1. Professor John Finnis, in his October 2015 lecture “Judicial Power – Past Present and Future”, said that “‘Past, present and future’ captures a good deal of the truth…about the distinctions between judicial, executive and legislative powers.” On this basis, concern about whether the exercise of judicial power is increasingly encroaching on issues beyond those for which the judiciary has constitutional responsibilities – and applicable skills – should concentrate on whether they have ceased to confine themselves to the past.
2. However, this concern should not be confined to the legislative function. It should relate more generally to any interference with the formulation of policy for future change. It is also immaterial whether the interference is during and in respect of the process of policy formulation or, using judicial hindsight, takes place subsequently in respect of the effects of the policies. Retrospective review interferes with the formulation of future policy because the extent of the risks it involves for different options becomes a factor that influences the choices between those options – often, perhaps, the determining factor.
3. For practical purposes, all primary legislation and most secondary legislation is about giving effect to policies for change. But not all such change requires legislation. Courts confining themselves to the past would not carry out the judicial review of primary legislation; but that is not all.
4. Governments are elected on the programmes they submit to the electorate for making things better. They define what they mean by “better”; and, in office, they work out the real-world changes that need to be produced to implement their promises. These changes can often be produced wholly or partly in reliance on existing powers. So the use of existing powers to give effect to change needs to be distinguishable from their use for the day to day operation of an existing system in accordance with its established rules. It is the substance of the matter that is important in identifying whether decisions are about the future, and not whether the policy also involves a change that takes a legislative form.
5. Legislation is needed to give effect to policies for change only where there are legal obstacles to the real-world change needed for making things better. Often the legal obstacle will be an absence of adequate incentives to change behaviour, whether generally or by particular decision-makers. But legislation’s intended contribution to change is, usually, to bring about behavioural change without the need for enforcement or litigation. Having an effective system for resolving disputes to which any proposed legislation may give rise is only a secondary consideration. Although it is an important consideration, the priority is always to secure the practical benefits that represent the ultimate policy objective.
6. At a subsequent election, a government will have to account to the electorate for the way they have discharged their mandate for change. That is the essence of the democratic system; but governments can only be held democratically accountable – and be expected to accept responsibility for their policies – if they have the freedom to formulate them. Democratic accountability is incompatible with a capacity to shift the blame for failures, or for inaction, onto inhibitions imposed by unelected watchdogs. Judicial functions need to leave room for political accountability by governments both for what they do and for what they do not do.
7. Moreover, judicial influence on policy formulation is problematic quite apart from the way in which it diminishes democratic accountability. The judiciary’s primary function of providing for the resolution of disputes between parties creates an approach – a way of thinking – that makes it ill-suited to the task of assessing the justification for policies for change.
8. All policy for change has an element of unpredictability and risk about it. It involves operating on a stochastic process. It always has to be analysed by reference to its likely impact on a whole system, as well as on those immediately affected; and it has to accommodate its interaction with the rest of a government’s programme. It nearly always involves losers as well as winners. It very often involves a potential “win” for what is seen as the wider public good at the expense of a certain “loss” for particular parties.
9. It may often involve “aiming off” or “overshooting” to account for irrational behaviour or for financial or other practical constraints. For example, the instincts of the naturally law-abiding citizen (often the main target) – or, perhaps, of the instinctively cautious regulator – are to leave a safe space between their own conduct and the parameters of what is allowed by the law. Policy has to strike a balance between this and the fact that there will always be others who will try to push those parameters to the limit, or beyond. Similarly, the effectiveness of a sanction may depend on balancing its impact and severity against any financial or other constraints on actually enforcing it in practice.
10. Time and money are important but not the only constraints. Another relevant factor is the capacity, by the exercise of leadership, to make the change stick with enough of those who have opposed it, often those most closely affected.
11. Finally, change is a thing at which policymakers may often, in practice, have only one shot. If the first shot falls short, the initial failure will certainly make a second attempt more difficult, both politically and practically, and may rule it out altogether. Allowing the process of change to become protracted can be enough to render it impossible, whatever its potential for producing real benefits.
12. None of this fits with the culture, training or methodology of the courts. The principles on which the rule of law is based and by which the courts test whether a legal obligation or liability is fair depend to a very large extent on the idea that the consequences of actions or omissions should be predictable. When choosing between policy options, the most predictable outcome and so the most difficult to challenge, is the “do nothing” option, which Civil Service practice requires in every policy paper. Where a policy of change is pursued, the public “win” is always going to be less certain, and its extent less predictable, than the “loss” to the party before the court. In the law, the best is seldom the enemy of the good. The same is not true of practical politics.
13. The legal concept of “proportionality” – a test which, it appears, is now regularly applied to policy decisions by the courts – seems to involve only a theoretical and abstract balancing of the significance of a legal mischief against the severity of the consequence that is aimed at it. No allowance appears to be made for the pragmatic and political factors that result in “aiming off” or “overshooting”. In any event, the significance of a mischief is always bound to be a matter for political judgement. It can only be assessed properly in the context of the system within which it exists and by reference to the political importance of the ultimate policy objectives and of the related benefits to be secured by curing it. Assessing it in isolation, or in retrospect by reference to its impact in a particular case, makes no sense at all.
14. Furthermore, the impartiality required of the judiciary disqualifies them from themselves exercising the leadership that a responsibility for change necessarily involves, and also – you would think – from reviewing the leadership of politicians and their arguments for change.
Where we are
15. So are the courts involving themselves inappropriately in the formulation of policy for change? And, if they are, why?
16. The answer to the first question seems to be a clear “yes”.
17. Prof Finnis’ lecture and other contributions to this site clearly show that the courts are increasingly going beyond the question whether rules made as a result of political decisions for change have been applied, and are finding ways to question the justification for the policy implemented by the rules themselves, whether directly or, by reviewing the process by which the policy is arrived at, indirectly.
18. A whole series of cases illustrates this. Examples of the direct questioning of the policy justification include R (PCSU) v Minister for the Civil Service  EWHC 2041 (where the court was prepared to assess the justification in the interests of the national economy for disappointing public servants’ expectations of what they might expect if they were made redundant, but did allow the Minister a high degree of “margin appreciation”) and Reilly (No. 2) (in which no similar allowance appears to have been made in respect of a retrospective piece of legislation overturning legislation the intention of which was known and clear when it was made.) There is also the recent ECJ case of Scottish Whiskey Association v Lord Advocate (in which the Scottish courts have been given the job of determining whether taxation would be as effective at protecting human life and health as minimum pricing.)
19. Unsurprisingly, given the law’s inherent partiality for the status quo, the justification for change seems too frequently to run into trouble in the courts. But even when it does not, the mere existence of the courts’ willingness to review policy justifications creates a significant chilling effect on political decision-making. Opportunities for review are frequent and can become protracted. The effect is obstructive and the application of different standards from those used by the electorate, on their different timetable, tends to incentivise a risk-averse approach to change, and to discount the relevance and value of political leadership.
20. Why do the courts now seem more willing to question policy justifications than perhaps they once were? There are several possible explanations, although they do not, either individually or taken together, provide a legitimate reason for creating inhibitions on policy-making for which there is already an established, and relatively effective, system of political accountability. In any event, it certainly cannot be the role of the courts to adjudicate on the merits of that system or to initiate constitutional change to compensate for its perceived defects.
21. Nevertheless, one explanation may be found in the impression that the courts and the legal profession now share what seems to be a wider populist disillusion with our political system. This disillusion seems, unfortunately, to be actually strengthened by the courts’ response to it; but, in the first instance, it seems to be built on a false premise. It is a view that assumes, wrongly, that Parliament is just “a rubber stamp”. This is thought, amongst other things, to be demonstrated by how infrequently the Government is seen to be defeated in Parliament – a view that represents a total misconception about how Parliamentary influence is actually exercised. Political processes tend, for the most part and in practice, to work through consensus and compromises in preference to victories and defeats.
22. So it is true that nearly all amendments made to Bills in Parliament are moved on behalf of the Government; but that disregards the fact that a very significant proportion of them represent concessions to points made by others. It also completely disregards the very significant, but invisible, influence which Parliament has on policy formation, as a result of the necessary acceptance by all governments of a practical need to anticipate Parliamentary difficulties and, where possible, to avoid those difficulties in the proposals they bring forward.
23. A second explanation for developments in the judiciary’s approach to policy may relate to failures properly to recognise the distinctions mentioned above between the formulation, the implementation and the operation of policy. These are important distinctions but they are inevitably less than clear-cut, because the processes merge and overlap. A policy for change will often leave some details to be worked out in practice in the course of implementation. A requirement for consistency may turn regular practice in operational activities into a policy for the future.
24. In addition, the difficulties of drawing distinctions between the formulation and implementation of policy and operational activity are aggravated by increases in the number of layers of decision-making – in other words by more “devolution”, in its widest sense, within the process of political decision-making.
25. When demarcations have to be made between the remits of different layers of the policy-making process, it is easier to see the decisions taken at the lower levels as merely implementing the policy formulated higher up, or even as the day to day operation of that policy. Instead, they may be better analysed as part of discrete elements of a distributed, and – ideally – collaborative, policy-making process.
26. When the demarcations between different layers of such a process are set out in legislation, the courts may assume that that creates a reason why everything within the established parameters should be justiciable. This applies whether the layers are domestic or are the result of international arrangements, such as those provided for by the EU and the ECHR. Some regulation of policy formulation within internationally agreed constraints may be unavoidable, from a legal point of view, where the constraints are incorporated into domestic law; but it is not inevitable or desirable, that this should be allowed to influence the way domestic law works in other areas.
27. An interesting example of the sort of difficulty that can arise in the domestic context is contained in the Government’s draft Wales Bill, published in October 2015. This sets out a proposal for a new, reserved powers model for devolution in Wales. For this purpose, it incorporates some flexibility into the reservations and restrictions it would impose on the legislative competence of the National Assembly for Wales. The flexibility allows provisions going beyond the basic competence of the Assembly where they have “no greater effect … than is necessary to give effect to” [a provision that is otherwise within devolved competence]. The enactments setting out the added flexibility are necessarily justiciable, because they set the limits on what the Assembly has power to do; but they appear to give the courts a task which they are ill-fitted to take on – namely, to determine all the potential options for implementing the devolved policy, to decide which of those are necessary for that purpose, to decide the effect of each and then to decide which has the “least effect” in the area where flexibility is allowed. (For more on this, see p30+ of Challenge and Opportunity: The Draft Wales Bill, a joint report by the Wales Governance Centre at Cardiff University and The Constitution Unit at UCL).
28. This example leads on to a consideration of the influence of the way legislation is drafted. There are no limits on what can be put in legislation. There is no formal requirement for legislation to contain only matters intended to be justiciable. That would deny the sovereignty of Parliament to use legislation however it wants. But it is certainly the case that putting things in legislation that are intended to clarify constitutional relationships without making them justiciable carries risks. In the cases where something needs to be said without any desire to transfer power to the courts, the risks need to be avoidable by the use of complete clarity about what is intended. But making things clear should, in turn, create an obligation on the courts to recognise the intention that has been clarified – something which, as Prof Ekins and Prof Forsyth argue in their report Judging the Public Interest, the Supreme Court appears not to have done in the case of Evans v AG.
29. The present situation has the potential for continuing constitutional conflict, at least until there is a greater understanding and respect for the political process within the judicial system. There needs to be a greater recognition of the way in which policy-making for change, and related law-making, are processes that can only be undertaken with political leadership, in a practical context and within a system of political accountability. Neither judicial review nor democratic political accountability can work effectively if they have to compete with each other, or if, when operating in parallel, they apply different standards. There needs to be more clarity about where the law cannot and will not go. The current situation creates a perception that the courts will never rule anything out in advance, so far as interfering with policy formulation is concerned. Keeping the options open in this way is something that it is difficult for policymakers to accept, not least because their own functions require them to make clear and often, in practice, irreversible policy decisions for the future, and to commit themselves to them.
Sir Stephen Laws KCB, QC(Hon), LLD(Hon) is a former First Parliamentary Counsel (2006-2012). In recent years he has served as a member of the McKay Commission on the consequences for the House of Commons of devolution and, most recently, as a member of the advisory panel for Lord Strathclyde’s review of secondary legislation and the primacy of the House of Commons. He is Senior Associate Research Fellow at the Institute of Advanced Legal Studies, an Honorary Senior Research Associate at University College London and an Honorary Fellow of the University of Kent Law School.