Dec 14, 2010
Ho ho ho. How many civil servants does it take to put up a tree?
[précis: link is to an Independent article mocking Treasury civil servants for making a meal of implementing the Chancellor’s directive to save money on this year’s in-house Christmas tree]
Surely just a bit of Christmas fun? But here lie a few juicy issues: how government buys stuff sensibly, what transparent spending data might do for us, and indeed the whole way we use models of things that we know about to try and understand things that we don’t.
The story of the Treasury Tree is a perfect example of the pitfalls that arise in scaling up concepts from experiences in our own lives to those in complex organisations, and particularly, when we try to do this in the public sector.
Yes. Of course the outcome of this story is absurd. But the actions and thought processes along the way aren’t. They’re just designed to do rather bigger things than buy a Christmas tree.
Let’s travel way back in time. A discipline called Government Procurement doesn’t exist yet. There is a small tin of petty cash under the stairs in Her Majesty’s Treasury. The under-under-secretary grabs a few shillings and pops out to the market. Up goes the tree. All is well. Until the day when someone spots that he’s buying the tree for a little more than he could in the market a bit further down the road – oh, and the guy selling the trees is also his mother-in-law’s gardener. Wouldn’t look good you see, and of course if you roll up all the tree and plant buying together, there’d be a much better deal on offer. Look whiter than white? Save a bit of money? Who would argue against that?
So begins the slow, insidious march under the mighty arches of procurement – the need to ensure probity in all dealings, robustness of supply, and the quest for the best deal available. Probity by ensuring there’s an unbiased and open competition for the contract, robustness by weeding out those who can’t show a healthy trading record (by filling in a big long pre-qualification questionnaire, say), and the best deal – well, that’s just a matter of the lowest price, surely?
I exaggerate for effect, but rational judgement that makes quite a lot of sense if you’re buying fighter aircraft, or street light maintenance for ten years, just doesn’t stack up when trying to buy a single bloody Christmas tree. How the Permanent Secretary and others involved in this sorry story must have come to hate even the words describing our adopted Yuletide symbol.
I don’t know how seriously that big list of questions in the article were posed – who would water the tree, get the decorations, dispose of the damn thing, but again, they’re actually quite enlightened, in a twisted, distorted, terribly out-of-place way. I know the odds of the Exchequer Partnerships contractor slipping and breaking her neck while stringing up the lights are infinitesimally small, but does that mean the risk can be ignored? That a strong letter from the Prime Minister to the insurance company (who’d be within their rights not to pay up, I’m sure, if they really dug their heels in) would be enough to make the problem go away? You see – already we’re having to make up processes and assumptions to solve the problem, because we’re dealing within a system built, first and foremost, to do things bigger and cheaper. You can see why the Perm Sec ended up going up the ladder himself, can’t you?
All those niggling little implications, extra duties, details that hadn’t been thought through – they’re the very reasons they outsourced trivial shit like tree care and plant watering, remember? To make it tidy, predictable, managed. They created this rod with which to be flogged in the newpapers. It’s quite hard to design in an off-switch for times like this, believe me.
A few years ago there was a project in which several tens of millions were spent on equipment by the government, with a lamentable lack of just those niggly little questions about the details. It was just assumed, as you would for a Christmas tree, that if equipment was purchased and given to public sector employees, then all the other issues of maintenance, logistics, renewal, disposal and so on would just sort themselves out. They don’t. It took several years of work, the design of entirely new financial processes and the creation of a new organisation to make sure that happened. Not cheap. Not trivial.
I don’t for a minute dispute that the tree story looks bonkers – I just argue that it’s based on reasoning designed for greater and more complex affairs. And I invite you to consider this: if the tree is at the ‘petty purchase’ end of the scale, and should just be paid for from the tin under the stairs (and watered by the office junior before they leave) and fighter aircraft are at the other end where you need to ensure that your billions are spent without fear of corruption, failure or waste, then where’s your cut-off point? For there’ll always be one. And any boundary creates edge effects, gaming (fixing things to ensure the boundary is under- or over-achieved) and other delightful consequences.
So where is that sensible edge? The edge of reason, if you will. £100? £1,000? But over what period? And with what discretion? Authorised by whom? If you can solve this one, you’ve solved one of the great wicked problems that hampers public sector innovation and agility. Web development in particular suffers here: we want simple commodities and flexible skills (at a cost much closer in magnitude to that of a Christmas tree than a Rural Payments system); we want to specify loosely and build quickly; we want a fast-track to innovation, unsaddled by the implications for scaling to a national level on existing infrastructure or bypassing heavyweight integration testing (for good or ill…). And we can’t, generally, have them. If we can, it’s through fudging; finding short-term-contract loopholes or underspent budgets in other areas.
We perform the equivalent of dipping into a tin under the stairs, in some cases to achieve some highly significant, public-facing outcomes. And as financial matters get more transparent, we’re going to get bitten pretty hard on the arse by the details that happen around that breakpoint between tree and fighter. Any playing around with payment mechanisms, timing or contract packages will be picked up on. Quite rightly? Perhaps. But the wicked problem isn’t going to get any simpler, that’s for sure. And with an overwhelming media interest in quirkiness and fudge, as we see in the woeful tale of the Chancellor’s Norway spruce, we can be sure of reading more headlines about how much the government spends on cheese, and perhaps rather fewer on how much is wasted on massive transactional processing engines.
There is a solution of course. It’s called the sorting hat. It’s the same one that’s being put to work when we read stories like “Ministers believe as many as one in four people claiming sickness benefit…” It’s the same one that people use to judge, in the comfort of their warm homes, how obese someone should be before they start to drop off NHS waiting lists. It’s the test for what sort of public architecture is the right sort, and whether or not we should widen the motorways. And sorting hats are, of course, completely fictional.
Our comfort in this type of judgement is, naturally, our own set of models. Where Christmas trees are put up by dad (and thrown away by mum). Where civil servants occupy their day moving memos from one tray to another. Where working class folk go out to work all day, and if they can’t, scrub steps or sit quietly in smoky clubs, respecting each other and getting up to no bother. Where people don’t fall off ladders on polished floors, because that’s just daft. And where any rule can be set to one side if the story’s silly enough. According to the papers, anyway.
Good luck finding those boundaries. But do remember they have to be there, and there’s usually a good reason why.
Update: picture below of the problematic pine, taken 15 December (with thanks to Dan Harrison).