The object of this paper is to illustrate how research can assist in the field of public policy. The problems inherent in measuring open-ended situations where demand can be infinite are discussed. The need to use a variety of secondary measures to assess the results of policy where direct primary measures are not possible is pointed out. To set the scene for two case histories the role of the Central Office of Information in London is described. Illustration of the use of indirect or secondary measures in assisting in the implementation and assessment of public policy is made by reference to the Housing Improvement Grants Scheme which began in 1969 and to the Legal Aid and Assistance Campaign which began in 1973. There is a brief conclusion pointing out the risks which can be run if secondary measures are wholly endogenous.