Don’t trust what you can’t understand

economics ▪ miscellaneous

Numbers is alluring. An economist can choose to spend hours explaining the mechanism through which a fuel subsidy is bad for the poor. Or else, s/he can try giving two numbers – the increase in fuel prices and the increase in the number of poor people – and say that by plugging the former in his/her “economic model” s/he gets the latter. More often than not, that person will face less arguments and leave the meeting earlier and a happier man (or woman) using the latter strategy.

But if you are in the audience, you should always suspect an economist (or anyone, for that matter) who tries to convince an audience of non-economists with such a strategy. I’ll let you in on a secret: We economists know that mathematical models scare non-economists. Having elaborate and opaque models to make policy arguments to non-economists (especially without explaining the underlying mechanism) is a sure sign of weakness.

So my advice is, if you are in the audience, feel free to intimidate the economist into explaining his or her model’s underlying mechanism and assumptions. You have the right to understand the model before trusting the numbers it’s churning out. Even if it takes hours, so be it!. After all, it’s the economist’s fault for trying to exploit the audience’s numerophobia.

Why this concern? Economists aren’t supposed to be paranormals. Models are supposed to illustrate, not predict – and even that must be done under certain assumptions. So in order to use it in front of a lay audience, the economist needs to explain the underlying assumptions and mechanisms clearly enough for a non-economist to understand. And, most importantly, the actual numbers should matter less than the mechanism behind them.

By the way, here is the article in The Economist that inspires this entry.