The man (the one clapping his hands) could be seen as a bit paranoid or simply as a fool. But he indeed knows (and can teach us) that sometimes (frequently) there can be something that – even if we don’t see it – is there! Obviously this is not the case. But take this others, more serious, examples (they are taken from Nassim Nicholas Taleb’s book The Black Swan) .

Until a certain point in history man believed that swans could only be white. Thousands and thousands of observations were, on a daily basis, confirming that belief. One day, though, when Australia was discovered (XVII century), also black swans were discovered. Thousands of year of observations were invalidated from a single observation!

Another interesting one goes like this: “There is a belief among gamblers that beginners are almost always lucky.” The fact, Taleb tells us, is that: “Those who start gambling will be either lucky or unlucky. [...] The lucky ones, with the feeling of having been selected by the destiny, will continue gambling; the others, discouraged, will stop and will not show up in the sample”[2]. Only those who continued gambling will remember the old days of luck!

Well now there is another interesting one and then we can get to the point. I once have read the results of a 2007 survey coming from a university claiming that something like 95% of its graduate students had found a job after a few months. Wow! Not bad, isn’t it? I don’t want to question the results and the rigour of such study, but I’d like to ask a question: “Couldn’t it be that only those who actually found the job answered your survey?” Maybe the others still desperately looking for a job were too hopeless or depressed to answer the questionnaire. Therefore they didn’t show up in the sample? I don’t know. It was only a doubt! But this would mean to give a distorted view of reality!

As you can see, there is a common denominator to the stories I reported above. There is always something that is not taken into account even if it is relevant to the matter. Silent evidence as Taleb calls it. Silent evidence refers to facts that are not taken into account. The consequence of ignoring such important data is that a distorted portrait of reality is given! Think about it, the matter is serious. It is not only something regarding nerdy scholars that are trying to falsify others’ work. It is about the way we are presented reality and the way we perceive it. We say that something happens because of something else, but we ignore part of the truth.

In economics research, for example, the same problem is known as survivorship bias. In fact, frequently, when studying a population of firms, only those that are still alive are taken into account. The consequence is that one can come the conclusion that these firms are doing particularly well because they have some particular attributes. But, what about those firms that didn’t survive? Maybe they had the same attributes and were only less lucky. This is similar to the gambler’s luck problem.

Think, for example, at the following affirmation: “Robert De Niro, Paolo Coelho and Coldplay have become famous and rich because of their incredible talent.” Well, it may be. Nobody here is claiming that these are not very good at doing what they do. But, what about the hundreds of thousands actors, writers and musicians (and whoever else) who did not become famous? There probably were loads of talented guys and girls among them. Probably equally talented. But they ended up doing something else. Again silent evidence at work.

Ok, some examples were very simple. But you get the point: it is dangerous to establish a line of causation (a strong one) when there are many other factors we may not be aware of and that can tell much more about the whole story. We have therefore to be careful of what we are said. There is (frequently) a side of the matter that is not being presented as it truly is. It could be because (whoops!) it is convenient to hide it or because  (damn it! I used it again!) one is not able to notice it (it can happen!). Therefore one has to be able to recognize that sometimes there is not a single cause for a fact. Reality is complex and fuzzy and it is difficult to understand.

A study done by Chicago University’s scholars Steven Levitt (co-author of the book Freakonomics) and John J. Donohue III, can be used to explain with an example the strange links that sometimes exist in reality[3]. In the nineties the U.S. witnessed an incredible drop in crime rates (especially murder rates). The fall was incredible: close to 40%! Explanations of every kind started being given by the media and various experts. It was because of the increasing use of incarceration some were saying. It is because of the increase of the number of police officers! It is because of better policing strategies! It is because of better economic conditions! And so on and so forth.

Though, the study found that the above advocated causes were only (a minor) part of the truth. They had played a role of course, but they could not have been the only reasons of the sharp decline in criminality. Levitt & Donohue got out with another explanation. It was also because of 1973 law (Roe v. Wade) that had legalized abortion! (why the hell?! One may ask...) In the paper they write: “States with high abortion rates in the 1970s and 1980s experienced greater crime reductions in the 1990s. [...] Legalized abortion appears to account for as much as 50 percent of the recent drop in crime.”[4] The reason, in a few words, is that unwanted children are more likely to grow up in an difficult environment (e.g.: abandoned teenager mothers, economically disadvantaged families, etc...). And “Decades of studies have shown that a child born into an adverse family environment is far more likely than other children to become a criminal”.[5] Interesting findings!

Obviously someone opposed to the findings of the study (maybe for political views on the matter or maybe for more scientific reasons regarding the method[6]). But I don’t care here. I just wanted to bring an interesting example to show how sometimes causes of phenomena are hidden (e.g.: delayed in time; mixed with other, more evident, causes; etc...) and that conventional wisdom cannot always explain it. Bear that in mind!