Infrastructure services and early childhood development in Latin America and the Caribbean: Water, sanitation, and garbage collection

Balza, Lenin H.; Cuartas, Jorge; Gomez-Parra, Nicolas; Serebrisky, Tomás Working Paper Infrastructure services and early childhood development in Latin America and the Caribbean: Water, sanitation, and garbage collection IDB Working Paper Series, No. IDB-WP-1612 Provided in Cooperation with: Inter-American Development Bank (IDB), Washington, DC Suggested Citation: Balza, Lenin H.; Cuartas, Jorge; Gomez-Parra, Nicolas; Serebrisky, Tomás (2024) : Infrastructure services and early childhood development in Latin America and the Caribbean: Water, sanitation, and garbage collection, IDB Working Paper Series, No. IDB-WP-1612, Inter- American Development Bank (IDB), Washington, DC, https://doi.org/10.18235/0012998 This Version is available at: https://hdl.handle.net/10419/299453 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/3.0/igo/ Infrastructure Services and Early Childhood Development in Latin America and the Caribbean: Water, Sanitation, and Garbage Collection Lenin H. Balza Jorge Cuartas Nicolas Gomez-Parra Tomás Serebrisky WORKING PAPER No IDB-WP-1612 Inter-American Development Bank Infrastructure and Energy Sector June 2024 Infrastructure Services and Early Childhood Development in Latin America and the Caribbean: Water, Sanitation, and Garbage Collection Lenin H. Balza Jorge Cuartas Nicolas Gomez-Parra Tomás Serebrisky Inter-American Development Bank Infrastructure and Energy Sector June 2024 Cataloging-in-Publication data provided by the Inter-American Development Bank Felipe Herrera Library Infrastructure services and early childhood development in Latin America and the Caribbean: water, sanitation, and garbage collection / Lenin H. Balza, Jorge Cuartas, Nicolas Gomez-Parra, Tomás Serebrisky. p. cm. — (IDB Working Paper Series; 1612) Includes bibliographical references. 1. Infrastructure (Economics)-Latin America. 2. Infrastructure (Economics)-Caribbean Area. 3. Water supply-Latin America. 4. Water supply-Caribbean Area. 5. Sanitation-Latin America. 6. Sanitation- Caribbean Area. 7. Refuse and refuse disposal-Latin America. 8. Refuse and refuse disposal-Caribbean Area. 9. Toddlers-Health and hygiene-Latin America. 10. Toddlers-Health and hygiene-Caribbean Area. I. Balza, Lenin. II. Cuartas, Jorge. III. Gomez-Parra, Nicolas. IV. Serebrisky, Tomás. V. Inter-American Development Bank. Infrastructure and Energy Sector. VI. Series. IDB-WP-1612 http://www.iadb.org Copyright © 2024 Inter-American Development Bank ("IDB"). This work is subject to a Creative Commons license CC BY 3.0 IGO (https://creativecommons.org/licenses/by/3.0/igo/legalcode). The terms and conditions indicated in the URL link must be met and the respective recognition must be granted to the IDB. Further to section 8 of the above license, any mediation relating to disputes arising under such license shall be conducted in accordance with the WIPO Mediation Rules. Any dispute related to the use of the works of the IDB that cannot be settled amicably shall be submitted to arbitration pursuant to the United Nations Commission on International Trade Law (UNCITRAL) rules. The use of the IDB's name for any purpose other than for attribution, and the use of IDB's logo shall be subject to a separate written license agreement between the IDB and the user and is not authorized as part of this license. Note that the URL link includes terms and conditions that are an integral part of this license. The opinions expressed in this work are those of the authors and do not necessarily reflect the views of the Inter-American Development Bank, its Board of Directors, or the countries they represent. Infrastructure Services and Early Childhood Development in Latin America and the Caribbean: Water, Sanitation, and Garbage Collection∗ Lenin H. Balza Jorge Cuartas Nicolas Gomez-Parra Tom´as Serebrisky June 2024 Abstract Access to essential infrastructure services such as water, sanitation, and garbage collection can considerably affect children’s environment and may play a significant role in shaping early childhood developmental and health outcomes. Using data from the Multiple Indicator Cluster Surveys (MICS) and the Demographic and Health Surveys (DHS) for 18 countries in Latin America and the Caribbean (LAC), we show a significant positive association between access to water and sanitation and early childhood development, as well as reduced instances of stunting. In addition, we identify a negative association between access to improved garbage collection services and the rates of stunting and underweight among children under five. Our findings are robust after using alternative measures for access and controlling for individual, maternal, and household factors, alongside considerations of household wealth and caregiver’s stimulation activities. Similarly, the economic relevance of the relationship is highlighted by the substantial gap relative to the size of the vulnerable groups, persisting even after adjusting for confounding variables. Our results also suggest that households may be able to lessen the potential impact of pollutants through mitigation measures such as treating water to make it safe for consumption, using handwashing cleansers, and storing household trash in lidded containers. The current findings underscore the importance of investing in basic infrastructure services as a critical component of comprehensive strategies to enhance early childhood development and health in low- and middle-income countries. We emphasize the importance of considering the quality and type of infrastructure services alongside their availability. Future research should incorporate more complete and detailed data to improve understanding of the causal relationship between water, sanitation, and garbage collection and early childhood development, as well as the mechanisms underlying the observed associations. JEL codes: H41, I15, J13, Q53. Keywords: Early childhood development, Water and sanitation access, Waste management infrastructure, Low- and middle-income countries, Mitigation strategies. ∗Balza: Inter-American Development Bank (lenin[email protected]); Cuartas: Harvard Graduate School of Education/Universidad de los Andes ([email protected]ard.edu); Gomez-Parra: Inter-American Development Bank (ngomez- [email protected]); Serebrisky: Inter-American Development Bank ([email protected]). The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of the Inter-American Development Bank, their Boards of Executive Directors, or the governments they represent. We thank Lisa Bagnoli, Daniela Collazos, Giuliana Daga, Florencia Lopez Boo, David Mat´ıas, Carol Terracina Hartman, and an anonymous reviewer for their comments and suggestions. Any errors or omissions remain the sole responsibility of the authors. 1 Introduction More than 250 million children younger than five living in low- and middle-income countries are at risk of not reaching their developmental potential due to adversity and lack of nurturing care (i.e., good health, adequate nutrition, early learning opportunities, safety and security, and responsive caregiving) (Black et al.,2017;Lu et al.,2020;McCoy et al.,2022). This is concerning as early childhood development (ECD) strongly predicts educational trajectories, future health, and economic and social well-being throughout an individual’s lifespan (Black et al.,2017;Heckman, 2006). Indeed, brain development and the acquisition of foundational skills are highly sensitive to environmental input early in life (Sheridan and Nelson,2009). Early experiences – positive or negative – thus have enormous long-term consequences for individual and societal welfare (Black et al.,2017;Cunha and Heckman,2007;Knudsen et al.,2006). Extensive evidence has shown that ECD is particularly sensitive to various nutritional, health, and psychosocial factors, including the home environment, access to learning opportunities, and exposure to violence (Attanasio et al.,2020;Cuartas,2021,2022;Cuartas et al.,2020,2021;Jeong et al.,2017;McCoy et al.,2016;Salhi et al.,2021;Wolf and McCoy,2019). While there is extensive literature within the field of public health assessing the impact of the physical environment (in particular, the relation between water, sanitation, and hygiene with nutrition) on ECD (Danaei et al.,2016;Sania et al.,2019), the bulk of this research has focused on stunting as an outcome and not child development specifically. Stunting has often been used as a proxy for child development, though recent evidence suggests that it is an inadequate proxy (Sania et al.,2019;Perumal et al.,2018). In this context, relatively little effort has been made to integrate developmental science with environmental health (Trentacosta et al.,2016). This work is nonetheless important as climate change, industrialization and poor basic infrastructure access have increased several specific environmental risks that may be particularly influential early in life. These are especially pronounced in low- and middle-income countries (LMICs), where millions of children lack access to clean water sources, improved sanitation, and adequate waste management (McCoy et al.,2022). Indeed, poor water, sanitation, and garbage disposal represent particular risks for children as they consume more water, food, and air relative to their body weight than adults, while their metabolism is less capable of coping with contaminants and their brain architecture is more sensitive to pollutants (Trentacosta et al.,2016). Inadequate or polluted water, poor sanitation, and garbage disposal can affect ECD through at least three mechanisms. First, poor hygiene, polluted water, and inadequate sanitation and waste management can lead to biological changes affecting the metabolic system with downstream consequences on children’s weight and nutrition (Heindel and Schug,2014). Second, ingesting pollutants and microbial contaminants can disrupt the normal development and pruning of synapses in key areas of the brain (e.g., the prefrontal cortex), impacting brain structure and function and consequently cognition, social-emotional development, and behavior (Walker et al.,2011). 1 Handwashing and combined water, sanitation, handwashing, and nutrition interventions might improve child motor development (Stewart et al.,2018).1Finally, pollution and the lack of clean water, adequate sanitation, and waste management sources can be a major source of stress for families, which can affect their behavior and the type and quality of care that they provide to children (Bornstein et al.,2015). In sum, evidence supports hypothesizing that inadequate water, sanitation, and garbage disposal can impact childhood health and development, with downstream consequences for lifelong individual and societal outcomes. Consistent with these findings, investing in improved water, sanitation, and waste management infrastructure can be a key strategy for promoting positive health and developmental outcomes early in life (Acuff and Kaffine,2013). Environmental problems such as odors and air and water pollution are caused by poor waste management and may be exacerbated as the climate crisis deepens (Janmaimool,2017;Vaccari et al.,2019). Poor waste management increases the risk of severe diseases, like dysentery and jaundice (Rathnamala et al.,2021). However, budget and financial constraints, socioeconomic factors, and negative attitudes towards waste management make it difficult to invest in waste management infrastructure in LMICs (Al-Khatib et al.,2016; Vaccari et al.,2019;S¸chiopu et al.,2007;Manga et al.,2019;Wang et al.,2018;Keser et al.,2012; Rathnamala et al.,2021). Furthermore, these policies are hard to target as responsibility for the generation of solid waste is shared across industry, households, and government (Abdulredha et al., 2018;Yamamoto and Kinnaman,2022;Hamilton et al.,2013;Agyei-Mensah and Oteng-Ababio, 2012). Addressing environmental infrastructure gaps is crucial for enhancing ECD in Latin America and the Caribbean (LAC) region (G¨unther and Lahoti,2021;Schady,2015;Korc and Hauchman,2021). While LAC has made significant strides in enhancing access to clean water, sanitation remains a pressing concern, especially in rural areas (Alvarez,2019;Jeuland et al.,2013). Similarly, the region has made significant progress in waste management, with an average solid waste collection rate higher than the global average. Yet, challenges persist in achieving consistent and equitable waste management services across different areas. The region grapples with variations in service quality and frequency, as well as a low level of municipal planning in solid waste management (Grau et al., 2015). These infrastructural challenges not only pose immediate health risks, but also could have long-term implications on cognitive and physical development. The region’s unique socio-economic dynamics, coupled with its diverse topography and climate, further accentuate these challenges. Vulnerable areas often are characterized by limited state presence and investment. Families in these locations often lack the resources or knowledge to adopt mitigative measures, exacerbating the health and developmental risks for children. In stark contrast, urban centers, despite better infrastructure, grapple with challenges posed by rapid urbanization and population density, often leading to overburdened facilities (Sparkman and Sturzenegger,2016). 1However, modestly improved sanitation practices do not necessarily enhance child health outcomes, underscoring the complexity of achieving health benefits through interventions (Patil et al.,2014). 2 This paper aims to provide evidence on the potential positive effects of improved water, sanitation, and waste management infrastructure on ECD in LAC. Our findings can help provide important evidence to support targeted investments that benefit the environment, local populations, and young children. In particular, we seek to address the following research questions. First, does access to improved water, sanitation, and waste management impact young children’s health and development? Second, to what extent do the links between access to improved water, sanitation, and waste disposal vary across countries, when controlling for individual and household characteristics, such as maternal education, wealth, health status, stimulation behaviors, and the availability of certain assets and goods (e.g., hand cleansers, refrigerators, access to bottled water, and lidded containers)? We use two main data sources to examine these questions. First, we employ UNICEF’s Multiple Indicators Cluster Surveys (MICS) from Belize, Costa Rica, Cuba, El Salvador, Guyana, Honduras, Jamaica, Panama, Paraguay, Dominican Republic, Saint Lucia, Suriname, Trinidad and Tobago, and Uruguay, which include information on access to improved water and sanitation and on child health, and development. We also use the Demographic and Health Surveys (DHS) from Bolivia, Colombia, Dominican Republic, Guyana, Nicaragua, and Peru, which provide information on child health and nutrition, and household waste management. We conducted a thorough multivariate statistical analysis, utilizing linear probability models to explore the correlations between infrastructure access and developmental and health outcomes, while taking advantage of the differences in infrastructure between countries. To mitigate potential endogeneity concerns, especially the confounding effect of household wealth, we incorporated control variables in our models. We further validated our findings by incorporating the stimulation index, which is a widely recognized predictor of ECD. Additionally, we used categorical variables to analyze access to water and sanitation, and applied a propensity score matching approach to the data. Despite these variations, our results remain consistent. The findings of our paper significantly emphasize the core associations between access to improved water, sanitation, waste management, and the developmental and health outcomes in young children. Children in households with better access to these facilities are more likely to be developmentally on track and less prone to stunting. Notable variability in these associations across countries calls for further exploration of other environmental factors affecting children’s health and development. Additionally, improved garbage collection services are linked with reduced rates of stunting and underweight among children under five. We also explore mitigating behaviors that households can adopt amid inadequate access to clean water and sanitation facilities. Our findings advocate for promoting household water treatment practices and using lidded containers for waste as provisional solutions towards universal access to water and sanitation. Examining predicted rates of stunting and underweight among children who have experienced recent illness episodes highlights potential underlying mechanisms affecting child development, emphasizing the health channel. 3 Our findings resonate with and extend the existing literature, marking an advancement in comprehending the infrastructure-health nexus. Gao et al. (2021), using a sample of more than 88,000 children aged 26-59 months in 20 countries in Sub-Saharan Africa, found a positive association between a composite index of improved housing (including access to better water sources and sanitation facilities) and early childhood development. In addition, Schady (2015) highlights the disparities in water and sanitation infrastructure access in LAC. It outlines the public health implications of inadequate facilities, linking microbial ingestion to health adversities such as diarrhea, a significant factor in under-five mortality in developing regions. While existing evidence substantiates the relationship between improved infrastructure and reduced child mortality, a further need exists to conduct new evaluations to assess the impact on child morbidity, nutritional status, and development. There is a lack of well-identified evidence within the LAC region, advocating for comprehensive evaluations to understand the multi-dimensional benefits of improved water and sanitation access for children over time (Schady,2015;G¨unther and Lahoti,2021). To the best of our knowledge, our paper is the first to explore the independent relationships between access to water, sanitation, and waste management, and their impact on ECD and health outcomes in LAC using between-country heterogeneity. We particularly emphasize the inclusion of waste management as part of access to basic infrastructure. This disaggregated information is critical for understanding specific early childhood developmental and health outcomes that may be impacted by physical environment features and for informing future investments in infrastructure to promote positive developmental trajectories from a young age. The remainder of this paper is organized as follows. In section 2we summarize the methodology, including the measurement tools and estimation method. Section 3describes the data used to assess the links between access to improved water sources, sanitation, and garbage disposal and early childhood developmental and health outcomes. Section 4summarizes the results. We conclude with a general discussion and implications for future research and practice. 2 Methodology: Measurement and estimation In this paper, we consider two main outcome variables: (1) child development and (2) child health/nutrition. We use the early childhood development index (ECDI) (Loizillon et al.,2017) as a measure of young children’s development. The ECDI is a 10-item parent-reported scale that aims to assess the basic developmental skills of children aged 3-5 years in four developmental domains: social-emotional, literacy-numeracy, approaches to learning, and physical. The social-emotional domain consists of three items asking whether the child (1) gets along well with other children, (2) kicks, bites, or hits other children or adults, and (3) is easily distracted. Two items cover literacynumeracy: (1) whether the child can identify or name at least ten letters of the alphabet, and (2) whether the child can read at least four simple words. Approaches to learning is measured by two items: (1) whether the child follows simple instructions and (2) whether, when given something to 4 Figure 4: Access to water and sanitation infrastructure by survey (a) Water access (dummy) BZ2006 (77.4%) BZ2011 (84.0%) BZ2015 (87.5%) CR2011 (99.2%) CU2019 (98.2%) DO2014 (95.1%) DO2019 (96.9%) GY2006 (50.9%) GY2014 (69.0%) GY2019 (70.0%) HN2019 (92.7%) JM2011 (89.6%) LC2012 (97.3%) MX2015 (97.9%) PA2013 (80.8%) PY2016 (92.4%) SR2006 (66.1%) SR2010 (54.0%) SR2018 (77.9%) SV2014 (95.7%) TT2011 (91.2%) UY2012 (99.7%) (b) Sanitation access (dummy) BZ2006 (92.1%) BZ2011 (95.4%) BZ2015 (92.1%) CR2011 (98.2%) CU2019 (89.7%) DO2014 (94.9%) DO2019 (92.6%) GY2006 (91.8%) GY2014 (91.2%) GY2019 (95.6%) HN2019 (85.2%) JM2011 (99.1%) LC2012 (95.9%) MX2015 (97.8%) PA2013 (76.9%) PY2016 (80.5%) SR2006 (80.0%) SR2010 (69.4%) SR2018 (92.2%) SV2014 (96.2%) TT2011 (98.3%) UY2012 (98.8%) Notes: This figure shows the percentage of children who have access to water and sanitation services in their homes across different countries in Latin America and the Caribbean (LAC). Data were obtained from the Multiple Indicators Cluster Survey (MICS) for Belize (BZ), Costa Rica (CR), Cuba (CU), Dominican Republic (DO), El Salvador (SV), Guyana (GY), Honduras (HN), Jamaica (JM), Mexico (MX), Panama (PA), Paraguay (PY), Saint Lucia (LC), Suriname (SR), Trinidad and Tobago (TT), and Uruguay (UY). We present statistics for water access in Figure 4a, and sanitation access in Figure 4b. Statistics correspond to simple national averages from each survey. 11 Figure 5: Developmental and health outcomes by survey and access to both water and sanitation (a) ECDI 86.52 83.73 76.19 90.84 81.54 82.32 75.85 81.19 74.57 71.35 90.67 74.40 69.95 77.32 87.50 59.60 70.95 96.30 81.82 89.13 85.13 79.55 93.87 87.94 87.45 82.04 88.86 88.85 74.77 87.82 81.75 82.52 84.15 90.35 73.60 82.06 92.36 89.34 BZ2011 BZ2015 CR2011 CU2019 DO2014 DO2019 GY2014 GY2019 HN2019 JM2011 LC2012 MX2015 PA2013 PY2016 SR2010 SR2018 SV2014 TT2011 UY2012 60 70 80 90 100 No access Access (b) Stunted 22.12 18.51 22.36 8.12 12.35 24.09 23.82 15.95 12.87 28.62 26.49 12.23 0.00 12.84 11.02 7.53 7.38 19.30 19.60 13.05 6.76 6.52 12.21 12.54 8.61 10.40 17.35 12.28 4.55 2.95 7.12 5.94 4.96 9.85 BZ2006 BZ2011 BZ2015 CU2019 DO2019 GY2006 GY2014 GY2019 HN2019 LC2012 MX2015 PY2016 SR2006 SR2010 SR2018 SV2014 TT2011 0 10 20 30 No access Access (c) Underweight 1.33 3.08 0.83 2.37 3.06 2.19 4.38 5.87 4.78 1.37 1.79 0.66 0.00 3.33 2.91 4.90 4.92 1.58 3.02 1.90 2.47 2.19 2.03 6.03 5.47 5.83 1.87 1.04 0.93 3.32 4.65 4.97 4.84 5.95 BZ2006 BZ2011 BZ2015 CU2019 DO2019 GY2006 GY2014 GY2019 HN2019 LC2012 MX2015 PY2016 SR2006 SR2010 SR2018 SV2014 TT2011 0 2 4 6 No access Access Notes: This chart displays the percentage of children who are considered developmentally on track according to the early childhood development index (ECDI), as well as the rates of stunting and underweight. Figures for each of these categories are shown in Figure 5a, Figure 5b, and Figure 5c, respectively. Statistics are unweighted averages calculated separately for each survey for children with and without access to infrastructure services. Data were obtained from the Multiple Indicators Cluster Survey (MICS) for Belize (BZ), Costa Rica (CR), Cuba (CU), Dominican Republic (DO), El Salvador (SV), Guyana (GY), Honduras (HN), Jamaica (JM), Mexico (MX), Panama (PA), Paraguay (PY), Saint Lucia (LC), Suriname (SR), Trinidad and Tobago (TT), and Uruguay (UY). 12 sanitation access, with a higher percentage of children with access to sanitation being on track with ECDI in most countries. In nearly all countries, the difference in the percentage of children who are developmentally on track is significant between those who do and do not have access. The data also show in Panel B that the percentage of children with stunting is higher among those without access to infrastructure in most countries, except Trinidad and Tobago and Saint Lucia. The largest absolute difference is found in Honduras, Mexico and El Salvador, between those with and without access, while Guyana in 2019 has the smallest difference. In Panel C, we provide descriptive information about underweight, which displays greater variation between countries compared to ECDI and stunting. Generally, children with access to both services report higher incidence of underweight than those without access. However, there is a remarkably low incidence of underweight across all countries in general. Overall, the results presented in Figure 5indicate that access to water and sanitation services might be linked to early childhood development outcomes, with children who have access to these services generally having better ECDI scores and experiencing lower rates of stunting. We try to explore the particular low incidence of underweight across countries during the empirical models. We employ data from the Demographic and Health Surveys (DHS) to examine households’ garbage management in LAC. DHS surveys are nationally representative household surveys that collect data on various aspects of health, including reproductive health, child health, and nutrition. They cover both low- and middle-income countries, enabling comparative analyses across different contexts. Moreover, these surveys provide detailed information on garbage collection, such as type and frequency, which is relevant for our hypothesis that children’s health outcomes are negatively affected by higher exposure to residential waste. As shown in Figure 6, we retrieved garbage collection data from 14 surveys with a total of 118,597 mother-child dyads. Access to garbage collection services varies significantly across countries. In Guyana, for instance, access to garbage collection was low at 12.1% in 1998, rising to 21% in 2005. Meanwhile, Nicaragua reported a low access rate of around 23% in both 1998 and 2001. The Dominican Republic showed progress, with the percentage of access increasing from 60.7% in 2007 to 74.6% in 2013. Notably, access to garbage collection services is steady in Peru across multiple surveys from 1991 to 2011. Even over two decades, the percentage of access to garbage collection services remained almost stagnant, hovering around 34%. Additional statistics for the pooled sample of countries/surveys are available in Appendix Table A5. 13 Figure 6: Access to garbage collection by survey BO2003 (43.5%) CO2005 (71.4%) CO2010 (55.6%) DR2007 (60.7%) DR2013 (74.6%) GU1998 (12.1%) GU2005 (21.0%) NC1998 (23.4%) NC2001 (23.6%) PE1991 (32.6%) PE2004 (34.4%) PE2007 (34.4%) PE2009 (34.2%) PE2011 (34.0%) Notes: This figure presents the percentage of children whose households had access to garbage collection services across different surveys. Data were obtained from the Demographic and Health Surveys (DHS) for Bolivia (BO), Colombia (CO), Dominican Republic (DR), Guyana (GU), Nicaragua (NC), and Peru (PE). These data only include countries for which garbage collection information is available from the DHS surveys. Statistics represent unweighted averages. Figure 7presents differences in stunting and underweight by the availability of garbage collection across surveys. The largest difference in stunting between children with and without garbage collection services is observed in Guyana (1998), at 30.98% and 62.72%, respectively, while the smallest difference in stunting was observed in the Dominican Republic (2013), at 7.49% and 10.1% respectively. Similarly, children with access to garbage collection services are less frequently underweight than those without access. The largest difference in underweight children between those with and without garbage collection services was observed in Guyana (1998), at 9.36% and 21.25% respectively. The smallest difference in underweight children was observed in the Dominican Republic (2013), at 3.14% and 4.8% respectively. In general, these statistics support our hypothesis that children’s health outcomes may be negatively affected by higher exposure to residential waste, highlighting the importance of investing in efficient waste management systems for promoting better health and early childhood development outcomes. 4 Results In this section, we provide evidence of the potentially significant role that access to basic infrastructure services, such as water, sanitation, and garbage collection, plays in early childhood health and development. Specifically, the MICS data demonstrate a strong link between access to water and sanitation and children’s likelihood of being on track in their development according to the ECDI, as well as a lower likelihood of stunting. In addition, the DHS data highlight a negative association between access to improved garbage collection services and rates of stunting and underweight among children under five. 14 Figure 7: Stunting and underweight by type of garbage disposal system (a) Stunting 41.48 23.98 26.82 14.55 10.10 62.72 29.30 37.20 31.90 47.91 37.09 37.09 34.58 29.72 23.89 12.59 11.50 9.64 7.49 30.98 16.15 20.83 11.94 19.76 13.84 13.84 13.36 10.92 BO2003 CO2005 CO2010 DR2007 DR2013 GU1998 GU2005 NC1998 NC2001 PE1991 PE2004 PE2007 PE2009 PE2011 10 20 30 40 50 60 Stunting (%) Non-collected Collected (b) Underweight 7.44 7.52 5.78 4.49 4.80 21.25 10.14 11.54 8.58 12.21 5.80 5.80 6.44 6.32 3.36 4.30 2.69 3.15 3.14 9.36 8.50 6.16 3.81 3.44 2.33 2.33 2.51 1.97 BO2003 CO2005 CO2010 DR2007 DR2013 GU1998 GU2005 NC1998 NC2001 PE1991 PE2004 PE2007 PE2009 PE2011 0 5 10 15 20 Underweight (%) Non-collected Collected Notes: This figure presents the percentage of children with stunting and underweight across different surveys by access to garbage collection. Data were obtained from the Demographic and Health Surveys (DHS) for Bolivia (BO), Colombia (CO), Dominican Republic (DR), Guyana (GU), Nicaragua (NC), and Peru (PE). The data presented here only include countries for which garbage collection information is available from the DHS surveys. Statistics represent unweighted averages. 15 Water and sanitation infrastructure using the MICS: We present the results of the linear probability models that explore the association between access to water and sanitation and various developmental and health outcomes. These outcomes include binary indicators for being on track in the ECDI, as well as stunting and underweight status in children. The data utilized for this analysis originate from 22 surveys conducted across 14 different countries. It is important to note that the specific countries and surveys included in the analysis vary based on the model and health outcome in question, as not all outcomes are available across all surveys.11 Table 1: Access to infrastructure and the relationship with the ECDI Dependent variable ECDI (dummy) (1) (2) (3) (4) (5) (6) (7) (8) Water access (dummy) 0.039*** 0.041*** 0.027** 0.019* (0.010) (0.010) (0.010) (0.010) Sanitation access (dummy) 0.059*** 0.058*** 0.037*** 0.029*** (0.009) (0.008) (0.008) (0.008) R2-squared 0.039 0.051 0.059 0.063 0.040 0.052 0.060 0.063 Mean of outcome 0.834 0.834 0.834 0.834 0.834 0.834 0.834 0.834 N 37964 37964 37964 37964 37964 37964 37964 37964 Surveys 19 19 19 19 19 19 19 19 Controls Country-region-year FE Yes Yes Yes Yes Yes Yes Yes Yes Children level No Yes Yes Yes No Yes Yes Yes Mother level No No Yes Yes No No Yes Yes Household characteristics No No No Yes No No No Yes Notes: This table presents results from a linear probability model for the correlation between child development indicators and access to water and sanitation infrastructure. The analysis sample is composed from 19 surveys included in The Multiple Indicator Cluster Surveys (MICS). The countries that have been included in this analysis and their respective sample sizes can be found in Appendix Table A1. The analysis controls for fixed effects at the country-region-year level. Additional controls at the child level include age (in months), birth cohort, and sex. Mother-level controls account for the mother’s education level. Household characteristics controlled for are urban residency status, access to electricity, total household size, and the number of household members who are under five years of age. Variable definitions are available in Appendix Table A2 and summary statistics in Appendix Table A3. Survey clustered standard errors are reported in parentheses (*** p<0.01, ** p<0.05, * p<0.1). In Table 1, we analyze the association between access to water and sanitation, and a child’s developmental progress as captured by the ECDI. The models, spanning columns 1 to 8, reveal a consistent positive association. For instance, the model in column 1, accounting for country-region- year fixed effects, suggests that access to water correlates with a 3.9% increase in the likelihood of a child being developmentally on track according to the ECDI. This relationship persists even when additional control variables are introduced, including child, mother, and household characteristics. The coefficients for water and sanitation access slightly diminish but remain significantly positive, 11The sample across the different models and outcomes is not harmonized due to substantial variations in the number and type of observations when comparing Tables 1and 2. Despite this, the results remain consistent and robust when using the same sample, as evidenced by the findings presented in Appendix Table A6. 16 emphasizing the robustness of these findings. Children with water access exhibit a 1.9% to 4.1% higher likelihood of optimal development, while sanitation access corresponds to a 2.9% to 5.8% increase. In Appendix Table A7, we refine this analysis, dissecting the ECDI into its constituent subdomains: learning, socio-emotional development, physical development, and literacy/numeracy. The models reveal that water access predominantly benefits learning, socio-emotional development, and literacy/numeracy, with associations as high as 7.6%, 1.5%, and 1.9% respectively. Sanitation access echoes these benefits, particularly enhancing learning (9.0%), socio-emotional development (2.0%), and literacy/numeracy (3.4%). The physical development sub-domain, while showing a positive trend, does not reach statistical significance, possibly due to short-term measurement limitations or variability in physical development milestones. In Table 2, we explore the associations between access to water and sanitation and health outcomes, particularly stunting and being underweight in children. The analysis reveals a noteworthy correlation: access to water is associated with a 1.3% to 3.8% reduction in the likelihood of stunting and a 0.3% decrease in the probability of a child being underweight. These associations are further strengthened with access to sanitation, which is correlated with a 4.8% to 8.1% decrease in the likelihood of stunting and a 0.3% reduction in being underweight. Although the percentage changes might seem modest, they are significant when considered in relation to the baseline prevalence of these health outcomes, thus highlighting their potential relevance for public health efforts. We tested additional models that include wealth to mitigate endogeneity issues and obtain more accurate estimates. Specifically, wealth may predict both access to improved sanitation facilities as well as developmental and health outcomes, leading to issues of selection or omitted variable bias. By controlling for wealth, we can account for the possibility that wealthier individuals are more likely to have better sanitation facilities while also simultaneously enjoying better health outcomes due to other factors that are correlated with wealth. These results are shown in Appendix Table A8. In our baseline model, we have opted not to include wealth due to the construction of the wealth index, which integrates proxies related to infrastructure access. This integration could potentially obscure the independent effects of water and sanitation access.12 The estimates in Appendix Table A8 show that even after controlling for household wealth, water and sanitation access are both positively associated with on-track ECDI scores and negatively associated with stunting and underweight. However, only sanitation is statistically significant. We confirm that the coefficients of the wealth quintile dummy variables show that children from richer households are more likely to be developmentally on track and have lower rates of stunting and underweight than those from poorer households. Previously, all our models control for maternal education in order to separate out the impact of maternal education on child developmental and health outcomes. Indeed, maternal education is linked to better hygiene, nutrition, and healthcare- 12See Appendix Table A2 for additional details about the construction of the wealth index. 17 Table 2: Access to infrastructure and the relationship with health outcomes Dependent variable Stunted (dummy) Underweight (dummy) (1) (2) (3) (4) (5) (6) (7) (8) Water access (dummy) -0.038***-0.038*** -0.026** -0.013 -0.003 -0.003 -0.003 -0.003 (0.012) (0.012) (0.010) (0.008) (0.003) (0.003) (0.003) (0.003) R2-squared 0.044 0.050 0.060 0.068 0.014 0.017 0.017 0.017 Mean of outcome 0.116 0.116 0.116 0.116 0.027 0.027 0.027 0.027 N 67248 67248 67248 67248 67248 67248 67248 67248 Surveys 17 17 17 17 17 17 17 17 Controls Country-region-year FE Yes Yes Yes Yes Yes Yes Yes Yes Children level No Yes Yes Yes No Yes Yes Yes Mother level No No Yes Yes No No Yes Yes Household characteristics No No No Yes No No No Yes (9) (10) (11) (12) (13) (14) (15) (16) Sanitation access (dummy) -0.081***-0.081***-0.061***-0.048*** -0.003 -0.003 -0.002 -0.002 (0.011) (0.011) (0.010) (0.007) (0.002) (0.002) (0.002) (0.002) R2-squared 0.047 0.053 0.062 0.069 0.014 0.017 0.017 0.017 Mean of outcome 0.116 0.116 0.116 0.116 0.027 0.027 0.027 0.027 N 67248 67248 67248 67248 67248 67248 67248 67248 Countries 17 17 17 17 17 17 17 17 Controls Country-region-year FE Yes Yes Yes Yes Yes Yes Yes Yes Children level No Yes Yes Yes No Yes Yes Yes Mother level No No Yes Yes No No Yes Yes Household characteristics No No No Yes No No No Yes Notes: This table presents results from a linear probability model for the correlation between child development indicators and access to water and sanitation infrastructure. The analysis sample is composed from the surveys included in the Multiple Indicator Cluster Surveys (MICS). The countries that have been included in this analysis and their respective sample sizes can be found in Appendix Table A1. The analysis controls for fixed effects at the country-region-year level. Additional controls at the child level include age (in months), birth cohort, and sex. Mother-level controls account for the mother’s education level. Household characteristics controlled for are urban residency status, access to electricity, total household size, and the number of household members who are under five years of age. Variable definitions are available in Appendix Table A2 and summary statistics in Appendix Table A4. Survey clustered standard errors are reported in parentheses (*** p<0.01, ** p<0.05, * p<0.1). seeking behavior for children, and may influence household wealth and investment in water and sanitation infrastructure. Lastly, we also tested the robustness of the results to the inclusion of the stimulation index, which is a strong predictor of ECD (Cuartas et al.,2023), obtaining consistent estimates. Recognizing the inherent interconnectedness of sanitation and water in determining health and developmental outcomes, it becomes essential to investigate not just their individual influences, but also their combined effects. The results of these interactions are illustrated in Figure 8. By including an interaction term between sanitation and water, we aim to distinguish whether their joint influence amplifies or diminishes the outcomes observed when they act independently. The 18 results for ECDI indicate a positive association with both water and sanitation access. Children with access to both water and sanitation have a predicted 83.8% likelihood of being on track in ECDI, which is higher compared to those with access to only water (81.7%), only sanitation (82.7%), or neither (77.7%). When it comes to stunting children with access to both water and sanitation have the lowest likelihood of being stunted, at 11.00%. Those without access to either have a higher likelihood of 16.75%, Similar trends are observed for children with access to only water (15.76%) or only sanitation (11.95%). The pattern continues for underweight, an indicator of acute malnutrition. Children with access to both water and sanitation have a likelihood of 2.69%, the lowest among the groups. Those without access to either are at a higher likelihood of 3.56%. Access to water only results in a likelihood of 2.62%, while sanitation only shows a likelihood of 2.77%. Figure 8: Predicted developmental and health outcomes by infrastructure access No access to water or sanitation Sanitation access only Water access only Access to both water and sanitation 0.74 0.76 0.78 0.80 0.82 0.84 0.10 0.12 0.14 0.16 0.18 0.02 0.03 0.04 ECDI (dummy) Stunted (dummy) Underweight (dummy) Proportion Notes: The coefficient plot visually displays the predictive margins for three distinct health and development outcomes in children: being on track according to the early childhood development index (ECDI), being stunted, and being underweight. The figure categorizes the results based on the children’s access to water and sanitation, delineating four scenarios: “No access to water or sanitation”, “Water access only”, “Sanitation access only”, and “Access to both water and sanitation”. The predictive margins are calculated from the linear probability models in Appendix Table A9. Variable definitions are available in Appendix Table A2 and summary statistics in Appendix Tables A3 and A4. Confidence intervals are represented by capped lines and set at 90%. In order to mitigate the negative impacts of inadequate access to clean water and sanitation facilities, households may adopt various behaviors or attitudes that promote health and well-being. These practices, ranging from treating water to enhance its safety for consumption, to ensuring the availability of soap or cleansers for handwashing, play a crucial role in safeguarding health, particularly in contexts where access to water and sanitation infrastructure is limited. As shown in Figure 9a, water treatment emerges as a significant mitigating factor, especially for developmental outcomes. Promoting household water treatment practices could be a viable interim solution while 19 working towards universal access to water and sanitation. The data also suggest that the availability of a cleanser might play a mitigating role, particularly in relation to ECDI and stunting (see Figure 9b). The robustness of our results is reinforced through a series of additional analyses. First, we confirm our results through the application of a preferential within propensity score matching approach (Arpino and Cannas,2016), as demonstrated in Appendix Table A13, providing further confidence in the validity and reliability of our findings. Second, we extend further our analyses with the results presented in Tables A11 and A12. By employing a categorical variable for both water and sanitation access, rather than relying on binary indicators, we explore the correlation of different types of access. Our findings interestingly highlight the significant role of packaged water in mitigating the adverse effects associated with other forms of water access. This observation could be indicative of the unreliable quality of piped water sources in some Latin American countries and the lack of appropriate storage options, underscoring the potential need for alternative safe drinking water solutions. On the sanitation front, flush systems stand out as particularly beneficial, showcasing greater effectiveness in promoting health and developmental outcomes when compared to both hygienic non-flush methods and unsanitary systems, as well as in comparison to a complete lack of sanitation facilities. These findings underline the necessity to recognize and address the varying impacts of different types of water and sanitation access. They also emphasize the importance of considering the quality and type of water and sanitation services alongside their availability. The data from the MICS compellingly illustrate the positive relationship that investments in water and sanitation infrastructure can have on the development and health of young children. This relationship holds strong even when taking into account various individual, maternal, and household factors, alongside considerations of household wealth and stimulation activities. To get more into this association, we expand the data analyses to include information on residential waste collection systems. Such data would provide a more comprehensive understanding of how waste management infrastructure services impact early childhood development. Since exposure to waste can affect the quality of an individual’s environment, this additional information may enhance the accuracy of the relationship between infrastructure services and early childhood development. Residential waste management infrastructure using the DHS: Table 3presents the results of the linear probability model to assess the correlation between garbage collection and child health indicators, specifically stunting and underweight. Our analysis begins with a simple correlation using year-country-region fixed effects, which is then extended by introducing additional controls, including children, mother and household characteristics. Controlling for household characteristics allows us to account for the potential impact of socio-economic status on the relationship between garbage collection and health indicators, as discussed previously. 20 Figure 14: Predicted stunting and underweight for recently sick children by type of garbage collection Diarrhea=0Diarrhea=1 Non-collected Collected Non-collected Collected 0.22 0.23 0.24 0.25 0.26 0.27 Cough=0Cough=1 Non-collected Collected Non-collected Collected 0.22 0.23 0.24 0.25 0.26 0.27 Fever=0Fever=1 Non-collected Collected Non-collected Collected 0.22 0.23 0.24 0.25 0.26 0.27 Stunting Diarrhea=0Diarrhea=1 Non-collected Collected Non-collected Collected 0.05 0.06 0.07 0.08 Cough=0Cough=1 Non-collected Collected Non-collected Collected 0.05 0.06 0.07 0.08 Fever=0Fever=1 Non-collected Collected Non-collected Collected 0.05 0.06 0.07 0.08 Underweight Sick recently × Garbage collection Proportion Notes: The coefficient plot visually displays the predictive margins for two distinct health outcomes in children: being stunted, and being underweight. The results are organized based on recent illness episodes such as diarrhea, cough, and fever in children. The figure then presents estimations according to the children’s access to garbage collection. The predictive margins are calculated from the linear probability models in Appendix Table A20. Variable definitions are available in Appendix Table A2 and summary statistics in Appendix Table A5. Confidence intervals are represented by capped lines and set at 90%. light of these insights, targeted policy interventions that address waste management and sanitation infrastructure, with a particular focus on households with children who have experienced recent episodes of illness may be effective. This approach will help mitigate the negative impacts of poor waste management on stunting as well as underweight status, thereby promoting improved child health and development outcomes. 5 Conclusion This paper provides insights into the link between access to essential infrastructure services and early childhood development (ECD) in Latin America and the Caribbean (LAC). Our findings show that access to water, sanitation, and garbage collection can significantly impact children’s developmental and health outcomes. Specifically, we document that access to water and sanitation is strongly associated with a higher likelihood of being developmentally on track as measured by the 27 early childhood development index, and with fewer instances of stunting and underweight. Additionally, the results reveal a negative association between improved garbage collection services and rates of stunting and underweight among children under five. Our findings hold strong even when taking into account various individual, maternal, and household factors, alongside considerations of household wealth and stimulation activities. Our findings are consistent with prior evidence (Piper et al.,2017;WaterAid,2016;Schady, 2015;Gao et al.,2021), and have important implications for policymakers in the region. First and foremost, the expansion of piped water and sanitation services is crucial for improving child outcomes. This expansion aligns with evidence indicating that improvements in water infrastructure can substantially reduce child mortality, with greater effects among socioeconomically disadvantaged groups. It is essential, however, to recognize the complexity of separating the effects of improved water from that of improved sanitation, as the magnitude of the relationship of clean water with child outcomes is often different than that of improved sanitation. A key dimension of our analysis that merits significant emphasis is the differential impact of types of access to infrastructure, a crucial aspect considering the Sustainable Development Goals (SDGs) have elevated standards for what constitutes adequate access to water and sanitation. Our analysis underscores the pivotal role of alternatives to piped water in either mitigating or exacerbating the effects of water quality received by households. This is especially pertinent in contexts where piped water sources may be unreliable, highlighting the need for alternative safe drinking water solutions. Regarding sanitation, flush systems demonstrate substantial benefits, promoting health and developmental outcomes more effectively than both hygienic non-flush methods and unsanitary systems, even in the absence of sanitation facilities. These findings stress the need to recognize and address the varying impacts of different types of water and sanitation access, emphasizing the necessity for specific measurements of quality and access to essential infrastructure. There exist challenges in accurately measuring real household access, a critical consideration for policy and infrastructure development. We underscore the importance of investing in waste management infrastructure as a critical but often overlooked component of comprehensive strategies to enhance early childhood development in low- and middle-income countries. This is particularly critical in light of previous work (Lu et al.,2020;McCoy et al.,2022) and our descriptive observations of persistent inequities in access to basic water, sanitation, and waste disposal services in LAC. By ensuring that all households have access to all basic services that promote good health and foundational skill development, policymakers can improve the developmental trajectories and well-being of children across the region and maximize the beneficial impact of other investments (including early childhood care and education and parenting programs) while promoting sustainable economic development for years to come. 28 Our paper also highlights the need for targeted policy interventions aimed at improving household practices in the region. The literature focuses on infrastructure hardware rather than on interventions that modify household behaviors and mitigation measures (Schady,2015;Rupasinghe et al., 2021). For example, households could be encouraged to protect their environment from potential pollutants through mitigation measures such as storing waste in lidded containers or participating in community-based recycling programs. Similarly, strategies like handwashing, and treating water hold promise, but their effectiveness often relies on behavioral changes. Policymakers could consider implementing public education campaigns to raise awareness of the importance of proper household practices for healthy child development. Parenting and ECD programs delivered through home visits and community meetings have been expanding in LAC and offer a potential means of providing information and support related to hygiene practices and the protection of children from environmental hazards (Leer and Lopez-Boo,2019). Importantly, our findings suggest that households in the region can reap significant economic benefits from investments in water, sanitation, and waste management infrastructure and their downstream benefits for child health and development. Indeed, recent simulation studies estimate that stunting alone causes learning and economic losses with a total cost of approximately $176.8 billion per birth cohort globally and $44.7 billion in Latin America (Fink et al.,2016), while the cost of not investing in critical interventions to promote ECD in LAC countries such as Guatemala, Nicaragua, Colombia, Peru, Ecuador, and Chile ranged from 0.05% to 3.6% of their gross domestic product (Richter et al.,2017). The cost of inaction is therefore substantial and has meaningful impacts on individual and societal outcomes. By promoting ECD and reducing rates of stunting and underweight among young children, practices to improve water, sanitation, and waste management can lead to lower healthcare costs for families, maximize the impact of subsequent interventions, and promote higher levels of educational attainment and workforce productivity over time. There is also a necessity for a multi-sector approach. We underline the need to jointly address a range of factors –environmental, nutritional, health-related– to enhance early childhood development effectively (Rupasinghe et al.,2021;Vargas-Baron et al.,2019;Black et al.,2017). Looking ahead, much remains to be learned about the complex relationship between access to infrastructure and early childhood development in LAC. Future research could explore the specific mechanisms through which access to essential infrastructure services affects child health outcomes, as well as the potential role of cultural and social factors in shaping household practices in the region. Additionally, further studies could explore the long-term economic benefits of investing in infrastructure for promoting healthy child development and sustainable economic growth. 29 References Abdulredha, M., R. Al Khaddar, D. Jordan, P. Kot, A. Abdulridha, and K. Hashim (2018). Estimating solid waste generation by hospitality industry during major festivals: A quantification model based on multiple regression. Waste Management 77, 388–400. Acuff, K. and D. T. Kaffine (2013). Greenhouse gas emissions, waste and recycling policy. Journal of Environmental Economics and Management 65(1), 74–86. Agyei-Mensah, S. and M. Oteng-Ababio (2012). Perceptions of health and environmental impacts of e-waste management in Ghana. International Journal of Environmental Health Research 22(6), 500–517. Al-Khatib, I. A., I. Abu Fkhidah, J. I. Khatib, and S. Kontogianni (2016). Implementation of a multi-variable regression analysis in the assessment of the generation rate and composition of hospital solid waste for the design of a sustainable management system in developing countries. Waste Management & Research 34(3), 225–234. Alvarez, L. (2019, June). Access to water and sanitation services in dispersed rural settlements: The path to universalization. Technical report. Arpino, B. and M. Cannas (2016). Propensity score matching with clustered data. An application to the estimation of the impact of caesarean section on the Apgar score. Statistics in Medicine 35(12), 2074–2091. Attanasio, O., S. Cattan, E. Fitzsimons, C. Meghir, and M. Rubio-Codina (2020). Estimating the production function for human capital: Results from a randomized controlled trial in Colombia. American Economic Review 110(1), 48–85. Black, M. M., S. P. Walker, L. C. Fernald, C. T. Andersen, A. M. DiGirolamo, C. Lu, D. C. McCoy, G. Fink, Y. R. Shawar, J. Shiffman, et al. (2017). Early childhood development coming of age: Science through the life course. The Lancet 389(10064), 77–90. Bornstein, M. H., T. Leventahl, and R. M. Lerner (2015). Handbook of child psychology and developmental science: Ecological settings and processes. Wiley. Cuartas, J. (2021). The effect of maternal education on parenting and early childhood development: An instrumental variables approach. Journal of Family Psychology. Cuartas, J. (2022). The effect of spanking on early social-emotional skills. Child Development 93(1), 180–193. Cuartas, J., D. McCoy, J. S´anchez, J. Behrman, C. Cappa, G. Donati, J. Heymann, C. Lu, A. Raikes, N. Rao, et al. (2023). Family play, reading, and other stimulation and early childhood development in five low-and-middle-income countries. Developmental Science, e13404. Cuartas, J., C. Rey-Guerra, D. C. McCoy, and E. Hanno (2020). Maternal knowledge, stimulation, and early childhood development in low-income families in Colombia. Infancy 25(5), 526–534. Cuartas, J., D. G. Weissman, M. A. Sheridan, L. Lengua, and K. A. McLaughlin (2021). Corporal punishment and elevated neural response to threat in children. Child Development 92(3), 821– 832. Cunha, F. and J. Heckman (2007). The technology of skill formation. American Economic Review 97(2), 31–47. Danaei, G., K. G. Andrews, C. R. Sudfeld, G. Fink, D. C. McCoy, E. Peet, A. Sania, M. C. Smith Fawzi, M. Ezzati, and W. W. Fawzi (2016). Risk factors for childhood stunting in 137 developing countries: A comparative risk assessment analysis at global, regional, and country levels. PLoS medicine 13(11), e1002164. 30 Engle, P. L., M. M. Black, J. R. Behrman, M. C. De Mello, P. J. Gertler, L. Kapiriri, R. Martorell, M. E. Young, I. C. D. S. Group, et al. (2007). Strategies to avoid the loss of developmental potential in more than 200 million children in the developing world. The Lancet 369(9557), 229–242. Fink, G., E. Peet, G. Danaei, K. Andrews, D. C. McCoy, C. R. Sudfeld, M. C. Smith Fawzi, M. Ezzati, and W. W. Fawzi (2016). Schooling and wage income losses due to early-childhood growth faltering in developing countries: National, regional, and global estimates. The American Journal of Clinical Nutrition 104(1), 104–112. Gao, Y., L. Zhang, A. Kc, Y. Wang, S. Zou, C. Chen, Y. Huang, X. Mi, and H. Zhou (2021). Housing environment and early childhood development in sub-saharan Africa: A cross-sectional analysis. PLoS Medicine 18(4), e1003578. Grau, J., H. Terraza, D. M. R. Velosa, A. Rihm, and G. Sturzenegger (2015). Solid waste management in Latin America and the Caribbean. G¨unther, I. and R. Lahoti (2021). Transitioning to No Poverty. MDPI-Multidisciplinary Digital Publishing Institute. Hamilton, S. F., T. W. Sproul, D. Sunding, and D. Zilberman (2013). Environmental policy with collective waste disposal. Journal of Environmental Economics and Management 66(2), 337–346. Heckman, J. J. (2006). Skill formation and the economics of investing in disadvantaged children. Science 312(5782), 1900–1902. Heindel, J. J. and T. T. Schug (2014). The obesogen hypothesis: Current status and implications for human health. Current Environmental Health Reports 1(4), 333–340. Janmaimool, P. (2017). Application of protection motivation theory to investigate sustainable waste management behaviors. Sustainability 9(7), 1079. Jeong, J., D. C. McCoy, and G. Fink (2017). Pathways between paternal and maternal education, caregivers’ support for learning, and early child development in 44 low-and middle-income countries. Early Childhood Research Quarterly 41, 136–148. Jeong, J., D. C. McCoy, A. K. Yousafzai, C. Salhi, and G. Fink (2016). Paternal stimulation and early child development in low-and middle-income countries. Pediatrics 138(4). Jeuland, M. A., D. E. Fuente, S. Ozdemir, M. C. Allaire, and D. Whittington (2013). The long-term dynamics of mortality benefits from improved water and sanitation in less developed countries. PloS one 8(10), e74804. Keser, S., S. Duzgun, and A. Aksoy (2012). Application of spatial and non-spatial data analysis in determination of the factors that impact municipal solid waste generation rates in turkey. Waste Management 32(3), 359–371. Knudsen, E. I., J. J. Heckman, J. L. Cameron, and J. P. Shonkoff (2006). Economic, neurobiological, and behavioral perspectives on building america’s future workforce. Proceedings of the National Academy of Sciences 103(27), 10155–10162. Korc, M. and F. Hauchman (2021). Advancing environmental public health in Latin America and the Caribbean. Revista Panamericana de Salud P´ublica 45. Leer, J. and F. Lopez-Boo (2019). Assessing the quality of home visit parenting programs in Latin America and the Caribbean. Early Child Development and Care 189(13), 2183–2196. Loizillon, A., N. Petrowski, P. Britto, and C. Cappa (2017). Development of the early childhood development index in MICS surveys. MICS Methodological Papers, No. 6, Data and Analytics Section, Division of Data. Research and Policy, UNICEF New York, 29. Lu, C., J. Cuartas, G. Fink, D. McCoy, K. Liu, Z. Li, B. Daelmans, and L. Richter (2020). Inequalities in early childhood care and development in low/middle-income countries: 2010– 2018. BMJ Global Health 5(2), e002314. 31 Manga, V. E., O. T. Oru, and M. N. Ngwabie (2019). Household perception and willingness to pay for improved waste management service in Mamfe, Cameroon. African Journal of Environmental Science and Technology 13(9), 354–364. Martel, P. (2016). Review of options for reporting water, sanitation and hygiene coverage by wealth quintile. New York: UNICEF Data and Analytics Section, Division of Data, Research and Policy. McCoy, D. C., A. L. Roy, and C. C. Raver (2016). Neighborhood crime as a predictor of individual differences in emotional processing and regulation. Developmental science 19(1), 164–174. McCoy, D. C., J. Seiden, J. Cuartas, L. Pisani, and M. Waldman (2022). Estimates of a multidimensional index of nurturing care in the next 1000 days of life for children in low-income and middle-income countries: A modelling study. The Lancet Child & Adolescent Health 6(5), 324–334. Patil, S. R., B. F. Arnold, A. L. Salvatore, B. Briceno, S. Ganguly, J. M. Colford Jr, and P. J. Gertler (2014). The effect of India’s total sanitation campaign on defecation behaviors and child health in rural Madhya Pradesh: A cluster randomized controlled trial. PLoS medicine 11(8), e1001709. Perumal, N., D. G. Bassani, and D. E. Roth (2018). Use and misuse of stunting as a measure of child health. The Journal of Nutrition 148(3), 311–315. Piper, J. D., J. Chandna, E. Allen, K. Linkman, O. Cumming, A. J. Prendergast, and M. J. Gladstone (2017). Water, sanitation and hygiene (wash) interventions: Effects on child development in low-and middle-income countries. The Cochrane Database of Systematic Reviews 2017(3). Rathnamala, G., R. Ashwini, and N. Babitha (2021). Domestic environmental destructions due to lack of solid waste management in rural areas. Adv. Math. Sci. J 10(3), 1807–1819. Richter, L. M., B. Daelmans, J. Lombardi, J. Heymann, F. L. Boo, J. R. Behrman, C. Lu, J. E. Lucas, R. Perez-Escamilla, T. Dua, et al. (2017). Investing in the foundation of sustainable development: Pathways to scale up for early childhood development. The Lancet 389(10064), 103–118. Rupasinghe, N., L. Benton, K. Mandeville, and M. Latypova (2021). Lessons from a review of international experience in early childhood development programming, part 1. Salhi, C., E. Beatriz, R. McBain, D. McCoy, M. Sheridan, and G. Fink (2021). Physical discipline, deprivation, and differential risk of developmental delay across 17 countries. Journal of the American Academy of Child & Adolescent Psychiatry 60(2), 296–306. Sania, A., C. R. Sudfeld, G. Danaei, G. Fink, D. C. McCoy, Z. Zhu, M. C. S. Fawzi, M. Akman, S. E. Arifeen, A. J. Barros, et al. (2019). Early life risk factors of motor, cognitive and language development: A pooled analysis of studies from low/middle-income countries. BMJ open 9(10), e026449. Schady, N. (2015). Does access to better water and sanitation infrastructure improve child outcomes? Evidence from Latin America and the Caribbean. Technical report, BID. S¸chiopu, A.-M., I. Apostol, M. Hodoreanu, and M. Gavrilescu (2007). Solid waste in romania: Management, treatment and pollution prevention practices. Environmental Engineering & Management Journal (EEMJ) 6 (5). Sheridan, M. and C. A. Nelson (2009). Neurobiology of fetal and infant development. Handbook of Infant Mental Health 3, 40–58. Sparkman, D. and G. Sturzenegger (2016). Fostering water and sanitation markets in Latin America and the Caribbean: How the public sector can support the private sector to bridge coverage gaps and improve service quality for low-income populations. 32 Stewart, C. P., P. Kariger, L. Fernald, A. J. Pickering, C. D. Arnold, B. F. Arnold, A. E. Hubbard, H. N. Dentz, A. Lin, T. J. Meerkerk, et al. (2018). Effects of water quality, sanitation, handwashing, and nutritional interventions on child development in rural Kenya (WASH Benefits Kenya): A cluster-randomised controlled trial. The lancet child & adolescent health 2(4), 269–280. Trentacosta, C. J., P. Davis-Kean, C. Mitchell, L. Hyde, and D. Dolinoy (2016). Environmental contaminants and child development. Child Development Perspectives 10(4), 228–233. UNICEF et al. (2020). A new tool to measure sdg indicator 4.2. Early Childhood Development Index 2030. United Nations, U. (2015). Transforming our world: The 2030 agenda for sustainable development. Vaccari, M., T. Tudor, and G. Vinti (2019). Characteristics of leachate from landfills and dumpsites in Asia, Africa and Latin America: An overview. Waste Management 95, 416–431. Vargas-Baron, E., J. Small, D. Wertlieb, H. Hix-Small, R. G´omez Botero, K. Diehl, P. Vergara, and P. Lynch (2019). Global survey of inclusive early childhood development and early childhood intervention programs. RISE Institute: Washington, DC, USA. Walker, S. P., T. D. Wachs, S. Grantham-McGregor, M. M. Black, C. A. Nelson, S. L. Huffman, H. Baker-Henningham, S. M. Chang, J. D. Hamadani, B. Lozoff, et al. (2011). Inequality in early childhood: Risk and protective factors for early child development. The Lancet 378(9799), 1325–1338. Wang, F., Z. Cheng, A. Reisner, and Y. Liu (2018). Compliance with household solid waste management in rural villages in developing countries. Journal of Cleaner Production 202, 293– 298. WaterAid (2016). The missing ingredients: Are policy-makers doing enough on water, sanitation and hygiene to end malnutrition? Field Exchange 53, 56. WHO, W. (2009). WHO child growth standards: Growth velocity based on weight, length and head circumference: Methods and development. World Health Organization. WHO, W. (2019). Water, sanitation, hygiene and health: A primer for health professionals. Wolf, S. and D. C. McCoy (2019). Household socioeconomic status and parental investments: Direct and indirect relations with school readiness in ghana. Child Development 90(1), 260–278. Yamamoto, M. and T. C. Kinnaman (2022). Is incineration repressing recycling? Journal of Environmental Economics and Management 111, 102593. 33 Appendix Figure A1: Predicted health outcomes by wealth quintile and type of garbage collection PoorestPoorerMiddleRicherRichest Non-collected Collected Non-collected Collected Non-collected Collected Non-collected Collected Non-collected Collected 0.10 0.15 0.20 0.25 0.30 0.02 0.04 0.06 0.08 Stunted (dummy) Underweight (dummy) Proportion Notes: The coefficient plot visually displays the predictive margins for two distinct health outcomes in children: being stunted, and being underweight. The results are categorized based on the household’s wealth quintile. The figure then presents estimations according to the children’s access to garbage collection. The predictive margins are calculated from the linear probability models in Appendix Table A14. Variable definitions are available in Appendix Table A2 and summary statistics in Appendix Table A5. Confidence intervals are represented by capped lines and set at 90%. 34 Figure A2: Predicted stunting and underweight with different household infrastructure access by type of garbage collection Off-gridOn-grid Non-collected Collected Non-collected Collected .22 .24 .26 .28 .3 .32 Electricity PipedNatureOther Non-collected Collected Non-collected Collected Non-collected Collected .2 .22 .24 .26 .28 Water OutsideInside Non-collected Collected Non-collected Collected .15 .2 .25 .3 Sewage Stunting Off-gridOn-grid Non-collected Collected Non-collected Collected .02 .04 .06 .08 .1 PipedNatureOther Non-collected Collected Non-collected Collected Non-collected Collected .02 .04 .06 .08 .1 OutsideInside Non-collected Collected Non-collected Collected .02 .04 .06 .08 .1 Underweight Proportion Notes: The coefficient plot visually displays the predictive margins for two distinct health outcomes in children: being stunted, and being underweight. The results are categorized based on the household’s wealth quintile. The figure then presents estimations according to the children’s access to garbage collection. The predictive margins are calculated from the linear probability models in Appendix Table A16. Variable definitions are available in Appendix Table A2 and summary statistics in Appendix Table A5. Confidence intervals are represented by capped lines and set at 90%. 35 Table A1: Sample size by dataset and outcome MICS DHS ECDI Health Health Country Year N Country Year N Country Year N Belize 2006 796 Belize 2011 785 Bolivia 2003 9,165 2011 1,945 2015 1,090 Colombia 2005 12,480 2015 2,537 Costa Rica 2011 906 2010 24,560 Cuba 2019 5,252 Cuba 2019 2,316 Dominican Republic 2007 9,221 Dominican Republic 2019 8,396 Dominican Republic 2014 7,781 2013 3,116 El Salvador 2014 7,337 2019 3,404 Guyana 1998 3,982 Guyana 2006 2,462 El Salvador 2014 2,986 2005 1,684 2014 3,335 Guyana 2014 1,307 Nicaragua 1998 6,935 2019 2,782 2019 1,184 2001 6,008 Honduras 2019 8,464 Honduras 2019 3,616 Peru 1991 7,757 Mexico 2015 8,065 Jamaica 2011 666 2004 7,990 Paraguay 2016 4,614 Mexico 2015 3,352 2007 7,990 Saint Lucia 2012 291 Panama 2013 2,305 2009 9,199 Suriname 2006 2,248 Paraguay 2016 1,830 2011 8,510 2010 3,301 Saint Lucia 2012 122 2018 4,225 Suriname 2010 1,265 Trinidad and Tobago 2011 1,198 2018 1,772 Trinidad and Tobago 2011 525 Uruguay 2012 752 Notes: This table shows the sample sizes for each country, grouped by outcome and dataset. The dataset used in the analysis are the Multiple Indicators Cluster Surveys (MICS) and the Demographic Health Surveys (DHS). 36 Table A4: MICS statistics for health analysis Variable Type N µ σ Min Max 2014 67,248 0.094 0.292 0.000 1.000 2015 67,248 0.130 0.336 0.000 1.000 2016 67,248 0.097 0.296 0.000 1.000 2017 67,248 0.082 0.275 0.000 1.000 2018 67,248 0.079 0.270 0.000 1.000 2019 67,248 0.041 0.199 0.000 1.000 Electricity dummy 67,248 0.903 0.296 0.000 1.000 Female dummy 67,248 0.492 0.500 0.000 1.000 Hands cleanser available dummy 59,142 0.931 0.253 0.000 1.000 Household members number 67,248 5.442 2.348 2.000 32.000 Household members ≤5 number 67,248 1.499 0.731 1.000 9.000 Mother’s education categorical Less than primary 67,248 0.047 0.211 0.000 1.000 Primary 67,248 0.286 0.452 0.000 1.000 Lower secondary 67,248 0.158 0.364 0.000 1.000 Upper secondary 67,248 0.071 0.256 0.000 1.000 Secondary 67,248 0.303 0.459 0.000 1.000 Tertiary/higher/university 67,248 0.130 0.337 0.000 1.000 Vocational/technical 67,248 0.000 0.013 0.000 1.000 Other or non-standard curriculum 67,248 0.002 0.042 0.000 1.000 Stimulation index categorical 36,760 4.240 1.934 0.000 6.000 Urban area dummy 67,248 0.506 0.500 0.000 1.000 Water treated for safe drinking dummy 67,171 0.308 0.462 0.000 1.000 Wealth index quintiles categorical Poorest 67,248 0.313 0.464 0.000 1.000 Second 67,248 0.215 0.411 0.000 1.000 Middle 67,248 0.186 0.389 0.000 1.000 Fourth 67,248 0.161 0.368 0.000 1.000 Richest 67,248 0.126 0.331 0.000 1.000 Notes: This table presents descriptive statistics for the variables from the UNICEF’s Multiple Indicators Cluster Survey (MICS). The analysis sample is composed from twelve surveys. In particular, Argentina 2011, Belize 2015, Costa Rica 2018, Dominican Republic 2014, El Salvador 2014, Guyana 2014, Haiti 2016, Jamaica 2011, Mexico 2015, Panama 2013, Paraguay 2016, and Uruguay 2012. 43 Table A5: DHS statistics Variable Type N µ σ Min Max Outcomes and waste management Stunting dummy 118,597 0.251 0.434 0.000 1.000 Underweight dummy 118,597 0.060 0.237 0.000 1.000 Garbage collection dummy 118,597 0.442 0.497 0.000 1.000 Controls Age in months number 118,597 29.770 17.216 0.000 59.000 Cohort categorical 1986 118,597 0.000 0.021 0.000 1.000 1987 118,597 0.012 0.110 0.000 1.000 1988 118,597 0.014 0.116 0.000 1.000 1989 118,597 0.013 0.112 0.000 1.000 1990 118,597 0.012 0.111 0.000 1.000 1991 118,597 0.014 0.116 0.000 1.000 1992 118,597 0.000 0.020 0.000 1.000 1993 118,597 0.010 0.100 0.000 1.000 1994 118,597 0.017 0.130 0.000 1.000 1995 118,597 0.018 0.134 0.000 1.000 1996 118,597 0.019 0.138 0.000 1.000 1997 118,597 0.029 0.167 0.000 1.000 1998 118,597 0.021 0.143 0.000 1.000 1999 118,597 0.028 0.166 0.000 1.000 2000 118,597 0.043 0.203 0.000 1.000 2001 118,597 0.043 0.204 0.000 1.000 2002 118,597 0.049 0.217 0.000 1.000 2003 118,597 0.068 0.252 0.000 1.000 2004 118,597 0.074 0.263 0.000 1.000 2005 118,597 0.088 0.283 0.000 1.000 2006 118,597 0.108 0.311 0.000 1.000 2007 118,597 0.100 0.300 0.000 1.000 2008 118,597 0.086 0.280 0.000 1.000 2009 118,597 0.072 0.258 0.000 1.000 2010 118,597 0.038 0.192 0.000 1.000 2011 118,597 0.013 0.111 0.000 1.000 2012 118,597 0.006 0.075 0.000 1.000 2013 118,597 0.003 0.058 0.000 1.000 Cough dummy 118,445 0.376 0.484 0.000 1.000 Diarrhea dummy 118,450 0.161 0.368 0.000 1.000 Electricity dummy 118,597 0.757 0.429 0.000 1.000 Female dummy 118,597 0.492 0.500 0.000 1.000 Fever dummy 110,652 0.256 0.436 0.000 1.000 Garbage container categorical Lidded 42,199 0.135 0.342 0.000 1.000 Lidless 42,199 0.359 0.480 0.000 1.000 44 Table A5: DHS statistics Variable Type N µ σ Min Max Bags 42,199 0.382 0.486 0.000 1.000 Other 42,199 0.123 0.328 0.000 1.000 Refrigerator 118,597 0.350 0.477 0.000 1.000 Household members number 118,597 5.962 2.531 2.000 30.000 Household members ≤5 number 118,597 1.732 0.859 0.000 10.000 Mother’s education quartiles categorical Low 118,597 0.280 0.449 0.000 1.000 Interquartile range 118,597 0.561 0.496 0.000 1.000 Top 118,597 0.159 0.366 0.000 1.000 Mother is head of household dummy 118,597 0.096 0.294 0.000 1.000 Toilet is inside dwelling dummy 118,597 0.357 0.479 0.000 1.000 Urban area dummy 118,597 0.552 0.497 0.000 1.000 Water access categorical Piped water source 118,597 0.616 0.486 0.000 1.000 Nature sources 118,597 0.252 0.434 0.000 1.000 Other 118,597 0.132 0.338 0.000 1.000 Notes: This table presents descriptive statistics for the variables from the Demographic and Health Surveys (DHS). The analysis sample is composed from fourteen surveys. In particular, Bolivia (2003), Colombia (2005, and 2010), Dominican Republic (2007, and 2013), Guyana (1998, and 2005), Nicaragua (1998, and 2001), and Peru (1991, 2004, 2007, 2009, and 2011). Variable definitions are available in Appendix Table A2. 45 Table A6: Access to infrastructure and the relationship with developmental and health outcomes: harmonizing the sample Dependent variable ECDI (dummy) Stunted (dummy) Underweight (dummy) (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) Water access (dummy) 0.041*** 0.042*** 0.030** 0.023 -0.040** -0.040** -0.028** -0.013 -0.001 -0.001 -0.002 -0.002 (0.013) (0.013) (0.014) (0.013) (0.013) (0.013) (0.011) (0.008) (0.005) (0.005) (0.005) (0.005) R2-squared 0.044 0.057 0.065 0.068 0.066 0.069 0.082 0.089 0.021 0.023 0.023 0.023 Mean of outcome 0.824 0.824 0.824 0.824 0.109 0.109 0.109 0.109 0.023 0.023 0.023 0.023 N 25599 25599 25599 25599 25599 25599 25599 25599 25599 25599 25599 25599 Surveys 14 14 14 14 14 14 14 14 14 14 14 14 Controls Country-region-year FE Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Children level No Yes Yes Yes No Yes Yes Yes No Yes Yes Yes Mother level No No Yes Yes No No Yes Yes No No Yes Yes Household characteristics No No No Yes No No No Yes No No No Yes (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) Sanitation access (dummy) 0.052*** 0.051*** 0.032*** 0.025*** -0.077*** -0.077*** -0.056*** -0.043*** -0.002 -0.002 -0.002 -0.002 (0.008) (0.008) (0.008) (0.008) (0.016) (0.016) (0.013) (0.010) (0.002) (0.002) (0.002) (0.002) R2-squared 0.044 0.058 0.065 0.068 0.069 0.072 0.083 0.091 0.021 0.023 0.023 0.023 Mean of outcome 0.824 0.824 0.824 0.824 0.109 0.109 0.109 0.109 0.023 0.023 0.023 0.023 N 25599 25599 25599 25599 25599 25599 25599 25599 25599 25599 25599 25599 Surveys 14 14 14 14 14 14 14 14 14 14 14 14 Controls Country-region-year FE Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Children level No Yes Yes Yes No Yes Yes Yes No Yes Yes Yes Mother level No No Yes Yes No No Yes Yes No No Yes Yes Household characteristics No No No Yes No No No Yes No No No Yes Notes: This table presents results from a linear probability model for the correlation between the early childhood development index (ECDI) and access to water and sanitation infrastructure. The analysis sample is composed from fourteen surveys included in the Multiple Indicator Cluster Surveys (MICS). The analysis controls for fixed effects at the country-region-year level. Additional controls at the child level include age (in months), birth cohort, and sex. Mother-level controls account for the mother’s education level. Household characteristics controlled for are urban residency status, access to electricity, total household size, and the number of household members who are under five years of age. Variable definitions are available in Appendix Table A2. Survey clustered standard errors are reported in parentheses (*** p<0.01, ** p<0.05, * p<0.1). Table A7: Access to infrastructure and the relationship with the developmental components of the ECDI Dependent variable ECDI learning (dummy) ECDI physical (dummy) ECDI socio-emotional (dummy) ECDI literacy/numeracy (dummy) (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) Water access (dummy) 0.076*** 0.078*** 0.054*** 0.039** 0.004 0.004 0.003 0.003 0.014* 0.015* 0.009 0.006 0.018** 0.019** 0.013* 0.008 (0.015) (0.015) (0.015) (0.015) (0.003) (0.003) (0.004) (0.003) (0.008) (0.008) (0.009) (0.008) (0.007) (0.007) (0.007) (0.006) R2-squared 0.159 0.197 0.215 0.218 0.008 0.009 0.010 0.011 0.042 0.049 0.052 0.055 0.023 0.026 0.033 0.035 Mean of outcome 0.275 0.275 0.275 0.275 0.985 0.985 0.985 0.985 0.815 0.815 0.815 0.815 0.970 0.970 0.970 0.970 N 37964 37964 37964 37964 37964 37964 37964 37964 37964 37964 37964 37964 37964 37964 37964 37964 Surveys 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 Controls Country-region-year FE Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Children level No Yes Yes Yes No Yes Yes Yes No Yes Yes Yes No Yes Yes Yes Mother level No No Yes Yes No No Yes Yes No No Yes Yes No No Yes Yes Household characteristics No No No Yes No No No Yes No No No Yes No No No Yes (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27) (28) (29) (30) (31) (32) Sanitation access (dummy) 0.090*** 0.088*** 0.050*** 0.035*** 0.004 0.004 0.003 0.002 0.020** 0.019** 0.010 0.006 0.034*** 0.034*** 0.025*** 0.020** (0.009) (0.009) (0.006) (0.007) (0.003) (0.003) (0.003) (0.003) (0.009) (0.009) (0.009) (0.009) (0.007) (0.007) (0.007) (0.007) R2-squared 0.159 0.197 0.215 0.218 0.008 0.009 0.010 0.011 0.042 0.049 0.052 0.055 0.025 0.028 0.033 0.036 Mean of outcome 0.275 0.275 0.275 0.275 0.985 0.985 0.985 0.985 0.815 0.815 0.815 0.815 0.970 0.970 0.970 0.970 N 37964 37964 37964 37964 37964 37964 37964 37964 37964 37964 37964 37964 37964 37964 37964 37964 Countries 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 Controls Country-region-year FE Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Children level No Yes Yes Yes No Yes Yes Yes No Yes Yes Yes No Yes Yes Yes Mother level No No Yes Yes No No Yes Yes No No Yes Yes No No Yes Yes Household characteristics No No No Yes No No No Yes No No No Yes No No No Yes Notes: This table presents results from a linear probability model for the correlation between the sub-components of the early childhood development index (ECDI) and access to water and sanitation infrastructure. The analysis sample is composed from 19 surveys included in The Multiple Indicator Cluster Surveys (MICS). The countries that have been included in this analysis and their respective sample sizes can be found in Appendix Table A1. The analysis controls for fixed effects at the country-region-year level. Additional controls at the child level include age (in months), birth cohort, and sex. Mother-level controls account for the mother’s education level. Household characteristics controlled for are urban residency status, access to electricity, total household size, and the number of household members who are under five years of age. Variable definitions are available in Appendix Table A2 and summary statistics in Appendix Table A3. Survey clustered standard errors are reported in parentheses (*** p<0.01, ** p<0.05, * p<0.1). Table A8: Access to infrastructure and the relationship with developmental and health outcomes: wealth and stimulation Dependent variable ECDI Stunted Underweight ECDI Stunted Underweight (dummy) (dummy) (dummy) (dummy) (dummy) (dummy) (1) (2) (3) (4) (5) (6) Water access (dummy) 0.017 -0.009 -0.001 0.011 -0.002 -0.003 (0.010) (0.007) (0.003) (0.010) (0.008) (0.003) R2-squared 0.067 0.090 0.022 0.065 0.072 0.018 Mean of outcome 0.834 0.117 0.026 0.834 0.116 0.027 N 37951 36760 36760 37964 67248 67248 Countries 19 17 17 19 17 17 Controls Country-region-year FE Yes Yes Yes Yes Yes Yes Children level Yes Yes Yes Yes Yes Yes Mother level Yes Yes Yes Yes Yes Yes Household characteristics Yes Yes Yes Yes Yes Yes Stimulation index Yes Yes Yes No No No Wealth quintiles No No No Yes Yes Yes (7) (8) (9) (10) (11) (12) Sanitation access (dummy) 0.024*** -0.051*** -0.004 0.018* -0.032*** -0.003 (0.007) (0.013) (0.003) (0.009) (0.007) (0.002) R2-squared 0.067 0.092 0.022 0.065 0.073 0.018 Mean of outcome 0.834 0.117 0.026 0.834 0.116 0.027 N 37951 36760 36760 37964 67248 67248 Surveys 19 17 17 19 17 17 Controls Country-region-year FE Yes Yes Yes Yes Yes Yes Children level Yes Yes Yes Yes Yes Yes Mother level Yes Yes Yes Yes Yes Yes Household characteristics Yes Yes Yes Yes Yes Yes Stimulation index Yes Yes Yes No No No Wealth quintiles No No No Yes Yes Yes Notes: This table presents results from a linear probability model for the correlation between child development and health indicators and access to water and sanitation infrastructure including wealth and stimulation factors as controls. The analysis sample is composed from multiple surveys included in the Multiple Indicator Cluster Surveys (MICS). The countries that have been included in this analysis and their respective sample sizes can be found in Appendix Table A1. The analysis controls for fixed effects at the country-region-year level. Additional controls at the child level include age (in months), birth cohort, and sex. Mother-level controls account for the mother’s education level. Household characteristics controlled for are urban residency status, access to electricity, total household size, and the number of household members who are under five years of age. Variable definitions are available in Appendix Table A2 and summary statistics in Appendix Tables A3 and A4. Survey clustered standard errors are reported in parentheses (*** p<0.01, ** p<0.05, * p<0.1). 48 Table A9: Access to infrastructure and the relationship with developmental and health outcomes: interactions between water and sanitation ECDI Stunted Underweight (dummy) (dummy) (dummy) (1) (2) (3) Water access (dummy) 0.040 -0.010 -0.009* (0.026) (0.016) (0.005) Sanitation access (dummy) 0.050* -0.048*** -0.008** (0.025) (0.011) (0.003) Water access ×Sanitation access -0.029 0.000 0.009 (0.029) (0.016) (0.005) R2-squared 0.063 0.069 0.018 Mean of outcome 0.834 0.116 0.027 N 37964 67248 67248 Countries 19 17 17 Controls Country-region-year FE Yes Yes Yes Children level Yes Yes Yes Mother level Yes Yes Yes Household characteristics Yes Yes Yes Notes: This table presents results from a linear probability model for the correlation between child development and health indicators and access to water and sanitation infrastructure including interactions between water and sanitation. The analysis sample is composed from multiple surveys included in the Multiple Indicator Cluster Surveys (MICS). The countries that have been included in this analysis and their respective sample sizes can be found in Appendix Table A1. The analysis controls for fixed effects at the country-region- year level. Additional controls at the child level include age (in months), birth cohort, and sex. Mother-level controls account for the mother’s education level. Household characteristics controlled for are urban residency status, access to electricity, total household size, and the number of household members who are under five years of age. Variable definitions are available in Appendix Table A2 and summary statistics in Appendix Tables A3 and A4. Survey clustered standard errors are reported in parentheses (*** p<0.01, ** p<0.05, * p<0.1). 49 Table A10: Access to infrastructure and the relationship with developmental and health outcomes: mitigation behaviors Dependent variable ECDI Stunted Underweight ECDI Stunted Underweight (dummy) (dummy) (dummy) (dummy) (dummy) (dummy) (1) (2) (3) (4) (5) (6) Water access (dummy) 0.055* -0.020 -0.008 -0.008 -0.053 -0.020 (0.028) (0.019) (0.005) (0.067) (0.033) (0.015) Sanitation access (dummy) 0.071*** -0.055*** -0.008* -0.042 -0.053 -0.015 (0.024) (0.015) (0.004) (0.057) (0.039) (0.016) Water access ×Sanitation access -0.042 0.007 0.008 0.079 0.017 0.010 (0.029) (0.018) (0.005) (0.072) (0.043) (0.019) Treats water (dummy) 0.077* -0.029 0.003 (0.038) (0.017) (0.010) Water access ×Treats water -0.038 0.040 -0.004 (0.046) (0.026) (0.013) Sanitation access ×Treats water -0.066* 0.030 -0.001 (0.036) (0.019) (0.010) Water access ×Sanitation access ×Treats water 0.040 -0.030 0.003 (0.049) (0.025) (0.013) Cleanser to wash hands (dummy) 0.011 -0.049 -0.018 (0.035) (0.029) (0.018) Water access ×Cleanser to wash hands 0.042 0.051* 0.015 (0.048) (0.024) (0.020) Sanitation access ×Cleanser to wash hands 0.082* 0.017 0.011 (0.045) (0.041) (0.020) Water access ×Sanitation access ×Cleanser to wash hands -0.103* -0.026 -0.004 (0.057) (0.045) (0.023) R2-squared 0.064 0.069 0.017 0.065 0.067 0.017 Mean of outcome 0.834 0.116 0.027 0.834 0.114 0.026 N 37907 67171 67171 31466 59141 59141 Countries 19 17 17 17 14 14 Controls Country-region-year FE Yes Yes Yes Yes Yes Yes Children level Yes Yes Yes Yes Yes Yes Mother level Yes Yes Yes Yes Yes Yes Household characteristics Yes Yes Yes Yes Yes Yes Notes: This table presents results from a linear probability model for the correlation between child development and health indicators and access to water and sanitation infrastructure including interactions with mitigation behaviors such as treating water for safe consumption and having hands cleanser available. The analysis sample is composed from multiple surveys included in the Multiple Indicator Cluster Surveys (MICS). The countries that have been included in this analysis and their respective sample sizes can be found in Appendix Table A1. The analysis controls for fixed effects at the country-region-year level. Additional controls at the child level include age (in months), birth cohort, and sex. Mother-level controls account for the mother’s education level. Household characteristics controlled for are urban residency status, access to electricity, total household size, and the number of household members who are under five years of age. Variable definitions are available in Appendix Table A2 and summary statistics in Appendix Tables A3 and A4. Survey clustered standard errors are reported in parentheses (*** p<0.01, ** p<0.05, * p<0.1). Table A11: Access to infrastructure and the relationship with developmental and health outcomes: decomposing water access Dependent variable ECDI Stunted Underweight (dummy) (dummy) (dummy) (1) (2) (3) Piped water sources (base category) Tubewell or borehole -0.051** 0.010 0.000 (0.021) (0.010) (0.005) Protected well/spring -0.011 0.014 -0.002 (0.007) (0.009) (0.002) Unprotected sources -0.016 0.000 0.002 (0.012) (0.010) (0.003) Transported water sources -0.021 -0.001 0.012** (0.013) (0.017) (0.004) Packaged water 0.009** -0.036*** -0.001 (0.004) (0.009) (0.003) Surface water -0.053* 0.030** 0.003 (0.029) (0.012) (0.005) R2-squared 0.064 0.069 0.017 Mean of outcome 0.834 0.116 0.027 N 37559 66279 66279 Countries 19 17 17 Controls Country-region-year FE Yes Yes Yes Children level Yes Yes Yes Mother level Yes Yes Yes Household characteristics Yes Yes Yes Notes: This table presents results from a linear probability model for the correlation between child development and health indicators and access to water when decomposing the main water access binary indicator in sub-categories. The analysis sample is composed from multiple surveys included in the Multiple Indicator Cluster Surveys (MICS). The countries that have been included in this analysis and their respective sample sizes can be found in Appendix Table A1. The analysis controls for fixed effects at the country-region-year level. Additional controls at the child level include age (in months), birth cohort, and sex. Mother-level controls account for the mother’s education level. Household characteristics controlled for are urban residency status, access to electricity, total household size, and the number of household members who are under five years of age. Variable definitions are available in Appendix Table A2 and summary statistics in Appendix Tables A3 and A4. Survey clustered standard errors are reported in parentheses (*** p<0.01, ** p<0.05, * p<0.1). 51 Table A12: Access to infrastructure and the relationship with developmental and health outcomes: decomposing sanitation access Dependent variable ECDI Stunted Underweight (dummy) (dummy) (dummy) (1) (2) (3) Flush systems (base category) Hygienic non-flush -0.025*** 0.026*** -0.002 (0.007) (0.007) (0.002) Unsanitary systems -0.039*** 0.048*** -0.001 (0.011) (0.007) (0.004) No facilities -0.043*** 0.076*** 0.002 (0.013) (0.013) (0.003) R2-squared 0.063 0.070 0.018 Mean of outcome 0.834 0.116 0.027 N 37795 66928 66928 Countries 19 17 17 Controls Country-region-year FE Yes Yes Yes Children level Yes Yes Yes Mother level Yes Yes Yes Household characteristics Yes Yes Yes Notes: This table presents results from a linear probability model for the correlation between child development and health indicators and access to sanitation when decomposing the main sanitation access binary indicator in sub-categories. The analysis sample is composed from multiple surveys included in the Multiple Indicator Cluster Surveys (MICS). The countries that have been included in this analysis and their respective sample sizes can be found in Appendix Table A1. The analysis controls for fixed effects at the country-region-year level. Additional controls at the child level include age (in months), birth cohort, and sex. Mother-level controls account for the mother’s education level. Household characteristics controlled for are urban residency status, access to electricity, total household size, and the number of household members who are under five years of age. Variable definitions are available in Appendix Table A2 and summary statistics in Appendix Tables A3 and A4. Survey clustered standard errors are reported in parentheses (*** p<0.01, ** p<0.05, * p<0.1). 52 Table A19: Access to waste management infrastructure and the relationship with health outcomes: lidded containers as a mitigation factor Dependent variable Stunted Underweight (dummy) (dummy) (1) (2) Garbage collection (dummy) -0.003 -0.003 (0.008) (0.006) Garbage collection ×Lidless (vs. lidded) -0.040 0.008 (0.021) (0.007) Garbage collection ×Bags (vs. lidded) -0.016 0.003 (0.010) (0.007) Garbage collection ×Other (vs. lidded) -0.036* -0.019 (0.015) (0.011) R2-squared 0.170 0.038 N 42199 42199 Countries 5 5 Dep. Var. Mean 0.269 0.048 Controls Country-region-year FE Yes Yes Children level Yes Yes Mother level Yes Yes Household characteristics Yes Yes Notes: This table presents results from a linear probability model for the correlation between child health indicators and access to waste management infrastructure including interactions with garbage container types prior to collection. The analysis sample is composed from surveys included in the Demographic and Health Surveys (DHS). The analysis controls for fixed effects at the country-region-year level. Additional controls at the child level include age (in months), birth cohort, and sex. Mother-level controls take into account the mother’s education level and whether she is the head of the household. Household characteristics include urban residency status, access to electricity, water, sewage, total household size, and number of members under five years old. Variable definitions are available in Appendix Table A2 and summary statistics in Appendix Table A5. Survey clustered standard error are reported in parentheses (*** p<0.01, ** p<0.05, * p<0.1). 59 Table A20: Access to waste management infrastructure and the relationship with health outcomes: health channels Dependent variable Stunted Underweight (dummy) (dummy) (1) (2) Garbage collection (dummy) -0.024*** 0.003* (0.004) (0.002) Garbage collection ×Diarrhea (dummy) 0.011* -0.010* (0.006) (0.005) Garbage collection ×Cough (dummy) -0.004 -0.008** (0.007) (0.003) Garbage collection ×Fever (dummy) -0.007 -0.007** (0.005) (0.003) R2-squared 0.172 0.049 N 110647 111124 Countries 13 13 Dep. Var. Mean 0.241 0.057 Controls Country-region-year FE Yes Yes Children level Yes Yes Mother level Yes Yes Household characteristics Yes Yes Notes: This table presents results from a linear probability model for the correlation between child health indicators and access to waste management infrastructure including interactions with recent child illness episodes such as diarrhea, cough, and fever. The analysis sample is composed from surveys included in the Demographic and Health Surveys (DHS). The analysis controls for fixed effects at the country-region-year level. Additional controls at the child level include age (in months), birth cohort, and sex. Mother-level controls take into account the mother’s education level and whether she is the head of the household. Household characteristics include urban residency status, access to electricity, water, sewage, total household size, and number of members under five years old. Variable definitions are available in Appendix Table A2 and summary statistics in Appendix Table A5. Survey clustered standard errors are reported in parentheses (*** p<0.01, ** p<0.05, * p<0.1). 60