March 16

Peter Cane: Separation of Powers and the Scope of Judicial Review of Administrative Decisions

I was much interested by Swati Jhaveri’s stimulating inaugural blog-post about the use by Singapore courts of the concept of ‘co-equality’ in the law of judicial review: on the one hand, to justify restricted review of prosecutorial decisions and, on the other, to justify unrestricted review of preventive detention orders. In Ramalingam Ravinthran v Attorney-General (2012), in reviewing a decision of the Attorney to charge an offender with a capital offence, the Court of Appeal said (paras 43-46),

the prosecutorial power…is constitutionally equal in status to the judicial power…[the organs that exercise each power respectively] have equal status under the Constitution …the separation of powers doctrine requires the courts [to]…presume that the Attorney-General’s prosecutorial decisions are constitutional or lawful until they are shown to be otherwise…This approach should not be regarded as the courts deferring to the Prosecution. It is…an application of the established principle that the acts of high officials of state should be accorded a presumption of legality or regularity…

By contrast, in Tan Seet Eng v Attorney-General (2015) the Court of Appeal, in holding that preventive detention decisions were reviewable on the normal grounds, said (at paras 1 and 90) that in a Westminster system of government (such as that in Singapore),

the specific responsibility for pronouncing on the legality of government action falls on the Judiciary…this…is not to place the Judiciary in an exalted or superior position relative to the other branches of the government. On the contrary, the Judiciary is one of three co-equal branches of government. But though the branches of government are co-equal, this is so only in the sense that none is superior to any other while all are subject to the Constitution… the Judiciary has the responsibility for the adjudication of controversies which carries with it the power to pronounce authoritatively and conclusively on the meaning of the Constitution and all other laws.

Whether it is used to justify restricted or unrestricted judicial review of administrative decisions, the idea that repositories of the various constitutional powers of government are co-equal is noteworthy. It is most closely associated with the US system of government. In Federalist 49 James Madison described the several ‘departments’ of government – legislature, executive and judiciary – as being ‘perfectly coordinate’. Each exercises power delegated to it by the People. For instance, when the judiciary pronounces upon the constitutionality of acts and decisions of the legislature and the executive, it does not exercise a power superior to theirs but, rather, enforces the Constitution, to which all branches are equally subject, on behalf of the People, in whom ‘sovereignty’ resides. The co-equal status of the branches of government is a corollary of the basic principle of distribution of public power in the US system, namely that power should be diffused by being divided and shared between organs of government in such a way as to force each to cooperate with the others to achieve their political goals (see generally my Controlling Administrative Power: An Historical Comparison (2016)).

The best-known aspect of diffusion of power in the US system is the sharing of legislative power between the Presidency and Congress, and between the two Houses of Congress. However, diffusion of power is a pervasive characteristic of American governmental arrangements. Ironically perhaps, diffusion does not ‘separate’ power in a strong sense. Indeed, as already noted, the whole point of diffusion is to force organs of government to share (overlapping) power.

One such power is the power of interpretation. The idea that the Constitution does not allocate that power exclusively to the courts provides a theoretical foundation for judicial ‘deference’ to administrative statutory interpretation (as required by the famous Chevron doctrine, for instance). In relation to constitutional interpretation, although ‘judicial supremacy’ is current practice in the US, ‘departmentalism’ – the idea that the various branches share the power to interpret the Constitution – has a long and respectable history in US legal discourse. Another distinctively US form of judicial control of administrative power – the doctrine of hard-look review – can also be partly explained by the diffusion principle. Under that doctrine, courts take a ‘hard look’ at whether administrative decisions are ‘rational’ in a strong sense that requires a close fit between the decision and the reasoning supporting it. This is deference in reverse: here, the sharing of power between the executive and the judiciary allows intrusive review of agency decisions by courts. Shared power is the natural home of the language of deference and intrusion – checks and balances.

What about ‘Westminster’ systems? In the medieval period in England, no clear distinction was drawn between types of governmental power. All power, whether legislative, executive or judicial, resided in the Monarch. Like other government officials, the judges of the common-law courts (who developed the judicial review jurisdiction in the 17th century), were general servants of the Crown who did more-or-less what the Monarch asked of them, participating in various ways in legislating, administering and judging. They were an integral part of the government machine. All this changed after the Glorious Revolution of 1688. The judges were ejected from the corridors of Whitehall and confined to the Royal Courts of Justice. Their main job now was to interpret, apply and enforce the will of the new sovereign law-maker, Parliament. No longer would the judges have a share in the running the country. In return for this reduction of status and responsibility the common-law judges were given security of tenure and salary. Just as importantly, they were to enjoy a monopoly of interpreting, applying and enforcing the will of Parliament. Deciding questions of law was their job alone. Judicial independence and the separation of judicial power became a distinguishing feature of the English system of government.

For some time after the Revolution the relationship between Monarch and Parliament was analogous to that between President and Congress in the US system today. Legislative power was divided and shared between them, and each needed the other. By the early 20th century, however, that relationship had been transformed by the transfer of ministerial responsibility from the Monarch to Parliament, the development of the party system, and the broadening of the franchise. Rather than being divided and shared, legislative, executive and bureaucratic power now became concentrated in the Government. The apotheosis of this new dispensation can be seen in the Australian federal system. Judicial power is corralled by complementary presumptions that judicial bodies may not exercise non-judicial functions (legislative, executive and so on), and that non-judicial bodies may not exercise judicial functions. Judicial supremacy in constitutional and statutory interpretation is unqualifiedly claimed and jealously guarded, and the language of ‘deference’ is absent from the judicial lexicon. This is separation of powers on steroids.

Does Singapore have a Westminster system? Yes and no. Adopting H Kumarasingham’s felicitous term, we may say that Singapore is an ‘Eastminster’ polity (H Kumarasingham (ed), Constitution-Making in Asia: De-Colonisation and State-Building in the Aftermath of the British Empire (2016)). Whereas the US system shares different types of power amongst the branches, under the Westminster model different types of power are concentrated in the various branches: legislative power (technically) in Parliament, executive and bureaucratic power in the Government, and judicial power in the courts. According to Kumarasingham, this tendency to concentrate power in the Government is even stronger under the Eastminster model, a legacy of imperial rule in non-settler colonies. In the US system, ‘sovereignty’ resides outside government, in the People. By contrast, in both Western and Eastern Minsters, it resides somewhere within the government machine – theoretically in Parliament and practically in the Executive. Other organs of government – notably, the courts – are not co-ordinate or co-equal with the sovereign, but ultimately subordinate to it. The judicial power is not shared between the branches but separated off into the judiciary. The operational logic of power-diffusion/sharing is co-operation. The operational logic of power-separation/concentration is contest. ‘Deference’ and ‘dialogue’ both sit uneasily with the operational logic of West/Eastminster systems. This is because the Revolution Settlement on which these systems are based combined judicial subordination to the sovereign with judicial supremacy over law. It is this supremacy over law that empowers the judiciary, though non-sovereign and ‘unequal’, to protect citizens against over-weaning executive power.

This account helps to explain several features of the earlier-quoted judicial dicta. When Singaporean courts (as in Ramalingham) engage in light-touch review, they do not (we are told) defer to the decision-maker but rather recognise different types and separate spheres of power. The separation of prosecutorial from judicial power is widely acknowledged to be a basic requirement of ‘the rule of law’. Similarly, when courts (as in Tan Seet Eng) review with a keener eye, they exercise (so the court tells us) their supremacy over law and the interpretation of the Constitution. The theory is the same in both cases: allocation of judicial power, but only judicial power, to the judiciary; and judicial supremacy over law, but only law. The language of ‘co-equality’ jars in this context.

Of course, all this is shadow play, smoke-and-mirrors. But in the exercise of public power, appearances matter! Doing the right thing is often much easier than explaining why it is the right thing to do. We need constitutional myths to bolster the state’s claim to legitimacy.


Peter Cane is Senior Research Fellow at Christ’s College, Cambridge.