Oh, yeah? Prove it! (Evidence and Arguments, Part I)
“Let’s assume for the sake of argument that [something is] intuitively accepted. That doesn’t relieve you from having to look at the evidence to find out if that’s true.” – Thomas Sowell
Imagine the following scene: a defendant sits in court, quivering as he waits for the trial to begin, knowing what is about to happen. The prosecution stands up and declares, “The defendant is guilty of murder.” The judge shrugs, and then states plainly, “Guilty as charged – court adjourned.” At this point, most readers may be wondering about the validity of the trial – acceptable procedure was not followed, the defence never spoke and, worst of all, absolutely no evidence was provided by the prosecution. While the particular setting of this scenario may be purely hypothetical, its essence is not: the use of evidence and logic as the primary determinants of argumentative soundness and subsequent decision-making has certainly taken a back seat in many of the contemporary discussions around economic, social and political matters. While poor evidence is often given, perhaps more concerning is the acceptance of claims for which justification and evidence are neither provided nor asked. Our focus here will be on two such claims.
The first claim is the myth of proportional representation. Using the representation of women as an example, the argument states that since women make up about half of the population in South Africa, they should naturally occupy half of all jobs across all categories; invariably, the focus is on high-level corporate management. Because women only make up about 20% and 30% of top- and senior-level management positions respectively, they must be facing some sort of institutional discrimination – or at least, that is the claim. Very little evidence, if any, is offered to this effect. In fact, as is usually the case, a cyclical argument is provided as a substitute for evidence:
There are too few women in certain positions because of discrimination. We see that there is discrimination, because there are too few women in these positions.
Irrespective of the order in which the previous two statements is read, the argument effectively assumes what it sets out to prove – much like the prosecution in the earlier illustration. Logical errors aside, the main point here is that no evidence is offered to back the claim of discrimination. This is certainly not a denial that discrimination may be taking place, in the same way that our defendant in the earlier illustration may have murdered the victim. Rather, it is important to first establish the truth of such claims – grounded in facts – before jumping to conclusions that we intuitively feel are correct. Note that everything said here is applicable to other demographic factors, such as race.
Let us, for the sake of argument, accept that discrimination is taking place in hiring against people of a particular sex or race. The most important question is whether, were it not for this discrimination, we would see people proportionally represented in all fields and hierarchies. In other words, is discrimination the only thing that causes jobs not to be occupied in proportion to a country’s demographics?
On this matter, Thomas Sowell, who has conducted much research in many different countries, is particularly insightful. In Malaysia, for instance, he compared the two largest ethnic groups at the time of his research: the Malays, who formed the majority of the population, and the Chinese, who made up the largest minority at about one third of the population. He found that these groups were proportionally represented in universities, but the trends in their choices of degree programmes were far more interesting points of information: Malays outnumbered the Chinese 3:1 in liberal arts degrees, while the Chinese minority outnumbered the Malays 8:1 in science and 15:1 in engineering degrees. Add to that the fact that harsh anti-Chinese discriminatory laws were present in Malaysia, it becomes very difficult to explain the fact that the average Chinese person in Malaysia earned more than double that of the average Malay by invoking discrimination alone (particularly when it worked in favour of the Malays who, historically, were always the politically and socially more powerful group). This is but one of numerous examples.
Sowell’s main point is that nowhere in human societies, whether historically or in the present day and whether by ethnicity or some other characteristic, do we see the proportional representation of groups in different fields. It is therefore curious that this expectation has emerged as a standard to work towards, specifically because there is no evidence to support it. Taken to its logical consequences, the proportional representation mindset implies that people are nothing but random variables, following a relatively predictable pattern of behaviour. It negates their talents, culture, personal values and interests. It discounts the value of individual choices in such important matters as education and skills development, which career to follow, and how far along the career path one wants to progress. It is also important to resist the urge to believe that Sowell and others are claiming that “group A does not engage in business” or “group B is not interested in corporate management”; rather, the claim is that it is logically untenable to say that demographic proportions are (or could be) a causal mechanism in determining who does what.
This brings the discussion to the second, albeit a much more brief, point. During a discussion about diversity and affirmative action recently, the well-known argument was raised that having a diverse management structure in business leads to the generation of more ideas and strategies, thus improving the success of the business. The conclusion on this basis was that affirmative action is a good policy. The problem here is twofold: the first is, as always, the lack of evidence to support this claim – but even if we concede that these benefits of diversity do occur in a business, there is no reason to conclude that a government-mandated policy should enforce hiring standards. After all, if the benefits were as voluminous as is claimed, the businesses that hire in a diverse manner would perform better and grow in market share relative to those that do not; the market mechanism would be sufficient to encourage diverse hiring practises, without which businesses and their owners would stand to lose. It therefore makes no sense to propose a politically-enforced solution, when the market ‘solution’ is adequate to the intended purpose. If one advocates the policy of affirmative action, there may well be good reasons for doing so, but using ‘diversity’ as justification is not one of them. All of this is not to say that diversity is undesirable, that it has bad effects or even that evidence does not exist for its potential benefits; it is merely highlighting the erroneous conclusions drawn from nothing but intuition and wishful thinking.
Until better evidence and arguments in favour of some of the positions that have been discussed arise, and provided one has critically engaged with the arguments, it is perfectly reasonable to think that such arguments are fruitless. Of course, all of this has given very little attention to why many people tend to omit evidence for their arguments. That may well be a discussion of interest, but the first priority should be getting people to think for themselves and approach all claims with a healthy – not toxic – dose of skepticism. As is the case when dealing with a fire, the more urgent course of action is putting it out rather than finding the root cause – or, to go back to our initial scenario, we should first make sure that the courtroom weighs up the evidence.
[Update (1/10/2015): As a prime example of what this article discusses, Emma Watson recently stated that she has experienced sexism in the movie industry, and cites the disproportionately high number of male directors and producers with whom she has worked. While her wording is somewhat clumsy (she refers to herself as having experienced sexism, as opposed to female directors and producers), the underlying absence of a logically-drawn conclusion is obvious.]
 According to the Employment Equity annual report for 2013/2014
 Note that the figures used are based on research conducted by Sowell in the 1980s and 1990s, but the conclusions are nonetheless still relevant.
 ‘Policy’ in the sense of being enforced by government, rather than a voluntary business practice.