New Russ Roberts video on income distribution statistics

Here is part 2 of his series.

The video tries to make clear the flaws in comparing the income of the median household in, say, 1970, with that of the median household today. In very important respects, they are not the same household, so that the question that you think you are answering is not the question that you are really answering.

Russ is trying to teach people to think for themselves about economic questions and data. That is absolutely admirable. I worry that most people would just prefer to use the credentials heuristic, which means that they will accept whatever interpretation of data comes from someone whose credentials they respect. The respected person might not be a careful economist. It might not even be an economist at all–it could be a journalist with very weak analytical skills who happens to get hold of some data.

Or worse, some people might use a confirmation-bias heuristic. If the interpretation of the data confirms the your bias, then you go with that interpretation. Otherwise, you tune it out.

2 thoughts on “New Russ Roberts video on income distribution statistics

  1. Once upon a time we learned in school that persuasive argumentation involves stating both the strengths and weaknesses of competing positions and then building a logical justification for preferring one over the other. Don’t see that being done in US publications very often. The public’s loss of confidence in journalists and academics provides a real opportunity for those like Russ Roberts who offer an alternative to appeals to authority.

  2. If people continue to use the credentials heuristic, then hopefully Roberts can influence current and future credential holders to be more responsible. But maybe that’s a pipe dream.

    I think the second part of his series was particularly well done. Liked it even better than the first.

Comments are closed.