Sign In

Correlation and causation – Econlib

The Economist has an interesting article on obesity.  You may know that in earlier centuries the rich tended to be more obese, whereas today the correlation has flipped in many developed countries.  What you may not know is that this new correlation is almost entirely driven by women:

That poor people are more likely to be overweight has often been explained by arguments that obesity, in the rich world, is a feature of poverty. Poor people may struggle to afford healthy foods. They may reach for processed or fast foods because they lack the time to prepare meals at home or have less time to exercise because low-wage jobs often involve working long shifts and can be less flexible than those performed by the “laptop class”. Or because low income is often a function of limited education, perhaps, so goes the thinking, that lack of education extends to a lack of knowledge about how to maintain a healthy weight.

The problem with all of these explanations is that the correlation between income and weight at the population level in advanced countries is driven almost entirely by women. In America and Italy the relationship between income and weight or obesity is flat for men and downward-sloping for women. In South Korea the correlation is positive for men but this is more than offset by the sharply negative correlation in women. In France the relationship slopes gently downwards for men, but the slope is much steeper for women.

The Economist discusses factors such as bias in the workplace, but I’d like to offer another explanation.  Let’s suppose that thinness is at least mildly correlated with some other factor that leads to success.  (BTW, I’ve known highly successful overweight women, so I’m not suggesting that the correlation is anywhere near 100%.)

If thinness is correlated with some sort of X-factor linked to productivity, then why don’t we observe the same relationship for men?  The Economist suggests that women have a more powerful incentive to be thin, due to the fact that overweight women are viewed more negatively than overweight men:

All women eventually recognise the importance placed upon their bodies. It is as though girls are walking through a forest unaware and are then shown the trees. They can wonder how the trees got there, how long they have been growing and how deep their roots really go. But there is little they can do about them and it is almost impossible to imagine the world any other way. And the fiction that clever and ambitious women, who can measure their worth in the labour market on the basis of their intelligence or education, need pay no attention to their figure, is difficult to maintain upon examination of the evidence on how their weight interacts with their wages or income. The relationship differs in poor countries where rich people are generally heavier than poor ones.

Bryan Caplan has argued that colleges don’t necessarily make students all that much more productive, and that the wage premium from education largely reflects signaling.  A degree from Harvard doesn’t signal that you learned a lot at Harvard, it signals that you are the sort of person that can achieve a Harvard education.  Similarly, being thin might signal that a woman succeeded in the difficult task of holding down their weight in a world where physical labor is increasingly rare and we are surrounded by tasty food options. Because men are less likely to be victims of fat shaming, wealthy men are less focused on holding down their weight.

PS.  This post is purely descriptive, I’m not suggesting that this state of affairs is desirable.

PPS.  The photo at the top of this post shows Alfred Hitchcock and Tippi Hedren, which can be viewed as an example of Hollywood’s double standard regarding weight.  Younger readers might recall Hedren’s daughter Melanie Griffith.  Much younger readers might recall Griffith’s daughter Dakota Johnson.

PPPS.  Slightly off topic, this tweet caught my eye:

[The blank states in 1987 are missing data.  Notice how far New York and Kentucky have diverged just since 2003.]

If find this perplexing.  I was 32 years old in 1987, and I don’t recall life being much different than today.  America was full of fast food even back then.   My weight hasn’t changed at all since 1987, and yet I don’t have an unusual amount of self control–I’m just an ordinary person.  Lucky genetics?  Very likely.  But the genetics of the overall population hasn’t changed much since 1987.  Like you, I’ve read lots of explanations, but none of them seem all that plausible.  Warm states like Hawaii, California and Florida are a bit on the low side, but much of the south is more obese than the north.  I suppose that any time series/cross sectional explanation will require multiple factors–some combination of climate, income, education, gender bias, sedentary jobs, decline in smoking, processed foods, cultural attitudes, etc.

PPPPS.  This tweet is also interesting:

Rayna Prime

Rayna Prime

Rayna Prime Editor