Wednesday, September 28, 2011

Financial Times numeracy check

This article from the Financial Times is unfortunately quite typical of the financial press (and yes, not only the financial press). Just ponder the plausibility of what is reported in the following paragraph, commenting on a proposal by José Manuel Barroso, European Commission president, to put a tax on financial transactions:
Mr Barroso did not release details of his plan, except to say it could raise some €55bn a year. However, a study carried out by the Commission has found that the tax could also dent long-term economic growth in the region by between 0.53 per cent and 1.76 per cent of gross domestic product.
The article doesn't mention who did the study, or give a link to it. But there's worse. If reported accurately, it seems the European Commission's economists -- or whoever they had do the study mentioned -- actually think that the "3" in 0.53 and the "6" in 1.76 mean something. That's quite impressive accuracy when talking about economic growth. In a time of great uncertainty.

I would bet a great deal that a more accurate statement of the confidence of their results would be, say, between 0 and 2 percent crudely, or maybe even -1 and 3. But that would be admitting that no one has any certainty about what's coming next, and that's not part of the usual practice.