Working Paper Series 2305 - Estimating effects of individual-level workplace mental

These points of contention form the beginnings of an empirical critique of individual- level mental wellbeing interventions, suggesting that there is insufficient evidence in support of the wide-spread adoption and promotion of such practices. [...] When analysing the dimensions of job quality, I recoded a series of variables in the BHW to resemble binary treatments for comparing with the wellbeing interventions. [...] For volunteering, the interaction term for work stress does not model the selection bias because the bias for the positive effect would theoretically be in the opposite direction, with happier workers more likely to participate. [...] For the negative estimates for resilience and stress management training, there are three possible explanations: (1) there is a negative effect from participating; (2) there is no effect and the negative coefficient reveals selection bias; or (3) a positive effect, but that it is small and not ameliorative enough to improve wellbeing up to the level of the control group. [...] Although most weaknesses of this analysis arise from the data used, the scale and cover- age of the same data also offers the main strength of this article.