Consensus and majoritarian democracies: Problems with under-informed single-level analyses

Introduction

I identify an academic version of the political gridlock we see in some democracies around the world. This happens in debates regarding what to do about a social system (say, a model of democracy) generating empirical problems with practice. I demonstrate that dismissing such systems on the lone basis of the empirical difficulties may arise from lack of appreciation of the normative objective leading to the establishment of such a system. I show that the dismissal is likely to be premature. In the same vein, proposing a normative idea is vulnerable if without any consideration about the empirical conditions in which it could be put into practice. I argue that the culprit is what I call an under informed model of single-level analysis. When done in a debate, this approach encourages gridlock. I explain what I call multi-level and informed single-level models of analysis as more productive alternatives.

In section 1 I explain that political philosophy deals with the espousing of normative ideas or objectives, and that these are ideas about how we ought to behave to achieve a certain kind of society that we all desire. Normative ideas vary according to their ambition (the enormity of what they expect of human behavior) and their remoteness or closeness to practicality. If they are so ambitious as to be impracticable, then they are irrelevant as normative objectives. It is therefore prudent to consider conditions in the real world when formulating a normative objective. Scholars who shun this approach are falling out of favour. Although there has been a trend in dismissing practicality concerns, I think the trend is harmful to normative theorizing, and this essay demonstrates it.

In section 2, I distinguish between consensus as a normative idea/objective and consensual democracy as a social system (we could design) for putting the idea into practice. It is taken as a given that consensus is better for group decision-making compared to voting. I recall that Kwasi Wiredu proposed a democracy by consensus due to his frustrations with majoritarian democracy in the experience of African states. I remark that Wiredu’s proposal is quite ambitious, although not too ambitious as to be completely impracticable. I argue that it is more difficult designing social systems to implement a normative idea than to conceive the idea. This is because there would be empirical conditions militating against a practice of the idea. In espousing an idea it is worse to ignore these conditions (the idea remains an idea) and it is better to take the conditions into account in designing systems for putting the idea into practice. This means studying existential obstacles to a normative idea and designing systems to minimize those obstacles. For this reason, normative theorists and the not-so normative (or the existential condition) theorists would need to listen more to each other. They also need to listen more to each other because social systems may not completely actualize normative objectives, but could get us somewhere between the status quo and the normative objective (a sort of mean).

In section 3, I demonstrate that it is inadequate to discuss either the normative in total blindness to empirical conditions or the empirical in complete oblivion to normative objectives. I call these under informed single-level analyses. I will cite Kwasi Wiredu, Helen Lauer and Martin Ajei’s philosophies as examples in the consensual democratic tradition, and Herbert Lionel Adolphus Hart, Ronald Dworkin and Warren Quinn as examples in the liberal majoritarian democratic tradition. There is a trend among the first set of scholars in dismissing majoritarian democracy based on an under informed empirical analysis, and an advocacy of consensus democracy based on an under informed normative appeal. The same happens with reverse to the second set of scholars, who present a theory of the individual that ignores the empirical circumstances of the community and makes majoritarian democracy resistant to community and consensual values. In both consensual and majoritarian traditions, such proposals contain incomplete evidence otherwise necessary for fair appraisal of the contending systems: for the consensus theorists, missing are both the normative objectives of (the supposedly opposing) majoritarian democracy and disquisitions of possible empirical obstacles to practicing a modern African consensus democracy. And for the liberal majoritarian democrats, missing is any consideration of the values of the community and the empirical obstacles to (and consequences of) asserting the exclusive and unlimited rights of individuals.

Instead of under informed single-level analyses, I propose what I call multi-level analyses as the most ideal, and at the very least, what I call informed single-level analyses. I demonstrate that if they had at least adopted informed single-level analyses, the essays of these authors would not exactly have been the same. In the African debate about consensus democracy, I give examples of informed single-level analyses with Emmanuel Chukwudi Eze, Kibujjo Kalumba, and Ademola Kazeem Fayemi’s approaches to Wiredu’s proposal, and an example of multi-level analyses with my treatment of Wiredu’s proposal. I also demonstrate that a debate in which participants on both sides (proponents and opponents) are all engaged in under informed single-level analyses will be grid-locked until some participants begin to adopt at least the informed single-level approach to analysis.

In section 4 I make concluding remarks about the one-sided nature of under informed single-level assessments (of either normative ideas or social systems designed to put the ideas into practice). I make these remarks in the light of my earlier analysis of the relationship between the normative and the empirical.

The normative and the empirical

Political philosophy deals with espousing normative ideas or objectives that seek to improve human behavior and the quality of life, especially in the area of political organization. Normative ideas are ideas about what we ought to do rather than what we do. Bridging the gap between what we do and what we ought to do is the main challenge of political philosophy. In this regard, we judge normative ideas by their level of ambition and their closeness to practicality. The ambition of a normative objective is measured by the level of moral maturity the objective requires from members of a society, and moral maturity is not a quality that is spread in appreciable quantities throughout the human society. The two aspects of a normative idea (the level of ambition and the closeness to practicality) are in a competitive relationship: the more ambitious a normative idea, the less practicable; the more practicable or closer to the practical world the less ambitious. Assuming there is a normative idea that is so ambitious as to be completely impracticable or utopian. Such an idea is not relevant as a normative idea: to be relevant, there must be the possibility of actualizing an idea at least partially. Sometimes we prefer more practicable and less ambitious normative ideas. This would be to save more time, energy and resources in actualizing the idea in practice. At other times we prefer more ambitious and less practicable ones (requiring more investment in the project of actualization).

The reason ambitious normative ideas are less practicable than their less ambitious counterparts is that they (the more ambitious ideas) are less fact sensitive. The more ambitious the idea the less fact-sensitive it is, and the less ambitious the more fact-sensitive. I think that normative ideas do not need to be very fact-sensitive: they would not be significantly normative if they were too close to the status quo. But I have hinted that normative ideas that are zero fact-sensitive are impracticable, utopian and irrelevant (if indeed there are any such ideas). It means normative ideas should be neither very fact sensitive nor zero fact-sensitive. Normative ideas need to be only minimally fact-sensitive enough to be bridgeable with practice. As such, normative ideas could be as ambitious as they can be, provided they are at least remotely practicable.

Due to the possibility of the irrelevance of excessively ambitious, completely impracticable normative ideas, it is prudent to consider the conditions of the empirical world when formulating a normative idea, in order to strike a certain balance between the ambition of the idea and the feasibility of practice. For this reason, normative theorizing in political philosophy cannot do without empirical investigations. There is a school of thought in Western political philosophy arguing that normative theorizing does not need any empirical considerations. G. A. Cohen (2003) is a chief proponent of this school. But over time and with growing frustration over the uselessness of excessively ambitious normative ideas, this school of thought has been losing influence.

As such, normative theorizing would do better working in collaboration with empirical feasibility checks. Normally, when a normative idea is proposed, criticisms of the idea emerge in the form of concerns about practicability. It has been a habit in both Western and non-Western political philosophy to dismiss such practicability concerns as mundane and pessimistic. But I think that such an attitude is harmful even to normative theory.

Consensus as a normative idea and consensual democracy as a social system

It is hardly disputed that consensus is better for group decision-making and action compared to voting. Kwasi Wiredu has proposed a democracy by consensus because of his frustrations with the majoritarian system of democracy practiced by many African states. Wiredu argues that the multiparty majoritarian democracy is adversarial, aggressive, divisive (1996, p. 179; p. 186; 2011, pp. 1059–1061, p. 1063), and ill-suited to the ethnic configuration of African states (2011, p. 1064). He argues that the relationship between political parties is constitutionally designed to be unhealthy (2011, p. 1061). He asks that we refresh our minds by considering the consensus political systems of pre-colonial African societies, where decisions were taken by consensus. He argues that consensus would ensure better representation of minorities (1996, p. 186). He proposes a non-party political system based on consensus (1996, p. 189; 2011, pp. 1061-1062). Wiredu admits that traditional political structures are outmoded as models for conceiving modern African political systems, but argues that consensus could at least be conceived as a normative idea informing political design and practice (1996, p. 189).

Although many of the supporters of consensus do not appear to realize this, consensus is a quite ambitious idea for group decision-making. I distinguish between consensus as a normative idea and consensus democracy as a social system designed (or we could design) to put the normative idea into practice. I have already discussed the relationship between ambition, fact-sensitivity, and practicability in normative ideas. I think consensus is a quite ambitious normative idea because of the amount of selfishness, negative individualism, dishonesty and impolite nature of many participants in many real-world deliberations, democracies and societies. The normative idea of consensus is also very ambitious because many issues lack the logical clarity to be discussed reliably toward common agreement, and the obstacles to consensus include conflicting subjective beliefs, values, and interests (see Ani, 2019). But in spite of its ambitious nature, the normative idea of consensus is partially achievable, not in terms of the number of unanimous decisions reached by groups but by the amount of compromises achieved, the amount of differences brought under some kind of control, improved optimality in decisions reached, how much more of the interests of everyone is taken into consideration building to a final decision, among other advantages over voting decisions.

Although routine unanimity is unachievable, a lot of consensus decisions are in fact taken in many group deliberations, especially on issues where the interests of participants coincide, and issues whose remedies are so logically straightforward that it would look foolish for a participant to object. But there are still many issues that are thorny, apart from varying attitudes of participants from one deliberation to another. I think the project of consensus democracy should be to increase the scope and number of consensus decisions, but this increase is constrained by the need to take group decisions that are genuinely unanimous. This is more reason why normative theorists need the input of the not-so-normative theorists concerned about issues of feasibility and practicability.

Let me come back to my distinction between consensus as a normative idea and consensus democracy as a social system designed to practice the normative idea. It is much easier to propose a normative idea compared to designing a social system for achieving the idea. A social system is designed to implement a normative idea, and the trouble is in the implementation. The reason why implementation is difficult is that certain empirical conditions are obstacles to putting the idea into practice. When a normative idea is proposed and existential obstacles to its actualization ignored, it remains what it is: an idea. To begin a journey to actualizing a normative idea is to first take account of existential conditions militating against the actualization of the idea. After such existential obstacles have been taken into account, they are reflected upon with a view to finding ways to minimize their features as obstacles. It is such obstacle-minimizing ideas that are used to design social systems for pursuing a normative idea. For this reason, a journey to actualizing (even partially) a normative idea only truly begins when normative theorists and the not-so-normative theorists concerned about obstacles begin to work together.

Normative ideas may also be seen as end-state ideas: they represent an end state toward which we desire to improve society. The project of gathering information about obstacles to the actualization of the end state, and reflecting on those obstacles with a view to overcoming them to get closer to the end state, is transitional theorizing. ^[1] This level of theorizing involves investigating factors of partial compliance (or non-compliance) to normative ideas. Consensus democracy is an end state, majoritarian democracy is the status quo, and we need transitional theorizing to get us from the status quo to the end state. We do not necessarily need to reach the end state (which may itself possess problems we do not wish to saddle). We may end up somewhere between the status quo and the end state. This would normally be a sort of mean between two extremes. Being a mean, it would be a compromise containing the best the two extremes can offer whilst expunging the worst of them. It only means that normative theorists should not project an air of superiority to the not-so-normative theorists. Both groups of theorists are like parties in a negotiation, trying to find a common ground that is workable. For all we know, the two groups are equal parties, none of whose positions would be the final one.

All of these mean that the normative and the empirical could be discussed in a complementary and hence gainful manner when discussing a normative idea in political philosophy. I seek to do two things with the rest of this essay. First, I wish to argue that it is quite vulnerable to evaluate the normative in complete isolation to the empirical world, and it is somewhat immobilizing to discuss the empirical alone without normative considerations. It is inadequate to evaluate an issue at only the empirical in total blindness to its normative objective, and it is also inadequate to discuss a normative idea in total blindness to its relationship with the empirical world. Each of these mistakes I will call under informed single-level analysis. I will cite examples of under informed single-level analyses with the single-level analyses characterizing Wiredu, Lauer, and Ajei’s philosophies in the consensus tradition and Hart, Dworkin and Quinn in the liberal (majoritarian) democratic tradition. I will argue that (what I call) multi-level analysis provides a more balanced evaluation of normative ideas and the empirical systems designed to put them into practice. I do not reject single-level analysis, but it should be an analysis done with an eye on what obtains at the other level. It should be what I call an informed single-level analysis.

Single-level and multi-level analyses

When a scholar criticizes a particular social system on the basis of its empirical failures, and rejects such a system due to these empirical failures alone, the analysis is done in complete blindness to any consideration of whatever may have been the normative idea leading to the establishment or fabrication of such a social system. Because such an analysis is only at the empirical level, it is a single-level analysis. Since such an analysis is done in complete oblivion to whatever may be the normative objective of such a social system, it is an under informed single-level analysis. To provide examples, I will begin with those who are guilty of the under informed model of single-level analysis in the liberal majoritarian tradition. But I would like to introduce them by saying some crucial things about the status quo, majoritarian democracy. In both Western and non-Western societies, there is little literature defending majority decision-making, for the simple reason that majority is the “natural” or “default” way of reaching group decisions emerging from difficult deliberation. Jeremy Waldron complains that there is lack of literature justifying majority decision because it is “taken for granted” (2014, p. 1695). According to Hannah Arendt, “the principle of majority is inherent in the very process of decision-making” and “likely to be adopted almost automatically in almost all types of deliberative councils and assemblies” (Arendt, 1963, cited in Waldron 2014, p. 1706). Majority is the solution to failure to reach natural unanimity. The most obvious justifications have been that majority decision-making respects the principle of equality (one person one vote), and it is efficient (Waldron, 2014, p. 1706). According to Rick Hills, “Where equality reigns, the weight of numbers is regarded as the best available proxy for quality of argument” (Waldron, 2014, pp. 1704–1705). Waldron has arguments against these justifications of majority principle (2014, pp. 1692–1723), but they are not relevant for my discussion here.

The principle of equality supporting the majority principle is coupled with the atomistic conception of the individual. In this conception, group decision is not necessarily the adjustment to the interests of individuals to an overall good, but the aggregate of the desires of the greatest number of individuals. According to the atomistic conception, the individual is ‘a small-scale sovereign’ (Hart, 1982, p. 183; cited in Ajei, 2016, p. 454), and the rights of this sovereign can ‘trump’ wider social goals (Dworkin, 1984, p. 153; cited in Ajei, 2016, p. 454). This conception advocates an unforgiving moral autonomy of the individual, and gives the individual the authority to pursue her own goals “not because such arrangement promotes overall human welfare, but because any arrangement that denied [the individual] that say would be a grave indignity” (Quinn, 1993, p. 170; cited in Ajei, 2016, p. 454). When we add the atomistic conception of the individual to the principle of equality (one person one vote), we have majoritarian democracy (the aggregating of the desires of the greatest number). But one notices that this liberal and atomistic conception bequeaths the individual with a veto power that encourages fragmentation and makes a consensus dispensation quite difficult. Since the individual will is the paramount, and collective decision is simply the aggregation of the coinciding will of the greatest number of individuals, there is no need in this theoretical scheme to consider the minority, or for any individual to adjust her interest to the overall welfare in a way that considers everyone in a more transcendent way.

Liberal majoritarian democracy has its (quite important) normative merit, which I shall explain when analyzing Wiredu’s under informed single level analysis. But the particular conception of the individual as a small-scale sovereign is an under informed single-level analysis of the normative value of individual autonomy, for the reason that it ignores the empirical fact that the individual usually lives with other individuals in a community and is almost never alone. Such a co-existence means that the individual’s rights are not absolute; her rights are attenuated by the existence of the rights of others in a community. And it is this atomistic conception of the individual that influenced the designing of majoritarian democracy in ways that encourage societal fragmentation and have led to calls around the world for consensus and community models of democracy.

It is ostensibly in response to such a conception of the place of the individual in the society (and a democracy) that consensual theorists such as Wiredu and his supporters have arisen. Wiredu, for instance, argues that consensual democracy is based on the idea of communalism (or the communal conception of a community), where you “adjust your interests to the interests of others even at the possible cost of some self-denial” (2011, p. 1056). But Wiredu and some of his supporters also commit the error of under-informed single-level analysis in advancing the theory of the importance of only the community, leading to criticisms that it all ignores a space for individual freedom (Jacques, 2011, pp. 1017-1030; Matolino, 2018, p. 119). Wiredu’s proposal has two basic parts that clearly show the under informed nature of his single-level analyses of both majority and consensus systems: a condemnation of the multiparty majoritarian democracy based only on its empirical difficulties (Wiredu, 1996, pp. 186, 188, 189; 2011, pp. 1059–1060, 1061–1063), and a recommendation of consensual democracy based only on its normative appeal (1996, p. 186, pp. 189–190; 2011, p. 1061, pp. 1063–1066). There is in fact a third part (which forms the bulk of the proposal): this is a description of the empirical successes of consensus decision making in traditional African societies, especially the Ashanti traditional society (1996, p.182–188; 2011, pp. 1057–1058, p. 1060) and a mention of the empirical success of a consensual system in Switzerland and Belgium (2011, p. 1064). But since the empirical conditions of modern Africa are different from those of both traditional Africa and modern Europe, Wiredu (and his supporters) astutely distance themselves from a recommendation of consensus for modern Africa based on its empirical success elsewhere. He (and they) argues that we should consider consensus not necessarily by its empirical success but on account of its normative value. So that leaves us with only two parts: a condemnation of majoritarian democracy based purely on its empirical difficulties, and a recommendation for consensus democracy based purely on its normative appeal.

Each of these parts is not just a single-level analysis, but an under informed single-level analysis. In the purely empirical condemnation of majoritarian democracy, there is not a word about the normative idea leading to the normative objective informing the designing of the multiparty majoritarian model. For fairness of analysis, I will mention that normative idea here. It is the idea that leaders, long standing in history as dictators, should be transformed into servants employed by the people through periodic elections. The normative value of this idea could be more appreciated if it were known that it was conceived to end the brutal tyrannies of feudalism. In the wake of the death of Charlemagne, the absence of effective strong central governments beginning from the 9^th century and the threat of constant invasion and famine made security necessary and elevated warriors who could protect patches of land. Weaker people needing protection pledged their loyalty to a dominant warrior in return for living on his (successfully defended) land. The feudal lord owned not simply his territory but those who lived on it. He owned their property and even their lives. Under that dispensation, ordinary people did not have any rights whatsoever. They did not have a right to life. This was a master-slave relationship, and is the background to protecting the individual from the abuse of power (liberalism) as well as the philosophical idea advanced during the Enlightenment that leaders should be the very opposite of slave-owning lords. Majoritarian democracy was thus the idea that this relationship should be reversed. The lordship should be transferred from the ruler to the ruled, and the tendency of leaders to feel like lords should be checkmated with a competitive arrangement in which the leader is constantly aware he could lose his position to a rival, courtesy of the wishes of the led. It was the idea that the led should be employers of their leaders. This idea has its own normative value. Dwelling, therefore, on only its current empirical difficulties (the competition among those jostling to win the hearts of the led) is a one-sided assessment. This is not to say that Wiredu should discuss the history behind the normative idea, but at least that he should have had the normative objective of majoritarian democracy in mind, to make his analysis fair, balanced, and complete.

Second, Wiredu shies away from discussing any empirical difficulties that his proposed non-party consensus democracy may itself encounter in modern practice, especially in the face of growing diversity. He is silent on this difficult task and exits his proposal at the blissful stage of presenting an idea with a normative appeal. This is somewhat too convenient. Of course, this kind of analysis is not strange. On the other hand, it is common, indeed so common that someone may argue it is perfectly legitimate. But the point of my essay is to demonstrate that it is hardly illuminating and does not move scholarship forward. I am, therefore, aware that this is how we mostly do our analysis in political philosophy. But the point of this essay is that we could do better. Let me now show this by applying my recommendation to Wiredu’s proposal.

Since Wiredu condemned majoritarian democracy in complete neglect of the normative idea leading up to its establishment (whose acknowledgment may have modified his approach and tone), and recommends consensual democracy at the normative level without a sincere stock-taking of empirical difficulties it may encounter today, what he has done throughout his proposal is a series of under-informed single-level analyses. Wiredu could have done one of two things in this proposal. It is either he conducted a multi-level analysis or, at the very least, a series of informed single-level analyses in place of the under informed model. The difference between the multi-level and the informed single-level is that the multilevel consists of attending explicitly to both the normative and empirical levels of analysis at each stage of a proposal and for every issue considered. An informed single-level analysis is similar to what we normally do (the trend has been single-level analyses), except that each single-level analysis is done with at least an implicit awareness of the content and importance of the other level. For a multi-level analysis, Wiredu could have highlighted the origin of the normative idea leading up to majoritarian democracy before discussing its current empirical difficulties in the light of its normative value. If this were done, I am convinced that his conclusions regarding the empirical difficulties of majoritarian democracy would have been more constructive than dismissive. This is because Wiredu could have seen that the tendency to tyranny leading up to the normative idea in Europe is also in Africa, and he would have needed to deal with the question of what happens if the idea (and the social system) of (re) electing leaders, and of different candidates vying to be leaders, is dispensed with. Wiredu would have needed to answer questions about whether leaders would remain servants of the people if they were no longer elected from a pool of alternative candidates, and particularly if they were no longer humbled by the awareness that they could lose their positions. Wiredu would have confronted the possibility that leaders who are not (re) elected may begin to take decisions that are informed by their awareness that they cannot be removed. He, therefore, would have been saddled with the task of finding a solution to this problem. It likely would have occurred to Wiredu that if political parties and elections were proscribed, alternative ways of instilling the spirit of service in political leaders would be needed. This problem would also have had its impact on Wiredu’s proposal of democracy by consensus. It could have presented to him the problem that accountability was a loose end he needed to tie up, and if he saw no novel way of doing this, he probably could have rescinded his non-party proposal and instead proposed a multi-party coalition democracy based on certain consensus- and compromise-encouraging provisions in the constitution.

All of these entail that Wiredu, in an imagined multi-level analysis, could have traced the normative merit of majoritarian democracy (which is a normative analysis), moved on to examining its difficulties (an empirical analysis), called for the need for a more consensual form of democracy (back to the normative level), and anticipated some of the real difficulties such a democracy may encounter (back to the empirical level). Each of these levels of analysis would have been informed by an awareness of the importance of the contents of the other level. If Wiredu wishes to do single-level analyses, this is also fine. But each level of analysis used should be conducted in a way that is not completely blind to the importance of the other level. This is an informed single-level analysis, and it cannot yield the same results as an under informed single-level analysis.

My next example is Helen Lauer. Like Wiredu, she focuses on only the empirical difficulties of majoritarian democracy and only on the normative merits of consensus. In other words, she focuses only on the under-informed model of single-level analysis. Her verdict on majoritarianism is not just purely empirical but also empirical in a way that actually conflates the normative objective of a majoritarian democracy. Read her:

... in those self-defeating technocracies, compromise is the mark jof political incompetence. The strength and desirability of elected representatives is measured characteristically by their capacity to stall, obstruct, thwart, if not obliterate the effectiveness of their opposition party rivals. Yet in those ‘culture[s] of conflict’, the very raison d’etre of a political party vanishes without a robust opposition to pit against and knock around. (Lauer, 2011, pp. 185-186)

Lauer’s empirical allegation is true that in a multiparty democracy, political parties often desire to knock around their rivals in their desire to gain or keep power. And, truly, anyone would dismiss a social system whose objective is to simply knock politicians and parties against one another. The problem with Lauer’s evaluation is that the knocking around is not the normative objective of majoritarian democracy. The normative objective is to transform political leadership from life-long and autocratic entitlements to periodically negotiated contracts. Those to award the short-lived contracts are those to be led, and those interested in leading would need to bid competitively. This is the normal bidding process that leads to the award of contracts. It is in the bidding that the knocking around occurs. Lauer is, therefore, not correct about the raison-de etre of the knocking around. She describes only the empirical difficulty without mentioning the normative design. In the absence of mentioning the normative objective, Lauer’s empirical criticism is marshaled in a way that gives the impression that knocking rivals around is actually the normative objective of majoritarian competitive democracy. If the true normative objective is factored into Lauer’s analysis, a reader may choose to assume a constructive rather than dismissive attitude to the empirical difficulties. There could be many more ways of minimizing this knocking around apart from simply rejecting the entire system. At least 80 something countries that have been searching for these more ways (around the world) have ended up in various shades of coalition multiparty democracy. Instead of rejecting majoritarian democracy in toto, the reader may ask how we could improve on the empirical difficulties without losing the normative objective. If it is in the bidding process for leadership that the knocking around of rivals occurs, we could think about ways of sanitizing the bidding process, instead of throwing away the whole idea of bidding. This is another example in which an under informed single-level analysis of a social system could be misleading.

Back to my point: Lauer does not consider that the normative idea leading to the establishment of competitive multiparty politics is to decentralize power from congealing on a single leader for too long by transforming leaders into employees of the people, who should seek re-employment periodically. As such, Lauer does not suggest what could replace the abolition of competitive politics. Indeed, she stops at merely suggesting its abolition. This shows an unpreparedness to confront the empirical consequences of such abolition. This is similar to Wiredu’s unpreparedness to confront what could be the empirical difficulties of practicing a democracy by consensus. But this kind of indecision arises from focusing on only one level of analysis to the neglect of what obtains at the other level. It technically leaves the evaluator in the dark regarding what position to take in the wake of such analysis.

Lauer in fact argues that Emmanuel Eze is wrong to criticize Wiredu’s consensus democracy on the grounds of empirical difficulties (Lauer, 2011, p. 184). And her praise of consensus is purely at the normative level. She writes that consensus democracy is normatively attractive, that its normative value is all that matters, and that this is what Eze should focus on (Lauer, 2011, p. 184). Defending her position (and the position of Wiredu), she writes that empirical conditions do not matter because “the legitimacy of any society’s governing institutions does not depend upon the fact of what this or that person actually believes at a given time” (2011, p. 183). The swing between levels of analysis (the empirical for one system and the normative for the other system) does not do justice to the matter of comparing both systems, for in Wiredu, Lauer (and Ajei, whom I shall soon highlight), we are spared the normative importance of one and the empirical difficulties of the other.

There is also a partiality of analysis. I will be guilty of this kind of analysis if, suppose I were proposing that people should have guns, and in my proposal, I focus on only the advantages of having a gun, and only on the disadvantages of not having a gun. My analysis conveniently ignores the disadvantages of having a gun and the advantages of not having a gun. My reason for suppressing these issues is my fear that acknowledging them would weaken my proposal. Lauer even argues that empirical conditions of traditional African consensus political systems or indigenous rule should be barred from the debate and we should focus only on the normative value of consensus (2011, p. 187). But arguing that the empirical conditions surrounding the traditional success of consensus should be barred from the debate is only an attempt to enforce the one-sided assessments.

A third example is Martin Ajei’s condemnation of majoritarian democracy and advocacy for consensual democracy (2016). He condemns majoritarian democracy only on the basis of its empirical difficulties and promotes democracy by consensus only on the basis of its normative merits. Missing from Ajei is any awareness of the normative value of majoritarianism and a sincere projection of what could be the difficulties in practicing a consensual democracy. In a related attack, Ajei dwells on the empirical failures of liberalism, and on the basis of this he concludes that liberalism has failed in Africa. He in fact argues (quite wrongly) that African political values are non-liberal, and that any attempt to reform majoritarian democracy into a social system with a potential to working well on the continent “can hardly be described as liberalism” (Ajei, 2016, p. 448). The most treacherous effects of under informed single level analysis is Ajei’s unequivocal dismissal of “basic civil and political rights” or “uncontroversial and very urgent human rights” (Ajei, 2016, p. 453). But Ajei would not have arrived at such a dismissive evaluation of liberalism if he considered the original normative idea leading to the growth of liberalism in the Western Hemisphere. The original idea leading to liberalism was to protect the individual from state abuse. This ultimately led to the emergence of all the inalienable rights (such as rights to life, speech, privacy, association and religion) that have been globally accepted, and which have been (at least indirectly) protecting many disadvantaged people all over the world from being disproportionately killed by tyrannical leaders. If this had occurred to Ajei, it would also have probably occurred to him that the subsequent adventures of liberalism in other areas of life such as classical economics (laissez faire), competitive politics (majoritarianism), secularism and so on, are later developments of liberalism that could be evaluated on their own individual merits with regard to how much they stray from the original normative idea leading to liberalism.

The pitfalls of under informed single-level analyses does not stop with merely being one sided in a debate: it also enables some participants to engage in a one-sided defense of an idea in a particular forum, and then engage in an equally one-sided dismissal of the same idea in a different forum. Bernard Matolino is fond of this erratic approach to the consensus debate. But my analysis of his behavior in this regard is too long to be included here: I shall treat him in detail elsewhere. ^[2]

I will not say that authors who are guilty of under-informed single-level analysis are not aware of the content of the level of analysis they are ignoring. Most of the participants in the consensus debate are at the top of their profession, and should expectedly be aware of both the normative objectives and the empirical difficulties of both majoritarian and consensus democracies. If they know the contents of the levels, then the one-sidedness may be a question of economizing with sincerity rather than lack of knowledge, or maybe they don’t always pay enough attention to what they know. Or, more likely in my opinion, it may be a desire for argumentative convenience (the desire to get away with an easier argument). Whatever the reason is for this kind of stubbornness, it is obvious that there is no way a debate can be constructively done when each side simply chooses to focus on a level of analysis that pleases it, and psychologically wishes the other level does not exist.

To the best of my knowledge, there are no clear majoritarians in the consensus debate in Africa, for the simple reason that no African scholar defends liberal majoritarian democracy with explicit arguments, much less in a way that completely ignores the ideal of consensus or communitarianism. All those involved in the consensus debate in Africa have shown some level of sympathy or the other for consensual democracy. This includes Eze, presumably the harshest critic of Wiredu’s proposal, whose biggest criticism is that consensus is only a moment in democracy (Eze, 2000, par 29). Eze’s main grouse is that we should cherish disagreements for their sake as much as we should cherish agreements for their sake (2000, p. 30). One must note, however, that Eze in fact commends Wiredu’s proposal, especially Wiredu’s proposal for substantive representation (2000, par 9). It means Eze appreciates the normative value of consensus. We can, therefore, say that Eze’s critique was an informed single level analysis.

Fayemi also shows signs of an informed single-level analysis in his critique of the workability of consensus. Fayemi commends Wiredu for espousing the idea of the right to decisional representation (2010, p. 6). In other words, Fayemi acknowledges the normative value of Wiredu’s proposal. But Fayemi doubts that a non-party democracy is the right social system for pursuing this normative objective (2010, p. 7). This means that Fayemi appreciates the normative value of Wiredu’s proposal but thinks that we should continue to debate what kind of social system is reasonably suitable (many things considered) for pursuing Wiredu’s normative objective.

A third example of the informed model of single level analysis is Kalumba. He appreciates Wiredu’s proposal, but argues that a unanimity-endorsement model seen in some of Wiredu’s writings is too high of a moral requirement. Kalumba prefers a reading of Wiredu in which the aim at unanimity, rather than its achievement, is the defining mark (2015, p. 107). Like Fayemi, Kalumba doubts that a non-party polity, as proposed by Wiredu, is a good way to implement Wiredu’s ideal of consensus (2015, pp. 108–111).

If critics have shown sympathy for the consensus ideal in one way or another, it means that critics appreciate the normative value of consensus, even in their criticisms. We can see this in Eze, Fayemi and Kalumba’s positions.

For an example of multi-level analysis, I will turn to my first response to Wiredu (Ani, 2014). I had said that my reaction to Wiredu’s proposal is that of “qualified acceptance” (Ani, 2014, p. 345). This means that I accept the normative value of a consensual dispensation, even though I may have problems with suggestions made in the designing of a particular social system for putting the normative objective into practice. I made it clear that my acceptance of the need to transition democracy in Africa to more consensual forms does not mean I accept many of what Wiredu said in his proposal. I wrote, “…I believe that consensus has its normative merits as a social and political idea” (Ani, 2014, p. 346). This statement about the normative value of consensus set the stage for a constructive rather than a dismissive critique of Wiredu’s suggestions for empirical practice. One may accept a normative objective but is not satisfied with the social system suggested. In that case, one may seek to suggest changes to the system or suggest another system, but one would not reject the entire package (normative objective as well as social system). So if I accept the normative value, I am not likely to reject the idea that we pursue it in practice. If I find that the social system suggested by Wiredu is too inappropriate for pursuing the idea, I am likely to suggest its improvement or replacement. For example, I had written that it is advantageous for groups to take consensus decisions because such decisions generate more inclusive value (Ani, 2014, pp. 353–354). But I had also argued that consensus decisions, although superior in inclusive value, are not necessarily superior in epistemic value: indeed, the drive for unanimity may undermine the epistemic quality of consensus decisions. The rest of my essay was then a suggestion about ways we could safeguard the epistemic quality of consensus decisions.

Due to the rarity of multi-level analysis in the African consensual democracy debate, some scholars think I am devious (or pretentious) to accept the normative value of consensus and criticize consensus at the same time. Bernard Matolino wrote that for taking (what he sees as) a dual position, I am “not innocuous” (Matolino, 2016, p. 40), indeed, I am “up to mischief” (Matolino, 2016, p. 43). Matolino chides me for not belonging squarely to either the camp of the supporters or the camp of the critics (Matolino, 2016, pp. 45–46). I had responded to this criticism by writing that he should abandon the idea of belonging to camps (by which I meant taking entrenched positions) (Ani, 2018, pp. 266–267).

All of these bring me to the top of my point about the disvalue of the under informed single-level analysis. It promotes ‘camps’ and entrenched positions. A little illustration should show its ability for stalling debates. If, for instance, we are in a debate about whether to liberalize and legalize the private ownership of submachine guns, and proponents of the legislation argue that such a gun could stop a large number of armed robbers charging into your compound, whist opponents argue that it could likewise be used to mow down a large number of innocent people. Suppose proponents are only concerned about the brighter uses of a submachine gun (as we see indeed in many gun debates), and many opponents are only concerned about the darker uses, it means the two sides will continue to talk past each other and the debate would last forever. In the same way, the debate about consensus and majoritarianism has the potential to come to a gridlock if proponents are only concerned about the normative appeal of consensus and critics only about empirical difficulties.

The one-sided nature of under informed single-level assessments

The ineffective nature of the under informed single-level approach to evaluating normative ideas and social systems is one of the reasons for this essay. Let me end by rehashing my arguments about the relationship between the normative and the empirical in researching ideas that could be transformed into social systems. When we evaluate social systems, we are in fact evaluating empirical attempts to transform certain ideas into practice. Such ideas have normative value in the sense that they are suggestions about how humans should behave in order to create a society that we could all enjoy. Some of such ideas could be quite ambitious, in terms of what they expect of human behavior in order to achieve an end-goal. Those too ambitious may be of no use. Moderately ambitious ideas are easier to transform into practice, although no normative idea is completely or perfectly translatable into practice. My point is that establishing a social system in order to actualize an idea brings humans into confrontation with real-world obstacles to actualizing the envisaged society. Scholars who evaluate such empirical obstacles would need to do so in conjunction with a parallel evaluation of the normative importance of the idea leading to the establishment of the social system in question. Without such a comparative assessment, evaluating empirical obstacles alone is a one-sided assessment. So is evaluating normative ideas alone in complete blindness to real world conditions. To be fair, real world conditions should not be given too much importance as this could induce higher levels of pessimism in the project of trying to give concrete life to a normative idea. But my point is that real-world conditions should be considered at all.

I would compare a normative idea envisaging a better society to desiring a destination, towards which one needs to journey. Deciding to embark on the journey to the destination is like creating a social system to attempt putting a normative idea into practice. One may encounter hitches along the journey. It is inadequate (and could be misleading) to evaluate the hitches encountered, and use the observation of hitches as a basis for deciding whether to abort the journey. It would be a more informed evaluation to evaluate the hitches alongside an evaluation of the importance or purpose of journeying to the destination. One then weighs the significance of the hitches to the importance of the journey. This informed assessment should lead to a better decision about whether to continue the journey or not. A simple criticism of the hitches, and a subsequent decision to abort the journey, would be a one-sided assessment. If an impression of the hitches alone is the basis for a subsequent decision to abort the journey, one may in retrospect regret the decision on realizing in future that the importance of reaching the destination outweighs the difficulties encountered in the hitches.

When we evaluate the empirical difficulties of a social system, the severity of the difficulties could lead us to be dismissive of the social system. But this is an under informed single-level assessment that forgets the whole idea leading to the social system. The most ideal would be a multi-level evaluation of the empirical difficulties of a social system alongside an evaluation of the normative objective (the end-goal) leading to the establishment of the system. One then weighs the significance of the empirical difficulties against the importance of the objective or envisaged end state informing the system. If the normative merit of the objective is as significant as the seriousness of the difficulties, it is unlikely that we would straightforwardly recommend an abolition of the system. We are instead more likely to engage in transitional theorizing to generate ideas about how to mitigate the difficulties to enhance the journey toward the end state. Like I mentioned, end states are not exactly designed to be reached, but to at least make a difference to our status quo.

If more participants in a debate embraced multi-level and informed single-level analyses, it is unlikely that some theorists would be calling for a total rejection of majoritarian democracy (Wiredu, Lauer and Ajei) whilst other theorists are calling for a complete rejection of democracy by consensus (Matolino, 2018, pp. 5, 8). What is more likely is that participants could progress to more advanced and sophisticated stages in the debate where they seek to reconcile the normative objectives of majoritarianism (taming political leaders) with the normative merits of consensus (enhancing unity through decisions), and to seek to put the best of the two systems into a hybrid system that transcends both initial systems, with a view to minimizing the empirical difficulties that each of the initial systems generates when put into practice. The multi-level analysis has even further uses: its keeps examining the new difficulties that may appear with a hybrid system, and attends to them with an awareness of the normative objective of the hybrid system. None of this dialectical movement in the direction of improvement is possible in a world of under informed single-level analyses.

Conclusion

I have argued that our assessments of either normative ideas or social systems need to be done in a way that puts all the evidence on the table at both normative and empirical levels, or at least in a way that does not completely neglect all the information at one of the levels of analysis. It is an error of exclusion, suppression of evidence, or one-sided assessment for participants to leave out information that would undermine their argument, and to ignore it when submitted by others. I have used examples to show that a debate in which most participants do this is a natural candidate for gridlock. The sad part is that many of our debates are done this way. Multi-level or at least informed single-level analyses offer an alternative to more constructive debates. I hope this essay has provided convincing demonstration of this point. I also hope the case studies I used offer lessons for a continuation of the consensus debate, and for other debates.