| Abstract: | A review of previous attempts to test hypotheses derived from the functional theory of stratification suggests that to test the theory as formulated by Davis and Moore, the unit of analysis must be the position rather than the individual: indices of the theoretical concepts must be measures of the stratification structure itself rather than perceptions of that structure: and a universe or representative sample of positions must be used to determine if the relationships specified by the theory are applicable throughout a society's labor force. This paper reports a test of the functional theory that attempts to meet these specifications. Three related models are deduced from the theory and tested: (a) a model in which functional importance and change in skill are presumed to antecede change in reward and change in prestige; (b) a second model in which the dependent variables are additionally influenced by the interaction of functional importance and change in skill; and (c) a third model in which possible reciprocal effects of functional importance and skill are estimated using longitudinal data. The predictions of the functional theory concerning the effects of functional importance are not supported by the data. |
