Will Setting Goals End Hunger? What’s Next for the SDGs…

From Council of Foreign Relations



Emerging Voices highlights new research, thinking, and approaches to development challenges from contributing scholars and practitioners. This post is from Dean Karlan, professor of economics at Yale University, president and founder of Innovations for Poverty Action, and founder of ImpactMatters, a newly-launched organization that helps nonprofits use and create evidence to assess their impact.


Earlier this fall the United Nations (UN) ratified the Sustainable Development Goals (SDGs), a new agenda for helping the poor around the world by 2030. Their ratification comes fifteen years after the Millennium Development Goals (MDGs), more than doubling the number of global commitments from eight to seventeen and breaking them down further into 169 targets.

The goals include ambitious and noble aspirations, from ending global poverty “in all its forms everywhere” (SDG 1) to ending hunger (SDG 2). But, do the goals matter?

Since last summer when the UN released a report on how the MDGs fared, debate has ensued about whether setting defined goals made countries more likely to achieve them. More precisely, did aiming at the MDGs lead politicians and donors to allocate more resources to poverty-fighting, thus leading to improved outcomes?

As we turn to the SDGs, here is what we know about how goals may help or hurt, and a modest proposal for applying it.

Behavioral economists and psychologists have shown that goal setting can help individuals lose weight, finish annoying work tasks, exercise regularly, save more, and so on. Politicians, being people too, may need a set of goals to help them focus on and invest scarce resources in social problems.

Research also shows that while goals help, setting too many of them can deter progress and limit motivation. If one pledges to lose weight, stop drinking coffee, read more books for pleasure, and ignore work e-mails on the weekend—all at the same time—something is going to give.

The 169 SDG targets (most likely added to placate various UN constituencies) mean a lot more goal setting than the MDGs. Can politicians and development experts keep track of all these commitments?

Unrealistic goals can also be a problem—not only do they not get met, but they may fail to motivate. We know that signing up for a marathon is not the best way to make good on a New Year’s resolution to get into shape. The same goes for pledging to end all global poverty instead of merely reducing it.

So when it comes to whether the SDGs matter, I have a proposal to find out: randomize the way the 169 targets are assigned.

Many development economists (myself included) have done rigorous studies on goal setting at the household and individual level. In the Philippines, we set up randomized control trials (RCTs) to test whether offering people a commitment “goal savings” account helped them to save more money, and encourage them to spend more of it on large, durable goods. It did. In Uganda, a study we did found that offering children a commitment “goal savings” account led to higher expenditures on school supplies and improved test scores.

Why not apply the same rigor to test what motivates the world’s politicians to reduce poverty and inequality? We can allocate each UN member country a number of targets on which to focus, and then track progress over time to learn whether those with particular targets did better on a specific goal than countries without them.

Of course, this would be logistically impossible, and would make for a badly-designed randomized trial (with the “control” goals already known).

Though my proposal is tongue-in-cheek, it highlights the challenge of measuring the new SDGs. Some ask: how are we going to know if any improvement was made because of the goals? That may be the wrong question.

Instead, we should ask whether a particular policy is a good or bad idea. Because while the SDGs motivate us towards aspirational outcomes, they do not tell us what to do. Learning what policies actually work is a much more useful basis for a study, and where we should focus our attention.