I attended the World Bank Spring Meetings in April to participate in the launch of the Global Findex—a new database of financial inclusion indicators based on nationally representative surveys of more than 150,000 adults in 148 economies. The Findex was created through a 10-year, $10 million grant from the Financial Services for the Poor Initiative at the Bill & Melinda Gates Foundation that funded a collaboration between the World Bank and Gallup to add questions to the Gallup World Poll.
Two key indicators will be collected every year: the percentage of adults owning a bank account and the percentage having taken a loan in each country; while the full set of 20+ questions was collected this year and will be collected every three years. The full set is much more nuanced and includes questions to determine: (i) active usage vs. simply owning an account (ii) whether the respondent is connected to a payment system or not (iii) mode of access to convert cash to digital and vice versa (e.g. ATM, branch, agent, etc.), and (iv) using mobile phones to transfer money. This data is a major step forward in addressing the needs of policy makers interested in improving financial inclusion, and will bring to the field the kind of data other fields have had for years.
What does the data tell us?
- Globally, approximately 2.55 billion adults do not have a formal account (50 percent of adults), with the majority – 2.37 billion – residing in developing economies.
- Unsurprisingly, the poor suffer the most - 77 percent of poor adults (<$2 day) do not have an account
- 9 percent of adults worldwide have taken a new loan from a bank, credit union, or microfinance institution in the past year
- Among the most commonly reported barriers to having an account are high cost, physical distance, and lack of proper documentation – all of which are potentially addressable through innovation and better policy.
- A surprisingly high percentage of account holders (e.g. 38 percent in Africa) appear to be using their accounts simply for receiving transfers and withdrawing the money, rather than saving.
While there are many more facets to the data—how people use accounts, what they save money for, where they get loans if not from banks—one of the major questions that came out in the discussions around the launch was: are we making progress?
The short answer is that we don’t really know whether financial inclusion has been improving over the past few years. There have been two recent attempts to estimate the number of unbanked people in the past few years—one by me and my former team when I was at CGAP (CGAP is a grantee from the Financial Services for the Poor (FSP) department of the Foundation, and is based at the World Bank), and one by Jonathan Morduch, Director of FAI (another FSP grantee), with help from McKinsey & Co. Both estimates came out in early 2010, and while the CGAP estimate featured data that we collected from 149 regulators over the course of 2009, the FAI/McKinsey estimate had publicly available data from 2004-2005. Despite using different data sets, both came up with similar top line numbers—that around 50 percent of the world had no bank account. The FAI/McKinsey team pushed the data and assumptions a bit further than our team did and estimated that 90 percent of the <$2/day poor were unbanked. At CGAP, we cut the data by developing countries and figured that 70 percent of people in developing countries had no account, but didn’t try to estimate whether they were poor or not.
The Global Findex data is much higher quality than either of these previous estimates, which were concocted from an unholy mix of supply side data reported to regulators, cross-country regressions, and byzantine Excel formulas. Nevertheless, the Findex data also shows 50 percent of the world unbanked, but only shows approximately 80 percent of the poor are unbanked (as opposed to 90 percent from FAI/McKinsey) and approximately 60 percent of adults in developing economies (as opposed to 70 percent from CGAP). Thus, the new data makes the picture look rosier than the previous estimates did by about 10 percentage points on each of these important numbers relating to the poor and developing countries.
So, the question is: is this progress, or just a more precise estimate in light of better data? There is no way to know; and given we had to make some pretty aggressive assumptions in making the previous estimates, we should be very hesitant to call this progress.
The good news is that next year (and each year following for the next decade) the Global Findex will release new, high quality data tracking these indicators and the field will have a clear answer to the questions of whether or not we are making progress and how fast. This in itself is progress.