Thursday, 10 October 2013

Outcomes, McKinsey and King Canute

Some years ago (2005, in fact), I was asked by the Environment Agency to review their new indicators, as we used to put it in those days.  They all seemed very sensible but one, in particular stood out.  It was the indicator which specified the number of centimetres the sea levels have risen.

There was a true outcome measure, at last.  I knew I would recognise one if I ever saw one.  Yet it was also a measure they had almost no power to affect.

It was as if King Canute had set it as a priority indicator for his own government (hence his picture here).

Most management consultants still talk about outputs and outcomes as if they were absolute categories, as if an outcome was a concrete thing beyond argument.

Nothing could be further from the case. What is the outcome of the NHS for example? The number of patients successfully treated? Or is it the health of the population? Because those are diametrically different ideas and require different measurements and probably different institutions altogether.

Outcome measurements assume that our institutions should be permanent. They are about organisational control. They don't let us imagine whether we might be better off with different institutions instead. There is a whole continuum of potential outcomes, none of which are definitive - except possibly sea level rises, but even then I have my doubts.

All these measures, outputs and outcomes alike, are susceptible to the little boy's question in the Emperor's New Clothes.  Yes, the sea levels are still low, but is the environment safe?  Yes, the school league tables are rising, but is their education any good?

At one end of the outcomes continuum, they are little more than the simple purpose of existing institutions. At the other end, they are usually outside the control of institutions anyway.

I know this is shocking to some people.  What, you don't believe in outcomes?  The jaw falls open.  But, no I don't.  The whole idea is a wrong-headed attempt to rescue clapped out targets from the wastepaper basket.

I've subjected you to this small rant because I have been sent an article by the management consultants McKinsey, who are after all the consultancy most associated with this nonsense.  And, staggeringly, it is still pedalling this old stuff.  It is called Delivery 2.0.

I long to take them by the scruff of the neck and point to our public services, hollowed out by the business of assigning measures to a handful of 'deliverables', and setting up elite delivery units to browbeat officials - and gargling with measures which, by their very nature, are one-dimensional, bowlderising, distorting and rendering ineffective.

Yet here is the McKinsey Museum Piece:

"It’s a cliché, but it’s true: what gets measured gets managed. Performance improves when it is managed. Internal performance management should begin by assigning accountability for outcomes to individuals. Once accountability is established, performance dialogues—regular conversations about each goal—are essential. One prime minister reviews the progress of six priorities every week; every six months, he holds a face-to-face performance dialogue with each minister.  These conversations must be based on standardized, clear management data (ideally available online) that can be reviewed and managed in real time. And the dialogues must be reinforced by rigorous evaluation and consequences (good and bad). Many governments are constrained in this regard; they may not be able to reward great performances with bonuses or condemn bad ones by firing the perpetrators. But they can publicly acknowledge outstanding people, promote highfliers faster, and move laggards to lower-profile roles."

What I find most enraging is that McKinsey seem unaware of Goodhart's Law, which explains why this approach has failed so miserably: a measure used to control will always be inaccurate.  However useless frontline staff and their managers may be, they will always know how to finesse the data.

In fact, this kind of approach makes finessing data their absolute priority, and far more important than delivering a good service.

McKinsey are desperately behind the times here.  The governments and organisations that can embrace complex interconnected objectives, by operating as close as possible to the frontline, are going to inherit the world.  Those which try to subject them to the McKinsey approach are going to slap themselves on the back and wonder vaguely why nothing works. 

1 comment:

ChrisJCook said...

Well, there is one outcome which is completely objective,and which the Danes use as the basis of policy.

That is 'least energy cost'.

This policy is implicit rather than explicit, which might frighten the horses.

After the Oil Shock of 1973 they simply and pragmatically legislated policies mandating the least carbon fuel input to the Danish energy supply chain for a given output of heat, electricity, and power.

The outcome has been that since 1980 Denmark's GDP has risen (in DK terms) by 78% while energy use has been stable, carbon fuel use has significantly declined,and a country of 6m souls has created the largest wind turbine company in the world, and an unmatched body of expertise in heat engineering.