Explaining Oxfam’s Projections for Extreme Poverty in 2022
By Max Lawson
This blog responds to Noah Smiths’ substack criticism of Oxfam’s extreme poverty numbers.
First of all, I would like to thank you for the scrutiny of our numbers, which I really believe increases accountability.
As we did a couple of years ago, I would have gladly provided the data that you’re after from the World Bank, as I do below. I was happy to be in touch in previous years, when you were looking into the wealth of the bottom 50% percent, and we made sure to connect you up with researchers and Credit Suisse entirely transparently and openly.
Firstly I want to say that it was a mistake that the paper we produced in March that you cite, unlike our Davos reports published in 2020 and 2021 which use the same World Bank dataset, did not have a full methodology note explaining all our calculations and sources. It did footnote all our calculations, but you are right that the citations could have been a lot clearer. Critically it does indeed not explain where the World Bank figures we used came from. This is in contrast to our Davos papers in January 2021 and January 2022 which are clear about this same source and all have full comprehensive methodology notes. Hopefully this response will clarify things and explain our calculations and their sources, and this is a learning we will definitely take for the future.
Summary: The 198 million extra people in extreme poverty figure
In summary: all these figures are taken from the same data set kindly shared with us by colleagues at the World Bank in March.
The difference between the World Bank figures that are published by the bank and the ones we use is that in the ones we use an increase in inequality of 2% is modelled by the bank. This has the impact of increasing the numbers of people living in poverty.
As laid out in the table below from the Bank, the projection of the World Bank for the numbers living in extreme poverty in 2022 if there had been no pandemic was 597 million.
The pandemic led to an increase in this projection to 686 million people living in extreme poverty in 2022.
But this estimate is based on no increase in inequality during COVID-19. The World Bank also modelled what poverty levels would be if instead inequality increased annually by 2%. This figure is 795 million. The difference between 795 million and 597 million is 198 million extra people living on less than $1.90 a day in 2022 because of COVID-19 and increasing inequality.
Our figures on the impact of the food price spike result largely because we chose to make our estimates from when price foods began to rise, and not the beginning of the Ukraine war. The Ivanic paper that we based our model on is here, following the same path as the Center for Global Development. The food prices were based on FAO food price indexes. As I describe below, given that food prices have risen again significantly, our estimates are likely to be on the low side, While the newer Bank estimates that you cite, published after our paper, do indeed take Ukraine into account, the dataset before does not. Using this methodology we estimate that the food price increases could push a further 65 million people into extreme poverty in 2022.
World Bank Data
Their data set looks at all three poverty lines, $1.90, $3.30 and $5.50. It looks at three scenarios, a distribution neutral one, a scenario where inequality reduces by 2% and one where it increases by 2%.
This is a clip from the excel sheet shared with us by the World Bank which we used. I have asked them for permission to include it here. The figures look at projections of what would have happened to poverty levels without COVID-19 and what is projected to have happened with COVID-19, up to 2030.
The figures that the World Bank publish, which you cite, are taken from the same data set they share with us but they only publish the distribution neutral figures.
At the time of writing our report, their projections for a distribution neutral projection of the numbers of people living on less than $1.90 a day in 2022 were 686 million, compared to 597 million if COVID-19 had not happened.
Our paper does explain this in the main body of the text, and in this table.
The first two numbers in this table correspond to the graph in your blog, taken from the World Bank.
They are slightly different because our paper was published just before the World Bank updated their figures. They had said 597 million people would have been in poverty if COVID-19 had not happened, and they revised this down slightly to 581.3.
The figure including the impact of COVID of 656.7 corresponds to our figure above of 686. Again, this slight difference is because the World Bank revised their figures after we published.
We then included the number of those projected to be living in extreme poverty in 2022 if inequality had not remained neutral but had increased annually 2% as per the World Bank projections that they had given us. This gives us a total of 795 million people.
The difference between 597 and 795 is 198 million people.
An increase in inequality has a big impact on poverty numbers because any economic growth of course leads to less poverty reduction if less of that income goes to the poorest because inequality has increased.
If you want me to put you in touch with the analysts at the World Bank who have kindly shared their projections with us over the years, I would be happy to do so.
We checked our 198 million figure with the Bank, and whilst they are less convinced than us that such an increase in inequality is likely, they agreed that it was possible, and if so the figure would indeed be 198 million. Interestingly their recent analysis (see page 7) of Viet Nam’s economy as an example does show an increase in inequality during COVID-19; other countries like the US will probably see temporary decrease in income inequality but an increase in wealth inequality.
This view was supported by Oxfam’s survey of 295 economists from 79 countries. They included leading global economists such as Jayati Ghosh, Jeffrey Sachs and Gabriel Zucman. 87% of respondents expected that income inequality in their country was either going to increase or strongly increase as a result of the pandemic. This included economists from 77 of the 79 countries
Nevertheless we can debate whether or not we think that COVID-19 has increased inequality, and by how much. These are projections based on simulations, and we will only know when the household surveys and other work is done. As we say in our paper, we think that an increase in inequality is likely.
We are I hope clear in the paper that we are comparing what would have happened in 2022 absent these three things- COVID-19 an annual increase in inequality of 2% and the further impact of the food price increases.
Food price increases
Our figures on the inclusion of the impact of the food price spike are higher than the Banks’ largely because we chose to make our estimates from when food prices started to rise, which in fact was in March of 2021, and not the beginning of the Ukraine war. The Ivanic paper is here: World Bank Document. We should have provided the hyperlink.
Regarding the application of the model Ivanic et al did we have followed the same path as Center for Global Development (which is also in citation in our paper). We didn’t guess what the food prices would be. We estimated based on FAO food price index for February. In February the index was 140,7 (later adjusted to 141,1). In March the index was up 159,7 and as of May the number is 157,4. So since we estimated the impact of food price inflation on extreme poverty, food prices have risen again significantly so our estimates are quite likely to be on the low side.
The World Bank figures we were working from did not include the impact of the Ukraine war and inflationary impacts; those figures were published after our paper, as you point out.
As you also correctly point out, we are not comparing the level of extreme poverty in 2022 with 2021, but with the level of extreme poverty that was projected to exist in 2022 if these crises had not taken hold. Our whole premise, as is that of the Bank, is to show the impact of these crises compared to the pre-crisis trend. We want to show that COVID-19 has had a huge and dramatic impact on the fight to end poverty. It has led to the first global increase in extreme poverty in over twenty years. Unless there is concerted action, it is quite possible that poverty will not even return to levels seen on the eve of the pandemic for many years to come, let alone meet the targets set by the global community for 2030. Key to this is whether we manage to finally address inequality.
If you choose instead to look at the difference in extreme poverty pre-pandemic (2019) compared with 2022, the figures would be 655 for 2019 and 795 for 2022, which would mean an overall increase during the COVID-19 period of 140 million people. Poverty shot up in 2020 and has since come down again a bit thank goodness, but it still remains shockingly higher than it would have been. And rising food and energy prices make continued declines in no way guaranteed.
At the same time, again for the first time in over three decades, the gap between the rich world and the developing world is widening, as global inequality increases, a point the World Bank has also made.
Finally, billionaire wealth during the pandemic saw an increase that was enormous by any measure, buoyed by governments pouring money into the economy.
This is the first time again in a very long time that we have seen both extreme wealth and extreme poverty increase so sharply and so dramatically at the same time.
It is true that since the peak in 2020 and 2021, both billionaire wealth and extreme poverty have reduced, but they both remain far higher than they were pre-pandemic. According to Forbes, Billionaire wealth increased almost 60% between 2020 and 2021 and has since come down around 10%- still an enormous rise by any measure and something all of those concerned about concentrated economic power should be worried about.
Whilst we could well be being too pessimistic in our projections, equally it does not look like we have seen the last of the food and energy price increases, or of inflation. The prospects for developing economies, especially in terms of government debt, are darker than they have been since the 1990’s. Sharp increases in interest rates by the Fed will only compound this. The war continues, and its impact on the global economy and especially on food prices looks set to continue throughout 2022 and into 2023 and beyond.
I hope we have convinced you that we are not in the business of making up numbers, that we work with the World Bank, The FAO, Forbes, Credit Suisse and others to use the best data available, and we share these findings with external stakeholders before we publish them.
We definitely agree that the citations in this paper were not up to the standard of our usual papers and the reference to the World Bank dataset and where it comes from are as clear as they could be.
As I said we welcome scrutiny, and debate, and indeed take your blog in this spirit. You have my email and please do get in touch anytime with any questions or concerns about our calculations.
In response to the previous criticisms of Oxfam’s numbers you have repeated in the remainder of your blog, we responded to these at the time, and we are happy to do so again.
In particular I have contacted in particular my colleagues working on the latest climate numbers for the emissions of the richest, released last year ahead of COP 26, to respond to your points on this.