Selection bias and endogeneity issues on the relationship between information technology and firm performance

This study extends prior research on the relation between information technology (IT) and firm performance by using both univariate and multivariate econometric models to assess the hypothesized relationships. Additionally, sample selection bias and endogeneity are examined to determine their effect, if any, on the results. The univariate results indicate that, on average, IT leaders outperform non-IT leaders. After controlling for sample selection bias and endogeneity, using Wooldridge (2002) 2SLS-IV, the coefficient of the endogenous variable is higher than suggested by ordinary least squares estimation and the Hausman F-Test is significant, indicating that the relationship between IT and firm performance is endogenous. Thus, it is important to control for sample selection bias and endogeneity to properly estimate the relationship between IT and firm performance.