Innovation and firm-level productivity: econometric evidence from Bangladesh and Pakistan

The labour productivity impact of innovation of manufacturing firms in Bangladesh and Pakistan, a highly neglected region for such studies compared with the developed world, is studied in this paper by using World Bank Enterprise Survey data conducted in 2006. To achieve this end, we apply the Cobb-Douglas production function, augmented with innovation-related (and other expected sources of productivity) inputs in a three-equation simultaneous equations system – connecting R&D to its determinants, innovation output to R&D, and productivity to innovation output – and in a two-equation system – connecting innovation output to its determinants, and productivity to innovation output – after correction for the biases attributable to the selectivity problem of R&D and to the endogenous nature of both R&D and innovation output. Our results reveal that Bangladeshi firms are more often innovators as compared to Pakistani ones; however, the productivity output appears to be relatively large in Pakistan. We are generally not able to reject the constant returns to scale assumption. In addition, our econometric analysis indicates a strongly positive influence of both material and capital inputs on firm productivity; moreover, the productivity effect of process innovation is straightforwardly positive, but product innovation seems to be less connected to productivity outputs. Finally, we notice that the traditional production inputs (material and capital) have more significant effects on productivity output as compared to non-traditional input factors (controls in our case). JEL classification: O31; O32; O33; L60