Constraints, satisfaction and residential mobility: Speare's model reconsidered.

Our study has shown that the stress threshold model, as formulated by Speare, only works partially. Consistent with Speare's model, we have found that subjective satisfaction is a strong predictor of thoughts about moving. Thoughts about moving is a good predictor of actual mobility. There are, however, three major problems with the model: stress as measured by satisfaction is not a particularly good predictor of actual mobility, although it does have some indirect influence through thoughts about moving; the "structural" variables have a strong independent impact on the mobility process beyond satisfaction; the satisfaction variables have little influence in mediating the effects of structural variables on mobility thoughts and behavior. The question of why our results differ from Speare's cannot be definitively answered here. We believe that our research has certain virtues in its direct measurement of satisfaction with home and community and also its much larger sample of movers. Speare's sample may have the virtue of being more representative of a large urban population since it was drawn from all segments of Rhode Island. Yet, while our sample is selective of areas within Seattle, we believe it provides a good representation of a wide variety of residential environments. Overall, our results are more consistent with the other studies which have addressed these issues (Bach and Smith, 1977; Lee, 1978; Michelson, 1977; Newman and Duncan, 1979), although the methods and approaches are not identical. Some of these studies were reviewed in the first part of the paper. We believe that understanding of the attitudinal predictors of changing residence is roughly at the same stage as research in the early post-World War II period on the attitudinal correlates of fertility behavior among American women. Research such as the Princeton study (Westoff et al., 1961; 1963) demonstrated that a variety of social attitudes about home, family, work and childbearing correlated poorly with levels of fertility. In contrast, such variables as objective religious affiliation, educational attainment, and race were clear correlates of fertility behavior. We still do not fully understand why these structural variables are important but we know that they are key predictors of behavior.(ABSTRACT TRUNCATED AT 400 WORDS)