Lord of the Files

 

I’m on a kick lately of trying to use the data in the Gapminder software to demonstrate or rebut claims I see elsewhere. (For example, see my recent post where Gapminder’s data seems to show that people in Asian countries drive well for their level of vehicle prevalence, compared to the population-weighted average of other countries).

I saw this book review at The Wall Street Journal’s website of Mara Hvistendahl‘s new book Unnatural Selection: Choosing Boys Over Girls, and the Consequences of a World Full of Men, about the selective abortion of females over the past few decades, in which she claims: “Historically, societies in which men substantially outnumber women are not nice places to live. Often they are unstable. Sometimes they are violent.” (I have not yet read the book.)

Jonathan Last, the author of the WSJ review, goes on to state: “There is indeed compelling evidence of a link between sex ratios and violence. … In Chinese provinces where the sex ratio has spiked, a crime wave has followed. Today in India, the best predictor of violence and crime for any given area is not income but sex ratio.

Gapminder contains data about sex ratios, broken down by country and by age category (though not for individual Chinese provinces nor Indian states, at this point). In Gapminder, each circle on a plot is sized proportionally to that country’s population and colored according to the country’s region: Asia being red, the Americas (North and South) being yellow, Europe and Central Asia orange, sub-Saharan Africa blue, the Middle East green, and South Asia light blue.

I made a graph from 2005 data of the sex ratio of people age 15-49 and found something surprising. (In each graph below, I have colored the region where males outnumber female light blue, and the region where females outnumber males light pink.)

For example, as the male-to-female sex ratio drops (that is, as there become more women relative to men), the murder rate tends to go up. (These are just correlations, so maybe the converse is true: as the murder rate goes up, the number of males goes down. Or, maybe some third variable is affecting both of these other factors, and so on.)

(click image to enlarge)

China and India both have high male-to-female sex ratios, and both have low murder rates. Many African (and some American countries, like Guatemala and El Salvador) have more women than men, and more murder.

Now, murder is a rare event, even in the places of the world where it happens most frequently. So, maybe some other measures of social well-being will show that places that are the best off tend to have a sex ratio for children that is near 105 (the biological normal) and those that are less well-off see the ratio go up, as daughters suffer, get aborted, or are just generally treated less-well than sons.

In fact, from Gapminder’s data, the opposite seems to be true.

As the Human Development Index drops, the sex ratio for people ages 0-14 (China and India being major exceptions) tends to drop. That is, girls don’t seem to suffer in low-HDI countries, boys do.

(click image to enlarge)

The same is true of per-capita income. In rich countries, the boy-to-girl sex ratio is about 105. But, as per-capita income drops, boys go missing (again, with China and India being the exceptions), not girls.

(click image to enlarge)

Also, according the Gapminder’s data files, the fewer years women in their 20’s and 30’s have in school, the more young girls there are. (Again with China and India being the glaring exceptions).

(click image to enlarge)

So, I don’t know which is wrong: Ms. Hvistendahl’s argument or Gapminder’s data files. Or both. Or maybe there’s some rationale that shows their is correct.

I’m not a sociologist nor developmental economist, so I’m not going to try to delve into explanations. Like I said, at this point I’m just collecting discrepancies between what people claim to be true and what Gapminder’s data files say.


  1. 1 Don’t look now honey but I think the internet just ate our economy. « excapite

    […] up is the insight Bradd provides when he sifts through the data (See Lord of the files) to discover that contrary to the ideas presented Mara Hvistendahl’s new book Unnatural […]




Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s



%d bloggers like this: