why and when what works won't (part 2)

In the first part of this series a couple of weeks ago, I summarised a few ideas from Malcolm Sparrow, the theorist of regulation and organisation.  This was the post about the concept of “respecting the problem”, the idea that you need to take each problem seriously as a unique entity with its own setting and characteristics, then make the decision about what information to gather about it. 

I suggested (kind of implicitly, mainly through the title!) that this might be one of the factors behind the often disappointing performance of “evidence based” initiatives.  Starting out with a mission statement to “do what works” often means failing to follow the Sparrow sequence of actions (identify a problem – gather evidence to understand it – solve the problem – tell everyone what you did). 

It’s a great idea, in principle to try and take successes and do more of the same sort of thing, but the implication of Sparrow’s work is that this isn’t a good idea in problem-oriented contexts.  It’s more suited to regular, “operational” activities, including the operational activities of regulatory bodies (processing normal tax returns, regular monitoring of outflows, etc).  But problems are almost always problems because they are things which are not being handled well by the current control environment, and you’d be very lucky indeed if it turned out they had a size and shape that exactly matched a solution that had worked somewhere else.

Sparrow’s work on problem-oriented regulation is meant to tackle what he sees as the big conundrum; that the amount of regulation in society keeps expanding, but the increase in the perceived burden doesn’t seem to deliver any decrease in the perceived useful results.  The reasons why this is the case are quite complex and get into politics, but my favourite part of “The Regulatory Craft” is the bit where he explains why his solutions don’t work either.

The basic issue is that it is very difficult to be the kind of organisation that can regularly and systematically take a problem-oriented approach.  The identify-understand-act-tell sequence requires you to respect the problem, and this means respecting the fact that once it’s identified, the rest of the sequence will probably require an organisational form which doesn’t respect your own current systems and ways of working.

Consequently, it’s a model which requires constant regular reorganisations.  And reorganisations are difficult – they require a significant amount of whatever the equivalent of energy is for administrative and cultural change.  The only way that this model can be followed is by having a significant amount of what looks like “overhead” in its systems; capability that is dedicated to identifying and understanding problems and establishing what kinds of new  information and communication is going to be needed to act on them.

Not only that, but respecting the problem means that there is going to be a lot of activity that looks like “reinventing the wheel”, and a lot of failed experiments. (The “port-running” operation at the US Customs had several false starts, some of them involving quite expensive things like Blackhawk helicopters).  A problem-oriented regulator is going to look like it has a lot of waste and inefficiency in it; it will definitely look like it could be streamlined and made more industry-friendly by standardising its processes and doing more of what works.

This is why the last stage in the sequence (“tell everyone what you did”) is so important; problem solvers need to get a constant series of Ws on the board to protect their budgets and maintain their popular legitimacy.  But it needs to be a process of telling everyone exactly what you actually did, avoiding the considerable risk that successes get packaged up into generic solutions and turned into “what works” stories to promote to other institutions. And of course, that itself is very demanding in terms of institutional strength and incentives - it’s asking quite a lot of someone who’s just solved an important problem to not say that they solved it because they’re so great and everyone else should be like them.