The empirical analysis shows that the for India emissions is a deterrent of economic growth but other factors are not statistically significant because India has not yet reached the threshold level of development where these factors play a bigger role in increasing productivity. Canada on the other hand shows a very different relationship with the predictors. Per capita GDP is a strong predictor of productivity and so is emission.
The above analysis is limited in its focus because it uses two countries only. Furthermore it is limited by data availability. An in-depth analysis with more variables, more countries and longer periods will give a deeper insight into the determinants and inhibitors of labour productivity and economic growth in the countries.The above model of India shows that neither government’s spending on education nor per capita GDP is statistically significant. Based on literature review, it implies that level of education in India is still very low, well below the threshold. As a result, despite government spending, the labour force remains widely uneducated or illiterate. In other words, the labour force does not realise the importance of education or is unable to derive benefits from education in the short run. This is why the model does not show any statistically significant relation between government spending on education and economic growth. Similarly, the threshold of per capita GDP is not yet reached for India because of which its relation with economic growth is not significant. The sign of emissions coefficient is negative which is as expected.
The model of Canada on the other hand offers very different results. While in the case of India, the square of log of per capita GDP did not show any significant contribution to the model and its removal actually increased the adjusted R2 of the model, in case of Canada it was exactly opposite. The addition of the squared term actually increased the explanatory power of the model significantly. GDP per capita has positive coefficient while its square term has negative coefficient. This is supported by literature. However, emission has a positive & significant coefficient in the model for Canada. This is because the correlation between the economic growth of Canada and its CO2 emissions is not statistically different from zero. However, the emissions in the country are quite low, much lower than India. As a result the people of the country have a stronger tolerance for small surge in the emissions of the country. Alternatively, the marginal cost of increased emissions in Canada is much lower than that of India. This is why the emissions coefficient of Indian model is negative while that of Canada is negative.