SPECIAL NOTE | November 2022 Floods in Pakistan: Human Development at Risk Juan D. Barón*, Mary Bend, Elena Roseo, Izza Farrakh Natural disasters not only cause death and KEY TAKEAWAYS destruction to physical capital but also have large Twenty-five percent of families with and long-term effects on children’s opportunities to children ages 3 to 17 report flooding accumulate human capital through factors such as in their area, with families with nutrition, education, health, and negative shocks to more educated parents showing more family income (Baez, de la Fuente and Santos 2010). resilience to the climate shock. Investing in human capital is the best way to enhance → Direct impacts on families resilience and adapt to climate change (Pangestu 2022). include damage to house, loss of crops and livestock as well as the loss of This note assesses the effects of the 2022 floods on human income sources. capital in Pakistan. It focuses on the results of a Pakistan-wide → Indirect impacts on families phone survey that gathered information on the experiences include limited access to roads, bridges, of approximately 4,000 families with children ages 3 to 17. The transport services, schools, health survey was originally designed to track children’s return to school centers, and increased food insecurity after school closures due to the COVID-19 pandemic, but the as well as increased health risks such team adapted the survey methodology to ensure that it could as dengue, malaria, and behavioral better document and understand the challenges families with problems in children. children are facing during the 2022 floods and that the country is experiencing as it seeks to protect and rebuild human capital → Direct and indirect impacts on families for the future. (For the methodology and caveats about the will likely decrease children’s interpretation of results, see final section of this note.) school attendance and learning as educational services will be limited and Starting on June 14, 2022, floods caused by unusually heavy as families cope with the challenges monsoon rains covered nearly one-third of Pakistan. While of reconstructing their homes and every province was impacted, estimates show that Sindh was addressing income loss. the worst affected. Satellite-based estimates for the flooded → Floods have exacerbated food areas suggest that approximately 4,000 km² in Sindh, 900 km² in insecurity, which families were Punjab, and 300 km² in Balochistan were under water for weeks already facing due to higher inflation, (UNOSAT 2022). leading to a likely negative impact on the nutritional and health outcomes for ACKNOWLEDGMENTS | This note is part of a series to monitor children’s educational children. experiences with the closure and reopening of schools due to the COVID-19 pandemic in support of the Ministry of Federal Education and Professional Training. The series The combination of all these factors is was written by the World Bank team consisting of Seo Yeon Hong, Manal Quota, likely to increase learning poverty and Mary Bend, Elena Roseo, and Juan D. Barón. For this note, the team benefitted from limit the opportunities of cognitive, comments and suggestions to the phone survey’s questionnaire by Amer Hasan, socioemotional, and healthy development Koen Geven, Lire Ersado, Aliya Kashif, as well as other colleagues from the Education, Poverty, Health, Transport, and Social Protection and Jobs Global Practices. of an estimated 1 million children in Pakistan who may not return to school. This note was made possible by generous funding from the United Kingdom’s Foreign, Commonwealth & Development Office (FCDO) through the Data and Results in Education (DARE) project and the Gates Human Capital Project Multi-Donor Trust Fund. Corresponding author. E-mail: jbaron@worldbank.org. * All authors belong to the Education Global Practice, the World Bank. More than 7,000 schools are being used to accommodate people displaced by flooding, meaning they are not currently providing access to education. Estimates of the humanitarian impacts shelters. As a result, an estimated 3.5 million of the floods are dire. It is estimated that children have had their schooling disrupted over 33 million people were affected by (NDMA 2022; UNICEF 2022). the floods, while more than 8 million were The economic impact of the disaster is displaced and nearly 600,000 were living substantial as estimated by the Post-Disaster in relief sites (NDMA 2022; WFP 2022). As of Needs Assessment (PDNA; Government of October 15, 2022, the floods had killed at least Pakistan 2022). PDNAs quantify the impact 1,717 people. An estimated 13 million people of disasters through three measures. First, remain directly exposed or living close to the direct cost of destroyed or damaged flooded areas, and livelihoods continue to be physical assets (Damage). Second, the severely and adversely affected. According to monetary value of changes in economic the revised UN Flash Appeal, approximately flows resulting from the disaster (Loss). Third, 21 million people require humanitarian recovery and reconstruction needs calculated assistance. Standing waters, food insecurity, as replacement costs for the short- and malnutrition as well as water- and vector- intermediate-term, and adding a premium borne diseases remain serious risks while for building back (Needs). The PDNA estimates immediate relief efforts are ongoing (NDMA a total damage of nearly US$ 15 billion, losses 2022; UNOSAT 2022; WFP 2022; OCHA 2022). of US$ 15.2 billion, and total needs of US$ 16.3 The floods have also caused extensive billion. Almost every sector of the economy damage to infrastructure. As of October, was impacted, but housing, transport, and reported infrastructure damages include communications were affected the most. approximately 2.1 million homes; 13,115 km In the education sector, total damage is of roads; and 436 bridges (NDMA 2022). estimated to be US$ 559 million, losses are Almost 26,600 schools have been damaged US$ 219 million, and needs for recovery or destroyed, most of which are in Sindh. amount to US$ 918 million. Moreover, the Currently, 7,062 schools across the country national poverty rate may increase, pushing are being used as shelters, rendering them between 8.4 and 9.1 million people into unusable for teachers and their students even poverty. However, these estimates do not after the emergency passes. In Balochistan, include the effects on human capital that will most schools are being used as emergency linger long after the floods recede. The impact of the floods will likely make it COMPOUNDING CHALLENGES even more difficult for children to attend school and receive quality education. Before Before the natural disaster began the floods, there were between 18.7 and 20.7 million children out of school (PAMS 2021, in June 2022, between 18.7 and Government of Pakistan 2019), and 75 percent of 10-year-olds in Pakistan were not able to 20.7 million children were out of read an age-appropriate text between 18.7 school. With the recent floods, and 20.7 million (World Bank 2019). With the recent floods, according to new simulations, estimates show that at least an learning poverty could increase to 79 percent (Saavedra and Sherburne-Benz 2022). additional 1 million children Estimates presented below show that at least could stop attending schools. an additional 1 million children could stop attending school. SPECIAL NOTE | Floods in Pakistan: Human Development at Risk | November 2022 2 I.  Overall flood impacts on families with children The floods have impacted a sizeable portion of households with children in Pakistan, threatening Pakistan’s future human development. Twenty-six percent of households with children ages 3 to 17 report living in areas that were somewhat, partially, or completely flooded (figure 1). Figure 1. Almost one-third of families with children Importantly, 14 percent of respondents report that ages 3 to 17 report some flooding in their area the area they lived in was completely flooded. This high percentage shows the strong immediate   Any flooding  Completely impact of floods on households and how these flooded floods have put their children at risk of dropping out   No flooding  Partially of school, food insecurity, child labor, and health flooded conditions. The phone survey also shows how floods  Somewhat flooded have a large detrimental effect on the future human development prospects of a substantial share of children in Pakistan. Figure 2. Floods have impacted Pakistan’s provinces and households unequally A. Households reporting flooding by province Floods have impacted Pakistan’s provinces and households unequally. Almost 1 out of 2 families with children in Sindh, 3 out of 4 in Balochistan, and almost 4 out of 10 in Khyber Pakhtunkhwa (KP) reported flooding in their area (figure 2, panel A). It is expected that the floods will further enlarge the provincial gaps in educational achievement in areas such as Sindh and Balochistan, which are known B. Households reporting flooding by education level to already have the most precarious education outcomes in the country. Disasters do not affect people equally as documented in the literature (Hallegatte et al. 2020; Kousky 2016); they impact the poorest hardest. Pakistan’s floods do not seem to be the exception. Households with children where the main caregiver has less education (a proxy for low income), that is, those with a middle school education or less, were on average 5 to 8 percentage points more likely (statistically) to report flooding in their area (figure 2, panel B). However, while floods have greater impact on lower-income households, still 20 to 25 percent of the most educated families surveyed have also reported flooding in their area. SPECIAL NOTE | Floods in Pakistan: Human Development at Risk | November 2022 3 Figure 3. Families with children report direct impacts to their household and income-related activities, and indirect impacts through limited access to public goods DIRECT INDIRECT In flooded areas, families with children reported suffering indirect and direct losses. As shown in figure 3, 39 percent reported losing agricultural crops, 32 percent reported losing livestock, and 15 to 29 percent reported damage to roads, bridges, and electricity supply. There were also direct hits to 20% households; 44 percent of families reported damage to their house and 26 percent reported losing their house. Also, 1 in 5 households reported losing the means to earn a living, casting doubt on the capacity of these families to quickly recover financially from the shock inflicted by the floods or to be able to ensure that their children will develop their cognitive, Percent of respondents social, and behavioral skills to succeed in life. in flooded areas The relief effort to reach families with children in reported losing the the affected areas remains insufficient. According means to earn a living to the phone survey, 80 percent of respondents reported that no flood relief efforts had taken place in their area, while 45 percent of those families who lost their houses reported living in temporary shelters. Among the 20 percent who noted relief efforts, government, foundations, the army, and local initiatives were listed as the top providers of support to flood victims. These results show that there is some recognition of efforts by the authorities to support flood victims, but more relief efforts are needed. SPECIAL NOTE | Floods in Pakistan: Human Development at Risk | November 2022 4 More targeted communication efforts, using multiple media to communicate in local languages, could help inform a greater number of the most vulnerable and most affected people about relief efforts. Information about availability of relief efforts does Figure 4. Educated people are more likely to be aware not reach all people equally. When asked about of relief activities to help flood victims their knowledge of relief activities happening in the area (figure 4), those with a higher education level (32 percent with a university degree, 21 percent with a secondary degree) were significantly more likely to be aware of relief activities than those with only a primary school education or less (only 14 percent). The reasons for this could be many, including: (a) those with more education are more likely to be literate in different means of communication and therefore are more connected to information, or (b) relief may be more widely available in areas where more educated individuals live since those areas might be easier to access. This data, combined with the earlier observation that less educated respondents were more likely to be inundated by the floods, highlights the need for more targeted efforts to identify the most vulnerable and most affected households and to use multiple media to communicate in local languages (e.g., television, radio, visual billboards, etc.) to ensure relief efforts are reaching all affected people. SPECIAL NOTE | Floods in Pakistan: Human Development at Risk | November 2022 5 II. Impact of floods on children’s education There is strong global evidence that disasters negatively impact education and have potentially cumulative and long-lasting effects (Baez, de la Fuente and Santos 2010). The immediate direct effect on education can be found in the destruction of schools or reduced access to them due to the destruction of other critical infrastructure (e.g., roads, bridges, etc.). When instruction is disrupted, quality suffers and children’s human capital accumulation opportunities are reduced, in both the immediate and longer term (Andrabi, Daniels and Das 2021; Husted, Opper and Park 2022). Figure 5. More than half of those interviewed The impact of floods on education access of children reported that the floods negatively impacted has been immediate. More than half of those families their children’s education with flooding in their area (51 percent) reported that the floods had negatively impacted their children’s   Some impact  Session education. Nearly one-third (28 percent) of respondents disrupted reported that their children’s schools were disrupted   No impact  Physical and/or damaged (figure 5), and almost one third damage (35 percent) reported that their children’s school was completely damaged. School damage due to flooding most often occurred in Sindh and Balochistan. Approximately 17 percent of schools were also used as temporary shelters for displaced families, which has further depleted school infrastructure. This situation was most prevalent in Sindh. Seventy-two percent of parents reported that their children were not studying during flood-related school closures (figure 6), while 28 percent of parents said their children were continuing to learn. Among this group, 28 percent reported that teachers visited their children and 20 percent reported that their children received printed educational materials. Nine percent indicated that their children continued their studies in temporary learning centers and just 4 percent participated in distance learning. Figure 6. Engagement of children in learning activities during school closures due to floods Has your child How have Studied at Studied home at home 44% continued engaging they continued in learning activities Yes engaging? TeacherTeacher visits visits 28% during school closures due to Received Received educational educational materials 20% floods? materials No Temporary Temporary learning learning centers centers 9% Distance Distance learning learning 1% SPECIAL NOTE | Floods in Pakistan: Human Development at Risk | November 2022 6 The impact of the floods could force families to reduce investments in the education of their children as families struggle with low prospects of earning a living in the short term, limited access to schools and other public services (e.g., health Figure 7. Almost 20 percent of children are and transportation), and an already high cost of unlikely to attend school in the future education. Reductions in attendance, retention, and progression have been documented after How likely is it multiple disasters in other countries in South Asia that your child will (Andrabi, Daniels and Das 2021; Jacoby and Skoufias attend school in 1997), Africa (Jensen 2000), and Latin America the future? (Ureta 2006; Santos 2007), among others. These examples demonstrate the real risk Pakistan faces   Not at all likely in the aftermath of the floods, with 66 percent of   Not very likely households with children reporting income losses.   Somewhat likely A significant number of children may not return   Very likely to school in the aftermath of the floods. Over half   Extremely likely (52 percent) of parents said their children would be extremely likely to attend school after the floods, and together with questions about whether their children were “somewhat likely” and “very likely” The impact of floods on education access cannot to attend school, the survey show that 83 percent be underestimated. If we consider that (i) before of households were at least somewhat likely to the floods there were 15 million children in schools send their children back to school (figure 7). When in the most affected provinces (Balochistan, KP, and families were asked about school attendance of Sindh); (ii) the floods have subsequently affected 25 children if schools were to open tomorrow, a similar percent of those in school in these areas; and (iii) 27 percentage (around 90 percent) expressed interest. percent of households report that they may not send These responses imply that at a minimum between their children back to school (including half of those 10 and 20 percent of parents are not likely to send who report that their children are somewhat likely their children back to school. to return to school), it is sensible to estimate that at least 1 million children might potentially become out of school due to the floods (approximately). REDUCING CONSTR AINTS This number is obtained by multiplying these three figures. To ensure parents send Among those who said it was unlikely their their children back to school children would attend school, 63 percent cited the prohibitive cost of education, both direct and it is important to reduce indirect, as the main factor. In addition, close to 1 in 5 (19 percent) of parents said the poor quality the biggest constraint to of schools and teaching was the primary reason education: its cost. Floods their child would be unlikely to attend school in the future. While rebuilding flood safe schools is have made education even a current top priority, improving the quality of the more expensive for families. education children receive while in class is also an urgent need to convince parents to send their children back to school and to limit the lasting effects of this natural disaster. SPECIAL NOTE | Floods in Pakistan: Human Development at Risk | November 2022 7 The phone survey also explored the gender of Figure 8. Some people say that it would be children in returning to school by asking parents difficult for female children to go back to school to agree or disagree with the following statement: in their area as roads, etc. are destroyed. Even “Some people say that it would be difficult for female though boys may be able to go, girls would not be children to go back to school in their area as roads, able to. Do you agree with this assessment? etc. are destroyed. Even though boys may be able to go, girls would not be able to. Do you agree with this assessment?” Approximately 33 percent of parents agreed with the statement that it would be more difficult for girls than boys to go back to school in the areas where roads had been destroyed Yes (figure 8). Parents were more concerned for girls than boys in regard to the deterioration of transport, roads, and other services. This aligns with results that point out that parents in Pakistan considered No transportation services to be, even before the floods, a key determinant of school attendance for girls and young women (Barón et al. forthcoming). Households might cope with the income shock of the floods by having their children work to help support the family. In the phone survey, 28 percent they envision their children working in the future (14 of families reported that they envisioned their percent), while caregivers with less education are more child working instead of returning to school due to likely to have their children work (between 28 and 32 the economic impact of floods. This is a concern; percent, see figure 9). Geographical location also affects evidence shows that children who stop attending whether children are likely to work instead of return to school as a response to shocks such as natural school; the rate of children who are expected to work is disasters are less likely to return to school compared highest in Balochistan (41 percent compared to around to other children (Beegle et al 2003; Santos 2007; Baez 27 percent in KP and Sindh). These results indicate that et al. 2007; Dong et al. 2020; De Janvry et al. 2006). more could be done to encourage families to support Also, social norms for boys and girls play a role in a return to school for their children, even while schools whether children return to school or work during and are under reconstruction, and especially with respect to after natural disasters (Takasaki 2017). future human development. The education level of caregivers within families is A significant proportion of households may require another factor. Families in which the caregiver has their children to support the reconstruction of their tertiary education are half as likely to report that home, potentially reducing children’s attendance to school. The phone survey seeks to understand whether Figure 9. Households expect their children to work house reconstruction is something children would be due to economic difficulty following the floods expected to help with, further preventing them from attending school. To those families who lost their home, the survey asks: “I will read out two statements regarding your child. Tell me which one you agree with: (a) Your child stays home from school to rebuild home; (b) Your child goes to school while other family/friends rebuild home.” Approximately 85 percent of flood- affected households who indicated that there had been damage to their house reported that they would keep sending their children to school while other household members rebuilt their home. However, a significant percentage of households, 15 percent, asserted that children are likely to stay at home helping with household chores and reconstruction. SPECIAL NOTE | Floods in Pakistan: Human Development at Risk | November 2022 8 III. Indirect impacts on human development There are less direct factors that might also reduce school attendance and school quality for children. The phone survey explored three of these key indirect factors: (a) health risks, (b) nutrition and food insecurity, and (c) transportation services. Figure 10. Health facilities Health risks were also damaged by floods The floods can affect human development through   Not damaged the destruction of health infrastructure. The and functional importance of health infrastructure goes beyond the  Damaged initial response of the emergency and extends to the recovery and support of healthy environments  Somewhat in situations where infection diseases and sanitation damaged issues become more prevalent (Baez, de la Fuente  Partially and Santos 2010). Standing water caused by floods damaged are sources for malaria and can increase the risk  Completely damaged of gastrointestinal diseases in children, which can become life threatening if left untreated. Like schools, health facilities have been damaged by the floods in Pakistan. As figure 10 shows, respondents said health facilities were either completely damaged (21 percent) or partially damaged (25 percent) in their area; 17 percent of Almost respondents said these facilities were “somewhat damaged.” In short, 1 in 3 health care facilities in areas covered by the phone survey seemed to 1 in 3 have survived the floods unscathed. The lack of health facilities can impact school attendance as underwater areas generate disease, further eroding educational achievement. Households lack regular Importantly, in response to the phone survey, 28 percent reported malaria and dengue outbreaks access to clean drinking in the flooded areas. Unlike the unequal impact of water, leaving them the floods on different households with different vulnerable to disease educational backgrounds, flood-related diseases show no preference—all levels of educational attainment bear the burden equally. In addition, 32 percent of households reported a lack of access to drinking water on a regular basis. Lack of drinking and clean water is one of most common sources of infectious diseases. SPECIAL NOTE | Floods in Pakistan: Human Development at Risk | November 2022 9 Mental health issues are also prevalent among Figure 11. Almost half of children exhibit changes children who have experienced natural disasters in their behavior (Lai and La Greca 2020). Children exposed to natural disasters exhibit a variety of mental health issues (Lai, Auslander, Fitzpatrick and Podkowirow Feeling sad 2014; Tang, Liu, Liu, Xue and Zhang 2014), including symptoms of depression (e.g., feeling sad, not Become more wanting to play, or becoming more aggressive). As aggressive shown in figure 11, the phone survey shows that half of all children exhibited behavioral changes that could be attributed to the impact of natural Become quiet disasters. While over one-third (37 percent) of parents noticed their children feeling sad, 19 percent Does not want said their children had become more aggressive, and to play 6 to 8 percent reported their children did not want to play or they became quieter than usual. Any policy response could consider complementing physical resources with socioemotional support for children who have endured the traumatic and stressful events of the floods. A comprehensive package of support to families exposed to natural disasters includes socioemotional support to help children cope with trauma experienced during the floods. SPECIAL NOTE | Floods in Pakistan: Human Development at Risk | November 2022 10 Figure 12. Food insecurity A. Was there a B. Was there a for households time when you time when the with children were worried No Yes household ran No Yes in the past 12 the household out of food? months would run out of food? Approximately 75 percent of families reported Increase in food insecurity damage to livestock or crops, impacting family income, children’s nutrition, and the energy intake of The floods have increased food insecurity for families and their children. families already impacted by inflation. The dual effects of floods on economic activity and food The floods have likely exacerbated child production have likely exacerbated food insecurity malnutrition in a country where stunting and all for all households, impacting the development forms of malnutrition were already high (UNICEF of children. Research shows a strong negative link 2018). With inflation projected to be 23 percent between natural disaster and long-lasting nutritional for the fiscal year (World Bank 2022), the impact outcomes for children (see Baez, de la Fuente and of floods on family incomes and the physical, Santos 2010). socioemotional, and cognitive development of children and youth could severely erode Over half (57 percent) of families with children opportunities of a generation of children in flooded reported being worried that they would run out areas. Targeting resources to the most affected of food due to lack of money or other resources families could mitigate the further deterioration (figure 12, panel A). The phone survey shows of food security and support children’s ongoing that less educated parents were almost twice as development as part of a comprehensive package of worried about the household running out of food support in the flooded areas. than the most educated parents (67 percent versus 37 percent). In fact, based on respondents to the phone survey, over one-third (46 percent) did run Disruptions to transportation out of food (figure 12, panel B). Food insecurity is a real risk for the healthy development of children in Due to the floods, transportation services have flooded areas, especially for children from less- been disrupted. Floods have also constrained educated families. mobility and prevented families from traveling to Over the past year, households’ concerns about regular destinations, including work and school; running out of food are likely linked to increases nearly 59 percent of families in flood-affected in food prices. Approximately 67 percent of areas say they cannot reach those destinations. households report food prices have recently risen, Travel has been more widely restricted due to with 50 percent of respondents reporting that floods in rural areas than in urban—61 percent of prices increased between 10 and 50 percent over rural respondents reported restricted movement the last year. Moreover, approximately 30 percent of compared to 53 percent in urban areas. Indeed, 83 respondents reported having gone without food for percent say flooded roads are the primary reason for at least a whole day due to financial insecurity. Food their restricted movement. As floodwaters recede insecurity is worse in rural than in urban areas. and after, policy makers can position transportation services to be ready when they do; 20 percent The destruction of crops by the floods has say it took more than a month for transportation increased food insecurity in households with services to resume in their area and 14 percent say children. In the phone survey, 46 percent of these services were still disrupted at the time of the households responded that they cultivated their survey. When families cannot get to work or school, own land (either for food or other products). their livelihoods, food security and overall human develoment are at risk. SPECIAL NOTE | Floods in Pakistan: Human Development at Risk | November 2022 11 IV. Methodology and survey design The data used in this note come from a nationally representative phone survey carried out by Gallup Pakistan. Data collection was carried out from September 29 to October 28, 2022. 4,044 The survey was carried out using random digit dialing (RDD) of mobile phones using all four telecom providers with active numbers across Pakistan. Approximately 93 percent of households have access to a mobile phone (Government of Pakistan 2021). Despite the high- penetration rate of mobile phones, lower-income households are overrepresented in the remaining Completed surveys in the study 7 percent of those who do not have access to mobile technology. This most likely means that The target population of the survey were parents in many cases the results presented show an or caregivers of children ages 3 to 17. If more than underestimation of the true impacts of the floods one child lived in the household, one child was on families and their children in Pakistan. randomly selected as the subject of inquiry. The sample was stratified by gender of the child. The To survey households, each random number was survey also oversampled households that reported called until a call was answered (with a maximum any impact of floods, aiming for a sample of at least of three attempts). To maximize the response rate, 1,000 households that suffered effects of floods calls were placed at different times on different in their area. The survey collected information on days of a week. Once an individual was contacted the education status of children, food security, on his or her mobile phone, consent was obtained, child work, health, environment, and household a screening questionnaire was administered, and a composition. A limited set of sociodemographic unique study identification number was generated characteristics was also collected, including for the respondent. Interviewers entered data education of parents, assets, gender, family into a tablet with Survey CTO software that had composition, rural, urban, district, and province. the preloaded questionnaire with automatic skipping patterns (Computer Assisted Telephone The survey randomly called 40,800 numbers, Interviewing, CATI). reaching 15,750 individuals who answered the phone, of which 5,420 agreed to the interview, 1,346 with incomplete surveys, and 4,044 with complete In the surveys. At the beginning of the survey, enumerators sample: 61% rural 39% urban offered a PKR 200 phone credit for completing the survey to 75 percent of those who answered the phone (selected randomly). To be able to better capture the aggregate impact, weights were created for the selection of province, rural/urban, sex, and education of household head. CAVEAT ON INTERPRETATION OF RESULTS | Phone surveys are known to miss families at the bottom of the income distribution who usually do not have mobile phones or live in hard-to-reach 16% 84% areas (where cellphone coverage is limited). In the current context, the damaged caused by floods may have exacerbated those issues as the floods could have increased the difficulty to reach poorer female male households or those that lost everything. Thus, the results and interpretations in this note should be taken with those factors in mind. In most instances, the results presented in this note are likely to underestimate the magnitude of the impact given the characteristics of those who might not be reachable via phone calls. SPECIAL NOTE | Floods in Pakistan: Human Development at Risk | November 2022 12 REFERENCES Andrabi, T., B. Daniels, and J. Das. 2021. “Human Capital Accumulation and Disasters: Evidence from the Pakistan NDMA (National Disaster Management Authority). 2022. “NDMA Floods (2022) SITREP–2022.” Islamabad, Pakistan: NDMA. Earthquake of 2005.” The Journal of Human Resources http://cms.ndma.gov.pk/storage/app/public/situation- (June 2021): 1–67. http://jhr.uwpress.org/content/ reports/October2022/9WVaA9l0lwTcmtxzP1fa.pdf. early/2021/06/02/jhr.59.2.0520-10887R1. OCHA (United Nations Office for the Coordination of Baez, J., A. de la Fuente, and I. V. Santos. 2010. “Do Natural Humanitarian Affairs). 2022. “Revised Pakistan 2022 Floods Disasters Affect Human Capital? An Assessment Based on Response: 01 Sep 2022–31 May 2023,” October 4, 2022. Existing Empirical Evidence.” IZA Discussion Paper 5164, https://reliefweb.int/report/pakistan/revised-pakistan- Institute for the Study of Labor, Bonn, Germany. 2022-floods-response-plan-01-sep-2022-31-may-2023- issued-04-oct-2022. Baez, J., and I. V. Santos. 2007. “Children’s Vulnerability to Weather Shocks: A Natural Disaster as a Natural PAMS (Pak Alliance for Maths and Science). 2021. “The Missing Experiment.” Working paper, Active Learning Network for Third: An Out of School Study of Pakistani 5-16 Year Olds.” Accountability and Performance in Humanitarian Action, Pak Alliance for Maths and Science, Islamabad, Pakistan. London, UK. https://mathsandscience.pk/publications/the-missing- third/ Barón, J., I. Trako, N. Ejaz, and M. Bend. Forthcoming. “Constraints to Girls’ Education in Pakistan.” Washington, Pangestu, M. 2022 “The Human Face of Climate Change,” DC: World Bank. World Bank Blogs, November 7, 2022. https://blogs. worldbank.org/voices/human-face-climate- Beegle, K., R. H. Dehejia, and R. Gatti. 2003. “Child Labor, change?deliveryName=DM161213 Income Shocks, and Access to Credit. Policy Research.” Working Paper, 3075, World Bank, Washington, DC. https:// Saavedra, J., and L. Sherburne-Benz. 2022. “Pakistan’s Floods openknowledge.worldbank.org/handle/10986/18222. are Deepening its Learning Crisis,” World Bank Blogs, September 28, 2022. https://blogs.worldbank.org/ De Janvry, A., F. Finan, E. Sadoulet, and R. Vakis. 2006. “Can endpovertyinsouthasia/pakistans-floods-are-deepening- Conditional Cash Transfer Programs Serve as Safety Nets its-learning-crisis. in Keeping Children at School and From Working When Exposed to Shocks?” Journal of Development Economics Santos, I. 2007. “Disentangling the Effects of Natural Disasters 79 (2): 349–373. on Children: 2001 Earthquakes in El Salvador.” PhD dissertation, Kennedy School of Government, Harvard Dong, Y., and H. Yang. 2020. “Natural Disaster and Child Labor: University, Cambridge, MA. Evidence from Indonesian Earthquake,” February 21, 2020. https://ssrn.com/abstract=3542086 or http://dx.doi. Takasaki, Y. 2017. “Do Natural Disasters Decrease the Gender Gap org/10.2139/ssrn.3542086. in Schooling?” World Development 94 (June 2017): 75–89. Government of Pakistan. 2019. “Policy Framework for Pakistan’s Tang, B., X. Liu, Y. Liu, C. Xue, and L. Zhang. 2014. “A Meta- Out of School Children.” Ministry of Federal Education and Analysis of Risk Factors for Depression in Adults and Professional Training: Government of Pakistan. Children after Natural Disasters.” BMC Public Health 14 (1): 623. https://doi.org/10.1186/1471-2458-14-623. Government of Pakistan. 2021. “PSLM — 2019–20: Pakistan Social and Living Standards Measurement Survey.” Pakistan UNICEF (United Nations Children’s Fund) Pakistan. 2018. Bureau of Statistics, Ministry of Planning Development and “National Nutrition Survey: Key Findings Report.” Special Initiatives: Government of Pakistan. https://www. UNICEF, Islamabad, Pakistan. https://www.unicef. pbs.gov.pk/sites/default/files//pslm/publications/pslm_ org/pakistan/media/1871/file/KeyFindings--- district_2019-20/Key_Finding_Report_of_PSLM_District_ NationalNutritionSurvey2018.pdf. Level_Survey_2019-20.pdf. UNICEF. 2022. “Pakistan: Humanitarian Situation Report, 5,” Government of Pakistan. 2022. “Pakistan Floods 2022: Post- October, 2022. https://www.unicef.org/media/129696/ Disaster Damage Assessment (PDNA).” Ministry of Planning file/Pakistan-Floods-Humanitarian-SitRep-October-2022. Development and Special Initiatives: Government of pdf. Pakistan. https://www.pc.gov.pk/uploads/downloads/ PDNA-2022.pdf. UNOSAT (United Nations Satellite Centre). 2022. “Preliminary Satellite Derived Flood Evaluations Assessment – Islamic Hallegatte, S., A. Vogt-Schilb, J. Rozenberg, M. Bangalore, and Republic of Pakistan” (map), October 12, 2022. https:// C. Beaudet. 2020. “From Poverty to Disaster and Back: a reliefweb.int/map/pakistan/preliminary-satellite- Review of the Literature.” EconDisCliCha 4: 223–247.  derived-flood-evolution-assessment-islamic-republic- pakistan-11-oct-2022. Husted, L., I. M. Opper, and R. J. Park. 2022. “The Impact of Natural Disasters on Human Capital.” EdWorkingPaper, Ureta, M. 2006. “Hurricane Mitch, Family Budgets and Schooling 22-548, Annenberg Institute for School Reform at Brown in Nicaragua.” Paper prepared for the Inter-American University, London, UK. https://doi.org/10.26300/bkah-we19. Development Bank conference, Washington, DC., February 10, 2005. https://www.cerge-ei.cz/pdf/events/ Jacoby, H., and E. Skoufias. 1997. “Risk, Financial Markets, papers/051114_t.pdf. and Human Capital in a Developing Country.” Review of Economic Studies 64 (3): 311–335. World Bank. 2019. “Pakistan: Learning Poverty Brief.” Washington, DC: World Bank. https://thedocs.worldbank. Jensen, R. 2000. “Agricultural Volatility and Investments in org/en/doc/214101571223451727-0090022019/original/ Children.” American Economic Review 90 (2): 399–404. SASSACPKPAKLPBRIEF.pdf. Kousky, C. 2016. “Impacts of Natural Disasters on Children.” World Bank. 2022. “Inflation and the Poor.” Washington, DC: The Future of Children 26 (1): 73–92. http://www.jstor.org/ World Bank. https://reliefweb.int/report/pakistan/ stable/43755231. pakistan-development-update-october-2022-inflation- Lai, B. S., B. A. Auslander, S. L. Fitzpatrick, and V. Podkowirow. and-poor. 2014. “Disasters and Depressive Symptoms in Children: WFP (World Food Programme). 2022. “WFP Pakistan Situation A Review.” Child and Youth Care Forum 43 (4): 489–504. Report,” September 19, 2022. https://reliefweb.int/ https://doi.org/10.1007/s10566-014-9249-y. report/pakistan/wfp-pakistan-situation-report-19- Lai. B. S. and A. La Greca. 2020. “Understanding the impacts of september-2022. natural disaster on children.” Society for Research in Child Development (SRCD) Child Evidence Brief 8, August. IN THIS NOTE | Source for all data graphics: Authors’ estimates based on phone survey. Photography: Insiya Syed. Graphic design: Elizabeth Salud. © 2022 International Bank for Reconstruction and Development / The World Bank 13