Regional and national socio-metabolic transitions in a globalized trading system

Abstract

In recent years the concepts of environmental sustainability, and of potential \planetary boundaries" with regard to Earth's ability to both continue to supply the material needs of human societies, and sink the waste products of those societies, have come increasingly to the fore. The extraction, processing and use of primary materials (i.e. those extracted directly from the environment such as biomass in the form of crops and wood, fossil fuels, metal ores, and non-metallic minerals) drive important environmental and economic problems such as deforestation, depletion of mineral resources and creation of toxic mine wastes, decreasing biodiversity, water pollution, and the build-up of greenhouse gases in the atmosphere, to name just a few. Crucially, as societies transition from basic agrarian societies to industrialized economies, they require much higher inputs of primary materials from the environment, and emit much more waste back into the environment. A large portion of the Earth's population is currently undergoing, or poised to undergo, this transition and so we are entering a period where the problems caused by the exploitation of primary materials should increase greatly. The ability to forecast the magnitude of this anticipated growth in material flows would be a valuable aid to planning to deal with these problems. Put simply, if you can't anticipate the likely size and timing of a problem, then it is very di cult to e ectively plan the size and timescale of the response required to mitigate the threat. To usefully inform potential policy making in this regard, the author judged that a relevant forecasting tool would need to be able to discriminate detail on primary material flows to at least the national level, and over multiple decades. The pursuit of such a tool provided the initial impetus for this project. One major existing theoretical framework, that of socio-metabolic transitions (SMT), appeared to have the potential to be developed into a practical forecasting tool, not least because some of the literature associated with it strongly suggested that countries at similar levels of development required similar per capita levels of material and energy inputs to sustain them. It became evident early on in this dissertation, however, that while the SMT framework provides a useful qualitative description of the evolution of material flow patterns during the process of industrialization, the idea that national material and energy flows were similar at similar levels of industrialization grossly oversimpli ed the reality. A number of reasons were subsequently identi ed for this, however the single most important one was the role of international trade in spatially separating the material and energy-intensive stages of production chains from end consumers. As a result, the SMT framework did not really have the potential to be developed into a useful forecasting tool for material flows. As there was no clear rival overarching framework similar to the SMT to replace it in this role, the emphasis of the project shifted, initially to investigating why the super cial potential of the SMT framework as a forecasting tool proved illusory, then on to exploring a second, more empirical path to forecasting, based on a much broader application of panel data modelling than had previously been used in this eld. An understanding of why the SMT framework could not be generalized into a predictor of material flows patterns was achieved via a series of case studies. Three of these examined the patterns and drivers of material flows, at the national level and over a period of decades, for three di erent World regions. In these studies, the relative importance of changes in population, a uence, and changing technological e ciencies in driving the growth of gross material flows, and in driving di erent mixes of individual materials categories (biomass, fossil fuels, metal ores, construction minerals), were important foci. A fourth study examined a set of three nations which were closely linked by mutual trade. This latter study detailed a major mechanism, the concentration of primary materials prior to trade, that in itself was su cient to rule out developing the SMT framework into an e ective forecasting tool. The exercise of extending statistical analyses revolved around combining the Stochastic Impacts by Regression on Population, A uence and Technology (STIRPAT) model, with panel data modelling techniques, using this combination to test a series of socio-economic variables for their explanatory power with regard to material ows. Both STIRPAT and panel analyses had been used in a limited fashion in previous studies on material flows, however they had typically been used separately, and restricted to a very limited range of potential explanatory variables. Some important ndings and questions arose from running these analyses and interpreting the results in the context of knowledge gained while performing the case studies. These include: While simple statistical models using only population and a uence perform quite well at explaining material flows at a global scale in cross-sectional models, they do not work well at a national level when time is also taken into account. They thus can't provide useful forecasts at a national level. None of the additional independent variables subsequently tested (including some relating to economic structure, biological productivity, climate, available land mass, and the level of education/knowledge of the population), improved performance of the models su ciently to change this conclusion. There is an important question over the validity of treating GDP as an independent explanatory variable in studies which seek to explain patterns of material and energy flows. The act of including GDP-based measures in regression analyses may in reality serve to mask more fundamental drivers of material flows. It also suggests that the apparent predictive power of a uence in cross-sectional analyses may rely on a partially circular logic, in part because the mere act of mobilizing material flows on a large scale will of itself contribute to GDP. The issue of circularity here appears likely to be very complex, and was not explored in detail as part of this project. Globalization of world trade means that the nation-state cannot be considered the appropriate unit for forecasting material flow trajectories. Not only is domestic extraction of primary materials often no longer governed by the demands of the domestic economy, but nations which have become major suppliers or importers of primary materials often develop in this way due to inherently unpredictable historical contingencies. Those historical contingencies, such as the emergence of a major export market for a particular commodity, often have their origin in factors exogenous to the nation where the extractive activity subsequently takes place. This further discounts the prospects for being able to forecast material flows from independent, national level variables.