There was a fascinating article in the Financial Times a few days ago, about the 'right kind of stupidity'.
It describes the idea of 'functional stupidity', conceived by Mats Alvesson and Andrew Spicer, describing the 'bovine state' which organisations can sink into - large organisations in particular. The idea is that organisations which attract particularly clever staff, like banks and management consultancies, are particularly prone to the kind of 'functional stupidity' that overwhelmed the Soviet Union.
Or as the UBS chief put it about the Libor rate fixing scandal: "A mechanistic reliance on risk processes".
The insight that which the FT's Andrew Hill brings to this debate is that organisations need a bit of functional stupidity just to get by, otherwise you have clever people questioning everything. You need agreed processes.
What he doesn't say is that agreed processes is, paradoxically, exactly the route by which organisations also become 'functionally stupid'. That is why it is the large organisations, which need processes just to survive, that are the most prone to it. Vision is clearer in small organisations. So is a sense of responsibility.
But the debate so far, at least as set out in the FT, doesn't cover the most ferocious method that organisations become functionally stupid. They boil their objectives down into numerical targets - or, worse, the government does it for them - that purport to measure everything. This is Taylorism, and Taylorism famously asks employees to leave their brains at reception, to be picked up on the way home.
The explanation for the staggering ineffectiveness of so many services in the first decade of this century was this very reliance on numerical targets. Like nearly everybody else, I haven't read the long-awaited report on the way older patients were abused at Mid Staffordshire Hospital, but it seems a clear cut case of functional stupidity brought on by an over-reliance on targets.
Targets appear to simplify the task for staff. It makes them concentrate on bowlderised versions of the job. Too simple means stupid.
This is important when it comes to working out how public services can be more effective. Management consultancies which have pushed the idea of breaking processes down into numbers have been, not just a terrible waste of money - but they have also been infectious carriers of functional stupidity.