NEW UPDATE: Completed trials may now upload and register supplementary documents (e.g. null results reports, populated pre-analysis plans, or post-trial results reports) in the Post Trial section under Reports, Papers, & Other Materials.
The role of informational and credit constraints in sanitation uptake
Initial registration date
August 04, 2017
October 31, 2017 10:33 AM EDT
University of Kent
Other Primary Investigator(s)
University College London and Institute for Fiscal Studies
Institute for Fiscal Studies
Institute for Fiscal Studies
Additional Trial Information
Achieving the United Nation’s ambitious target of eliminating open defecation by 2025 demands a thorough understanding of binding constraints to toilet uptake by households. In particular, adoption of safe toilets by poor households is thought to be restricted by credit and informational constraints. Safe toilets entail large investments, accounting for several months of household income for even basic low-cost options. Moreover, households may be unaware of the costs and benefits of safe sanitation. In this study, we consider the impacts of alleviating these constraints on the uptake of safe sanitation in rural India. We do so using a cluster randomised control trial in rural Maharashtra, where we randomly allocate 120 Gram Panchayats (GPs) to one of three arms: (i) access to micro-credit loans for sanitation; (ii) access to micro-credit loans for sanitation, and awareness creation activities; and (iii) a control arm where neither of these interventions was offered. Micro-credit loans could be accessed by the clients of our partner microfinance institution only. In addition to the uptake of safe sanitation, we will also consider whether the interventions improve usage of safe sanitation, and toilet quality, whether they change perceptions of costs and benefits of safe sanitation, and whether microcredit for sanitation crowds out investments on other dimensions, e.g. business. Finally, the design will also allow us to shed light on whether poor households are credit constrained, and not simply credit rationed, for sanitation investments.
We evaluate the effectiveness of two interventions: (i) access to micro-credit for sanitation; and (ii) sanitation credit combined with awareness creation activities.
The credit intervention provides access to collateral-free, joint liability micro-credit loans for sanitation to clients of a partner microfinance institution. Only individuals who have been clients of the microfinance institution for at least 1 year are eligible to apply for a sanitation loan. The loan covers a maximum amount of Rs. 15000, charging a 22% interest rate annually at a declining balance over a 2-year repayment period. The loans are intended to be used to construct a new toilet or to repair an existing toilet. The awareness creation activities are delivered through an NGO and disseminate information on sanitation issues to staff and clients of the microfinance institution, and to the broader community through community-wide activities. Clients of the microfinance institution receive information sessions during their weekly meetings with the loan officer, and were also invited to attend half-day branch-level workshops (Note that each branch covers a radius of around 25 km from the branch office). The community-wide activities included street plays and wall paintings, as well as engagement with community leaders to gain support for their activities and to strengthen awareness of sanitation and hygiene issues, and training of masons on how to construct different low-cost models of safe toilets. The information sessions, wall paintings and street plays disseminated the benefits of safe sanitation on health and non-health dimensions (e.g. safety of females); as well as on the availability and costs of different types of low-cost safe toilets.
Intervention Start Date
Intervention End Date
Primary Outcomes (end points)
The primary outcome will be safe toilet uptake (separately among clients, and non-clients of the microfinance institution).
Primary Outcomes (explanation)
A safe toilet will be defined to be an improved facility for which excreta is safely disposed of in situ or off-site.
Toilet quality will be measured using an index combining household responses to materials used to construct the toilet, and surveyor observations of the toilet. An indicator for a good quality toilet will also be constructed from these variables.
Toilet usage will be measured from questions asking where different groups of household members (e.g. male adults) defecate. They will be coded as using a safe toilet if they use a facility considered to be safe according to the definition above. Business profits will be calculated as the difference between revenues and expenditures on the two main enterprises of the household
Secondary Outcomes (end points)
Secondary outcomes will be toilet quality, toilet usage and open defecation (separately among clients and non-clients of the microfinance institution).
Other outcomes we will study are effects on total household credit, sanitation credit, business investments and profits, education investments, perceptions of costs and benefits of toilets, and uptake of a subsidy for toilet construction offered by the Government of India's Swaccha Bharat Mission policy. All analyses will be conducted separately for clients and non-clients of the microfinance institution.
Secondary Outcomes (explanation)
Toilet quality will be measured as an index combining household reports and surveyor observations of the toilet.
Toilet usage and open defecation at the household level will be measured as a weighted average of these behaviours across different age-gender groups within the household.
Total household credit will be calculated as the sum of the three largest loans of value > INR 500 taken by the household since the intervention began; and the three largest loans of value < INR 500 taken by households.
Sanitation credit will be measured as a binary variable, and by the value of credit for sanitation purposes
Business investments will be measured as a binary variable for any such investment since baseline
The study covers two districts in the Indian state of Maharashtra.Within these districts, we cover 120 gram panchayats (GPs), the smallest administrative unit by the Government of India.They have been selected according to two main criteria: (i) they should fall
within currently active operational areas and (ii) neither sanitation loans nor health insurance products should have been offered at any point in time to community members by our implementing partner. 120 GPs were selected from 133 eligible GPs to be part of the study. The 120 GPs were then assigned to one of the 3 study arms (2 treatment, 1 control) through stratified random sampling, with strata defined according to the district and whether the village was large (> 480 households).
For the baseline survey, we randomly sampled around 30 households per GP stratified by (i) whether a client of our implementing partner ( Grameen Koota) lives in the household and (ii) whether a child under the age of two years lives in the household. The intervention design establishes treatment being clustered at GP level, randomly assigning GPs to different treatment groups, i.e. control, GK (provision of sanitation loans by Grameen Koota) and 'GK + ND' (provision of sanitation loans by Grameen Koota + awareness creation activities by an NGO - Navya Disha). For the endline survey, we re-survey the 3600 households sampled at the baseline (1800 clients of GK, and 1800 non-clients), and supplement this sample with a further 2400 clients sampled from women who were clients at baseline in the 120 study communities.
Experimental Design Details
Further details on the trial design and interventions can be found here: https://www.ifs.org.uk/publications/7957
Randomisation done in office by a computer using Stata 13
Was the treatment clustered?
Sample size: planned number of clusters
120 Gram Panchayats
Sample size: planned number of observations
The evaluation sample will consist of 4,200 households who had a member who was a client of the microfinance institution before baseline (1,800 interviewed at baseline and endline, and 2,400 at endline only), and 1,800 households without a member who is a client of the microfinance institution before baseline (interviewed at baseline and endline).
The panel sample of 3,600 households was also stratified by the presence of children aged 0-2 years at baseline.
Sample size (or number of clusters) by treatment arms
41 Gram Panchayats (GPs) control,
40 GPs sanitation loans only,
39 GPs sanitation loans + awareness creation activities
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
We provide minimum detectable effect sizes for uptake of safe toilets among the whole endline sample, client households only, and non-client households only for a power of 0.8, a significance level of 0.05 and assuming two-sided tests. We also assume attrition of around 10% of the panel sample, evenly distributed among client and non-client households.
For the full sample, with 47 households per cluster on average at endline, and assuming a intra-class correlation of 0.15, this design will allow us to detect increases in safe toilet uptake of around 12.87 percentage points. For clients only, with an ICC of 0.15, the minimum detectable effect size will be 12.42 percentage points, while that for non-clients will be 14.2 percentage points. Inclusion of controls such as the education of the household head greatly reduces the ICC, and allows us to detect smaller effect sizes.
INSTITUTIONAL REVIEW BOARDS (IRBs)
Institute of Sustainable Development Institutional Review Board
IRB Approval Date
UCL Research Ethics Committee
IRB Approval Date
IRB Approval Number
Post Trial Information
Is the intervention completed?
Is data collection complete?