Comment: On “Long-Run Growth of Nondefense Government Expenditures in the United States”

Abstract

The paper by Tussing and Henning (1974), which recently appeared in this journal, falls into two parts. One reviews the “determinants” literature: the other presents the authors' own empirical results. The first part properly emphasises the need for a comprehensive approach to the determinants of public expenditure, although some causal influences (e.g., political considerations) appear to be neglected. This emphasis is a welcome recognition that in this field of research we are still concerned essentially with hypothesis formulation: attention needs to be directed primarily to the identification of the important explanatory variables. Unfortunately, despite the breadth of their perspective, Tussing and Henning appear overly ambitious in interpreting their own results. Causal interpretation requires a preconceived idea of what the significant determining variables are, which is difficult in the absence of an operational theory of the fiscal decisionmaking process. In place of this they have adopted a market analogue, which leads to a demand interpretation of expenditure growth. The growth in public spending is thus a function of income, of relative prices (“Baumol's disease”), and of numerous other specific demands (e.g., urbanization, the collectivization of social security, transportation technology, the cold war).