The gender impact of COVID-19 will not be known for some time

2022 Gender Report

Download PDF

GEM Report/Arete Photos

Globally, in the period between March 2020 and October 2021, schools remained closed or partly closed as a result of COVID-19 for 55% of instruction days, on average, with differences ranging all the way from countries that kept schools closed throughout the period to those that did not close them at all. Learning outcomes have undeniably been negatively affected, although the variety of experiences in school closure duration and in learning continuity opportunities means the impact will be unequally distributed among countries and among learners. There is little to suggest that governments have invested in remedial classes, especially to help those at a disadvantage, and there is uncertainty over learners’ ability to make up the losses. Data on enrolment are insufficient, but isolated examples do not suggest enrolment losses in primary and secondary education.

In the case of the gender impact of COVID-19, various factors suggested potential differentiation. Boys and girls did not face the same consequences in all countries in terms of access to devices, time use and early pregnancy risks. Some parents in Bangladesh, Jordan and Pakistan were reluctant to give girls access to smartphones (UNESCO, 2021c), which, on the other hand, have proved to be a less common aid in learning continuity than expected. Phone surveys of 19-year-olds during the pandemic showed that 70% of young women in Ethiopia but 35% of young men spent more time than before the pandemic doing household chores, while 42% of young women in Peru but 26% of young men spent more time looking after children (Ford, 2021).

Few studies have compared early pregnancies before and during the pandemic. The use of antenatal visit records from clinics is hampered by the estimated 39% decline in such visits due to lockdowns and related restrictions (Townsend et al., 2021). In a Kenyan study, teenage girls’ antenatal visits fell by 16% between the period from March 2020 to February 2021 and the same period before the pandemic (UNESCO, 2021c). A related concern is the increase in gender-related domestic violence reported by hotlines worldwide (Viero et al., 2021). In all, it will take years to assess whether the largest disruption of global education to date has a clear gender dimension implying a setback in the progress achieved towards gender equality in the past two decades.


Over the past 20 years, gender gaps in enrolment and attendance have been declining. Globally, the malefemale gender gap in out-of-school rates is close to zero in all three levels of education, although regions have progressed at different rates. There was rapid progress towards parity in primary and lower secondary education in Central and Southern Asia during the 2000s, while progress continued among youth of upper secondary school age in the past decade. Northern Africa and Western Asia is yet to achieve parity, as progress stagnated after 2012 in both primary and secondary education. Sub-Saharan Africa is the region furthest from parity. It also has experienced stagnation in the last 10 years. By contrast, in Eastern and South-eastern Asia there is a growing level of disparity that reached eight percentage points at the expense of young men in 2020, with disparity increasing also among adolescents towards the end of the decade (Figure 1).

Source: UIS database.

As mentioned above, these estimates are based on data up until the onset of COVID-19 and do not reflect its potential impact on education systems. There are significant concerns for low- and lower-middle-income countries that fully closed schools for more than two thirds of the time, notably Bangladesh (86%), Honduras (73%), Myanmar (80%), Philippines (93%) and Uganda (68%). While it is difficult to predict the medium- to longterm impact of school closures, such crises tend to more strongly affect those already behind. In four of the five countries, there is a gender gap at the expense of boys, while in Uganda girls are at a disadvantage (Figure 2).

Source: UIS database

Behind the global numbers, pockets of exclusion remain. In sub-Saharan Africa, the countries with the largest gender gaps at the expense of young women of upper secondary school age, where out-of-school rates for females are 20 percentage points higher than for males, include Guinea and Togo. A gap of about 15 percentage points is observed in Cameroon, Chad, Uganda and Zambia. (Figure 3).

Source: UIS database and World Inequality Database on Education.

The gender gap at the expense of young women in sub-Saharan Africa tends to be higher when overall out-ofschool rates are high; it is likely to close or even reverse when out-of-school rates decline. Even so, at similar levels of educational development, countries have followed different trajectories, even within Africa’s subregions. For instance, the Central African Republic, Côte d’Ivoire, and Guinea have seen almost no change in the gap, while neighbouring West African countries, such as Burkina Faso, Gambia and Mauritania, have closed or even reversed it. Still, closing the gender gap may not be accompanied by an overall improvement in school attendance. In Burkina Faso, the effects of conflict and displacement, which have halted progress, have been significant since 2017 (Figure 4).

Source: UIS database.

Governments need to examine the reasons and barriers that keep children, adolescents and young people out of school to understand the potential impact of policies, including those that may affect boys and girls differently (Box 1).

Source: GEM Report team analysis based on data from the 2016–17 Malawi Fourth Integrated Household Survey, the 2018–19 Nigeria Living Standards Survey and the 2018 Sierra Leone Integrated Household Survey.

New analysis for this report highlights the fact that progress towards parity in out-of-school rates does not necessarily mean parity in completion rates is achieved. The official definition of the completion rate is the percentage of those aged three to five years above the official graduation age from a particular education level who reach the last grade of that level. This is considered ‘timely’ completion since, in many countries, children start school late and repeat grades. Globally, at each education level, females enjoy an advantage of two percentage points in ‘timely’ completion rates at each of the three levels.

However, in many parts of the world, children, adolescents and youth complete each level of education even later. While globally there is near parity in ‘ultimate’ completion at each level, substantial regional differences are observed, especially for sub-Saharan Africa. At the lower secondary level, the gender gap is just one percentage point in ‘timely’ completion but eight percentage points in ‘ultimate’ completion. At the upper secondary level, the corresponding gender gaps increase from three to seven percentage points (Figure 6). This finding reflects the fact that boys can afford to complete secondary school later than girls, who are forced by gender norms to marry and have children early.

Source: GEM Report and UIS estimates available at the VIEW website

The analysis also shows that progress towards parity over the past 20 years has been much slower in sub-Saharan Africa than in other regions where females were at a major disadvantage in 2000. For instance, the gender gap in timely lower secondary completion fell by 13 percentage points in Central and Southern Asia and by 10 percentage points in Northern Africa and Western Asia between 2000 and 2020 but by less than 4 percentage points in sub-Saharan Africa. The gender gap in timely upper secondary completion fell by 6 percentage points in Central and Southern Asia and by 9 percentage points in Northern Africa and Western Asia but by 2 percentage points in sub-Saharan Africa.



Recent years have seen an increase in evidence on learning outcomes in primary and secondary education from long-established international and regional assessments (Table 1). These assessments’ proficiency levels in reading and mathematics correspond to the minimum proficiency levels used in SDG global indicator 4.1.1 in early grades, at the end of primary and at the end of lower secondary education. The gender gap in the percentage of students achieving minimum proficiency varies by subject and level.

In primary education, more girls achieve minimum proficiency in reading than boys. There are only five countries where the gap is at the expense of girls in the early grades, all of them in sub-Saharan Africa, two of them low-income (Chad and the Democratic Republic of the Congo), two lower-middle-income (Benin and Côte d’Ivoire) and one upper-middle-income (Gabon). Of those only the two low-income countries have a small gap at the expense of girls at the end of primary (Figure 7).

Source: GEM Report team estimates based on the 2019 rounds of the LLECE, PASEC and SEA-PLM surveys, the 2016 PIRLS and the 2018 PISA.

In the other 95 countries with data, the gap is at the expense of boys at the end of primary or the end of lower secondary education. The largest gap is observed in Saudi Arabia, where 77% of girls but 51% of boys in grade 4 achieve minimum proficiency in reading, according to PIRLS results. The gap in favour of girls is 19 percentage points in Bahrain, the Islamic Republic of Iran and Oman, also in PIRLS; 14 points among grade 6 students in Pacific countries taking part in PILNA (PILNA, 2019); 15 points in Malaysia according to grade 5 data from SEA-PLM; and 11 points in the Dominican Republic, according to grade 3 data from LLECE. The gender gap in reading tends to increase towards the end of primary and into lower secondary education. Even in countries where no gap is observed in early grades, including Lithuania and Norway, the gap in favour of girls rises to roughly 15 percentage points by age 15, according to PISA.

Boys tend to perform better than girls in mathematics, particularly in early grades (Figure 8). For instance, the gender gap favouring boys in grade 2 was 13 percentage points in Chad, 10 points in Benin and 8 points in the Democratic Republic of the Congo, according to PASEC results, and 8 points in Canada among grade 4 students taking TIMSS. This gender gap is reduced later in students’ trajectories. It is reversed to favour girls, among grade 8 students in TIMSS, in Bahrain (8 points), Jordan (7 points), Oman (13 points), Romania (8 points) and Turkey (5 points). The gap favouring girls is 6 percentage points among grade 6 students in Pacific countries taking part in PILNA (PILNA, 2019). Overall, the gender gap in mathematics has shrunk over time, further strengthening the evidence that gender gaps are social constructions and unrelated to gender-specific disposition factors (Borgonovi, 2021; Meinck and Brese, 2019).

Source: GEM Report team estimates based on the 2019 rounds of the LLECE, PASEC, SEA-PLM and TIMSS surveys.

There are no science assessment data in low-income countries, but results in middle- and high-income countries suggest girls have an advantage (Figure 9). Large gaps are observed in the Arab States, with the percentage of girls achieving minimum proficiency exceeding that of boys by 15 percentage points in Bahrain and Saudi Arabia and by 8 points in Kuwait among grade 4 students, and by 21 points in Bahrain, Jordan and Oman and 12 points in Kuwait among grade 8 students, according to TIMSS results. This region-specific gap, it has been argued, is consistent with the idea that girls have more interest and confidence in learning, and value it more, than boys do. It is posited that boys in the region are less motivated than girls because they are more likely to eventually attain a higher social status independently of their efforts, whereas girls face greater barriers and fewer choices (Ghasemi and Burley, 2019).

Source: GEM Report team estimates based on the 2019 rounds of the LLECE, PASEC, SEA-PLM and TIMSS surveys.

Looking beyond average levels of learning achievement for boys and girls highlights three significant patterns. First, the distribution of learning scores is more variable among boys than among girls (Figure 10). For example, in grade 4 the variability in reading scores is 23% higher for boys than for girls in Bahrain and 17% higher in the Islamic Republic of Iran, PIRLS results show. In the same grade, variability in mathematics scores is 19% higher for boys than for girls in Saudi Arabia and 12% higher in Australia and the Republic of Korea, according to TIMSS results. The excess variability of boys’ scores relative to girls’ increases slightly in secondary education. For instance, among grade 8 TIMSS participants in science, it is 22% higher in Kuwait and 16% higher in Malta. Among 15-year-olds in the PISA reading test, it is 20% higher in Israel and 16% higher in Finland.

Note: A ratio value of 1 indicates equal distribution of scores for boys and girls. Values above 1 point to higher variability among boys and lower variability
among girls. Richer countries are indicated with smaller data points.
Source: GEM Report team estimates based on the 2013 LLECE, 2019 PASEC, 2018 PISA, 2019 SEA-PLM and 2019 TIMSS.

Second, in mathematics the higher variability of boys’ scores is usually the result of more boys achieving high scores. Girls are under-represented at the top of the mathematics skills distribution, even though they perform better than boys on average (Baye and Monseur, 2016). For instance, in TIMSS the share of boys above the advanced benchmark in mathematics exceeds the share of girls for almost all countries, both at grade 4 and 8 levels. The difference is much more pronounced if expressed in relative terms (the ratio) than in absolute terms. For instance, among grade 4 students in Australia there is near parity, with 69% of girls and 70% of boys achieving minimum proficiency in 2019. But there is a large gender gap in that 37% more boys achieved advanced proficiency (12.3% vs 8%).

In some cases, the relative gap may be misleading: in Pakistan, only one girl achieves the top proficiency level for every three boys but the absolute gap between them is only 0.1 percentage points because overall performance levels are low. But in France and Spain, where fewer than one girl achieves the top proficiency level for every two boys, the absolute gap between them is more sizeable, at two and three percentage points, respectively.

Third, when girls perform better than boys in mathematics and science, they perform even better than boys in reading; in other words, gender gaps in reading are positively correlated with those in mathematics and science. For instance, in countries including Malaysia and Thailand, where girls perform better in mathematics, their advantage in reading is also higher. Similarly, in countries where girls perform better than boys in science, including Albania and North Macedonia, the gap in reading favouring girls is even larger. The correlation of gender gaps in reading and science (0.86) is higher than the correlation of gaps in reading and mathematics (0.61) (Figure 12).

Source: GEM Report team estimates based on the 2019 rounds of the LLECE, PASEC, SEA-PLM and TIMSS surveys.

What these data do not tell, however, is that these relative gaps, which eventually transform into career expectation gaps, are ultimately inextricably linked to stereotypical norms about the types of careers that best ‘suit’ men and women.

Girls perform better in mathematics in more gender-equal societies, as PISA data have shown (Guiso et al., 2008). This correlation has been demonstrated at primary school level in low- and middle-income countries, where gender gaps in learning are lower when economic, political, educational and health opportunities are more equally distributed between women and men, as summarized by the World Economic Forum’s Gender Gap Index.

This finding probably underlies the lower probability of girls opting for science, technology, engineering and mathematics (STEM) careers despite their an advantage over boys in mathematics and science in many countries. As they enjoy a comparative achievement advantage and higher self-concept in reading in relation to mathematics and science (Breda and Napp, 2019), this may be a more important factor in choosing employment. Conversely, in countries including Argentina and Colombia, where the gender gap in reading is smaller or non-existent, boys outperform girls in mathematics.

Likewise, girls perform better in mathematics in countries with higher participation rates in STEM subjects in tertiary education (van Langen et al., 2006). In lower secondary education, the presence of girls at the top of the mathematics skill distribution is positively associated with the proportion of female graduates in STEM careers.


The number of students enrolled in tertiary institutions increased from 100 million in 2000 to 235 million in 2020. The share of women in the student population rose from 48.8% to 51.9% in this period. There was already gender parity in the global gross enrolment ratio in 2000. Since then, the gender parity ratio has increased nearly every year, reaching 113 women for every 100 men enrolled in 2020. There is now disparity at the expense of men in

all world regions except sub-Saharan Africa, where the 2000s were a lost decade, with the gender parity index remaining constant at 67 women enrolled for every 100 men until 2011. Since then, it has steadily increased, reaching 76 women enrolled for every 100 men in 2019 (Figure 13).

Source: UIS database

Of the 94 countries with data for both 2000 and 2020, there was disparity at the expense of women in just 15. In some countries, including Cambodia and Nepal, where 40 to 50 women were enrolled for every 100 men in 2000, parity was almost achieved by 2020. In the United Republic of Tanzania, the number of women enrolled for every 100 men increased from 31 in 2000 to 84 in 2020. Yet despite progress, in some countries the gender gap in tertiary education remains wide, with only 47 women in Benin, 55 in Burkina Faso and 60 in Ethiopia for every 100 men enrolled. At the opposite extreme, there are about 50 men for every 100 women enrolled in the British Virgin Islands, Iceland and Namibia – and as few as 40 in Tonga and 14 in Qatar (Figure 14).

Source: UIS database.


Globally, it is estimated that 771 million adults lacked basic literacy skills in 2020, among which 98 million were aged 15 to 24. Females accounted for 63% of all adult illiterates and 55% of youth illiterates. Among adults, 83% of women and 90% of men were literate, a gap of 7 percentage points, whereas the gender gap was only 2 percentage points among youth (Table 2). The gender gap in adult literacy was largest in Central and Southern Asia (15 points) and sub-Saharan Africa (13 points). In Benin, Central African Republic, Guinea, Liberia and Mali, there were 60 literate women for every 100 men. Women living in rural areas were left even further behind. In rural Guinea, about 14% of women were literate in 2018 compared with 39% of men, while in urban areas about 52% of women were literate compared with 77% of men, which means the gender parity index was almost twice as high in urban areas (0.68 vs 0.35).


Progress over time has remained slow. In sub-Saharan Africa, female youth literacy increased by less than one percentage point a year between 2015 and 2020. More than one in four young women in sub-Saharan Africa are still illiterate. Slow progress in raising literacy rates means that, in absolute terms, the number of illiterate people has hardly changed. Globally, the substantial decline in Eastern and South-eastern Asia to 52 million illiterate women was offset by an increase in sub-Saharan Africa to over 127 million.

Acquisition of literacy skills is crucial for effective functioning in society, including for developing health literacy skills, which became critical during the COVID-19 pandemic. A study in India found that women who had participated in an adult literacy programme – conducted before COVID-19 spread and containing no specific material on the pandemic – had considerably higher COVID-19 knowledge than their illiterate counterparts. Over 80% of the newly literate women were aware of the symptoms, compared with 16% among the illiterate control group (Das et al., 2021). Adult literacy and numeracy programmes were hit hard by COVID-19, however. A rapid assessment by UNESCO in mid-2020 suggested that 90% of adult literacy programmes were partly or fully suspended (UNESCO, 2020d). Moreover, such programmes were mostly absent from countries’ initial education response plans (UNESCO, 2020a). Exceptions included Chad, which incorporated adult and non-formal education in its COVID-19 response plan, and Senegal, which, after developing a distance learning plan for children and youth, established a working group to focus on basic education for youth and adults (UNESCO, 2020d).


Globally, there were 83 million teachers in pre-primary, primary and secondary education in 2020, 27 million more than in 2000. Overall, women are over-represented in the teaching force. In pre-primary education, their share increased from 92% to 94% in this period while in primary education it rose from 59% to 67%, with the largest increases, of about 17 percentage points, observed in Central and Southern Asia and in Eastern and South-eastern Asia. Sub-Saharan Africa was the only region where women made up less than half the primary school teaching force. The imbalance was even greater in secondary education, where women’s share was 32% and had increased by only one percentage point in 20 years (Table 3).

As teaching is widely seen as a woman’s profession, gender segregation in the teaching force tends to take place in both state and non-state institutions, although some variations are observed. In Brazil, Japan and the Republic of Korea, there are more male teachers in private schools, while in France, Kazakhstan and Turkey such schools have more female teachers (Cherng and Barch, 2021). In Burkina Faso, 48% of public and 40% of private school teachers are women, but the gap is 30 percentage points or more in some regions, such as Plateau Central and Nord (Lange et al., 2021). In India, women account for 62% of teachers in unaided private schools and up to 73% in urban unaided private schools. The overall feminization of teaching in the private unaided school sector should be seen in relation to the fact that their teachers are paid much less (UNESCO, 2021b).

Around the world, teachers have been directly affected by the COVID-19 pandemic. In the United States, well over 1,000 educators had died as of September 2021 (Maxwell, 2021). In India’s Uttar Pradesh state, more than 1,600 teachers died, and South Africa had a similar toll (Ndaba, 2021; Rashid, 2021). The pandemic also posed challenges to teachers’ professional lives, partly because of the digital divide. School closures found many teachers unprepared for the move to remote learning, uncertain about their role and unfamiliar with the technology. This caused stress particularly among female teachers. In Slovenia, female primary school teachers reported high levels of stress when delivering online lessons, associated with the preparation of teaching materials (Loziak et al., 2020).

In India, jobs losses were a concern for teachers, especially those in private schools, most of which run on tight budgets and depend on fee collection. The feminized nature of the private school teaching force and the impact of the pandemic on private schools greatly increased the likelihood of women teachers having lost their jobs during the COVID-19 crisis (UNESCO, 2021b).