Promoting Innovation or Exacerbating Inequality? Laboratory Federalism and Australian Age Discrimination Law

According to laboratory federalism, federal systems can promote governmental innovation and experimentation, while containing the risks of innovation to only one jurisdiction. However, it is unclear whether these benefits are realised in practice and whether states are actually effective ‘laboratories’. This article evaluates the extent to which laboratory federalism is occurring in practice, focusing on a case study of age discrimination law in Australia. Drawing on related ideas of democratic experimentalism; legal doctrinal analysis of age discrimination law in the Australian states and territories, and at the federal level; and qualitative expert interviews with 66 Australian respondents, I map the potential and limits of laboratory federalism in advancing age equality. I argue that, in this particular context, the benefits of experimentation may be outweighed by the resulting difficulties of enforcing age discrimination law, exacerbating inequality in practice. The federal structure has led to a confused and confusing patchwork of legal regulation. There is therefore a need for stronger federal structures to facilitate mutual learning and better realise the benefits of laboratory federalism.


I Introduction
The practical operation of Australian federalismits benefits, advantages and limits as a form of governmenthas been thrown into the public eye by the management of the COVID-19 pandemic. 1n a period of crisis, the benefits and flaws of government systems are thrown into stark relief.the needs and community expectations of particular states. 3Individuals have a louder 'voice' when engaging with state governments, who are therefore likely to be more responsive to constituents' concerns.Further, individuals can choose to 'exit' particular jurisdictions if policies are not to their liking. 4In a federal model, states can adopt different responses to the tensions between autonomy and security, liberty and privacy, dignity and efficiency; this potentially facilitates innovation and competition between jurisdictions, 5 as states compete 'to attract desirable citizens and businesses'. 6tates may also be able to act more agilely to emerging social concerns, meaning policy change can be more immediate and responsive. 7Finally, divesting power to the states can minimise the risk of tyranny 8 at the federal level.These represent substantial regulatory advantages over those offered by a unitary system of government.
While these benefits are associated with federalismand not laboratory federalism specificallythey are likely enhanced and supported by jurisdictional experimentation.In the United States, Justice Brandeis's long-quoted (perhaps optimistic) view of federalism is where 'a single courageous State may, if its citizens choose, serve as a laboratory; and try novel social and economic experiments without risk to the rest of the country'. 9This serves as the basis for ideas of laboratory federalism; that is, where states can experiment and test novel policy solutions, learn from the innovations of other states, and where the risks of (failed) experimentation are contained within one particular jurisdiction.This model engenders an 'image of policy percolating up and through the federal system', 10 eventually reaching the federal level once it is tried and tested at the state level.
Experimentation is arguably critical for realising the posited benefits of federalism as a model of governance.It can ensure diverse policies across states, adapted to the needs of particular jurisdictions.It also ensures 'exit' offers residents meaningful choices between jurisdictions.Laboratory federalism may also be particularly useful in addressing 'wicked' problems.'Wicked' problems emerge in policy areas that are highly resistant to resolution and have many interdependencies, where the problem is complex, unstable and difficult to define, and where there is no clear solution, thereby requiring a value-based management strategy, 11 'trade-offs between conflicting goals', 12 and difficult normative decisions to be made. 13Where there is no 'right' answer, devolving decisions to the local level can minimise and contain the damage caused by failed experimentation, 14 and can respond to differing values and priorities in each jurisdiction. 15In contemporary society, many of the problems faced by government are 'wicked' and resistant to resolution; thus, a federal structureand experimentationis even more pertinent for addressing modern governmental and policy challenges, including those posed by growing inequality.

B Risks of Laboratory Federalism
These potential benefits of federalism, and laboratory federalism in particular, are balanced by the risks or potential limits of experimentation.It is questionable whether states will actually innovate; laboratory federalism may not occur to a desirable or optimal degree in practice.The risk of 'exit' may actually stymie social policy reform: states may be unwilling to introduce progressive reforms that risk putting business off-side, in case they choose to leave the jurisdiction. 16States may lack the capacity or inclination to innovate: in the context of environmental policy, for example, Engel argues that states are better seen as 'scale innovators' than 'policy innovators'; most ideas originate at federal or international level, so rather than creating new ideas, states just adapt and scale initiatives to their own circumstances. 17Further, it may not be in states' interests to innovate: if the costs and risks of an experiment are internalised, and it is easy to imitate successful experiments, then it is in states' interests to let other states experiment, rather than doing so themselves. 18Oates describes this as 'a standard sort of incentive for free-riding'. 19his may just mean we need to adjust our expectations of what states as 'laboratories' means in practice.As Rom argues, the metaphor of states as laboratories should be taken 'seriously but not literally'. 20States mostly do not 'experiment': instead, they 'innovate', that is, adopt policies or programmes that are new to the states themselves, not necessarily new in a global sense. 21Policy innovations are not developed or designed as laboratory experiments, seeking to find 'good policy': they are designed to 'serve the interests of the politicians who design them'. 22Policies are driven more by ideology, tradition and passion than by theory and evidence. 23Further, innovations can be assessed for both their policy effectiveness (do they address the problem?) and political effectiveness (is it politically attractive enough to be implemented and sustained?). 24Thus, assessing policy innovations against non-political criteria applied to actual laboratories is inappropriate.
Even if states are willing to experiment, and laboratory federalism occurs, it is unclear how successful those experiments will be.As Gelbach notes, 'laboratories are only as good as the experiments their operators conduct' 25 (if the operators conduct any experiments at all), and those experiments then need to be evaluated in a meaningful way. 26Successful policy experimentation and evaluation can consume significant resources (both fiscal and human): expecting states to bear this burden may be unrealistic, particularly where they are responsible for a wide range of policy areas, with limited resources and ways of raising governmental revenue. 27onversely, too much innovation can jeopardise the ability to evaluate or study particular reforms, to see what works.As Gelbach concludes in the context of United States welfare reform, the uncontrolled experimentation of the last two decades has left us with a crazy-quilt collection of state programs that do not lend themselves to the kind of empirical research that would be necessary to benefit from the Brandeisian idea of states as laboratories of democracy. 28perimenting in multiple areas simultaneously can make programme evaluation difficult or impossible.Instead, effective evaluation requires a long-term, strategic and measured approach to policy reform, which may be uncommon (or unrealistic) in practice.
Further, if states are diverse in terms of their constituents' needs and expectations, as laboratory federalism suggests, then they are also likely to have diverse institutions, histories and political groups.This may make it difficult or inappropriate to transplant innovations from one state to another, 29 or impossible to isolate the factors that contributed to policy success or failure. 30Thus, serious questions must be asked about the transferability of policy lessons from one jurisdiction to another.
Experimentation also imposes additional administrative costs on businesses and those who operate across state borders, as they are forced to navigate and comply with different rules in each different jurisdiction. 31This is particularly a challenge in small countries like Australia, where the population in each state is generally small, making economies of scale limited or non-existent.
Finally, and more fundamentally, there may be areas in which state experimentation is inappropriate.As Conkle writes, 'these potential benefits of federalism [do not] outweigh the need for national protection and enforcement of fundamental constitutional rights'. 32Even those who support experimentation recognise that it has normative limits: Azmy, for example, draws on the example of predatory lending reform in the United States at the state level to illustrate the practical and theoretical benefits of state experimentation. 33However, Azmy distinguishes this economic and social policy context from that where 'there [is] a need for the imposition of national moral norms [as] against racial discrimination'. 34In that normative context, there is value in uniformity, including as a means of creating political bonds across the country. 35hus, there is a live question as to when experimentation is (in)appropriate, and whether there are particular fields or policy areas (such as addressing inequality and discrimination) where uniformity is preferable and should be pursued.Indeed, in some fields, experimentation might actually lead to a 'race to the bottom'. 36If states are competing for businesses and investment capital, there is a temptation to become a low-regulation, 'business friendly' jurisdiction, secured via minimal protection for individual rights.Thus, competition can pursue both a race to the top; and to the bottom. 37The question becomes, then, whether there are some fields in which minimum standards are required, and where experimentation is likely to undermine individual rights and liberties.

C The Practical Operation of Laboratory Federalism: Existing Comparative Studies
Building on these theoretical concerns, we can also question whether laboratory federalism and experimentation work in practice: are states effective 'laboratories'? 38Federalism may mean that states ('the vanguard') experiment with, test and evaluate particular policy ends, which can then be assessed and followed, avoided or improved upon, by other states ('the laggard'), 39 and eventually adopted at a national level.However, as Althouse argues, it can be difficult to determine which states are in the 'vanguard' and which form the 'laggards': 'The entire vanguard/laggard distinction creates a somewhat false, or at least sometimes false, picture of progress, depicting one right answer waiting for us in the future with the only task being to discover it and put the answer into action'. 40Laggards' could instead be depicted as traditionalists, having an important role to play in federalism, 41 and, consistent with the notion of 'wicked problems', there may be no 'right' answer to complex policy dilemmas.Thus, laboratory federalism may present an overly simplistic picture of policy change and progress.
At a practical level, as Fenna points out, the potential for effective laboratory federalism is dependent on a number of pragmatic considerations, such as: 1.The number of jurisdictions; 2. The range of policy responsibilities exercised by the states; 3. The resources at the states' disposal; 4. The level of difference and similarity between the states, such as to ensure policy divergence, but within the 'spectrum of relevance'. 42 this, we might add: 5. The political will and inclination of states to innovate; 6.The degree to which states are able to implement and evaluate policy reforms (including through the provision of appropriate human and financial resources); and 7.The presence of a long-term legislative and reform agenda, to ensure policies can be implemented and reviewed across an appropriate timeframe.
As Fenna notes, these characteristics are not well established in the Australian federal system, with a small number of jurisdictions, legislative powers of the states which have been 'whittled away', and limited state revenue. 43Thus, Fenna concludes that 'the potential for laboratory federalism to flourish in Australia is limited, and robust examples of laboratory federalism at work are disappointingly hard to find'. 44hat said, even in a jurisdiction like Australia, there is still some evidence of policy experimentation and learning between states.For example, in his study of state betting laws, using a historical case study approach, Painter traces a shift across the Australian states from policy uniformity, to diversity, and back to uniformity. 45Painter attributes this shift to political factors, shifting public opinions, and the changing position of key stakeholders.However, there was also evidence of mutual learning between and across jurisdictions, as evidenced by the series of inquiries and reviews conducted in each state. 46n her scholarship, Appleby has considered the extent to which leaving law and order within the Australian states' legislative autonomy has allowed for local diversity and experimentation. 47ppleby argues that the definitional difficulties that plague the Kable principlewhich sets out the 42.Fenna (n 27) 185-6.43.Ibid  requirements of institutional integrity for state courts exercising federal judicial powers, as part of an 'integrated court system'have unnecessarily hampered experimentation in the area of law and justice, and led to states 'replicat[ing] known-to-be-valid constitutional provisions, at the price of developing new, locally tailored, regimes'. 48Thus, requiring states to comply with federal (judicial) norms has limited experimentation in practice. 49urther, Appleby argues that Australian states adopt law and order measures due to political pressure, often before a policy has even commenced in another state, let alone been subject to evaluation and assessment. 50Appleby therefore concludes that: The law and order arena provides an example of the complexity of federal theory in practice.It illustrates experimentation with innovative and novel measures and the tailoring of measures to local needs and community expectations.However, it also illustrates that local needs and community expectations may be but one reason why states adopt new policies.Successful political rhetoric is also ripe for adoption from state to state within the federation. 51us, rhetoric and political pressure may be more influential than policy evaluation and mutual learning between states.
Even in a larger jurisdiction like the United States, scholars are increasingly doubting the practical effectiveness of laboratory or experimentalism federalism for advancing particular policy ends.Cunningham-Parmeter argues that Justice Brandeis (quoted in Part 2(a) above) poses two necessary conditions of effective experimentation: internalisation and replication. 52First, internalisation requires that states internalise the costs of their policies, so that other states can effectively analyse and assess their success. 53Second, replication requires that states take different approaches to addressing similar problems. 54For Cunningham-Parmeter, immigration law in the United States does not satisfy either of these conditions, meaning it is an example of failed experimentation.In this context, states can only experiment within a narrow field of federally defined norms, limiting the level of diversity and experimentation that is possible. 55Further, states do not internalise the costs of their policies, tending to shift the social, economic and human costs of immigration to other states. 56Thus, Cunningham-Parmeter concludes that 'not all subjects are ripe for local experimentation and not all tests produce valid results.'Despite the appealing image of states as laboratories, today's immigration experiments will not advance the nation's ongoing search for sounder immigration policies'. 57oncerns about the effectiveness of laboratory federalism have also been raised in the context of United States health reform.In their study, Sparer and  policy laboratories: 'although state and federal officials look curiously at approaches developed elsewhere, and occasionally replicate and emulate, such diffusion is rare'. 58Thus, there is limited mutual learning and evaluation between states, undermining the scope for laboratory federalism to advance policy reform and development in practice.
Similarly, drawing on a case study of the legal recognition of same-sex relationships, Knauer notes that federalism can offer states freedom to experiment with innovative solutions to social issues. 59However, she argues that laboratory or experimental federalism is inappropriate for securing minority rights: federalism is not a political philosophy, and experimentation can lead to both progressive and conservative outcomes. 60Federalism is not an appropriate mechanism for securing broad-based rights, as rights are not portable from state to state, and rights can be overturned. 61A lack of uniformity across states has resulted in same-sex relationships having 'confusing and conflicting status' 62 which 'weighs heavily on same-sex couples'. 63As Knauer concludes, 'samesex couples who reside in a jurisdiction with some degree of relationship recognition must travel [between states] at their own risk'. 64Knauer's study, then, offers a practical example of the inappropriateness of experimentation in particular policy areas.
Overall, these studies illustrate the practical limits of laboratory federalism.In a wide range of policy areas, across both Australia and the United States, laboratory federalism is not occurring to a desirable or optimal degree.Democratic experimentalism might offer some suggestions as to why this is the case.

III Lessons From Experimentalism
There are noticeable synergies between laboratory federalism and democratic experimentalism. 65xperimentalism sees legal change as a process of 'directly deliberative polyarchy' supported by an experimentalist architecture. 66While experimentalismlike laboratory federalismfocuses on the potential for bodies to learn from difference to solve similar problems and improve performance, experimentalism does this through a process of bench-marking, goal-setting, reporting and information-sharing, 67 facilitated and monitored by a central institution.Experimentalism therefore emphasises the role of both central institutions (in the labour law context, for example, these might include government agencies, legislatures and/or courts) and local institutions (such as employers) in achieving change.If applied to the public law context, experimentalism would see states as lower units and federal governments as the central unit.In an experimentalist design, regulatory and policy goals are set collaboratively between central and lower units; while lower units can pursue these 58.Sparer  goals as they see fit, 68 they are required to report to a central body against the goals and standards.Central units, then, monitor the performance of lower units, pool information and promote learning between local units via a process of 'deliberative engagement'. 69xperimentalism therefore offers a more structured and focused means of learning from difference than laboratory federalism, and one which revolves around a process of goal-setting, reporting and information sharing.Further, experimentalism is grounded in an architecture that requires a relatively powerful central institution to promote mutual learning and engagement between lower level units.
This offers some insights as to why laboratory federalism may promote policy development to only a limited extent.First, laboratory federalism does not prescribe any institutional architecture to promote mutual learning between states.There is no formal or coordinated process of goal-setting, benchmarking or information sharing.It is therefore unsurprising that previous studies have found any inter-state learning to be ad hoc and unstructured: this is nearly inevitable given there is no structure in place that requires or facilitates this process.Of course, in some policy and national contexts there may be processes established to promote communication and coordination between governments in a federal structure; however, this is not prescribed by laboratory federalism.
Second, and relatedly, laboratory federalism often posits a very limited role for the central federal government. 70At most, the federal government might adopt successful policy reforms that have been trialled and tested at state level; more commonly, though, laboratory federalism requires the federal government to leave states to their own (experimental) devices.Drawing on experimentalism, Wiseman and Owen challenge this assumption and instead emphasise the potential for federal or central governments to be key drivers of experimentation, including in cooperation with the states. 71Even this expanded role stands in stark contrast to the relatively powerful central unit envisaged by experimentalism.In experimentalism, central coordination is required to promote learning and improvement, and to develop appropriate responses to challenging social problems.Laboratory federalism typically makes no provision for centralised control or coordination of this nature. 72he experimentalism literature is also informative in recognising the risks and limits of experimentation.In an experimentalist architecture, the central unit plays an essential role in ensuring that regimes are protective, rather than a new risk to worker wellbeing. 73Given the role of the central unit is largely absent in laboratory federalism, this degree of protection is also missing. 74ven despite the protective role of the central unit, experimentalism is still criticised for appearing silent as to minimum standards and, as Deakin notes, does not necessarily rule out 'experimentalist solutions of a deregulatory type'. 75Thus, as with laboratory federalism, experimentalism arguably has limited normative content and can facilitate both progressive and conservative agendas.Empirically, however, Sabel does not believe that experimentalism leads to a race to the bottom: rather, firms aspire to rise to the top. 76It is possible that this is also the case in federal countries: states may aspire to rise to the top as well.The question, though, is whether this happens in practice.
Overall, then, experimentalism offers a number of lessons that can be used as points of critique of laboratory federalism.Mutual learning is unlikely to occur without a structured process to engage states in peer and policy review, analysis and sharing.Further, omitting the role of the central unit in enabling and facilitating this process puts a larger burden on the states themselves to manage the process of policy review.While similar structures might be adopted in practice by some governments on an ad hoc basis, the omission of these features from laboratory federalism may undermine its efficacy in practice.
These lessons from experimentalism are broadly consistent with Rom's view of the four prerequisites for federal states to be effective policy innovators: state capacity to develop innovations; flexibility to innovate; a clear set of policy goals; and motivation to experiment. 77Thus, Rom sees a 'federal approach to problem solving'which is particularly appropriate when broad policy goals can be agreed on, but methods are subject to disagreementas involving federal governments establishing standards, and providing states with 'incentives to experiment, evaluate these experiments, and adopt those that are effective … in ways that are politically suitable'. 78This is in line with an experimentalist approach.
It also reflects the approach adopted by the EU in the open method of coordination ('OMC') as a new form of governance, designed to facilitate mutual learning between Member States.The OMC involves a process of benchmarking at the EU level to evaluate and make recommendations about Member State policies. 79The OMC therefore represents a more centralised model than laboratory federalism.However, Kerber and Eckardt question whether it is possible or desirable to centrally evaluate policies, particularly if there is limited consensus about the criteria to be chosen for the evaluation. 80Selecting criteria is a political decision; if the wrong criteria are chosen, this risks entrenching the wrong policies. 81Kerber and Eckardt therefore see laboratory federalism as offering more scope for jurisdictions to adopt their own evaluative criteria, and as being 'advantageous in the long run', as it minimises the risks of policy path dependency and lock-in. 82This assumes, however, that states conduct any evaluation at all, without the oversight of some centralised process.Kerber and Eckardt also argue that the OMC fails to encourage policy innovation, as it only focuses on the diffusion of successful policies, not innovation in itself. 83 benchmarking generally. 84Thus, the challenge is to facilitate learning and evaluation between states in a way that does not inhibit ongoing innovation.In the context of the OMC, Kerber and Eckardt argue that this could be achieved by focusing on the initial purpose of the OMCsupporting policy learningrather than emphasising policy convergence. 85Thus, a centralised body that 'provide[s] a central institution that helps to establish processes of policy learning on the level of the member states' can be a positive development 86 that may overcome some of the major deficiencies of laboratory federalism. 87

IV Case Study: Australian age Discrimination Law
To develop these ideas further, it is informative to consider a case study of one particular policy area and how it has evolved in a federal context.To this end, this part considers the development of Australian age discrimination law, particularly as it relates to work and employment, and the extent to which it reflects laboratory federalism, drawing on findings from qualitative expert interviews with 66 Australian respondents from equality bodies, legal practice, age lobby groups, government and unions.

A The Federal Legal Framework
Age discrimination in employment is regulated in Australia by 'a patchwork' 88 of age discrimination legislation at federal, 89 state and territory level, 90 and as both an equality issue generally, and in relation to employment particularly. 91This complex legislative context has resulted in a proliferation of bodies that might deal with discrimination issues, including federal equality bodies such as the Australian Human Rights Commission ('AHRC'), state equality bodies such as the Victorian Equal Opportunity and Human Rights Commission ('VEOHRC'), federal employment bodies such as the Fair Work Commission ('FWC'), 92 state administrative tribunals such as the Victorian Civil and Administrative Tribunal ('VCAT') and the federal Fair Work Ombudsman ('FWO'). 93Discrimination law in Australia is a 'crowded space' 94 with 'a lot of players in the field'; 95 it is 'just very complicated'. 96he Australian anti-discrimination regime has been developed in the context of a federal division of powers between state, territory and Commonwealth legislatures. 97The Commonwealth Parliament has no express power to legislate in relation to discrimination matters.Instead, the Age 84.Ibid  Discrimination Act 2004 (Cth) ('ADA') (and other federal discrimination legislation) has been created on the basis of a number of heads of law-making power, including the external affairs power (to implement international treaty obligations); 98 and powers to make laws for the territories, 99 Commonwealth employees, 100 foreign, trading and financial corporations, 101 banking, 102 insurance, 103 trade and commerce, 104 and incidental matters. 105This hodgepodge of powers is reflected in s 9 of the ADA, which specifies its geographical application, and s 10 of the ADA, which links the application of the Act to the relevant constitutional powers. 106iven this complicated division of powers, it is difficult to determine the outer reaches of the federal discrimination Acts, and there has been limited case law on this issue. 107In many instances, state and territory discrimination statutes will operate as a parallel scheme to the federal law, with similar (though not identical) wording to the federal legislation.This will likely minimise any difficulties with determining which legislation applies in any given case. 108here the state and federal schemes are inconsistent, Commonwealth legislation will generally take precedence under s 109 of the Australian Constitution.However, this is not always clear cut.For example, in Jemena Asset Management Pty Ltd v Coinvest Ltd, 109 which related to long service leave provision, it was held that a Victorian state scheme that provided for pension portability in the construction industry was not inconsistent with federal awards and agreements; rather, the two schemes were complementary. 110The state scheme did not alter, impair or detract from the federal provisions; further, there was no evidence that the federal scheme intended to cover the field of pension portability.
In the context of age discrimination law, while compulsory retirement is prohibited by ADA s 18(2), some state and territory legislation still contains mandatory retirement as an exception to the principle of age equality. 111This is likely a direct inconsistency; the passage of the ADA has therefore likely rendered those state and territory exceptions invalid (and almost certainly in the case of the Northern Territory legislation), 112  Beyond these examples of direct inconsistency, and following Jemena, there is perhaps an argument that state equality laws do not detract from federal laws; rather, the two schemes can be seen as complementary and mutually supportive, with potentially different scopes and tests.This should minimise any scope for inconsistency.There is also an argument (again, as in Jemena) that 'it is possible to infer from the beneficial nature of the federal [statute] that the Commonwealth legislature did not intend to exclude a compatible State law'. 114That said, there is far more substantive overlap between state and federal equality laws than was the case in Jemena, raising a real risk of a substantive conflict between the regimes. 115urther constitutional complexity in this space arises from the use of state administrative tribunals to hear age discrimination complaints in the states and territories. 116Under Chapter III of the Constitution, only 'courts' may be vested with the capacity to exercise judicial power with respect to federal matters. 117Ananian-Welsh concludes that, while the Queensland Civil and Administrative Tribunal ('QCAT') is a 'court', the New South Wales Civil and Administrative Tribunal ('NCAT'), VCAT and South Australian Civil and Administrative Tribunal ('SACAT') are not, leaving their constitutional character 'open'. 118This means, then, that 'disputes that raise constitutional questions, involve parties from different states, or are otherwise listed in ss 75 or 76 of the Constitution, may be determined by the courts of the states but not by state administrative tribunals'. 119iven the potential overlap between state and federal discrimination law, it is critical that tribunals are capable of resolving questions relating to the interplay between state and federal statutes; and between discrimination law and constitutional law. 120In the disability discrimination case of Citta Hobart Pty Ltd v Cawthorn, 121 though, the High Court held that the Tasmanian Anti-Discrimination Tribunal ('ADT') could not consider any potential inconsistency between state and federal discrimination law under s 109.The majority concluded that: The constitutional defence [under s 109] was genuinely raised in answer to the complaint in the Tribunal and was not incapable on its face of legal argument.That being so, the complaint and the defence together formed parts of a single justiciable controversy comprising a matter within the description in each of s 76(i) and s 76(ii) of the Constitution.The Tribunal was on that basis correct to order that the complaint be dismissed for want of jurisdiction. 122e claimin its entiretytherefore had 'the characteristics of a matter' described in s 76 of the Constitution, and had to be referred to the Supreme Court for determination, even if the constitutional defence under s 109 was ultimately dismissed. 123The strengths and merits (or otherwise) of the defence were largely irrelevant: 'it is enough that the claim or defence be genuinely raised and not incapable on its face of legal argument'. 124This has the potential to severely constrain the ability of state tribunals to hear and determine matters relating to discrimination law.
Further, state tribunals are likely unable to hear cases where the federal government is a party, 125 or those that cross state borders.Thus, in Commonwealth v Anti-Discrimination Tribunal (Tasmania), 126 Kenny J opined that the Tasmanian ADT lacked jurisdiction to resolve a discrimination complaint against the federal department Centrelink; 127 regardless, the Commonwealth was not bound by the Anti-Discrimination Act 1998 (Tas). 128Instead, then, these matters involving federal jurisdiction will typically need to be transferred to a properly constituted 'court', occasioning additional expense and delay, and likely inhibiting access to justice. 129o address these complex issues of jurisdiction, Ananian-Welsh recommends that state tribunals be reconstituted as Chapter III 'courts', by designating these bodies as courts of record. 130This would not necessarily mean reducing the flexibility and informality of tribunal proceedings (which is a major advantage of the current state tribunal system): as Ananian-Welsh concludes, 'The powers, practices and procedures of a tribunal will play a secondary and distinctly nondeterminative role in ultimate determination of constitutional character'. 131urther complexity arises from the constitutional differences between the states and territories, particularly as they affect the status of territory tribunals.In the territories, the field is also shaped by s 122 of the Constitution.Ananian-Welsh confines her discussion to state tribunals and does not consider in detail how territory tribunals might differ; 132 it is possible that the same Chapter III rules apply to territory tribunals, 133 or that s 122 of the Constitution brings territory tribunals beyond the reach of Chapter III. 134In Bottrill v Sunol 135 (decided prior to the High Court's decision in Burns), the ACT Civil and Administrative Tribunal ('ACAT') rejected the argument that s 75(iv) of the Constitution prevented ACAT from determining a matter between a resident of the ACT and a resident of NSW. 136Section 75(iv) refers to residents of 'different States'; and the ACT is not a state. 137Since Burns, then, ACAT has adjudicated vilification disputes between parties from the ACT and NSW 138 and Queensland and the ACT. 139This arguably does not resolve the question of whether ACAT is capable of exercising judicial power in federal matters when, for example, 124.Ibid [35] (Kiefel CJ, Gageler, Keane, Gordon, Steward and Gleeson JJ).See also [75], [78] (Edelman J). 125.Ananian-Welsh (n 116) 865-6.126.(2008) 169 FCR 85. 127.Ibid 142 [236], 145 [246]-[248] (Kenny J). 128.Ibid 117 [151] (Weinberg J), 127 [189]  adjudicating discrimination claims that raise constitutional issues. 140The character and powers of ACAT and the Northern Territory Civil and Administrative Tribunal ('NTCAT') are therefore still matters of significant contention.
The experience in the ACT and NT is likely also shaped by the risk that the Commonwealth Parliament might disallow or overrule territory legislation. 141In the NT Anti-Discrimination Amendment Act 2022 (NT), for example, the new vilification provisions 142 are drafted to mimic s 18C of the Racial Discrimination Act 1975 (Cth), despite most states adopting a different approach, 143 and a Victorian inquiry putting forward models to overcome perceived shortcomings of s 18C. 144It may be, then, that legislative reform in the NT is being constrained by federal legislation or the fear of federal intervention; or, perhaps, that there is limited knowledge of developments in other jurisdictions (see below).
The federal system also poses a risk of 'double-dipping', where individuals who experience discrimination might make claims under both state/territory and federal discrimination legislation, or under both employment and discrimination legislation.This is expressly prohibited by s 12 of the ADA, which disentitles individuals from making a claim under the ADA if a claim has already been made under state or territory legislation.The FWA also prohibits 'double-dipping' in ss 724 and 734.

B Individual and Organisational Impacts of the Federal Structure
Before considering the extent to which there is mutual learning and experimentation in this field, it is important to recognise the organisational and individual risks arising within this complex federal structure.These are not problems of laboratory federalism as such; but they are important counterpoints when considering the practical impact of jurisdictional diversity in a federal country.
At an individual level, for those who experience discrimination, the most noticeable practical impact of the federal structure relates to the need to choose where proceedings or a complaint should be brought. 145This may have particular significance in practice, particularly in relation to relevant time limits, costs that might be incurred and any limit on damages. 146Once a claim is filed, it can be difficult (or impossible) to shift between jurisdictions. 147Claimants who have quality legal representation can use this 'choice' strategically to maximise chances of success, minimise exposure to costs orders and maximise the efficiency of the process. 148This duplication and variety of avenues for making a claim may therefore be a healthy sign of democracy, 149 enhancing access to justice.
For claimants who are unrepresented, though, or represented by practitioners unfamiliar with equality law, it is difficult to make a meaningful 'choice' of jurisdiction. 150State and federal overlap is 'fairly confusing for a lot of people'. 151Even with skilled representation, it can take hours to decide which route to pursue, 152 and the decision is 'not easy', 153 often requiring a full assessment of a claim's merits 154 that is highly context-specific. 155Vulnerable individuals may be more inclined to use 'local' state or territory processes, 156 especially in small jurisdictions where there are personal connections between equality agencies and referring NGOs, 157 or where there is scepticism of the federal government. 158Where claimants are more resilient (or legally represented) they might pursue a complaint to the more 'scary' federal system, where the threat of proceeding to a federal court can 'keep an employer on track'. 159Overall, though, confusion regarding where claims can and should be made might lead to claims not being brought at all. 160he limited case law on age discrimination in Australia 161 shows that some claimants have (unsuccessfully) attempted to bring both human rights and FWA claims. 162Others have sought to bring claims at both state and federal levels, 163 or to challenge federal legislation under state discrimination law. 164The complexity of these jurisdictional decisions is reflected in the Tasmanian case of B v Naval Reserve Cadets. 165In that case, part of the alleged conduct occurred on federal government property and part of it did not.Further, the Tribunal had to consider whether the Naval Reserve Cadets were an emanation of the Commonwealth or whether the Commonwealth might be covered by state discrimination legislation.Resolving these questions necessarily entailed further submissions and an additional hearing 166 (and, now, would likely be held to be beyond the ADT's jurisdiction).
Overall, then, this raises the question of whether a 'choice' of jurisdiction is meaningful, if it can only be exercised effectively with skilled, detailed (and expensive) legal advice, accessible to only a small proportion of claimants. 167Indeed, most claimants who file and pursue an age discrimination complaint with an equality agency in Australia are not legally represented. 168There are serious questions, then, about whether the possibility of state-based experimentation, which might strengthen legislative protection in some jurisdictions, is sufficient to balance these risks of federalism for individuals.While well-resourced claimants may be able to optimise the benefits and opportunities of experimentation, the practical impacts of regulatory diversity for less fortunate claimants are far less positive.
Multiple systems and statutes across jurisdictions may also result in a duplication of enforcement, 169 fragmenting and undermining protection from discrimination: 'you've got a relatively small population base but within that a ludicrous … amount of duplication'. 170It might fragment the funding of statutory equality agencies, undermining the effectiveness of enforcement in small jurisdictions. 171Having multiple state, territory and federal structures might also undermine the collection of data in relation to claims, and therefore the evidence base for effecting change in this area. 172A lack of age discrimination complaints, for example, may fall into the 'Bermuda triangle' between state and federal agencies, 173 and not be noticed.These are issues of federalism, not laboratory federalism as such.However, they are potentially compounded by the differences in the legal frameworks between jurisdictions, which make understanding the law, and monitoring its implementation, more complex.For organisations and employers, experimentation creates a 'minefield for organisations to navigate', as different exceptions and exemptions apply at different levels. 174This is a particular problem for companies operating across jurisdictions. 175This legal complexity is illustrated by my previous study of exceptions to age discrimination law in Australia and the United Kingdom. 176Most Australian jurisdictions adopt similar, albeit differently worded, exceptions to age discrimination law.However, the exceptions often vary in their breadth across jurisdictions, and additional exceptions exist in some states and territories and not others.This represents a complex legal field for both organisations and claimants to navigate.
That said, this diversity and multitude of players within a federal system can enhance protection from discrimination: state, territory and federal equality bodies address equality issues 'from quite different angles' and in different legislative frameworks. 177Multiple bodies and agencies can therefore fulfil complementary roles; state and territory bodies are more connected with their communities and can therefore focus on targeted capacity building particular to that community. 178iven the cultural, normative and structural differences between Australian jurisdictions, a 'one size fits all' approach to advancing equality is inappropriate. 179Agencies with limited resources can also address some of these constraints by 'building on the back of what other Commissions have done and recognising their work', 180 though this essentially outsources costs of enforcement to other jurisdictions.Understandably, then, 'the four smaller Commissions -WA, Tasmania, NT and SA -… are much more inclined to work collaboratively together.Because none of us have any resources', 181 though both large and small equality agencies can learn from each other's experiences and processes. 182This flags, then, the potential for mutual learning and experimentation across and between the states, both in relation to the legislative framework and in their approach to enforcement.

C Experimentation and Policy Percolation?
To what extent, then, is there evidence of experimentation and mutual learning in this space?Can the different approaches and structures in each state and territory drive innovation and produce better solutions for addressing age discrimination? 183The broad similarity between age discrimination statutes in some states and territories implies that jurisdictions are at least taking legislative 'inspiration' from other jurisdictions. 184Inter-state comparison can play a key role in equality law reform processes: 'you know the question that they're going to ask in the Cabinet room or Parliamentary Council or something is, "What are the other states and territories doing?"' 185 This 'copycat' 186 approach may support policy percolation and diffusion over time.
There is also growing evidence of experimentation in equality law among the Australian states and territories, and some evidence of mutual learning.For example, statutory equality duties have been adopted in equality statutes in Victoria and Tasmania, and equality duties under human rights statutes have been implemented in Queensland, Victoria, and the ACT (see further below). 187The ACT has made statutory provision for dual or multiple discrimination, 188 and a shifting burden of proof has been adopted in the ACT and under the FWA. 189Diversity and experimentation may therefore breed innovation and a race to the top. 190 process of inter-state learning and engagement was also reflected in the original emergence of age discrimination legislation in Australia.Age was first added to the grounds of discrimination in the Human Rights and Equal Opportunity Commission Act 1986 (Cth) from 1 January 1990; 191 and prohibited in South Australia from 1 June 1991.Compulsory retirement was progressively abolished in NSW from 1 January 1991, and other states soon followed.192 What is notable about this history, however, is that it was the federal government that led developments in this area; states and territories were following the central government's lead, the opposite to what is envisaged by laboratory federalism.Indeed, in relation to discrimination law generally, one of the justifications for the 1984 consolidation of equality law in South Australia was that the Commonwealth had indicated that it would amend federal equality laws to not apply to South Australia '[i]f the Commonwealth is satisfied that the legislation of this State implements the objectives of their Acts'.193 This makes explicit the influence the federal government has had over the development of state discrimination law, and the limited field within which states have been free to experiment.Similar concerns drove the passage of equality law in Western Australia: [A]s Western Australians we need our own equal opportunity legislation so we can be in charge of the affairs of our State.This is not an area where we will be happy with Commonwealth control.If we establish our own legislation, the Commonwealth has agreed it will allow its own law to 'roll back' so that the power of the State legislation will be preserved … We must face up to the fact that if we do not provide our own legislation in this area it will be provided anywaybut by the Commonwealth.The choice is not whether or not we will have legislation to outlaw discrimination, but whether we will have State legislation or Commonwealth legislation. 194us, fear and mistrust of Commonwealth interferenceand a Commonwealth threat to legislatedrove equality legislation in Western Australia.
In a contemporary setting, discrimination law appears to be entering a new and renewed period of experimentation and cross-jurisdictional learning.Since 2020, substantial reviews and inquiries into discrimination law have been conducted in the NT, 195 ACT, 196 Queensland, 197 WA 198 and federally. 199As Painter argues, these inquiries and reviews provide an important source of evidence regarding the degree to which '[i]mitation, emulation, or learning are important in shaping the evolution of policy in each jurisdiction'. 200While the process of legislative change is ongoing, and is not limited to age discrimination law specifically, these reviews and their recommendations illustrate a strong degree of cross-jurisdictional influence.
A critical example of this is the percolation and spread across jurisdictions of the Victorian positive duty to eliminate discrimination. 201In Victoria, this duty applies to all protected grounds, including age, and requires those who have a duty not to engage in discrimination, sexual harassment or victimisation to '[t]ake reasonable and proportionate measures to eliminate that discrimination, sexual harassment or victimisation as far as possible'. 202Through this series of inquiries, reviews and reforms, a positive duty like that in Victoria has been proposed in the NT, ACT, WA, Queensland and federally; and, to date, adopted in the NT, 203 ACT 204 and federally in relation to sex discrimination and sexual harassment. 205These provisions have largely replicated the framing and wording of the Victorian positive duty.Indeed, the AHRC's Respect@Work report specifically cited a case study of the Victorian provisions in informing its recommendation for adopting a positive duty in the context of sexual harassment and sex discrimination. 206In its Free and Equal position paper, the AHRC further recommended extending a federal positive duty, modelled on that in Victoria, to cover all protected grounds. 207The WA Law Reform Commission also recommended the adoption of a positive duty, again modelled on and closely reflecting the Victorian provisions, 208 and referring to the Victorian provisions and the Respect@Work report as comparative examples. 209The Respect@Work recommendations 210 and Victorian positive duty 211 also informed the Queensland review of discrimination law: the options canvassed in the Queensland Human Rights Commission's discussion paper relating to the implementation of a positive duty closely reflect the legislative framework in Victoria. 212The Queensland Human Rights Commission ultimately recommended a positive duty be introduced, replicating that in Victoria and as recommended by Respect@Work. 213In the ACT, the discussion paper on equality law reform drew extensively on the Victorian experience in considering the adoption of a positive equality duty; 214 with a positive duty ultimately being adopted in the Discrimination Amendment Act 2023 (ACT).
This provides a clear and documented example of state-based experimentation percolating across the states and territories, and up to the federal level, as envisaged by laboratory federalism.State and federal inquiries and reviews have provided a forum, albeit an ad hoc one, for jurisdictions to engage in peer and policy review, analysis and sharing.Victoria's decade-long 'experiment' with a positive duty has provided other jurisdictions with a degree of certainty that the risks of experimentation are minimal, allowing other states to simply imitate the successful Victorian experiment 215 and 'allaying some of their worst fears'. 216hat said, these developments have not come from nowhere.The AHRC has been recommending the adoption of positive equality duties in Australia since at least 2008, though it has also repeatedly noted that there is a need for further consideration of how such a duty might be drafted and implemented. 217Victoria's experimentation has likely helped to resolve some of the practical difficulties of adapting positive duties to the Australian context, offering a workable example of a positive duty that can be replicated in other jurisdictions. 218There are also numerous examples of positive equality duties internationally, as in Great Britain, 219 Northern Ireland, 220 and Canada, 221 meaning the stateseven Victoriaare better seen as 'scale innovators' than 'policy innovators', 222 adapting international ideas to their own circumstances, rather than creating new ideas as such. 223This also reflects Rom's argument that states are unlikely to 'experiment' with policies that are new in a global sense; their 'innovation' involves adopting policies or programmes that are new to the states themselves. 224iven this long history, the impetus for change in 2020 perhaps came from the telling findings of the AHRC's Respect@Work report, which chronicled the perilous state of workplaces for women, and the chronic failings of discrimination law to achieve meaningful change.Successful experimentation, then, is likely not enough to prompt mutual learning; replication across jurisdictions likely requires a compelling catalyst for change.Further, learning is not so much about finding the 'right' solution, but about sorting convincing from unconvincing arguments. 225Again, Respect@ Work offered a critical positioning for these reviews and inquiries, by cataloguing a compelling argument that existing discrimination lawrelying on individual complaints and individual enforcementhad failed to achieve change. 226In this context, advocating against a positive duty became unconvincing. 227here is a risk, though, that these developments reflect the replication and duplication of interstate developments, without mutual learning or evaluation.For example, the enforcement mechanisms for the positive equality duty in Victoria were watered down in 2011, 228 removing the VEOHRC's powers to ensure compliance via compliance notices, enforceable undertakings and public inquiries. 229The VEOHRC can now only enforce the positive duty via persuasion; or, in exceptional circumstances, by conducting an investigation, 230 and thereafter taking any action it sees fit, including entering into an agreement with the organisation. 231These lesser enforcement powers no longer reflect what was recommended in the Gardner review; 232 the VEOHRC has repeatedly called for its original, broader suite of enforcement powers to be reinstated. 233Despite these concerns, these lesser enforcement mechanisms have been largely replicated in the NT, where the duty will be enforced by the Commission undertaking an investigation, and taking any subsequent action considered appropriate. 234However, unlike in Victoria, in the NT the Commission can enter into an enforceable undertaking. 235In the ACT, the positive duty is not to be enforced by the Commission; instead, compliance with the duty can be taken into account in conciliating and adjudicating discrimination complaints. 236At the federal level, the AHRC will be empowered to inquire into compliance with the positive duty; 237 in cases of non-compliance, this can lead to a compliance notice, 238 or enforceable undertaking. 239The WA Law Reform Commission goes further, recommending both escalating compliance powers for the WA Equal Opportunity Commission and individual enforcement mechanisms for a positive duty. 240This may illustrate mutual learning and experimentation across and between jurisdictions, particularly in rectifying some of the enforcement gaps in Victoria.It remains to be seen whether the individualised, complaint-driven approach to the enforcement of positive duties in the ACT proves more effective than the investigative, Commission-driven approach in Victoria, the NT and federally.Thus, even in adopting successful experiments from other jurisdictions, there is still evidence of diversity, evaluation and innovation in this space.
This, then, flags the importance of enabling mutual learning and evaluation across and between jurisdictions, particularly in the absence of a structured process of peer review or a central unit that facilitates mutual learning, as envisaged by democratic experimentalism.In a contemporary setting, processes to facilitate information exchange and learning between the states have been established through the government agencies operating in each jurisdiction: the Australian Council of Human Rights Agencies ('ACHRA') provides a forum for engagement between human rights bodies at the federal, state and territory level, to consider issues of mutual concern and emerging developments. 241ACHRA has become a forum for the sharing of resources and the development of joint submissions, creating broader synergies and efficiencies across jurisdictions. 242he Australian Discrimination Law Experts Group ('ADLEG'), a group of academics with a shared interest and expertise in equality law, has also come to represent a critical enabler of mutual learning across jurisdictions.Through scholarly submissions to government inquiries, reviews and consultations, ADLEG has facilitated the transmission of ideas and 'experiments' across Australian jurisdictions, including positive equality duties.In the WA Law Reform Commission review of discrimination law, for example, ADLEG's submissions were cited 42 times. 243By working collectively, ADLEG has been able to pool scholarly knowledge across jurisdictions, identify and communicate critical experiments to policymakers, and encourage innovation in the reform process.That said, relying on the individual goodwill of time-poor academics to overcome capacity constraints of individual states and territories is potentially not sustainable in the longer-term, especially given the sheer number of reviews, consultations and inquiries being conducted in the field.A long-term, structured process to enable mutual learning is required.

V Conclusion
Consistent with Appleby's study of law and order policy in Australia, then, state and territory governments and human rights agencies appear attuned to developments in other jurisdictions and use interstate developments as inspiration, justification and a source of political pressure to develop their own equality law framework.This is both enabled and facilitated by the involvement of groups such as ADLEG in inquiry and consultation processes.Thus, it appears that developments in each state and territory do inform legal developments elsewhere, though this process is sporadic and ad hoc, strongly influenced by political factors, and may depend on a catalyst for change.Laboratory federalism is clearly occurring in practice.
That said, the federal government is occupying a more prominent position in this area than might be envisaged by laboratory federalism.Indeed, federal interventions like the Respect@Work report have arguably acted as a catalyst for change in the states and territories, by raising the profile of equality law and its limitations, acting as a critical example of regulatory possibilities and making state equality law reform politically desirable and expedient. 244While technically confined to sexual harassment, the implications of Respect@Work have already been wide-reaching in state and territory reform processes.We may need, then, to revise our ideas of laboratory federalism.Rather than experimentation percolating up to the federal level in a one-way transmission of ideas, the process of change appears far more dynamic, with experiments and innovations travelling from the states, to the federal level, and back, across and within the federal structure.
This federal 'dialogue' echoes experimentalism's emphasis on a strong central body playing a key role as part of an experimentalist architecture.Equally, it might reflect a context in which states have limited scope to experiment due to both limited resources and being confined by federal norms that prohibit age discrimination.These federal norms could be seen as both undermining laboratory federalism and, conversely, acting as an important regulatory backstop that protects individual rights.Laboratory federalism, therefore, continues to raise fundamental tensions in the context of individual rights protection and in equality law in particular.Building on this study, future research should continue to examine the normative limits of laboratory federalism, to build a more comprehensive picture of where diversity and innovation is and is not normatively desirable.
More generally, too, this case study of age discrimination law allows us to reflect on the emerging challenges for contemporary federalism.Indeed, this study illustrates the difficulty of distinguishing the impacts of federalism from an evaluation of laboratory federalism and the degree of experimentation and mutual learning occurring across jurisdictions.As Fenna notes, Australian federalism (and that in the United States and Canada) was originally modelled on the idea that: The two levels of government were assigned their own respective policy domains (though of course not without overlaps and ambiguities), within which they were to exercise full legislative authority.In the main, they were expected to operate autonomously and wield sovereign power within their respective areas of jurisdiction or competence. 245is notion of the division of powers becomes increasingly complex as the overlaps and ambiguities of legislative responsibility become more pronounced.The social and economic problems of our timesinequality, climate change, global pandemics, to name a fewdo not lend themselves to simple, unified solutions. 246No one government can hold the key to policy reform; issues and problems do not confine themselves to one level of government.
This growing complexity and overlap of policy responsibility is reflected in the structure of Australian age discrimination law, which is characterised by multiple pieces of (potentially conflicting) legislation across state, territory and federal governments.While this may strengthen the protection of individual rights for well informed, well-resourced claimants, it is likely to be confusing and bewildering for those without expert legal assistance and may actually inhibit individual claims, particularly given the constitutional limits placed on the administrative tribunals that hear most discrimination claims.This may serve to undermine the practical operation of age discrimination law and its enforcement, particularly for those in the most vulnerable circumstances.
As Fenna argues, while federal theory seeks to prevent 'unnecessary harmonisation', 247 harmonisation should be sought where diversity generates costs and the benefits of diversity are limited. 248At present, equality law may be a case where the costs of diversity outweigh its benefits.To address this, future research should critically consider how ideas of laboratory federalism can be further developed, drawing on democratic experimentalism, to better specify the systems and processes required to enable effective experimentation and mutual learning in federal systems.
As Gelbach has mapped in the context of welfare reform, there can be too much variation in policy reform to make evaluation meaningful: Gelbach (n 25) 157.In this case, too much reform can be as troublesome as not enough.Gelbach therefore concludes: 'From a policy-design perspective, the story of welfare policy devolution is a cautionary tale concerning what can be learned from a simultaneous explosion of state-level experimentation.By my lights, anyway, welfare reform has been no exemplar of Justice Brandeis's laboratory': at162.27.In Australia, for example, states have been 'allocated the majority of expenditure responsibilities', yet the key sources of governmental revenue (customs and excise duties, and income tax) are now controlled by the federal government, which has led to 'fiscal inequality': see Sam Reinhardt and Lee Steel, 'A brief history of Australia's tax system' (Working Group Paper, 22nd APEC Finance Ministers' Technical Working Group Meeting, Khanh Hoa, Vietnam, 15 June 2006).See further Alan Fenna, 'The Division of Powers in Australian Federalism: Subsidiarity and the Single Market' (2007) 2(3) 25. Jonah B Gelbach, 'Uncontrolled Experiments from the Laboratories of Democracy: Traditional Cash Welfare, Federalism, and Welfare Reform' in Jonathan Klick (ed), The Law and Economics of Federalism (Edward Elgar, 2017) 110, 156.26.29.Sparer and Brown (n 10) 193-4.See also James A Gardner, 'The "States-as-Laboratories" Metaphor in State Constitutional Law' (1996) 30(2) Valparaiso University Law Review 475, 481-2.30.Sparer and Brown (n 10) 190-1.31.See, eg, Kenneth Wiltshire, 'Australian Federalism: The Business Perspective' (2008) 31(2) UNSW Law Journal 583, 603-4, 608.

186. 44. Ibid. Instead of focusing on federal outcomes -which have generally 'not produced significant, enduring differences
Brown conclude that states are generally poor 48.Ibid 674, 680-3.49.See the similar concerns in the context of Cunningham-Parmeter's study of United States immigration law, below; Cf Brendan Lim, 'Laboratory Federalism and the Kable Principle' (2014) 42(3) Federal Law Review 519.Lim argues that the apparent conflict between laboratory federalism and the Kable principle is illusory.50.Appleby (n 3) 678-9.In the United States, see the detailed discussion in Andrew Karch, Democratic Laboratories: Policy Diffusion among the American States (University of Michigan Press, 2007) ('Democratic Laboratories').
The authors see this as a problem of 75.Simon Deakin, 'Learning or Diversity?Reflections on the Future of International Labour Standards' in Proceedings of the International Colloquium on the 80th Anniversary of the ILO Committee of Experts on the Application of Con- Fair Work Act 2009 (Cth) ('FWA').92.The FWC is responsible for conciliation and arbitration of disputes.93.The FWO is responsible for compliance and enforcement of employment legislation.94.Blackham, 'Addressing Age Discrimination in Employment' (n 2) A3. 95.Ibid A101.96.Ibid A5. 97.See, eg, Australian Constitution ss 51, 52, 122.
and employers rarely rely on the exception in practice.113TheFWA is also empowered by the referral power under s 51(xxxviii) of the Constitution.Thus, it may more comprehensively cover the states than discrimination law, at least to the extent that the states have referred their powers to the Commonwealth government.107.Neil Rees, Simon Rice and Dominique Allen, Australian Anti-Discrimination Law (Federation Press, second ed, 2014) Tom Pauling and Sonia Brownhill, 'The Territories and Constitutional Change' (2007) 28 Adelaide Law Review 56.See further the extensive discussion in GJ Lindell, Cowen and Zines's Federal Jurisdiction in Australia (Federation Press, 4th edition, 2016) ch 5. Lindell, while challenging Pauling and Brownhill's reasoning, concludes: 'There was and continues to be a tendency by the High Court to construe s 122 in the context of the Constitution as a whole, including Ch III, rather than as a disparate power': Ibid 251.134.See Stephen McDonald, 'Territory Courts and Federal Jurisdiction' (2005) 33(1) Federal Law Review 57.135.[2017] ACAT 81.136.Ibid [62].See also Pauling and Brownhill (n 133) 70.137.Ibid [62]-[63].See also Scott v Bowden [2002] HCA 60 [14] (McHugh J). 138.See Kerslake v Sunol (Discrimination) [2022] ACAT 40 ('Kerslake').139.Rep v Clinch (Appeal) [2021] ACAT 106 (an argument based on Burns v Corbett was abandoned: [22]).