The Greens’ proposal that election campaign promises be costed has drawn the praise it is due: it is an excellent proposal.

But it is a harder job than it might seem.

Even during the normal course of events outside of election periods, adequate costings or cost-benefit assessment of policy seems more the exception than the rule.

If the government is to build civil service capacity to cost all political party promises during the election period, why not use that opportunity to build an agency able to review policy over the course of the government’s term?

Let’s start with the difficult part on election policy costings.

Parties wanting costings on their policies can get them through the Treasury – although that potential is far more notional than real.

The State Services Commission’s Guidance for the 2014 election included guidelines for costing party political policies; the Treasury also provides a high-level guidance document.

Going through minister

But requests for costings must go through the minister of finance. This is no fault of current or previous governments – the Treasury is responsible to that minister.

A party outside of government requesting a policy costing could well worry the minister might request sensitive details. So it seems unlikely opposition parties would seek such costings under the existing guidelines.

Were an independent agency of Parliament to undertake the costings, what the agency might lose in expertise would be made up for in credibility: asking an agency responsible to a minister to undertake work that would be kept confidential from that minister does not work well. The Greens’ proposal consequently could be better housed elsewhere.

Suppose that agency provided costings on a voluntary basis. If everything were working properly, parties wanting policy costings from that kind of independent agency would request them early in the election period.

Time to do the research

The agency, and the ministries with which it might need to coordinate, would then have sufficient time to undertake the work. The parties could still time their policy releases throughout the election campaign but they would have had to have developed them early enough to have them costed.

Policies that parties would prefer not be costed would not be costed, while those that are costed have the cooperation of the party requesting it. Policies would be costed when parties expect they would fare well under evaluation.

That changes when an agency is charged with costing all policies.

Imagine you were a political consultant who came up with some policy that you, and your party, reckoned would be great politics, even if it were terrible economics.

Currently, you would never seek a sound costing on that policy – and you would hope the Taxpayers’ Union would not hire someone to run an eye over it.

Were all election promises costed by an independent agency, you would be well advised to delay submitting your policy for costing so that only a cursory take on it could be provided.

You would also be tempted to wrap the policy in enough complexity that any costing provided in a hurry would have to make simplifying assumptions. Or, conversely, you could avoid providing any of the details necessary for a costing, saying the finer points would be worked out while in office.

You could then attack any assumptions the costing unit might be forced to make as being Treasury ideological “burps” – as former Finance Minister Michael Cullen once dismissed sound Treasury advice.

An agency doing this job would need to have a fair bit of in-house expertise in costing. It would also need to have a strong reputation for down-the-line analysis. And, for it to do any good, there would have to be underlying voter demand for sound policy costings.

That last requirement is a bit of a worry. You might think that, if voter demand for policy costings were strong, parties would already be seeking independent and credible costings of their main policy planks.

Switching the default might make a difference. Currently, policies are only costed where parties want them costed. In the alternative, policies would only fail to be costed if the party had something to hide, or if the party came up with its policy proposals late in the game.

Voters who came to expect costed policies might draw the proper inference from ones that were not able to be costed.

And so while I really rather like the idea of all policies’ being costed, it will not be a small job where parties might want to throw sand into the works – for some policies.

But the attention to the costs of policy is strongly welcome.

The cost of existing policies, unfortunately, draw rather less attention. Even in the absence of independent costings, political parties have reasonably strong incentive to draw attention to the costs of their opponents’ platforms.

New policies proposed after the election at least draw cursory assessment as part of standard regulatory impact analysis – even if Sapere’s analysis last year found that 20% of regulatory impact statements completely failed to meet quality assurance guidelines and only 30% fully met them. Strengthening independent evaluation there might not go amiss.

Usual spending untested

But the bulk of the government’s spending – the big line items that do not change much from budget to budget – receive rather less attention.

Take the government’s student loan programme. Labour’s zero-percent student loan policy drew attention during the 2005 election campaign and again in 2008. But we do not often hear discussion of the $5 billion gap between the nominal value of the loan and their carrying value.

The vast majority of government spending simply carries on, from year to year, without much notice.

Perhaps an agency established to cost new political party proposals could, between elections, spend its time running rolling reviews of existing spending programmes, ensuring they continue to deliver value for money.