Blockchain-Driven Financial Inclusion in India: New Evidence from ARDL based Time Series Analysis (2016-2023)

Abstract

This study investigates how blockchain awareness, digital public infrastructure, and macroeconomic activity influence financial inclusion in India over the period January 2016 to December 2023. A monthly Financial Inclusion Index was constructed using Principal Component Analysis (PCA) from key variables: UPI transaction volume, Jan Dhan account penetration, and ATM density. To examine both short-run and long-run dynamics, the Autoregressive Distributed Lag (ARDL) bounds testing framework was applied, incorporating a structural dummy for the COVID-19 pandemic period. The results confirm a cointegrated relationship among financial inclusion and its determinants. In the long run, UPI usage and internet penetration emerged as the most influential drivers, while blockchain awareness proxied by Google Trends showed a modest but statistically significant effect. Industrial production also had a positive long-run association. The error correction term was negative and highly significant, indicating stable convergence to equilibrium. Diagnostic and stability tests, including CUSUM and CUSUMSQ, confirmed model robustness. These findings emphasize the role of digital connectivity and behavioural interest in new technologies as levers for inclusive finance. The paper contributes novel insights by combining behavioural proxies, macroeconomic fundamentals, and structural shocks in a unified high-frequency time series framework, and recommends targeted policy support for both digital infrastructure and emerging fintech solutions. Importantly, this is the first known application of the advanced time series technique ARDL (Auto Regressive Distributed Lag Model) technique to model financial inclusion using high-frequency monthly data for India, addressing a critical methodological gap in existing research. Additionally, the study employs the Toda-Yamamoto Granger causality framework to establish the direction of influence among the variables, providing further robustness to the time series analysis."