Monday, May 21, 2018

Let's outlaw she'll-be-right budget projections

The practice of including in the budget 10-year “medium-term” projections of the budget balance and net debt is pernicious. It should be abandoned in the interests of responsible economic management.

It’s supposed to increase transparency and accountability, but in practice does more harm than good, presenting the government of the day with an almost irresistible temptation to portray the future as more assured than it is.

The future is unknowable. We can’t forecast the economy even a year ahead with any accuracy, but what we can be most sure of is that, even with pure motivations, a mechanical projection out 10 years is highly likely to be way off-beam.

We know the economy never moves in straight lines, but each year’s 10-year projection shows it on glide path to where we want to be. Obviously, no account is taken of unexpected shocks to the economy – even though it’s a safe bet there’ll be more than a few in the space of a decade.

Treasurers' unworthy intention is to leave non-economists with the impression everything’s under control and on the improve. But I think it’s likely to leave even our economists and econocrats with a false sense of comfort. If you doubt that, you haven’t read Daniel Kahneman’s Thinking Fast and Slow.

I remember how proud the Hawke government’s hard-man finance minister, Peter Walsh, was after persuading the cabinet to include in the budget papers not just the figures for the budget year, but also for the following three years of “forward estimates”.

This would improve transparency and accountability, making it harder for governments to hide the budgetary consequences of their decisions in later years.

But when Labor’s Wayne Swan came under pressure to get the budget deficit down in the early years of this decade – struggling with the big-spending proclivities of his successive prime ministers – he soon realised the way to make the deficit look like it was headed steadily in the right direction was to “re-profile” big spending commitments into more convenient years.

In particular, he was always using his “fiscal bulldozer” to push spending commitments beyond the three-year forward estimates, where they couldn’t be seen.

As commentators started drawing attention to this trick it became clear he’d have to bolster the budget’s credibility by providing some sort of answer to the question of what would happen beyond the forward estimates.

Thus the greater weight put on medium-term projections. Thus has the budget’s purview been inflated from one year to 10 – all with without succeeding in curbing treasurers’ temptation to mislead. From wherever he’s watching on, I doubt the late hard-man of Finance is cheering.

Sad to say, the medium-term projection has been about deception – both numerical and visual – from the off. In all the budgets since Swan introduced them, never once has the budget balance failed to glide smoothly up to a healthy surplus, nor the net debt failed to glide smoothly down towards zero.

How’s it done? By making assumptions, of course. Assumptions are the unavoidable basis for all “projections”. But the proof that the budget’s projections have always been more for support than illumination is that never have the assumptions been fully and clearly spelt out.

In this budget, what little explanation there is has been sprinkled through three different chapters (“statements”) of budget paper No.1. Buried in statement three is the warning that “projections of the receipts impact over the medium term are subject to higher levels of uncertainty and are sensitive to changes in economic conditions, underlying assumptions and forecasts”.

And this year Treasury seems to have slipped into statement two the additional warning that assuming the spare capacity in the economy is absorbed over five years from the first year of the projections is “a well-established approach but it is not without drawbacks”.

The key assumptions are: the rate at which government spending will grow – which will be based on any new (financial) year’s resolution the government has made to be frugal in future – and the economy’s medium-term “potential” rate of growth, when we’ll get back to it and how quickly we’ll use up the (estimated) spare capacity once we have.

This year’s fine print acknowledges that the assumed potential growth rate of 2.75 per cent a year is based partly on the assumption that labour productivity grows each year at its 30-year average rate of 1.6 per cent. But former top econocrat Dr Mike Keating notes that the average over the past decade is only 1.35 per cent – which makes a big difference.

Even without the ever-present temptation to fudge, projections are a device for deluding ourselves we know more about the future than we do. By ignoring all the uncertainties, they breed not understanding, but complacency. An honest government would abandon the practice.