The subtitle of this book is misleading, because its doesn’t show that success always starts with failure. But that’s one of the few things I would really criticise about this book. Essentially much of what the author shows is that the best way to solve problems in complex environments is not by following some grand masterplan dictated from above. Instead, led by many examples, he shows that a bottom up approach, allowing room for trial & error tends to work better.
Having said that the first real example of the US Army, and its approaches in Iraq, doesn’t seem to quite fit. In that case much of what was required was already known, it was just being ignored by the powers that be who were convinced of their own ability to control things. I couldn’t help but think of the drone led wars that the US is pursuing which share much in common with the failed Iraq strategy – a reliance on intelligence, remote command structures and an almost overwhelming desire to avoid US casualties and so put ‘boots on the ground.’ Its clear that the current approach will never solve the problem.
The book is exceptionally readable, and on occasions this seems to be at the expense of more balanced discussions around the issues he examines. In the chapter on solving the problems of the bottom billion he advocates experiments and randomised trials, citing the benefits where they have been done and the failures where they haven’t. But some of the nuances that were revealed in, say, Poor Economics are lost. That’s not to say that our knowledge isn’t improving and that randomised trials aren’t playing a large part on that, but that we are not as far along the road as the author suggests.
I could raise a similar issue about the global warming chapter. At the risk of oversimplifying, the book says impose a carbon tax and that will solve the problem. He acknowledges its not without its difficulties, but skips the most important one which is that it won’t be implemented either in the near term or on a meaningful scale. The idea certainly has an intuitive appeal, but a significant country is highly unlikely to implement it – due to loss of industrial competitiveness – and international agreements seem even less likely. It may still be an idea worth promoting, but a realistic plan B would have been welcome.
Given my background in analysing the financial services industry and the ongoing economic difficulties in European banking I looked forward to the chapter on the financial services industry. One of the core elements of his proposition is that experiments, and elements within them, must be able to fail, cutting to the heart of much of the problem in financial services. His core idea, of introducing firebreaks to allow individual institutions to fail has much to commend it in theory. And some of the ways he mentions are being implemented e.g ‘living wills’ to allow work on winding up a bank to begin quickly. Ringfencing personal banking in the UK may fall into that category too. What I find interesting is that the standard approach in the US, led by the FDIC, is for a troubled bank to be taken over by a larger institution i.e. knocking any firewalls down and pushing other banks more towards being ‘too-big-to-fail’. While effective when a small bank is absorbed into a big one, the indigestion some suffered during the crisis suggest its not the best approach for a large bank.
There is an unanswered question, and perhaps it is unanswerable (and probably beyond the authors scope.) Firewalls are great when a single company fails. However it may not help in situations such as Spain, where the system is at risk because many banks effectively carried exactly the same risks – property in particular. In that case the firewalls are of little use, but the alternative is not clear.
As books go it is interesting, readable, educational and thought provoking. And that’s as good a recommendation as I can think of.