There’s No Place Like China
Some of the discussion pertaining to Mara Hvistendahl’s new book Unnatural Selection: Choosing Boys Over Girls, and the Consequences of a World Full of Men has unfortunately gotten sidetracked, in my opinion, in the politically incandescent topic of abortion.
There are several means of selecting the sex of a child, prior to conception (for example, through sperm sorting), during pregnancy (abortion) and after birth (infanticide). But regardless of how that sex selection occurs, Ms. Hvistendahl’s thesis (to the best of my understanding) includes:
1. Sex selection was promoted by Western influences, both governmental and non-governmental, primarily in the 1960s and 1970s as a means of reducing population growth
2. Sex selection it is a major and growing problem in China, India and the rest of the non-Western world, and
3. This is both a moral issue and potentially has dire demographic and social consequences
Fortunately, Dr. Hans Rosling of the Karolinska Insitute in Sweden (the organization that selects the recipient of the Nobel Prize in Medicine each year) has developed a software tool called the Gapminder that pulls together social data from various public sources. Using this software, we can check to see how much of this thesis holds up.
Sex selection is (predominantly) a Chinese problem
In a recent column at Salon, Ms. Hvistendahl says: “[W]omen should account for half of the human population, and in parts of the world they now account for far less. That alone justifies moral outrage.”
But what Ms. Hvistendahl does not say is that her phrase “in parts of the world” means basically “China”. The graph below, made from Gapminder, shows the sex ratio in children ages 0-14 vs. per-capita income levels in the countries of the world for the year 2005. Births of baby boys naturally outnumber baby girls by a ratio of about 105 to 100, so this line is indicated on the chart as the ‘biological norm’.
In the wealthiest countries, where sex selection is not a significant problem (the US and the small orange circles representing Western Europe being prime examples), the sex ratio for young children is close to the biological norm. Some wealthy countries show large deviations from this level, such as South Korea (where boys are more common than expected) and Saudi Arabia (where girls are).
The general trend is a decreasing sex ratio (that is, more young girls than expected biologically) as per-capita income drops. In only a handful of countries does the sex ratio lean more heavily male (say, greater than 108) than expected biologically. In far more places (mostly sub-Saharan Africa) does the ratio lean more heavily female (say, less than 102). That is, if we exclude China and India from consideration for just a few seconds, worldwide, there are more young females than expected biologically. In most of the world, it is boys who are missing, not girls.
However, the two elephants that can’t be ignored, of course, are China and India. Let’s look at their sex ratios over time, not only for young children, but also for adults and older people.
(click image to enlarge)
Western influences don’t seem to have been very influential
Ms. Hvistendahl claims that Western influences engaged in efforts to promote sex selection in the non-Western world. (Note: I have not yet read Unnatural Selection, so I am not familiar with the details and evidence behind this.) “I make very clear in my book”, she says in her Salon piece, “that the victim [of sex selection] is women — women who in the 1960s and 1970s were used as pawns in a Western drive to reduce birth rates in Asia”.
India’s data shows that ‘Western influences’ throughout the 1960s and 1970s do not seem to have had any detectable influence on the sex ratio of young children, which according to Gapminder has stayed essentially constant in that country for the past 50 years (and, if Gapminder’s projections are correct, should remain stable until at least 2050).
The only place where trends look in any way worrisome are in China, where the ratio of young boys to young girls began rising in the 1980s, after China’s 1979 introduction of the ‘One Child’ policy. There is little evidence suggesting a rise in the sex ratio of young children during the 1960s and 1970s, when ‘Western influences’ in China were supposedly pushing sex selection. This ratio is (according to Gapminder’s projections) at its peak right now and is expected to decline from this point forward. Of course, this means that the ratio for 15-49 year old adults will only peak in the coming decades, probably the 2040s.
Better development is associated with higher sex ratios
Ms. Hvistendahl offers (to the best of my knowledge) ancient Greece and the settlement of the American West by white people in the 19th Century as examples of ‘the Consequences of a World Full of Men’. However, in a previous post I used Gapminder’s data to look at correlations between the sex ratios seen in the world and metrics of social well-being. That post showed that countries with sex ratios closer the biological norm tend to be the wealthier countries. Poorer countries tend to have sex ratios that indicate that harm is falling disproportionately on boys, not girls.
Countries with higher sex ratios of adults (that is, higher levels of adult males relative to adult females) tend to have lower murder rates, not higher ones. However, I was not able to look at factors like sex trafficking, which might legitimately be issues.
At the risk of repeating myself, based on what Gapminder’s data shows, three keys points of Ms. Hvistendahl’s thesis seem to be in question:
1. If Western influences made a concerted effort in the 1960s and 1970s to encourage sex selection in Asia, they were largely (if not entirely) unsuccessful.
2. Only China is potentially a “powderkeg” in terms of the sex ratio of adults. India’s sex ratios among people under 50 have been stable for decades. Almost all other countries have sex ratios near or below the biological norm.
3. Gapminder’s data suggests that a high boy-to-girl ratio is more correlated with high income, high HDI and low murder rates than the other way around.
It does not seem possible to me (though maybe that’s just some fault on my part) that both Gapminder’s data and the thesis behind Unnatural Selection are correct.