Convergence in obesity and overweight rates across OECD countries: evidence from the stochastic and club convergence tests

Obesity has become a serious public health problem in the last two decades and affects not only developed countries but also the developing world. To date, despite all attempts to stop its increase, obesity remains a serious social problem, and obesity-related diseases have a significant cost on the health system. In this paper, we investigate the convergence of obesity rates, overweight rates, and body mass index (BMI) across 35 OECD countries over the period 1975-2006. The empirical findings of the paper are expected to have important policy implications to better inform policymakers. In addition to the traditional convergence tests, this paper uses a newly developed LM based panel unit root test with endogenously determined structural breaks to test for the stochastic convergence. Given the shortcomings of traditional and stochastic convergence tests in light of the possibility of multiple equilibria associated with groups of countries following different convergence paths, the club convergence algorithm is also employed. Traditional cross-sectional tests show that both beta- and sigma-convergence of the variables of interest exist across sampled countries. Moreover, the univariate LM unit root test results provide support for convergence in the relative BMIs, obesity rates, and overweight rates for a majority of the OECD countries. The club convergence test results, however, suggest the rejection of full panel club convergence only in BMI variables and overweight rates for females and the presence of a certain number of clubs for these variables