Legitimation is at once the most difficult and the simplest component of the policymaking process. It generally involves the least complex and technical forms of policy analysis, and the number of actors is relatively limited, except in initiatives and referendums. On the other hand, the actors involved are relatively powerful and have well-defined agendas of their own. Consequently, the task of the policy analyst seeking to alter perceptions and create converts to new policies at the legitimation stage is difficult. The type of formal evidence used at other stages of the process may not carry much weight at this stage, and political factors become paramount.

        The barriers that the policy analyst faces in attempting to push through his or her ideas are sometimes individual and political, as when members of Congress must be convinced through partisan analysis or vote trading to accept the analyst’s concept of the desirable policy alternative. Conversely, the task may be one of altering substantial organizational constraints or mediating turf wars on a decision that would facilitate the appropriate policy response to a problem. The problem may also be a legal one, of persuading the courts to respond in the desired fashion to a set of facts and to develop the desired remedy for the perceived problem. Or, finally, the problem may be a political one in the broadest sense--that is, to persuade the voters (through the political mastery of the analyst) to accept or reject a particular definition of an issue and its solution. This is a great range of problems for the analyst, and it demands an equally great range of skills.

        No individual is likely to have all these skills, but someone must make strategic choices as to which skills are the most appropriate for a particular problem. If the problem is to get a dam built, then Congress is clearly the most appropriate arena. If the problem is a civil rights violation, the best place to begin is probably the court system. If the problem is a specialized environmental issue, then the regulatory process is the appropriate locus for intervention. Policies do not simply happen; they must be made to happen. This is especially true given the degree of inertia in American government and the number of points at which action can be blocked. It frequently occurs that the major task of the policy analyst is to define clearly the problem that must be solved. Once that is done, the solution may not be simple, but it is at least potentially analyzable, and a feasible course of action may become more apparent.