کارایی محیط زیست و بهره وری نیروی کار : تجارت کردن یا دینامیک های مشترک؟ تحقیقات نظری و شواهد تجربی از ایتالیا با استفاده از NAMEA

In this paper we test an adapted EKC hypothesis to verify the relationship between ‘environmental efficiency’ (namely emissions per unit of value added) and labour productivity (value added per employee). We exploit NAMEA data on Italy for 29 sector branches and 6 categories of air emissions for the period 1991–2001. We employ data on capital stock and trade openness to test the robustness of our results. On the basis of the theoretical and empirical analyses focusing on innovation, firm performances and environmental externalities, we would expect a positive correlation between environmental efficiency and labour productivity — a negative correlation between the emissions intensity of value added and labour productivity — which departs from the conventional mainstream view. The hypothesis tested is a critical one within the longstanding debate on the potential trade-off or complementarity between environmental preservation and economic performance, which is strictly associated with the role of technological innovation. We find that for most air emission categories there is a positive relationship between labour productivity and environmental efficiency. Labour productivity dynamics, then, seem to be complementary to a decreasing emissions intensity in the production process. Taking a disaggregate sector perspective, we show that the macro-aggregate evidence is driven by sector dynamics in a non-homogenous way across pollutants. Services tend always to show a ‘complementary’ relationship, while industry seems to be associated with inverted U-shape dynamics for greenhouse gases and nitrogen oxides. This is in line with our expectations. In any case, EKC shapes appear to drive such productivity links towards complementarity. The extent to which this evidence derives from endogenous market forces, industrial and structural change, and policy effects is discussed by taking an evolutionary perspective to innovation and by referring to impure public goods arguments.

In this paper we test the hypothesis of an ‘adapted’ ‘Environmental Kuznets Curve’ (EKC) in which the correlation between labour productivity (value added per employee) and environmental efficiency (emissions per unit of sectoral value added) is the link being analysed. The dynamic relationship between the abovementioned ‘efficiencies’ is a core, if not the primary, element behind the observed macro EKC trend. The role of technological (eco-) innovation as a latent factor in this relationship has been highlighted in empirical and theoretical contributions (Karvonen, 2001). Here we specify an empirical model for an examination of an original NAMEA (National Accounting Matrix with Environmental Accounts) sector-level time series panel dataset. Emissions per added value is used as a proxy for environmental efficiency/productivity (environmental intensity of value added generated). The underlying assumption is that the core direction of economic change is towards higher mechanisation (capital/labour ratios) (Pasinetti, 1981) and higher labour productivity, testing whether environmental efficiency is positively or negatively related to labour productivity dynamics (Femia and Panfili, 2005). Empirical analyses of joint economic and environmental productivity at sector level are quite rare due to the paucity of (panel) environmental data. This constitutes an added value of our paper. We argue that firm-based studies (Mazzanti and Zoboli, 2008 and Mazzanti and Zoboli, in press) and sector-based analyses provide highly complementary evidence, given that the former focus on specific issues and allow greater detail, whereas the outcomes of the latter are more of general flavour (Table 1a).The paper is structured as follows. Section 2 provides a theoretical framework for the empirical analysis and describes the dataset. Section 3 presents the panel-based regression results. Section 4 discusses the factors that support the emerging stylised fact of a joint economic and environmental productivity, offers some interpretations and discusses some open issues. We expect to find robust statistical evidence of the ‘double productivity/efficiency hypothesis’, i.e. an inverse relationship between emissions intensity and labour productivity, although an articulated set of differences across different pollutants and between industry and services may emerge.

We find that for most NAMEA emissions there is a positive relationship between ‘labour productivity’ and ‘environmental productivity’ (emissions efficiency) in the Italian experience. We show that this macro-aggregate evidence is driven by sector dynamics in a non-homogenous way, across pollutants. If services tend to show always a ‘complementary’ relationships between efficiency of emissions and labour productivity, industry is to some extent characterised by inverted U-shaped dynamics for GHG and NOx. This evidence fits with our expectations. The prevailing technological dynamics is one in which the intensification of capital in the Italian economy has led, ex post, both to increasing value added per employee and to reducing air emissions per value added, which corresponds to the descending part of an EKC in these two variables, or to an EKC pattern in which a jointly increasing productivity has substituted for a trade-off between value added and environmental efficiency. This stylised fact on joint economic–environmental productivity for NAMEA emissions across production branches seems to depart from the conventional neoclassical trade-off between ‘optimal’ allocations in terms of labour productivity, and allocations aimed at reducing emissions. Of course, our results cannot exclude that, ex ante, single firms face a trade-off in allocating investments, and there are opportunity costs of investing in environmental efficiency. However, these trade-offs are not observable outside of a firm-based information set. The evidence we have provided is on ex post joint productivity trends at the level of production branches, but it also calls into question the existence of systematic trade-offs between different kinds of productivity at the level of firm strategies, as well as the separability between optimisation with and without the environment. However, our results are independent on any a priori assumption about the relationship between capital, energy and labour possibly arising either from neoclassical approaches or heterodox approaches. We discussed certain factors behind the stylised fact that emerges from our analysis. The key suggestion is that at the roots of the (joint) dynamic between environmental and labour productivity different types of innovation play different roles, with a possible key role of the ‘impure public good content’ of R&D processes. Major restructuring processes in the economic system, and environmental policies are further credible ‘drivers’, with the latter possibly provoking ‘anticipating’ innovation strategies by some firms and sectors, both for air pollutants and GHGs. The dynamics of exogenous energy market forces can be added as an underlying determinant of energy/emission savings through capital intensification. The motivations of a joint productivity dynamics may also depend on the links between different innovations, which make the one conditional on the other and can prevent full separability of production factors and innovative strategies. These mechanisms should be clarified by further empirical research at firm and industry level, to provide a fuller explanation of the joint dynamics we have presented above.