1. Introduction


The relationship between environmental degradation and economic growth is a long standing debate. Victor (1972) emphasizes that economic growth should be regarded as a threat to the environment, as economic activities use environmental resources as inputs for production of goods and services, and produce waste beyond the environment’s absorptive capacity. Meadows et al. (1972) emphasize that, without major changes in socio-economic patterns, economic growth would not be sustainable because of exhaustibility of natural resources.


The main findings of the EKC literature reverse these conclusions: environmental damage first increases with income and eventually, after a turning point, decreases. As stated by Panayotou (1993) “At higher levels of development, structural change towards information-intensive industries and services, coupled with increased environmental awareness, enforcement of environmental regulations, better technologies and higher environmental expenditures, result in levelling off and gradual decline of environmental degradation”. Economic growth, in this view, is not a threat to global sustainability. Not only are there no environmental limits to growth, but economic growth is itself a means to attain environmental improvements (Stern et al., 1996). The factors that may offset the “scale effect” are crucial in determining the inverted–U relationship between income and environmental degradation. They are identified in (IBRD, 1992): (a) the structure of the economy (i.e. the goods and services produced in the economy), (b) the efficiency of resource use (i.e. inputs used per unit of output in the economy), (c) the substitution of resources which become scarce with clean technologies which can reduce the environmental damage.


The EKC debate has important policy implications. If the EKC hypothesis is correct, a solution to environmental problems could be economic growth. As noted by Ansuategi et al. (1998) “fostering economic growth should remain the paramount policy objective of the international community, and calls for concerted global action to curb environmental problems are misplaced”.


The EKC literature mostly consists of an empirical strand of specifications and estimation of a reduced form equation relating an environmental indicator to income per capita. For many pollutants there is not agreement on the EKC existence and for higher stages of economic development an increasing relationship is found (N-shaped curve), (Ansuategi et al., 1998).


Differences in these results reflect differences in methodology employed and model specification. Stern et al. (1996) stress the main problems with the estimation of an EKC. They concern the simultaneity of the model, data problems and international trade. Regarding the first aspect, the EKC does not consider the feedback from the state of the environment to economic growth. If this kind of feedback exists, i.e. economic activities depend on environmental resource use (Arrow et al., 1995), growth in the earliest stage of development could compromise the ability of a country to grow in the future. This determines the need for environmental policies and the crucial role of regulation in influencing the PIR. Furthermore, the use of observations which are aggregations over a number of subunits could give rise to heteroskedasticity problems. Finally, with the international trade some countries (normally the developed ones) will specialize in human capital and manufactured-capital-intensive activities, whereas others (the developing ones) specialize in labour and natural resource-intensive activities. As a consequence, part of the reduction of environmental damage in developed countries could reflect this specialisation.


If the EKC seems to be confirmed for lack of clean water and lack of urban sanitation, it does not hold for river quality, municipal waste and CO2 emissions. On the basis of these results, Shafik and Bandyopadhyay (1992) state that: “actions tend to be taken where there are generalised local costs and substantial private and social benefits”. In case of air protection, the EKC does not hold and policy intervention might be needed. Furthermore the PIR analysis does not allow policymakers to choose between policy alternatives because it focuses mainly on the link between economic growth and the environment, thus not considering the factors behind this relationship. For this reason in this paper we are interested in studying the factors determining the EKC in order to infer some useful guidelines for policy.


Another strand of the EKC literature consists of theoretical models, which focus on the interaction between environmental degradation and economic growth and include optimal growth, endogenous growth and overlapping generation models. In short, theoretical and empirical literature is aimed at verifying the EKC hypothesis for some pollutants and possible explanations for the turning point existence.


An interesting finding of the EKC literature is that the PIR is influenced by inequality among individuals (Magnani, 2000; Torras and Boyce, 1998). Torras and Boyce (1998) stress the importance of considering the income distribution for testing the EKC hypothesis. The link is investigated by considering environmental quality as a luxury good, whose demand increases with income. In these authors’ view, considering only average income is insufficient. With increasing income, greater consumption could bring increased pollution. Growing income, if we expect the EKC hypothesis to hold, should be channelled through environmental goods. Magnani (2000) states that “voters” preferences over consumption of private goods and public goods, such as environmental amenities, depend on their relative position in income distribution function. While growth in per capita income may increase the capacity to pay for environmental amenities (the absolute income effect), income inequality may drastically reduce a country’s willingness to pay (the relative income effect) by shifting the median voter’s preferences away from consumption of the public good “environmental amenity”. The study shows that a high demand for pollution abatement depends on an absolute income effect and a relative income effect, which impact on the ability and the willingness to pay for environmental protection.


This short overview clearly shows that the EKC literature encompasses the three sustainability dimensions (economic, environmental, social), which became more popular after the publication of the Brundtland report (WCED, 1987). The aim of this paper is to assess the economic, environmental, and social impacts of different EKC factors.


EKC determining factors are well captured by the Grossmans’s decomposition (1995). Grossman argues that the EKC could be decomposed into a scale effect (the polluting effect of the increasing economic activity scale), a green technology effect (the emissions reducing effect of the green technological change) and a composition effect (the emissions reducing effect of more fossil free oriented economy structures).


Technology plays a polluting and an environmental friendly role in the Grossman’s EKC factors decomposition. On one side industrial technology increases the added value generated by the conventional inputs, improves the scale of the economic activity and consequently increases the level of emissions. On the other side green technology is environmental friendly and reduces the pollution intensity over time. In this view the EKC is not a “black box” but opportune policies can be set about the EKC factors in order to tackle pollution (Panayatou, 1997). The crucial role of environmental policy in determining the PIR through technological development is opportunely highlighted by Arrow et al. (1995).


A wide literature in the EKC issue has been devoted to investigate the role of technical progress in determining the PIR. Lantz and Feng (2005) find for CO2 emissions that an inverted U-shaped relationship exists between CO2 emissions and the level of technology. Some decomposition studies (Howarth et al., 1991, Sun 1998, Bruha and Scasny 2006) show that the composition effect is small and may even increase emissions, whereas the technique effect is large and always decreases emissions. For this reason in this paper we will specifically focus on the role played by technology in determining CO2 emissions.


To reach this target we will not adopt econometric tools as the most previous studies, but we will use climate change integrated assessment models. The analysis by numerical simulations of climate change integrated assessment models to investigate the EKC is almost a novelty in environmental literature. Galeotti (2003) investigates two factors determining the EKC in order to assess their contribution on the PIR for developed countries: environmental friendly technology (expressed as the carbon intensity) and the environmental policy (viz. Kyoto Protocol) by two modified versions of RICE 96 including endogenous technical change. Our contribution in this paper will be to investigate another crucial factor determining the EKC: the industrial technology included in the production function and increasing the scale of the economic activity. Specifically our analysis tool will be the RICE99 model. Differing from Galeotti we link those factors not only to the PIR but also to their economic, environmental and social impacts.


Whereas an integrated analysis of economic, environmental and social issues has been widely debated by previous studies linking the EKC to the social dimension (Gangadharan and Valenzuela 2001, Jha and Murty 2003, Wu et al. 2006, Costantini and Monni 2006) only a few papers have been devoted to the climate change scenarios assessment on the basis of a wide set of sustainability criteria (Cantore 2005, Bosetti and Buchner 2005). This paper is aimed at filling this gap. The next section presents the RICE99 model, the section 3 shows the results, finally the conclusions.