article published in Quadrant Magazine June 12 2011
Plotting a course for disaster
Economic modelling of climate change – what does it mean?
In his 7 June address to the Press Club Treasurer Wayne Swan claimed that new Treasury modelling projects economic growth per head to 2050 only 0.1 per cent a year less than otherwise under a carbon (dioxide) tax. This modelling is subject to serious questioning – or should be. Although that poses difficulties because the modelling has not been released, there are some obvious questions considered below.
But note first that it is passing strange that details have not been released because the outcome is the same as that previously modelled and published in detail in October 2008 by the Treasury and also by Treasurer Swan and the then Minister for Climate Change Wong. That modelling was published when the emissions trading scheme was in vogue, that is, before Swan and Gillard persuaded Rudd to drop that approach and Gillard took over and then switched to a carbon tax policy for the first 3-5 years. However, given that either the ETS or the carbon tax would target a similar reduction in emissions, there is no reason why the details of the modelling of economic growth should differ significantly from one to the other.
Treasurer Swan and government advisers have argued that their acceptance of the dangerous warming thesis requires structural changes in the Australian economy – and such changes necessarily affect economic decision-making by both businesses and consumers.
It is important therefore to assess the likely basic assumptions underlying any projection of economic growth to 2050.
For a start there is the question of at what rate the carbon tax will be set in the future and what other emission-reduction policies will continue. Swan adopts a rate of $20 per tonne of carbon dioxide emitted (but only as one possibility) and he says nothing about reducing other inefficient emission-reducing policies. Because such a rate of tax is highly unlikely, by itself, to effect significant changes in investment or consumer behaviour, to be emissions-effective it would clearly have to be increased over time. But now that the just published Productivity Commission report on Carbon Emission Policies in Key Economies shows Australia already has a middle range policy within that group of major emitters, why does the Government envisage Australia going ahead on its own and adopting a carbon tax that would put existing policies well out in front? Any such outcome would become even more questionable when account is taken of the fact that countries with whom Australia competes as a supplier of resources have very limited emissions reduction policies i.e. we are already way “ahead” of them.
The Productivity Commission analysis of key economies also exposes as false the attempts by Climate Change Minister Combet and other Ministers to portray China as having a “tougher” emissions-reduction policy than Australia. The Productivity Commission report shows this is only the case if “normal” increases in investment in electricity generation, including replacements of inefficient or worn out existing plants by plants which are almost all using coal, are counted as emission reducing because the improved efficiency provides that. Clearly, China is acting no differently to other countries in this regard. The aim of these policies is to reduce generation costs and the reduction in emissions is incidental.
The Productivity Commission analysis is consistent with other analyses indicating that many policies adopted by other countries have extensive exemptions, particularly for exporters. Indeed, in referring to the large number of emissions-reduction measures adopted in key economies, the Productivity Commission notes that despite having “material effects in terms of costs” to taxpayers ... “few appear to have had significant impacts on abatements” (emphasis added).
This means that it is just nonsense for Swan to claim that “seven of our top ten trading partners have adopted major policies to reduce pollution and support clean energy” (emphasis added). The analysis in the Productivity Commission report suggests that, while a $20 carbon tax might mean Australian policies would still be less “tough” than those adopted in the British and German outriders, neither of those countries has an economy-wide price measures and both source significant proportions of their electricity from nuclear power. But as the IPA’s Alan Moran has pointed out, our policies already produce abatement levels that are three times China’s, five times Japan’s, and seven times Korea’s. Why does the Government want Australia to become a leader in a team of players who, particularly if extended to include competing suppliers, are having little significant impact on abatements?
Returning to the modelling, the absence of detailed assumptions is of importance for a number of reasons. For one, the 2008 modelling stated that “carbon capture and storage (CCS) technology combined with coal and gas generation is assumed to be available on a commercial scale from 2020 in both Australia and the world.” This astonishing assumption has received no public attention despite the fact that the availability of CCS on a commercial scale would put a virtual end to the scare of dangerous warming because it would mean a major reduction in emissions. It is obviously necessary to know if that assumption forms part of the latest modelling and, if so, why Australia could not achieve emissions reductions by developing CCS rather than imposing a carbon tax.
Details of other assumptions regarding the variables in economic growth are also important. For instance, a key question regarding future economic growth under a carbon tax (or price) is how much productivity growth will slow. A basic object of an emissions reduction scheme is that it requires a switch from using the lowest cost form of energy (coal and oil) to one using a higher cost form (ranging from gas to solar power). Unless the efficiency of the latter sources improves (that is, unless they reduce the quantity of capital and labour now required to produce a unit of energy), productivity growth must slow. But the 2008 modelling seemed to make no attempt to take account of this issue or of analyses by prominent overseas economists. IPCC lead author Richard Tol, for example, claimed then that the cost of mitigatory action to 2100 would be about 40 times greater than the benefits. Depending on policies adopted, the possible adverse productivity effects still seem to be escaping attention in considering economic growth under the proposed tax.
Key to an assessment of all major growth variables is the question of what level the carbon tax is likely to reach. As already noted, the $20 per tonne canvassed as an initial rate will be far too low if the policy is to have any substantive effect in reducing emissions. But at the same time how high the end tax rate will likely be is important for forward planning. Unless that rate is known private investors will continue to experience the uncertainty many have complained about in recent times. Moreover, a rate that is likely to change every (say) 3-4 years after an election would make the economy more subject to fluctuations – and possible transitional periods.
Rival economic modeller Warwick McKibbin has also drawn attention to the possible effect of a carbon tax on employment, the growth in which Swan claims would not alter. Advice by the leading Treasury modeller to the Senate Estimates Committee is that the model simply assumes that real wages will fall and that full employment will then be restored. If that is the case it implies a lower rate of growth in wages, which is a not unimportant detail that unions (and others) should know about before considering the tax proposal. More importantly, serious questions must be raised about the validity of the advice. The modeller’s thesis is that, in any structural change, “there will be a transitional period in which employment may in principle move away from what is called full employment level but that adjustments in the economy bring it back to that level”.
This, of course, dodges the issue in a very serious way. The accompanying assessment given by the modeller that “the unemployment rate was not affected” in periods of past big structural changes overlooks the fact Australia has been through lengthy periods of less than full employment. Given the inevitability of structural change under a carbon tax (including at possible different levels), any modelling should at least include an assessment of the likely extent of the “transitional” period – or periods. Treasurer Swan’s emphasis on economic growth being little different whether there is a carbon tax or not undoubtedly reflects advice from his modellers that any problems with structural changes will be quickly overcome in a short transitional period.
In effect the modelling appears to assume away any “structural” problems on the basis that, over time, we humans adjust our behaviour to be consistent with the model. My recollection of modelling undertaken in Treasury when I was there is that it was generally reasonably accurate over a short period – but not beyond that, let alone over forty years. Even in the short term, “shocks” or sudden changes in the behaviour of a major component, such as has recently occurred with the GFC and the more recent sudden jump in the saving rate, can quickly destroy forecasts.
But the adoption of a carbon tax would be a structural change directly designed to cause changes in behaviour and produce “transitional” periods of adjustment. For this purpose the modelled outcome promulgated by Swan over 40 years is next door to useless.
Finally, Swan’s attempt to simply brush aside the not insubstantial loss of income over time from the modelled slower growth under a carbon tax implies that he believes there is no substantive economic case for reducing emissions. In turn, this means that the basis for such a policy rests almost entirely on acceptance of “the” science of climate change, to which Swan’s only reference is that it “is convincing”. Of course, economist Garnaut, who has become a political adviser not an independent economist, also claims it is now accepted “beyond reasonable doubt”.
Although as an expert economist Garnaut should have had the capacity to analyse many aspects of "the" science, his 2011 report includes no analysis of them. That report should have explained, for example, why, contrary to the science, no increase in temperatures occurred in about half of the last 60 years when emissions were increasing; why derived measurements have shown slightly higher temperatures in the past when fossil fuel use was low; and why Gillard and Combet are now projecting a “scare” increase in sea levels of up to 1.1 metre to 2100 when recent increases indicate a level lower than even the IPCC projection of its maximum 59 cms; and so on.
The meaninglessness of the economic modelling, and the failure to recognise the now widespread scepticism with the science, offer another example of this government’s failure and inability to change when the circumstances obviously require that.