Depends on who you ask – there's a certain school of economists that insists that a sufficiently "rational" argument is in fact preferable to evidence.