Prof. Martin Feldstein favors a carbon tax to reduce global emissions of carbon dioxide, an important greenhouse gas but by no means the only one ("The Case for a World Carbon Tax," editorial page, June 4). He does not discuss whether such a reduction is urgent and necessary; scientific views are divided on this issue. But economists generally agree that CO₂ control will be very costly; they favor a carbon tax only because it allows for an economically efficient solution that lowers this cost somewhat.

Some further observations are appropriate:

Stabilizing emissions at the 1990 level, the general goal announced by European nations - will not stabilize atmospheric concentration; it merely reduces its rate of growth slightly. According to the United Nations scientific reports on climate change, stabilization requires that CO₂ emissions be reduced world-wide by 60% to 80%. No one has yet dared to estimate the economic cost of such drastic measures - never mind the political cost. In 1990 the Congressional Budget Office reported costs to the U.S. economy of about $150 billion a year for only a 20% emissions reduction. Additionally, consumers would face doubled electric bills and a doubling of gasoline prices.

Contrast these numbers with the European Community proposal for "a tax on all carbon fuels that would start at the equivalent of $3 a barrel and rise to $10 a barrel by decade's end." That's a tax of only eight cents to 25 cents a gallon, compared with the current European pump price of about $3 - and hardly likely to be effective in enforcing energy conservation. The EC Commission may simply be exploiting the greenhouse scare to raise tax revenues.

But the greatest problem for any carbon tax proposal is how to divide up an agreed-to global emission level into national quotas. Relating them to 1990 emissions (which are roughly equivalent to national gross domestic products) is clearly unfair to developing countries and won't fly. It would stunt their development and keep them forever poor. Third World nations would rightly consider any quota that's locked into 1990 emissions as a form of "eco-imperialism."

On the other hand, applying emission limits to only the major industrial countries - comprising less than 20% of the current world population - means that their emissions would soon become insignificant compared with those of the rapidly growing developing nations.

An often-mentioned compromise - a uniform per-capita quota - would be unworkable for the same reason. In addition, it would provide the wrong incentives. If carbon tax revenues from rich countries are used to buy emission permits from poor countries, what's to stop Third World kleptocrats from keeping their subjects poor and encouraging rapid population growth to maximize this income transfer?