time lost following the different procedures for establishing a new firm is proportional to the income per capita of the country. Since the income per capita is measured for one 14 In our subsample the degree of correlation increases.
30 year, the natural indicator is the number of days divided by 365. We add this calculation to the monetary cost as a proportion of income per capita informed by Djankov et. al. to get a unique measure 15 . The higher the value of this indicator the less efficient is the government. In column 4 we show the results of the regression that includes this variable in the list of regressors. Not surprisingly the coefficient of this variable is negative. It is also statistically significant. The less efficient the government the lower the rate of growth of TFP. Finally in model (5) we try to move our analysis to a more precise measure of the impact of technology on growth. The claim of most endogenous growth theorist is that their every reason to believe that the growth of technology depends on economic decisions at least as much a does factor accumulation. We implicitly have identified this growth in technology with the quality of institutions and economic policies but the ability of the various countries to innovate and catch up will be affected not only, for example, by the quality of the human capital but also by their direct efforts in doing those innovations and adaptations. The investment in R & D may be a good approximation to those efforts. We include the expenditures in R & D as a percentage of GDP averaged for the years 1984 to 1997 in our regression. The data comes from the World Bank. The estimated parameter is positive and statistically significant suggesting that increases in R & D may boost the rate of growth of TFP. To do a minimum check of the robustness of our results we run our model with two additional variables. The first one is an indicator of openness. We use the one built by Sachs and Warner (1995). Their indicator takes the value of 1 if the economy is open and the value of 0 if the economy is closed for each of the years between 1950 to 1992. We use as independent variable the proportion of years that the economies are open between 15 These probably underestimate the true costs for the different firms since their opportunity cost is probably much larger that the one reflected by the income per capita. Also, surely it differs from one firm to the other. In absence of a better alternative we keep this one but remain aware that this variable has to be improved. On the other hand, and in spite of these caveats we think that the variable remains a good indication of the efficiency of governments. It is really this dimension that we would like to capture.
31 1980 and 1992 16 . The other independent variable uses the Index of Economic Freedom.
- Summer '18
- Sagar Arora
- The Land, TFP