Water Insufficiency,Health Hazards and Rainwater Harvesting in North 24 Parganas,West Bengal,India: Results of a Socio-economic Survey

Abstract

Rural households suffer from various health hazards due to unsafe water. Rooftop rainwater harvesting (RRH) has been recommended by various experts as a safer alternative to contaminated ground and surface water. The rural households, however, for various reasons, may not be willing to adopt RRH. The present study was based on primary data collected from 923 rural households in 4 blocks of North 24 Parganas district, West Bengal, India to identify the socio-demographic factors influencing the occurrence of health hazards and willingness to adopt RRH (WRRH). The study was focused on how health hazards and WRRH were related to water insufficiency, water awareness, poverty level and other socio-demographic variables. RRH not only depends on its feasibility but also on the willingness of the household to install it. A set of indices, namely health hazard index (HHI), water insufficiency index (WII), water awareness action index (WAAI), willingness to adopt RRH index (WRRHI) and Poverty Level Index (PLI) were developed from the relevant indicators. A binary logistic regression of HHI and WRRH was carried out on these indices along with some other socio-demographic variables. Most of these indices were found to have a significant effect on HHI. WRRH, however, was not found to depend on these indices, rather on religious belief and awareness of the adverse effects of arsenic poisoning in the village. The study also highlights the significance of undertaking awareness programmes on the consequences of using unsafe water by the government and non-governmental organisations.

Keywords

Rooftop rainwater harvesting (RRH)health hazards socio-demographic factors binary logistic regression rural India

1. Introduction

Water allocation, its use, uncertainty and socio-economic complexity encompass the magnitude of water stress. According to the Census of India (2011), more than 68.8 per cent people live in rural areas. Millions of Indians are designated as ‘poor’ as per the national and international standards (Majumder, 2008). The economy is not progressing fast enough to maintain the standard of living to cope up with population growth in the developing countries. It is reported that India has a population affected by water stress, which is three times more than the combined population of 17 other extremely high water-stressed countries (Pandey, 2019). Rural India mostly depends on groundwater, as the major livelihood option for rural household is agriculture (Sarkar & Ray, 2019). However, the standard of living depends on various multidimensional aspects such as education, income level and housing status. (Banerjee, 2018). Income is not supposed to be considered as the single most important indicator, rather indicators such as health, education, housing, water and sanitation should also be considered with due importance as measures of poverty and well-being of society (Datt et al., 2016). As the withdrawal of groundwater would be 87 per cent by 2050, water would become a fugitive resource (UN habitat, 2015). Fast economic growth is very often associated with degradation of the land, soil, water and air resources (Zhou, 2016). Karfakis et al. (2012) argued that sometimes water exploitation happens due to many traditional, socio-economic factors. It is not only the water demand alone, rather the severity of the water crisis tends to aggravate climate change (Fenwick, 2010). The key issue of the impact of qualitative water stress in populated countries is a big challenge due to contamination by arsenic, iron, fluoride and other heavy metals (Roy, 2008). According to Mahimairaja et al. (2005), arsenic is an extremely toxic and carcinogenic metalloid, which is found in rocks, soil, water, sediments and air. More significantly, this toxicity of arsenic enters the human body through the food chain, that is, rice, vegetables, fish, fruits, milk, other dairy products, meat, etc. Through the open market, this danger spreads from local to regional and ultimately to a global scale (Bhattacharya, 2010).

Arsenicosis is a chronic illness developed from drinking water containing high levels of arsenic within a period from 5 to 20 years (Bhattacharya et al., 2019). In West Bengal, India’s first case of arsenicosis was reported in 1983 (Mukherjee et al., 2009a). The Central Ground Water Board (CGWB, 2019) reported that a total of 9 districts with 111 blocks and 3,417 villages are contaminated by arsenic in West Bengal. The Geological Survey of India confirmed that the severity of arsenic affected areas is found to be linear in the eastern part of River Bhagirathi, West Bengal (Elangovan & Chalakh, 2006). Arsenic toxicity may lead to severe health hazards; it may even lead us to death. Another consequence of consuming unsafe water is that it invites socio-economic and cultural complexity. West Bengal has an average rainfall of 1800–3000 mm occurring yearly (Mukhopadhyay et al., 2016). Much of the precious reliable rainwater drains out through surface run-off, even in the areas with contaminated groundwater. The present world is paying more interest on water harvesting for sustainability. In this article, the authors try to establish the relationship between the magnitudes of water insufficiency and health by using some indices. Further, the authors analyse the effects of a water crisis on rural households to adopt rooftop rainwater harvesting (RRH). This article aims to find out the scale of magnitude of water stress in terms of affected population and its mitigation by adopting potential rooftop RRH in the contaminated study area as well as in other parts of India. Thus, the objectives of the present article are as follows:

To find out the adverse effects of unsafe water in rural households;

to identify the health hazards in rural households and the variables influencing the health hazards; and

to find out the willingness of households for the installation of RRH gadgets.

II. Background of the Study

Numerous studies have so far been carried out on health risks associated with arsenic contamination in the Padma–Meghna–Brahmaputra (PMB) Basin (Bhattacharya et al., 2019; Bhowmick et al., 2018; Elangovan & Chalakh, 2006; Mahimairaja et al., 2005). These health implications include cardiovascular problems, cough, gastrointestinal problems, anaemia and leucopenia, hepatic effects, renal effects, neurological effects, dermal effects to carcinogenic effects, even foetal loss and premature delivery among women and many more. Several researches have focused on the economic aspect of filtration technology (CPCB, 2002; Singh et al., 2014; Udayanapillai & Kaliammal, 2016; Zakhar et al., 2018); many of them concentrated on site-specific recharge structures or tank size and its cost-benefit analysis (Anand et al., 2017; CGWB, 2013; Okoye et al., 2015; Martinson et al., 2002). Though RRH is reckoned to be a feasible alternative for decontamination of surface and groundwater, the current practices and performances of RRH in the contaminated areas of West Bengal have been largely unexplored (Özdemir et al., 2011). Moreover, none of them addressed the behavioural approach behind the non-adoption of RRH. This happened due to two reasons. First, presently, there is no such low-cost, well-planned mitigation strategy available for millions of arsenicosis affected rural people (Bhowmick et al., 2018). Numerous private vendors of water filtration systems maximise their business profit without even maintaining proper standards. Hence, the quality of water matters a lot among rural people. Second, none of them addresses the household of the direct and indirect effects of such water insufficiency. Arsenic threats are very much associated with malnutrition; thus, poor rural people eventually must pay more. Infants, children and women are more vulnerable than others, and they constitute about 90 per cent of the vulnerable patient slot (Bhattacharya et al., 2019). Understanding the gravity of the problem, the government took some initiatives, namely Multi-sectoral Development Programme (MSDP)/Border Area Developmental Programme (BADP) through local government (Panchayat Raj Institutions) and implemented some community-scale water treatment plants in the affected areas. An Arsenic Task Force was formed, who suggested a master plan in 2005–2006 for the entire affected areas in West Bengal. In this master plan, they chalked out to implement 338 groundwater-based Piped Water Supply Schemes (PWSS) and 12 surface water-based PWSS and installation of 165 arsenic removal plants on the existing groundwater-based schemes (Bhowmick et al., 2018). RRH project is also there, though not yet popular. To be precise, RRH is still not appreciated by the government, especially in rural South Bengal of India.

The present study area is situated over the unconsolidated alluvium of Pleistocene-recent age, which is the source of excessive iron and arsenic at a depth of 15 m and 50 m, respectively, in the PMB plain. The slope and underground groundwater flow are 2–5 m towards southward and towards the south-east direction, respectively (Bhattacharya et al., 2019). Most of the shallow tube wells in the research area, located in the North 24 Parganas district, have more than 300 ppb of arsenic concentration, well above the permissible limit of 10 ppb as prescribed by the World Health Organization (WHO). In this district, more than 2.19 million people are suffering from arsenic-related problems. Among them, more than 42.4 per cent are under high risk, that is, above 10 ppb (PHED, 2018). Several lists of affected people have been prepared by different organisations that may differ from each other (Bhattacharya, 2010). The Department of Health, Government of West Bengal, reported arsenicosis patients to be <15,000, while the non-governmental sources put these numbers between 200,000 and 300,000. Sometimes complexity of the water crisis ultimately ended with social exclusion (Wutich & Brewis, 2014). Water stress is always associated with human capital loss and ultimately social loss. Hunter (1875) stated that during the British period, females used to collect water from a river or tanks for domestic uses. After that, the then British government switched over from surface to groundwater for safe water sources. From that time onwards, farmers usually practised Aman (winter rice) much more extensively than Aus (summer rice). Hence, traditionally, there is a huge requirement of water for lean period cultivation. It is further aggravated by high-yielding rice varieties, that is, Boro rice (based on shallow tube well (STW) irrigation in winter season), which have more arsenic accumulated rice than the local varieties (Bhattacharya et al., 2019). At the same time, after the partition of Bangladesh, a huge number of refugees came and settled here. Soon, poor and helpless people simply encroached the levee of River Bhagirathi and settled here permanently. Some colonies were reclaimed from marshy land and consequently, recharge areas gradually phased out. However, PWSS has been started recently in some villages but in meagre numbers. Under this circumstance, most inhabitants still depend on groundwater and still consume contaminated water.

Figure 1.

Source: The authors.

III. Methods

The methodology consists of three parts: selection of study area and data collection, calculation of indices and analysis of data.

Sample Selection and Data Collection

As stated above, the study was conducted in 4 blocks in the district of North 24 Parganas. These were Amdanga, Basirhat I, Barrackpur II and Barasat II (Figure 1). The selection of the four arsenic-affected blocks (administrative units below districts) was based on the ‘stage of groundwater’ (CGWB, 2013). The CGWB of India categorised all the blocks into three groups, namely ‘safe with arsenic contamination’, ‘safe with water table decline’ and ‘semi-critical’. At first, the authors wanted to select one block from each group. But after inspection, it was observed that the group ‘safe with arsenic contamination’ contained some blocks with saline water. Consequently, the authors chose Barrackpur II (‘semi-critical’), Barasat II (‘safe with arsenic contamination’ and non-saline water), Basirhat I (‘safe with arsenic contamination’, and saline water) and Amdanga (‘safe with water table decline’) blocks. The selection of blocks was not taken randomly. Three blocks, other than Basirhat I, were selected based on their proximity to the district administrative headquarters, that is, Barasat. Basirhat I was selected as it had saline water in both deep and shallow tube wells. However, the surface water in this block was not saline.

Table 1.

List of Selected Blocks and Block-wise Numbers of the Selected Households

Block	Number of Surveyed Villages	Households (Census of India, 2011) (For Selected Villages Only)	Number of Surveyed Households	Percentage of Households
Amdanga	6	4,079	179	4.388
Barasat II	7	4,439	247	5.564
Basirhat I	8	4,001	286	7.148
Barrackpur II	12	2,791	211	7.560
Total	33	15,310	923	6.029

Source: The authors (Primary data 2019) and Census of India (2011).

After selecting the blocks, villages were selected randomly from each block, and households were also randomly selected from each selected village. The number of sampled households in each of these villages is shown in Table 1.

A draft questionnaire was made based on a pretest of 15 respondents in the study villages. The questionnaire was then finalised and canvassed. The questions were translated into the local language (Bengali) for the sake of convenience of the respondents. Targeted respondents had generally been the heads of the households. In the absence of the head, the next senior member of the family, that is, spouse, elder son or daughter, had been approached. It was a door-to-door survey conducted to gather key information on socio-economic status. The first section of the final questionnaire was about the respondents’ background. The second part comprised of water quality, quantity, choice and the demand for drinking and domestic water. The third part dealt with water-related issues and the direct and indirect effects of water. The last section was about the quantity of land, current use of rainwater and willingness of drinking harvested water and adoption of RRH. The survey was conducted between October 2018 and March 2019. The sample size was 923 households. However, male and female respondents were 509 and 414, respectively, of which 55.1 per cent belonged to non-Hindu and the rest fell under the Hindu category.

IV. Methodology

Deciding on the installation of RRH gadgets did not only depend on physical characteristics but also on socio-economic features. The entire calculations were made in two stages. In the first stage, we calculated five indices, which were developed based on the responses of the households to the perception-based statements made by the respondents on the socio-economic aspects. For each index, a set of 3–7 perception statements related to the theme of the index was taken into consideration. The response to each question was converted to an indicator comprising only two values—‘0’ and ‘1’. The code ‘1’ was given to the answer, which positively contributed to the index value, and was taken ‘0’ otherwise. Thus, ‘1’ stood for more of the value of the index, and ‘0’ stood for less or absence (Özdemir et al. 2011). The indices were obtained by adding the indicators (binary values) to the questions related to the specific indices, namely, ‘health hazard index (HHI)’, ‘water insufficiency index (WII)’, ‘water awareness action index (WAAI)’, ‘willingness to adopt RRH index (WRRHI)’ and ‘poverty level index (PLI)’. These indices thus took only non-negative integer values starting from 0; the maximum was the number of questions corresponding to the index. for example, for HHI, we combined responses of seven questions. Thus, for each household, we had seven answers. Suppose the answers to these questions were 1, 0, 0, 1, 1, 1, 0 and 1, we just considered the sum of these values. For this particular household, the sum was 5. Thus, we had a number in between 0 and 7 for each household. A cut-off value, say 3, was then considered, and each household having the sum of the answers to the 7 questions less than or equal to 3 was given the value 0, and the households having the sum greater than 3 were given the value 1. The cut-off value was considered in such a way that there were almost equal number of households in each group.

We then found the interrelation between these indices. Most of the bivariate tables showed significant dependency between the indices.

The second stage consisted of logistic regressions to see (a) the combined effect of four indices on HHI and (b) the combined effect of three of these indices, namely WII, WAAI and PLI, along with HHI and some socio-demographic variables on WRRH. Thus, in these two logistic regressions, the target variables were HHI and WRRH, respectively.

Table 2 shows the questions and the possible answers for the five indices. The value of the index is just the sum of the values in the third column of Table 2 corresponding to the responses of the questions in each index.

During the field survey, it was found that the percentage of households suffering from waterborne disease type I had been more than those suffering from waterborne disease type II, which was followed by vector-borne disease. On the other hand, the number of households suffering from arsenic-related diseases and waterborne diseases had been more or less the same. More than 49.2 per cent households had been suffering from gastrointestinal problems, which raised questions on the quality of water. It was observed that about 65.9 per cent households suffered frequently from water-related diseases. The percentage of the households that complained about the occurrence of vector-related nuisances during the previous year was 77.9 per cent. However, only 7.5 per cent households suffered frequently from vector-related diseases.

When asked about the taste of water, 69.3 per cent households told that it is good. The villagers meant sweet water as ‘good’ and salty or water with bad odour as ‘not good’. About 85.5 per cent households showed dislike for their water due to reddish colour. The households that had been using either boiled water or filtered water for drinking purposes were only 2.6 per cent. Only 14 per cent households had been using pond water for domestic purposes. Only 40.4 per cent households had been collecting rainwater occasionally, and that too in a very crude manner. They also collected it during the rainy season but did not store it for using it afterward.

Table 2.

The Perception Questions Towards Finding HHI, WII, WAAI, WRRHI and PLI

Index	Perception Questions	Possible Answers	Percentage of ‘1’
HHI	1. Have any of your members suffered from vector-borne diseases (malaria, dengue and chikungunya)? 2. Are you frequently suffering from waterborne disease type I (dysentery and diarrhoea)? 3. Have any of your members suffered from waterborne disease type II (jaundice and typhoid)? 4. Are you suffering from arsenic-related diseases (thickening of the skin, bronzing of the skin, pigmentation on skin and skin lesion)? 5. Are you frequently suffering from gastrointestinal problems? 6. Does anybody suffer from constipation in your family? 7. Do you face vector-related nuisance throughout the year?	Yes = 1, no = 0 Yes = 1, no = 0 Yes = 1, no = 0 Yes = 1, no = 0 Yes = 1, no = 0 Yes = 1, no = 0 Yes = 1, no = 0.	7.5 20.9 14.5 20.3 49.2 14.2 77.9
WII	1. How does the water taste? 2. Do you like the colour of the water? 3. Did waterborne diseases attack you frequently? 4. Did vector-borne diseases attack you within the last 5 years? 5. Is the water demand–supply gap large enough?	Good = 0, not good = 1 Like=0, dislike = 1 Yes = 1, no = 0 Yes = 1, no = 0 Large = 1, small = 0	69.3 14.5 65.9 7.5 46.5
WAAI	1. Are you treating drinking water? 2. Do you collect rainwater for domestic purposes? 3. Do you use pond water for domestic uses?	Yes = 1, no = 0 Yes = 1, no = 0 Yes = 1, no = 0	2.6 40.4 14.0
WRRHI	1. Is the decision of adoption of RRH conditional? 2. Are you willing to drink or use RRH water directly? 3. Are you ready to contribute (cash or kind) for the start of RRH? 4. Do you have any positive expectations from RRH?	Yes = 1, no = 0 Yes = 1, no = 0 Yes = 1, no = 0 Yes = 1, no = 0	77.8 48.0 33.8 19.7
PLI	1. Do you have any savings? 2. Do you have costly assets? 3. Do you have cattle? 4. Do you purchase fuel? 5. Do you have a pucca house? 6. Are you in service or doing business or in other occupations? 7. How much homestead land do you have?	Yes = 0, no = 1 High = 0, low = 1 Yes = 0, no = 1 Yes = 0, no = 1 Yes = 0, no = 1 Service or business = 0, other = 1. Small land = 0, large land = 1	24.1 50.1 18.4 43.6 41.3 41.5 47.7

Source: The authors.

Almost 100 per cent of women did not know about the daily requirement of water for domestic or drinking purposes. According to the Public Health Engineering (PHE) department guidelines, 70 litres per capita per day (lpcd) water supply has been standard for rural areas (Dave, 2016). However, the gap between water demand and supply was classified as per the existing guidelines of PHE, which states that if the gap is ≤160 lpcd, it is considered as small gap, while for gap greater than 160 lpcd, the same is treated as a large gap. We found a wide gap between water demand–supply to the extent of 46.5 per cent. Generally the water demand had never been a fixed phenomenon, it varying over time, space to cultural or socio-economical values (Mahmood & Hossain, 2017). Almost 77.8 per cent respondents had been conditionally willing to adopt RRH though we found 48 per cent households had been willing to use rainwater directly for domestic or drinking purposes. In addition to that, 33.8 per cent respondents had been willing to invest money or work as labour for the installation of RRH. Among them, 19.7 per cent respondents had positive expectations from RRH. In such cases, if households supported any of the five options or more, that is, low cost, in situ, arsenic-free, easy, water security, then it was given the value 1 otherwise, it was considered 0. The types of assets consisted of television with cable, car, bike, computer/laptop, refrigerator and water filter. If the household possessed these assets, it was given the value 1 otherwise, it was given the value 0. These values were subsequently added. If the sum was ≤4, it was considered low, while if the sum was greater than 5, it was considered as high. Biomass and coal were considered under the category of ‘purchased fuel gas’. If the households possessed it, it was given the value 0, otherwise it was 1. In the case of PLI, households with no monthly savings, not able to purchase fuel, not possessing any cattle, residing in ‘kacha’ house, having occupation ‘other than service and business’ and having small homestead land were considered as indicators of low poverty level.

Analysis of Data

Since HHI is the primary concern, we computed the bivariate contingency tables, taking HHI on the one side and each of the aforementioned indices on the other side and performed chi-square test ( $χ^{2})$ test to see whether their relations were significant. They were significant at 5 per cent level in three out of four cases. HHI was not significant at all with PLI. However, we retained it in the logistic regression of HHI on the indices to observe the simultaneous effect of these indices. Logistic regression was carried out because the response variable, HHI, was considered as binary. The variables were analysed with SPSS software (version 16).

V. Results and Discussion

People in South Bengal experience a hot and humid tropical climate almost throughout the year. Annual ranges of temperature and rainfall had been 16–42°C and 1,295–3,945 mm, respectively. The average relative humidity had been greater than 65 per cent (Chatterjee et al., 2009). Generally, South West Monsoon (June-Sep) had been slightly less than 1,000 mm and north-east monsoon had been slightly above 200 mm of rainfall on the average. Hence, the average annual rainfall had been about 1,200 mm. The trend of annual average rainfall over 100 years (1901–2001) indicates the adequate availability of rainwater. Over these 100 years, the mean rainfall had been fluctuating in June, but during August and September, it had always been reasonably stable. Despite having this generous rainfall, the inhabitants of the district had been facing serious qualitative and quantitative water crisis due to lack of proper conservation.

Relation Between Indices

To understand the associations among four indices, that is, WII, WAAI, WRRHI and PLI with the index HHI, we found the bivariate contingency tables of each of these four indices with HHI and performed chi-square test to observe the degree of associations. For each case, the p-value was noted.

Table 3.

Relation Between HHI and WII

	Less Water Insufficiency	More Water Insufficiency	Total
(1)	(2)	(3)	(4)
Low health hazard	417	229	646
Acute health hazard	114	163	277
Total	531	392	923
Proportion of acute HHI	0.215	0.416	0.300

Source: The authors.

Note: p-Value = 0.000.

Table 4.

Relation Between HHI and WAAI

	No Water Awareness Action	Water Awareness Action	Total
(1)	(2)	(3)	(4)
Low health hazard	308	338	646
Acute health hazard	159	118	277
Total	467	456	923
Proportion of acute HHI	0.340	0.259	0.300

Source: The authors.

Note: p-Value = 0.004.

There has been a significant relation between HHI and WII, as the p-value of the chi-squared test is observed to be 0.000, which is much less than the level of significance of 0.05 and 0.01. Thus, this relation is statistically significant at 1 per cent level of significance (Table 3). The relation is positive because the product of the diagonal frequencies is larger than the product of the off-diagonal frequencies, that is, higher HHI signifies higher WII. The local people usually viewed the term ‘water insufficiency’ quantitatively, and consequently, they could not associate water insufficiency with health hazards. In this article, water insufficiency has been considered qualitatively. Thus, the relation has been positive. This may be further analysed by considering the percentages of households with acute health hazards in each column of the table. The percentage of households with acute health hazards among households with less water insufficiency (21.5%) is much less than that of people with more water insufficiency (41.6%). Similarly, we compute the cross tables for other indices.

There has been a significant relation between HHI and WAAI, as the p-value of the chi-squared test is 0.004, much less than 0.01 (Table 4). It means that the relationship is statistically significant at 1 per cent level. HHI and WAAI have a negative relation because 34.0 per cent households were found with acute health hazards among households that were not taking water awareness action, whereas this was 25.9 per cent among households taking water awareness action.

Table 5.

Relation Between HHI and WRRHI

	Less Willing	More Willing	Total
(1)	(2)	(3)	(4)
Low health hazard	261	385	646
Acute health hazard	94	183	277
Total	355	568	923
Proportion of acute HHI	0.311	0.322	0.300

Source: The authors.

Note: p-Value = 0.037.

Table 6.

Relation Between HHI and PLI

	Less Poor	Poorer	Total
(1)	(2)	(3)	(4)
Low health hazard	362	284	646
Acute health hazard	155	122	277
Total	517	406	923
Proportion of acute HHI	0.299	0.300	0.300

Source: The authors.

Note: p-value = 0.519.

A significant relation has also been there between HHI and WRRHI at 5 per cent level (Table 5). When we compare the proportion of acute HHI, we found that HHI and WRRHI have been positively related. Though no significant relation is found between HHI and PLI (Table 6), poorer households were observed to have more health hazards.

Besides this, we have also computed several cross tables. We mention here only those tables that have been more relevant to the present article. We need to find out whether the households taking action were aware of the problems. It is evident from Table 4 that the households, who had more health hazards, acted less, which seemed to be contradictory. Action is related to awareness. People, who had been suffering, needed to be made more aware, which is reflected in the data presented in Table 7.

Table 7.

Relation Between Awareness of Arsenic and Water Awareness Action

	Action Not Taken	Action Was Taken	Total
Not aware	362	309	671
Aware	105	147	252
Total	467	456	923
Proportion of awareness	0.225	0.322	0.273

Source: The authors.

Note: Chi-square value = 11.057, p-value = 0.001.

There has been a significant positive relation between ‘Awareness of arsenic’ and ‘Water Awareness Action’. It implies those who had been aware of arsenic had taken appropriate water awareness action.

Logistic Regression Analysis

The logistic regression was carried out taking the log-odd ratio of the HHI on the four indices to observe the combined effect of the indices. HHI was taken as a binary dependent variable (y), which was given the value ‘0’ for low health hazards and ‘1’ for acute health hazard. Each of these indices at the right side of the regression was also taken as binary, with the value ‘0’ for low and ‘1’ for high.

In logistic regression, except PLI, all the indices, namely, WII, WAAI and WRRHI have statistically significant effect on HHI. WII and WRRHI have a positive effect on HHI, while WAAI harms HHI (Table 8). The reason behind adding the PLI is to see if there has been any relation between the poverty level and health hazard when other variables are included. Here, the health hazard had been a proxy of the low quality of water and below standard lifestyle. Most of the inhabitants used contaminated water, as they used shallow tube wells for drinking purposes. The PHE department of government regularly monitored those tube wells, and some of the tube wells were marked in red, so that people would avoid using those. Without having any alternatives, people continued to consume unsafe water. Also, the depth had been within 80 m. Since almost everybody was forced to use the water, there was not much difference between rich and poor in terms of water intake. Also, there was not much difference in income between the poor and the so-called nonpoor. From the above regression, we can conclude that household health had been largely affected by the quality of the available water.

The above results imply that the quality of drinking water matters a lot. Drinking unsafe water further indicated that either the people were too poor to afford the safer quality of drinking water, or they were not much conscious about it.

The excess water available during the rainy season is to be managed properly, and there has been a need for alternative water storage. Thus, RRH is one of the feasible and simple alternatives. However, the installation cost has been reasonably high. Only 17.8 per cent households, that were aware of arsenicosis in their locality were willing to accept RRH. Only a few households used a filter or boil water for drinking purpose, and many of them used pond water for domestic purposes.

Table 8.

Result of Logistic Regression of HHI on WII, WAAI, WRRHI and PLI

	B	S.E.	Wald	Sig.	Exp(B)
WII	0.941	0.149	40.011	0.000	2.562
WAAI	−0.481	0.155	9.615	0.002	0.618
WRRHI	0.409	0.161	6.472	0.011	1.506
PLI	−0.120	0.152	0.627	0.428	0.887
Constant	−1.268	0.173	53.853	0.000	0.281

Source: The authors.

Note: −2 Log likelihood = 1071.055, percentage of correct predictions: 69.9%.

The intention of the households to accumulate in-house water during the rainy season had been another important aspect to investigate. 38.7 per cent of the households complained that water entered into their houses due to heavy rain, whereas 30.2 per cent informed that houses were damaged during consecutive heavy rains. Logistic regression was carried out to see how people decide about RRH. We regressed WRRH on the socio-economic characteristics, an attitude of acceptance and knowledge of water conservation, etc. Among the indices, WRRHI was not included as a regressor as one of the components of WRRHI was the response variable.

It is evident from Table 9 that most of the variables have an effect on WRRH in the expected direction, but only religion, presence of arsenicosis patients in the locality and possession of water tank have significant influence, at least at 5 per cent level of significance. RRH was less acceptable to Hindus compared to Muslims. The section of the household that already had a water tank would accept the proposal. The households with members suffering from arsenicosis would naturally opt for the implementation of RRH. Surprisingly none of the indices, that is, HHI, WII, WAAI, and PLI, have any significant effect on WRRH. Apart from the three regressors having significant influence, there are three variables, namely below poverty line (BPL) households, households with water accumulation problem and the houses which were damaged by heavy rain have a significant effect at 10 per cent level of significance. BPL and water accumulated houses did not opt for RRH, while the houses damaged by rain still wanted that. It is worthy to be mentioned that if a person earns less than ₹27,000 per annum (₹2,250/month) they would be categorised as BPL (MEITGI, 2016).

Table 9.

Result of Logistic Regression of Willingness to Rooftop Rainwater Harvesting on the Related Independent Variables

	B	S.E.	Sig.	Exp(B)
Religion	−0.415	0.187	0.027	0.661
BPL or APL	−0.409	0.240	0.088	0.665
Own house	−0.354	0.386	0.359	0.702
Known arsenicosis patient	0.746	0.251	0.003	2.109
Water accumulation in house	−0.342	0.201	0.088	0.710
Heavy rain damages the house	0.397	0.222	0.073	1.488
Own water tank	0.796	0.304	0.009	2.216
HHI	−0.012	0.188	0.949	0.988
WII	0.146	0.171	0.393	1.158
WAAI	0.000	0.166	0.998	1.000
PLI	0.075	0.171	0.659	1.078

Source: The authors.

Note: −2 Log likelihood = 945.483, percentage of correct predictions: 77.8%.

Religion: Hindu = 1, non-Hindu = 0; own water tank: yes = 1, no = 0; own house: yes =1, no = 0; BPL: yes =1, APL = 2; known arsenicosis patient: yes = 1, no = 0; water accumulation in house: yes = 1, no = 0; heavy rain damages the house: yes = 1, no = 0; HHI: less = 0, more = 1; WII: less = 0, more = 1;WAAI: yes = 1, no=0;, PLI: less = 0, more = 1.

About 95 per cent households had their own house, and among them, 51.2 per cent households had a concrete roof, and 82.3 per cent fell under the APL category. They were in a position to adopt RRH. On the other hand, 47.3 per cent had a smartphone; 58.1 per cent had anormal phone; and a combination of both phones had 22.2 per cent. But there had been a lack of awareness about safe water. Only 5.9 per cent households had water filters. Most of them used tube well water, submersible pump water or well water (89.1%) as a source of drinking water. We found that more than 78 per cent of the households having income below US$132.21 (₹10,000) used shallow tube well (STW). Water uses varied widely between rich and poor. Farmers always considered themselves insecure due to contaminated water (Singh et al., 2017). Despite this fact, people kept on consuming unsafe water.

More than 72 per cent households had proximity to the source of water. Pandit and Biswas (2019) stated the Government of India gives higher priority to water conservation mainly in the areas where there is water scarcity—qualitatively or quantitatively. An alternative water supply scheme must be explored where surface water is inadequate to meet the demand. Arsenic is difficult to be removed in a simple, economic way. Kulkarni (2016) found that RRH could fulfil more than 50 per cent domestic demand of an 8–10-member family and the public needs. RRH does not depend on physical or environmental phenomena; rather, it depends much on social attitude and acceptability (Debusk & Hunt, 2012). RRH helps to improve the hygiene and sanitation of households (Baguma et al., 2010). There had been a lack of knowledge about the proper process of RRH among rural people.

The percentage of households that used different sources of water for drinking and domestic purpose was 55.7 per cent. Even households consuming treated water, either filtered or boiled, were also found to suffer from arsenicosis. Again, we found 2.1 per cent, 0.5 per cent and 11.8 per cent had been suffering from thickening of skin, skin lesions and bronzing of skin, respectively. Overall, a total of 87 per cent households used tube well water for drinking purpose, and among them, 25.4 per cent households also suffered from acute health hazards. Hence, it might also be one key inclination towards RRH. Generally, better literacy and basic education help public awareness to build up, which encourages demand for basic services (Mukherjee et al., 2009b). Education and awareness were not the one-sided relations; rather, sometimes it had been reciprocal (Bhattacharya et al., 2019). Early detection of arsenicosis got delayed due to a lack of information or awareness. Mahimairaja et al. (2005) cited many cases where educated people were also affected in the same way by arsenic. From the surveyed data, we found that 35.5 per cent people had been literate up to 12th standard and had been suffering from water- and vector-related diseases.

VI. Conclusions

Unsafe drinking water in North 24 Parganas has a cascading effect on the households facing great risk of arsenic toxicity. Water pollution sometimes relates to a lack of social education as skin lesions are mistaken as leprosy and painful cough with blood-sputum is mistaken as tuberculosis. Rural people took it as a curse of God. In this way, early symptoms of arsenicosis were not recognised (Bhattacharya et al., 2019). The awareness of water quality, especially in terms of arsenic and iron, among those who had been suffering from acute health hazards was 8.2 per cent and 7.7 per cent, respectively. In the survey, it was found that 27.3 per cent and 25.2 per cent people were aware of arsenic and iron contamination, respectively.

Unsafe water has always a social cost. Due to arsenicosis, female patients sometimes remain unmarried; even married women get divorced. People lose their working ability and become liabilities for their families, and sometimes, they need to borrow money from the local money lenders to run their families (Bhattacharya et al., 2019).

The cumulative effects of unsafe water on households had so far been neglected continuously. Recently, the Government of West Bengal started supplying surface water. The supplied water had been sometimes more than its requirement. In that case, water misuse should be restricted by the installation of a water meter. Tax may be levied on water consumption beyond 70 per lcpd. Most of the middle class, wealthy and low-income households were willing to adopt RRH if the government would have provided it at free of cost or offered some incentives or subsidy. From the above discussion, it can be concluded that the potential of RRH is still not appreciated properly. Every work, rather every moment, matters a lot. Thus, it is the need of the hour to introduce simple and smart plans for water risk management in the form of RRH, which will make people aware of the necessity of water conservation through social knowledge followed by its real-life application. The households, under all circumstances, should be aptly educated for acting towards RRH.

Footnotes

Acknowledgements

We acknowledge the valuable comments and suggestions of the reviewers who went through the first draft of the article and helped a lot to improve the quality of the text.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

References

Anand

, Navin

, Shankar

, Kumar

, Ali

, Verma

, Ghosh

, & Kumar

(2017). Arsenic contamination of groundwater and human blood in Vaishali district of Bihar, India: Health hazards. International Journal of Advanced Research , 5(8), 2092–2100. https://doi.org/10.21474/ijar01/5276

Baguma

, Loiskandl

, & Jung

(2010). Water management, rainwater harvesting and predictive variables in rural households. Water Resources Management , 24(13), 3333–3348. https://doi.org/10.1007/s11269-010-9609-9

Banerjee

A. K.

(2018). Multidimensional indices with data-driven dimensional weights: A multidimensional coefficient of variation. Arthaniti: Journal of Economic Theory and Practice , 17(2), 140–156. https://doi.org/10.1177/0976747918792644

Bhattacharya

A. K.

(2010). Artificial groundwater recharge with a special reference to India. IJRRAS , 4(2), 214–221.

Bhattacharya

A. K.

, Sarani

, Kumar Roy

, Karthik

, Singh

, Kumari

, Daksh

, Kumar

, & Anurag & Mishra

A. K.

(2019). An Analysis of Arsenic Contamination in the Groundwater of India, Bangladesh and Nepal with a Special Focus on the Stabilisation of Arsenic-Laden Sludge from Arsenic Filters. Electronic Journal of Geotechnical Engineering , 24(Bund. 01), 1–34.

Bhowmick

, Pramanik

, Singh

, Mondal

, Chatterjee

, & Nriagu

(2018). Arsenic in groundwater of West Bengal, India: A review of human health risks and assessment of possible intervention options. Science of the Total Environment , 612, 148–169. https://doi.org/10.1016/j.scitotenv.2017.08.216

Census of India. (2011). District Census Handbook, North 24 Parganas , part XII B, Series 22, (pp. 26–28).

Central Ground Water Board (CGWB). (2019). Arsenic hot spot in ground water in India (pp. 1–23). http://cgwb.gov.in/WQ/WQMAPS/Arsenic.pdf

Central Ground Water Board (CGWB). (2013). Master plan for artificial recharge to ground water In India . http://cgwb.gov.in/documents/MasterPlan-2013.pdf

10.

Chatterjee

, Nath

, Jana

, Goswami

, Sudipto

, Mukherjee

, Shome

, Bagchi

, Sarkar

M. J.

, Bhattacharya

, Jacks

, & Ahmed

K. M.

(2009). Geochemistry and speciation of solid and aqueous phase arsenic in the Bengal Delta Plain Aquifers. Bhujal News Quarterly Journal , 24(2–3), 59–70.

11.

CPCB. (2002). Arsenic contamination in groundwater . CPCB.

12.

Datt

, Ravallion

, & Murgai

(2016). Four questions and answers on growth and poverty reduction in India. Arthaniti: Journal of Economic Theory and Practice , 15(2), 1–10. https://doi.org/10.1177/0976747920160201

13.

Dave

S. N.

(2016). Rural drinking water situation: Challenges and opportunities in West Bengal . https://www.mdws.gov.in/sites/default/files/annualre-

14.

Debusk

, & Hunt

W. F.

(2012). Rainwater harvesting : A comprehensive review of literature . Water Resources Research Institute, University of North Carolina.

15.

Elangovan

, & Chalakh

M. L.

(2006). Arsenic pollution in West bengal. Technical Digest , 9, 31–35.

16.

Fenwick

(2010). Identifying the water poor : An indicator approach to assessing water poverty in rural Mexico . University College London.

17.

Hunter

W. W.

(1875) A statistical account of Bengal (pp. 1–291). Trubner and Co.

18.

Karfakis

, Lipper

, & Smulders

(2012). The assessment of the socio-economic impacts of climate change at household level and policy implications (pp. 133–150). Agricultural Development Economics Division, FAO.

19.

Kulkarni

S. J.

(2016). Review on studies, research, and surveys on rainwater harvesting. International Journal of Research and Review , 3(9), September 2016. (www.gkpublication.in).

20.

Mahimairaja

, Bolan

N. S.

, Adriano

D. C.

, & Robinson

(2005). Arsenic contamination and its risk management in complex environmental settings. Advances in Agronomy , 86, 1–82. https://doi.org/10.1016/S0065-2113(05)86001-8

21.

Mahmood

, & Hossain

(2017). Feasibility of managed domestic rainwater harvesting in South Asian rural areas using remote sensing. Resources, Conservation and Recycling , 125, 157–168. https://doi.org/10.1016/j.resconrec.2017.06.013

22.

Majumder

(2008). Capability based approach and India’s poverty. Arthaniti: Journal of Economic Theory and Practice , 7(1–2), 26–40. https://doi.org/10.1177/0976747920080103

23.

Martinson

D. B.

, Ranatunga

N. U. K.

, & Gunaratne

A. M. C. H. A.

(2002, November). Reducing rainwater harvesting system cost. Proceedings of the 28th WEDC Conference. https://www.researchgate.net/publication/237287171_THE_ROOFWATER_HARVESTING_LADDER

24.

Ministry of Electronics & Information Technology Government of India (MEITGI). (2016). Below poverty line certificate . https://www.meity.gov.in/content/uid

25.

Mukherjee

, & Banerjee

(2009a). Education and health: Strategic sequencing in social development. Arthaniti: Journal of Economic Theory and Practice , 8(1–2), 25–49. https://doi.org/10.1177/0976747920090103

26.

Mukherjee

S. C.

, Pati

, Dutta

R. N.

, & Saha

K. C.

(2009b). Groundwater arsenic contamination situation in West Bengal, India : A nineteen year study. Bhujal News Quarterly Journal , 24, 1–40.

27.

Mukhopadhyay

, Kulkarni

, & Kulkarni

(2016, January). Geostatistical and geospatial approaches for the characterization of natural resources in the environment. https://doi.org/10.1007/978-3-319-18663-4

28.

Okoye

C. O.

, Solyali

, & Akintuʇ

(2015). Optimal sizing of storage tanks in domestic rainwater harvesting systems: A linear programming approach. Resources, Conservation and Recycling , 104, 131–140. https://doi.org/10.1016/j.resconrec.2015.08.015

29.

Özdemir

, Elliott

, Brown

, Nam

P. K.

, Hien

V. T.

, & Sobsey

M. D.

(2011). Rainwater harvesting practices and attitudes in the mekong delta of Vietnam. Journal of Water Sanitation and Hygiene for Development , 1(3), 171–177. https://doi.org/10.2166/washdev.2011.024

30.

Pandey

(2019). India world’s 13th most water-stressed country: WRI . https://www.downtoearth.org.in/news/water/india-world-s-13th-most-water-stressed-country-wri-66066

31.

Pandit

, & Biswas

A. K.

(2019). India’s National Water Policy: ‘Feel good’ document, nothing more. International Journal of Water Resources Development , 35(6), 1015–1028. https://doi.org/10.1080/07900627.2019.1576509

32.

PHED. (2018). West Bengal drinking water sector improvement project: Arsenic and fluoride in drinking water in West Bengal—Characteristics, implications, and mitigation . https://www.adb.org/projects/documents/ind-49107-006-rrp

33.

Roy

(2008). Economic benefits of arsenic removal from groundwater—A case study from West Bengal, India. Science of the Total Environment , 397(1–3), 1–12. https://doi.org/10.1016/j.scitotenv.2008.02.007

34.

Sarkar

, & Ray

(2019). Collective action and tragedy of tank water. Arthaniti: Journal of Economic Theory and Practice . https://doi.org/10.1177/0976747919868696

35.

Singh

S. K.

, Gosh

A. K.

, Kumar

, Kislay

, Kumar

, Tiwari

R. R.

, Parwez

, Kumar

, & Imam

M. D.

(2014). Groundwater arsenic contamination and associated health risks in Bihar, India. International Journal of Environmental Research , 8(1), 49–60. https://doi.org/10.22059/ijer.2014.693

36.

Singh

L. K.

, Jha

M. K.

, & Chowdary

V. M.

(2017). Multi-criteria analysis and GIS modeling for identifying prospective water harvesting and artificial recharge sites for sustainable water supply. Journal of Cleaner Production , 142, 1436–1456. https://doi.org/10.1016/j.jclepro.2016.11.163

37.

Udayanapillai

A. V

, & Kaliammal

(2016). Groundwater characterization and quality assessment by using GIS and Geo statistics from Pandalgudi region, Virudhunagar District, Tamil Nadu, India. International Journal of Research in Environmental Science , 2(4), 1–16. https://doi.org/10.20431/2454-9444.0204001

38.

Wutich

, & Brewis

(2014). Food, water, and scarcity. Current Anthropology , 55(4), 444–468. https://doi.org/10.1086/677311

39.

Zakhar

, Derco

, & Čacho

(2018). An overview of main arsenic removal technologies. Acta Chimica Slovaca , 11(2), 107–113. https://doi.org/10.2478/acs-2018-0016

40.

Zhou

Z.-Y.

(2016). What determines food security in China: A review of literature. Arthaniti: Journal of Economic Theory and Practice , 15(1), 55–84. https://doi.org/10.1177/0976747920160104