## Abstract

India hosts two-thirds of the global population defecating in the open and faces a worrying trend of districts declared ‘water scarce’. This paper aims to assess the costs and health benefits of sanitation interventions undertaken by the National Rural Drinking Water Security Pilot Project in India between 2012 and 2015. To achieve this goal, a pretest–posttest control group study was undertaken in two study areas located in Karnataka and Uttar Pradesh states. Full software and infrastructure costs were included as well as health endpoints, sourced from primary health centers. In Karnataka, latrine coverage of households increased from 16% to 59% in villages with high level of interventions, and from 7% to 18% in villages with lower levels of intervention. In Uttar Pradesh, coverage increased from 33% to 70% in high intervention villages and from 27% to 39% in the low intervention villages. We found health-related net benefits of USD 13 and USD 10 per person per year and benefit/cost ratios of 2.5 and 5.0 in Karnataka Uttar Pradesh, respectively. Given the positive economic returns on the intervention in culturally heterogeneous sites of southern and northern India, this intervention has potential for bringing significant benefits to the Indian population.

## INTRODUCTION

Worldwide, 2.3 billion people do not use a basic sanitation facility and of those, 892 million practice open defecation in 2015. India accounts for 25% of those lacking basic sanitation and 45% of those practicing open defecation (JMP 2017).

There is growing scientific literature linking poor water, sanitation and hygiene (WASH) to numerous diseases, such as diarrhea (Wolf et al. 2014), soil-transmitted helminth infection (Strunz et al. 2014), trachoma (Stocks et al. 2014), and schistosomiasis (Prüss-Üstün et al. 2008; Clasen et al. 2014) nutritional status (Victora et al. 2008; Dangour et al. 2013; Freeman et al. 2017), stunting (Checkley et al. 2008), environmental enteropathy (Humphrey 2009; Lin et al. 2013), and impaired cognitive development (Bain et al. 2014; Clasen et al. 2014; Sclar et al. 2017). In addition, the cost savings and economic impacts associated with improved WASH have been well established (Hutton 2013).

To improve the WASH situation and generate benefits for its population, in 2011 the Indian Ministry for Drinking Water and Sanitation initiated its ‘National Rural Drinking Water Security Pilot Project’ (NRDWSPP). It aimed to address the specific needs of the rural inhabitants through a combined set of interventions covering water resource management, agricultural productivity and WASH which supported them with knowledge, finances, infrastructure and management competencies.

A review of the health economics literature (Hutton & Chase 2016, 2017) revealed few cost–benefit analyses (CBA) have been undertaken in the field of water and sanitation. In general, economic studies have modeled the costs and benefits of a mix of WASH interventions and used generalized unit cost estimates representing state or national level (Hutton 2012; Cronin et al. 2014; Hutton et al. 2014; Weis et al. 2017). This pilot project, on the other hand, provided a unique opportunity to assess the costs and benefits of WASH interventions in specific locations and assess how successful these interventions were. A key question is whether the experiences of the pilot project have implications across India, especially in the light of the more recent sanitation programme of the Government of India – the Swachh Bharat Mission (GoI 2016), initiated in 2014.

The main objective of this study therefore is to investigate what are the costs and benefits of the WASH interventions, to understand their determinants, and assess implications for adopting a similar project design in other districts and states of India. This study presents the detailed results of a health analysis done as part of a more comprehensive cost–benefit study of the NRDWSPP (Weis et al. 2017).

## METHODS

### Study design

The study design follows the pretest-posttest control group of Campbell and Stanley (Campbell et al. 1963). In epidemiology, this form of study is also known as controlled before and after study (Cochrane, no date). Here the differentials of two points in time and two scenarios are measured: the situation at the end of the project (factual) is compared with the situation at the start (counterfactual) and high-intervention villages (study group) are compared to low-intervention villages (control group). While the comparison between the two points in time (factual–counterfactual) reveals changes over time, the comparison between the high- and low-intervention villages (study–control) cancels any changes that are not related to the project. Factors that are canceled out through this approach are for example maturation of the individuals, historic tendencies, pretesting and measuring instruments (Campbell et al. 1963).

### Location and groups

The two locations were selected to represent both different socio-economic contexts and different performance levels of pilots implemented under the differences NRDWSPP, and including both northern and southern states in India. In total, the NRDWP targeted around 2.23 million people located in 15 blocks covering ten states of India. For the present study, six villages were chosen within the Mulbagal Block in Kolar District, Karnataka State in southern India and six villages were chosen in Barauli Ahir Block in Agra District, Uttar Pradesh (UP) State in northern India. The six villages were separated into two groups of three villages each; one group consists of villages where the project intervened with high intensity (study group); the other group only received a limited number of interventions (control group). The control group comprised low-intervention villages instead of no-intervention villages for several reasons. First, the collaborating Non-Governmental Organizations (NGOs) were active in all villages of the block and had established connections with farmers, health practitioners, local political leaders and other key persons. These connections were essential for data collection. Second, the study follows the principle of overestimating costs and underestimating benefits in order to avoid statistical type one errors. Type one error occurs when assuming a hypothesis as true when it is false in reality. Type two error occurs when a hypothesis is labeled false when it is actually true (World Bank 2008). This approach ensures results produced are conservative of the impact and thus adds credibility when making recommendations to policy makers. Third, it was known that villages would implement national programmes (e.g. Swacch Bharat Mission, Mahatma Gandhi National Rural Employment Guarantee Scheme and Integrated Watershed Management Program) and hence would not remain the same as the start year.

For selecting the locations, attention has been paid to comparability between the study and the control group in order to assure that observed changes can be ascribed to the intervention (Boardman et al. 2011). In Karnataka, the villages in Mulbagal Block pre-intervention arm were matched for: number of villages, number of households and inhabitants, access to medical services, social status and water availability (see Table S1 in Supplementary Information, available with the online version of this paper). In UP, Barauli Ahir Block was more diverse and hence urbanization level and proximity of villages were included to ascertain the similarity of villages under investigation (see Table S2 in Supplementary Information, available online). Minor differences regarding the population size were unavoidable. However, the presentation of results in per capita terms overcomes this problem. The study period was three years from January 2012 to January 2015.

### Interventions

The sanitation part of the intervention adopted the community-led total sanitation (CLTS) (World Bank WSP 2014). It followed a grassroots approach with decisions made at community level to ensure the specific needs of each village were addressed. To minimize the intervention cost while assuring development outcomes, the sanitation technology adopted by most households was the single pit poor-flush latrine, based on community demand and an intention to become open defecation free (ODF). Community mobilizers and regular follow-up of the project team helped ensure actual use of toilets built. Further community mobilization was created through general knowledge-sharing events and through group visits to other locations that successfully implemented new structures and achieved a high level of ODF (AFPRO 2013; MDWS 2013; World Bank WSP 2014). In the analysis, actual usage was assumed if toilets were not used for other purposes (e.g. no evidence of being used for storage), if water and/or soap were available and/or if other signs of usage existed.

The project also aimed to strengthen the management competencies and local accountability through educating communities on how to operate and maintain existing and new infrastructure systems and by encouraging existing Village Water Security Committees (VWSC) to take a leading role in management. Where no traditional VWSC existed, the project helped set up new local management structures. A further structural intervention of the pilot project was the coordination of agencies and funding schemes in order to create sufficient financial support, and avoid duplication of effort and financial support (World Bank WSP 2015a, 2015b).

### Health data

Based on the findings of the Economics of Sanitation Initiative (Prüss-Ustün et al. 2014; World Bank WSP 2015c), the analysis included twelve water-related morbidity endpoints (diarrhea, helminthes, malnutrition, hepatitis A & E, malaria, chikungunya, dengue, acute lower respiratory infections (ALRI), measles, scabies, trachoma and other related diseases) and six endpoints potentially leading to mortality (diarrhea, helminthes, malnutrition, hepatitis A & E, malaria and other related diseases). Note that chikungunya and dengue were included due to local morbidity patterns in the southern study sites and interventions to improved water storage and environmental management to reduce mosquito breeding. Furthermore, the pathway of effect between poor WASH and measles, respiratory infection and malaria is through malnutrition and the resulting compromise of immunity (chances of contracting disease) and the ability to recover from these diseases (chances of case fatality), especially for children under five years of age (Fishman et al. 2004).

In Mulbagal Block the morbidity data were gained through a survey of the Public Health Centers that treated the inhabitants of the study and control villages. Comparing the factual situation of 2015 of the study group to its counterfactual situation in 2012, the dataset showed an improvement in four categories (diarrhea, malnutrition, malaria and dengue) and a worsening in three (helminthes, ALRI and scabies). No data were available for three endpoints (hepatitis, chikungunya and trachoma). The control group in comparison showed a worsening in all categories (Table 1).

Table 1

Morbidity cases collected from public health centers of Eragamuthanahalli, Pethandlahalli and Perumakanahalli village (control group) and Oorkunte, Kothur and Mittur village (study group), authors' investigation

Control group
Study group
DiseaseCounter-factualFactualCounter-factualFactual
Diarrhea 155 175 26 24
Helminthes 245 291 40 60
Malnutrition 29 38
Hepatitis A & E 10 24
Malaria
Chikungunya 30 32
Dengue
ALRI 85 105 20 22
Measles 41 50
Scabies 80 90 10 12
Trachoma
Other
Control group
Study group
DiseaseCounter-factualFactualCounter-factualFactual
Diarrhea 155 175 26 24
Helminthes 245 291 40 60
Malnutrition 29 38
Hepatitis A & E 10 24
Malaria
Chikungunya 30 32
Dengue
ALRI 85 105 20 22
Measles 41 50
Scabies 80 90 10 12
Trachoma
Other

In Barauli Ahir, data were only available on the block level. Here, a direct comparison between high and low intervention sites was not possible. A detailed explanation on how the health data has been processed further can be found in the calculation section below.

Impact of the intervention on mortality data could not be ascertained through field surveys, as the study population was too small to deliver reliable results. Instead, we used mortality data available for India (World Bank WSP 2011; Prüss-Ustün et al. 2014) and processed it as described below. Regarding the transferability of this data, it could be argued that many studies only investigated the effect of water and sanitation on children below five years of age. However, a meta-analysis of Wolf et al. (2014) on health impacts of drinking water and sanitation has found great similarities with other age groups. This study therefore follows the conclusion that the estimates derived from children under five can be used for all age groups.

## ANALYSIS

### Monetizing health effects

For monetizing the health effects we aimed to include all costs and benefits that are directly linked to the interventions of the project (primary effects). Those that only showed indirect links (secondary effects) are unfit for cost–benefit analysis (Boardman et al. 2011) and have therefore been excluded.

In general, costs and benefits can occur for the government or for individuals and can be financial (market value exists) or intangible (no market value exists). Costs included were hardware costs for toilet construction and software costs spent for staff, exposure visits and diverse training, equipment to measure the project's progress (such as geo information systems and field surveys) and administrative costs. Due to the short time horizon of the project, some costs that occur after the three-year period such as emptying latrines and replacement of infrastructure were not included. To measure relevant costs, this analysis builds on official Government estimates (MDWS 2015) as well as field observations.

The estimation of benefits required a more elaborated approach, given many of them were intangible in nature (i.e. without direct market values) or were not available from existing data sources. To more fully reflect the societal perspective, this study attempted to capture a broad range of health and non-health impacts of water and sanitation interventions (Hutton 2000). All the benefits except the productive value of higher water quality were measured through the total economic value (TEV) approach (Figure 1). The TEV approach differentiates between who bears the cost (overall society versus the project ‘beneficiary’ receiving improved water and sanitation services) and the type of benefit (administrative savings, medical care savings, economic savings in production and consumption, and intangible costs for individuals) (WHO 2008) (Figure 1).

Figure 1

Economic valuation of total health costs by COI and WTP (WHO 2008).

Figure 1

Economic valuation of total health costs by COI and WTP (WHO 2008).

The measurement of the ‘costs borne by the society’ is possible through finding market prices for the different health services (financial costs). This approach is called cost of illness (COI). The ‘costs borne by the victim’ are more complex to measure as no market prices exist (intangible costs). Among the available approaches, the human capital approach (HCA) proved to be unsuitable for this analysis as it focuses mainly on financial losses rather than the benefits to individuals which include intangible aspects. The remaining two methods both build on the willingness to pay method (WTP), where people's preferences are measured. The value of a statistical life (VSL) method attaches a monetary value to premature death through observations of risk in real markets (Pearce & Howarth 2000). Attention was paid to avoid double-counting of benefits due to the different valuation methods capturing (part of) the same benefit, such as between the cost of illness and the value of statistical life or between costs paid by patients and costs paid by government (marked grey in Figure 1).

### Calculation of costs and benefits

Building on above monetarization, we follow the method described by Cronin et al. (2014) and calculated differentials of: (1) the two points in time (2015 and 2012) and (2) two scenarios (high intervention villages (study group) and low intervention villages (control group)). The net benefits of these two differentials are shown by the upper grey triangle in Figure 2 and are calculated through the following equation:
(1)
where f stands for the factual, in 2015 existing situation, and c for the counterfactual situation, existent in 2012. The indicates that two situations have been compared, namely the high intervention group (study group) and the low intervention group (control group) (Cronin et al. 2014). Such method follows the pretest-posttest control group design of Campbell and Stanley (Campbell et al. 1963). In contrast to studies with only one group and a pre- and post-test, this study design introduces a control group in order to cancel out effects which might arise from general changes such as implementation of national programmes, underlying historic tendencies, and year-to-year climatic variations (Campbell et al. 1963).
Figure 2

Concept of the social and economic cost-benefit analysis of sanitation, source: authors.

Figure 2

Concept of the social and economic cost-benefit analysis of sanitation, source: authors.

For calculating the morbidity cases of the factual situation f, it is important to bear in mind that not all observed morbidity cases can be attributed to water and sanitation issues. Following the investigation of the Economics of Sanitation Initiative (ESI) (World Bank WSP 2015c), Table 2 shows the attributable fraction, i.e. the proportion of morbidity cases attributable to poor WASH. The health benefit of sanitation alone is assumed to be 28% of cases, taken from meta-analysis (Wolf et al. 2014).

Table 2

Diseases and attributable fractions, source: authors

DiseaseAttributable fraction
Diarrhea 88%
Helminthes 100%
Malnutrition 50%
Hepatitis A 88%
Hepatitis E 88%
Malaria 36%
Chikungunya 36%
Dengue 36%
ALRI 34%
Measles 29%
Scabies 50%
Trachoma 100%
Other 35%
DiseaseAttributable fraction
Diarrhea 88%
Helminthes 100%
Malnutrition 50%
Hepatitis A 88%
Hepatitis E 88%
Malaria 36%
Chikungunya 36%
Dengue 36%
ALRI 34%
Measles 29%
Scabies 50%
Trachoma 100%
Other 35%

The medical costs saved were based on the illness type, the expected number of outpatient visits and hospital admissions and the level of the health facility. In Mulbagal Block, it was observed that people first pass through the primary health centers before consulting health institutions of higher levels. Hence, the attributable number of cases was spread according to the health facility use and in a last step monetized by the average payment for the corresponding health facility.

Given the lack of mortality data at local level, risk of fatality from each disease was based on national numbers for India, including diarrheal deaths and deaths medicated through poor malnutrition (World Bank WSP 2011; Prüss-Ustün et al. 2014) (Table 3). The total mortality from poor sanitation was adjusted to each study area according to the population size, and the assumed reduction from implementing sanitation programmes was 28% (Wolff et al. 2014). The generalization of these mortality risks from India to the study areas was assumed correct given the similar coverage of sanitation to national rates. The value for the VSL calculation was taken from a national study of Cropper and colleagues (Cropper et al. 2013) which gave a VSL of USD 280,000 (2013 prices), and extrapolated to the study areas based on the average product of the two states to keep internal consistency in the VSL for this study, as it ensures an equal valuation of a life independent from the place of living (Pearce & Howarth 2000).

Table 3

Attributable deaths due to poor sanitation (India), numbers from Prüss-Ustün et al. (2014) and WSP (2011)

Attributable deaths due to poor sanitation (India)
Diarrhea 115,404
Malnutrition 37,830
ALRI 53,044
Malaria 2,217
Measles 9,036
Other 33,491
Attributable deaths due to poor sanitation (India)
Diarrhea 115,404
Malnutrition 37,830
ALRI 53,044
Malaria 2,217
Measles 9,036
Other 33,491

## RESULTS

The results demonstrate that villages receiving high levels of intervention were closer to achieving ODF, measured by the percentage of households having and using a private latrine. High intervention villages more than doubled their coverage from 33% to 70% in the case of Barauli Ahir and from 16% to 59% in Mulbagal Block (Table 4).

Table 4

Latrine coverage in limited and high intervention villages, source: authors

Limited intervention villages
High intervention villages
Latrine coverage in Mulbagal BlockEragamuthanahalliPethandlahalliPerumakanahalliTotal%OorkunteKothurMitturTotal%
2012/2013 15 27 7.42 15 15 25 55 15.58
2014/2015 20 27 20 67 18.41 65 60 85 210 59.49
Latrine coverage in Barauli Ahir BlockSaimariBadBisera KalaTotal%RajraiKakuaBahentaTotal%
2012/2013 181 159 26 366 26.68 80 254 54 388 32.99
2014/2015 236 225 69 530 38.63 252 378 196 826 70.24
Limited intervention villages
High intervention villages
Latrine coverage in Mulbagal BlockEragamuthanahalliPethandlahalliPerumakanahalliTotal%OorkunteKothurMitturTotal%
2012/2013 15 27 7.42 15 15 25 55 15.58
2014/2015 20 27 20 67 18.41 65 60 85 210 59.49
Latrine coverage in Barauli Ahir BlockSaimariBadBisera KalaTotal%RajraiKakuaBahentaTotal%
2012/2013 181 159 26 366 26.68 80 254 54 388 32.99
2014/2015 236 225 69 530 38.63 252 378 196 826 70.24

Applying the calculations described above and using the model assumptions on health impact of improved sanitation, this increased use of latrines reduces the burden of disease leading to economic benefits. In the study in Mulbagal, the reduced burden of disease is equal to USD 30 per person in high intervention villages, and only USD 7 in low intervention villages. In the case of Barauli Ahir, the health impacts of improved sanitation are valued at USD 15 in high intervention villages, and USD 2.5 in low intervention villages. The incremental benefit of high over low intervention villages is estimated at USD 22 in the Mulbagal case study and USD 12.5 in the Barauli Ahir Block case study (Table 5).

Table 5

Benefits (USD) per capita from reducing the burden of disease, source: authors

Mulbagal Block
Barauli Ahir Block
DiseaseHigh interventionLow interventionBenefitsHigh interventionLow interventionBenefits
Diarrhea 14.04 3.12 10.92 7.32 1.21 6.11
Helminthes 5.86 1.60 4.26 2.75 0.45 2.29
Malnutrition 3.59 0.77 2.82 1.91 0.32 1.59
Hepatitis A & E 0.26 0.07 0.19 0.14 0.02 0.12
Malaria 0.44 0.09 0.35 0.23 0.04 0.19
Chikungunya 0.00 0.02 −0.02 0.00 0.00 0.00
Dengue 0.01 0.00 0.01 0.00 0.00 0.00
ALRI 0.14 0.07 0.07 0.09 0.02 0.08
Measles 0.00 0.02 −0.02 0.00 0.00 0.00
Scabies 0.10 0.09 0.01 0.12 0.02 0.10
Other 4.48 0.93 3.55 2.41 0.40 2.01
Total 28.93 6.78 22.15 14.98 2.48 12.50
Mulbagal Block
Barauli Ahir Block
DiseaseHigh interventionLow interventionBenefitsHigh interventionLow interventionBenefits
Diarrhea 14.04 3.12 10.92 7.32 1.21 6.11
Helminthes 5.86 1.60 4.26 2.75 0.45 2.29
Malnutrition 3.59 0.77 2.82 1.91 0.32 1.59
Hepatitis A & E 0.26 0.07 0.19 0.14 0.02 0.12
Malaria 0.44 0.09 0.35 0.23 0.04 0.19
Chikungunya 0.00 0.02 −0.02 0.00 0.00 0.00
Dengue 0.01 0.00 0.01 0.00 0.00 0.00
ALRI 0.14 0.07 0.07 0.09 0.02 0.08
Measles 0.00 0.02 −0.02 0.00 0.00 0.00
Scabies 0.10 0.09 0.01 0.12 0.02 0.10
Other 4.48 0.93 3.55 2.41 0.40 2.01
Total 28.93 6.78 22.15 14.98 2.48 12.50

Looking at the project's cost, in the Mulbagal study, USD 9 was spent per resident and in the case of Barauli Ahir Block USD 3 per resident (Table 6). These costs include the construction of toilets, payments to the support organization, financing of exposure visits and training, as well as the acquisition of measuring equipment (MDWS 2015).

Table 6

Costs (USD) per capita for different investments, sources:MDWS (2015) and authors

MulbagalBarauli Ahir
InvestmentsCosts ($)Costs ($)
Toilet construction 5.49 2.15
Tap connections 3.34 0.00
Support organization 0.10 0.23
Exposure visits 0.05 0.10
Training 0.03 0.06
Measuring equipment 0.06 0.10
Total 9.08 2.65
MulbagalBarauli Ahir
InvestmentsCosts ($)Costs ($)
Toilet construction 5.49 2.15
Tap connections 3.34 0.00
Support organization 0.10 0.23
Exposure visits 0.05 0.10
Training 0.03 0.06
Measuring equipment 0.06 0.10
Total 9.08 2.65

A benefit/cost ratio of 2.5 is found in Mulbagal and of nearly 5 in Barauli Ahir. This means that each dollar invested in Mulbagal creates a value of 2.5 dollars in health savings, in the case of Barauli Ahir Block it even creates a value of 5 dollars in health savings. The net economic benefit, amounts to USD 13 per resident in Mulbagal and USD 10 per resident in Barauli Ahir (Table 7).

Table 7

Benefit/cost ratio and net benefit (USD, per capita), source: authors

MulbagalBarauli Ahir
Benefit/cost ratio 2.44 4.72
Net benefit ($) 13.07 9.85 MulbagalBarauli Ahir Benefit/cost ratio 2.44 4.72 Net benefit ($) 13.07 9.85

Although more net benefits are generated for an average person in Mulbagal, the relative impact of these benefits are very different for the two regions. A plausible reason for this is the high cost difference between the two blocks concerning tap connections. It was one of the NRDWSPPs targets to improve the domestic water supply to increase hand washing and reduce sanitation-related infections. Greater efforts were made in Mulbagal made to install a tap system for the households. Unfortunately, Mulbagal block was hit by a drought in the years of this pilot project and the groundwater table sank drastically. It was hence not possible to use the newly installed infrastructure, and benefits that could have arisen from better water supply to households did not materialize. However, the analysis controlled for the impact of the drought and estimated what the benefits would have been had the taps been used by households.

To test the validity of the above findings, a Monte Carlo sensitivity analysis was undertaken. It showed that roughly 99% of all values for net benefit lies between USD 3 and USD 24 in Mulbagal and between USD 5 and USD 15 in Barauli Ahir (Table 8). Only 0.13% of simulation results surpass this maximum value in Mulbagal, while less than 0.46% of simulation results are less than the minimum value. In Barauli Ahir, these results are 0.40% and 1.13%, respectively. In Barauli Ahir, the probability of receiving negative values for net benefit is zero while in Mulbagal the probability of a negative net benefit is 0.02%.

Table 8

Expected values from a Monte Carlo sensitivity analysis, source: authors

Mulbagal ($)Barauli Ahir ($)
Max 99% 23.91 15.37
Min 99% 3.42 5.12
Average 13.66 10.25
St. Dev. 3.42 1.71
Mulbagal ($)Barauli Ahir ($)
Max 99% 23.91 15.37
Min 99% 3.42 5.12
Average 13.66 10.25
St. Dev. 3.42 1.71

## CONCLUSION AND IMPLICATIONS

This study has shown the potential health benefits from the NRDWSPP, the economic benefits that can result and the high return on investments. It has been demonstrated that sanitary infrastructure and behavioral interventions led to net benefits per inhabitant of USD 13 in Mulbagal and USD 10 in Barauli Ahir as well as a benefit/cost ratio of 2.5 in Mulbagal and 5.0 in Barauli Ahir.

These results have furthermore been found to be robust to a comprehensive sensitivity analysis using Monte Carlo simulation. A broader analysis has previously shown that the overall societal benefits exceed costs by 10 to 11 times (Barauli Ahir Block vs. Mulbagal Block), when including agricultural, time-saving and environmental effects of water supply and sanitation interventions (Weis et al. 2017).

The implications of the high benefit/cost ratios of 2.5 and 5.0, as well as the high net benefits show that the actors involved have achieved benefits for a large part of the rural population in the pilot study sites. It furthermore indicates that the grassroots approach of the pilot project successful in addressing the water and sanitation problems that were present at the rural intervention sites in 2012. We therefore recommend that government and community actors continue to strive for universal sanitation access and usage in those sites.

Given the very different locations and cultural heterogeneity of the two study sites (one in southern and one in northern India), it suggests that the health impacts and returns on investment might be similar across different locations in India. As the pilot project targeted around 2.23 million people (World Bank WSP 2015a, 2015b) and the average net benefits of the two study sites lies between USD 10 and 13 per person, a simplified estimate would suggest that the pilot project created net social health benefits between USD 22 million and USD 29 million. However, for scaling up the pilot approach, a more detailed assessment is needed of the baseline variables in new sites and the potential to benefit.

In summary, the significant benefits associated with sanitation interventions suggest that the Indian Government could increase population welfare through scaling up the pilot project or investing in similar interventions such as the Swacch Bharat Mission.

## ACKNOWLEDGEMENTS

The authors are grateful to the World Bank Water and Sanitation team, especially Sonal Gaurishanker, Laura Inha and Mili Varughese for their valuable input. Furthermore, they recognize the essential support of the governmental officers and non-government organizations responsible for the NRDWSPP in the Blocks of Mulbagal (Kolar district, Karnataka state) and Barauli Ahir (Agra district, state of Uttar Pradesh).

## FUNDING SOURCES

This work was supported by the Water and Sanitation Program of World Bank India as part of the National Rural Drinking Water Security Pilot Project. The financial support included staff payments and travel expenses related to the pilot project. The authors affirm not having entered into an agreement with the funder that may have limited their ability to complete the research as planned, and had full control of all primary data.

## REFERENCES

REFERENCES
AFPRO
2013
National Drinking Water Security Pilot Project Annual Report: July 2012–June 2013
.
Action for Food Production
,
Mulbagal
,
India
.
Bain
R.
,
Cronk
R.
,
Hossain
R.
,
Bonjour
S.
,
Onda
K.
,
Wright
J.
,
Yang
H.
,
Slaymaker
T.
,
Hunter
P.
,
Prüss-Ustün
A.
&
Bartram
J.
2014
Global assessment of exposure to faecal contamination through drinking water based on a systematic review
.
Trop. Med Int Health
19
(
8
),
917
927
.
Boardman
A. E.
,
Greenberg
D. H.
,
Vining
A. R.
&
Weimer
D. L.
2011
Cost-benefit Analysis: Concepts and Practice (The Pearson Series in Economics)
, 4th edn.
Prentice Hall (Massachusetts)
,
Boston
,
USA
.
Campbell
D. T.
,
Stanley
J. C.
&
Gage
N. L.
1963
Experimental and Quasi-Experimental Designs for Research
.
Houghton Mifflin
,
Boston
,
USA
.
Checkley
W.
,
Buckley
G.
,
Gilman
R. H.
,
Assis
A. M. O.
,
Guerrant
R. L.
,
Morris
S. S.
,
Mølbak
K.
,
Valentiner-Branth
P.
,
Lanata
C. F.
&
Black
R. E.
, &
The Childhood Malnutrition and Infection Network
2008
Multi-country analysis of the effects of diarrhoea on childhood stunting
.
Int. J. Epidemiol.
37
,
816
830
.
Clasen
T.
,
Boisson
S.
,
Routray
P.
,
Torondel
B.
,
Bell
M.
,
Cumming
O.
,
Ensink
J.
,
Freeman
M.
,
Jenkins
M.
,
Odagiri
M.
,
Ray
S.
,
Sinha
A.
,
Suar
M.
&
Schmidt
W.-P.
2014
Effectiveness of a rural sanitation programme on diarrhoea, soil-transmitted helminth infection, and child malnutrition in Odisha, India: a cluster-randomised trial
.
The Lancet Global Health
2
,
645
653
.
Cochrane
no date
Non-randomised Controlled Study (NRS) Designs
.
Cochrane Childhood Cancer
,
Amsterdam
,
The Netherlands
. .
Cronin
A. A.
,
Ohikata
M.
&
Kumar
M.
2014
Social and economic cost-benefit analysis of sanitation in Odisha State, India
.
J. Water Sanit. Hyg. Dev.
9
,
521
.
Cropper
M.
,
Gamkhar
S.
,
Malik
K.
,
Limonov
A.
&
Partridge
I.
2013
The Health Effects of Coal Electricity Generation in India
.
Agricultural & Applied Economics Association
,
Washington
,
USA
. .
Dangour
A. D.
,
Watson
L.
,
Cumming
O.
,
Boisson
S.
,
Che
Y.
,
Velleman
Y.
,
Cavill
S.
,
Allen
E.
&
Uauy
R.
2013
Interventions to improve water quality and supply, sanitation and hygiene practices, and their effects on the nutritional status of children
.
Cochrane Database of Systematic Reviews
8
.
https://doi.org/10.1002/14651858.CD009382.pub2
.
Fishman
S. M.
,
Caulfield
L. E.
,
deOnis
M.
&
Blössner
M.
2004
Childhood and maternal underweight
. In:
Comparative Quantification of Health Risks
.
World Health Organization
,
Geneva
,
Switzerland
. .
Freeman
M. C.
,
Garn
J. V.
,
Sclar
G. D.
,
Boisson
S.
,
Medlicott
K.
,
Alexander
K. T.
,
Penakalapati
G.
,
Anderson
D.
,
Mahtani
A. G.
,
Grimes
J. E. T.
,
Rehfuess
E. A.
&
Clasen
T. F.
2017
The impact of sanitation on infectious disease and nutritional status: a systematic review and meta-analysis
.
Int. J. Hyg. Environ. Health
220
,
928
949
.
GoI
2016
Swachh Bharat Mission
.
Government of India
,
New Delhi
,
India
.
Available at: https://swachhbharat.mygov.in/ (Accessed 21 April 2016)
.
Humphrey
J.
2009
Child undernutrition, tropical enteropathy, toilets, and handwashing
.
The Lancet
3754
,
1032
1035
.
Hutton
G.
2000
Considerations in Evaluating the Cost-Effectiveness of Environmental Health Interventions
.
World Health Organization
,
Geneva
,
Switzerland
. .
Hutton
G.
2012
Global Costs and Benefits of Drinking-Water Supply and Sanitation Interventions to Reach the MDG Target and Universal Coverage
.
World Health Organization
,
Geneva
,
Switzerland
. .
Hutton
G.
&
Chase
C.
2016
The knowledge base for achieving the sustainable development goal targets on water supply, sanitation and hygiene
.
Int. J. Environ. Res. Public Health
13
.
doi: 10.3390/ijerph13060536
.
Hutton
G.
,
Chase
C.
2017
Water Supply, Sanitation, and Hygiene (Chapter 9)
. In:
Disease Control Priorities
, 3rd edn,
Vol. 7
(
Mock
C. N.
,
Nugent
R.
,
Kobusingye
O.
&
Smith
K. R.
, eds).
Injury Prevention and Environmental Health. World Bank
,
Washington
,
USA
, pp.
171
198
.
Hutton
G.
,
Rodriguez
U. P.
,
Winara
A.
,
Viet Anh
N.
,
Phyrum
K.
,
Chuan
L.
,
Blackett
I.
&
Witz
A.
2014
Economic efficiency of sanitation interventions in Southeast Asia
.
Journal of Water, Sanitation and Hygiene for Development
4
,
23
36
.
JMP
2017
Progress on Drinking Water, Sanitation and Hygiene
.
WHO
,
Geneva
,
Switzerland
& .
Lin
A.
,
Arnold
B. F.
,
Afreen
S.
,
Goto
R.
,
Huda
T. M. N.
,
Haque
R.
,
Raqib
R.
,
Unicomb
L.
,
Ahmed
T.
,
Colford
J. M.
Jr
&
Luby
S. P.
2013
Household environmental conditions are associated with enteropathy and impaired growth in rural Bangladesh
.
Am. J. Trop. Med. Hyg.
89
,
130
137
.
MDWS
2013
Toolkit for Preparation of Drinking Water Security Plan
.
Ministry of Drinking Water Supply and Sanitation
,
New Delhi
,
India
.
MDWS
2015
Water Security Planning and Implementation Pilot Programme – Budget Estimates for Two Years, Khajuraho, India
.
Ministry of Drinking Water and Sanitation
,
New Delhi
,
India
.
Pearce
D.
&
Howarth
A.
2000
Technical Report on Methodology: Cost-Benefit Analysis and Policy Responses
.
European Commission, DG Environment
,
Brussels
,
Belgium
. .
Prüss-Ustün
A.
,
Bartram
J.
,
Clasen
T.
,
Colford
J. M.
Jr
,
Cumming
O.
,
Curtis
V.
,
Bonjour
S.
,
Dangour
A. D.
,
De France
J.
,
Fewtrell
L.
,
Freeman
M. C.
,
Gordon
B.
,
Hunter
P. R.
,
Johnston
R. B.
,
Mathers
C.
,
Mäusezahl
D.
,
Medlicott
K.
,
Neira
M.
,
Stocks
M.
,
Wolf
J.
&
Cairncross
S.
2014
Burden of disease from inadequate water, sanitation and hygiene in low- and middle-income settings: a retrospective analysis of data from 145 countries
.
Trop. Med. Int. Health
19
,
894
905
.
Prüss-Üstün
A.
,
Bos
R.
,
Gore
F.
&
Bartram
J.
2008
Safer Water, Better Health: Costs, Benefits and Sustainability of Interventions to Protect and Promote Health
.
World Health Organization
,
Geneva
,
Switzerland
. .
Sclar
G. D.
,
Garn
J. V.
,
Penakalapati
G.
,
Alexander
K. T.
,
Krauss
J.
,
Freeman
M. C.
,
Boisson
S.
,
Medlicott
K. O.
&
Clasen
T.
2017
Effects of sanitation on cognitive development and school absence: a systematic review
.
Int. J. Hyg. Environ. Health
220
,
917
927
.
Stocks
M. E.
,
Ogden
S.
,
D.
,
D. G.
,
McGuire
C.
&
Freeman
M. C.
2014
Effect of water, sanitation, and hygiene on the prevention of trachoma: a systematic review and meta-analysis
.
PLoS Med.
11
.
doi: https://doi.org/10.1371/journal.pmed.1001605
.
Strunz
E. C.
,
D. G.
,
Stocks
M. E.
,
Ogden
S.
,
Utzinger
J.
&
Freeman
M. C.
2014
Water, sanitation, hygiene, and soil-transmitted helminth infection: a systematic review and meta-analysis
.
PLoS Med.
14
.
doi: https://doi.org/10.1371/journal.pmed.1001620
.
Victora
C.
,
L.
,
Fall
C.
,
Hallal
P. C.
,
Martorell
R.
,
Richter
L.
&
Sachdev
H. S.
, ,
Maternal and Child Undernutrition Study Group
2008
Maternal and child undernutrition: consequences for adult health and human capital
.
The Lancet
371
,
340
357
.
Weis
D.
,
Hutton
G.
&
Kumar
M.
2017
The national rural drinking water security pilot project in India: a cost–benefit analysis
.
Water Economics and Policy
3
,
1
31
.
WHO
2008
Economic Valuation of Transport-Related Health Effects
.
Review of Methods and Development of Practical Approaches, With a Special Focus on Children
.
World Health Organization, Regional Office for Europe
,
Copenhagen
,
Denmark
. .
Wolf
J.
,
Prüss-Ustün
A.
,
Cumming
O.
,
Bartram
J.
,
Bonjour
S.
,
Cairncross
S.
,
Clasen
T.
,
Colford
J. M.
Jr
,
Curtis
V.
,
De France
J.
,
Fewtrell
L.
,
Freeman
M. C.
,
Gordon
B.
,
Hunter
P. R.
,
Jeandron
A.
,
Johnston
R. B.
,
Mäusezahl
D.
,
Mathers
C.
,
Neira
M.
&
Higgins
J. P.
2014
Systematic review: assessing the impact of drinking water and sanitation on diarrhoeal disease in low- and middle-income settings: systematic review and meta-regression
.
Trop. Med. Int. Health
19
,
928
942
.
World Bank
2008
An Economic Framework for Analyzing Social Costs and Benefits of National Water Reuse Strategy – A Case Study of Morocco
.
The World Bank
,
Washington, DC
.
World Bank WSP
2011
Flagship Report Economic Impacts of Inadequate Sanitation in India
.
The World Bank, Water and Sanitation Program
,
New Delhi
,
India
. .
World Bank WSP
2014
Report on A 5-Day ToT Workshop on CLTS
.
The World Bank, Water and Sanitation Program
,
Nalgonda, Telangana
.
World Bank WSP
2015a
Report – Draft II, The World Bank
.
Water and Sanitation Program
,
New Delhi
,
India
.
World Bank WSP
2015b
Support to Drinking Water Security Pilot (Final Report No. P131954)
.
The World Bank, Water and Sanitation Program
,
New Delhi
,
India
.
World Bank WSP
2015c
Economics of Sanitation Initiative
.
The World Bank, Water and Sanitation Program
,
New Delhi
,
India
.