How Australia Banned Non-Competes

The reform in question centres on the use of non-compete clauses, and other post-employment restraints, in work contracts in Australia.

As part of the 2025-26 Budget, the Federal Government announced it would ban the use of non-compete clauses (NCCs) for workers earning less than $175,000. 

In one sense, this announcement is just the beginning. It sets off a process of detailed policy and legislative implementation. But in another important sense, it is the culmination of an equally painstaking journey: of detailed research and evidence gathering, public advocacy and coalition building. It is a journey involving Government Ministers, public servants, global experts, policy commentators, the Bureau of Statistics, and a relatively new economic think tank, the e61 Institute.

This essay describes how these players and forces combined to bring about what we at the e61 Institute believe is a positive change.

Restricting NCCs is an important policy. It is by no means simple, and it has its opponents. That is not to overstate its macro-economic significance. NCCs are one part of a broader labour market mosaic. The intent of this essay is not to equate NCCs with other major large scale reforms of the last forty years. Rather it is to showcase a prototype for successful policy change. In a time when reform seems hard to do, what was the combination of factors that made this one possible? Can it be replicated in other areas? Can it be done on a larger scale? 

We believe it can. So the story is worth telling. And it’s a fitting feature for this exciting new policy publication Inflection Points, because it supports what is a pro-growth and pro-abundance measure. Also because, if there is a consistent tone or dominant sentiment that runs through this story, it is optimism about our ability to make change. 

Post employment restrictions are elements in a contract between firm and worker that purport to limit what a worker can do once they leave the firm. They can take many forms, but our research suggests four prevalent ones:

• Non-compete clauses: restricting the ability of an employee to work for other firms of a defined nature (those within a geographic area, those within the same industry, etc.)
• Non-solicitation of clients: restricting the ability of an employee to leave and take the firm’s client business with them
• Non-solicitation of co-workers: restricting the ability of an employee to leave and encourage other of the firm’s workers to join them
• Non-disclosure agreements: restricting the ability of an employee to disclose details about the workings of the firm to others

Until recently, little has been known about the prevalence of these restrictions in the Australian labour market. By contrast, a recent OECD paper by Dan Andrews and Andrea Garnero shows that NCCs and other post-employment restrictions have been a significant policy issue in the United States and a few European countries, with considerable debate about the appropriate regulatory response (if any).

One challenge regarding NCCs is that, at a conceptual level, the issue is not clear cut. In principle, economic arguments can be mustered to justify or negate the legitimacy of NCCs. As is often the case, detailed empirical work is needed to shine light on which conceptual argument is stronger or more relevant to the facts.

The traditional, more benign view argues that NCCs are justified to protect legitimate business interests. A firm might wish to protect its trade secrets or client relationships. It may also wish to train staff, but hesitate to invest in training if workers might leave the firm shortly after. A NCC could thus support more staff training and potentially more innovation in general.

For high paid staff, the NCC might represent a legitimate part of the explicit contract: more money in exchange for agreeing not to leave and work for competitors. 

But in recent years, a more critical view has emerged. By this account, NCCs are used by employers to limit worker mobility – enhancing the market power of the boss and diminishing that of the worker by restricting their outside options.

The effects can spread beyond the firm to the broader economy. If labour mobility is restricted, and barriers are put in the way of new firm creation, this can stifle the economy’s ‘diffusion machine’ by stopping the spread of new ideas and business models, and the emergence of new competition. It also distorts the Schumpeterian creative destruction process by which resources like labour move from lower value uses to higher value ones, as when a worker goes from a low productivity to a high productivity firm.

To these competing economic views we could perhaps add a third, essentially legal, view:that NCCs, whether good or bad in theory, are irrelevant because in practice they are hard to enforce. 

By early 2023 the weight of empirical evidence overseas was tilting in favour of the more critical view of NCCs. US evidence suggested that:

• NCCs had spread to low wage occupations, like fast food workers and hairdressers
• NCCs were rarely an explicitly bargained outcome, with fewer than 10 per cent of workers negotiating higher pay in return for the NCC. Around one third of workers were asked to sign an NCC after already accepting the job
• NCCs could have a chilling effect even if not technically enforceable. Many workers were turning down job offers from competitors due to a NCC in their contract, even when they worked in US states which had deemed NCCs to be non-enforceable

But in Australia, even by early 2023, this was all still fairly obscure to the domestic policy debate. There was no empirical evidence on the prevalence of NCCs or on the effects of their use. And the issue was decidedly not on the policy agenda. Until it was.

In May 2022, the Albanese Government was elected. Part of the incoming Ministry was Dr Andrew Leigh, appointed Minister for Competition, Charities and Treasury and also – significantly – Assistant Minister for Employment. By 2022, Dr Leigh had been in Parliament for 12 years, spending time in government and opposition. He had a strong understanding of the practical and often murky world of regulatory policy. But he also brought his experience as a Professor of Economics with particular research interests in the use of micro-data. 

That first-hand understanding of the full value chain of evidence based policy was to prove vital in the process leading to reform of NCCs. 

In addition, Dr Leigh brought a strong conviction about the importance of social justice alongside economic efficiency and growth. Often these goals can be in conflict but, as Dr Leigh has often noted, some policy issues can bring them into alignment. And NCCs had a good claim to do so.

If there is a date at which we can say NCCs were firmly placed on the Australian policy agenda, then 23 March 2023 is as good a candidate as any. It was then that Dr Leigh wrote an opinion piece in the Australian newspaper titled How Uncompetitive Markets Reduce Wages.

He explained the perils of monopsony power in labour markets, in language both wonkish and folksy. Literally folksy – the opening para cited Johnny Cash’s song Sixteen Tons as a poetic account of life under the yoke of a dominant employer in a company town. Leigh raised the issue of NCCs. He said respected labour lawyers had told him they were now commonplace. But he was unable to find any survey evidence on their prevalence in Australia. He announced that he would seek Treasury and ACCC advice on the matter.

Later that year, the Government announced a Competition Taskforce supported by a team within the Treasury. That team, led by experienced officials like Jason McDonald and Owen Freestone, became pivotal drivers of the agenda to reform NCCs. 

As Dr Leigh had pointed out, a key deficit was the lack of data. A good policy process would start with strong data on how widespread NCCs were, followed by empirical evidence as to the effect they had on wages, mobility and productivity. This would be followed by evidence-based assessment of policy alternatives, and consulting with business, lawyers, and workers about practical effects. One of the distinctive features of the process was the way the Treasury officials and the Minister worked as part of a broader ecosystem that included the Australian Bureau of Statistics and the independent, non-partisan e61 Institute – still a new kid on the block, with a strong interest in micro-data research and policy impact.

The e61 Institute was founded in 2022 by Dr Andrew Charlton and Professor Greg Kaplan. The founding vision for e61 was that rigorous data-driven economic research could better inform policy making. The name reflected aspects of the approach. The ‘e’ in e61 stands for economics. The ‘61’ is the international telephone code for Australia. From its inception, part of e61’s model has been to bring the best of international research to Australian policy problems – something that became particularly important in the case of NCCs.

Andrew Charlton and Greg Kaplan were both PhD trained economists. They also represented two complementary traditions that run through e61 today: the public policy profession and the academic profession. Both are fundamental to the ambition of rigorous research for policy impact. Early hires at e61, like Dan Andrews and Gianni La Cava, reflected this mix. 

e61’s early work on productivity, led by Andrews, focused on the diminished dynamism of the Australian economy – with indications that new firm entry, job mobility and the creative destruction that characterises a healthy market economy had all slowed. This built on Andrews earlier work at the Federal Treasury and implied a policy program to reduce or remove barriers to firm entry and growth, and labour mobility across firms and locations. 

Andrews’ experience at the OECD also provided important links to the international academic literature on the issue, and policy trends in other developed economies.  

But as Andrew Leigh had pointed out, the first problem was data, or the lack of it. 

The e61 Institute was founded with a strong emphasis on data. In part, this reflected a key development in Australia at the time – the embrace of administrative micro-data by the Australian Bureau of Statistics to provide a rich complement to their survey and census products. Linked administrative data provides the ability for researchers to track, in de-identified form, how firms and individuals interact across the economy and over time. 

The use of the Tax Office’s Single Touch Payroll data during COVID had shed light on employment and earnings trends across the economy and thereby shown the value of administrative data to inform timely policy responses. The data is based on the full population, rather than a survey sample. It can be fast – almost real time in the case of STP during COVID. And it is longitudinal by nature, telling us something about how individuals and firms behave over time.

The real value of these administrative datasets continues to be unlocked by the ABS as they link other data – including their own survey data – into the integrated environment. Hence business survey data can be linked to de-identified administrative data based on tax records to provide further insight than the survey could provide on its own.

When it came to NCCs, we were not even close to this position in 2023. There were simply no reliable numbers on the use of NCCs, let alone their economic consequences. At that point there was no question in the ABS’s Business Surveys that asked about the use of post employment restraints.

But another option presented itself. The e61 Institute’s primary funder is the Susan McKinnon Foundation (now McKinnon), a body set up in 2015 by Sophie Oh and Grant Rule to improve Australia’s democracy through high quality political leadership, public service capability, and policy innovation.

One aspect of the Foundation’s work was the McKinnon poll – a survey of 3,000 respondents, weighted to census data, on a range of issues. Dan Andrews at e61 teamed up with Bjorn Jarvis of the ABS to devise questions about the use of post-employment restrictions to be incorporated as part of the McKinnon poll. This provided an insight from the employee’s perspective: how many respondents said they were covered by some form of restriction such as a NCC.

One challenge was that many workers are unaware that they might be covered by a NCC. As a result, the survey directed the questions about post-employment restrictions to those who had changed jobs in the last 12 months. Of course, to the extent that NCCs limit job mobility, this could lead to an understatement of true prevalence of NCCs.

The poll was conducted in May 2023 and the results published by e61 (authored by Andrews and Jarvis) in June 2023. The headline results reflected that an estimated 22 per cent had a NCC, 16 per cent had agreements prohibiting the poaching of clients and 7 per cent had agreements prohibiting the poaching of co-workers.

Interestingly, the prevalence of NCCs was fairly evenly distributed across the workforce. There were only modest variation by income or between millennials and Gen X (with lower numbers for Baby Boomers). The NCC numbers were slightly higher for managers (at 39 per cent) but also high for community and personal service workers, at 26 per cent.

So for the first time we had addressed one of the problems outlined by Andrew Leigh: we had an estimate of the prevalence of NCCs across the Australian workforce – around 1 in 5. The results of the McKinnon poll diffused rapidly through Australian policy circles and influential media sources, thanks partly to e61’s strong engagement and relationship with leading journalists.

But we knew little about the types of firms that used NCCs and the ways in which they used them. And we could say little about the effects of NCCs on worker or firm outcomes, like pay or job mobility.

For that, we would need a firm-side estimate, and ultimately for this firm-side information to be integrated into the linked data environment overseen by the ABS.

The McKinnon poll did more than provide an initial snapshot of the numbers. It provided a reliable prototype for survey questions on this issue. The next step would be to explore including questions on NCCs into one of the ABS’s existing business surveys.

Statistical agencies like the ABS need to exercise great caution with the content and approach to their surveys. There have been celebrated international cases where official statistics have lost credibility due to low survey response rates, including on important economic indicators such as labour force data (e.g. the unemployment rate). The ABS has managed to maintain high confidence and integrity in its survey products. 

There is a high threshold for getting new questions included. The precedent of the McKinnon poll and the emerging public interest in the issue were useful, along with the proactive stance of the ABS.

The ABS worked with e61 to devise the firm-side questions which would be part of the Short Survey of Employment Conditions. The results were released in February 2024. On the day of release, the e61 Institute released its analysis of the results, entitled The Ties That Bind: Five Facts on Post Employment Restraints in Australia.

The first fact – both interesting and reassuring – was that firm-side responses, once weighted for size, suggested a very similar prevalence of NCCs to that found by the worker-side questions in the McKinnon poll. e61 estimated from the ABS data that around 21 per cent of workers were covered by NCCs.

Other forms of post employment restrictions were somewhat higher than shown in the McKinnon poll. Around 23 per cent were estimated to be covered by no-poaching rules in respect of co-workers. Around 29 per cent were estimated to face a no-poach clause in respect of clients. Around 58 per cent were estimated to be covered by a non-disclosure agreement of some type (compared with 26 per cent in the McKinnon poll).

Other important trends emerged. One was that NCC use seemed to be increasing, with 11 per cent of firms saying they had increased their use in the last five years and 2.3 per cent decreasing their use. Of the firms who did not use NCC’s, one in five said they were likely to do so in the future – higher numbers than suggested they would cease to use them.

Another finding was that firms used NCCs in a fairly blanket, indiscriminate form. The overwhelming majority of firms using NCCs were applying them to something between three quarters and all of their workforce. This suggested that the restraints were not targeted at top, high-end talent, and were unlikely to reflect a genuine bargaining process in return for better pay or other conditions.

Finally, the survey provided a first glimpse at the parts of the economy where post employment restraints like NCCs were particularly intensively used. Unsurprisingly, the incidence was highest in service sectors, including knowledge industries like finance, professional services and the real estate and rental industry. These are parts of the economy where firms often have little in the way of physical capital or even distinctive intellectual property. What holds the firm together is the unique combination of people working together at the firm, and often, the client list.

It is thus understandable that these are the sectors in which firms will go the extra mile to protect their client base and lock workers in. But they are also sectors that rely on new firm entry and competition to drive innovation, unlike other industries where innovation occurs in established business via corporate research and development. As is often the case, the incentive to use NCCs was strongest in the parts of the economy where they could potentially have the greatest economic cost.

A final step would be for the ABS to integrate this survey data into the linked administrative data environment, known as the Business Longitudinal Analysis Data Environment (BLADE). This would allow researchers to assess the effect of NCCs by comparing firms that use them to otherwise similar firms that do not, and to try and assess the effects on job mobility and wages over time.

This integration occurred in the course of 2024, and in October of that year, e61 released further analysis entitled Non-Compete Clauses, Job Mobility and Wages in Australia, authored by Dan Andrews, Ewan Rankin and Jack Buckley. 

As the note outlined, this task was, in many ways, the hardest. There is a natural challenge in identifying the causal effect of NCCs on firm and worker behaviour from what we observe in the data. There is an obvious problem of reverse causality: some firms will increase their use of NCCs because of high worker turnover, affecting the relationship between NCCs and turnover observed in the data. The relationship between NCC use and turnover will therefore not give us a clear sense of the causal effect of the former on the latter.

More broadly, NCC use and firm turnover could be jointly caused by other firm characteristics, which could see NCCs positively correlated with pay, even if the true causal relationship is the opposite. 

These forms of selection bias are difficult to overcome. The results were presented with that caution. To mitigate possible selection bias, regression results controlled for a range of observable variables (worker age, tenure, gender, occupation, firm size, industry, state and remoteness). In addition, the control group was confined to firms that used non-disclosure agreements, but not NCCs. This was to try and control for some unobservable differences between firms that use NCCs and those that don’t. Those using NDAs are likely to have something to protect, and thus be different to firms that do not use NDAs at all, hence a better control group with which to compare those firms using NCCs.

There were two main results. Firms that said they had increased their use of NCCs over a five-year period were found to have lower worker turnover by the end of that period. Workers at firms that increased their use of NCCs saw a decrease in their likelihood of leaving of around 1.8 percentage points.

Regarding pay, the results found that workers with a NCC did not receive a higher initial starting salary than those without (suggesting that there was no explicit compensation for the restriction imposed). But over time, workers with a NCC saw slower wage growth, particularly lower skilled workers. The results suggested that after five years of tenure, lower skilled workers with a NCC could experience a wage level 10 per cent lower than otherwise.

Moreover, the impact of NCCs on high skilled workers was different. While low skill workers saw larger declines in job mobility and wages, high skill workers spent more time in between jobs when leaving a firm using NCCs. That latter fact suggested that NCCs induced a misallocation of high skilled labour at a time when skill shortages were rife.

And so, for the first time, we had evidence of both the prevalence of and effect of NCCs. The selection bias problems have to be taken seriously, and results interpreted accordingly. But this is usually the case in complex public policy problems. Researchers often seek topics and data that provide neat ‘identification’ – a quasi-experimental setup that can remove sources of statistical bias and provide clear causal inference. Policy makers rarely have that ability. They don’t choose the questions they have to answer, and the data often will not furnish simple causal relationships.

The solution is not to give up. It is to approach the estimation of effects with humility and caution and find the best available ways to generate unbiased results.

Moreover, the empirical analysis is but one input into the policy process. Policy makers also triangulate through consultation with stakeholders, and harnessing expert opinion. The policy process overseen by Treasury had these collaborative features.

In parallel to the development of the evidence base, Treasury put in place a policy process to raise the profile of NCCs as an issue and seek out Australian and global experts. This was done in partnership with others.

In October 2023, Treasury and the e61 Institute organised an online seminar for Australian policy makers, bringing together a number of global perspectives on NCCs and recent regulatory approaches. Consistent with e61’s philosophy, the webinar combined the perspectives of academics and policy makers, and brought the best global thinking into an Australian context.

Andrew Leigh opened the webinar – held at a time to accommodate presenters from Europe and the US as well as an Australian audience. Professor Evan Starr from the University of Maryland provided a survey of recent academic research. Perspectives were provided from the OECD, the US Department of Justice and the US Federal Trade Commission. The Australian Treasury provided insights from the domestic policy perspective. Dan Andrews moderated a discussion among the experts and policy makers on the call.

Audience participants included representatives from Treasury, Department of Prime Minister and Cabinet, the ACCC, the Department of Employment and Workplace Relations and several members of the Competition Taskforce appointed by Minister Leigh.

In many ways, the bringing together of this group was indicative of the approach taken by Treasury throughout the process.

In April 2024, Treasury released an Issues Paper on the issue of NCCs, seeking input from experts and affected parties. The paper included references to the data now becoming available, and the international research literature and overseas approaches to regulation.

Central to Treasury’s deliberations was the choice of optimal policy approach. One such approach was a blanket ban on the use of NCCs. Another was a ban above a specified wage threshold. A further option would be to nudge or force firms to offer explicit compensation to workers who were asked to sign an agreement with a NCC.

One dilemma was that high skilled workers were more likely to have consented to NCCs in return for more money. Low skilled workers were more likely to suffer from the removal of outside options and resultant loss of bargaining power. These suggested a wage threshold below which NCCs would be regulated or banned. On the other hand, the gain to economy-wide productivity might be higher from allowing greater freedom of movement for high skilled workers. 

In the end, the Government’s announcement on Budget night in March 2025 reflected a judgment about the relative costs of NCCs for different workers, and the strong emphasis on fairness for low and middle income earners – evident from the very beginning in Minister Leigh’s opinion piece in March 2023.

Such a reform was also attractive from a practical implementation perspective: it can be readily executed via an amendment to the Fair Work Act (FWA), noting that a high-income threshold already exists that limits the eligibility of employees earning above $175,000 to be protected from unfair dismissal under the terms of the FWA.No doubt the Government’s internal deliberations towards a final announcement are a story in themselves (or an important chapter). 

As the forgoing discussion suggests, there are many ingredients to reform. Many necessary conditions. Arguably the policy outcome could not have been achieved without many players and factors working together. Every reform is different, but this episode clearly offers lessons for future efforts.

First, this reform was possible in part because of an acceptance of a bigger contextual problem. Specifically, the sluggish productivity growth in the Australian economy, arguably linked to a lack of dynamism and the slowing of job mobility and new firm creation. This is not easy to solve, but provided a clear sense of urgency about finding potential barriers to mobility and addressing them.

Second, as reflected in Andrew Leigh’s opinion piece (and several times since) this was a reform that did not create a tension between equity and efficiency. It was a form of workplace flexibility that a Labor Government could embrace. Not all reforms offer this. Often there will be trade-offs and a need to think hard about distributional considerations. Even this reform has its opponents. But at least reform advocates could be satisfied that in boosting job mobility, this reform would not increase income inequality.   

Third, the need for good data was a key factor. It was hard to make the case for reform with only an anecdotal sense that NCCs were becoming more widespread. Without data it was impossible to know where in the economy they were being deployed, and whether they really were an issue for low income workers. Having a pro-active ABS, plugged into the policy world and willing to support better policy decisions, was a key advantage. But the ability to start the ball rolling with the McKinnon poll was of critical importance. In both the McKinnon poll and the subsequent ABS survey, the Bureau was effectively working hand in hand with the e61 Institute and the Susan McKinnon Foundation. 

Fourth, beyond data sits detailed empirical research. The conceptual arguments around NCCs were ambiguous. Only through empirical work could we test the competing conceptual accounts. This was greatly aided by the linked administrative data available through the ABS’s BLADE dataset. 

It also requires a lot of thought. Most real world policy questions do not provide the researcher with a neat ‘identification strategy’ to isolate the causal effect of a policy measure. Problems of endogeneity (like reverse causality) abound and there are not always simple ways out. Having researchers with strong empirical technique helps; so too understanding the underlying economics of the situation helps identify all the ways in which an empirical strategy might succeed or fall short. But giving up is not an option. We have to do the best we can with the data and the real world experiments at our disposal. And then be honest and cautious about the limitations. 

Given this, it is important that those doing the empirical work are, and are seen to be, non-partisan and independent. This has been an abiding principle for both the e61 Institute and the McKinnon Foundation. 

Fifth, the international connection played an important role. As noted, it is part of e61’s DNA that it sought to bring the best of the global research and policy effort to an Australian policy issue. Because the e61 Institute combines both academic and policy orientations, this is a natural fit: the academic profession is global by nature. The policy profession is less so. But the ability to bring in overseas academic researchers, the OECD and US regulators to provide insight into NCCs was informative. It also provides comfort to domestic policy makers to know that others have grappled with similar issues and trade-offs.

Sixth, and perhaps most important, was the open collaboration that occurred between the Minister, bureaucracy, ABS and a non-government research institute, backed by its funder. The distinct roles of different parts of the value chain were respected, but also subjected to pragmatic judgment in order to get the job done. One example of respecting the roles of others is that e61, despite its deep involvement in the NCC issue, never made a concrete recommendation to the Government about policy. The Institute retained its role as a researcher, contributing findings to inform the policy process and leaving the policy experts to fashion a well calibrated response. 

In other instances, collaboration was the order of the day. The joint webinar between e61 and Treasury was a prime example. So too the repeated cooperation between the ABS and e61 to develop new data to inform the policy debate. Treasury’s approach throughout was to seek input and share early thinking with candour and trust. Minister Leigh gave regular updates to a range of non-government economists about the progress of the competition agenda. 

The result was that at no point did progress stall because of an artificial governance boundary or sensitivity about information flows. This combination is somewhat atypical, but can be replicated with strong trusting relationships.

Seventh, no reform happens without champions. If nothing else, it takes resilience to push for change once opposition or scepticism emerges. Minister Leigh was that champion – both internally within government and as the public face of the reform. 

Finally, a thread that runs through all of these success factors is that this was a very public process. There was no attempt to solve a problem in-house and then announce it to the world. The deliberative process was itself public. In a synthetic way, this mirrored the policy process built into a Productivity Commission or similar inquiry. The problem was defined. Information sought about its extent and precise contours. Issues papers released. Relevant global precedents explored. Research work put in the public domain to inform the debate. 

All this occurred in a public and observable way. Public processes almost always work best for detailed policy deliberation, especially on contentious issues. They are much more effective at gathering relevant information, but also allow stakeholders to be and feel part of the process. 

As noted at the beginning, the reform of NCCs is not done. There is a process of implementation now underway, with its own story to write. But there is a lot in the story so far, and much to give us optimism about the potential to deliver better policy in the national interest.