Background
As middle-income countries become more affluent, economically sophisticated and productive, health expenditure patterns are likely to change. Other socio-demographic and political changes that accompany rapid economic growth are also likely to influence health spending and financial protection.
 
Methods
This study investigates the relationship between growth on per-capita healthcare expenditure and gross domestic product (GDP) in a group of 27 large middle-income economies and compares findings with those of 24 high-income economies from the Organization for Economic Cooperation and Development (OECD) group. This comparison uses national accounts data from 1995-2014. We hypothesize that the aggregated income elasticity of health expenditure in middle-income countries would be less than one (meaning healthcare is a normal good). An initial exploratory analysis tests between fixed-effects and random-effects model specifications. A fixed-effects model with time-fixed effects is implemented to assess the relationship between the two measures. Unit root, Hausman and serial correlation tests are conducted to determine model fit. Additional explanatory variables are introduced in different model specifications to test the robustness of our regression results. We include the out-of-pocket (OOP) share of health spending in each model to study the potential role of financial protection in our sample of high- and middle-income countries. The first-difference of study variables is implemented to address non-stationarity and cointegration properties.
 
Results
The elasticity of per-capita health expenditure and GDP growth is positive and statistically significant among sampled middle-income countries (51 per unit-growth in GDP) and high-income countries (50 per unit-growth in GDP). In contrast with previous research that has found that income elasticity of health spending in middle-income countries is larger than in high-income countries, our findings show that elasticity estimates can change if different criteria are used to assemble a more homogenous group of middle-income countries. Financial protection differences between middle- and high-income countries, however, are not associated with their respective income elasticity of health spending. `
 
Conclusion
The study findings show that in spite of the rapid economic growth experienced by the sampled middleincome countries, the aggregated income elasticity of health expenditure in them is less than one, and equals that of high-income countries.