The Promise and Peril of Information on the Returns to Education: What We Learn from a Dozen Interventions

By David Evans and Amina Mendez Acosta

How much is an education worth? In addition to an enormous scholarly literature reporting income returns to education, many websites seek to provide simple information for students about how much those who finish high school will earn compared to those who don't, how much those who finish university will earn, and which university degrees deliver the largest average salaries. In Latin America and the Caribbean, many children still do not complete secondary education, and fewer have information to make informed choices about post-secondary enrollment choices.

But many children and youth, especially those in low- and middle-income countries, may not have access to detailed information on the returns to education. That yields the idea for a very simple but potentially powerful intervention: what if you just told people what the true returns to education are, so that they could make the most informed decisions for their lives?

The most famous example is a study by Jensen, in which eighth-grade boys in the Dominican Republic believed that the return to finishing high school was very low, when the data showed that it was high. Jensen reports that "students at randomly selected schools given information on the higher measured returns completed on average 0.20-0.35 more years of school over the next four years than those who were not"! That's a big effect.

Who else has tried this? Amina Mendez Acosta and I looked for every study we could find that evaluated providing information on the returns to education to children or youth. We include both studies that provided information about the returns to different levels (high school versus primary only; college versus high school) and those that provided information about the returns to different fields of study or colleges at the post-secondary level. We report what we find in a new working paper, "Changing Perceptions of Educational Returns in Low- and Middle-Income Countries: A Meta-Analysis."

The studies

In addition to the Jensen study, we found a dozen more. Two provided information to students in primary school (one in Madagascar and one in Peru), five in lower secondary (in China, Colombia, the Dominican Republic, and Peru), and nine in upper secondary (in Chile, China, Colombia, Georgia, Mexico, and India). If you think that adds up to more than thirteen, you're right! Several of the studies spanned multiple levels of education, as you can see in Figure 1.

Figure 1: The thirteen studies on providing returns

Source: Evans and Mendez Acosta 2024.

The range of studies demonstrate the number of decisions that these programs have to make. Do we provide information to students or parents or both? Do we provide simple estimates of earnings at different levels? How much do you disaggregate the data (for example, by gender)? Do you share the information by email, through in-school presentation, online videos, text messages, or in some other way? The studies we have are different enough that we can't conclude which delivery methods or audiences are most effective, but they do give us a glimpse of the range of possibilities.

The good news

Even when you take an average of all these effects, across studies that were very successfully implemented and others that were less successfully implemented, we find positive effects. Specifically, we see positive and statistically significant average effects both on learning outcomes and also on measures of whether kids stay in school. The average effects aren't big (between 0.02 and 0.05 standard deviations) compared to other educational interventions, but they are statistically significant, and since these tend to be very low cost interventions, they are likely to be extremely cost effective. Several of the studies had larger effects (in the Dominican Republic and in Peru), particularly on keeping kids from dropping out of school.

The caution

First, implementation is crucial for any policy, and these are no exception. In one case, videos were mailed to schools but many didn't receive them and some that did receive them didn't show the videos. The following year, monitoring calls plus modest school incentives improved implementation dramatically. In another case, schools couldn't show videos for lack of electricity. One program set up a call center to troubleshoot challenges with showing videos. In settings with limited technology, old fashioned (or shall we say classic) delivery mechanisms like visiting speakers or posters may be wise options. Going forward, policymakers may enlist artificial intelligence and mobile communications like WhatsApp in their efforts.

Second, the returns to education reported are in averages and few students, if any, will receive the average return. Some might earn more; some might earn less. Boys and girls might receive different returns. Students in rural areas might receive different returns than students in urban areas. Interventions can share the returns at a more disaggregated level (girls versus boys), but they still won't be perfect predictors of individual students' returns.

Third, these programs are often premised on the assumption-at least at the level of basic education-that people underestimate the returns to education. But if people overestimate the returns, they may reduce their educational investments. That may be wise at the post-secondary level (realizing that studying a certain major at a certain college is unlikely to land you a lucrative job is crucial) but is more concerning if, for example, it means some parents reduce support for their children in basic education. To be clear, we do not observe an average negative effect in the studies we review: but we do observe significant variation among students and parents as to whether they initially under or overestimate the returns to education.

Take aways

Last year's high-level report on Smart Buys in global education based on cost-effectiveness analysis put interventions that provide information on the benefits of education among just three "great buys" (the highest category). Our study backs that up: we see evidence from many settings suggesting that these can deliver positive impacts at low cost. But pay attention to implementation, and make sure you're sharing information so that people understand that education is a great investment, just one that still comes with at least some risks in terms of the payout.

Public link

Please use the above public link if you want to share this noodl on another website.