This article has been authored by Ujjaini Biswas is a fourth-year B.A.,LL.B (Hons.) student at NALSAR University of Law,
Hyderabad. Her academic interests lie in criminal law and constitutional law. She has interned in various law chambers, focusing on criminal, family, and commercial litigation across various courts, including the Delhi High Court and Supreme Court.
Introduction
This doctrine of legitimate expectation has been firmly established through a catena of judicial decisions, one of which is the 2023 Supreme Court (“SC”) ruling of Sivanandan CT and others v. High Court of Kerala and others (“Sivanandan”).[1] The administration can override this doctrine only if compelling reasons of public interest exist.[2] However, what made this case of interest was that the Court held in favour of the petitioners, yet they were left without any relief. In this case comment, I analyse the ruling on three aspects. Firstly, I briefly explain how the SC correctly ruled the violation of legitimate expectation. Secondly, I explain how the ruling opens the door to inconsistent outcomes across administrative law. Lastly, I argue that there was a scope for the application of the theory of the second actor to address the tensions arising and strengthen the Court’s reasoning.
What the Ruling was about
In Sivanandan, a group of petitioners challenged the Hon’ble Kerala High Court’s (“HC”) decision to change the examination rules for the higher judicial service of Kerala. The existing rules allowed selections based on an aggregate of two written test papers and a viva voce, for which there was no cut-off. Following the 2016 written test, the selected candidates underwent a viva voce in January 2017. It was after this that the HC’s Administrative Committee decided to apply a minimum cut-off for the viva voce, impacting the final selection list. The same was justified on the grounds of ensuring that only those candidates possessing the necessary personality and knowledge were selected. As a result, candidates who scored higher overall were not selected, as they did not meet the cut-off marks for the viva voce.
The SC ruled that the original recruitment procedure created a legitimate expectation among the candidates that the process would be fair and non-arbitrary. It held that the HC’s actions amounted to an arbitrary exercise of power. However, the Court could not unseat the candidates who got recruited over the aggrieved candidates, because the list was issued in 2017, while the dispute was resolved in 2023. The already selected candidates were otherwise qualified for judicial office, and so unseating them after 6 years, in the court’s opinion, would be contrary to the public interest.
Doctrine of Legitimate Expectation: What the Court got right
Returning to the core of the doctrine, as was articulated in cases of Union of India v. Hindustan Development Corporation[3] and Ram Pravesh Singh v. State of Bihar,[4] the SC held that the expectation must be based on “sanction of law, or custom, or an established procedure followed in regular and natural sequence”, has to be reasonable, logical and valid, based on an established practice of a decision-making authority, and not a mere anticipation.[5] In the present case, it was not a promise expressly made, but an established practice of selecting candidates based on the highest aggregate.
Alongside violating legitimate expectation, the Kerala HC’s rationale for imposing a via-voce cut-off was also unreasonable. Almost all administrative decisions are rational and made for intelligible reasons, but the question is whether they meet the legal standard of reasonableness.[6]What transpired in the present case is that two equally qualified groups of candidates were treated differently: one group that met the newly introduced cut-off began serving in the judicial service for six years, while another group, who would have qualified under the long-standing existing rules, was excluded solely due to retrospective application of a new criterion. This disparity not only undermines fairness, but also defeats the purpose of the judicial service recruitment process itself, rendering the HC’s decision unreasonable.
It is the response of the SC to this unreasonable decision that is problematic. Once the SC found that the administrative decision was ultra vires, arbitrary, and in breach of legitimate expectation, ideally, it should have provided for straightforward remedial consequences. In the conventional framework, the invalidity of the first act impacts the second act as well, but in reality, the Court used public interest as the rationale for not unseating the selected candidates. This reasoning is not analytically rigorous. The doctrine of legitimate expectation only explains why the first action (of the HC) was unlawful, but it does not explain when or why the consequences of that unlawful act should not unravel the subsequent administrative decisions. This is precisely where the theory of the Second Actor becomes essential.
The Missing Link in Sivanandan
Articulated by Christopher Forsyth, the second actor theory addresses a structural problem that lies at the core of administrative law, that is, how courts should treat subsequent administrative actions, which rely on an earlier act that has now been declared unlawful.[7] An unlawful administrative act is void in law, but it still exists in fact, and because it exists in fact, people and authorities inevitably rely on it for their actions.[8] Central to this theory is a crucial conceptual shift. Instead of focusing on the invalidity of the first act, courts must check whether the statute empowering the second actor authorises the actor to act even if the first act is invalidated. Joan Mendes, while explaining EU administrative law cases, explained how discretion is embedded in institutional and organisational settings, and not just in single acts or actors.[9] The actions of the second actor are thus shaped by legal norms interacting with the administrative structure. Therefore, the second action is not always automatically invalidated if the first one is. Afterall, all the actors act within their own legally structured roles.
This analysis is entirely absent in Sivanandan, despite the theory squarely applying to the facts. The Administrative Committee of the Kerala HC, the first actor, introduced the viva voce cut-off in clear violation of the 1961 rules. This decision directly shaped the recruitment process. The second actor, that is the selection authority, then published the list of selected candidates, making appointments based on the unlawful change in rules. When the SC held that the change in the cut-off was ultra vires, it effectively declared the first action of the Administrative Committee to be void in law. According to doctrine of ultra vires, an action is void ab initio if the law is declared unconstitutional.[10] Would that then invalidate the second actor’s actions too? Courts in India have not answered that yet. A look at rulings from other jurisdictions indicates that the answer is in negative. In White & Anor v. South Derbyshire District Council, the local authority granted licenses which were later invalidated.[11] The England and Wales High Court held that unlawful administrative acts have no legal effect once they are quashed, but before that, they may have some legal effect, particularly to protect innocent third parties relying on them.[12] Hence, the court quashed the prosecution initiated against the licensees, on grounds, that they cannot be punished for the fault of the first actor. In R v. Wicks, the accused, who was charged with the breach of a planning ‘enforcement notice’ argued that the notice was invalid. The House of Lords had held that ‘enforcement notice’ meant that the notice issued by the planning authority was formally valid, but it did not have to be a legally valid order for a valid conviction of the accused to ensue.[13] Therefore, here, Forsyth explained that the second actor could act validly notwithstanding the validity of the first one.[14]
In Sivanandan, the SC refused to unseat the appointed candidates. It cited “public interest” and the passage of six years as reasons.[15] What the SC failed to recognise was that the theoretical justification for preserving the appointments of the candidates despite the invalidity of the first act could have been based on the theory of the second actor. Had the Court invoked this theory, it would have given rise to a separate line of inquiry. The Court would have to examine whether the statutory scheme governing judicial appointments required that the validity of the appointment process depend on strict compliance with the 1961 rules, or whether the law is broad enough to allow appointments to stand even when the preceding procedural step was invalid. If the statute mandates compliance with the Rules as a necessary prerequisite, then the second act (of selecting the candidates based on retrospective application of the rules) stands invalid, but if the selection authorities have the discretion once the selection list is published, the appointments could be preserved even when the initial decision was void.
Why this Reasoning Mattered
The second actor theory requires a structured evaluation of the consequences of invalidating the second action. Administrative decisions, as Eyal Benvenisti explained, are embedded in a chain of institutional actions.[16] They are not isolated. Different actors in this set up, function under different limitations and incentives, even when responding to the same legal defect, and so, courts must assess the institutional roles and capacities, and not merely the legality of the first act. Courts need to be able to manage the normative concerns of the Second Actor theory, that is, the need to avoid unthinking collapse of later decisions solely based on the illegality of the earlier one.
The Court has to consider not only the “public interest” factor, but also other risks such as unfairness meted out to the innocent candidates, institutional disruption, administrative chaos, and, most importantly, the practical impossibility of restoring the status quo ante. Although the SC discusses the last factor, there is no principled reasoning behind its ruling on why the petitioners were left without any remedy. The remedy for an illegal act is not automatic, as courts routinely balance the rule of law with fairness to affected individuals, institutional stability, reliance and good faith. Courts have to select which remedy to provide. Individuals who relied on the administrative decision ought not to be punished for defects within the administrative machinery. Forsyth emphasised that second actors should not have to bear the consequences of an invalid action, which they neither contributed to nor could have detected. Therefore, the SC could have used this as a reason as to why the petitioners would not get practical remedies.
The second actor framework essentially structures the remedy that an aggrieved party would receive. Courts could proceed in a three-step remedial inquiry, with the first being whether the second actor acted in good faith, relying on the first act, the answer to which is positive in Sivanandan. Secondly, whether the statutory scheme conditions the validity of the second act on the first, the answer to which is again affirmative, and lastly, if the second act is invalid, what remedy best balances legality and reliance. This is where SC could acknowledge the invalidity, while also choosing not to quash the appointments. By not following this remedial path, the ruling has potentially opened the door to inconsistent outcomes across administrative law. The next time appointments are made based on an unlawful rule change, or past actions are preserved despite being invalid, aggrieved parties would be left without remedy in an unjustified fashion. Using the second actor theory would have provided answers to them that the Court’s public interest justification otherwise does not.
Conclusion
Sivanandan is a compelling illustration of the strengths and silences of Indian administrative law. On the one hand, the affirmation of the doctrine of legitimate expectation by the SC was sound, precise, and convincing. It correctly recognised the retrospective alteration as the action violating reasonable expectations of the candidates, and that such arbitrariness strikes the core of a fair administrative process. Yet on the other hand, the Court stopped short of applying the theory of the Second Actor, and essentially left the petitioners without any remedy. Had the Court delved into the analysis of the case using the second actor theory, rather than relying on a broad and vague public interest consideration, it could have provided a legally and logically coherent justification to the aggrieved petitioners. It would have also prevented future inconsistencies in similar administrative disputes. Ultimately, Sivanandan demonstrates the need for doctrinal clarity, prompt judicial intervention, and principled reasoning when public administration errs.
[1] Sivanandan C.T. & Ors v High Court of Kerala & Ors (2024) 3 SCC 799
[2] Foram Patel & Rishin Patel ‘The Doctrine of legitimate Expectation: From Development in England to Indian Scenario’ (2021) ILI LAW REVIEW <https://ili.ac.in/pdf/win21_8.pdf > accessed 03 December 2025
[3] Union of India v Hindustan Development Corporation (1993) 3 SCC 499
[4] Ram Pravesh Singh v State of Bihar (2006) 8 SCC 381
[5] Sivanandan C.T. & Ors v High Court of Kerala & Ors (2024) 3 SCC 799, ¶ 25
[6] Hasan Dindjer, ‘What makes an administrative decision unreasonable?’ (2020) 84(2) THE MODERN LAW REVIEW < https://doi.org/10.1111/1468-2230.12581 > accessed 04 December 2025
[7] Dr. Christopher Forsyth, ‘The legal effect of unlawful administrative acts: the theory of the second actor explained and developed’ (2001)35 AMICUS CURIAE 20 <https://sas-space.sas.ac.uk/3747/1/1330-1448-1-SM.pdf> accessed 03 December 2025
[8] Id.
[9] Joana Mendes, ‘ Law and administrative discretion in the EU: value of a comparative perspective’ YALE LAW SCHOOL <https://law.yale.edu/sites/default/files/area/conference/compadmin/compadmin16_mendes_law.pdf> accessed December 04 2025
[10] Arvind Datar, ‘Legal Notes by Arvind Datar: The theory of the second actor’ (Bar and Bench, 31 Aug 2024) < https://www.barandbench.com/columns/legal-notes-by-arvind-datar-the-theory-of-the-second-actor#:~:text=The%20theory%20of%20the%20%E2%80%9Csecond,the%20declaration%20of%20its%20illegality> accessed 04 December 2025.
[11] White & Anor v South Derbyshire District [2012] EWHC 3495 (Admin)
[12] White & Anor v South Derbyshire District [2012] EWHC 3495 (Admin)
[13] R v Wicks [1997] 2 WLR 876
[14] Forsyth (n 7) at 22
[15] Sivanandan C.T. & Ors v High Court of Kerala & Ors (2024) 3 SCC 799, ¶ 53
[16] Eyal Benvenisti, ‘The Interplay between Actors as a Determinant of the Evolution of Administrative Law in International Institutions’ (2012) 68(3) RESEARCHGATE < https://www.researchgate.net/publication/228175644_The_Interplay_between_Actors_as_a_Determinant_of_the_Evolution_of_Administrative_Law_in_International_Institutions#:~:text=This%20analysis%20provides%20the%20background,by%20disagreements%20among%20member%20states.&text=Content%20may%20be%20subject%20to%20copyright.&text=symposium%20of%20%22Law%20and%20Contemporary%20Problems%22.&text=disagreements%20among%20member%20states.&text=commentator%2C%20Lewis%20Kornhauser%2C%20and,most%20helpful%20comments.&text=completely%20and%20finally%20reduced%20to%20rules.%E2%80%9D > accessed 15 December 2025
