COVID-19 (coronavirus), first detected in Wuhan, China, in late December 2019, spread around the world, culminating in a global health emergency. It was quickly followed by a deep economic contraction in virtually every country.

The world tumbled into recession, and the overall global gross domestic product (GDP) shrank by 4.3 per cent in 2020. These figures hide immense human suffering. By December 2021, there were 14.9 million excess deaths globally—deaths attributable to the pandemic.

Poverty increased dramatically. There were 70 million more people living in extreme poverty in 2020 than in 2019—an 11 per cent increase. To put things in context, this increase in extreme poverty is roughly four times larger than the spike in poverty during the Asian financial crisis of 1997–98.

The consequences of the pandemic, however, were not limited to its effects on mortality, economic growth, or poverty. Households were ridden with stress, mental illness, domestic violence, teen pregnancy, and early marriage spiked in some settings.

Millions of children lost a caregiver or were orphaned. Many more missed out on vital nutrition and health care and suffered declines in early childhood development. Nearly one billion children missed a year or more of schooling and learned little, if anything, while schools were closed.

Tens of millions of young people were shut out of the job market or entered it with fewer skills and diminished prospects. Taken together, these effects represent a profound loss of human capital. Unless they are reversed, these losses will lead to declines in productivity and earnings as the children and youth of today become the labour force of tomorrow.

Human capital losses not only affect individuals through declines in their future earnings. They also can have negative economy wide effects. Human capital is one of the main drivers of economic growth, and so, anything that erodes it could result in lower growth rates for many years to come.

Indeed, the long-term costs of the pandemic—working through the reductions in human capital caused by the pandemic—are likely to dwarf the short-term costs. The erosion of human capital from the pandemic was greatest among poorer households. This erosion could lead to a sharp increase in inequality in the future—an increase that would compound the rising inequality already observed in many countries in recent decades.

Lower wages, more poverty, more inequality and less growth are an explosive mix. So what should be done? After quantifying the present collapse of human capital among young people under the age of 25, this report describes interventions that governments must put in place quickly to limit and reverse the damage.

Pandemic destroyed human capital at critical moments

The pandemic led to a sharp decline in human capital at critical stages of the life cycle. This report focuses on changes in human capital during early childhood (0–5 years), among school-age children (6–14 years), and in early adulthood (15–24 years). People younger than 25 today—those most affected by the erosion of human capital—will make up 90 per cent of the prime-age workforce in 2050.

Lost opportunities: The protracted effect of the pandemic on youth and young adults Youth (ages 15–24) is the period when people make the transition from mainly accumulating to utilising human capital. Young people may be in school, employed (whether formally or informally and in high or low-wage jobs) or idle. They can also engage (or not) in behaviours such as unprotected sex, drug use, criminal activity and gang membership. The decisions made by young people have long-term consequences—and the pandemic affected them in critical ways.

First, at the outset of the pandemic youth suffered deep employment losses. In 10 of the 12 countries analysed, there was a decline in youth employment in the second quarter of 2020, ranging from one percentage point in Vietnam to 11 percentage points in the Philippines. The exceptions are the two lower-income countries in the sample, where youth employment increased—by one percentage point in Ethiopia (2021) and by three percentage points in Pakistan.

It also reveals substantial differences in the pattern of recovery. By the end of 2021, youth employment had recovered fully and exceeded pre-pandemic levels in Brazil, Mexico, and Türkiye. On the other hand, there is no evidence of a recovery in South Africa, while in Bulgaria, Jordan, and—especially—Vietnam, youth employment continued to decline throughout 2021.

These job losses were compounded by declines in wages for young people in many countries. The implications of job losses are very different if young people who in normal times would have been employed seek more schooling instead.

To gauge whether this occurred in practice, figure ES.4 plots the effects of the pandemic on the proportion of young people, by gender, who are Not in Education, Employment, or Training (NEETs) in six countries that collected data on both employment and school enrolment.

The share of NEETs increased in most countries, even after schools reopened. Although the results of the analysis suggest that the differences in pandemic effects across gender were modest, structural impediments to women’s participation in the labour market are far higher in many countries than they are for men’s participation.

Declines in employment not matched by increases in school enrolment are a grave concern for two reasons. First, time out of the labour force is time spent without acquiring on-the-job experience, and such experience is a key way to build human capital.

Second, time spent out of work can lead to scarring in the labour market. In the United States, for example, individuals entering the labour market in a typical recession (associated with a 4–5 percentage point rise in unemployment rates) have initial earnings that are 10–15 per cent lower than those of similar cohorts entering labour markets in “normal” times. These negative effects may not fade out for a decade.

Finally, evidence suggests that beyond the labour market, the pandemic worsened a variety of outcomes for young people in some settings, including higher rates of teenage pregnancy, worse mental health, and declines in the development of key social-emotional skills and executive functions. However, fewer data are available on these outcomes than for schooling and employment.

Policies to reverse human capital losses

The pandemic eroded human capital at critical ages. Whether this erosion leads to a permanent reduction in future stocks of human capital depends on both the size of the initial drop in the level of human capital as well as the rate at which human capital accumulates thereafter. This point is illustrated in figure ES.5, which shows three possible paths for an individual.

Because human capital is built in a sequential, cumulative fashion, in the absence of remediation policies, early deficits will increase over time. This is the worst-case scenario in figure ES.5 (red line), which shows an increasing divergence in human capital accumulation relative to the pre-pandemic path.

It is also possible that, after the initial drop, human capital grows along a trend exactly parallel to its pre-pandemic path (yellow line). Two points are important about this scenario. First, because the effect of shocks tends to compound over time, setting out on this recovery path implies substantial investments in human capital over and above those that would have been made in the absence of the pandemic.

Second, this scenario only results in a partial recovery of human capital losses. Thus, in this scenario the stock of human capital would still be lower in the future.

Finally, as figure ES.5 illustrates, the only path that does not entail permanent losses in human capital is the one in which human capital grows at a higher rate after the initial shock (green line).

Achieving this convergent path is a tall order. It would mean, for example, that children would have to learn more in every grade in school than they were learning before the pandemic. However, this is the only path by which individuals and countries can achieve the levels of human capital they would have had in the absence of the pandemic.

Figure ES.5 clearly indicates the magnitude of the task ahead. It is possible to recover human capital losses, but it will require a substantial sustained effort—including, but not only, additional expenditures. With this framework in mind, this section points out what actions should receive priority at each stage in the life cycle. Some of the most important ones are summarised in table ES.1.

Young children have missed critical investments in health and preschool, and their levels of cognition, vocabulary, and early learning in math and language have in many cases dropped dramatically.

To prevent this poor start from amplifying into greater human capital losses as these children progress along the life cycle, policies should prioritise transfers for households whose income has not recovered, catch-up campaigns for vaccination and nutrition, parenting programmes to encourage more cognitive and social-emotional stimulation in the home, restored and expanded coverage of pre-primary education, and mental health counselling programmes for parents.

School-age children suffered from unprecedented school closures. They learned little if anything while schools were closed, and so their learning losses are massive. In addition, some children may be at risk of dropping out of school altogether, especially in lower-income countries.

To reverse learning losses, decision-makers should keep schools open and increase instructional time, assess learning and match instruction to students’ learning level, implement targeted catch-up policies such as tutoring for children who have fallen the furthest behind, and streamline curricula to focus on foundational learning.

To minimise dropouts, countries should track students at risk of dropping out, especially in transition years, and alleviate the financial constraints to school attendance. Youth have suffered from sharp declines in their job prospects, and the extent to which employment has recovered varies a great deal across countries.

Appropriate policies will therefore vary by country—in particular, by the extent to which there has been a recovery of both adult and youth employment. For countries where neither adult nor youth employment has recovered, policies should primarily be geared toward demand-side interventions that spur firms to start hiring again.

For countries where adult employment has recovered but youth employment has not, support for supply-side policies such as adapted training, job intermediation, entrepreneurship programmes, and new workforce-oriented initiatives for youth are all important.

Countries where both adult and youth employment have recovered should monitor developments in the labour market to ensure that the recovery has been equal across groups. In all countries, policies should recognise that youth are a diverse group and that skills are the best insurance against a crisis.

Building agile, resilient and adaptive human development systems for future shocks

The COVID-19 pandemic has arguably been the largest global shock to human capital in the past century. Moreover, countries will continue to face shocks in the future—health and climate emergencies, natural disasters, and macroeconomic crises—that, like the pandemic, can erode human capital across the life cycle.

In addition to impeding human capital accumulation at each stage of the life cycle, the pandemic has revealed systemic weaknesses in how governments integrate efforts across sectors to address the multidimensional nature of human capital deficits. In some cases, interventions in health will be most appropriate to address specific losses in human capital; in others, it may be that education or social protection policies are most effective.

But in most cases, countries need solutions that bring these sectors together into a holistic human development system. Evidence suggests that during the COVID-19 crisis, very few countries responded with integrated approaches, and most countries lacked the capacity to collect and link data from programmes in different sectors.

Such a human development system should build on existing sector-specific systems and individual programmes to take a broader look at how investments in human capital could be coordinated and how complementarities could be exploited. In a crisis, human development systems can help policy makers

resolve trade-offs across many competing needs in a constrained fiscal environment. To be effective, such systems should have three key characteristics:

1. They should be agile, resilient and adaptive and able to expand and contract quickly during crises to reach vulnerable groups.

2. They should have a mandate and authority to coordinate across sectors, identify interventions that are complementary and resolve trade-offs.

3. They should be data-driven, effectively use technology, and identify problems and “pain points” as a crisis unfolds.

To build these systems, countries need to invest in data collection and information systems to provide targeted support when required. They also need to leverage technology to deliver services (including cross-sectoral beneficiary registries, platforms and payment systems) and to invest in coordination mechanisms (including joint committees with representation from all ministries involved in aspects of human capital).

Finally, countries should invest in flexible payment systems and contractual mechanisms that allow for the rapid reallocation of resources in response to an evolving crisis. During the pandemic, most countries were able to expand existing programmes in all sectors, but especially in health and social protection.

For example, Argentina relied on expansions of Programa Sumar to ensure access to health care for the unemployed. Social protection programmes grew rapidly during the pandemic, reaching nearly 1.4 billion people (17 percent of the world’s population) over 2020–21. Notably, countries that had made earlier investments in technology, such as in interoperable beneficiary identification and payment systems, were able to expand social assistance coverage faster.

Some countries were also able to retool, redirect or reactivate programmes built in response to earlier shocks.

Sierra Leone adapted the social safety net systems set up to respond to Ebola, flooding, and landslides for use in rolling out cash transfers and providing additional types of support during the pandemic. Some countries were able to engage service providers beyond the traditional public sector to deliver services.

For example, in 2020 and 2021, the Indian state of Kerala contracted with more than 300 private hospitals so it could add them to the publicly funded insurance scheme for the poor and vulnerable to help sustain service delivery during the pandemic. This expansion—effectively more than doubling the number of private hospitals in the scheme—built on years of previous engagement of the private sector by the state government.

Uruguay was able to move from in-person instruction to remote learning during the pandemic thanks to Plan Ceibal, a functional remote-learning programme launched in 2007 that has helped ensure access to free laptops for students and teachers, has provided them with Internet connections and, critically, has trained teachers in remote instruction over the last decade.

By contrast, truly cross-sectoral approaches were rare. By and large, countries failed to assess the costs and benefits of specific sectoral policies jointly. The timing and length of lockdowns and mobility restrictions generally did not take into account the coverage of social protection, which made it difficult for households to comply with the restrictions.

Keeping schools closed for as long as many countries did, even after restrictions on the use of public transportation had been lifted, and markets, stores, movie theatres, and restaurants had opened, reflected a failure to balance competing risks—the risk of infection in schools (low) relative to the risk of learning losses (high)—and to update policy choices as new information became available.

A human capital recovery: What will it take?

Table ES.1 lists, for each phase of the life cycle, policy recommendations for recovering from pandemic- induced human capital losses and building resilience for the next shock. The list is long, particularly in the context of competing crises and tight fiscal space.

Which policies should countries put at the top of their human capital recovery list? First, an emphasis on transition periods in the life cycle—from early childhood to school age, from one level of education to another, from school to work, and from youth to adulthood—can help stem the accumulation of losses.

Transitions are defining moments: what happens during these stages may generate deficits that interfere with the entire trajectory of human capital accumulation in subsequent stages of the life cycle.

Second, the report also provides evidence on the full cost of each proposed policy. This full cost includes fiscal costs, as well as costs stemming from implementation complexity and political commitment required.

It highlights that many of the proposed policies do not require important fiscal efforts. Rather, they require institutional capacity building and political will.

Recovery is possible, and many countries have already made progress in recovering losses in human capital that arose from the pandemic. In Pakistan, for example, more than 1.2 million children missed immunisations during the first year of the pandemic, but intensive outreach efforts, in part enabled by an electronic immunisation registry, contributed to successful catch-up efforts that had immunised 76 per cent of these children by March 2021.

In Tamil Nadu, India, school-age children exhibited severe deficits in reading and math when they returned to in-person learning after pandemic-induced school closures. After six months, two-thirds of the losses had been recovered, with 24 per cent of the recovery attributable to government-run after-school catch-up programmes.

The pandemic may have led to a clear-cut collapse in human capital, but what is needed to get on a path of recovery is equally clear. Transforming the collapse into a recovery should begin now.