Events at CERGE-EI

Dissertation Committee:
Jan Hanousek (chair)
Randall K. Filer
Štěpán Jurajda

Abstract:

Microfinance emerged in the 1970's, aiming to help lift people out of poverty and promote economic growth by providing financial services to low-income households. Despite its global popularity, evidence supporting its net benefits is mixed; therefore, more comprehensive empirical testing is called for. This Dissertation aims to contribute to the on-going debate in microfinance in the following three dimensions:

In the first chapter, the impact of microfinance on macro economies is analysed. Motivated by limited knowledge of the economy-wide effects of microfinance, we aim to measure its impact on economic growth, financial sector development and reductions in income inequality. Measuring aggregate effects including those on non-recipients of microfinance programs constitutes an important contribution. We identify the “promised” impact of microfinance on economic growth measured by real GDP per capita, financial intermediation captured by broad money per capita and income inequality measured by the Gini coefficient. We also estimate the reverse feedback of macro fundamentals on microfinance itself. To estimate the dynamic equations, we use panel vector autoregressions (VARs) based on Arellano-Bover's (1995) and Blundell-Bond's (1998) instrumental variable system estimator, which enables us to control for potential endogeneity of microfinance and macro fundamentals. To address potential parameter heterogeneity in a dynamic panel, we divide countries into three broad clusters based on economic development, poverty, financial sector development, and levels of control of corruption. Such clustering also enables us to address the external environment for microfinance, which is multidimensional. In general, the results indicate an important and significant impact of microfinance at the macro level. In particular, the growth of microfinance is found to be positively and significantly associated with economic growth. We further find support for the impact of microfinance on financial sector development captured by broad money circulation in an economy. A one percent increase in microfinance borrowers in a country leads to a USD 314 increase in broad money per capita, which is equivalent to 13.8% of the mean value for the whole sample. The effect clearly differs across clusters in response patterns of microfinance. Finally, we find a positive impact of microfinance on reductions in poverty, as indicated by measurable lessening of income inequality. The result is stable for sample integrity of the Gini coefficient and trimming for outliers. Overall results indicate a significant role of microfinance and its potential to affect the broader economy. The impact and transfer dynamics, however, differ substantially by macro-institutional environment.

The second and third chapters are devoted to analysis of the microfinance environment in Uzbekistan, based on the primary dataset, which constitutes the first evidence of the impact of microfinance from the Central Asia region.

The second chapter describes the microfinance environment in Uzbekistan, emphasizing two types of non-bank microfinance institutions - Credit Unions and Microcredit Organizations. The specific nature of these institutions provides new evidence of the commercially oriented microcredit model and SME lending, which is an emerging trend in mainstream microfinance. The chapter offers two important contributions. On the supply side of microcredits, we analyse the determinants of initial placement of MFIs in districts of Uzbekistan. We find that MFIs follow general economic principles when choosing locations. On the demand side, we analyse the actual margins of excess demand for microcredits by considering only the pool of eligible applicants. We find that the total probability of microcredit approval is, on average, only 0.5, which implies that actual margins of untapped market may be much lower than projected when the narrow definition of eligible applicants is taken into account.

In the third chapter, the causal impact of improved access to microcredits is measured in terms of geographical distance to the nearest non-bank MFI. Proximity matters because of travel, time and other pecuniary costs and, more importantly, in terms of knowledge diffusion and (dis)connection from the microfinance network causing “signal dissipation”. The methodology is based on propensity score matching as a second best solution for program evaluation in the absence of experimental intervention. We match the 25% of households residing the closest to MFIs with the 25% residing the farthest, which defines the “treatment” as having easy or difficult geographical access to microcredits. To ensure the validity of the matching technique, we re-create pre-treatment covariates using a set of retrospective questions embodied in a single cross-sectional design. The accuracy and memory recall of retrospective data is ensured by the use of “fundamental events” analogues to event studies in general finance literature. We therefore provide an additional contribution to impact assessment and program evaluation. Overall results indicate a positive and significant effect of improved access to microcredits. For the business channel, we find that having better access to microcredit results in clients running more efficient enterprises in terms of higher business income and profits, and reducing the numbers of employees under competitive pressures. We also observe significant changes in household consumption patterns. Households with better access to MFIs tend to invest more in human capital, captured by expenditures on health and education, and to reduce spending on non-durable items such as weddings. The findings suggest that better access to microcredits improves household consumption patterns, which is in line with theoretical predictions.

Full Text: “Measuring the Impact of Microfinance” by Nargiza Alimukhamedova