The map that so many people hate

by Sasha Alyson

A tweet of the above map, along with a link to our story about gender parity in education, drew more replies than anything we’ve done, heavily from Australia and New Zealand.

Some were thoughtful, from people who’d read the story. From others, who clearly had not read it, the most printable replies said “That’s a lie!” or “Go back where you came from.” Whatever happened to the good old days when people read the message before they shot the messenger?

Some of the thoughtful ones, however, pointed out real concerns. As I did further research, I found three quite different issues that merit another look. Let’s take them one by one.

1. Comparing regions on the map

UNESCO published a gender-parity chart based on what percentage of children, in six regions, failed to achieve Minimum Proficiency Levels (MPLs) in math and reading. The chart was hard to take in at a glance, so I converted the data into a map, and suggested using it to compare overall gender parity in education in various regions.

That was a mistake. This index is only one part of the picture, and it can be a misleading part. Parity figures involve dividing one number by the other. When doing that, if the numbers are small, a slight difference can look more important than it really is. This is easiest to see with an example, using actual UIS data (Minimum Proficiency in Reading, end of secondary school) for three countries.

The first two columns show the percentage of females, and males, who achieved this level. In nearly all countries where data exists, females do better in this category.

Then I calculated the gender parity index in two ways. (For ease of comparison, I made the calculations such that GPI less than 1.00 means boys are disadvantaged. Some tables calculate the ratio in the opposite way.)

The first shows gender parity amongst those who DID reach MPLs. Australia was closest to being equal, with about 76 boys for each 85 girls, or 89%.

Then it shows the gender parity index based on those who did NOT reach MPLs. The country order is reversed. Algeria looks much better: 72 girls compared to 85 boys do not achieve the MPLs.

Neither way is “right.” Both show a part of the picture; neither shows it all.

2. Did I make up the data?

Many people charged me with making up the data, even though the source – UNESCO – was clearly stated. I used a report from the UNESCO Institute for Statistics (UIS), and then averaged 4 datapoints per region to make a simpler chart. There were a couple of judgment calls involved so below, I’ll explain exactly what I did. There’s nothing controversial here.

3. UNESCO did pull some fast stuff

It seems quite clear, on the other hand, that the UNESCO Institute for Statistics did manipulate and misuse data, as well as scramble numbers, and is quite free about estimating data that it doesn’t have while not being upfront about that. Some of this I’m still researching; in The Rabbit Hole of UNESCO Statistics I’ve documented several highlights, and a story linked below documents previous deceptive behavior.

Data source

I created the map using data published by UNESCO.(1) There’s good reason to be skeptical of U.N. statistics but that’s what is available. Even by its own numbers, the U.N. and the big NGOs are deceiving the world about gender parity. They use this issue to claim the high moral ground. It is sheer, self-serving hypocrisy.

Data for six regions comes from the UNESCO chart shown below. Inexplicably, UNESCO didn’t include Oceania on their chart so I drew it, using UNESCO data from the same report.

The chart shows four categories of the gender gap per region:
Blue: Math, lower secondary age
Red: Math, primary age
Green: Reading, lower secondary age
Orange: Reading, primary age

Perfect gender parity gets a score of 1.00. Sub-Saharan Africa hit that with math, lower-secondary age, so there is no blue line to see. Numbers lower than 1.00, with lines going to the left, mean that boys are disadvantaged; to the right, girls are disadvantaged.

When you take data from one source and display it in a new way, there are always judgment calls involved. There’s a lot of room for hidden biases to creep in, and it’s not always accidental. In the interests of transparency, I’ll explain two decisions I made.

First, by U.N. definitions, gender parity is achieved if the actual number is within 3 points of 1.00; that region is gray, on their chart. For the map, I took the average only of how far the line extended beyond the gray. So if the score was 1.07, that became 4 because it went only 4 points beyond the gray. (The chart wouldn’t look much different if I hadn’t made this adjustment.) Here are the average values, by region:

Second, suppose that in one region the blue line goes 3 points to the left, and orange goes 3 points to the right. How do you handle that? Do you add 3+3 and get 6? Or treat the first number as a negative, and get 0? Neither approach is perfect, but the second seems far better for describing inequality. Under the first method, a region where all the lines went 5 points left would be treated as comparable to one where half went left, half went right. But inequality isn’t about random fluctuations; it’s about consistently putting one group at a disadvantage.

Even if you used the the first method, however, not much changes:

Notes and Sources

1. Chart and data are from More Than One-Half of Children and Adolescents Are Not Learning Worldwide, Unesco, and Unesco Institute for Statistics, Fact Sheet No. 46, September 2017 (UIS/FS/2017/ED/46). Chart is on page 9. (I added white space and a line between the regions; UNESCO had put them flush against each other, making it harder to read.) Oceania data is in tables which, again inexplicably, are far apart. Reading is on page 3, Math is in an annex at the end.

Related stories

Left: Gender parity in education. The aid industry tells us that girls are short-changed on education, especially in Africa. They’re twisting the data, to bring in the dollars. Here’s the story which gave rise to this follow-up.
Right: The Rabbit Hole of UNESCO Statistics. Why did UNESCO drop Oceania from its chart? We found that UNESCO repeatedly manipulated things to avoid upsetting Western countries, and to disguise its lack of actual data.
Left: Schools in the global South are getting worse. The aid industry doesn’t ask why, and for good reason. It created much of the problem.
Right: Funny numbers. “Aha!” Sherlock Holmes exclaims, examining a UNESCO chart that shows education levels by region. “This isn’t a statistics institute, it’s a propaganda department!” How did he know?
Left: The Global Partnership for Education. For nearly two decades, GPE has claimed to improve education in the global South. It’s got hundreds of datapoints. Is it doing any good? On that, GPE says “Insufficient data.”
Right: It’s still colonialism. Today’s foreign aid and charities look different from colonialism of the past. But the old goals and attitudes haven’t changed.
Left: Cellphones and literacy. UNESCO took money from big tech to publish a deceitful report that benefited the company that gave it the money. If an African president had done that, what would we call it?
Right: The high cost of meddling. Aid organizations claim they want to “fix” problems, but they’re really driven to create jobs for themselves, while pleasing foreign donors. They end up meddling, and it carries a high cost.

Other stories of interest

Left: Rethinking education. In many countries, the schools simply don’t work. We can’t fix this until we ask: What should be the top priorities of education? Here are some ideas.
Right: World Vision undermines local economies by giving away leftover merchandise where it’s not needed. It makes no sense — until you understand the financial incentives.
Left: The brain drain. As the coronavirus pandemic spreads, wealthier countries are trying to extract a particularly valuable resource from the others: Doctors.
Right: What would make a better future? There are ways that wealthier countries can genuinely help others, if they want to. Give the aid money directly to the poor, for example. Here are ideas.