Daniel Kalderimis* Chris Nixon# and Tim Smith§

4.1 Introduction

In this cross-cutting theme paper, we consider the relative importance of certainty and discretion in regulatory design. We use as a starting point for our analysis the research streams in the New Zealand Law Foundation Regulatory Reform Project.[1]

It is helpful to begin by roughly defining what we mean by certainty and discretion.

By certainty we simply mean ex ante predictability as to how a given regulation will be applied and enforced by regulatory agencies. A total lack of certainty in regulation can be assimilated to arbitrary and capricious power: the anathema of the rule of law. The objective of certainty is hard to doubt and is reflected in the recent Treasury publication, Best Practice Regulation Model,[2]which states:[3]

The regulatory system should be predictable to provide certainty to regulated entities, and be consistent with other policies.

Certainty is desirable from an economic perspective since it has a central role in promoting efficient economic activity – i.e. low cost economically inventive behaviour. The more certain property rights, the domain of legal activities and the more predictable the actions of institutions, the more confidence business has to invest. And, perhaps more importantly, the more ‘room’ there is to develop innovative ways of doing things better. Finally, there is a reduction in social costs when the bounds of acceptable actions are clearly known.

Uncertainty cannot, however, be eradicated because regulation must be formulated and applied in a complex world, in which not all relevant facts are known, causal relationships are uncertain, risks are unable to be precisely quantified, and the impact of regulation on behaviour is unable to be predicted with complete confidence.

Discretion, in contrast, is the freedom or authority of a regulatory agency to act according its own judgment. It confers choice as to whether (or, in some cases, how) to exercise regulatory powers. Regulatory discretions typically require the decision-maker to exercise its own value judgement within the parameters of legally acceptable options.[4][5]

Discretion permits focused, tailored, case-specific solutions. Many problems have more than one legally, economically, socially or environmentally acceptable outcome. Discretion can be beneficial from an economic perspective since it allows regulators and institutions to deal appropriately and fittingly with conditions that are unique, or to devise high level rules than can assist in best matching the treatment to the issue – dealing innovatively with a series of one-off events or processes, within the ambit of a broad purpose or principle.

On a classical legal account of rule of law values, the tendency has been to place certainty and discretion as diametric opposites. Economic literature, in contrast, has not been so dogmatic, as even the brief discussion above illustrates. At the same time, the economic perspective does not always capture (or, perhaps, always give sufficient weight to) societal values such as the rule of law, which necessarily inform choices of regulatory design in relation to certainty and discretion.

This is a substantial topic, on which much has been written. Rather than seek to join issue with the debates in that literature, this paper has the more modest aim of assisting in focussing discussions of certainty and discretion, as seen through the prism of the research streams of the New Zealand Law Foundation, by considering the relationships between regulatory values, objectives and mechanisms.

In this paper, we suggest that certainty and discretion should not be seen as diametric opposites; nor, indeed, the same type of concept. Whereas certainty is an objective of regulatory design, the degree of discretion afforded to institutions is a mechanism by which various objectives of regulatory design, including but not limited to certainty, may be achieved. Thus, the use of discretion will, in some circumstances, increase certainty; in other cases, discretion may reduce certainty but serve other objectives (such as flexibility or durability), and the reduction in certainty caused by the presence of discretion may be mitigated by other mechanisms.

The extent to which the objectives of certainty and durability are sought in a particular regulatory design, and the choice of mechanisms by which those objectives are to be achieved, will depend in part on the values of the society in which the regulation is placed, and the factual context of the regulation. In this paper, we focus on values of legitimacy and efficacy, and consider the linkages between those values, the objective of certainty, and the mechanism of discretion.

As the examples in the research streams of the New Zealand Law Foundation Project illustrate, there is no single analysis that can work in all cases and there is no optimum balance between certainty and discretion. This is partly because regulatory problems are not all the same size and shape. It is also because regulatory problems have different difficulties associated with their amelioration.

4.2 A traditional legal account

On a traditional legal account, certainty is said to be inherently desirable from the perspective of the rule of law; conversely, “the uncertain and crooked cord of discretion” is said to be the antithesis of “the golden and straight metwand” of law.[6] A V Dicey provides the classical statement:[7]

[The rule of law] means, in the first place, the absolute supremacy or predominance of regular law as opposed to the influence of arbitrary power, and excludes the existence of arbitrariness, of prerogative, or even of wide discretionary authority on the part of the government. Englishmen are ruled by the law, and by the law alone…

Lest this be thought to be a view of historical significance only,[8] Lord Bingham’s 2010 book, The Rule of Law, incorporates a chapter “Law not Discretion”,[9] which argues for the proposition that: [10]

questions of legal right and liability should ordinarily be resolved by application of the law and not the exercise of discretion

In contrast perhaps, it has been suggested within the last decade that “scholars of regulation rarely pay much attention to the rule of law.”[11] Scholars of regulation, typically take a realist perspective, arguing that discretion is always present, and that even formal rules are shot through with discretion due to the inevitable judgements that must be made in applying legal standards to factual situations. Dr Julia Black, one of the foremost scholars of regulation in recent times, argues that the proposition that: [12]

… discretion can be ‘managed’ both to confine its exercise to certain actors and to limit the way that those actors use the discretion given, is thus ambitious, if not misguided.

In practice, neither side of the debate takes their position to its logical extreme, by which discretion is either excluded (which is impossible) or unmanaged (which is undesirable). Lord Bingham concludes his discussion by confirming that: [13]

… the rule of law does not require that official or judicial decision-makers should be deprived of all discretion, but it does require that no discretion should be unconstrained so as to be potentially arbitrary.

Dr Black concludes her discussion by noting that persons exercising discretion should: [14]

… act within the law, and in accordance with principles of rationality, proportionality, consistency, non-discrimination, transparency and due process.

Nonetheless, within the divergent perspectives of those who can be categorised, in admittedly oversimplified terms, as ‘traditionalists’ – who support certainty for rule of law reasons; and the ‘pragmatists’ – who support discretion for outcome-oriented reasons – a number of implicit claims about the relationship between regulatory values, objectives and mechanisms can be seen. We illustrate these claims in Figure 4.1.

Figure 4.1: Linkages between values, objectives and mechanism on a traditional legal account

On our simplified account, traditionalists tend to assimilate certainty with legitimacy. Pragmatists tend to assimilate flexibility with efficacy. However, neither association always holds. Modern literature and experience indicates that, in some contexts, greater overall legitimacy may be achieved by more flexibility; and, in other contexts, that greater efficacy can be achieved by increased certainty.

A number of further implicit claims are made by traditionalists and pragmatists as to the regulatory mechanisms most apt to give effect to the desired values. At the risk of making further simplifications, traditionalists tend to assimilate ex ante rules with certainty; pragmatists tend to assimilate standards and discretion with flexibility. Again, however, neither association is entirely robust: standards and discretion, particularly when layered with other mechanisms, may provide more certainty than rules in some contexts; by contrast, rules are not necessarily antithetical to the long-term adaptability of a regulatory system. Constitutions and religious texts are amongst the best-known deontologies, or rule-based systems of regulation.

Contemporary debates on regulation, whilst recognising the importance of legality to the value of legitimacy, have moved away from doctrinaire claims that only semantically precise rule-based forms of regulation are valid and legitimate. As Braithwaite has argued, it is frequently unclear which techniques of regulatory design (for instance, the use of principles, standards, or a layered combination of the two) are more likely to achieve certainty, given that this involves predicting the response of law subjects in a real world situation, faced with various legal and social constraints.[15]

Accordingly, perhaps the most frequently encountered, but also most difficult, issue for contemporary regulatory design is arguably not concordance with the rule of law, but the way in which the concepts of certainty and discretion interact with the values of efficacy and legitimacy. This is so because:

· it is frequently unclear which of certainty or discretion is more likely, as a matter of regulatory design, to achieve the regulatory object or objects; and

· even if the path to efficacy can be identified with clarity, efficacy is not merely a matter of regulatory words, but of the practical implementation and administration of regulatory systems.[16] Thus, efficacy requires an investment and comes at a cost. The true cost required to achieve efficacy may be unknown at the outset and grow significantly over time.

For a country such as New Zealand seeking to address the full range of national regulatory problems with limited resources, this last matter is an especially important issue.

4.3 Regulatory values, objectives and mechanisms

As our brief discussion of the traditional legal account indicates, in seeking to assess the relative virtues of certainty and discretion as a matter of regulatory design, it is essential to identify benchmark values against which the competing costs and benefits can be assessed. Crucial to this process is the ability to obtain the data needed to do this analysis.[17]

To strike an appropriate balance in individual cases between certainty and discretion, we need a value framework enabling a principled discussion about regulatory design to take place in each context. This must revolve around how society seeks to regulate its members. Therefore, it follows that most beneficial regulation will describe and embed society’s values within regulation as far as possible.

That balance must reflect the reality that there is always a tension between those doing the regulating and those being regulated – otherwise there is no point in the regulation – and that this tension is dynamic. Situations change, new technology enables new possibilities and if the incentives are large enough some will always try to find a way around the rules. The optimal approach to giving effect to regulatory values in any particular context may therefore change over time.

As signalled above, for the purposes of this paper, we have identified legitimacy and efficacy as two primary values which inform issues of certainty and discretion in regulatory design. We do not claim that these are the only relevant regulatory values that might be identified.[18] For example, as an OECD nation – in itself a politically mandated value – New Zealand has as one of its objectives economic growth. Thus, as the Treasury has argued, one criterion new regulations should be tested against is their tendency to support or restrict a growth agenda.[19] Other high-level values which regulation might seek to reflect might be to increase societal diversity or tolerance, or to limit inequality.

However, for the purposes of keeping this paper within a rigorous and confined scope, we look to the values of legitimacy and efficacy. To serve those two values, different objectives of regulation may be identified. One of these is certainty. To offer a counter-point to this (although the objective is not, as we shall see, inevitably in tension with certainty), we have selected the objective of durability, that is, the ability of the regulation to remain both legitimate and efficacious over time. Durability is akin, but not identical, to flexibility. Long-term durability may often require flexibility in the regulatory design, to enable the regulation to adapt to changed circumstances; this may, in some circumstances, be in opposition to the objective of certainty, but this is not necessarily the case.

Finally, a plethora of mechanisms – including discretionary powers and discretions conferred on administrators, enforcement agencies and courts – may be incorporated into regulation to achieve the objectives of regulation, and thereby serve regulatory values. Critically, as the research streams of the New Zealand Law Foundation Project illustrate, the available responses of regulatory design to meet and optimise regulatory performance against the values of efficacy and legitimacy are not limited to the classical distinction between ‘rules’ and ‘standards’.[20]

The particular mechanisms, when implemented by the administrative agencies and courts concerned, will then produce outcomes, which are to be tested against the objectives and values the regulation was designed to promote.

This broad framework we have developed is illustrated in the diagram below.

Figure 4.2: Value, Objective, Mechanism framework

We address each of these concepts in greater detail, and by reference to the individual research streams, in the next part of this paper. By way of introduction, however, it is appropriate briefly to set out some further definitions of the terms legitimacy, efficacy, durability and flexibility, as these are all core concepts to this paper.

4.2.1 Legitimacy

In the context of New Zealand’s constitutional arrangements as a democracy under the rule of law,[21] legitimacy has at least two aspects:[22]

(a) democratic legitimacy: that is, what choice of regulatory design will best ensure the democratic legitimacy, and constitutionality, of the exercise of power by the state; and

(b) legality: New Zealand is a democracy under the rule of law. A legitimate regulatory system will ensure that the exercise of power by the state is exercised consistently with the rule of law.

In adopting the above definition of legitimacy, we seek to at least partially avoid the jurisprudential debate as to whether the rule of law has a “substantive content”.[23] Irrespective of whether the “thin” account of the rule of law of Raz[24] or the “thick” account of Lord Bingham is adopted, in the context of New Zealand’s democratic constitution, consistency with democratic values is an appropriate and necessary objective of regulatory schemes. Legitimacy, in our usage, thus incorporates both consistency with democratic values and with the rule of law, to the extent that those concepts are different.[25]

4.2.2 Efficacy

Efficacy measures efficiency, effectiveness, practicality and proportionality. For the purposes of assessing efficacy, all regulations should be seen as an investment by the state in achieving certain types of conduct from citizens or other entities within its jurisdiction. Thus, the total cost and the overall proportionality of the entire regulatory system which will be required to ensure a regulation is effective, should be taken into account.

Efficiency measures are aimed at detailing whether or not regulation can be delivered at least cost. Components of efficiency include examining unit costs of regulation deliverables (productive or scale efficiency), whether resources are matched with the issues that need to be addressed (allocative or matching efficiency), and the ability of regulation to foster new approaches that can reduce the cost of regulation (dynamic efficiency or innovation). Typically, these judgements are set out as quantitative measures with accompanying context.

Other questions that require answering include to what degree does the regulation fix or mitigate the problem it was introduced to address (effectiveness); does it solve the regulatory problem in a practical manner (practicality); and is the regulatory response proportionate to the size of the problem (proportionality).

4.2.4 Durability and flexibility

The durability of regulation revolves around the fitness for purpose of the regulation over time and its ability to respond to change. As is recognised in Treasury’s Best Practice Regulation Model, durability of regulation may be related to its flexibility:[26]

Flexibility: … A regulatory regime is flexible if the underlying regulatory approach is principles or performance-based, and policies and procedures are in place to ensure that it is administered flexibly…

Flexibility and durability can be two sides of the same coin; a regime that is flexible is more likely to be durable, so long as the conditions are in place for the regime to “learn”. Indicators of durability are that feedback systems are in place to assess how the law is working in practice; decisions are reassessed at regular intervals and when new information comes to hand; and the regulatory regime is up-to-date with technological change. …

Durability: closely associated with flexibility; the regulatory system has the capacity to evolve to respond to new information and changing circumstances.

The Treasury’s association of the objective of flexibility and durability may, however, be overstated. In some cases too much flexibility in regulation may itself lead to a lack of durability, for example if regulatory standards are degraded over time in a manner that leads to a loss of public confidence in the regulatory design.

4.4 Exploring the linkages between regulatory values, objectives and mechanism

In this part of the paper we consider further how objectives of certainty, durability and flexibility may serve the relevant regulatory values – being democratic legitimacy and legality (as twin aspects of legitimacy), and efficacy. We also consider the role of discretion as a mechanism of regulatory design in giving effect to those objectives.

4.4.1 Democratic legitimacy

The democratic legitimacy of regulation derives from its formulation by, or consistent with the delegated authority granted and supervised by, a duly elected representative body.[27] Plainly, in a democratic state, the democratic legitimacy of the exercise of state power against its citizens is of particular importance. However, due to New Zealand’s autochthonous constitutional structure, democratic accountability cannot be seen as an overriding criterion in regulatory design, nor the only basis by which to assess the constitutional legitimacy of regulation.

Māori rights, existing under the Treaty of Waitangi, at common law, and increasingly in legislation, are a part of New Zealand’s constitutional fabric. The way in which expression is given to those rights is part of an ongoing constitutional dialogue, which frequently engages concepts of certainty and discretion. For instance, the virtues of durability and the lack of certainty were a key factor behind the incorporation of deliberately imprecise language into section 9 of the State-Owned Enterprises Act 1987. The implications of that imprecision were then explored in a sequence of well-known court decisions,[28] and remain in place during the present adoption of a mixed-ownership model for existing state-owned enterprises. Indeed, within the Supreme Court’s present resolution of the challenge to that mixed-ownership model is an invitation for ongoing dialogue in multiple fora.[29]

This highlights how uncertainty can be used as a tool to deliberately leave open and unresolved delicate constitutional and political debates. These types of debates, which do not have a simple, final, answer, can be exacerbated rather than solved by a premature excess of certainty and precision.[30] Thus, the certitude with which the Privy Council announced in Te Heuheu Tukino[31]that the Treaty of Waitangi was not a part of New Zealand law has come to be seen as an invitation for continued attempts to make it so through other routes.[32] The Chief Justice, Dame Sian Elias, has recently made similar observations about the doctrine of parliamentary sovereignty, which Her Honour argues has, in its traditional Diceyan form, “defined away much of the proper subject of constitutional law” and stifled legitimate and important constitutional dialogue.[33] In other words, there is a respectable position that some debates must be allowed to develop and uncertainty provides this opportunity.

Democratic accountability is also only one value within New Zealand’s constitutional arrangements and that, in certain contexts, other values may inform the choice of regulatory design.[34] With that important caveat, we turn to consider the relevance of democratic legitimacy to questions of certainty and discretion. At first blush, that relevance may be minimal. Democratic legitimacy does not require that every regulatory act be the subject of direct or representative vote. As Alexander Bickel put it:[35]

… no democracy operates by taking continuous nose counts on the broad range of daily governmental activities. Representative democracies – that is to say, all working democracies - function by electing certain men for certain periods of time, then passing judgment periodically on their conduct of public office. … The elected officials, however, are expected to delegate some of their tasks to men of their own appointment, who are not directly accountable at the polls. The whole operates under public scrutiny and criticism – but not at all times or in all parts.

In New Zealand, with the exception of certain Crown prerogatives, reserved powers, and the originating jurisdiction of the High Court, all powers emanate from Parliament. Accordingly, the exercise of discretion by public officials in accordance with authority delegated by duly enacted legislation can be said to be consistent with a democratic system, as Bickel defines it.[36]

That cannot be, however, the end of the analysis. As a descriptive matter, democratic accountability and associated public scrutiny is plainly an important factor in determining the question of who exercises discretion within a regulatory system.[37] The choices are numerous, but include:

a) Parliament;

b) a government department or other Crown entity responsible to a Minister, who in turn is responsible to Parliament;

c) another representative body whose membership is elected by a sub-set of the broader electorate, either by reference to geographic characteristics (e.g., City, District and Regional Councils, and District Health Boards) or by reference to interests (e.g. Boards of Trustees);

d) a Crown entity or tribunal whose membership is appointed by a responsible Minister, but who is required to act independently of the government in discharging its functions; and

e) a Court.

At least two ways in which matters relevant to democratic legitimacy may affect that choice emerge from the research streams of the Law Foundation Regulation Reform Project.

4.4.1.1 Rights

First, the choice of which level of government policy-making discretion should be exercised may be considered in terms of the “rights” dimension. This dimension is examined by Dr Petra Butler in her paper, “The Appropriate Form of Regulation”.[38]

In this context, the 2012 Legislative Advisory Committee (LAC) Guidelines state that:[39]

Provisions which affect fundamental human rights and freedoms should always be included in primary legislation. Examples of these rights and freedoms include … rights under the New Zealand Bill of Rights Act 1990 generally …, provisions which expropriate property …, social and economic rights …

The LAC Guidelines are consistent with the traditional proposition that the greater the infringement of regulation on fundamental human rights and freedoms, the greater should be the democratic character of the decision infringing the rights. For example, courts in the past applied a canon of construction against doubtful penalisation. Such canons of construction serve the function of a “preference eliciting rule”.[40] Legislators, faced with the predictable default rule were incentivised to calibrate their choice of legislative language to define the scope of the infringement.

It should be noted that less weight is now given to such canons by the Courts. So, for example, the New Zealand approach to section 6 of NZBORA is that the Courts will not consider giving legislation a meaning other than that produced by ordinary legislative techniques unless that normal meaning constitutes an unjustified incompatibility with the principles and an alternative meaning is available.[41]

An initial question arises as to whether this preference for legislative determination of potentially rights-infringing policy judgements is an effective way of protecting rights. As Bickel observed, rights instruments are typically counter-majoritarian in nature.[42] That is not to say that rights instruments are anti-democratic. The counter-majoritarian nature of rights instruments has been defended on the basis that rights, and in particular civil and political rights, enhance accountable and democratic government.[43] Indeed, this justification was at the forefront of the White Paper urging adoption of a bill of rights as supreme law:[44]

The Bill would reaffirm and strengthen the fundamental procedural rights in the political and social spheres – rights such as the vote, the right to regular elections, freedom of speech, freedom of peaceful assembly, and freedom of association.

However, it is almost invariably the case that the particular persons and groups whose rights may be compromised by the democratic process are those with least influence on it. Professor Andrew Geddis, in his sobering study “The Comparative Irrelevance of the NZBORA to Legislative Practice” has suggested that the relevance of the rights and parliamentary mechanisms contained in NZBORA to legislative decision-making is limited:[45]

it is at its weakest when it comes to protecting those rights most likely to be overlooked in the legislative process, while situations in which it may actually have some effect are those where Parliament was already disposed to uphold the rights in question.[46] Nonetheless, absent a supreme law, it is likely that the best mechanism for ensuring that any infringements are demonstrably justified in a free and democratic society is to ensure that, if rights are to be infringed, it is only by the direct decision of a democratically elected body.

This leaves the question of what infringements with what rights should require Parliamentary consideration. In this regard, Dr Butler persuasively argues that, as a formulation, the LAC Guidelines almost certainly go too far. Given that almost any regulation will affect rights, particularly given the incorporation of social and economic rights into the definition of the rights concerned, it is difficult to see how any matter could be dealt with other than by primary legislation if the LAC Guidelines were strictly adhered to.[47] This is clearly unrealistic.

The question is therefore of the threshold to be applied: what level of interference with what rights should trigger a requirement that the matter be the subject of an act of Parliament? Dr Butler proposes four thresholds:

1) a matter that so significantly infringes a right in NZBORA that it subverts the scheme of NZBORA has to be regulated by an act of Parliament;

2) a matter that significantly infringes a right should be regulated by an act of Parliament. However, in some exceptional circumstances, such as the response to the Canterbury Earthquakes, it may be appropriate to leave a wider discretion to the Executive;

3) the subject, content, purpose and scope of subsidies and benefits have to be at least tied to a budget; and

4) the powers of any autonomous government body need to be carefully circumscribed by an act of Parliament.

The last two of Dr Butler’s thresholds are less directly tied to the central rights. However, importantly, the thresholds indicate that the democratic legitimacy of legislation is not to be considered independently from the rule of law. Thus, her fourth principle indicates that the less democratic accountability a body has (because it is established to be autonomous from Ministerial responsibility), the more important it is that its powers are circumscribed by an act of Parliament.

4.4.1.2 Public participation

Second, public participation in regulatory decision-making is addressed in Mark Bennett and Joel Colón-Ríos’ paper “Public Participation in New Zealand’s Regulatory Context”.[48] Bennett and Colón-Ríos identify the following factors as relevant to the level of public participation required in any particular decision, by reference to the examples of the New Zealand electricity industry and resource management:

a) whether regulation is a key electoral issue (suggesting that regulation should be by Parliament);

b) whether the regulation has distributional consequences (as opposed to determining the most efficient means of meeting a defined objective), which may indicate that the decision should be made by those with either direct or indirect democratic accountability;

c) whether a technocratic approach to regulation is required, for example, because:

i. it deals with a complex subject matter, such as economics;

ii. the objectives of the regulation can be defined by Parliament in economic terms; and

iii. objectives of competition and efficiency are uniquely suitable for a mean-ends economic analysis (that is, with the regulator merely giving effect to the standards previously debated and agreed on, and set down in law);

d) whether the relevant judgements involve choices that are political, social, and cultural as well as economic, or where “local knowledge” of an issue may be important (each of which may point towards more democratic accountability for the regulator);

e) whether the decision affects the whole of the community, or only affects a geographically localised set of people, or some other sub-set of the community.

This analysis is similar to that proposed by Rose-Ackerman, to match the process with the substance of the regulation at hand. In the figure below, the technical issues are matched with scientific method. Where individual rights are the focus, court-like processes may be the best match. In the political sphere, regulations with distributive outcomes are decided in a legislative process; where net benefits are possible an administrative balancing approach is possibly the best fit; and where all interests need to be represented a consensual approach may be required. This matching process may enhance not only the democratic legitimacy of regulation, but also enhance its ultimate efficacy.

Of course this stylised representation does not reflect all regulatory problems and approaches since many do not fit nicely into one box. In more difficult cases which have different elements hybrid mechanisms/processes could be developed e.g. independent technocratic findings could feed into a bureaucratic-led process or even inform consensual decisions.

Figure 4.3: Matching regulatory processes with substance

Source: Rose Ackerman S (1996), Economics, Public Policy, and Law. Victoria University of Wellington Law Review Volume No. 1 pp1-16

Bennett and Colón-Ríos’ insight is to emphasise that the electoral and parliamentary democratic process is not the only mechanism for ensuring public participation and legitimacy for regulatory decision-making. Other mechanisms include designated consumer advocates to represent the interests of those who are likely to face a collective action problem in having their interests represented before the regulatory decision-maker on any consultation.

This insight is important because it means that the relevance of democratic accountability to questions of regulatory design does not end with the choice of the decision-maker. Even in the case of independent decision-makers, an issue of what mechanisms would best supplement the democratic content of the decision-making process should be considered. Not the least of which are the relevant societal values and how those societal values are expressed through public participation in the regulatory process.

Nor are these issues independent: the need for, and appropriateness of, various alternative mechanisms for enhancing public participation is dependent in part on the nature of the decision-maker, as well as the extent and kind of discretionary decision-making power being exercised. Equally, the choice of decision-maker may itself be informed by whether, and to what extent, alternative mechanisms for public participation are available.

Inevitably, as Bennett and Colón-Ríos acknowledge, the cost of such mechanisms must be considered, a matter of particular weight in the context of New Zealand’s relatively small, centralised economy. Equally, that size and centralisation may permit experimentation in public participation in the regulatory process to a degree that has not previously been undertaken.

4.4.1.3 Legality

It is uncontroversial and well-accepted that a primary value of our legal system is that the rule of law should guard against the exercise of arbitrary power and provide a system characterised by stability, certainty and prospectivity. These insights have been developed by a range of well-known scholars, philosophers and judges, including Joseph Raz, Lon Fuller and, recently, the late Lord Bingham. Without addressing the details of this rich vein of jurisprudence, the essential insight is that the inherent virtue of law is its ability to adequately guide behaviour. As Gerald Postema has recently put it:[49]

[B]y its nature, law does whatever it seeks to do through offering normative guidance to those who are subject to law. Law normatively guides conduct of law subjects by addressing to them rules or standards, which they as rational, self-directing agents can grasp and apply on their own, taking the rules or standards as (at least) partial reasons for doing what they call for. To do its ordinary work, law must be intelligible to those who are subject to it; it must make practical sense to them at least to the extent that they can, across a wide range of application, grasp what kind of behaviour the law calls for and how it’s doing so might give them some reason for complying. Further, the practical social space within which law-subjects act is relatively continuous and open-ended, rather than divided into discrete, small modules sealed off from other such modules. So, law-subjects will need to consider how whole sets or combinations of rules or standards mean to guide their conduct. They must face the requirements or opportunities that the law utters more or less as a whole and not discretely one directive at a time.

For the law to be able to act as a practical normative guide to behaviour it must be sufficiently public, accessible, clear, stable, predictable and prospective. At the same time, certainty is not an absolute value of the rule of law, and may be outweighed by other matters (such as the effectiveness of the regulation in achieving its purpose).

In many cases, the value of legality may not speak decisively to a particular choice of regulatory mechanism. This is because the value of legality forms an accepted background with which most regulatory design choices are consistent.

There are a number of reasons why this is so: the most important of which is the broad acceptance of the rule of law in informing regulatory design. However, it is also necessary to reflect the dynamic nature of the regulatory process. Technology and societies preferences change and therefore regulation has to adjust to reflect those changes. Flexibility, to the extent it promotes durability, may itself be argued to be required by a “thick” account of the rule of law, and therefore promote the value of legality. This may be so, for example, where different treatment of persons in differently situated circumstances is required to provide equality over time.

Additionally, the linkage between certainty and particular regulatory forms is highly context dependent. Contemporary debates have moved beyond binary preferences for ‘law over discretion’ or ‘rules over standards’. It is now widely accepted both that the ‘rules/standards’ dichotomy does not capture the full range of regulatory forms outside of the classical form of ‘command and control’ regulation,[50] and that the level of certainty provided by a particular regulatory form is dependent on the broader context in which the regulation is situated.[51]

Regulation with different characteristics may be best to promote certainty depending on different contexts. For example, one aspect of the context considered by the literature is the subject matter (or type of action) to be regulated. Here, the framework proposed by Braithwaite is helpful:[52]

a) When the type of action to be regulated is simple, stable and does not involve huge economic interests, rules tend to regulate with greater certainty than principles.

b) When the type of action to be regulated is complex, changing and involves large economic interests:

i. principles tend to regulate with greater certainty than rules;

ii. binding principles backing non-binding rules tend to regulate with greater certainty than principles alone;

iii. binding principles backing non-binding rules are more certain still if they are embedded in institutions of regulatory conversation that foster shared sensibilities.

Other contextual factors which may lead to different forms of regulation providing greater certainty are discussed further in this paper, in relation to the efficacy of different forms of regulation, below.

In addition to the characteristics identified by Black and Baldwin, it is important to appreciate that aspects of legality – in particular, checks against the exercise of arbitrary power - may be addressed by secondary rules or systems, including ex ante procedural requirements and ex post appellate structures.[53] For example, Dr Gill has summarised the experimentation with Regulatory Impact Statements as an ex ante mechanism for securing transparency and accountability:[54]

... a regulatory management regime targeted behaviour change in bureaucrats by giving the NZ Cabinet collectively greater information with which to control bureaucrat-driven rule making. The RIS regime, when it was first introduced, placed great emphasis on providing ministers with a short (2-3) page problem definition and assessment of options and consequences so that ministers could make more informed decisions about regulatory quality.

In many cases, a “tiered approach” to regulation will provide an optimal way to promote certainty and accountability, particularly where the regulated matter is complex and the regulation is targeted at a large group.[55] A tiered approach may involve binding principles that take precedence over detailed, safe-harbour rules that are subordinate, and may include indicative guidelines that have no legal status. These types of strategies are recognised as indicators for appropriate certainty in the Treasury Best Practice Model and include:[56]

1) Safe harbours are available and/or regulated entities have access to authority advice.

2) Decision-making criteria are clear and provide certainty of process.

The complex linkage between legitimacy and certainty, and the variety of mechanisms by which certainty may be promoted in particular contexts is unlikely to be a decisive factor in regulatory design. This is not, however, to understate its importance: this view of the rule of law is possible only because, in New Zealand at least in recent times, the rule of law almost inevitably forms an accepted context against which regulatory choices are made within. Forms of regulation which raise serious questions of inconsistency with the rule of law are for political and cultural reasons unlikely to be advanced, absent special circumstances.[57] At the same time, in a country where the Economic Stabilisation Act 1948 and the Public Safety Conservation Act 1932 remain relatively recent history,[58] vigilance that the boundary on acceptable regulatory forms is maintained is both necessary and appropriate.

It is also necessary to acknowledge, in the context of a discussion of legality and the rule of law, the increasing significance of what is sometimes described as the globalisation of the rule of law. This globalisation is primarily (although not exclusively – the globalisation of human rights being itself a significant subject) a feature of the development of transnational economic architecture such as the World Trade Organisation, regional trade agreements and bilateral investment treaties. This increasingly important source of substantive law and legal process may, in particular contexts, provide additional constraints on regulatory form and substance.[59] To take a recent, and high profile, example discussed in the paper by Professors Boast and Frankel,[60] bilateral investment treaties may constrain regulatory takings, or at least subject such regulations to the scrutiny of investor-state arbitration. This development has been, and remains, a subject of legal and political controversy.[61] It is outside the scope of this paper to address that controversy, save to comment that we do not see such transnational architecture as modifying the value of legality, as we have discussed it. Rather, transnational architecture itself represents a choice of regulatory design – to accede to and be bound by the relevant international treaty obligations.

4.4.2 Efficacy

It follows from our analysis above that very often the determinative factor in weighing the objectives of certainty and flexibility and durability, and how best these matters are to be promoted and balanced as a matter of regulatory design, including through the deployment of discretion, is pragmatic – what is the efficacy of these choices in the context in which they arise?

The arguments for discretion being delegated by Parliament to administrators as a mechanism for ensuring effective regulation are well known. A primary rationale for delegation is that it is difficult for ex ante rules to achieve a required purpose in a complex society; in particular, a number of the individual research streams have emphasised the importance of a dynamic regulatory system that is able to receive feedback on existing regulation, evaluate outcomes against regulatory objectives, and respond accordingly.[62]

However, it is important to emphasise that the effectiveness of regulation is not always in tension with certainty. Indeed, in some circumstances, certainty may assist with regulatory effectiveness. A brief example illustrates the point. Section 36 of the Commerce Act 1986 creates liability for firms with substantial market power that take advantage of that power for an anti-competitive purpose. The wording of the section provides no explanation as to the distinction between conduct which does, and which does not, constitute taking advantage of substantial market power.[63]

However, in interpreting that section, the New Zealand courts have, consistent with their United States and Australian counterparts in this area,[64] emphasised that: [65]

 …it is important when addressing the statutory concept of use of market power to take an approach which gives firms and their advisers a reasonable basis for predicting in advance whether their proposed conduct falls foul of s 36 and risks a substantial financial penalty.

This approach is consistent, not only with promoting rule of law values (the matter emphasised in the passage cited above), but also – although we acknowledge that this claim is contestable – with promoting the effectiveness of the regulation. Certainty of application arguably promotes the effectiveness of regulation in two ways. First, assuming that most firms with substantial market power do not wish to act anti-competitively, certainty in the application of the statutory standard reduces enforcement costs. As has been observed by Thomas Leary, a member of the US Federal Trade Commission:[66]

... antitrust laws are not really enforced by bureaucrats like me; we … review only a minute fraction of the business strategies that are considered every day in areas of potential concern. The people who really enforce the antitrust laws, day-to-day, are private counsellors employed either as “inside” or “outside” lawyers.

Ensuring that the application of the standards is understandable and predictable to those with the prime responsibility for applying them on a day-to-day basis, since it impacts on the way firms conduct their business and reduces enforcement costs. Second, Australasian and United States competition law has traditionally been premised on an expectation that competition, and the long term interests of consumers, are promoted when large firms, including firms with substantial market power, compete vigorously against their rivals.[67]

If judgements as to the application and enforcement of s36 are unpredictable and inconsistent, the s36 standard may chill the very object it is intended to promote – competitive activities[68] − by deterring firms with substantial market power from competing vigorously lest their conduct be adjudged later to have crossed an uncertain boundary of conduct, exposing them to the expense of a proceeding for pecuniary penalties.

Similar considerations have been applied in other contexts. Thus, in defining the defence of qualified privileged to an action for defamation in the context of news media reporting, Lord Hoffman emphasised the importance of a cautious approach to ex post review of decisions, particularly under flexible standards, where the conduct that may be deterred by the risk of false positives is of particular value to society.[69]

In what circumstances certainty may enhance the efficacy of regulation, and in others prevent it, and the role of discretion in providing that certainty, is a complex subject. A more comprehensive list of regulatory design considerations are set out in another chapter.[70] For present purposes, we note certain aspects that are of particular relevance to the use of discretion within regulatory design.

First, of central importance is the impact of a regulatory system to consider the effectiveness of regulation as it is applied. As observed by Dr Gill, “effects depend on how regulators apply rules and discretion”.[71] Both section 36 of the Commerce Act[72] and the qualified privilege defence to defamation are, on their face, standards rather than bright line rules. Given this, it might be thought that both provide significant discretion to the courts in applying the standards to parties’ conduct, and equally produce significant uncertainty. However, in both contexts the courts have provided secondary rules as to how the standard will be applied that are designed to provide more certainty to parties subject to the regulatory standard.

Second, the value of certainty in ensuring effective regulation depends on at least two matters: first, the subject matter of the regulation, and whether there is a risk that uncertain regulation may adversely affect behaviour that the regulator wishes to promote (in the case of section 36, competitive activity by firms with substantial market power; in the case of the qualified privilege defence, aggressive journalism); second, the expected behaviour of the regulated parties.

A further example of the relevance of these matters to the relationship between efficacy, certainty and discretion is usefully explored by Professor John Prebble in his examination of the General Anti-avoidance Rule (GAAR).[73] The GAAR, as set out in section BG1 of the Income Tax Act 2007, is a standard which overrides the specific tax rules set out in the Income Tax Act and provides that, if an arrangement has the effect of avoiding tax, then the arrangement is void against the Commissioner of Inland Revenue for tax purposes even though the arrangement complies with the rest of the Act. The GAAR is “necessarily imprecise”, and how it may be applied in any particular case lacks certainty, so much so that the authors of “Does the Use of General Anti-Avoidance rules to Combat Tax Avoidance Breach Principles of the Rule of Law? A Comparative Study” conclude that the imprecision is a breach of the rule of law.[74]

Professor Prebble concludes that the lack of certainty in the context of GAAR is justified by two matters:

a) the first is the nature of tax law. In particular, the characteristics of tax law include a significant “ectopia” or gap between the law on the one hand and the economic substance to which it relates;

b) the second is the nature of tax practice. In Professor Prebble’s words:[75]

[T]here is an entire industry devoted to manipulating fiscal laws with a view to obtaining tax advantages without incurring a corresponding economic cost.

The first matter raised by Professor Prebble indicates that adopting an uncertain standard in relation to GAAR is unlikely to have any chilling effect on behaviour that might be thought socially desirable. Because of the gap between tax law and the economic substance to which it relates, a cautious approach to tax planning is unlikely to have any negative implications for parties structuring their economic affairs in a way that they consider optimal, putting tax implications to one side. Rather, any chilling effect is likely to be limited to the aggression with which tax planning is carried out. Unlike competition, or journalism, it is unclear that minimising tax impost has any societal benefit per se.

The second matter raised by Professor Prebble illustrates the importance of the expected behaviour of the regulated entities as a determinant of whether certainty can be expected to assist effectiveness. Certainty may promote regulatory effectiveness where regulated entities can be expected, in the main, to co-operate with the regulatory structure (that is, the entities are “political citizens” or “well intentioned and well informed” in Dr Black’s terminology).[76] Conversely, where groups of regulated entities seek to circumvent the objective of the regulation – that is, they are amoral calculators – the provision of certain rules is likely only to enable those entities to better achieve that circumvention, although general principles and standards will not necessarily secure compliance.[77]

4.5 Examples of regulatory strategies

Against the above background, we briefly assess some of the examples of regulatory design raised in the individual research streams against the framework we have attempted to outline in this paper for considering issues of certainty and discretion. In assessing these examples, we do not purport to provide a comprehensive schematic of different models of regulation. Given the number of possible dimensions for characterising different types of regulation,[78] this is not achievable. However, some different models of regulation that have been highlighted in the literature are:

a) pure “command and control” regulation, enforced by sanctions for non-compliance. On the traditionalist account, this is best served by a rules-based approach which strives to promote certainty;

b) at the other end of the spectrum of intervention, an approach based on self-regulation (or a “regulatory conversation”) supported by a benchmark set of rules against which any self-regulation is measured. For example, the US Environmental Protection Agency granted flexible performance-based permits as long as the firm demonstrated in negotiations that the results would equal or exceed the applicable standards set by the agency;[79]

c) a rules-based approach which also incorporates non-binding principles. This is the approach of most New Zealand legislation, where a principles or purposes section is common, and the Court is mandated to take a purposive approach to its interpretation: Interpretation Act 1999, s 5;

d) a principles-based approach with rules or guidelines that are not binding, but do provide a safe-harbour, such that compliance with the rules will guarantee compliance with the principle; and

e) a principles-based approach where compliance with detailed rules does not provide a safe-harbour. An example of this is the GAAR, discussed above.

Examples of most of these forms of regulation appear in the individual research streams. The command and control approach is evident in criminal law,[80] where certainty is traditionally maximised by construing any ambiguity in favour of the accused. As has been discussed above, a similar approach has been traditionally taken in the field of competition law. As discussed by Paul Scott in this context, certainty may also be provided by ensuring that guidance can be taken from comparable jurisdictions. [81] Even in these cases, however, a rules-based approach (or, standards with guidance approach) is supplemented by significant discretion in enforcement strategy, exercised primarily by the enforcing authority (i.e., the Police and Commerce Commission) but also through the process of determining the level of sanction to be imposed on infringing conduct.

The leaky homes crisis is often attributed to a change from prescriptive, detailed buildings codes to more flexible “performance based regulation”. However the discussion by James Zuccollo and Mike Hensen in relation to weather tight buildings illustrates the challenge of measuring efficacy between regulatory types.[82] On the common account, the discretion of the building code reforms essentially failed to provide acceptable levels of quality assurance. Flexibility, therefore, did not lead to durable regulation. However, NZIER has argued that the productivity gains foreseen as possible in the building sector from the removal of prescriptive regulation may have been of sufficient social value to significantly off set, or even outweigh, the cost to building owners of repair when subsequently adopted building techniques were found to be flawed.[83] This illustrates that, even in assessing relatively stark regulatory choices, the assessment of efficacy may be highly contested.

The combination of principles and rules reflected in New Zealand tax legislation has been discussed above. As noted there, the GAAR provides a principles based overlay whereby compliance with the detailed rules of tax legislation does not provide a safe harbour. Interestingly, Professor Prebble not only endorses the flexible nature of the GAAR (for reasons which appear unimpeachable), but goes further to opine that it is neither desirable nor sensible for the Commissioner of Inland Revenue to provide non-binding guidance on its interpretation of the GAAR.[84] On the account of tiered regulation proposed by Black, the provision of a further tier to the regulatory structure might be considered desirable. However, in the particular context of the GAAR, it may be that, given the primary targets of the GAAR are amoral calculators, a further tier is simply inefficient. Further, it is relevant that rule of law concerns as to administrative accountability are addressed in this context through the provision of rights of appeal to the generalist Courts.

Perhaps the most sophisticated form of regulation discussed in the individual research streams is the regulation of the New Zealand electricity industry, discussed in the paper by Mark Bennett and Joel Colón-Ríos.[85] Under the Electricity Industry Act 2010, a tiered system of regulation, rules and non-binding information has been put in place. In the context of the wholesale electricity market, this includes a code of market participant rules, in the form of the Electricity Industry Participation Code 2010. The Code includes precise and detailed rules, breach of which are subject to sanctions.[86] Enforcement actions may be commenced either by industry participants, or by the Electricity Authority,[87] and are determined by an independent tribunal (the Rulings Panel).[88]

In addition to these precise rules, Part 5 of the Code provides an additional jurisdiction on the Electricity Authority to intervene in the wholesale market if it considers that an “undesirable trading situation” has arisen, and take remedial steps. The definition of undesirable trading situation is relatively broad, and provides the Authority with discretion to protect the functioning of the market even absent breaches of the specific rules of the Code.[89]

Finally, and in addition to the precise rules and binding standards, as Bennett and Colón-Ríos note, the Electricity Authority also engages in a number of participatory processes in its decision-making, as well as performing broader market education functions through publications, advisory groups, and forums.

Taken together, the electricity regime thus combines a number of mechanisms and decision-makers, each of which have elements of discretion, but overall provides both participants and the public with a significant degree of certainty within a framework intended to be flexible and durable. That certainty is not absolute; indeed the level that should be provided remains contested.[90] However, the regulatory scheme is at least a sophisticated effort to balance democratic legitimacy, legality, and efficacy in a complex and multi-faceted industry.

The sophistication of the regulation provided by the Electricity Authority comes at a price. The 2012/13 appropriation for the Electricity Authority was approximately $68 million, of which $63 million was associated with electricity industry governance and market operations.[91] In addition to the members of the Authority, the Authority employs over 30 full time staff. The level of the investment in regulation of the electricity industry no doubt reflects the critical function of electricity supply to New Zealand, and its contribution to the economy. It also reflects a history of electricity reform in New Zealand since at least 1984,[92] a process which has seen the development of electricity regulation from simple administrative discretion (in the form of price setting by a central authority), to more market orientated reforms, and to the current tiered model.

4.6 New Zealand and efficacy – small size, same costs

The latter emphasis on cost is of real significance to regulatory design in New Zealand. A fact of being a small developed country is that our regulatory system is expected to have systems in place to deal with the same issues which arise in a large developed country, but with fewer market players and a smaller budget.

This challenge does not scale down proportionately such that it is equivalent on a per capita basis to the challenge faced by having more market players and a larger budget. This is because New Zealand is rarely in a position to reap economies of scale[93] (although, its relatively small size, may, in some circumstances, enable regulatory solutions to be implemented (or amended if found wanting) more quickly than in larger jurisdictions).

The perennial New Zealand regulatory challenge is to provide world-class regulation across the board using a relatively modest investment in infrastructure and bureaucracy.[94]

As observed earlier by Postema, it is important to view regulatory systems as a whole and in context. Thus, the expected efficacy of a law is inseparable from the mechanisms which will be utilised to monitor and, in some cases, enforce, compliance. Obtaining scale has been one reason for New Zealand to seek to create better synergy within and between its regulatory agencies to improve efficiency and decrease costs. Thus, whereas, for instance, the United Kingdom may require (or be able to justify) a separate serious fraud office, financial markets regulator, and competition regulator, New Zealand may not. Similarly, where the United States may have a preference for regulating through bespoke regimes produced and administered entirely within its borders, there may be advantages to New Zealand applying uniform or model laws or international conventions in many areas to reduce the cost of reinventing the wheel.

On the other hand, New Zealand’s small market economy lends itself to oligopolistic behaviour. Fewer players in the market can lead to entrenched power, with relatively infrequent new entrants to restrain behaviour. Small size is no excuse, however, for ineffective regulation. The need to ensure effectiveness in a small market is also a factor in the design of an applicable regulatory regime.

A certain rules-based regime may involve higher enforcement costs than a more discretionary regime. On the other hand, ensuring that a more discretionary regime is effective may require continued oversight and adjustment by the regulator and also central agencies. Where the balance is to be struck will depend upon the specific context. If there is any general statement which might be made, it is that discretionary and principle-based regulation is likely to require being situated within a sophisticated and evolving regulatory structure. In some circumstances, it may be argued that the error costs of bad regulation are such that the administrative costs of discretion (thereby enabling fast feedback loops) are justified. However, to our mind this merely illustrates the need for context-specific assessments of the most appropriate regulatory formula.

4.7 Conclusion

This paper has sought to explore the linkages between values, objectives and mechanisms and to place considerations of the importance of certainty as an objective of good regulatory design, and to what extent the mechanism of discretion should be deployed, within that framework.

In discussing the relevance of legality within the value of legitimacy, we have largely sided with the pragmatic approach to regulatory development. That is, we do not consider that rule of law appeals to certainty are likely to be determinative of many of the difficult questions of regulatory design which New Zealand, as a modern democratic society, faces. To avoid any doubt, that is not because the rule of law is unimportant. To the contrary, it is part of New Zealand’s constitutional bedrock. Rather, it is because: (a) for the most part, New Zealand regulatory design options do not abrogate the rule of law; and (b) very many difficult design issues are encountered which cannot be answered by reference to traditional rule of law values.

Our account does not mean that certainty has no significant role to play in regulatory design. Where all other things are equal, a regulatory system which is more certain is perhaps preferable. We simply do not consider that it is appropriate to consider certainty on its own as a higher-order value. It is an objective which may improve the legitimacy or efficacy of a regulatory scheme. Our account has suggested that other aspects of the value of legitimacy, such as democratic legitimacy of regulation, are also relevant both to choices between certainty and flexibility, and to choices of who exercises discretion, and how. The benefit of flexibility and discretion in providing for ongoing constitutional dialogue was noted in particular. We have also sought to explain how certainty, as well as discretion, may play an important role in assessing the likely efficacy of regulation in particular contexts, and how different mechanisms may achieve this.

Our concluding thought is that it is the arena of efficacy, and the implications for New Zealand as a small democratic country, in which effort to consider questions of certainty and discretion are best explored.