The Supreme Court’s Madhya Pradesh Government Formation Judgment: Round-Up

Below is a round-up of the six posts discussing the judgment of the Supreme Court in the Madhya Pradesh government formation case.

  1. A Question of Jurisdiction (by Rishav Ambastha)
  2. On the Powers of the Governor (by Anmol Jain)
  3. On the Powers of the Governor: A Response – I (by Amlan Mishra)
  4. On the Powers of the Governor: A Response – II (by Nivedhitha K)
  5. On the Powers of the Governor: A Rejoinder (by Anmol Jain)
  6. Some Concluding Remarks

Coronavirus and the Constitution – XX: Parliamentary Accountability

We have already discussed on this blog how the government’s measures to contain the coronavirus outbreak at both the state and union levels have bypassed legislative accountability (here). In this post, I want to expand the discussion on legislative accountability by exploring three points: (1) the legislature’s role in placing temporal limits on the executive’s emergency powers; (2) how involving the legislature incentivises policy scrutiny and increases transparency; and (3) understanding whether Indian legislatures could have continued to function. I use the term ‘parliamentary’ accountability, but the argument is equally applicable to legislative assemblies in the states. The goal of this discussion is not to suggest that parliament continuing to function would have served as a silver bullet to bad policy or governmental overreach. Rather, the idea is to explore how the democratic structures of our Constitution can act as a restraint on government power – making it imperative that we demand more from our elected officials.

The legal sources of the government’s measures are the Disaster Management Act 2005 (“DMA”) and the Epidemic Disease Act 1897 (“EDA”). Under these two laws, the government has issued several ‘guidelines’ that form the legal framework of India’s ongoing ‘lockdown’. We often think that legislatures, with their lengthy debates, committee procedures, and voting are cumbersome bodies not suited to the decisive action required in an emergency. The truth of this is debatable. For example, the numerous clarifications and addendums to the Ministry of Home Affairs’ circulars demonstrate the value of debate and committee scrutiny. However, the choice of the DMA and the EDA become especially significant when we consider the two courses of action the Constitution itself provides for dealing with extreme situations calling for immediate action: (1) declare a constitutional emergency; or (2) pass an ordinance. Both a constitutional emergency and the passing an ordinance require that the legislature (at some future point) ratify the government’s actions. Therefore, we can conclude that even the most extreme situations contemplated by the Constitution involve some legislative oversight. The DMA and the EDA however bypass the legislative branch altogether resulting in a “rule by executive decree”. This has some important consequences.

The temporary nature of emergencies

Emergency powers are fundamentally temporary. The term ‘derogation’ is often used to explain the relaxing of some legal requirements during periods of crisis – but such relaxations are essential exceptions to the ordinary rule of law and all derogations have a recognised beginning and most importantly an end. When the crisis subsides, the political and legal system must return to normal. When a constitutional emergency is proclaimed under Article 352 of the Constitution, the emergency automatically ceases after one month if not placed before Parliament and (if approved) again automatically ceases after six months unless re-approved by Parliament. Even ordinary legislation granting wide emergency powers such as the U.S. PATRIOT Act (enacted in 2001 to fight terrorism post the 11 September attacks) contained a ‘sunset’ clause providing that large parts of the Act would cease to operate in 2005 unless renewed. An even more poignant example is the U.K.’s Coronavirus Act. Before it was passed, the (then) Coronavirus Bill granted the government emergency powers for two years. After the opposition objected, the government agreed to amend the Bill – providing that the House of Commons must debate and renew emergency powers granted by the Bill every six months. In evaluating the Coronavirus Bill, the House of Lords’ Constitution Committee noted: “Two years would have been too long for these powers to have operated without reapproval by Parliament and we welcome the cross-party agreement to reduce the period to six months” (here). The rationale behind granting the executive sweeping emergency powers is therefore always contingent on (1) the existence of a crisis; and (2) a return to normalcy at the earliest possible instance.

The following questions now become relevant. Who declares the start of an emergency? How long will the emergency last? Does it need to be renewed? Who is responsible for its renewal? Can the legality of the declaration be challenged in a court?

Precisely because an emergency vests significant powers in the executive, it makes sense that another body should be responsible for the declaration of an emergency. Otherwise what stops the executive from invoking an emergency in manner that is at best frivolous, and at worst self-serving. For example, a government could declare an emergency and use its emergency powers to silence political dissidents. In the U.S. even though the executive is tasked with all operational aspects of fighting a war, the executive cannot take any action unless the legislature (Congress) first passes a declaration of war. The U.K.’s unwritten constitution contains no notion of emergency powers and the executive is entirely reliant on the legislature first passing a legislation enabling the executive to exercise any additional powers. The same principle applies to renewals, it makes sense that a body other than the executive is responsible for renewing the executive’s emergency powers. Requiring another body to determine whether a situation is truly exigent and how long it will likely last for is an essential check against the abuse of emergency powers by the executive. This is a significantly stronger check than having courts adjudicate on the legality of the emergency proclamation after it is made as the damage may already be done by the time the court delivers a verdict (especially at the pace the Indian Supreme Court currently hears politically sensitive cases).

The Indian Constitution does envision Parliament playing a crucial role in the declaration of constitutional emergencies by requiring that all emergency proclamations be placed before Parliament within one month. Admittedly Parliament’s record of acting as a check on the executive with respect to emergency proclamations is a poor one – in July 1974 Parliament ratified the Indira Gandhi government’s emergency proclamation leading to three years of systemic governmental overreach. One may assume given India’s model of parliamentary democracy (where the executive’s party commands a majority in the legislature), legislative ratification is a forgone conclusion. While the individual judgement of parliamentarians has been severely curtailed by the anti-defection law, there still exists the chance (especially in coalition governments) that the government will have to work harder to appease the various factions of its own support base. Moreover, having even a token debate in parliament is a significant improvement on the current situation.

It is important to understand that the DMA requires no formal declaration of emergency (even under the Act’s own framework). Therefore, the ground reality is that the executive has been granted sweeping emergency powers, the courts are virtually at a standstill and public gathers have been outlawed – all without an emergency being ‘declared’ or any justification for when exactly the coronavirus became an emergency or any indication of when the emergency will end. (The use of the term ‘notified disaster’ was used widely in the media but has no relation to the beginning or ending of the government’s emergency powers and only concerns the use of disaster relief funds.) It is particularly important to recognise that the coronavirus outbreak may be with us for a while. The nationwide ‘lockdown’ has already been extended twice. Much like the ‘war on terror’, what seemed necessary as an immediate response can perpetuate a permanent derogation from the rule of law. In a Parliament approved emergency, ratification by Parliament may have been (and with good reason) a foregone conclusion in March, but three, six or nine months later the government may have faced some pressure to justify a renewal of its emergency powers. Under the DMA however, they face no such pressure to justify a continued resort to emergency powers. Therefore, it is crucial to create temporal boundaries on the invocation of emergency powers and it is submitted that ratification by parliament is one method to do so.

Incentivising parliamentary scrutiny

The ratification of emergency proclamations and ordinances by Members of Parliament creates a powerful incentive for them to scrutinise government action and can require the government to moderate its position. During the coronavirus outbreak Italy used ‘Decree-Laws’ – a decree issued by the government which must be placed before the Italian Parliament within sixty days (here). Similarly, in India, an ordinance cannot be enacted unless Parliament is not in session and must be placed before Parliament upon its reassembly. The ordinance expires unless expressly passed as a piece of legislation by Parliament within six weeks of Parliament reassembling. What this means is that parliamentarians are on the hook for the legal content of these measures. While this may be less of an incentive for members of the ruling party, it creates an incentive for members of the opposition to scrutinise the legislation. Parliamentarians are held electorally accountable for their votes for or against a legislation and draconian measures by a government are unlikely to pass without some form of debate and scrutiny when brought before Parliament. For example, recently the opposition forced multiple adjournments in the Lok Sabha until the government agreed to discuss the communal violence in Delhi (here).

This brings us neatly to the point of parliamentary questions and debate – where members of the opposition have unfettered access to government ministers. Although the actual mechanisms of questions and the debate (or ‘zero-hour’) in Parliament are worth an independent exposition a few key points may be made here. The first is the nature of the threat we are currently facing. Unlike a war with another nation or the fight against ‘terrorism’ where increased transparency may hamper the government’s efforts to defeat the threat – in a public health crisis more transparency is always better. The government should disclose the scientific data on which it bases its decisions. This will not hamper the fight against the coronavirus but will expose bad decision making. Second, unlike questions asked in a newsroom or addressed on social media, questions asked in Parliament form part of the official record of the House and can be used to hold ministers accountable (here). Third, unlike the news media, Members of the legislature cannot be silenced by the courts (see Articles 105 and 194 of the Constitution under which Members enjoy legislative privileges). The Supreme Court’s recent efforts to silence ‘fake news’ and instead mandate reliance on information produced by the government and the statements of the Solicitor General of India make this particularly relevant. Lastly, parliamentary proceedings are broadcast live on national television and on the internet.

There is a deeper point to be made here. Government transparency, and ministerial accountability is fundamentally tied to the broader question of electoral accountability. During proceedings in parliament, it is government ministers who answer questions. This allows voters to evaluate the performance of the government first-hand in an unfiltered manner. To date, a Joint Secretary from the Ministry of Health has given almost all the coronavirus press briefings. The Joint Secretary is an unelected official and making such an official the face of the crisis disassociates ministerial responsibility from the actions of the government in combatting the crisis. Even in the U.S. where the President is not politically accountable to the legislature (outside the extreme case of impeachment), as the head of the executive the President has continued to deliver daily press conferences and answer questions – in stark contrast to the head of the executive in India. Neither the Health Minister not the Prime Minister is legally obligated to give press briefings, but their refusal to do so makes Parliament one of the last forums where the government can be asked hard questions.

Could Parliament have continued to function?

This post would not be complete without addressing the elephant in the room. On 23 March 2020 Parliament was adjourned over fears that the gathering of Members would act as a vector for transmission. Two questions must be answered: (1) can Parliament legally meet outside its official seat; and (2) what the alternative options that Parliament can adopt are. Article 85 of the Constitution permits the President to summon Parliament “at such time and place as he thinks fit” and Rule 11 of the ‘Rules of Procedure and Conduct of Business in the Lok Sabha’ states that “A sitting of the House is duly constituted when it is presided over by the Speaker or any other member competent to preside over a sitting of the House under the Constitution or these rules.” (Rule 10 of the Rajya Sabha rules contains an analogous provision.) Therefore, the short answer is that the place of meeting does not matter so long the President summons Parliament and the Speaker, or other competent person, presides over Parliament. (Interested readers may refer to Shubhankar Dam’s article on precisely this point. He cites historical instances where the ‘place’ of meeting has been in dispute.)

Therefore theoretically, Parliament can meet at an alternative location that is more conducive to social distancing norms or even potentially online. Some comparative context is useful here. Several countries, Australia, New Zealand and Germany amongst them, have struggled to keep their legislature’s open and have adjourned them during the present outbreak. However, Congress in the U.S. has met several times to pass emergency economic legislation. The House of Commons in the U.K. has met virtually, with Members asking questions from remote locations. Similarly, the Canadian Parliament has managed to meet virtually. Perhaps most tellingly, the Indian Supreme Court and various High Courts have managed to implement protocols to allow virtual hearings for thousands of litigants since the beginning of the ‘lockdown’. There would certainly be some teething troubles, but it would not be beyond the realm of possibility to assume that the Indian Parliament could continue to function during the ‘lockdown’. I will end this discussion with two points. As I noted earlier, Parliament sitting is not a silver bullet to all the country’s ailments – given the legislative frameworks which already exist under the DMA and the EDA and the ever present anti-defection law Parliament’s role would certainly be limited. However, eliminating Parliament from governance during an emergency is neither contemplated by the Constitution nor is should it be condoned by the voters who elected this Parliament. At the very least, the limited benefits of Parliament sitting highlighted here could be secured.

Concluding thoughts

Notions of accountability flowing from a separation of powers model focus on preventing the centralisation of power in one body. For example, the body making the law should be distinct from the body implementing or interpreting the law. However, this model fails to incentivise those in power to listen to citizens. Legislators in the minority can blame those in the majority, the government can blame the judiciary for curbing its measures, and the judiciary can blame an overzealous legislature or the executive (see the eternal dispute over judicial appointments).

Competition for power through democratic structures creates a vital link between citizens and their agents in government. It forces the three wings to look beyond horizontal competition inter-se the executive, the legislature and the judiciary and consider a vertical balance of power where rival power seekers must convince citizens of their ability to effectively govern. This is particularly effective in a plural society such as India where there exists a multiplicity of groups with cross-cutting interests and membership, forcing those in power to cater to a wide range of interests. Parliamentary accountability is one of the best examples of how competition for power can further the interests of citizens. ‘Politicking’ during a crisis may be frowned upon but is also an excellent method to ensure that the citizen’s preferences are accounted for in governance. It forces those in power to accommodate the needs of more diverse interests or risk losing the mandate to govern. A healthy legislative body should ensure this robust competition for power. The strength of the Indian parliament as a deliberative body which can hold the government accountable has been in decline for some time, but the present crisis should not be its death knell.

The Supreme Court’s Madhya Pradesh Government Formation Judgment – IV: A Response to Anmol Jain (2) [Guest Post]

[Editor’s Note: Justice is an indivisible concept. We cannot, therefore, discuss contemporary Supreme Court judgments without also acknowledging the Court’s failure – at an institutional level – to do justice in the case involving sexual harassment allegations against a former Chief Justice. This editorial caveat will remain in place for all future posts on this blog dealing with the Supreme Court, until there is a material change in circumstances.]

[This is the fourth post in the series examining the Supreme Court’s judgment in Shivraj Singh Chouhan v Speaker, Legislative Assembly of Madhya Pradesh and Ors. This is a guest post by Nivedhitha K.]

In this essay, I respond to Anmol Jain’s post on the recently delivered judgment in Shivraj Singh Chouhan & Ors. v. Speaker, Madhya Pradesh Legislative Assembly & Ors. The facts of the case were summarised in the previous post. The crux of the case was: whether the 14th March 2020 communication of the Governor directing a floor test on 16th March was constitutional. Thus, the question that needed to be answered was whether the Governor had the power under the Constitution to direct a floor test to be conducted.

I will divide the post into two sections. In the first section, I will rebut the argument that the discretionary power of the Governor under Article 163 can only be exercised concerning those functions that are expressly within his discretion, under the Constitution. I will argue, invoking the doctrine of necessary implications, that the Governor can exercise his discretionary powers even when such powers have not been expressly conferred under the Constitution. In the second section, I will argue that the submission of the author that the bench’s reliance on Bommai and Rebia is misplaced is erroneous. I will argue that Bommai is a precedent for the exercise of guided discretion by the Governor, when he is met with a situation of political uncertainty such as in this case. I will then argue that in the subsequent reliance on Rebia, the author in contravention to his earlier stand, admits an exception for the exercise of discretionary power when the situation is unworkable. I will conclude by referring to the observation in Shivraj on the importance of the due process of no-confidence motion, as opposed to conducting a floor test through the direction of the Court/Governor.

Power to Direct a Floor Test under Article 174 r/w 175(2)

Anmol’s preliminary point of argument is that the Court holds that the power to direct a floor test falls under Article 163 of the Indian Constitution. I rebut this point. The Governor’s power to direct a floor test is inherent under Article 174 r/w 175(2) and not Article 163.

An excerpt from the communication of the Governor to the Chief Minister is below:

“………, it is necessary for you to gain the trust vote in Vidhan Sabha immediately after my speech on 16.03.2020. In this regard, I by exercising the powers conferred by Article 174 r/w 175(2) of the Constitution and other Constitutional powers vested in me…..”

When the Court held that the communication of the Governor was not ultra vires the Constitution, it accepted the plea that Article 174 r/w 175(2) is the repository of power to direct a floor test.

Article 174 gives the Governor the power to summon, prorogue, or dissolve the Assembly. He used this power to direct that the ‘Session of Madhya Pradesh Vidhan Sabha will begin on 16th March 2020 w.e.f. 11 a.m. in the morning.’ Article 175(2) states that the Governor may send messages to the Assembly with ‘respect to a Bill then pending in the Legislature or otherwise and the house shall with all convenient dispatch consider any matter required by the message to be taken into consideration.’ The Governor relied on this provision to direct that ‘after my speech, only one work will be done i.e. trust vote.’

My argument is that the Governor has the power to direct a floor test under Article 174 r/w 175(2) and that Article 163 does not provide the power to the Governor but merely guides the exercise of power. Article 163(1) states that “There shall be a Council of Ministers with the Chief Minister at the head to aid and advise the Governor in the exercise of his functions, except in so far as he is by or under this Constitution required to exercise his functions or any of them in his discretion.” A plain reading of Article 163 is sufficient to conclude that the Governor must exercise his power usually at the aid and advice of the Council of Ministers unless ‘he is by or under this Constitution required to exercise his functions ….in his discretion’. Therefore, aid and advice is the rule and discretionary exercise of power is the exception. To identify the ambit of discretionary exercise of power, it is necessary to interpret the words ‘except in so far as he is by or under this Constitution required to exercise his functions or any of them in his discretion.’

The Drafting History of Article 163

Anmol refers to the Constituent Assembly debates where the members observed that the Governor shall exercise discretion only when specific functions are expressly stated to be exercised through discretion. Indeed, an amendment was introduced by Mr. Kamath to remove ‘except in so far as he is by or under this Constitution required to exercise his functions or any of them in his discretion’ from Article 163 (Article 143 in the draft Constitution). Mr. Krishnamachari, Mr. Alladi Krishnaswamy Iyer, and Dr. Ambedkar argued that it was only a question of drafting and that Article 163 was a general provision for the exercise of discretionary powers by the Governor which have been specifically provided under other provisions of the Constitution. Alladi Krishnaswamy Iyer observed:

Sir, there is really no difference between those who oppose and those who approve the Amendment. In the first place, the general principle is laid down in Article 143 namely, the principle of ministerial responsibility that the Governor in the various spheres of executive activity should act on the advice of his ministers. Then the Article goes on to provide: except in so far as he is by or under this Constitution required to exercise his functions or any of them in his discretion.’ So long as there are Articles in the Constitution which enable the Governor to act in his discretion and in certain circumstances, it may be, to over-ride the cabinet or to refer to the President, this Article as it is framed is perfectly in order.

In this context, there was repeated reference to Articles 175 and 188 of the draft Constitution. Dr. Ambedkar stated that if the House decided to divest the Governor of his discretionary power when provisions that specifically provided discretionary power to the Governor with regard to certain functions such as Articles 175 and 188 were taken up for debate, the general discretionary clause under Article 143 (now Article 163) could be deleted:

I personally myself would be quite willing to amend the last portion of clause (1) of Article 143 if I knew at this stage what are the provisions that this Constituent Assembly proposes to make with regard to the vesting of the Governor with discretionary power. My difficulty is that we have not yet come either to Article 175 or 188 nor have we exhausted all the possibilities of other provisions being made, vesting the Governor with discretionary power. If I knew, I would very readily agree to amend Article 143 and to mention the specific Article, but cannot be done now.

The proviso to Article 175 of the draft Constitution (now Article 200) stated that the ‘Governor may, in his discretion, return the Bill together with a message requesting that the House will reconsider the Bill or any specified provisions thereof … and that if the Bill is passed again with or without the amendment, the bill will have to be assented to by the Governor.’ Dr. Ambedkar introduced an amendment to remove the phrase ‘discretion’ from the provision. Commenting on the introduction of the amendment, he stated that ‘in a responsible government there can be no room for the Governor acting on discretion’. The amendment was accepted, and it led to the present form of Article 200 without the phrase ‘discretion’. Similarly, Article 188 specifically provided the Governor with the discretionary power to exercise his functions without the aid and advice of the Council of Ministers for two weeks if a grave emergency that threatened the peace and tranquillity of the State had arisen. Then, the President might – if he was satisfied – assume to himself the functions of the State. Dr. Ambedkar moved an amendment to delete Article 188 of the draft Constitution. He was of the view that the discretionary power for a short period was ‘futile’ if the President was going to take over after two weeks. This amendment was also accepted. The Governor was also – earlier – expressly provided discretion in the appointment and dismissal of his ministers (Art 144(6)), summoning and dissolution of the legislature (Art 153), the appointment of the Provincial Auditor –in- chief (Art 210) and members of the Public Service Commission (Art 285). Discretionary clauses in all these provisions were deleted.

Mr. Kamath brought to the notice of the House the reassurances of Dr. Ambedkar that Article 163 would be appropriately amended if the specific discretionary powers in the subsequent provisions were removed. However, Article 163 was intact in its entirety in the final text of the Constitution. The only matter in which the Governor was expressly required to act in his discretion was with regard to the administration of tribal areas in Assam for a transitional period. Therefore, the final text of the Constitution provided for complete ministerial responsibility. It is thus clear that the debates and the final text of Article 163 have no nexus between them.

The interpretation of the phrase ‘by or under the Constitution’ solely relying on the Constituent Assembly debates is erroneous for the following reasons. Firstly, though through a period, the Constituent Assembly debates have been accepted as an aid for interpretation, it cannot override the text of the Constitution. In this case, the debates on Article 163 and the final text of the provision share no nexus. And secondly, an interpretation that makes the Constitution workable will have to be used. If Anmol’s interpretation is accepted, it would render the clause providing general discretionary power in Article 163 redundant since the unamended Constitution did not specifically provide for discretionary exercise of power concerning any of the Governor’s functions.

Reports of Sarkaria Commission and Punchhi Commission

The Sarkaria Commission had interpreted the discretion exercisable by the Governor by the Constitution to include two classifications: (1) through express provisions; (2) through necessary implication. According to the Commission, the phrase under the Constitution means the discretion exercisable by virtue of the rules made under the Constitution. By the amended Constitution, the Governors have been required to expressly act in their discretion with respect to the administration of Arunachal Pradesh, Assam, Meghalaya, Mizoram, Nagaland and Sikkim. The Commission, echoing the words of Pandit Das Bhargava in the Constituent Assembly, observed that there might be situations in which it would not be possible or practicable for the Governor to receive the advice of the Council of Ministers. In such cases, the Governor’s discretionary power can be interpreted through necessary implication. A few of the examples mentioned in the Commission’s report are the appointment of the Chief Minister after an election under Article 164(1), submission of report under Article 356 on the breakdown of the constitutional machinery in the State, and recommendation of a bill for the consideration of the President under Article 200. These views have been accepted by the Punchhi Commission as well.

Power to Exercise Discretion through Necessary Implication

Let us now look at the text of the Constitution independent of the Commission’s report to ascertain as to whether the discretionary power exercisable by the Governor can be read through necessary implications. Article 163(2) states that ‘if any question arises whether any matter is or is not a matter as respects which the Governor is by or under this Constitution required to act in his discretion, the decision of the Governor in his discretion shall be final, and the validity of anything done by the Governor shall not be called in question on the ground that he ought or ought not to have acted in his discretion.’ If the discretion was only purported to be restricted to express discretionary provisions, the question of the ambiguity of exercisable discretion itself would not arise.

Let us examine two factual situations to identify if excluding the reading of the discretionary power of the Governor through necessary implication would lead to an unworkable Constitution. For the first situation, let us take the example of the latest political crisis in Maharashtra in December 2019. None of the parties had the majority to form the Government, and the coalition among the parties was in the nascent stage in the first week after the election. In that case, there was no council of ministers to aid and advise the Governor on whether there was a breakdown of the Constitutional machinery for him to send a report to the President under Article 356 of the Indian Constitution. For the second situation, let us take the example of the political crisis in Madhya Pradesh in March, 2020. It seemed that the Government no more exercised a majority after twenty-two MLAs submitted their resignations, of which only six were accepted. The numbers were enough for the Governor to form a prima facie objective opinion that the Government might not be enjoying a majority. Let us assume that the Assembly was not in session for another month – unlike the instant case where the Assembly was already summoned for 16 March 2020. In that case, if interpreted to mean only expressly provided discretionary power, the Governor would not hold the power to summon the House unless he was advised by the Council of Ministers – which might be too far-fetched to expect, since it was their majority that would have to be tested in the house. Pandit Das Bhargava brought these contingent situations to the knowledge of the House when he observed: ‘It is quite right that so far as our conception of a constitutional Governor goes he will have to accept the advice of his ministers in many matters but there are many other matters in which the advice will neither be available nor will he be bound to accept that advice.’

The majority in Rebia recognised the fallacy of restricting the discretionary power of the Governor to express provisions and observed that there might arise a situation when the Chief Minister no more enjoyed the majority in the House. If the Governor could arrive at an objective opinion that a doubt was cast on the numbers of the ruling party, the Council of Ministers would lose their legitimacy to advice the Governor. In this context, it was observed:

The above position (of aid and advice) would stand altered if the Government in power has lost the confidence of the House. …….. However, where there is reason to believe, that the Government in power no longer enjoys majority support, it is open to the Governor, to take steps to determine the issue of majority by a floor test……….. We find no justification in taking a different view, than the one expressed by the Justice Sarkaria Commission report, conclusions whereof were reiterated by the Justice M.M. Punchhi Commission report. We endorse and adopt the same, as a correct expression of the constitutional interpretation, insofar as the present issue is concerned.

Reliance on Bommai and Rebia

Anmol argues that reliance on Bommai and Rebia by the Court was misplaced. With regard to Bommai, he argues that it is not a precedent for the direction of the floor test by the Governor. He argues that the decision leaves open the question of the process through which the Chief Minister is to prove his majority. Even if assuming for the sake of argument that Bommai did not support a Governor-directed floor test as the only means of testing if the Chief Minister enjoyed the confidence of the House, Bommai would not be immaterial for the adjudication of a factual situation such as Shivraj. According to Bommai, the Governor through his discretionary power could send a report to the President under Article 356 only after he fulfilled his duty of summoning the House and calling the Chief Minister to prove his majority- except when he was faced with a violent situation in the House. Therefore, in Bommai it was observed that the Governor could summon the house even without the aid and advice of the Council of Ministers. If anything, Bommai is a validation of interpreting the discretionary power of the Governor through necessary implications.

No-confidence Motion and Floor Test

Anmol argues – based on Rebia – that even when the Assembly is not in session, ‘no-confidence motion remains significant for initiating the process, and based on this motion, the Governor is empowered to act even against the aid and advice of the Council of Ministers.’ His argument here is two-fold: one, the initiation of the process to prove the confidence of the House must only be through the no-confidence motion, both when the Assembly is in session and is not in session. Second, when the Assembly is not in session, the initiation of the process should be through a no-confidence motion, but the Governor can act against the aid and advice of the Council of Ministers, if through the advice, the Council of Ministers attempts to by-pass proving the majority. In this argument, the author concedes that there may arise situations in which the advice of the Council of Ministers cannot be accepted. Accepting the advice in these situations would abrogate the principle of ministerial responsibility. Therefore, the author has himself carved out an exception for exercise of discretionary power if the situation is unworkable. This is the basis of deducing discretionary power through necessary implications.

His argument that even if the Assembly is not in session, the process of proving the majority has to be initiated through a no-confidence motion is erroneous. According to Rule 143 of the MP Assembly Rules, to introduce a no-confidence motion a member shall seek leave from the Speaker to introduce the motion and on the very same day give a written notice to the Secretary General. For grant of leave for motion, atleast 1/10th of the total number of members must vote in favour of the motion. On grant of leave, the Speaker shall allot a date within 10 days from the date on which the leave is sought. Therefore, it is evident that the no-confidence motion can be initiated only when the Assembly is in session. Thus, if there is a surge of political uncertainty when the Assembly is not in session, the first step is not to initiate a no-confidence motion but rather to summon the Assembly. If the interpretation of the author on the express discretionary power of the Governor is accepted, then neither can a no-confidence motion be initiated by the member nor can a floor test be directed by the Governor.

The bench in Shivraj seemed to be wary of the possible misuse of the discretionary power of the Governor to direct a floor test. It stated that the usual mode of proving the majority is only through a no-confidence motion:

In exercising the constitutional authority to demand a trust vote, the Governor must do so with circumspection in a manner that ensures that the authority of the House to determine the existence or loss of confidence in the government is not undermined. Absent exigent and compelling circumstances, there is no reason for the Governor to prevent the ordinary legislative process of a no confidence motion from running its due course. (paragraph 45)


However, the bench did not expound the meaning of the phrase ‘exigent and compelling circumstance’. It was ultimately held that the communication of the Governor was intra vires the Constitution. Therefore, it can be presumed that the Court regarded the instant factual situation to fall within the ambit of ‘exigent and compelling circumstance’. If the Governor had not directed for a floor test, then by virtue of the procedure under the MP Assembly Rules, even if leave was granted for no-confidence motion (since the session was in anyway to begin on 16th March), a 10 day cap for the no-confidence motion would be too long a period for the Assembly to function during the COVID-19 pandemic. In the instant case, the Assembly was also adjourned on 16th May, preventing the possibility of granting a leave for no confidence motion. Though there is no such reasoning in the judgment that the ‘exigent circumstance’ in the instant case was the surge of the COVID-19 pandemic, one can merely presume that these were the factors that guided the court.

[Disclaimer: The author is an intern in the office of Dr. DY Chandrachud J. Views are personal, and she did not substantially assist in the case.]

The Supreme Court’s Madhya Pradesh Government Formation Judgment – III: A Response to Anmol Jain [Guest Post]

[Editor’s Note: Justice is an indivisible concept. We cannot, therefore, discuss contemporary Supreme Court judgments without also acknowledging the Court’s failure – at an institutional level – to do justice in the case involving sexual harassment allegations against a former Chief Justice. This editorial caveat will remain in place for all future posts on this blog dealing with the Supreme Court, until there is a material change in circumstances.]

[This is the third post in the series examining the Supreme Court’s judgment in Shivraj Singh Chouhan v Speaker, Legislative Assembly of Madhya Pradesh and Ors. This is a guest post by Amlan Mishra.]

The judgement of the Supreme Court on the Madhya Pradesh political crisis, which occurred in March, has been discussed on this blog. This piece by Anmol Jain seeks to critique the judgement for bringing the question of ‘whether confidence is lost or not’ under the purview of the Governor’s discretion. He argues that Art. 174 and 175 r/w Art. 163 (sending messages/directions and summoning the Legislature using the Governor’s discretion) provide for exercise of discretion only in those situations explicitly provided by the Constitution. Anmol stresses that while directing a government to face a no-confidence motion, initiated inside the house, is within the scope of the Governor’s ‘discretionary directions’, independently directing a trust vote when no such motion exists is beyond his discretion. In this post, I seek to argue that Anmol’s strict reading is not well founded. For there could be scenarios (like in this case) where the government may adopt delay tactics in order to deter any political process (like a no-confidence motion) aimed at holding it accountable from taking place. Lastly, I would also assess how the court navigated the political thicket in this case.

On Discretion

It is well-recognised that the Governor’s role is to form a government which enjoys support in the legislature. It has been held in Rameshwar Prasad v. Union of India, that the mere individual opinion of the Governor cannot be a ground for imposing President’s rule (under A. 356), but an attempt should first be made to test the government on the floor of the house. This means that the Governor has the duty to determine support on the floor of the house, before recommending invocation of A. 356. Here, the court also held that in a ‘parliamentary democracy of a state’, there should be at all times, either a democratic ‘popular’ government or the state should be under President’s rule. It stressed that no legislative assembly can be ‘live’ in the absence of an executive government. Thus the Governor should have means to resolve a situation when the ‘majority’ of a government is in question so that he can either invoke Art 356 or explore other means of forming a democratic government. This is a discretionary power inherent in A. 356 and in the collective responsibility of the Council of Ministers (A. 75). This discretionary power has been recognised more explicitly in Nabam Rebia v. Deputy Speaker. That is, of course, not to say that the power to make such a determination should be exercised at his whims or fancy.

This begs the question: should a Governor ‘interfere’ to resolve the question of majority or should he let political processes inside the legislative assembly resolve itself? Notice, if the process inside the legislature is likely to present an answer (eg. a no-confidence motion has been accepted by the Speaker and is to be scheduled soon), he need not interfere in the legislature. But where the political processes, for whatever reason, are unlikely to present an answer, he may issue directions ordering a floor test. He must do so, because political processes may not present an answer (i.e., establishing a majority), making it impossible for him to secure the continuance of a ‘popular government’. The above test of: ‘Are the political processes likely to present an answer?’ should then serve well in determining the boundaries of Governor’s exercise of discretion in cases where majority is suspect.

In this light, let us appreciate this extract from Nabam Rebia, which Anmol cites to buttress the central role of a ‘no-confidence motion’ before the Governor can ‘interfere’ by exercising his discretion:

The Chief Minister and his Council of Ministers lose their right to aid and advise the Governor, to summon or prorogue or dissolve the House, when the issue of the Government’s support by a majority of the Members of the House, has been rendered debatable. … And in such a situation, if there is a non-confidence motion against the Chief Minister, who instead of facing the Assembly, advises the Governor to prorogue or dissolve the Assembly, the Governor need not accept such advice.

Anmol highlights ‘the non-confidence motion’ to draw attention to the fact that Governor’s discretion kicks in only when a non-confidence motion is pending in the house. This I argue is a simplistic reading of this paragraph. Of course, the situation envisioned in the paragraph is one scenario where the Governor may refuse to dissolve/prorogue the house. But a purposive reading also means that he may use his discretion in directing a floor test, if the government uses delay tactics to not let the no-confidence motion be introduced. In other words, if political processes inside are proving inadequate or are being circumvented, he may use discretion in directing a floor test. Notice this follows logically from the paragraph. Putting a strict reading on it misses the purpose of the paragraph, which is to stop delay tactics by the government (like prorogation, adjournment etc) to avoid a test of its confidence. What if instead of prorogation (explicitly mentioned in this extract), the government resorts to adjournment to avoid testing its confidence, such that processes inside the legislature cannot present an answer? I argue that this being the situation in this case, the Governor’s discretionary directions were rightly upheld by the court.

Adjournment as a Delay Tactic: The Case of MP

Here, I show that in the MP Scenario, the government used adjournment to avoid the test, thereby creating a situation where political process could not throw up any answer to the question of majority. This legitimised THE Governor’s interference to settle the question. Notice this paragraph, in J. Chandrachud’s judgement, where he traces how adjournment made determination of ‘majority’ difficult:

The Chief Minister, adverting to the turmoil in the state, addressed a communication to the Governor on 13 March 2020 stating that the convening of the floor test would be a sure basis for resolving the conundrum. This is a strong indication that the Chief Minister himself was of the opinion that the situation in the state had cast his government‘s majority in doubt. However, upon the convening of the Legislative Assembly, no floor test was conducted, and the House was adjourned till 26 March 2020. These facts form the basis on which the Governor advised that a floor test be conducted. Based on the resignation of six ministers of the incumbent government (accepted by the Speaker), the purported resignation of sixteen more Members belonging to the INC, and the refusal of the Chief Minister to conduct a floor test despite the House having been convened on 16 March 2020, the exercise of power by the Governor to convene a floor test cannot be regarded as constitutionally improper.


This means that had the government not adjourned the house, and kept open the prospect of a no-confidence motion on the floor, the exercise of Governor’s discretion in this case would have been unwarranted. Since that did not happen, and instead, delay tactics were used, the Governor’s ‘interference’ in the legislature to check the political accountability of the government was justified.

Now, one can still argue that the Governor had sent directions on March 14th even before the house could hold its first session (on 16th March), and before any political process inside the legislature could begin. Thus, the argument may conclude that the Governor’s directions were untimely, as no delay tactics had been employed till them. However, the Governor kept reiterating his directions even after the adjournment of the house (made on 16th March till 26th March). This lent legitimacy to his directions, once the government started adopting delay tactics.

Avoiding the Political Thicket

Now, I wish to analyse how the court navigated the political thicket in this case. Courts traditionally have been wary of taking decisions which sway the balance in favour of a particular political party, and rightly so. Likewise in this case, the Congress Party and the Speaker argued that ordering an immediate floor test would ‘short-circuit’ the power of the Speaker in deciding the question of resignations of MLAs. This is important because unless the Speaker makes a decision about resignation before the floor test, all decisions about the disqualification/ resignation of MLAs may become irrelevant. Once the Government falls (as it did), the Speaker has very little time (before he is replaced) to decide on disqualification and resignation of MLAs. This chain of events may allow rebel MLAs to vote against their party whip, and still survive their disqualification, once the government changes (and there is a favourable Speaker).

It must be recognised (as has been argued in the blog here and here) that a remedy should have been fashioned which allowed the Speaker adequate time to decide these questions properly without ‘short-circuiting’ his decision. However, the court in this case failed to fashion such a remedy and merely noted that there is no explicit bar on the Speaker’s decision, and the floor test and these decision could run parallalely. Such a balance as Bhatia notes elsewhere is ‘not any balance at all’.

However, the court did try to equalise the setting by ordering a floor test the next day. The court noted that this would decrease the prospect of rebel MLAs sealing deals with the new government and thereby lessen the chances of them violating the Tenth Schedule (on defections). Readily ordered floor-tests are increasingly becoming a great, though unequal tool, to lessen subversion of democratic commitments and stop horse-trading. This also lessens unhelpful accusations of mala-fide in the functioning of the Governor (in sending directions to the Assembly) or the Speaker (in adjourning the assembly). Inherent in ordering an immediate floor test is the idea that despite bad faith by constitutional functionaries, the floor of the house if the place for determination of these questions and not the court-room.

Another attempt by the court was to allow interactions between the Speaker and the rebel MLAs through video-conferencing at a ‘neutral setting’ so that Speaker could take a decision on the resignations. This suggestion was declined by Sr. Adv. Singhvi because he did not have ‘instructions’ from his clients. However, this presents innovative ways of resolving political crisis, while allowing political processes to continue inside the legislature.

Overall, in my view, the Court did a decent job of delineating the discretionary powers of the Governor and tows a sensible line in navigating the political thicket.

Postscript: The Supreme Court’s Problematic Order in the Karnataka Case

Yesterday, I wrote that the ongoing Karnataka controversy represents a breakdown of constitutional conventions. This breakdown creates a space for inevitable judicial intervention – but a space that is fraught with risk for the Court. In fashioning a remedy, the Court ought to make it as difficult as possible for the warring political functionaries to subvert constitutional conventions, while leaving the final solution to the existing democratic processes.

Today morning, the Court passed an order in the case. Noting that it had to maintain a “constitutional balance between the competing and conflicting rights”, it refrained from issuing any directions to the Speaker to decide upon the resignation and disqualification petitions. However, the Court also held that “until further orders the 15 Members of the Assembly, ought not to be compelled to participate in the proceedings of the ongoing session of the House and an option should be given to them that they can take part in the said proceedings or to opt to remain out of the same.”

In an article on Livelaw, Manu Sebastian has written that this second part of the order effectively conflicts with the Tenth Schedule, as it effectively authorises the rebel MLAs to disregard the party whip. That point aside, does the order meet the two-part test set out above – of allowing the democratic process to decide the issue, while making its subversion more difficult? On the first count, it certainly does, a point made particularly evident by the Court’s own observation that there is a trust vote scheduled for tomorrow.

On the second point, however, I would argue that the Order comes up notably short. The Court’s attempted “balance” is to give both parties freedom to act: the Speaker has the freedom to decide on the petitions, while the rebel MLAs have the freedom not to attend the proceedings of the House. However, on closer scrutiny, this balance is not a balance at all, as the second part of the order – on the issue of attending the proceedings of the House – effectively and presumptively holds the resignations to be valid until and unless the Speaker decides otherwise. This is because it is only if the resignations were valid would the party whip – and thereby the Tenth Schedule – cease to apply. In all other circumstances, the rebel MLAs defiance of the whip would be subject to disqualification under the Tenth Schedule.

The matter grows murkier when we consider the fact that the Court expressly notes in its order that its “balance” is occasioned by the fact that there is a trust vote tomorrow. This being the case, the Court’s apparent granting of freedom to the Speaker becomes effectively chimerical: because the whole point is that the ruling combine is likely to lose its majority in the circumstances that the rebel MLAs are able to defy the party whip without being disqualified – which is precisely what the Court’s order allows. In effect, therefore, the Order – while purporting to grant the Speaker unlimited time – effectively grants the Speaker time until the trust vote to decide, after which any decision the Speaker makes will, for all practical purposes, be infructuous.

As I had mentioned in my last post, the two subversions of constitutional conventions at stake here are the Speaker abusing his powers on the one hand, and large-scale horse-trading on the other. The Supreme Court’s order, unfortunately, is framed in a way that makes the former far more difficult (in a similar manner to how the Supreme Court fettered the governor’s ability to abuse his powers the last time around), but at the same time, actively allows for the facilitation of the latter, by judicially noting that the rebel MLAs “ought not” to be subjected to the party whip.

This, it should be obvious, is no balance at all.

Judicial Supremacy amid the Breakdown of Constitutional Conventions: What the Karnataka Controversy Tells Us about our Parliamentary Democracy

It has long been observed that the smooth functioning of parliamentary democracy depends upon constitutional conventions. Put simply, a constitutional convention refers to a set of uncodified norms that are sanctified by a long tradition of unbroken practice. Political functionaries tend to adhere to these norms either out of a sense of public duty, or out of fear of paying a political cost by breaking them.

A written Constitution can reduce the extent to which governance relies upon conventions. It cannot, however, eliminate them. The range of human behaviour can never completely be captured in a text. In a written Constitution with judicial review, an extra wrinkle is added to the situation: it creates situations where courts may be asked to rule upon the scope and the content of these conventions, and – in exceptional circumstances – even asked to guarantee their enforcement. This will require the Court to enter the “political thicket” (see this recent article by Mukund Unny), along with all its attendant dangers.

All this is difficult enough. In India today, however, there is an even further layer of complexity. Constitutional conventions and judicial review depend upon one basic premise: that constitutional functionaries tasked with implementing constitutional conventions act in good faith. For example, parliamentary democracy vests substantial power in the office of the Speaker of the House. The Speaker of the House conventionally comes from the ruling party, but once they occupy the Chair, they are expected to shed their partisan affiliation, and impartially administer the rules of the House (including its conventions). The presumption of the Speaker’s impartiality is the underlying basis for another very important constitutional convention: that Courts shall not be called upon to adjudicate disputes relating to what goes on in Parliament. The Parliament has its own adjudicating authority – the Speaker – and the doctrine of the separation of powers requires Courts to defer absolutely to how the Speaker manages the affairs of the House.

However, once it becomes clear – as it arguably has become in India – that Speakers repeatedly and blatantly act according to partisan motives (the conduct of the last Lok Sabha speaker in certifying money bills and refusing to hold votes of confidence is a case in point), a judicialisation of the Speaker’s conduct becomes inevitable. If opposition parties have good reason to believe that the game in the House is rigged, they have little choice but to go to Court. And the Court is then faced with an impossible situation: constitutional conventions require it to stay out of Parliament, but at the same time, staying out would result in another set of conventions being violated with impunity. There is no clean – or good – answer in such a situation.

What is happening in Karnataka represents a classic example of the breakdown of constitutional conventions, and its knock-on effect upon the judiciary. Recall that the ruling Congress-JDS combine in Karnataka has a thin majority. Recently, a number of MLAs of the ruling combine offered their resignations to the Speaker. The result of this would be to deprive the ruling combine of its majority, and offer the chance to the opposition BJP to stake a claim to form the government. The MLAs have argued that they are resigning of their own free will, while the Congress-JDS argues that they have been bribed and threatened by the BJP to do so.

At this point, Article 190 of the Constitution comes into play. Article 190 provides that MLA resignations are to be offered to the Speaker. It also allows the Speaker the discretion to reject the resignations if, in her view, they are not “voluntary or genuine.” Article 190, therefore, presumes that legislators act in good faith when resigning, and makes the Speaker the judge of that. What Article 190 does not do – indeed, what it cannot do – is to guarantee that the Speaker herself will act in good faith (that presumption is a constitutional convention).

Before the Supreme Court, the legislators have argued that the Speaker is deliberately delaying deciding on the resignation letters, and therefore violating his duty to act in good faith. They have asked the Court to direct the Speaker to decide upon the resignations in a “time bound manner” (notice that the idea of a judicial authority “directing” the Speaker of the House to do anything would be unheard of in most parliamentary democracies in the general course of things, and indeed, that is what the Speaker himself effectively said after the Supreme Court passed an interim order). The legislators have also argued that if the Speaker is acting out of partisan motives: basically, he is waiting until the ruling combine issues a three-line whip to its party members, at which point, the anti-defection Tenth Schedule will kick in. The moment the rebel legislators vote against the whip, their resignations will become infructuous, because disqualification will kick in.

As mentioned above, this puts the Court – which will hand down its order in the case tomorrow – in an impossible situation. The existence of partisan Speakers is an indisputable fact (indeed, there is already a pending petition before a Constitution Bench on the issue of Speakers deliberately sitting on disqualification decisions in order to allow ruling parties to maintain their majority). But the existence of horse-trading and defections in order to secure ministerial berths or for other similar reasons, is equally indisputable. But while both these facts are indisputable, for obvious reasons, and to avoid a complete breakdown of governance, neither of these can be acknowledged in the open, and in Court. The Court, thus, has to pretend that constitutional functionaries act in good faith, while – in specific cases – carve out remedies that are meant to operate in a world in which they do not.

What is the Court to do in a case like this? One – tempting – solution that it must avoid is full-scale intervention. That will swiftly drag the Court into the political weeds, and will make accusations of partisanship inevitable. Already, the Court has been placed in a situation where whatever it does will have the direct effect of favouring one set of political parties over the other. That is a very dangerous position for a constitutional court to find itself in.

The contours of a solution, however, might be visible from the Court’s own precedent: in particular, what it did in Karnataka last year, when the controversy was about government formation. In that case, the tables were somewhat reversed: the issue concerned the actions of the Governor in allowing the BJP to form the government despite the Congress-JDS’ claims to having a majority, and then allowing the Chief Minister fourteen days to prove his majority (it was alleged that this inordinately long time was given to enable the BJP to use its superior financial power to buy out opposition MLAs). The Court refused a full-scale intervention (i.e., setting aside the Governor’s decision), but it did reduce the time given to the Chief Minister to 48 hours, by ordering a videographed floor test. The BJP was unable to prove its majority, and ultimately, the Congress-JDS combine came to power.

The Supreme Court thus accomplished two things: first, it simply made it more difficult for the parties involved to act in bad faith, by reducing the time period to 48 hours; and secondly, its solution was not judicial (setting aside or upholding the governor’s action as valid), but parliamentary – a floor test. The blueprint, therefore, seems to be this: the task of the Court in cases like these is to fashion a remedy where the solution to the crisis is found through the existing democratic processes, but where it becomes far more difficult for constitutional functionaries to subvert the process and break conventions by acting in bad faith. In the present controversy, that might be accomplished by the following solution: the Court asks the Speaker to decide upon the resignations within a reasonable time (but enough time for the Speaker to make an enquiry as envisaged by Article 190), but makes it clear that the Speaker’s decision will be subject to judicial review under the Bommai standard (relevance/existence of material and an absence of mala fides). If it is later found that the Speaker acted wrongly, his decision on the resignations will be set aside, and – as happened in the Arunachal Pradesh case – status quo ante as of today will be restored, with the resignations now being treated as valid. In the meantime, the other democratic processes (the trust vote, the operation of the anti-defection law etc.) can go on as per their own logic.

This solution, it is submitted, would respect the constitutional authority given to the Office of the Speaker, while also subjecting him to judicial oversight in case he decides to act in bad faith. At the same time, it would allow the Speaker to form an assessment of whether the rebel legislators are acting in good faith or not, with the knowledge that his decision can – and will – be challenged. And the Court is saved from wading into murky political disputes (for now) in a way that will open it up to accusations of partisan bias.

This is, of course, an imperfect solution; and there may be other potential solutions that may strike the balance better (should the Court insist that the decision on resignations precede the trust vote/three line whip? Would that involve a direction to delay the state budget? Etc.). But I want to make one final point: the very fact that we are here today discussing the range of alternatives open to the Court demonstrates a disturbing development. The repeated bad faith actions and breaches of constitutional conventions by political functionaries have created a gaping, open space that is being filled by judicial supremacy. This has been going on for a while now: Speakers’ partisan decisions on certifying money bills has made court challenges inevitable; Governors’ partisanship and horse-trading has made judicial interventions into government formation inevitable; and so on. The beginning of all this, of course, was the repeated and unprincipled imposition of President’s Rule, which first dragged the Court into such questions.

But dragging the Court into this domain presents a deep threat to judicial independence: a Court whose decisions will regularly have such huge political ramifications presents a ripe and tempting target for capture, to unscrupulous political parties. It is for this reason that, in every case of this sort, the Court must be profoundly careful about what it is doing, and what the consequences of that are – because, ironic as it may sound, judicial supremacy in the political process is the shortest road to a compromised judiciary.

(Postscript: An additional point – and an additional way in which the Supreme Court, in particular, can avoid being tainted by a partisan brush – is the importance of sticking to procedural rules in cases like this. It is unclear how an Article 32 petition is maintainable in the present case – and even more unclear why the Supreme Court did not ask the parties to approach the Karnataka High Court as the jurisdictional forum (recall that a similar case from Tamil Nadu, involving the AIADMK was argued before the Madras High Court). This becomes particularly pertinent because the present Court has indeed sent constitutional cases back to the High Courts recently (the challenge to the Aadhaar ordinance being a good example). Ensuring that such cases come to it through proper channels will help the Supreme Court – as an institution – to avoid one particular Article 32-shaped pitfall. Of course, that issue is now infructuous, in the present case. I am grateful to Suhrith Parthasarathy for having pointed this out to me.)

Guest Post: The Absence of Deliberative Democracy – The Fetters of the Anti-Defection Law

(This is a guest post by Udit Bhatia. It is based on the article, Cracking the Whip: The Deliberative Costs of Strict Party Discipline in the journal, Critical Review of International Social and Political Philosophy).

The anti-defection law, introduced by the 52nd amendment to the Indian Constitution, prohibits legislators from voting against their party’s whip on any legislation, or voluntarily giving up membership of their party. Legislators who violate their party’s whip stand to lose their place in parliament. Much of the rationale for such law – indicated by the Parliamentary Debates that went into its framing – seems to have been the prevention of horse-trading. This was evidenced in recent fears about corruption in the aftermath of the Karnataka election. But the scope of the amendment is wider than that, since it does not just forbid legislators from voting against their party during a trust motion or no-confidence vote. Rather, it also seeks to stop them from voting against the party line on any legislative matter where a whip is issued. Following the Supreme Court’s 1996 ruling in G. Viswanathan vs The Honourable Speaker, Tamil Nadu Legislative Assembly, the law binds a member of parliament (MP) to her party’s directives even if that party has expelled her. India is among a handful of states—the others being Pakistan, Bangladesh and Fiji—that bind MPs to the will of the party leadership in this manner.

The anti-defection law, thus framed, has important negative effects on parliament’s capacity for discussion. Why, one might ask? After all, the anti-defection law merely prohibits legislators from voting against the party. But this would ignore ways in which restrictions on the vote affect legislators’ voices as well. Constraints on how legislators vote can restrict the formation of opinions contrary to their party’s line. If the only position that an MP is required to endorse is the one mandated by their party, this leaves them with little incentive to engage in the demanding task of scrutinising alternative ideas. Fetters imposed by the anti-defection law can also restrict the expression of dissent through their chilling effect on backbench MPs. If a legislator criticised her party’s stance and publicly expressed disagreement, she would ordinarily be expected to demonstrate consistency by voting against it. This is particularly true where disagreement runs deep or revolves around an issue considered central to her political project. But if one criticises the party on a wide range of issues, or vehemently so on some particular issue, electors might ask why one yet continues to remain member of that party and vote in accordance with its whips. ‘Why not just vote against your party if you disagree that much’, electors could ask. Party leaders could raise the cost of cross-voting by penalising dissenting voters. At the same time, legislators do not wish to be viewed as hypocrites. As a result, putting up a façade of consistency means that one is unlikely to express opinions contrary to that one is required to vote for. Finally, the anti-defection law also restricts the uptake that legislative dissent is likely to receive. In well-functioning parliaments, even when backbenchers ultimately vote with their party, they can exercise considerable influence behind the scenes. Party leaders have to anticipate their backbench’s reaction, and put effort into allaying their concerns so as to secure their consent. In India, on the other hand, the anti-defection law takes away any such incentive for party bosses to do so, by enabling them to rely on whips rather than discussion with their backbench colleagues.

Such constraints on legislation have important implications for legislative discussion. In the first instance, they undermine the benefits of the legislature’s numerical size. There is a reason why parliaments, compared to judiciaries, are large-sized bodies. With their total strength of 545 and 250 members, the Lok Sabha and the Rajya Sabha are considerably larger institutions than the highest court in the country. This is because legislatures are supposed to pool information widely dispersed across a society. Their role is to ensure that varied interests in a large, heterogeneous country like India’s find representation through an appropriately sized body of persons. Following recent innovation in the social sciences and political theory, we conceive the primary strength of parliaments as their ability to manifest the wisdom of the multitude. For instance, following Scott Page and Lu Hong’s experimental work, Helene Landemore’s book, Democratic Reason draws upon their ‘diversity trumps ability theorem’, and argues that inclusive groups of diverse decision-makers outperform small groups of expert ones. However, the anti-defection law undermines the numerical benefit of the legislative chamber, restricting effective decision-making to a small, relatively homogenous group of party leaders.

Further, the anti-defection law can also deprive us of the benefits of having two chambers. There is a reason why the framers of the constitution insisted on establishing two legislative houses: the Lok Sabha and the Rajya Sabha. One of the primary thrusts behind this design was that law-making bodies, like any other collective, are fallible. They can make errors, pass laws in haste, or fail to adequately consider some relevant interests. Bicameralism offers the opportunity to obtain a second opinion on legislative affairs. At the moment, the ruling party lacks a majority in the Rajya Sabha. But what if this were not the case? In that case, the government, with its majority in parliament, could enact laws in the Lok Sabha, and issue a whip to see those measures through in the Rajya Sabha as well. In such cases, second chambers cannot really offer a second opinion on decisions adopted by the first chamber. For second chambers to serve this purpose, they must be constituted of a distinctive set of persons. To see why, imagine that all members of the ruling party in the second house were chosen from that party’s MPs in the first house. In this hypothetical case, some legislators would serve a dual role as members of the first and the second chamber of Parliament. This situation would be deemed unacceptable as a way of offering a second opinion rather than allowing members of the first chamber to re-affirm their own decision. After all, MPs already do so in the course of multiple readings of a bill in the first chamber. Deliberative autonomy and the ability to form independent judgments is crucial for the distinctiveness of legislators. But distinctiveness cannot be simply about the physical presence of two different sets of legislators. If the only permissible view they can voice is the one sanctioned by the party’s leadership, and if they lack the capacity to form opinions that differ from that view, then distinctiveness no longer obtains. We, then, lose, an important benefit that bicameral division of the legislature offers.

Finally, there is a third negative cost that the anti-defection law imposes on legislative discussion. Part of the role of a well-functioning legislature is to clarify for the wider public the different shades of opinion that exist on any topic. Parliamentary discussion should be oriented at demonstrating the underlying dimensions on which various political disagreements exist, locating different political parties on the space of political reasons. By stifling the expression of dissenting views, the anti-defection law undermines parliament’s ability to offer this map of opinion. In doing so, it can also foster sharp polarisation, because we fail to see how internally variegated parties are, and how there are political actors across the aisle who might actually agree with us. Consider, for example, a debate on a healthcare policy. It is valuable for us to see how, despite the opposition party leaders’ criticism of that policy, there are members within that very party who agree with the government.

In India, the deliberative costs of the anti-defection law are exacerbated by its lack of adequate intra-party discussion. Were parties to offer vibrant internal mechanisms for deliberation, we might think that the constraints they impose within are parliament are less troublesome. However, the dictatorial state of our parties precludes such a stance. So far, much of the commentary on the Indian parliament has focused on the lack of adequate discussion caused by frequent disruptions. This is understandable. The scale of disruptions and the rowdy scenes we have come to witness in recent years are worrying symptoms for the health of our democracy. But we should not assume that all our legislators need to do is behave themselves and conduct themselves in an orderly fashion. As I have tried to show, the problem lies much deeper. What they can and are likely to say is conditioned by the anti-defection law. Even if disruptions stopped, and legislators could speak more freely, parliamentary discussion is likely to fall short of deliberative ideals.

Anti-Defection and Expelled Members: Text, Structure, History, and Theory

Yesterday, a two-judge bench of the Supreme Court issued notice and referred to a larger bench the case of Amar Singh vs Union of India. Amar Singh was expelled from the Samajwadi Party in December, and was designated an “unattached member” in the Rajya Sabha. In his petition, he argued that, since he had been expelled, he should now be entitled to vote against the Party whip, or join another political party, without being disqualified from parliamentary membership under the Tenth Schedule of the Constitution. Standing in his way was a judgment of a two-judge bench of the Supreme Court in 1996, G. Vishwanathan vs The Speaker, in which the Court had held that the Tenth Schedule was applicable to expelled members. With the case now being referred to three judges, the Court can consider the issue afresh.

Interestingly, this is not the first time the Court was asked to reconsider the correctness of G. Vishwanathan. Amar Singh had been previously expelled from the Samajwadi Party in 2010. He, along with two other expelled members (of other parties), filed an identical petition asking that the 10th Schedule not be applied to them. The Supreme Court referred the case to a three-judge bench, but that bench only heard the case in 2016. At that time, all the three petitioners’ terms in parliament was over; therefore, despite hearing the matter at length, the Supreme Court finally declined to answer the referral question about the correctness of Vishwanathan. But now, with Amar Singh expelled all over again, the case is back in Court for a second round.

What makes Amar Singh vs Union of India so interesting is that it lies at the intersection of constitutional text, the scheme of the Tenth Schedule, its framing history, and democratic theory. Broadly, the Tenth Schedule provides for disqualification from Parliament in case of “defection”. Section 2 of the Schedule provides for two circumstances: where a person has “voluntarily” given up the membership of his party; and where a person votes against her party whip. The Explanation to the Section states that “an elected member of a House shall be deemed to belong to the political party, if any, by which he was set up as a candidate for election as such member.”

Vishwanathan and the Text of the Tenth Schedule

In Vishwanathan, the Supreme Court relied upon the text of Section 2 to hold that the only circumstance contemplated by the Schedule, in which a member no longer belonged to her political party, was if she “voluntarily” gives up his membership. The Court also relied upon the Explanation, which stated that a member was deemed to belong to the party who set her up as a candidate for election. For this reason, the Court held that:

“Even if such a member is thrown out or expelled from the party, for the purposes of the Tenth Schedule he will not cease to be a member of the political party that had set him up as a candidate for the election. He will continue to belong to that political party even if he is treated as `unattached’. The further question is when does a person `voluntarily give up’ his membership of such political party, as provided in paragraph 2(1)(a)? The act of voluntarily giving up the membership of the political party may be either express or implied. When a person who has been thrown out or expelled from the party which set him up as a candidate and got elected, joins another (new) party, it will certainly amount to his voluntarily giving up the membership of the political party which had set him up as a candidate for election as such member.”

In other words, despite his expulsion, for the purposes of the Tenth Schedule, an MP remained a member of the political party that had expelled her, until she joined another party – at which point, she was deemed to have “voluntarily” given up her membership, and subject to disqualification.

Of course, this pure textual reading – that leads to a seemingly counter-intuitive result – is not the only way of reading the Tenth Schedule. In Vishwanathan, the Court based its entire interpretation upon the deeming provision in the Explanation (extracted above); however, it is well-settled that an Explanation to a Section can only clarify its scope; it cannot expand it. In this context, if we start our reading with the main Section (instead of starting with the Explanation, as the Court did), we arrive at the following reading: the word “voluntary”, in its ordinary, common-sense meaning, obviously cannot apply to a situation of expulsion. An expelled member has been compelled to leave her party. Section 2(1)(a) of the Schedule, therefore, does not apply to her. Nor does an expelled member “belong” to the party that expelled her any more. Consequently, 2(1)(b) of the Schedule does not apply either. Since Sections 2(1)(a) and 2(1)(b) exhaust the circumstances in which an elected member can be disqualified for defection, the Tenth Schedule does not apply to expelled members.

Parliamentary History

If both these readings are equally convincing, which of them ought to prevail? It is here that Parliamentary history comes in. The original draft of the Tenth Schedule had an additional “Section 2(1)(c)”, which specifically made the Schedule applicable to expelled members. During the course of the debates, however, this additional section was dropped from the final version of the Schedule. And indeed, a perusal of the 1985 Parliamentary debates reveals that the Lok Sabha members were explicit about the reasons why it was dropped: multiple members stood up to draw a distinction between what constituted defection, and what constituted dissent, and argued that the elimination of Section 2(1)(c) was important to preserve intra-party dissent. There was no contrary voice. Consequently, it seems clear that, at the very least, Parliament did not intend the Tenth Schedule to apply to expelled members. Of course, to what extent the Court can take this into account – and to what extent the Parliamentary Debates can modify the Court’s interpretation of the Tenth Schedule – is another question.

The Basis of Anti-Defection 

Consider also the question of democratic theory. Anti-defection is premised on the theory that the basic unit of parliamentary democracy is the political party, and that therefore, a legislator’s responsibility to her party is more important than her responsibility to her constituents. This arguments comes through clearly in Kihoto Hollohan vs Zachillhu (which G. Vishwanathan’s Case relies upon). In that case, the constitutionality of the 10th Schedule was challenged. The petitioners argued that:

“… the very concept of disqualification for defection is violative of the fundamental values and principles underlying Parliamentary democracy and violates an elected representative’s freedom of speech, right to dissent and freedom of conscience and is, therefore, unconstitutional as destructive of a basic feature of the Indian Constitution.” (Emphasis supplied)

The challengers advanced a theory of representative democracy where the primary responsibility of a Member of Parliament was towards the constituents that had elected her. To substantiate this theory, they relied upon Edmund Burke’s 1774 speech to the Electors of Bristol and the observations of Lord Shaw sitting in the British Appeals Court in 1910, both of whom had stressed that under the British “Parliamentary Constitution” and system of “representative Government”, an individual legislator’s conscience ought to be the supreme. The Tenth Schedule, that subordinated the legislator’s conscience (presumably guided by the demands of her constituency) to that of party ideology, on the pain of disqualification, was therefore contrary to the basic feature of “democracy”.

The Supreme Court rejected this argument. Its reasoning proceeded as follows: first, applying principles of judicial deference, it accepted the legislative judgment that “unethical political defections” constituted “a canker eating into the vitals of democracy”. It then acknowledged the importance of debate and discussion within the Parliament, but also went on to observe:

“… a political party functions on the strength of shared beliefs. Its own political stability and social utility depends on such shared beliefs and concerted action of its Members in furtherance of those commonly held principles. Any freedom of its members to vote as they please independently of the political party’s declared policies will not only embarrass its public image and popularity but also undermine public confidence in it which, in the ultimate analysis, is its source of sustenance – nay, indeed, its very survival. Intra-party debates are of course a different thing. But a public image of disparate stands by Members of the same political party is not looked upon, in political tradition, as a desirable state of things.”

Yet what of the legislator’s responsibility to her constituents, and the possibility that the exercise of this responsibility might come into conflict with the party line – and, correlatively, the requirement of dissent and democracy within political parties? The Court recognised this problem, responded by extracting a lengthy passage from Rodney Brazier’s Constitutional Reform and, in particular, seized upon a passage from the book that advocated a right to recall in case a legislator crossed the floor of the House. It finished by holding that compulsory disqualification was nothing more than a “statutory variant of… [the] justification underlying the power of recall.

Thus, in Kihoto Hollohan, the Supreme Court explicitly endorsed a model of parliamentary democracy which privileged the party line – and therefore, by extension, the vehicle of the political party – over the MP’s sense of responsibility to her constituents, and her right to dissent within the party. However, it is important to note that this does not negate the MP’s responsibility to her constituents or her right to dissent. In fact, in other judgments, the Supreme Court has indicated that the relationship between an elector and a candidate is normatively at least as important as that of the role of the political party within the Indian Constitutional scheme. A good example of this is a set of cases brought to Court in the last decade, which sought to make disclosure of personal details about political candidates compulsory. In Union of India vs Association for Democratic Reforms, for instance, the Supreme Court noted that:

“For health of democracy and fair election, whether the disclosure of assets by a candidate, his/her qualification and particulars regarding involvement in criminal cases are necessary for informing voters, may be illiterate, so that they can decide intelligently, whom to vote? In our opinion, the decision of even illiterate voter, if properly educated and informed about the contesting candidate, would be based on his own relevant criteria of selecting a candidate. In democracy, periodical elections are conducted for having efficient governance for the country and for the benefit of citizens – voters. In a democratic form of government, voters are of utmost importance. They have right to elect or re-elect on the basis of the antecedents and past performance of the candidate. He has choice of deciding whether holding of educational qualification or holding of property is relevant for electing or re-electing a person to be his representative. Voter has to decide whether he should cast vote in favour of a candidate who is involved in criminal case. For maintaining purity of elections and healthy democracy, voters are required to be educated and well informed about the contesting candidates. Such information would include assets held by the candidate, his qualification including educational qualification and antecedents of his life including whether he was involved in a criminal case and if the case is decided – its result, if pending – whether charge is framed or cognizance is taken by the Court?”

In this case, therefore, the Court placed a substantial amount of importance upon the relationship between the citizen-elector and the candidate qua candidate. This suggests that the Indian Constitutional scheme contains elements of both the legislator-centric and the party-centric theories of representative democracy.

Now, if this is the case – i.e., that Indian parliamentary democracy contains elements both of the part-centric model as well as the legislator-constituent-centric model – then the Xth Schedule ought not to be interpreted in a way that entirely privileges the former over the latter. In other words, given that the Xth Schedule prescribes the extremely draconian punishment of disqualification for “defection” – i.e., it subordinates the interests of the party over the relationship between the legislator and her constituents – when interpreting the Xth Schedule, the Court should do so strictly, so as to not entirely efface the latter. In other words, if there are two possible interpretations open to the Court – one that expands the scope of the Xth Schedule, and one that does not – all other things being equal, the Court should choose the former.

The Finality of the Speaker

Lastly, consider this. The Tenth Schedule makes the speaker’s decision on the issue of whether defection attracts disqualification final. The rationale for this is that – following common law traditions – those acts that have purely to do with the business of the House are to be regulated by the Speaker. Now, in the Parliamentary Debates, one of the reasons for removing Section 2(c) from the Draft Xth Schedule Bill was that expulsion of members is something that takes place “outside the House”. Consequently, an interpretation of the Tenth Schedule that would bring expulsion within its meaning would effectively mean extending the jurisdiction of the Speaker to activities outside the legislature. Could such an interpretation stand?


Therefore, when the three-judge bench assembles to hear Amar Singh vs Union of India, it will have to contend with a range of arguments centred around the text of the Tenth Schedule, its scheme, the Parliamentary history, and competing theories of democratic legitimacy in a parliamentary system. The outcome should be fascinating.

(Disclaimer: The author was one of the lawyers working on the brief on the behalf of the Petitioner, Amar Singh.)



The Uttarakhand Incident: Thinking through the Anti-Defection Law

(In this Guest Post, Udit Bhatia, a doctoral candidate at the University of Oxford, examines the question of anti-defection laws, in light of the recent political and legal tussle in Uttarakhand)

The anti-defection law, introduced by the 52nd amendment to the Indian Constitution, prohibits legislators from voting against their party’s whip on any legislation, or voluntarily giving up membership of their party. Legislators who act otherwise are expelled from the legislative body. This rule was formulated in response to what was perceived as large-scale dubious floor-crossing by legislators in response to monetary incentives to bring down governments. What the amendment forbade, however, was not just the practice of legislators voting against their party in a trust motion, or vote of no-confidence, but from voting against it on any legislative matter where the party chooses to issue a whip. The anti-defection law was at the forefront of a recent Supreme Court ruling over President’s Rule in the state of Uttarakhand. While much of that debate focused on the centre’s seemingly unjustified intervention in the state’s legislative affairs and its trust vote, it seems to have been yet another missed opportunity to deliberate upon the justifiability of an anti-defection law in the first place.

Before I examine that debate further, I wish to emphasize that such strict policing of party boundaries might be somewhat less problematic if parties themselves were internally democratic units. As with legislation, then, the coercive nature of party positions on those legislations could derive legitimacy through their being voted upon after a discursive process open to all members. Political parties in India, however, remain far from this model of intra-party democracy. The selection of candidates for elections remains a highly centralised process for a large majority of the national parties. Additionally, the executive committees themselves are products of insufficient, if not non-existent, procedures in most cases. Although there exists little research on this, the opaqueness with which parties respond—when they do finally respond—to the Election Commission’s requests for information about internal elections is telling. Their reports say little about the method through which the executive committees were chosen. It remains unknown whether such selection takes place through unanimous nomination, or election; and if the latter, how many votes were polled in favour of those selected.

The debate over defection can be located in the wider one over the relationship between elected representatives and their constituents. Political theorists have long debated whether representatives are best understood as trustees or delegates. The trusteeship model sees representatives as free to engage in political actions with adequate leeway on behalf of their constituents provided that they shall be held accountable ex post through elections. On the delegate model, representatives are elected to try and execute a mandate given to them by their constituents. They are not free to engage in whatever political actions they see fit, but must advocate and attempt to secure their constituents’ wishes. There might of course, be a clash with other constituents’ wishes, or turn out to be contextual limitations to how far representatives can do this. However, they are required to adhere as closely as possible to what their constituents elected them to do.

One might defend the anti-defection law as a component of the delegate model. On this explanation, representatives are elected by constituents to demonstrate adherence to political parties through which they have been chosen. The mandate of the representative lies in ensuring that the party’s platform, as determined by the organisation’s leadership is fulfilled as far as possible. On the other hand, we may even defend the anti-defection law as a component of the trusteeship model. On this reading, the agent to which leeway extends in fulfilling constituents’ wishes is the party and not the representative herself. Thus, we ought to give adequate room to parties to operate in whichever way they see fit in order to execute their platform insofar as they are accountable later through elections. The party’s control over the representative is a part of the leeway afforded to it.

I suggest that such defences of the anti-defection law would be misguided for at least two reasons. Firstly, any appeal to either a pure trusteeship or delegate model appears misguided. As constituents, we often work with both models in choosing our representatives. Thus, I might choose the Aam Aadmi Party, viewing my vote as having delegated it to enact a Jan Lokpal law. On the other hand, I may remain indifferent or unsure of my stand on its policies with respect to monitoring of public schools. I find it sufficient that I have a vote that I can exercise later to reward or punish what it decides to do, but do not hold a fixed view on the matter for myself. Further, it might also be that I amend my delegate-view as a result of the Aam Aadmi’s political actions. The party may decide to not execute the Jan Lokpal law, and explain that there are good reasons for not doing this. I may change my mind as a result of their explanation even though I originally viewed my vote as having delegated the party to enact that law. Therefore, the appeal to pure delegate or trustee models to defend the anti-defection law are mistaken because (a) it is unclear that constituents subscribe exclusively to one view (b) whatever stance constituents take on their vote is itself subject to change as a result of the political process.

Secondly, the emphasis on the party’s wishes to the complete exclusion of the choices of individual legislators does not square with elements of the democratic process. We do, after all, elect individual legislators belonging to certain parties, rather than voting for a party and allowing it to choose parliamentarians in its own wisdom depending on its vote-share. We also care about whom we vote for, as is evident in processes requiring scrutiny of individual candidates, their antecedents, and their wealth. It seems strange that we would invest as much in the individual characteristics of representatives when, at the end of the day, they are bound by the wishes of their party’s leadership.

One might now suggest that my argument does not provide compelling evidence for favouring complete autonomy for legislators. It might be true that we do not adopt pure trustee or delegate models in electing representatives. But it might be that on this particular choice—adherence to the party’s platform—we did, indeed, view our vote as an act of delegation. It might be that we do care about individual features of representatives, but we also elect them qua members of a party. My argument so far only seems to have led to further doubt in relation to the relationship between parties, representatives and their constituents. I will now argue that there is one way of resolving the debate: through a democratic process rather than a philosophical one.

It seems that the reason why debates over trustee versus delegate models are inconclusive is because the matter is not one that should be entrusted to a purely philosophical debate. Rather, it is precisely the kind of matter that democratic processes are intended to resolve. Whether legislators should lose their seat for voting against party lines or not seems precisely the kind of matter that voters ought to be entitled to determine since their can be reasonable disagreement on this issue. So far, it seems that my argument leads to the conclusion implied by the anti-defection law: if legislators defect from their party, they ought to face re-election. However, this does not follow from the principle sketched above. Contrary to the the Supreme Court’s assertion in Kihoto Hollohan, that “What might justify a provision for recall would justify a provision for disqualification for defection”, there is a distinction between the two.

Giving persons the ability to determine whether or not I have acted in accordance with their wishes does not require asking them this on every occasion. After all, it might be that my wish is precisely that I do not have to offer my judgment on each particular occasion. Rather, one need only provide persons the choice of offering their judgment. Thus, there is a difference between constituents having to give their judgment on whether their defected MP still represents them, and between constituents having the choice to give their judgment. The former results in an election being triggered as soon as the MP defects from the party. The latter allows constituents to trigger an election if they believe they are no longer represented as a result of the MP’s defection. The former is more consistent with democratic principles since it does not take a stand on the delegate or trustee model, instead leaving matters to the electorate. Unlike the pure trustee model, it does not leave constituents merely with the option of offering their judgment at the next electoral cycle. Rather, it affirms that a defection might be a worthy ground for loss of the MP’s ability to represent constituents at a particular moment. On the other hand, unlike the delegate model, it does not suggest that defection necessarily disqualifies MPs from being seen as representative of their constituents’ wishes. Instead, it affords them leeway to claim that their defection was an extension of their role as their constituents’ representatives—as long as their constituents do not repudiate this claim and ask to hold them to account.

As a result, one way out of the debate on defections seems to be awarding citizens a right to recall representatives where defection counts as one of the grounds for triggering recall. Does this move not threaten the political stability and social utility of political parties as the Court argued in Kihoto Hollohan? Firstly, such fears seem unfounded once one surveys the several democratic systems where parties continue to thrive in spite of defections. Secondly, such an argument ignores the considerable power parties anyway enjoy over their legislators, for instance, through the possibility of denying ministerial berths of future tickets for elections. Finally, my proposal sketched here also ensures that defection comes with an additional cost: the fear of facing a recall election. This makes it likelier that legislators would expose themselves to such risk only when such defection is perceived as consistent with their constituents’ demands rather than on extraneous grounds.

Specifying the proportion of constituents needed for triggering recall elections, and other legitimate grounds for  doing so is beyond the scope of this piece. Further, I will also refrain from assessing whether or not defection in a trust vote ought to be subject to a possible recall or the more drastic measure of by-election. This is because one might suggest that there are good practical reasons for disqualifying defectors in trust votes since these have more severe consequences for the stability of government, and it isn’t possible to curb corrupt practices. One might also argue that even though legislators can reasonably claim disagreement upon particular moves of the party, wholesale rejection of party membership ought to be subject to a greater threshold of scrutiny. However, even in this more limited form, the option of recall elections on grounds of defection would be more consistent with democratic principles than automatic disqualification for such a move.