Consumer stimulus vs Income shock - who wins?

Record declines in employment and hours worked have been met with record policy stimulus. In this note, I delve into the scale of the income shock facing the consumer and match it against the scale and timing of the fiscal and monetary stimulus. Who wins? Who loses? And what else is needed?

An income shock without parallel

What does it require to be counted as employed in Australia? It’s a pretty low bar. If you happen to be one of the 26,000 people in the labour force survey conducted each month, to be counted as employed you need to have worked one hour or more for pay in a business, or for free in family/farm business. If you are on leave or on strike it counts, and you are also counted even if you didn’t work that week but are self-employed.

When thinking about the scale of the employment losses and the rise in the unemployment rate that COVID-19 will impose on the Australian economy, it is useful to reflect briefly on the structure of the labour market. The diagram below provides the basic structure. Of the 20.8 million people over the age of 15 and not in the defence force, two-thirds are in the labour force – comprising 13.0 million employed and 720k unemployed.

Of the one-third not in the labour force, 80% either don’t or can’t work as they are retired, in education or providing caring duties. But 20%, or 1.4 million people, are not in the labour force who claim they would like a job – double the actual number unemployed! Not all are discouraged workers who have given up officially looking, in fact only 90k identify as ‘discouraged’ workers.

If you were wondering why the RBA prefers to talk about hours worked rather than the unemployment rate, the first reason is because a near doubling of the JobSeeker Payment (formally the Newstart Allowance) for six months has the potential to draw a large number of the 1.4 million people who indicate they would like to work back into the workforce by formally registering as unemployed. If Australia ends up with far higher unemployment rates than current government projections, it may perversely be attributed to the size of the JobSeeker package.

What else stands out from this diagram? Of the ranks of the employed, part-time workers make up one-third, and casual workers (a sub-set of part-time workers who typically don’t have access to superannuation entitlements) represent 22% of total employment.

Interestingly, only 10% of part-time workers claim they couldn’t find full-time work and a further 15% said it’s a requirement of the job. For the rest, part-time work fits their studying, lifestyle or caring duties. That is, three-quarters of part-time workers actively choose this form of employment.

The above charts clearly illustrate which industries are the large employers. Health care is now the dominant employer in the country, aided by the National Disability Scheme and the aging population, employing 14% of all workers. Historically, health care, education, utilities and public administration were thought to be dominated by the public sector and thus more insulated from the economic cycle. However, those days are long gone: public sector employment currently makes up just 15% of total employment, with 80% of the utilities sector privately employed, 75% of the health care sector and half the education sector. Even 17% of public sector administration is provided by private sector workers.

 However, it is still the large employing sectors within the private sector that are most exposed by the current crisis. Retail (1.3mn employees), construction (1.2mn employees), professional services (1.2mn employees) and accommodation/food (935k employees) are directly impacted by the COVID-19 restrictions.

Moreover, as the charts above highlight, accommodation/food and retail have very high part-time employment shares, currently 61% and 49% respectively. In contrast, construction and professional services are dominated by full-time workers, with just 15% and 22% part-time employment shares.

Indeed, it may surprise to learn that Australia has the third highest share of part-time workers in the developed world. A rising female participation rate, the shift towards a user-pays university model, and policies designed at retaining older workers in the workforce have all contributed. However, it is also true that emerging from every economic slowdown, the share of part-time workers has increased, resulting in lower income certainty, job security and pension entitlements. This is likely to again be the case as we emerge from the current crisis as firms seek to utilise a cheaper and more flexible workforce. In the interim, it is the part-time and casual employees in service-based industries that are most vulnerable to job losses.

 

This has been reflected in the ABS’s preliminary labour markets statistics, utilising new ATO data. The release reveals the enormous damage done to employment and wages from the restrictions in just 3 weeks to the 4th of April. An update of this will be provided on May 5. Employment declined by a record 6% in the 3-week period, with a staggering 26% decline in employment in the accommodation/food and a 19% decline in arts/recreation industries. This data also reveals the impact on wages is slightly larger, having fallen 7%, and that the burden of employment and wage declines is falling heavily on those aged below 29 years where employment has declined 9%.

Although there is uncertainty on how these figures will translate to the official labour force survey, we estimate the unemployment rate is already at 10.4% at the end of the first week of April and has likely moved higher over the balance of April.

 

As we look forward, our forecasts suggest that the unemployment rate will peak around 12% in mid-2020, before declining relatively slowly through 2021 with the unemployment rate likely to remain above 8% at the end of 2021.

With such an employment decline in prospect, we are focussed on the combination of five key factors:

• The split between employment losses that are full-time vs part time;
• The proportion of workers by age and by average income that have lost employment;
• The aggregate decline in hours worked, mindful that hours worked will differ greatly by industry;
• The response of those outside of the workforce that may be encouraged to register for the JobSeeker payment and its impact on the unemployment rate; and
• The take-up rate of the JobKeeper program and the proportion of employers who top-up the payment.

“We estimate the COVID-19 hit to labour income in 2Q2020 to be -17% (q/q) – 8x the next biggest quarterly decline in the post-war period. By comparison, it took 3 years in the Depression for average earnings to decline 20%.”

Our forecasts for compensation of employees assumes the job losses remain predominately in lower income part-time roles in the private sector industries most heavily impacted by the restrictions. Nevertheless, we estimate a 9% decline in average hours worked per employee and an 8% decline in employment in 2Q 2020, suggesting compensation of employees (i.e. excluding any fiscal offsets) will decline 17%qoq. This is 7.7-times the next largest quarterly decline, recorded at the height of the 1983 recession. There simply is not a comparable analogue to compare the current income shock to in the post-Depression period. Indeed, even during the Depression it took three years from 1930-33 for average employee earnings to decline 20%. 

During COVID-19 the decline to employee earnings is of a similar scale, but occurring in just three months. That is not a typo.

A fiscal package without peer

Set against a labour income shock without a parallel in terms of its size and velocity, has been a government fiscal package without peer in Australian history and second only to the US in terms of the size of the fiscal package as a share of the economy.

The bulk of the Australian response is via five key initiatives.

The largest is the $130bn JobKeeper package, designed to pay a salary of $1,500 per employee per fortnight. To be eligible, firms with revenue greater than $1bn need to report a 50% decline in revenue and for smaller firms a 30% decline. The $130bn was estimated on the basis that 6 million employees would be captured under the scheme. Basic arithmetic suggests that even if all 6 million employees gained access to the payment for the six-month life of the program, that $117bn not $130bn would be distributed. Moreover, the government indicates that as of the 28^th of April, 540k businesses had registered for the scheme (and a smaller number have actually completed the paperwork) which covers approximately 3.3 million employees. Given this is 55% of the projected 6 million employees, this represents a saving of ~$9bn out of the $130bn Budget in the month of April.

"Australia’s fiscal package is the 2nd largest global response to COVID-19, providing a 10.5% of GDP stimulus. We believe several of the stimulus measures will not be fully allocated in the alloted time…"

Will the Federal government reach its 6 million employee coverage under JobKeeper? We continue to believe it is unlikely. We suggested over a month ago that: (i) the combination of high revenue hurdles for businesses to gain access to the scheme, and (ii) the ineligibility of casual employees that have been with the same firm for less than 12 months (i.e. the most exposed employees as detailed above) made it unlikely the full $130bn would be deployed in the six-month time-frame.

Moreover, our modelling of the peak infection rate informed our view that the easing of restrictions would occur by mid-June and, therefore, some recovery in employment through 3Q2020 would make it unlikely that the initial pace of take-up of the JobKeeper package would be sustained beyond the initial months of the program. At best, we estimate 70% of the program will be utilised over the six-month period. The remainder will likely be utilised to over the following 12 months, either by extending the scheme or repurposing the funds to new programs.

In terms of the remaining programs:

• The “Cash payments to business” initiative has so far distributed only $3bn of the $32bn in the program, or 14%, to approximately 177k firms
  

• “Raising income support payments to individuals” (the JobSeeker payments) have been more effectively distributed, with $9bn of the $14bn already allocated for the next six months and the Prime Minister yesterday noted that 800k applications for the payment have now been made. At $1,100 per fortnight, it implies that soon 81% of the $14bn will have been allocated.

• The “One-off income support payment” is two payments of $750, the first paid between 12 March to 13 April and the second paid from 13 July. Given $5.1bn of the $8.8bn package has already been distributed to 6.8 million people, this indicates that the Government has under-budgeted for the second payment by ~$1.4bn.

• “Early access to Super” has a small Budgetary cosy of $1.1bn, as the scheme facilitates the income boost to come via voluntary withdrawals from the superannuation balances of those impacted by the crisis, with withdrawals of up to $10k in 2019-20 and a further $10k in 2020-21. The Treasury had estimated approximately $27bn would be withdrawn under the initiative. If each individual who applied chose to withdraw the full $20k then Treasury was assuming 1.35m individuals would avail themselves of the policy. Currently 757k individuals have been granted access to their Super withdrawing $6.3bn, with the average withdrawal being $8.3k. Although it is early days for this scheme, it appears that less than 50% of Treasury’s estimate for fund withdrawals has been realised. While we continue to assume that the full $27bn will be withdrawn, the risks are clearly to the downside as both the average withdrawal and number of applicants prove to be lower than Treasury’s estimate.

"Even if the full package is not delivered on schedule, in comparison to any other fiscal transfer to the household sector the COVID-19 response is by far the largest ever delivered by an Australian government."

Even if the full fiscal package does not get deployed within the set time frames, it is important to realise that this fiscal package supporting household income is by far the largest of any period in Australia’s history and 3-times the payments as a share of GDP delivered during the Great Depression. 

A surprising net result for household income 

In this titanic battle between huge negative shocks to labour income and generous fiscal payments, thinking through the relative strength of these opposing forces in the aggregate and over time is not without its challenges.

However, our analysis of each of the fiscal packages, inclusive of voluntary withdrawals from superannuation, when netted against the size of the income shock yields a surprising and somewhat pleasing result.

For the remainder of 2020 the fiscal stimulus more than offsets the hit to labour income. This is an important result. Labour income is expected to decline $13bn in May, $13bn in June and $9bn in July. While these are large numbers, they are swamped by the fiscal transfers to the household sector totalling $14.5bn in May, $17.5bn in June and $22.6bn in July. July will mark the peak stimulus month, with labour income better off by an estimated $15bn in that month. Labour income remains in net surplus until November before slipping into net deficit in December and through 2021.

Although consumption will still decline sharply in 2Q2020 due to restrictions on mobility and industry shutdowns, the hit to labour income is more than offset by the combination of peak fiscal stimulus arriving in July simultaneous with an ongoing easing of restrictions. This will be a powerful force for a consumer recovery in 2H2020, and the Government should be congratulated for the design and implementation of the policy.

Note we are assuming that 30% of the $130bn JobKeeper package is used to smooth the ‘fiscal fade’ that would be associated with the conclusion of the fiscal packages in 3Q2020. If the full package was implemented within the time allocated, then the net labour income benefit will be larger in the near-term and the fiscal fade larger through 2021.

There are additional benefits to household income via RBA rate cuts and the sharp decline in income and capital gains tax paid during sharp economic downturns. Both of these factors also need to be allowed for in order to gain the broader picture for household disposable income – the more relevant income measure for consumption growth.

Our expectations for both are shown in the figures below, and although interest payments and taxes paid are not immaterial in any discussion of household income, they pale in comparison to the fiscal transfers.

 

Will it be enough to sustain a consumer recovery through 2021 and beyond?

 The problem with large fiscal stimulus packages is that eventually they lapse and fiscal fade creates a headwind for future economic growth. The task for policy makers is to provide sufficient stimulus to sustain a recovery without leaving public finances in a parlous state.

Without any further fiscal stimulus, the profile for annual household disposable income growth for Australia is shown below. As noted above, despite a sharp hit to income in March and April, the fiscal stimulus through May to November helps sustain annual disposable income growth through the balance of 2020.

If we are correct, this will be an extraordinary outcome in the face of the scale of the economic challenge. 

However, when looking out to 2021 household income growth suffers from the combination of three factors. Firstly, the lapsing of the fiscal programs and lack of further monetary easing. Secondly, the tapering of the employment and hours worked gains from the relaxation of COVID-19 restrictions. And, lastly, the reality of a labour market with ~460k more people unemployed at the end of 2021 relative to pre-crisis.

Moreover, income growth is not consumption growth. In normal economic times trends in household saving matter little. Typically, the household saving ratio follows the broad trend in household wealth, with households happy to consume more out of current income when asset prices are rising. During periods of employment uncertainty and asset price declines, however, the saving rate tends to rise sharply and remain high for several quarters or indeed years.

Our forecasts assume financial assets eek out modest gains over the period to end-2021 and house prices decline 12% in 2020 before rising 6% in 2021. If those assumptions are broadly correct, then net wealth to income will follow the path shown in the chart above. As such, it is reasonable to assume the household saving ratio rises from 3.5% currently to 6.5% through 2021.

The combination of these factors presents the challenge for the sustaining of a consumer recovery beyond the 2H2020 fiscal boost. Even under a more optimistic asset price assumption, it remains unlikely that income growth will be sufficient in the absence of further government assistance. Nevertheless, the Government does have options. The more prominent of these are the extension of some of the existing fiscal packages, the implementation of new one-off measures or the bringing forward of packages that are already ‘paid’ for in the Budget.

The biggest card up the Government’s sleeve is to pull forward income tax cuts scheduled for the 2022-23 and 2024/25 years. A sceptic might note that the tax cuts are aligned to the likely election years, however, given the importance of ensuring a durable recovery that rationale is now redundant. The collective focus is on building a credible path to recovery.

For the relatively modest cost of approximately $5-6bn, the first phase of the income cuts could be pulled forward to commence on 1 January 2021. Compared to the economic benefit of avoiding a double dip in consumer demand and persistently high unemployment rate, this would be a well worth the investment.