OK, this may be a bit pointy headed, but it has got me thinking. I ran an early draft of this post past Ben Ramalingam (see pic), who thinks a lot about this kind of thing, and include some of his comments here.
Fact one: we NGOs are always calling for the regulation of what are called ‘negative externalities’ – where the actions of one actor cause much wider ‘pollution’ – of the environment, or society, or the economy, for which they are not normally held responsible. That can involve ‘pricing in the externality’ eg by taxing polluting activities, or simply trying to stop or reduce it through regulation.
Fact two: a number of development wonks, such as Ben and Owen Barder, are trying to get our colleagues to take complexity and systems thinking more seriously. One aspect of that is that in tightly coupled, complex systems, it’s very hard to predict when one minor action (a new government in Greece saying ‘hey, we’ve got no money’, or a Tunisian stallholder setting himself on fire, or chaos theory’s best known example/conundrum, ‘can the beat of a butterfly’s wing in Brazil set off a tornado in Texas?’) might trigger large-scale and unforeseen (and unforeseeable) changes.
So what happens when you put these two together – how on earth can you regulate or tax butterflies and the disastrous externalities they may cause? This is discussed with relation to the financial crisis on the FT’s alphaville blog:
“Part of the difficulty in designing better regulations is that it’s hard to price the sum of the actions of individuals. No one looks at themselves as the butterfly that caused the hurricane in Europe.
How many people on structured credit desks thought they were doing a bad thing by trading financial products that were going to blow up in the crisis? And you can bet that most of the people at AIG Financial Products thought that they were doing a good job at something that was a low risk business. Sure, a few people clocked it, and AIGFP made the decision to get out of some segments of the businesses when worrying signs emerged. How many people who worked there would have believed their actions would result in a $182bn bailout financed by taxpayers?”
The answer would probably be to reduce the overall level of fragility of the system – putting in firebreaks such as dividing up retail and investment banks (the so-called narrow-banking approach) or putting limits on size.
Finally, Ben says we might also need to think about more use of scenarios (think about Rumsfeldian known knowns, unknown unknowns etc) and create more dynamic forms of regulation, that change as the context changes, and that kick in when certain thresholds seem to be appearing (I guess that includes things like the circuit breakers on the stock market that kick in when prices vary by more than a certain amount in the course of a day – any other examples?)
Ben has blogged about whether catastrophes can be predicted – and reckons the jury is still out. He quotes Ben Bernanke’s reflections, which are hard to argue with:
“Like weather forecasters, economic forecasters must deal with a system that is extraordinarily complex, that is subject to random shocks, and about which our data and understanding will always be imperfect… To be sure, historical relationships and regularities can help economists, as well as weather forecasters, gain some insight into the future, but these must be used with considerable caution and healthy skepticism.” (emphasis added by Ben R)
It might be very hard to do much more – you can’t regulate for specific butterflies, or even prove causality when the hurricane hits, but you can improve anticipation in general, and try to have adaptive measures in place. Any other thoughts?
The full Alphaville blogpost is worth a browse, and the comments contain this lovely analogy of the Robin Hood Tax:
‘Think of spam email: when sending emails is essentially free, sending out millions of spam emails can be profitable even if a fraction of respondents would take the bait. But if you had to pay even a nominal charge, even less than a penny, per email sent, that ‘business model’ would essentially become loss making in many cases. The Tobin tax would have a similar effect on a lot of this ‘phantom liquidity’ we get across many markets through high frequency traders – who, after all, are playing a zero sum game mostly, with their profits essentially being losses for a lot of other players. A small transaction cost might just be enough to discourage a lot of this socially useless activity.’
[h/t Rob Nash]