2.1
Over the course of the inquiry, the committee heard a number of views on the prevalence and effects of insecure work—including opposing views.
2.2
Inquiry participants representing industry argued in favour of the 'flexibility' offered by casual, contract, labour hire and on‑demand employment models, and offered evidence to suggest that work has not become less secure, or more precarious, over time.
2.3
Conversely, unions, labour academics and participants from the social sector argued that insecure and precarious forms of work are damaging to individuals and society, and have in fact become more prevalent in Australia, accelerated by the COVID-19 pandemic.
2.4
The committee sought evidence from a variety of sources to understand the true nature and prevalence of insecure work, and its impacts on our health and wellbeing, including our mental health.
2.5
In this chapter, the committee goes beyond the widely-quoted headline Labour Force data, which can sometimes present an incomplete picture when it comes to insecure work. Using data and analysis prepared especially for this inquiry, resources provided by the Australian Bureau of Statistics (ABS), and research provided by the Parliamentary Library, the committee looks at the increase in four categories of work that are indicators of increasingly precarious and insecure work:
indirect employment, outsourcing and labour hire;
on-demand and "independent" contractor jobs; and
fixed term contract roles.
2.6
The committee finds that insecure and precarious forms of work have increased, and are increasing. The committee's conclusions are based on a rigorous analysis of the best available evidence.
2.7
The next chapter—Chapter 3—will build on this by looking at the impacts of insecure work on physical health and wellbeing, mental health, relationships and family life. Later, Part 3 of this report presents policy options aimed at reversing the trend towards precarious and insecure work.
Insecure and precarious work are increasing
2.8
The committee heard conflicting evidence on whether work has become less secure since the mid-1990s. The mid-90s are considered a benchmark period for comparison because there was a clear and dramatic increase in the proportion of total employees who were casuals over the 1980s and early 1990s—from around 13 per cent in the early 1980s, to 24 per cent in 1996. Then from 1996 to 2019, the proportion of the workforce employed as casuals remained fairly consistent, hovering around 24 to 25 per cent.
2.9
Over the course of 2020, measures associated with COVID-19 dramatically impacted Australia's workforce. The proportion of casuals in the workforce fell from 25 per cent, to around 20 per cent in May 2020. This was because in April 2020, almost 350 000 people (mostly casuals) 'left or lost a job'. By May 2021, the number of casual employees had recovered to 2.6 million. However, by August 2021, the figure had dropped again—likely due to the impact of lockdowns associated with the Delta outbreak. In August 2021, there were 2.4 million casual employees, representing 23 per cent of all employees.
2.10
According to industrial relations and Occupational Health and Safety academic, Dr Elsa Underhill, the last 15 years has seen a shift in 'the composition of casual employment', with more primary income earners, and more men now employed as casuals. More people who would 'ordinarily have been employed as a permanent worker in the past' are now being engaged on casual contracts: 'So casual employment, in a sense, has become more acute even though the numbers may not reflect that. The impact of it upon those doing the work has become more'.
2.11
Evidence from the Household, Income and Labour Dynamics in Australia (HILDA) Survey report published in 2021 indicates that casual employment is not necessarily the temporary form of employment that it is often thought to be. In fact, HILDA data shows that:
over half of people who were employed as casuals in one survey year were still employed as casuals in the following year;
the 'persistence rate in casual employment appears to have increased in the most recent decade'—rising from 54 per cent to 57 per cent;
the 'rate of movement' of casuals into permanent employment has been lower in the second decade of this century than the first decade;
there has been a 'slight rise in the proportion of casual employees exiting to fixed-term employment' (as opposed to permanent employment); and
the majority of people who were casual employees in any one HILDA survey year were 'still not in permanent employment five years later'.
2.12
Some participants in the committee's inquiry argued the fact that casual employment has not increased in decades indicates that work in Australia is as secure (or as insecure) now as it was in 1996. The Australian Chamber of Commerce and Industry (ACCI) submitted that:
Despite media and community interest in so-called 'insecure work', all forms of alternative employment, be it casual or part-time employment, independent contracting, labour hire arrangements or on-demand work remain stable and at the margins, non-statistically variable or insignificantly changing to justify concerns about the Australian labour market.
2.13
However, Professor of Employment Relations at Griffith University, Emeritus Professor David Peetz FASSA submitted that '[i]nsecurity does not require casualisation'. The National Foundation for Australian Women (NFAW) similarly argued that the 'displacement of secure by insecure work' is not caused by 'a single factor', but is driven by 'multiple strategies' which enable employers 'to maximise numerical flexibility, maintain a constant downward pressure on wages and side-step the responsibilities of the National Employment Standards'.
2.14
Evidence received during the course of the inquiry did not support the ACCI's assertion that 'all forms of alternative employment' are 'stable and at the margins, non-statistically variable or insignificantly changing'. A number of inquiry participants provided evidence to show that these 'alternative forms of employment'—specifically part-time, independent contracting, labour hire, and on-demand work—have increased and continue to increase.
2.15
Professor Peetz observed that, while the proportion of those engaged in casual work has remained stable, '[o]ther forms of cost reduction that involve shifting risk to employees … have grown in prominence (contracting, labour hire, franchising, the 'gig economy'), and part-time work has increased markedly. There is also evidence suggesting other indicators of insecure work, such as underemployment, underutilisation and multiple job holding have increased, and are increasing. This evidence is detailed below.
Non-standard employment has increased
2.16
The NFAW noted that there are many ways to define insecure work, but regardless of how one defines it, 'it is important to note that in 2018—for the first time ever', the proportion of employed Australians in permanent full-time paid jobs with leave entitlements became less than 50 per cent.
2.17
'Non-standard employment' describes 'any form of employment other than permanent full‑time dependent jobs'. Not all non-standard employment is insecure. For instance, a permanent part-time job with adequate, predictable hours and fair pay may be 'non-standard employment', but not insecure work. However, the rise of non‑standard employment still provides a useful starting point for measuring the increase in insecure work.
2.18
According to Melbourne University labour academics, Inga Laß and Mark Wooden:
Evidence suggests the incidence of non-standard employment is on the rise in many Western countries. … In Australia, the incidence of non-standard employment rose dramatically during the last three decades of the 20th century, driven mainly by growth in both part-time and casual employment.
2.19
Laß and Wooden estimated the proportion of employees in non‑standard jobs increased from 24 per cent in 1971 to 47 per cent in 2000, and reached just under 57 per cent by 2017. ABS estimates are similar: in 2016 the ABS found 52 per of employees were split among non-standard employment types: casuals (about 19 per cent); self-employed (17 per cent); permanent part-time (about 13 per cent); fixed-term contracts (almost 3 per cent); and fixed-term casuals (about half a per cent).
2.20
According a 2021 Parliamentary Library Research Paper, there was 'a slow steady decline' in standard employment between 1992 and 2017 (from around 70 per cent to about 60 per cent)—the downward trend can be clearly seen in see Figure 2.1. There was a slight increase to almost 62 per cent in 2021. However, the increase was due to 'the pronounced shedding of casual employees [during the pandemic] rather than strong growth in permanent full‑time employment', and does not indicate a trend.
2.21
This evidence contradicts the view submitted by the ACCI, that:
… the share of workers engaged in alternative forms of employment, whether it be as casuals, independent contractors, and under labour hire arrangements, has not changed substantially in over a decade.
2.22
The ACCI argued that the increase in non‑standard forms of work is a legitimate and natural result of more women entering the workforce, seeking flexibility, and the changing nature of businesses and technology. The ACCI further argued that these forms of work are 'no less appropriate than other forms of genuine and consensual labour engagement'.
2.23
While women entering the workforce and seeking part-time work is certainly a factor in the increase in the proportion of casual and part-time employment, 'male casual employment' has also been 'gradually increasing'. This is because industries that are 'traditionally male', such as Construction and Manufacturing, have seen 'increasing casualisation', while at the same time, 'increasing numbers of men' are now 'working in service industries which are characterised by casual work'.
2.24
While describing non-standard forms of work as statistically 'stable' and 'at the margins', the ACCI also said these forms of work 'have an increasingly important role to play in ensuring an agile and productive workplace' in Australia because 'the needs of the modern economy cannot and will not be met by employing only permanent employees working between 9am and 5pm Monday to Friday'.
2.25
Evidence gathered during the course of the inquiry indicates that the increase in non-standard forms of work is made up by an increasing proportion of jobs in the following categories:
part-time jobs—which the committee observed are being used in some sectors as 'quasi casual' jobs;
indirect employment, outsourcing and labour hire jobs;
on-demand, 'gig' and independent contractor jobs; and
fixed term contract roles.
2.26
The committee asked the Australian Bureau of Statistics and the Parliamentary Library for advice and analysis to assist it in digging deeper into the various data sources on Australia's workforce arrangements. This analysis was combined with the evidence received during the course of the inquiry to paint a more detailed picture. Here are the committee's findings.
Insecure part-time work has increased
2.27
At almost a third of the workforce, Australia's proportion of part-time work is the third-highest in the OECD, behind the Netherlands and Switzerland.
2.28
There was 'a surge' in permanent part-time employment between 1996 and 2016. Part-time employment more than quadrupled for men, and more than doubled for women over this period. While only 10 per cent of workers were part-time in 1996, by 2018 the proportion had risen to almost a third of all employees, at 32 per cent.
2.29
Professor Peetz observed that the growth in permanent part-time work between 2009 and 2016 (an increase of 36 per cent) outpaced the growth in casual part‑
time work (which increased by 13 per cent). This was likely due to retailers switching to this form of employment after realising that permanent part-time workers tend to be 'more committed' to the role, and more reliable, than casuals.
2.30
The NFAW explained that part time work is not 'in itself … a measure of precarity'. However, 'rapid increases' in part time work, as have been observed over recent decades, are often indicators of 'underemployment and underutilisation'. For instance, in 2017, 27 per cent of part-time workers wanted to work more hours, suggesting that for many part-time work is not a choice but the result of 'a lack of fulltime opportunities'.
2.31
Insecure and precarious part-time work has increased. This is evidenced by an increasing use of permanent part-time workers as a 'quasi-casual' workforce—such as in aged care. The NFAW observed that:
… at the same time that massive increases in casualisation began to taper off, the industrial relations system had widened the scope for using permanent part-time employment as casual work without the necessity of a casual loading, and often without overtime.
2.32
The Centre for Future Work at the Australia Institute (Centre for Future Work) submitted that many permanent part-time workers 'do not have week-to-week income security' and have fluctuating and unpredictable hours and pay. While these workers 'fall outside the traditional understanding of "casual" labour', their 'underemployment and week-to-week uncertainty' is ostensibly the same.
2.33
The Centre for Future Work reported that part-time work has increased markedly during the pandemic-related employment recovery, where nearly 60 per cent of new jobs created since May 2020 are part-time positions:
Part-time employment (both waged and self-employment) increased by over 518,000 positions from May 2020 through February of 2021–compared to 358,000 fulltime positions created in the same period.
2.34
The committee's second interim report identified a number of sectors in which part-time employment contracts are being used in a way that denies employees 'working time security', economic security and fair pay. Many of these sectors are undervalued, low-paid female-dominated sectors—including aged care and disability care—with a high proportion of Award-dependent workers.
2.35
The industrial Awards in these sectors allow employers to engage workers on low minimum-hours contracts, then roster additional hours at ordinary rates—'flexing' up and down. This leaves these workers with little job security and inadequate income security. For instance, a worker may be unable to secure a home loan, despite working over 35 hours a week in the same job for many years, because their contract only guarantees them eight or ten hours a week.
Indirect employment and outsourcing have increased
2.36
A major contributor to the increase in insecure work over recent decades is a rise in 'indirect employment' through contracting and sub-contracting, and outsourcing—practices often adopted by corporations to 'minimise the costs and risks they face, and avoid accountability when things go wrong'.
2.37
According to Professor Peetz, the number of 'franchise business units increased by 80% between 1998 and 2014'. Many corporations have set up 'spin-offs or subsidiaries' or outsourced work to contracting firms, minimising risk for the parent company. This method of outsourcing has the effect of 'enabling companies to avoid accountability, it cuts costs—which means wages are lower—and it transfers risk to workers—which means jobs are more insecure'.
2.38
This outsourcing model negates the need to engage employees casually, because employers are able to make employees redundant 'when the work dries up', and do not have to pay redundancy benefits if 'the contractor firm goes bankrupt'. According to Professor Peetz, this model has also led to an increase in 'insecure small business owners'.
2.39
There were differing views expressed during the inquiry as to the appropriateness of these business practices. Professor Anthony LaMontagne from the Institute for Health Transformation at Deakin University was asked to respond to the idea that job security is 'inextricably linked with enterprise security'—he said:
… why should the worker employed by a business bear the risks of that business undertaking if they're not sharing in the rewards? So the businessperson is … entering the market and taking the risk of surviving and thriving or whatever the case may be for that business, and the employees … are not sharing in the rewards along with those risks. … So it's not fair for the business to externalise that risk onto the worker without sharing the rewards.
Labour hire is more prevalent than the commonly-used data suggests
2.40
This inquiry has demonstrated that labour hire is more prevalent than the most commonly relied-upon data indicates. According to IBISWorld, in 2021 there were almost 11 000 labour supply businesses in Australia, with the industry being worth over $30 billion.
2.41
The data used by most commentators to identify the number of people engaged in labour hire or temporary third-party employment in Australia is the annual Characteristics of Employment supplement from the ABS. Labour hire data is published in the Working arrangements release. This survey data provides the number of employees who report being 'registered with a labour hire firm or an employment agency', and 'whether they were paid by a labour hire firm or an employment agency'. According to the ABS, this source is the 'best source of information on labour hire as a working arrangement in Australia'.
2.42
In August 2021, according to the Working arrangements release, 3 per cent of employees reported being registered with a labour hire firm or employment agency, and 30 per cent of those reported they were paid by a labour hire firm or employment agency. Similarly, in 2020, the figures were 3 per cent registered with a labour hire firm, and 33 per cent of those paid by a labour hire firm—a total of 112 600 individuals.
2.43
These figures suggest that labour hire in Australia has not increased over recent years. For instance, in 2018 the ABS found the total number of employees reporting they were paid by a labour hire firm or employment agency was around 126 000 people.
2.44
While the Characteristics of Employment supplement represents high‑quality, employee‑level data collection, this methodology results in figures suggesting there are only around 120 000 people employed in labour hire or temporary agency roles in Australia—and that the sector has not grown, and is not growing. Other sources of evidence suggest this does not reflect the reality of the industry.
2.45
While the ABS Characteristics of Employment data does not show an increase in temporary third-party employment across Australia, IBIS World said this of the industry in June 2021:
Firms across the private and public sectors have favoured outsourcing non-core activities over the past two decades, fuelling growth across the Temporary Staff Services industry. Industry revenue increased over the three years through 2018–19, supported by low unemployment and increased business outsourcing.
2.46
The Characteristics of Employment data is not capturing the true number of labour hire jobs, or workers engaged in the industry, across Australia at any one time. Industry itself quotes higher numbers, and academics regularly question how closely the ABS data represents reality—especially when commentators regularly rely on the number of people who report being paid by a labour hire firm or an employment agency as the figure used to estimate the size of the labour hire industry in Australia.
2.47
In her report, Labour Hire & Contracting Across the ASX100, Dr Katie Hepworth from the Australasian Centre for Corporate Responsibility acknowledged the difficulty in measuring 'the exact size' of the labour hire workforce in Australia. Dr Hepworth said one key reason for the current under-reporting in mainstream datasets is likely to be 'confusion on the part of labour hire workers regarding their employment relationship', with workers not always knowing whether they are directly-hired or 'on-hired' to work for a company, and whether they are being paid by a labour hire firm or not.
2.48
Program Manager of the Labour Statistics Branch at the ABS, Mr Bjorn Jarvis, provided evidence that supported this view. Mr Jarvis suggested that, alternatively, the Labour Account data—a collection of data from industry and other sources—'provide[s] the best indication of the size of the [labour hire] industry' from an industry-level perspective.
2.49
Mr Jarvis explained that labour hire falls within the Administrative and Support Services Industry Division, under 'Labour Supply Services', and to get a more accurate picture of growth in the industry, the Labour Account:
… gives us the best sense of the size of the industry, how many people are employed by the industry, the jobs within that industry, the hours that are worked within that industry, and also the total remuneration for people employed within that particular industry.
2.50
According to ABS Jobs in Australia data, the number of jobs in 'Labour Supply Services' (labour hire) has increased from 584 312 in 2011–12 to 797 710 in 2018–19 (latest available data). As reported in this committee's third interim report, this figure is higher in real terms than the commonly relied-upon Characteristics of Employment data, and even higher than industry estimates, which put jobs in the industry around 360 000 to 430 000.
2.51
In June 2021, IBISWorld reported that demand in the labour hire and temporary staff services industry 'is forecast to rise steadily over the next five years, as the negative economic effects of the COVID-19 pandemic wane':
Many firms that require temporary staff do not have the scale or funds to hire specialist full-time staff, and instead rely on industry firms to supply skilled and professional workers on short-term contracts or for project‑based work. Using contract and temporary staff provides client companies with greater flexibility, particularly for short-term work and during periods of economic uncertainty.
2.52
This data also indicates consistent growth in this precarious and insecure form of employment over time. A graph from the committee's third interim report has been reproduced below to again highlight this growth (Figure 2.2).
2.53
There are also sources of data specific to certain industries that show a significant rise in the use of labour hire workers over time.
2.54
The Construction, Forestry, Maritime, Mining and Energy Union—Mining and Energy Division (CFMMEU Mining & Energy) (now the Mining and Energy Union) claimed that labour hire in the coal industry 'has grown dramatically as the proportion of workers directly employed by mine operators has fallen'. The union said the ABS Labour Force survey data fails to provide an accurate picture of casualisation in the industry, because:
The data comes from the Labour Force Survey, which is around 50,000 workers in 26,000 households. But that's out of a workforce of 13.9 million, when the coal mining workforce is shown as 43,000 (a number that bounces around a lot too!). To reach that conclusion, the ABS interviewed approximately 138 workers who said they worked for a coal mining company.
2.55
The union was able to supply more accurate, industry-specific data from the Coal Long Service Leave Corporation, which 'is accurate because all employers in the coal mining industry have to supply monthly returns' which include information on the employment category of all employees and their rates of pay. The data shows that yes, over 30 per cent of the total coal workforce are casuals and the hourly rate paid to casuals is around 40 per cent less than that paid to permanent employees (Figure 2.3).
2.56
The graph shows a consistent and moderately increasing gap between the wages of casual workers—many of whom are labour hire—and permanent workers in the coal mining industry.
2.57
The CFMMEU—Mining & Energy union also highlighted evidence provided at the recent Queensland Coal Board of Inquiry, which heard that in 2017, less than 50 per cent Queensland coal mining workforce were directly employed by mining companies—with the rest engaged through labour hire—compared with rates of 65 per cent in 2002, and 94 per cent in 1996.
On-demand platform work and Uberisation have increased
2.58
While reliable workforce data is not yet available on gig workers, a number of sources indicate that work in the gig economy—precarious by nature—has increased, accelerated by the pandemic.
2.59
As discussed in the committee's first interim report, the 'gig' or 'platform economy' is 'presently small as a portion of the overall workforce', but is growing and 'has the potential to grow'.
2.60
Professor Peetz described this rise in platform work as the 'biggest challenge to the employment relationship', saying 'virtual platforms provide a new, cheap way of control that may replace the need for the employment relationship'.
2.61
In a December 2020 Green Paper, The Rise of the Gig Economy and its Impact on the Australian Workforce, the Actuaries Institute estimated that the gig economy had increased 'over nine times in market size' from 2015 to 2019, capturing over $6 billion 'in consumer spend'.
2.62
The Actuaries Institute used 'transaction data' to estimate the size and scope of the gig economy in Australia. This data is sourced from de-identified 'electronic bank transactions for more than three million individuals in Australia over five years'.
2.63
COVID-19 led to an initial drop in the use of rideshare, then a massive rise in the food delivery sector. The Actuaries Institute mapped an initial decline of around 70 per cent in the overall gig economy during the early COVID-19 lockdown period, due to 'a sharp decline in private transport', then charted 'a surge' in food delivery services from March to October 2020, leading to a recovery in the sector:
The gig economy has recovered since the beginning of May 2020, capturing over 40% more consumer spend in October 2020 compared to the pre‑lockdown period in February 2020.
2.64
This growth has been confirmed by a number of industry research companies. Hospitality Magazine reported on IBISWorld data which indicated:
Delivery has experienced a boom across the country. More than 4 million Australians use delivery services to order food … Uber Eats, Deliveroo and Menu Log continue to lead the market, with new player DoorDash launching earlier in the year.
2.65
IBISWorld's September 2020 industry report for the 'Online Food Ordering and Delivery Platforms' sector reported annual revenue for the sector of $847.9 million, with annual growth in revenues of 43.8 per cent between 2016 and 2021.
2.66
The Centre for Future Work referred to ABS data showing that self‑employment increased by over 145 000 positions between May 2020 and February 2021, which 'more than offset the decline' in self-employment seen at the start of the pandemic (a loss of around 100 000 positions). However, 80 per cent of the new self-employed jobs are 'owner-managers in relatively insecure situations' (not incorporated, without employees, or both):
This growth of insecure self-employment represents a continuing shift in business practices toward nominally 'independent' contractors and sole providers (including people working through digital 'gig' platforms), who have little infrastructure, support, or security in their jobs.
2.67
Uber itself has reported that the number of drivers on its rideshare platform has exceeded 60 000, with another 59 000 people working on its Uber Eats platform. This combined total would make Uber the second largest employer in Australia, albeit its workers are engaged as contractors without access to rights in the National Employment Standards, rather than hired as employees.
2.68
While arguing that gig work is marginal, the ACCI also reported that decreases in unemployment seen in late 2020—with the creation of '110 000 new jobs'—was 'largely driven by non-employees':
During the initial weeks and months of the COVID-19 crisis many workers turned to alternative forms of employment, such as roles through agency work or in the on-demand economy…
2.69
The ACCI submitted that, 'Uber alone experience[ed] a 59% increase in delivery workers joining the platform to support themselves financially during the pandemic'. Gig platform, Freelancer, also reported that the number of job ads posted on their site increased by 14 per cent between April-June of 2019 to April-June of 2020—'an increase from 429 000 job posts to 605 000'.
2.70
The ABS reported 'strong employment growth for non‑employees' during the pandemic, citing an increase in the number of 'owner managers (self‑employed people) without employees, most of whom were working in an unincorporated enterprise' (sole traders). Between July and August 2020, the number of sole traders increased by 50 200 people; in contrast, over the same period, there was 'minimal growth in the number of employees'—up by only 2 600 people.
2.71
Australian labour academics, Inge Laß and Mark Wooden have noted that the proportion of people working part-time hours 'in their own business without any employees' has been rising 'slowly but steadily', which 'may reflect growth in gig work'. The authors note that 'sole trader status and part-time hours' characterise many of the new on-demand platform sector jobs, such as food and parcel delivery. Existing methods of data collection on employment status can fail to capture some of these jobs, as collection tends to focus on the workers' status in their 'main job', whereas many platform jobs are secondary employment.
2.72
The NFAW observed that 'the proportion of self-employed individuals working part-time has grown markedly in recent years, reaching 35 per cent in 2017'.
2.73
The problems with enumerating gig workers are well known, and governments are beginning to take steps to resolve these issues. These steps are further discussed in the section on data later in this chapter.
Fixed-term contracting has increased
2.74
According to the ABS, there were 413 100 employees on a fixed-term contract in Australia in August 2020, which accounted for 3.3 per cent of all employment. This represents an increase of 24 600 people, or 6.3 per cent, since August 2019—an increase likely related to COVID restrictions.
2.75
The NFAW pointed out that many industries with a high proportion of fixed term contracts are dominated by women.
2.76
Since 2014, employees on fixed-term contracts have maintained a fairly consistent share of total employment of around 3 per cent of just above. In 2020, unusually the number of employees on fixed-term contracts working in the private sector slightly outnumbered those in the public sector—220 600 compared with 193 200). However, a much greater proportion of public sector employees were on fixed term contracts than private sector employees—9.9 per cent compared with 2.1 per cent.
Other indicators of insecurity
2.77
As well as an increase in the above insecure or precarious work status categories, evidence shows an increase in certain other indicators of insecurity—from multiple job holding and low wages growth, to increasing underemployment and underutilisation.
Multiple job holding has increased
2.78
Evidence indicates that an increasing number of Australians are working multiple jobs to make a living.
2.79
In September 2021, the ABS declared that multiple job holding had hit a 'record high' in the June quarter, with 6.5 per cent of employed people working more than one job. Figure 2.4 shows that, aside from a brief dip due to the pandemic, the number of multiple job holders has been increasing since 1995, even if the rate has remained relatively steady.
2.80
According to Mr Jarvis, the biggest increases in multiple job holding were in Administrative and Support Services (incorporates labour hire), Arts and Recreation Services (heavily-impacted by the pandemic), and Education and Training. However, Mr Jarvis also noted that the rate of multiple job holding 'was higher than pre-pandemic levels' across 'all industries':
The growth in multiple job holding coincided with a faster increase in secondary jobs, which increased by 1.4 per cent during the June quarter, compared with 1.2 per cent for main jobs.
Secondary jobs increased by 33 per cent over the 2020-21 financial year, from a low in the early stages of the COVID-19 pandemic in the June quarter of 2020. By June quarter 2021 they were 9 per cent above pre‑pandemic levels.
2.81
The NFAW argued that the official ABS data on multiple job holding may not include gig economy workers, most of whom could be said to hold multiple jobs, and as such is under-counting multiple job holders.
2.82
The committee asked the ABS about the limitations of its data in relation to multiple job holding and the ABS replied that it is 'confident' that its Labour Account data 'provides a comprehensive picture of the number of multiple job holders and the industries in which main and secondary jobs are worked'.
2.83
However, the ABS also acknowledged gaps in its data collection, saying it 'will be undertaking additional data collection activities to address data gaps', including in relation to 'on-demand/digital platform workers'.
Low wages growth is a side effect of insecure work
2.84
Professor Peetz submitted that record low wages growth observed since the Global Financial Crisis (GFC) is another sign of job insecurity:
The growing insecurity of workers, including through underemployment, and the concentration of economic power on the other side of the table help explain why wage growth is stagnating. Indeed, that successful 'war on wages' may be the biggest single sign of worker insecurity.
2.85
The NFAW made a similar claim, saying casualisation 'arguably exerts downwards pressure' on wages and conditions, the quality of jobs for 'all employees, not just casuals'. In addition, the NFAW identified a 'clear association' between increasingly precarious employment and deteriorating wages and working conditions.
2.86
According to the NFAW, many workers feel uncertain 'about their economic future' due to Australia's stagnating wages, with annual wage increases decelerating 'to their slowest pace in decades' since 2012: 'Among women working casually, for example, wages increased by only 1.7% between 2010 and 2016'.
De-unionisation is linked to job insecurity
2.87
A number of inquiry participants cited research showing the link between de‑unionisation and job insecurity.
2.88
The Australian Institute of Employment Rights pointed to international research which has established a link, and said that de‑unionisation 'has undermined workers organising protection'.
2.89
Professor Paula McDonald from the Queensland University of Technology Business School explained that 'decreasing density of union membership' makes 'collective efforts' to improve the conditions of work for the benefit of employees more difficult:
If we think about young people, for example, in tourism, hospitality, retail et cetera, there are really low rates of unionisation and hence problematic opportunities for them to exercise voice in a way that doesn't personally disadvantage them in the workplace or such that no punitive consequences come as a response.
2.90
A reduction in union power and influence over recent decades has also led to 'a pronounced increase in [income] inequality'.
Unemployment is higher than headline figures suggest
2.91
Prominent economist, Mr Saul Eslake said the definition of 'unemployment' used by the ABS in its monthly Labour Force survey 'is often criticized as being too narrow':
It's certainly not the definition used by Centrelink to determine eligibility for the JobSeeker payment: single people can currently earn up to $1,217 a fortnight … from casual employment and still be eligible for at least some payment whilst looking for work … It's also increasingly recognized, by economists and others, as being an inadequate and incomplete measure of labour market 'slack'—especially given the greater flexibility which employers now have to manage variations in their labour requirements by adjusting hours worked, as opposed to 'head-count'.
2.92
Mr Eslake provided evidence showing that, in the wake of the initial COVID lockdowns in 2020, the official unemployment statistics 'resulted in an unusually large number of people being recorded in the monthly labour force surveys as being 'employed'', who fell into categories that could be otherwise considered unemployed. These categories were:
People working 'zero hours' because there was 'no work, or not enough work available' or because they had been 'stood down'—never higher than 80 000 over the five years before the pandemic, this number 'shot up to almost 767 000' in April 2020.
People working zero hours but 'counted as "employed" for the reference week of the survey because they 'began, left or lost a job' during that week—never more than 27 000 people 'in any month' since this category was first measured in 2014, 'it leaped to 357 000' in April 2020.
People who became unemployed during COVID but did not 'actively look for work' because 'they were precluded from some forms of doing so … by restrictions', 'saw little point in doing so given the apparent economic situation', or were ('temporarily') not required to look for work in order to receive the JobSeeker payment—estimated to have risen from 'zero' in March 2020 'to a peak of almost 675 000 in May 2020.
2.93
Mr Eslake explained that adding these categories of people to the unemployment rate creates an 'effective unemployment rate' in April 2020 of 17.8 per cent—significantly higher than the official unemployment rate of 7.4 per cent.
2.94
Mr Eslake argued an effective unemployment rate 'arguably, more "accurately" captures the impact of COVID-19, and the measures deemed necessary to control it, on the labour market'. (See Figure 2.5.)
Underemployment and underutilisation have increased
2.95
A number of sources indicate that underemployment and underutilisation have increased over the decades.
2.96
The ABS measures underemployment in relation to two groups of employed people:
part-time employed people who would prefer to work more hours, and were available to do so; and
full-time employed people who worked part-time hours (i.e. less than 35 hours) in the reference week for economic reasons.
2.97
According to the ABS, in February 2020, before the pandemic, there were around 13 million people employed in Australia ('almost two-thirds of the civilian population aged 15 years and over'), and around 1.2 million—or 9.1 per cent—of these workers were underemployed. Almost all of those who were underemployed were part-time workers (1.1 million).
2.98
The underemployment rate among part-time workers increased during the recession of the early 1990s (from 4.1 per cent of workers in 1991, to 6.3 per cent in 1993). The underemployment rate for part-time workers increased again during the GFC (from 4.5 per cent in September 2008, to 7.2 per cent in December 2009). These increases only impacted part-time workers.
2.99
Part-time underemployment never returned to 'pre‑downturn levels', meaning the increase in part-time underemployment 'became a permanent shift'. The ABS concluded:
The increases in part-time underemployment reflect the increasing share of part-time employment in the labour market, as well as the increasing prevalence of underemployment amongst part-time workers.
2.100
In February 2020, 43.8 per cent of underemployed part-time workers reported being 'underemployed a year or more', and 49.7 per cent of underemployed part-time workers had taken 'active steps to look for additional hours'.
2.101
In addition to the underemployment rate, the ABS tracks the 'expanded underemployment rate'. This data incorporates additional groups of employed people who 'work less than their usual hours for economic reasons or have a preference for more hours'. These groups are:
part-time employed who worked less than their usual hours, for economic reasons;
full-time employed who worked less than their usual hours for economic reasons, but still worked 35 hours or more; and
full-time employed who would prefer (and are available) to work more hours.
2.102
Figure 2.6 shows the underemployment rate alongside the expanded underemployment rate from 2014 to 2020. The expanded underemployment rate rose to a high of 22.6 per cent in April 2020 due to the impacts of COVID‑19, then fell throughout the year to reach 14.1 per cent by December 2020.
2.103
In addition to underemployment statistics, the ABS collects data on persons who are not in the labour force. In February 2020, 6.8 million people were not in the labour force. Of these, 1.1 million (16.3 per cent) were 'marginally attached'. This means they 'wanted to work, and were either available to start work, or had actively looked for work'.
Hours worked
2.104
Inquiry participants were interested in the issue of hours worked and the impact of the pandemic on underemployment and underutilisation.
2.105
Professor Peetz explained that, long before the pandemic, underemployment was increasing, as captured in hours worked:
Between 2010-11 and 2016-17, the number of hours sought, but not worked, by underemployed people grew by 31%, five times the total growth in hours worked.
2.106
Similarly, the NFAW submitted that hours worked per employed Australian had 'dropped by over an hour a month for the five years' leading up to 2017.
2.107
The ACCI stated that underemployment 'hit a historical high of 13.8% (1.8 million people working reduced or zero hours) in April 2020', and youth unemployment 'hit 23.6% in April 2020'.
2.108
The impacts of the pandemic on casual workers, women, young people and older workers are discussed in greater detail further on in this report.
The need for better data and statistics
2.109
Inquiry participants were generally complementary about the quality of Australia's labour statistics and data collection. Participants acknowledged the hard work of statisticians, and the limited resources at their disposal, along with their dedication to this work. However, it was acknowledged that there are certain areas where there are gaps in that data, or where data and statistics could be made more accessible.
2.110
This section reviews evidence and proposals for reforms or refinements to labour data collection and distribution that arose during the inquiry, and makes some observations and recommendations.
2.111
The NFAW argued that Australia's existing data collections 'are failing to elucidate the current situation for some groups of insecure workers'. The NFAW argued that, because they inadequately capture gig workers, multiple job holders and other 'segments of significantly insecure work', current Labour Force surveys may be undercounting insecure work, as well as 'obscuring an increasingly bifurcated workforce, with clear winners and losers'.
2.112
Professor Peetz suggested the ABS should collect data on self-reported work insecurity, including: insecurity of employment; insecurity of hours; and insecurity of income (compared with 12 months ago). The ABS could also collect data on other 'indicators of insecurity', such as:
the extent of 'employment in franchise employment' (better collected through employer surveys than employee surveys); and
'authoritative data on the extent of the platform economy'.
2.113
The committee asked the ABS about these suggestions. The ABS responded that it:
… must ask questions that can be understood and answered consistently by all people, in order to produce official statistics that enable people to draw effective inferences about the state of labour market and changes over time. 'Secure' is a term that would require respondents to make a subjective assessment and would be interpreted differently by different respondents.
2.114
When asked on notice, 'Has work become less secure over the last 40 years and what are key indicators?', the ABS replied simply: 'The ABS does not produce measures of how secure work is'. However, the ABS also pointed out that existing ABS surveys currently collect data about changes in people's hours, and whether people are intending or expecting to change jobs in the near future. These—and a number of other sources of ABS data—are arguably indicators that could be used to track changes in (at least people's perceptions of) job security over time.
2.115
Some participants argued for a greater focus on disadvantaged workers, through better disaggregation of data, and greater data collection among disadvantaged cohorts. For instance, the NFAW recommended that the government:
ensure better data collection to capture alternative and insecure forms of work, including labour hire, sham contracting and gig work; [and]
better monitor and report on risks of insecure work and exploitation among particular groups in Australia’s workforce, including disaggregating data by disability, indigeneity, ethnicity and visa status with a view to better protecting those segments of the workforce that are particularly vulnerable to insecure work.
2.116
Other suggestions were focussed on making data and statistics more readily‑available and more 'usable' by average people. Professor Peetz said:
Increasingly, [the ABS] is collecting data that it does not publish, but makes available through online access to microdata in various forms. That is useful for very serious and singularly focused researchers who have the time and inclination to chase those things up. But for a data-literate journalist or other observer, for example in a community group, who wants to look something up in a reasonably short period of time, it becomes very hard—many will not or cannot do it. So a more accessible approach to publication would help.
2.117
Emeritus Professor Michael Quinlan commented on the difficulty in measuring fluctuations in the use of migrant labour, saying he couldn't provide 'an accurate figure of how much of an increase has occurred lately'. Describing the data collected by the ABS on short-term migrant labour as 'one of the major issues', Professor Quinlan said: 'I think there could be significant improvements to that to identify the extent and nature of these work arrangements'.
2.118
The ABS was asked to comment on reports that it is investing in new data collection in relation gig economy workers, and was able to confirm that this is indeed the case:
The Government provided $2.8 million over four years to the ABS in the 2020–21 Mid-Year Economic and Fiscal Outlook to collect additional information on industrial relations related topics from people (through the annual Characteristics of Employment supplement to the Labour Force Survey) and businesses (through the two-yearly Survey of Employee Earnings and Hours).
The ABS is in the early stages of developing additional questions for the Characteristics of Employment supplement … around on-demand/digital platform work and its nature and extent.
… All of these data will be new data.
Committee view
2.119
Inquiry participants broadly agreed that there were 'data deficits' when it came to insecure workers—including gig workers—and that there was a growing 'consensus' among academics, unions, and many government bodies that insecure work is on the rise.
2.120
In the 1980s and early 1990s many employers moved from permanent to casualised workforces to reduce costs and overheads, and transfer risks from the company onto individual workers. Once this mechanism had reached the limits of its usefulness, other mechanisms came into play.
2.121
Mechanisms like using part-time workers in a way similar to casuals, outsourcing, subcontracting, using labour hire services or agency staff—these measures have all grown in ways that are difficult to capture in official statistics, increasing employers' 'flexibility', while disadvantaging workers impacted by them.
2.122
The pandemic has accelerated a trend that has been apparent for the last decade at least, in which the most disadvantaged workers have become even more disadvantaged. Some workers who lost casual work during the pandemic were forced to turn to gig work, losing even the minimum safety net of award pay rates; some regained work in the traditional employment market; some remain in the gig economy; and some have simply given up.
2.123
Many discouraged workers are invisible, many women have dropped out of the labour force and are now being ignored in the numbers—underemployment is still up and underutilisation has grown slowly over time.
2.124
On-demand platform work has grown but there is no real way to measure its true extent at this stage. Some gig workers are illegal, and some are working on the black market—certainly many of the multiple jobs that gig workers hold are not captured by official statistics.
2.125
The committee is encouraged to see that the Government is taking steps to rectify these gaps in the data. However, we caution that the data must be collected at both the employee and industry levels, and made freely and widely available. The industry data must be transparent.
Problems with labour hire data
2.126
Two key sources of data on employment in Australia are the Australian Bureau of Statistics Labour Force Survey and Characteristics of Employment Supplement, along with other ABS data sources, and the University of Melbourne's Household and Labour Dynamics (HILDA) Survey. Both sources have large sample sizes, sound methodologies, and are of a high quality and standard.
2.127
However, these key surveys are employee surveys, and do not always capture the size and details of specific industries or working arrangements, like labour hire. They have also failed to adequately capture, categorise and enumerate some emerging work types, like on‑demand platform work. It is likely that this has led to an undercount of these work types in the data over the past ten years.
2.128
In particular, ABS data on labour hire is confusing at best—misleading at worst. Survey report after survey report—correctly cited—allows commentators to argue that labour hire and temporary third-party employment are not increasing in Australia. All the while, the industry grows, increasing in size and value, and reporting more and more jobs in its industry-level data.
2.129
At a minimum, the ABS should start publishing the labour hire statistics from the Characteristics of Employment supplement with caveats around the reliability of this self-reported employee data. However, it is the committee's view that the ABS should provide clearer and more coherent statistics on labour hire.
2.130
The Labour Account data provides a more realistic indication of the number of jobs across the sector in Australia, and a better source of information on earnings, hours worked, and other indicators. However, few people will have ever heard of this data, of know of its existence.
2.131
Most people are not statisticians, and do not have the skills, time or inclination to trawl through the ABS's databases or try to interpret the many graphs and data cubes available for download. As such, what is chosen to feed into the 'Key statistics' on the 'Working arrangements' page for labour statistics is all that many people will ever see.
2.132
The committee is also conscious of the potential for politicisation of the collection and reporting of data and statistics by governments who wish to paint a more flattering picture, or support particular agendas.
2.133
It is particularly concerning that the Attorney-General's Department, which is responsible for policy development dictating Australia's industrial relations system, has little to no information on the number of people working on gig platforms, and insists upon relying on the inferior labour hire data contained within the Characteristics of Employment dataset, rather than the ABS Labour Account.
2.134
As such, the committee is recommending that the ABS is funded and supported to undertake some additional work reconciling and better presenting its labour hire data going forward.
2.135
Labour hire is complex as a working arrangement, and indeed, even as an industry category. However, the committee believes that with a modest amount of additional funding and support, the ABS could ensure that the statistics it publishes on labour hire provide a more realistic indication of the number of labour hire jobs and workers in Australia (no matter which of the ABS's data sources they come from).
2.136
It would also be beneficial for the ABS to publish changes in labour hire employment over time, as it does with casual employment and a number of other indicators. This information is critical to inform legislative and policy responses relating to the industry.
2.137
The committee recommends that the Australian Government provides funding and support to the Australian Bureau of Statistics (ABS) to enable the ABS to reconcile information from the Characteristics of Employment, Labour Account, and Jobs in Australia releases in order to provide clearer and more coherent information on labour hire employment in Australia.
The revised labour hire information should be published on the ABS website and include:
data reconciling the number of people who are registered with a labour hire firm or employment agency with industry level estimates of people employed directly by those businesses; and
data and graphs/tables showing changes in labour hire employment over time.
Supporting independent data and research
2.138
Frank and fearless advice to Government begins with high‑quality information, data and research. The committee commends the statisticians at the ABS, the researchers at the University of Melbourne who conduct the HILDA Survey, and the many labour academics and research bodies that have contributed to this inquiry, for their hard work and dedication.
2.139
It is critical that the ABS be supported to collect and publish data in a robust, professional and independent way, and that policymakers, the public, the Australian Parliament and its committees, have free and unfettered access to ABS data and advice.
2.140
The committee sincerely thanks the ABS for its assistance during the inquiry, but notes that responses to questions on notice were held up excessively in Ministerial clearance processes, and appeared truncated when they did arrive.
2.141
The committee urges the Australian Government to commit to protecting the professional autonomy of the Australian Bureau of Statistics and its statisticians to ensure the highest standard of statistics and advice can be published and provided freely—especially to policymakers, and the Australian Parliament and its committees.
2.142
The Australian Government should not pick and choose the ABS labour hire data it acknowledges, nor should it be hiding critical data from the Australian public.