The Treasury should plan more than four years ahead

When the federal budget is handed down on Tuesday much of the focus will be on the deficits or surpluses over the next four years. This so-called “forward estimate period” is supposed to be far enough into the future to be meaningful, but close enough to today that the predictions are accurate.

In reality the forward estimates are neither meaningful nor accurate.

A better approach would be to require Treasury to present an estimate of the “steady state” deficit or surplus, as a percentage of GDP, based on the policies in place and outlined in the budget.

It might seem paradoxical, but economists are actually much better at predicting the long-run steady state than the near future, which is subject to all sorts of random fluctuations. The steady state is precisely the time at which these short-run fluctuations have calmed down and washed out.

The big drivers of the size of this year’s budget deficit will be assumptions about variables like nominal GDP growth and the iron ore price. Nobody knows what these are going to be in the coming year with even the slightest bit of accuracy. If you doubt that, then just look at how accurate the budget assumptions have been compared to the actuals over the past, say, 20 years. It’s not a pretty picture.

Worse still, our collective obsession with the four-year forward estimates period means that governments have every incentive to bury the true costs of policies just outside that window. Both sides of politics are serial offenders in this regard.

Have you wondered why the Coalition’s enterprise tax plan was described as costing $50 billion over 10 years last year but now apparently costs $65 billion? It’s because we are now one year later into the phase-in (with 2026-27 replaced by 2027-28).

Playing us for fools

Now I’m on record as being strongly in favour of cutting the corporate tax rate for all companies – not just small businesses – to 25 per cent. In the face of aggressive international tax competition, a small(ish), capital-thirsty country like Australia would be crazy not to do so. Even President Obama wanted to cut the US company tax rate by 7 percentage points.

But I don’t like the glacial 10-year phase-in, and I don’t like politicians playing us for fools by pretending that something that will happen for sure in five years’ time doesn’t count because of some stupid budgeting convention.

The Gillard Labor government was just as guilty of these sorts of shenanigans with policies like university funding and the National Disability Insurance Scheme. Again, I am for these policies, but we should own up to what they cost and the trade-offs involved in implementing them.

Perhaps the canonical case of where the steady-state is instructive, and the short-run is meaningless, is the effect of immigration.

Immigration boosts total GDP in the short run because there are more folks in the economy. In steady state, basic economic theory tells us that there is no effect of immigration on GDP per capita, and the best empirical evidence bears this out. In fact, immigration tends to reduce per capita GDP in the short run because it lowers the amount of capital per worker.

The bizarre accounting of immigration

How does the federal budget treat immigration? It takes the bizarre stance of counting the short-run total (nominal) GDP benefits, but ignores the costs from additional health, welfare and education expenditures that surely come with more people in the country. Why? Because those expenditures are assumed to be fixed in the short run.

Even a Mob accountant would be reticent to try and pull a fast one like that.

In steady state, however, we know there is no per-capita revenue benefit from immigration unless the skill mix shifts to increase average human capital. On the expenditure side, these new immigrants could increase or decrease per capita health, welfare and education spending – it depends, again, on who they are. People coming from a war zone are going to be different than people who have not suffered in that way, for instance. Older immigrants will have a different profile than younger ones.

Again, I am generally in favour of immigration, but the federal budget should treat the costs and benefits of immigration symmetrically, and it should focus on the long-run not the short-run. Steady-state government accounting does just that.

The chief purpose of the federal budget is to quantify tax and spending policies and inform debate around them both in the parliament and the community. The four-year forward estimates fail to achieve that and encourage gaming and obfuscation.

A good budget process forces a focus on the underlying assumptions behind policies and their likely impact in the long run. The current process simply leads to heroic assumptions about growth rates and a focus on the short run; to gaming and guesswork.

Perhaps more than ever we need government to be transparent and focus on the long term. Steady-state government accounting would be a step in that direction.

This article was originally published by the Australian Financial Review and reproduced by UNSW.