Abstract
The courts have hitherto played a remarkably limited role in the working of devolution in the United Kingdom. In this respect, the UK has been an exception to practice in most federal or decentralised systems, where the courts have tended to assume a prominent role in adjudicating intergovernmental disputes. This article examines whether political congruence contributed to this minimal role for the courts in IGR. It argues that additional factors are also at play: the limited role accorded to lawyers and legal issues in the UK's system of government, the flexibility of the devolution settlements, and the political nature of the UK's constitution, which make it both possible and desirable to resolve intergovernmental disputes by different means. While the courts are likely to become more prominent as a result of third-party litigation, particularly as devolved legislatures become more active, the limited role of the courts in most intergovernmental matters appears unlikely to change.
Introduction
Before devolution in 1999, there was an expectation that one consequence would be an increased, and increasingly political, role for the courts. They would be called upon to resolve a variety of disputes arising from the new constitutional settlement, as had been seen in many other federal and decentralised systems around the world. This role would increase the political profile of the courts, and put pressure on their perceived impartiality. In practice, however, little of this has happened; the amount of devolution-related litigation has been modest, and there has been no intergovernmental litigation whatever to date. This limited role for the courts is a classic case of a dog that did not bark (Conan Doyle 1950, 28), and like that dog, its silence tells us a good deal about devolution in the UK, not least because it marks a significant difference between the UK and other systems elsewhere.
This article will seek to explain this phenomenon, and in doing so will address two questions. First, what factors mean that the courts have played such a minimal role in devolution, and the resolution of devolution disputes, since 1999? While explaining a negative is always difficult, it is possible to identify a number of aspects of devolution's institutional design, and of the practice of government and constitutional politics in the UK more generally, that explain this. Second, how durable are those factors? How susceptible are they to changes in the constitutional and political context, such as the increased party political incongruence that has emerged since 2007?
The article first discusses the role of courts in federal and decentralised systems. Next, it considers the expectations of the role they would play before UK devolution, and the role they have actually played. It then explores reasons for the courts’ limited role. As well as published materials including court judgments and official documents, the article draws on interviews with civil servants and politicians involved in devolution, carried out since 2001. 1
Courts and Politics in Federal and Decentralised Systems
There is considerable debate about the nature of constitutional adjudication, and indeed the nature of judicial decision-making more generally. In the political science literature, it is regarded as inherently political, with the result that the highest court of appeal in any system will assume a prominent role as a political actor, as well as its legal one (e.g. Shapiro 1981; Stone Sweet 2000). Traditionally, lawyers have taken a different view, treating the exercise of judicial authority as a technical exercise in which political considerations do not figure—a view, in the UK, first seriously challenged by J. A. G. Griffiths (1977), and now accepted by very few scholars. Certainly, judicial decisions in constitutional cases are political by their nature. This role is particularly pronounced in federal and decentralised systems. As Daniel Halberstam (2008, 143) puts it:
The precise content of the federal bargain will necessarily be incomplete. The authors cannot foresee all the contingencies that an effective system of governance must confront. Federal bargains will be struck on particulars without consensus on an underlying theory; and federal bargains will be struck on more abstract principles without agreement on the particulars that the principles entail. … A successful federal system must remain flexible enough to allow for effective governance and yet be stable enough to prevent radical centripetal or centrifugal shifts of power that undermine the principle of shared rules.
The constitutional arrangements of most federal and decentralised states mean that such flexibility is a scarce commodity. Written constitutions with formal amendment procedures commonly require unanimity among the main institutional actors, or a very high level of consensus among them. They may also require the holding of referenda (as notably in Australia). Where parties cannot agree, there can be no change to the constitution and so amendments cannot be made. In such circumstances, the burden passes to the courts to determine what the constitution ‘means’. The decisions made by the courts in pursuit of this are often not only controversial in themselves, but also make the role of the courts the subject of wider political controversy.
One consequence of this is the emergence of an academic literature exploring the role of the courts in the working of a particular federal or decentralised system. This literature is usually concerned with only a single system, and involves perspectives both from lawyers and from political scientists. Examples in English include John T. Saywell (2002) for Canada; Donald P. Kommers (1997) and Philip Blair and Peter Cullen (1999) for Germany; or Geoffrey Sawer (1967) or Leslie Zines (1997) for Australia. In Canada, the trend toward emphasising provincial rights has halted with the ending of appeals to the Privy Council and the Supreme Court of Canada becoming the final court of appeal. In Australia, the High Court has consistently taken a view supportive of the federal order rather than the states, most notably in cases concerning income tax powers (Saunders 2003). In Germany, the Constitutional Court has developed the principle of Bundestreue or ‘federal trust’, creating a legal basis for the co-operative and integrated nature of German federalism and requiring both federal and Land governments to co-operate with each other. Institutionally speaking, there are wide variations between such systems, with multiple purpose top-level courts being usual in the common-law world, while specialist constitutional courts are more usual in civil-law ones. The only common thread is that the courts assume a prominent role in determining significant issues in federal and decentralised systems with written constitutions. How they exercise this role tends to be specific to the system concerned.
Comparative discussions are rarer in the academic literature, and mostly relate to the common-law world (see Johnston 1969; Baier 2006). Much of this literature is concerned with doctrinal analysis; its chief concern is to consider the judicial doctrines used by the courts themselves to explain what conclusions the courts have reached, in general and in particular cases, rather than to explore whether or to what extent the courts play a political role, or assess the impact of the courts’ exercise of their powers on the federal system as a whole. This approach is particularly suitable for common-law systems, where the doctrine of legal precedent set out in earlier judgments (also known as stare decisis) means that judicial determination is an iterative process, each case building on others. Even in such similar systems, there is relatively little cross-pollination of legal doctrines, with such doctrines developing incrementally but each remaining largely distinct from that of other countries. Such an approach can be criticised, however, for overlooking the cumulative impact of decisions and their overall impact on a particular country's system (as does Bzdera 1993), in a critique that has much in common with legal-realist approaches.
The Courts’ Role in a Devolved United Kingdom: Expectations and Reality
In this context, it was reasonable to expect that the courts in the UK might assume a similar role following devolution to Scotland, Wales and Northern Ireland. That was the expectation of policy-makers, and the devolution legislation included extensive and similar provisions for the resolution of ‘devolution issues’ (carefully defined in the acts). These procedures are set out in Schedule 6 to the Scotland Act 1998, Schedule 10 to the Northern Ireland Act 1998 and (now) Schedule 9 to the Government of Wales Act 2006 (formerly Schedule 8 to the Government of Wales Act 1998). Their provisions include the rapid transmission of such issues to the Judicial Committee of the Privy Council for determination, and for the law officers of the governments that have them (UK and Scottish governments, since 2007 the Welsh Assembly Government, and since 2010 the Northern Ireland Executive) to have formal standing to take part in those proceedings. The definition of a ‘devolution issue’ is not exhaustive, however, and it isquite possible for cases that fall outside its scope nonetheless to raise questions affecting the working of the devolution settlements. The definition is, rather, a way of highlighting a particular class of cases of concern, and creating procedural mechanisms to enable them to be determined with relative speed by the Judicial Committee (or now the UK Supreme Court).
In the run-up to devolution in 1999, legal scholars devoted considerable attention to the legal aspects of devolution and the legal procedures (e.g. Bates 1997; Craig and Walters 1999). During parliamentary and wider debates, there were predictions that devolution would result in extensive and protracted litigation. One point that attracted debate was the choice of the Judicial Committee of the Privy Council to deal with ‘devolution issues’—not the Appellate Committee of the House of Lords, which otherwise was the UK's highest court for all three of the UK's legal jurisdictions (Scotland, Northern Ireland, and England and Wales, though with very limited jurisdiction for Scottish criminal matters). The choice was odd as the Judicial Committee had a small and shrinking jurisdiction, having once sat on appeals from the UK's various colonies (which led it to play an active and controversial role between 1867 and 1948 in adjudicating federalism issues from Canada: Cairns 1971). Moreover, although it was possible to add more judges to the pool from which the Judicial Committee was drawn, its normal membership was the Law Lords who also sat in the House of Lords. An important factor appears to have been Scottish concerns about the principle of a legislative chamber also playing a judicial role, although attempts to establish the reasoning behind this decision have proved largely fruitless (author's interviews). However, this approach created a different problem, by establishing a ‘dual apex’ at the summit of legal adjudication in the UK. While the formal jurisdiction of the two courts might be clearly distinct, the effects of their judgments are not; each is so authoritative that it would be highly persuasive for all other courts (including the other top court), even if it were not formally binding. The risk this created was that differently composed final courts of appeal of similar authority and standing might determine the same issue in different ways, leading to serious confusion (Le Sueur and Cornes 2001).
However since 1999, the courts have in practice played an extremely limited role in relation to devolution. Until 2010 that could scarcely even be characterised as minimal. While it has increased since then, it still remains of limited impact. There have been very few devolution cases of any sort. Most of the cases that have been decided arose because of the incorporation of the European Convention on Human Rights (ECHR) into the devolution statutes, and have concerned alleged infringements of Convention rights in the prosecution of criminal offences in the Scottish courts (discussed in O'Neill 2001). 2 They would therefore be likely to have arisen regardless of devolution, and have not in substance concerned the respective powers of the Scottish Parliament and Executive/Government. 3 One minor irony of this has been that UK-level courts have been involved in Scottish criminal matters in a way that was impossible before 1999, a matter that has become an increasing source of debate and controversy in Scotland. 4
None of the cases decided by the courts has involved intergovernmental litigation, with one tier of government alleging that legislation passed or some other action taken by the other tier exceeded its powers (Gee 2005). There have been no challenges to legislation made by the National Assembly for Wales or Welsh Assembly Government (though there have been a couple of challenges to planning decisions taken by the Assembly Government, neither of which has reached the highest courts). In Northern Ireland, there has been a good deal of litigation about the terms of the Belfast Agreement, but concerning the rights inter se of various ministers (discussed in Anthony 2002). There have been no legal challenges to legislation passed by the Northern Ireland Assembly, whether when it sat in 2000–02 or since devolution resumed in 2007. Until June 2011 only four cases, all Scottish—Anderson, Reid and Doherty v. Scottish Ministers [2001] UKPC D5, DS v. Lord Advocate [2007] UKPC 36, Whaley v. Lord Advocate [2007] UKHL 53, and Martin and Miller v. Lord Advocate [2010] UKSC 10 (discussed in more detail below)—have involved the UK's highest courts considering whether devolved legislation was within the powers of the legislature that passed it. 5 The three earlier cases raised questions of whether devolved legislation complied with human rights obligations under the ECHR; only the Martin and Miller case concerned the purely domestic division of powers and the interpretation of Schedule 5 to the Scotland Act 1998, setting out the scope of matters reserved to the UK parliament. In each of those cases, the validity of the devolved legislation was upheld. 6
The article will argue that this limited role for the courts is not merely a consequence of the institutional framework of devolution (though that is a factor), or of the fact that devolution is still a novel phenomenon which is ‘bedding in’, but that it reflects wider aspects of the UK's constitution and constitutional politics. As such, the impact of political incongruence as it has developed since the May 2007 elections has been limited, and the existing ‘de-judicialisation’ of devolution appears strongly founded, although some increase in the impact of the courts is likely in the coming years.
Explaining the Lack of Devolution Disputes: Institutional Factors
It is necessary to examine the institutional framework of devolution to appreciate why it is so hard for justiciable issues to arise in the UK. These factors are particularly important for governments, so help to explain the lack of intergovernmental litigation. Litigation brought by third parties is something that is much harder to control, of course, and these factors have much less impact in relation to that.
Constitutional Structure
The first key feature of devolution is that its constitutional structure means that only the devolved legislatures can exceed their powers, and not the UK level. 7 Legally speaking, devolution involves conferring legislative and executive powers from Westminster on the devolved institutions, which are bound by the limits of that authority thereafter. The UK parliament remains (explicitly) sovereign (see section 28(7) of the Scotland Act 1998; sections 93(5) and 107(5) of the Government of Wales Act 2006; section 5(6) of the Northern Ireland Act 1998). To the extent that its powers are constrained by devolution (for example, by the Sewel convention that Westminster should not legislate on devolved matters without the consent of the devolved legislature affected), those constraints are unlikely to be regarded as legal in nature by the courts, however far-reaching they may be in political terms. Thus, while much of the litigation in other systems arises from the constituent units alleging that the federal/central government has exceeded its powers, such claims would be very hard to sustain in the UK. Only one party will be in a position to legislate beyond its powers, not both.
Restraints on Exceeding Devolved Legislative Powers
Second, it is hard for a devolved legislature to pass legislation that exceeds its powers, and almost impossible to do so inadvertently or unknowingly. Substantively, devolved legislation must be within the express powers of the National Assembly (in Wales), and not relate to reserved matters (in Scotland) or reserved or excepted ones (in Northern Ireland, though legislation on reserved matters is permissible with the consent of the secretary of state). Devolved legislation must also comply with requirements of European Union law, and the ‘Convention rights’ arising under the ECHR protected by the Human Rights Act 1998. The UK government also has power to intervene if proposed devolved legislation interferes with UK international obligations, interests of defence or national security or (in Wales) water supplies in England.
Procedurally, as Robert Hazell (2007) emphasises, there are numerous checks to enforce these requirements. The presiding officer of the relevant legislature must make a statement as to whether the legislation is within that legislature's powers on its introduction. All government legislation in Scotland and Northern Ireland must similarly be certified by the minister introducing it (though not that of private members or committees), while in Wales all legislation must be certified by whoever introduces it—whether a minister, committee or back-bench member (see section 31 Scotland Act 1998; sections 9 and 10 Northern Ireland Act 1998; sections 97 and 110 Government of Wales Act 2006). 8 It will therefore be immediately apparent if legislation exceeds that body's powers before it is debated, let alone passed into law (or purport to be passed into law). In addition, each of the devolution statutes contains provisions allowing the UK Attorney-General and other law officers (of both UK and devolved governments) to refer proposed devolved legislation to the Privy Council (now the UK Supreme Court) to resolve the question of whether it is within devolved legislative competence. The UK secretary of state has further powers to prevent devolved legislation receiving royal assent if issues of its legislative competence have not been resolved by the courts.
These statutory provisions are notable for two reasons. One is their detail, and the extent to which they create procedural hurdles through which devolved legislation must pass. This contrasts with the approach under many federal constitutions, which do not create such requirements of process to vet legislation by constituent units but simply provide that legislation beyond a legislature's competence is invalid. These hurdles in turn reduce the likelihood that a devolved legislature will seek to pass legislation that is beyond its powers. The second is the extent to which these provisions underpin intensive work by officials of all the institutions involved to ensure that legislation is in fact within competence. Thus, all legislation proposed by the Scottish Government will not only be the subject of debate within government, involving its own legal advisers and legislative drafters, but also consideration by lawyers working for the Scottish Parliament (regarding the certificate the Presiding Officer must issue) and the UK government (in the Office of the Solicitor to the Advocate General). It is not yet clear what processes will apply to Assembly Acts in Wales, but the complex process of conferring legislative powers on the Assembly under Part 3 of the Government of Wales Act 2006 means its effect was limited. However, the legal resources of the Wales Office (which has just two lawyers) are limited compared with the Scotland Office. The lawyers will, of course, discuss questionable provisions with each other (author's interviews). Legally questionable provisions are unlikely to survive such scrutiny readily.
The Flexibility of the Devolution Arrangements
Third, if there are areas where devolved powers are not adequate, it is relatively easy to extend them. The details of these mechanisms vary according to which settlement is involved, and whether the powers involved are legislative or executive. Extending devolved legislative powers can be done by secondary legislation; orders can amend the list of reserved powers for Scotland (in Schedule 4 to the Scotland Act 1998) or of reserved matters for Northern Ireland (in Schedule 3 to the Northern Ireland Act 1998). For Wales, a Legislative Competence Order could add particular matters to Schedule 5 to the Government of Wales Act 2006 (and Schedule 7 regarding Assembly Acts can similarly be varied). All such changes require the assent of the devolved legislature as well as of the UK parliament, and in Northern Ireland cross-community consent is needed as well. Changes to legislative powers can also be made by primary (UK) statute, which requires the assent of the devolved legislature for Scotland and Northern Ireland (though there it has been signalled by the Executive, not the Assembly), but not for Wales. 9 Devolved executive powers can similarly be extended, in a variety of ways, chiefly by orders transferring functions to the Scottish, Welsh or Northern Ireland ministers. There are further statutory powers to remedy ultra vires acts undertaken or legislation passed by the devolved institutions, to make consequential modifications to Westminster legislation and other necessary changes. In any event, Westminster's continuing sovereignty means that primary legislation can also redress any shortcomings in the scope of devolved powers.
The upshot is that the devolved constitutions remain very flexible instruments. In this respect, they are quite unlike the constitutionally rigid systems identified by Halberstam in the passage quoted on p. 304 above. Both by virtue of Westminster's ongoing sovereignty and this array of express statutory powers, it is easy technically to alter the scope of devolved and non-devolved powers—if the parties agree. The key issue becomes the political one of whether the parties will agree, rather than the legal one of whether they are able to agree or to implement their agreement. This contrasts profoundly with systems with written constitutions, which can only be amended by slow and complex processes. (In practice, those procedural requirements make many such constitutions incapable of amendment.) That in turn means that the application of a written constitution often requires a court to determine the issue, while in the UK an alternative approach is possible, and often considered preferable.
The Limited Role of Law and Lawyers in the Practice of Government in the UK
A fourth set of factors that limit the role of the courts in the devolved UK concerns the role of law within governments in the UK. In principle, the role of giving legal advice belongs to a government's law officer: the Attorney General, Solicitor General or Advocate General for Scotland in the UK government, the Lord Advocate and Solicitor General for Scotland in the Scottish Government, and the Counsel General in the Welsh Assembly Government. In practice, the bulk of the work is undertaken by officials, and the law officers only advise on the most politically sensitive of issues. Particularly in the middle ranks of the civil service, lawyers are relatively scarce on the ground. According to official civil service statistics, in 2008 only 5,640 officials working in the Home Civil Service, of a total headcount of 515,660, belonged to the ‘legal profession’ category—barely over 1 per cent (Office for National Statistics 2009, table 8). All lawyers, even trainees, belong to the ‘management grades’ of executive officer and above, which account for about half of the total civil service (Office for National Statistics 2009, table 1); but even 2 per cent is a very small part of an organisation. They are more heavily weighted toward senior levels of the civil service; of the 38 officials of permanent secretary rank in October 2008, two were lawyers (5.3 per cent): the Treasury Solicitor, who is head of profession for lawyers in the UK government, and the First Parliamentary Counsel who is in charge of legislative drafting. In addition, the Director of Public Prosecutions, whose responsibilities now extend only to England and Wales, is of similar status, though not formally a grade 1 official (Cabinet Office Civil Service Capability Group, June 2009). The most senior lawyer-civil servants in the devolved administrations in Scotland and Wales are at Director-General level or its equivalent (the Solicitor to the Scottish Government, and the Director of Legal Services in the Welsh Assembly Government).
More important than the numbers of lawyers in government is what they do. Lawyers in government do legal work, which is narrowly conceived and defined. Very few officials undertaking general policy development work have a legal background or training, except in specialist areas like land tenure. Instead, government lawyers are chiefly concerned with managing litigation to which government is a party (including public-law actions such as judicial review, and private-law ones such as employment tribunal cases), dealing with interests in land and other property, and advising on technical aspects of the drafting and framing of legislation (Daintith and Page 1999, chs 7 and 9; author's interviews confirm this has remained the case in the Welsh and Scottish governments, as well as for the UK government). Advising on the impact of European Union law and the constraints it imposes is an important area, and more recently the Human Rights Act 1998 means that similar issues have arisen with the ECHR. Yet this remains a relatively narrow area of work, compared with the range undertaken by the civil service as a whole. Much work that might be considered ‘legal’ in nature, both in developing policy and administering provisions already put in place, is in fact undertaken by policy officials, often relatively junior ones (Page and Jenkins 2005). The role of lawyers is particularly prominent when it comes to the preparation and drafting of legislation, which is the main area where they come to engage with the political aspects of government (see Hazell and Rawlings 2005, for discussion of how this applies to devolution arrangements in Scotland, Wales and Northern Ireland). Even then, however, their work is conceived in relatively narrow terms—to frame legally workable approaches to policy, to ensure that it operates within such binding constraints as apply (notably EU law and the ECHR), and to apply a relatively small number of areas of ‘legal policy’ (Page 2002 and 2009). Those areas of legal policy do not, importantly, include any questions relating to devolution but rather involve matters such as rights of entry into private property (Daintith and Page 1999, ch. 8; Trench 2005).
This means that most officials concerned with general policy development in the UK do not naturally think in legal terms, nor do they look to the courts as a way of obtaining definitive resolution of difficult issues. Indeed, their inclination is usually to find solutions that do not involve the courts, but are worked out by other means and then given legal form so that they can be implemented effectively. Lawyers play a technical but not policy-making role. In this, the UK may be contrasted with systems such as in Germany or the United States, where large numbers of people in ‘policy’ roles in the civil service have a legal education or training, and are more prone to seek legal solutions to problems.
Non-judicial Resolution of Legal Issues
The fifth factor is the way that even legal disagreements can be resolved without recourse to the courts. The discussion above highlights the extent to which such legal issues are construed very narrowly in the UK's system of government. In essence, they must relate to existing legislation, or binding external constraints such as EU law or the application of the ECHR, and will usually concern the interaction of two existing statutes or established legal rights. Issues of policy are not usually considered ‘legal’, unless they concern legal policy such as rights of entry into property. When such issues do arise, private parties would need, and expect to use, the courts. Government, however, has a wider choice, thanks to the position of the law officers—the Attorney General (and Solicitor General and Advocate General for Scotland) in the UK government, the Lord Advocate in the Scottish Government and the Counsel General in the Welsh Assembly Government. Of these, by far the most important is the UK Attorney General, whose opinions are binding across UK government in the absence of a contrary opinion from the courts (Edwards 1984; also Daintith and Page 1999, ch. 7). This provides an alternative way of resolving legal issues within the UK government without taking them to the courts, with the attendant risks of publicity and uncertainty. Law officers’ opinions are widely disseminated across government through an online database, but as far as the outside world is concerned they remain confidential and are only published in exceptional circumstances and with the consent of their author as well as the minister who sought them.
Law officers’ opinions have been used to resolve at least one difficult issue, of whether the Scottish Parliament's legislative powers extended so that it could bind bodies concerned with reserved matters that operated in Scotland, on subjects like land law that are not reserved. This had proved the subject of considerable debate in 2000–02 (author's interviews). The need to disseminate this opinion to policy officials as well as lawyers (and perhaps beyond the UK government too) meant that its substance was subsequently incorporated in a Devolution Guidance Note, DGN 15 on ‘Scottish legislative proposals giving devolved powers and functions to UK bodies’. For practical purposes, this means that this issue has been conclusively resolved unless a case is taken to the courts, which is unlikely to happen.
At one point in 1999–2000 the idea was even mooted of using joint law officers’ opinions to tackle certain issues—so the Attorney General and the Lord Advocate would produce a common opinion, binding on both the UK and Scottish governments, resolving a particular matter. Subject to any third-party challenge (which would be unlikely), that ruling would provide an authoritative determination of the issue, without involving the courts at all, and would amount to the creation of a system of ‘private public law’: private, because it was dealt with extra-judicially, but relating to matters of public law. This suggestion was, however, rejected by the Lord Advocate, and never proceeded with (author's interviews). The mere fact that it was considered, however, indicates how powerful such law officers’ opinions can be as a tool of government. They also reflect a more political aspect—the desire to deal with such matters in private, away from the public gaze, even though matters of considerable public import were at stake.
The Existence of Political Mechanisms for Dispute Resolution
The arrangements for intergovernmental relations after devolution were designed chiefly with a view to being used to resolve disputes. This reflects an expectation on the part of politicians in the UK that they are the prime decision-makers, and are reluctant to use the courts if they can avoid doing so. These arrangements were set out in Memorandum of Understanding (originally 1999, now 2010, discussed in Trench 2007a and Rawlings 2000). The key institution here was the Joint Ministerial Committee (JMC), particularly in its plenary format, which was conceived first and foremost as a forum for the resolution of such disputes. Its failure to meet between 2002 and 2008 was often explained as happening because there were no disputes needing to be resolved (author's interviews). The Agreement establishing it emphasised the desirability of direct, bilateral negotiations to resolve any disagreements, involving if necessary the territorial secretary of state as well as the Whitehall department and devolved administration concerned, and stated:
The JMC is a consultative body rather than an executive body, and so will reach agreements rather than decisions. It may not bind any of the participating administrations, which will be free to determine their own policies while taking account of JMC discussions. Nonetheless, the expectation is that participating administrations will support positions that the JMC has agreed (Memorandum of Understanding 1999, para A1.10).
There is therefore a clear line of escalation contemplated in the Memorandum of Understanding (and which has survived unchanged through various amendments of it): first, direct bilateral attempts to resolve an issue; second, involvement of the ‘good offices’ of the secretary of state; and third, use of the JMC (Trench 2007a and 2007b). These arrangements were further refined in Memorandum of Understanding (2010), which included provisions relating to ‘dispute avoidance and resolution’ that provided a slightly more structured approach, but largely followed the principles already in place.
Even these procedures did not apply to one potent possible source of disagreement: finance. The UK has retained a highly centralised system of finance, in which a high level of discretion is conferred on HM Treasury to determine whether, for example, a change in spending in England should trigger a consequential increase in the devolved governments’ block grants (Trench 2007c). Here, however, a different set of procedures applied, with any disagreement determined by Treasury ministers, subject to review by the UK cabinet—but not the JMC. 10 They would, moreover, be highly unlikely to give rise to any justiciable issue, even if there were an irreconcilable disagreement about a financial matter.
What was therefore created as part of the institutional framework of devolution was a system that created extensive flexibility at the margins, to enable disagreements to be resolved politically and then implemented in a legal form, without need for the courts to be involved. These are, however, only necessary conditions for limiting the role the courts have played in devolution, not sufficient ones. To identify those, it is necessary to look at the political character of devolution as it has developed since 1999.
Political Factors, 1999–2007
As has been discussed elsewhere in this volume, Labour dominated all three governments in Britain during the ‘first phase’ of devolution (1999–2007), while for much of that period, particularly between 2002 and 2007, devolved government in Northern Ireland was suspended. This prolonged period of single-party domination had a number of consequences. One was the emergence of a political climate in which there was an expectation that devolution would be harmonious, and that intergovernmental differences would not emerge. While this may not have corresponded with the legal or institutional reality, it meant that close media attention focused on any difference between devolved and UK governments, increasing pressure on politicians to avoid such differences. A second, related, effect was the creation of strong pressure on elected politicians to avoid disagreements and disputes, particularly in public. Instead, the tendency was to resolve such matters in private, even when there were strong differences between the two tiers of government. Moreover, such differences would be approached in the knowledge that there would, ultimately, be a resolution of them—they would not turn into more serious disputes, subject to formal consideration in more public settings like the JMC, or litigation (Trench 2007b). This was, in its turn, facilitated by a high level of civil service co-operation and co-ordination, usually at fairly low levels (of those directly working on a particular bill or project), which resulted in comparatively few issues needing ministerial involvement, but a large amount of effort to ‘oil the cogs’ of governmental machinery. Such co-operation in turn meant that the posts of secretary for state for Scotland and Wales in the UK government required decreasing amounts of time, and could be combined with other posts such as Transport, Leader of the Commons, Northern Ireland or even Defence.
In this climate, disagreements seldom turned into public disputes, and those that did were not susceptible to resolution through the courts. They were thrashed out between politicians, sometimes even without civil servants in attendance, and certainly away from the glare of public scrutiny. Once a solution had been agreed upon, it could be implemented by one of the technical mechanisms discussed above, without the binding constraints that a written constitution would impose and which would prevent similar resolutions being found.
While this account may explain the absence of intergovernmental litigation, it does not explain the absence of devolution-related litigation more generally. Such litigation as there has been has related to actions brought by third parties; numerous challenges to prosecutions for breach of Convention rights, for example. The absence of further litigation to challenge unpopular policies pursued by the Scottish Executive and Parliament—the ban on smoking in public places, for example—is harder. This failure of interest groups to pursue legal challenges may be due to two factors. On one hand, many of the civil society interest groups might wish to support the wider project of devolution, or receive financial help or other assistance from devolved institutions. They would regard taking legal action against the devolved institutions, and legal action that challenged the fundamental powers of those bodies, as likely to endanger that relationship and so avoid doing so. On the other hand, the resources to support a legal challenge might not be readily available—whether the funds to support an expensive legal action, or the knowledge of how Scots law and legal practice worked (for commercial organisations, whether ones, operating across the UK or further afield). Such explanations remain no more than hypothetical, however; the only provable point here is the small number even of third-party actions. The emergence of challenges from commercial actors such as general insurers or tobacco companies to devolved legislation that undermine their business interests suggests that this may start to change, as devolved legislatures start to challenge their interests.
Political Factors after 2007
The political landscape changed significantly, of course, in 2007. Partial or wholesale changes in the administrations in Scotland and Wales, and the resumption of devolution in Northern Ireland, coincided with several additional changes, including a change of UK prime minister and the assumption of legislative powers by the National Assembly for Wales. This ‘second phase’ of devolution was characterised by a number of features, including the emergence of (disjointed) constitutional debates in each part of the UK, increased importance of finance in intergovernmental relations, and a renewed emphasis in London on the role of the UK level and ‘Britishness’ (Trench 2008). Subsequently, this has been taken a stage further by the election of a Conservative–Liberal Democrat coalition UK government. It is too early to draw any conclusion on the impact of the new UK coalition, but what is striking is the limited role the courts have played since 2007, notwithstanding the weakening of the glue of party ties that had played such an important role before 2007.
Explaining this is, nonetheless, not difficult. The institutional mechanisms to facilitate solutions to disagreements about the existing constitution remained in place. While the considerations at work since 2007 are different to those before then, the overall outcome has been very similar.
Given the relatively reactive approach of successive UK governments, the central consideration has been how the devolved governments have chosen to approach intergovernmental relations. Such relations have always been central in the case of Wales, but the distinct nature of Welsh devolution means that these tend to be pursued bilaterally rather than multilaterally (Rawlings 2003; Trench 2007d). Administrative factors mean that Northern Ireland has been much less affected by much ordinary UK legislation. The key player has therefore been Scotland. The SNP Scottish Government, in particular, has been keen to put intergovernmental relations on a more formal and quasi-diplomatic footing, and in particular to make greater use of the revised JMC, including a new protocol on dispute resolution encouraging disputes to be resolved at the political level (see Gallagher, and Cairney, in this issue). Furthermore, as Paul Cairney explores in more detail, the SNP has clearly made a political choice to demonstrate its competence as a party of government, differentiating itself from the UK government while restraining any urge to find disagreement for the sake of it.
The combination of political pragmatism and use of technical mechanisms to resolve a difficult issue can be seen in the issue of compensation to prisoners who had had to ‘slop out’ their own bodily wastes in Scottish prisons. The practice was abolished in prisons in England and Wales by 1996, but not in Scotland, thanks to a decision made by the pre-devolution Scottish Office. Several cases had established that the principle of ‘slopping out’ was contrary to the Convention rights, and in Somerville v. Scottish Ministers [2007] UKHL 44, the Scottish Government's last attempt to avoid having to pay compensation (by arguing that these claims were barred by a time limit that applied to such claims made under the Human Rights Act but not the Scotland Act, and that the Human Rights Act time limit should apply to all such claims) was dismissed by the House of Lords. The sum of £76 million had reportedly been set aside to pay such claims, which could go back many years in the absence of a time limit. There was, however, political support from the UK government for the principle that the Scottish Government should not have to pay claims dating back more than one year. In March 2009 a joint statement by the first minister and secretary of state for Scotland announced their agreement on this, and how it was to be achieved. An order under section 30 of the Scotland Act was made amending the reservations of powers in Schedule 4 to the Act, and following that legislation was passed at Holyrood to bar such claims (see Scotland Office Press Release, ‘Somerville: Joint Statement from Secretary of State and First Minister. Agreement to End Slopping Out Anomaly’, 19 March 2009, available at: http://www.scotlandoffice.gov.uk/scotlandoffice/11541.html; The Scotland Act 1998 (Modification of Schedule 4) Order 2009, SI 2009 no. 1380; and the Convention Rights Proceedings (Amendment) (Scotland) Act 2009). 11
In the case of Wales, the very fluidity of the National Assembly's legislative powers has meant that a large amount of political effort has gone into the issue of conferring legislative powers on the Assembly. This has been particularly the case for those conferred by Legislative Competence Orders, even though the conferral of powers directly by primary Westminster legislation (so-called framework powers) has been arguably more important substantively. Because Westminster's powers are used to confer all such powers, no issue capable of determination by the courts can be presented by them (indeed, the Government of Wales Act 2006 is careful to exclude any appeal to the courts over the making or refusal to make a Legislative Competence Order). However, one trend has emerged as a result of the process: in the quest for legal certainty, and in order to make sure that the powers conferred on the Assembly will withstand scrutiny by the courts, their drafting is highly detailed and precise. However, the price of legal certainty is that the scope of devolved powers is hard for the general public to find or understand. This may be simplified by the majority vote in favour of the wider legislative powers set out in Part 4 of the 2006 Act, at the March 2011 referendum.
Beyond this, the devolved governments have taken pains to find ways of separating constitutional debates from ones about day-to-day policy matters. In Scotland, this revealed itself in two separate processes, the Scottish Government's ‘National Conversation’ and the Commission on Scottish Devolution, supported by the three unionist parties. In Wales, this took the form of the All Wales Convention (2009) chaired by Sir Emyr Jones Parry. While these constitutional debates were fragmented and in different ways flawed (see Trench 2009), they moved constitutional debate out of the arena both of day-to-day politics and of matters involving the courts or issues that might be justiciable. A partial exception might relate to a bill calling a referendum on Scottish independence, expected in the current Scottish parliamentary term. Such a legal challenge would involve not only choices about whether to pursue litigation, with its political disadvantage of seeming to deny the Scottish people a choice about their future, but also technical issues of what is and is not within the parliament's existing legislative competence, which itself would depend on the referendum question proposed (discussed in Murkens et al. 2002, chs 2 and 3; for proposed questions, see Scottish Government 2010).
The Impact of Changing Judicial Architecture
The UK Supreme Court started sitting on 1 October 2009, though the legislation creating it (the Constitutional Reform Act 2005) had been on the statute book for some four years. The impact of the new court on devolution is already perceptible, although it is likely to take some time for all its implications to emerge.
The establishment of the Supreme Court puts an end to one problem of devolution, already discussed above—the overlapping jurisdictions of the Judicial Committee of the Privy Council and the Appellate Committee of the House of Lords, and the problems of a ‘dual apex’ to the judicial system when it came to matters concerning devolution. This had already become apparent in the December 2007 judgment in the case of Spiers v. Ruddy [2007] UKPC D2, where the Judicial Committee overruled its earlier opinion with which the House of Lords had disagreed (and which had also been the subject of an opinion of the European Court of Human Rights). However, the establishment of the new Court had itself been the subject of behind-the-scenes controversy, as the UK government's proposal to establish it (Department for Constitutional Affairs 2003) was made as part of what became known as the ‘botched reshuffle’, without any prior consultation or discussion with the devolved administrations. While the Scottish Executive had no substantive objection to the proposal, the fact that its very direct interest in the issue of how devolution disputes might be resolved could be overlooked was a source of major concern. The Court itself has been at pains to emphasise the extent to which it represents all parts of the UK, such as by ensuring that symbols of Scotland, Wales, Northern Ireland and England all figure equally in its logo and decoration.
Two particular sources of increasing pressure on the courts, particularly the Supreme Court, to take a more active role in relation to devolution can be identified. One is increasing litigation brought by third parties. While, in a sense, criminal prosecutions ‘happen’, other parties may decide actively to use the courts in pursuit of a political or commercial objective. An indicator of this tendency is the challenge brought in the (Scottish) Court of Session in June 2010 to the restrictions on the sale of tobacco in the Tobacco and Primary Medical Services (Scotland) Act 2010.
The second is an increasingly prominent role for the courts more generally. This has various causes, but prominent among them is a perceived increasing assertiveness of the judges following the establishment of the UK Supreme Court, and the idea that this will make judges more conscious of their independent role. Whether this will materialise remains to be seen. However, the impact of human rights may create some sort of model, expanding both judicial activism and a view on the part of the judges that this is necessary and legitimate (Nichol 2006). Coupled with a more general tendency for judicial resolution of public law matters, the likelihood is strongly that the courts will take a more active role in future (Le Sueur and Malleson 2008).
Against this background, Martin and Miller v. Lord Advocate [2010] UKSC 10, decided in March 2010, is likely to prove a landmark case. Although this was the third time the UK's highest courts had been asked to determine whether devolved legislation was within competence or not, it was the first time the task had fallen to the UK Supreme Court since 2009, and it revealed serious differences in the views of the judges. The case concerned an extension of the sentencing powers available to sheriffs sitting on summary business (not with a jury), which were increased by the Criminal Proceedings etc. (Reform) (Scotland) Act 2007 (an act passed in the second session of the parliament, and proposed by the Labour–Lib Dem coalition). When the use of enhanced powers to sentence for the offence of driving while disqualified was challenged, the High Court of Justiciary (sitting as an appeal court) found the new provisions within competence. The Scottish Executive and the Presiding Officer had earlier done so too, when the legislation was before the parliament. The court decided by a three to two majority that the legislation was within competence, but as well as being a majority decision the two Scottish judges on the court took very different views not just on the substantive issue but on how to approach considering such a case. Lord Hope of Craighead, and the majority, applied a purposive approach, looking to the overall intended effect of the legislation in question. They discussed (though did not actually apply) the statutory ‘pith and substance’ test’ (section 29(3) of the Scotland Act 1998, requiring the courts to take into account by ‘having regard … to [the provision's] effect in all the circumstances’). 12 Lord Rodger of Earlsferry (in the minority with Lord Kerr of Tonaghmore) took a more conventional approach for Anglo-Saxon judges, using a fine-grained legal analysis of the wording of the Act to determine its meaning, and noting the existence of the order-making powers that exist to enable the Scottish Parliament to legislate where (as here) it lacks the power to do so but there is support from the UK tier for what it wishes to do. These differences of legal view on such an important question are a source of concern in themselves; they may also encourage litigation, as parties (particularly those seeking to overturn devolved legislation) may hope that a differently composed court might come to a different conclusion.
With the revived Scottish constitutional debate since 2007 as well as the establishment of the new UK Supreme Court, it is perhaps unsurprising that the Scottish Government has sought to review the arrangements for dealing with devolution litigation and the relationship between Scottish courts and those elsewhere in the UK. The report it commissioned from Professor Neil Walker of Edinburgh University's Law School (Walker 2010) found that the present arrangements were unsatisfactory in themselves, and given the overall nature of constitutional relations between Scotland and the UK recommended a ‘quasi-federal’ court to deal with Scottish cases raising common UK issues, with other cases to be considered only by the Scottish courts. These debates have been fuelled by the Supreme Court's decision in the case of Cadder v. Lord Advocate [2010] UKSC 43, concerning the availability of legal advice when suspects were being questioned. This led to a significant change in Scottish practices for police questioning, and also caused the Scottish Government to establish a wider review of the issues chaired by a Court of Session judge. Other devolved governments have taken a less prominent role in such debates, although there is ongoing debate about the need for a separate legal jurisdiction in Wales as a result of the legal differentiation arising from devolution (Jones and Williams 2004), a position supported by a number of members of the judiciary as well as Plaid Cymru. Welsh thinking, however, does not seem to have addressed specifically the nature of a court system, whether or not separate legal jurisdiction for Wales might be established.
Conclusion
The courts have played a limited role in the working of devolution, and intergovernmental relations, since 1999. This reflects deep-seated features of the UK's political constitution, its administrative culture, and the strategic and tactical choices made by politicians about how they would manage such issues. This approach has been made possible by the relative ease with which the division of powers between UK and devolved tiers of government can be adjusted, and the frequency with which that has been used. These features have very largely persisted through the changes of party control of governments that took place in 2007 and 2010. In this respect, the UK has been a marked exception to patterns of intergovernmental relations in most federal and decentralised states, where the courts become significant political actors to which governments as well as third parties often have recourse.
The limited role of the courts is coming under some degree of pressure. In addition to litigation arising from third parties, the number of technical legal issues needing judicial resolution may increase. Nonetheless, it is hard to see an upsurge in intergovernmental litigation, particularly on issues of political sensitivity, as that is unlikely to advantage any particular government whatever party or parties are in office. Each, after all, competes for support from the same electorate, and to resort to the courts implies both a serious failure on the part of all the politicians involved and a loss of control about matters that politicians regard as their province. The pursuit of constitutional debates in other arenas accentuates this.
One particular factor might change the situation radically, however. That would be a UK government that took a restrictive view of devolution, and was determined to limit the autonomy of the devolved governments and legislatures by rigorously policing the existing arrangements. Such an approach would be a high-risk strategy politically, and has been eschewed by both Labour and Conservative–Lib Dem UK governments to date. Yet a restrictive approach from central government toward local government in the 1980s and 1990s led to an upsurge in litigation, giving the courts an increasingly prominent role as political actors and leading to a ‘juridification’ of central–local relations (Loughlin 1996). The many grey areas and areas of overlapping functional competence in the present arrangements make it very doubtful whether the present system of devolution could survive a similar change.
Footnotes
An earlier version of this article was presented at an ESRC-funded seminar on ‘Dispute Prevention and Resolution: The Role of the Civil Service and the Courts’, Edinburgh, July 2009. Interviews between 2001 and 2004 were carried out while I was working at the Constitution Unit at University College London on projects on ‘Intergovernmental Relations in the United Kingdom’ (part of a research programme on ‘Nations and Regions: The Dynamics of Devolution’ funded by the Leverhulme Trust), and also on ‘The Law and Devolution Disputes’ funded by the ESRC, grant L219 252 015. I am grateful to both bodies for supporting that phase of my work, and to the two anonymous referees who provided helpful comments on an earlier draft. The text covers legal developments up to 30 June 2011.
1
Interviews were carried out on the basis that they would be non-attributable, and no interviewee would be identified publicly. Unless otherwise noted or made in qualified terms, statements drawing on interviewing have been confirmed by at least two interviewees.
2
Human rights issues have a greater importance for legislation made under devolved powers than by the UK parliament, as any legislation or action that infringes the Convention rights is automatically void: Scotland Act 1998, section 29; Government of Wales Act 2006, sections 94 and 108; Northern Ireland Act 1998, section 6.
3
The executive arm of devolved government in Scotland was called the Scottish Executive on its creation in 1999. In 2007, the SNP administration rebranded it the ‘Scottish Government’. The correct title is used depending on the time referred to.
4
This controversy arises because of the strong tradition of ‘legal nationalism’ in Scotland, which has emphasised the importance of the distinctive legal nature of Scottish law and the Scottish legal jurisdiction. See, generally, Meston et al. (1991) and for the wider character of legal nationalism, 
, particularly ch. 5). This distinctiveness pre-dates but survives the Union of Scotland and England in 1707, and (as Kidd points out) by no means overlaps with political nationalism.
5
Anderson, Reid and Doherty concerned the Mental Health (Public Safety and Appeals) (Scotland) Act 1999, and whether this was acceptable in the case of those detained in a secure mental hospital whose mental condition was considered to be untreatable. DS v. Lord Advocate concerned section 10 of the Sexual Offences (Procedure and Evidence) (Scotland) Act 2002, which provided for disclosure of a defendant's prior convictions when being tried for sexual offences. Whaley v. Lord Advocate concerned the Protection of Wild Mammals (Scotland) Act 2002. Convention rights as set out in the ECHR were at issue in all three cases, though Whaley also raised issues of compliance with wider UK obligations. It is notable that two cases were raised as a ‘devolution issue’ and were heard by the Judicial Committee, and the other was not and was heard by the House of Lords. There have been other ‘devolution issue’ cases, but these have involved challenges by individuals, usually on human rights grounds, to actions of devolved executives.
6
In September 2011, the UK Supreme Court gave judgment in a further devolution case, Axa General Insurance v. Lord Advocate [2011] UKSC 46. The case concerned the validity of Scottish and Northern Ireland devolved legislation which reinstated civil liability for the industrial disease known as ‘pleural plaques’, removed by a 2004 House of Lords decision. The case raised legal issues of whether such legislation breached the ECHR because it had limited retrospective effect, and whether devolved legislation was subject to a test of its reasonableness at common law. The Court decided ECHR obligations were not breached, and that the threshold for determining the lack of ‘reasonableness’ was a very high one indeed.
7
Executive devolution works rather differently. Here, with the exception of a small number of functions which are exercised jointly (and are expressly defined), powers are exclusive: one government, and its ministers, have the sole power to determine the matters involved. Legislative devolution creates concurrent powers, but not executive devolution.
8
There is an important difference between the provisions for Northern Ireland and those for Scotland and Wales; in Northern Ireland, a bill may not be introduced if the Presiding Officer decides that any of its provisions are beyond the Assembly's legislative competence. In Scotland and Wales, the statement is for information only and does not debar consideration of the bill.
9
This arises by virtue of the application of the Sewel convention, as set out in the various Devolution Guidance Notes (DGNs) which are now the responsibility of the Cabinet Office. The notes have the status only of internal documents, but reflect discussions and (on this point at least) agreement with the devolved administrations, and in the case of Scotland are echoed by published internal guidance there too. The most pertinent notes are DGNs 8, 9 and 10 (Post Devolution Primary Legislation for, respectively, Northern Ireland, Wales and Scotland). They are available at:
11
The press release also refers to the UK government seeking parliament's support ‘to bring forward a comprehensive solution extending the same protection to the devolved administrations in Wales and Northern Ireland putting all of the devolved jurisdictions on a consistent footing and consolidating the changes to the Scotland Act’, which has not happened to date.
12
The ‘pith and substance’ test is an approach developed by the common law courts, particularly the Judicial Committee of the Privy Council, to determining whether legislation was within the powers of a legislature. The Judicial Committee particularly applied it to the Canadian constitution, while it was the court of final appeal until 1948. It was also applied to the powers of the Stormont Parliament in Northern Ireland by the House of Lords sitting judicially, in Gallagher v. Lynn [1937] A.C. 863 (HL).