Multidimensional health modeling: Association between socioeconomic and psychosocial factors and health in Latvia

This research aims at estimating association between socioeconomic and psychosocial factors on the one hand and health in Latvia on the other hand. While information on association between socioeconomic determinants of population health in Latvia is scarce, effect of psychosocial resources on individual health in this country hasn’t been estimated before. We find empirical support for the association between different psychosocial factors and physical health in Latvia. This paper proposes new approach for modelling self-assessed health. We find that the concept of health is too complicated to measure effects of health determinants using a one-dimensional econometric model. We apply two-dimensional stereotype logistic model that allows capturing nonmonotonicity in effects of factors and revealing significant effects that remain unseen if single dimension models, such as ordered logit or ordered probit, are used. Modelling self-assessed health using multi-dimensional stereotype logit provides higher model goodness of fit and quality measures in comparison to ordered probit model.