Dear Shyamal Chowdhury,
Thanks for sharing this interesting study. It important to understand how to address gaps in sanitation in particularly vulnerable groups such as migrants that may be overlooked in traditional assessments of household sanitation. I was wondering what the implications of the second experiment are? Given that migrants were much more likely to practice OD when returning home after migration to small towns where OD was high, this seems like an important finding with rapid growth of these areas. How can this be addressed by sanitation promotion programmes?

I also have a couple questions to better understand what was done. The data collection process for this study seems challenging, how did you track behaviors of returning migrants? In addition, why do you think that the migration incentive groups were much more like to OD? I am unclear if the control group were other migrants that you also tracked, or other people that stayed at home.

Thanks in advance,
Sarah

I have been corresponding with Shyamal Chowdhury from the University of Sydney about a sanitation grant by the Gates Foundation which he is co-leading and which ended recently after two years duration. He has asked me to post the following information about this project:

Title of grant: Does sanitation behavior migrate? Evidence from seasonal migrant workers in Bangladesh

• Name of lead organization: Innovations for Poverty Action (IPA)
• Primary contact at lead organization: A. Mushfiq Mobarak (Co-researchers: Shyamal Chowdhury, Senior Lecturer at School of Economics, The University of Sydney, Australia) and Raymond Guiteras)
• Grantee location: IPA, New Haven, CT 06510, USA
• Developing country where the research is being or will be tested: Bangladesh
• Start and end date: July 2013- June 2015
• Grant type: Other
• Grant size in USD: 396,655 (as per grant database here )

Short description of the project:
The researchers expanded on two existing randomized controlled trials in Bangladesh – which had previously been implemented - to study the sanitation behavior of season migrants and their families (several thousand people were included in the trial).

Seasonal migration could be an important tool to address seasonal poverty documented in many countries such as Bangladesh (Bryan, Chowdhury and Mobarak, 2014), Ethiopia (Dercon and Krishnan, 2000), Mozambique and Malawi (Brune, Gine, Goldberg, and Yang, 2011), Madagascar (Dostie, Haggblade, and Randriamamonjy, 2002), Kenya (Swift, 1989), Indonesia (Basu and Wong, 2012), Thailand (Paxson, 1993), India (Chaudhuri and Paxson, 2002), and inland China (Jalan and Ravallion 2001).

Project #1: Effects of Sanitation Interventions on Sanitation Decisions during Migration
In the first experiment, the researchers randomly assigned rural villages to sanitation campaigns, and tracked the sanitation behaviors of migrants from those villages when they moved to cities and towns.

The interventions included:
i. Latrine Promotion Program (LPP) – A CLTS-style community awareness program that teaches the importance of proper sanitation and charts the status of a community’s current sanitation practices.
ii. Subsidies – 50% subsidies for latrine purchase randomly assigned with the poorest 75% of the population
iii. Supply – Selected and trained supply agents acted as liaisons between latrine parts dealers and provided information to households on how to install and maintain a latrine.

See also here on the forum where a publication from this project was discussed ("Encouraging sanitation investment in the developing world: A cluster-randomized trial"): forum.susana.org/forum/categories/5-clts...ve-without-subsidies

Project #2: Effects of Migration on the Sanitation Behavior of Returning Migrants
In the second experiment, the researchers randomly assigned rural residents incentives to seasonally migrate, which encouraged these residents to move to cities and towns for work during the pre-harvest lean season (the term “lean” is often used to characterize the particular season that we are referring to). The researchers tracked the sanitation behaviors of returning migrants, both at their destination and at their origin.

The interventions included:
i. Migration Incentive – An 8.50 USD incentive (covering a round-trip bus ticket & food) to households to temporarily out-migrate during the pre-harvest lean season
ii. Destination Assignments – Random assignment to one of four specific migration destinations (a big city, versus smaller towns offering agricultural work)

Goal(s):
The research contributes to our understanding of migrant workers hygiene and sanitation behavior at origins and destinations, and if migrant works sort into groups with similar hygiene practices. Our data will also facilitate the evaluation of large scale sanitation interventions that a number of large NGOs including BRAC are currently undertaking in rural areas of Bangladesh.

Objectives:
In this study, researchers seek to answer two related questions: do rural sanitation campaigns impact migrant (worker) behaviors when they move to cities and towns? Conversely, do migrant workers bring back sanitation habits acquired in cities and towns to their home villages?

Research or implementation partners:
Innovations for Poverty Action (IPA) Bangladesh hosts the project and was responsible for data collection and data integrity, and to ensure that the project is run properly. RDRS Bangladesh, a local non-profit organization was responsible for implementing interventions.

Results to date:
Project #1: Effects of Sanitation Interventions on Sanitation Decisions during Migration

1. While LPP alone did not decrease open defecation among migrants, LPP in combination with the other marketing treatments lead to decreased Open Defecation (OD) rates among migrants. LPP along with the subsidy treatment decreased OD rates among migrants by 11.5% and LPP along with both the subsidy and supply treatments decreased OD rates among migrants by 10.9%, representing 45.5% and 43.1% decreases in OD compared to migrants in control households, respectively.
2. Using the proportion of increased latrine ownership that the researchers can attribute to the interventions, migrants from households who invested in latrines are less likely to practice OD during migration. Assuming linearity in these results, a 10% increase in latrine ownership leads to a 5.8% reduction in OD, representing a 22.9% decrease in OD compared to migrants in control households.

Project #2: Effects of Migration on the Sanitation Behavior of Returning Migrants

1. Adults in a household with an individual induced to migrate by the incentive are 44 percentage points more likely to OD, making them over two times more likely to OD than adults in households who did not receive this treatment.
2. Furthermore, this effect is driven by households with migrants induced to migrate to high-OD areas such as small towns (OD-rate: 40.2%) as opposed to urban areas (OD rate: 20.6%). Adults in households with an individual induced to migrate to a high-OD area are 64.7 percentage points more likely to OD, which is roughly a three-fold increase over the OD propensity of adults in control households.
3. We find no distinguishable differences in adult OD rates in between households with an individual induced to migrate to a low-OD urban area and control households.

Links, further readings:
Documents in SuSanA library: www.susana.org/en/resources/library/details/2247
(contains one poster so far)

Documents from a related project called “Bolstering demand for improved sanitation when adoption decisions across households are inter-linked“: www.susana.org/en/resources/library/details/2246

Forum discussion for that related project: forum.susana.org/forum/categories/5-clts...ve-without-subsidies

Two-page summary in Hanoi Convening report by the BMGF in July 2015:



Current state of affairs: The researchers are working on the final project report .

Biggest successes so far: Field works have been completed and data have been collected and analyzed.

Main challenges / frustration: Due to the pro-longed political unrest and violence in Bangladesh since 2014 general election, researchers’ visits and field works including surveys had to be postponed in multiple occasions .