Sometime late in the next century, our heirs will discover whether global warming produces dramatic changes such as the disappearance of coastlines and an upsurge in tropical diseases. The task facing us now is how best to address the possibility of human-induced climate damage on a global scale.

The real issue is what steps countries should take to control greenhouse gas emissions that contribute to global warming. At the negotiations in Kyoto, Japan, more than 150 countries have been focusing on "targets and timetables." While such discussions can provide good political theater and photo ops, they miss a fundamental point: Carrying out significant reductions in greenhouse gases will require an institutional capability that simply does not exist now.

Consider these three political facts:

1. A winning short-term strategy for politicians in developed countries is to talk tough on climate change but do relatively little. The appearance of doing something can go a long way toward assuaging the concerns of most citizens while helping to neutralize pressures from environmental groups.

2. If anything is done in the short term, developed countries are likely to bear virtually all the costs. Climate change is generally not a high priority for developing countries. Putting food on the table is--not to mention the annual toll of 2 million children who die of diarrheal diseases from poor sanitation.

3. It is unlikely that anything significant will be done to reduce greenhouse gases for at least the next decade and possibly longer. Developing countries will not do much unless the developed countries show them the money. Developed countries will continue to blow hot air while recognizing the impossibility of achieving much without the cooperation of the developing countries.

The stated U.S. policy of reducing greenhouse gas emissions to 1990 levels by the year 2000 illustrates the political problem. President Bush agreed to this policy at Rio in 1992, but it is now clear that the United States and most other developed countries will fail to meet it. The Clinton-Gore proposal extends this deadline to at least 2008--conveniently not on their political watch. Why?

The costs of stabilizing emissions of carbon dioxide--a major greenhouse gas--by 2010 are likely to be substantial. Estimates center around $80 per ton of carbon reduced, which translates to a 20-cents-per-gallon gasoline tax. Not long ago Congress balked at a much smaller BTU tax. Few legislatures elsewhere in the world are likely to be any bolder. So the most likely outcome of international negotiations over the next decade is that a select group of developed countries will agree to do something that has a relatively low cost.

What is to be done?

There may be economic justification for doing something now. While estimates of the costs and benefits of climate change are highly uncertain and changing all the time, economic models suggest a modest carbon tax--on the order of 1 to 10 cents per gallon of gasoline--might make sense, assuming most countries cooperate in an agreement.

But there's the rub. To get results will require widespread cooperation among countries. Even if developed countries froze their emissions, worldwide emissions would still grow around 150 percent by 2050 because of the increase from developing countries. Moreover, if China or India does not eventually agree to binding targets, some greenhouse-gas producers will simply set up shop there because it is cheaper.

How to get most countries to participate is the $64,000 question that no one can really answer. It may simply require waiting for some of them to get rich enough to be willing to spend their own money on this problem.

But even rich nations will be more serious about getting the job done if they can do it cheaply. Virtually every economic study suggests that there could be substantial savings to designing a market-based program for greenhouse gases, similar to the U.S. allowance-trading to reduce sulfur dioxide emissions, which gives business flexibility in meeting prescribed targets.

Unfortunately, we do not yet have the institutional know-how and political will to implement such a cost-effective approach for reducing greenhouse gases on a global scale. Instead of rushing to go global, we should be looking to set up modest multi-country tests--call them "case studies"--that might be applied globally in the decades ahead.

For example, the Scandinavian countries, which have already implemented carbon taxes, could continue on that path. The United States and other countries interested in tradable permits, or a system combining permits with a cost cap, could use that approach. Other European countries might want to try a combination of command-and-control regulations and market-based approaches.

The case studies should focus on developing national institutions, such as those needed for monitoring emissions, enforcing nations' obligations, facilitating international trading of permits and resolving disputes. Such national institutions will be crucial to carrying out novel, low-cost approaches.

The case studies would set the stage for achieving necessary reductions later on, when developing countries will be ready to join an agreement without receiving significant side payments from the developed world.

The last thing we should be doing now, in our state of ignorance about the warming problem and institutional responses, is to rush into an unworkable worldwide policy. That is why we should focus on the quiet work of learning how to build institutions that can effectively address the problem, and not on public grandstanding.