Long-Term Growth Modeling of Healthcare Spending in OECD Countries, 1970-2004

Rationale: This paper develops and compares long-term healthcare expenditure growth models for the US and other industrialized countries. In the US, the Centers of Medicaid and Medicare Services (CMS) estimates long-term healthcare expenditure growth with the rule GDP+1, which is the annual growth rate of gross domestic product (GDP) per capita plus one percent. Developed by a team of leading health economists, it is used to make national policy decisions. Other industrialized countries and theoretical economists have developed models also. The various models have not been compared. Objective: To understand, model and predict the long-term growth in healthcare expenditures for the US and other selected countries in the Organization for Economic Co-operation and Development (OECD). Study Population and Data: Data from 1970 to 2004 for six OECD countries are modeled to gain insight into the variations of aggregate and disaggregate healthcare expenditure growth across countries. Theoretical Framework: The Solow long-term economic growth model is adapted to estimate healthcare expenditure growth. We separate capital into physical and human, and use the Mincer wage model to estimate human capital. Parameters of the resulting model provides information about the contribution levels of capital, labor and technology, income demand elasticity, and the long-term asymptotic share of healthcare expenditures to GDP. Validation: The resulting model explains the variation in growth paths of healthcare expenditures within and across OECD countries. The remaining residuals are fitted by splines to identify healthcare policy changes. The Chow test is used to determine significant differences in coefficients across countries and differences across disaggregate categories of expenditures. Results: We have found that the aggregate healthcare expenditures are well-modeled as a linear function of potential GDP, or full-employment GDP, per capita. The relationship suggests that the drivers of healthcare expenditures are related to the drivers of potential GDP (i.e., capital, labor, and technology). The dominant explanatory factors for the growth path of healthcare expenditures are the tangent hyperbolic growth of healthcare labor and the different returns to education. Additional variation in disaggregate healthcare expenditures can be explained by healthcare policies modeled with splines