Standard measures of inter-generational social mobility distort actual patterns of mobility and health behaviour: evidence for a better methodology

Abstract

Objective: Researchers commonly ignore initial social level (e.g. low, middle, and high) when classifying inter-generational social mobility as upward, stable, or downward. This article examines the consequences of this practice in the study of social mobility's relationship to tobacco and alcohol use. Design: Mobility was tracked using Norway's 1970 (mother's education) and 1998 (offspring's education) national education registry's, linked to a 1998 national survey providing data on offspring's tobacco and alcohol use. Setting: Community. Participants: One thousand and forty-five randomly selected Norwegians aged 24–31 years, and their mothers. Main outcome measures: Tobacco and alcohol use. Results: Within upward-stable- and downward-mobile groups, sub-groups of offspring defined by the education level of their mothers had significantly different levels of tobacco and alcohol use. Stable social mobility at the intermediary level was associated with alcohol consumption of less than one to three times per month, and stable social mobility at the high level was associated with alcohol intake of two to three times a week. Upward mobility from low to high and middle to high levels was associated with consumption levels once a week, but upward mobility from low to middle level was associated with a consumption level of less than once a month. The associations between social mobility and tobacco use were similarly complex. Conclusions: A literature characterised by mixed findings suffers from classification imprecision, following from the erroneous assumption that all forms of upward, stable, and downward social mobility are equivalent with respect to health behaviour. If future research is in accord with the findings of this study, then aggregation of diverse social mobility patterns into a few gross categories is not commendable.

Keywords:

• socio-economic
• health behaviour
• health inequalities

Acknowledgement

The Norwegian Research Council funded this research through a grant to MBM. Both authors are independent of the funding source.

Notes

Note

1. The estimate of association for the category HL was unreliable due to too small numbers in the cell (n = 1).