This study aims to provide empirical evidence on the causal relationship between bribery and firm innovation. To this end, we use a micro-dataset of small and medium firms in Vietnam surveyed in 2015. Given the binary nature of the dependent variable, a simple probit regression model is employed. However, as bribery variable is potentially endogenous, a simple probit regression may give biased estimates. We deal with the potential endogeneity by making use of the bivariate probit model. A property of the bivariate probit model is that it can produce efficient estimates of a typical probit model with endogenous binary explanatory variable. A Hausman-like likelihood ratio test is implemented following the estimation to test the existence of endogeneity. We find that bribery significantly undermines firm innovation. Also, firms run by household appear less innovative. The probability of innovation diminishes significantly if firm owners or managers have previous experience in firm products. As expected, larger firms seem to be more innovative. Exporters tend to be more innovative compared to non-exporters. Our findings provide support to the hypothesis that bribery is detrimental to firm innovation and, thus, innovation may be a mediating channel, through which, bribery impedes firm long-term performance.

This research aims to revisit the hypothesis that bribery hurts firm performance in the context of a perceptibly corrupt country. Specifically, we use micro-data from Vietnamese small and medium firm surveys in 2013 and 2015 to examine whether bribery impedes firm revenue growth and labor productivity growth. An issue arising in this type of research is the potential endogeneity between firm bribing behaviors and firm performance. To go around the issue, we follow the literature to instrument bribery variable with the average probability of bribery in other provinces. We further employ the Analysis of Variance technique (ANOVA) to unveil if the effect of bribery is dependent on bribing purposes. The regression results show that firm performance is significantly influenced by firm size, firm age and firm bribing behavior. Larger firms are more likely to grow faster while firm performance tends to be negatively related to firm age. Particularly, we find that bribery significantly impedes firm revenue growth and labor productivity growth. The analysis of variance shows that the effect of bribery on firm performance may vary across bribing purposes. Our findings, therefore, support the sand-the-wheels hypothesis that bribery hurts firm performance even in a highly corrupt business environment.