To test the relationship between financial inclusionand stability, Sahay etal. (2015) used panel regression with country fixed effects for thetimeframe from 2004 to 2011. Dependent variables were bank Z-score, taken fromthe Global Financial Development database. Financial inclusion variables fromIMF’s Financial Access Survey (ATMs per 100,000 adults, ATMs per 1,000km2,commercial bank branches per 100,000 adults, commercial bank branches per1,000km2, registered mobile money accounts per 1,000 adults, deposit accountswith commercial banks per 1,000 adults, depositors with commercial banks per1,000 adults, household depistors with commercial banks per 1,000 adults,household deposit accounts with commercial bank per 1,000 adults, loan accountswith commercial banks per 1,000 adults, household loan accounts with commercialbanks per 1,000 adults, borrowers from commercial banks per 1,000 adults). Thevariableswere lagged by one year in the regression. The explanatory variables were alsointeracted with the variable BCP, which approximates the quality of bank supervisionby measuring the degree of compliance with Basel Core Principles (BCP).
Twomeasures of BCP were tested: a composite of all the principles, and a subset ofBCP principles relevant to financial inclusion (Core Principles 1, 3, 4, 5, 8,9, 10, 11, 14, 15, 16, 17, 18, 24, 25, and 29). Control variables were thelagged values of the Financial Institutions Depth index (FID) from Sahay andothers (2015a), real GDP per capita, excess of credit growth above nominal GDP;contemporaneous variables of population, FDI-to-GDP ratio, trade-to-GDP ratio,inflation, government balance, a dummy for banking crisis, and the Lernerindex. The coefficient on the variable “number of borrowers per 1,000 adults” wasfound to be negative and significant for both X and X2. Thecoefficient of the interaction with both measures of BCP was positive.
Forother variables of financial inclusion, the relationships were found to be insignificantor inconclusive.