That figures produce a feeling of accuracy, of rigorous knowledge, is something as well known as it is disputed. But without taking it to the more philosophical point of whether measuring is knowing or not, it must be said that the figures given, especially in social matters, should always be applied with a critical sense to avoid false illusions. And since we are now more sensitive than ever to fake news, it is worth noting that it is also with figures that it is particularly easy to deceive, whether intentionally or not. Therefore, without the intention of being exhaustive in such a short space, I will show five mistakes that information often falls into when it is based on data, and what confusions of analysis and perception they can create.
First of all, one of the most widespread errors is the confusion between facts and opinions. The study of how public opinion is manufactured has a long history. The reason is that what we believe we "think" - Bourdieu would say that, in fact, "we are thought" - is often situated in a different sphere from what our factual decisions are, which can be quite discordant. Passing off opinions as the result of reasoned, informed, critical and conscious thought, or as the sincere expression of behaviour, is a way of hiding the forms of social coercion that shape them and the self-deceptions in which we take refuge. This is the very common case of saying that one is left-wing in a survey, and of deciding and acting with criteria that would objectively be considered right-wing.
A second way of confusing social reality with the statistics we have is not taking into account that one thing is what can be registered and another is everything that escapes the institutional control systems. The precision of the figure that IDESCAT gives of how many inhabitants there are in Catalonia on January 1 of each year only refers to those who are registered and not to those who are in irregular situations, either because they do not appear in the census or because they are there but live elsewhere, and there are quite a few cases. Nor, for example, can the figures on the so-called poverty rate include the entire underground economy – around 20% of the GDP – which would reduce it very significantly. Nor is it easy to make transparent the data distorted by the fraud that exists in all areas and especially in taxes.
Something similar, thirdly, occurs with the statistics that are the result of voluntary declarations of certain facts. Data on the evolution of criminal acts, for example, can only include those that are reported, not those that actually occur. This is especially relevant when an increase in reports is confused with an increase in cases. After all, if effective awareness campaigns are carried out to report certain abuses, or means are put in place to make it easier, it is logical that such reports will increase regardless of whether the actual cases have also increased or even decreased.
Even more delicate is the case of statistics that we could call funded or sponsored. The fact that a coffee brand funds a university study that concludes that drinking coffee is good for your health should make the journalist who reports it raise an eyebrow. The fact that an organization that fights pollution gives data on how many deaths would be saved if we lived near a forest, or that a suicide prevention association tells us that cases have increased, should make us look very critically at its sources. And, very particularly, we should be careful when the data are covered by the white coat of the scientist.
Finally, and fifthly, there is the great difficulty of moving from concepts – often loaded with ambiguities and ideological biases – to a calculation system that will inevitably contain arbitrariness that is often ignored in the presentation of the results. What is meant by "risk of poverty" and how is it measured? How many times is the number of "immigrants" confused with that of "foreigners", without taking into account naturalizations? We already know that well-known economists, such as Joan Martínez Alier, discuss very well-foundedly the misleading value given to the GDP as a measure of wealth, apart from discussing how it is calculated? And what happens with technical issues such as the error margins of surveys that are never taken into account?
This is not a call to distrust statistics in general, but to be critical in each specific case. And, above all, it wants to be an invitation to be rigorous when using the figures and when publishing them.
