A higher GDP means more emissions

Decoupling so far: at best only partial. Despite the fact that decoupling has been on the sustainability agenda for nearly thirty years, the attempts to compensate, through increased eco-efficiency and sub-stitution, for the increased pressure against the environment resulting from eco-nomic growth have only been able to slow down the environmental degradation. Within most sectors, and for most impact categories, the annual environmental load is higher today than thirty years ago, both in total and when measured per capita. The accumulated environmental degradation and resource depletion, for example in terms of concentration of CO2 in the atmosphere, extinction of species, loss of farmland, and exhaustion of finite energy sources and minerals, are far more exten-sive today than in 1987 when the Brundtland Commission published its sustain-ability report.Admittedly, rising GDP levels accompanied by reduced emissions have been found for certain pollutants at a local scale. However, for a number of other types of environmental impacts there is far less evidence of such decoupling. Minimizing The illusion of green capitalism 177the consequences on one impact category may also sometimes result in increased environmental load on another impact category. According to Spangenberg (2001), no indication of an environmental Kuznets curve can be found when using meas-ures accounting for the total use of energy, materials, and land. During the period 1974–2013, global CO2 emissions increased steadily along with economic growth, while there was non-growth or reduction in the emis-sions in three periods of economic downturn (in the early 1980s, 1992, and 2009). For the first time during the whole forty-year period in which the International Energy Agency has been collecting data on carbon dioxide emission, 2014 showed no increase in CO2 emissions despite economic growth (IEA, 2015). It remains to be seen whether the 2014 result, which IEA attributes to greater generation of elec-tricity from renewable sources in China, and increased energy efficiency and more renewable energy in OECD countries, will continue. Anyway, zero-growth in CO2emissions is far from sufficient in a situation where a reduction of 90 percent or more within the next fifty years is called for in order to avoid “dangerous climate change.” Moreover, without the 3 percent global growth in 2014, the emissions would probably have decreased instead of remaining constant, as they did in 2009 when the global economy also had a downturn due to the financial crisis.In a somewhat older study, Azomahou and Van Phu (2001) investigated economic growth and CO2 emissions in one hundred countries during the period 1960–1996. Using non-parametric regressions, they found a correlation between economic growth and increased CO2 emissions among poor countries, middle-income coun-tries, and wealthy countries. Within all three groups of countries, the CO2 emissions tend to increase with higher gross domestic products.To the extent that correlations between affluence levels and emissions of some pollutants (notably NO2, carbon monoxide, and sulfur dioxide) consistent with the EKC hypothesis have been found, this may in part be due to the limitation of many such studies to a confined geographical scale (e.g. cities). According to Stern (2004), the total pollution emissions tend to rise with increasing income rather than follow an inverted U-shaped curve. Pollution concentrations may still follow the curve predicted by the EKC hypothesis due to urban and industrial decentralization, taller smoke stacks, and so forth. Moreover, reduced emission levels in wealthy countries may be a result of outsourcing polluting industries to third world countries. Another case in point is the fact that the emission reductions obtained for impact categories where some evidence for the EKC hypothesis can be found are invariantly resulting from strong governmental regulations. While the EKC hypothesis is often used in the environmental debate to create an image of a “self-regulatory” market economy, the empirical examples referred to in supporting this hypothesis were dependent on precisely the type of regulations that are usually fiercely opposed by its proponents.The fact that dematerialization to a degree sufficient to counterweigh the impacts of economic growth has not taken place until now does not in itself rule out the possibility for a stronger and more successful dematerialization in the future. The prospects for this still do not seem very promising in the light of the nearly forty-five years that have passed since the publication, and the nearly thirty years since the publication of the Brundtland report. Decoupling growth from environmental degradation has been on the agenda during the very same period when only partial or local evidence of the EKC hypothesis has been found. Moreover, many of the lowest-hanging fruits have probably been picked in this period. Improving eco-efficiency further will often require technologically more challenging and socially more controversial measures. This may be the reason why some more recent studies indicate that in the longer term, the Environmental Kuznets Curve tends to be N-shaped instead of following the originally hypothesized inverted U shape. This suggests that pollution increases as a poor country becomes industrialized, decreases once a threshold affluence level is reached, and then begins increasing again as national income continues to increase.In much of the debate about economic growth and decoupling, there has been a tendency to focus solely on emissions of various pollutants (and in recent years a rather one-sided concentration on CO2 emissions only), whereas nega-tive environmental impacts resulting from land take and land fragmentation have tended to be overlooked. The prospects for decoupling growth from the latter kinds of impacts may be even less bright. This is particularly the case for the direct impacts on biodiversity, natural areas, food-production soil, and landscape qualities from converting undeveloped land into building sites or technical infra-structure such as roads and power plants. The few studies carried out on loss of biodiversity suggest that further economic growth in an already wealthy country or region is accompanied by higher biodiversity losses (Mills and Waite, 2009; Tevie et al., 2011).Theoretically, it is possible to imagine a “Factor 4” or “Factor 10” reduction of the land use per produced unit by building more densely and higher, as well as pla-cing most of the buildings in areas that are already technically affected, such as traffic areas and derelict industrial estates/brownfields. However, unless the new buildings or infrastructure replace existing, less environmentally favorable parts of the built environment, construction of new, eco-efficient buildings or infrastructure will at best be ecologically friendly in a relative sense, i.e. compared with environmentally more straining solutions (Xue, 2014; see also Chapter 8). In the USA, farmland is reduced ten times faster than the influx of new cultivated land, and in China the reduction of farmland is 30–40 times faster than the supply of new soil for food production (Steigan, 2011). Phosphorus is another natural resource of key import-ance to food production. According to Cordell et al. (2009), the peak of phosphorus production may be reached already within 30–40 years, and the known sources of phosphorus may be exhausted within 50–100 years.Some of the raw materials used in renewable energy production, such as lithium, are also becoming increasingly scarce (Defra and BIS, 2012). Along with a number of other so-called rare minerals (including beryllium, cobalt, and tantalum) used in computers, mobile phones, aircrafts, communication satellites, and rechargeable bat-teries, lithium is on the “critical list” identified by a panel of sixty-nine executives of key global companies in seven different industries (PricewaterhouseCoopers, 2011).