Religious Suasion in Financial Decision-Making: Evidence from the Punjab in Pakistan
How to cite
Barnard, Jordan. 2018. "Religious Suasion in Financial Decision-Making: Evidence from the Punjab in Pakistan." AEA RCT Registry. AEARCTR-0003282. https://doi.org/10.1257/rct.3282-1.0
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
This study examines the role of religion in financial decision-making by utilising a field experiment with borrowers of the largest Islamic microfinance provider. There are three treatment arms: those who were selected to attend a loan disbursement in Pakistan's most famous mosque; those who received their principal in their local mosque; and those who received their principal in a branch. Changing the way that loans are disbursed provides exogenous variation in the religious conceptualisation of loans. Exploiting these differences, first the role of religion in repayments is considered. Next, the role of religion in charitable contributions is investigated. Finally, using survey data multiple behavioural hypotheses pertaining to individual and intra-household decision-making are tested. For instance, by receiving one's loan in a place of worship, it could create behavioural impediments for family members wishing to appropriate a borrower's loan. \\
Intervention
There are three treatment arms: (1) those who were selected to attend a loan disbursement at Pakistan's most famous mosque; (2) those who received their principal in their local mosque; and (3) those who received their principal in a branch. Changing the loan disbursement venue provides exogenous variation in the religious conceptualisation of the loans by the beneficiaries.
Experimental design
22 branches (Primary Sampling Units) were selected with probability proportional to the number of loan disbursements in the month of July 2018 from the branches in the Punjab region. The list of first time borrowers was stratified by gender and beneficiaries randomly selected for each treatment group. In each branch, 23 males and 23 females (if possible) were selected per treatment. This resulted in a sample size of 3000 individuals. 1000 per treatment. At baseline, general demographic and financial questions were asked. We will conduct extended surveys of all 3000 beneficiaries in the sample with their consent six months after intervention.
Design details
Cited by
No citation data available for this registration yet.