x

Please fill out this short user survey of only 3 questions in order to help us improve the site. We appreciate your feedback!
Back to History Current Version
Market Structuring of Sludge Management for the Benefit of Vulnerable Households in Dakar (Demand-side Trial)
Last registered on August 12, 2014

Pre-Trial

Trial Information
General Information
Title
Market Structuring of Sludge Management for the Benefit of Vulnerable Households in Dakar (Demand-side Trial)
RCT ID
AEARCTR-0000344
Initial registration date
August 12, 2014
Last updated
August 12, 2014 6:02 PM EDT
Location(s)
Region
Primary Investigator
Affiliation
University of Virginia
Other Primary Investigator(s)
PI Affiliation
University of Pennsylvania
PI Affiliation
University of Virginia
PI Affiliation
University of Wisconsin-Madison
Additional Trial Information
Status
On going
Start date
2013-08-28
End date
2015-06-30
Secondary IDs
Abstract
Poor sanitation is an important cause of childhood diarrhea, which often leads to child mortality. We will study how to increase demand and decrease prices for an improved sanitation technology, mechanical desludging. On the demand side, we will measure the effects of social and behavioral factors (social pressure, learning, and procrastination) on household demand. In a related trial focused on the supply side, we will study the effect of different auction mechanisms on collusion and prices paid by consumers.
Registration Citation
Citation
Houde, Jean et al. 2014. "Market Structuring of Sludge Management for the Benefit of Vulnerable Households in Dakar (Demand-side Trial)." AEA RCT Registry. August 12. https://doi.org/10.1257/rct.344-2.0.
Former Citation
Houde, Jean et al. 2014. "Market Structuring of Sludge Management for the Benefit of Vulnerable Households in Dakar (Demand-side Trial)." AEA RCT Registry. August 12. http://www.socialscienceregistry.org/trials/344/history/2382.
Sponsors & Partners

There are documents in this trial unavailable to the public. Use the button below to request access to this information.

Request Information
Experimental Details
Interventions
Intervention(s)
This project seeks to identify methods of increasing the use of mechanized desludging by increasing household willingness to pay for improved sanitation services. We investigate two broad mechanisms through which willingness to pay may change: social network impacts and liquidity. We test the impact of several social network interventions that encourage take-up through learning and social pressure, and the impact of payment and deposit interventions which affect the liquidity of the households.

The treatment group is comprised of four thousand households, grouped into 400 neighborhoods of 10 households each. A further 800 households (two per neighborhood) will be surveyed to evaluate the effects on those nearby who were not offered the subsidized mechanized desludgings.
Four thousand households will be randomly assigned to receive a subsidized price of either $34 or $48 for mechanized desludging services, which normally cost around $50. The subsidy is valid for a maximum of 2 desludgings over a nine month period. To measure learning from doing, researchers will see whether those who received subsidized services continue using mechanical desludging after the discount ends.

To understand the extent to which social pressure influences use of mechanized desludging, the attribution of the discounts will be made public for half of the clusters through the distribution of discount lists, and for the other half it will be offered privately. The impacts of learning from others and coordination will be measured when 1000 randomly chosen households in 200 clusters are first told either how many or specifically which of their neighbors have signed up, and are then given the opportunity to sign up.

A number of payment structures will also be tested. Eighty-seven percent of households (3500) will be asked to leave a deposit at the time of the survey if they would like to sign up for the subsidized mechanized desludging. The remaining 500 treatment households will receive a subsidy for mechanical desludging, but will not be asked to leave a deposit. One third of households will be asked to pay the remainder at the time of service, one third will be given a savings account earmarked for desludging and billed monthly, and one third will be given the same earmarked savings account, but allowed to contribute whenever they wish. Half of households will also be offered a general, non-earmarked savings account. Varying the frequency of payments and savings options will test the relative importance of commitments and mental accounting to encourage payment and usage.
Intervention Start Date
2014-02-10
Intervention End Date
2015-06-10
Primary Outcomes
Primary Outcomes (end points)
prices paid for mechanical desludging, takeup of mechanical desludging, household health, willingness to pay for desludging services, savings account balances, follow-through on plans to use mechanical desludging
Primary Outcomes (explanation)
Willingness to pay is measured as the takeup rate (or follow-through rate) of mechanical desludging at different prices (subsidy levels). Our project will also have access to data from a related study that collects auction data from a call center that matches clients and desludgers. We will observe if callers accepted or rejected the auction winner's price for different bids.

Health outcomes will be measured using diarrhea rates through baseline and endline survey data, as compared to incidence of cold or cough symptoms. Sanitation will affect diarrhea rates, while respiratory problems (cold/cough) should see no effect.

Follow-through rate is important as realized take up is key to understanding the true effect of the treatment on willingness to pay. We will compare the commitment to take up the improved sanitation service and the realized take up at the end of the study.
Secondary Outcomes
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
Experimental Design
Experimental Design
We use a randomized controlled trial (RCT) to investigate social effects on the adoption of mechanized desludging. First, we will offer randomly selected neighborhoods subsidized desludging services coupled with various social pressure treatments to measure direct social effects.

We will construct 400 groups of twelve neighboring households each. These neighborhoods will be far enough apart that, in general, their sanitation decisions will not affect one another and the households will not know one another. In our 400 selected neighborhoods, each neighborhood will include 10 randomly chosen treated households and 2 untreated households; these 2 untreated households will allow us to measure indirect spillovers.

We will first conduct a baseline survey of demographic information, including household composition, education, health, membership and participation in associations and cooperatives, and savings habits. We will also collect GPS data on the locations of the households.

We then conduct a second survey on willingness to pay for improved sanitation with the household member who is in charge of making decisions regarding desludging ("the decider"). The decider survey will cover savings and loans, wealth and durable assets, brief questions on income and spending, sanitation practices, and social networks. The social network component of the survey will include questions asking who in their neighborhood they talk with about waste disposal, who they would choose to lead a neighborhood sensitization on health, who is a member of the same association or cooperative as them, from whom they would borrow or to whom they would lend money, who they did borrow from or lend to within the past year, with whom they are related, which households use mechanized desludging, and where each household dumps its sludge.

At the time of the second survey, in the 10 treatment households, we will offer the decider his randomly assigned treatment. There are several treatment arms (please see experimental design) but one main treatment involves randomizing discounts of different sizes to households that sign up for a subscription of two desludgings, and randomizing whether this discount is private or public information.

At the end of the decider survey, households in the "deposit" group which would like to sign up for the subscription will be asked to pay a deposit of roughly US$6 – an amount equal to the respondent's participation gift. The deposit will be credited towards the second desludging, and will be unavailable to them until the end of the nine-month subsidy period. After the subsidy period ends, they will have the option to continue using the subscription for an unsubsidized third desludging depending on the interest of the Senegalese Ministry of Sanitation in continuing the program following the main research period.

After the year in which treated households have access to subsidized desludgings, we will re-interview both the 4000 treated households and the 800 untreated households, allowing us to measure their sanitation practices and relationships with neighbors.
Experimental Design Details
Not available
Randomization Method
We use cluster randomization across neighborhoods in Dakar, with each cluster and household assigned to treatment through randomization in Matlab.
Randomization Unit
Cluster-level randomized treatments:
Public / private treatment: the values of neighbors' subsidy prices are shared with others in their cluster in half of the clusters in order to observe the effect of pressure from the neighbors on the take up of desludging services.
Learning from others treatment: half of neighbors are told who (half are told how many) of their neighbors chose to take up the desludging subscription.

Individual-level randomized treatments:
Subsidy levels: subsidy levels are randomized at the household level, and the number of high versus low subsidies varies across the different clusters.
Payment frequency: households receive reminders to save for their desludgings, and the frequency with which they are asked to pay in the reminders varies between at will, monthly, and payment at the time of service.
Earmarking treatment: some households are offered one earmarked desludging account while others are offered two accounts (one earmarked and one general savings account)
Commitment deposit treatment: 87% of households are asked to pay a deposit toward a subsidized desludging, a randomly selected 13% are not asked to pay a deposit.
Spillover estimation: 2 households per cluster are selected to be surveyed but not receive the treatment or the subsidies. We observe whether these households are more likely to take up when there are more households with high discounts in their neighborhood.
Was the treatment clustered?
Yes
Experiment Characteristics
Sample size: planned number of clusters
400 neighborhood clusters
Sample size: planned number of observations
4800 households
Sample size (or number of clusters) by treatment arms
Cluster-level randomized treatments:
Public / private treatment: 200 public clusters (2000 households) vs. 200 private clusters (2000 households)
Learning from others treatment: 100 number clusters (1000 households) vs. 100 name clusters (1000 households) vs. 200 no info clusters (2000 households)

Individual-level randomized treatments:
Subsidy levels: 2000 high vs. 2000 low subsidies
Commitment deposit treatment: 3500 deposit households vs. 500 no-deposit households
Earmarking treatment: 1000 single account households vs. 1000 two accounts households vs. 2000 no account households
Households receiving at least one account are enrolled in one of three payment plan options: Payment frequency: 666 at will vs. 667 monthly vs. 667 time of service
Spillover estimation: 4000 households receiving subsidy vs. 800 spillover households (2 per cluster)
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Cluster-level randomized treatments (the public / private treatment and the learning from others treatment), as well as the individually randomized earmarking treatment, have an MDES of 0.17. The commitment deposit treatment has an MDES of 0.14, and the spillover estimation has an MDES of 0.13.
Supporting Documents and Materials

There are documents in this trial unavailable to the public. Use the button below to request access to this information.

Request Information
IRB
INSTITUTIONAL REVIEW BOARDS (IRBs)
IRB Name
Comite National d'Ethique pour la Recherche en Sante (Senegal)
IRB Approval Date
2012-02-01
IRB Approval Number
000227
IRB Name
Innovations for Poverty Action
IRB Approval Date
2012-09-14
IRB Approval Number
574
IRB Name
University of Virginia
IRB Approval Date
2013-08-23
IRB Approval Number
2012-0272-00