There are no exact sciences of human behaviour. Economics, criminology, sociology, and the various branches of psychology are amenable to study through observation, data collection and statistics but the results are invariably subject to interpretation and experts differ without being able to resolve their differences through critical experiments.
The human world is less predictable than the physical world. Very few can make fortunes backing horses or investing in the stock exchange. It is impossible to predict with certainty re-offending behaviour in the criminally insane. This problem was captured for all time by a former medical superintendent of Broadmoor Hospital who observed ‘We all know that half the patients in Broadmoor could be released, the trouble is we don’t know which half. Patrick McGrath was the Medical Superintendent and I met him as a young psychiatrist when I visited the hospital.
There is a particularly valuable generalisation which arises from these limitations in our understanding expressed as Goodhart’s Law. My preferred way of expressing this is ‘Any robust statistical pattern breaks down when people rely on it for forecasting.’ Reductionists think this is to do with the great complexity of human beings, because they think everything, in principle, can be reduced to physics. This is an astonishing dogma. In fact it is to do with volition or intentional behaviour. If we were in a deterministic straitjacket (or indeed a random and unpredictable world), why would not human history be as monotonous as history on the moon.