Environmental Kuznets Curve Revisit: Role of Economic Diversity in Environmental Degradation

8 Background: Unlike the classical view, a new path of economic growth and development among 9 the emerging and developing nations seems to have distinct impact on environment. Customary 10 patterns of production and consumption have undergone significant changes and t he new “growth 11 with non-smoke-staks ” has put the developing economies on a path that can change Environmental 12 Kuznets Curve (EKC) fundamentally. With this view, the current study attempts to examine how 13 these growth patterns among developing world have impacted the degradation of environment. We 14 argue that including income per capita and share of manufacturing would not capture the full 15 growth dynamic of developing and emerging countries and therefore it masks the real impacts on 16 environmental degradation. To this end, we introduce the Economic Complexity Index (ECI) to 17 the model to reflect the full impacts of new growth approaches on CO2 emission levels by using a 18 panel data analyses of 100 emerging and developing countries over 1963-2018 period. 19 Results: The results indicate that complexity of the economies of developing and emerging 20 countries has added to the CO2 emission levels in absolute terms but it has helped to reduce the 21 CO2 intensity. 22 Conclusions: The implications of the findings for developing and developed countries could be 23 quite significant. For advanced economies, a downwardly-shifted Kuznets curve implies that, on 24 one hand, technology transfers have been successful in curbing the environmental degradation of 25 developing economies and, on the other hand, the economic transformation strategies of developing world is working in a sustainable way. (2016) investigated the EKC for 139 Indian cities using NO2 emissions as the proxy of environmental degradation for 2001 – 2013 period. The results confirm the EKC hypothesis, with reemphasis on the impact of growth-catalyzing economic policy decisions on environment.

developing economies and, on the other hand, the economic transformation strategies of 26 developing world is working in a sustainable way.  The classical view on industrialization of economics pioneered by Nobel prize laureate, Arthur 54 Lewis, was basically recommending on moving farmers from subsistence farms into factories and 55 commercial farms. This view, however, has undergone considerable changes as the scope for 56 "classic labour-intensive, export-oriented industrialisation" is narrower now. The model now is in 57 between a classic manufacturing and a subsistence farming for many developing countries, labeled 58 as "industries without smoke-stacks". Rodrik (2016) argues that automation, competition and 59 shifting demand are closing the door to countries wanting to copy Asia's miracle. Economic 60 transformation under this model is quite distinct from the Asian model and this has strong 61 implications for green growth and climate change. 62 The pattern of industrialization has undergone serious changes in nature. The classical pattern of 63 industrialization is not the common practice anymore in the way that the agriculture-64 manufacturing-services sequence shifted to agriculture-services-manufacturing or for some simply 65 agriculture-services. In other words, many developing countries are jumping over the second stage 66 of industrialization such that service sector constitutes more than 50 percent of the economy with 67 a very thin manufacturing sector. This arguably influences the shape of the classical environmental 68 Kuznets curve 4 (EKC) as manufacturing sector is the main contributor to the environmental 69 degradation compared to other sectors. As service sector is less polluting than manufacturing, it is 70 expected that developing countries experience a different version of the classical inverted U-shape 71 curve.

72
According to EKC theory there is an inverted U-shaped relationship between environmental 73 degradation and per capita income, simply indicating that at early stages of growth countries will 74 be facing higher environmental degradation, but as the economy grows the intensity of 75 environmental degradation will ease up.

76
Structural change and economic transformation to service and information-based economic 77 activities which are less emission-intensive than physical production and the growing ecological 78 efficiency of production and consumption by means of a 'greening' technical and technological 79 progress are perceived to be the main channels of reducing the intensity of damages to the 80 environment. On the other hand, product diversification, product complexity and economic 81 complexity are the main driving forces of structural transformations in the economy. Therefore, 82 economic growth and development with less adverse impact on the environment seems to be 83 happening and the relationship between GDP per capita and environmental quality depends on 84 scale, composition and technology effects of such transformation.

85
In this context, this study examines the Environmental Kuznets Curve for 100 emerging and 86 developing countries over the period 1990-2018 from a different perspective than other studies 87 had on the issue. More specifically, it attempts to see how changing the pattern of industrialization, 88 in particular, and economic growth, in general, affect the inverted U-shape curve between 89 environmental degradation and level of income of these countries.

90
To the knowledge of the authors of this study, this research is among the first studies to analyze 91 the relationship between economic degradation and structural transformation. More specifically, 92 our study is unique in the sense that it benefits from a highly cited indicator of economic 93 transformation known as Economic Complexity Index (ECI) developed by Harvard University.

94
To this end, this research aims at investigating the interactions between economic complexity and 95 economic growth and it impact on environmental degradation.

96
The remainder of this paper is organized as follows: Section 2 provides a literature review. Section 97 3 surveys economic complexity and environmental degradation. Section 4 describes the data and 98 variables definitions, as well as methods for empirical analysis. Section 5 presents the empirical 99 results and discussions, and Section 6 concludes the paper with some key policy implications. As far as the authors of this study know, the majority of the exiting literature on environmental 103 Kuznets curve has used share of manufacturing in GDP as a representation of industrialization to 104 study the impacts of industrialization on economic degradation or they simply looked at the 105 association between GDP per capita and environmental degradation without digging into on how 106 the growth takes place. GDP per capita per se can not reflect the dynamics of the growth dynamics 107 happening differently across countries.

108
A big body of theoretical models can be found in the literature aiming to explain the environmental 109 Kuznets curve (EKC). The EKC hypothesis was empirically tested for many countries using   In another study Grossman and Krueger (1995) find that for a number of environmental variables, 127 the relationship between per capita income and environmental degradation takes an inverted U-128 shape confirming that environmental quality initially worsens but ultimately improves with  Germany, for the period of 1960 to 1993. They found that the EKC is not generally fit for all the 136 countries, as each country has its own technological, structural, energy price and economic growth 137 path, and therefore the emission situation and path would not be the same for all.

281
The empirical model that will be tested is shown in the following form: to the model to capture the probable impacts of ECI on the shape of the Kuznets curve. The model 300 also includes region dummies to capture any region-specific effects.

301
The effect of ECI on the shape of the EKC will be assessed by derivate of the equation (2) above 302 with respect to Income per capita as follows: For which, any significant figures for 1 and/or 2 indicates a change in either intercept 305 or the slope (or both) of the EKC.

306
A random Generalized Least Square Model was used on an unbalanced panel data covering 100 307 developing and emerging economies across the globe over 1963-2018. 6 The availability of the data 308 for various variables dictated the choice of the countries include in the model.

309
It is worth noting that in our model and for the Emerging group we included most populous 310 countries of BRICS plus Malaysia, and Turkey called BRICS+. The reason for adding these 311 countries to the BRICS is that their economies are different from their counterparts in their regional 312 grouping either size-wise and or economic-structure-wise.    As shown in Table 1 below, on a grouping basis, Africa with an average ECI of -1.002 has the 332 lowest score of ECI, followed by South Asia with -0.6063, Central Asia with -0.3308, and MENA 333 with -0.3155. On the other end, Eastern Europe with 0.4455 has the highest score of ECI, followed 334 by , BRICS with 0.1767 and South America with -0.2335.  At first glance and as it can be seen in Figure A.1 in appendix, ECI is associated with higher 345 environmental degradation evidenced by the simple two way scatter diagrams. Our model below 346 investigates this correlation in a more rigorous way.

370
Our variable of interest, ECI, has been associated with more adverse environmental impacts, 371 implying that the whole growth dynamics of developing countries have worsened the environment.

372
In other words, the economic transformation process of developing economies, on average, have 373 deteriorated their environment.

374
To dig more into the case, we included the interaction of ECI and income per capita. It is interesting 375 to see that ECI has flatten the Kuznets curve by downward shifting of the EKC evidenced by 376 negative sign of the interaction term between ECI and the income per capita and more clearly 377 presented in Figure 2 below. Put it simply, the economic dynamics of developing and emerging 378 economies has still led to inverted U-shape curve but it has shifted the curve downward, implying 379 that the process of environmental degradation has been lessened and reduced. The change is 380 significant evidenced by its t-value. This impact is only observed for the case where CO2 per capita 381 is used as the proxy for degradation.

382
The effect is opposite for the case where CO2 in kilo tonnes represents the environmental        Note: *, + and † represent 1%, 5%, and 10% level of significance, respectively.