Do national IQs predict differently across regions?

Not as far as we can tell

I am paywalling every post in December as an experiment. It turns out that families cost money and more so when you add a baby, so I am trying to see how much revenue a month of paywalling will raise.

Going back to Lyman Stone’s tweet:

in fact, national IQ’s predictiveness of GDP is often very weak within-region, i.e. within-ancestry-groups. measurement error is obviously a plausible part of this, but it also just tells us that whatever is happening with national IQs, they *might* explain why Europe and East Asia are rich and Africa isn’t, but they probably *won’t* explain why some East Asian countries are richer than others. they don’t much outperform just having categorical variables for major regions.

In my last post, I looked mainly at within country predictions of level of development from cognitive ability scores. Now these aren’t strictly speaking the same as within regions, or even within-ancestry-groups, which is what Lyman Stone was talking about. So let’s also do the worldwide regions to test his claim. Here I decided to use the premade Social Progress Index (not my composite variable from it, their own index). This is because Lyman Stone has concerns that I have been cheating with the construction of the composite index. In his words:

do you genuinely intend us to believe you did not engage in some kind of creative measurement to generate the finding that 2 SD of implied IQ generates exactly 2 SD of “socioeconomic development”?

how many different times did you respecify the model before you landed here

Alright, so here’s an index I definitely didn’t make. I didn’t even select the variables for inclusion (there are 51 variables). It was in fact made by some left-wing sociologists who made it because (according to their own promotion article in The Economist in 2013):

“GDP is simply too one-dimensional to provide a complete measure of a nation’s progress,” says Michael Green, who heads the Social Progress Imperative, a non-profit created to promote the new index. He mentions Nigeria and Ghana as examples. Nigeria has a higher GDP but has had slower social progress. Ghana, however, has a smaller GDP but ranks higher on the SPI. Moreover, says Mr Green, metrics improve accountability and create healthy competition. “They provoke action, the desire to be on top,” he argues.

Anyway, so let’s predict the SPI from national IQs. As before, I will be using the Seb Jensen NIQs (our most recent compilation). For regions, I like the United Nations ones which I have put here. These look like this: