South Korea did not suffer the Asian Financial Crisis
We’re getting to a point, as a society, in our ability to collect and present data that I think it will soon be foolish to talk about, say, average GDP growth in the US instead of simply showing GDP growth numbers by state, or by county, or by ZIP code, regularly instead.
Averages tend to hide important insights rather than reveal them, and in general the lower the level of aggregation we can get, the better.
Thursday, May 8, is International Women’s Day, so I revisited the body mass index (BMI) data I looked at in my previous post, ‘Looking Up Everywhere‘, and graphed the growth in BMI for men and women separately for South Korea over the past few decades, just as I had with Russia.
For Russia, both men and women tended to gain and lose BMI in parallel. When women suffered, so did men. But South Korea tells a different story.
In the 1980s, both women and men gained BMI (that is, they likely gained weight as a result of improving economic conditions), but women tended to gain faster than men, with women averaging about 0.50% growth per year.
Economists like to say that South Korea went into recession when the Asian financial crisis hit in 1998, but the BMI figures say that this is not strictly true. Starting in that year, women stopped gaining BMI and, for the next few years, actually lost BMI (that is, lost weight).
But men’s waistlines, in comparison, didn’t even notice the troubles. Men continued to gain BMI after 1998 at the same rate they did before the recession. That is, it’s not really true to say that South Korea experienced a recession in the late 1990s. It’s more accurate, from BMI data, to say that South Korean women experienced a recession in the late 1990s.
By 2005 the economy had apparently fully recovered, and both women and men were gaining BMI at about the same rate again. So it’s my hope that, over time, we report average rates as little as possible and break down the data like this to the point where interesting differences like these are apparent.