egy. Nonetheless, the committee believes that the proposed strategy, if adequately implemented, could reasonably be expected to achieve a significant reduction in underage drinking and the associated social costs. The exact decrease that could be expected is speculative. Available information on the social costs of underage drinking is a starting point for gauging the potential cost-effectiveness of the recommended strategy. If annual social costs attributable to underage drinking (conservatively estimated to be $53 billion per year in 1996; most likely higher now due to inflation) were reduced by only 2 percent after 10 years, or if a 1 percent reduction were sustained for 2 years, an expenditure of approximately $1 billion over that period would be economically justified. If social costs were reduced by 5 or 10 percent after 10 years, the economically justifiable cost would be significantly higher. While the committee believes that the enormous social costs of underage drinking warrant an investment in the proposed strategy, specific efforts to collect cost data and to quantify the proposed outcome measures should be built into strategy implementation in order to obtain more precise measures of cost-effectiveness.