The impact of commitment savings on seasonal poverty dynamics in northern Bangladesh
How to cite
Takahashi, Kazushi; Shonchoy, Abu. 2021. "The impact of commitment savings on seasonal poverty dynamics in northern Bangladesh." AEA RCT Registry. AEARCTR-0007862. https://doi.org/10.1257/rct.7862-1.2000000000000002
Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.
* Primary investigator
Abstract
It is widely discussed that the rural poor suffer seasonal starvation during the agricultural lean period. Unlike unanticipated shocks such as natural hazard and economic crisis, this seasonality appears to be anticipatable, hence would be manageable. To investigate whether time-locked commitment savings schemes can prevent the seasonal fall in consumption, we implement a randomized controlled trial to offer households commitment-savings products, either with market (8% per annum) or premium (50%) interest rates. We target 180 households with limited access to agricultural land that comprise the poorest segment in the Greater Rangpur region of northern Bangladesh. We pay particular attention to well-documented seasonal deprivation in the agricultural lean period from the mid-September to the end of October, locally known as Monga; households can receive the pre-specified interest if they can save money in the experimental account until Monga. To trace the impact of the commitment-saving product, we conduct bi-weekly surveys 18 times a year from April to January for two years, 2018/2019 and 2019/2020, where the first and second year correspond to pre- and post-intervention period, respectively.
Intervention
We introduce a savings scheme in June 2019 in collaboration with a local NGO, Gaibandha Rural Development Foundation (GRDF). This savings scheme involves a commitment device, in that the households are allowed to withdraw money at any time, but interest is not paid if they break the commitment to save until the beginning of the Monga period, that is, mid-September.
Experimental design
Out of the 180 sample households, 72 randomly selected households are eligible to receive the premium interest rate of 50% (Treatment 1: T1), while 54 households are eligible to receive the market interest rate, i.e., approximately 8% per annum (Treatment 2: T2). The remaining 54 households serve as the control group (Control: C).
Both savings schemes provide a free account with no opening, maintenance and withdrawal fees. In execution, the stuff of GRDF visit eligible households individually to collect savings twice a month independently from the household survey team, organized by MOMODa Foundation.
We differentiate the interest rate to examine whether access to savings schemes alone is sufficient or whether additional incentives are required for people to save for the anticipated seasonal deprivation shocks during Monga. The maximum saving amount is set at 4,000 BDT, which is about a half of the monthly household expenditure.
Design details
Cited by
No citation data available for this registration yet.