Introduction
Disagreement is common within political parties. Party members routinely ‘conflict over the direction the party or its candidates will take in elections and in the course of governing’ – ‘or the proper stance in opposition’ (Rosenblum, Reference Rosenblum2008: 361). Even an appeal to the ‘foundational’ values of the party may prove insufficient to resolve such conflicts, for partisans may continue to disagree about the nature of such values and what they demand in practice. Intra-party conflict over these and other issues can, and often does, become heated. It can lead some members of the party to emphatically dissent from the party’s official stance, criticising the leadership for its decisions, perhaps even raising doubts about the leaders’ overall credibility. And sometimes those dissenters face severe sanctions. Indeed, the expulsion of party members for expressing views that party leaders deem objectionable is far from uncommon in democratic systems around the world.
Consider the case of Suheyl Batum, who was a member of Turkey’s leading opposition party, the Republican People’s Party (CHP). Batum belonged to a group of backbenchers, who were critical of their party’s attempt to attract conservative religious voters, instead espousing a return to Turkey’s secular and nationalist outlook. In 2014, he criticised the party leader, Kemal Kilicdaroglu, asking him to step down in the aftermath of the latter’s poor performance in the national presidential elections. He was eventually expelled by the CHP for ‘acting in a manner damaging to the party’s integrity’ (Anadolou Agency, 2014). Or consider Gufran-e-Azam, who was a member of the Congress party, the leading opposition party in India. In 2014, he criticised the party president, Sonia Gandhi, and her son the party’s Vice President, Rahul Gandhi, for their poor performance in the recently held general elections. The Congress had dropped to its poorest ever tally of seats in parliament in the face of a convincing victory by Narendra Modi. Azam was expelled for ‘indulging in anti-party activities and making unnecessary statements against the party leadership’ (Economic Times, 2014). Or consider that in 2019, a former advisor to Prime Minister Tony Blair was expelled by the Labour party after he publicly declared on Twitter that he had cast his vote for the Liberal Democrats in elections to the European Parliament. Justifying its decision, Labour stated that ‘publicly declaring or encouraging support for another candidate or party is against the rules and is incompatible with party membership’ (BBC, 2019).
To cite an even more recent example, in October 2020, Jeremy Corbyn was suspended from the UK Labour party, which he had led until only 6 months ago. The suspension was a result of his comments on a report on antisemitism in the party under his leadership, which Corbyn insisted, had been dramatically overstated for political reasons. The suspension was lifted after 3 weeks, following an internal disciplinary process, although Corbyn continued to find himself excluded from the party’s legislative caucus. In the period between his suspension and reinstatement, the party was threatened by what commentators described as a potential ‘civil war’, with Corbyn raising funds for a legal challenge, and some of his followers resigning from Labour in protest. In the end, the episode marked, according to some commentators, the seemingly definitive end of one of the more spectacular recent attempts to change the Labour party and its ideology from within. In the context of Corbyn’s suspension, the
Corbynistas have been cleared from the shadow cabinet, and now loiter on the backbenches, unhappy at the direction of travel. Veterans of Labour’s last spell in office, a decade ago, would like a fight: better a cleansing clear-out than squalid unity. Nor is it apparent that the membership who once chanted Mr Corbyn’s name will put up much of a fight (The Economist, 2020).
While the consequences of this particular incident for the Labour party are certainly debatable, there are cases where similar circumstances have had clear ramifications for a political party and the political system at large. Take, for example, the 1987 expulsion of Vishwanath Pratap Singh from the Indian National Congress by then Prime Minister, Rajiv Gandhi. Gandhi justified the move by citing an interview that Singh – a major rival within the party – had given to a newspaper, which included adverse comments about corruption in his government (LA Times, Reference Times1987). Singh would later form a new party and return as Prime Minister 2 years after his expulsion from the Congress party. Although Singh’s government was short-lived, the party he founded after his expulsion went on to become a major force in national and regional politics in the years to come.
Whatever the particular consequences of suspending or expelling party members for their speech acts, a more fundamental question concerns the permissibility of limiting party members’ freedom of speech. To be sure, party constitutions generally prescribe procedures that must be followed in disciplining members in disputes like the kind this article addresses. But the principles that ought to regulate such procedures remain in need of clarification.Footnote 1 In other words, the problem of free speech in the party is not just one of a fair procedure for arbitrating disputes. Rather, it consists of identifying the normative concerns that should affect how parties respond to disputes generated by their members’ exercise of speech. So, should party members really have to mind their speech, or failing that, face disciplinary action from their organisation?
Arguably, many will find the sanctioning of party members for, say, openly criticising the party leadership democratically suspect, in part because parties are often evaluated according to the same democratic standards that govern society at large, where free speech typically enjoys constitutional protection as valuable political liberty. But there are also serious objections to free speech within the party, to do with the party’s nature as a competitive association whose ability to contest elections may be undermined by internal dissent as well as party members’ freedom to exit the party (and join another one) in response to internal speech regimes. Taking these considerations seriously, we want in this article to systematically discuss the problem of freedom of speech within the party, clarifying why and when it is problematic to limit party members’ free speech. We develop our argument on the basis of a normative understanding of parties that sees them as broader membership organisations (rather than teams of political elites) that allow citizens to exercise political agency in ways that no other form of political organisation can, as well as insights from the theoretical literature on organisations (esp. Anderson, Reference Anderson2017; Herzog, Reference Herzog2018).
Our main focus in the article is on limitations on freedom of speech (because it is a central political liberty without which the exercise of political agency would be severely impaired) and expulsion from the party (because it is arguably the harshest form of disciplinary action available), but our argument has the potential to travel further inasmuch as we mount a broader defence of party members’ moral agency according to which there is a presumptive case that any interference with their political freedoms is problematic. We will also bracket the issue of when criticism expressed by party members may be morally objectionable because of its particular substantive content (e.g. because it qualifies as ‘hate speech’). This is not because the issue is unimportant, but because arguments for why certain forms of speech are morally impermissible are independent of the arguments that we advance. Moreover, while we pay heed to the fact that party members may be committed to the party in very different ways, we set aside the role obligations that holders of specific posts (e.g. a party spokesperson) might bear. Suffice to say that our argument is compatible with the notion that, even if party members should normally enjoy the freedom of speech, certain role bearers should be subject to greater constraints.
Our aim in this article is to highlight the dimensions relevant in evaluating how claims for freedom of speech within the party should be treated. We do not attempt to offer a bright line or threshold that delineates the precise scope of permissible speech within parties. Rather, we seek to shed light, first, on the importance of freedom of speech within the party, and second, the kind of contextual factors that bear consideration in assessing intra-party disputes. In doing so, we aim to contribute both to a wider normative debate about parties and partisanship (e.g. White and Ypi, Reference White and Ypi2016; Bonotti, Reference Bonotti2017; Wolkenstein, Reference Wolkenstein2019), and a more specific debate on the regulation of parties’ internal affairs.
The latter debate is frequently tied to the relationship between political parties and the state. For the most part, common law has treated parties as purely private institutions, outside the domain of public law (see Orr, Reference Orr2014 for an overview of this position; Bhatia, Reference Bhatiaforthcoming). On this account, parties are ‘assemblages of persons gathered together to advance their political interests […] governed by the law of contract’ (Morris, Reference Morris2012: 107), which in turn places sharp limits on the degree to which public law standards may be applied to parties’ internal affairs. Since parties are self-governing private bodies, they ought to be free to conduct their affairs as they see fit (Rosenblum, Reference Rosenblum2000; Brody, Reference Brody2002). Yet, the ever-growing proximity between parties and the state in many established democracies has led many to argue that they are better characterised as quasi-public entities (van Biezen, Reference van Biezen2004: 701–722; Gauja, Reference Gauja2010: 92–94; also see Katz and Mair, Reference Katz and Mair2009). Indeed, rather than being mere creatures of contract, parties are now increasingly embedded in national constitutions or legislation around the democratic process (for an overview of these trends, see van Biezen, Reference van Biezen2012). In many jurisdictions, they also enjoy privileges like access to state funding. As such, goes the argument, parties may legitimately be required to comply with public law standards, such as non-discrimination laws or requirements of procedural fairness in their disciplinary procedures [a classic defence of this claim is Kelsen 1929 (Reference Kelsen2013)].
While our argument is predicated on the idea that, in most jurisdictions, it is more plausible to treat parties as quasi-public entities, and thus require them to comply with public law standards, the below analysis also suggests that the party–state relationship constitutes merely one dimension that bears on the issue of parties’ handling of internal dissent. More particularly, settling how parties should respond to dissent within their ranks requires us to ask not only how parties relate to the state, but also offer an account of their relationship with the wider party system, the nature of their relationship with the party member in question, and where a party stands in the life cycle of the electoral process at a given moment. No doubt, this list is far from exhaustive. But we hope that this article contributes towards reorienting and expanding the debate over parties’ internal affairs, by calling attention to the multiplicity of dimensions it involves, over and above the party–state relationship.
What is a party?
We begin by outlining the normative conception of a party that we employ in this paper. This is predicated on the empirical observation that parties are complex membership organisations consisting of party elites and officials, a diverse administrative staff, as well as ‘ordinary’ members and activists (there is an extensive political science literature that uses this understanding of party, and key recent contributions are Katz and Mair, Reference Katz and Mair2009; Mair, Reference Mair2013; Scarrow, Reference Scarrow2014; Ignazi, Reference Ignazi2017). Empirical scholars using this conception of party routinely draw attention to the fact that actors other than party elites – most notably party members and activists – often play an important role in parties (e.g. mobilising voters, deciding on candidates or policy, etc.), though they are naturally careful about drawing normative conclusions from this observation (Hazan and Rahat, Reference Hazan and Rahat2010; Scarrow, Reference Scarrow2014). Our own understanding of party does not specify specific roles for specific actors within the party, but assumes that the agency of those who engage in parties ought to be protected.
Before further explicating what we mean by that, it is worth noting that this view of the party is far from unfamiliar. It figures both in classical (e.g. Gramsci, Reference Gramsci1971; Kelsen, Reference Kelsen2013) and contemporary (e.g. White and Ypi, Reference White and Ypi2016; Ebeling and Wolkenstein, Reference Ebeling and Wolkenstein2018; Wolkenstein, Reference Wolkenstein2018; Biale and Ottonelli, Reference Biale and Ottonelli2019; Wolkenstein, Reference Wolkenstein2019) theoretical scholarship on parties, and its proponents typically argue that parties should be organised in an internally democratic fashion. Some scholars belonging to this tradition, such as the famous constitutional theorist Hans Kelsen, even suggested that parties should be subject to constitutional regulation so as to guarantee individual members (esp. MPs) a ‘degree of democratic self-determination’ vis-à-vis party leaders (Kelsen, Reference Kelsen2013: 41). In this way, the agency of party members should be protected.
This image of the party marks itself off from another popular normative understanding of the party, one that sees them as leader-centred teams of politicians that are rationally motivated to win office. Wedded to a minimalist conception of democracy as ‘competition for [political] leadership’ (Schumpeter, Reference Schumpeter2008: 271), this view is often simply assumed to be the correct view of the party in political science scholarship (on this point, see White and Ypi, Reference White and Ypi2016: 9–14). Accounts that argumentatively defend it (typically against versions of the view of the party that we endorse) tend to highlight the destabilising consequences of internal party democracy, conjuring up the risk of activist capture – parties being controlled by radical minorities whose preferences and values are out of sync with those of the voters – and internal conflict that makes parties unresponsive or produce bad policy (e.g. Rosenbluth and Shapiro, Reference Rosenbluth and Shapiro2018; Manow, Reference Manow2020). The fear is that protecting (or promoting) the agency of party members, to use the language we introduced above, will have high costs for parties and democracy at large.
A major problem with most defences of this ‘minimalist’ view of party is that they tend to operate with a specific political system, or perhaps more accurately: an idealised image of that system, in mind. More particularly, both recent (e.g. Rosenbluth and Shapiro, Reference Rosenbluth and Shapiro2018) and classic (e.g. Schumpeter, Reference Schumpeter2008) accounts treat the mid-20th-century British Westminster system as an ideal embodiment of a well-functioning party democracy. This strategy of theory building overlooks that the ostensible risks of empowering party members are not equally great across political systems; under Proportional Representation (PR), for example, individual parties are under less pressure to be responsive to society as a whole than in the majoritarian Westminster system (they instead form big-tent coalitions to speak to a broader set of voters) (Wolkenstein, Reference Wolkensteinforthcoming). Hand in hand with this idealised image of a particular political system comes a heavily idealised understanding of party competition, according to which the decentralisation of power within parties, say through more inclusive and democratic decision-making procedures, undermines their capacity to provide voters with meaningful policy alternatives. This view of party competition overlooks that, at least according to one influential empirical literature, parties’ growing tendency to ‘limit the degree to which they ‘out-bid’ one another’ is not a function of increased internal democratization, but the collusive behaviour of elites from multiple parties (Katz and Mair, Reference Katz and Mair2009: 758; Ignazi, Reference Ignazi2017; Mair, Reference Mair2013).
This leads to another problem that plagues defences of the ‘minimalist’ view of parties: they often exaggerate the extent to which parties in an established democracy are actually internally democratic, foregrounding individual examples of internal democratisation that in the view of the authors had adverse effects for the party in question (both Rosenbluth and Shapiro, Reference Rosenbluth and Shapiro2018 and Manow, Reference Manow2020 mention the election of Jeremy Corbyn as leader of the Labour party as a paradigmatic case of this) while turning a blind eye to a great deal of evidence that suggests that most parties are still very much controlled by small groups of elites (see, e.g. the comprehensive volume Cross and Katz, Reference Cross and Katz2013 and Ignazi, Reference Ignazi2017). Yet, perhaps, parties struggle to appeal to voters for other reasons than ‘decentralizing democratic reforms’ (Rosenbluth and Shapiro, Reference Rosenbluth and Shapiro2018: 2). Alternative explanations are certainly available, ranging from political elites’ general acceptance of a neoliberal politics of necessity (e.g. Mair, Reference Mair2013; Mudge, Reference Mudge2018) to a weakening of traditional partisan identities due to widespread ‘cognitive mobilisation’ (e.g. Invernizzi-Accetti and Wolkenstein, Reference Invernizzi-Accetti and Wolkenstein2017) – but these are rarely systematically considered by defenders of the minimalist view of parties.
Finally, the normative account of parties that sees them as hierarchically organised teams of professional politicians that act on the basis of self-interest is insensitive to the fact that parties de facto consist of ‘ordinary’ members and activists too, as the above-mentioned empirical research on parties highlight – and that it is reasonable for those individuals to associate with parties (White and Ypi, Reference White and Ypi2016: 13–14). The minimalist view cannot explain this and thereby fails to take seriously the perspective of a great number of people who identify with parties and for whom associating with a party makes sense for reasons that cannot be reduced to electoral self-interest. The more comprehensive conception of party that we endorse can explain this, in addition to explaining why the party members’ agency is worthy of protection. It thus can offer a more accurate normative account of the party that is sensitive to the perspective of those who engage in parties.
Moral agency and the party
With these preliminary arguments for our normative account of the party in place, we can now proceed to the more detailed normative argument for protecting party members’ agency. This forms the backbone of the remainder of the paper. The first thing that needs to be unpacked is the category of party membership itself. While party membership is not a monolithic category (for a cutting-edge overview, see Scarrow, Reference Scarrow2014), for our purposes, it suffices to distinguish between (1) party members who have joined the party in order to influence political decisions, for instance, by joining a local party branch or trying to run for offices within the party (call them activists) and (2) party members who have joined the party for other reasons and expect other kinds of benefits from their membership (call them the loyalists).Footnote 2
This is a stylised distinction, and one can certainly be both activist and loyalist at the same time. But, as conceptual research that differentiates between the social and political benefits of party membership highlights, it makes good sense to separate these two kinds of party membership analytically (Scarrow, Reference Scarrow2014: chs. 7 and 8). Accordingly, party members that we classify as activists are driven by ‘selective’ (obtaining political information and exercising influence) and ‘exclusive’ (office-holding) political benefits, while loyalists are those whose main motivation to engage in a party is ‘collective’ (affirming and sustaining a sense of identity and community) and ‘selective’ (finding friends and raising one’s social status) social benefits. While the problem of freedom of speech and its regulation concerns first and foremost party elites (meaning elected MPs and officials, cf. the cases mentioned in the introduction), and activists, the arguments we develop also have implications for loyalists.
Party elites and activists typically take an active interest in shaping the political decisions that are made within the party. To that effect, they seek to exercise agency within the party, and it is our argument that such an agency should be a matter of moral concern and deserves protection. To understand this, it is useful to first remind ourselves what the nature of the agency at stake is. Some might be inclined to think that individual agency tends to be undermined by participation in collective structures: to submit oneself to the decisions of a collective organisation, they might say, is to trade one’s individual autonomy for the pursuit of shared goals. But though exercising agency qua membership of a group differs from doing so as an individual, there can be no doubt that one can also meaningfully be an agent within a group. Kutz (Reference Kutz2000: 105–106) helpfully distinguishes here between exclusive and inclusive authorship:
I am the exclusive author of the actions I perform myself, as well as of the events caused by those actions. My authorship is exclusive because I and only I can say of an action or event ‘I did it’, or ‘I caused it to be done’. By contrast, I am an inclusive author of the actions of the group in which I participate, inclusive because I am one among those who can say ‘We did it’.
The agency of active party members is best conceived in terms of inclusive authorship.
An important component of inclusive authorship that must be further unpacked is the notion of ‘participation’. In general, we suggest the most appropriate way of thinking about ‘participation’ is to conceive of it as an individual party member’s ability to ensure fair consideration of her preferred course of action. This requires more than mere inclusion in the decision-making process (it does not suffice merely to have a formal right to vote on, say, a party manifesto), but less than the actual influence on the outcome (one need not, say, have personally decided the formulations of the key passages of the manifesto). It requires that party members can secure uptake of their views, without having to have the final word on each decision (see Moore and O’Doherty, Reference Moore and O’Doherty2014). This conception of participation allows members to exercise agency in a way that is consistent with respect for others’ agency, such that they can refer to their joint decisions as something ‘we’ did.
To see why it is a matter of moral concern that active party members can exercise agency in this way, consider first that parties are organisations that are based on a division of labour between different kinds of party members who contribute on multiple fronts (acting as one amongst its public faces, mobilising on the party’s behalf, and so on). Because of this, no single party member is the exclusive author of the party’s actions: each member’s actions are elements of larger processes. This implies that party members have limited knowledge about morally relevant aspects of their work. ‘Responsibility is divided, and … there are often gaps or ambiguities in the division of responsibility, so that no one feels responsible for the outcome … As a consequence, moral wrongs can happen without anyone intending them, and without anyone feeling an immediate responsibility for them’ (Herzog, Reference Herzog2018: 71). This point applies to complex organisations more generally, and we argue that it is true for parties as well. In fact, one of the great dangers of organising parties in a way that leaves no space for their members’ exercise of agency and decouples them from the consequences of their actions is that ‘one can make them complicit in wrong-doing much more easily than if they had to face the consequences of their actions directly’ (Herzog, Reference Herzog2018).
To better understand this point, imagine a group of party activists who seek to convince fellow citizens to vote for their party. The activists spend long days canvassing in their local community, going from door to door and telling people about the party’s new manifesto and the many desirable aims formulated in it. Yet, the activists themselves had no influence on the creation of the manifesto: they were not asked to contribute to it, nor did they have the opportunity to object to or revise any of the proposals it contains. In short, the activists were barred from participating in the process in which the manifesto, whose realisation they want to secure, was discussed, framed, and written. This, we suggest, eventually places serious limits on their capacity to act as morally responsible agents and subject the potential or actual consequences of their contributions to sustained reflection. To put it baldly, they might not know what sort of political agenda they are actually promoting.
Normative scholarship on organisations draws attention to two further ways in which the absence of opportunities to participate in the decisions of the party may pose a threat to active party members’ moral agency (Herzog, Reference Herzog2018: 176–182). The first is complete identification, which occurs when party members uncritically buy into the norms and obligations that others have defined for their role in the party. It is not difficult to interpret some cases of partisan polarisation along these lines, where party members intransigently hold on to their own normative commitments and ignore valid concerns raised by opponents (within and without the party). The second threat to the moral agency is disengagement, where party members gradually become alienated from the party and begin to see their roles in the party merely as ‘coats’ they don temporarily, to be shed once they retreat into the private realm of their ‘real’ lives – think, for example, of party officials who have served the party over many years in administrative and coordinative tasks, without gaining any influence on the decisions, the implementation of which they were meant to oversee. Both of these attitudes pose dangers to party members’ moral agency because they risk making them mere instruments for the realisation of aims determined by others. This is antithetical to the idea of relating to one’s actions as a responsible agent rather than one who is merely acted on or made to act in certain ways.
Agency, deliberation, and free speech
Having established why the agency of active party members deserves moral concern, we move now to the more specific question of what is the link between the exercise of agency by party members and intra-party dissent. One possible conclusion readers might draw from the above considerations is that insofar as parties’ internal democratic processes adequately accommodate members’ concerns – say by establishing effective internal deliberative forums that grant members a degree of what Biale and Ottonelli (Reference Biale and Ottonelli2019) call ‘reflexive control’ over the party’s decisions (on this, also see Wolkenstein, Reference Wolkenstein2019) – party members have no claim to voice dissent and their freedom of speech may permissibly be restricted by the party leadership. For when members’ opinions have received fair consideration, their exercise of agency within the party has been protected, and protected in a way compatible with respect to their co-partisans’ agency.
But this conclusion is spurious. It would make little sense to say, on the one hand, that party members’ agency deserves moral concern, and on the other hand accept that their free speech may be limited so long as parties allow for the fair consideration of members’ views in the making of internal decisions. It would make little sense because if it is valuable that party members take responsibility for the party’s actions, as we suggest it is, then they must also be permitted to criticise those actions after they were taken, rather than being sanctioned for acting as responsible agents who continually reflect on their actions’ possible or actual consequences and perhaps even come to question the decision they contributed to. Indeed, from the point of view of party members’ agency, the outcome of a (however fair) intra-party decision-making procedure can never be treated as a kind of ‘closure’ that ends the debate. For, as deliberative democratic theorists remind us, ‘every democratic decision, even if it takes all existing preferences into account, creates itself a new historical reality which elicits new reactions and new preferences which can be used to question, again, in a never-ending process, the previous decision’ (Rummens, Reference Rummens2012: 28).
There is indeed an internal connection between freedom of speech and free deliberation within parties that this latter point underlines. As with citizens’ deliberations in the public sphere, it is precisely for reasons of respect for the agency of party members that deliberation should not be subject to constraints, which in turn means that they should be ‘free to express their beliefs and values’ even after preliminary decisions have been made and votes were taken (Rostbøll, Reference Rostbøll2011: 7). Limiting free speech within parties thus implies limiting deliberation. Note also that sanctions on free speech need not be implemented in order to be effective. As scholarship on freedom of speech reminds us, the mere threat of sanctions, even when these remain unused, can deter individuals from saying things that they fear could attract penalty – this is typically called the ‘chilling effect’ (for a recent overview treatment, see Townend, Reference Townend, Tumber and Waisbord2017). This applies not only to potential sanctions from the state, but also to other kinds of organisations, such as business firms (see Anderson, Reference Anderson2017: esp. 39–40; Herzog, Reference Herzog2018: 204–205), and indeed political parties.Footnote 3
All of this, we think, makes for a strong presumptive case for why active party members’ freedom of speech deserves greater concern, and hence protection, than might at first appear. We need not at this point go into the more detailed questions of whether free speech protection should be absolute, or whether there are particular thresholds for when speech should be regulated or not regulated. It is anyway difficult to see how these issues could satisfyingly be settled by a generalised theoretical argument about free speech within parties; proposing bright-line thresholds of (im)permissible speech in the name of argumentative precision holds the risk of decoupling theoretical reasoning too much from concrete political contexts. Similarly, it can sometimes be difficult to determine whether a party member exercises free speech to voice constructive criticism or just deliberately wants to act against the party line. We acknowledge that there are clear-cut cases where party members simply want to delegitimise the status quo, and the argument of this article does not extend to these sorts of actions, regardless of whether or not there may be additional reasons for protecting them from sanctions.Footnote 4 So, for now, we will simply state that all party members should have the freedom to speak up when they have concerns, without fearing sanctions . In the remainder of the article, we discuss a number of arguments that pull in the other direction, suggesting that freedom of speech within parties may permissibly be limited.
Before proceeding to these arguments, a brief clarification is in order. Some might wonder whether there is any noteworthy relationship between dissent as an exercise of moral agency and publicness. In short, does it matter whether dissent is voiced publicly or not? As the examples cited in the introduction indicate, in practice, publicness is often necessary in order for dissent to receive uptake, since many parties do not provide members with appropriate means to exercise voice internally. To suggest that party members – elites and ‘ordinary’ members alike – could simply settle their conflicts and disagreements behind closed doors, in intra-party decision-making procedures, would seem to presuppose a rather idealised image of internal party democracy. As many have noted, in reality, internal decision-making procedures tend to be dominated by a few powerful party elites that seek to systematically ‘drown out’ any dissenting opinions (Katz and Mair, Reference Katz and Mair2009: 759; Invernizzi-Accetti and Wolkenstein, Reference Invernizzi-Accetti and Wolkenstein2017: 102–103). Notice, furthermore, that parties are typically territorially extended entities, where local or regional members can only win the attention of the national leadership by expressing critique in the public sphere. The party’s membership base is typically dispersed across the country. To exercise agency within the party, in such circumstances, is to speak to one’s fellow party members, and to persuade them about one’s assessment of the leadership’s actions or policy positions. Such dialogue cannot be restricted to narrow intra-party forums which preclude attempts to engage with the wider membership base, and through them, with the party’s leadership. It has to be public.
The exit rights argument
One powerful argument against. protecting party members’ freedom of speech turns upon the voluntarist nature of the party. It holds that party members’ freedom of speech may permissibly be restricted because they are free to exit any party that is unwilling to accommodate their views, and join another party (or even found a new one). Not only have they, by joining a specific party, willingly entered an organisation that might not always be able to accommodate their views. Because they can always leave the party, they are not ‘coerced into silence’ either. Rather, they remain free to exercise their deliberative agency as described, just not within one particular party . Note that arguments of this sort have been advanced in discussions around ‘workplace constitutionalism’, in connection with the rights of employers to limit employees’ speech (see, recently, Anderson, Reference Anderson2017). Here, we apply this exit rights argument to political parties.
Harming party members’ political rights?
The first reply to this argument is that restrictions on party members’ freedom of speech cannot be defended on the grounds that party members are free to exit the party, since the costs of losing the specific political rights that come with being a party member can be unduly burdensome. This concern would seem to apply to party elites as well as activists, whose participation in the party is closely bound up with the exercise of political rights. It does not apply to party loyalists, who, as noted, join the party for other reasons than exerting influence on political decisions. For those we class as party elites, losing political rights would in many cases go hand in hand with losing employment, since it is their job to exercise political rights within a particular party. Of course, some of them will be sufficiently well off not to suffer economically from the loss of employment, and they will likely quickly find new employment (e.g. in advisory jobs or lobbying). But there is nonetheless a serious impact on an individual’s life plans involved in losing one’s job.
Activists, whose loss of political rights within a particular party would typically not involve losing their employment, would at first seem to be less impacted in terms of their life plans. But one must be careful with granting economic rights relative priority over political rights. The question, we think, is not whether losing one’s job or ‘just’ one’s ability to actively engage in and on behalf of a specific party has greater immediate impact on individuals, but whether the pursuit of our respective rights is harmed. If the claim is that having to withdraw from one’s party can harm the pursuit of political rights, then the absence of immediate, wide-ranging (economic) impact on an individual’s life is not a counterargument we can admit.
A sceptic might counter that membership of a political party is not necessary for the enjoyment of political rights to begin with. Innumerable scholars have emphasised the growing importance of less or non-institutionalised ways of exercising political rights, notably through social movements that are rooted in civil society. For many, social movements perform largely the same function as parties, and they tend to be more attractive to citizens at that (e.g. Tilly and Tarrow, Reference Tilly and Tarrow2007; for discussion and critique, see Dean, Reference Dean2016). So, political rights can certainly be enjoyed and exercised in other ways than through party membership. Yet, given that virtually all democratic states organise democratic decision-making around elections and parliaments, which in turn are organised by parties, partisan affiliation seems to remain the most viable route for ordinary citizens to meaningfully exercise their political rights so as to be heard in the polity’s decision-making process. Those left outside political party competition are considerably less likely to contribute to how decisions are made, even when these decisions affect them in significant ways [for a classic defence of this point, see Kelsen, 1929 (Reference Kelsen2013): 39].
Of course, none of this is to say that those who do not join a political party have no influence at all. Indeed, activists who engage in social movements, in street-level protest, or actions of civil disobedience can achieve considerable impact on political decisions. But it remains the case that they depend to a large degree on being heard by those partisan actors who are ultimately responsible for translating collectively formed wills into binding decisions through the relevant legislative channels; the ‘most they will be able to project therefore is the capacity to influence the governmental agenda – not the capacity to design it and execute it independently’ as parties do (White and Ypi, Reference White and Ypi2010: 818; Dean, Reference Dean2016). Notice furthermore that, depending on the regulatory context in one’s jurisdiction, there can be significant barriers, especially in raising finances, for entry for those who wish to run as independent candidates or start their own party.
Nor are we suggesting that any party is fit to enable the effective exercise of political rights. The widely debated literature on the ‘cartelisation’ of parties suggests in fact that many parties are organised in such a way as to minimise the influence of their members and maximise the party elite’s control over the agenda (Katz and Mair, Reference Katz and Mair2009; Mair, Reference Mair2013). Yet countervailing tendencies are also observable, with parties trying to increase the impact of their members and multiply the ways in which they can exercise voice (Scarrow, Reference Scarrow2014; Gauja, Reference Gauja2015). At any rate, we think that there are good reasons to think that affiliation to a party is an important way of enacting political rights.
To be sure, it is necessary to distinguish between (1) membership of a party and (2) membership of a particular party. This distinction is important because it might be argued that everything that has been said up until this point implies only that the effective exercise of political rights requires membership of a political party; it does not follow that membership of a particular – one’s current – political party is necessary for that purpose. Opinions that are heretical in one party may in fact be very welcome in other parties, perhaps even central to their platform. So, dissenters within a party are likely to have meaningful alternatives available.
That said, one must be cautious with overstating the number of meaningful alternatives. Much depends on the number and kinds of parties that are available, which is mainly a consequence of a country’s electoral system (Rosenbluth and Shapiro, Reference Rosenbluth and Shapiro2018). In some countries – usually those with First-Past-The-Post (FPTP) electoral systems – there are only two parties that alternate in power. These typically represent quite different values and visions for society, and so joining the other party will hardly be an attractive option for a dissenting party member (unless, perhaps, she joins the other party just to ‘make a point’, emphasising that she has turned the humiliation of not being allowed to voice her view within her old party into newly found confidence in the old party’s arch-rival). In other countries – usually those with PR electoral systems and low electoral thresholds – a greater palette of parties might be available. Thus, party members may be able to exit their party and join an alternative party organisation in order to exercise agency. But note that it would be rash to equate the presence of multiple parties with the existence of multiple meaningful alternatives to exercise agency. For the presence of multiple parties only facilitates agency so long as it translates into a wider range of substantive political options: as noted above, there is an extensive political science literature that shows that parties in PR systems sometimes try to outbid each other and limit competition over issues (Katz and Mair, Reference Katz and Mair2009; Mair, Reference Mair2013; Ignazi, Reference Ignazi2017).
But what about founding a new party? On the face of it, this might seem like an effective general alternative that enables party members to exercise political rights without having to mind their speech. Yet again, limitations apply. And again, this has much to do with electoral systems. However easy it might be to found a new party in FPTP systems, the chances of new parties to gain parliamentary representation (let alone govern) are notoriously small in those systems. But even under PR, it is not easy for new parties to establish themselves electorally, unless the effective electoral threshold is close to nil, as in Israel or the Netherlands. Features other than the electoral system, moreover, bear on the flexibility of a party system and the viability of new parties. Polarisation, for instance, can sharply raise the costs of defecting from one’s preferred party, and tends to entrench the existing party system (Goff and Lee, Reference Goff and Lee2019). Similarly, where wealth and social status are narrowly distributed across existing political elites, it is much harder for outsiders to consolidate the economic capacity and social power base needed to challenge the existing order (North et al., Reference North, Wallis and Weingast2009). Often, therefore, the costs of founding a new party will far exceed the benefits, rendering this an unviable option. However, this is not to say that founding a new party can never be a fruitful way forward for dissenting party members. It is simply to say that one must be very careful with presenting this option as a knock-down argument in favour of the proposition that parties are justified in limiting their members’ freedom of speech.
Harming party members’ identity?
In light of all this, we suggest that nothing conclusive follows from discussing these arguments with respect to the permissibility of restricting party members’ freedom of speech. Yet, another argument is available. This holds that involuntarily losing one’s membership of a party adversely impacts individuals’ social status and self-esteem, turning on a close connection between political affiliation, social standing, and evaluations about self-worth. Accordingly, party members suffer personal harm from exercising exit rights, and in that their broader identity and self-understanding are compromised.
Now, it is certainly not unreasonable to assume that, for some party members, partisan affiliation is a constitutive element of their identity. As the political science literature acknowledges, this is undoubtedly the case for loyalists who join parties to affirm and deepen their sense of belonging to a particular partisan community. But the same may be true for activists and party elites who joined their party in order to shape collective political decisions in accordance with shared principles and aims (White and Ypi, Reference White and Ypi2016). Indeed, many activists and elites invest their personal identity with the party label and view their fate as closely connected with their party’s. As former UK Labour MP Ian Austin stated, as he resigned from the party, ‘The Labour Party has been my life, so this has been the hardest decision I have ever had to take’ (Reuters, 2019). Statements like these emphasise that decisions to leave a party are not taken lightly, for party members’ personal identity is often closely entwined with their political affiliation; thus, we should be wary of imposing on them the choice between free expression of their opinions and exit.
Some might object that the cultivation of such ‘identitarian’ dispositions is not obligatory on behalf of party members (an argument inspired by the Schumpeterian ‘minimalist’ view that party membership is purely instrumental to asserting material interests, a view we rejected in the first section). However, drawing attention to this fact does little to chip away at the claim about the importance of identity and self-esteem in connection with party membership. Nor would it suffice to point out that not all members adopt an attitude of the kind outlined above towards their respective parties – think, for example, of passive members who joined a long time ago and are now alienated from their party. These responses are insufficient because for many party members, membership is, after all, a matter of deeply held commitment, closely tied to their self-worth; and for those members at least, the exercise of exit rights is a very burdensome option.
This burden does not, by itself, ground a claim to freedom of speech within the party. As we have argued, the grounds for speech are afforded by the significance that exercising moral agency within the party bears. Our point, rather, is that, in evaluating whether or not some members can genuinely exercise – or be expected to exercise – exit rights, the burdens this option imposes deserves attention. If, as is the case with some members, the relation between them and their party runs sufficiently deep, exit may be so burdensome so as to render their exclusion from the party a disproportionate response. In such cases, affording due recognition to the nature of members’ relationship with their parties and the significance it bears for them affects how the proportionality of expulsion (or exit) is weighed (for more on proportionality and speech, Parr and Billingham, Reference Parr and Billingham2020: 378–383).Footnote 5
Additional non-material costs?
The final argument against the exit rights argument is that for many who engage in parties, leaving the party would be an extremely hard decision because of further non-material costs this would entail. One might argue that such costs are relatively unique to the particular party a person supports. While it is certainly true that most members of contemporary parties do not normally spend their lives attending party forums, making friends, or developing relationships, for some, paradigmatically (not exclusively) for those we call loyalists, the relationships they develop through long-standing partisan affiliation are an important part of their lives. Perhaps, these relationships are even the primary reason why they still engage in the party. Many such individuals may have spent longer time supporting, campaigning, or working for a party than most individuals have spent at their present place of work. It is not unreasonable to imagine, then, that their social lives may have become closely linked with their partisan affiliation. Of course, party activists and elites may equally develop such forms of attachment to the party. A former Labour Foreign Secretary stated while resigning from the party, that ‘The Labour Party has been my family; most of my friends are in it’ (Metro, 2009). He is hardly the only long-standing member of a party whose partisan affiliation matters for this reason.
Conscious of the friendship and indeed family-like relationships that engaging in a party can engender, several scholars have gone so far as to suggest that partisanship as such is a form of ‘political friendship’ (Muirhead, Reference Muirhead2014; Ypi, Reference Ypi2016). In this view, partisanship, which is institutionally stabilised by membership in a party, is an ‘associative relation established when the interest in such projects is shared with other people who (like friends) support each other in their pursuit’ (Ypi, Reference Ypi2016: 605). For our purposes, the primary takeaway from these theoretical arguments is that party members’ ‘awareness of the worthiness of one’s political commitment … draws … confirmation from the day to day engagement with concrete others who contribute to that shared project with their knowledge and efforts’ (Ypi, Reference Ypi2016). And with this ‘day to day engagement with concrete others’ comes the establishment of close personal relationships that become valuable through, and eventually independent of, the shared commitment to particular political goals that initially brought the relevant individuals together.
Once again, our arguments about the significance of partisan affiliation for one’s self-esteem and social relationships must be qualified. While partisan commitments may have such significance for many, not all members may relate to their party in the same way. Consider, for instance, relatively new members, or those members whose engagement with their party has been intermittent rather than sustained. For such individuals, partisan commitments are less likely to be connected to their sense of self or their social lives in ways that could render exit from the party unduly burdensome. As such, parties may be justified in offering greater protection against disciplinary sanctions to those who have been party members for a specified period of time or have demonstrated a given level of engagement with the party.
To sum up: the exit rights argument cannot conclusively settle claims for the exercise of free speech within political parties. At least for those party members who exhibit a deep commitment to the party, exit rights arguments seem insufficient in order to justify restrictions to limits of speech. But it is difficult to see how one could advance a more generalised argument. In many (perhaps most) cases, whether exit is costly depends on specific individual circumstances, to do, for instance, with the depth of an individual’s commitment to a given political party, or the thickness of her relationships with other party members. And there are also many cases where our judgment about the costs of taking the exit route depends on systemic factors, such as whether meaningful alternative parties exist. What is certain is that party members at least in some situations cannot simply withdraw from their organisation without bearing significant burdens. But again, we cannot plausibly draw the strong conclusion that parties may never permissibly constrain their members’ freedom of speech.
The competitiveness argument
Consider yet another argument for restricting party members’ freedom of speech. This holds that electoral success is a non-negotiable objective that any political party must pursue, and so any conduct that undermines a party’s ability to pursue this objective – such as dissent by party members – may permissibly be restricted.
An initial objection to this argument might be that, apart from electoral success, partisans have a compelling interest in the advancement of shared ends: the propagation of collective ideas and persuading others of their validity. In this respect, parties are less like firms and more like members of a religious denomination. The realisation of shared ends arguably presupposes some discipline within the party, and the exercise of free speech by some might jeopardise this collective good for others. But note that the notion of ‘shared ends’ also presupposes a level of inclusive deliberation, however unfinished or episodic, over the proper aims and purposes of a party. And this means that, without sufficient safeguards for members’ speech, we could not rely on the party’s discursive processes to point towards ends that could be taken as shared.
Now, how exactly might freedom of speech within parties pose a problem to their pursuit of electoral success? First, if party members are able to freely voice whatever criticism they have of the party’s general direction, this might mean that ‘intra-party dissension flares all the time, unsuppressed’ (Rosenblum, Reference Rosenblum2008: 361), making it very difficult for the party to agree on a shared platform. The typical result of this, programmatic vagueness, ‘is clearly costly to political parties’ (note that this is categorically different from deliberate and strategic position blurring, which can in fact be advantageous to parties, see Rovny, Reference Rovny2014: 272). Unable to communicate where they stand in terms of policy, voters are likely to desert them. Second, free speech in the form of individual members vocally criticising the party leadership might send to voters the potentially costly signal that the party leadership has ‘lost control’ over the party, casting doubts on their credentials as leaders more generally. This, too, can be electorally costly.
What must be borne in mind is that ‘electoral success’ might mean very different things for different parties, and how much dissent a party can admit may depend much on its understanding of electoral success. Clearly, electoral systems create an important enabling and constraining environment here. In FPTP systems, electoral success typically means winning elections and, by extension, office. Because of that, one might be inclined to say that parties under FPTP can accommodate less internal disagreement. To win, they simply need to minimise contingencies and speak with one voice. Historically, this is reflected in high levels of party discipline and centralised leadership (Kam, Reference Kam2009; Rosenbluth and Shapiro, Reference Rosenbluth and Shapiro2018).
In PR systems, by contrast, electoral success can mean a range of different things, from being able to form new coalitions, to further empowering the opposition, to increasing the total number of votes of one’s ideological ‘bloc’, as in the Scandinavian parliaments (Green-Pedersen and Thomsen, Reference Green-Pedersen and Thomsen2005). There is a broad palette of goals parties can set for themselves. Under those circumstances, it seems, there is also more space for dissenting opinions within the party, not only because for many parties their stakes are lower than under FPTP – it is not always a matter of winning or losing – but also because parties have more room to experiment with internal democracy and alternative forms of member participation (Wolkenstein, Reference Wolkensteinforthcoming). They might even develop a distinctive ‘party brand’ around internal discussion and bottom-up participation, just as many of the Green parties did when they first emerged in Western European party systems.
In addition to electoral systems, the territorial organisation of parties is likely to affect the extent to which they can accommodate internal disagreement without compromising electoral success. In federal states like Germany, parties form ‘territorially integrated organisations which are internally subdivided along the federal polity’ (Benz, Reference Benz2003: 34–35). And although the regional party organisations tend to generally support the national party’s values and goals, they ‘still have room to maneuver because of their own legitimation through regional elections’ (Hadley et al., Reference Hadley, Morass and Nick1989: 95). This means that regional party organisations may well freely voice their disagreement with the national party line and – depending on the size of the region and the electoral support they enjoy – even exercise pressure on the national party. There is little reason to think this must be detrimental for the party’s overall electoral success, however. Indeed, if the relative electoral stability of many of the established parties in the above-named countries is any indication, then there is no necessary connection between the federal organisation and the level of disagreement it brings with it, and a party’s electoral fortunes.
These considerations certainly provide no knock-down argument against the proposition that parties may permissibly constrain their members’ freedom of speech in order to ensure that they retain electoral competitiveness. But they chip away at the underlying premise that allowing freedom of speech within parties necessarily makes it harder to succeed in the electoral arena. The next thing we want to question is whether electoral success is always and necessarily the most important goal parties pursue.
Remember that there are many political systems where (mainstream) parties have formed ‘cartels’ in order to limit competition. Party scholars have suggested that, in those systems, the interests of the cartelising parties lie as much in winning elections as ‘in having the possible costs of losing reduced as much as possible. After all, always winning is unlikely’ (Katz and Mair, Reference Katz and Mair2009: 756). In practice, this mainly involves that the parties in question secure greater access to public funds; ensure that their members obtain important positions in the bureaucracy; and distribute more and more policy responsibilities to unelected bodies that are somewhat aligned with their political objectives (think of central banks, constitutional courts, etc.) (see Hopkin and Blyth, Reference Hopkin and Blyth2019). Thus, for parties that cooperate in cartels, electoral success might indeed not be the single most important goal. They might invest just as much effort and energy into upholding their system of mutual cooperation with other parties and the state more generally. This considerably reduces the force of the argument that parties may permissibly limit their members’ freedom of speech in order to remain electorally competitive.
Besides the more specific case of cartel parties, there is a real question of whether electoral success should be seen as always having priority over other aims pursued by parties. By this, we mean that it may well be that there are certain periods, most obviously election campaigns, where electoral success is of primary importance. During such periods, and all else equal, parties may be said to have a stronger claim to impose sharper restrictions on their members’ speech so as to ensure a coherent message for voters. An analogy from the law may help clarify what we mean here. In many jurisdictions, election law recognises the distinctive status of campaign periods, carving out exceptions that would not apply under ordinary political moments. For instance, parties and candidates are frequently permitted to distribute election materials and posters only during specified campaign periods. Such regulations recognise the distinctiveness of electoral campaigns and indicate how norms that govern such periods should not become routine practice. The question of strict party unity in the context of the need for a coherent electoral message could plausibly be approached along similar lines.
We think, indeed, that it is crucially important that parties do not always act as they do during campaign periods. Doing so would reduce them to mere electoral machines, and they would likely fail to take seriously their broader role as collective agents that channel and articulate societal grievances, advance public justifications for their aims and policies, and socialise citizens into the political process (on this, see White and Ypi, Reference White and Ypi2016). Arguably, parties’ capacity to serve these other democratically important functions would be severely impaired if they always imposed strict speech limits on their members. For example, it is hard to see how parties could effectively channel and articulate societal grievances if members are sanctioned for criticising the party line for being insufficiently attentive to citizens’ most pressing concerns. The cases of Suheyl Batum and Gufran-e-Azam, mentioned at the beginning of this article, can help illustrate this point. These are senior party members who were critical of their party leaders’ performance in their election campaigns, and subsequently urged a post-mortem of their parties’ campaign strategies and incumbent heads. The timing of their criticism meant that they could hardly be said to damage their parties’ electoral prospects except if the relevant temporal lines were drawn so widely that the parties were seen as perpetually in campaign mode.Footnote 6
In sum, we think the correct conclusion is that, first, the argument from electoral competitiveness has much less weight than might at first appear. Depending on the political system within which parties operate, free speech within parties need not be costly for them, plus it is hardly the case that winning elections is always or necessarily the most important objective parties pursue. Second, insofar as parties seek to limit free speech on the grounds of electoral aims, their claim to do so is much stronger during election campaigns. Election campaigns are extraordinary periods, and we believe there are good reasons to think that this also (pro tanto) permits a more restrictive approach to internal freedom of speech.
Conclusion
We have argued that preservation of party members’ moral agency constitutes a significant interest, and that this interest is best served by the protection of their speech vis-a-vis the party. As such, there is a compelling presumptive case in favour of freedom of speech within the party. While we have acknowledged that there may be circumstances where this case may be overridden, we have tried to argue that it may be harder to defeat than one might initially think. Although members may seem free to exit their respective parties, we have argued that exit from one’s party is not a plausible substitute for exercising moral agency within the party. For it may, depending on the context, negatively impact one’s exercise of political rights to democratic participation, impose costs on an individual’s social status and self-esteem, or carry non-material costs attached to the severing of close ties and relationships with co-partisans.
To be sure, such arguments are not decisive in grounding a universal right against sanctions for dissent. However, they warn against conceptualising exit rights as a default strategy for the exercise of moral agency, demonstrating why partisans may deserve greater protection for internal dissent. Nor is it clear, as we have argued, that parties’ interests in electoral success always trumps their members’ interests in the exercise of moral agency. First, depending on the electoral system in place, internal dissent may not be so costly for parties. Second, there are ways of balancing parties’ electoral objectives with their members’ claim to exercising moral agency by, for instance, taking a stricter view of dissent during a limited campaign period. In sum, to settle how parties should respond to dissent within their ranks, we need to take into account a wide range of circumstances, including where a party stands in the life cycle of the electoral process at a given moment and the party’s relationship with the wider party system. Indeed, there is a multiplicity of normatively relevant factors that the ongoing debate over parties’ internal affairs should heed, and we hope that this article can contribute to reorienting and expanding the debate by drawing attention to some these factors.
Acknowledgements
We were fortunate to have the opportunity to present earlier versions of this article in workshops and seminars at the universities of Aarhus, Cambridge, Groningen, and Oxford. The feedback we received on these occasions proved immensely helpful. In addition, we would like to thank Nick Barber, Sam Bagg, Tarunabh Khaitan, and Vatsal Naresh, as well as four anonymous reviewers and the editors of this journal, for their insightful comments and criticisms.
