Implementation

Larry Summers says something important and overlooked here: "it is much easier to design policy than to implement it."

This is surely true. It's a cliche of ministerial memoirs that a man arrives in Whitehall with high hopes only to find that reform is harder to achieve that he hoped. Take for example universal credit. Most of us would agree that it is a good idea in principle to simplify benefits and reduce withdrawal rates (let's leave aside the level of the benefit). However, the implementation of the scheme has proved trickier than the basic idea. A similar thing might be true of free schools and I suspect that in foreign policy - in peacetime and war - implementation is everything and top-level policy design relatively trivial.

However, media reporting of politics underweights this. Political reporting, and especially comment, dwells upon either soap opera or policy initiatives. Of course, there is much reporting of implementation but this is mostly after-the-fact descriptions of individual failure rather than analyses of the structures which produce that failure - which corroborates the old jibe that a journalists is someone who watches the battle from the mountaintop and then rides down and bayonets the wounded.

This is to be regretted because there are strong reasons to suspect that the policy implementation process is sub-optimal.

One is that politics is dominated by what I've called cargo-cult management. There's a fetish of "leadership" and "boldness" which encourages a neglect of the unglamorous gruntwork of proper management: tracking progress, achieving small partial targets and overcoming problems. This neglect will be magnified by cognitive biases such as overconfidence and groupthink.

Perhaps the most grievous problem, though, is a lack of information. "Make sure you get real-time, high-quality data" says the Cabinet Office Implementation Unit. Giles Wilkes' experience suggests this advice has, well, not been implemented:

Much of the time, Whitehall throngs with officials struggling just to find out what is going on. The sound of dysfunction is not the cacophony of argument, but the silence of suppressed documents and unreturned phone calls.

There are many reasons why this is the case. Some are exacerbated by politicians' own stupidity and arrogance; if they think bad news challenges their egotistical self-perceptions they'll be loath to hear it and underlings will censor themselves, and if whistleblowers are threatened with disciplinary action they'll keep quiet. Other problems, though, lie in the very nature of hierarchy. It easily produces silo mentalities and a lack of trust which impedes the information flow necessary for proper policy implementation. It also yields perverse incentives: it's much more pleasant to give your superiors good news than bad - so guess what they'll hear?

In saying all this I'm merely echoing something Kenneth Boulding wrote back in 1966 (pdf):

Organizational structure affects the flow of information, hence affects the information input into the decision-maker, hence affects his image of the future and his decisions, even perhaps his value function. There is a great deal of evidence that almost all organizational structures tend to produce false images in the decision-maker, and that the larger and more authoritarian the organization the better the chance that its top decision-makers will be operating in purely imaginary worlds.

The questions he posed back then have not satisfactorily been solved. It is not just the England football team that has failed to progress since 1966.

Comments

[Luis Enrique]

I have been told that the people who devise policy in the UK govt often have almost no contact with the people who will have to implement it. I think this is puzzle because I cannot see anything in the nature of a hierarchy that stops planners consulting implementers. Planner bosses can command planner minions to talk to implementers.

[rogerh]

A feature of the English class system. Once worked for a consultancy with strategists and implementers. Some strategists spoke nicely, wore good suits and shiny shoes and specialised in very high level 'roadmaps' and 'concepts'. They went down well with some Ministers and the snootier commercial firms. Did not get on with the implementers who tended to find their roadmaps and concepts unusable. Best kept apart.

The more earthy commercial firms and a some govt departments preferred the remaining strategists who were not much different from the implementers. They were rougher types who did not suffer fools and were inclined to upset apple carts or a least give them a shove.

The trouble is that an organisation that is fragile at the top will tend to opt for the first kind of strategist who will tend to perpetuate the fragility. Horses for courses and as a wise man said 'Strategy is nothing without tactics'.

[Churm Rincewind]

Well,er,yes. But haven't we known this ever since Bismark's remark over a hundred years ago that politics is the art of the possible?

[Edis Bevan]

The work of Geoffrey Vickers on what he called 'Appreciative Systems' springs to mind...

[Guano]

Also Appreciative Inquiry.

A appreciative systems' approach starts from the perspective that the cooperation of the various stakeholders would be useful. Most change management starts from the opposite perspective; that the cooperation of some of the stakeholders is neither useful nor desirable, because change requires running roughshod over the interests of those stakeholders. The image of a leader (in this approach) is someone who is willing to do that (like Blair annoying his own party and key constituency).

[Tim Blackwell]

It's been suggested that the universal credit regulations were written with automation in mind: they would be fed into a rule based system which would output the appropriate user dialogs etc. This was not necessarily a terrible idea, but it may be indicative of a faith based approach.

Hypothetical systems are almost always more attractive than actually existing systems, on the one hand because actually existing systems are those which have survived their encounters with reality, on the other because unattractive hypothetical systems are generally cheap to discard (except, sadly, politically). It's a shame our absurd tax credits system was not discarded before implementation.

[Jim M.]

"Make sure you get real-time, high-quality data" says the Cabinet Office Implementation Unit."

And from the ONS, with regard to the effect of Drugs and Prostitution on the UK economy...

"Prostitution: Finally, extensive data gaps have been filled with assumptions, recognising that this area of the economy is very difficult to measure."

Hmm....

[rogerh]

With respect I don't think the discussion so far has really addressed the implementation problem. Take a concrete example - passports. So far as I can see there was a system that was known to be vulnerable to loading pressures and that system was changed (probably to make cheaper) and has now shown loading pressure problems. Assuming that is what happened why were measures not taken to ensure the well known loading problems did not recur. Perhaps the strategy said 'make it cheaper' so someone did - regardless of consequences. Perhaps there was a disconnect between those cutting cost and those who could see the likely result. Perhaps the already poor staff relations were aggravated by the changes. Perhaps 'security' had grown from a sacred cow into a sacred elephant that no-one dare shoot. My tentative diagnosis is the top is disconnected from the bottom and problems do not have serious consequences so careless or non-caring (but cheap) implementation is the likely result. Unfortunately this disconnect is a design feature of government work.