World Development Report 2018: Learning to Realize Education’s Promise

Education builds human capital, which translates into economic growth. If improvements are faster among the disadvantaged, the additional growth will reduce poverty, reduce inequality, and promote social mobility. Through its effect on civic agency—meaning high levels of political engagement, trust, and tolerance—education can create the building blocks for more inclusive institutions. Greater civic agency can create a political constituency for inclusive institutions, strengthening the social contract between the state and its citizens. A more engaged citizenry can also provide political support for the reforms needed to realize the promise of education.

At the national level, education underpins growth. Human capital can boost growth in two ways: first, by improving the capacity to absorb and adapt new technology, which will affect short- to medium-term growth, and, second, by catalyzing the technological advances that drive sustained long-term growth. Widespread basic education may provide a bigger boost for countries far from the global technological frontier—a group that includes most low- and middle-income countries. These countries do not need to push that frontier out through innovation, but they do need widespread basic education to absorb and adapt the technologies that are already available globally. In countries close to the technological frontier, mainly high-income countries, higher levels of education can boost growth through innovation. Although data limitations make empirical analysis of this relationship challenging, many influential studies have concluded that higher levels of education do drive more rapid growth. Growth accounting analyses also suggest that education can explain a significant share of growth—a share that may be even larger if unskilled workers are more productive when there are more skilled workers in an economy.

But this statistical evidence is not the only—or even the most compelling—evidence on the impact of education on growth. Countries that have sustained rapid growth over decades have typically shown a strong public commitment to expanding education, as well as infrastructure and health. Although the relationship flows the other way as well—in that rapid growth allows greater investment in all three sectors—research on the East Asian miracle countries in particular flags education and human capital as factors in their rapid growth. Countries such as the Republic of Korea reaped the benefits of their “progressive universalism” approach to education, in which they ensured high-quality basic education for all children early on, followed by expansion of high-quality secondary and tertiary opportunities. These cases reinforce the idea that strong foundational skills drive growth early in development, but also that as countries approach the global technological frontier, they need to invest more in higher education and in research and development.

As education coverage expands, poor people typically benefit the most at the margin, and so income inequality should fall. A review of more than 60 studies reveals that greater education coverage is associated with substantial reductions in the income gap between households across the income distribution. Specifically, going from a primary enrollment rate of 50 to 100 percent is associated with an 8 percentage point increase in the share of income going to households in the poorest decile.

Education strengthens the political development of nations by promoting the civic engagement of their populations. People with more education consistently participate more in political activities than those with less education: education increases awareness and understanding of political issues, fosters the socialization needed for effective political activity, and increases civic skills. Evidence from a variety of settings shows that this relationship is causal. In the United States, getting more education—for example, as a result of preschool programs, high school scholarships, or smaller class sizes—leads people to vote more often (table 1.2). Using changes in compulsory school laws to identify the causal impact of education confirms these findings for the United Kingdom and the United States, while using access to community college or changes in child labor laws does so for the United States. In Benin, receiving more education made people more politically active over their lifetimes. In Nigeria, too, educational expansion substantially increased the civic and political engagement of its beneficiaries decades later.

As with the other effects of education, context matters in how education affects political views and engagement. In an indicator of perceptions of one common mechanism for political participation, surveys in 30 developing countries show that more-educated citizens are more likely to believe that living in a democracy is important (figure 1.3). But in Kenya, although more education caused young women to have more political knowledge, it also led them to be more disenchanted and more accepting of political violence, perhaps because democratic institutions were particularly fragile at the time of the research.

Education increases trust, tolerance, and civic agency. Evidence from member countries of the Organisation for Economic Co-operation and Development (OECD), as well as from developing countries, indicates that more-educated individuals are more trusting and tolerant of people they know and even of strangers. Although such cross-sectional evidence cannot prove a causal relationship, historical analysis suggests a mechanism: the spread of literacy may have contributed to a generalized decline in violence after the Middle Ages, because the ability to read others’ viewpoints promoted empathy. Some educational environments appear to promote trust especially well. Data for 28 countries reveal that the openness of a classroom climate, or the “degree to which students are able to discuss political and social issues in class,” is positively linked to trust and tolerance. Similarly, teaching styles that encourage teamwork rather than a more top-down pedagogy appear to promote social capital: students are more likely to believe in the importance of civic life and the value of cooperation.

Education makes institutions work better and improves public services. Educated parents are better able to leverage decision-making authority at the school level. In The Gambia, a school-based management program improved student learning—but only when there was a high level of literacy in the village. A more educated population generally demands more transparent use of public resources, better service delivery, and government accountability. Recent cross-country research identifies citizen complaints as a primary mechanism: educated citizens complain more, inducing officials to behave better. Education also appears to improve dimensions of governance: countries that had achieved mass education by 1870 had less corruption in 2010.

Growth built on human capital rather than other sources (such as natural resources) may lead to fewer incentives for conflict, for three main reasons. First, because human capital is difficult to appropriate, conquest of a well-educated population may be less rewarding than seizure of natural resources or even physical capital. Second, education raises the opportunity cost of fighting: it is easier to recruit people who have poor job prospects. Third, as discussed, education can promote tolerance and cooperation, thereby reducing the propensity to turn to violence to resolve conflicts.

Learning and the promise of education

Education can be a powerful tool for individual and societal empowerment, but its benefits are not automatic. It is not just that education cannot do it alone, in that much also has to go right in other sectors of the economy and society (box 1.2). Another problem is that if an education system is managed poorly, it can promote social “bads” instead of social “goods.” First, education can deepen cleavages between favored and disadvantaged groups. Young people from poor, rural, and otherwise disadvantaged households not only complete less schooling, but also learn much less while in school (see part II of this Report). In such cases, education does little to enhance social mobility. Second, leaders sometimes abuse education systems for political ends and in ways that reinforce autocracy or the social exclusion of certain groups.

Finally, schooling is not the same as learning. Education is an imprecise word, and so it must be clearly defined. Schooling is the time a student spends in classrooms, whereas learning is the outcome—what the student takes away from schooling. This distinction is crucial: around the world, many students learn little (figure 1.4). To be sure, many students learn something, even in settings facing huge challenges. And students enjoy some benefits from education regardless of whether they are learning. When schools serve as oases of security in violent areas, or when participation in schooling keeps adolescent girls from becoming pregnant, these are real societal benefits. When graduates can use their degrees to open doors to employment, that opportunity changes their lives, even when the degree represents less learning than it should.

Intuitively, many of education’s benefits depend on the skills that students develop in school. As workers, people need a range of skills—cognitive, socioemoafter controlling for the years of schooling completed, tional, technical—to be productive and innovative. As years of schooling do not predict growth once test parents, they need literacy to read to their children or scores are taken into account (figure 1.5), or they to interpret medication labels, and they need numerbecome only marginally significant. In other words, acy to budget for their futures. As citizens, people need what matters is less the years of education completed literacy and numeracy, as well as higher-order reathan the knowledge that students acquire while in soning abilities, to evaluate politicians’ promises. As school. Simulations show that providing all students community members, they need the sense of agency with basic cognitive skills could massively boost ecothat comes from developing mastery. None of these nomic outcomes, especially in developing countries capabilities flows automatically from simply attend- (figure 1.6). This finding suggests that cross-country ing school; all depend on learning while in school.

comparisons of the years of schooling completed—

Research on the benefits of education has begun especially when used to explain economic phenom- to reflect this distinction between schooling and ena—could be misguided if they do not account for learning. In the past, most empirical research equated the differences in skills acquired during those years education with schooling—whether measured by (box 1.3).

school enrollment, number of years of schooling, or

At the micro level, too, growing evidence shows degrees acquired—in part because of lack of other that skills acquisition determines how much individgood measures of education. But as the focus on uals gain from schooling. For example, learning—not learning has grown, some studies have explored the just schooling—matters in how education affects effects of the skills that students acquire. The results earnings. Across 23 OECD countries, as well as in a confirm the intuition: skills matter.

number of other countries, simple measures of founThe channel by which schooling accelerates ecodational skills such as numeracy and reading profinomic growth appears to be through boosting learnciency explain hourly earnings over and above the ing and skills. Thanks to the growing availability of effect of years of schooling completed. These effects large-scale student assessments, it is now possible to extend beyond the labor market. Across 10 low- and explore how learning mediates the relationship from middle-income countries, schooling improved measchooling to economic growth. While the relationsures of financial behavior only when it was associship between test scores and growth is strong even ated with increased reading ability. When people had acquired more schooling but not more literacy— such as “Parents love their children” or “Farming is which was common in these countries—financial hard work.” Yet even these highly imperfect meabehaviors did not change. Socioemotional skills sures of skills have considerable predictive and matter as well: various measures have been shown explanatory power. If better measures of skills were to significantly predict earnings over and above the available, skills would likely explain even more of the effects of schooling and cognitive skills.

impacts of education—and the role remaining for the

Learning matters for health, too. Numerous studsimple schooling measure (which typically retains ies have documented the benefits of girls’ schooling predictive power in these analyses) would be further on outcomes such as lower fertility or better child surdiminished.

vival, but these studies do not typically distinguish

Finally, learning promotes social mobility. The between learning and schooling. There are excepresearch cited earlier on intergenerational social tions, however. In Morocco, research showed that mobility in the United States also investigated which maternal education improved child health through educational mechanisms were responsible. One canits effects on the ability of mothers to acquire health didate is school quality based on inputs, such as school knowledge. Globally, data from 48 developing counspending and class size, and these measures did have tries show that learning is responsible for much of some predictive power. But learning outcomes turn these gains. Each additional year of female primary out to be especially important: the test scores of the schooling is associated with roughly six fewer deaths community in which a child lives (adjusted for the per 1,000 live births, but the effect is about two-thirds income of that community) are among the strongest larger in the countries where schooling delivers the predictors of social mobility later in life.

most learning (compared with the least).

The literature on the benefits of learning is still

Even limited measures of skills explain a lot. The growing, with much more research needed. But both measures used in the studies just noted are often common sense and the emerging research literature narrow, capturing only simple numeracy or reading make it clear that if investigators care about the benproficiency. Sometimes, the measures are coarse. For efits of education, they should focus on whether stuexample, the 48-country study of the relationship dents are learning—not just on how well schools are between schooling and health uses as its measure of equipped or even how long students stay in school.

literacy whether a woman can read a single sentence

Part II of this Report takes up this issue.