Life expectancy gaps related to income/education
I’m currently writing a topic on ‘the causal effect of education on health’, so this is a topic I’ve looked at a bit – consider this post a ‘workblog’-post, even though it’s only tangentially related to what I’m working on.
This kind of stuff – health disparities related to education and income – pops up in the public debate every now and then, see e.g. this recent article (in Danish), or this analysis by AE-rådet (also in Danish). This is ‘politics’ to some extent (see the previous post), but it’s also a question about what’s actually going on in the world, and the latter type of question is the type of question I tend to be interested in answering. I’d like to make some general points here which are sometimes overlooked:
i. People with lower education are fatter. And being fat is bad for your health.
ii. People with lower levels of education smoke more: “Well-documented declines in smoking prevalence over time have not occurred evenly throughout society (12, 13). They have been most substantial among the most educated. Thus, the least educated form increasing proportions of those who remain smokers.” Regarding alcohol the picture is more complicated (as I’ve talked about before), however it should be noted that if the variance of the quantity consumed by the highly educated is lower than for the lower educated groups, as they claim in the article I link to at the beginning of this paragraph, then it would make sense if the highly educated people who die from alcohol-related diseases die later and lose fewer years of their life to the alcohol than does the group with low education (‘the uneducated alcoholic loses 20 years, the educated alcoholic loses five…’). Either way alcohol matters much less than smoking, and the differences aren’t that big in the former case. Incidentally the causal pathways of the smoking link are still unclear: “The causal pathways between education and smoking are both complicated and contested in the literature.” (link)
iii. Lifestyle differences among different educational groups make up a big part of the difference in health outcomes: “the mediating effects of health behaviors – measured by smoking, drinking, exercising and the body mass index – account in the short run for 17% to 31% and in the long run for 23% to 45% of the entire effect of education on health, depending on gender.”
iv. An additional point related to point iii.: I haven’t looked for studies on this because it’s obvious, but the health gradient is more sensitive to stuff like income level and employment status in countries like the US than it is in Denmark. So international (non-Scandinavian?) estimates of the magnitude of educational effects and income effects on health outcomes are likely to be biased upwards, compared to what the magnitude would be in a country like Denmark where ability to pay for medical services problems are unlikely to have much influence on life expectancy at this point.
v. I’ll spell out this point even though it should be obvious by now: Many of the reasons why people with a low education on average die too soon relate to the fact that they on average make poorer choices when it comes to their health. And the stuff mentioned above is just a small part of what’s going on; you also have related stuff like information channels and compliance differences, on top of stuff like ‘likelihood of seeking proper medical attention conditional on you actually needing it, and ability to verbalize complaints so that the doctor makes the correct inferences’ (e.g. a lot of T2 diabetics don’t get diagnosed, and this lowers their life expectancy significantly).
vi. Note that whereas it’s true that some jobs are still more unhealthy than others (a traditional mechanism most people think of when they’re thinking about these things), if the connection between type of work and health risks is known people employed in such jobs would be expected to earn a risk premium – this is not super relevant when you look at education and health, but it is something to have in mind when analyzing health and income stuff.
vii. It should be noted that if you get better over time at treating people for stuff that isn’t lifestyle-related and so stop a lot of people from dying early on of other causes, then lifestyle-stuff is going to become a big driver of health disparities.