On Sunday, the Prime Minister discussed the forthcoming overhaul of the benefits system with Andrew Marr.  A big-deal change challenge.  The discussion was reported on the Guardian's website and the piece drew all manner of comment: some of it amusing and some of it useful (but hardly any of it both). I found myself somewhat agreeing with one of the commenters who, rather rudely in my view, doubted the achievability of Mr Cameron's ambitions. Wanting to be more constructive, I found myself mentally spooling through a list of advice that I would wish to share with Mr Cameron were he perched on my sofa with Sunday cake and tea. Dammit I wrote it down and stuck it there amongst the other 300 comments. I read it again on Monday night and, despite it only receiving 6 Recommends (the Guardian equivalent of Likes), I was pleased.  The piece wasn't intended to be exhaustive or even useful: it was just what I felt at the time.  Copied here...

In a statement about The driving purpose of the Coalition Government on 2 August 2010, Mr Cameron said,

"this government, unlike previous governments, will govern for the long term".

If he really means this, here's what he should do about welfare reform (and probably pretty much every other reform too)...

Acknowledge that in complex and dynamic economies like ours, solutions to the really big challenges can only be grown and not manufactured.

• Accept that genuinely radical welfare reform will take many, many years to achieve and forget the idea of doing something politically big in the short term.
• Completely forget any idea of using 'Big IT' to drive the change - this would cost loads and get nowhere.
• Get going - right now - on some small, localised experiments to model and learn about the long-term reform he wants to grow. Let the learning from these experiments inform more experiments. And so on.
• Put genuine collaboration with the citizen (supported by small and smart IT) at the heart of these experiments.
• Build a team of passionate, smart and emotionally intelligent people to promote these changes. If necessary ignore his recent public sector salary hair-shirtery (many big firms are in the market for these sort of folks again).
• Do not, under any circumstances, allow the intellectual centre of gravity of these experiments to reside in a consultancy.
• And last: let the answer come - don't force it and a solution with the right support, policies, laws, technology will emerge.