Last week, Thomas Coughlin reported that “the wellbeing framework that puts the ‘value of a statistical life’ at $4.7 million is coming under fire.”

There is a lot to criticise about the wellbeing framework, and I am hardly one of its cheerleaders. But it is absurd to criticise it for trying to apply proper cost-benefit assessment – and even more absurd to criticise it for putting a value on statistical lives.

Every policy choice the Government makes that affects safety, health, and risk implicitly puts a value on lives. This is not a new feature of the living standards framework. The Ministry of Transport developed its value of a statistical life figure back in 1991 and has used it ever since in helping to decide how much it should spend on road safety measures. If straightening a curve in the road costs $1 million dollars and is expected to save a life, then the investment is worthwhile – but it would not be if the project cost $20 million.

Having the figure helps the Government to decide how much it should spend on projects that save lives as compared to projects that improve them. Government only has a limited budget – even if tax rates increased substantially. Spending everything on safety improvements might make sense if we thought that the value of a statistical life really were limitless, but it would not leave much money for everything else the Government does.

Without a hard figure on things like this, it is easy to descend into policy absurdities. Whether you think that the current valuation on a statistical life is too high or too low – and I think it should be increased – at least having a single figure helps set a constant benchmark. If Government spent $20 million per statistical life saved with some interventions, but refrained from making investments that could save lives at a tenth of the cost, Government would be achieving far less than it could be. Pharmac drug-funding decisions would be worse without a razor-sharp eye on whether the benefits of some new wonder drug treating a media-friendly illness were really worth the cost.

But that does require discipline. The previous National government implemented scaffolding regulations that, according to NZIER work commissioned by BRANZ, imposed about $1.40 in cost for every dollar’s worth of safety and other benefits. Government could have done far more good at less cost by pursuing other alternatives. When politics means that a government is steadfastly determined to ignore decent cost-benefit assessment, it can cost lives. And it can make some kinds of measurement a bit risky.

And that brings us back to the Government’s living standards framework for wellbeing.

Amy Adams criticised Treasury and the Finance Minister for absurdities in the values placed on different outcomes under the framework. The fiscal cost to the Government of avoiding a case of diabetes was quoted at $3900; the benefit to the public of improved contact with your neighbours is valued at $8500, and gaining a friend is worth $592.

The relative figures do seem absurd – although Treasury Secretary Makhlouf was right to note that the personal costs of diabetes would have to be added to the Government’s fiscal costs to make those kinds of comparisons.

But even more absurd is that Government policy should aim at helping people gain friends or interact more with the neighbours. The absurdity comes from trying to put values on outcomes that are really none of the Government’s business. If the Government offered me a case of diabetes, or more chances to interact with one of my neighbours, well, I’d have to think about it – your mileage may vary.

Unfortunately, not having values on those outcomes is also risky.

Imagine being the poor Treasury official tasked with telling the Minister that his preferred policy does not stack up under any reasonable cost-benefit assessment. You lay out all the facts and figures, showing that the policy is a colossal waste of limited funds, and that the money could be put to far better use elsewhere – only to have your analysis overturned when a Ministry official points out that the policy would improve social capital, or help neighbours make friends with each other, and that the value of those uncounted benefits should tip the scales.

It can then be understandable why Treasury would want to have standardised figures – and hope that there’s at least some rigour in estimating the effects of policy on such nebulous things as friend-making.

But we might also worry that, where Government is not always all that committed to rigour in cost-benefit assessment, even having these sorts of numbers invites creative accounting.

It is not hard to imagine the Provincial Growth Unit explaining to Minister Jones that he could justify giving $100,000 to his favourite small-town lawn bowls club if it encouraged even as few as two hundred new friendships. Increasing club membership from 35 to 41 could be enough to do it, under certain assumptions. It could then far too easily encourage a sham version of the living standards framework, rationalising political decisions rather than ensuring that every dollar of fiscal or regulatory cost does as much good as possible.

And that is where the opposition really must hold government to account. The problem is less whether having an additional friend is worth one seventeenth as much as having additional contact with your neighbours in Treasury’s CBAx assessment. The problem rather is whether the government bothers putting its bigger ticket items through the framework at all.

It would be interesting to know just how the Government manages to justify the Provincial Growth Fund, tax breaks for pretty race horses, or the Taranaki oil ban within any kind of rigorous assessment in the wellbeing framework.

Running those numbers might cost Treasury some friends – but I bet the CBAx tool would show it’s worth that cost.