A woman carries a bag of rice donated by the U.S. Agency for International Development in Leogane, Haiti, in January 2010, days after a powerful earthquake hit the country. (Lynne Sladky/AP)

The good news is that global poverty is falling, according to World Bank data. The bad news? It’s getting harder to reach the world’s poor — 80 percent of global citizens living on less than $1.90 a day, and 75 percent of those living on less than $3.10 per day live in rural areas.

This presents a challenge for foreign-aid donors, and to their goals of eradicating extreme poverty by 2030 and lowering inequality within countries. Meeting these ambitious goals will require donors to direct aid to more remote areas — where the poorest live — rather than urban areas, where aid projects are often easier and cheaper to execute.

Total development aid, meaning concessional loans and grants for development purposes from governments of high-income countries, stood at about $170 billion in 2016. Although that might seem like a lot of money, it’s less than what Samsung earned in revenue that year and not much when compared with the magnitude of the goal of ending extreme poverty in less than 15 years. Given these resource constraints, aid needs to be targeted carefully.

I recently wrote two articles examining whether aid from some of the most poverty-sensitive donors, the World Bank and the African Development Bank, was actually reaching the places where the poorest people live. It isn’t.

So where does foreign development aid go?

Instead of going to poorer places, I found that aid tends to go to the places that:

  • hold more of the country’s richest 40 percent of the population.
  • are more urban.
  • have more light at night (a proxy for development levels).
  • have lower estimated rates of child malnutrition.
  • have lower estimated rates of infant mortality.

Poorer countries tend to receive more aid than richer countries. However, within the poorer countries, richer cities and regions are more likely to receive at least some aid. And in dollar terms, the aid these places get typically outstrips aid headed to poorer regions. It’s not a question of population density — after controlling for differences in the number of people across places within countries, I found that foreign aid still flows to wealthier places.

What does this tell us?

For one, these findings suggest that most aid may not be directly helping the world’s poorest people. Foreign aid targets different goals — and to be effective, some types of aid must be directed to relatively rich places. For instance, it makes sense that aid to develop port facilities would go toward a coastal city, even if that city was already relatively prosperous.

But aid to fund schools, roads, electrification, improved sanitation, health clinics and other targets could be directed to poorer places within countries. Typically, it seems this is not the case.

Does this mean that aid is being spent badly? Not necessarily. Donors are under a lot of pressure to produce results. Building a school in a remote and impoverished region, for instance, will probably be more expensive than building a similar school on the outskirts of a country’s capital. This is because it costs more to move supplies and equipment to a rural area that may be reachable only by unpaved roads — or only on foot.

Perhaps we think it is good for donors to try to get “more bang for their buck” by focusing on helping relatively well-off people in poor countries — people who nevertheless are poor by global standards. This is a debate worth having, but it starts with acknowledging that aid is not reaching the poorest.

What’s more, pressing donors for more efficiency and impact may create incentives against their helping the poorest. Unfortunately, there may be a trade-off between helping those most in need and helping the most people per dollar. If donors want to eradicate extreme poverty and reduce inequality, they will need to shift more of their resources to the places where the extremely poor live. Such a shift, however, may come with the cost of helping fewer people overall.

Ryan Briggs is an assistant professor of political science at Virginia Tech. His research focuses on the political economy of poverty alleviation. Follow @ryancbriggs.