Workplace determinants of social capital: cross-sectional and longitudinal evidence from a Finnish cohort study.

To examine which contextual features of the workplace are associated with social capital.This is a cohort study of 43,167 employees in 3090 Finnish public sector workplaces who responded to a survey of individual workplace social capital in 2000-02 (response rate 68%). We used ecometrics approach to estimate social capital of work units. Features of the workplace were work unit's demographic and employment patterns and size, obtained from employers' administrative records. We used multilevel-multinomial logistic regression models to examine cross-sectionally whether these features were associated with social capital between individuals and work units. Fixed effects models were used for longitudinal analyses in a subsample of 12,108 individuals to examine the effects of changes in workplace characteristics on changes in social capital between 2000 and 2004.After adjustment for individual characteristics, an increase in work unit size reduced the odds of high levels of individual workplace social capital (odds ratio 0.94, 95% confidence interval 0.91-0.98 per 30-person-year increase). A 20% increase in the proportion of manual and male employees reduced the odds of high levels of social capital by 8% and 23%, respectively. A 30% increase in temporary employees and a 20% increase in employee turnover were associated with 11% (95% confidence interval 1.04-1.17) and 24% (95% confidence interval 1.18-1.30) higher odds of having high levels of social capital respectively). Results from fixed effects models within individuals, adjusted for time-varying covariates, and from social capital of the work units yielded consistent results.These findings suggest that workplace social capital is contextually patterned. Workplace demographic and employment patterns as well as the size of the work unit are important in understanding variations in workplace social capital between individuals and workplaces.