Financial inclusion, poverty and income inequality in low- and middle-income countries: A mixed-method investigation

In the mid-2000s, the World Bank and other international institutions endorsed financial inclusion (FI) and microcredit as important development tools for reducing poverty and income inequality in low- and middle-income countries (LMICs). Whereas microcredit programs have been widely criticised as a development strategy, FI lacks similar scrutiny. This thesis aims at closing this gap using mixed methods. Overall, our results suggest that FI is unlikely to reduce poverty and income inequality in LMICs. Instead, it may lead to over-indebtedness, especially of informal workers.

We first undertake a rigorous inquiry into the conceptual basis of FI. Chapter 2 systematically examines 67 studies and offers a new definition of FI, clarifying its objectives, elements and the nature of financial intermediaries. We find that the dominant definitions of FI are based on mainstream hypotheses that neglect the macroeconomic particularities of LMICs.

Chapter 3 aims to overcome this shortcoming by examining the macroeconomic conditions that shape FI in LMICs utilising hypothesis from the Post-Keynesian approach. This chapter develops a model of the relationship between individuals and financial institutions and incorporates power mechanisms, such as social shame, that underlie FI processes in LMICs.

Chapters 4-6 present mixed-method empirical evidence on the relationship between FI, poverty and income inequality. Chapter 4 reports on a case study in Brazil comprising 30 interviews with low-income individuals. Chapter 5 utilises microdata from 451,372 individuals to create a multi-dimensional index of FI using multiple correspondence analysis. Finally, Chapter 6 employs this new index to econometrically estimate the effects of FI, poverty and income inequality, and to investigate causal relationships. Our findings indicate that poverty reduces the level of FI, but that FI presents no robust effects on poverty.

In sum, this thesis provides a rigorous conceptualisation of FI and makes theoretical and empirical contributions that challenge the expected effectiveness of FI in LMICs.