On the Inaccuracies of Economic Observations