Why carbon taxes won’t work

by Cliff Stainsby

Second in a series on climate policy

“Ban Carbon Emissions, Don’t Price them: Why Cap and Dividend is the Best Approach” identified five criteria for evaluating various plans to reduce heat-trapping greenhouse gas emissions: scale, urgency, certainty, simplicity/transparency, and fairness. A response suggested that my claim that carbon taxes do not meet these five criteria was mistaken, and in particular that carbon taxes can be simple and transparent.  I maintain they cannot, for the following reasons.

The context is that climate science strongly indicates that the survival of our civilization requires us to stop dangerous global warming by entirely eliminating greenhouse gas emissions by 2050.  Getting ‘close’ to that target is no better than missing it by a great deal; anything less than a 100-percent or at worst a 95-percent reduction by 2050 is a failure, because of the risk of positive feedback loops.  Many analysts, including Dr. James Hansen of NASA, think these are already happening and that we are in ‘overshoot’ with a very limited time to get atmospheric CO2 levels back down to 350ppm from the current 385 ppm.

In this urgent context, my response to any assertion that ‘simple and transparent carbon taxes’ can achieve the necessary goal, is the following sincere challenge: please describe a non-complex, transparent and fair carbon tax regime that can deal with the problem. For the following reasons I believe it cannot be done. 

Most carbon tax regimes and proposals set prices far too low. The Princeton economists who developed the Cap and Dividend proposal referenced in “Ban Carbon Emissions, Don’t Price them” have suggested that a $200-per-tonne ($200/tC) tax would obtain just a 7% reduction in emissions. A recent Australian study suggests that, for transportation, a $600/tC carbon tax will get in the order of a 50% reduction.  Based on these estimates, what tax would be required to achieve a 95-to-100-percent reduction?

For reasons I’ll expand on next, we cannot reasonably say, but we can readily tell that the figure would be far, far too high to be palatable to anyone, particularly politicians.  And if such a tax were applied, the consequential tax amendments for equity adjustments would be so convoluted as to void any hope of simplicity.

There are many other reasons why carbon taxes won’t work.  Economists (including those in government finance departments) determine taxes and their impacts using economic models that rely on many basic inputs including, for example, forecasts of changes in Gross Domestic Product, Employment, and the Consumer Price Index.  In a best case scenario, any forecasted reduction in emissions related to a carbon tax, is only as good as these various input forecasts. And remember, in these economic forecast models, errors in any of the inputs such as GDP become compounded.

Early last year, I reviewed three forecasts that had been made by the BC Provincial government in February for the provincial Budget for fiscal 2008/2009. Later I reviewed and compared the 2008 September First Quarterly Report forecasts for the same three variables, which were all done and completed well before the general financial collapse. 

What did I find? For the three basic forecasts of changes in GDP, CPI, and Employment over the seven-month period, the government had changed its forecasts substantially. Remember, these forecasts were made for basic economic parameters, and they were made in 2008 for the year of 2008, i.e. they were reviewed in the same year the forecast was made, before the year was even complete. One couldn’t select a much shorter review period. Yet the predictions varied enormously

GDP                        +29%
CPI                          -22%
Employment          -47%

I used to do this kind of analysis regularly for Union bargaining conferences and the actual results were always unpredictable. The variances shown above occurred over the very short period of seven months. Given the recent financial turbulence, who could derive a short-term or long-term forecast now?... With what expected accuracy? Roll the dice...

These are only three of many basic assumptions that go into economic modeling and forecasting. Modeling provides the economic context for tax rate setting, in this case carbon tax rates, and for expectations for emission reductions and carbon tax revenues.  Unfortunately, one reliable rule of forecasting is that forecasts become increasingly unreliable as one goes further into the future.

In the case of carbon taxes and targets for 2050, we are talking about forecasting 41 years ahead. Has there ever been an economic forecast that was even remotely successful over such a long period? 

Let’s try to imagine (seriously) an economic forecast that accurately predicted the impact today on government revenues, resource use, or on any variable for that matter, of a tax imposed in 1969!  Then let’s ask ourselves, why would an economic forecast made today provide any better reliability for outcomes in 2050?

And then let’s introduce the inevitable impacts of already unavoidable global warming—impacts of more severe and more frequent events even than the current BC Mountain Pine Beetle epidemic, the deadly European heat wave in the Summer of 2003, the devastation of New Orleans by hurricane Katrina, and the hundreds of thousands of deaths in Darfur due to drought and starvation induced by global warming. These three latter events were all attributed to global warming by Sir David King, who was until recently the chief Science Advisor to the UK government.

Add to all this the prospect that any errors in the impacts of carbon emission reductions will be compounding, year after year.

If we’re not yet convinced about the non-efficacy of a carbon tax in curtailing global warming in the face of all this mounting uncertainty, consider that all of the unknowns reviewed so far seem almost trivial against the fact there is no established reliable relationship between a carbon tax and a change in carbon emissions, beyond the experienced generality that as prices rise emissions can be expected to fall.

There are some estimates of a correlation, but they are varied and speculative, even in the short term. They have no track record, even in more-or-less normal circumstances. In fact, a mere 10 years ago, many, if not most, economists were saying that gasoline demand was largely insensitive, or inelastic, to price (a notable exception in the early 1990s being Ernst von Weizsacker of the Wuppertal Institute for Climate, Environment, and Energy). And the sensitivity of demand to price is completely speculative in the context of long-term, continuous requirements for dramatically reduced emissions.

To review, the uncertainties related to establishing appropriate values for carbon taxes are immense. They include:

• Uncertain forecasts of basic economic parameters, which are used as inputs to models upon which tax rate setting are based;
• The highly uncertain relationship between carbon prices and carbon emissions, which also changes as emissions decrease, reflecting increasing public resistance to reducing emissions as further reductions begin to affect not just casual, luxury or ‘easy-to-forego’ expenditures, but basic lifestyles;
• Heightened fluctuations in the already uncertain overall economic context, due to already entrained global warming impacts—more pine beetle kills and damage from other species, other diseases and pests and interactions with other parts of the world; more Katrinas; more storms and droughts and floods; more heat waves on the 2003 European scale; increases in food prices, less certain supplies of food with more food riots and more starvation; increasing energy costs with impacts on the reliability of supply due to peak oil and higher costs of trade due to increasing costs of fossil fuels.

The survival of civilization is now dependent upon definite, dramatic reductions in greenhouse gases, not just for the next 41 years, to 2050, but well beyond, out 91 years to 2100.  Certainty, one of my five criteria, is a central, indispensable feature of any survival plan. Can we imagine the chaos if we subject that survival to all of the uncertainties associated with reliance on taxing carbon?

Unfortunately in this regard, many in the environmental community have seemingly become infatuated by such economic mechanisms—which have no touchstone with biophysical reality—and by the associated but categorically compromised appeal toward ‘getting the price of carbon right’.  Wrong!!  We must get the emissions of carbon right! And that is a task that economics is simply not equipped for; for the reasons cited here, carbon taxes in particular cannot provide anything remotely like the certainty we need to quickly implement dramatic quantifiable reductions in carbon emissions.

To add to the difficulty, a recent paper by scientists from the Tyndall Centre for Climate Change and published by the Royal Society, “Reframing  the climate change challenge in light of post-2000 emission trends”, by Kevin Anderson and Alice Bows,  reports that reduced carbon emissions from agriculture will be extremely difficult to achieve; they could see only a 50% reduction in that sector which, given the unrelenting requirement for the elimination of emissions, is very ominous. If this is the case, the rate of reductions elsewhere must be greater to compensate. In effect, this requires negative emissions, which are suggested also by Dr. Weaver’s work at UVic.

The Anderson and Bows paper stated: ”Even atmospheric stabilization at 650 ppmv CO2e demands the majority of OECD nations begin to make draconian emission reductions within a decade. Such a situation is unprecedented for economically prosperous nations. Unless economic growth can be reconciled with unprecedented rates of decarbonization (in excess of 6% per year), it is difficult to envisage anything other than a planned economic recession being compatible with stabilization at or below 650 ppmv CO2e.“

The Anderson-Bows paper was published last fall. In some of the literature,  including that paper, 650 ppm equates to about 4 degrees C of warming. There is no science to explain how, with that much warming, we could avoid the positive feedback loops that lead to even higher temperatures.  To state the obvious, 650ppm is almost double the 350ppm considered to be the maximum safe level.

Our plans to reduce emissions must endure reliably through the already inevitable impacts of now unavoidable global warming, which are going to severely stress our ability to govern ourselves. Growing political and social tensions will make it difficult for  a carbon tax regime to succeed.  Imagine for a moment pine beetle kills exacerbated three or four times, the consequent reductions in employment and government revenues, and the accompanying demand for increased government expenditures. How would we fund our health care, education and the legal system? Moreover, our infrastructure, which is already falling apart, needs to be rebuilt to be carbon emission free—this applies to urban centres, transportation and food systems.

Fitting any workable carbon tax into all this (never mind ‘simple and transparent’) presents an insoluble problem when we already know how difficult it is to forecast the impacts of taxes in recessions vs. boom times, let alone when the world begins to realize (if it ever does) that economic growth itself is a big part of this problem. 

In summary, carbon taxes can’t work, they don’t meet the crieria. The related uncertainties I have identified are insurmountable. Economic modeling and forecasting are already rife with inherent uncertainties that will compounded by the unprecedented unknowns of unavoidable global warming both in the biophysical world and the world of politics and international governance. Any attempt to solve any of these problems means making the tax regime more complicated, and so less transparent.  Worse, any government’s credibility will become progressively stretched every time it has to change a tax structure that hasn’t achieved the necessary emission reductions—probably every year or so for the next 40 years.

By comparison, the ‘Cap and Dividend’ system does a far better job of meeting the five criteria, readily lending itself to effective implementation of a scheduled series of known specific emission reductions over time, and with a much greater likelihood of  maintaining transparency, fairness and social acceptance throughout. Now is the critical time to choose a path that works over one that demonstrably will not.

Cliff Stainsby
Cobble Hill

Click here for insight into the politics of carbon policies