Turning data into impact in Bangladesh

By making sure people in Bangladesh received necessary information about impending disasters, communities were able to act to prevent damage from floods. 

Want a $38 return on every $1 invested? Want to turn data into impact? Make sure people get the information they need to act before a disaster strikes, and that they know what actions they can take. Communities in Bangladesh were 4.5 times more likely to relevant information to prevent damage from floods early enough to do something about it. Acting before the floods, rather than reacting after, saved thousands of dollars in damages.

Golapi Begum talks about how working with the project changed her perspective: “I did not believe them first, neither did many of our community people. We thought that they were not God, how they could predict the exact flooding time and water level!! When the first wave of flood hit at the exact time and water level they predicted, we were really surprised and started believing them.’’ 

Golapi says it made a big difference for women, too. “Usually, we do not even get usable latrines in shelter, let alone separate latrine for women. We were quite happy that we could use the latrines instead of going far from shelters during flood.”

The Supporting Flood Forecast-based Action and Learning (SUFAL) project ran from 2019-2021 in Bangladesh with $241,000 in funding from ECHO. The project reached 42,898 people directly. Partners included Assistance for Social Organization and Development (ASOD), Islamic Relief Worldwide, and Concern Worldwide.

What changed?

  • People suffered less damage during floods. On average, people in the project were able to prevent 43% more damage—about $542 more dollars of assets and livestock saved—if they were part of the project in 2020 than if they weren’t part of SUFAL.
  • People have more money and less debt. Project participants had 12% more income ($140 or 12,000 BDT) and were 21% less likely to lose income during the 2020 floods than the control group. They were also 40% less likely to have to take on debt to survive and rebuild after the flood.
  • People have better information sooner. People in SUFAL were 4.5 times more likely to get information early enough to take action—6 days ahead of a flood compared to 3 days for the control group. It’s 15 times better than they had before the project started. They were also 3-5 times more likely to trust the information they got. They were also 5 times more likely to get information that helped them decide what actions to take to prepare for the flood.
  • People act faster. SUFAL participants were 2.5 times more likely to take action before the day of a crisis—77% of participants compared to 31% in control.
  • Communities act together. It’s not just families and individuals—in SUFAL, whole communities were more likely to work together. They were 68% more likely to take early action together than communities where SUFAL didn’t operate.
  • Local governments are buying in long term. The Union Parishads (local governments) are setting up their own early flood response funds, and are much more likely to be spending money before the flood to reduce damage rather than react to it. “In previous years, we used to allocate the fund and support for flood management to LGIs after the flood hit. In 2020, they submitted their requirements much earlier and we could allocate the resources 7 to 10 days earlier flood.’’ - Mehedy Hasan Titu, Project Implementation Officer, Jamalpur

How did it happen?

  • Work with local leaders to get buy in. The project worked with local governments to improve shelters, flood barriers, and rescue boats before the flood, and local groups financed 9% of these costs. Local leaders also felt more comfortable sharing and promoting early warning messages because they understood them and what to do next based on what the message said.
  • Make information transparent. The project used a range of tools—from solar powered digital forecast boards to community volunteers going door to door to share early warning info with women to mobile phone messages—to make sure that people got the info they needed early enough to do something about it. 8,800 people got info through mobile phones, and 40,000 got info through community volunteers.
  • Build forecasts together. The project worked with several different government committees, ministries, and information from local communities to build early warning predictions that were more accurate and more localized so they could inform action plans.
  • Get cash to folks who need it. The project gave cash grants to the poorest and most vulnerable families--$50 as a grant to respond to COVID-19, and $50 as a flood grant. The project also offered cash for work opportunities for people to repair flood barriers, shelters, and other critical infrastructure. In a context where 93% of people had lost some or all of their livelihoods because of COVID-19 and 43% of markets weren’t functioning, this was a critical step so people could act early.
  • Get clear on actions and responsibilities. The local teams put together standard operating procedures and early action matrices—clearly laying out who was going to do what based on different data points and conditions as the crisis evolved.

Want to learn more?

Read the project evaluation.