Listening to Young Lives at Work Covid-19 Phone Survey: First Call shows widening inequality

With the virtual fieldwork just completed on our second Covid-19 phone survey, we reflect on seven key findings from the first call that point to widening inequality. 

Young Lives at Work adapted to the coronavirus situation to implement a Covid-19 phone survey about the pandemic's short and medium term impacts. You can read more about Young Lives' response to the pandemic here and here. In June, we wrote about the design of the first call and what we were hoping to find out.  After virtual training in four countries, seven weeks of on-line fieldwork, almost 40,000 phone calls, data cleaning, coding, merging, and preliminary analysis, the headline reports of our first Covid-19 Phone survey were released in August and the data are available here.  In this blog we share our approach to the first call and key findings. 

We interviewed a total of 9,541 individuals in the Older and Younger Cohorts, reaching almost 91% of those who we were aiming to reach (see our attrition report here).  We discovered seven common findings, as well as key differences between the four countries, Ethiopia, India, Peru and Vietnam (below). The crisis is impacting the poorest most severely and widening inequalities, with poorer people less able to protect themselves, get information, work from home or stay in education.

Seven Findings from the Young Lives at Work Covid-19 Phone Survey


1. Poverty and a lack of information has impacted people’s ability to take precautionary measures against Covid-19 infection, especially in Ethiopia.

2. The economic effects of lockdown policies have been more significant than the health impacts to date - although many households in Peru and India are likely to have been exposed to Covid-19.

3. Across all countries, except Vietnam, many young people are going hungry. Although government support is well targeted, it is not sufficient in Peru and Ethiopia.

4. Remote working has been the exception, not the rule. Job losses or suspension without pay are widespread, even in Vietnam, the least affected country.

5. Education of 19-year-olds, in all countries, has been severely disrupted while access to online learning has been highly unequal.

6. Caring responsibilities increased for 25-year-olds with the burden still tending to fall on young women most of all.

7. Levels of anxiety about the current situation are high, especially in India​.

Poverty and a lack of information impact people’s ability to take precautionary measures against Covid-19 infection

According to the WHO, the most common symptoms associated with COVID-19 are a dry cough, fever, and tiredness. Public information in each country has emphasized a list of symptoms which overlap with this but there are some differences (e.g. in Peru difficulty breathing is highlighted on the official government website). Most respondents were able to identify at least two of the symptoms, and those with internet access were most likely to be informed. 

We asked respondents about the five behaviours which are widely recommended as a means of preventing infection: social distancing, washing hands more frequently, avoiding handshakes or physical greetings, avoiding groups and wearing protective gear when outside. Only slightly more than half (56%) of the Ethiopia sample adhered to all five, rising to 69% in Vietnam, 72% in India and 84% in Peru. Those with internet access or residing in urban areas showed a higher degree of compliance with these measures, and overall, females tended to comply more than males. In Peru, social distancing is the behavior with the least adherence, especially among vulnerable households.

We also assessed the resources available in households to comply with the World Health Organization's (WHO) recommendations on self-isolation, through an adapted version of the Home Environment for Protection Index (HEP) developed by Brown et al., 2020. The HEP measures the ability to receive reliable information on virus protection and the presence of available space and facilities suitable for implementing social distancing within the household.  

Young Lives Households: Home Environment for Protection


Information/ communication device

2 or less people per room

Household toilet

Household piped water

HEP score all indices

























Notes: Proportion of households. Adapted from Brown et al, 2020. Detail on the Young Lives HEP index can be found here.

Peru and Vietnam have relatively high averages for the protective index, but in Ethiopia, under a third have access to a piped water source, and the number of people sharing a room also makes it difficult to implement self-isolation when someone is believed to be infected with Covid-19. We find that young people who are the most vulnerable, are living in households with lower protection. Households in the higher HEP group (who are wealthier, on average) are also more likely to follow all behaviours, relative to those in the lower HEP group. 

A cause for concern is the number of those employing ineffective (though benign) preventative measures. A large number reported eating garlic or ginger to protect themselves against the virus, as well as drinking lemon, or adding hot pepper to food to prevent infection.

So far, the health impact of the crisis has been higher in Peru and India than in Ethiopia and Vietnam. In both Peru and India, approximately 6% believed someone in their household had been infected. In contrast, this figure was fewer than 1% in Ethiopia, and almost zero in Vietnam. Whilst our sample are not representative of the national populations, the rates do reflect the situation in each country. Of those who were believed to be infected in Peru and India, only around one-in-three were tested for the virus in both countries.

Many young people are going hungry in all countries except Vietnam: although government support is well targeted it is not sufficient in Peru and Ethiopia. 

The crisis has impacted food security in Vietnam notably less than in the other countries. One in six Young Lives households in Peru, India, and Ethiopia reported running out of food at some point since the beginning of the crisis. This percentage was even larger among households that faced food shortages (food insecurity) in our last visit in 2016 (about twice as high in India). In Vietnam, the overall proportion was much lower, at 4%.

[[{"fid":"10059","view_mode":"default","fields":{"class":"media-element file-default cke_widget_element","data-delta":"4","format":"default","field_file_image_alt_text[und][0][value]":false,"field_file_image_title_text[und][0][value]":false},"type":"media","field_deltas":{"4":{"class":"media-element file-default cke_widget_element","data-delta":"4","format":"default","field_file_image_alt_text[und][0][value]":false,"field_file_image_title_text[und][0][value]":false}},"link_text":null,"attributes":{"class":"file-default media-element","data-delta":"4"}}]]

Source: Young Lives COVID-19 phone survey. We defined food insecure households as those reporting “sometimes do not eat enough” or “frequently do not eat enough” and food secure households those reporting “eating enough but not always what they would like” or “eat enough of what we want”.

Government assistance has reached our respondents to very different degrees. About 92% of the households in India received at least one form of support from the government during the lockdown, although in many cases the support consists of a small basket of food or face masks.  This compares to around half of those in Peru, falling to just 6% in Ethiopia. In all countries it was relatively well targeted, reaching proportionately more of those households that reported food insecurity in a previous visit. However, in Peru the proportion of the most vulnerable households that received a direct cash transfer was far from universal. This could signal either a targeting problem, a delay in payments, or both. Moreover, the size of the transfer appears insufficient (as its value corresponds to about 82% of a minimum wage per family for the entire period).  

Job losses or suspension without pay are widespread even in Vietnam, the least affected country, and remote working is the exception rather than the rule

Many of our 25-year-old respondents lost their jobs. This was particularly severe for respondents who had been informal workers with no written contract in our last visit. In Peru and India, 7 out of 10 respondents had reduced or lost their source of income due to lockdown, 6 in 10 in Vietnam, and 4 in 10 in Ethiopia.  A concentration of income losses among those in the informal sector is an indication of this group’s additional vulnerability to the economic consequences associated with the pandemic. However, it is important to state that everyone, even those who were formal workers prior to the crisis, was severely affected.

The proportion of those who lost income or employment was also relatively higher in urban areas compared to rural areas and a higher proportion of males experienced these losses in both locations. 

Remote working has been possible only for a lucky minority of 25-year-old workers living in urban areas. The highest proportion (28%) in India were able to work from home during the outbreak, falling to 20% in Vietnam, 18% in Ethiopia and 17% in Peru. The percentage is much higher within households who are better equipped for protective measures against Coronavirus (High-HEP). Presumably, this is due to the availability of better infrastructure (e.g. access to internet, computer ownership) and the nature of the work activities performed. 

Education of 19-year olds in all countries has been severely disrupted and access to online learning has been highly unequal

With schools and universities closed very early on in the outbreak in all countries, the interruption to education was striking. Inequalities in those whose studies were interrupted are clear both across countries, gender and wealth. Access to study from home was slightly higher for females than males in all countries, and wealth and parental education almost doubled the chances of being able to study at home. In Vietnam, the vast majority of our 19-year-old cohort (almost 90%) accessed remote learning, falling to 70% in Peru, and 38% in India. In contrast, only 28% in Ethiopia continued to learn remotely, this fell to 14% if their parents had no education. 

This echoes the findings of another Young Lives study that interviewed headteachers in Ethiopia and India, see here for more information

Caring responsibilities increased for 25-year olds and the burden still tends to fall on young women

Although slightly more 19 year-old women have been able to continue their studies online, wide disparities are clear when looking at caring. In all countries except Peru, more than double the number of young women, relative to young men, have had to take on extra caring responsibilities during the lockdown.  The disparity is particularly striking in India and Ethiopia.

[[{"fid":"10056","view_mode":"default","fields":{"class":"media-element file-default cke_widget_element","data-delta":"1","format":"default","field_file_image_alt_text[und][0][value]":false,"field_file_image_title_text[und][0][value]":false},"type":"media","field_deltas":{"1":{"class":"media-element file-default cke_widget_element","data-delta":"1","format":"default","field_file_image_alt_text[und][0][value]":false,"field_file_image_title_text[und][0][value]":false}},"link_text":null,"attributes":{"class":"file-default media-element","data-delta":"1"}}]]

Source: Young Lives COVID-19 phone survey.

Levels of anxiety about the current situation are high, especially in India

How is this impacting on young people’s stress levels? We asked respondents whether the statement "I am nervous when I think about current circumstances" applies to them. We found that stress levels are worryingly high – in India more than 90% of all the young people indicated it applied or strongly applied to them. In Vietnam and Ethiopia 65% agreed with these statements. Peru had surprisingly the lowest anxiety levels, with just under 50% feeling nervous. In the second call we have asked more detailed questions about mental health based on validated scales.

What happens next? 

The data are available on open access (see here).  Due to the nature of the confidentiality agreement that Young Lives has with the families, the datasets are anonymized, and information on geographical location is limited. 

The second call fieldwork has now been completed. We will release headline findings in November. This second phone call has gone into more depth about young people’s labour market experiences, to understanding the medium-term impacts of the pandemic on their work life, their home life, and their education. More specifically, it contains information on the household socio-economic status, food security, labour, education, time use, health (including mental health) - the main themes of the YL survey that can be implemented over the phone.

This is a longer version of our blog first published in The Conversation in August, here.  Follow us on Twitter @yloxford for news on Young LIves at Work. 


Smarter through Social Protection? Evaluating the Impact of Ethiopia's Safety-Net on Child Cognitive Abilities

Social protection
Working paper

In this working paper, the authors provide new estimates of the impact of Ethiopia's Productive Safety Net Programme (PSNP) in Ethiopia on child cognitive achievement in the medium term. The programme is the second larget in Africa, and has been rolled out to almost 10 million beneficaries since 2005. The authors exploit four rounds of the Young Lives panel data spanning 2002-2013 and find a small but significant effect of the programme on cognitive achievement as measured by numeracy skills. They examine heterogeneity of impacts via "graduation" from the scheme, and find a positive effect on children in households that graduated just before 2013, but none for children in households that remain in the programme. 

To access the paper, please visit:

Social Protection For All Ages? Impacts of Ethiopia’s Productive Safety Net Programme on Child Nutrition

Social protection
Journal Article


  • An Ethiopian public works program improves child nutrition in the medium term.
  • We do not find significant differences in impact between children aged 5, 8, and 15.
  • Impacts associated with improved household food security and fall in child labor.
  • Height-for-age captures medium-term program impact better than consumption.


We investigate the impact of a large-scale social protection scheme, the Productive Safety Net Program (PSNP) in Ethiopia, on child nutritional outcomes. Children living in households that receive cash transfers should experience improved child nutrition. However, in the case of the PSNP, which for the majority of participants is a public works program, there are several potential threats to finding effects: first, without conditionality on child inputs, increased household income may not be translated into improved child nutrition. Second, the work requirement may impact on parental time, child time use and calories burned. Third, if there is a critical period for child human capital investment that closes before the age of 5 then children above this age may not see any improvement in medium-term nutritional outcomes, measured here as height-for-age.

Using a cohort study that collected data both pre-and post-program implementation in 2002, 2006 and 2009, we exploit several novel aspects of the survey design to find estimates that can deal with non-random program placement. We present both matching and difference-in-differences estimates for the index children, as well as sibling-differences.

Our estimates show an important positive medium-term nutritional impact of the program for children aged 5–15 that are comparable in size to Conditional Cash Transfer program impacts for much younger children. We show indicative evidence that the program impact on improved nutrition is associated with improved food security and reduced child working hours. Our robustness checks restrict the comparison group, by including only households who were shortlisted, but never received PSNP, and also exclude those who never received aid, thus identifying impact based on timing alone. We cannot rule out that the nutritional impact of the program is the same for younger and older children.



Catherine Porter and Radhika Goyal (2016) 'Social Protection For All Ages? Impacts of Ethiopia’s Productive Safety Net Programme on Child Nutrition', Social Science & Medicine 159: 92–99.

Download from the journal website

Childhood Shocks, Safety Nets and Cognitive Skills

Poverty and shocks
Social protection
Working paper

A growing body of evidence emphasizes that shocks in early childhood can have irreversible effects on long-term child welfare and poverty. A number of studies have investigated the effects of shocks on child nutrition and health. However, evidence on the effects of shocks in early childhood on child cognition, particularly when measured after the early childhood window during preschool and beyond, is scarce. Given its history of recurring natural and economic shocks, Ethiopia presents a compelling context in which to seek a better understanding of this question. Using child-level panel data from rural areas of Ethiopia, this paper analyzes effects of both economic and non-economic shocks on child cognition skills measured after the early childhood age window.

The identification strategy for the study exploits the timing of shocks that occurred between the two periods in which cognitive scores of children were measured - between 2006, when the children in the study were between 4 and 6 years old, and 2009, when they were between 7.0 and 9.6 years old. Using difference-in-differences analysis and controlling for child, household, and village-level baseline characteristics, we find that exposure to these shocks significantly decreased child cognitive skills. Specifically, exposure to drought reduced child cognitive skills by 0.18 standard deviations, while food price inflation undercut cognitive skills by more than one standard deviations (0.98 due to cereals price inflation and 0.47 due to inflation in meat prices). Divorce contributes to a reduction in child cognitive abilities by another 0.39 standard deviations. On the other hand, the safety net program put in place in 2005 to protect households from the economic effects of such shocks mitigated the reduction in cognitive skills by 0.18 standard deviations. These results are in line with comparable studies from similar contexts and suggest that policies that aim at mitigating these shocks are crucial for child welfare and for future human capital development. The results also suggests that social safety nets, if well designed and implemented, can help check the long-lasting detrimental effects of shocks experienced during early childhood.

Keywords: Shock, safety nets, cognitive skill, children, difference-in-differences, Ethiopia

The Impact of Social Protection Schemes on Girls’ Roles and Responsibilities

Social protection
Children's work and time-use
Journal Article

The focus of this article is the effect on adolescent girls' roles and responsibilities of public works schemes or cash transfers, which are the main forms of social protection in developing countries. Increasing participation in social protection is intended to enhance the development of girls in participating households, but evidence on their school participation and workloads suggests that the reverse may be happening. The article probes what happens to girls' roles and responsibilities when households participate in social protection schemes in rural Ethiopia and Andhra Pradesh. It argues that effects are complex, and often context-specific; however, the assumption that "beneficiaries" benefit means that negative impacts are rarely acknowledged. The article combines a review of other papers addressing the effects of social protection on children's work with analysis of quantitative and qualitative data, recognising that this question cannot be answered with a methodology that considers girls' schooling or workloads in isolation.

Available on the journal publisher's website.

Growing Up in Ethiopia and Andhra Pradesh: How Is Increasing Participation in Social Protection Schemes Affecting Girls’ Roles and Responsibilities?

Social protection
Children's work and time-use
Working paper

The focus of this paper is the effect on adolescent girls' roles and responsibilities of public works schemes or cash transfers, which are the main forms of social protection in developing countries. Increasing participation in social protection is intended to enhance the development of girls in participating households, but evidence on their school participation and workloads suggests that the reverse may be happening. The paper probes what happens to girls? roles and responsibilities when households participate in social protection schemes in rural Ethiopia and Andhra Pradesh. It argues that effects are complex, and often context-specific, however, the assumption that "beneficiaries" benefit means that negative impacts are rarely acknowledged. Nonetheless, the most important question to ask is not "do schemes increase girls' work?" but "how do they change the nature of girls' work and its relation to other valued dimensions of their lives?" The paper combines review and analysis of quantitative and qualitative data, recognising that this question cannot be answered with a methodology that considers girls? schooling or workloads in isolation.

Beyond Food Security: Transforming the PSNP in Ethiopia for the Well-being of Children

Yisak Tafere
Social protection
Working paper

This paper investigates the possible impacts of the Productive Safety Net Programme (PSNP) on children's well-being and recommends child-focused social protection that goes beyond the PSNP. It draws on data from the Young Lives survey of 569 rural households and qualitative case studies of 32 households and children living in four rural communities in 2006 and 2009. The quantitative analysis finds that, despite an increase in the incidence of economic shocks (such as drought and food-price inflation) and idiosyncratic family-related events such as the illness or death of family members, the value of cash and food transfers in real terms from the PSNP did not improve from 2006 to 2009, and even declined.

Therefore the contribution of the PSNP to risk reduction is limited because transfers did not increase in the face of shocks. Moreover, the substitution effect of the Public Work component of the PSNP dominates the income effect and this has caused children to spend more time on paid and unpaid work. The survey data also show that the Public Work component did not increase the time children spent on schooling and studying at home, while the qualitative data suggested that it had a negative impact on their learning. Insufficiency of PSNP transfers forced households to send their children to work for wages. The schooling of children engaged in Public Work and wage labour has been affected, and in some cases they have been forced to drop out of school altogether. The existing PSNP could be improved in such a way that it provides Direct Support for schoolchildren so that their schooling may not be hampered. But we argue that the PNSP on its own cannot ensure children's overall well-being.

Though it protects many children from hunger, the PSNP fails to ensure food security, contributes little to poverty reduction and does not guarantee that children attend school. Ensuring children's well-being and reducing their poverty require thinking beyond the PNSP. The paper concludes that, amid limited resources and contexts of vulnerability to protracted shocks, there is a need for child-focused social protection.

The Impact of Social Protection Programmes on Child Labour and Education in Ethiopia

The study investigates the impact of participation in Social Protection Programmes on child labour and education in Ethiopia, the largest social protection program in Sub-Saharan Africa outside of South Africa, Specifically PSNP and Pension. Social Protection programmes are recently widely recognized in developing countries as main means to tackle poverty, reduce vulnerability and to smooth consumption of households. It is very important to assess intra-households impact of the programmes. Most of the previous studies assess the explicitly objective of the programs. My study evaluated the effect of SP programmes on children outcomes which in turn important to break intergenerational translation of poverty.Using Propensity Score matching method to identify program impacts, we find evidence of both processes at work and schooling. Results are presented by for each programme by place of residence and gender. I find that participation in Public Works leads to significance increases school enrolment of child however PW also increases boys time on paid work outside home and girls on unpaid work outside home in rural areas. PW is better than EGS for enhancing boys school enrolment. The direct support is effective in reducing child labour and increasing education in rural and urban areas. DS reduces time devoted to any activities by 40 minute per a typical day for children for both urban and rural resides. When I Disaggregated based on gender direct support reduces time spent domestic chores in turn increases time devoted on schooling. Besides enhances grade completed and enrolment rate. For boys DS reduced time spent on total work by 57 minutes per day. Similarly Pension is effective in reducing child time spent on work and increasing schooling. For girls it reduces time spent on study at home and highest grade completed. For boys significantly reduce time spent on total work almost by 2 hours per typical day and increases highest grade completed by 1 grade and enrolment rate by 9%.

Social Protection and Children: A Synthesis of Evidence from Young Lives Longitudinal Research

Catherine Porter
Social protection
Policy paper

This paper sets out the key findings from Young Lives research into the ways that major social protection policies are impacting on children, their families and communities in Ethiopia, India and Peru. Most research and policy debate focuses on effects of social protection on households, with children assumed to be passive beneficiaries of programmes to reduce vulnerability. Here we concentrate on children.

Young Lives research finds that social protection schemes have both intended and unintended consequences for children, mostly positive but highlighting areas of concern that should be prioritised in programme design. For example, in Latin America, social protection has strongly moved in the direction of conditional cash transfers (CCTs) and the conditions attached to receipt of benefits are often directly aimed at child ‘welfare’, for example school attendance or visits to the health clinic. However, insights from qualitative analysis and child perspectives need to also enter into programme design.

Social protection, commonly understood as government-led schemes that either mitigate risk or reduce vulnerability and/or chronic poverty, has moved up the policy and research agenda in recent years with cash transfers now present in 45 countries, covering 110 million families in the global south (Hanlon et al. 2010). In this paper, we explore three important social protection schemes in Ethiopia, India and Peru. While operating in quite different contexts, all three represent formal national schemes being fairly rapidly expanded, and all contain elements of `conditionality’ that affect children’s lives. In Ethiopia and India, this is a work requirement, and in Peru, there are conditionalities on beneficiaries that involve visiting health centres and school attendance.

From Young Lives qualitative and quantitative analysis of social protection policies in these countries we find:
Evidence of the public works scheme in India acting as a cushion, and possibly providing insurance affects:

  • In Ethiopia, we find that schooling outcomes are improved by certain components of the Productive Safety Net Programme, though there are different impacts on boys and girls.
  • In Peru there is evidence that conditional cash transfers are mainly reaching the intended beneficiaries.

However Young Lives research also uncovers some of the unintended consequences of social
protection for children:

  • While children in Peru are attending school there are concerns that these increased demands on schools (through increases in class size) had not been adequately matched by investment. There is also evidence that programme placement has raised some tensions in the community – views that have been expressed by children and their communities in Young Lives sites.
  • In Ethiopia and India, there is some evidence that public works schemes increase the work demands on children, either directly or through children substituting for adults in the household who are involved in the programmes.

These findings must also be considered along with other evidence from qualitative work that documents how children are contributing to the household economy and managing risks themselves and with their families, and limited evidence on the extent to which increased work impacts on children’s schooling.

Overall, child-focused research on social protection can provide important insights to make social protection more inclusive of children’s needs. This can improve programme design to make better use of scarce resources, and invest in the future of children in very poor communities. From these findings we draw a small number of key conclusions relevant to those seeking to ensure policymakers give close consideration to children’s needs and are increasingly child sensitive.

Productive Safety Net Programme and Children’s Time Use Between Work and Schooling in Ethiopia

Tassew Woldehanna
Social protection
Working paper

Government, non-government and donor organisations have developed a social assistance programme known as the Productive Safety Net Programme (PSNP) which has two subprogrammes, namely the Public Work Programme (PWP) and Direct Support Programme (DSP). PSNP is designed to reduce the vulnerability of poor people to drought. It targets households in most cases without considering ex ante the issue of intra-household resource distribution. This paper assesses, using Young Lives survey data, the impacts of PSNP and Agricultural Extension Programme (AEP) on time use between work and schooling, as well as the highest grade completed by 12-year-old children in rural and urban Ethiopia. Empirically the study used propensity score matching techniques to estimate the impact of PSNP and AEP on child welfare measured by time use in various types of work, schooling and studying. We found that PWP in rural areas increases child work for pay; reduces children's time spent on child care, household chores and total hours spent on all kind of work combined; and increases girls spending on studying. The DSP in rural and urban areas reduces time children spent on paid and unpaid work, and increases the highest grade completed by boys in urban areas. On the other hand, AEP in rural areas was effective in reducing child work for pay and total work, increasing time girls spent on schooling and the highest grade completed by girls.