Judicial Review of Policy Decisions: The Four Grounds and the Limits of Interference

A policy decision is a choice about what the State should do: which sectors to open, which schemes to fund, how to allocate resources, whether to privatise, what standards to set. Such decisions are reviewable, but on a narrow footing. The courts have repeatedly held that they will not examine whether a policy is wise, sound or preferable to another, because that judgment belongs to the organ the Constitution entrusted with it and because the court has neither the material nor the responsibility to make it. What the courts do examine is whether the policy is constitutional, whether it conflicts with a statute, whether it is arbitrary or irrational, and whether it was adopted in bad faith. This topic sets out that fourfold test, the change-of-policy problem, and the situations in which a policy nevertheless falls.

1. The Scope of Review Stated

📖 Directorate of Film Festivals v. Gaurav Ashwin Jain, (2007) 4 SCC 737

Facts: The scheme governing national film awards, including the categories of awards and the criteria for them, was challenged, and the High Court had issued directions modifying aspects of the policy. The Directorate appealed, contending that the framing of such a scheme was a matter of policy.

Held: The Supreme Court set aside the directions and stated the scope of review of policy with precision. The scope of judicial review of governmental policy is now well defined: courts do not and cannot act as an appellate authority examining the correctness, suitability and appropriateness of a policy, nor are they concerned with whether a better, fairer or wiser alternative is available. Legality of the policy, and not the wisdom or soundness of the policy, is the subject of judicial review. The Court held that a policy decision may be interfered with only where it is unconstitutional, or contrary to a statutory provision, or arbitrary, irrational or in abuse of power.

Ratio: The clearest modern formulation. Policy is reviewable on four grounds only: unconstitutionality, conflict with a statute, arbitrariness or irrationality, and abuse of power. Wisdom, suitability and the availability of a better alternative are outside review.

2. The Four Grounds Applied

Ground	What must be shown
Unconstitutional	The policy infringes a fundamental right, offends Article 14, or violates any other constitutional provision or the distribution of legislative power
Contrary to statute	The policy conflicts with an Act or with rules made under it, or purports to do by policy what only rules could do
Arbitrary or irrational	The policy has no determining principle, rests on no material, or is one no reasonable authority could adopt
Abuse of power	The policy was adopted for a collateral purpose, in bad faith, or as a device to favour or exclude particular persons

Two points about the application of these grounds are worth stating. First, a policy is not arbitrary merely because it produces hardship in individual cases, since any general rule does; the question is whether the rule itself has a rational basis. Second, the ground of abuse of power is where most successful challenges to policy actually succeed, because it is concerned with purpose rather than with merit: a policy that recites a public object but in substance serves a private one is bad without the court needing to form any view about what the right policy would be.

3. The Leading Illustration: Disinvestment

📖 Balco Employees Union (Regd.) v. Union of India, (2002) 2 SCC 333

Facts: The Union Government decided to disinvest its majority shareholding in Bharat Aluminium Company Ltd. and to transfer management control to a private purchaser. The employees' union challenged the decision, contending that the process was flawed, that the price was inadequate, that the workers had not been consulted and that public interest had been sacrificed.

Held: The Supreme Court dismissed the challenge. It held that the decision to disinvest and the manner of doing so are matters of economic policy, and that in such matters the court has neither the expertise nor the material to substitute its own judgment; it is not for the court to determine whether a particular public sector undertaking should be privatised, nor to sit in judgment over the valuation arrived at by experts. Judicial review is available where the decision is contrary to any statutory provision or to the Constitution, or where the decision-making process is vitiated by mala fides or irrationality, but not otherwise. The Court also held that employees have no vested right to be consulted before a policy decision to disinvest, and that a change in the identity of the employer does not by itself affect their service conditions.

Ratio: Economic policy decisions such as disinvestment are reviewable only for illegality, unconstitutionality, mala fides and irrationality in the process. The court does not assess valuation, commercial wisdom or the desirability of the policy itself.

4. Change of Policy

A separate set of questions arises when a policy is changed. The starting point is that the executive is entitled to change its policy: a government cannot bind its successors, or itself, to a policy for all time, and the power to make policy includes the power to alter it. Four qualifications apply.

The change must be prospective in effect. Rights and benefits already accrued under the earlier policy are not ordinarily divested by the new one.
Those affected may be entitled to be heard. Where a settled practice has founded a procedural legitimate expectation, an opportunity to make representations may be required before the change is applied to them, as Navjyoti Co-operative Group Housing Society v. Union of India, (1992) 4 SCC 477 illustrates.
The change must be in the public interest and bona fide. A substantive legitimate expectation may be defeated by a change of policy in the public interest, but the change is itself reviewable on Wednesbury principles, per Punjab Communications Ltd. v. Union of India, (1999) 4 SCC 727.
Promissory estoppel may operate. Where a clear promise was made and acted upon, the State must place material before the court showing why the public interest requires it to resile, per Motilal Padampat Sugar Mills Co. Ltd. v. State of U.P., (1979) 2 SCC 409.

5. Policy, Law and the Boundary between Them

Not everything called a policy is one, and the label matters because it determines the standard of review.

Instrument	Character	Review
A policy decision proper	A choice about objectives and priorities	The four grounds only (Gaurav Ashwin Jain)
A policy embodied in statutory rules	Delegated legislation	Full ultra vires review against the parent Act and the Constitution
An executive instruction giving effect to a policy	Administrative direction	Cannot override rules; enforceable through Article 14 and legitimate expectation
A policy applied to an individual case	Administrative decision	Ordinary grounds: relevant considerations, fairness, reasons, non-fettering
A policy applied inflexibly to every case	Fettering of discretion	Bad, since the authority must keep its mind open to the individual case

⚠ Policy is a shield against merits review, not against legality review

When the State describes a decision as one of policy, it succeeds in excluding one question and not others. It excludes the question was this the best course, which no court will answer. It does not exclude the questions whether the decision was within the power, whether it conflicted with a statute or the Constitution, whether it rested on any material or determining principle, and whether it was taken for the purpose for which the power exists. A challenge to policy that argues the decision was unwise will fail however strong the argument; the same challenge framed on purpose, material or statutory conflict may succeed on much thinner material.

6. The Position in Summary

Policy decisions are reviewable, but the courts do not act as an appellate authority on the correctness, suitability or appropriateness of a policy, and are not concerned whether a better alternative exists.
Interference is confined to four grounds: unconstitutionality, conflict with a statutory provision, arbitrariness or irrationality, and abuse of power (Gaurav Ashwin Jain).
Economic policy decisions such as disinvestment are reviewed for illegality, unconstitutionality, mala fides and irrationality in the process, and not for commercial wisdom or valuation (Balco Employees Union).
A policy may be changed prospectively in the public interest, subject to accrued rights, procedural legitimate expectation, Wednesbury review of the change and promissory estoppel.
The label of policy does not control: a policy embodied in rules attracts full ultra vires review, an instruction cannot override rules, and a policy applied inflexibly to every case is a fettering of discretion.

7. Related Topics and Provisions

Judicial Review of Economic Policy (Topic 96): the economic field in detail.
Judicial Restraint (Topic 93) and Judicial Deference (Topic 94): the disciplines underlying this standard.
Doctrine of Legitimate Expectation (Topic 54) and Promissory Estoppel (Topic 55): the limits on changing policy.
Failure to Exercise Discretion (Topic 50): fettering by inflexible application of policy.
Public Interest and Administrative Discretion (Topic 57): what the courts examine when public interest is pleaded.
Constitution of India: Articles 14, 19, 21, 32, 226 and Part IV.