Financial inclusion and economic growth: Evidence from Zambia

Résumé

Financial exclusion remains a major developmental challenge in Zambia despite substantial policy initiatives and reforms within the financial sector. The problem disproportionately affects rural households, smallholder farmers, women-owned enterprises, informal workers, and micro, small, and medium enterprises. Consequently, limited access to formal financial services continues to constrain savings mobilization, productive investment, enterprise expansion, and household resilience, thereby weakening the country’s broader economic transformation agenda. As economies increasingly depend on efficient financial systems, digital payment methods, and easier access to affordable credit for inclusive growth, financial inclusion has become a key policy concern rather than merely a social welfare issue. The purpose of this study was to investigate the relationship between financial inclusion and economic growth in Zambia using annual time series data covering the period from 1989 to 2019. The study was anchored on financial intermediation theory, endogenous growth theory, and technology acceptance theory, which explain the channels through which improved access to financial services can promote investment, innovation, productivity, and long-term economic expansion. Real Gross Domestic Product (RGDP_t) was used as the dependent variable, while financial inclusion was proxied by Outstanding Loans from Commercial Banks (OLB_t), Automated Teller Machines per one hundred thousand adults (ATM_t), and the Getting Credit Distance to Frontier score (DF_t). Human Capital (HC_t), Total Factor Productivity (TFP_t), Foreign Direct Investment (FDI_t), and Labour Force Participation Rate (LF_t) were incorporated as control variables. The Augmented Dickey–Fuller results showed that RGDP_t, OLB_t, ATM_t, HC_t, TFP_t, and FDI_t were integrated of order one I(1), while DF_t and LF_t were stationary at levels I(0). The final long-run Error Correction Model therefore retained RGDP_t, OLB_t, ATM_t, HC_t, and FDI_t. Residual-based Engle–Granger testing confirmed a stable long-run cointegrating relationship (Engle & Granger, 1987). The normalized long-run coefficients indicate that ATM_t, HC_t, OLB_t, and FDI_t are positively associated with RGDP_t over the sample period. More importantly, OLB_t remains the most policy-relevant financial inclusion variable in the short run because it is the only financial inclusion indicator that remains statistically significant in the final parsimonious Error Correction Model (ECM). The parsimonious ECM, selected using the Akaike Information Criterion and small-sample parsimony, shows that variations in OLB_t exert a statistically significant positive short-run effect on economic growth, while the lagged error correction term is negative and statistically significant, implying convergence toward long-run equilibrium. In the broader general ECM, the quality indicator Getting Credit Distance to Frontier score (DF_t) and Labour Force Participation Rate (LF_t) were not retained in the long-run system because they are integrated of order zero I(0), while Total Factor Productivity (TFP_t) did not remain in the final parsimonious short-run ECM once AIC-based reduction and small-sample parsimony were applied. This is an important result because it indicates which dimensions of financial inclusion matter most in the short run. The study concludes that financial inclusion should not be assessed only in terms of the expansion of access points or credit volume, but also in terms of the quality, allocation, and productivity of financial intermediation. Policy implications emphasize improving credit allocation, strengthening productive financial usage, expanding effective access infrastructure, and linking inclusion policies to wider productivity-enhancing reforms.