Policy and the evolution of uncertainty

Dr Steven Kennedy PSM
Secretary to the Treasury
Australian Business Economists


I would like to acknowledge the Ngunnawal people who are the Traditional Custodians of the Land where I am today. I would like to pay my respects to their Elders – past, present and emerging – and extend my respect to any Indigenous Australians who are with us today.

I would like to thank the Australian Business Economists for inviting me to speak today to provide a post-Budget briefing.

I have been the Treasury Secretary for a year now, with the October Budget my first as Secretary.

In the past, the Treasury Secretary often gave a post-budget speech to this forum and I am keen to reinstate the tradition.

You have all had a month to analyse the Budget framing and measures, so I am not going to talk about individual measures in detail.

What I want to do is reflect on the experience of Treasury in designing economic and fiscal policy during a period of extreme uncertainty, caused by a ‘once in a hundred year’ pandemic.

I also want to talk to you about Australia’s underlying macroeconomic context and productivity growth in a post-COVID world, both of which present unique challenges.

The initial COVID-19 shock

Assessing the COVID-19 shock and responding has been a challenging exercise for Treasury. More challenging than any episode I can remember.

In January and February, when the virus first emerged overseas, we were mostly worried about a trade shock, with no international students returning from China and fewer tourists coming to Australia.

Things moved quickly in March. Once it was clear that the virus was present and spreading here in Australia, governments put in place restrictions to limit its spread.

However, even ahead of the formal restrictions there was already a rapidly deteriorating willingness for people to move around freely as they sought to minimise the risk of being infected.

Uncertainty became the predominant issue.

We were unsure how quickly the virus would be contained.

We were unsure how firms and households would respond.

We were unsure how the financial system would respond – but the early signs from some parts of the markets were not reassuring.

In this environment Government policy needed to resolve as much of this uncertainty as possible to retain confidence.

Treasury developed forecasts based on which sectors were being shut down and used our macroeconometric model to assess the spill over effects. We expected there to be large forecasting errors, but falls in GDP of around 20 per cent were being seriously contemplated.

Fortunately, and unlike previous shocks, we had access to a number of real-time data sources to track the effects of the shock.

The Weekly Payroll Jobs and Wages series developed by the ABS using Single Touch Payroll Data enabled us to track the labour market in close to real time. We also used data on unemployment benefit claims from the Department of Social Services, data on spending from the banks, customs data from the Department of Home Affairs, data on insolvencies from ASIC and novel data such as from supermarkets, Google, Apple and Seek.

These information sources proved to be invaluable, and we have drawn on them heavily as we have adjusted our forecasts.

And during that initial period, we were updating forecasts multiple times a month.

The initial onset of the virus was unique in economic terms, in that it had elements of both a supply and demand shock.

The most direct and significant impacts were on the supply side, with many businesses restricted from operating and supply chains disrupted.

Because businesses could no longer operate and households were unable to work and earn an income, this had the potential to lead to a swift contraction in demand and cascading levels of business failures, unemployment and other financial difficulties.

The JobKeeper Payment and the Coronavirus Supplement were designed and introduced to replace the lost incomes for businesses and households who were substantially and directly affected. Importantly, JobKeeper was specifically designed to keep a connection between employees and employers.

They helped to resolve the financial uncertainty faced by the private sector and transferred it to the public balance sheet where it could be more easily absorbed.

Both policies were deliberately designed to be need or demand based because the size of the shock to the economy was difficult to predict.

The size and speed of the initial shock necessitated a large fiscal response. Conventional monetary policy support – even if available – would not have been sufficiently targeted in the way that fiscal policy was in March. Monetary policy cannot replace incomes or tie workers to jobs.

Nonetheless the RBA also played an important role in reducing uncertainty. The Bank’s policy response – extending liquidity operations followed by a package of measures including bond purchases and a term funding facility – ensured there was ample liquidity in the banking system and that financial markets continued to function effectively.

Together with actions by the financial regulators, including APRA’s regulatory capital concessions, these policies ensured that the financial system could act as a shock absorber by extending repayment deferrals to firms and households.

Collectively, these macroeconomic policy actions, in conjunction with other Government support measures and the broader success of Australia’s health response in containing the virus, were successful in putting a floor under the economy.

While gross national income fell in nominal terms by 7.0 per cent in the June quarter 2020, household disposable income rose by 2.2 per cent.

Treasury analysis of Australian Taxation Office (ATO) Single Touch Payroll data and the Household, Income and Labour Dynamics in Australia Survey indicates that higher-income households on average saw close to no change in disposable income between the March and June quarters, while the average income of the lowest income households increased by around 20 per cent.

After the announcement of JobKeeper, business confidence increased by 20 points in April and by another 25 points in May. Consumer confidence also jumped significantly.

And Australia’s GDP and unemployment outcomes have fared better than most and better than we initially feared.

The economic recovery

I want to turn to the economic recovery and the design of the fiscal measures included in the Budget.

We know much more about the virus and how the economy is travelling than we did in March.

We have also seen significant early steps towards recovery.

Between May and September, around half of the employment lost in the initial months of the pandemic was recovered – a faster rebound than many would have predicted.

Although female employment fell more sharply than male employment between March and May, it has also recovered more quickly.

Treasury analysis of ATO Single Touch Payroll data indicate that around 40 per cent of the employment recovery between mid-April and mid-June was people returning to their old jobs, a higher proportion than usual.1

In sectors such as Food and Accommodation and Arts and Recreation that were particularly affected by government restrictions, the percentage of the recovery in employment that was people returning to their old jobs was around 70 to 80 per cent.

That said, as the recovery continues we are still dealing with more uncertainty than normal.

We do not know precisely when we will get a vaccine, how effective that vaccine will be and how much faster the recovery will be once a significant proportion of the population is vaccinated.

We also don’t know how treatments will improve over time, how the virus will evolve, or whether improvements in testing and tracing regimes will simplify the management of the virus to allow the easing of remaining restrictions. 

Until the risks of the virus subside significantly, there is always a risk of further outbreaks.

There is also a great deal of uncertainty about the global recovery.

A large number of countries did not manage to fully control the virus in its initial stages and are now largely forced to live with significant community transmission. This means that there is significant scope for second and third waves, as we are seeing particularly in Europe.

Slower global growth will drag on Australia’s recovery.

Despite this uncertainty we expect that over time, as restrictions are progressively removed in Australia and assuming any outbreaks are quickly controlled, the COVID-19 shock will become more like a traditional demand shock, with a significant output gap to close.

In usual circumstances, monetary policy would have provided significant support to boost the recovery and close the output gap.

In response to the 1990s recession, the RBA cut the cash rate target by over 1000 basis points, and by over 400 basis points in response to the GFC.

In March, the RBA quickly lowered the cash rate by 50 basis points to 25 basis points, and announced the initial unconventional policy package.

On Tuesday, they lowered the cash rate to 10 basis points and announced a further program of bond purchases.

While the RBA’s actions will provide important additional support, the scope for these policies to provide sufficient stimulus is limited and has necessitated the large levels of fiscal support both at the onset of the pandemic and in the Budget.

Since the onset of the pandemic the Commonwealth Government has provided $257 billion in direct economic support or 13 per cent of 2019-20 GDP.

In comparison, in response to the GFC the Commonwealth Government provided $72 billion in economic stimulus or 6 per cent of 2008-09 GDP.

Given the lack of monetary support available, the states and territories can also play an important role in supporting the economic recovery, especially in the areas where they have primary responsibility and are better placed to design and implement policy, including social and other infrastructure.

The design of the fiscal policy package in the Budget took into account two sources of uncertainty.

First, the ongoing uncertainty surrounding the health recovery.

Second, uncertainty around the effectiveness of individual policy measures at stimulating demand and reducing unemployment in the current environment.

The Budget deliberately used a wide range of levers and incentives to support aggregate demand, providing incentives for all actors in the economy to contribute to the recovery.

Measures in the Budget include:

  • Support to lower income households through lower taxes and once-off payments.
  • Incentives to encourage businesses to bring forward investment and hiring.
  • Direct government support for infrastructure, construction and manufacturing.
  • Direct government support for essential services, particularly health, aged care and disability services.

At the same time, the Government augmented the automatic stabilisers and allowed them to work.

Additional welfare payments and lower tax receipts were significant elements of the $281 billion estimated deterioration in the underlying cash balance since MYEFO on account of parameter and other variations.

Given the lack of conventional monetary support available, the recovery could falter without a strong fiscal policy response leading to years of anaemic growth. More and more unemployment would become entrenched reducing the productive capacity of the economy. Lower growth also means that inflation and wages would likely remain lower for longer.

An IMF working paper by Blanchard, Cerutti and Summers from 2015 looked at 122 recessions over the past 50 years in 23 countries and found that 69 per cent of recessions were followed by a sustained period of output below the pre-recession trend.2  

Higher levels of unemployment, lower levels of output and lower prices would also have significant long-term implications for government revenue, spending and debt.

This means that the real driver of increased debt is the shock itself, not the discretionary fiscal policy response.

A broader justification for discretionary fiscal policy support is that there are significant welfare costs associated with unemployment and the loss of incomes that can persist for years.

In a slowdown we typically see the NAIRU drift up, but it can be pushed back down with sustained economic growth.

Fiscal policy can help limit these welfare losses by generating economic growth, and then pay for it in a way that spreads the burden equitably over time and across society.

This ties directly into the Government’s fiscal strategy.

The first phase of the strategy focuses on supporting the economy and promoting economic growth until the recovery is assured. This phase of the strategy will remain in place until the unemployment rate is comfortably below 6 per cent.

In the longer-term, the strategy will transition to being focused on stabilising and reducing debt as a share of the economy. This will ensure that we will rebuild our fiscal buffers and are well-placed to respond to future economic shocks.

The mix of fiscal and monetary policy

The lack of conventional monetary space available right now is not solely due to the COVID-19 shock.

The neutral or natural interest rate (the real cash rate where monetary policy is neither expansionary nor contractionary) has been steadily falling globally over the past 40 years.

Lower neutral interest rates mean there is less space for monetary policy to operate before hitting the effective lower bound.

The declining neutral rate is due to structural developments that drive up savings relative to the willingness of households and firms to borrow and invest.

While the academic research is not settled on the relative importance of different structural drivers, it is likely due to some combination of population ageing, the productivity slowdown and lower preferences for risk among investors.

A number of central banks have suggested that interest rates will not rise for many years.

Even when cash rates return to currently estimated neutral rates, monetary policy will have a reduced capacity to respond to negative shocks. If the neutral rate of interest falls further there will be even less space.

This raises fundamental issues for the operation and coordination of macroeconomic policy including the appropriate role of fiscal policy as a cyclical stabilisation tool.  

Fiscal policy has always responded to large shocks but there is now a question about whether the threshold for intervention should be lowered.

Monetary policy setting through central banks has traditionally been the preferred macroeconomic stabilisation tool with fiscal policy focused on structural and sustainability issues. This has been the accepted wisdom in advanced economies for the past 30 years.

This delineation was preferred because central banks can make timely decisions using an appropriately nimble instrument which helped to manage demand but was considered to have a broadly neutral impact on resource allocation in the long-run.

Central banks also benefit from high levels of external credibility, based on independence, transparency and accountability against a clear mandate.

There have been reviews by central banks thinking within the current framework but less has been done on the broader framework.  

Making timely decisions requires information and evidence.

Treasury is already forecasting more frequently than it ever has and we have incorporated the use of real-time data during this crisis.

Treasury is actively considering how our processes can be adjusted to make more up-to-date assessments of the appropriate stance of fiscal policy.

However, fiscal policy tools are more complex than the cash rate.

There is effectively an infinite suite of different fiscal policy tools.

Fiscal policy instruments for stabilisation ideally could be implemented very quickly in response to a shock – however, a complex policy development process could create additional lags.

Semi-automatic stabilisers are one option proposed by Claudia Sahm, Olivier Blanchard and Larry Summers (among others) that seeks to address these issues.

In many ways, the design of the Government’s initial response to COVID-19 was to put in place temporary automatic stabilisers.

Any move towards more active fiscal policy needs to be pinned to credible long run anchors.

In the case of Australia, this is achieved through a continued commitment to sound public finances, underpinned by the tax to GDP cap, balanced budgets and stabilising and reducing debt in the longer-term.

Productivity growth is still key

It is important to remember that macroeconomic stabilisation policy can only do so much. The living standards of Australians will ultimately depend on the drivers of long-run economic growth – population, participation and productivity growth.

The Intergenerational Report (IGR) will be released next year and it will touch on each of these drivers but I want to briefly focus on productivity growth.

Prior to COVID-19, Australia – like many advanced countries – had modest productivity growth.

Australia’s multifactor productivity growth over the five years prior to COVID – between 2013-14 and 2018-19 – averaged 0.7 per cent.3 In comparison, multifactor productivity growth between 1994-95 and 2004-05 averaged 1.4 per cent.

There are many suggested causes for this.

Treasury research has highlighted reduced dynamism, including low levels of firm entry; less job switching; slower adoption of frontier technologies and processes; and less labour reallocation from low to high productivity firms.4

This decreased dynamism may have also made scarring effects from downturns more persistent, as it takes workers longer to climb back up the job ladder.5

Forthcoming Treasury research highlights declining competition as a cause of part of the decline in dynamism. Estimates suggest that reduced competition can account for at least one-fifth of the slowdown in labour productivity growth in the private non-financial sector in Australia that occurred between the 2000s and the mid 2010s.

But it is fair to say that many of these factors have been found in other economies, so it appears to be a global phenomenon.

It is not clear how the COVID-19 shock will affect Australia’s long-run productivity growth.

Clearly COVID-19 has the potential to usher in some large structural changes.

There has been some forced innovation – necessity is a great ramrod for breaking down the barriers to technological adoption that perhaps were holding us back from fully leveraging new technologies.

Remote work is one obvious example.

ABS Survey Data from September indicates that 31 per cent of Australians with a job reported working from home most days in the prior four weeks, compared with around 12 per cent of people before March.6

More people working from home has the potential to improve welfare for many Australians, but it could also have significant implications for transport infrastructure planning and for the functioning of CBDs.

Treasury analysis of CBA bank transaction data indicates that total spending in CBD postcodes over the 6 months to September was 34 per cent lower than the comparable period in 2019, but only 5 per cent lower in non-CBD postcodes.

We have also seen many businesses being forced to quickly pivot to be able to operate and trade online.

ABS Survey Data from September found that 36 per cent of businesses had changed the way products or services are provided to customers.7

The ability of many businesses to pivot in response to COVID-19 indicates the potential for innovation and adaptions that can see dynamism return to support productivity growth.

On the other hand, the COVID-19 shock has the potential to drag on productivity growth.

For example, there is a risk that firm closures among smaller firms will lead to even more market concentration.

ABS payroll data have shown larger employment losses for smaller employers due to the COVID-19 shock.8

Preliminary Treasury analysis of similar data has shown that employment within industries has tended to become more concentrated among the largest firms.

Workers may become more risk averse and less willing to switch into jobs where they are potentially more usefully employed, further weighing on dynamism. Businesses may also be less willing to expand and invest.

The Government’s JobMaker plan is designed to increase productivity and promote long-term growth, with a focus on measures that will promote a flexible and dynamic economy.

It included a range of regulatory reforms designed to encourage digital adoption, reduce the cost of doing business, and encourage the reallocation of capital and labour to growing firms and sectors.

For example, insolvency reforms will assist viable small businesses to restructure and where restructure is not possible, it will help businesses to wind up faster, promoting a more efficient allocation of resources.

Final remarks

All budgets involve a degree of uncertainty. However, this is usually contained.

We are normally debating small amounts of macroeconomic variation from the trend and small changes to tax and spending. For example, will growth will be 2 or 2 ½ per cent?

This Budget has been unusual in that there is so much uncertainty about the future of the economy – both domestic and international.

But fiscal policy has adapted and been successful in supporting the economy.

The recovery will take time and there are new challenges for macro policy, but Australians can be confident that while bumpy, there is a clear path to recovery.

One final comment, COVID-19 reminds us of the importance of quality governance and institutions in responding to larger shocks.

Evidence-driven decision making supported by high levels of public sector capacity has seen Australia through this crisis better than most and will continue to do so.

1 While ATO Single Touch Payroll data is not available before 2020, analysis of ABS Labour Force Survey microdata indicates that the proportion of unemployed individuals returning to their old job is significantly higher than at any time since 2014, when data is first available.

2 Blanchard, O., Cerutti, E., and Summers, L. (2015). ‘Inflation and Activity - Two Explorations and their Monetary Policy Implications’, IMF Working Papers, vol. 15(230).

3 Australian Bureau of Statistics. (2020 October 30). Australian System of National Accounts. Australian Bureau of Statistics, Australian Government. Market sector, hours worked basis.

4 See for example: Andrews, D. and Hansell, D. (2019), ‘Productivity-enhancing labour reallocation in Australia’, Treasury working paper, 2019-06.
Deutscher, N. (2019), ‘Job-to-job transition and the wages of Australian workers’, Treasury working paper, 2019-07.
Quinn, M. (2019), ‘Keeping pace with technological change: the role of capabilities and dynamism’, Speech to OECD Global Forum on Productivity, Sydney 20 June 2019.

5 Andrews, D., Deutscher, N., Hambur, J., and Hansell, D., (2020). ‘The career effects of labour market conditions at entry’. Treasury working paper, 2020-01, June 2020

6 Australian Bureau of Statistics. (2020 October 13). Household Impacts of COVID-19 Survey. Australian Bureau of Statistics, Australian Government.

7 Australian Bureau of Statistics. (2020 September 24). Business Indicators, Business Impacts of COVID-19. Australian Bureau of Statistics, Australian Government.

8 Australian Bureau of Statistics. (2020 October 20). Weekly Payroll Jobs and Wages in Australia. Australian Bureau of Statistics, Australian Government.