Correlation does not imply causation; if you really want to show that A causes B, you have to show why it does, find the actual mechanism, which is often somewhere between difficult and impossible.
Politics and the media use false correlations all the time and nobody seems to question this. I don’t know if the newspapers are not wanting to question too hard lest the politician not return for an interview, or the reporters simply don’t understand; in this increasingly polarised world I rather think that nobody cares.
A good example is the claim that Oxford University does not admit enough black students and therefore is racist. I don’t know the figures, but I, and particularly my partner, have worked with university dons and I simply cannot believe it; you never met a more liberal minded set of people genuinely trying to be fair to everyone. To me it seems much more probable that black students are more likely to come from a poorer area because their parents could not get such good jobs, and were therefore not so well educated.
This is a complex area, there are sociological, historical and cultural factors at work. My partner is from a working class (white) background and strongly discouraged by her parents from going to university because “people like us don’t go to university”. Her school said, outright, “we don’t send people to Oxbridge”. She was the first in her family to go to university and ended up with degrees in law, English, French, history and teaching having gone to both Oxford and Cambridge universities.
Statistically, by proportion of entrants for children leaving school with three A grade A-levels, Chinese come top, white in the middle and black near the bottom1. Quite how many factors are at work here is impossible to say, but you cannot call Oxford University racist because of a single correlation such as that.
This complexity is not new. In 19th century France, Sophie Germain worked on the theory of elasticity, and a lot of her work was incorporated into the design of the Eiffel Tower, yet her name is not one of those honoured by inclusion on the tower itself. That is primarily because she was a woman, but those who excluded her didn’t need to look far to find an excuse – her work was rather full of errors. Is this because as a woman she was unable to understand rigour? Of course not, it was because as a woman she never received the formal training in mathematics and had to teach herself.
Figure 1 – The names of the great and good men who designed the Eiffel Tower
This brings us on to the equality of opportunity vs. equality of outcome argument, essentially the same as positive discrimination. To cure Oxford university of their ‘racism’, should you:
- Bias Oxford’s intake (for example by setting quotas, i.e. positive discrimination) so that there are the same number of graduates between minorities2 with the same proportion of first class degrees etc. (equality of outcome)3, or
- Try to get all schools to ensure that the candidates achieve A-levels which are broadly comparable, irrespective of ethnicity (this is the same as the original problem, just pushed down an educational level), or
- Ensure that all suitable schoolchildren are encouraged to apply (equality of opportunity), or
- Find out why black schoolchildren get fewer good A-levels (ideal, but very complex and even slower to solve)?
I can see the arguments for positive discrimination, but I don’t think it works well. In Japan, women are allowed one day off per month for their feminine needs. It’s not a country that’s heavily bought in to the sexual equality paradigm, so men, wanting an excuse, regard them as unreliable because of this.
I also think those who benefit from positive discrimination will probably be less able to do the job than others, not through any lack of native ability or talent, but because in receiving the boost, they have by definition missed out on earlier education or training. Anyone looking for confirmatory bias (and there will be plenty) will certainly find it.
The last refuge of biased media is the phrase “is associated with”, implying causation without actually saying so. Coffins are associated with death – does that mean that coffins are hazardous for your health?
Notes and references
[1] https://www.ethnicity-facts-figures.service.gov.uk/education-skills-and-training/a-levels/draft-percentage-of-students-achieving-3-a-grades-or-better-at-a-level/latest
[2] And how many categories of minority are there? Skin colour, parental income, school, gender, sexuality … Inasmuch as we are all individual, this approach, taken to an extreme would have us all awarded the same degree and avoid the trouble of taking examinations.
[3] India does this to counter their caste system.