Designing microcredit loans for agricultural farmers

How to cite

Kono, Hisaki; Shonchoy, Abu; Takahashi, Kazushi. 2023. "Designing microcredit loans for agricultural farmers." AEA RCT Registry. AEARCTR-0010732. https://doi.org/10.1257/rct.10732-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Hisaki Kono* , Abu Shonchoy , Kazushi Takahashi

* Primary investigator

Registry ID
AEARCTR-0010732
DOI
https://doi.org/10.1257/rct.10732-1.0
First registered
Last updated
Status
Completed
Trial dates
2015-05-01 – 2016-07-31
Countries
Bangladesh (Dinajpur)
JEL codes
G21, O16, Q14
Registry record
http://www.socialscienceregistry.org/trials/10732

Abstract

Despite the expansion of microcredit access, its outreach is still limited among farmers. Standard microcredit causes a timing mismatch between cash flow and credit flow for farmers. They have little income until harvest while standard microcredit requires weekly installment. This mismatch will cause underinvestment and borrowing for repayment, resulting in lower uptake rates. Further, agricultural investment is sequential while credit is disbursed in a lump-sum, inducing present-biased (PB) farmers to fail in setting aside a sufficient amount of the fund for later investment. We randomly offered three microcredit programs that differ in repayment and disbursement timing to tenant farmers: (1) standard microcredit, (2) crop credit that disburses credit in a lump-sum and requires a one-time repayment after harvest, or (3) sequential credit that disburses credit sequentially and requires a one-time repayment after harvest.

Intervention

Traditional credit (T1): This is the standard microcredit product in which the full loan amount was disbursed at the beginning of the crop season, and borrowers were liable to repay the loan in regular weekly installments of equal amount, beginning from the first month after loan disbursement. The loan matured after the harvest, when farmers were supposed to pay the last installment.

Crop credit (T2): This product removes the weekly installment from Traditional credit (T1). The borrowers were required to repay the full amount at the end of the harvesting period, which corresponded to the due date of the last installment in T1.

Sequential credit (T3): This product modifies T2 by changing the schedule of loan disbursement. To match the sequence of agricultural investment, the disbursement was divided into three phases. Based on the bookkeeping exercise in the pilot survey, we set the limit of the first disbursement at 60% of the maximum loan size so that the remaining 40% was still available for late-stage investment. At the time of the second disbursement (one month after the first disbursement), borrowers could receive up to 20% of the loanable amount in addition to the unused loanable amount at the first disbursement. The third and final disbursement was made one month after the second disbursement. At each disbursement, borrowers could decide the amount they would like to receive as long as it was within the specified limit.

Sequential in-kind credit (T4): This product intends to strengthen the commitment function of T3 by disbursing the credit in kind such as seed and fertilizer (valued within the loanable amount).

Experimental design

After collecting the baseline data, we randomly assigned 200 members (4 members per group) to each of the four credit products (T1-T4). The remaining 200 members served as the control group. We stratified the individuals based on the score of economic status that would be correlated with the latent productivity. Specifically, we computed the score by factor analysis, where we include indicators for owning agricultural lands, owning livestock, owning productive assets, borrowing money in the last three years, having electricity connection, having latrine toilet, and housing conditions (if the house is made of mud), the area of agricultural land, and the years of education. We constructed a strata of five households with similar scores, and randomly divided the five households into five different treatment statuses.

Design details

Randomization method
Randomization done in office by a computer
Randomization unit
Individual
Clusters
1000 farmers
Observations
1000 farmers
Arms
200 farmers T1, 200 farmers T2, 200 farmers T3, 200 farmers T4, and 200 farmers control
Primary outcomes
Uptake, credit size, investment
Secondary outcomes
profit, other borrowings, default

Cited by

No citation data available for this registration yet.

Keywords