Ludger Woessmann tells us about some dire facts:
It’s about their new 75 page working paper which they summarize thus:
How far is the world away from ensuring that every child obtains the basic skills needed to be internationally competitive? And what would accomplishing this mean for world development? Based on the micro data of international and regional achievement tests, we map achievement onto a common (PISA) scale. We then estimate the share of children not achieving basic skills for 159 countries that cover 98.1% of world population and 99.4% of world GDP. We find that at least two-thirds of the world’s youth do not reach basic skill levels, ranging from 24% in North America to 89% in South Asia and 94% in Sub-Saharan Africa. Our economic analysis suggests that the present value of lost world economic output due to missing the goal of global universal basic skills amounts to over $700 trillion over the remaining century, or 11% of discounted GDP.
Ah, so they have produced a new set of world IQs, I mean world ‘basic skill’ scores for 159 countries. Their methods will be familiar to people who have read these kinds of studies before:
The quality of educational performance information varies significantly across countries and unsurprisingly tends to correlate with their current level of development. Recognizing the varying limitations of available data, we separate the database construction into five layers of decreasing reliability that indicate different degrees of certainty and precision in the comparability of available international test information. Layer 1 includes countries that have participated in any wave of the Programme for International Student Assessment (PISA) or PISA for Development (PISA-D) – a total of 90 countries. Layer 2 adds countries that have participated in the Trends in International Mathematics and Science Study (TIMSS) but not in PISA – 14 additional countries. Layer 3 incorporates countries that have participated in regional achievement tests – TERCE and SERCE in Latin America and SACMEQ and PASEC in Sub- Saharan Africa – but not in PISA or TIMSS, an additional 20 countries. Layer 4 merges in the two countries for whom sub-national regions have participated in PISA – India and China.
These 126 countries with direct assessments of students represent 84.8 percent of the world population and 95.7 percent of world GDP. For an additional 33 countries that have not participated in any internationally comparable achievement test (Layer 5), we impute achievement based on measures of educational enrollment and the achievement of similar countries in terms of world region and income level. This allows us to provide estimates of achievement deficits in 159 countries with a population of at least one million or a GDP that is at least 0.01 percent of world GDP. These 159 countries cover 98.1 percent of the world population and 99.4 percent of world GDP.
A central element of the analysis is the development of a method for reliably combining the available assessment information to place the countries of the world on a common achievement scale. Even though the different tests were not designed with that objective in mind, we show that it is possible to transform student-level achievement on all tests into a PISA-equivalent score while introducing minimal constraints on the underlying score distributions. Our method equates the scales of the different tests by using the student-level distributional information found in the group of countries that participate in each pair of test regimes. We rescale the performance of countries participating only in TIMSS (Layer 2) or in one of the regional tests (Layer 3) onto the PISA scale using the underlying distributional information from countries jointly participating in these and in PISA. From the resultant database on the distribution of achievement, we can produce measures of the share of students (not) reaching basic skill levels.
The full underlying achievement distributions provide common support at the student level which is fundamental to our harmonization of scores across tests. This is particularly relevant for countries in Sub-Saharan Africa that perform outside the observed range of average achievement on the broad international tests. Previous transformation methods based on linear extrapolation from country mean scores tend to overestimate these countries’ true achievement levels.
How rich would the world be if every country had the same high intelligence as the top Western ones?
We use our skill measures to quantify the economic gains that the world could reap from reaching the goal that every child achieves at least a basic skill level. Using estimates of the association between skills and long-run growth rates from existing empirical growth models with worker skills, we project country by country the future path of GDP with improved skills. The discounted added world GDP amounts to over $700 trillion compared to the status quo GDP trajectory over the remaining century. This economic gain from reaching the goal of global universal basic skills is over five times the current annual world GDP, or 11.4 percent of the discounted future GDP over the same horizon. Put the other way around, this amount documents the lost economic output due to missing the goal of global universal basic skills. Importantly, the gain from lifting all students who are currently in school to at least basic skill levels turns out to be more than twice as large as the gain from enrolling the children currently not attending school in schools of current quality levels.
But what do they really mean with “basic skills”?
This definition of basic skill levels may be thought of as a modern definition of functional literacy. Without the necessary skills to compete and thrive in the modern world economy, many people are unable to contribute to and participate in development gains. Literacy was once defined in terms of the ability to read simple words. But in today’s interconnected societies, it is far more. It is the capacity to understand, use, and reflect critically on written information, to reason mathematically and use mathematical concepts, procedures, and tools to explain and predict situations, and to think scientifically and draw evidence-based conclusions. For development, citizens around the world will need the basic skills that industrial employers seek and that the formal labor market rewards. While some developing countries may today appear unprepared to employ even basic skills fully, past analyses (described in section 5.1 below) suggest that even subsistence agriculture can benefit from basic education and that the natural evolution of economies involves expansion of technologies that employ the available skills.
Compare with this famous definition of intelligence from Linda Gottfredson:
Intelligence is a very general mental capability that, among other things, involves the ability to reason, plan, solve problems, think abstractly, comprehend complex ideas, learn quickly and learn from experience. It is not merely book learning, a narrow academic skill, or test-taking smarts. Rather, intelligence is a very general mental capability that, among other things, involves the ability to reason, plan, solve problems, think abstractly, comprehend complex ideas, learn quickly and learn from experience. It is not merely book learning, a narrow academic skill, or test-taking smarts. Rather, it reflects a broader and deeper capability.
Anyway, here’s their ‘not world IQs’ map:
It look the same as every other such map. Cue your favorite meme.
But mathematically, it’s unsatisfactory because it’s a nonlinear transformation telling us which proportion of children are below some arbitrary ‘basic skills’ threshold. We could transform the data, but the authors did it for us:
And the mean scores by region:
However, these scores appear to be uncorrected for differences in enrollment. You see, no one really enrolls all children in the standard education system, as this would be impossible due to the requirements of the far left tail and children with serious disabilities. But countries differ in how many they enroll and since the children who are not included are below average, one has to adjust for this sampling bias. The authors make this adjustment for their data table of skills, but only for the “% below basic skills”, not for the “mean ability”, so it seems we have to do our own conversion after all. I put their data in this spreadsheet for ease of use. Here’s my transformed version of their data, scaled to national IQs using Lynn’s 2012 dataset.
Sub-Saharan African countries had a mean IQ of 71.6 according to Lynn 2012, and according to Woessman et al, they have 71.2 IQ. This shouldn’t surprise readers as every dataset produces numbers like this.
Their working paper is otherwise unremarkable. It:
- Does not mention Richard Lynn, David Becker, Heiner Rindermann
- Does not mention intelligence, world IQs, or anything else of the sort in order to interpret the results in familiar constructs. There is one mention of IQ, but only in passing while discussing Chinese regional variation.
- Does not cite any research by intelligence researchers produced over the last 20 years despite these being economists who are doing intelligence research.
We can ask ourselves: are the authors not familiar with the large body of work by intelligence researchers? Seems unlikely. In that case, why don’t they cite it? It is somewhat of an academic violation to deliberately avoid referring to relevant work, but we know why, because anyone who goes into that territory gets harassed. In that view, the authors are playing it smart. Spreading knowledge of the results, though under different names. The different names add to the conceptual confusion but at least the data are accumulating, ready until a more sensible time when there are less Wokes around to get people fired.
