I’ve changed my mind. After a lot of thinking, I’ve come to believe that neither motor fuel taxes nor carbon taxes are viable.

On the surface, higher fuel taxes are attractive. In theory they should lead buyers to choose fuel-efficient vehicles. And over the past couple of decades, it’s become obvious that federal mandates on fuel efficiency in the automotive fleet haven’t been effective. Higher taxes would signal to motorists that their fuel costs are unlikely to go down. That type of message is preferable to command-and-control rules like the Corporate Average Fuel Economy standard.

But fuel taxes are regressive. America’s poor would get hit harder by new energy taxes than the rich, and they would pay far more taxes – as a percentage of their income – than the wealthy.

Another problem: demand in the U.S. is fairly inelastic. America’s sprawling cities, combined with the lack of good public transportation, leave many Americans no choice but to drive. Thus, higher fuel taxes would merely be a punitive levy that would do little to change behaviors or cut consumption. Further, fuel taxes would have to be inordinately high, likely $2 or even $3 per gallon, to make a significant dent in demand. This can be seen by looking at historical data. In 1998, regular gasoline sold for about $1.03 per gallon, and Americans consumed an average of 8.4 million barrels per day. By May 2007, gasoline prices were consistently above $3 per gallon – triple the 1998 price – and yet consumption that month was nearly 9.2 mmbd, an increase of almost 9.3 percent over 1998 levels.

There are similar problems with a carbon tax. It would also hit the poor much harder than the wealthy. And although many economists believe that a carbon tax is the most logical way to address the issue of rising carbon dioxide emissions, the key problem is that for carbon taxes to work, they must be imposed globally so that individual countries are not disadvantaged. That point was made clear in a July 2007 report, “The Challenge of Global Warming: Economic Models and Environmental Policy,” by William Nordhaus, an economics professor at Yale University. Nordhaus advocates a carbon tax that would start at about $30 per ton (about $0.30 per gallon of gasoline) which would then be raised by 2 or 3 percent per year.

Nordhaus says his computer models “point to the importance of near-universal participation in programs to reduce greenhouse gases.” If only half of the world’s countries were to participate in the carbon tax program, Nordhaus estimates that would result in an “abatement cost penalty of 250 percent.” In other words, unless every country joins the carbon tax program, the countries that opt in will need to more than double their carbon tax rates to compensate for the countries that opt out.

But it’s unrealistic to expect consensus on carbon taxes. There are still 38 countries that have yet to sign the international treaty to ban land mines. (The non-signers include the U.S., Russia, China, and Saudi Arabia.) If getting international consensus on a non-economic issue like land mines is difficult, imagine how hard it will be to forge an agreement to tax energy.

So what should be done? Three things: embrace natural gas, solar, and nuclear. If the countries of the world are going to agree that carbon dioxide is bad, then they should be using more natural gas, the cleanest of the fossil fuels. Burning gas produces about half as much carbon dioxide as coal and creates far fewer air pollutants. Plus, it’s a pretty good transportation fuel. Solar and nuclear are low- or no-carbon alternatives that can provide significant increments of new electric power.

But even if gas, solar, and nuclear use surges, oil will still be a dominant fuel in the 21st century – and that’s going to be true regardless of whatever tax scheme gets put into place.