United Nations General Secretary Ban Ki Moon is convening a high level meeting on global warming at the U.N. headquarters on September 24. The idea is to jump-start the climate change negotiations for the 13th Conference of the Parties (COP-13) of the United Nations Framework Convention on Climate Change (UNFCCC). COP-13 is scheduled for December 3-14 in Bali, Indonesia.

President George W. Bush is also inviting representatives from the major industrial countries and large developing countries to come to Washington, DC to discuss climate change on September 27-28. The goal of both meetings is figure out what to do about reducing the emissions of greenhouse gases, especially carbon dioxide, after 2012 when the Kyoto Protocol expires. Under the Kyoto Protocol most industrialized nations—with exception of the United States and Australia—have agreed to cut their greenhouse gas emissions by 5 percent below what their 1990 levels.

What is the optimal climate change policy—the one that sets future emissions reductions to maximize the economic welfare of humans? Yale University economist William Nordhaus,perhaps the world's leading expert on the economics of climate change, has just released a new study, The Challenge of Global Warming: Economic Models and Environmental Policy,which estimates the costs of various proposed trajectories for limiting carbon dioxide over the next couple of centuries.

Nordhaus and his colleagues have developed a small but comprehensive model that combines interactions between the economy and climate called DICE-2007, short for Dynamic Integrated model of Climate and the Economy. Nordhaus first computes a baseline that assumes that humanity does essentially nothing to limit its output of carbon dioxide. By 2100 CO2 atmospheric concentrations would rise from the pre-industrial level 280 parts per million (ppm), to 380 ppm today, to 685 ppm in 2100. Global average temperature would rise by 2.4 degrees Celsius by 2100. In this baseline scenario, the DICE-2007 model estimates that the present value of climatic damages is $22.6 trillion. DICE-2007 includes damage to major sectors such as agriculture, sea-level rise, health, and non-market damages.

Nordhaus then uses his model to assess the ambitious CO2 reduction proposals made by British economist Nicholas Stern and former Vice President Al Gore. Nordhaus calculates that the Stern and Gore proposals for steep immediate emissions reductions produce very similar cost/benefit results. Nordhaus also evaluates explicit temperature and concentration goals, e.g., limiting average temperatures to 1.5 degrees Celsius above current levels or greenhouse gas concentrations to no more than 1.5-times pre-industrial CO2 atmospheric concentrations.

So what did Nordhaus find? First, the Stern proposal for rapid deep cuts in greenhouse gas emissions would reduce the future damage from global warming by $13 trillion, but at a cost of $27 trillion dollars. That's not a good deal. For an even worse deal, the DICE-2007 model estimates that the Gore proposal would reduce climate change damages by $12 trillion, but at a cost of nearly $34 trillion. As Nordhaus notes, both proposals imply carbon taxes rising to around $300 per ton carbon in the next two decades, and to the $600-$800 per ton range by 2050. A $700 carbon tax would increase the price of coal-fired electricity in the U.S. by about 150 percent, and would impose a tax bill of $1.2 trillion on the U.S. economy.

In addition, scenarios which attempt to keep the future average temperature increase below 1.5 degrees Celsius and concentrations below 1.5-times pre-industrial atmospheric concentrations are also not cost-effective. The DICE-2007 model calculates that both would cost more than $27 trillion in abatement costs and provide only about $13 trillion in reduced damages.

The optimal policy? Nordhaus reckons that the optimal policy would impose a carbon tax of $34 per metric ton carbon in 2010, with the tax increases gradually reaching $42 per ton in 2015, $90 per ton in 2050, and $207 per ton of carbon in 2100. A $20 per metric ton carbon tax will raise coal prices by $10 per ton, which is about a 40 percent increase over the current price of $25 per ton. A $10 per ton carbon tax translates into a 4 cent per gallon increase in gasoline. A $300 per ton carbon tax would raise gasoline prices by $1.20 per gallon.

Following this optimal trajectory would cost $2.2 trillion and reduce climate change damage by $5.2 trillion over the next century. "The net present-value global benefit of the optimal policy is $3.4 trillion relative to no controls," writes Nordhaus. "While this is a large number absolutely, it is a small fraction, about 0.17 percent, of the discounted value of total future income." Keep in mind that in this optimal scenario climate change damages would still accumulate to $17 trillion (lower than $22.6 trillion in the baseline case), but they are not abated because to do so would cost more than the benefits obtained.

A more optimistic scenario envisions the invention of a low cost zero-carbon technology. Such a technology would have a net value of around $17 trillion in present value. As Nordhaus notes, "The net benefits of zero-carbon substitutes are so high as to warrant very intensive research." Setting a price on carbon through a rising tax will encourage the development of such technologies. Another good way to hurry the process along would be to offer a substantial prize to the inventor of a cheap low carbon energy technology, e.g., perhaps a better battery, or paint-on solar cells.

Nordhaus cogently argues that neither doing nothing nor trying to halt global warming immediately are sensible policy targets. Nordhaus's study is certainly not the final word on climate change policy, but it would be a excellent starting point for climate change negotiators when they gather in New York, Washington and Bali this fall.

Ronald Bailey is Reason's science correspondent. His book Liberation Biology: The Scientific and Moral Case for the Biotech Revolution is now available from Prometheus Books.