Global Forum on Food Security and Nutrition (FSN Forum)

Matthew Fisher-Post

the World Inequality Lab of the Paris School of Economics

Dear CFS and HLPE-FSN colleagues,

Please let me share below 10 points of feedback for the v0 draft, whether these fit within the scope of the e-consultation or beyond.

  1. Let me first congratulate you for the range, depth, and level of effort in this important work. There is a clear need for this overview of distributional aspects of food systems, and your team is answering the call.
  1. While it would be impossible to comment on all aspects of this report—and my colleagues and I (at the World Inequality Lab of the Paris School of Economics) would perhaps not be the right people—I hope we can offer some small kernel of input to help in the rounds of revisions.
  1. Indeed, the first thing we would note is the wide range of this report. Clearly there are many political pressures to discuss every potential driver and outcome of every possible inequality, and the report may suffer from trying to be everything to everyone. The herculean effort to cover so much ground recalls the adage, that if everything is a priority, nothing is a priority. Among so many priorities, it may be too easy for readers and policymakers to lose track of (or never receive) certain points of emphasis. It is hoped that subsequent revisions—despite even more feedback in public e-consultation—can foreground a specific and limited set of priorities, both in the content of the analysis and the policy actions that flow from this analysis.
  1. If we may suggest a few points of emphasis, they would be the following: Focus on incomes, on data quality, and on social protection policies. In each of these areas, some further empirical grounding could help drive the report and help its emphases resonate.
  1. Focus on incomes and wealth. You note several studies (including the systematic review of Alao et al 2021) that establish a robust finding, and one that is perhaps not surprising: the same individuals and households that experience food insecurity, are those that are poorest in the income distribution. This is not to say that all of multidimensional poverty measures can be collapsed to a simple income measure, nor that systemic inequities do not matter—far from it. Rather, it is only to suggest that incomes are easiest to observe, and that they are suggestive of the other elements of inequality and inequity that this report mentions and shines some light on. It may be rather the exception than the rule if there are some cases where income-rich can be nutrition-poor (e.g., historically excluded minorities in high-income countries). These cases are important, and deserve attention, and indeed it is one of the strengths of this report to shine a light (with text boxes, etc.) on such cases. However, there remains room for a further empirical foundation in some areas of this report—and the study of and focus on income and wealth (esp. land) inequalities would facilitate further empirical solidity of the findings and conclusions in this report.
  1. Income poverty and inequality is also perhaps more important as a measure than is consumption poverty/inequality. While consumption measures are frequently used for measuring absolutely poverty—and detailed food consumption analysis shows the prevalence of under- and malnutrition—we would argue that income measures are superior as an indicator of the true extent of inequality. Indeed, among the poorest, consumption is usually on average several times higher than income. This is not surprising given low incomes, but the consumption measure then underestimates inequality because it shows what low-income households have already had to do to overcome poverty. Consumption measures are taken after survival and resilience strategies and government transfers. Income measures show the market outcomes individuals face, prior to public and private transfers that facilitate the most basic levels of consumption among the poor.
  1. That said, it is also worth noting that data is scarce, even on consumption and nutrition and non-income measures of inequality and poverty. Income and wealth data remain in their ‘prehistory.’ Where they exist, they are often inadequate, incomplete, and of low quality. The 50x2030 initiative has drawn attention to the importance of measuring agricultural income as a pillar of household income, especially among the rural poor. Labor force and other surveys that do not use as robust a method to estimate household income, can underestimate inequality and give an inaccurate picture of poverty and food insecurity. Perhaps less relevant in this context, it should also nonetheless be mentioned that existing survey data sources do not cover all forms of income, and notoriously miss top incomes specifically and capital income more generally. This report should be able to complement the previous HLPE report on data—and particularly some concrete actions among NSOs and international organizations for increasing the coverage, timeliness, and quality of surveys and income/inequality measures, especially among rural and agricultural households and among actors (including business entities) up and down the food systems value chain. Not enough is known about inequalities in food systems, and more and better data would go a long way to help close these gaps in our knowledge.
  1. Social protection is briefly discussed in the report, but its role in ameliorating inequity and inequality deserves further emphasis and depth of discussion. Social protection policies remain the single most powerful policy tool for redistribution and poverty reduction. In the first instance, progressive taxation (e.g., of income and wealth; rather than indirect taxes on consumption) can reduce the fiscal burden on the poor—and then targeted social protection policies can similarly aim fiscal policy at the poorest and food-insecure. Social insurance largely misses the poor, when it is tied to formal employment. And social assistance is too rare, and frequently inadequate even where it does exist. Health and education spending are significant investments in wellbeing and human capital, but cash transfers and direct (or conditional) cash transfers have an important role to play, as well. This is of course not news to you, but the place of social protection policy among actions (and transformative actions) for reducing poverty and inequality should not be understated or obscured.
  1. The chapter on drivers of inequality (chapter 4) is perhaps less inspired than the discussion of actions to confront inequality (chapter 5), but this may be a result of the material the team is forced to work with. As discussed above, the former might not be well enough understood. And the latter may remain a wishlist, hoping that everyone can do everything all at once. The vignettes and text boxes that pepper this report are helpful, and maybe some more of these short discussions could help the reader and also sharpen the emphases of the report.
  1. If the e-consultation is meant to be as comprehensive as the report itself, I hope it is no disappointment to refrain from commenting on the definitions and literature that underpins this report. Some further discussion of the literature on the points of emphasis above (income measures, data scarcity, and social protection) would be welcome, but it is understood that this report may suffer from too many voices rather than too few. It is hoped that these ideas may help as you build toward a consensus in delivering the final product and influencing related policy processes. We stand ready to elaborate further on any of the above points, and even to help include further empirical grounding as you proceed to revise the report.