Modeling firms' innovation decisions in developing countries.

摘要

This dissertation aims to study, understand and analyze firm level innovation in developing countries; with a special focus on Pakistani private firms. To do so, we present a model of heterogeneous firms making decisions regarding undertaking innovation. The model allows for firms to interact with each other in three different way. Secondly, it does not assume an immediate dispersion of profit upon innovation. The model is simulated using Non-Linear Model Predictive Control. This is used to calibrate the model with Pakistan's data. We use data from the Innovation Module of the Enterprise Survey conducted by World Bank for the country over the period 2013-2015. Further, we simulate different scenarios for the country, including fiscal action rewarding the adoption of innovative technology by new entrants. The model predicts market share and research effort by the heterogeneous firms. The research effort is explained by estimating the Knowledge function. This is done by applying a doubly censored Tobit regression on Total Research Expenditure per worker. This is the sum of intra & extra mural research expenditure, cost of training employees and the cost of acquiring equipment for innovation. We next estimate the probability of product and/or process innovation being conducted by Pakistani firms by estimating the Innovation function. This is done by estimating a bivariate Probit regression on the two types of innovation. These two dependent variables are binary. The predicted research effort, from the Knowledge function, is treated as a regressor here. We find that size and age of a firm significantly and positively affect research investment while corruption, measured as an index, is a significant inhibitor of research investment. Our estimations show that an exporting and medium sized firm is more likely to innovate. On the whole, research effort is low in Pakistani firms and the nature of innovation can be described as incremental. The current research effort by the firms exerts little influence on the probability to conduct innovation. The dissertation presents a detailed account of innovation in Pakistan. This exercise is done for the first time ever. Moreover, a comprehensive literature review of firm level innovation is also presented.