Determinants of Profit Persistence and Total Factor Productivity

Abstract

This study examines the determinants of total factor productivity (TFP) using the US Compustat database during the period 1976-2009. Various econometric approaches are followed. Initially, a Stochastic Frontier Analysis (SFA) for a production function using two conventional inputs-labour and capitalis applied in order to calculate the TFP. Secondly this study uses panel data analysis and more precisely the Generalized Methods of Moments (GMM) system. Both linear and quadratic models are examined in order to test for non-monotonic relationship. Besides the drivers of TFP traditionally employed in other researches, a relatedness measure of technology based on patents is explored, in order to capture the information spillovers from the effects of networks and business activities through in more or less related technologies. In addition, the determinants of TFP are examined during economic recession and non-recession periods. The overall results indicate that both firm and industry characteristics can be important factors for TFP. Therefore, managers should account for drivers who affect both the individual firm and the sector or industry where the firm is active. As such, organizations can monitor productivity for strategic reasons such as corporate planning, organization improvement, or comparison to competitors. It can also be used for tactical reasons such as project control or controlling performance to budget.