Recent Findex findings show that account ownership has increased, globally. But more people than ever are not using their accounts. And while digital payments accounts are on the rise, the number of savings accounts has declined since 2014. 

These findings have provoked debate about the future of financial inclusion and even the survey methodology itself. Beth Rhyne and Sonja Kelly from the Centre for Financial Inclusion (CFI) raised concerns in CFI’s recent publication Financial Inclusion: Hype vs Reality. Greta Bull, CEO at the Consultative Group to Assist the Poor (CGAP), responded with a two-part blog series on why she is more optimistic about the numbers.  

For us at insight2impact, the findings aren’t necessarily surprising. In fact, they confirm our conviction that, though access and uptake are necessary preconditions for meaningful financial inclusion, usage is the ultimate objective. 

Over the past three years, our basic premise has been that people do not use financial services because they want to use financial services, but because they want to meet an underlying need for which financial services may be helpful. For example, they want to pay school fees and use a payment service to meet this need. 

But understanding needs, by itself, does not explain usage. Needs do not explain why so many people still choose to pay for school fees in cash when they have access to formal payment services. 

As my colleague Jeremy Gray explained in his earlier blog Closing the gap, usage is based on a complex set of considerations; and, in some cases, consumers aren’t even cognitively aware of all the important factors that influence their decisions. His blog is based on our conceptual framework on Drivers of usage, which draws on a wide range of decision-making literature. The conceptual framework introduces and categorises the complex set of considerations that influence consumer usage behaviour. It forms part of our i2i #FinNeeds measurement frameworks aimed at advancing financial inclusion.  

Testing the framework

Locating or isolating the most relevant driver can be tricky. Not all the drivers can be measured at once, and various methodologies and data sources are required to inform different parts of the framework. 

We’ve started to test this framework as part of our needs-based measurement work with the AFI Inclusion Data Working Group (AFI FIDWG). For the pilot exercise that we’re running in Mexico, we built a question on self-reported reasons for uptake into a demand-side survey undertaken in Mexico’s Puebla state. Our findings provide food for thought for policymakers and providers alike.

Consumers have different motivations for using formal and informal financial services. They tend to prefer formal financial services, such as credit cards, bank loans or insurance, for functional reasons (gauged in terms of the value provided, costs incurred or convenience of using the service). But they use informal services, such as borrowing from family and friends or belonging to savings groups, for relational reasons (perceived trust and a sense of belonging). In fact, the respondents place more importance on the relational factors when making their usage decision than they do on the functional factors. This was common across both banked and unbanked individuals. No wonder, then, that even the banked continue to turn to family and friends, or informal services, for many of their financial needs. 

A team from CFI recently did some qualitative interviews with merchants in Mexico to better understand their financial lives. Their findings confirm that access to financial services is not the issue, but rather the misalignment between individual’s needs and the financial products on offer.

To explore a different angle, we applied our usage framework to the transactional data of a large retail bank in Mexico. After clustering the customers into groups based on intensity of usage, we were able to build a regression model to compare how different demographic characteristics drive intensity of usage. 
As expected, we found significant links between education and income levels and usage of debit and credit cards. What was unexpected was that income effect goes in different directions for debit cards (where lower socio-economic classes are likely to be low users) versus credit cards (where lower socio-economic classes are more likely to be high users). Further, men were less likely to be in higher-usage clusters than women. 

As a next step, we're merging demand-side and supply-side data from Mexico to build a broader picture of bank customers' financial lives outside of the bank, looking at their need and their use of cash, social and informal financial devices. 

We'll also apply this approach in our upcoming pilot in Nigeria, as we look particularly at drivers of digital payments. 

Drawing the map

Ultimately, drivers are levers. We want to understand drivers to answer the fundamental underlying question, what can be done to change behaviour?

Over the coming year, we’ll be working with AFI to develop a toolkit for our #FinNeeds measurement framework to set out which data to use when, and how to analyse it to help policymakers and financial service providers get closer to “solving” the usage issue. We will also present our findings from the Zimbabwe and Mexico pilots at the AFI Global Policy Forum in Sochi, Russia, in September 2018. 

Please let us know your feedback and suggestions on the drivers framework, as well as particular questions that you’d like to see it answer. We can’t claim that the usage issue has been solved, but we’re attempting to draw a useful map.