Over the past 60 years women and girls in developing countries have made enormous progress. Plenty of data illuminates this trend. Take life expectancy at birth: it went from 54 years in 1960 to 72 years in 2008. During the same period, we experienced the world’s fastest ever decline in fertility. These changes reflect gains for women on many fronts, including education, employment, access to reproductive health care and decision-making power, and it all happened much faster than it did in today’s rich countries. It took India 44 years and Iran just 10 to reduce the number of children born to a woman from six to three; in the U.S., it took 123 years. Two thirds of all countries have reached gender parity in primary education enrollment, and in more than a third, more girls than boys are in school. In a striking reversal of historical patterns, women now represent the majority of university graduates. And more than half a billion women have joined the labor market since the 1980s, which means that today four in 10 workers globally are women.

Yet amid all this progress, the remaining gaps are stubbornly resistant to closure. Although women on the whole live longer than men, in parts of the world such as sub-Saharan Africa they are as likely to die in childbirth as northern European women were in the 19th century. Women still hold fewer positions of power in politics and business. And although many women work for pay, they do so under circumstances that weigh their talent, abilities and education differently from those of men.

These inequities are outrageous, and addressing them is both a rights issue and a fundamental objective of development. Just as we think development translates into less poverty and improved access to services for everyone, we also see it as a process of expanding freedoms. It promotes the ability to seize opportunities and decide on a life path. For us, pursuing gender equality—especially in women’s access to income-generation opportunities—is a smart policy that itself leads to better development, not just the other way around.

Fully closing the well-being gap between the sexes requires specific, timely and intentional action. There are a few primary ways of doing this. First, remove the barriers that prevent women from having the same access as men to economic opportunities; this can generate higher productivity, and hence higher incomes, for all people. Second, increase women’s education, health and overall agency, which results in better outcomes for both mothers and their children. Third, put more women in positions of power. Giving women a representative voice shifts policies and spending toward issues such as sanitation, schooling and health. If only implementing these changes were as easy as identifying them.

A Persistent Gap

Unequal access to economic opportunity is one of the major hurdles preventing progress. This is a problem in all countries, rich and poor, and in all industries, from farming to entrepreneurship.

The first roadblock is simply a barrier to entry: women must be able to access the economic space to participate in it. Although the female labor force has grown considerably in most parts of the world, a substantial participation gap is visible everywhere between men and women—56 percentage points in the Middle East and North Africa. Even if women can manage to break through that barrier, they are probably not competing on a level playing field. Female farmers have a more difficult time obtaining fertilizers, machinery and improved seed varieties, so their yields are often lower. Likewise, female entrepreneurs often have restricted access to capital and credit. Sometimes that is because they are less likely to own land or other assets required as collateral; other times it is because application procedures require a male co-signer or because banks assign female applicants a higher risk rating than male ones. This means that women-owned businesses are often less profitable, creating a chicken-and-egg cycle that is difficult to break. When these issues are corrected, overall productivity dramatically increases. A 2016 study showed that closing the gender gap in entrepreneurship would raise productivity and incomes by 12 percent in sub-Saharan Africa and 38 percent in the Middle East and North Africa.

Even in places such as the U.S. and Canada, where the labor-participation gap between sexes may fall below 15 percentage points, other factors hinder equality. Men and women tend to concentrate in different economic sectors, and it is easy to observe that women are more likely to work in education and social services, whereas men are more likely to work in construction and transport. But what is less noticeable is that women often occupy the lower-paid roles in whatever sector they are in. For instance, women are often teachers, nurses and clerical workers instead of principals, doctors and supervisors. Even as entrepreneurs, they tend to concentrate in traditionally female sectors such as food or clothing production.

Although we could discuss the preferences women (and men) might have for certain sectors, these patterns are not random. The problem is that those “preferences” reflect the influence of ideas and norms about “women’s work” and “men’s work,” as well as other gendered attributes, such as the notion that women are better caregivers, whereas men are better suited for heavy physical work. The critical point is that heavily “feminized” sectors tend to have, on average, lower wages. And, of course, the gender difference in earnings is well known: Globally, a woman earns roughly 82 cents for each dollar a male worker earns. In Jordan and Côte D’Ivoire, the difference between incomes is about 40 percent. Rich countries are not exempt: New Zealand has a low of 9 percent; it is more than 30 percent in South Korea. But this difference reflects the different economic positions women hold. Even when you factor in characteristics such as education, sector and age that will make a female and male worker otherwise equal, the income gap persists. Female teachers in rural Pakistan, for instance, earn about 30 percent less than male teachers. Understanding—and addressing—the entrenched social norms behind these persistent gaps is key to closing them.

Stubborn Social Norms

One of the major causes of lower earnings and productivity stems from constraints on women’s time. Women devote far more of their days to care and household work than men, which in turn means that women have less time for paid work. Deep-rooted social norms drive the differences in “domestic” roles. What is most striking is that these norms, and hence patterns of time use, do not change even as women take up a larger share of market work. In Ghana, a wife still does more than 80 percent of the housework even when she brings home all the income. This imbalance is largely true elsewhere as well, including the U.S. Even in the most progressive places, this pattern reflects assumptions about the division of roles, identifying them as natural and biological instead of cultural. As a result, we all live as though this is just the way things are. You can see the effects everywhere in the workings of societies, households and markets. Look at school schedules, which are not consistent with a full-time employment day, or policies that allow mothers but not fathers to use sick days to care for an ill child. Breaking these patterns requires upending default expectations.

Credit: Jen Christiansen; Source: Ana Maria Munoz Boudet (data from World Development Indicators, World Bank)

Unfortunately, policies used to address the time-constraint issue primarily work around these norms instead of confronting them head on. Although the results are promising, it’s not a complete solution. The most popular example of this is providing increased access to child care services and improving parental-leave policies. As expected, the expansion of access to early child care and preschool services consistently leads to higher labor-market participation of women across all countries. Expanding child care accessibility (location, times of operation), affordability (direct and related costs) and capacity (removing waiting lists, including different ages of children) has a positive impact on the mother’s labor engagement.

There is ample evidence that this policy change often alleviates constraints on working mothers’ time. In the early 1990s Argentina embarked on a program to expand early childhood education. It resulted in the creation of 175,000 preschools over seven years. Researchers observed labor patterns as the program expanded and found a positive impact of between 7 and 14 percentage points in mothers’ employment. Most notably, these positive effects hold even when the preschool is only part-time.

But policies such as allocating more time to maternity leave—and adding paternity or parental leave to include fathers—have not always been as successful. In Germany, the expansion of maternity-leave regulation and coverage led more women to ultimately return to work with their prebirth employer. But when neighboring Austria shifted its maternity-leave duration in 1990 from one year to two years, it significantly reduced the percentage of women who returned to work at all.

Then there are the parental-leave programs in Scandinavian countries, which give fathers an incentive to share the load after the birth of a child. When they were first introduced, they were rarely used by men. But when such a policy in Sweden specifically gave fathers dedicated, nontransferable leave, it delivered positive results in terms of getting fathers to spend more time with their children. Studies have shown that these men remain very involved in parenting over time but slowly transition out of other domestic tasks as the leave effect ends.

Similarly, differences in human capital often affect women’s career trajectories. The root of the problem is not a difference in ability or capacity between the sexes—it is discrepancies in how we invest in and value women as workers. In developing countries, these differences appear as access to and completion of education; in more developed countries, where educational attainment is higher and both men and women attend university in large numbers, gender differences show up in fields of study. In the U.S., for example, women represent less than 35 percent of the degrees in the science and technical fields even though they account for almost 60 percent of college graduates.

When it comes to bridging human capital between men and women, governments must make investments beyond traditional education to include training, apprenticeships and other labor policies. Young job seekers are an important target of these policies. One notable example is the series of Jóvenes (youth) programs implemented across Latin America in the early 2000s, which encompassed a combination of training, internships and incentives to employers with a goal of breaking biases in hiring young workers. Across the board, these programs increased the probability of young women becoming employed and boosted their earnings.

But efforts to translate the lessons from the Jóvenes programs to other settings have had uneven results. An evaluation of a similar vocational-training program in Malawi found that family obligations limited female participation. The Jordan New Opportunities for Women pilot involved more than 1,000 young women from community colleges. Demand for the program was extremely high: many of these women successfully completed the vocational training, and half of those graduates found jobs thanks to wage-subsidy vouchers provided for employers. But the effects were short-lived, and no changes were found in employment or earnings. Here again, it appears that social norms and employers’ views about women heavily impeded successful replication of the initiative.

The third driver of gender inequality in economic opportunities is what we would typically call discrimination: the differential treatment of women by markets and institutions. When few women are employed in a certain sector, employers may hold biased beliefs about their qualifications. They may be reluctant to hire women because they associate extra costs, such as maternity leave, with female employees, or they may assume that women are not the primary breadwinners and therefore lack motivation. A number of studies that compare the reactions of employers with otherwise identical female and male CVs found substantial evidence of gender discrimination when it comes to recruitment and hiring. Moreover, access to jobs often involves gendered networks: when women are poorly represented in an occupation, they are less likely to become aware of opportunities and may be unable to find mentors. In many countries, gender-specific job advertisements and gender-biased selection criteria and recruitment are still common. In a surprisingly high number of countries in the Women, Business and the Law database, employers are not prohibited from asking job applicants about family status and family planning.

Government institutions can also treat men and women differently, often in ways that play against women’s interests. For example, in 45 percent of countries, women cannot work the same jobs or hours as men. Many former Soviet countries are particularly restrictive. In Russia, there are 100 occupations currently forbidden to women, including steelworker, firefighter and oil-well worker. Other countries require women to get male permission to accept jobs, open bank accounts or operate businesses. In Chile, husbands have the sole right to administer marital property, regardless of whose income or savings was used to purchase the property. In Pakistan, a married woman cannot register a business in the same way as a married man. In Mongolia, many women cannot work at night. In Yemen, a husband can object to a wife’s employment.

These three sets of differences between men and women—in responsibilities for care and housework, in human capital investments, and in treatment by markets and institutions—not only get in the way of equitable economic opportunities but often do so in ways that reinforce one another. Devoting much more time to care and housework may encourage women to self-select occupations that offer greater flexibility in hours but trap them into lower pay. This is especially true if formal employment options come with restrictions on part-time work, as is the case in many developing countries. Preemptively knowing it will be hard to get a job—and fit in—in a male-dominated field such as engineering or construction may discourage girls and young women from acquiring the education and skills necessary to pursue those opportunities. And repeated experiences of discrimination in job applications and hiring may push women toward informal self-employment or discourage participation in the labor market altogether, which starts the cycle of inequality all over again.

Fine-Tuning the Solutions

Why, if we have been able to clearly identify the roots of the gender gaps in economic empowerment, does change remain so elusive? For starters, the multiplicity of factors requires many actions coordinated toward the same goal. And given that the nature, structure and functioning of markets, institutions and norms vary widely across countries, a one-size-fits-all policy approach is impossible. As we have shown, policies that were successful in one context may face significant resistance in another.

Still, a few general principles have proved critical for successful policies. To be effective, policies and interventions need to target the multiple underlying factors that drive gender differences in access to economic opportunities. That is, they need to directly address the constraints on women’s time that arise from gendered social norms about care and housework, not circumnavigate them. To increase productivity, they must fill the gaps in information, skills and access to professional networks that constrain the opportunities of female wage workers, farmers and entrepreneurs. And they need to help build a more level playing field in markets and institutions by targeting discriminatory preferences.

Women perform the lion’s share of household duties, even in the most progressive countries. Here a woman prepares food on a farm in Bangladesh. Credit: Thomas Trutschel Getty Images (farm in Bangladesh)

One such promising example of a dynamic intervention hits many of those targets. The Empowerment and Livelihood for Adolescents program (ELA) implemented in Uganda by the nongovernmental organization Building Resources Across Communities provided young women aged 14 through 20 with life skills and vocational training at “girls’ clubs” where a mentor was present to lead activities. The clubs also served as a safe space in which adolescent girls could meet, socialize and recreate. A quarter of the girls in the 100 communities selected for the evaluation participated in the program, and the results were impressive. Girls in ELA communities were 72 percent more likely to be engaged in an income-generating activity four years later. But most important, the program shifted the girls’ attitudes and aspirations. For example, the participants were more likely to believe that women should earn money for their families and less likely to be worried about finding a good job in adulthood. Strikingly, those shifts helped teen pregnancy rates fall by 34 percent, and early entry into marriage or cohabitation dropped by 62 percent.

Policy makers, too, must learn from their own successes and failures and those of others. To achieve that, they must be able to clearly articulate both what goes into the policy and its expected impact. These initiatives also need to be piloted and properly evaluated before taken to scale. Too many interventions designed to support gender equality rely only on good intentions and intuitions. Whereas some might deliver in terms of outputs—say, the number of women who received training from a vocational program—many fail to deliver the expected final outcome, which in this case means getting more women to enter (and stay in) the labor market.

Policies and programs that show the greatest potential should be expanded and intensified. They must not be stopped even once the desired objective seems to be achieved—long-term follow-through is especially critical when tracking social changes as well as logistical ones. Progress requires constant, concerted and coordinated action to make markets, institutions and societies as a whole work more fairly for women, and changing the long-standing norms about women’s and men’s roles both at home and at work is a tall order. Resources for gender-specific interventions compete with other urgent development priorities, such as food security and poverty reduction. And goals like those can have no specific effects on closing gender gaps. That is why it is essential to rely on tested and proven tactics instead of leaning too heavily on advocacy and good intentions.

Greater gender equality is achievable with a combination of political will and evidence-based policies. Women and men of the future all stand to benefit from a more equitable society.