Saturday, February 4, 2012

The Mystery of Merangin's Missing Women

Picture from here.
In 1990 economist and future nobel laureate Amartya Sen wrote a landmark article in the New York Times Review of Books entitled "More than 100 Million Women are Missing".  The essence of the article, based on demographic statistics and reason, was that in several parts of the world the normal proportion of men to women has been skewed by certain "cultural traits".  Sen points out that at birth, pretty much universally, boys outnumber girls by a ratio of around 105 boys to 100 girls.  However, since girls tend to be heartier than boys, boys experience greater childhood mortality.  In addition, women tend to be more resistant to disease.  This combined with a number of other factors means that women tend to have a longer life expectancy.  All of these conditions taken together means that in normal populations there are more women than men.

Sen's Logic

Amartya Sen photo from here.
Sen argues that in most of Asia (EXCLUDING SOUTHEAST ASIA and JAPAN) and North Africa this natural population bias has been disrupted.  He provides several examples; at the time the article was published the ratios in India and Pakistan were 94 and 90 girls born per 100 boys, respectively.  According to Sen there are a number of reasons for these unnatural disparities, ranging from female infanticide to sex-selective abortion to the neglect of female children, leading to a greater susceptibility to disease and a higher incidence of malnutrition, resulting in a far higher under-age-5 mortality for girls than for boys.  Moreover, a general disempowerment of women creates the social conditions necessary for the aforementioned dynamics to emerge.  In short, Sen argued that male children are preferred.  In all Sen estimates that there are more than 100 million "missing women".  Although Sen has his critics and some have attempted to refute his numbers and logic, there certainly is some truth to his thesis.

This brings us to Merangin district, where I've been doing some work over the past week.  As I was pouring over the socioeconomic statistics for the district, I noticed something strange about the population numbers.  I'll explain further in a moment, but first I want to introduce the population pyramid

Population Pyramids

Year 2000 pyramid for Mozambique from here.
A population pyramid is a graphical tool used by demographic geographers to display the breakdown of a population by age cohort and gender.  Normal population pyramids divide the population into 5 year age cohorts or blocks arranged chronologically with the youngest cohort on the bottom.  The cohorts are further divided by gender, so you can see how many males and females there are in each group.  Horizontal bars indicate the size of the cohort.  The graph is called a population pyramid because that's the shape of a normal growing population, as you can see with the example I've provided to the right.

I can't remember where I got this graphic.  Sue me.
Population pyramids are useful because they indicate whether a population is growing, holding steady, or shrinking.  These trends are important for planning; if you have an aging population then, as a policymaker you probably want to think about putting more public resources into improving healthcare.  If the population is growing, the leadership needs to think more about providing educational opportunities for the increasing numbers of youths.  It is important to remember that not all populations are growing.  Under normal circumstances, regardless of the overall structure of the population, there should be more women than men in every cohort.  Population pyramids can also show us the effects of certain historical events, as you can see from the graphic of the populations of East and West Germany.  Sometimes we see weird anomalies in population pyramids, and it's fun to try and come up with an explanation as to why there might be a weird gender distribution in certain cohorts.  Have a look at the population pyramids below and see if you can come with an explanation for the weird bulges in certain cohorts.  Once you make your guesses you might look at the wikipedia page for each place to see if you can confirm your answer.  I took the graphic from Rubenstein's Human Geography textbook which I use when I teach introductory geography.

The occurrence of war and epidemics can have a pretty significant impact on a population at certain times, which is reflected in the overall structure of the pyramid.  Similarly, the presence of colleges, military bases, prisons, and other facilities can account for demographic weirdness.

Merangin's Missing Women

Returning to my story, as I was going over demographic information I noticed something odd.  Unlike other districts in Indonesia, there are more men than women in Merangin.  At first I thought this might be accounted for by the presence of encroaching farmers that come from other districts to open up new land in the national park (presumably these would be mostly men working seasonally away from their families; I think that this probably accounts for some of the imbalance in the older cohorts), but a quick check of the age breakdown disproved this hypothesis as the disparity shows up among younger cohorts as well as old.  I made the population pyramid below using Excel; to make your own population see this great tutorial.

As I mentioned previously, Southeast Asia is an exception to Sen's observation of the missing women trend in Asia.  As far as I know, there's no evidence of any sort of sex-selective abortion or infanticide in Indonesia, and since the Merangin anomaly doesn't show up in other districts, there had to be some other explanation.  I asked my new friends at the planning office about this, and they were as vexed as I was.  One suggested that it could be explained by a lack of health facilities in the more far-flung parts of the district, but if this were the case boys would be affected as well.  We weren't able to come up with a good answer.  I had a colorful and imaginative conversation with my good friend Agung as we tried to come up with a possible solution, but because the various adolescent myths of the gender-determining merits of various copulative positions we'd heard were at odds (1), we didn't come up with a resolution.  As an open minded scholar, I can't rule out the influence of some hitherto-undocumented environmental factor or even the effects of thaumatology (2).

The only thing I've been able to come up with that makes any sense is that maybe encroaching farmers tend to bring male children with them to help work their newly-opened land, but I'm still looking for a definite answer to the mystery of Merangin's missing women.  If you have any ideas, drop me a line.

(1)  Most likely due to the fact that Agung is a product of the southern hemisphere, whereas I come from the northern hemisphere, and so the coriolis effect probably has some sort of influence.

(2)  The dark arts.

1 comment:

  1. In Sen's analysis, sex selection and infanticide contribute to the gender imbalance, but not as the primary reason. He points out that in a patriarchal society women are dependent on their male relatives (father, brother, husband..) - as they usually don't own property, and earn lower wage (or have limited paid work opportunities), As a result, women have fewer safety nets especially if they are widowed or childless. It is difficult to be poor, but it is doubly difficult to be poor AND female - leading to a shorter life expectancy among women.