as their workforce and informality distributions across broad sectors (industry, services, and agriculture). From these, Only Chad reports having a “Low” (less than 50% replacement of the lost income) income support policy for both formal and informal workers, according to the Oxford COVID policy changes database. The other countries shown report no income support from the government.
The total informality share in these countries lies close to 80%. It can be seen that agriculture is one of the most relevant economic sectors in these regions (especially in Western and Sub-Saharan Africa). As shown in Figure 5, except for Nigeria, countries in this region have smaller shares of highly impacted sectors than most of the other countries worldwide. However, because of its relevance and its informal nature, agriculture should be particularly considered in terms of policymaking.
Asia
Figure 6 shows the Asian countries that, as of May 1, 2020, are under full lockdown measures, as well as their workforce and informality distributions across broad sectors (industry, services, and agriculture). All these countries lie in the South-eastern or Southern regions of Asia. From these, only Pakistan reports having a “Low” (less than 50% replacement of the lost income) income support policy for both formal and informal workers, according to the Oxford COVID policy changes database. The other countries report no income support from the government.
As shown in Figure 6, the total informality share in these countries lies close to 80%. Also, Figure 7 shows that these countries also have a bigger share of the labor force in highly impacted sectors than the previously seen African countries in full lockdown.
A regional comparison
Finally, after seeing the countries that classify as most vulnerable, considering their full lockdown state, and their informality and exposure to highly impacted sectors, we would like to present an overview of how this looks like across regions worldwide.
Table 2 shows the world’s regions (whose country distribution is based on the same distribution made by the ILO in [5]), their average share of highly impacted sectors, average informality rate and their product, which is our joint indicator Impact Weighted Informality Rate. Table 3 shows the same ranking, but considering the indicator Impact Weighted IHDI. In both tables, the third column is the result of the product of the first two, showing how they are joint indicators of our considered vulnerability factors. The tables show only countries under full lockdown and with little or no income support, to reflect countries within these regions where the working population is more restricted and less supported financially through the crisis.
From the tables, a few things can be seen:
● Both joint indicators used to compare the vulnerability across regions for countries under full lockdown show similar results, showing that there is a certain correlation between regions and countries with high informality rate and low IHDI, which reflects more inequality, less resiliency to deal with the economic crisis, and a serious threat of a large sector of the population being possibly disregarded by the measures taken to combat the crisis. The main outlier to this trend is South-Eastern Asia, where for instance Indonesia lies in the category of high human development according to the IHDI, but still has a large total informality rate (85.6%).
● Some regions (Central and Eastern Asia, North America, Western and Northern Europe, and Australia) are not shown. This is because there is no country in either of these regions, with both full lockdown policies and no significant income support from the government. All countries in regions not represented in Table 2 have either milder lockdowns, or significant income support to workers, or both.
● Western Africa, (Represented mainly by Nigeria in both tables) is the region that, according to the Impact Weighted Informality Rate, shows the most significant vulnerability, combining extremely high informality rates (92.9%) with significant shares of the population in highly impacted sectors (36.42% is a high share, considering that the region with the highest figure in southern Europe, with 44.36%). Informal workers seem to be in deep trouble to get an income because of the lockdown policies, jeopardizing dramatically the region’s economy and level of employment (being the informal employment the most relevant figure for this case).
● Other African regions (Sub-Saharan Africa and Northern Africa) lie lower particularly in Table 2. In the Case of Northern Africa (Tunisia, Algeria) the informality rate is significantly lower than in other African regions. This is important while structuring policies, because many stringency relaxation policies usually address the formal sector, leaving those in the informal economy more affected and disregarded. As for Sub-Saharan Africa, even if the informality rate is a bit higher than that of Western Africa (93.3%) this region has a lower share of highly impacted sectors. In this region, one of the main factors to consider is how policies, both lockdown, and economical ones, affect the agriculture sector, which represents a high share of workers, almost all of them informal.
● Countries under full lockdown in Central America, South-Eastern Asia, and Southern Asia, also show a high vulnerability. Even though at a bit smaller percentages than Western Africa, these regions show a very significant share of informality (around 80% in all three) and a high share of people working in high impacted sectors.
● Finally, two regions in Europe do have countries under full lockdown and with little or no income support reported as of May 1, 2020. These regions lie at the very bottom of both tables mainly because of their relatively low informality rates and high IHDI indices. The fact that formal economy prevails in these regions makes workers vulnerable to the crisis more easily accounted for by policies taken by their respective governments. Doing this poses much more of a challenge in the rest of the regions described above.
Summary
This article discusses the vulnerability of countries around the world to employment loss due to COVID-19 policies. In order to quantify vulnerability, four factors are considered viz. stringency of lockdown policies, the share of informal workers in the workforce, the IHDI, and the share of adversely impacted sectors in the economy. For countries under strict lockdown, vulnerability to employment loss is quantified by the ‘Impact Weighted IHDI’ and ‘Impact Weighted Informality’. According to these indices, countries like Nigeria, Guatemala, Honduras, Indonesia, Bolivia, and Bangladesh emerge among the countries most vulnerable. On a regional basis, Western Africa, Central America, Southern Asia, South-eastern Asia, and South America emerge as most vulnerable.
It should be mentioned here that other measures of vulnerability such as the multidimensional poverty indices, competitive industrial performance, etc. can also be considered. Also, countries with less stringent lockdown policies and/or substantial government support for the economy were excluded from this analysis. Every country in the world has been touched by this pandemic, and every country has been dealing with their own issues and vulnerabilities. This work touches on a few aspects of those vulnerabilities and attempts to shed some light on which countries/regions are expected to be especially hard-hit from the economic effects of this pandemic.
Credits — Omdena AI pandemic challenge
The work presented in this article is part of a task dedicated to analyzing the effects of the current COVD crisis and unemployment, considering particularly those that are more vulnerable in the context of the Omdena AI pandemic project. This work would not have been possible without the help of all team members. Special thanks to the domain experts that helped to shed light on how to try to quantify vulnerability for data analysis.
For the development of this article, a dashboard was created in the context of this Omdena project, with the purpose of visualizing the situation of all countries where data is available in terms of the vulnerability variables described here, further considering all economic sectors, and geographical regions.
The link to the dashboard is provided here.
References
[1] https://www.bsg.ox.ac.uk/sites/default/files/2020-05/BSG-WP-2020-032-v5.0_0.pdf
[2] https://www.ilo.org/wcmsp5/groups/public/@dgreports/@dcomm/documents/briefingnote/wcms_743146.pdf
[4] http://hdr.undp.org/en/content/inequality-adjusted-human-development-index-ihdi
[5]https://www.ilo.org/wcmsp5/groups/public/---dgreports/---dcomm/documents/publication/wcms_626831.pdf
