Highlights the dangers of using unrealistic mathematical models of human, organisational & market behaviour to guide policy prescriptions. Though real markets work better than known alternatives, the author makes the crucial distinction between the real & the speculative-financial, where totally different realities apply.