Yesterday, I attended a Productivity Commission roundtable on behavioural economics and public policy. The papers are on the web. My takeaway from it was how poorly people outside academia understand at the moment what behavioural economics is all about. Many advocates saw it as economists getting into understanding what choices people really make and as a rationale for government intervention in those choices. Others saw the complexity as a reason to stick to normal economics.
That may be an implication, and when taken seriously, as I have noted before, it makes policy evaluation almost impossible. However, the recent research in economics is not about that at all. What it does is takes models from psychology and elsewhere and then looks at how firms might exploit this and how policy-makers should take it into account. It is not about understanding human behaviour but understanding the implications of human behaviour. This is incredibly useful but the main implication is that when you design policy you need someone psychologically trained to set in on the process and, more often than not, you need to consider real evidence — especially, randomised trials — to be assured you are doing the right thing.