The world's biggest grift

I was labelled, officially and institutionally, of mediocre intelligence at age 13. It took me another 17 years to understand that this labelling was the product of perhaps one of the most pervasive cases of pseudoscience in history. This is not a story of my intelligence (or lack of it). This is a story of how humans want to believe so badly in simplicity and neatness, they’ll fall for nearly anything.

My labelling occurred between the limbo of arriving at school and first period. 15% of my classmates weren’t there, including ——y and ——y (names redacted, nickname syntax of appending a y to the surname included). I was asking questions. Some (usually quiet, and therefore more convincing) nerd told us her brother had taken a coach to Oxford University, along with about 100 other students on the ‘Gifted and Talented’ (G&T) programme. Presumably, the trip's intent was to ignite the required passion and grit to secure the grades, and thus position at Britain’s most prestigious university. I looked around ‘To being neither gifted or talented’ I joked. A few laughed, you didn’t need to be a genius to see the logical inevitability of the statement.

What struck me was not that the trip was taking place, but that there had been an attempt to keep it under wraps. This school was a rural country one, not much of anything happened, even a big sports game would lace conversations for weeks before, it was statistically unlikely that ——y or any of the pupils hadn’t mentioned the trip. Even more damning was that none of the teachers once mentioned the clear, large absence in each class. No, there had been explicit instructions not to talk about the visit, the teachers had been specifically instructed not to mention the absences in class. There was a conspiracy.

The obvious explanation is that the poorly executed subversion was a clumsy attempt to protect the normies' feelings. Probably that was part of it, but I now have a deeper explanation, the basis of the selection for the whole trip was based on a complete house of cards, and I think some of the more enlightened teachers were, rightly, ashamed. But before we dismantle this house of cards, we need to discuss a few examples of human fallibility, relating to our need to classify.

The grift of classification

Humans want to classify things, it’s something innate in us. Without pattern recognition your great^20 grandmother could have inadvertently hoovered up nightshade, thinking it black cherries, ending your life and her 4 million or so other direct descendants.

Classification, of course, has its uses in modernity, ordering your drill bits according to size is a better method than pushing them into their clips at random. But our primal need for classification can make people do very questionable things, like the (very offensive) trend of organising your books by colour. While dodgy aesthetics are relatively harmless, what isn’t harmless is when we use small s science to classify things that have no right to be classified. Data can be manipulated to tell any story and there is no shortage in history of noisy data being weaponised by capitalists to make people do, and believe, some very silly things.

It’s about their writing

In 1965, Francis T. Hilliger founded a company based on a 100-year-old science that was now gaining significant traction among corporations across the globe. Hilliger’s company provided tuition, analysis and expert testimony to some of the 15,000 companies in the UK and USA that utilised his science in their hiring decisions. Hilliger’s company Handwriting Analysis Ltd, was one of many that proffered to arm employers and employees in the art of Graphology, the practise of analysing handwriting to determine personality traits or suitability for a role.

There has still yet to be any scientific paper published that concludes handwriting can predict anything about a person, let alone their employability for a specific role. Graphology was a grift, perpetuated by entrepreneurs to extract money from tens of thousands of corporations, willing victims in the lie that something as simple as handwriting could help them make better decisions.

No it’s about a personality quiz

If you think we are beyond this - we aren’t. 8 years ago your author attended a week-long programme of team building and career planning when working for a large automotive company. The basis for the entire programme (estimated cost £1m) was a test we completed a few weeks before, a test you’ve probably taken at some point in your life - the Myers-Briggs Type Indicator (MBTI). The MBTI was developed by Isabel Briggs Myers and her mother, Katharine Cook Briggs, based on Carl Jung's theories of psychological types. Not only is the work of Jung completely lacking any scientific rigour, Jung himself stated that his work was not meant to categorise people rigidly.

The test is, unsurprisingly, pseudoscience. 50% of people get a different result just weeks apart (I’ve taken the test three times and had three different results). Even if you are in the cohort of people who score the same, there has not been a single peer-reviewed scientific study that demonstrates you can reliability predict any life outcomes from this test. The reason for MBTI proliferation is the same as Graphology, corporate executives with a mania for simplicity and categorisation and a willing group of entrepreneurs willing to sell overpriced testing and services to these turkeys. Currently, the Myers Briggs Company (the exclusive publisher of the MBTI test) does $63m in annual revenue and 80% of Fortune 500 companies use some form of the Myers-Briggs test.

The biggest grift

Ok where are we going with this. What do handwriting and personality testing have to do with classifying students as either gifted or not? It was researching the MBTI that got me thinking. If some very smart people can fall for such a gift, what other grifts are hiding in plain sight? What was actually used to select people for the G&T programme?

Reading up on the G&T programme, reveals, unsurprisingly a shitshow. Schools were asked to pick their top 5% or so students and enrol them in the programme. The schools were left to their own devices to select these students, some let teachers pick, so you’d have a mix of people who were gifted in say the arts, sciences, or sports, others based it on standardised testing, others on a combination.

Our school used standardised testing. I know this because not a single person who moved in or out of the programme. I also know this because while getting a dressing down over some sartorial misdemeanour in Mr J—— office, I had it explained to me that my ‘CAT scores were nearly gifted and talented’. I guess he thought that this should somehow correlate to the state of my shirt's top button.

The CAT (Cognitive Ability Test) is a test students take at age 11. All I remember of the exam is seeing a series of diagrams of folded paper, and then being instructed to label what the resulting structure would be. I remember a) having never seen anything remotely similar to this in my life, b) wondering what possible relevance this could have to anything c) freaking out a bit and probably spending way longer on this section than I should have.

A CAT is an IQ test. They label them CAT tests because apparently calling it an IQ test would be, to use the colloquial parlance ‘too triggering’, and here’s where we arrive at possibly the biggest grifts, a grift used by racists, intellectuals, politicians and CEOs alike for over 100 years. The grift of IQ.

IQ

IQ testing is pervasive. It’s used by governments, corporations and, as we’ve seen, in deciding the ability of 11-year-old children. I don’t need to explain to you how embedded the word IQ is when we talk about intelligence, calling someone ‘high IQ’ is common praise, calling someone ‘low IQ’ is a common degradation. It’s also pseudoscience. There is no evidence that IQ has any real correlation to real-world outcomes. IQ is so pervasive I don’t expect you to believe this, which is why I’m going to take the rest of the article explaining step by step why. I’ll be drawing heavily from Nassim Taleb’s paper Fooled by Correlation: Common Misinterpretations in Social "Science" as the scientific backbone.

IQ a brief history

IQ was developed in 1905 by Alfred Binet and Théodore Simon in France. It was never intended as a test to measure intelligence, it was designed to measure unintelligence (this may seem semantic, we’ll explore later, it’s not). Its purpose was to identify slow children who needed extra support. The US military saw this testing as an opportunity to mass screen recruits efficiently during WW1 when traditional methods of evaluation like interviews were considered too slow. 1.7 million people were sorted into 5 categories, ranging from category A - ‘officer potential’ to category E - unfit for service.

Noah Hurari has a quote - “You learn history not in order to predict the future, but to free yourself of the past and imagine alternative destinies”. WW1 set the use of IQ as a mainstream practice to measure intelligence, not because of scientific rigour, not because it was an effective predictor, but because, in a time of crisis, the largest employer in the world needed a way to, near instantly, categorise its recruits. Had WW1 not happened it's unlikely we would even use the word IQ, and the world would be a better place for it.

IQ doesn’t predict any real-world outcomes

The major flaw with IQ is it doesn’t make any accurate predictions of much of anything. It’s hard for us to quantify what is meant by success, but the ‘hardest’ measure we have is wealth. If the US Army thought IQ could identify their most capable leaders, clerks and logisticians, it would also stand to reason it should predict our best managers, engineers and other roles that would require above-average intelligence, and that, by the nature of supply and demand, also command the highest salaries. Indeed peddlers of IQ tests, would point to research that argues ‘there is no better predictor of employment outcomes as IQ’, let’s put that to the test.

Below is a raw plot of a study comparing IQ with net worth from Zagorky, 2007, who concluded ‘that each point increase in IQ test scores raises income by between $234 and $616 per year’, based on a sample size of around 7000 people. Social scientists are not known for their grasp of basic statistics but this conclusion from the data is so factually inaccurate you have to question the motivation of the study.

IQ and wealth at low scale (outside the tail). Mostly Noise and no strikingly visible effect above $40K, but huge noise. Psychologists responding to this piece do not realize that statistics is about not interpreting noise. From Zagorsky (2007)

At first glance, it might appear that there is indeed some kind of correlation between IQ and net worth, but as we’ll see this correlation is only at the left tail. Remove either the bottom quadrant of earners or IQ and it’s clear there is no correlation at all.

The only conclusion we can take from this study, if any at all, is that at the left tail end (very low IQ) there is a relationship between IQ and traditional measures of success. But any test of intelligence will tell you this, give me a sample of 100 people and in a five-minute conversation on say, why Mensa nerds are so obsessed with intellectual prestige and dressing poorly, and I’ll be able to give you the bottom 10% of intellect. Another stat for good measure for those more statistically inclined, the correlation between the same individual taking an IQ test is smaller than the correlation between the tests and the output, in other words, your IQ score on two separate tests has less of a relationship than the average IQ of a millionaire and someone with nothing.

IQ Circularity

Take a look at the below chart. Some would argue this is proof that higher IQ leads to more desirable jobs with higher intelligence requirements, no.

First, notice the variance. 25% of janitors have a higher IQ than 25% of college professors. Secondly, IQ testing is so embedded in society we cannot discount the circularity of IQ testing. To get into the schools necessary to become a professor, doctor, engineer or accountant you must have scored highly on a test similar to IQ, to quote Taleb ‘If you renamed IQ, from “Intelligent Quotient” to FQ “Functionary Quotient” or SQ “Salaryperson Quotient”, then some of the stuff will be true. It measures best the ability to be a good slave confined to linear tasks. “IQ” is good for (the late great) @davidgraeber’s “BS jobs”.

Aside on correlation

After circulating this argument among a few friends, it seems people still do not understand the crux of the argument. To any response along the lines of "Yes, but there is a correlation of IQ to outcomes of 0.3, which is really high," my reply is simple: You do not know what you are talking about. Even worse, you are likely parroting Jordan Peterson, Sam Harris, or some other alt-right snake oil salesman.

First, a correlation of 0.3, even if correct, is not significant. I do not understand why sociologists operate outside the realm of normal scientific standards. Bring a correlation of 0.3 to any hard science discussion, label it as significant, and you will be laughed out of the room. A correlation of 0.3 explains less than 10% of the variance—less than the variance observed when the same individual takes the same test twice.

Finally, correlation is not linear. Suppose there is a correlation of 1 between an IQ of 70 and 100 and income, but a correlation of 0 between an IQ of 100 and 130. The blended correlation would appear as 0.5. People then use this blended correlation to argue that IQ predicts real-world outcomes. It does not. What it actually shows is that extreme unintelligence (which has little to do with an IQ test) corresponds to real-world outcomes, while for those of normal intelligence, it does not.

The graphs below demonstrate this point clearly. A correlation as high as 0.8 can, in fact, indicate no meaningful correlation at all within the normal range.

IQ as a leveller

IQ is touted as some fixed variable, in a way it’s marketed as a leveller, giving rich and poor the same opportunities. Even this is a lie. IQ is not only useless, it’s also easily gamed. I was told by my school I couldn’t prepare for the CAT exams, it’s not true. Those weird paper folding tests I did – they’re called non-verbal reasoning, just practising the test once before takes your IQ up 8 points. That would move you from an IQ in the 85th percentile (normal) to the 95th (’gifted’). 62% of students have help preparing for the 11+ (a test very similar to the CAT taken at certain schools in the UK), while the stakes for the 11+ are a lot higher – a place at a great school, you have to think that a certain % of state school students had done at least some preparation for the CAT.

To conclude, IQ testing is a near-zero indicator of real-world outcomes, it perhaps shows some correlation (if we completely stripe the second-order effects of circularity) of how good someone is at taking IQ-based exams, designed by other IQ nerds. It selects for exam takers, bureaucrats, and other types of ‘intellectuals’ who are not adapted for real life, as evidenced by the zero correlation to wealth. People are fooled by IQ because it does correlate with extreme unintelligence, which if not removed from the whole sample, generates statements that are easily disproven such as ‘each point increase in IQ test scores raises income by between $234 and $616 per year’. IQ is a scam. If you need more convincing, please take a read of Taleb’s IQ whitepaper

The institutionalisation of exceptionalism

Despite writing nearly 3000 words on the subject, I’m not bitter about my CAT test or not being part of the G&T elite. If I’d thought the best universities were a possibility I may have worked harder for my exams, ended up at a good uni, and mimetic desire would have inevitably pulled me (along with seemingly 95% of our most promising intellectual class) into a lifetime of servitude to PWC.

I do, however, think it’s criminal we are using such an antiquated, game-able, unscientific measure to decide, in a small or large way, the fate of children. IQ is at best, a lazy slapdash way of assigning a label, at worst an institution to ensure people who don’t have the opportunity to prepare for such a test remain in their position.

It also opens a bigger question, why are we always looking for shortcuts to classify people? I have nothing against identifying exceptional people and doing things to promote their exceptionalism, if anything I think it’s a moral duty. But surely, by definition, exceptional, is not something you can measure on a standardised test. I have worked with one exceptional person, A——, (who incidentally didn’t pass the 11+, failing on ‘logic’). I identified him through watching him work, I hired him based on his ability to solve engineering problems on the fly. I never once gave him a test, I’ve never once given a test that isn’t an exact replication of real work that will be done by any person I’ve hired. I need the hire to work, so I take the time to assess them properly. Skin in the game.

Skin in the game is also a massively overlooked driver for exceptionalism. I was dumb until I started my own company, at which point the necessity to survive bought out something in me I didn’t know existed. The same is true for everyone I have hired. A—— has skin in the game, his identity is tied to his work, and he learns and performs at a 10x rate than those who view their work as simply a job.

The shortcuts in classification aren’t just around classifying a certain type of exceptionalism in a way that has no validity, it also stretches to what we classify as exceptionalism. We want simplicity. We want neatness. We want rules. You can be exceptional, but only so much so. You can be one among 20 people in your year who are all equal outliers, neatly grouped into a bucket of gifted and talented students. Students who are taken on the same field trips, funnelled into the same programme, with the same narrowly defined view of success.

The worst thing is, that identifying actual exceptionalism isn’t even hard, it just requires us weaning ourselves away from pop-psychology, Graphology, Myers-Brigs, IQ, and all these other very fallible, and very capitalist grifts. Grifts exploited (by actually smart people) to appeal to the needs of our amygdala to see patterns, correlations and trends where none exist, blinding us to the untapped human potential right in front of us.

Exceptionalism is obvious at five. Spend 30 minutes with any child and you’ll see where they have skin in the game, the elements of life they put their soul into, the places where they stand head and shoulders above their peers. By thirteen, exceptionalism is blurry. By twenty-five it’s buried. Not because it disappears, but because we’ve trained ourselves not to show it, and not to see it.

We’re victims in the institutionalisation of exceptionalism, trapped in the IQ system designed over a hundred years ago to reward bureaucrats, pen pushers and conformists. To say exceptionalism can be measured or categorised is double-speak. Great talent is not found at the right end of a bell curve – it’s found in late-night obsessions, unstructured curiosity. We need this talent to solve our biggest problems, it’s a moral duty, and if we truly wanted to nurture it, we’d leave it well alone.