|
Field
Trial Title
|
Before
Socio-Economic and Behavioral Effects of Improved
Urban Drainage in Bangladesh
|
After
Fostering Climate Resilience: Socio-Economic Effects of Improved Urban Drainage in Bangladesh
|
|
Field
Abstract
|
Before
We study the effect of improvements in urban drainage infrastructure for flood prevention on affected households with a survey in Barishal, Bangladesh. Specifically, our project analyzes the socio-economic and behavioral effects of drainage improvements on affected households. Improvements in drainage systems are a key component for flood control and climate change adaptation in many urban settings around the globe. We will use a spatial regression discontinuity design with the distance to the boundaries of the area benefiting from the project as a running variable. We will complement this analysis with a grid-cell level analysis on the likelihood and length of experiencing a flooding event during the rainy season using satellite imagery as an objective measure of risk exposure.
|
After
This study evaluates the socio-economic impact of urban drainage infrastructure improvements as a climate resilience measure in Barishal, Bangladesh, a city where waterlogging poses severe risks to urban livelihoods. We measure the effects by collecting a household survey immediately following the rainy season when benefits are expected to materialize. Leveraging high-resolution elevation data, we develop a two-dimensional flood hazard model to estimate household-specific changes in the duration of waterlogging induced by the intervention. We show that most of the expected benefits are not local, but are experienced as indirect network effects by households not directly targeted by the intervention. We exploit these spatial spillovers to identify the causal impact of the intervention on household-level socio-economic outcomes, including damages to house structure and assets, work disruptions, transportation and connectivity, health and nutrition, mental health, and child-related outcomes.
|
|
Field
Trial End Date
|
Before
November 30, 2024
|
After
October 31, 2025
|
|
Field
Last Published
|
Before
June 15, 2023 04:24 PM
|
After
September 24, 2025 09:49 AM
|
|
Field
Intervention (Public)
|
Before
Our project studies the effects of an improvement of urban drainage infrastructure funded by the KfW in the city of Barishal, Bangladesh, on climate resilience of affected households. The intervention consists of improvement of the existing urban drainage network in the commercial center of the city, since the current network is not able to handle the demands under heavy rainfall which leads to frequent flooding.
|
After
Our project studies the effects of an improvement of urban drainage infrastructure funded by the KfW in the city of Barishal, Bangladesh, on climate resilience of affected households. The intervention consists of improvement of the existing urban drainage network in the urban center of the city, since the current network is not able to handle the demands under heavy rainfall which leads to frequent flooding.
|
|
Field
Intervention End Date
|
Before
February 29, 2024
|
After
June 30, 2025
|
|
Field
Primary Outcomes (End Points)
|
Before
Flood risk, household level:
- Observed flooding (satellite images)
- Reported flooding
- Stated flood risk perceptions
Socio-economic outcomes, household level:
- Spending behavior
- Food security and dietary adequacy
- Assets
- Aspirations, life satisfaction
Education:
- Health outcomes, individual level
- Likelihood of suffering from waterborne/vectorborne diseases
- Frequency of visits to health centers
- Child health (early childhood, children under 5)
Behavioral outcomes, individual level:
- Changes in littering behavior
- Changes in norm enforcement
|
After
H1: Flood risk, household level:
- Self-reported flooding
H2: Socio-economic Outcomes
- damages to house structure and assets
- work disruptions
- transportation and connectivity
- health and nutrition
- mental health
-child-related health and education outcomes
|
|
Field
Experimental Design (Public)
|
Before
To identify the causal effect of the improved drainage network, we will use the supply-side determinants of drainage infrastructure in a spatial regression discontinuity design (RDD). Specifically, we will exploit discontinuity arising from variations in access to the gravity fed canals of the drainage network based on the flow direction of the drainage canals. We will complement this analysis with a difference-in-differences analysis for a subsample.
|
After
Leveraging high-resolution elevation data, we develop a two-dimensional flood hazard model to estimate household-specific changes in the duration of waterlogging induced by the intervention. We show that most of the predicted reductions in flood risk are not close to the targeted drains as intended by the policy-makers, but are experienced as indirect network effects by households not directly targeted by the intervention. We exploit these spatial spillovers to identify the causal effects of the intervention
|
|
Field
Randomization Unit
|
Before
Households
Grid cells
|
After
Households
|
|
Field
Planned Number of Clusters
|
Before
2,000 households
|
After
2,600 households
|
|
Field
Planned Number of Observations
|
Before
2,000 households
|
After
2,600 households
|
|
Field
Sample size (or number of clusters) by treatment arms
|
Before
1,000 households treatment and 1,000 households control
|
After
The treatment is defined as a continuous variable (reduction in predicted flood risk) therefore, there are no traditional "treatment arms".
|
|
Field
Power calculation: Minimum Detectable Effect Size for Main Outcomes
|
Before
Regression Discontinuity with uniform distribution around the boundary: 0.25 standard deviations (alpha = 0.05, (1 - beta) = 0.8).
|
After
|
|
Field
Keyword(s)
|
Before
Behavior, Environment And Energy, Health
|
After
Education, Environment And Energy, Health, Labor, Other
|
|
Field
Intervention (Hidden)
|
Before
|
After
Treatment definition: Treatment is defined as the predicted flood risk reduction due to the infrastructure improvement based on a two-dimensional flood hazard model.
The auxiliary hypothesis (H1) tests whether the treatment correctly captures a reduction in flood risk. It tests whether the treatment variable correlates with reported flood experience at household level.
H1: Households with a larger predicted reduction in flood risk will be less likely to report local flooding.
The main hypothesis (H2) tests whether the treatment had an effect on socio-economic outcomes.
H2: Households with a larger predicted reduction in flood risk will be less likely to experience negative indirect effects of flooding on
a) damages to house structure and assets
b) work disruptions
c) transportation and connectivity
d) health and nutrition
e) mental health
f) child-related health and education outcomes
|