Quantitative Economics How sustainable are our economies by Peter Bartelmus_9

194 10 Diagnosis: Has the Economy Behaved Sustainably? 40 30 20 10 0 1987 1988 1989 1990 1991 1992 −10 years 1993 1994 1995 1996 1997 1998 efficiency structure volume total −20 Fig. 10.4 Decomposition of annual changes in carbon dioxide emission, the Netherlands, 1987– 1998 Source: de Haan (2001), fig.2; with permission by the copyright holder, Taylor & Francis: http:// www.informaworld.com. The calculation of direct and indirect environmental impacts and the decomposi-tion of their annual changes into driving forces provide information about the origin and causes of pollution and natural resource use. Such information is useful for set-ting priorities for environmental protection and natural resource management. It also permits to relate environmental impacts and their changes to the results of economic activities responsible for these impacts. However, any assessment of sustainability by showing the delinkage of economic growth from environmental concerns requires comprehensive analysis of all significant environmental impacts. So far almost all input-output studies deal only with selected pollutants and a few energy sources. The reasons are lack of data and time/cost constraints for compiling or estimating the data. An aggravating factor is the notorious reluctance of corporations to provide informa-tion about their production processes and environmental impacts. On the other hand, the policy relevance of input-output analyses and the close-ness of input-output tables to the supply/use accounts seem to have seduced national accountants into embracing input-output modelling for greening their physical accounts. The revised SEEA-2003 presents thus the backward modelling (‘backcasting’) of direct and indirect environmental impacts and their decomposi-tion as part of hybrid accounting, without discussing the underlying model assump-tions (United Nations et al. in prep., ch. 4). There is perhaps some realization that simply juxtaposing environmental statistics and economic indicators in hybrid accounts does not really green the accounts themselves. Further Reading 195 It is important, however, to clearly distinguish between more or less objectively observed statistics and modelled information. Even backcasting represents a dis-tinct step away from reality into a hypothetical situation. An even greater step is modelling that reaches into the unknown future. The next chapters will make this step as transparent as possible by remaining closely linked to the national accounts and their input-output tables. Further Reading FR 10.1 Input-Output Tables and Analysis Leontief (1951) developed input-output analysis for describing and explaining the structure of a market economy. He was also one of the first to introduce environ-mental pollution and its control into the input-output model (Leontief, 1970). The United Nations System of National Accounts (SNA) (United Nations et al., 1993) devotes one chapter (ch. XV) to the relationships between the supply-and-use table of the national accounts and input-output tabulations. A handbook of the SNA (United Nations, 1999) elaborates on these relationships (including squaring tech-niques). The handbook also presents a clear, practical review of the concepts, methods and compilation of input-output tables, including the calculation of a greened GDP (see Section l2.1). Many economic and mathematical textbooks include descriptions of methods and uses of input-output analysis. Lahr and Dietzenbacher (2001) present exten-sions of input-output models into regional and environmental analysis. The European Network of Environmental Input-Output Analysis focuses on the use of input-output techniques for life cycle analysis: http://www.leidenuniv.nl/cml/ssp/ projects/envioa/proceeding1.pdf. FR 10.2 Decomposition Analysis Decomposition techniques, applied to input-output tables, have become known as structural decomposition analysis (SDA). SDA can in fact be traced back to early work of Leontief: see for a brief history and methodological review Rose and Casler (1996). Typically, SDA explains changes in output and employment, but more recently the focus has been on natural resource use (especially energy) and pollution. Dietzenbacher and Los (1998) describe the techniques applied in the illustrative example of pollution in the Netherlands (de Haan, 2001). Rørmose and Olsen (2005) apply SDA to CO , NO and SO emissions in Denmark; they also give a detailed description of the applied input-output and decomposition methods. 196 10 Diagnosis: Has the Economy Behaved Sustainably? Review and Exploration What do the data tell us: has economic growth been sustainable? Evaluate the results of green (physical and monetary) accounts for the assessment of sustainability. What is the purpose of structural decomposition analysis of environmental impacts? Why is it important to distinguish between accounting and modelling? Is the calculation of total (direct and indirect) emissions from an input-output table a modelling or an accounting exercise? Explain the difference between ex post (descriptive) and ex ante (predictive) modelling. Chapter 11 Prediction: Will Economic Growth Be Sustainable? Looking back in Ch. 10 at the causes for past environmental impacts makes it possible to stay close to the observed data while using the powerful tools of input-output analysis. The real challenge for sustainability policies is however to predict future trends and anticipate the success or failure of policy options. The trade-off of taking the analysis to the future is the need for additional assumptions that remove the models further from reality. The focus of this book on quantitative assessment justifies concentrating on those analyses that build upon empirical data. Econometrics, as its name implies, is modelling that remains closest to measure-ment (of parameters and trends). The well-known and fiercely contested Environmental Kuznets Curve hypothesis about the correlation of economic growth and improvement of the environment provides a good example for econometric tests. The testing of this hypothesis is mostly applied at the national level. At the global level, the popular Limits-to-Growth model also uses some econometric parameter testing. It is a dynamic simulation model, criticized, however, for mostly relying on untested feedback loops and exponential growth assumptions. The model has become known for its rather pessimistic baseline scenario of ‘overshooting’ environmental limits and consequential social collapse. Chapter 12 analyses how these limits can be addressed while seeking optimality and sustainability of economic activity. 11.1 Econometrics: The Environmental Kuznets Curve Hypothesis Do we need ‘economic growth that is … socially and environmentally sustainable’ (WCED, 1987), or is it ‘qualitative’ development (Daly, 1996)? The first statement pleads for sustainable development with, the second without, ‘quantitative’ economic growth. As discussed in Ch. 2, the two views reflect the prevailing dichotomy between ecological economists, who see economic growth as the cause of environ-mental degradation, and environmental economists, who see it – with some modification – as the solution. The continuing discussion of the Environmental P. Bartelmus, Quantitative Eco-nomics, 197 © Springer Science + Business Media B.V. 2008 198 11 Prediction: Will Economic Growth Be Sustainable? Kuznets Curve (EKC) hypothesis (cf. Section 2.2.2) reflects the search for evidence in support of either view. The protagonists of the hypothesis claim that there is ‘no evidence that environmental quality deteriorates steadily with economic growth’ (Grossman & Krueger, 1995). Rather, they find an inverted-U relationship where environmental impact increases at low levels of national income and decreases at higher ones. This section reviews critically the testing of the EKC hypothesis since it is implicitly, and sometimes explicitly, at the heart of the environmental-economic dispute. The relatively simple and transparent ‘metric’ analysis points to the crucial role of empirical data; it reveals also the limits of correlational methods for explaining the causes of environmental trends. 11.1.1 Regression Analysis: Testing the Hypothesis Figure 10.2 puts question marks behind the decoupling of primary materials use from economic growth. Regression analysis is the tool of assessing empirically the potential linkage or delinkage of economic growth and environmental quality. For testing the EKC hypothesis, cross-country data typically establish the parameters of the regression function. The next step is to apply the parameters to time series of the explanatory variable of GDP (or national income) per capita. The emissions or con-centrations of different pollutants E represent usually the environmental impacts during a year of income (Y = GDP p.c.) generation. A quadratic equation then specifies the EKC hypothesis as Ei =b1 +b2Y +b3Y2 +ei (11.1) As in most regression analyses the error term e represents all those influences (in this case on changes in environmental quality from pollution), which are excluded from the analysis as unknown, peripheral or due to data deficiencies. This simplifi-cation, inherent in a ‘reduced-form’ equation of one dependent and one explanatory variable, has been the focus of critique and rejection of the EKC. Adding a further cubic term for Y allows to test the case of relinkage, introduc-ing a possible second turning point after an initial EKC phase: Ei =b1 +b2Y +b3Y2 +b4Y3 +ei (11.2) Figure 11.1 illustrates the cases of EKC confirmation (the inverted U) and rejection (because of relinkage). Assuming e to be constant, we Obtain, for b = 0, b > 0 and b < 0, the EKC of graph A May obtain, for b , b and b ¹ 0, the so-called N-curve of relinkage (graph B) Reject the EKC hypothesis for b ¹ 0 and b ,b = 0, owing to a linear positive or negative association of environmental impact and economic growth. ... - tailieumienphi.vn