I acknowledge the Gadigal people of the Eora Nation, and all First Nations people present today.
Thank you for the chance to speak to the McKell Institute. William McKell’s election of New South Wales Premier in 1941 marked the start of an era – 24 years of Labor government. McKell’s government was farsighted – enacting annual holidays, improving workplace safety laws and establishing the Office of the Public Defender. His government looked to shape the future.
In the same spirit, I’ve aimed to speak with the McKell Institute about issues that are ahead of their time. Last year, you were kind enough to host me to discuss competition and artificial intelligence.
Today, I am pleased to share the latest on restraint of trade clauses, an issue that has raised competition concerns around the world. Indeed, in McKell’s provocative Easing the Burden report, prepared in January as a pre‑budget submission, you named non‑compete clauses as one of your top six priorities for the Australian Government this year (Douglass & Cavanough 2024).
Non‑compete clauses
As The Economist points out, when it comes to non‑compete clauses ‘The clue is in the name’ (The Economist 2023). Non‑compete clauses restrict an employee from working for a competitor or setting up a competing business when they leave. Non‑compete clauses typically apply for a certain time, in a set geographic location, and within a defined industry. Let me provide some real examples to get things started.
Breakdancing
‘Charlotte’, a 17‑year‑old landed her first casual job as a dance teacher. It was her dream job, but it wasn’t perfect. Far from it – Charlotte was forced to quit after experiencing harassment. Months later, she took a job at a different dance studio and immediately received a warning letter from her former employer. The letter said Charlotte had breached a restraint of trade clause to not work or volunteer for a competing business for 36‑months within a 15 kilometre exclusion zone – noting the company had several studios. Putting her in a difficult position, Charlotte’s former employer also contacted the new dance studio.
Client disservice
‘Mia’, a disability support worker was doing what she loves – working in the community and helping people. She developed trust and rapport with her clients over many years.
Mia worked for a registered National Disability Insurance Scheme provider but wasn’t too happy about being offered a new contract with a lower hourly rate. So Mia went out on her own as an independent and later joined a rival registered provider. Without any cajoling, several former clients decided to transfer their care plans and follow her to the new provider. Mia received a letter from her former employer stating that she had breached restraint of trade clauses relating to non‑competition and non‑solicitation of clients.
Unbending
Patrick, a 21‑year‑old boilermaker, built a solid resumé and decided it was time to take his skills and pay packet to the next level. His new opportunity involved not working for a competitor but working in‑house for a former client. It was a rural town and customers were hard to find so Patrick’s previous employer wasn’t pleased about losing a client.
Patrick was branded a troublemaker and was sent a letter saying he breached his post‑employment obligations.The letter further applied the blowtorch by threatening court action seeking thousands of dollars in damages plus costs.
Assessing the effect of Worker Restraint Clauses Issues Paper
Businesses might use restraints of trade to help protect their trade secrets and confidential information, as well as their customer lists – and have done so for decades.
The simplest and broadest restraint that businesses can use to protect this information is a non‑compete clause. This is the bluntest tool in the shed. It stops a worker from moving to any competing business, or from starting a new one.
However, a growing body of domestic and international evidence suggests the increasing use of non‑compete clauses is harming job mobility, innovation and wages growth. Non‑compete clauses are being applied to hairdressers, early childhood workers and security guards. Once upon a time, only the best‑paid corporate executives were required to spend a period of ‘gardening leave’ between jobs. Now, gardeners are being forced to take gardening leave.
Our government established the Competition Taskforce in 2023 to examine whether Australia’s competition laws, policies and institutions remain fit for purpose. As an immediate priority, we directed the Taskforce to investigate the use of non‑compete and related clauses that restrict workers from shifting to better‑paying jobs. Working Future – ourEmployment White Paper – also acknowledged concerns about non‑compete clauses, and it confirmed the government’s commitment to investigate their use in Australia (Treasury 2023).
Today, I am pleased to launch the Taskforce’s Issues Paper Non‑competes and other restraints: understanding the impacts on jobs, business and productivity (Treasury 2024). The Issues Paper is an open invitation for everyone – businesses, employees, academics, think‑tanks – to provide their views and help build the evidence base on restraints of trade in Australia. To help understand the impact on workers, businesses and our economy, the Issues Paper shines a light on the different restraint clauses workers can be subject to – such as non‑competes, non‑disclosure and non‑solicitation clauses.
I note the recent contribution of former President of the Fair Work Commission and current RBA board member Dr Iain Ross to this debate where he concludes that the weight of evidence favours a regulatory response to deal with the unfairness inherent in the existing law and practice, and puts forward a number of possible options (Ross, 2024).
The government’s Issues Paper also deals with no‑poach and wage‑fixing agreements between businesses that restrict workers without them even knowing.
Today, I’m going to focus on non‑compete clauses, since research suggests these clauses are inefficiently restricting workers and harming businesses and the broader economy.
I want to make 7 observations.
1. Job mobility is important to the whole economy
First, job mobility matters to the entire economy. In fact, job switching is something economists use as an indicator of a well‑functioning, competitive and dynamic labour market (Treasury 2023 p55).
A high run rate in cricket tells us the batting team is scoring efficiently based on the number of overs faced. Job switching is similar. A high job switching rate is a good sign because it shows the economy is working efficiently to match employees with employers. This creates a double dividend. For businesses, it means improved productivity as they can attract the talent and skills they need. This is especially important for start‑ups and firms looking to expand. For workers, it means job satisfaction and higher wages as they move to more productive firms.
As you know from your own career, the biggest wage gains typically come from switching firms. The e61 Institute recently put a figure on those higher wages (Wong 2024). Researcher Aaron Wong found job switching and the pay increases that come with it are worth $5,700 per year for typical workers. It is worth even more for younger job switchers who can earn on average $7,500 more per year than job stayers. But don’t hand in your resignation just yet – there are benefits to job mobility even if you are happy in your current job. Previous Treasury research found just having outside options puts workers in a stronger bargaining position (Deutscher 2019). Job mobility is also critical to the flow of innovation. The diffusion of new technologies and new ideas occurs when workers join new firms and bring better ways of doing things with them (Lobel 2013).
Despite these benefits, Australia has seen a general decline in job mobility during the past 30 years, similar to other advanced economies (Treasury 2023). This partly reflects an ageing population where older workers tend to shift jobs less frequently. But it’s also part of a broader fall in dynamism since the early 2000s.In particular, entry and exit rates of employers, job reallocation, proportion of high growth firms, and proportion of employment by young firms all declined over this period (Leigh 2024). So it makes sense to carefully assess any barriers that may be limiting people from moving jobs, limiting businesses from expanding and limiting the flow of innovation in the economy.
2. Non‑compete clauses are more widespread than we thought
My second observation is that the use of non‑compete clauses has become commonplace in Australia. For the first time, the Australian Bureau of Statistics conducted a survey of employers on their use of restraint clauses (ABS 2024). The survey found 1 in 5 Australian businesses used non‑compete clauses for at least some of their employees in 2023.
Non‑compete clauses were most common in financial services with 40 per cent of employers using them. Rental, hiring and real estate services were next with 33 per cent of employers using non‑compete clauses. Employers in retail were least likely to use them at 13 per cent.
The Australian Bureau of Statistics survey also suggests that more businesses are likely to use non‑compete clauses in the future. These results build on the employee survey conducted by Dan Andrews from the e61 Institute and Bjorn Jarvis from the Australian Bureau of Statistics (Andrews & Jarvis 2023). That survey found that 22 per cent of Australian workers had a non‑compete clause. This is comparable to estimates in other countries, with:
- 18 per cent of workers covered by non‑compete clauses in the United States
- 26 per cent in the United Kingdom, and
- 37 per cent in the Netherlands (up from 19 per cent in 2015).
3. Non‑compete clauses apply to lower‑paid workers too
My third observation busts a common myth. Non‑compete clauses are often associated with higher‑paid professionals. But that’s often not the case. The Australian Bureau of Statistics survey found the use of restraint clauses – including non‑competes – was not limited to upper‑level managers or executives but included all workers of varying incomes (ABS 2024). It found the majority of businesses that used restraint clauses applied them to more than 75 per cent their employees.
So restraint clauses don’t just apply to the boardroom. They apply to workers in the mailroom too. They are just as likely to apply to the person guarding the carpark as they are to the person guarding trade secrets. The earlier survey of employees found non‑compete clauses apply to workers in lower‑paid jobs such as early childhood workers, yoga instructors and gig workers (Andrews and Jarvis, 2023).
The ABS survey results also highlight a significant difference when it comes to the size of the employer using non‑compete clauses. Businesses with more than 1000 employees were twice as likely (40 per cent) to have employees with non‑compete clauses than those with fewer than 20 employees. Big, incumbent businesses can use non‑compete clauses and other tools to maintain their market power and prevent their competitors from growing or new start‑ups emerging. The fact that non‑compete clauses apply to lower‑paid workers too and they are more likely to be dealing with big business is significant when negotiating agreements.
Overseas evidence suggests workers tend to have relatively lower bargaining power in the negotiation of non‑compete clauses. With low bargaining power, negotiating an employment contract is hard, and awkward. A study in the United States found only 10 per cent of employees negotiate over their non‑compete with many assuming the terms were not negotiable (Starr, Prescott, Bishara 2021). Another issue is awareness, with the Taskforce hearing reports about Australian workers signing contracts without even being aware of the terms.
4. Enforcement threats can create a chilling effect
That leads to the fourth observation from the Issues Paper: enforcement threats can create a ‘chilling effect’ on worker mobility. Common law governs restraint of trade clauses in Australia, with the partial exception for New South Wales.
Common law
Under common law, restraint of trade clauses are presumed to be void and unenforceable unless:
- the employer can demonstrate that the clause is reasonably necessary to protect a legitimate business interest, or
- the worker can show it is otherwise against the public interest.
Over the centuries, courts have determined that reasonableness really depends on the case‑by‑case circumstances.
This has its strengths, since the courts will really dig into a clause to work out what was fair and reasonable depending on the facts.
But this means there are no clear, bright lines for workers, or businesses, to know exactly whether a non‑compete clause is even legally enforceable without going to court. The lack of a bright line means that these clauses have been written into employment contracts, and workers may just assume they’re enforceable. This has downstream impacts on workers and businesses being matched together productively.
Cascading clauses
Rules of severance are another factor in common law. The court can ‘blue pencil’ or ‘sever’ part of a restraint clause leaving the rest of the restraint to stand if it is reasonable.
This has seen businesses take a cautionary approach to non‑compete restraints by using cascading clauses. As the name implies, this strategy involves listing the broadest restrictions at the top of a list and cascading down to the narrowest restriction. For example, a restraint clause agreed between an insurance broking firm and one of its senior employees said: (New South Wales Court of Appeal 2010, NSWCA 267):
The Restraint Period means, from the date of termination of your employment:
(a) 15 months
(b) 13 months
(c) 12 monthsRestraint Area means:
(a) Australia
(b) The State or Territory in which you are employed at the date of termination of your employment;
(c) The metropolitan area of the capital city in which you are employed at the date of termination of your employment.
It was then up to the court to determine each clause in dispute based on the particular facts and circumstances. In this example, the court found the insurance broking firm had a legitimate interest in protecting strong client connections. The court found that a 12‑month restraint, limited just to metropolitan Sydney was reasonable based on the fact, striking out the wider restraints.
New South Wales
It’s a different scenario in New South Wales. Businesses don’t need to rely on a restraint with cascading clauses. Because of the Restraints of Trade Act 1976 (NSW), the starting point here in Sydney is that a restraint of trade is valid to the extent it doesn’t contradict public policy, whether it is in severable terms or not.
The Act permits New South Wales courts to amend – or rewrite – the restraint such that it is not against public policy. This means that restraint of trade clauses being more frequently upheld in some form in New South Wales (56.1 per cent) compared to the average for all other Australian jurisdictions (33.3 per cent) (Chia and Ramsay 2016). But most non‑compete agreements go unchallenged in court. And the Taskforce heard that some businesses seek to enforce restraint of trade clauses in cases where a court would deem them to be clearly too broad.
The Issues Paper says worker enquiries to employment lawyers about restraints of trade are common but are rarely continued beyond initial letters or reminders. Understandably, workers often adjust their behaviour to avoid further escalation. Legal Aid NSW told the Taskforce that, of the more than 100 matters relating to restraints of trade between 2020 and 2023, only one commenced to court proceedings (Treasury 2024).
The cost and uncertainty of litigation are a deterrent for both workers and businesses. But uncertainly tends to weigh more heavily on workers (Arup and others 2013). Workers often lack the legal knowledge and resources to bargain and undertake litigation. Contesting one of these clauses in court can cost tens, if not hundreds of thousands of dollars. This, in turn, creates a chilling effect. Workers may be too afraid to face the risk of a bout of unemployment or a court dispute and won’t move to a better‑paying job.
Job offers from rivals are declined (Starr and others 2020). Job searching is re‑directed to other industries. The new employer won’t be able to benefit as much from the worker’s more specific experience. They won’t be able to pay as much. Plans to start a competing business go out the window. It hurts other workers looking for a new possible employer, and can hurt consumers, who may miss out on lower prices from competition.
The impact of this chilling effect on the economy
There may be significant economic consequences from stopping workers from seeking out a better‑paying job, or from starting a new business.
In theory, workers trade away their right to compete when they have a non‑compete clause, and so their wages should be higher. But studies in the US suggest the opposite.
When non‑compete clauses for low‑wage workers were banned in Oregon in 2008, wages increased for this group by 2–3 per cent (Lipsitz and Starr 2022).
A 2015 ban of non‑compete clauses for tech workers in Hawaii had the same result: new hire wages in the sector grew by 4 per cent (Balasubramanian and others 2022).
This wage growth was likely driven by the large increases in job mobility observed in both states. Improved job mobility means workers move to the jobs they are most suited for. This helps everyone. Non‑compete clauses should also, in theory, support businesses to invest more in their staff and to innovate more.
Research in the US did find that training was 14 per cent higher in states with greater enforcement of non‑competes, but without the pay day for workers, with wages being 4 per cent lower (Starr 2019).
The jury is still out on the net impact on innovation. Creativity is hard to measure, but that doesn’t stop economists from trying. Businesses need to protect trade secrets, but increased competition and the flow of workers can actually boost innovation.
Banning non‑competes for tech workers in Hawaii was associated with a 10 per cent increase in the number of tech firms, with skilled workers spreading across the labour market (Glasner 2023).
5. Businesses have other options to protect their interests
My fifth point is around alternative options to non‑compete clauses. Whether it’s the recipe to the secret sauce, code to latest algorithm or access to a celebrity client book, businesses have reasons to protect their turf. But they have other statutory, contractual and common law protections that may be less restrictive on job mobility.
For example, businesses can use the Corporations Act 2001, which prohibits employees during or after employment in a corporation from improperly using a company’s information for personal gain, third‑party gain, or to cause detriment to the company.
Or other targeted restraints that protect what businesses really care about.
Non‑disclosure agreements restrict a former worker from disclosing information for a certain time period after leaving the business. The Australian Bureau of Statistics found non‑disclosure clauses were the most common restraint clause, used by 45.3 per cent of Australian businesses in 2023 (ABS 2024).
There are also non‑solicitation agreements that restrict a former worker from soliciting clients or customers, for a certain time period after leaving the business.
The Australian Bureau of Statistics survey found around 25 per cent of Australian businesses used non‑solicitation agreements (ABS 2024). But a word of caution, these other agreements can sometimes be broad enough to have the same effect on job opportunities as a non‑compete clause.
And can we ask why businesses rely on non‑compete clauses? Views are welcome. The Issues Paper notes protection of business’ goodwill and intellectual property, discouraging staff turnover or investment in workers may be reasons. But largely comes down to practical reasons. From the business perspective, non‑compete clauses are a relatively low‑cost option.
And in many ways, it’s easier to restrict someone working for a competitor than proving a former employee has spilled the beans. But what’s good for one business might not be good for the economy.
6. No‑poach agreements between businesses could be harmful
The sixth observation is that no‑poach agreements between businesses warrant closer consideration. It doesn’t surprise me that employers want to hang on to their staff. High staff turnover increases the average cost of a worker when you consider spending on recruiting, onboarding, and training. But unlike non‑compete clauses, employees aren’t involved in negotiating the agreements.
Agreements are often made in secret and verbally, so workers don’t know they are impacting on their wages and mobility. In a no‑poach scenario, two or more businesses agree to not solicit or hire each other’s current or former workers.
No‑poach agreements can occur where rivals are competing for the same often limited talent pool. For example, in the United States, the Department of Justice successfully went after a group of tech companies including Apple, Google, Adobe and Intel after it was revealed they had agreements not to recruit or solicit each other’s software engineers (United States v. Adobe Systems Inc and others 2010).
These tech firms had a ‘gentlemen’s agreement’ where they agreed not to cold call each other’s employees about job openings.
No‑poach agreements can also emerge in other contexts, like franchise agreements.
Major franchises such as McDonald's, Bakers Delight and Domino’s have standard clauses that prevent franchisees from hiring workers in other stores (Leigh 2023).The Issues Paper raises questions about the enforceability of no‑poach agreements.
Exemptions for certain anti‑competitive agreements may mean these no‑poach agreements don’t fall foul of competition laws. This is in contrast to other countries such as the US, UK and the European Union, which consider agreements between competitors about their workforce as cartel behaviour, and illegal.
The Issues Paper raises the question as to whether the current exemption in the Competition and Consumer Act 2010 should be reconsidered so that the competition watchdog has power to address agreements between competitors that seek to suppress wages, but not agreements that seek to increase wages.
7. Australia is well‑placed to learn from other countries
My seventh, and final point, is that relative to other countries, Australia is at an early stage in investigating non‑compete clauses. This allows us to observe the different regulatory approaches taken overseas. The Issues Paper notes that many jurisdictions have either banned or restricted the use of non‑compete clauses and many others are contemplating similar moves.
The United States has proposed a nation‑wide ban on employers imposing non‑competes. Five states already have bans in place. The United Kingdom is planning to limit the length of non‑compete clauses to three months – reducing the average duration from around six months while maintaining flexibility for businesses. Austria has banned non‑competes for lower‑paid workers and put in place rules for higher‑paid workers to ensure they don’t unreasonably impede careers. Spain has requirements where agreements are not able to exceed two years for technical employees or six months for other employees. Further, the Spanish system requires adequate compensation for workers where restraints apply. In Finland, employers must compensate employees for all non‑compete agreements starting at a minimum of 40 per cent of the employee’s regular salary for up to six months.
Closing remarks
In 1956, William Shockley shared the Nobel prize for silicon semiconductors. A brilliant scientist, he was also a bad boss. At his US company, he spied on employees and was both racist and paranoid. So eight of his top engineers left, founding Fairchild Semiconductor in 1957. Shockley called them ‘the traitorous eight’, but couldn't stop them. A decade later, two of them left to create Intel. Then another departed to create AMD. One reason why Silicon Valley came to dominate technology was that workers could walk out to create new firms. Since the 1870s, non‑competes have been unenforceable in California (Gilson 1999, Schwartz 2023). Banning non‑competes is one of many factors behind Silicon Valley’s success.
Today, millions of Australian workers don't have the freedom to quit their jobs and start a competing firm. There is a lot happening in the world of restraints on trade.
Australia is at the fact‑finding stage, but we know:
- Job mobility is important to the whole economy
- Non‑compete clauses are more widespread than we thought
- Non‑compete clauses apply to lower wage workers too
- Clauses can have a chilling effect on workers, impacting on labour mobility
- Businesses have other options to protect their interests
- No‑poach agreements between businesses could be harmful
- Australia is well‑placed to learn from other countries
Feedback on the Issues Paper will inform our consideration of whether reform is needed. The Competition Taskforce have also launched a survey for workers and businesses to better understand their experiences with these clauses. In addition, the ABS is linking its employer restraint clauses survey data with the broader ABS business‑side microdata. This will enable a deeper understanding of how these clauses are affecting job mobility, wages, and business dynamism.
We will take a measured approach, informed by evidence. If we decide further reform is required, we will further consult on potential reform options. In the meantime, our government is progressing its competition agenda in a raft of other areas.
I thank the McKell Institute for your interest in these issues and I look forward to your questions.
References
Andrews D and B Jarvis 2023 The ghosts of employers’ past: how prevalent are non‑compete clauses in Australia? e61 Institute.
Arup C, Dent C, Howe J, Caenegem W 2013 Restraints of Trade: The Legal Practice UNSW Law Journal 36(1).
Australian Bureau of Statistics (ABS) 2024 Restraint Clauses, Australia, 2023 Results of the Short Survey of Employment Conditions, containing experimental estimates on the use of restraint clauses in Australia. Released 21 February 2024.
Balasubramanian N, Chang JW, Sakakibara M 2022 Locked in? The enforceability of covenants not to compete and the careers of high‑tech workers Journal of Human Resources 57(S)S349:S396.
Chia H and Ramsay I (2016) Employment Restraints of Trade: An Empirical Study of Australian Court Judgements, Australian Journal of Labour Law, 29(3):283–304.
Deutscher N 2019 Job‑to‑job transitions and the wages of Australian workers, Treasury Working Paper, Treasury.
Douglass M & Cavanough E 2024 Easing the Burden The McKell Institute.
Glasner B 2023 The effects of noncompete agreement reforms on business formation: a comparison of Hawaii and Oregon, Economic Innovation Group.
Leigh A 2024, Game changer: harnessing microdata for a fairer competition landscape, address to the Chifley Research Centre, 30 January 2024.
Leigh A 2023, How uncompetitive markets hurt workers Australian Journal of Labour Economics 26(1).
Lipsitz M and Starr E 2022, Low‑wage workers and the enforceability of non‑compete agreements, Management Science 68(1):143–170.
Lobel O 2013, Talent wants to be free: why we should learn to love leaks, raids and free riding, Yale University Press.
New South Wales Court of Appeal 2010, Hanna v OAMPS Insurance Brokers Ltd [2010] NSWCA 267.
Ross I 2024, Non‑compete clauses in employment contracts: The case for regulatory response, Tax and Transfer Policy Institute – Working Paper 4/2024, March 2024.
Gilson R 1999, The Legal Infrastructure of High Technology Industrial Districts: Silicon Valley, Route 128, and Covenants Not to Compete, New York University Law Review 74(3):575:629.
Schwartz M (host) 22 June 2023, Did California’s Noncompete Ban Fuel Silicon Valley Innovation? [podcast] Uncommon Law Bloomberg, accessed March 2024.
Starr E 2019, Consider this: training, wages, and the enforceability of covenants not to compete, ILR Review 72(4):783–817.
Starr E, Prescott J, and Bishara N 2021, Noncompete Agreements in the U.S. Labor Force. Journal of Law and Economics 64, no. 1 (2021): 53–84.
Starr E, Prescott J and Bishara N 2020, The Behavioral Effects of (Unenforceable) Contracts University of Michigan Law School.
The Economist (12 January 2023), America’s trustbusters plan to curtail the use of non‑compete clauses. Good.
Treasury 2023 Working Future: The Australian Government’s White Paper on Jobs and Opportunities, Treasury.
Treasury 2024, Job Mobility and a Dynamic Economy: Assessing the Effect of Worker Restraint Clauses, Treasury Issues Paper.
United States v. Adobe Systems Inc et al 2010 D.D.C 1:10‑cv‑01629
Wong A 2024, Climbing the wage ladder: Linking job mobility and wage e61 Research Note No.11.