This year’s comment on the Government 2018 Budget is in three parts. Today, I comment on the proposed tax cuts, which are the signature feature of this year’s Budget. I conclude that there are real doubts about whether those tax cuts are sustainable in the longer-run. The second part of this series discusses why tax relief focussed on low-income earners is desirable and then considers the change in policies required to make that tax relief sustainable. The final Part 3 of the series will consider the government’s overall priorities as revealed in the Budget.
Part 1: The Economic Impact of the Tax Cuts and their Sustainability.
The Government’s proposed income tax cuts are the signature feature of this year’s Budget. They have been widely interpreted as a political response to help win the next election. However, Fairfax economist, Ross Gittins, thinks that “This Budget is too good to be true”[i]. I agree, and my contribution here is only to consider in somewhat more detail why both Ross and I have drawn the same conclusion that the Government’s numbers cannot be believed. Furthermore, I therefore consider that the Government’s proposed second round of tax cuts should not be legislated now because of the shaky nature of the Government’s projections, and if we are serious about Budget repair, it would be better to cover the cost of the first round of these tax cuts by additional revenue raising.
As the Government widely signalled in advance, this Budget’s critical feature is tax cuts. Prior to the release of the Budget, the Government was heavily criticised by most economic commentators for putting tax cuts ahead of returning the budget to a sustainable surplus of the order of 1 per cent of GDP; a target that the Government had promised to achieve “as quickly as possible”.
Now that we actually have the Budget numbers, it would appear at face value, that the Government has managed to pull off the double: tax cuts and a return to budget surplus. While the return to a surplus equivalent to 1 per cent of GDP is not projected to be achieved before 2026-27, which is longer than desirable, that is probably acceptable in all the circumstances.
It appears that the Government has been greatly assisted by a rebound in revenue. Since the previous 2017 Budget last year, revenue has been revised upwards because of what are called ‘parameter revisions’. These revisions can include the impact of a stronger than expected economy, but also are affected by forecasting errors in assessing the relation between revenue and economic activity. In 2017-18 it is estimated that revenue has increased by as much as $12.1bn. because of these parameter revisions, and in the following years by $9.9bn. (2018-19), $5.1bn. (2019-20) and $3.8bn. (2020-21). This additional revenue is more than sufficient to pay for the tax cuts which are estimated to cost $4.1 bn. in the first year (2019-20) and then $4.4 bn. and $4.5 bn in the two following years.
But how credible are the new forecasts, which only cover the next two years (2018-19 and 2019-20), and even more so, how credible are the projections for all the years beyond the forecasts? These projections are only based on a set of assumptions; assumptions that unfortunately are not adequately spelt out in the Budget documentation, and which I will try to recreate here as best I can.
My reasons for doubting the revenue projections are as follows.
First, although revenue has surged in the current financial year, and this surge is very likely to be sustained into 2018-19, the respected independent budget analysts, at Deloitte, have argued much of this high rate of increase is likely to represent a one-off addition. In particular, mining companies have used up the various tax write-offs that they had accumulated and are now starting to pay the normal rate of taxation again. That makes a considerable difference to the rate of increase in revenue in the year of the changeover, but not to succeeding years. However, the Treasury modelling, as the Deloitte analysis points out, “bakes in this good news into the medium-term outlook”, as the Treasury modelling assumes that this higher trend continues, whereas Deloitte argues that some of the present extra revenue represents a timing shift in favour of 2017-18 and 2018-19 and that revenue growth will slacken in the later years. Thus, if we compare the average annual rates of increase of nominal GDP and revenue for the three years after 2018-19, we find that the Treasury is projecting an increase of 5½ per cent for revenue compared to an increase of only 4½ per cent for nominal GDP. Frankly, an examination of this relationship between revenue and growth for all years since 1970, strongly suggests that this difference of a percentage point looks too high, and even more so if allowance is made for the revenue lost by the Government’s tax cuts.
Second, as has been widely recognised, the key reason for the slow economic growth over the last decade has been the slow rate of wage growth, which has translated into a slow increase in disposable household incomes. This in turn has resulted in a slow increase in consumption, notwithstanding that households are now consuming a higher proportion of their incomes, with the household saving rate falling from 8.3 per cent in 2009 to 4.6 per cent at the end of 2017. However, the Government is now predicting that wage growth will accelerate from an annual average rate of 2.1 per cent over the last four years to an average annual rate of nearly 3 per cent over the next four years out to 2021-22; which would represent pretty much a full return to historic rates of wage growth.
Clearly some of the recent slow wage growth in Australia has been a response to the cyclical downturn in the economy, and to that extent wage growth can be expected to pick-up as unemployment comes down. Indeed, the tenor of much official thinking seems to be that most of the last decade’s slow wage growth is cyclical, and a full return to past higher rates can be expected eventually. To my mind, however, this analysis is mistaken for two reasons:
- We now know that in most advanced economies wages have been lagging productivity growth since at least the beginning of the 1990s. Indeed, in some countries real wages have actually fallen, and in the US the real wages for a typical male employee in 2015 were no more than they were 45 years ago. To me this suggests a structural problem in most of the advanced economies. So, although Australian wages have matched productivity growth over the longer term (with some fluctuations around that trend), we cannot be sure that we have not been affected by the structural change in the labour market that seems to have affected the distribution between wages and profits in most other advanced economies; indeed there is disturbing evidence that in the last few years, Australian wage growth has started to lag productivity.
- The main problem of low wages, however, is not that they lag productivity growth, but that the dispersion of earnings between high and low wage-earners has become much more unequal in most advanced economies. This is also true of Australia, although Australian earnings inequality has not increased as much as in most other OECD countries. But the economic problem with increasing income inequality is that the low-income households typically have a higher propensity to consume than their high-income equivalents, and a shift in income against low-income households can then hold back consumption growth. As Stephen Bell and I show in our recent book, Fair Share[ii], this tendency for consumption demand to fall in response to increasing inequality was offset prior to the Global Financial Crisis by increasing consumer debt, but with consumer debt in Australia now running at 190 per cent of household disposable income this avenue of avoiding stagnation is much more constrained.
In short, I am therefore sceptical that earnings and consumption will increase as fast as the Government is projecting, and that in turn will impact on the projections for economic growth and government revenue as well.
Third, the Treasury projections that deliver a return to surplus assume that in the long run productivity will increase at a trend rate of 1.6 per cent annually, which contributes to an annual potential growth rate of 2¾ per cent for the economy as a whole. In addition, it is also assumed that over the next five years, the economy will grow faster than its trend rate as the present spare capacity is absorbed over the next five years. Again, however, these projections could well turn out to be over-optimistic. In particular, the average annual rate of productivity growth over the last decade was only 1.35 per cent, compared to the Treasury projection of 1.6 per cent. A scenario analysis by the Parliamentary Budget Office shows that if the average annual rate of productivity increase is a quarter of a percentage point lower than assumed in the Budget, potential GDP will then only grow at an average annual rate of 2½ per cent and not the 2¾ per cent projected in the Budget. This in turn impacts heavily on receipts; indeed, my estimate is that if nominal GDP grows by a quarter of a percentage point less than projected by the Government from 2020-21 onwards, then by 2026-27 the loss of projected tax receipts would be equivalent to 0.4 per cent of GDP, and instead of the budget surplus being equivalent to 1 per cent of GDP it would in that case be only 0.6 per cent of GDP. And, of course, this impact of lower GDP growth on tax receipts accumulates over time; so that by the mid-2030s the budget could be back in deficit, if nothing else changes.
Finally, there are also doubts about the projection of budget payments. I have no reason to question the estimates of expenditure for the forward estimates period, however, beyond the estimates for subsequent years are based on assumptions regarding the underlying growth of existing programs. Unfortunately details regarding these assumptions are not available from the Budget documentation, but from the limited information available it appears that expenditures are now projected by the Government to grow at a slower rate than GDP for ever more. On the other hand, in its 2015 Intergenerational Report – the Government’s last proper explanation of its long-term budget projections – the Government projected real spending under its ‘currently legislated scenario’ to increase at an average annual rate of 3.1 per cent, which is significantly faster than the likely rate of economic growth. Furthermore, the Budget documentation shows that since the 2014 Budget the policy decisions made by the Government affecting its expenditures have not made much difference to the total and consequently to their rate of growth. Another source of information about the longer-term outlook for government expenditures is the Parliamentary Budget Office[iii]. In a report released last November, the PBO’s ‘central scenario’ projects outlays to rise as a proportion of GDP after 2020-21, with the increase being about 0.3 per cent of GDP by 2026-27.
Frankly, if one considers the drivers of government expenditures, the big-ticket items are welfare, health education, and defence. All of these seem likely in the absence of policy changes to increase faster than GDP because:
- we have an ageing society that will impact on welfare and health expenditures,
- there is a high income elasticity for health and education, and their cost per service tends in all countries to rise faster than for the economy in general,
- Defence expenditures are tied to a constant proportion of GDP.
In sum, it seems most likely that the Government has underestimated the long-run underlying rate of growth of Budget outlays, probably by at least 0.3 per cent of GDP and very likely by more. Of course, this underestimation must carry through to the bottom line, and have the same impact on the budget balance.
So, putting these various qualifications of the Budget forecasts and projections together it must be extremely doubtful that the projected surplus will be as big as the Government has estimated. Furthermore, it is even more likely that this surplus cannot be sustained in the longer run, unless the Government’s policies are changed. I will address the necessary changes in policy to return the Budget to a sustainable surplus in Part 2 of this series to be posted tomorrow. What is immediately obvious, however, is that more policy changes will be required if taxation relief is to be provided to low income households, as would be desirable; probably by raising extra revenue by dropping the proposed company tax cut. And it would be irresponsible to legislate at this time for the government’s proposed second round of tax cuts that, anyway, are not planned to take effect until 2024, given the uncertainties about getting the budget back into surplus.
Michael Keating is a former Secretary of the Departments of Prime Minister and Cabinet and Finance. In this capacity he was closely involved in the preparation of many Budgets in the past.
[i] See Pearls and Irritations, 9 May 2018
[ii] Stephen Bell & Michael Keating, 2018, Fair Share: Competing Claims and Australia’s Economic Future, MUP.
[iii] Parliamentary Budget Office, 2017, 2017-18 Budget medium-term projections: economic scenario analysis, Report no. 05/2017.