Ideas to Action:

Independent research for global prosperity


Views from the Center


There is a little-noticed but important difference between the World Bank’s original goal for poverty reduction and the first of the subsequent UN Sustainable Development Goals (SDG1). Both target the Bank’s “$1.90 a day” poverty rate. The difference is that the Bank’s goal was to reach a 3 percent poverty rate by 2030, while the SDG1 is to “eradicate” poverty by 2030, where “eradicate” means zero. Yet that 3 percent could well make a big difference, as I explain in a new CGD working paper published today, “SDG1: The Last Three Percent.”

A simple linear projection of current progress against extreme poverty in the world does suggest that we are on track to attaining the UN’s SDG1. However, linear projection is deceptive if development does not reach the poorest as effectively. And there are some signs that is true. I have also been tracking (as best I can) the living standards of the very poorest in the world—those living at the floor, which is somewhere around $1.00 a day (at 2011 purchasing power parity)—well below $1.90 a day. I don’t see much progress there. Roughly speaking the world’s poorest today are just as poor as they were 30 years ago; what has changed is that there are fewer of them. Indeed, at the present rate of progress for the poorest people, it will take them another 200 years or more to get to $1.90 a day. (See my post here for more detail.)

Why is it so hard to reach the last 3 percent?

When we think about it, there are a number of a priori reasons why the last few percent could be harder to reach with current development policies:

  • It is well known that there is transient poverty—people flow in and out of poverty in any time period, reflecting the imperfections of risk markets and the institutions (including governmental) for social protection. Insisting on “eradicating” poverty may then be a much harder goal than 3 percent.
  • Aside from transient poverty, the chronically poorest may live in remote places with characteristics, such as poor infrastructure and/or natural conditions, that lower the productivity of their labor and capital. Thus, they may be caught in geographic poverty traps, whereby a poor location has a causal role in retarding the prospects of escaping poverty. And these can be costly problems to fix.
  • There may be social, political, and/or economic constraints facing the poorest, associated with their identity and the discrimination they face in specific social contexts. Standard policies that work well for the majority may fail for these groups. Relatedly, refugees, undocumented migrants, and stateless people may be harder to reach, and probably both poorer than average and undercounted in our survey data used to monitor poverty. (Indeed, when a poor person migrates but stays poor, and is not counted in the survey data for the destination, the measured global poverty count falls even though the true value is unchanged.)
  • The poorest may be caught in low-level attractors (a “dynamic poverty trap”) such that large idiosyncratic gains are needed to get onto a sustainable path out of poverty. Small gains will not succeed in assuring sustained poverty reduction.
  • Exposure to theft and violence associated with weak legal enforcement and other deficiencies in state capacity may create extra hurdles to escaping poverty, with reduced efficacy of standard policy packages for reaching the poorest.

By implication, our progress in reducing numbers of poor may not actually come with much progress in reaching the poorest globally, once a country reaches the last 3 percent (say). We can call this the “poorest left behind hypothesis.”

Where might we look for evidence for or against this hypothesis? There could be no surprise if countries that have not been very successful against poverty also had a harder time reaching their poorest. (For example, a number of observers have pointed to the challenges facing many countries in Sub-Saharan Africa in attaining SDG1.) The more interesting place to look is the set of countries that have made substantial progress against poverty over the longer term. To achieve SDG1, do we just need to get today’s poorest countries on the same path as these past star performers (which could be a big challenge in its own right)? Or does the problem go even deeper, as suggested by the arguments summarized above?

My new CGD working paper documents recent signs of a levelling off in progress for the poorest in many of the countries that have been most successful against poverty over the longer term.

An example is Malaysia—one of the most successful countries in reducing poverty over the last 50 years or so. Malaysia’s “$1.90 a day” poverty rate had reached 3 percent in 1984. Yet 32 years later, it was still not quite zero (0.1 percent). The last 3 percent took more than three decades! Possibly the world’s last 3 percent could take a similar time even if other developing countries were as successful as Malaysia has been in reducing poverty.

I find signs of a similar pattern in East Asia as a whole. As is well known, East Asia has been the star performer regionally over the longer term. However, I find signs that the region’s progress against poverty has slowed down in recent times. This is evident in both the region’s lack of progress in lifting the floor—thus reaching the poorest—and in its slower progress in reducing the poverty rate.

Looking beyond East Asia, I also find signs of this levelling off on average for the 18 developing countries globally that have reduced their poverty rate from over 10 percent (around the current global rate) to under 3 percent during 1981-2017. Similarly to East Asia, progress in reaching the poorest declined once the last 3 percent had been reached, though some countries did better than others.

This slowdown in progress against poverty once one reaches the last 3 percent is not due to lower growth. Rather, the cause is the incidence of growth, which has not reached the poorest. In short: rising inequality is the main culptit.

A pro-poorest approach

The new paper’s results suggest that, for many developing countries, success in eliminating poverty will not come with a “business as usual” approach that relies heavily on economic growth. While growth has helped assure huge progress in reducing counts of poor people in the developing world, it appears that eliminating extreme poverty will require that the future growth process in market incomes is more deeply pro-poor. These approaches must be targeted to reach the poorest, and the growth process in market incomes must be accompanied by more active, and effective, redistributive efforts. Current social protection and antipoverty policies have helped, but not much (as my recent paper with Juan Margitic shows). “Business as usual” will not do the job there either. 

We also need to acknowledge that addressing the factors that make it harder to reach the poorest can generate ethically challenging trade-offs, notably if success in truly eradicating extreme poverty were to slow progress for others who are also poor, but not as poor. This trade-off may well arise in some circumstances. Yet there is surely ample scope for largely avoiding such a trade-off, notably by efforts to assure that it is the rich who carry the bulk of the extra cost. And the rise in market-income inequality that appears to underly the slowdown in progress for the poorest suggests that there is even more scope for such financing in today’s world.  


CGD blog posts reflect the views of the authors, drawing on prior research and experience in their areas of expertise. CGD is a nonpartisan, independent organization and does not take institutional positions.