MULTIPLE-OBJECTIVE METHODS IN REGULATORY DECISION-MAKING - A CONCEPTUAL-FRAMEWORK APPLIED TO US AUTOMOBILE SAFETY STANDARDS

Public policy making is a complex matter. Policy makers are charged wi th balancing a broad spectrum of competing objectives, reflecting in v arying degrees the interests and aspirations of a diverse range of con stituencies and stakeholders. Policy decisions have differential impac ts on differing constituencies and the contributions of these impacts to objectives are frequently uncertain, difficult to quantify, and hot ly disputed. Formal methods of decision analysis have been advanced as aids for coping with complexity and have been applied to some public policy issues, most notably the management of water resources. While f ormal methods have the apparent benefit of rationalizing the policy pr ocess and improving the efficacy and equity of policy decisions, serio us questions remain concerning the acceptability and ultimate usefulne ss of formal analyses in the public arena. In this article we examine these questions in the context of policy making relating to government regulation of automobile designs for safety. We consider what would b e involved in attempting to use a relatively modern, multiple-objectiv e approach in this context. The key questions are how and, more import antly, why multiple-objective methods might be used. The article begin s with an illustrative case study, describes the current policy-making process, identifies the parties involved in and affected by this proc ess, elicits key objectives, looks briefly at some scalar-objective ap proaches, and then outlines a multiple-objective approach. The framewo rk developed arguably is useful in assisting policy making, at least a t a qualitative level. Issues which impede a more quantitative resolut ion of this framework are discussed. This article is intended as a pil ot study which may promote research toward the achievement of a useabl e multiple-objective procedure applicable in the public domain. (C) 19 97 Elsevier Science Inc.