What do people think?

In a typical interview or press conference with an athlete or a coach, he or she may be asked something like “do you plan to be in the same club next season?” He or she may find the question uncomfortable, but if in the answer she includes, hidden somewhere, the word “yes,” chances are that the headline in the media a few minutes later will be: “Athelete/coach X plans to continue in the same club next season.”

Smart people would use this as an example of the intellectual mediocrity or opportunism of the Sports media, because the athlete was not “planning” anything, or if he or she was planning anything, it was perhaps to avoid being asked about the topic, to even avoid having to decide. However, it is not very different from the kind of expert analyses that are derived from opinion polls and surveys.

This is important because how people truly form their opinions is a key input in the analysis of democratic societies and of particular aspects of them, such as politics or business. Economists such as Harvard’s Stéphanie Stantcheva have recently devoted significant efforts to learn, through online representative surveys, how people form their opinions on a variety of subjects, from taxation to international trade or inflation. I am also working on this with colleagues, especifically about how people think about public vs private ownership or about the importance of competition.

As Stantcheva and her co-authors have illustrated, it is important to complement closed questions with a few options, with open questions where subjects can openly express their views without limitation. It is interesting how people develop their thinking when they are less primed by the options given by the researchers.

The example of Catalan Independence and Catalan identity (or similar identity problems) illustrates my concern. Lots of political debate hang around surveys where, for example (taking the last one by an official sociological body of the Catalan government), around 40% of respondents say they are in favor of Catalan Independence when there is just another option (no to Independence), and the figure goes down than by more than 10 points when there are more than two options. This is an example of the well known fact that the framing of the questions determines the answers to a substantial degree in many areas. In both cases, it is not clear what independence means (in or out of the EU or the euro-zone, for example), or what the other options are (what is a state in a federal Spain as opposed to an autonomous community).

But pundits and experts use the results of these surveys to say things like “support for Independence” has declined (or increased), when they should just say that the answer to these questions in surveys has increased to decreased in some way or other. Similarly when they are asked whether they feel only Spanish, only Catalan, more Catalan than Spanish, etc. What if they don’t feel anything about this, or they care little. Surveys do not capture vague preferences or the intensity of them.