ABSTRACT
Despite progress in Sustainable Development Goal 6, which envisages clean water and sanitation for all, enormous disparities in access to water and sanitation services between and within countries continue to be a significant challenge. Evidence on the spatial heterogeneity of water, sanitation, and hygiene practices among the Scheduled Tribe (ST) population is limited in India. This study estimates the spatial heterogeneity in ST people's access to improved water and sanitation facilities and its correlates at the district level. Geospatial techniques like Moran's I, univariate and bivariate local indicator of spatial association, and spatial regression models were carried out for 707 districts covered in the fifth round of the National Family Health Survey, 2019–21. Stata was used for descriptive analysis, and ArcMap and GeoDA were used for spatial analysis. Only about half of ST households had access to improved water and sanitation facilities in India. Spatial heterogeneity across districts was evident, with 130 districts from Uttarakhand, Himachal Pradesh, Punjab, Haryana, Delhi, Sikkim, Arunachal Pradesh, Nagaland, Mizoram, and some portion of Karnataka forming the hot spots. Gender and age of the household head, family type, and region were significantly associated with improved water and sanitation facilities among ST households.
HIGHLIGHTS
Only 50% of the ST households had access to improved water and sanitation.
Spatial heterogeneity across districts was evident in ST household's access to improved water and sanitation.
There were 130 hot and 131 cold spots in tribal-dominated highland and plateau zones.
Gender and age of the household head, family type, and region were associated with improved water and sanitation facilities among ST households.
INTRODUCTION
Access to improved water and sanitation is a fundamental human right, a Sustainable Development Goal (SDG), and crucial for achieving several other SDGs (Hutton & Chase 2016; Bayu et al. 2020; UN-Water 2023). Despite progress in SDG-6, which envisages clean water and sanitation for all, enormous disparities in access to water and sanitation services between and within countries continue to be a significant challenge (Sanico 2019). In 2020, 26% of the world's population (2 billion) did not have access to safely managed drinking water services, and an estimated 46% (3.6 billion) lacked access to safely managed sanitation (UN-Water 2023). Poor water, sanitation, and hygiene (WASH) practices can lead to increased susceptibility to morbidity and mortality (Pradhan 2015; Palo et al. 2021; Lin et al. 2022) and economic losses (Fuente et al. 2020). 80% of all deadly diseases are brought on by a lack of safe drinking water (Niemeyer 2018). Unsafe WASH is responsible for the deaths of around 400,000 children under five each year, or 1,000 every day (UNICEF 2023). Poor sanitation reduces socio-economic development due to anxiety, risk of sexual assault, and lost opportunities for education and work. It is also linked to the transmission of diarrhoeal diseases and malnutrition (WHO 2023).
The Indian government has implemented several programs to improve water and sanitation services, including the Central Rural Sanitation Programme (1986), National Rural Drinking Water Supply Programme (1969), Accelerated Rural Water Supply Programme (ARWSP) in 1972–1973, National Drinking Water Mission (1986) that was succeeded with the Rajiv Gandhi National Drinking Water Mission (RGNDWM) in 1991, Total Sanitation Campaign (1999), Swajal Dhara (2002), Bharat Nirman Programme (2005), National Rural Drinking Water Programme (NRDWP) that emerged from the ARWSP in 2009, and Nirmal Bharat Abhiyan (2012) that was rebranded as ‘rural sanitation’ (Biswas et al. 2022). India has made significant progress in sanitation over the past few years through the Clean India Mission (SBM-2014), the largest sanitation program globally. However, India's diverse socio-economic landscape presents unique challenges in achieving universal access to improved WASH.
In India, caste is a form of social stratification, and the castes are usually divided into four categories: Scheduled Castes, Scheduled Tribes (STs), Other Backward Classes, and General castes. STs are defined as ‘such tribes or tribal communities or parts of or groups within such tribes or tribal communities as are deemed under article 342 to be Scheduled Tribe for the purpose of this constitution’ (MoHFW & MoTA 2023). Around 90% of the tribal population in India lives in rural areas, and a large proportion of STs are collectors of forest produce, hunter–gatherers, shifting cultivators, pastoralists, nomadic herders, and artisans (MoHFW & MoTA 2023). Among the most vulnerable groups are the STs, constituting approximately 8% of India's population. Historically, STs have faced developmental disparities due to various factors, including geographical isolation, socio-political complexities, and limited resources (Hays 2018). As per Census 2011, only 10.7% of the ST population has access to tap water, as against 28.5% of the non-ST population. The availability of tap water in tribal households ranges from 2% in Odisha to 54% in Goa. Again, in rural areas, only 10% of ST households have access to improved sanitation facilities, compared to 61% of urban households (MoHFW & MoTA 2023).
Tribal populations in different states/regions of the country differ in socio-economic status and have uneven access to developmental programs. Evidence on determinants and spatial clustering of WASH practices among the tribal households is limited, and the available literature is based either on a specific region or particular tribes or on a small sample size. Existing small-scale studies indicate higher levels of unimproved sanitation and drinking water among the tribal population due to remote locations, limited resources (Pradhan 2015; Saha et al. 2020), and inadequate program outreach (Vila-Guilera et al. 2022). Considering the large tribal population and their socio-cultural and geographical diversity, it is pertinent to assess their access to improved water and sanitation facilities from nationally representative data. Additionally, an in-depth exploration of the spatial heterogeneity and its correlates, which contribute to their restricted access, is helpful for district/state-level targeted policy and programs. Against this backdrop, this study estimates the spatial heterogeneity in ST people's access to improved water and sanitation facilities and its correlates at the district level in India.
METHODS
The current study used data from the fifth round of the National Family Health Survey (NFHS-5), 2019–21. The NFHS-5 is a nationwide, cross-sectional, and large-scale demographic health survey conducted under the stewardship of the Ministry of Health and Family Welfare (MoHFW), Government of India. It used multi-stage stratified systematic sampling to draw representative estimates at the national and sub-national levels (state/district level). In the first stage, villages from rural areas and census enumeration blocks from urban areas were selected through the probability–proportional-to-size method. Households were selected through the systematic sampling in the second stage. The survey collected information from 636,699 households (with a 98% response rate), covering 724,115 women and 101,839 men nationwide. The published survey report provides details of the design, data collection strategy, and ethical considerations (IIPS & ICF 2021). Of the total households, 123,443 ST households were used for analysis.
Outcome variable
The primary outcome variable of this study was the availability of sources of drinking water and sanitation facilities among the ST households in India. It was further divided into three categories: ‘improved’ (a. improved sources of drinking water and b. improved sanitation facilities), ‘partially improved’ (a. improved sanitation and unimproved drinking water or b. improved drinking water and unimproved sanitation), and ‘unimproved’ (a. unimproved sources of drinking water and b. unimproved sanitation facilities). The categorization of the source of drinking water and sanitation facilities as ‘improved’ and ‘unimproved’ was based on the WHO/UNICEF Joint Monitoring Programme for Water Supply and Sanitation for the SDG monitoring period guideline (WHO/UNICEF JMP 2018). Access to an ‘improved’ source of drinking water consists of piped into dwelling, piped to yard/plot, piped to a neighbor, public tap/standpipe, tube well or borehole, protected well, protected spring, tanker truck, cart with small tank, bottled water, and community reverse osmosis plant. In contrast, the unimproved source of drinking water consists of unprotected wells, unprotected spring, river/dam/lake/pond/stream/canal, rainwater, and others. Similarly, access to improved sanitation facilities includes flush-to-piped sewer system, flush-to-septic tank, flush-to-pit-latrine, flush-don't know where, pit-latrine-ventilated improved pit, pit-latrine-with slab, and composting toilet. In contrast, unimproved sanitation facilities include flush-to somewhere else, pit-latrine-without slab/open pit, bucket toilet, hanging toilet/latrine, open defecation, and others.
Predictor variables
The predictor variables used in the study were the gender of the household head (HH), which are categorized as (male, female); the age of the HH (below 25, 25 and above); educational status of the HH (illiterate, literate); place of residence (urban, rural); religion (Hindu, Christian, others); type of house (kachha, semi-pucca, pucca); family structure (nuclear, extended); region (north, south, east, west, north-east, central); and wealth index. The wealth index measures a household's living standard and was calculated using data on the household ownership of selected assets, materials used for housing construction, and types of water access and sanitation facilities. The NFHS-5 dataset already contains the variable ‘wealth quintile’, which is categorized into poorest, poorer, middle, richer, and richest. For this analysis, the five categories of wealth quintile were recoded as ‘poor’ (poorer and poorest), ‘middle,’ and ‘rich’ (richer and richest) to generate the wealth index variable.
Statistical analysis
Statistical analysis of the present study commenced with a descriptive statistic of the percentage of three outcome variables (unimproved, partially improved, and improved water and sanitation facilities) among the ST households by selected socio-demographic characteristics. Furthermore, multivariable analysis in the form of multinomial logistic regression was conducted to determine the factors associated with the availability of water and sanitation facilities, and the results were presented as adjusted odds ratio (AOR) with 95% confidence intervals (CIs). The goodness of fit of the model was assessed through pseudo R2 statistics. A variance inflation factor was generated to check the multicollinearity, and no evidence of multicollinearity was found in the variables used. In all the analyses, weights were used to restore the sample's representativeness. The analyses were performed using Stata (V16).
The present study also utilized spatial tools and techniques such as univariate and bivariate Moran's I statistics and regression models to measure spatial autocorrelation (Khan et al. 2019; Singh & Vishwakarma 2021) in the distribution of improved water and sanitation facilities. Spatial clustering of improved water and sanitation across the districts was identified with the help of univariate and bivariate local indicator of spatial association (LISA) maps. Regression models were used to identify significant correlates of improved water and sanitation facilities among the tribal households. At first, ordinary least square (OLS) regression was used to estimate the extent of spatial autocorrelation in the error term. Given that the OLS has confirmed the presence of spatial autocorrelation in its error term, we subsequently conducted estimations for the spatial lag model (SLM) and spatial error model (SEM). The SLM operates on the assumption that the local geographical proximity influences observations of the dependent variable. In contrast, the SEM considers variables that are not included in the regression model but have an impact on the outcome variable. The primary distinction between the two models is that, in contrast to the SEM, the SLM does not consider the spatial dependence of the error term (Khan & Mohanty 2018; Khan et al. 2019; Singh & Vishwakarma 2021). Diagnostics for spatial dependence of the regression models were also carried out to find the best-fitted model for the dataset. After analyzing the Akaike information criterion (AIC) value, it was found that the AIC value was lowest for the SEM, which further denotes that the spatial error was the best-fitted model for the study. The spatial analyses were conducted using ArcMap (10.8) and GeoDa (v 1.12).
RESULTS
Access to improved water and sanitation and its determinants
Table 1 depicts the percentage distribution of ST households with unimproved, partially improved, and improved water and sanitation facilities by selected background characteristics. Among the surveyed ST households, 6.29% of households had unimproved, 43.53% had partially improved, and the rest had improved water and sanitation facilities. More than half of the households with improved water and sanitation facilities were headed by literate persons (56.29%) and those aged 25 and above (50.53%). More urban households had improved water and sanitation facilities (72.83%) compared to rural areas (46.44%). Greater diversity was observed in access to improved water and sanitation facilities among the religious groups and the region. Christian households (64.43%) and households in the north-east region had the highest percentage of improved water and sanitation facilities (66.40%). A higher percentage of rich households had access to improved water and sanitation facilities (83.05%) compared to poor (38.12%) and middle-income households (63.90%).
Access to sources of drinking water and sanitation facilities among tribal households by selected background characteristics in India, 2019–21
Background characteristics . | Unimproved water and sanitation facilities . | Partially improved water and sanitation facilities . | Improved water sanitation facilities . | Total number of tribal households . |
---|---|---|---|---|
Household headship | ||||
Male | 6.54 | 43.28 | 50.18 | 101,950 |
Female | 5.07 | 44.75 | 50.18 | 21,489 |
Age of the HH | ||||
Below 25 | 7.29 | 56.31 | 36.40 | 120,443 |
25 + | 6.26 | 43.21 | 50.53 | 3,000 |
Education | ||||
Illiterate | 7.93 | 50.52 | 41.56 | 51,164 |
Literate | 5.12 | 38.59 | 56.29 | 72,279 |
Place of residence | ||||
Urban | 1.72 | 25.45 | 72.83 | 17,496 |
Rural | 7.04 | 46.52 | 46.44 | 105,947 |
Religion | ||||
Hindu | 6.12 | 45.11 | 48.77 | 105,913 |
Christian | 5.72 | 29.85 | 64.43 | 10,450 |
Others | 9.60 | 40.14 | 50.25 | 7,080 |
Type of house | ||||
Pucca | 9.66 | 54.39 | 35.95 | 12,406 |
Semi-pucca | 7.87 | 49.19 | 42.94 | 64,423 |
Kachha | 3.00 | 31.93 | 65.07 | 43,432 |
Type of family | ||||
Nuclear | 6.86 | 45.30 | 47.84 | 73,363 |
Extended | 5.44 | 40.94 | 53.62 | 50,080 |
Region | ||||
North | 5.16 | 48.43 | 46.41 | 11,626 |
Central | 7.32 | 42.29 | 50.40 | 28,526 |
East | 7.58 | 52.79 | 39.63 | 30,977 |
North-east | 5.35 | 28.25 | 66.40 | 12,686 |
West | 6.58 | 43.35 | 50.07 | 23,930 |
South | 3.01 | 36.52 | 60.47 | 15,697 |
Wealth index | ||||
Poor | 8.68 | 53.19 | 38.12 | 81,640 |
Middle | 2.77 | 33.33 | 63.90 | 20,329 |
Rich | 0.49 | 16.46 | 83.05 | 21,474 |
Total | 6.29 | 43.53 | 50.18 | 123,443 |
Background characteristics . | Unimproved water and sanitation facilities . | Partially improved water and sanitation facilities . | Improved water sanitation facilities . | Total number of tribal households . |
---|---|---|---|---|
Household headship | ||||
Male | 6.54 | 43.28 | 50.18 | 101,950 |
Female | 5.07 | 44.75 | 50.18 | 21,489 |
Age of the HH | ||||
Below 25 | 7.29 | 56.31 | 36.40 | 120,443 |
25 + | 6.26 | 43.21 | 50.53 | 3,000 |
Education | ||||
Illiterate | 7.93 | 50.52 | 41.56 | 51,164 |
Literate | 5.12 | 38.59 | 56.29 | 72,279 |
Place of residence | ||||
Urban | 1.72 | 25.45 | 72.83 | 17,496 |
Rural | 7.04 | 46.52 | 46.44 | 105,947 |
Religion | ||||
Hindu | 6.12 | 45.11 | 48.77 | 105,913 |
Christian | 5.72 | 29.85 | 64.43 | 10,450 |
Others | 9.60 | 40.14 | 50.25 | 7,080 |
Type of house | ||||
Pucca | 9.66 | 54.39 | 35.95 | 12,406 |
Semi-pucca | 7.87 | 49.19 | 42.94 | 64,423 |
Kachha | 3.00 | 31.93 | 65.07 | 43,432 |
Type of family | ||||
Nuclear | 6.86 | 45.30 | 47.84 | 73,363 |
Extended | 5.44 | 40.94 | 53.62 | 50,080 |
Region | ||||
North | 5.16 | 48.43 | 46.41 | 11,626 |
Central | 7.32 | 42.29 | 50.40 | 28,526 |
East | 7.58 | 52.79 | 39.63 | 30,977 |
North-east | 5.35 | 28.25 | 66.40 | 12,686 |
West | 6.58 | 43.35 | 50.07 | 23,930 |
South | 3.01 | 36.52 | 60.47 | 15,697 |
Wealth index | ||||
Poor | 8.68 | 53.19 | 38.12 | 81,640 |
Middle | 2.77 | 33.33 | 63.90 | 20,329 |
Rich | 0.49 | 16.46 | 83.05 | 21,474 |
Total | 6.29 | 43.53 | 50.18 | 123,443 |
The adjusted effect of selected covariates on the likelihood of having partially improved and improved water and sanitation among ST households is presented in Table 2. The female-headed households had slightly higher odds of having a partially improved (AOR = 1.19; 95% CI: 1.11–1.29) and improved (AOR = 1.17; 95% CI: 1.09–1.26) water and sanitation facility compared to male-headed households. Households headed by individuals aged 25 and above had a higher probability of having an improved water and sanitation facility (AOR = 1.84; 95% CI: 1.58–2.14) compared to households headed by individuals below 25 years. Households headed by literates were likelier to have an improved water and sanitation facility (AOR = 1.28; 95% CI: 1.22–1.36) than non-literate ones. Rural households were less likely to have both partially improved (AOR = 0.45; 95% CI: 0.39–0.51) and improved (AOR = 0.18; 95% CI: 0.15–0.20) water and sanitation facilities compared to urban households. Households residing in pucca houses had higher odds of improved water and sanitation facilities (AOR = 1.02; 95% CI: 0.93–1.13) than those living in kachha houses. Extended families had a higher probability of having an improved sanitation facility (AOR = 1.19; 95% CI: 1.13–1.26) than nuclear families. Households from all other regions had higher chances of improved water and sanitation facilities than the northern region. However, households from the north-eastern regions had the highest odds (AOR = 3.10; 95% CI: 2.79–3.45) of having improved water and sanitation facilities compared to the northern region. Rich households (AOR = 22.94; 95% CI: 19.07–27.59) were likelier to have improved water and sanitation facilities than poor households.
Multinomial logistic regression showing the risk factors for partially improved and improved water and sanitation facilities (reference to unimproved facilities) among tribal households in India, 2019–21
. | Partially improved water and sanitation facility . | Improved water and sanitation facilities . | ||||
---|---|---|---|---|---|---|
. | . | 95% CI . | . | 95% CI . | ||
Background characteristics . | AOR . | Lower . | Upper . | AOR . | Lower . | Upper . |
Household headship | ||||||
Male® | ||||||
Female | 1.19*** | 1.11 | 1.29 | 1.17*** | 1.09 | 1.26 |
Age of the HH | ||||||
Below 25® | ||||||
25 + | 0.98 | 0.85 | 1.14 | 1.84*** | 1.58 | 2.14 |
Education | ||||||
Illiterate® | ||||||
Literate | 1.11*** | 1.05 | 1.17 | 1.28*** | 1.22 | 1.36 |
Place of residence | ||||||
Urban® | ||||||
Rural | 0.45*** | 0.39 | 0.51 | 0.18*** | 0.15 | 0.20 |
Religion | ||||||
Hindu® | ||||||
Christian | 0.75*** | 0.69 | 0.82 | 1.07 | 0.99 | 1.17 |
Others | 0.62*** | 0.57 | 0.68 | 0.87*** | 0.79 | 0.94 |
Type of house | ||||||
Kachha® | ||||||
Semi-pucca | 1.05 | 0.98 | 1.13 | 0.94 | 0.87 | 1.00 |
Pucca | 1.19*** | 1.08 | 1.32 | 1.02 | 0.93 | 1.13 |
Type of family | ||||||
Nuclear® | ||||||
Extended | 1.04 | 0.99 | 1.10 | 1.19*** | 1.13 | 1.26 |
Region | ||||||
North® | ||||||
Central | 0.89* | 0.81 | 0.99 | 1.81*** | 1.64 | 2.01 |
East | 0.92 | 0.84 | 1.02 | 1.18** | 1.07 | 1.31 |
North-east | 1.08 | 0.97 | 1.20 | 3.10*** | 2.79 | 3.45 |
West | 1.17* | 1.04 | 1.32 | 1.63*** | 1.44, | 1.84 |
South | 1.95*** | 1.64 | 2.33 | 2.95*** | 2.46 | 3.52 |
Wealth index | ||||||
Poor® | ||||||
Middle | 1.95*** | 1.76 | 2.16 | 5.15*** | 4.66 | 5.69 |
Rich | 3.79*** | 3.14 | 4.58 | 22.94*** | 19.07 | 27.59 |
. | Partially improved water and sanitation facility . | Improved water and sanitation facilities . | ||||
---|---|---|---|---|---|---|
. | . | 95% CI . | . | 95% CI . | ||
Background characteristics . | AOR . | Lower . | Upper . | AOR . | Lower . | Upper . |
Household headship | ||||||
Male® | ||||||
Female | 1.19*** | 1.11 | 1.29 | 1.17*** | 1.09 | 1.26 |
Age of the HH | ||||||
Below 25® | ||||||
25 + | 0.98 | 0.85 | 1.14 | 1.84*** | 1.58 | 2.14 |
Education | ||||||
Illiterate® | ||||||
Literate | 1.11*** | 1.05 | 1.17 | 1.28*** | 1.22 | 1.36 |
Place of residence | ||||||
Urban® | ||||||
Rural | 0.45*** | 0.39 | 0.51 | 0.18*** | 0.15 | 0.20 |
Religion | ||||||
Hindu® | ||||||
Christian | 0.75*** | 0.69 | 0.82 | 1.07 | 0.99 | 1.17 |
Others | 0.62*** | 0.57 | 0.68 | 0.87*** | 0.79 | 0.94 |
Type of house | ||||||
Kachha® | ||||||
Semi-pucca | 1.05 | 0.98 | 1.13 | 0.94 | 0.87 | 1.00 |
Pucca | 1.19*** | 1.08 | 1.32 | 1.02 | 0.93 | 1.13 |
Type of family | ||||||
Nuclear® | ||||||
Extended | 1.04 | 0.99 | 1.10 | 1.19*** | 1.13 | 1.26 |
Region | ||||||
North® | ||||||
Central | 0.89* | 0.81 | 0.99 | 1.81*** | 1.64 | 2.01 |
East | 0.92 | 0.84 | 1.02 | 1.18** | 1.07 | 1.31 |
North-east | 1.08 | 0.97 | 1.20 | 3.10*** | 2.79 | 3.45 |
West | 1.17* | 1.04 | 1.32 | 1.63*** | 1.44, | 1.84 |
South | 1.95*** | 1.64 | 2.33 | 2.95*** | 2.46 | 3.52 |
Wealth index | ||||||
Poor® | ||||||
Middle | 1.95*** | 1.76 | 2.16 | 5.15*** | 4.66 | 5.69 |
Rich | 3.79*** | 3.14 | 4.58 | 22.94*** | 19.07 | 27.59 |
®Reference category; *, **, *** refer to <0.05, <0.01, and <0.001 level of significance.
Spatial heterogeneity of improved water and sanitation
District-wise availability of improved water and sanitation facilities among the tribal households in India, 2019–2021.
District-wise availability of improved water and sanitation facilities among the tribal households in India, 2019–2021.
Table 3 contains the univariate and bivariate Moran's I statistics depicting the spatial autocorrelation of improved water and sanitation and the district-level predictors. The univariate Moran's I value for improved water and sanitation among the tribal households was 0.518 (p = 0.001), showing significant spatial autocorrelation between districts. Bivariate Moran's I value of being an illiterate head of the household (−0.302), residing in rural areas (−0242) and kachha house (−0.183), Hindu (−0.158), and from a poor economic background (−0.456) showed negative spatial autocorrelation with improved water and sanitation among tribal households. On the other hand, an HH of more than 25 years old and being in a nuclear family showed a positive spatial autocorrelation.
Moran's I statistics showing the spatial dependence for improved water and sanitation and its correlates among the tribal households across the districts of India, 2019–21
District-level covariates (%) . | Univariate . | Bivariate . | ||
---|---|---|---|---|
Moran's I . | p-value . | Moran's I . | p-value . | |
Improved water and sanitation | 0.518 | 0.001 | ||
Illiterate | 0.386 | 0.001 | −0.302 | 0.001 |
Age of the HH_25 + | 0.035 | 0.065 | 0.073 | 0.001 |
Female HH | 0.261 | 0.001 | −0.023 | 0.094 |
Rural | 0.319 | 0.001 | −0.242 | 0.001 |
Kachha | 0.380 | 0.001 | −0.183 | 0.001 |
Nuclear family | 0.187 | 0.001 | 0.044 | 0.006 |
Hindu | 0.726 | 0.001 | −0.158 | 0.001 |
Poor | 0.611 | 0.001 | −0.456 | 0.001 |
District-level covariates (%) . | Univariate . | Bivariate . | ||
---|---|---|---|---|
Moran's I . | p-value . | Moran's I . | p-value . | |
Improved water and sanitation | 0.518 | 0.001 | ||
Illiterate | 0.386 | 0.001 | −0.302 | 0.001 |
Age of the HH_25 + | 0.035 | 0.065 | 0.073 | 0.001 |
Female HH | 0.261 | 0.001 | −0.023 | 0.094 |
Rural | 0.319 | 0.001 | −0.242 | 0.001 |
Kachha | 0.380 | 0.001 | −0.183 | 0.001 |
Nuclear family | 0.187 | 0.001 | 0.044 | 0.006 |
Hindu | 0.726 | 0.001 | −0.158 | 0.001 |
Poor | 0.611 | 0.001 | −0.456 | 0.001 |
HH, household head.
Univariate local Moran's I and LISA cluster map showing clusters of improved water and sanitation facilities (with p-value) among the tribal households across the districts of India, 2019–21.
Univariate local Moran's I and LISA cluster map showing clusters of improved water and sanitation facilities (with p-value) among the tribal households across the districts of India, 2019–21.
Bivariate LISA cluster map showing spatial autocorrelation between improved water and sanitation (WAS) facilities among tribal households and its selected correlates, i.e., (a) illiterate HH, (b) age of the HH ≥25, (c) female-headed household, (d) rural, (e) kachha house, (f) nuclear family, (g) Hindu, and (h) poor households.
Bivariate LISA cluster map showing spatial autocorrelation between improved water and sanitation (WAS) facilities among tribal households and its selected correlates, i.e., (a) illiterate HH, (b) age of the HH ≥25, (c) female-headed household, (d) rural, (e) kachha house, (f) nuclear family, (g) Hindu, and (h) poor households.
Spatial regression model
Table 4 presents the estimated results of the OLS, SLM, and SEM models for improved water and sanitation facilities in the districts of India. The OLS estimation revealed the initial examination of the association between improved water and sanitation facilities and the district-level correlates without considering the spatial structure of the data. On the other hand, spatial auto-regressive models (SLM and SEM) provide the association between predictors and outcome variables by considering the spatial effects. According to OLS estimation, illiterate heads of households, rural background, Hindu religion, and poor wealth status significantly formed a negative association with improved water and sanitation. However, based on the model diagnostics of the spatial models, the SEM appears to be the best-fitted model to the data under study, with the lowest AIC value (5,887.54) compared to every other model. Along with the covariates significantly associated with the OLS model, another new covariate, i.e., kachha house in the SEM model, also became significantly associated with improved water and sanitation facilities. The estimated coefficient was −0.353 (p-value <0.001) for poor households, −0.211 (p-value <0.001) for those households belonging to the rural area, −0.211 (p-value <0.001) for kachha houses, and −0.134 (p-value <0.001) for those households having an illiterate head formed a statistically significant negative association with improved water and sanitation facilities among the tribal households. Thus, the coefficient estimates of poor households confirmed that a 10-point increase in the proportion of poor households was associated with a 3.53-point decrease in the accessibility of improved water and sanitation facilities. Similarly, a 10-point increase in the proportion of kachha houses across the districts was associated with a 2.11-point decrease in the improved water and sanitation accessibility, whereas a 10-point increase in the proportion of illiterate HHs was supposed to bring down the accessibility of improved water and sanitation facilities by 1.34 points. An increasing proportion of Hindu households (10 points) across the districts was also associated with decreasing accessibility of improved water and sanitation (0.73 points). The SEM model suggested a lag coefficient of 0.498 (p-value <0.001), and the corresponding R2 value of the model was 0.613, explaining the better model accuracy.
OLS, SLM, and SEM to assess the association between improved water and sanitation facilities and selected background variables among tribal households in India, 2019–21
District-level covariates (%) . | Improved water and sanitation . | |||||
---|---|---|---|---|---|---|
OLS . | SLM . | SEM . | ||||
Coef. . | p-value . | Coef. . | p-value . | Coef. . | p-value . | |
Illiterate | −0.204 | 0.000 | −0.173 | 0.000 | −0.134 | 0.001 |
Age of the HH_25 + | 0.061 | 0.695 | 0.006 | 0.970 | 0.044 | 0.752 |
Female HH | 0.076 | 0.128 | 0.086 | 0.064 | 0.077 | 0.120 |
Rural | −0.219 | 0.000 | −0.183 | 0.000 | −0.211 | 0.000 |
Kachha | −0.090 | 0.107 | −0.115 | 0.025 | −0.211 | 0.000 |
Nuclear family | −0.007 | 0.872 | −0.042 | 0.300 | −0.045 | 0.277 |
Hindu | −0.089 | 0.000 | −0.071 | 0.000 | −0.073 | 0.004 |
Poor | −0.411 | 0.000 | −0.304 | 0.000 | −0.353 | 0.000 |
Lamda (λ) | 0.498 | 0.000 | ||||
ρ (Rho) | 0.324 | |||||
AIC value | 5,887.54 | 5,809.55 | 5,787.09 | |||
R2 | 0.528 | 0.588 | 0.613 | |||
N | 707 | 707 | 707 |
District-level covariates (%) . | Improved water and sanitation . | |||||
---|---|---|---|---|---|---|
OLS . | SLM . | SEM . | ||||
Coef. . | p-value . | Coef. . | p-value . | Coef. . | p-value . | |
Illiterate | −0.204 | 0.000 | −0.173 | 0.000 | −0.134 | 0.001 |
Age of the HH_25 + | 0.061 | 0.695 | 0.006 | 0.970 | 0.044 | 0.752 |
Female HH | 0.076 | 0.128 | 0.086 | 0.064 | 0.077 | 0.120 |
Rural | −0.219 | 0.000 | −0.183 | 0.000 | −0.211 | 0.000 |
Kachha | −0.090 | 0.107 | −0.115 | 0.025 | −0.211 | 0.000 |
Nuclear family | −0.007 | 0.872 | −0.042 | 0.300 | −0.045 | 0.277 |
Hindu | −0.089 | 0.000 | −0.071 | 0.000 | −0.073 | 0.004 |
Poor | −0.411 | 0.000 | −0.304 | 0.000 | −0.353 | 0.000 |
Lamda (λ) | 0.498 | 0.000 | ||||
ρ (Rho) | 0.324 | |||||
AIC value | 5,887.54 | 5,809.55 | 5,787.09 | |||
R2 | 0.528 | 0.588 | 0.613 | |||
N | 707 | 707 | 707 |
Coef., coefficient; HH, household head; OLS, ordinary least square; SLM, spatial lag model; SEM, spatial error model; AIC, Akaike information criterion.
DISCUSSION
Only about half of ST households had access to improved water and sanitation facilities in India. Spatial heterogeneity across districts is evident, with only one-fourth of tribal households in 25 districts, mainly in Bihar, Jharkhand, Odisha, and Tamil Nadu, having access to improved water and sanitation facilities. The univariate LISA cluster and significance maps revealed 130 districts as hot spots and 131 districts as cold spots in tribal-dominated highland and plateau zones, with 26 districts identified as spatial outliers. The bivariate LISA cluster map reveals that households with heads aged 25 and above, illiterate, Hindu, nuclear, and poor formed significant hot spots with high access to improved water and sanitation facilities, and cold spots with low access. The study found that areas with more rural, poor, kachha, illiterate, and Hindu tribal households had decreased access to improved water and sanitation facilities. Additionally, household headship, age of the HH, family type, and regions were significantly associated with partial and improved facilities.
Our study found that improved sanitation facilities were more likely in tribal households headed by females, similar to prior research highlighting women's crucial role in ensuring safe sanitation facilities' availability (Armah et al. 2018; Indarti et al. 2019). Women's control over household income provides them more opportunities to demonstrate their abilities in the family leadership role (Donacho et al. 2022) and their greater responsibilities within the households associated with high water use (Agbadi et al. 2019). Another possible justification is that in remote areas, especially where the ST population lives, female heads of households ensure that their families have adequate access to water and sanitation facilities to alleviate the burden of fetching water in remote locations (Gaffan et al. 2022) as women typically take on the responsibility of managing WASH (Gurung et al. 2023). The north-eastern region had better access to improved water and sanitation facilities than other regions. A past study also found significant regional and state-level differences in access to toilet facilities – much better in the north-eastern region and worse in the central and eastern regions (Shukla 2020). One conceivable explanation for high access to improved water and sanitation facilities in the north-eastern region could be higher literacy levels among the tribes of that region (RGI 2011). Moreover, some tribal communities in the north-eastern region, like the Khasi, Jaintias, and Garo tribes, practice matriarchy where women have power in resource allocation and exchanges, which could explain the prevalence of high accessibility of water and sanitation facilities in that region (Banerjee et al. 2016).
Tribal households with non-literate heads have less access to improved water and sanitation facilities, which is consistent with previous studies (Okurut et al. 2015; Abubakar 2017) and could be explained by the fact that educated individuals are aware of the drawbacks and benefits of using improved and unimproved water sources or sanitation services (Armah et al. 2018). Our study also found that rural tribal households have decreased access to improved water and sanitation facilities. Compared to urban areas, rural areas are more deficient in water supply and sanitation, while, in urban areas, small and medium towns are more deficient than significant cities. Therefore, sanitation access corresponds to our social and economic hierarchy (Bhagat 2014). Again, kachha tribal households have lower access to improved water and sanitation facilities, consistent with an earlier study that found semi-pucca and pucca households were more likely to have improved water sources (Roy et al. 2023). The reason for this could be that pucca houses (improved houses made with cement and bricks) have better household environment conditions than kachha houses (houses made with poor quality materials for floor, roof, and wall) and that pucca households have better toilet facilities than kachha households (Coffey et al. 2014). Poor households are less likely to have access to improved water and sanitation facilities. The possible explanation might be that wealthier households know about sanitation services and their impact on health (Prakash et al. 2022). Improved sanitation facilities were less likely among the Hindu tribal households. An earlier study (Adukia et al. 2021) emphasized a crucial contrast between a household's religious belief generating disparities in sanitation practices and a household's economic situation influencing differences among religious groups. The possible reason for less access to improved water and sanitation among Hindu tribal households may be related to their orthodox religious belief that having a toilet within the house would be ritually polluting (Jacob 2023). Given Christians' adherence to the doctrine of ‘cleanliness is next to Godliness,’ it is not surprising that they tend to utilize improved water and sanitation facilities more frequently (Immurana et al. 2022).
Another finding of our study is that living in an extended family and a HH over 25 years old are positively auto-correlated with improved water and sanitation. This is consistent with a study that found that increasing the size of a household decreases the likelihood of using an improved water source, implying that household wealth decreases with increasing size (Armand & Fotuè 2013). One possible justification for extended ST households having better access to water and sanitation practices is that their collective resources and support network enable them to prioritize and invest in improved infrastructure and hygiene facilities. Again, it revealed that households with middle-aged and older-aged adult heads were more likely to have access to better sanitation facilities than houses with young adult heads. The possible reason might be that due to their higher economic standing, older people could afford basic services compared to younger people (Armah et al. 2018). In contrast, due to financial constraints, younger people, especially those under 25, may opt for cheap houses that may not have basic water and sanitation facilities (Gaffan et al. 2022).
The main strength of this study is the complete description of the household water and sanitation facilities among the tribal people using data from a recent nationally representative household survey with a robust sampling design. This research is current and relevant for policy and program interventions prioritizing necessary steps and investments in enhancing access to and sustainability of water and sanitation. The spatial heterogeneity identified in the present study can guide better allocation of resources and the implementation of programs to ensure equitable access. However, the cross-sectional methodology of the NFHS-5 restricts any causal relationship between the predictors and outcome variables. The first phase of the NFHS-5 survey, which covered half of the nation, was conducted before COVID-19, and its second phase covered the other half during and after the COVID-19 pandemic when sanitary practices were at their peak. This might have also affected the study's conclusions.
CONCLUSIONS
In India, nearly half of the tribal households have access to improved water and sanitation facilities, and spatial heterogeneity exists across districts. The high–high cluster of improved water and sanitation facilities is primarily concentrated in the north and north-eastern regions, while the low–low cluster is concentrated in the plateau and highland zones of the central and eastern regions of the country. Areas with more rural, poor, kachha, illiterate, and Hindu tribal households have lower access to improved water and sanitation facilities. Tribal households headed by literates, females, and those in the rich wealth category are positively associated with improved water and sanitation facilities. The findings underscore the urgent need for targeted interventions, infrastructure development, and policy interventions to enhance access to improved water and sanitation among tribal households in India. Moreover, efforts toward improved water and sanitation facilities among ST households may contribute to achieving SDG-6 and ensure the welfare of the ST population.
FUNDING
The authors received no financial support from any funding agency, commercial entity, or not-for-profit organization for the research, authorship, and/or publication of this article.
AUTHOR CONTRIBUTIONS
P.K.M. and M.R.P. conceptualized the article, designed the methodology, wrote the article, and reviewed the article. D.S. carried out methodology and statistical analysis, and wrote the article. P.D. did methodology and spatial analysis, and wrote the article.
ETHICS STATEMENT
The study is based on secondary, publicly available survey data that has been identified, and survey agencies that conducted the field survey for the data collection have also collected a prior consent from the respondents. The NFHS was approved by the Institutional Review Board of the Institutions involved, and the datasets are available on registration at https://www.dhsprogram.com/data/new-user-registration.cfm for broader use in social research. The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008. They ruled that no formal ethical consent was required to conduct research from this data source.
DATA AVAILABILITY STATEMENT
All relevant data are available from https://www. dhsprogram.com/data.
CONFLICT OF INTEREST
The authors declare there is no conflict.