Non-compete clauses in employment contracts: The case for regulatory response

Abstract

In the employment context, non-competes are contractual terms which provide that once the employment ends the employee cannot work for another employer in the same industry or field (i.e. a competitor), within a specified geographic area, for a specified time. The existing law and practice regarding non-competes in Australia is plagued with confusion and uncertainty. Non-competes also have adverse economic consequences; they are associated with reduced employee mobility and consequent negative impacts on wages and productivity.

The use of non-competes by Australian business has increased over the past 5 years and absent a policy response; this trend is likely to continue. Non-competes are no longer limited to highly paid executives but now apply to about one in five Australian workers, across income, age, occupational, and education groups.

The distribution and prevalence of non-competes in Australia are broadly consistent with data in other developed economies. A number of jurisdictions within the Organisation for Economics Co-operation and Development have imposed restrictions on the use of non-competes. The US Federal Trade Commission is considering a ban on their use and, in the UK, the government has announced its intention to limit the term of non-competes to 3 months. In Australia, the Competition Minister has recently asked the Australian Competition and Consumer Commission and Treasury for advice on the competitive aspect of non-competes. The Australian Government released an Issues Paper (https://treasury.gov.au/sites/default/files/2024-04/c2024-514668-issues-paper.pdf#page40) – ‘Non-competes and other restraints: understanding the impacts on jobs, business and productivity’ in April 2024. The Issues Paper outlines the existing research and evidence in Australia and overseas on the use and effects of non-competes. A public consultation process will conclude in May 2024 and Treasury will advise the Government on the outcomes of the consultations by the end of 2024.

After reviewing the arguments for and against restricting the ‘reach’ of non-competes, I conclude that the weight of the evidence favours a regulatory response to ameliorate the unfairness inherent in the existing law and practice. A number of possible regulatory responses are considered.

About non-competes

Background

Non-competes are terms in an employment contract which provide that once the employment ends the employee cannot work for another employer in the same industry or field (i.e., a competitor), within a specified geographic area, for a specified time.

Non-competes have become increasingly common, in part because they are easier to enforce than other post-employment restraints dealing with confidential information and customer solicitation. Further, as noted by Stewart et al (2016), ‘This is a strategy that has been overtly encouraged by courts in recent years’. ¹

Like other post-employment restraints, non-competes are subject to the common law doctrine of restraint of trade ² and as such are presumptively invalid on the basis that they are contrary to the public interest. To rebut that presumption, the employer benefiting from the non-compete must establish that the restriction imposed is no wider than is reasonably necessary to protect a legitimate interest.

In theory, the practical application of the doctrine should seek to balance three overlapping and competing interests:

1. 1. Employers have an interest in protecting their businesses.

2. 2. Former employees have an interest in plying their profession or trade.

3. 3. The public has an interest in the promotion of competition and the efficient allocation of economic resources.

A non-compete which does nothing more than protect the original employer from competition from a former employee is unenforceable as it is the very thing the doctrine prohibits. ³ However, a non-compete is lawful if the employer can point to an interest they are entitled to protect, such as confidential information or customer connections and if the restraint goes no further than is reasonable to protect that interest. ⁴

In practice, the common law’s primary concern is to assess whether the employer has a legitimate interest and to determine if the non-compete ⁵ is commensurate with that interest. The employees’ position is generally treated as irrelevant. ⁶ As Stewart notes, ‘it is rare for a court to concern itself in any detail with the relative bargaining power of the parties; or with the overall “fairness” of the agreement; nor is it necessary that the employee receive any additional consideration for entering into the restraint’. ⁷ Consequently, in Australian courts, the majority of non-competes are upheld. ⁸

The doctrine of restraint of trade is also intended to protect the public from the anti-competitive consequences of private contractual arrangements. Yet the public interest is often forgotten in contemporary judgments in which the central focus is on the interests of the employer seeking to enforce the non-compete. ⁹ As Stewart (1997 at 182) notes: ‘in practice, reasonableness in the public interest is almost never considered as a distinct issue and it is hard to find a single case in which a restraint has been struck down on that ground’.

Even those who contend that the common law, if properly applied, strikes an appropriate balance between the post-employment interests of employers and their former employees, acknowledge that ‘there is a tendency in some recent cases to give greater weight to the terms of the contract as opposed to the public policy that militates against enforcement’. ¹⁰

The critique of non-competes can be reduced to four broad issues:

1. (i) confusion and uncertainty is created by the use of overly broad non-competes, particularly the use of ‘stepped’ or ‘laddered’ clauses (Confusion and Uncertainty);

2. (ii) ready access to injunctive relief disadvantages employees and fails to take sufficient account of the public interest (Injunctive Relief);

3. (iii) the use of non-competes has become widespread including low-wage occupations, such as hairdressers and the types of interests protected by a non-compete have expanded (Proliferation of Non-competes); and

4. (iv) non-competes reduce competition and job mobility and have consequential negative impacts on wages and productivity (Economic Consequences).

Confusion and uncertainty

To enforce a non-compete, the employer must show that the restraint is reasonable in the sense that it is commensurate with a legitimate interest and is no wider than strictly necessary to protect that interest. The ‘reasonableness’ of a non-compete is assessed having regard to the three elements common to such terms:

• the nature and extent of the particular activities sought to be restrained;

• the geographic area in which those activities must not occur; and

• the duration of the restraint.

There are no set rules about what constitutes a ‘reasonable’ geographic area or a ‘reasonable’ duration and the scope of each element is considered having regard to the other elements of the non-compete. Each non-compete must also be assessed with particular reference to the legitimate interest the employer is seeking to protect. Whether a court upholds the validity of a non-compete is dependent on the facts in each case and past cases provide little guidance. ¹¹

The subjective nature of the test has been the subject of academic criticism; for example, Riley describes it as a ‘very flabby test’ ¹² and Pivateau (2008) describes the test as ‘a standard [that] holds minimal value in the construction of non-compete agreements’. ¹³ Fell and Rudz (2023:1254) argue that the test has given rise to ‘inconsistency in the enforcement of post employment non-compete restraints…to the detriment of employees’. One court characterised the task of assessing the reasonableness of a non-compete as ‘a swampy morass of conflicting interests and policies into which a court may eventually need to plunge to resolve the problems these covenants present’. ¹⁴

The fact-dependent nature of the proceedings and the general nature of the test give rise to a diversity of judicial opinion. An authority can usually be found to support any particular point of view; as is apparent by comparing the cases relied on by Riley (2012) and Stewart (1997), as opposed to Neil and Saady (2018).

In addition to the uncertainty generated by the very nature of the test and its application in particular cases, the ability of courts to ‘sever’ offending aspects of a non-compete creates an additional layer of uncertainty.

At common law, an invalid part of a non-compete may be severed if a ‘blue pencil’ can be run through the offending part without affecting the original nature of the clause and contract. ¹⁵

The capacity to sever offending clauses encourages employer overreach in drafting non-competes and increases the use of overly broad clauses which, absent the blue pencil doctrine or, in New South Wales, the Restraints of Trade Act, would be unenforceable.

It has also become common for non-competes to be drafted to facilitate the process of severance by the adoption of ‘cascading’ or ‘laddered’ non-competes. Fell and Rudz (2023 1270) describe the practice as ‘ubiquitous’. Such terms set out a series of overlapping or cumulative restraints comprising multiple versions of different non-competes which vary as to geographic breadth, duration, and/or activities affected. Any restraint that is too wide can be severed, leaving what remains enforceable.

Cascading or laddered non-competes have been the subject of judicial ¹⁶ and academic ¹⁷ criticism for creating uncertainty. For example, Stewart (1997 at 218) argues:

Employers and other covenantees should be compelled to be clear as to what activities they wish to restrain, in what location and for what period. If they exceed the limits set by the law, and find that they cannot enforce the restraint in the face of conduct that could on any basis have been the subject of a reasonable covenant, then so be it. That is the price to be paid for taking insufficient care or being overly ambitious as to the scope of the restraint. The risk of losing out in that way surely does no more than balance the natural advantage that most employers enjoy through superior resources, access to legal advice and the intimidatory effect of the mere presence in a contract of a restraint, valid or not.

The problem for an employee bound by a laddered non-compete is that, absent a judicial determination, they have no way of knowing which of the restraints will be found to be reasonable and which go too far and be unenforceable. The impact of such uncertainty is amplified by the inequalities between the employer and employee.

Neil and Saady (2018) contend that the uncertainty inherent in the current law is a testament to its fairness:

The lack of uniformity in enforcing restraints illustrates the fairness of the test, as it ensures restraints are enforced in a manner that balances employer and employee interests in each case. While some judgments may give the impression of “elasticity”, most differences of approach result from judges attempting to apply the law to the commercial context of each case. No result will ever be the same because the facts of each case are unique.

However, this view ignores the chilling effect created by the uncertainty inherent in the application of the current law. The cost of legal proceedings, in financial and emotional terms, presents a significant deterrent to employees seeking to challenge the validity of a non-compete. ¹⁸ In terms of the financial cost, the legal practitioners interviewed by Arup et al in 2013, put the cost of opposing an application for an interlocutory injunction at between $20,000 and $100,000. Such costs are a significant disincentive for most employees.

For many employees, the mere threat of legal action by the employer is enough to secure compliance, whatever the lawfulness of the non-compete, as Stewart has observed:

One should not underestimate the in terrorem impact of invalid covenants on employees who, even if they have access to legal knowledge or advice, possess insufficient resources to pursue a legal challenge to those restrictions. ¹⁹

An in terrorem effect occurs when individuals adhere to a contractual term in anticipation of legal action if they choose not to comply. In the case of non-competes, the in terrorem effect is such that former employees comply with more non-competes than would be upheld if they went to trial ²⁰ and settlements are generally made before proceedings are issued. ²¹

Starr et al (2020) provide empirical evidence of an in terrorem effect, including with respect to unenforceable non-competes. Using nationally representative data for 11,500 US labour force participants, the authors found that non-competes are associated with reductions in employee mobility and changes in the direction of that mobility (i.e. towards non-competitors) in States that do and do not enforce non-competes. Indeed, they found no statistical differences in mobility patterns between States where non-competes are enforceable and States where they are not. ²² When asked how important their non-compete was in determining whether or not to leave their employment and go to work for a competitor, some 58.1% of respondents in States that do not enforce non-competes answered that it was ‘somewhat, very or extremely important’, compared to 53.5% of respondents from states that do enforce non-competes. The authors conclude: ‘a non-compete is associated with both a longer tenure and a reduced propensity to leave for a competitor even when the non-compete in question is unenforceable under state law’. ²³

In sum, uncertainty is an inherent by-product of the current legal framework for assessing the validity of non-competes. The generality of the legal test and the fact-dependent nature of the proceedings lead to a diversity of judicial opinion. This uncertainty is compounded by the ‘blue pencil’ approach to severance which encourages employer overreach in drafting non-competes and has given rise to cascading non-competes. Consequently, in most cases, the parties cannot be certain that their particular non-compete will be enforceable; absent a judicial determination. This uncertainty weighs more heavily on employees and, for many, the mere threat of legal action is enough to secure compliance, irrespective of the legal merits or enforceability of the particular non-compete. The US evidence suggests that in States where non-competes are unenforceable, employers still use them; they have an in terrorem effect on employees and are associated with reduced employee mobility. The unfairness of such an outcome is manifest and is contrary to the public interest.

Interlocutory injunctive relief

The initial enforcement action in respect of a non-compete is typically an interlocutory proceeding for injunctive relief. The interlocutory decision effectively determines the matter as very few cases proceed from the interlocutory to the trial stage. ²⁴ This is particularly problematic for defendants, as a former employer need only show that it has an arguable case and the balance of convenience favours injunctive relief. ²⁵ Few cases give serious consideration to whether damages would constitute an adequate remedy such that injunctive relief ought not be granted. ²⁶

The role of injunctive relief in the enforcement of non-competes has been the subject of academic criticism. For example, Riley ²⁷ argues:

There is an argument for introducing more rigid requirements for assessing the balance of convenience in these cases. In defamation cases, where the policy of the law is to guard freedom of speech jealously, courts have had no difficulty in finding that the balance of convenience almost invariably favours refusal of an injunction restraining the allegedly defamatory imputation. ²⁸ Perhaps it is time to identify a similar presumption that applies in cases where the enforcement of a negative covenant would prevent an individual from pursuing a chosen career. At least our courts might adopt the approach taken by courts in New York faced with similar questions:

“A party seeking the drastic remedy that a preliminary injunction confers must establish a clear legal right to that relief under the law and upon undisputed facts set forth in the record

… To prevail on a motion for preliminary injunction relief, the movant must clearly demonstrate a likelihood of success on the merits, the prospect of irreparable harm or injury if the relief is withheld and that a balance of the equities favours the movant’s position.” ²⁹

Arup et al make a similar point, proposing that ‘it would be preferable for the courts to require the employer to make out a stronger case before an interlocutory injunction is granted’. ³⁰

In sum, injunctive relief is typically used to enforce non-competes and few cases proceed to the trial stage at which the merits of the employer’s case can be fully tested. In interlocutory proceedings, the employer need to only establish an arguable case and that the balance of convenience favours injunctive relief. A review of the relevant cases suggests that the balance of convenience is weighed in the employer’s favour.

Proliferation of non-competes

The ‘reach’ of non-competes has expanded in recent times, both in the number and type of employees covered; and the type of interest said to be protected by a non-compete.

Australian data

The most recent Australian data on the use and prevalence of employment restraints is from the ‘Short Survey of Employment Conditions’ (the short survey) conducted by the Australian Bureau of Statistics (ABS) in late 2023. Utilising the Employee Earnings and Hours responding sample, about 7000 businesses were invited to answer a series of questions on employment restraint clauses. While not compulsory, the survey achieved a relatively high response rate of around 70%. ³¹

The results of the short survey were released on 21 February 2024, ³² revealing that about half (46.9%) of Australian businesses reported using at least one type of restraint clause in their employment contracts; with 20.8% using non-competes and 18% using clauses prohibiting the solicitation of co-workers.

The results also suggested that the use of restraint clauses was not limited to highly paid executives; rather they were used somewhat indiscriminately. Most businesses that used non-competes (68.2%) or an employment term prohibiting the solicitation of co-workers (67.2%) reported that the clause applied to between 76 and 100% of their employers.

The data suggest there is an increasing tendency to use restraints and that this trend is likely to continue. While the majority of businesses reported that their use of non-competes and terms prohibiting the solicitation of co-workers had generally remained the same over the past 5 years, a higher proportion of employers reported an increased use than a decrease. Of businesses currently using non-competes, 9.3% reported an increase in the use of such terms over the previous 5 years and only 1.9% reported a decrease. Of businesses currently using non-solicitation of co-worker restraints, 8.4% reported an increase and only 1.1% reported a decrease. Further, in response to a question regarding the expected future use of restraint clauses, about one in five businesses (19.9%) who do not currently use a non-compete reported that they would include such a term in future employment contracts. There was a similar response (22.5%) in respect of terms prohibiting the solicitation of co-workers. Such terms are sometimes referred to ‘non-recruitment’ restraints.

The ABS short survey does not report the proportion of employees bound by non-competes or terms prohibiting the solicitation of co-workers; it only provides data on the proportion of businesses using such restraints. A subsequent e61 publication ³³ uses ABS data to provide a lower, central, and upper estimate of the proportion of employees subject to particular restraint clauses. The authors estimate that roughly 20 to 25% of Australian workers are subject to non-compete and non-solicitation of co-worker terms, with a central estimate of 21 and 23%, respectively.

The ABS report that non-competes were used across all industry sectors with the highest use in the services sector, in particular financial and insurance services; rental, hiring, and real estate services; electricity, gas, water, and waste services; and administrative support services.

The Arup et al study is broadly consistent with the results reported by the ABS and notes that: ³⁴

Some sectors of the economy are more strongly represented than others. The interviewees suggested that the financial services sector has figured most prominently in recent years, as well as other service sectors including real estate, recruitment, retail, distribution, medical services, information, insurance, media, and hairdressing. ³⁵

To that list, one could add horticulture workers. A standard form contract used by labour-hire firms to engage ‘contractors’ to perform work (such as strawberry picking) in the horticulture industry in Victoria contains the following terms:

Post-agreement restraints non-compete ³⁶

13.

1. (a) For a period of 1 month after the Contractor’s engagement with the Company has been terminated for whatever reason, the Contractor agrees that it will not, within Victoria, be engaged as an employee, independent contractor, adviser, or in any other capacity in any business which, in the reasonable opinion of the Company, is in competition with the Company.

2. (b) The Contractor acknowledges that any breach by the Contractor of this clause would cause irreparable harm and significant damage to the Company and accordingly that the Company has the right to seek and obtain immediate injunctive relief in relation to any such breach.

3. (c) The Contractor acknowledges that the covenants in respect of non-competition contained in this clause are fair and reasonable and that the Company is relying upon this acknowledgement in entering into this agreement.

Plainly, non-competes are no longer confined to highly paid senior executives, as Riley (2012) observes:

By the power of the word-processed precedent document, restraints that were once considered appropriate only to preserve the value of goodwill purchased from a business owner are now appearing in contracts for moderately paid salary earners. ³⁷

It is also important to note that non-competes are not confined to individual employment contracts – they are also a feature of a number of enterprise agreements approved under the Fair Work Act 2009 (Cth) (the FW Act).

The inclusion of non-competes in enterprise agreements is an important development as they provide a mechanism for expanding the number of employees covered by non-competes. Some 3993 agreements, covering 700,000 employees, were approved by the Fair Work Commission (FWC) in 2022/23. ³⁸

Further, unlike employment contracts, enterprise agreements do not require the consent of each employee covered by the agreement. An agreement is ‘made’ when a majority of the employees who will be covered by the agreement cast a valid vote to approve the agreement. Further, once an agreement is approved by the FWC, it applies not only to those employees who had an opportunity to vote on the approval of the agreement, but to all future employees who fall within the scope of the agreement. ³⁹ Such ‘future employees’ will have no say in whether they agree to be bound by the non-compete term.

The 1070 single employer enterprise agreements approved between 1 July and 30 September 2023 were searched using the following key words and phrases: ‘confidentiality’; intellectual property’; ‘restraint’; ‘restraint of trade’; and ‘restriction of trade’. Only the first 100 results were examined in detail and post-employment restraints were found in 8 of those 100 agreements. Two of these agreements contained non-competes. ⁴⁰

Despite the presence of post-employment restraints (such as non-competes) in enterprise agreements, it is likely that such terms are unenforceable.

While not free from doubt, it is likely that a court would find that a term seeking to impose a post-employment restraint (such as a non-compete or a term prohibiting the solicitation of former co-workers term) is not a matter ‘pertaining to the relationship between an employer … covered by the agreement and the employer’s employees … covered by the agreement’, and hence is not a ‘permitted matter’ and is of no legal effect.

A consequence of the present legal framework is that a non-compete may be included in an enterprise agreement despite the fact that the term will likely have no legal effect. The potential for such a term to have an in terrorem effect on the employees covered by the agreement is obvious. There is nothing on the face of the agreement to indicate that the non-compete term has no legal effect. In all likelihood, the employees covered by such an agreement would assume that the non-compete term is valid. This is an entirely unsatisfactory state of affairs.

Data from other developed economies

The above observations about the distribution and prevalence of non-competes in Australia are broadly consistent with data in other developed economies.

Using nationally representative employee US survey data, Starr et al (2021) found that non-competes are more likely to be found in high-skill, high-paying jobs but were also common in low-skill, low-paying jobs and in States where non-competes are unenforceable. As to the prevalence of non-competes, Starr et al conclude that ‘38.1 per cent of US labour force participants have signed a non-compete at some point in their lives and that 18.1 per cent … currently work under one’. ⁴¹ These results may underestimate the prevalence of non-competes in the US as they are based on a survey of employees who may not know or recall if they are subject to a non-compete.

In the UK, the Department for Business and Trade estimates that non-competes ‘are widely used across the labour market, with around 5 million employees [out of a total workforce of 30 million; so about 17 per cent] subject to a non-compete clause in Great Britain, with a typical duration of around 6 months’. ⁴²

Non-competes are also common in Europe. For example, in the Netherlands, around 19% of employees are covered by non-competes ⁴³ and in Austria, non-competes apply to over 35% of private sector workers. ⁴⁴

The interests protected by non-competes

Historically the type of interests legitimately protected by a non-compete have been trade secrets, confidential information, and ‘customer connections’. ⁴⁵ More recently, courts in England and Australia have also been willing to accept that an employer has an interest in a ‘stable workforce’ sufficient to justify a reasonable restraint prohibiting the solicitation of co-workers. ⁴⁶

In summary, the ‘reach’ of non-competes in the Australian workforce is extensive and is no longer confined to highly paid executives. Non-competes apply across income, age, occupational, and education groupings. Further, a consequence of the present legal framework is that a non-compete can be included in an enterprise agreement, despite the fact that the term will likely have no legal effect; such a term will however have an in terrorem effect. The prevalence of non-competes in Australia is broadly consistent with data in other developed countries. There has also been an expansion in the type of interests which may be legitimately protected by a non-compete to include an interest in a stable workforce – which provides the legal basis for terms which prohibit the solicitation of former co-workers (sometimes referred to as no poaching restraints).

Economic consequences

A number of Australian labour law scholars have criticised non-competes as ‘legal handcuffs’ that reduce competitiveness, labour mobility and, hence, productivity. ⁴⁷ The impact of non-competes on productivity is particularly important.

Productivity matters; labour productivity growth is the main driver of higher living standards over the medium term. In recent decades, in Australia and other advanced economies, growth in labour productivity has slowed markedly. If sustained, lower productivity growth will negatively impact living standards and will mean lower potential growth.

The research literature consistently highlights two factors associated with Australia’s productivity slowdown:

1. 1. Non-mining investment was weaker than expected over much of the 2010s despite declines in interest rates and the end of the mining investment boom. The decline reflected a broad-based weakness in average investment across firms and industries (Hambur & Jenner, 2019). Generally speaking, if the amount of capital available to workers increases – capital deepening – they will be able to produce more output and hence be more productive.

2. 2. There has been a decline in business and economic dynamism in Australia (Deutscher, 2019; Andrews & Hansell, 2021; Andrews et. al 2022) and overseas (Calvino et. al 2020), as evidenced by the slower adoption of new technologies (Andrews et al 2019; Akcigit & Ares, 2019); declining firm entry rates (Gutierrez & Phillippon, 2019); and a slowdown in resource allocation efficiency. Job mobility and labour reallocation rates have declined, ‘trapping’ scarce labour in lower productivity firms. Andrews & Hansell (2021) found that the pace of labour reallocation from low-to-high-productivity firms has slowed since 2012, lowering labour productivity growth by around 0.15 percentage points per year in the mid-2010s compared to the early 2000s and accounting for about a quarter of the slowdown in non-mining non-financial private sector labour productivity growth from 2012 to 2014.

Non-competes are associated with reduced employee mobility and with changes in the direction of that mobility (ie towards non-competitors); ⁴⁸ with consequential negative impacts on wages and productivity. A recent e61 research note ⁴⁹ examined the relationship between job mobility and firm entry rates and the prevalence of employment restraints at an industry level in Australia. The authors found that:

1. 1. Job mobility appears to be lower in industries where the use of restraint clauses is higher.

2. 2. There is a negative correlation between the current prevalence of restraint clauses and the change in firm entry rates over the past 15 years.

Prohibiting or restricting non-competes enhances employee mobility and knowledge spillovers between firms. Gilson (1999) examined the economic effect of the law with respect to non-competes by comparing two high-technology industrial districts, California’s Silicon Valley and Massachusetts’ Route 128. He contends that the continued success of Silicon Valley and the decline of Route 128 is due to the fact that California prohibits non-competes and Massachusetts enforces them.

A paper by Marx et al (2009) examines the impact of non-competes on employee mobility by using Michigan’s 1985 reversal of its non-compete enforcement policy as a natural experiment. The authors found ‘a strong decrease in average Michigan mobility once non-competes began to be enforced’. ⁵⁰

A recent paper published by the US National Bureau of Economic Research (Gottfries & Jarosch, 2023) assessed the consequences of an outright ban on non-competes in the US labour market using a new model on wage-posting and on-the-job search, drawn from the Burdett–Mortenson model. The model used was validated by testing with existing empirical evidence on firm mergers and non-competes. The model predicts that wage gains are typically 2% to 6% but can be as high as 15% when non-competes are common and hiring costs are higher.

The baseline analysis by Gottfries & Jarosch (2023) suggests that a ban on non-competes between workers and firms in the US would increase wages by 4% as a consequence of the rise in competition, with strong wage spillovers to firms that did not initially use non-competes and a large increase in worker turnover. The authors also note that while a ban on non-competes is likely to increase wages and reduce labour misallocation, it also generates a sharp rise in worker churn, increased turnover costs, and reduced labour demand.

Similarly, Johnson et al (2023) posit that average earnings of all US workers would likely increase by 3.2% to 14.2% (midpoint of the interval: 8.7%) nationally, if non-competes were made unenforceable.

An outlier is Young (2021), who found that a ban on non-competes in Austria for low-wage workers was associated with a 1.7% increase in annual job-to-job transition rate, disproportionally to higher quality and higher paying jobs but had limited effects on earnings. It is worth noting that the ban on non-competes in Austria only applied to those earning below the 52nd percentile of the earnings distributions.

As mentioned earlier, the slowing in Australia’s productivity performance in recent decades has coincided with a decline in job-to-job transitions. ⁵¹ Job-to-job transition, sometimes referred to as job switching, occurs when a worker changes jobs without going through a period without paid employment. Andrews and Hansell (2021) suggest that the observed decline in productivity-enhancing labour reallocation is holding down overall productivity growth and is consistent with a rise in adjustment frictions. Non-competes are a source of friction in labour markets, creating a barrier to labour mobility.

International research suggests that job-to-job transitions can increase aggregate productivity by moving workers to more productive firms. ⁵² In the Australian context, Buckley (2023) finds that the share of job switching that involves workers moving to higher productivity firms has fallen over time, from 54.2% between 2003 and 2006 to 52.8% between 2015 and 2019. ⁵³ Buckley contends that making it easier for workers to switch jobs could boost productivity:

Australian workers currently face a number of barriers when looking to move to a more productive firm. Some of these barriers have gotten worse in recent years. Non-compete clauses (NCCs), for instance, have become more prevalent over the last 15 years and are now a default option in many employment contracts. Removing or limiting the use of NCCs would help remove one source of friction behind the decline in job mobility. ⁵⁴ (citations omitted).

Competition among firms for employees, as reflected in higher job-to-job transition rates, is an important driver of higher wages. In the Australian context, Deutscher (2019) found that a 1 percentage point increase in the rate at which workers switch jobs is associated with a 0.5 percentage point increase in growth in average wages.

A recent Australian study found that switching jobs is associated with pay increases that are nine percentage points higher on average than for non-switchers and that the largest beneficiaries are workers aged between 21 and 34 years old. When such workers switch jobs they earn, on average $7,500 more per annum than ‘job stayers’. ⁵⁵

The next section considers whether a regulatory response is needed to address the deleterious impact of non-competes and, if so, what policy options are available.

Policy options

Is there a need for a regulatory response?

The case in favour of a regulatory response to curtail the use of non-competes rests on two broad propositions. The first is that the existing law and practice in respect of the use of non-competes in Australia is manifestly unfair and contrary to the public interest. In most cases, the parties to a non-compete cannot be certain of its enforceability without a judicial determination and such uncertainty weighs more heavily on employees than employers. The uncertainty inherent in the application of the current law has a chilling effect and for many employees, the mere threat of litigation is enough to secure compliance, irrespective of the enforceability of the non-compete. Further, employers have ready access to injunctive relief to enforce non-competes, to the disadvantage of employees and the public interest. Few cases proceed to the trial stage, where the merits of the non-compete can be fully tested.

The second proposition in favour of a regulatory response is that non-competes reduce employee mobility with consequential negative impacts on wages and productivity.

The case against a regulatory limitation on non-competes rests on four broad propositions.

1. (i) Non-competes are necessary to protect trade secrets, particularly in high-skilled labour markets.

Two points may be made in response to this proposition.

First, the ‘reach’ of non-competes extends well beyond the highly skilled and highly paid. Starr et al (2021) report that 14.3% of US workers without a college degree and 13.3% of US workers with annual earnings of less than $40,000 are covered by a non-compete. A similar pattern emerges in Australia. About one in five Australian workers are subject to a non-compete and they are prevalent across different societal groups, including low-paid employees. In short, the ‘reach’ of non-competes is inconsistent with their stated justification; as Gottfries & Jarosch (2023) observe ‘it seems hard to argue that the millions of low-skilled, low-pay workers have access to important trade secrets’ (p22).

Second, there are other ways to protect trade secrets. An employer wishing to restrain the post-employment use or disclosure of confidential information can rely on the common law duty of confidentiality and the equitable doctrine of breach of confidence. As Stewart (1997) points out (at p 188):

the equitable obligation not to misuse information communicated or acquired in confidential circumstances falls upon any person into whose hands information comes as a result of another’s breach of confidence. Hence the duty of confidentiality may be enforced not only against the ex-employee but also against any third party (such as another employee, or rival firm set up by the ex-employee) who acquires the information as a result of the ex-employee’s breach.

It is acknowledged that these alternate means have limitations. The duty of confidentiality is only breached by the misuse of ‘secret’ information and may not cover information which is only ‘confidential’. ⁵⁶ Further, an injunction will only be granted to restrain the use of trade secrets by a former employee if the precise material to be covered by the injunction can be specified ⁵⁷ and to claim damages for such a breach the employer must show that any loss of business is attributed to the disclosure of the secret information. ⁵⁸

As mentioned earlier, one of the reasons that non-competes have become more common is that they are easier to enforce than other post-employment restraints dealing with confidential information and customer solicitation. Indeed courts have encouraged this approach, for example in Littlewoods Organisation Ltd v Harris ⁵⁹ the court said:

experience has shown that it is not satisfactory to have simply a covenant against disclosing confidential information. The reason is because it is so difficult to draw the line between information which is confidential and information which is not and it is very difficult to prove a breach when the information is of such a character that a servant can carry it away in his head. The difficulties are such that the only practicable solution is to take a covenant from the servant by which he is not to go to work for a rival in trade. ⁶⁰

1. (ii) Any further limitation on non-competes undermines the sanctity of contract.

This argument is advanced by Neil and Saady (2018), who contend that employees ‘consciously and freely chose to be bound’ by non-competes when they entered into a contract with their employer and refusing to enforce such a contractual obligation ‘undermines the authority and sanctity of contract’.

There are two problems with this argument. The first is that the common law doctrine of restraint of trade already ‘undermines the … sanctity of contract’ as non-competes are invalid unless the non-compete is no wider than reasonably necessary to protect a legitimate interest. Further, both the common law and a range of statutory provisions routinely imply and prohibit certain contractual terms.

The second, and more substantive point, is that the premise of the argument – that employees ‘consciously and freely chose to be bound’ by non-competes – ignores the inequality in bargaining power inherent in most employment relationships and is contrary to the available evidence.

Starr et al (2021) found that only 10% of employees report attempting to negotiate over the terms of a non-compete or asking for additional compensation or benefits in exchange for agreeing to such a condition. When presented with a non-compete, most respondents report just reading and signing it, and some not even reading it. ⁶¹ Starr et al (2021) conclude that ‘the evidence … indicates that employers present (or employees receive) non-compete proposals as take-it-or-leave-it propositions’. ⁶²

Further, as mentioned earlier, the proposition that employees ‘consciously and freely chose to be bound’ by non-competes is plainly wrong in respect of non-compete terms in enterprise agreements and their application to ‘future employees’.

1. (iii) Workers are better off being bound by a non-compete.

There are two elements to this argument, namely that workers bound by non-competes earn higher wages and that they receive additional training.

Whether or not workers are better off being bound by a non-compete is somewhat contentious. A number of US studies have found that workers bound by non-competes earn higher wages than those without non-competes. ⁶³ There is an obvious tension between these studies and those discussed earlier in Section 2.5 of this paper which report an increase in wages following the banning of non-competes. This tension is discussed by Starr (2023) who observes that those asserting non-competes result in higher wages are ‘confusing correlation with causation’:

There are many reasons why workers with non-competes may earn more than workers without non-competes that have nothing to do with the noncompete. For example, non-competes are more common for more educated workers, and more educated workers tend to have higher earnings. So it’s not surprising that non-competes are associated with higher earnings, but this may just have to do with the types of workers that agree to non-competes or the types of firms that deploy them.

Similarly, Balasubramanian and Yamaguchi (2023) find that positive wage relationships with non-competes are likely driven by selection and should not be interpreted causally and that a negative average wage effect is more likely.

Further, as Starr (2023) points out, if it is the case that enforcing non-competes is likely to lead to better outcomes for workers, we would expect non-competes to be associated with relatively higher wages where they are enforceable compared to where they are not. But this is the opposite of the results of Rothstein and Starr (2022) who find that the non-compete-wage differential in states that might enforce non-competes is 6% lower than in states that will not enforce them.

Starr (2023) convincingly concludes:

Commentators suggesting that non-competes cause higher wages are largely misinterpreting cross-worker correlations. Rather, recent economic evidence suggests that the positive correlations between non-competes and wages is likely spurious and that non-competes (potentially alongside non-solicits and NDAs) likely reduce earnings by increasing retention and shielding firms from labor market competition. These findings align with studies of state policy shocks which cover state-level bans on non-competes for low-wage workers, high-tech workers, and a variety of other changes to state non-compete laws.

The proposition that workers bound by non-competes receive more training is supported by two studies, which find that non-competes and their enforceability are associated with more training. ⁶⁴

While it may be accepted that non-competes are associated with the provision of additional training the reasons for that association are less clear. Starr et al (2018) suggest that one reason there is more training in US states where non-competes are enforceable is because non-competes prevent firms from hiring experienced workers thus causing them to hire inexperienced workers who require more training. ⁶⁵ Further, the existence of such an association does not mean that imposing further restrictions on the use of non-competes will necessarily reduce the extent of training provided to employees.

1. (iv) Non-competes foster innovation.

A number of commentators have suggested that non-competes give firms incentives to develop and share valuable information with workers that may make those workers more productive because they provide protection against their workers providing that information to a competitor. ⁶⁶ However, the evidence from recent studies suggest that non-competes do not foster innovation.

Johnson, Lipsitz, and Pei (2023) suggest that an average-sized increase in non-compete enforceability reduces patenting by 16–19% over the ensuing 10 years. Other recent studies report similar findings. Reinmuth and Rockall (2023) find that an average increase in the enforceability of non-competes ⁶⁷ reduces patenting by 11.6% after 5 years.

In summary, there are factors weighing for and against curtailing the ‘reach’ of non-competes and the type of interest which may be protected by post-employment restraints, but the weight of the evidence favours a regulatory response to ameliorate the unfairness inherent in the existing law and practice with respect to the use of non-competes.

Despite the fact that the doctrine of restraint of trade is intended to protect the public from the anti-competitive consequences of non-competes and to take account of the interests of the affected worker, these considerations are often forgotten in contemporary judgments in which the central focus is on the interest of the employer. A regulatory response is required to correct this imbalance.

The options

A number of jurisdictions within the Organisation for Economics Co-operation and Development have imposed restrictions on the use of non-competes ranging from the general prohibition of such terms to more limited regulatory interventions, such as the payment of mandatory compensation to the former employee (by their former employer), while a non-compete operates to prevent them working for a competitor. ⁶⁸

A number of reforms have been proposed in the US, including banning non-competes for some or all employees and regulating the non-competes contracting process, ⁶⁹ and in recent years, many States have passed laws limiting employers’ ability to impose non-competes on their employees. ⁷⁰

At present, the restriction of non-competes across the US is under active consideration by the Federal Trade Commission (the FTC).

In January 2023, the FTC released a ‘Notice of Proposed Rulemaking’ to prohibit employers from imposing non-competes on workers. The proposed rule would:

• provide that non-competes are an unfair method of competition and would ban employers from entering non-competes with their workers, including independent contractors; and

• require employers to rescind existing non-competes with workers and actively inform their employees that the non-competes are no longer in effect.

The FTC estimates that the proposed rule would increase workers’ earnings by nearly $300 billion per year. FTC staff are currently considering comments received on the proposed rule and will present any proposed revisions to the FTC for its determination.

The regulation of non-competes has also received attention in the UK where the government has announced its intention to introduce a statutory limit on the length of non-competes of 3 months. ⁷¹

Australian labour law scholars have suggested a range of measures to mitigate the deficiencies in the current legal framework. For example, Arup et al suggest placing an upper limit on the duration of non-competes and requiring employers to pay their former employees for the period the non-compete operates. ⁷² They also propose a policy of ‘no modification of restraints – or, at least only modification to make good the restraint if the parties can show they had genuine doubts about the appropriate restraint to apply in a particular position’. ⁷³

Riley (2005) argues that clauses which prohibit the solicitation of former co-workers be ‘entirely unenforceable’ and that more rigid requirements be introduced for assessing the balance of convenience in applications for injunctive relief to enforce non-competes. Arup et al (2013) suggest that a ‘cleaner reform… would be to withhold interlocutory injunctive relief altogether from employers and require them to go to trial (on an expedited basis) to plead the merits of the restraint’. ⁷⁴

In April 2024, the Australian Government released an Issues Paper, ‘Non-Competes and other restraints: undertaking the impact on jobs, business and productivity’. ⁷⁵

The Issues Paper outlines the existing research and evidence in Australia and overseas on the use and effects of restraints of trade on workers, and no-poach and wage-fixing agreements made between businesses. Several issues are identified relating to the use and impact of non-compete clauses including:

• the ‘chilling effect’ of restraint clauses on worker mobility, particularly among lower-income workers, to choose better-paying jobs, and the ability for businesses to start up, recruit talent, and grow;

• the high cost of litigation, the lack of clear guidance and ‘bright line’ rules, and the use of cascading clauses or the ‘blue pencil test’, which can leave both workers and businesses with an unclear understanding of whether an agreed restraint will be upheld as reasonable and enforceable; and

• the economic consequences of potentially inefficient allocation of labour and information, which may be hampering productivity growth and innovation.

The public consultation process concluded in May 2024 and Treasury will advise the Government on the outcomes from the consultations by the end of 2024.

Some potential reform options in the Australian context are discussed below.

The various policy options canvassed can be located on a two-axis matrix, depending on the complexity of the option and the extent to which it is likely to encounter resistance (see Figure 1).

Figure 1. Policy options – complexity vs likelihood of resistance.

Monetary thresholds

A common regulatory response is to render unenforceable non-competes with employees who earn less than a prescribed amount. The argument advanced in favour of such measures is that ‘low-wage’ non-competes are unfair, as Hiraiwa et al (2023) put it:

‘low-wage workers typically do not benefit from the investments that firms use to justify [non-competes], often do not review their [non-compete] and may be unable to afford legal protection to defend against even frivolous [non-compete] lawsuits.’ ⁷⁶

To these arguments, one could add that ‘low-wage’ employees often lack the bargaining power to resist a non-compete or to negotiate its terms.

Non-competes with ‘low-wage’ employees are unenforceable in 11 States in the US, ⁷⁷ but there is no particular consistency in the threshold chosen or the method for determining that threshold. The criteria adopted in each state are summarised in Table 1. ⁷⁸

Table 1. Non-competes limited by wage threshold

The selection of an appropriate non-compete enforcement threshold in the Australian context is a policy choice.

As shown in Table 1, the higher the threshold, the more jobs affected.

Chart 1. Share of Australian jobs earning below any given level.

The ‘high income threshold’ is a concept used in the FW Act. If an employee (defined as a ‘high income employee’) has a guarantee of annual earnings which exceed the high-income threshold, the unfair dismissal provisions of the FW Act do not apply to that employee (s.382). The high-income threshold is indexed and is currently $167,500. If that figure was used as a non-compete enforceability threshold, it would have the practical effect of prohibiting the enforcement of non-competes in over 90% of Australian jobs. A threshold set at ‘average earnings’ covers just under 80% of jobs.

The evidence from Hiraiwa et al (2023) suggests that a monetary threshold, which covers about 80% of workers, does not reduce average firm value and firms do not appear to value the enforceability of non-competes for workers earning around the threshold.

In the event that an enforcement threshold is adopted, it should be indexed.

Term limitations

Regulatory responses to non-competes can include the imposition of a limit on their duration; a requirement that the former employee be compensated for the duration that the non-compete applies; or a combination of these two measures.

As mentioned earlier, the UK has announced an intention to introduce a 3-month term limit to all non-competes. In Germany, employers are obliged to pay the restrained employee 50% of their remuneration for the entire restraint period (a maximum of two years).

The advantage of a requirement that the former employee be compensated for the duration of the non-compete is that it, in effect, applies a market solution to the problem. The employer is required to place a value on the restraint and to pay that amount to the worker who is subject to the restraint.

An appropriate policy intervention may be to introduce a non-compete term limit of up to 6 months subject to the restrained employee being compensated for the duration of the restraint.

Such a term limit broadly aligns with how Australian courts have actually applied the reasonableness standard in particular cases. Fell and Rudz (2023) analysed all Supreme and Federal Court decisions between 1 July 2016 and 22 February 2023 in which the court made a final (as opposed to interlocutory) determination about the reasonableness of the duration of a non-compete. As the authors caution, the determination that a particular duration was reasonable in one case does not mean the same conclusion automatically applies in a different context, the relevant considerations are highly fact-dependent. Nevertheless, the results indicate that long periods are more likely to be unenforceable:

• all 24 months restraints were unenforceable;

• only half of the 12-month restraint were enforceable and nearly all of those involved senior employees were likely to possess significant confidential information, form connections with significant clients, and receive substantial benefits under their employment contract.

• in no case was a 12-month restraint upheld against an employee who had no significant management or strategic responsibilities.

• 6-month restraint are likely to be enforced against senior employees but likely not enforceable against non-senior employees.

The amount of compensation provided also presents a policy choice – ranging from the remuneration the employee would have received had the employment relationship continued (i.e. full compensation), to some proportion of that amount (partial compensation). Any determination of the appropriate amount of compensation should consider:

1. ➢ the fact that the former employee will not be performing any work for their former employer during the restraint period; and

2. ➢ the former employee will not be able to commence working for a competitor (or to operate their own business) during the restraint period. Further, job switching usually results in higher remuneration and so the former employee will be denied access to that higher remuneration for the duration of the restraint.

Consideration should also be given to the interaction between the compensation paid during the restraint period and any payments in lieu of notice of termination or any ‘gardening leave’ arrangement. ⁷⁹

Injunctive relief

The criticism of the current law with respect to injunctive relief is that it is too readily available to employers seeking to enforce non-competes, to the detriment of the public interest, and the employee to whom the non-compete applies.

The reform options range from prohibiting access to injunctive relief altogether, which would require employers to go to trial (on an expedited basis) where the merits of the non-compete can be fully tested; to ‘lifting the threshold’ for the grant of injunctive relief.

As Arup et al (2013) note, prohibiting access is a ‘cleaner reform’ but would be a much bigger step than requiring employers to make out a stronger case before injunctive relief is granted or ‘tilting’ the balance of convenience requirement towards the affected employee.

A middle path may be to expressly require courts to take into account the public interest in the promotion of competition and the interests of affected employees in deciding whether the balance of convenience favours the grant of injunctive relief. There is, however, some complexity and risk in constraining judicial discretion. Depending on the other policy responses adopted, it may not be necessary to pursue this particular option.

Prohibition

There are a range of options within the prohibition policy mode, including:

• prohibiting all non-competes;

• prohibiting particular aspects of non-competes – such as terms prohibiting the solicitation of former co-workers;

• prohibiting the severance of unlawful aspects of non-competes by the courts; and

• prohibiting the inclusion of non-competes in enterprise agreements.

Each of these options is considered below.

1. (i) Prohibiting all non-competes

Subject to limited exceptions, some States in the US provide that non-competes are unenforceable. For example, s.16600 of Californian Business and Professions Code states:

‘Except as provided in this chapter, every contract by which anyone is restrained from engaging in a lawful profession, trade or business of any kind is to that extent void.’ ⁸⁰

The compete prohibition of non-competes may undermine the capacity of employers to take reasonable steps to legitimately protect their business and accordingly is likely to encounter strong resistance.

1. (ii) Prohibiting employee non-solicitation terms

Riley convincingly argues that such terms should be ‘entirely unenforceable’, ⁸¹ on the following basis:

an employer has no legitimate interest in keeping staff. The case law (largely English single judge decisions ⁸² ) holding that an employer has a legitimate interest in maintaining a stable workforce is misconceived. These cases focus – illegitimately – on the value of a stable and trained workforce as some kind of business ‘asset’, but forget that a workforce is not a thing of property. Staff are legal persons. They are subjects, not objects; they are owners, not owned. As independent legal actors they enjoy a liberty to receive, consider and should they so desire accept offers of employment. No contract made between other persons should be allowed to limit that liberty. To allow otherwise is to tolerate what the common law doctrine of privity of contract has always prohibited: burdening people behind their backs with the consequences of another’s bargain.

There is a strong case for prohibiting the use of terms which prohibit the solicitation of former co-workers.

1. (iii) Limiting judicial discretion

The common law approach to severance encourages employer overreach in drafting non-competes and has led to the adoption of cascading or laddered non-competes.

Pivateau (2008) proposes that ‘courts put an end to the blue pencil doctrine… because it creates an agreement that the parties did not actually agree to [and] does nothing to address the underlying problems of non-compete agreements’. ⁸³ Pivateau also contends that employers should be required to nominate precisely the position from which the employee would be barred, the interest that is worthy of protection, and the extent of the restraint that is considered reasonable. ⁸⁴

Some US states such as Georgia, Virginia, and Wisconsin follow a ‘no modification’ approach to non-competes, also known as the ‘all or nothing’ rule. Under these provisions, courts must refrain from either rewriting or striking out terms in non-competes which are wider than strictly necessary to protect the employer’s legitimate interest.⁸⁵

There is a reasonable case for limiting the capacity of courts to sever unlawful aspects of non-competes, but such an approach brings with it some risk and complexity. It would, in effect, mean overriding parts of the Restraints of Trade Act and constraining the operation of a common law judicial discretion is not a task to be undertaken lightly.

Careful thought would need to be given to the drafting of such a provision. Depending on the other policy responses adopted, it may not be necessary to pursue this option.

1. (iv) Prohibiting the inclusion of non-competes in enterprise agreements

As mentioned earlier, the inclusion of non-compete terms in enterprise agreements is an important development. It is likely that such terms have no legal effect, as they are not about a ‘permitted matter’, but they have an in terrorem impact. On the understanding that non-competes are not ‘permitted matters’, it follows that the FWC cannot refuse to approve an enterprise agreement merely because it includes a non-compete. Nor can the FWC require an employer to give an undertaking in relation to a non-compete or amend an agreement to remove a non-compete.

One means of addressing this unsatisfactory state of affairs is to amend the FW Act, as follows:

1. 1. Amend s.194 to make a non-compete an ‘unlawful term’ (the FWC can only approve an enterprise agreement if it is satisfied that it does not contain any ‘unlawful terms’: s.186(4).

2. 2. Amend s.191A (or insert a new section) to provide that the FWC must amend an enterprise agreement to remove a non-compete term as part of the approval process.

These amendments would remove the risk that employees may mistakenly believe that a non-compete term in an enterprise agreement is effective and enforceable.

Conclusion

The existing law and practice with respect to the use of non-competes in Australia is manifestly unfair and contrary to the public interest. Non-competes create a barrier to labour mobility. Prohibiting or restricting non-competes is associated with increases in workers’ earnings and job switching. Job switching enhances knowledge spillovers between firms and is likely to increase labour productivity by moving workers to more productive firms. The weight of the evidence favours regulatory reform in respect of the use of non-competes in Australia.

Reasonable minds may differ about how a particular option is characterised and, of course, the devil is in the detail.

Taking all these matters into account, four particular policy interventions should be considered as an initial response to the issues with the current legal framework which have been identified in this paper.

1. 1. The adoption of an indexed monetary threshold for non-compete enforceability.

2. 2. Placing a term limit (up to six months) on non-competes in respect of employees who earn more than the monetary threshold, subject to the employer compensating the employee for the duration of the restraint.

3. 3. Render unenforceable terms which prohibit the solicitation/recruitment of former co-workers.

4. 4. Amend the FW Act to provide that non-competes and terms which prohibit the solicitation of former co-workers (sometimes referred to as ‘non-recruitment’ terms) in enterprise agreements are unlawful and require that they be removed as part of the approval process.

Consideration should also be given to extending the reforms beyond ‘employees’ to contractors and ‘employee-like’ workers given recent amendments to the FW Act.

The adoption of these policy responses would also address the problems inherent in ‘laddered’ non-competes. The imposition of a 6-month term limit coupled with a requirement to compensate the worker subject to the restraint effectively removes the incentive for a ‘laddered’ restraint.

In implementing a policy response, five other matters should be considered.

First, the simplest mechanism for implementing any of the policy options mentioned above would be by an amendment to the FW Act.

Second, the evidence suggests that employers regularly require employees to enter into non-competes that are unenforceable and that such (unenforceable) non-competes exert an in terrorem effect on employees and restrict mobility. ⁸⁶

This suggests that it is not enough to merely restrict the enforceability of non-competes. A policy intervention must also limit the use of non-competes by increasing the ‘cost’ of using unenforceable non-competes. ⁸⁷ One way of doing this would be to adopt a legislative response modelled on the pay secrecy provisions in ss 333B to 333D of the FW Act, namely:

1. ➢ provide that non-competes with employees who earn less than a prescribed amount (say $100 000 per annum) are unenforceable; and

2. ➢ expose employers to a civil penalty in the event that they enter into a non-compete with an employee who earns less than $100 000 per annum.

The same approach can be taken with respect to other policy options. For example, with respect to a term limit, any non-compete which purports to operate for longer than the prescribed limit would be unenforceable and employers entering into such-non-competes would be liable to a civil penalty.

Third, any regulatory response should consider providing an exemption in relation to the sale of a business or the dissolution of a partnership. Such exemptions are common in jurisdictions which limit or prohibit non-competes and recognise the fact that the purchaser in such transactions is often buying ‘good will’ and in such circumstances a reasonable non-compete is appropriate.

Fourth, depending on the regulatory responses adopted, some thought may need to be given to the interaction with relatively common contractual provisions, such as ‘gardening leave’ (Coulthard, 2009).

Finally, any regulatory response should be evaluated two years after implementation in order to assess its effectiveness and to determine if further measures are necessary.