By Dinesh Dodamgoda –

Minister Lal Kantha has called for the abolition of the presidential pardon, apparently in response to recent controversies over its use. Although those concerns merit serious consideration, abolishing the power would remove an important constitutional safeguard. A better response is to identify why the pardon process has failed and how its administration can be strengthened.

This article explores the importance of the executive pardon as a democratic “institution”. It examines the failures of the political “organisations”—including politicians and public officials—that have developed around it in Sri Lanka. To address these failures, I recommend strengthening these organisations through education to shift attitudes and behavioural norms regarding the presidential pardon, thereby improving “informal institutional frameworks”.

The Executive Pardon

The executive power to grant pardons traces back to the monarch’s historic prerogative of mercy. Over time, democratic constitutional systems have shaped this power, creating a more thoughtful balance between executive discretion and the rule of law. No judicial system is perfect: human error, investigative overreach, and unreliable evidence can sometimes lead to unjust outcomes. By helping correct judicial mistakes, extending mercy and fairness, offering political amnesty, and serving as a check within the system of checks and balances, the power to grant pardons can make a meaningful contribution to justice and democracy.

The scope of executive pardon powers varies across democracies. Under the United States’ plenary model, the President may grant reprieves and pardons for federal offences, except in cases of impeachment. By contrast, the United Kingdom’s Westminster model places the royal prerogative of mercy within a framework of established constitutional conventions, ministerial advice and procedural limits. In Commonwealth and hybrid models, such as those of India and Sri Lanka, written constitutions expressly define the pardon power, and these systems combine the Westminster tradition of ministerial advice with formal safeguards to uphold the rule of law.

Article II, Section 2 of the U.S. Constitution gives the President broad discretion to grant reprieves and pardons for federal offences, but not for state offences. This authority does not extend to cases of impeachment and may be exercised at any stage of the legal process, even before formal charges are filed. For example, President Ford pardoned President Nixon in 1974, intending to bring an end to the “American tragedy” of the Watergate scandal.

In the U.K.’s Westminster model, the Sovereign grants pardons on the advice of the justice minister or the Cabinet. Procedural frameworks govern this power and are subject to judicial review and political accountability to Parliament. The courts have held that, while they would not second-guess the wisdom of a pardon, they can review whether a minister failed to properly consider the relevant legal criteria or acted unlawfully by refusing to assess the evidence.

In India’s codified Commonwealth and hybrid system, Articles 72 and 161 of the Constitution, respectively concerning the President and the Governors, provide that the power to grant pardons rests with the President and the Governors. Although pardon decisions are generally subject to limited judicial review, the Supreme Court of India has held that the President or the Governor acts on the aid and advice of the Council of Ministers. Courts may review these decisions on limited public law grounds, including arbitrary decision-making, mala fides (bad faith), reliance on irrelevant material, or political, caste, or religious bias.

In Sri Lanka, Article 34 of the Constitution authorises the President to grant pardons, respites, or commutations, but only through separate formal processes. For example, in death penalty cases, the President receives reports from the trial judge who heard the case, the Attorney-General, and the Minister of Justice, who submits a recommendation. The Supreme Court may also review the matter in a fundamental rights case to ensure compliance with Article 12(1) of the Constitution (right to equality).

Compared with the less restrictive US model, courts in the United Kingdom, India and Sri Lanka subject executive pardons to judicial scrutiny. Under established Commonwealth jurisprudence, courts do not assess the wisdom or merits of a pardon. Instead, they examine whether the decision-making process was lawful: whether the required constitutional procedures were followed; whether the decision was arbitrary; whether the relevant facts were properly considered; whether it involved mala fides or irrelevant considerations, such as personal or political favouritism; and whether it infringed fundamental or victims’ rights.

Failures in Sri Lanka

Recent public criticism, including Minister Lal Kantha’s remarks, reflects growing concern about the exercise of the presidential pardon in Sri Lanka. This concern has intensified following Supreme Court decisions in 2024 on the pardon in the Royal Park murder case, the pardon granted to Gnanasara Thero in 2026, and the 2026 clemency case under President Wickremesinghe’s administration.

In these cases, the Supreme Court found that the executive had failed to follow the procedure under Article 34 of the Constitution: that pardoning individuals convicted of contempt without reasoned justification was impermissible; and that the executive had failed to demonstrate a rational assessment or adherence to proper procedure. Taken together, these rulings demonstrate how arbitrary clemency can undermine the rule of law and the aggrieved parties’ right to equality under Article 12(1) of the Constitution.

Commenting on the 2026 Gnanasara Thero decision, former President Maithripala Sirisena said that officials at the Presidential Secretariat had misled him and that a technical error had contributed to the outcome. The explanation highlights possible administrative weaknesses but does not challenge the constitutional responsibility attached to the exercise of presidential power. Because the relevant constitutional and procedural requirements were violated in the cases discussed above (despite differing opinions), the shortcomings do not appear to stem from “institutions” (the constitutional provisions) or the “rules of the game,” as the Nobel Prize-winning scholar Douglass C. North defines “institutions” in his 1990 work on Institutions. Rather, they seem to lie with “organisations” or the “players”—the political bodies or individual actors who hold executive and administrative powers. I therefore find North’s theoretical framework of “institutions” and “organisations” a helpful way to understand and address shortcomings in Sri Lanka’s use of the presidential pardon power.

Institutions and Organisations

“Institutions” are the “rules of the game” in a society: humanly devised constraints that shape how people interact. They can be formal, such as constitutions, laws, and property rights, or informal, such as sanctions, customs, traditions, codes of conduct, and shared mental models. In this context, Article 34 of the Sri Lankan Constitution is a “formal institution”. By contrast, unwritten constraints—such as the customary obligation to protect loyalists and signals that they will be rewarded; norms around the use of pardons, political reciprocity, and elite bargaining; selective enforcement and the use of the law as a weapon; and religious, cultural, and commemorative traditions such as ‘Wesak Poya Day prisoner release’—are “informal institutions”, conventions, or codes of behaviour that govern how people interact on this issue.

Formal and informal institutions operate together. Article 34 establishes the constitutional authority and, in relevant cases, the procedure for granting pardons. Informal norms shape how political and administrative actors interpret and exercise that authority, while judicial review provides an enforcement mechanism against arbitrary, discriminatory or procedurally unlawful decisions.

The principal “organisations” or “players of the game” include the President, the Presidential Secretariat, the Ministry of Justice, the Attorney-General’s Department, political parties and Parliament. Their effectiveness depends on whether they understand their responsibilities, provide independent advice, maintain accurate records and comply with the constitutional process.

An organisation’s governance structure, the skills of its people, and the lessons it learns through experience all contribute to its long-term success. In practice, some organisations thrive while others face challenges. In the Gnanasara Thero case (2026), the Royal Park murder case (2024), and the Wickremesinghe clemency case (2026), the organisations failed to exercise the pardon power lawfully, and the enforcement mechanism—the Supreme Court’s judicial review—ultimately invalidated the presidential pardons.

Minister Lal Kantha’s opinion

Minister Lal Kantha’s proposal treats the constitutional power or the “institution” itself as the source of the problem. However, recent cases support a more nuanced conclusion: the danger arises when weak procedures, inadequate advice, political influence and harmful informal norms shape the exercise of that power. Thus, abolishing the institution will remove not only the risk of abuse but also the power’s legitimate capacity to correct injustice, extend mercy and respond to exceptional circumstances.

As North (1990) explains, institutions and organisations influence one another, shaping outcomes towards either success or persistent inefficiency. When they succeed, both direct and indirect demand for educational investment can encourage further development. This investment helps players or “organisations” strengthen their strategies and skills, increasing their chances of winning the “game” (granting the pardon). In contrast, North observes that limited investment in education can reduce productivity and lower the chances of winning the game.

Two Presidents granted these pardons. It is reasonable to assume that, as players, both may have been misled and lacked the necessary understanding of the pardon procedures. The administrative staff and other players may also have lacked the relevant knowledge or failed to advise the Presidents appropriately. Minister Lal Kantha may also have been misled or failed to fully understand that presidential pardons are a vital safeguard for justice and democracy.

Conclusion

Minister Lal Kantha’s call to abolish the presidential pardon therefore goes beyond the evidence. A more constructive response would preserve the constitutional power while strengthening the organisations responsible for its exercise. Reforms should include written procedural guidelines, specialist legal advice, complete records of the material considered, reasoned recommendations, conflict-of-interest safeguards, and appropriate public reporting, subject to legitimate confidentiality requirements.

Furthermore, education and professional training investment should form part of this reform, particularly for political office-holders and officials involved in the pardon process. Over time, clearer procedures and stronger accountability could also reshape the informal norms surrounding executive clemency. The presidential pardon should be understood not as an unrestricted personal privilege but as an important democratic constitutional responsibility that must be exercised lawfully, transparently and in the public interest.

*Dinesh Dodamgoda, a former Member of Parliament and Attorney-at-Law, holds an MSc in Global Security from Cranfield University in the United Kingdom. He has also conducted research at Newcastle University and at the Centre for the Study of Terrorism and Political Violence at the University of St Andrews. He was a Fulbright Scholar in U.S. National Security at the University of Delaware in the United States.