- Developing nations require cash flow if they are to grow
- Mobile money operations allow communities to obtain the currency they need
- Training mobile money agents could help them make better decisions
Money is a strange concept. Depending on the currency, the cash value of an entire home can be carried in a suitcase. But this transportability comes at a price. It can take a person an entire lifetime to save an amount of money a thief could steal in less than five minutes.
This is one of many financial challenges facing developing communities. Improving your lot in life often requires accumulating capital, but keeping a large stash of cash under your mattress isn’t a secure option. Mobile money could solve this and other issues. Chris Parker, assistant professor of supply chain management at Penn State, is looking for ways to optimize these solutions.
“Mobile money allows people without access to a bank account to easily transfer money,” says Parker. “They put money into an account with their mobile phone company and can then electronically transfer that to family and friends or have access to other financial services. The system was designed to work with simple feature phones (as opposed to smartphones) so that the bulk of customers can access the system.”
Mobile money could make life easier for a lot of people. Borrowers may currently have to travel long distances from rural areas to urban centers to obtain traditional cash—journeys which can be dangerous and time consuming. Mobile cash helps cushion the blow of economic shocks like a failed crop or an injury by allowing quick access to resources. Finally, mobile money systems help borrowers build credit history.
While such a system extends the reach of financial resources, the money doesn’t appear from thin air. Parker’s recent work looked into mobile money agents in Tanzania, and specifically investigated the decisions they make in terms of stocking up on mobile cash.
How much is enough?
All of these benefits hinge on a mobile money agent’s ability to supply cash in the first place. Unlike a bank, contractors aren’t flush with resources. Agents have to decide how much currency to hold on any given day, and the wrong choice could cost them.
“If the agents hold too little, they won’t be able to perform transactions and will miss out on the commission they receive for each transaction,” says Parker. “If they have too much, they are either paying interest on that capital investment or losing out on other ways they could have been using their money, such as investing in other business ventures. The randomness of transactions each day makes it extremely difficult to predict how much cash to have without access to sophisticated forecasting algorithms.”
Parker wants to help solve some of these problems by providing mobile money agents with daily guidance via text message. These could be explicit recommendations (“You should have 185,000 shillings in electronic currency tomorrow”) or information about expected demands (“On a typical Sunday you have received 90,000 shillings in deposits, and on a busy Sunday you have received 275,000 shillings”).
Parker also provided training in person or via voice message. This research spanned over eight weeks and included 4,771 agents. The study derived data for this experiment from their mobile money operator partner in Tanzania.
A second source was questionnaires and sign-in sheets from the in-person training. The recommendations for how much money to stock required a forecast of the distribution of cumulative demand throughout a day based on each agent’s historical transaction data.
Parker discovered that the agents who ran out of money less frequently were those who received explicit recommendations and were trained in person. The first part makes sense, as a direct recommendation may be taken more seriously than a general range. In-person training could become an invaluable resource—if scientists can cross the trust barrier.
“The in-person training was much more expensive and could become even more expensive to roll out to areas more rural than Dar es Salaam where we were working,” says Parker. “However, there is more time to explain the system we created and to develop trust.”
Overall, Parker estimates that agents ran out of money on about 30% of the days contained in his data. The system he developed brings this number down closer to 27%, and he believes that 7% is a more optimal level.
“Clearly there is a lot to be done here,” says Parker. “First, we need to figure out why people aren’t listening to the recommendation fully. Are there still trust issues? Second, we may not observe reduced stockouts via the notification treatment because the voice message delivery doesn’t convey enough information.”
Parker suggests it could be possible to create other cheap training systems that would benefit agents. He also wants to investigate how these results translate to rural areas or to different countries.
Climbing out of poverty is difficult, but a lack of available cash only compounds the issue. Hopefully, research like Parker’s will help cash-strapped regions continue to grow and thrive.