Compromise and Consensus: Toward a Two-Track Model of Deliberation

Abstract

In this article, we discuss the relationship between deliberative democracy, compromise, and consensus. According to one view, deliberation should aim exclusively at compromise (compromise-only view). According to another view, compromise processes are incompatible with with the epistemic logic of deliberation (no-compromise view). Against the compromise-only view, we argue that it relies on a misunderstanding of the role of consensus and fails to capture how consensus-oriented deliberations are conducive to important learning processes. Against the no-compromise view, we argue that it mistakenly envisages deliberation in purely epistemic terms and that this misunderstands the point and purpose of deliberative democracy as a practical political ideal. By adopting a distinct second-order perspective, we demonstrate how compromise processes can be rendered compatible with a deliberative orientation to finding the best political outcomes. Based on these discussions, we articulate a two-track model of deliberation that shows how deliberative democracy can be dually committed to compromise and consensus.

Introduction

On the Danish broadcast, Deadline, the now-retired politician and member of Venstre (Denmark’s liberal party), Karsten Lauritzen, offered the following reflections on democratic politics:

politics is great when we strike the good compromise. This is achieved by getting together and discussing in an orderly and respectful fashion, and then listening to one another. Then we find what I consider to be the best solutions. Finding the best political solutions neither implies that I [read: “my party, Venstre”] get to decide everything, nor that the Social Democrats do.

The statement is thought-provoking for at least two reasons. First, it flies in the face of the conventional and widely accepted definition of compromise, according to which compromises are delineated as sub-optimal outcomes from the viewpoint of the compromising parties. Indeed, from this perspective it appears puzzling that Karsten Lauritzen would describe compromises as “the best solutions,” knowing quite well that such solutions, by definition, will diverge from and only partially realize the political ideology that he has fought for throughout his political career. Second, by stressing the need for respectful discussion and listening, he firmly situates compromise within a deliberative understanding of politics. Again, this might strike us as surprising, given that a compromise situation is commonly depicted as a process in which parties grant concessions to each other based on their relative bargaining power—as opposed to a deliberative process in which participants evaluate different positions on their merits, with the aim of arriving at the substantively best political decision.

In this article, we have several aims. One is to vindicate Lauritzen’s claim. We show that compromises can be viewed—in the here and now—as the better outcome from the perspective of the compromisers, provided they have been struck under favorable deliberative conditions. To make this argument, we attempt to establish two claims. The first follows a particular second-order logic: political actors recognize that it would be counter-productive to have their first-order political judgments unilaterally implemented into law because the purported positive effects of the law, to which the first-order judgment correspond, are conditional on meeting practical conditions of general acceptability that depend on the conflicting political attitudes of a pluralist public, where these conditions of political pluralism are an intrinsic feature of democratic societies. The second addresses a problem that is commonly associated with compromises—namely that they are constituted through an arbitrary composition of clashing political judgments and interests, thus lacking a coherent rationale with which citizens can identify—by arguing that deliberation (or deliberative negotiation) offers a response to this problem by facilitating a collective discovery of policies that combine conflicting positions in ways that the different parties to the compromise have reasons to endorse given their first-order judgments and acknowledgment of the importance of public acceptability.

It is an important feature of our argument that reconciling deliberative democracy with compromise does not jeopardize the traditional understanding of the deliberative ideal. This jeopardy claim can be advanced in two ways—the compromise-only view and the no-compromise view—and we provide reasons to reject both versions. The compromise-only view claims that deliberation should aim at compromise at the expense of consensus. We argue that the criticism of consensus that motivates this call for exclusively compromise-oriented deliberation rests on mistaken premises and show that renouncing the ideal of consensus either relies on misunderstandings concerning the role of consensus within the deliberative democratic paradigm or fails to capture how consensus-oriented deliberations are conducive to learning processes—learning processes that critics of consensus and defenders of compromise-oriented deliberations themselves rely on when they claim that there are certain views with which one should not compromise. The no-compromise view states that compromise is incompatible with the rationale of deliberation. We show that this view takes the form of envisaging deliberation in purely epistemic terms as a commitment to truth and that this misunderstands the point and purpose of the ideal of deliberative democracy as a practical political ideal. We then proceed to demonstrate how compromise processes, pace the no-compromise view, are compatible with the deliberative ideal properly conceived.

On the basis of our rejection of these two views, we articulate a two-track model of deliberation that aims to show how the deliberative democratic paradigm can be dually committed to compromise and consensus. As we will demonstrate, this two-track model enables us to appreciate the different attitudes of political actors (making a good political decision under conditions of disagreement vs winning the battle of hearts and minds), the different temporal perspectives (short-term vs long-term), and the different sites in the deliberative system that pertain to compromise- and consensus-oriented deliberations, respectively. ¹

Deliberative Democracy

Deliberative democracy is a paradigm that places emphasis on inclusive and rational discussion among citizens as a condition for legitimate decision-making. In deliberation, all participants are included on equal footing and committed to the norms of mutual justification, which require that they state their views truthfully and offer reasons in support of them in a way that renders their claims to correctness or rightness intelligible and reasonable, even to those who may not believe that they are the best ones when judged purely on their merits carefully (Wolkenstein, 2020: 108). Participants are also expected to listen to others and display a willingness to change one’s views in the light of better reasons.

Deliberative democrats thus conceive of democracy as much more than just the aggregation of individual preferences during elections and decision-making by means of majority rule. They recognize that citizens’ preferences are endogenous to the political process (Rostbøll, 2008: 80) and envision them as products of public deliberation, where citizens exchange their views and opinions on political matters. Thus, in deliberative democracy, the quality of opinion- and will-formation in the public sphere is a central concern, with the aim being that citizens‘ opinions are transformed into „considered public opinions” and channeled into the center of formal decision-making (Lafont, 2019: 171–175). This is intended to ensure that citizens are not only subjects of the laws but also in some sense authors (if only indirectly), as required by the ideal of democratic self-government (Lafont, 2019: 17–32). This discursive approach, moreover, reflects the importance that deliberative democrats attribute to the epistemic dimension of democracy. Political deliberation should evince the argumentative search for and weighing of reasons such that “the force of the better argument” can come to the fore and help to shape collective decisions (Habermas, 2006: 413). Thus conceived, deliberative democracy is a paradigm concerned with enabling and securing the conditions for rational and inclusive communication.

One criticism of deliberative democracy is that the ideal of deliberation is too demanding and unrealistic. Deliberation—which, as we have seen, requires that one states one’s views truthfully, offers reasons to support them, carefully listens to others, and is willing to change one’s views in light of better reasons—may seem overly demanding. When we look at interactions in real politics, citizens oftentimes fail to live up to deliberative ideals. In part to accommodate these realist concerns, deliberative democratic theorists have developed a systemic approach to deliberative democracy (Kuyper, 2015; Mansbridge et al., 2012; Neblo, 2015; Parkinson, 2018), elements of which were already present in Habermas’ more holistic account, briefly described above. According to this approach, deliberation is not exclusively conceived of as a singular and synchronic event or process but also as a complex amalgam of communication that involves different actors and occurs at various sites and institutions within the political system. There may be many deliberative moments which, when considered in isolation, fail to live up to the deliberative ideal, but, when considered from the perspective of the overall political system, can serve the overall ideal of bringing reasons to bear on politics. In other words, single instances of communication do not have to live up to the ideal standards of deliberation of inclusiveness, equality, and rationality, as long as the broader system displays the capacity to approximate these ideals. Thus, the systemic approach not only provides a basis for responding to realist concerns, it also offers a useful framework for analyzing different existing sites, practices and actors based on the deliberative functions they serve for the political system as a whole. The two-track model of deliberation that we articulate in the final section can be read as a contribution to this approach insofar as it attempts to situate consensus and compromise oriented-deliberation within the different sites, practices, and institutions of the deliberative system.

A final important point concerns the role of consensus within the deliberative paradigm. Deliberative democrats have characterized deliberation as being oriented to consensus or rational agreement in the sense that participants in deliberation, once they raise political claims in discursive exchanges, implicitly must assume that these claims can be assented to on the basis of reasons adduced in support of them. It is this inescapable commitment and orientation to the correctness of the claims that people raise in political discourses that ties the deliberative paradigm to the notion of consensus as “a matter of conceptual necessity” (Martí, 2017: 560; see also Habermas, 2018: 873; Finlayson, 2005: 34–35).

The Criticism of Consensus and the Compromise-Only View

The deliberative democratic paradigm has been criticized from a number of different theoretical standpoints. In particular, the idea of deliberation as being committed or oriented to consensus has been a recurring target of criticism, not only among theorists who reject the paradigm altogether (Mouffe, 1999) but also among more sympathetic commentators who seek to fundamentally revise it, either by significantly downplaying or altogether rejecting the relation between deliberative democracy and the notion of consensus (e.g. Chambers, 2003: 321, 2017; Mansbridge, 1999; Warren, 2007). The criticism leveled against the notion of consensus has come in different variations. Some argue that it makes the deliberative democratic paradigm susceptible to a particular form of ideological domination because the search for consensus can exert hegemonic pressure on marginalized groups in society to conform to dominant majority opinions (Przeworski, 1998). This concern led Jon Elster (1983: 40) to claim that he, in fact, would have more confidence in a democratic decision short of full consensus (a decision with which a minority would continue to disagree) than one with which everyone appeared to agree. Others have raised concerns about the feasibility of aiming for consensus under conditions of pluralism and deep disagreement (Bellamy, 1999; Waldron, 1999). If disagreement is a fundamental fact of politics to be reckoned with rather than transcended, then it seems implausible, so the argument, to stipulate consensus as the aim of deliberation. Aiming for consensus against a societal background of insurmountable pluralism and political conflict is seen as futile according to this line of criticism.

These issues have led a number of democratic theorists to argue that democratic politics and deliberation should settle for and be aimed at compromise rather than consensus (e.g. Bellamy, 1999: 110). In a recent piece, Weinstock (2017: 648) puts it this way:

Better, it would seem at first glance, not to attempt the impossible, and to satisfy oneself with solutions to political problems that do not require of any party that they change their minds as to the questions of fundamental value.

He backs up his call for compromise by articulating a different form of deliberation oriented to compromise. In compromise-oriented deliberations, participants will no longer pursue first-order evaluations of the merits of the various positions advanced in the course of their exchange but rather use deliberation to reduce the insurmountable differences that exist between them (Weinstock, 2017: 643–644). Deliberation so understood will be less centered around finding the correct answer by having “the uncoerced force of the better argument” prevail and more around accommodation between conflicting positions. This compromise-only view would, it seems, avoid the aforementioned problems of ideological domination and unfeasibility associated with deliberation aimed at consensus. It does not exert pressure to conform because compromises do not require that actors come to substantive agreements on the policy issue under consideration. The disagreement at the level of the first order remains. For the same reason, it also does not run into the problem of feasibility. As a realistic collective response to the fact of disagreement, a compromise gives everyone something in return for the losses that compromisers incur, rather than unfeasibly requiring that everyone substantively agrees to one thing (consensus).

Although we share the vision of making compromise consistent with the logic of deliberative democracy, we believe that the almost full-scale rejection of deliberation oriented to consensus runs into problems that make the compromise-only view unattractive. Below we list three of the most important problems, which will be taken up again in the last part where we spell out a two-track model of deliberation.

Problems with the Criticism of Consensus and the Compromise-Only View

The first problem is that defenders of the compromise-only view tend to underdetermine exactly what is meant by the proposition that deliberation should be almost exclusively oriented to compromise rather than consensus. In the case of Weinstock, his repeated mentioning of “decision making” could suggest that he is exclusively concerned with formalized and decision-oriented deliberations. While it may be plausible to hold that deliberation should generally (though not necessarily always) be oriented to compromise (rather than consensus) in synchronic instances of decision-making in parliaments, it is questionable whether deliberative democrats actually defend the opposite claim; that parliamentarians by default should aim at consensus in their daily legislative work. In a decision-making context, in which decision-makers of different political orientations come together, it seems hopelessly naïve to expect consensus to be forthcoming, which is why deliberative democrats are willing to settle for something less in these contexts (see, for example, Habermas, 1996: 165). The criticism of consensus which motivates the compromise-only view actually rests on a common misunderstanding. It suggests that since deliberative democrats claim that deliberation in some sense is oriented to consensus (in the above sense that actors participating in political discourses are oriented to the correctness of their claims), it must follow that they solely endorse decision-making by consensus. However, this is not a view that can be attributed to deliberative democrats and it is also unclear why the latter position would follow from the former one (cf. Lafont, 2019: 38; Martí, 2017: 60–61; Warren, 2017: 48). Deliberation oriented to consensus should be situated at another level, which brings us to the second point.

Another problem with the compromise-only view is that it relies on a mistaken view of the relation between the deliberative orientation to consensus and pluralism and disagreement, suggesting that the former runs counter to the latter (Shapiro, 1999: 31–32; Urbinati, 2014: 99). As several deliberative democrats have noted, this picture fails to capture how deliberation oriented to consensus actually feeds rather than depreciates conflict and pluralism:

[T]he assumption that political discourses are also oriented to the goal of reaching an agreement in no way implies an idealistic conception of the democratic process as a peaceful university seminar. On the contrary, it is safe to assume that it is precisely the participants’ orientation to the truth or rightness of their beliefs that first adds fuel to the fire of political disputes and lends them a polemical character. To argue is to contradict (Habermas, 2018: 873; see also Lafont, 2019: 63; Rostbøll, 2008: 24–25).

Thus, the relationship between deliberation oriented to consensus and pluralism and conflict should be construed differently: it is because rather than despite the orientation to consensus that politics comes to be imbued with persistent conflict and disagreement in contexts of pluralism. In other words, it is because political actors genuinely consider their first-order political positions to be worthy candidates of substantive endorsement, that they show up to defend them against criticisms and object to those views and laws that contradict these positions. “To argue is to contradict” and contradictions are the very stuff of which disagreement is made. ²

The final problem we would like to highlight is that the compromise-only view is internally unstable. Here is why. Most defenders of compromise admit that certain substantive considerations trump the value of compromises that we might otherwise ascribe to them (Margalit, 2009; Overeem, 2017: 121; Richardson, 2002: 152). As Weinstock (2017: 652) puts it, some views “may be so odious that there is no question of compromising with them.” He is certainly right about this point. It is difficult, however, to square this commitment with his (almost) full-scale rejection of deliberation as being oriented to consensus. The question that looms large is the following: how did a widely shared agreement on the odiousness of certain positions come about, if not because past political actors were oriented to consensus in their political struggles? Why, for instance, is it the case that we nowadays are unwilling to compromise with unashamedly patriarchal views? Presumably, it is because political actors (minorities at the time) were unwilling to satisfy themselves with solutions that would “not require of any party that they change their minds as to questions of fundamental value” (Weinstock, 2017: 648) that we eventually came to have a (broadly) shared view on matters of gender equality. For these processes to succeed it was actually of crucial importance that citizens “changed their minds as to questions of fundamental value” (Weinstock, 2017). This goes to show that the orientation to consensus can be conducive to learning processes and, by extension, that the rejection of that orientation would be impervious to such processes. If we relate these considerations back to the issue of odiousness of compromises, we find that the problem for Weinstock is that he treats this demarcation criterion (odiousness) as a pre-deliberative constraint, thus failing to appreciate how it is genetically dependent on a deliberative orientation to consensus with which his arguments seek to dispense. ³ This makes the compromise-only view internally unstable.

These considerations suggest that the criticism of consensus either relies on misunderstandings or entails a position that, if internalized by citizens, would be deleterious to learning processes—learning processes that advocates of the compromise-only view themselves rely on when they claim that there are certain views with which one should not compromise.

The No-Compromise View

If deliberation is in some sense constitutively oriented to consensus, where does this leave us with respect to compromise? The relationship between deliberation and compromise has been subject to much controversy. Indeed, it is not difficult to see why compromise might be in tension with the deliberative ideal. As Gerald Gaus (1997: 231) notes, negotiation (which typically results in compromises) seems to be at odds with the truth-oriented nature of deliberation. His point is that deliberation cannot simultaneously display sincere reasoning and secure definite agreements (through compromises) because deliberation typically results in intractable disagreement rather than consensual agreement under conditions of pluralism for the very reasons we noted in the last section. The tension between deliberation and compromise solutions has also been recognized by deliberative democrats. Habermas (1996: 165f), for instance, conceives of compromise merely in terms of bargaining—a purely strategic, non-deliberative form of communication. In his view, compromise solutions are the outcomes of power struggles and thus can only be aligned with the deliberative ideal through an equal distribution of bargaining power, such that “fair” compromises can be struck (Habermas, 1996: 166f).

In a recent article, Ian O’Flynn and Maija Setälä (2022) have pushed the argument that compromise is inconsistent with the logic of deliberation. In their view, deliberation evinces an epistemic logic, according to which political decisions are to be based purely on their substantive merits (see also O’Flynn, 2022: 39–41). This presupposes a shared conviction among deliberators that political problems admit of epistemically correct answers and the function of deliberation accordingly is to steer participants toward these answers (O’Flynn and Setälä, 2020: 5). Compromise, by contrast, comprises an agreement based on mutual concessions, a splitting of differences between political adversaries. This distributive logic of compromises, so the argument goes, is unrelated to the deliberative search for the best political decisions (O’Flynn and Setälä, 2020: 9). Thus, their point

is not just that deliberation and compromise are different activities with different goals. Rather, it is also that there would be no good deliberative reason to opt for compromise as a means of dealing with the disagreement that deliberation leaves unresolved (O’Flynn and Setälä, 2020: 11).

In other words, the epistemic logic of deliberation seems incompatible with compromise. Call this the no-compromise view. In the following sections, we want to contest this view.

Rethinking Good Political Outcomes and the Role of Deliberation in Achieving Them

one will often bring about substantively worse results if one tries to implement a substantively better decision (. . .) than if one tries to implement a decision that most others agree with (Kolodny, 2014: 202–203).

An initial problem with the no-compromise view is that it relies on a one-sided understanding of how societies achieve good political outcomes. It gives the impression that the latter achievement is merely a function of the quality of the reasons and knowledge generated among deliberators. This becomes clear once we consider how O’Flynn and Setäla dissociate deliberation understood as justification to others from deliberation understood as being oriented to making the best decision, claiming the former to be subordinate to the latter:

There can, admittedly, be ancillary reasons for (or other desirable features) doing things deliberately, such as making people aware of why a particular decision has been made and making decisions in a way that makes them justifiable to all. But the primary rationale for deciding through deliberation is to make the best decision or judgment we can in the circumstances (O’Flynn and Setälä, 2022: 1, our emphasis).

Unlike O’Flynn and Setätä, we insist that deliberation aimed at “making people aware” of why decisions have been made and at decisions that are to some extent “justifiable to all” cannot easily be disentangled from deliberation oriented to making the best decisions (Forst, 2012: 185–186; Gutmann and Thompson, 2004: 45; Habermas, 2015: 48). In many cases, achieving good political outcomes requires that those outcomes become manifestly good in the eyes of those who are expected to comply with them (Lafont, 2006).

To see why, imagine a polity in which an expert council suddenly is given dictatorial discretion over public schooling policies. The expert council intends to use this power exclusively with a view to making the best decisions. The experts have gathered evidence about public schooling systems across the world, read many of the most important treaties on public education, and so forth. After lengthy and careful deliberation, they arrive at a final judgment that, from their perspective, has a strong presumption of correctness in its favor. Much to their frustration, however, their epistemically well-grounded judgment is out of sync with most citizens,’ and crucially also with parents’ and teachers,’ views on what good public schooling would require. Epistemically motivated by the aim of achieving the best public schooling policy, will the expert council dictate the implementation of their preferred school reform, in the awareness that it enjoys little popular support? We submit that they shouldn’t (given their motivation) because the good effects of the school reform will not materialize if met by resentment from those actors whose behavior and attitudes ultimately will determine the fate of the reform. If they went ahead, it is easy to imagine dissatisfied parents placing their kids in private schools and teachers becoming increasingly alienated from their profession, resulting in poor teaching performances.

What this shows is that attitude-based responses to political outcomes decisively affect whether they turn out to be good political outcomes. For this reason, the expert council should realize that turning a purportedly correct judgment on public schooling into a good political decision crucially involves demonstrating, through the force of arguments, the correctness of it to (enough of) those actors upon whose behavior its successful materialization hinges. The example also shows, against O’Flynn and Setälä, why deliberation understood as justification to others should not be seen as playing second fiddle to deliberation understood as being oriented to finding the best outcomes: the two are intimately linked. If this argument is sound, then it seems to follow that the goodness of outcomes not only depends on the epistemic quality of the reasons that underpin them but also on the extent to which they are aligned with the deliberatively refined beliefs of those expected to comply with them (cf Lafont, 2019). This perspective enables us to repudiate the charge that compromise is incompatible with the search for the best political decisions.

Rephrasing Compromise as the Best Political Decision

We now want to argue that compromises can be motivated by a distinct second-order judgment that renders them compatible with the search for the best political decisions: political adversaries recognize that it would be counter-productive to have their first-order political judgments (those they believe to be correct) unilaterally implemented into law because the purportedly good outcomes, to which their first-order judgments correspond, will not easily materialize if met by disapproval and resentment by their political counterparts who hold diverging first-order judgments. ⁴ A political actor may defend alternative A as the best political solution, say against B and C. However, if unsuccessful in persuading others—that is, if B and C continue to enjoy significant support—she will recognize at the second-order level that the best solution, here and now, consists of a reasonably worked-out compromise between the three political positions. This line of reasoning about compromise bears some resemblance to the arguments recently advanced by Fabian Wendt (2017: 50), whose justification of compromise is based on a similar second-order logic:

Assume you think that law X would be the morally best (most just, efficient etc.) law, while law Y is actually in place and widely accepted. My claim is that the fact that Y is widely accepted is a pro tanto reason to keep it in place and hence to make a compromise by supporting it in one way or another. The same holds when law Y is not actually in place, but is expected to be widely accepted if enacted, while law X is actually in place, but is not widely accepted. Again I claim that this fact provides a pro tanto reason to make a compromise by supporting or voting for Y, even though you think law X would be the better law.

The emphasis on the value of having “widely accepted laws” relates to our point that alignment oftentimes is required for achieving good political outcomes. However, in contrast to Wendt, our argument actually suggests that X should not be seen as the better law, even from the perspective of the political actor who defends the first-order judgment to which law X corresponds. These might sound like contradictory formulas, but the example of the expert council and the public schooling reform should give us reason to recognize that they are not: X should be viewed as the better law from the perspective of the political actor defending its corresponding content, only under the condition that she succeeds in demonstrating and justifying its goodness to enough of those political adversaries who hold it to be an inferior law. In other words, for X to be rationally judged to be a better law (and have its purportedly positive effects materialize) it must become a law that can enjoy wide acceptance. Concomitantly, if X cannot generate wide acceptance and a compromise Y is widely accepted because it aligns with the opinions of a broader segment of citizens, then compromisers have pro tanto reasons to conceive of Y as the better law, here and now, even if it diverges from and only partially realizes their first-order political opinions. Our argument thus challenges a core assumption in the literature on compromise which states that a compromise is perceived as sub-optimal from the perspective of the compromising parties (Baume and Novak, 2020a: 6; Bellamy, 1999: 102, 2012: 442; May, 2005: 318; Overeem, 2020: 49). According to the second-order approach adopted here, compromises can be viewed as the best outcome because they are more likely to be widely accepted to the extent that they align (if only partially) with the beliefs and interests of a broader segment of the citizenry (Overeem, 2020: 62) and, thus, by extension, their benefits are more likely to materialize as intended.

To substantiate this point, it is helpful to draw a distinction between the first-person evaluation of the quality of the content of the law and the first-person evaluation of the quality of the law as a collectively binding act. While the traditional definition of compromise seems correct in depicting compromises as sub-optimal in terms of the first-person evaluation of the content of a decision, it seems wrong, or at least problematically one-sided, once we consider the first-person evaluation of the decision as a collectively binding act. Indeed, under certain conditions, it is plausible for compromisers to conceive of compromises as the best decisions, for the time being, to the extent that they recognize that the diverging political judgments and opinions of others bear directly on the question of what the best political decision is for a given polity at a particular point in time. Thus, we can sum up this second-order appraisal of compromise as follows: what a particular political actor personally believes to be the most justified political judgment on a particular policy matter will not always coincide with what she believes would be the politically best thing to do for the society as a whole and in the here and now.

Against these considerations, one could argue that if compromises could constitute the best political outcomes from the perspective of the compromisers, how can they still be considered compromises? If Karsten Lauritzen believes that the best solutions for Denmark consist in striking good compromises rather than having his party get to dictate matters, in what sense would these solutions still count as compromises from his perspective? Here is the answer: they would still count as compromises because political actors will not forgo the burdensome deliberative task of trying to change the hearts and minds of citizens and have them all (or enough of them) end up supporting their first-order political opinions. However, they should recognize that to turn their purportedly justified first-order political opinions into good laws, it is necessary to justify them and render their correctness manifest to those expected to comply with them. Until then—that is, as long as significant political disagreement prevails on the relevant issues—it is reasonable for the compromising parties to believe that the best outcomes, here and now, consist in compromises that combine the conflicting elements into well-reasoned policies. In the final section where we articulate a two-track model of deliberation, we spell out these two different political temporalities in more detail. ⁵

Why Deliberation is Important for Striking Good Compromises

The above considerations support the view that compromises under certain conditions can be viewed as the best options from the perspective of the compromising parties insofar as they appreciate the significance of public alignment for good political outcomes, even if these options only lead to a partial realization of their first-order political opinions. There is, however, a significant blind spot in this argument that we have not yet addressed. The problem is that compromises would seem to consist of an arbitrary composition “of the range of views on what the right answer is”(O’Flynn and Setälä, 2022: 10), thus lacking an overall ratio with which political actors can identify (Goodin, 2008: 231–232). George Santayana (1922: 83) makes the point pithily when he notes that a compromise is “odious [. . .] to intellectual natures because it seems a confusion.” This issue is important because it could be read to defeat the aforementioned second-order reason for endorsing compromises as the best outcomes: alignment may be of little importance if the policy as a whole—constituted through the partial realization of the first-order opinions of the compromisers—lacks rational coherence.

At one level the incoherency charge is correct—and, perhaps, trivially so—because a compromise combines conflicting and contradictory principles. Thus, when judged from the perspective of one of these principles, a particular compromise inevitably will display a degree of incoherency (Boot, 2020; Gutmann and Thompson, 2012: 37). ⁶ However, it would be wrong to infer from this that compromises are entirely arbitrary and impervious to rational evaluation. To see why consider an example in which two compromisers champion conflicting first-order views on how to reform the public schooling system. Actor A defends a back-to-basics, conventional approach that places emphasis on rote learning and regurgitation, with the main objective being high test scores. Actor B is in favor of a progressive and more holistic approach to public education that stresses capacities for problem-solving and critical thinking. Now, if a compromise is reducible to an arbitrarily composed mix of diverging views—that is, if no meaningful attempt can be made to rationally combine the diverging views—then, in principle, nothing would stand in the way of a political compromise the realization of which would entail full implementation of the back-to-basics approach in year 1, followed by full implementation of the progressive approach in year two, and so forth. It would be as arational as any other combination. This certainly appears implausible. One does not have to be a believer in rational consensus to insist that there are better solutions (better combinations) out there than one that would require that teachers and students radically shift their methods, conduct and expectations on an annual basis. Compromises thus can be of better or worse quality. Moreover, good compromises do not announce themselves. While we may not want to go as far as saying that compromises can be truth-tracking (Dalaqua, 2020; Richardson, 2002) (whatever that means in the political realm), the mere fact that we can criticize or appraise them for being more or less practically workable, efficient, and optimal in terms of the partial realization of the various conflicting views and interests would seem to suggest that compromises admit of some degree of rational evaluation (Fumurescu, 2020: 20; Zanetti, 2022: 39). Judgments on different issues pertaining to a particular compromise will feed into this evaluation: Does it offer an improvement over the status quo in terms of welfare? Is the combination of the conflicting interests and principles implementable? Will it be effective? Is it optimal in terms of the partial realization of the conflicting interests and principles? Such questions are bound to generate disagreement. Deliberation, even if it fails to deliver coherency at the level of principles, can serve a genuine function in this context by steering compromisers toward better and more reasonable answers to these questions. Thus, we want to add an important proviso of our second-order appraisal of compromises: they can be viewed as the best outcomes, provided they have been struck under deliberative conditions that facilitate the attempt to work out how the conflicting positions can be combined into well-reasoned policies.

An example of what we have in mind can be found in the joint work of Mark Warren et al. (2013) where they develop the idea of deliberative negotiation which they argue is aimed at partially integrative agreements, inter alia. Deliberative negotiation refers to those interactions that are neither purely deliberative nor mere instances of bargaining, but which include elements of both modes. They are deliberative in the sense that they are based on “mutual justification, respect and reciprocal fairness” (Warren et al., 2013: 93); and they are strategic in the sense that they occur against a background of recognized conflicting interests and struggles for influence, with each actor trying to exert their interests on the final outcome. When deliberative negotiation culminates in partially integrative agreements, parties have managed to trade items (policies or parts of a policy) with each other based on their different valuations. One policy item that is of lower importance to one party may be of higher importance to another and vice versa (Warren et al., 2013: 95; see also Baume and Novak, 2020b: 83; Carens, 1979: 127). An exchange in that case would generate a joint gain while also contributing to minimizing the felt losses of the parties to the agreement. What we take to be their main point is that the mutual advantage that results from partially integrative agreements is facilitated by virtue of the deliberative features of the negotiation context. Conditions under which actors generally state their views truthfully, explain why what they want is important, listen to and try to understand the reasons that motivate the views of their adversaries (even if they do not share them), and display respect toward their counterparts enable the cooperative discovery of partially integrative agreements (see also Spang, 2021: 63–66). By contrast, if parties were to misrepresent their valuations, it is more difficult to see how such improved outcomes could be achieved because how could differently valued policy items be traded if actors can have no confidence that their counterparts are stating their valuations truthfully?

Partially integrative agreements provide a good example of what we have in mind when we claim that (good) political compromises ⁷ do not announce themselves but have to be discovered collectively in an (at least somewhat) cooperative framework. The possibility of achieving such agreements also adds additional support to the alignment argument that we advanced in the previous section. To the extent that negotiators negotiate on behalf of their constituents, the lower and higher values that they assign to different outcomes can be expected (roughly, of course) to map onto that of their constituents. Thus, when deliberative negotiation results in partially integrative agreements, citizens will mostly experience losses along dimensions they care less about whereas gains mostly will be experienced along dimensions they care more about. This may well be an important source for inscribing a ratio into compromise solutions—one that deliberative negotiation makes possible and that may otherwise be lacking. ⁸

These considerations add up to the conclusion that the no-compromise view is plausible only if deliberation is construed solely in terms of an epistemic logic of truth-orientation. However, we have no reason to endorse this as an adequate view of practical and political deliberation. If we move beyond this view (a view that implies that most decisions lack a deliberative rationale because most decisions are compromises), we find that there is a deliberative reason for compromise. It consists in recognizing the significance of disagreement as a constraint on how far a political community can go in terms of the policies and laws that it enacts at a particular point in time. In this sense, the best political decision at a particular moment oftentimes will consist in a compromise and thus will diverge from the true solution (if we assume that it exists). ⁹ Short of truth-orientation, deliberation or deliberative negotiation serves a genuine function of improving compromise processes by guiding compromisers toward outcomes that are more effective and easily implementable, that offer improvements over the status quo, and that minimizes the losses that accrue from compromises.

Toward a Two-Track Model of Deliberation

Against the background of the two discussions above, we now want to show that deliberative democracy should neither dispense with compromise nor with consensus and, in doing so, articulate a two-track model of deliberation that incorporates both. ¹⁰ To make space for compromise and consensus simultaneously within the deliberative paradigm, we need to appreciate that deliberation oriented to compromise and consensus, respectively, rely on different motivations and constraints and take place at different sites and practices in the deliberative system. When we think of compromise, we are immediately drawn into a decision-making context in which political actors take note of the fact of disagreement and accept that achieving good political decisions may imply, here and now, incorporating elements of the positions of their political adversaries into collective decisions. This suggests that the compromise-track pertains to the formal institutions at the center of the deliberative system, where binding laws are formed and enacted. Here members of adversarial political parties come together, typically behind closed doors, to work out how their different views can be integrated into compromises. As already mentioned, the aim of participants on this compromise-track is not primarily to change hearts and minds. ¹¹

However, deliberative negotiation will still display some of the key features that we traditionally associate with deliberation, such as the willingness to listen, to explain why what one wants is important, to state one’s position truthfully, and to display respect toward others. These features make it much easier for the compromising parties to reach decisions that combine their conflicting interests and views in a well-reasoned and optimal manner.

It is often claimed that compromises or coalition governments constituted through them sever the link of accountability between representatives and their constituencies (e.g. Kirshner, 2018: 293). Representatives typically strike compromises behind closed doors. This element of secrecy makes it possible for legislators to “hide behind” compromises so to speak, claiming that other parties to the compromise are responsible for unpopular elements of a certain policy, thus making it increasingly difficult for voters to hold their representatives to account. This points to the importance of having a deliberative system in place with discursive channels between representatives and their constituencies through which the rationale behind specific compromises can be transmitted. ¹² In other words, it must be possible for voters to get to know why certain elements of their political views were prioritized at the expense of others. Only under this condition will voters be able to endorse compromises as their own and, by extension, display the kind of amenable conduct upon which the successful materialization of the corresponding policy hinges. Legislative debates, it has been shown, provide a discursive channel conducive to this end, enabling parties to defend and justify compromises to their constituencies (see Martin and Vanberg, 2008). Other channels may include media appearances, public statements and press releases, social media, and so forth.

Turning now to the consensus track. By consensus, we refer exclusively to political actors’ orientation to the correctness of their political claims and the forms of political action that underpin this orientation. We do not mean to suggest that because political actors are oriented to the correctness of their claims, we should expect to see an increasing number of political issues to be settled through consensus (Habermas, 2022: xv; Lafont, 2021: 4), while the number of issues requiring compromise will correspondingly decrease. As already suggested by the cited paragraph from Habermas (2022: 6), this orientation to consensus actually amplifies political conflict and disagreement. So, where do we detect this orientation? We see it when party members campaign and offer arguments in favor of their political visions for society, trying to get as many citizens as possible to see the correctness of these visions and rally behind them. We see it when voters make choices between these proposed visions according to their considered judgments. We see it when young Friday for Future protesters march in the streets, attempting to exert public pressure on legislators to pass much-needed legislation on climate change (see also Gutmann and Thompson, 2012: 42). And we see it in everyday political talk in which discussants try to persuade each other about the correctness of their political positions (Mansbridge, 1999: 211–239). These forms of political action and activity occur at different informal sites within the public sphere (Bächtiger et al., 2018: 19). Although not qualifying as deliberation in the narrow sense, they can still, as made clear in section “Deliberative Democracy,” serve deliberative ends insofar as they contain propositional content and bring new and diverse voices and concerns into the political arena.

This would suggest that the consensus track can only be appreciated by adopting a capacious understanding of deliberation (Lafont, 2019) that is not limited to face-to-face deliberation which typically occurs in small groups. Deliberation according to this more capacious understanding briefly also refers to the anarchic and decentralized flows of communication that move between different arenas and agents in the public spheres as well as other parts of the deliberative system (Bächtiger et al., 2010: 35–38; Habermas, 1996: 287–387; Lafont, 2020: 106). It is through the connection and interaction between these different parts “that an anonymous public conversation” takes place (Benhabib, 1996: 74). Another point worth emphasizing in this context is that the consensus track, unlike the compromise-track, includes decision-free deliberations (Rostbøll, 2008: 187). The protestors, the discussants engaging in everyday political talk, and even the political campaigners are all relieved of the immediate burden of decision-making in their deliberative entanglements.

To get a better grasp of the distinct logics of these two tracks, it is helpful to consider the role of political parties that are unique in the sense that they operate on both tracks. On the one hand, political parties are vehicles of certain principled political projects that bring together people of shared political values and attitudes (Ragazzoni, 2017: 100). On the other hand, political parties are institutional actors, whose members come together, typically in parliament, to hammer out legislation, which in the great majority of cases comes in some form of compromise, given political disagreement (White and Ypi, 2016: 149). We can see the difference between these two roles and how they map onto the compromise- and consensus track by considering party members in their roles as legislators and campaigners, respectively. In their roles as legislators, their deliberations will be decision-oriented and constrained by a number of factors, one of which is disagreement and the conflicting opinions of others that we have focused on in this paper. Here, members of parties are operating on the compromise-track. By contrast, in their role as campaigners, the defense of principled political projects is foregrounded. They are less constrained by the abovementioned factors, as their deliberative activities (the attempt to justify and have more people rally behind their principled political projects) are relieved of the burden of decision-making: they defend their political projects, independently of whether they will be able to muster the kind of support that would enable them to fully implement them ex-post elections. Here, party members are operating on the consensus track. ¹³

Building on this, we also need to add a temporal dimension to tease out the difference between the compromise- and the consensus track. If we adopt a synchronic perspective and see how political disagreements and conflicts present themselves to us in the here and now, then in many cases aiming for consensus seems inappropriate, given that societies are under constant pressure to make decisions. This constraining feature thus explains the obvious appeal of compromise over consensus in synchronic instances of political decision-making. On the other hand, if we take a holistic and diachronic perspective that considers the complex interaction between different deliberative agents, sites, and practices over time, then the fact of disagreement becomes less significant, not in the sense that it disappears but in the sense that it assumes a less fixed and stable form. In this context, it helps to remember that a consensus, unlike a compromise, rarely is achieved in synchronic instances of deliberation but rather emerges as the result of long and tedious learning processes during which a political position shifts from being in the minority to the majority, to finally come to enjoy the full support of the citizenry in a given policy. John Dewey (2016 [1927]) was right when he wrote that “all valuable as well as new ideas begin with minorities.” These different temporal perspectives are not merely theoretical constructs. We believe that they roughly correspond to the short- and long-term perspectives that political parties adopt in their conduct. When faced with the fact of disagreement and the pressure to make decisions, party members are pushed toward adopting a synchronic perspective and endorsing compromises as the better outcomes, for the time being. On the other hand, this endorsement does not imply that they will forgo their “principled, collective project unfolding in time” (Ragazzoni 2017: 99) and pursue the long-term deliberative aim of persuading as many citizens as possible about the correctness and rightness of their respective projects.

In closing, let us provide some important insights that flow from this two-track model. To begin with, as we have seen, the long-term effects of political discourses oriented to consensus may be that the lines come to be drawn differently when it comes to the question of which views are acceptable candidates for compromise and which aren’t. Successful historical battles for hearts and minds (along the consensus track), for instance, imply that measures such as the abolition of slavery, the prohibition of marital rape, the prohibition of corporal punishment in school, and so forth no longer are objects of serious political disagreement in some societies. By extension, political views that contest such measures would fall outside the domain of views with which political actors could acceptably compromise in these societies.

Moreover, the two-track model enables us to respond to the so-called agonistic objection to compromise, according to which “[c]ompromises take the heat out of public debate and ultimately harm democratic politics, which requires that competing views be traceable” (Baume and Papadopoulos, 2022: 480). ¹⁴ As our model makes clear, however, compromises need not take “the heat out of public debate” and “leave competing views [un]traceable” because parties can (and typically will) continue to pursue the long-term and conflict-generating aim of changing hearts and minds (which we have characterized as the paradigmatic mode of the consensus track). Moreover, given the way in which we have construed consensus-oriented deliberation (as an orientation to the correctness of the claims raised in political discourses), the omnipresence of compromises in democratic politics actually would seem to flow from the deliberative and conflictual orientation to consensus. Put simply, the need for compromises arises because one actor raises a particular claim (thus implicitly assuming it to be a worthy candidate for consensus) to which another actor responds with an opposing one (thus implicitly assuming it to be a worthy candidate for consensus). Compromises take place against this background of conflicting political claims to correctness or rightness. Gutmann and Thompson make a related point when they write: “Without the uncompromising, compromise would need no defense. With the uncompromising at their flanks, compromisers gain bargaining power” (Gutmann and Thompson, 2012: 42). The agonistic objection implausibly assumes that the compromising and conflictual modes of politics cannot coexist, thus failing to recognize how the two modes differ in terms of temporal constraints and motivations and how compromises actually presuppose a conflictual orientation in politics.

This is not to suggest that this dual structure of politics cannot give rise to a particular tension once we consider it in the context of representative democracy. Eric Beerbohm (2018: 33) characterizes voters in elections as “purists,” which, in our vocabulary, means that their choices are consensus-oriented (they pick what they perceive to be the best or most correct political program). ¹⁵ Elections do not offer citizens the option to choose mixed packages (constituted through compromises). Thus, Beerbohm (2018: 34) continues, “what ordinarily distinguishes the citizen from the official who speaks for them is the latter’s access to the practice of compromise.” This asymmetry between voters and legislators may, as Patrick Overeem (2020: 61) puts it, give rise to democratic disappointment. Elections raise expectations because citizens vote in accordance with their first-order political commitments, but what they end up getting in terms of policies are compromises that (at most) only partially realize these commitments. As a result, legislators may become reluctant to compromise because they anticipate that compromises will not go down well with their constituents to whom they want to appear faithful to their shared political agenda. This pattern can lead to legislative deadlocks and the incapacity of a society to act collectively (Warren et al., 2013: 89). On the other hand, it seems problematic, as some authors do, to stipulate that citizens generally are adverse to and fail to recognize the desirability of compromises (see Baume and Novak, 2020b on this topic), especially if one considers the harms incurred to societies where legislators are unable to respond collectively to problems. Indeed, if we follow the logic of our two-track model, there is no necessarily problematic tension between recognizing compromises as the better solutions, in the here and now, while continuing to participate in the long-term battle for hearts and minds. ¹⁶

Conclusion

In this article, we articulated a two-track model of deliberation that simultaneously leaves space for compromise- and consensus processes. To pave the way for this model, we discussed and repudiated two views, one of which asserts that deliberation should be (almost) exclusively oriented to compromise, the other of which claims that compromise is incompatible with deliberation. Against the compromise-only view, we argued that it relies on a misunderstanding of the role of consensus and that it fails to appreciate how consensus-oriented deliberation is conducive to important learning processesf. Against the no-compromise view, we argued that it portrays deliberation in purely epistemic terms and that this misconstrues the point and purpose of the ideal of deliberative democracy as a practical political ideal. We then outlined a distinctive second-order perspective from which compromises can be endorsed as the best outcomes in the here and now and then demonstrated how deliberation or deliberative negotiation can improve compromise processes. Upon this basis, we introduced a two-track model that aims to demonstrate how deliberative democracy can be dually committed to compromise and consensus without running into internal inconsistencies. This model enables us to appreciate the different aims of political actors (making a good political decision under conditions of disagreement vs winning the battle of hearts and minds), the different temporal perspectives (short-term vs long-term), and the different sites in the deliberative system that pertain to compromise- and consensus-oriented deliberations, respectively.