I have an interest in exploring neglected and "strange" opportunities for growth in developing settings. This one is the most recent one occupying my mind, I would appreciate a comment about which one of the "possible interventions" seems more interesting / impactful.

Background

Access to Mobile Money might have direct implications in poverty reduction. Because most of the developing world is not banked, mobile money is reaching many households as the first opportunity to obtain all the benefits of traditional finance, with almost no entry level costs. So whatever benefits a household gets from financial services (keeping money safe, making savings easier by restricting availability of cash, motivating one to save, saving transport costs when doing payments, giving out or receiving money) they get them from a mobile money account. 

There is some research about the issue. One study is displayed in Innovations for Poverty Action, and also here´s a broader analysis of research on Mobile Money. 

https://www.poverty-action.org/study/long-term-effects-access-mobile-money-kenya 

In the study, which is one of the most conclusive so far, they found a moderate effect in poverty reduction in Kenya that could be related to the expansion of Mobile Money. They observed different areas of Kenya between 2008 and 2014, and found that in those areas that had greater access to mobile money:

• There had been a general 2% reduction of extreme poverty
• Had 22% fewer female-headed households in extreme poverty as per consumption rates
• And 13% of those female-headed households living on 2$ a day (almost extreme poverty)
• No similar effect for males

Interventions.

So let´s assume the research is sound. I read it some weeks back and it got me thinking about possible interventions that would capitalize on this effect.

Intervention 1: Create mobile money accounts for poor people, speeding up the process of adoption of Mobile Money. This would be really cheap, it costs between nothing and 2 $ to create an account. If someone is interested on playing with these numbers, I could make a rough budget for 1,000, 10,000, 100,000 and 1,000,000 dollars. 

Cons: 1) It might be costly to move and coordinate with villages without Mobile Money (which are generally remote), or to find people without Mobile money in urban settings, it requires targeting, transport costs, and it is not that attractive, if you want to distribute food in a cheap way, you can give out thousands of tons in a matter of hours, but "mobile money accounts" are something that most people already have, and that does not derive any immediate positive effect... we are looking at a fairly small and slow growth process, so we need to be mindful of this, we might outweight the benefits by spending too much on logistics. 

2) It seems as if we are taking over the work of the company in promoting their business for them. Also, some mobile money companies must be worse than others, so we should be careful in extrapolating conclusions. These considerations apply to the other options.

Intervention 2: 

Creating mobile money accounts for poor people and giving out a small cash transfer (7$-17$). This  justifies better the resources previously invested in targeting (finding people without these accounts) and makes it more likely unbanked people to report themselves. 

Con:

It makes the process significantly more expensive

It also has potentials for fraud.

Intervention 3:

A Research and advocacy with mobile money companies for easier entry level restrictions, or tailored promotion to the extreme poor, to facilitate access to accounts for the unbanked. This is the least costly. 

Con: 

It is not clear there are actually very high restrictions, it seems there are just logistical considerations, and companies are already doing their best at reaching more people, as it is in their best interest.

Intervention 4:

Create mobile money accounts for refugees /Internally Displaced Persons who have lost their National Identification Cards. This is not just paying for a service, but following-up on the person’s documentation, making sure an ID is retrieved or created, and then topping it up by buying a SIM card, registering it, creating an account, and if necessary buying a phone. (In Cameroon, this whole package would cost roughly 50$).

Cons: 

Again, it is getting more expensive.

This is also much more complicated for refugees. IDPs are citizens of the country and most of the times they already have the phone account, making the transition very easy and cheap for them. 

Intervention 5: 

Supporting individuals living in extreme poverty to become entrepreneurs and create a mobile money booth in an underserved community. These services often spread with individual operators who are entrepreneurs, and many of them are also living in poverty or slightly above. Training extreme poor individuals and deploying them in communities without Mobile Money should have much more direct impact in two directions: First, the person supported will move from unemployment to a very profitable venture, that family is fairly likely to come out of poverty. Second, other people in the village are likely to use more mobile money if they can cash in and out easily, so effectively you would be bringing banking services to small villages (150 to 2,000 persons) that did not have them before and might not have had it for 6 months to 10 years without the intervention. This could be measurable as indirect impact. 

Cons: 

This is more expensive, by several order of magnitude, than simply creating an account. Creating an account costs between nothing and 2 dollars, this involves setting up a small “box”, providing some capital so the business can work, buying a phone, training… we would be talking of at 100 to 400 USD per beneficiary. 

Intervention 6: 

Contribute to more research

It seems most research is captured here https://www.annualreviews.org/doi/10.1146/annurev-economics-063016-103638. So the body of evidence might be insufficient. It would be interesting to replicate the studies in other contexts within the developing world, and to have a strong evaluation and research arm on any of the above mentioned possible interventions. 

I find particularly challenging the fact that no progress was made with men, if we follow the “science” so far, we it won’t be efficient in any of the programs above to target men, but this seems counterintuitive (eg. Not supporting a disabled IDP without an ID on creating his account, because he won’t take proper advantage of the mobile money seems counterintuitive, the theoretical benefits are many) I would prefer to look at families, prioritize females within the families, use female-headed as a scoring criteria, but not stop working with men altogether.

Note that in Africa, most households would be male-headed, so when the study looked at the “female-headed” the sample size became much smaller, and random variance more important. There could also be a particular effect in female-headed households for some reason. 

Perspectives of funding 

Negative: Nobody would fund any of this. The effect is too small; it might not be observable under traditional funding cycles (6 months – 2 years). The only one that seems ready for funding under the current environment is Intervention 5, we could get a traditional development actor to fund it, maybe 4 in a crisis setting, as a protection service. But the first option, which could be more impactful because it´s scale, would have a really hard time in getting any funding from traditional sources. 

Positive: 

The fact that it is connected to a company opens up opportunities for profitable Corporate Social Responsibility. In that context, something that was a bad idea (using development funds to promote the profit of a private corporation), becomes much more attractive (using profits of a company to fight poverty).  There could be a synergy somewhere, it is possible some areas remain unbanked because they are just slightly unprofitable, but if they are “colonized” they become profitable in the long run, so a company might be interested in helping an NGO bring banking to these areas under a CSR grant, and reaping the benefits 10 or 15 years after. I am willing to explore this in the Cameroonian context by submitting a concept note to the https://mtn.cm/foundation/projects/submit-a-project/

General Implications for development and humanitarian interventions

• Make all cash transfers mobile money transfers. This is generally good to improve traceability too, this is not a very bold change, and it is already the preferred method for the big players (GiveDirectly, the World Bank, the World Food Programme). 
• Make all payments mobile money payments in the course of work, and  if recipients don´t have account, either spend the extra time making accounts for them or (the least nice option) forcing them to create the account in order to receive the payment (some institutions with 0 cash policies already do this, no excuse accepted).
• Make any documentation social work complete by not only creating/retrieving the ID, but doing a Mobile Money account at the end, you are already working with the person/family and there are almost no costs, it could literally add less than 5 minutes to the time spent on case-management. 

Please comment! Feel free to criticize the ideas. Most importantly, drop your opinion about which of these different interventions seems more impactful and necessary, even an extremely short comment would be greatefuly appreciated. 

Thank you for reading!