The Normative Theory of Partisanship

The recent normative literature on parties and partisanship identifies two main dimensions along which parties and partisanship, at their best, contribute to democracy. The first, horizontal dimension captures what we refer to as the collegiality of parties and partisans: their ability to work and cooperate across partisan lines to sustain liberal democratic institutions and advance the common good. Parties and partisans are collegial and therefore desirable when they behave responsibly towards the other parties in their party system, which in turn demands that they support the principle of non-violence and, by extension, the values of legitimate opposition and regulated rivalry (Rosenblum Reference Rosenblum2008). Furthermore, collegial parties abide by certain norms of civility – for instance, concession speeches following an electoral loss and other rituals of protocol and etiquette (Bonotti and Zech Reference Bonotti and Zech2023)Footnote ¹ – and advance public justification; that is, they ‘attempt to move beyond a particularist viewpoint with the aim of demonstrating how a certain claim has public appeal’ (White and Ypi Reference White and Ypi2016, 61) or, to put it in more Rawlsian terms, justify political rules by appealing to the ideal of public reason (Bonotti Reference Bonotti2017).

The collegiality dimension of the normative theory of partisanship presents some overlaps with the ideas of deliberation and deliberativeness. Although there is limited work on parties and deliberative democracy (Muirhead and Rosenblum Reference Muirhead and Rosenblum2020, 102–4), parties can clearly play a key role in shaping the terms of deliberative democratic debate (Muirhead Reference Muirhead2019; Rosenblum Reference Rosenblum2008) and, depending on their internal processes, they can even be forums where deliberation takes place (Wolkenstein Reference Wolkenstein2016). However, while deliberativeness is closely linked to the public justification dimension of collegiality (cf. Bohman and Richardson Reference Bohman and Richardson2009), it does not perfectly overlap with collegiality itself. Indeed, deliberativeness may be lacking while the other two dimensions of collegiality, non-violence and civility, are present.

The second, vertical dimension of the normative theory of partisanship captures parties' ability to provide ‘linkage’ between citizens and the state (Dalton, Farrell, and McAllister Reference Dalton, Farrell and McAllister2011, 7; Lawson Reference Lawson, Lawson and Merkl1988). Party linkage plays a key role in representative democracies, particularly via candidate recruitment (Sartori Reference Sartori2005, 63), citizen mobilization (Verba, Schlozman, and Brady Reference Verba, Schlozman and Brady1995), the aggregation of voter interests (Almond Reference Almond, Almond and Coleman1960, 40), representation (Powell Reference Powell, Dalton and Anderson2010), and policy implementation (Klingemann, Hofferbert, and Budge Reference Klingemann, Hofferbert and Budge1994). While this is what, empirically, parties actually do, a number of democratic theorists have pointed out that it is also normatively desirable that parties provide this kind of linkage between citizens and the state. As Wolkenstein cogently puts it, ‘[p]olitical parties serve a number of important functions in representative democracies. Connecting citizens to government is perhaps the most important one. This is how parties were traditionally conceived, and it continues to be the main standard according to which their legitimacy as representative institutions is evaluated’ (Reference Wolkenstein2016, 297).

We believe that, ultimately, the linkage function of parties can be reduced to two elements: voice and accountability. Voice refers to citizens' ability (individually or via civil society organizations that pool their demands) to influence parties' manifestos.Footnote ² However, voice per se may not be sufficient to guarantee effective linkage since parties and partisans may include pledges that reflect citizens' demands in their manifestos but then disregard those pledges once elected. Therefore, linkage also requires accountability, which, following Pitkin (Reference Pitkin1967), presents two elements: first, the uptake of citizens' voice and, more specifically, the expression of their demands (for example, during parliamentary debates), the implementation of manifesto pledges (when a party is in power), or principled opposition to governmental proposals and decisions (when a party is not in power); second, the sanctions that generate incentives for representatives to perform uptake. The latter include formal ‘punishments’ such as the decision by an individual voter to abandon a previous party allegiance (for example, by voting for another party or candidate at the next election) and informal mechanisms through which citizens can attribute blame – especially (but not solely) in a targeted, individualized way – to representatives whose uptake falls short of expectations.

It is important to stress that our focus on party systems (rather than on individual parties and partisans) requires a systemic shift in the normative theory of partisanship, though only with regard to linkage. Indeed, collegiality is inherently systemic because when parties and partisans comply with the demands of collegiality, the party system as a whole will become more collegial. When it comes to linkage, however, a systemic shift means focusing not only on the linkage that individual parties and partisans supply to their constituents but also and especially on the overall level and quality of linkage across the entire party system.

To capture this systemic shift, we relabel the linkage dimension as systemic linkage and describe its two sub-dimensions as systemic voice and systemic accountability. Ideally, systemic voice should be both clear and proportional. Clarity is present when citizens can effectively communicate their demands to their preferred party and/or candidate, which is to say they have meaningful opportunities to be heard and to shape their preferred party's manifesto (as well as the more specific pledges made by individual candidates); for example, via direct contact with their representatives and, especially, via various mechanisms of intra-party democracy and deliberation (Wolkenstein Reference Wolkenstein2016). Proportionality requires that a clear voice should be guaranteed for as many citizens as possible, proportionate to the distribution of their demands across society. Likewise, a party system displays systemic accountability when all individual parties and partisans are subject to formal and informal sanctions that incentivize their uptake of citizens' demands.

Before applying our framework, which is summarized in Table 1 above, to the analysis of party systems and political finance arrangements, we would like to clarify three features of our approach.

Table 1. The concepts of the normative theory of partisanship

First, to make our analysis tractable, we focus on established liberal democracies and set aside the issues that may arise when considering democratizing and/or deeply divided societies, and those where populist or other illiberal actors seriously challenge liberal democracy.

Second, though we present a normative analysis, we engage closely with real-world institutions and empirical evidence to avoid an idealized approach detached from practice. At the same time, we also depart from the hyper-empirical (and typically statistical) approach of much contemporary research on political finance and party systems. We aim to provide an ‘applied democratic theory’ (Wolkenstein Reference Wolkenstein2021, 314), which enables systematic normative thinking about political finance reform in different party system contexts.

Third, we acknowledge that the normative theory of partisanship is somewhat derivative of more fundamental democratic principles, especially concepts and theories of democratic representation, deliberation, and public justification (Muirhead and Rosenblum Reference Muirhead and Rosenblum2020). However, our analysis uses the normative theory of partisanship because it is more empirically sensitive than highly abstract concepts and theories. As a result, the conjectures that emerge from our analysis are more amenable to empirical testing. Furthermore, the normative theory of partisanship draws together an array of concepts that are central to other democratic theories in a synthesis that focuses on parties and their distinctive role in representative democracies; that is, the fact that, unlike other representative organizations (for example, NGOs and advocacy groups), ‘[p]arties and partisans … seek to govern’ (Muirhead and Rosenblum Reference Muirhead and Rosenblum2020, 99).

Party Systems and the Normative Theory of Partisanship

We apply the normative theory of partisanship to analyze party systems of three kinds: concentrated systems, moderate pluralist systems, and polarized pluralist systems. While this schema is a simplification of the complex terrain of party competition, it nonetheless distinguishes between systems based on properties – especially fragmentation and polarization – that are widely recognized in classic and contemporary empirical scholarship on party systems (for example, Bardi and Mair Reference Bardi and Mair2008; Duverger Reference Duverger1954; Kalmoe Reference Kalmoe2020; Katz and Mair Reference Katz and Mair1995; Sartori Reference Sartori2005). Party systems also differ in terms of their stability (for example, Mair Reference Mair1997; Nwokora and Pelizzo Reference Nwokora and Pelizzo2018), but to keep our analysis tractable we assume that a country's party system approximates one kind over time. As we explain later, the framework should be reapplied when this assumption does not hold (that is, a party system has transformed) because it may be necessary to revise what should be considered the ideal political finance arrangement for that context.

In the subsections that follow, we draw on empirical literature to present a stylized summary of each of the three party system types we consider and then generate conjectures (based on existing empirical literature) regarding its performance against the three normative dimensions of partisanship: collegiality, systemic voice, and systemic accountability.

Concentrated Party Systems

We describe as ‘concentrated’ a party system with low fragmentation, corresponding to a two-party system (Sartori Reference Sartori2005) or one with two ‘effective’ parties (Laakso and Taagepera Reference Laakso and Taagepera1979).Footnote ³ Examples include Australia, the United Kingdom, and the United States. Concentrated party systems often result from majoritarian (for example, single-member plurality) electoral systems (Duverger Reference Duverger1954; Riker Reference Riker1982). They tend to exhibit centripetal tendencies and ideological convergence (Downs Reference Downs1957; Sartori Reference Sartori2005), with the major parties competing to win the support of pivotal voters with moderate preferences. This, in turn, leads these parties to cast themselves as ‘catch-all’ organizations (Katz and Mair Reference Katz and Mair1995; Kirchheimer Reference Kirchheimer, LaPalombara and Weiner1966; Sartori Reference Sartori2005), appealing both to strong ideological partisans on their side of politics and to more pragmatic moderates. Yet, ideological convergence in concentrated systems is accompanied by high levels of competitiveness since these are ‘winner-takes-all’ systems with an expectation of single-party governments and (particularly in the United Kingdom and other Westminster democracies) constitutional arrangements that concentrate power in the executive without any clear division of powers between the executive and legislative branches of government.

We expect that collegiality will be low in these systems. High stakes and intense, all-encompassing competition increase the likelihood that partisans from different parties will regard and treat each other as enemies rather than competitors. This process may even result in ‘partisan warfare’ (Kalmoe Reference Kalmoe2020). When it comes to public justification, given the low number of parties in this system and the fact that most governments are formed by a single party, once they have been elected, legislators will not be under much ‘justificatory pressure’ (Bonotti Reference Bonotti2017, 34) to take into account a broad and diverse range of demands (for example, demands advanced by coalition partners and/or opposition partisans) beyond those arising from their constituents when justifying laws and policies.

We expect that systemic voice, like collegiality, will also be low in concentrated party systems. While it is true that in single-member districts (which are associated with these systems), constituents may have regular opportunities to visit and speak to their local MP (see Dalton, Farrell, and McAllister Reference Dalton, Farrell and McAllister2011, 67–8), this access may not necessarily translate into actual impact on the party's manifesto. Indeed, the main parties in a concentrated system are usually selective and tend to prioritize the preferences and demands of the median voter to the detriment of those of other constituents (including, perhaps, the enthusiastic ones who frequent an MP's office) (Magni-Berton and Panel Reference Magni-Berton and Panel2018). Therefore, many citizens' demands will be neglected, resulting in a disproportionate distribution of voice across the system. Furthermore, even those demands that make it into a major party's platform are likely to be diluted into an indistinct, ‘grey’ manifesto. Due to this loss of clarity, those demands are likely to have no more than a weak impact on the party's political agenda.

We expect that systemic accountability will generally be high in concentrated party systems.Footnote ⁴ The main parties in this system regularly perform all three uptake roles (expression, implementation, and opposition). Furthermore, the presence of single-member districts provides voters with powerful formal (that is, electoral) and informal (that is, public scrutiny) sanctioning tools because parties have limited power to protect individual MPs from punishment by voters (Carey and Shugart Reference Carey and Shugart1995).

Moderate Pluralist Party Systems

‘Moderate pluralist’ party systems, as seen in Belgium, Germany, and the Netherlands, present between three and five ‘relevant’ parties (Sartori Reference Sartori2005), which span a moderate range in the ideological spectrum. The extremes of the system are closer and receive less support than the poles in a polarized pluralist system, which we will discuss shortly. Yet, despite the number of parties exceeding two, competition has centripetal characteristics defined by stable ideological blocs (Sartori Reference Sartori2005, 159). Moderate pluralist systems tend to result from a proportional electoral system, usually with large multi-member districts (Riker Reference Riker1982). Elections in moderate pluralist systems tend to produce indecisive outcomes, with no one party securing sufficient support to govern on its own. Therefore, coalition governments are common. Unlike polarized systems, though, the coalition governments in moderate pluralist systems tend to be stable more often than not, and they are predictably formed by ‘friendly’ parties that occupy the same side of the major ideological cleavage (Sartori Reference Sartori2005, 163).

We believe that collegiality is generally high in moderate pluralist systems. For a start, competition between parties tends to be narrower and less intense than in concentrated party systems, and parties are normally required to work together in coalitions to govern. With fairly predictable coalition arrangements, repeated interactions over time between partisans of different parties are likely to foster trust and reciprocity and to generate ‘deep-rooted traditions of compromise … [and] … cooperation’ (Bille Reference Bille1989, 57), thus reducing inter-party conflict and promoting collegiality. Furthermore, the presence of a more diverse range of parties in parliament (than in concentrated party systems) will exert multifarious justificatory pressure upon those in power (and upon legislators more generally), forcing them to take the voices of those parties (and their constituents) into account when formulating policies, which in turn advances public justification (Bonotti Reference Bonotti2017).

When it comes to systemic voice, we believe that moderate pluralism presents contrasting tendencies. On the one hand, given the presence of a greater and more diverse range of parties than in concentrated party systems, the voters' voice is more likely to be clearly and proportionately reflected in party manifestos across the system under moderate pluralism. On the other hand, multi-member districts mean that voters in a moderate pluralist system are less likely to be in close contact with their representatives than they would be in a concentrated party system (Dalton, Farrell, and McAllister Reference Dalton, Farrell and McAllister2011, 67–8). All things considered, we expect a medium level of systemic voice in moderate pluralist systems.

We believe that systemic accountability for major parties under moderate pluralism will be lower than it is for such parties in concentrated systems. As in concentrated systems, these parties will still regularly perform each of the three uptake roles, though to a diminished extent because of the compromises that coalition politics demands. However, the presence of large multi-member districts, especially alongside closed-list proportional representation (through which party leaders can distribute electoral risk among candidates), gives MPs some protection from their constituents' sanctions. Therefore, the accountability of individual MPs can be very low under moderate pluralism (Lancaster Reference Lancaster1986). However, the collective accountability of the major parties can still be high because structured competition between stable party blocs means that a decline in voter support for one or more coalition partners, as a punishment for their uptake failures, can result in the coalition's replacement. In sum, we expect a medium level of systemic accountability under moderate pluralism.

Polarized Pluralist Party Systems

Polarized pluralist systems (for example, Italy, Austria, Denmark, and Spain) present between five and eight relevant parties (Sartori Reference Sartori2005). At this level of fragmentation, there is often a centrifugal dynamic that fuels polarization because the system's fringe parties can remain electorally viable while promoting extreme positions (Pelizzo and Babones Reference Pelizzo and Babones2007). New parties often emerge; thus, the system will tend to present higher volatility than other types, especially at its fringes (Evans Reference Evans2002). As with moderate pluralism, governments are typically formed from coalitions, though these are likely to be less predictable and stable than coalition governments in a moderate pluralist system.

What are the implications of polarized pluralism for collegiality? Since coalition dynamics are less predictable in polarized pluralism than under its moderate counterpart, and due to the presence of ‘many small relevant parties in keen competition’ (Bille Reference Bille1989, 57), there will be fewer opportunities for elected partisans from different parties to engage in repeated interactions over time with the same coalition partners to cultivate collegial relationships grounded in trust, reciprocity, and a spirit of compromise. When it comes to justificatory pressure, we would expect that this will be greater than under both concentrated and moderate pluralist systems due to the presence of a more diverse range of parties in the legislature; however, the greater ideological distance between parties, which is typical of polarized pluralism (Powell Reference Powell and Holler1987), will often make it difficult for the parties in government to provide policy justifications that all other parties can accept. In sum, we expect that collegiality in a polarized pluralist system will be at a medium level – higher than in concentrated party systems but lower than under moderate pluralism.

In terms of systemic voice, like under moderate pluralism (and unlike in concentrated party systems), this is likely to be reasonably proportional; in fact, more proportional than under moderate pluralism since there will be more (and more diverse) parties that citizens can vote for and identify with. Furthermore, due to the short-lived nature of coalition governments in polarized pluralism, parties' time horizons are also likely to be short. Parties will, therefore, try to maximize support among their core voters on the left and right of the political spectrum more forcefully than they would under moderate pluralism (Evans Reference Evans2002), where time horizons are longer and a willingness to compromise is more important. Accordingly, citizens' demands are likely to be more closely reflected in parties' manifestos than they would be under a moderate pluralist or concentrated party system. Furthermore, although citizens will have less direct contact with their MPs than they would in a concentrated system, we expect that contact in a polarized pluralist system will be higher than under moderate pluralism because the presence of more parties and candidates in any given constituency supplies greater opportunities for contact across the political spectrum. Therefore, we expect that systemic voice will be high under polarized pluralism.

When it comes to systemic accountability, due to the volatility of polarized pluralism, all three dimensions of uptake – implementation, opposition, and expression – are likely to be more widely but also less stably distributed across parties under this system than in moderate pluralist and concentrated systems. However, formal electoral sanctions are likely to be weaker (than in both concentrated and moderate pluralist systems) across the political spectrum since entering and leaving government often does not depend on winning or losing elections – coalition negotiations between elections are, practically speaking, just as important as elections in the process of government formation (Kam, Bertelli, and Held Reference Kam, Bertelli and Held2020, 752). Therefore, it will be more difficult for voters to use electoral sanctions to punish elected partisans and parties who fail at uptake. For all these reasons, we expect that systemic accountability will be low under polarized pluralism.

Table 2 below summarizes our assessment of the party system types in terms of the dimensions of the normative theory of partisanship.

Table 2. Party system types and the dimensions of the normative theory of partisanship

Party System Types in the Real World: Variants and Change

Our analysis in the previous section focused on stylized models of party systems, an approach that is necessary, we argued, to enable systematic normative analysis. While our models aim to capture a substantial range of practical scenarios, the features of real-world party systems may diverge from these stylized descriptions in several ways.

First, a party system may present some but not all of the features associated with a particular model, making it a distinctive variant of the basic model. For example, although we have described concentrated systems as ideologically moderate, the contemporary US party system displays a high level of polarization (while also corresponding to the basic concentrated model in other key respects) (Campbell Reference Campbell2018). Second, a real-world party system may not remain as one model over time. The occurrence of ‘party system change’ (Nwokora and Pelizzo Reference Nwokora and Pelizzo2018) may, for instance, involve the party system becoming more fragmented (as well as changing in other respects) and, therefore, move from moderate pluralism to polarized pluralism. Third, a party system can also ‘change’ less significantly by moving from one variant to another, as happened in the US as its party system became increasingly polarized (while remaining fundamentally a concentrated system).

Taking account of such real-world complexities is possible within our framework, consistent with its goal of supporting empirically-informed normative analysis. In particular, when a real-world party system diverges from the assumptions of a basic model, the conjectures associated with that model should provide a point of departure from which amendments can be made to account for the real-world divergence. For example, if a real-world concentrated party system displays greater collegiality than the basic model suggests – this seems a reasonable description of the mid-twentieth century US party system, when polarization was lower and cross-party cooperation (bipartisanship) was fairly common (Han and Brady Reference Han and Brady2007) – the party system might need to be scored as ‘medium’ instead of ‘low’ in this dimension. As we explain later, this revised party system scoring may have implications for the assessment of political finance arrangements and for identifying the political finance setup that should be regarded as the best fit for the party system.

Political Finance and the Normative Theory of Partisanship

We now apply the three dimensions of the normative theory of partisanship – collegiality, systemic voice, and systemic accountability – to evaluate political finance arrangements. As we noted in the introduction, the existing normative research on political finance evaluates political finance rules against first-order principles, especially political equality, liberty, freedom of expression, representation, participation, transparency, and anticorruption (for example, Cagé Reference Cagé2020; Dawood Reference Dawood2015; de Sousa Reference de Sousa2001; Grant and Rudolph Reference Grant and Rudolph2004; Pevnick Reference Pevnick2016; Tham Reference Tham2010). Additionally, empirical studies suggest further normative principles. Most notably, La Raja and Schaffner (Reference La Raja and Schaffner2015) explain how US political finance arrangements have fuelled (harmful) polarization, suggesting that ideological moderation can be regarded as another basic principle. The normative theory of partisanship is ultimately rooted in such values – these connections are summarized in Table 3 below. Consider, for instance, Pevnick's (Reference Pevnick2016) argument that political finance reformers in the US have been motivated by anti-corruption goals and two equality rationales (equal opportunities for citizens' political influence and equal visibility of competing party positions). Each motivation can be tied to the normative theory of partisanship. Systemic accountability is correlated with anticorruption (and transparency) (Schwindt-Bayer and Tavits Reference Schwindt-Bayer and Tavits2016), while the two equality theories identified by Pevnick can be linked with the proportionality and clarity dimensions of systemic voice. The equality of citizens' influence can also be relevant to public justification and, thus, collegiality. This equality promotes a more diverse range of views in parliament, subjecting the government to justificatory pressure. Collegiality might also be enhanced through representation and ideological moderation. More generally, we believe that it is possible to link each of the first-order principles that feature in political finance debates with one (or more) of the dimensions of the normative theory of partisanship, which demonstrates our point that such a theory distils first-order principles of political finance (and, as we explained earlier, of democratic theory) into a coherent framework for assessing political finance in practice. As a tool for this purpose, the normative theory of partisanship integrates a wider range of concerns than other normative approaches; for instance, it includes aspects of collegiality that go beyond Pevnick's (Reference Pevnick2016) account (that is, non-violence and civility), making it well-suited for comparative analyses.

Table 3. Applying the normative theory of partisanship to political finance

While we use this framework to analyze a variety of political finance arrangements, there is no consensus on how to classify such systems. The most recent efforts to do so highlight the source of funding as a major distinguishing criterion (for example, Koß Reference Koß2010; van Es Reference van Es, Norris and van Es2016; Wiltse, La Raja, and Apollonio Reference Wiltse, La Raja and Apollonio2019).Footnote ⁵ Therefore, we consider models based on public, private, and mixed funding. As with our treatment of party systems, this schema simplifies a complex reality to make it amenable to systematic normative analysis.

Public Funding

We first consider a model in which political financing is dominated by the state. According to Ohman's (Reference Ohman2013) global survey (spanning 180 countries), most countries (66 per cent) provide state subsidies to political parties for election campaign purposes. This figure has been rising over time, driven by the practical fundraising challenges faced by parties as their memberships decline and the belief that public funding can reduce corruption (cf. Casal Bértoa et al. Reference Casal Bértoa2014). Generally, the state refunds parties based on their expenditure and subject to a minimum electoral performance threshold. The extent of state generosity – how much is disbursed and what activities it covers – varies considerably, as does the relative importance of state subventions compared to private donations (van Es Reference van Es, Norris and van Es2016). For example, the state supplies the vast bulk of funding in countries like Israel, Germany, and Sweden. Idealizing from such cases, we assume a model in which parties receive all their income from state coffers (while all other potential revenue sources are banned) and the state allocates funds to parties in proportion to their electoral popularity based on a low threshold – for example, gathering a certain number of signatures (McCabe and Heerwig Reference McCabe and Heerwig2019, 326).

We expect that collegiality will be high under this model, for two reasons. First, public funding reduces barriers to political participation and, specifically, citizens' ability to form political parties (Scarrow Reference Scarrow2006). Second, public funding limits competition between parties to policy issues, thus preventing a financial ‘arms race’ (van Biezen Reference van Biezen2008). Other things being equal, both factors are likely to increase the number and range of parties in parliament. With a larger number of parties in parliament, collegial practices such as negotiation and compromise are likely to become more attractive, and the justificatory pressure on governing parties will be heightened because those parties will need to defend and explain their policies in terms that take into account a broad range of interests and demands.

We believe that systemic voice is likely to be both clear and proportional, and therefore high, under this model of political finance. Clarity will result from the fact that fundraising obstacles are low, and citizens can more easily form, join, and vote for parties whose manifestos closely reflect their preferences (Scarrow Reference Scarrow2006). But voice will also be proportional because, given the idealized conditions that we assume, nearly all citizens can access funding to create their own parties; therefore, voice will not be monopolized by a few powerful private donors.

We expect systemic accountability to be low under the public funding model because elected partisans in this system will not be obligated to financiers who could sanction them by reducing their financial support. A risk of this detachment of parties from societal funding sources is the pathology of cartelization (Katz and Mair Reference Katz and Mair1995), whereby parties come to prioritize the maintenance of ‘club’ benefits for the political elite and neglect their constituents' demands.

Private Funding

We also consider the scenario in which the state plays no role in funding parties and campaigns. This model is mainly seen in economically developing countries (Wiltse, La Raja, and Apollonio Reference Wiltse, La Raja and Apollonio2019), where fewer resources are available for public funding of parties, and the norms that fueled its spread in western democracies in the 1970s and 1980s have yet to take root (van Es Reference van Es, Norris and van Es2016, 226). Parties in the US receive money from the state, but private donors play a greater role in political financing in that country than in most others (La Raja and Schaffner Reference La Raja and Schaffner2015). More generally, in the absence of state subventions, parties will likely rely heavily on private donations from, for example, corporate entities, rich individuals, or a large group of small contributors. Sometimes, the ‘fat cats’ may give to all major parties to hedge their bets (Tham Reference Tham2010).

Our analysis treats the relationship between parties and such external donors as the main mechanism of non-state-funded political finance, even though some parties may also be able to fund a large part of their activities through other means (for example, party membership dues, investment vehicles, etc.) (Hopkin Reference Hopkin2004), and private money may be donated to candidates instead of parties (La Raja and Schaffner Reference La Raja and Schaffner2015). Furthermore, while private funding may be constrained by caps that exclude fat cats, we assume (for simplicity) a private system without caps (and, therefore, its likely domination by a small number of donors who donate large sums).

Collegiality, we believe, is likely to be low under this model. This is because campaign financiers are likely to exercise significant pressure on elected representatives, encouraging them to advance their sectarian interests, therefore hindering the process of public justification that is central to collegiality. Moreover, competition for donations is likely to compound competition over policy (Katz and Mair Reference Katz and Mair1995), creating the conditions for non-collegial interactions among partisans from different parties. Collegiality within parties may also be low (or at least lower than it would be otherwise) if private donors regularly give money to individual candidates and not just parties because this is likely to introduce an intra-party aspect to the competition for funding (La Raja and Schaffner Reference La Raja and Schaffner2015).

Under the private model, systemic voice is likely to be clear only for those who can afford to donate large sums, who will be able to capture the policy-making process and ‘crowd out’ a wider range of perspectives (Ackerman and Ayers Reference Ackerman and Ayres2004; Cagé Reference Cagé2020). As a result, systemic voice is likely to be non-proportional and thus low. Of course, systemic voice could potentially be high if, instead of the fat cats model which we assume, private funding came from membership dues and/or from a large number of small donations.Footnote ⁶

Similar considerations, we believe, also apply to systemic accountability. More specifically, systemic accountability in a private model dominated by fat cats will tend to be low because parties and partisans will only be accountable to a small group of large donors. But, again, the private model's performance in this dimension would be higher if the distribution of financiers approximated the distribution of voters across the party system.

Mixed Funding

We also consider three mixed models that, in our view, combine public and private elements in distinctive and feasible ways. In the first of these models (dual-source funding), parties receive funding from state and private sources, with varying degrees of state and private contributions. This model captures most of the political finance arrangements we see in practice (Ohman Reference Ohman2013; van Es Reference van Es, Norris and van Es2016; Wiltse, La Raja, and Apollonio Reference Wiltse, La Raja and Apollonio2019). Exemplifying a second kind of mixed model, Ackerman and Ayres (Reference Ackerman and Ayres2004) propose a scheme whereby the state is the only funding source. It provides citizens with vouchers (public component) that citizens can then distribute to parties and candidates as they choose (private component) (see also Cagé Reference Cagé2020; McCabe and Heerwig Reference McCabe and Heerwig2019). Building on this idea, we consider a third mixed model that also combines public and private financing principles, though in the opposite way. In this case, private donors are the only funding source (private component). However, they must donate to parties via an electoral commission or, as Ayres and Bulow (Reference Ayres and Bulow1998) suggest, a system of blind trusts through which the identity of donors and the size of individual donations are anonymized before being transferred to the proper recipients (public component).

Mixed Type 3 – Anonymous Donations Scheme

Under the anonymous donations scheme, we expect a high level of collegiality. Assuming that the scheme guarantees strict anonymity, large donors cannot credibly demand or expect anything from politicians in return for their donations. Competition for donations is unlikely to compound competition over policy via a financial arms race and thereby encourage non-collegial interactions. Furthermore, the anonymization element of this mixed model can contribute to public justification by preventing a few private donors from capturing the policy-making process to advance their sectarian interests.

When it comes to systemic voice, we expect a medium level in this dimension for reasons that echo our arguments in relation to the private model. That is, systemic voice could be high if the distribution of (anonymous) financiers approximated the distribution of voters across the party system. But in an unequal society (assuming no caps on anonymous donations), some parties and candidates will likely receive much more than others. Nevertheless, thanks to anonymity, we also expect that voice will be more proportional under this model than would be the case in the private model since very wealthy donors will be less able to directly influence party manifestos.

Finally, we expect that systemic accountability will be low under this model since parties and elected partisans will not know to whom, precisely, they are accountable. However, while anonymity undercuts accountability in some ways, it at least does not promote the perverse kind of accountability that characterizes the private model, where certain parties or candidates are captured by a few (non-anonymous) private donors.

Table 4 below summarizes our assessment of the political finance models in terms of the dimensions of the normative theory of partisanship.

Table 4. Political finance models and the dimensions of the normative theory of partisanship

The Civic Model of Political Finance and Its Dynamics

Having considered the various party system types and political finance models against the normative dimensions of partisanship, we now draw together these lines of analysis to rank order their different combinations. Our goal, ultimately, is to identify the most promising political finance arrangement for each party system type. We describe this series of ideal (or quasi-ideal) pairings as the civic model of political finance. As this framing of the problem suggests, our analysis holds constant the party system while treating political finance as a variable. Indeed, we assume that political finance arrangements are more alterable than party systems, which is a plausible assumption.

First, while party systems sometimes change fundamentally in established democracies, this seems to happen not as regularly as major political finance reforms (see, for example, de Sousa Reference de Sousa2001). Second, and more fundamentally, party systems are rooted in change-resistant social and institutional structures. Lipset and Rokkan (Reference Lipset and Rokkan1967) pointed to their mooring in class and religion. Though these traditional cleavages have become less significant in many countries, the patterns of competition they established often remain locked in long after a society has changed, for example because those with the power to change the electoral system – the other determinant of party system type – do not want to see any real change. Cases such as Portugal suggest that elites may reform political finance precisely to deflect attention from the wider failings of the party system (de Sousa Reference de Sousa2001).

At this point, one might argue that political finance rules may sometimes influence party systems; for example, public funding can help small new parties survive and produce ‘at least a slight diversification of the party system’ (Scarrow Reference Scarrow2007, 204). However, even if this is the case – and the matter can be disputed since studies suggest that public funding has uncertain effects on party systems (see, for example, Pierre, Svåsand, and Widfeldt Reference Pierre, Svåsand and Widfeldt2000; Scarrow Reference Scarrow2006) and that it can sometimes actually help to sustain the status quo (Paltiel Reference Paltiel and Malbin1980) – our framework is not undermined. Instead, in this scenario, policymakers should consider this potential effect when designing political finance rules and be prepared to revisit and potentially modify those rules should the party system change.

Towards Balanced Collegiality

So, what setup should be selected when the opportunity arises to reform a country's political finance arrangements? We argue that the answer depends on the party system in place at the time and, specifically, its qualities and deficiencies in terms of the three dimensions of the normative theory of partisanship. The party system deficiencies on the horizontal dimension can take two forms. When collegiality is too weak, inter-party relations can become problematically conflictual, while the opposite excess, the pathology of cartelization, occurs when parties become too collegial or ‘cozy’ with each other (Katz and Mair Reference Katz and Mair1995; Scarrow Reference Scarrow2006). Thus, there are intuitive reasons to defend the location between these poles, where the system is responsible, that is, providing the ‘best basis for public-regarding policies’ (Rosenbluth and Shapiro Reference Rosenbluth and Shapiro2018, 236).

The design of political finance can move a political system toward this ideal if it includes public funding when the party system generates a low level of collegiality, as does the concentrated party system. The voucher scheme would also be a sensible choice for a concentrated party system as it would foster collegiality, as would the anonymous donations scheme. A privately funded model would exacerbate the conflictual tendencies already present in this type of party system. Private funding can curb collegiality when the latter risks becoming excessive, so it is most appropriate for a moderate pluralist system. A dual-source funding system can match the moderate level of collegiality associated with polarized pluralism. By contrast, private funding, on the one hand, and voucher and anonymous donations schemes, on the other hand, may push the system toward too little or too much collegiality.

Towards Balanced Systemic Voice

The systemic voice dimension can also be presented along a spectrum. At the low end, when systemic voice is weak and attenuated, parties do not adequately capture and project the views of citizens, and the system is relatively detached from ordinary citizens. As explained earlier, this is an inherent risk in concentrated party systems, which political financing can exacerbate (for example, see Rosenbluth and Shapiro Reference Rosenbluth and Shapiro2018, 249). Though a higher level of voice than this is desirable, it is also possible that this function will be amplified too much, and that the party system will become hypersensitive. In this scenario, which is characteristic of polarized pluralism, ‘government is … hypertuned to microlevel needs yet lacks the capacity or incentives to deliver broad, sustained, high-level performance’ (Kettl Reference Kettl2015, 70).

When it comes to systemic voice, the ideal lies between these poles in an expressive system that captures and communicates citizens' opinions (to a large extent) without creating the expectation that the private views of every citizen will be fully represented in the party system. These conditions are approximated in the workings of a moderate pluralist system. Hence, under moderate pluralism, either a dual-source model or the anonymous donations scheme can help to preserve a moderate level of systemic voice. By contrast, the other political finance arrangements might disrupt this balance in either direction. Public funding can boost voice in a concentrated party system, while private funding will tend to exacerbate the system's inherent problems in this area. The various mixed funding arrangements, especially the voucher scheme, will also improve the concentrated system's performance in terms of systemic voice. Private funding and, to a lesser extent, the dual-source and anonymous donations arrangements can help move a polarized pluralist system in a more balanced direction in the systemic voice dimension. By contrast, the public funding and voucher scheme models tend to introduce or solidify a situation of excessive systemic voice.

Towards Balanced Systemic Accountability

Finally, let us consider the normative positions along the systemic accountability dimension. When systemic accountability is substantially lacking, the democratic system is unresponsive since it fails to translate citizens' demands into governance outputs. This problem often accompanies the detachment pathology discussed in the previous section. However, as with the systemic voice sub-dimension, a political system can also display an excess of accountability, resulting in a ‘pandering’ scenario (Gersen and Stephenson Reference Gersen and Stephenson2014, 188) in which politicians try to constantly appease their constituents, reducing the system's capacity to effectively deal with collective dilemmas and address voters' long-term interests (Rosenbluth and Shapiro Reference Rosenbluth and Shapiro2018). The balanced position between these poles is a responsive system that satisfactorily delivers on citizens' demands without creating the expectation that every private demand should cause a change in government policy.

Public funding, private funding, and two of the three mixed systems (the dual-source model and the anonymous donations scheme) can help counter the tendencies of an excessively accountable party system, such as our idealized conception of the concentrated party system. By contrast, the voucher scheme would likely worsen the properties of the party system in this dimension. However, since the voucher scheme will increase accountability, it can compensate for the deficiencies of polarized pluralism in this area. Our analysis suggests that there is no perfect match in this dimension for the already well-balanced tendencies of moderate pluralism. Instead, assuming that a party system of this type is in place, the various political finance arrangements that we consider will tend to weakly push the democratic system in the direction of either too much or too little accountability.

The Civic Model of Political Finance

Bringing together these observations, we can identify the most promising political finance arrangements for each party system type, a series that we describe as the civic model of political finance.Footnote ⁷ As this suggests, the model is context-dependent. It is not represented by one financing arrangement across all circumstances but by the best-fitting political finance arrangement(s) given the properties of the party system. This financing setup can help counter the party system's dysfunctional tendencies, in the right direction and to an appropriate extent, resulting in higher-quality democracy – understood through the lens of the normative theory of partisanship – than would otherwise be the case (i.e. if some other financing arrangement were to be used). When this condition is satisfied, the party system and the best-fitting political finance arrangement(s) combine to generate a balanced profile along the three normative dimensions.

As Table 5 below shows, given the existence of a concentrated party system, public funding as well as the anonymous donations schemes would yield a balanced profile. Private funding, by contrast, would consolidate two of the inherent pathologies of a concentrated system – conflictuality and detachment – whereas the voucher scheme would produce a pandering scenario. The other financing setups would ameliorate this system's problematic tendencies without fully counterbalancing them. Therefore, these political finance arrangements would have what might be termed a weakly balanced effect. For example, the anonymous donations scheme would counterbalance the systemic accountability weaknesses of a concentrated party system but would only partially offset its systemic voice deficiency.

Table 5. Tendencies of party system type and political finance model pairings

None of the financing systems that we consider is a perfect complement to moderate pluralism, as shown in Table 5. This is because moderate pluralism has a mostly balanced democratic complexion to begin with, so the operations of a political finance arrangement can distort the inherent balance of this party system. However, the distortion will not be equivalent across all the financing setups. The least distorting among them will be the private and dual-source models, which each generate a weakly balanced tendency when used in the context of moderate pluralism. Instead, when combined with moderate pluralism, public funding as well as the vouchers and anonymous donation schemes will result in dysfunctional scenarios. In all three cases, there is a high risk of cartelization.

Polarized pluralism lacks a complementary political financing arrangement. Four of the financing setups that we consider are likely to consolidate a serious system shortcoming; that is, unresponsiveness. The voucher scheme could compensate for this deficiency but will likely consolidate the system's weakness in the systemic voice dimension, resulting in a hypersensitive tendency.

A striking conclusion from this analysis is that there is no perfectly complementary (that is, yielding a fully balanced profile) political finance arrangement for two of the three party system types that we consider. This conclusion somehow testifies to the public frustration that often surrounds discussions about political finance reform in contemporary liberal democracies. Yet, while our framework may not offer perfect solutions, it provides normative guidelines for reforming existing political finance arrangements to avoid worst-case scenarios. This can help set normatively favourable terms for debates about reform, presuming that policymakers and citizens alike decide to concentrate their attention on choosing between the various political finance arrangements that would result in a balanced or weakly balanced outcome.

Table 5 includes examples for most of the party system/political finance pairings alongside the evaluative descriptions derived from our framework. These are identified using Koß's (Reference Koß2010, 17) mapping of twenty countries' political finance arrangements on the basis of main income source, which can be used to identify examples of public funding, private funding, and dual-source models. We classify the party systems of these countries on the basis of their recent election results and following Sartori's criteria. To supplement Koß's mapping, we also include a few US States to capture some (though not all) of the empirically rare scenarios that feature prominently in our theoretical analysis. The positioning of a country (or state) beside an evaluative description should not be interpreted as our judgement that its politics matches that description. Rather, it suggests that due to its party system and political finance characteristics, the country could be a useful case study for examining the conjectures underlying that evaluative description.