Liberalization, bankers' motivation and productivity: A simple model with an application

- An evaluation of the effects of liberalization on bankers' motivation and productivity is vital.
- The literature, so far, examines banking efficiency and productivity mainly through DEA.
- We model banks as profit cum welfare maximizing; derive bankers' optimal effort and productivity.
- As a test case, we implement our model on Nepalese banks; find liberalization to be a success story.
- We provide a micro-founded common analytical framework and believe it will motivate future work.

Proponents of financial liberalization argue that deregulation motivates bankers to increase their effort and operate at a higher level of efficiency and productivity. Sceptics, however, see that liberalization engenders economic instability and banking crises, and impedes growth. Bank efficiency and productivity, following liberalization, is extensively examined. Nonetheless, the core issue of bankers' self-motivation remains implicitly assumed and unaddressed. Does liberalization self-motivate bankers and increase their efforts and productivity? This paper models bank productivity from this perspective and evaluates what proportion of banks' total factor productivity is accounted for by the self-motivated productivity of bankers. We provide a micro-founded framework for the analyses of bankers' optimal level of effort and effort-driven productivity. Our model also captures banks' unit input-output prices, optimal wages, bank spread and the overall cost of bank services - measures that are important in evaluating reform policies. We assess the financial liberalization of Nepal as a test case and find that (i) bankers' efforts and productivity have notably improved in Nepal, although banking services have become costly, and (ii) bank spread has moderately declined in recent years. Our approach is parametric which differs from DEA, hence complements the literature. We hope this analytical framework will be useful to evaluate reform episodes elsewhere.