Insulation from Observability and Types of Social Conformity

between size and administration that would otherwise occur. In particular, some special bureaucratic constraint may operate within government hospitals which inhibits the free growth of the administrative component. It is also important to recognize that these propositions, even if confirmed empirically, are not necessarily sufficient to explain all structural and temporal variations in the relative size of the administrative component. Specifically, the inclusion of a proposition concerning the routinization of roles would undoubtedly improve the general explanation of variations in the relative size of the administrative component. Routinization is not discussed in this paper because the apparent contradiction between the two sets of findings could be resolved without taking it into account. On the other hand, these propositions appear to explain adequately the apparent overall increase in the relative importance of administrative activities within organizations during the past several decades (as evidenced by the rapid rise in the per cent employed in clerical occupations, for instance). It is often suggested that this increase in administration is a function of the sheer growth in organizational size. If our propositions are correct, however, the explanatory variable is organizational complexity rather than organizational size. Clearly, more systematic studies are required before any such conclusions can be considered to be substantially confirmed.