Figure 1.
Impact pathways of mobile money.
Table 1.
Use of mobile phones and mobile money among sample households.
Figure 2.
Types of activities performed with mobile money among sample households.
Table 2.
Descriptive statistics for variables used in regression models.
Table 3.
Determinants of mobile money and mobile phone use (probit model estimates).
Table 4.
Determinants of household income.
Table 5.
Determinants of remittances received.
Table 6.
Determinants of input use in banana production.
Table 7.
Determinants of banana sales and profits.
Table 8.
Treatment effects on household income with extended models.
Figure 3.
Treatment effects on household income with alternative estimators.
Notes: Original FE refers to the fixed effects model shown in Table 4, column (1). IV1 is based on an instrumental variable estimator where mobile money was instrumented with the percentage of households using mobile money at the village level. IV2 is based on an instrumental variable estimator where mobile money was instrumented with the percentage of households owning a mobile phone at the village level. IPW1 is based on an inverse probability estimator where the original probit model shown in Table 3, column (1), was used to calculate propensity scores. IPW2 is based on an inverse probability estimator where the original probit model was extended by variables measuring prices of banana, fertilizer, and pesticides. *,*** significant at the 10% and 1% level, respectively.