Study of the Kuznets environmental curve hypothesis from a global perspective 1960–2019: a semi-parametric panel data proposal

Abstract

This work uses parametric and semiparametric panel data analysis methodologies to test the hypothesis of the environmental Kuznets curve, in 186 countries in the period 1960–2019. The main results reveal the acceptance of this hypothesis in the relationships of CO2 emissions (kt) and economic growth (GDP) and urbanization (% population) in the parametric models. Using semiparametric methods, the polynomial relations of fourth degree between CO2 emissions and GDP and of third degree between it and urbanization are verified. The economic policy implications derived from these results seem to indicate the need to continue making efforts in the reduction of CO2 emissions, through greater efforts in innovation and research and development, in search of clean and less polluting energies. The relationship between CO2 and economic growth is a major challenge, in terms of achieving a flattening of this relationship.

Appendix

(1) Spatial analysis: The existence of spatial autocorrelation of carbon emissions was explored through Moran’s index and through Geodata software (Luc Anselin). In eq. 7, the formula of Moran’s index is shown

$$ \mathrm{I}=\frac{\sum_{i=1}^N{\sum}_{i\ne j}^N{w}_{ij}\left({x}_i-\overline{x}\right)\left({x}_j-\overline{x}\right)}{\sum \limits_i{\left({x}_i-\overline{x}\right)}^2{\sum}_{i=1}^N{\sum}_{i\ne j}^N{w}_{ij}} $$

(7)

where $\overline{x}=\frac{1}{N}{\sum}_{i=1}^N{x}_i$; w_ij is the element on the ith row and ith column of the spatially weighted matrix w and N is the number of countries and x is the indicator of interest. The statistical significance of this index is checked with the z-score (comparison of Moran’s index and his expectation) and its p value. With the Geodata software used in this work, the index is obtained with bootstrap methods. If the index is found to be significant, the proposed methodology is followed.

Table 7 shows the values obtained for the initial, intermediate, and final decades 1960, 1990, and 2019, respectively.

Table 7 Moran Indexes on CO2 by decades

The lack of statistical significance of Moran’s indices through the statistical values of z scores and p values shows the absence of spatial correlation effects in relation to the CO2 variable. The maps in Fig. 1 show the CO2 levels of the countries analyzed. There is no grouping or clustering.