Medicare For All is often described as being one way to achieve universal healthcare, a system in which all residents are assured access to healthcare. Such descriptions remind me of the tired but true adage that there is no such thing as a free lunch.
A while back, the Lone Economist conceded that the USA spends much more money on healthcare per capita than other large economies while suffering a higher death rate, but that this was an irrelevant statistic from an individual perspective (It’s the Perspective, Stupid). The posting was light on details and might have given the impression that I am dismissive of the shortcomings of the USA healthcare system. So, let me better explain my position.
The evidence that the USA spends much more on healthcare per capita than other countries is very well established. But the evidence that the USA death rate is much higher than other countries is less so. Facile international comparisons often draw unjustified inferences. Underlying cultural differences in countries might be driving the results and obscuring the true relationship between healthcare provision and death rates.
To prove my point, I compare the USA death rate to those of two other countries, Canada and Great Britain (GB). [Technically, GB is not a country, but I could not find death statistics for Northern Ireland.] Although the healthcare systems of these countries are different from that of the USA (i.e. they both have universal healthcare), they are English-speaking with similar economic and legal systems. Consequently, the potential for confounding, i.e. mistakenly measuring the effect of health insurance coverage on death rates due to unobserved reasons, is reduced, but not eliminated.
Figure 22.1 illustrates the number of deaths per 100,000 people in Canada, the USA, and GB in 2015. At 846, the USA’s death rate is comfortably between those of Canada (i.e. 740) and GB’s (i.e. 988); however, there are two underlying factors that confound their comparability.
Figure 22.1 Raw and Adjusted Death Rates per 100,000 persons in 2015
National Center for Health Statistics (https://www.cdc.gov/nchs/data_access/vitalstatsonline.htm), Mortality Multiple Cause Files, mort2015us.zip;
Office for National Statistics (https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/datasets/deathsregisteredinenglandandwalesseriesdrreferencetables), drtables15.xls;
Office for National Statistics
Statistics Canada (https://www150.statcan.gc.ca/t1/tbl1/en/tv.action?pid=1310071001)
First, death rates rise dramatically with age and both the populations of Canada and GB are older on average than that of the USA. Therefore, other things the same, the death rates of Canada and GB should be higher than the USA’s. If we adjust the populations of Canada and GB so that they have the same age distribution as that of the USA, their death rates shrink considerably. In fact, after this adjustment the USA has a slightly higher death rate than GB.
The second reason the raw death rates are not comparable is that death is counted equally whether the decedent is old or young. I think most people would admit that learning of the death of anyone causes sorrow, but that learning of the death of a child is especially horrifying. I also think these feelings are not entirely subjective. The younger one is, the more years of potential life one has, making death at a young age a worse – and objectively measurable – phenomenon than death at an old age.
If we accept the premise that the death of a child is worse than the death of an adult, then we should weight deaths of children more than deaths of adults. Differentially weighting the age groups results in the third set of death rates in Figure 22.1 labeled YPLL(100)-adjusted. After this last adjustment, the death rate of the USA is 75% greater than Great Britain’s and more than double Canada’s.
This weighting factor is based on Years of Potential Life Lost (YPLL). PYLL is normally calculated as 75 minus the age at death for those who die before age 75. But 75 is an arbitrary number that is supposed to indicate the age at which death is no longer considered premature. The average life expectancy in the USA, GB and Canada is greater than 75 – especially for people in their 60’s – so I used 100 instead of 75.
To understand why this adjustment makes such a large difference, see Figure 22.2. This graph shows GB/USA and Canada/USA death rate ratios by age group in 2015. If the distribution of death rates were the same for different age groups in each country, the bars in the graph would have approximately the same height. In fact, the death rate ratios for both Canada and GB are approximately 70% to 80% in the youngest age groups, i.e. 0-14, and are especially low in the working-age groups, i.e. 15-64. It is only in the retirement-age groups, i.e. 65 and older, that the death rate ratios are close to 100% or greater. This illustrates that comparative death rates in Canada and GB vis-à-vis the USA are especially low in the younger age groups when years of potential life are at their highest.
Figure 22.2 GB/USA and Canada/USA Death Rate Ratios by Age Group, 2015
The different death rates by age group between countries are a result of many factors, not just the availability of health insurance. For example, relatively permissive gun control laws in the USA explain some of the difference, but even after excluding all deaths due to external causes, such as gunshot wounds and suicides, the adjusted death rate for the USA is much larger than that of the other two countries. Immigrants have relatively long life-expectancies and Canada has a 50% higher proportion of immigrants than the USA; however, Great Britain has the same proportion of immigrants as the USA, so the “immigrant effect” does not explain the very significant differences in adjusted national death rates.
Back to my main point, no country can afford to provide unlimited healthcare to everyone. There are not enough doctors, nurses, and pharmacists in existence or even potentially in existence to do this. Even countries with universal healthcare must limit the provision of healthcare in some way.
Figure 22.2 is particularly telling about the consequences of the different ways GB and the USA limit healthcare. GB limits healthcare by intensity of treatment while the USA limits it by age. Great Britain’s National Health Service does not pay for several expensive procedures and medications that the USA’s Medicare and private health insurers pay for, but a significant percentage of the USA population under age 65 have only the relatively stingy Medicaid or no health insurance at all. Consequently, GB has significantly lower death rates for people under age 65 and significantly higher death rates for people age 75 and older.
From a national perspective, there is an undeniable arithmetic logic to universal healthcare. It results in fewer years of potential life lost. From an individual perspective, however, the math is much different. People who would otherwise not be able to obtain private health insurance, a minority of the population under age 65, gain while the majority loses. The fact that the sum of the gains is far greater than the sum of the losses does not change this distributional reality.
For many cultural and historical reasons, getting people to look at our healthcare system from a national perspective is a heavier lift in the USA than it is in the rest of the world. Individualism is our defining national characteristic. Calling single-payer systems, like Medicare For All, “universal” implies that we can have a system that imposes no individual sacrifices. Unfortunately, there is no such system.