Impacts of the British National Health Service: A Quasi‐experimental Study*

Abstract

This paper tests for possible effects of the implementation of the British National Health Service (NHS) upon the health of the British population. Although analysis does reveal statistically significant improvements in Great Britain's infant mortaUty rate concurrent with the implementation of the NHS, a comparison with other western European nations reveals similar improvements simultaneously with the British poUcy intervention. Thus, we cannot conclude that the NHS had a significant effect upon British health. A more important result, however, is the demonstration of the utiUty of both the Box-Jenkins techniques and, in conjunction, the nonequivalent control group quasi-experimental design as important tools for performing poUcy evaluations. *I would Uke to thank Robert Albritton and John Scholz for their comments and suggestions. A special word of thanks goes to Helmut Norpoth for his assistance with the statistical analyses performed in this paper. I, of course, retain fuU responsibiUty for any errors of judgment or interpretation. 706Health care programs are among the most frequently evaluated of all groupings of pubiic policies. Whether viewed as a consumer good (Paglin, 1974; Kiarman, 1965), as a public investment related to subsequent economic gain (Denison, 1962;Fuchs, 1973;Barlow, 1967;Mushkin, 1962), or as an end in itself -important if for no other reason than as a final product for human welfare (Mushkin and Collings, 1959) -public health has traditionally been regarded as one of the substantively most important areas of policy evaluation research.Frequently, however, evaluation research in the field of public health falls short of its nominal goal -evaluating health. This may happen for any one of several reasons. At a level far removed from the policy evaluator, the problems of public h^lth may prove so intractable as to necessitate a reconsideration of purpose by the agency charged with the task of policy implementation. Wildavsky (1979, Ch. 2) speaks of an agency's frequent need to make a "strategic retreat" from unattainable objectives. The most desirable goals sought by an agency entrusted with policy implementation often prove unrealizable because they require drastic changes in human behavior. Attainable objectives, on the other hand, tend to be those that can be met simply by a manipulation of government resources. An example of a strategic retreat from objectives in the health area is the collection of hea Ith-related policies of Lyndon Johnson's Great Society. When the goal of producing an improvement in the general level of health appeared to be unrealizable, it was decided that the program's emphasis wouid instead be shifted to the provision of equal access to medical care for all persons. Thus in this instance the object of study for the health policy evaiuator is no longer health per se, but rather the provision or distribution of health services (Beck, 1973).A second reason for the failure to evaluate the effectiveness of policies has to do with temporal constraints, it may take years...