Economic development and health outcomes: Evidence from cardiovascular disease mortality in Europe

Highlights

• Inverted U-shaped relationship between cardiovascular mortality and GDP per capita.

• Results robust to inclusion of various control variables and confounders.

• Relationship consistent with increasing returns from healthcare spending.

Abstract

The impact of economic growth on health outcomes has generated various conflicting results in recent years, particularly with the emergence of cardiovascular disease as one of the most salient global health issues. This paper looks at the relationship between economic development and health outcomes. We first estimate an empirical model using data on annual cardiovascular disease mortality in a panel of 27 European countries, over the period 2003 to 2014, together with per capita GDP levels in each country. We also include a number of additional determinants of health outcomes as prescribed by the literature, including various socio-economic, lifestyle and contextual variables. The results show that there is a statistically-significant inverted U-shaped relationship, whereby as income increases mortality rates first rise, before declining at subsequent levels of development. This trajectory persists even after the inclusion of the other determinants of health, as well as other robustness checks. We then seek to explain the persistence of the inverted U-shaped income-mortality relationship using a microeconomic model of consumer choice, with our findings suggesting that this relationship is consistent with a scenario of increasing returns to scale from healthcare investment. The results emphasise the importance of investment in the latest healthcare technologies and continued improvements to public healthcare systems, particularly within developing countries, in conjunction with efforts to boost economic growth.

Introduction

Cardiovascular diseases have emerged as one of the key global public health issues of the last few decades. Recent estimates suggest that cardiovascular diseases are directly responsible for around 3.9 million deaths in Europe each year, accounting for 45% of all deaths within the region (Wilkins et al., 2017). Furthermore, the annual healthcare costs associated with such diseases are estimated at €111 billion in the European Union, with a further €54 billion in productivity losses.

The growing concern surrounding cardiovascular disease and wider health outcomes is occurring within a broader debate regarding the influence of economic development on society at large. This debate also extends to the relationship between growth and health outcomes, and in particular the prevalence of cardiovascular diseases. This is because although there is plenty of evidence that economic development can lead to improved health outcomes (Pritchett and Summers, 1996), cardiovascular disease has often been characterised as a ‘disease of affluence’ since many of its risk factors, like obesity and hypertension, are typically more prevalent in developed countries (McKeown, 1988), although this viewpoint has been severely challenged in recent years (e.g. Dahlöf, 2010).

The objective of this paper is to analyse the empirical relationship between economic development and health outcomes within the context of cardiovascular diseases. Specifically, we ask whether the income-health trajectory is non-linear in nature, and investigate the reasons behind the existence or otherwise of this relationship. We use data on annual cardiovascular disease mortality rates among a sample of 27 European countries over the period 2003 to 2014, together with data on annual GDP per capita. We also extend this baseline model to include a number of additional explanatory variables to explain variation in mortality rates, including health expenditure per capita, body-mass-index, education levels and smoking prevalence, in order to analyse potential channels through which the income-health relationship may operate. We then seek to delve deeper into this relationship by specifying a simple microeconomic model of health, whereby a representative agent must allocate resources across consumption and health expenditure, and where health outcomes positively-impact on utility both directly and indirectly through increased income.

This paper fits in with the voluminous economics literature on the various determinants of health outcomes, starting with Michael Grossman's seminal work on the demand for health (Grossman, 1972a, 1972b). Over the years, several authors have sought to analyse these broad determinants empirically, within the scope of the health production function framework (Rosenzweig and Schultz, 1983; Kenkel, 1995; Thornton, 2002), with various additional variables proposed to explain health outcomes including stress levels (Thoits, 2010) and air pollution (Chay and Greenstone, 2003).

One relationship that has captured significant attention across various academic disciplines is that between health and economic development. Within the endogenous growth framework, health status has been modelled as a key input in generating economic prosperity (Ehrlich and Lui, 1991; Bhargava et al., 2001), with various empirical results suggesting that differences in health, generally captured via life expectancy at birth, can account for a significant proportion of variation in growth across countries (Barro, 1996; Bloom et al., 2004; Well, 2007). By contrast, the impact of economic development on health outcomes has received comparatively-less attention within the theoretical literature, although in recent years a vibrant empirical research agenda within this field has emerged, with several interesting results reported (Granados and Ionides, 2008; Granados, 2012; Babones, 2008).

The contrasting results reported can be attributed to several factors, including the countries included within the sample (Klasen, 2008) as well as the actual measures of health outcomes under consideration. Nonetheless, there is also growing evidence that the relationship between growth and health may be distinctly non-linear in nature. For example, Clark (2011) reports a negative and significant relationship between income and infant mortality, with the effect stronger at higher levels of income, as well as a positive impact of development on life expectancy that, by contrast, weakens as income grows. Ezzati et al. (2005) find the presence of an inverted U-shaped relationship between mean national income and BMI and cholesterol in a sample of 100 countries, which in turn may imply that cardiovascular disease incidence will primarily be concentrated in low to middle income countries, in line with Grecu and Rotthoff (2015).

Our study contributes to the literature on economic development and health outcomes in a number of ways. Firstly, we utilise a sample of 27 European countries over a 12-year period in order to estimate a relationship between income per capita and mortality rates from cardiovascular diseases, allowing for non-linearity in this relationship. The panel structure of our data allows us to control for unobservable factors which may be specific to each country and which may have a bearing on the hypothesised relationship estimated. We also estimate an extended model to incorporate a wide variety of social, economic and contextual factors as prescribed in the original health production function by Grossman (1972a) and subsequent work in the literature. Apart from analysing each determinant in its own right, this approach also allows us to assess the empirical robustness of the income-health outcomes relationship by scrutinising the various channels through which economic development may be linked with cardiovascular disease mortality. In addition, we develop a simple microeconomic model in order to explore the conditions under which a relationship between per capita income and health outcomes can exist and persist, and analyse the various forms that this may take.