This paper seeks to clarify the current debate surrounding individual accounts by exploring systematically the variety of structural and administrative arrangements that either have been proposed for individual account systems in this country or have been adopted in other parts of the world. Plans from Chile, Latin America, Australia, the United Kingdom, Sweden and the United States are examined, including the Roth, personal security accounts, Ball, Porter, Moynihan and Feldstein/Samwick proposals. The paper focuses on two aspects of arrangements: (1) the mechanisms employed under each plan to move the money from contributors to pension fund managers and back to retirees and (2) the variations among the plans in the financial management choices available to workers. The purpose is to analyze the logic underlying different structural arrangements, and to help quantify the linkage between structure and outcomes. In particular, available information can help to illustrate some of the potential relationships between the structure of an individual account plan, gross investment returns, administrative costs, and net returns. The goals of individual accounts systems include an improved overall economy, higher rates of return on Social Security contributions, greater individual choice and responsibility, and reduced government liabilities. No single plan is likely to be the best one to achieve all of these goals. Constructing an individual account plan requires making choices in which one objective must be sacrificed in order to pursue another. The administrative costs associated with the plans implemented in Latin America and the United Kingdom have caused some analysts to look for alternative structures that promise lower costs and (potentially) higher net returns to workers. The alternatives tend to suffer from one or more drawbacks of their own, however. Employers may have to shoulder a higher burden, worker choice may be constrained, or the government may have to be relied upon to play a major role in collecting contributions, maintaining account information, or even managing the accumulating funds. The trade-off between worker choice and administrative costs is also present in the design of withdrawal options, where choice is likely to increase costs as a result of both adverse selection and the marketing and promotional costs of private insurers. A further issue that divides plans concerns the treatment of gender differences in life expectancy under the different payment options. Before Americans can consider proposals for individual accounts they must first develop more of a consensus about the relative importance of the objectives being pursued and the social values being understood
