Grameen Bank made a folk legend of the poor woman catapulted to prosperity by becoming a telephone-wielding entrepreneur.
By selling minutes on her phone plan (purchased with a Grameen loan) women like her, the story goes, moved markets and brought valuable information to the masses.
Craig McIntosh and Michael Futch set out to test the impact of rural telephones in a totally different setting, Rwanda. Unfortunately, as often happens in randomized trials, reality struck:
The intended design of this study was a randomized controlled trial, in which the phones would be phased in to a group of 380 pre-selected villages in an experimental fashion.
In the end, competitive pressures in the MFI sector as well as operational glitches in the rollout process undermined the randomized structure.
When we returned to do a follow-up study 15 months after the baseline we found that 94 communities had received phones, but that the actual rollout bore little resemblance to the intended design.
For the empirical economist, such glitches are almost inescapable. The candid disclosure, and the honest attempt to still analyze the (non-experimental) impacts, is one of the reasons I like Craig’s work so much.
Unfortunately, the observational analysis bodes ill for rural telephony as a business model in Rwanda:
We see a moderate but significant increase in the percentage of local farmers who report arranging for their own transport to market (from 27% to 40%) and local entrepreneurs are more likely to pass news via cell phone.
However, the community analysis shows absolutely no impact of the phones on trading activity or availability of goods in local markets, and household-level impacts are repressed by the fact that airtime usage in Rwanda is so low as to indicate that few VPOs realize sufficient profits to pay off their six month loans from that source alone.