This new paper, with Martina Viarengo, addresses a narrow technical question in order to get traction on some very large consequential (and currently hotly debated) questions about development and development assistance.
The technical question is pretty easily posed, with a little background. A global lower-bound poverty line of “dollar a day” (which is now P$2.15 per person per day in 2017 PPP) is widely accepted. It is also widely accepted that the dollar-a-day threshold is not “the” poverty line but only the lowest line (often called “extreme poverty” because it is an extremely low threshold) and that one can legitimately measure and report national and global poverty at a wide variety of poverty lines, the World Bank data platform routinely reports poverty at P$3.65 and P$6.85. In addition, many countries have their own national poverty lines, nearly all of which are higher than P$2.15, and the World Bank reports a country specific “societal poverty line.” The first Sustainable Development Goal is “End poverty in all its forms everywhere” which obviously allows for a wide array of measures of poverty. The question is: Is there a global upper-bound poverty line for development related poverty? Or, put another way, if development assistance is about helping developing countries improve their wellbeing and among those goals for wellbeing is the reduction of poverty, what is the highest poverty line, beyond which one would say, people above this threshold are prosperous, not poor?
Our paper argues, with an analytical framing and empirical estimates that a good candidate for a global-upper bound poverty line is P$21.5. This is a good candidate in part because focal points matter and this is exactly ten times higher than the lower-bound, dollar-a-day line for extreme poverty of P$2.15.
This is much higher than the current highest line used by the World Bank (P$6.85) but consistent with other proposed poverty or prosperity lines. The World Bank prosperity gap measure uses a threshold of P$25 per person per day (here, here). Max Roser has powerfully argued (here and here)–extending and improving on earlier augments of mine (here)–that the poverty lines of the richest countries should be the upper bound, on the pretty reasonable grounds that :the poor” should mean “the poor” no matter where they live. P$21.5 is around the societal poverty line (here, here) of the lower consumption per capita but fully “developed” countries, like Spain.
The larger point of proposing and defending as reasonable a high upper-bound for a global poverty line (10 times the low bound) is to change the discourse on what “development assistance”–from multi-lateral organizations, bilateral organizations, and development related philanthropy–is about.
The original vision as “development” as a distinctive field and endeavor came into existence in the 1950s was that, as the new post World War II world order was inexorably leading to decolonialization, these newly sovereign states would develop as countries to become “developed” by which was meant a four-fold transition to (i) high productivity, (ii) capable public sector organizations, (iii) a responsive polity, and (iv) social cohesion and equal treatment. Francis Fukuyama as called this “getting to Denmark.” It was envisioned that “development assistance” could accelerate the progress so that rather than “getting to Denmark” happening at the centuries long-pace of the actual Denmark, this transition to national development could happen in decades. This “national development” process was not seen as an end in itself, but believed that this four-fold transformation of national development and “getting to Denmark” would lead to higher levels of human flourishing, as in Denmark.
The entirety of Article I of the World Bank (International Bank for Reconstruction and Development) which lays out its purposes are worth reproducing:
The purposes of the Bank are:
(i) To assist in the reconstruction and development of territories of members by facilitating the investment of capital for productive purposes, including the restoration of economies destroyed or disrupted by war, the reconversion of productive facilities to peacetime needs and the encouragement of the development of productive facilities and resources in less developed countries.
(ii) To promote private foreign investment by means of guarantees or participations in loans and other investments made by private investors; and when private capital is not available on reasonable terms, to supplement private investment by providing, on suitable conditions, finance for productive purposes out of its own capital, funds raised by it and its other resources.
(iii) To promote the long-range balanced growth of international trade and the maintenance of equilibrium in balances of payments by encouraging international investment for the development of the productive resources of members, thereby assisting in raising productivity, the standard of living and conditions of labor in their territories.
(iv) To arrange the loans made or guaranteed by it in relation to international loans through other channels so that the more useful and urgent projects, large and small alike, will be dealt with first.
(v) To conduct its operations with due regard to the effect of international investment on business conditions in the territories of members and, in the immediate postwar years, to assist in bringing about a smooth transition from a wartime to a peacetime economy.
The Bank shall be guided in all its decisions by the purposes set forth above.
The world “productive” (or variants) occurs six times. The stated goal of “development of productive resources” is raising “the standard of living and the conditions of labor.”
The word “poverty” is never mentioned as among the purposes of the IBRD (nor in the articles of IDA the concessional lending arm).
The World Bank’s 1990 World Development Report had a one word title: Poverty. This report needed a headline number: how many people in the world were poor? The decision, which was entirely political (I was in the larger unit of the World Bank producing the report and peripherally–as a very junior staff–involved in these discussions), was made to choose for this “headline” measure of poverty the lowest defensible poverty line (the poverty line of the poorest countries (here)) and hence report the smallest number of people poor.
This choice of reporting “global poverty” at the “dollar-a-day” consumption per person per day poverty line was thought to be innocuous, but set in a motion a very strange process of “downward mission creep.” Usually when people say “mission creep” they mean that an organization takes on a broader and more expansive agenda. But in this case, the spread of the combination of ideas that “the goal of development is poverty reduction” and “poverty is defined by a dollar-a-day poverty line” led to tendencies to downward mission creep among development organizations, in which rather than seeing development as a process of helping countries achieve high productivity and hence high standards of living and “conditions of labor” (as in the IBRD articles), development was seen as an effort to help “the poor” where “the poor” were defined on a penurious, lowest possible, poverty line.
This has set off an increasing fracture among people and organizations working on “development” between those who maintain the original vision of national development and those who want to frame the endeavor of development as a mitigation of the worst consequences of the lack of development. This latter approach looks to frame the promotion of development as the implementation of interventions (projects, programs) that are cost effective at reaching the poor, while essentially bracketing the question of what helps a country in “getting to Denmark.”
This approach, creeping “defining development down” or embrace of “kinky development” since 1990, has reached several reductio ad absurdum moments but which, having swallowed the dollar-a-day’ poverty pill, people do not see as ad absurdum.
For instance, in 2016 Bill Gates proposed that promoting chickens would be an effective investment for poverty in Africa. That statement was both perhaps sensible and completely absurd. If one was asking the question: “what is a discrete intervention that can be implemented as a cocooned project that would be cost-effective at brining about modest increases in the sustained income of selected households, targeted based on their low-income (even within African countries) and hence reduce some very low-bar measure of poverty (and for which the impact was visibly attributable to the funding)?” Then perhaps programs to promote higher incomes from chickens is a sensible claim (one would need much more detail on design, implementation, and evidence, to be sure, but perhaps sensible). But, at the same time, if one means poverty reduction in the way that “getting to Denmark” means poverty reduction–getting to Denmark’s level of poverty–then chicken projects as a any serious component of a poverty reduction are clearly absurd. One might think an incredibly intelligent person like Bill Gates, having reached the conclusion that “chickens are the path out of poverty for Africa” might stop and reflect: “How did I come to accept a definition of poverty such that this statement might be even remotely plausible?”
A second example is what the latest World Bank data now say about extreme poverty. In the latest (April 2025) data the poverty rate in Pakistan is 4.93 percent. More astoundingly, poverty in Nepal is .37 percent–at the dollar-a-day poverty line less than 1 in 100 people in Nepal are poor. So if “development is about eliminating dollar-a-day poverty” then Nepal is a developed country. At which point anyone sensible says “any indicator that says Nepal has no poverty is an absurd standard for poverty.” That is right. Child mortality is still 27 per thousand (a level Denmark reached 65 years ago). Child malnutrition is 18.3 percent. So somehow about about 18 percent of Nepali children are not poor but are malnourished. 49 percent of households are not using safely managed sanitation and yet are also not poor. It is not absurd to care about a standard for the “poorest of the poor” but to act as if that is “the” development goal is absurd.
We are not arguing to give up on poverty as a goal for development. The opposite in fact. In order for development assistance to be relevant to development it needs to use an array of standards for global poverty as a goal. Certainly extreme poverty measured as dollar-a-day is one of those standards, but not the only one, and if there is a low bound and an array their either is an upper-bound poverty line (or, if not, infinity). A sensible upper-bound poverty line as a development goal is 10 times higher, P$21.5, a threshold that far more accurately reflects the aspirations of people and countries for their own lives and futures.