Financial Development and Economic Growth: Evidence from Heterogeneous Panel Data

Abstract

The sharp disagreement in economics literature about the nature of the relationship between financial development and economic growth is widely known. Empirical evidence is also mixed as has been documented by Levine (1997, 2003b). Most empirical studies focused either on indirect finance or on direct finance. Previous panel data studies also failed to permit heterogeneity in slope coefficient. Past empirical studies even ignored the inflation effects on the relationship between finance and growth. This dissertation examines the empirical relationship between financial development and economic growth while incorporating the inflation rate effect on financial development; dividing countries into panels of Low, Lower Middle, Upper Middle, and High Income Countries. It focuses on both the indirect finance and the direct finance, separately as well as collectively. The econometric methodology of Weinhold (1999) and Nair-Reichert and Weinhold (2001) is applied for causality analysis in heterogeneous panel data which is based upon the Mixed Fixed Random Effects model of Hsiao et al. (1989). Two sets of results are reported: First, the relationship between financial development and economic growth from contemporaneous non-dynamic fixed effects panel estimation can be interpreted as mixed. Negative and statistically significant estimates of the coefficient of the inflation and financial development interaction variable, in the case of Low and Lower Middle Income Countries, indicate that financial development may be harmful to economic growth when inflation is rising. Such evidence is not found from the data for Upper Middle and High Income Countries. Second, in contrast to the recent evidence of Beck and Levine (2003), use of a more appropriate econometric methodology of dynamic heterogeneous panel for causality analysis and a refined model reveals that there is no definite indication that finance spurs economic growth or that growth spurs finance. These findings are in line with Lucas’s view on finance that the importance of financial matters is over-stressed in popular and even professional discussion. The only exception is the activity in stock markets in High Income Countries, where the result supports the Robinson (1952) view that finance follows where enterprise leads.