Is health care a luxury or a necessity?

I ask this question in economic terms.  Luxury goods are defined as goods that increase as a share of a person’s (or country’s) budget as income increases; necessity goods are defined as goods that decrease as a share of a person’s (or country’s) budget as income increases.  There is actually mixed evidence on the topic with some studies saying luxury (Kleiman, 1974; Leu, 1986; Newhouse, 1977) and others saying necessity (Baltagi & Moscone, 2010; Di Matteo, 2003; Farag et al., 2012,; Parkin, McGuire, & Yule, 1987).

A paper by Shaikh and Gandjour (2018) tries to answer this question using an instrumental variables strategy outlined in a paper by Brückner (2013).  A simplistic approach would run a regression of GDP per capita on pharmaceutical spending.  However, there may be reverse causality if higher pharmaceutical spending improves long-term health and thus GDP improves.  To address this issue, the authors use international tourist spending as an instrument for GDP.  The authors use data on pharmaceutical spending from the WHO’s World Medicines Situation 2011—Medicine Expenditures’ annex and GDP per capita data from World Bank Open Data or the IMF’s World Economic Outlook database.  The authors find that: 

income elasticity of public pharmaceutical expenditure is greater than unity in the full sample…[however], GDP per capita has a statistically significant positive effect on pharmaceutical spending only for high‐income countries…

In short, public pharmaceutical spending is a luxury good.

The authors also find that pharmaceutical spending has a negative effect on GDP per capita.  At first glance this is surprising.  If people are healthier, shouldn’t GDP per capita improve.  In some cases, the answer is likely yes.  In other cases, while increased pharmaceutical spending would increase GDP, it may not increase GDP per capita.  Consider the example where all people die at age 70.  Assume that a new drug expands life expectancy to age 80 and also increases productivity by 20%.  In the table below, under the low pharmaceutical spending case, GDP is $100,000 and GDP per capita among the 2 people alive is $50,000.  With high pharmaceutical spending, overall GDP increases by 20% to $120,000, but GDP/capita falls by 20% to $40,000 since the elderly man who would have died at age 70 is now alive.  Further, the likely effect of pharmaceutical spending on GDP may take a long time to appear, for instance treatments that improve the health of children will increase school achievement and likely long-run earnings.  In short, the story is more complicated than it seems.

Source:

• Shaikh, Mujaheed, and Afschin Gandjour. “Pharmaceutical expenditure and gross domestic product: Evidence of simultaneous effects using a two‐step instrumental variables strategy.” Health economics (2018).