The Health Transition in Hungary after the Regime Change, Convergence of Sub-Regions – an Eigenvector Spatial Filtering Approach

Lajos Bálint , Hungarian Central Statistical Office (HCSO)

Introduction: To understand well-being, the allocation of health care resources and the dynamics of health inequalities, it is important to know whether regional mortality inequalities have decreased or not. The convergence of health status has been rarely investigated in a former communist countries. Method: Convergence was tested using Sigma and Beta methods. Sigma convergence characterises the evolution of dispersion over time. Different dispersion indices and their population-weighted versions were applied. Beta-convergence examines whether regions with lower life expectancy have experienced faster growth than regions where the initial level was inherently higher. Different OLS models have been applied to deal with the of different population size (weighted OLS) and strong autocorrelation of the error terms (Eigenvector Spatial Filtering). The convergence tests were carried out for men and women separately. Results: Life expectancy has improved significantly over the 25 years in Hungary. The improvement affected all regions, but not to the same extent. Beta convergence models support the catching-up of regions with poor initial conditions for both sexes. Beta convergence might be caused by diminishing returns of the inputs factors in the health production function, which might lead to convergence in general conditions (e.g education and income). Changes in lifestyle, in particular a reduction in alcohol consumption, healthier diets, better health care could also contribute to catching up in more disadvantaged regions. As regards sigma convergence, the results are mixed. I found no evidence of convergence of regions for men, while for women there is strong evidence of divergence.

See extended abstract

 Presented in Session P1. Postercafe