Skip to Main Navigation
FEATURE STORY October 22, 2020

Who on Earth Can Work from Home? A global comparison sheds light on the importance of ICT infrastructure


  • A new estimate of the share of jobs that can be performed from home takes into account access to the internet, and finds that globally, one in every 5 jobs can be done from home – but in low-income countries, it’s only one in every 26 jobs.
  • Given the variation across the globe or within countries, policies to address negative labor market impacts of COVID-19 need to be well-targeted and tailored to local conditions.
  • The overall labor market burden of COVID-19 is bound to be larger in poor countries, where only a small share of workers can work from home and social protection systems are weaker.

The COVID-19 pandemic is expected to inflict severe pain in labor markets across the globe, and policy makers will need to target support to workers who are hit hardest. To inform such efforts, a recent working paper Who on Earth Can Work from Home presents an estimate of jobs that can be done from home across the globe. What’s new is the emphasis on access to the internet, which has become even more critical during the pandemic. Çağlar Özden, co-author and lead economist in the World Bank’s Development Research Group, unpacks the analysis and explains why it matters.

The COVID-19 pandemic has changed the way we think about working from home. But the situation varies widely depending on a country’s income level and type of economy, as well as availability of critical infrastructure – in this case the internet. What was your motivation for this research and what did you find?  

Analyzing labor markets is the bread and butter of the World Bank’s work. The road out of poverty passes through finding well-paying, secure and productive employment. We constantly ask – how health outcomes and education impact labor markets, and how infrastructure or macroeconomic policies affect labor market outcomes across the world. And what the impact is on inequality.

When COVID-19 hit, we wanted to measure the impact of the pandemic on labor markets and how it differs across countries. Policy makers need to counter negative effects inflicted by shutdowns, social distancing, and other efforts to contain the disease, and understanding whether jobs can be performed from home helps to determine labor market vulnerability. Also, many governments in developed countries enforced lockdowns to contain the virus, but we questioned the feasibility of doing the same in low-income countries.

There were already several papers that had come out, but many focused on high-income countries with less emphasis on low-income countries. We started by applying the metrics they used to low-income countries, examining the type of tasks carried out by the worker, and explored the extent of working from home in developing countries.

We then asked how access to infrastructure, in this case the internet, affects labor markets. Before COVID-19, research on remote work looked mostly at the composition of work and types of jobs across countries. Failing to account for internet access causes overestimation of the number of jobs that can be performed from home across the globe by around 27 percent on average.

The disparity we found is striking: In high-income countries, it’s 1 in 3 jobs that can be done from home while in low-income countries it’s only 1 in 26 jobs. The overall labor market burden of COVID-19 is bound to be larger in poor countries, where only a small share of workers can work from home and social protection systems are weaker.

You also found that COVID-19 has made income inequality worse. Does lack of access to the internet play a role here?  

In terms of income inequality, we found that the type of job one has and their access to the internet matter – a lot. And there are big differences across the world or within countries.

In high-income countries, where the internet is readily available, the type of job matters more. Educated people tend to have high-skilled jobs that are amenable to working from home. Low-skilled jobs that are temporary or noncontractual are hit harder.

In developing countries, the internet is more relevant because access is highly unequal across income groups, education group levels and geographic regions.

Lack of internet access creates another layer of distributional inequality, especially in middle-income countries. While rich people tend to have internet access, poor people may not have access at home. Even if you are a relatively well- educated person whose job can be done from home, internet access determines your fate.  

On the other hand, in low-income countries where internet access is limited, the situation is somewhat reversed. The majority of people work in low-skilled service jobs like agriculture, and people with high-skilled jobs who depend on internet access in the office are generally most affected.

One of the challenges you encountered during this research was lack of good data on ICT infrastructure, especially in low-income countries. Why is this so important?  

As I mentioned, one of the main reasons why we wanted to write this paper is to inform government policies and to make the case for more targeted interventions.  Policies cannot be uniform. They need to take into account local conditions.  

But the absence of reliable data on IT infrastructure from some low-income countries was a much bigger problem than we thought. Labor force data or infrastructure data are important in so many different contexts. We need better data on digital access and services, which is one of the determinants of poverty and inequality in the information age.

The digital divide is certainly not a new concern, but the working paper underscores the significance of ICT infrastructure during this pandemic for labor market outcomes. How are you taking this research forward?  

Yes, ICT infrastructure is important for high quality jobs, especially for younger people, so we have to make it our priority. We have started a new project on cell phone access in India, especially in rural areas, and how it affects income levels and labor mobility among others.

My colleagues are also working on research to help governments conduct more targeted interventions. When people are staying at home, and you are planning an unemployment subsidy, which group - age, education level, occupational sector - should you be targeting? This is an important question.

This research has also shown us what kind of data we should be collecting and how. Nobody would have guessed, a year ago, that we would be in this situation but it’s important to enable us to answer questions on how to address the negative impacts of COVID-19.