Is technology still a major driver of health expenditure in the United States? Evidence from cointegration analysis with multiple structural breaks

Abstract

Using a longer span of available time series data and employing powerful unit root and cointegration tests that allow for multiple structural breaks, developed recently by Carrion-i-Silvestre et al. (Econ Theory 25:1754–1792, 2009), Perron and Yabu (J Bus Econ Stat 27:369–396, 2009), Kejriwal and Perron (J Econ 146(1):59–73, 2008; J Bus Econ Stat 28(4):503–522, 2010a; J Time Ser Anal 31:305–328, 2010b) and Maki (Econ Model 29:2011–2015, 2012), this paper empirically investigates, whether technology continues to be a major driver of real per capita health expenditure, along with some control variables such as per capita income and life expectancy, in the United States, during the period 1960–2012. Specifically, the paper applies the most recent cointegration tests under multiple structural breaks and extends the work of Okunade (J Health Econ 21(1):147–159, 2002) with the possibility whether a linear cointegration model with multiple structural breaks would provide a better economic model to quantify the impact of some major determinants of US real per capita health expenditure. This paper presents evidence to show that per capita real income, technology as indicated by four proxy measures and life expectancy at birth are some major drivers of real per capita health expenditure in the United States. Contrary to the available evidence in the literature, the finding of this paper is that the point aggregate income elasticity of health expenditure estimate is less than one, indicating that health care has evolved to become a necessity in the United States. Policy implications of the empirical findings are discussed in the paper.