Abstract Industrialized countries may reduce their costs of meeting carbon constraints if they penalize fuels not only on the basis of their carbon intensity but also on the basis of their import–export status. Simulations of these policies show that participating industrialized countries can reduce their costs and hence increase their willingness to participate. However, they will impose higher costs on the world, because the most carbon-intensive fuels will not be taxed most heavily. Such a bias creates a “how” inefficiency in addition to the “where” and “when” inefficiency created by current international agreements to control greenhouse gas emissions. Although countries have always had such incentives, these considerations must be more fully acknowledged in today's energy markets, after September 2001.