There are a lot of promising initiatives in the coalition Government's first "Wellbeing" Budget. But delivering on that promise requires closing a currently open loop in the wellbeing policy process. We need evaluation to help us tell which of the Government's initiatives have proven most successful, and Budget 2019 delivered far less than we might have hoped for on that front.  

Finance Minister Grant Robertson told assembled media and economic analysts at this year's Budget lockup that 2019 represents a first foray into a new approach to budgeting.

Figure 1 in Budget 2019, one of the first slides in Robertson's presentation, illustrated the wellbeing policy loop. The policy process begins with an evidence-based evaluation of the areas in need of attention. It works its way through to evaluation of the outcomes of the programmes that follow from that evaluation. And that evaluation informs the next budget rounds.

It's a beautiful vision.

But it is not a vision unique to the current Budget. It is integral to any decent policy process. Good government requires ensuring that every budgeted dollar delivers real and valued outcomes. If it does not, those dollars should flow instead to other promising initiatives. There is no shortage of them.

And neither was the vision here newly announced; it has rather been a mainstay of this Government's communications around the wellbeing agenda. We were to expect an evidence-based approach leading to evaluated outcomes and enhanced wellbeing.

Very nice.

The Government announced $1.9 billion for initiatives targeting improved mental health. It was no surprise that mental health would receive additional attention in this Budget. But despite 73 different references to "mental health" in Vote Health – the Budget allocation document for health – there was no obvious funding for measures to evaluate how well any of the different initiatives achieve their objectives.

If overall statistics on mental health improve, which programmes should take the credit? If they do not, which programmes are most at fault? Surely some will prove more promising than others.

The Government has put $455 million towards improving access and choice among mental health and addiction services providers. Figuring out what works and what doesn't work allows closing the policy loop – scaling down programmes that are less effective and using the funding to boost more effective alternatives.

Initiatives targeting family violence at least came with a $5m increase in funding for evaluation. But quite how that evaluation will be undertaken remains a bit up in the air.

Rigorous evaluation work relies heavily on Statistics New Zealand's integrated data Infrastructure. Statistics NZ has quietly developed the world's best linking-up of administrative records held by various government agencies and ministries to allow rather better ways of figuring out what works.

But that infrastructure is already bursting at the seams, with increasingly heavy demands on it from academic researchers, think tanks, and government agencies. The performance indicators noted in Vote Statistics expect the agency to deliver a 50 per cent increase in the number of microdata users in the next year – but a decline in funding.

Treasury often undertakes this kind of programme evaluation work, or at least assists in overseeing it. But Treasury's core economic and analytical capabilities have been seriously eroded under its secretary, Gabriel Makhlouf.

Treasury was awarded $5m a year over the next four years to "deliver core functions and the wellbeing approach". Rebuilding Treasury to be able to deliver the evaluation capabilities necessary for achieving Robertson's vision may be a bigger job than that.

It would have perhaps been the wrong day to ask the minister of finance whether he has any confidence in Treasury's capability to deliver. I have doubts.

Programme evaluation has always been the poor cousin to programme announcements. When voters too easily infer a government's depth of caring by the quantity of its spending, rather than the outcomes that spending achieves, that outcome is not surprising.

I hope the wellbeing approach yet proves to be something more than that. It may otherwise be difficult to distinguish from business as usual.