MICHAEL KEATING. How Useful and Reliable are the Budget Projections?

May 29, 2018

The Treasurer wants us to believe that the Government has a credible plan to restore the Budget to a surplus and cut taxes at the same time. This conclusion is based on the projections for revenue, expenditures and the budget balance incorporated in this year’s budget. However, both Ross Gittins and I have separately criticised the credibility of these budget projections. I also agree with Gittins’ conclusion that the ‘Treasurer’s unworthy intention is to leave non-economists with the impression that everything is under control and on the improve’, and even economists and econocrats can be left ‘with a false sense of comfort’ (Sydney Morning Herald, 21 May). In this article I now want to go further than Gittins and show why the Treasurer’s claim that the Government has a credible plan to restore the Budget to a surplus, while also cutting taxes, is demonstrably false.  

My starting point is that, of course, a budget surplus requires revenue to be greater than expenditure. That equally means that if we start from a deficit position, then a surplus can only be achieved if revenue increases faster than expenditure. In the current budget taxation receipts are projected to rise to 23.9 per cent of nominal GDP by 2021-22, and after that the Government has committed to cap taxation receipts at that percentage for ever after. Thus from 2021-22 onwards the rate of increase in taxation revenue projected by the Government must be the same or less than the rate of increase in nominal GDP.

The Government is projecting a budget surplus equivalent to 0.8 per cent of GDP in 2021-22. I am sceptical about that forecast[i], but in any event, if that surplus is going to increase after 2021-22 to achieve the Government’s claimed surplus of 1 per cent of GDP in 2026-27 then payments will need to rise more slowly than GDP. Indeed, hanging on to any budget surplus will not be possible even if payments rise only modestly faster than GDP. Accordingly, to test the validity of the Treasurer’s claims we need to consider the likely future trajectories of both GDP and government payments.

Unfortunately, the 2018 Budget provides almost no information regarding the assumptions underpinning its projections for GDP and payments; which of itself doesn’t increase confidence in these projections. This has led me to provide my own projections for future expenditure and GDP/revenue growth over the next decade. The detailed assumptions underpinning these two projections are provided in an appendix to this article, but in short, my projections show that:

  1. nominal GDP will increase at an average annual rate of 4.6 per cent between 2019-20 and 2027-28; and
  2. government expenditure will increase at an average annual rate of 5 per cent over the whole period from 2016-17 to 2027-28, and at an average annual rate of 5.5 per cent for the later period from 2020-21 to 2027-28.

The important point to note about these two projections is that the rate of increase in GDP is projected to be almost 1 percentage point lower than the projected growth in government outlays. That means that if these projections are a reasonable representation of the future, then there is no way that the budget will return to a sustained surplus, without either a tax increase or a substantial reduction in government expenditure.

But how plausible are my ten-year budget projections and why should anyone believe my projections and not the Treasurer’s? First, the size of the difference between the projected rates of increase in expenditures and receipts is so large, that I think we can conclude that any sensitivity analysis based on plausible variations to my projections’ key assumptions would not alter my conclusion. For example, while I prefer my original projections, even if we increase the projected rate of increase in prices and productivity in my projection by a ¼ of a percentage point each over the period 2019-20 to 2027-28, the average annual rate of increase in GDP (and therefore revenues) would still be around ½ a percentage point less than the increase in expenditures.

Second, my projection of government expenditures is based on the same projections by the independent Parliamentary Budget Office, and the underlying drivers of this expenditure are clearly shown by the PBO. Furthermore, similar growth in government expenditures was projected in the then existing policy scenario shown in the Government’s 2015 Intergenerational Report. However, in the present Budget the ten-year expenditure growth must be much lower, although unfortunately we are not shown exactly how fast the Government projects its expenditure to rise after the forward estimates period ending in 2021-22, nor why these latest expenditure projections are apparently much lower than previously.

Instead, what we know about government expenditure is that in advanced economies, including Australia, this expenditure tends to rise faster than nominal GDP, because:

  • as incomes rise the demand for many of these services (most notably health and education) usually rises faster than the demand for goods; and
  • prices for these services also tend to rise relatively, because their provision offers fewer opportunities for productivity improvement than goods manufacture.

In addition, as Stephen Bell and I argue in our recent book, Fair Share[ii], the increase in inequality has been the key reason why most countries have experienced economic stagnation over an extended period, and in future there will be pressure to increase income support and other forms of assistance to people disadvantaged by structural adjustment, rather than reduce the expenditure on welfare.

Furthermore, everyone who has any experience of government expenditure control, knows that it is very difficult to reduce the growth of public expenditure to less than GDP unless major policy changes are introduced to the main areas of government expenditure, such as health, education and welfare, infrastructure spending, and defence. However, ever since the failure of its 2014 Budget, the Government has not shown any interest in seriously trying to change policies so as to achieve a significant reduction in expenditures in these major spending areas. While the Government has continued cutting the budgets for relatively small programs, such as culture and foreign aid, this nickel and diming approach cannot seriously be expected to return the budget to surplus in the foreseeable future.

Thus for these various reasons I think we can be reasonably confident that my ten-year projections are a fair representation of the challenges the Budget faces in future. In particular, these projections refute the Government’s claim that we can have both the Government’s proposed tax cuts and achieve a budget surplus as well. While on the other hand, the Government’s unwillingness to provide any details to support their projections is suspicious in itself. Accordingly, I think we can be reasonably confident that people like the Australia Institute are right when they contend that economic growth and therefore revenue are unlikely to be sufficient over the next decade for Australia to be able to ‘guarantee the essential services that Australians rely on’ without some increase in government revenue relative to GDP. To my mind this is a very important conclusion, which should guide future policy development.

And to those who still reman unconvinced of this conclusion, I think even they would have to admit that there is sufficient uncertainty about the Government’s projected budget surplus, that it would be both unnecessary and irresponsible to legislate now for tax cuts that won’t take effect for more than two years.

Michael Keating is a former Head of the Departments of Finance and Prime Minister and Cabinet. In this role he was heavily involved in the preparation of Budgets for twenty years, including three years while he was Head of Finance, when real government outlays actually fell – the only time that has ever occurred.


Appendix: Assumptions underlying the alternative ‘Keating’ Budget projections

Of course, all ten-year projections of the budget and the economy are necessarily based on assumptions after the first couple of years, which typically is the length of the forecasts based on modelling relationships and recent information. But the critical problem with this year’s ten-year Budget projections is that the necessary information regarding the underlying assumptions is for the most part missing.

In a previous article on the Budget (Pearls & Irritations, 18 May), I argued that:

  1. The recent surge in government revenue this year cannot be expected to continue at the same rate when projecting the increase in government revenue over the longer term.
  2. The Government projections for wage increases were too high, given the history of lower rates of wage increase in all the advanced countries over the last couple of decades, and recent experience in Australia[iii].
  3. Productivity growth was likely to be less than the annual growth rate of 1.6 per cent projected by the Government.
  4. While no details were given in the Budget documentation, the implicit rate of increase in government expenditures assumed by the Government must have been unrealistically low to achieve the surplus projected.

In the light of these criticisms I have now done some more research and produced alternative projections for the future growth of Australian Government expenditures, revenue and the budget balance over the next ten years, based on the following methodology and assumptions.

First, the latest 2018 Budget estimates show that taxation receipts rise to 23.9 per cent of GDP by 2021-22, and the Government has committed to cap revenue at that percentage forever after. Although in my view there is nothing sacrosanct about this essentially arbitrary revenue cap, for my alternative projection I have also adopted this cap. That effectively means that the level of revenue in the later years is determined by the level of GDP. Thus what matters for future revenue in the later years is the rate of growth in nominal GDP.

Again the Government provides almost no detail regarding its projected GDP growth rate. It seems that the Government has assumed that the average annual rate of growth of potential productivity is 1.6 per cent. I have previously argued that this assumption is too high (Pearls & Irritations, 18 May). Instead, in my projection for the years after 2019-20 I have assumed an average annual rate of productivity increase of 1.3 per cent, which is about the average over the last ten years. Nothing is known about what the Government assumed for the increase in prices after 2020-21, but the Budget forecasts for the GDP deflator show an average annual rate of increase of 1¾ per cent between 2016-17 and 2019-20. For the following years out to 2027-28 I have assumed an average annual rate of increase of 2 per cent for the GDP deflator. After combining my projections for productivity, price inflation, and the Government’s employment projections I find that the nominal GDP is then projected to increase at an average annual rate of 4.6 per cent between 2019-20 and 2027-28.

While some might feel that this average annual growth rate of 4.6 per cent for future GDP growth is on the low side, I think it is in line with what might be called the ‘new normal’ as defined by the rate of growth over the last decade in Australia and other advanced economies which we normally compare ourselves with. As shown in the Table below, Australian economic growth has been faster than in all other major advanced economies, but what I am projecting for Australian real growth over the next decade is in line with the experience of the last decade. Furthermore, while real economic growth is currently picking up in most countries, that is from a very low base, and there are considerable concerns about the fragility of this recovery. For example, the OECD in its latest update (March 2018) stated that ‘Medium-term growth prospects remain much weaker than prior to the financial crisis’. Similarly, the rate of inflation has been subdued – close to zero for many countries over the last decade, and even now prices for the US and the EU are only forecast to increase by only 1 per cent in the next year. Accordingly, I suggest that my projection for an average annual increase in the Australian GDP deflator of 2 per cent between 2019-20 and 2027-28 is reasonable.

Real GDP average annual growth rates

Country 2008-2016 2017
Australia 2.6 2.3
Canada 1.5 3.0
France 0.6 2.0
Germany 1.0 2.5
Italy -0.8 1.5
Japan 0.6 1.5
UK 1.2 1.7
US 1.5 2.3

Source: OECD Data base.

Second, we know that the rate of increase in government payments is largely independent of the rate of increase in economic activity. Unfortunately, the Treasury is again silent on its expenditure projections, however the Parliamentary Budget Office (PBO) did produce ten-year expenditure projections last July based on the 2017 Budget. Since then, an examination of the Reconciliation Tables in the recent Budget documentation suggests that revisions to expenditure estimates because of policy changes and changes in Budget and program parameters are not sufficiently material to make much difference to the rates of expenditure growth shown in that previous 2017 Budget. Accordingly, I have used the forecasts of expenditure shown in the most recent Budget for the first few years, and the PBO expenditure growth rates (published last July) for the out-years after 2020-21. This methodology shows an average annual rate of expenditure growth of 5 per cent for the whole period from 2016-17 to 2027-28, and an average annual rate of expenditure growth of 5.5 per cent for the later period from 2020-21 to 2027-28.


[i] See ‘MICHAEL KEATING.2018 Budget Comment (Part 1), Pearls & Irritations, 11 May, 2018.

[ii] Stephen Bell & Michael Keating, 2018, Fair Share: Competing Claims and Australia’s Economic Future, Melbourne University Press.

[iii] Subsequent to the appearance of my previous 11 May article (link above), the Deputy Governor of the Reserve Bank has also raised this possibility that the long-run relationship between the increase in wages and the state of the labour market has changed.

Share and Enjoy !


Receive articles straight to your Inbox

How often?

Thank you for subscribing!