Madhav Khosla and Milan Vaishnav
 
The recovery of representative institutions—and the promise of legislation—is a central theme in studies on constitutional democracy. Even though much writing underscores the shift in power away from the legislature towards courts and the executive, relatively less attention has been paid to the internal workings and practices of the legislative branch. In a new article, we focus on a somewhat remarkable feature of several parliamentary democracies—namely, legal prohibitions (“anti-defection laws”) that curb, or ban, the practice of floor crossing by a legislator during their term in office, and that, in their most extreme form, prevent legislators from voting as they wish. The countries that have experimented with such laws are diverse, including India, Israel, Pakistan, South Africa, and New Zealand.
 
One way to think about anti-defection laws that limit the independence of legislators is through the lens of political fragmentation. There is widespread appreciation of the risks that political fragmentation poses to democracy, not least by disabling party leaders from enforcing discipline. This burgeoning literature is clear that political fragmentation threatens democratic government. In aggregate, fragmentation is thought to hamstring party leaders, who cannot enforce party discipline or control factions. Anti-defection laws promise to mitigate fragmentation by empowering party elites.
 
In parliamentary democracies, floor crossing can have devastating consequences. With no separation between the legislature and the executive as in presidential systems, party switching can bring down governments, often forcing fresh elections. But, in attempting to ensure government stability, anti-defection laws simultaneously subvert the character of the legislature. In the case of India, which has among the most extreme version of an anti-defection law, a legislator must follow the party dicta while voting on a bill or else suffer disqualification, from their party as well as the legislature. As such, their yes-or-no vote is a foregone conclusion based upon the decision of the party leaders, and various features that we associate with the act of legislation and the representative chamber—not least the value of deliberation and debate—are all irrelevant to the final vote that is cast. Notice the distance between this reality and Waldron’s foundational observation that legislators “do not just assemble and vote,” but rather “deliberate.”
 
The Indian law arose out of the very real problem of horse trading and consequent government instability. As such, it emerged out of a genuine crisis of representation. The law promised to bind a legislator to their party. By allowing governments to function, it would make democracy possible. But it has created its own crisis of representation. Now, even though an individual legislator has a role to play, they have no real role in their individual capacity as a legislator. They matter as an aggregate group controlled by the political party, which becomes the ultimate repository of power. If a defining feature of parliamentary government is the accountability of the executive to the legislature, we now witness an inversion: in many respects, the executive (which typically overlaps with the party leadership) controls the legislature.
 
In addition to representation and accountability related concerns, the practical challenges of implementing anti-defection laws offer further reasons against them. In the Indian case, three separate workarounds have arisen. In the first, the speaker of the legislature acts strategically upon disqualification petitions, often leaving them undecided to manipulate voting outcomes. In the second, a lawmaker resigns before violating the whip, thereby avoiding disqualification and bringing down the numbers needed for a majority vote. The lawmaker then contests the consequent by-election with the support of the party to which they effectively defected. The third workaround is the most creative. Here, a faction splits from a party and claims to represent the genuine authority of their political party rather than a group of dissidents, raising existential questions about the identity and ownership of parties.
 
Notably, this is not an India-specific narrative. The experience in Israel and South Africa similarly raise concerns arising from political practice. In Israel, the anti-defection law has been repeatedly amended to suit a specific political need. For example, a 1994 amendment stipulating that a Member of the Knesset who splits from their party to create a new party is exempt from the ban on cabinet appointment allowed the Rabin government to secure the defection of a three-member breakaway faction from an opposition party. To offer another instance, a 2004 amendment on defections and mergers was driven by the Sharon government’s need to secure one member of a three Member of the Knesset breakaway faction that differed with the faction’s decision to merge with the Labor Party.
 
South Africa offers a similar picture of self-dealing. Here, the ban on defections was first set aside for political reasons and subsequently introduced to satisfy new needs. The 1996 Constitution initially prohibited defections both for reasons of government stability and because of fears that the “likely parliamentary majority of the African National Congress could be used to woo minority legislators and over-concentrate political power.” Subsequently, the ANC sought to break an emerging opposition threat that combined two parties, the Democratic Party and the New National Party. A law was enacted to authorize defections, and the minimum required threshold for defection meant that large parties like the ANC could easily absorb members from smaller parties and thereby cement their hegemony. Eventually, as the dominance of the ANC was secured and anger over defections grew, the ban was reintroduced.
 
The problem with regulating defections is simple. A legislature typically has considerable control over its internal processes, and so floor crossing provisions are interpreted, amended, and enforced by the very people they seek to regulate. But perhaps the problem is the scope of the law. Might we resolve matters with a tailored anti-defection law? India, Israel, South Africa—and other countries like New Zealand—each vary on the extent to which they permit legislators to vote independently. But imagine a case where an anti-defection law applies only to the most existential of votes, such as no-confidence motions. This could potentially be a happy middle ground. It recognizes the value of legislator autonomy, thereby constraining the use of the law, while appreciating the importance of party discipline.
 
Pakistan offers such an example. But it shows us that concerns arise even with such a limited law. First, a vote of no-confidence is the principal tool of accountability legislatures possess; if lawmakers forfeit agency in this domain, their voice is only relevant for subordinate matters. Second, as the 2022 no-confidence motion against then-Prime Minister Imran Khan shows, a narrow law presents a dilemma. In 2022, the question arose whether, in a trust vote, the votes of the defecting members should be counted in the final tally. The plain text of Article 63A of the Pakistan Constitution suggests that they should. After all, that is the point of defecting—and it relates to the basic right of a legislator to cast their vote. But the Pakistan Supreme Court argued otherwise. It claimed that the only way to end defections once and for all was to disallow the votes even of those who are willing to suffer disqualification.
 
The dilemma is plain. If the votes of defectors are counted, defections persist, and the goal of government stability is not realized. A legislator may still choose to defect even if they are disqualified because of other incentives, as the practice in countries like India and Pakistan amply confirms. However, if the votes are not counted, legislators cannot hold the executive accountability in their capacity as legislators. In most cases, no-confidence motions could not arise, and the primary tool of executive accountability is removed. Where there is no intra-party democracy, the only real option for a dissident would be to resign and to establish a new party.
 
How then shall we deal with defections? One tempting answer, taking parties as the new legislatures, is to regulate parties more effectively. There are sound reasons for stronger rules relating to financial disclosures and the like, but there are reasons to be cautious about regulating the internal governance of parties. As Samuel Issacharoff notes, “requirement[s] that parties have open and democratic internal structures would put at risk ideological and religious parties that may be organized around certain fixed principles not amenable to internal majoritarian override.”
 
The answer instead may lie by turning to Edmund Burke. Although Burke is, in this context, best known for his Bristol address on the importance of legislator autonomy, he theorized a great deal about defections given the cabinet instability during King George III’s reign. Burke’s answer to corruption and patronage—to the endless upturning of governments during the 1760s—was the entrenchment of party ideology. A politics of strong party commitments could limit floor crossing because the ideological transition would increase the democratic costs of crossing over.
 
This insight encourages us to revisit the relationship between political polarization and political fragmentation. As Richard Pildes shows, polarization can greatly harm the functioning of legislatures, and—in the United States especially—polarization can obscure the problem of fragmentation. In the American context, one can see the reasons for addressing fragmentation to improve government performance. But, in cases like India, with non-programmatic multiparty systems, greater political polarization may well result in less political fragmentation. The world’s largest democracy has witnessed an increase in political options without deep changes in substantive ideological alternatives. The regular defections from the principal opposition party, the Indian National Congress, to the ruling Bharatiya Janata Party is a vivid case in point. In a country like India India, the similarity between parties, rather than their differences, explains the prevalence of defections—and the harmony between limited polarization and considerable fragmentation.