begins somewhere in the middle of the loan term when the client
stops making the scheduled loan payments. Because our dependent
variable is a binary variable we use a logit regression to estimate
the effects of social capital on repayment performance. The binary
dependent variable (Y1/2) is one for all households that did not
pay on time (principal or interest) or in case of Vietnam stated that
their loan has been rescheduled already. We regressed each country
sample separately. Thus, we have two dependent variables:
Y1/2: Have you been late in paying the principal or interest on
time or has the loan been rescheduled already (Y1 Thailand, Y2
Vietnam)?