A bombardment of ideas presented through real life accounts of people, companies and events makes for stimulating reading. But the book never convincingly develops the theme of success always starting with failure.
Yes, Donald Rumsfeld's handling of the aftermath of the Iraq war adopted a flawed approach. Porteus and HM McMaster, who succeeded Rumsfeld, responded to conditions on the ground, adapted their policy to the circumstances and were judged to succeed. A top down approach similar to Rumsfeld's may have previously have been adopted by Lyndon Johnson and McNamara in Vietnam, but I am not clear about where the success was achieved. These examples are examples of similar misguided top down approaches being inappropriate - not necessarily that success always starts with failure.
On the argument that effective top down planning in the modern economy is rare, Tim Harford is more convincing. He holds that expert, professional judgement only has limited success in predicting outcomes. The market economy is like the biological, evolutionary process proceeding by trial and error and survival of the fittest. It responds to the ever shifting landscape and is too complicated for pundits to accurately predict outcomes.
Harford maintains that if trial and error and selection of the successful actions is key, then variation is important in providing multiple opportunities for organisms, or companies, to find the best reponse to a changing external environment. Planned, centralised organisations, like the NHS, operate at a huge disadvantage. A decentralised process of trial and error is the solution, bringing in pluralistic approaches and encouraging innovation. But then the throwaway observation that Walmart and Tesco are centralised but able to manage experimentation undermines the absolute necessity for this approach.
Harford tackles the obvious necessity for careful planning is some capital intensive industries, such as North Sea oil exploration, nuclear power stations or banking where trial and error is not an option. He approaches his solution with a richness of case studies. The solution he sees is making failure survivable by decoupling the system in which organisations operate.
An interesting section on why the patent system may produce a steady stream of high quality scientific research but does not lend itself to the revolutionary breakthrough is followed by a section on the use of prizes, as per Longitutude Prize in 1714 or prizes offered by Bill Gates today, in stimulating multiple approaches to a given problem.
Whilst thought provoking on a variety of novel ideas, the catchy title is not proven.