Emily Feenstra
Associate, Investments

Examining the Returns of Tertiary Education in Brazil

June 20, 2017

By Emily Feenstra, Investment Associate, Omidyar Network, and Princess Adentan

Read this article in Portuguese here

Central to Omidyar Network’s education investment strategy is the belief that all children deserve quality educational opportunities, yet a persistent opportunity gap exists between high-income and low-to-middle-income (LMI) families globally. Although this gap takes different shapes in each country, the effects are similar—opportunity gaps manifest as income gaps later in life.

Brazil, a country of focus for Omidyar Network, frequently ranks among the most unequal countries in the world. In 2014, the OECD released a report highlighting that Brazilian workers experience one of the largest differences in earnings according to their level of education; tertiary-educated Brazilians earn more than 2.5 times the income of those with upper-secondary education. But do all learners have access to the tertiary education that is so crucial to higher earnings later in life?

In order to answer this question and implement our strategy successfully, strong baseline data is critical—yet in this case the available data was lacking. So, to understand the nuances of the educational opportunity and income gap in Brazil, in late 2016, Omidyar Network conducted a survey in conjunction with the RIWI Corporation. The following is a summary of our key findings.

Tertiary graduates showed significantly higher incomes than non-graduates.

In our sample, the average monthly income of a tertiary graduate (R$2.7K) was 1.7 times that of a non-graduate (R$1.5K). While this is slightly lower than the OECD estimate, our survey was conducted in December 2016 at the height of Brazil’s economic crisis and a time when unemployment and underemployment rates peaked. These factors aside, our data confirms that tertiary graduates experienced a sizeable earnings premium compared to non-graduates in Brazil.

Even among tertiary graduates, students from lower-income households earn far less than tertiary graduates from higher-income households—suggesting low economic mobility. 

There was a strong correlation between graduates’ current income and their household income. While this correlation is not inherently surprising—and is sadly consistent around the world—the starkness of our data warrants a closer look. The average current income of graduates from households in the highest income bracket (defined as those earning over R$4,000/month) is 4.4 times the average current income of graduates from households in the lowest bracket (defined as those earning less than R$500/month). Unfortunately, this finding is very consistent with studies of Brazil’s macroeconomic environment. In 2016, the Stanford Center on Poverty and Inequality published a report on economic mobility comparing the rigidity of intergenerational earnings across middle-income and high-income countries. Brazil ranked fourth out of 24 countries on the strength of the correlation between father and son earnings.

Graduates of public university programs in the fields of education, healthcare, and mathematics earned more than their counterparts from private university programs, while incomes were comparable for graduates in the fields of law, engineering, computer science, and management.

It is a widespread perception in Brazil that many public university programs are very high quality, relative to more variable quality in the private sector. However, there had been no data available to validate or refute this belief. When controlling for household income, we found that public graduates earned 21 percent more than private graduates in education and healthcare, and 11 percent more in mathematics. Incomes were more comparable for graduates of public and private university programs in the fields of law, management, computer science, and engineering earning between 1 and 7 percent more than their private university counterparts. From our data it is not entirely clear what is driving these differences, however we invite others to dig in and help investigate this further.


Among private university graduates, the cost of the program was correlated with future earnings in most cases.

Even when controlling for household income, in most fields the more a student pays for university the more they can expect to earn in the future. This is good news, as it would suggest that a program’s pricing does reflect, at least to some extent, the labor market value of that degree. Graduates from medium-cost programs earn R$560 to R$1,240 more than their peers who graduate from the lowest-cost programs. This data highlights the critical role that educational financing can play in expanding access for low- and middle-income students to higher quality, more expensive programs.

Finally, technology plays an important role in increasing access to tertiary education for low-income people but these programs need to be high quality and connected to the labor market.

Today, there is a significant income disparity between graduates of in-person, blended, and online programs, which is particularly acute for students from low- and middle-income communities. Low-income graduates of in-person programs earn almost 50 percent more than their peers graduating from online programs, and 35 percent more than graduates of blended programs. This effect persists even within a given degree program.

That said, it is important to note that graduates of online programs still earn more than non-graduates, and given the lower cost of online programs in many cases these programs are still a good investment for students. Graduates of online programs in our sample earned R$1.9K per month compared to R$1.4K for non-graduates (across income levels). Thus, we believe that this data speaks to the opportunity in the market to continue to improve the quality of online and blended offerings, to go beyond video classes streamed to remote locations, to more engaging, interactive, and adaptive offerings that also leverage technology to directly connect graduates with the labor market. We at Omidyar Network believe we are at the earliest stages of leveraging technology to advance education quality and learning outcomes, and that the solutions available will continue to improve through continued investment and applying lessons learned.

Omidyar Network recognizes that the goal of accessible and equitable quality education requires a collective effort. We invite entrepreneurs and innovators, researchers and practitioners interested in education in Brazil—to not only consider these highlights but to contact us to discuss further and to access the raw survey data to conduct additional analyses.

Methodology: This survey was conducted with RIWI Corporation’s random domain intercept technology (RDIT) which delivers anonymous opt-in surveys to random web users who make input errors by typing in website URLs that no longer exist or mistyping non-trademarked websites that RIWI owns or controls. We reached over 45,000 respondents ages 22-45, of which 10 percent completed the full 19 question multiple choice survey. The sample was not statistically reweighted based on demographic characteristics to match Brazil’s population.

It looks like there's some information missing
By clicking, you agree to the Terms and Conditions


High-Volume Recruitment Start-Up Giraffe Closes Second Funding Round

Giraffe, the automated high-volume recruitment start-up, announced that it has closed a second round of equity funding.



Researchers to Put Facebook's Role in Elections Under Magnifying Glass

Omidyar Network is joining six other funders to support a new initiative to study Facebook's role in elections and democracy.



Facebook announces commission to examine social media's effects on elections

Omidyar Network has joined an effort to fund a new initiative that will enable independent researchers to perform an assessment of the role that Facebook plays in elections.