Updated estimates of the impact of COVID-19 on global poverty


This page in:

Note: A more recently updated version of the numbers and chart below is available here.

In April we estimated that COVID-19 is pushing between 40 and 60 million into extreme poverty. Since then, the epicenter of the pandemic has shifted from Europe and North America to the global south. This has increased the death toll in low- and middle-income countries, induced longer shutdowns, and increased the economic costs of the pandemic. As a result, our estimates of the impact of the virus on global poverty have shifted as well. 

Using the newly launched growth forecasts of June from the Global Economic Prospects, we can update the estimate of the pandemic’s impact on global poverty. The new growth forecasts contain two scenarios—baseline and downside—allowing us to explore two different scenarios for how the pandemic may impact poverty. The baseline scenario assumes that the outbreak remains at levels currently expected and that activity recovers later this year, while the downside scenario assumes that outbreaks persist longer than expected, forcing lockdown measures to be maintained or reintroduced. Should the downside scenario materialize, vulnerable firms would exit markets, vulnerable households would sharply reduce consumption, and several low- and middle-income countries would see heightened financial stress. The baseline scenario has global growth contracting by about 5% in 2020 while the downside scenario presents a global growth contraction of 8% in 2020. 

Utilizing the same method as in our last post, we estimate the impact of COVID-19 on poverty by comparing poverty projections that use the new GDP forecasts with poverty projections that use the GDP forecast before COVID-19 took off, in this case the GEP forecasts from January. Under the baseline scenario we estimate that COVID-19 will push 71 million into extreme poverty , measured at the international poverty line of $1.90 per day. With the downside scenario, this increases to 100 million.

Projecting what happens in 2021 and beyond comes with even more uncertainty. The GEP forecasts expect that global economic output will increase by about 4% in 2021, yet our poverty forecasts suggest that the number of people living in extreme poverty will be broadly unchanged between 2020 and 2021. How can that be? A lot has to do with the growth rates of the countries with the most poor. Nigeria, India, and the Democratic Republic of Congo -- three countries which we project are home to more than a third of the world’s poor—are predicted to have per capita growth rates in real GDP of –0.8%, 2.1% and 0.3%, respectively. With population growth rates of 2.6%, 1.0% and 3.1% this is hardly enough for sustainable decreases in the poverty headcount. 

In the last blog we suggested that Sub-Saharan Africa might be hit hardest. The new GEP forecasts give a particularly sobering picture for India, which is home to many of the world’s poor. As a result, though the picture is broadly unchanged for Sub-Saharan Africa compared to our last update, South Asia may see a larger increase in the number of poor as a result of COVID-19. A big caveat to this finding is that the latest poverty estimates we have from India are from 2011-12. This makes it very difficult to get an accurate picture of poverty there before the pandemic took off, let alone a picture of poverty today. 

At higher poverty lines, the regional distribution of the added poor changes markedly. Of the 176 million people expected to be pushed into poverty at the $3.20 poverty line under the baseline scenario, two-thirds are in South Asia. Of the 177 million expected to be pushed into poverty at $5.50, many newly poor are in East Asia & the Pacific, and few are in Sub-Saharan Africa, simply because few people there have living standards at this level. 

Note that the number of new poor at $1.90 is not a subset of the new poor at higher poverty lines. If someone in Sub-Saharan Africa sees their daily income fall from $2.00 to $1.50 as a result of COVID-19, they will be an added poor person at the $1.90 line but not at the $3.20 line, where they were counted as poor both before and after the pandemic.

As we expressed last time, and as these revised estimates make clear, poverty projections carry a lot of uncertainty and are likely to develop further as more information becomes available and as the pandemic develops. More analysis on how the numbers change with changing growth rates, inequality, and assumptions about how economic growth translates into poverty reduction is available in a working paper. In that working paper, we also look at what all of this means for our ability to meet the first target of the first Sustainable Development Goal – to end extreme poverty by 2030.

1 We measure the impact of COVID-19 on poverty as the difference in millions of poor in 2020 with the two growth vintages, to which we subtract the difference in millions of poor in 2019 with the two growth vintages. The latter assures that changes in poverty due to revisions to 2019 growth rates, which cannot have been due to COVID-19, are not driving the results. Here, this becomes (684-595)-(632-614) =71 million.


July 09, 2020

Thank you for these updates to your projections. Will you make the whole dataset with country projections available that you refer to in the text ?

August 07, 2020

Hi Dean,

Thank you for your comment. You may reach the team directly for this concern: data@worldbank.org

July 09, 2020

We look forward to establishing exchanges and cooperation with the World Bank and looking forward to getting the latest World Bank newsletters.

Ambe Festus N.M.
July 09, 2020

I agree

July 09, 2020

You are relating poverty to GDP. You should do it to Gini index to make it more realistic.

August 07, 2020

Dear Elvio,

We definitely agree that relating it to the Gini index is important. We do so in the accompanying working paper, which you can find here: http://documents1.worldbank.org/curated/en/765601591733806023/pdf/How-M…

July 09, 2020

How can poverty be reduced

Chadli Belarbi
July 09, 2020

The Impact of the COVID-19 on Tunisia Economy and the Social impact on the health Situation from early sign of the Corona Virus. Not only Tunisia a small country of 12 million citizen with 70% are below the poverty line, The new Government lead by the President of Government Ilyas Fakhfakh who was stunned on the cost of the fight against the invisible disease from the poor equipment used to testing and the lack of the regular suppliers from the face masks to the cloves to available data of the people who have symptoms of the corona disease and who don’t. The budget already set by the financial committee was done prior to exemption of the disease s to mainly large Tunisian cities, especially those who are in the South or the North West of Tunisia

R. Friedman
July 09, 2020

Would love to see this data in terms of population percents by region.

Ali Kombo Ali
July 09, 2020

NGOs should be given chances in development projects

Prof Morgan Chetty
July 09, 2020

I am most interested in data on how covid 19 will affect the African continent
This article does cover a lot. Do we have more data on South Africa

August 05, 2020

exactly South Africa needs more assistance many people suffer during this period.

August 07, 2020

Hi Prof Chetty,

Thank you for your comment. You may reach the team directly regarding the availability of the data you are asking for here: data@worldbank.org

July 09, 2020

People in the poor countries do work on daily wages also they are in suffering from poverty on its high peak even before the covid-19 dear really at formidable risk

Tim Conway
August 05, 2020

Hi Christoph, hi all. This is really useful and interesting - but there's one bit which is puzzling me so I'd appreciate your help. How can the total numbers which will fall into poverty at the $3.20 and $5.50 lines be so similar (176m cf 177m)? Does this not imply very small numbers of people in the band between $3.20 and $5.50? I get the specific example in your point "Note that the number of new poor at $1.90 is not a subset of the new poor at higher poverty lines. If someone in Sub-Saharan Africa sees their daily income fall from $2.00 to $1.50 as a result of COVID-19, they will be an added poor person at the $1.90 line but not at the $3.20 line, where they were counted as poor both before and after the pandemic." - but not sure how to interpret it in the aggregate. If the new $1.90 poor are not a subset of the new $3.20 poor (and similarly the new $3.20 poor are not a subset of the new $5.50 poor), does that mean that the total increase in global population under $5.50 = 71 + 176 + 177 = 424 m? In addition to the numbers of new poor, could you maybe add to the note some columns to show new *totals* under each of the three poverty lines (under all the scenarios for growth contraction and distribution), just to make interpretation easier? (Also, helps in that there are a lot of baseline, pre-pandemic nowcasts for 2020 poverty out there, and I realise I am not 100% sure which ones you are using.) Apologies if it's in there somewhere, but I'm reading it at speed and can't find it. Many thanks!

August 07, 2020

Hi Tim, We definitely understand that it can seem counterintuitive that the number of added poor is almost identical at $3.20 and $5.50. Yet, it does not imply that the population mass between $3.20 and $5.50 is small nor that the number of added poor at $5.50 is the sum of 71, 176 and 177. Though it is true that the number of poor at $1.90 is a subset of the number of poor at $5.50, the number of *added* poor at $1.90 is not a subset of the number of added poor at $5.50. Here is a stylized numerical example, which, we hope, illustrates the point. Suppose we consider a country of four individuals with the pre-COVID incomes of $1, $2, $4, and $6. Before COVID the $1.90, $3.20 and $5.50 headcount is 1, 2 and 3, respectively. Suppose as a result of COVID all four individuals lose $1. The new headcounts at $1.90, $3.20 and $5.50 are 2, 3, and 4, respectively. In other words, the added poor is 1 for all three poverty lines, and the number of added poor at $5.50 is not the sum of the added poor at lower poverty lines. Happy to discuss further as needed . As the pre-pandemic baseline we use country-level poverty nowcasts that we generated in January. Great idea about column totals, we will definitely keep this in mind for the next update of the blog.

Ojwang K.
August 05, 2020

Need to know the impacts of Covid 19 exhaustably in East Africa, especially Kenya.