Issues with Mortality as a Measure of Freedom
Early in Chapter 1, Sen presents us with the story of Maitreyee and her husband from the Sanskrit text Brihadaranyaka Upanishad. Maitreyee is curious if amassing all of the wealth in the world would make her immortal, to which her husband denies: “there is no hope of immortality by wealth” (13). However, Sen takes this exchange as helpful to explain how wealth may provide “the capability to live really long (without being cut off in one’s prime) and to have a good life while alive (rather than a life of misery and unfreedom)” (14). A strong measure of freedom is having access to opportunities that allow you to increase your life expectancy. Many of the empirical cases Sen presents in the following chapters align with this: comparing the life expectancy of richer African Americans to poorer Chinese and Indians in Kerala (Fig. 1.1) or the lack of correlation between GNP per capita and life expectancy in a number of countries (Fig 2.1). All of this is to show that Sen uses life expectancy as a strong measure of freedom to support his argument comparing freedom to income.
My main issue with this approach is that life expectancy as a measure of freedom only supports some types of freedom. Sen helpfully names five different types of freedoms: political freedoms, economic facilities, social opportunities, transparency guarantees, and protective securities. Take a citizen of Hurleylandia (a country I just made up) where there is immense freedom for social opportunities (healthcare, transportation, education) and protective securities (disaster relief, unemployment benefits) that directly increase my life expectancy. However, Hurleylandia restricts all access to political freedoms, economic facilities, and transparency guarantees. I may be free in some regards, and I may live a long life because of it, but I am not free. Some ancient societies, such as Rome, gave slaves access to the best healthcare resources available at the time to maximize their life expectancy for the sake of maximizing their efficiency as workers in their prime physical condition. But, even with high life expectancy, it is hard to argue that these slaves have any freedoms—including social opportunities.
One rebuttal I anticipate Sen may make to my critique of this measure is that because of the “interconnectedness of freedoms” he refers to, promotion of one freedom will inherently promote others. But in Hurleylandia, ancient Rome, Singapore, or elsewhere, policy makers may create barriers that restrict the interconnectedness of the different types of freedom. For example, if I have the freedom of an adequate education to make myself literate and an effective orator, then by Sen’s interconnected argument I should be better able to practice my political freedom by, say, publicly protesting the government. However, in Hurleylandia, anyone who publicly shares their political opinions is immediately arrested. Now, promotion of social opportunities does not promote other freedoms.
I also want to point out that I do not think that Sen’s larger argument would suffer without mortality as a measure of freedom, as he does name other measures of freedom. This post is simply to point out flaws I see in his focus on mortality as one of these measures.
Hi Zach! I think you bring up some good concerns about using life expectancy as a type of freedom index. One way to reframe your argument could be to look at life expectancy in comparison to GNP and see which one is a better proxy for freedoms, as Sen seems to make the same argument you do about life expectancy and apply it to GNP. They both are not completely comprehensive of all types of freedom but we could ask which one best reflects the interconnectedness of freedom".
ReplyDeleteA response to my question could easily be, "Why don't we find an index that incorporates all kinds of freedom?". I think we could, but I also think there's a lot of merit in both GNP and Life expectancies. They are very easy for people to understand and it could be argued that they give lawmakers, development agencies, and states more autonomy as they can easily understand how to increase their ratings on one of these measurements. If a state is rejected from a trade agreement because they aren't "developed enough", using life expectancy or GNP makes it really clear to see what they would have to change if they really want to be a part of a trade agreement or reach a certain status on the world stage. (This example doesn't really make economic sense but ignore that for now.)
This argument could be countered by saying that we shouldn't just take the easy-to-understand metric but the accurate metric (which I think you're getting at Zach). The creator of GDP as a measurement, Simon Kuznets, agreed, warning against using it as a sole metric saying, "The welfare of a nation can scarcely be inferred from a measurement of national income as defined by the GDP". And yet despite that, we want to avoid Annapurna's indecision, so it is much easier to only have one metric. I think that having a larger informational base is important but maybe not realistic on an international scale. So while life expectancy has its own flaws, the question should be, "Is it slightly better than GNP?" as I think slightly better is a much more feasible change than implementing a metric with much fewer flaws. I think the fewer flaws to complexity relationship looks exponential, making it harder to argue for a much more accurate metric.