Data: Medical treatment loans and their effects on health care utilization and out-of-pocket expenditure: Evidence from an experiment in northern Bangladesh

We modified a health credit scheme to lower its transaction cost and improve its access. Using a cluster-randomised controlled trial, we piloted the modified and standard version of the scheme in 24 branches of a microcredit program in northern Bangladesh. We compared the loan disbursement records of the treatment and control branches before and after launching the modified scheme. Using a propensity score matching technique, we matched the users of the two schemes based on their disease profiles. Then, we compared the utilisation of formal healthcare and out-of-pocket health expenditure among the users (matched and unmatched) of the two schemes. We find the modified scheme increasing access to health credit. Moreover, contingent on health shocks, higher access to credit leads to higher formal healthcare visits and greater out-of-pocket health expenditure.

We provide the loan disbursement records of the 24 branches (i.e., branch_loan_uptake.dta), and the survey data of the matched (i.e., matched_incidents.dta) and unmatched health events (i.e., unmatched_incidents.dta) which were financed using the two health credit schemes. We also provide the code (i.e., data_analysis_mtl.do) we used to get the results. All the files are in Stata format; however, they can be read in other statistical softwares by converting them into appropriate formats. All the variables and values of categorical variables are labelled.