The Household Water Insecurity Experiences (HWISE) Scale: comparison scores from 27 sites in 22 countries

Household-level water insecurity – the inability to access and benefit from adequate, reliable, and safe water for well-being and a healthy life – affects billions of people globally, but until recently, there were few metrics that could facilitate its household-level monitoring and evaluation (Jepson et al. 2017; Wutich et al. 2017). Freshwater availability has traditionally been reported at the community, watershed, or national level using resource-based metrics, but advances in water insecurity measurement have led to the creation of at least 13 different scales that measure household water experiences with availability, accessibility, reliability, and use, in contrast with at least 67 resource-based metrics that assess freshwater availability at larger geographic scales (Octavianti & Staddon 2021). These experiential metrics are particularly crucial for understanding human adaptation to natural resource stressors associated with climate change (Maja & Ayano 2021).

The Household Water Insecurity Experiences Research Coordination Network (HWISE-RCN; www.hwise-rcn.org) was formed in 2018 to promote scholarship and practice related to mitigating household-level water insecurity. Investigators collected data from 27 sites in low- and middle-income countries between 2017 and 2018 as part of a larger project that created and validated a cross-culturally comparable household water insecurity scale (Young et al. 2019b). This 12-item experiential scale, known as the HWISE scale or HWISE-12 (Young et al. 2019a), has received interest throughout the international development communities as a monitoring and evaluation tool that can complement the household-level metrics produced by the World Health Organization's (WHO) Joint Monitoring Programme for Water Supply, Sanitation and Hygiene (JMP) (Slaymaker et al. 2020). A shortened 4-item version of the scale (the HWISE-4) was recently validated to facilitate rapid deployment in field survey settings with constrained resources (Young et al. 2021).

Several HWISE-RCN-affiliated studies have focused on how these new metrics are associated with water governance (Miller et al. 2020), water sharing (Rosinger et al. 2020), financial expenditures on water (Stoler et al. 2020), injuries (Venkataramanan et al. 2020), and self-reported health (Jepson et al. 2021). Research on the measurement of water insecurity, however, has not fully considered how the HWISE scale varies within and across sites. For instance, while studies have demonstrated the validity of the HWISE scale and its relationship with economic, health, and social outcomes, none have provided site-level scores. This brief provides a site-wise summary of the different HWISE scale scores as points of comparison for future research that uses these metrics to study household water insecurity – and better understand these scales' utility – in diverse global settings.

Cross-sectional survey data were collected from 7,709 households in 27 sites across 22 countries in two waves. Sites were selected through existing professional networks to maximize variation in local climate, water infrastructure, and typical water problems (Young et al. 2019b). Most sites targeted approximately 250 households and used simple random sampling to select households, with four exceptions (purposive sampling in Singida, Tanzania; Kampala, Uganda; and Upolu, Samoa; parallel assignment in Pune, India) for ongoing studies that required different sampling strategies to achieve their research objectives. Adults were eligible respondents if they reported being ‘knowledgeable about their household's water situation’ (Young et al. 2019a). The HWISE survey provided a much more detailed view of household water insecurity than other standard household surveys, such as Demographic and Health Surveys or Multiple Indicator Cluster Surveys, because it also probed experiences that demonstrate the consequences of inadequate, unreliable, or unsafe water. All participants provided verbal or written informed consent in the respective local language, and all study activities were reviewed and approved by the appropriate ethical review boards (Young et al. 2019b).

Enumerators used paper- and tablet-based surveys to collect data on sociodemographic characteristics and experiences with water availability, accessibility, reliability, and use, which are core components of household water insecurity (Jepson et al. 2017). Table 1 presents the composition of the HWISE-12, HWISE-11, and HWISE-4 scales with the full wording of each item. Each item reflects an experience related to water adequacy (having sufficient quantity for drinking and household consumption), reliability (the availability of water when needed), safety (water that is fit-for-use, such as drinking or bathing), or psychosocial experiences related to these problems.

Each survey item elicited the frequency of household experiences related to water in 4 weeks prior to the survey and categorized responses as: never (0 times) scored as ‘0’, rarely (1–2 times) scored as ‘1’, sometimes (3–10 times) scored as ‘2’, and often (11–20 times) or always (more than 20 times) which were combined and scored as ‘3’. The HWISE-12 is calculated by summing the scores of 12 items, yielding a range of 0–36 (Young et al. 2019a), and the HWISE-4 is the sum of four items from the HWISE-12 with a range of 0–12 (Young et al. 2021). One HWISE-12 item, related to experiencing shame about one's water situation, was introduced in late-2017 during the second wave of data collection. As a result, the HWISE scale in study sites surveyed during the first wave has typically been represented as the HWISE-11 without the shame item (range 0–33); the HWISE-11 accounted for 99.3% of the variation in HWISE-12 scores with minimal additional error (Stoler et al. 2020).

This research brief presents the mean, standard deviation, and additional descriptive information of 27 sites for (1) HWISE-12 (where available), (2) HWISE-11, (3) HWISE-4, and (4) the Household Food Insecurity Access Scale (HFIAS, range 0–27) (Coates et al. 2007). We provide HFIAS scores for additional context given the wide availability of HFIAS reference data globally and its cross-cultural applicability, and because HFIAS and HWISE scores have been associated with each other in prior studies (Young et al. 2019a; Stoler et al. 2020).

Table 2 presents the four indicators as well as season (wet, dry, both, or neither) and sample size information. There were 3,293 households from 13 sites with complete HWISE-12 data (i.e., the 11 sites used to derive HWISE-12 in Pune, India, and Dhaka and Chakaria, Bangladesh), yielding an aggregate mean score of 9.32 (standard deviation [SD] = 8.81). Site-level mean HWISE-12 scores ranged from 1.64 (SD = 4.22) in Pune, India, to 20.90 (SD = 7.50) in Cartagena, Colombia. HWISE-12 can offer a more nuanced view of water insecurity than WHO JMP measures of drinking water service level, which do not consider sufficiency for all household uses. For instance, the HWISE-12 identified water insecurity even among households with ‘basic’ water services, as classified by the JMP drinking water service ladder (Young et al. 2019a).

Because we could only compute HWISE-12 for 13 sites, Table 1 also contains HWISE-11 estimates for all 27 sites. Among 6,484 households, the aggregated mean HWISE-11 score was 6.95 (SD = 7.50). Site-mean HWISE-11 scores ranged from 1.54 (3.78) in Pune, India, to 19.56 (5.65) in Punjab, Pakistan.

HWISE-4 scores could be computed for 7,351 households across the 27 sites, yielding an aggregate mean score of 2.84 (SD = 3.08). Site-mean HWISE-4 scores ranged from 0.51 (SD = 1.50) in Pune, India, to 8.21 (SD = 2.55) in Punjab, Pakistan.

We present a bubble plot of the mean site scores in Figure 1, and the frequency of affirmation for each item (never, rarely, sometimes, or often) by site in Supplementary Material, Figure 1, to aid interpretation of the variation within and between sites that is summarized by the means and standard deviations in Table 2. HWISE scale scores exhibited greater variation in sites surveyed during the rainy season, perhaps because household-level characteristics (e.g., differences in wealth and financial access to water storage technologies) modified households' ability to take advantage of relatively higher water availability. Sites surveyed during the dry season tended to have higher HWISE scale scores as expected, and less variability, indicating that seasonal decreases in water availability may have affected households more uniformly. Although the sites with the five highest scores across all HWISE measures were surveyed in dry conditions, household water insecurity is shaped by local context as well. For example, the marginalized community where the survey was implemented in Cartagena was, in 2018, awaiting a long-delayed piped water service expansion by the municipal water authority. The bar chart for Cartagena in Supplementary Material, Figure 1 reveals the frustration that characterized this community's water insecurity: nearly 75% of respondents reported ‘worry’ or ‘anger’ about their water situation often or always, by far the highest prevalence of these experiences among all sites.

Finally, we computed HFIAS scores for 7,077 households across 26 sites (HFIAS was not implemented in Upolu, Samoa), yielding an aggregate mean score of 6.10 (SD = 6.58). Site-mean HFIAS scores ranged from 1.03 (SD = 2.55) in Kathmandu, Nepal, to 16.08 (SD = 8.06) in Gressier, Haiti. The overall range was typical of HFIAS scores observed in similar low- and middle-income settings (e.g., De Cock et al. 2013; Roba et al. 2019).

This research brief presents site-level HWISE-12, HWISE-11, HWISE-4, and HFIAS scores for 27 study sites across 22 countries. Overall, there was substantial variability within and between sites, as well as by season, reflecting the many ways in which households or communities experience and adapt to water insecurity (Jepson et al. 2017). These scores provide comparison data for future studies that use, adapt, or improve these metrics for novel contexts. We hope that future deployment of the HWISE scale will establish test–retest or inter-rater reliability, attempt to validate a version for high-income countries, or assess its utility in dynamic scenarios such as disaster recovery. The scale is limited in that it does not illuminate differences in the underlying causes of water insecurity, which may be individual in nature (e.g., the elderly being unable to lift large containers of water) or structural (e.g., exclusion due to gender or ethnic discrimination) (Slaymaker et al. 2020), or convey the degree of household disruption from a given water insecurity experience. Because the HWISE-12 and the HWISE-4 were intended to represent universal, but not comprehensive, experiences of household water insecurity, studies may benefit from including additional complementary metrics of water insecurity sub-domains that are not necessarily applicable to all households in a given community, such as experiences with water conflict, effects on children, or water used for agricultural livelihoods. Nonetheless, the ability to rapidly screen communities cross-culturally makes the HWISE scale an important tool that can help public health practitioners track progress toward achievement of Sustainable Development Goal 6 ‘clean water and sanitation for all’.

We acknowledge the support of the Household Water Insecurity Experiences Research Coordination Network (HWISE-RCN) funded by the National Science Foundation under grant no. BCS-1759972. We also thank Hala Ghattas, Hugo Melgar-Quiñonez, and Nathaly Triviño for additional support of this project. The HWISE study was funded with the Competitive Research Grants to Develop Innovative Methods and Metrics for Agriculture and Nutrition Actions (IMMANA). IMMANA is funded with UK Aid from the UK government. This project was also supported by the Buffett Institute for Global Studies and the Center for Water Research at Northwestern University; Arizona State University's Center for Global Health at the School of Human Evolution and Social Change and Decision Center for a Desert City (National Science Foundation, No. SES-1462086); and the Office of the Vice Provost for Research of the University of Miami. S.L.Y. was supported by the National Institutes of Health (Nos NIMH R21 MH108444 and NIMH K01 MH098902). W.E.J. was supported by the National Science Foundation (No. BCS-1560962) and the Texas A&M University-CONACYT Research Collaborative Grant. C.S. was supported by the Lloyd's Register Foundation. M.C.F. was supported by the World Bank Strategic Impact Evaluation Fund (Award No. 7175829). Funders of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report.

J.S. is involved in the conceptualization, formal analysis, investigation, writing – original draft, and writing – review and editing. J.D.M. is involved in formal analysis, investigation, and writing – review and editing. All other authors are involved in the investigation and writing – review and editing.

This research does not contain any conflict of interest.

Data cannot be made publicly available; readers should contact the corresponding author for details.