The ugly truth behind transitional justice in the post-revolution phase: A constitutional law and economics analysis

Abstract

This article explores the transitional justice (TJ) dilemmas after revolutions have overthrown autocratic regimes through developing a model that uses a law and economics methodology. The article seeks to answer two questions: Why do post-revolution regimes resort to or ignore TJ policies towards former autocratic regimes? And why is it difficult to adopt and apply welfare-enhancing TJ mechanisms in practice, including popular suggestions within the TJ literature to portray the civil society organizations as the key solution to TJ dilemmas? To answer these questions, the article provides a theoretical positive analysis of the scenarios and dilemmas of TJ. It argues that TJ should function both as an internalization mechanism of negative externalities of the violations of the past-regime, and a form of constitutional arrangements as an ex ante incentives structure to prevent the repetition of these violations. However, due to asymmetric information problems, behavioural biases and the constitutional nature of TJ, the ‘TJ momentum’ precludes most of the traditional solutions for this principal–agent problem.

I. Introduction

In the aftermath of a crisis involving one group oppressing another, one of the first questions faced by society is how to deal with the former oppressors and their legacy in a way that achieves justice for the victims and prevents a repetition of the past. This question has been the subject of lengthy debates within different social fields. The world has recently witnessed two waves of revolutions in the Middle East and North Africa (MENA) region (France 24, 2020b), which resulted in wars in many places. The fears of backsliding towards authoritarianism in Eastern Europe are also growing. Moreover, protests are roaring again –sometimes violently – in the streets of Latin American countries. These are just some examples of crises involving oppressive regimes and human rights violations worldwide, which again attract attention to the questions of post-crisis strategies. Under which law should former dictators be tried? Who should be held accountable – only the leaders or each individual of the past-regime? Should the new regimes include the criminals of the past, or should they be vetted? How can the institutions responsible for the past policies be transformed to prevent their repetition in the future? The process that addresses the relationship between the new and the past political orders, starting from negotiations until implementing transformative policies, is referred to as transitional justice (TJ).

This paper aims to contribute to the debate within the TJ literature over one of the primary questions following a revolution: TJ determinants. What leads a nation to resort to a meaningful TJ process after a revolution over an autocratic regime, or precludes it from resorting to such a process? To answer this question, the article presents a threefold analysis. First, it adds to the legal scholarship that uses the constitutional nature of TJ as a starting point to better understand its dynamics and problems. Second, it adheres to the interdisciplinary approaches to TJ that benefit from fields other than law and politics to explain TJ after revolutions by using constitutional economics for the first time. And third, it questions a common suggestion in the TJ literature that civil society can be the answer to many of the TJ dilemmas away from typical state-centric and traditional legal solutions.

The United Nations dedicated part of its Rule of Law Initiative to the transitional justice project, which covers the broad aspects of TJ and provides a range of mechanisms that can be used by different societies (Olsen, Payne and Reiter, 2010b; United Nations, 2010). According to these guidelines, the United Nations provides a normative rule book that indicates the mechanisms a nation should follow to achieve its transition into a system that respects the rule of law and promotes human rights. The UN guidelines define TJ as ‘the full range of processes and mechanisms associated with a society’s attempt to come to terms with a legacy of large-scale past abuses, in order to ensure accountability, serve justice and achieve reconciliation’ (United Nations, 2010, p. 2). ¹ This definition is in line with the other available definitions for TJ in the literature (e.g. Anderlini, Conaway and Kays 2005: 1; Kassem 2013: 47–48); Kritz 1995; Olsen, Payne and Reiter 2010b: 9–10; Sriam and García-Godos (2013: 2); Stan and Nedelsky 2013; Teitel 2003: 69). There are two main domains for TJ within this definition. The first aims at achieving peace in the aftermath of an armed conflict, while the second focuses on achieving political transformation after oppressive regimes (Fisher 2020: 291). This article is part of the scholarship that discusses the second category. ²

As many studies argue, achieving the previously mentioned goals through adopting some or all of the TJ mechanisms is much more problematic in reality than it sounds in a rulebook. The dilemmas of adopting and then applying TJ mechanisms have been under investigation from different social fields, including the different veins of law – for example, criminal, constitutional and international law – along with economics, political sciences, history, sociology, psychology and anthropology, among others (Bell 2008, pp. 9–10; David 2017: 152; McAuliffe 2011: 90; Palmer and Clark 2012: 1). In this article, I first use the constitutional economics theories to discuss why and how the adoption of successful TJ mechanisms after revolutions over autocratic regimes is remarkably rare and challenging in reality, before asking why resorting to classic strategies under the public choice theory and within the TJ literature to solve these dilemmas would not help this setup. ‘Public choice’ refers to the research field using economic tools to analyse the traditional problems of political sciences. While the traditional assumption is that governmental actors strive to maximize the people’s utility, public choice works under the assumption that these actors – like all other humans, including voters – aim to maximize their own utility. When using the public choice theory – also known as political economy – to identify constitutional law problems or suggest solutions to address these problems, this branch of analysis is called ‘constitutional economics’. ³

I first use a novel methodology to reconfirm what other interdisciplinary approaches have been claiming. I then use this same methodology to reveal why the traditional suggested solutions to TJ’s challenges, and classic solutions of constitutional economics to other principal–agent problems, would not be valid in the case of post-revolutionary TJ. These realizations become clear once we deal with: (1) the parties to the TJ processes as choice makers who are rational, but also subject to behavioural biases; (2) TJ as a constitutional arrangement that differs from any other legal policy because it cannot have a third-party enforcer; and (3) the particular circumstances of what I call the ‘TJ momentum’. ⁴

I will first map the available literature on TJ challenges before attempting to present a more accurate model of how TJ works within democratization processes after revolutions over autocratic regimes. I use the law and economics methodology to explain TJ as a bargaining process between the rulers of the past and the rationality of the transitional regime’s existence before discussing the roots of TJ dilemmas that result from asymmetric information and behavioural biases. I finally explore the classic solutions to these economic problems and investigate whether they could serve a purpose in solving the TJ dilemmas after a revolution. I argue that the instantaneous nature of TJ after revolutions in the absence of any trusted institutions that could enforce a ‘contractual’ arrangement between the parties, or supply accurate information in the market, precludes the effectiveness of most of these solutions. This extraordinary nature of the TJ momentum hence requires innovative methods to treat this exceptional situation. While civil society has been suggested by modern TJ literature as the solution to TJ challenges after revolutions, constitutional economics help us understand why this is not in fact the case.

Since this research focuses solely on the democratic transitions after revolutions over autocratic regimes, I limit my analysis to this setup, which excludes other situations that may involve TJ, such as civil or international wars, or any other form of conflict. ⁵ For the purposes of this research, the concept ‘revolution’ refers to a popular uprising that was followed by a change in the leadership of the state. This includes both the cases of regime collapse and negotiated transition. This definition may depart from other definitions in the literature that consider revolutions as a strict subset of regime changes (Colgan 2012: 446). However, it fits the purposes of this research better, as it includes even the failed transformations, which still constitute existing cases of TJ after revolutions, showing variant scenarios of post-revolution TJ setups. To this end, I apply a positive economic analysis, also referred to as ‘the political economy of law’ (McNollgast 2007: 1654, 1661–64), through the tools and theories of microeconomics, including the rational choice theory, asymmetric information problems as a market failure, and the behavioural biases emerging from a principal–agent relationship.

The remainder of this article proceeds as follows: Part II provides an overview of the relevant literature and how this analysis contributes to it. Part III presents the mapping of TJ actors and scenarios and the rationale of TJ as a constitutional arrangement that plays different roles according to the revolution scenario it falls into. This is a necessary preamble to Part IV, which explores the dilemmas of TJ policies in general and their suggested traditional solutions. Finally, Part V offers concluding remarks.

II. Literature review

Although TJ is a multidisciplinary field, as explained in the introduction, it was long dominated by legal scholarship (Bell 2008: 6; Fisher 2020: 299) while also receiving major contributions from political scientists (Olsen, Payne and Reiter 2010b: 2; McAuliffe 2011: 90l Posner and Vermeule 2003: 762). Despite the variations in the tools and theories used by these two scholarly fields, their central questions remain the same: How should new regimes deal with past regimes’ crimes? Should they follow policies of punishment or forgiveness? Restorative or retributive justice? What are the most effective mechanisms of TJ for achieving its different goals? Which should prevail: the moral or the practical considerations? How can the new regime use TJ as a transformative set of policies to promote peace, the rule of law, human rights or democracy as various goals of political transformations? Furthermore, for the interests of this article, what leads a state to pursue TJ, or what might preclude it from seeking or achieving TJ (Posner and Vermeule, 2003, p. 762; Teitel, 2000, p. 3)?

The legal literature on TJ tends to focus on the legal complications of the rule of law and criminal justice implications in the context of dealing with TJ as a part of what is called ‘transformative law’. ⁶ Additionally, this literature focuses on how these concepts might change or persist in the context of TJ (David 2017; Dyzenhaus 2003; McEvoy 2007; Posner and Vermeule 2003; Safjan 2008; Teitel 2000). The political science field is generally more focused on studying the connection between TJ and political transformations, democratization and balance of power, and how these factors may influence each other (Olsen, Payne and Reiter 2010b: 9–25). On a final note, the TJ dilemmas can differ distinctively, depending on which domain of TJ is discussed – conflict to peace or authoritarianism to democracy. ⁷ This is not only because the mechanisms of focus differ, as some scholars and practitioners point out (Arthur 2009: 360), but also because transforming abusive state authorities into democratic institutions that respect human rights alters the calculus of the relevant actors in a way that creates its own dilemmas. This will be the topic of this article. Under this section, I locate the article’s multi-fold contribution: first, within the literature discussing TJ as a constitutional arrangement; second, within the interdisciplinary approaches to TJ; and third, within the literature on TJ determinants.

Not only is TJ multidisciplinary, including different social fields, but within the legal scholarship, TJ is internally interdisciplinary (Bell 2008: 19). Its disparate aspects are studied by public international law (Chang and Lee 2016), international criminal law (Iverson 2013), international humanitarian law (Salmo 2006), international human rights law (Szoke-Burke 2015), criminal law, administrative law and constitutional law (Teitel 2000: 7–9). However, the focus of this article is on TJ as a constitutional arrangement that falls under the public choice problems.

There is a close two-way relationship between TJ policies and constitutional law. On the one hand, constitutional reforms after wars, revolutions or conflicts are generally considered part of the TJ process (Olsen, Payne and Reiter 2010b: 1). This is known as the constructive role of constitutionalism ⁸ in times of political change within the concept of transitional constitutionalism (Teitel 2000: 191–211). The discussions within the legal literature therefore focus on constitutional justice as a part of reform that operates in tandem with its counterparts, such as criminal and administrative justice, to achieve political transformation. In other words, constitutionalism in this sense takes a normative transformative approach that addresses not only the future regime, but also the past one (Teitel 2000; Turner 2015). Teitel (2011) provides a further detailed discussion on the interactive relationship between TJ and the transformation of constitutionalism and the state responsibility in contemporary constitutionalism. This tension between the backward and forward perspectives of constitutional transformation in transition is also considered to be one of the dilemmas of TJ, which is reflected not only in the constitution-writing process, but also in the treatment of constitutional courts to the new cases before them (Safjan 2008: 235–39).

On the other hand, the mechanisms of TJ other than institutional reforms that have the constitution as their subject raise constitutional objections and arguments related to their violation of established liberal democratic constitutional values such as the ban on retroactivity, equal treatment by law and the rule of law. ⁹ There has always been a debate over whether such violations should be permitted and even considered by the constitutional principles (Dyzenhaus 2003; Graver 2015; Posner and Vermeule 2003; Safjan 2008: 239–49). However, this debate is not the subject of this paper. Instead, this research focuses on how the viewing of TJ policies as constitutional arrangements, in the light of constitutional economics, reveal the roots and extents of their dilemmas.

Interdisciplinary approaches to TJ aim at providing a new understanding of it that is more realistic than the rigid traditional legal approaches that used to monopolize the field (Bell 2008: 17; McEvoy 2007). This research points in the same direction by using economic thinking to reconsider some typical portraits regarding TJ, and particularly its relation to civil society, to explain what precludes them in practice.

In general, TJ – and especially its constitutional aspects – has not been subject to economic methodologies. Some scholars looked into the economic aspects in the means of financial restrictions of TJ (Olsen, Payne and Reiter 2010a). Others investigated the connection between TJ and economic development (Duthie 2008; Mani 2008). There are also studies that consider economic inequality and socioeconomic factors as roots of the conflict and as aspects of the TJ process (Carranza 2008; Laplante 2011; Miller 2008; Schmid and Nolany 2014). However, except for these aspects related to economics in its finance sense, TJ has remained out of the reach of economics when it has come to the applied methodology – in other words, using microeconomic tools and theories to understand TJ problems and suggest solutions to these problems. Even the connection between property rights or economic transformations, especially in post-communist societies, and TJ came from a purely legal perspective. Law researchers tried to both compare the different possible strategies for a state to transform these property rights systems and search for a theory behind these changes, using legal rather than economic approaches (Atuahene 2010; Damşa 2016; Riedel 2018; Stan 2013). As far as I am aware, the reasoning, difficulties and search of solutions to TJ still lack economic thinking and the ability to search one step further beyond the surface of the power setup, especially through the use of behavioural economics.

The following analysis will build on the work of other law and politics scholars when they started looking beyond classic legal understanding and adding a realistic flavour to the TJ analysis, using constitutional economics for the first time. Although the economic thinking and political analysis have the common considerations of political dynamics, power and information asymmetry in the TJ setup, constitutional economic analysis is still needed to add a missing aspect to the current understanding of TJ. This aspect is the additional layer of the dilemma that disrupts the currently available solutions – a layer that is composed of the constitutional nature of the setup and the expected behaviour of the relevant players under economic theories.

Some initiatives within the legal literature have started to reconsider the role that legal solutions can offer to solve the TJ problems and invite perspectives from other social fields. One of the most important is McEvoy’s (2007) proposal. Not only did McEvoy spell out the limitations of the classic legal theories of TJ in reality, but he also invited law scholars and practitioners to look beyond legalism and to use other social sciences to reach a more realistic understanding of TJ. However, McEvoy’s proposal is more focused on political sciences and criminology as necessary disciplines to interact with conceptualizing TJ without considering the rational choice of economic theory. To some extent, this limitation might endanger the main recommendation of his analysis, which focuses on empowering and including ‘grass-roots’ organizations such as local communities and civil society. ¹⁰

His proposal goes hand in hand with further analyses, coming specifically from the political sciences field – for example, Arthur and Yakinthou (2018) – especially after the way civil society was pictured for its role within democratic transitions in the 1980s as separate from the state and empowering for the transitional process (Foley and Edwards 1996; Jeffery, Kent and Wallis 2017). However, it ignores what economic analysis captures: that civil society organizations themselves are a party to the TJ constitutional setup. They have their own preferences that might diverge from or unite with any of the other parties, and they incur costs and expect benefits from whatever TJ approach is adopted. That is why they cannot be considered a third party enforcer of the TJ process. As a constitutional arrangement, TJ cannot have a third party enforcer because it is not a usual contract between parties; even international parties are not third parties to the process because they are part of it and have their own interests.

The view of civil society organizations in the TJ context as neutral, secular and homogeneous parties was recently challenged by Kent, Wallis and Cronin (2019). ¹¹ Their analysis reinforces the argument of this article against the idea of civil society as a third-party enforcer to TJ, as contractual agreements between parties of the post-revolution phase have diverse priorities, sources and backgrounds. However, while they use case studies in the Asia-Pacific region to show the manifestation of this argumentation, especially in religious and ethnic civil society organizations, I use a theoretical analysis that shows the roots, dynamics, and consequences that can apply to different contexts and different forms of civil society.

Modern literature on TJ acknowledges how the development of TJ shows its origins as a social need for a group that includes the ‘victims’ that aim ultimately to achieve a social, rather than just political, transformation, even if this transformation relies on state-level actions (Fisher 2020, p. 288). This conceptualization of TJ calls for more focus on the relationship between society-level and state-level actors in the context of TJ adoption and application, including the allegedly intermediate actors – that is, civil society. Accordingly, viewing this through the lens of the principal–agent problems between the victims (the principals) and the state-actors (the agents), which is an established theory in constitutional economics, can lead us to necessary realizations for understanding the TJ dynamics. ¹²

This research relates to a growing body of literature that tries to define the determinants of TJ. Whether a nation adopts TJ or not, and which TJ mechanisms it tends to select, could be reasoned through many theories. Some scholars relate these choices to economic limitations (Olsen, Payne and Reiter 2010a: 61). After all, TJ is financially costly, and it usually faces societies that are already economically exhausted. In a close connection, for example, Grodsky (2010) argues that a meaningful accountability towards the past human rights violations is conditioned on the powerful primary actors’ ability to secure essential economic and political needs for a citizenry that demands TJ. Other studies connect the TJ arrangements to the distribution of power, which depends on many variables, including the past level of repression, autocratic regime duration, the timing of the transition, the type of transition, the leadership type of the past-regime, the background of the new leaders, the democratic past of the society and the strength of the civil society (Nalepa 2010: 370–72; Olsen, Payne and Reiter 2010b: 43–59). All these factors should decide whether the past-regime is weak enough to be held accountable before the new one, and whether the demanders of TJ have enough power to spoil any settlement that excludes this process. But what about the limitations of the momentum?

The idea of the momentum was referred to by David C. Gray (2010) in his reference to a ‘liminal moment’ that needs to be seized in order to achieve a sort of parity between the victims and abusers. While Gray is interested in a specific aspect of TJ, i.e., reparations, and liminalism in his argumentation relates to the whole transition from an abusive regime to one that respects human rights and the rule of law, the ‘liminal moment’ for this article concerns the momentum right after a political change where three conditions apply. First, there is asymmetric information between the principals and the agents; Second, low levels of trust between the parties because of the absence of political credit due to the pre-existence of an autocratic regime; Third, the lack of a commitment device; And, fourth, deciding a constitutional arrangement under the veil of uncertainty. ¹³ Within this research, the TJ momentum is what profoundly constitutes its extra-traditional dilemmas as a distinctive form of principal-agent-problems.

In his invitation to those who are interested in democratization to remove the ‘transition lens’, Carothers (2002: 17–18) says that what is seen as an exceptional transitional middle ground between autocracy and democracy is, in reality, the case for most of the political systems in the contemporary world. One assumption regarding democratic transitions that Carothers invites us to ‘let go’ of is the idea that achieving a successful democratization process depends primarily on the intentions and actions of its political actors and elites, with no significant influence from its social, economic and institutional legacies (Carothers 2002: 17). Although this argumentation is sound and essential, one cannot ignore that when it comes to the calculus of the TJ momentum, it can seem flawed once you look at it from a rational choice perspective. First, the intentions and actions of the public elite are connected and result from the social, economic and institutional legacy of the past-regime; they are not two different matters. Second, while it can take years before one can deconstruct and analyse the economic, social and institutional legacy of a past-regime to build a strategy in order to reverse or reform it, it looks more rational to offer strategies that deal with the instant constitutional transaction taking place right after a political change. I try to offer insights in this regard through the following analysis. I agree that there is no one-size-fits-all scenario of the political landscape of countries in transition, which is why I will present three scenarios of this transition. However, I differentiate between ‘nations in transition’ and nations in the ‘momentum’.

III. TJ scenarios after revolutions explained

TJ as a constitutional policy

As explained earlier, from a legal perspective TJ pertains to different legal veins, including constitutional law. The constitutional matters regulated by TJ include the formation of the state authorities during and after the transition, the civil rights limitations of some of the citizens, partial organization of the electoral rules and, most importantly, extraordinary judicial procedures that could not be valid under the ‘standard’ constitutions. TJ policies can be formalized in different legal forms, including constitutional texts, basic laws, regulations, or merely legal actions or decisions. The analysis of this article is not limited to one of these forms; rather, it tackles the concept of TJ regardless of its formality. Nevertheless, beyond its legal nature and formality, how can we explain the economic rationale of TJ generally?

There is a close two-way relationship between a successful democratic transition and successful TJ in legal and political thinking. Some may find the latter as a result of the former (Kassem 2013: 53), since achieving justice is completely conditional on the presence of political will and the proper environment, as indicated earlier. Others argue that transitional justice itself is one tool for completing a democratic transition after authoritarian rule (Arthur 2009).

The central rationale of transitional justice is putting an end to the ideology of the past regime, not just for provoking accountability or compensating the victims so that ‘justice’ is restored, but also to promote the political, economic, legal and social change (Ehito 2015: 75) for which the revolution aimed in the first place. In other words, TJ seeks to avoid bearing the roots of the authoritarian regime and its affiliated unsolved inefficiencies into the newly formed system, which can threaten the achievement of a genuine transformation. Finally, TJ policies usually address crimes and use mechanisms that deviate from the standard laws in normal situations (David 2017). Accordingly, TJ is basically, a tool to do the following:

1. 1. Internalize the negative externalities ¹⁴ of the past autocratic activity that has ongoing spillover effects at both the individual and collective levels. Although the past regime’s harmful activity is supposed to be in the ‘past’, the externalities of this activity still impact the present – for example, discriminative policies that led to the poverty of an entire geographical area of the state, or torture in prisons that deprived a family of its member or a person of their health.

2. 2. Minimize the risk of a counter revolution by the past-regime, while at the same time minimizing the risk of its supporters spoiling the transition if they are under attack, depending on their power.

3. 3. Avoid a costly direct interaction between the victims and violators, which could lead to welfare-decreasing ends such as violence or further social polarization. ¹⁵

But do these incentives apply to all post-revolution cases?

The TJ parties

A prerequisite for answering this question is an explanation of the incentives structure of the democratic transitions after the possible scenarios of revolutions. I argue that there are three basic scenarios for the post-revolution setups. Before presenting them, I first organize the relevant actors in this analysis into three groups: ¹⁶

1. 1. The past-regime. This group would include (a) the main figures of the past-regime, such as the head of state, the head of the parliament and the executive, and the leaders of the governing political party; (b) the secondary figures, such as the high-ranking officials who occupied leading positions under the past government, such as ministers, governors and the heads of the public authorities, as well as members of the regime’s political party.

2. 2. The rulers of the transitional phase. These do not necessarily have to be the same powers which started the revolution, but they should be the party which has sufficient power on the ground to end the past-regime and control the transitional phase. For example, in the Egyptian case in 2011, the revolution started by youth movements over social media. However, ruling the transitional phase – even during negotiations before the moment of Mubarak’s removal - was in the hands of the Supreme Council of the Armed Forces (SCAF) (Barany 2011). ¹⁷

3. 3. The victims. These are not necessarily only the persons who were subject to a direct crime that involved human rights violation by the past-regime, such as torture, but also the indirect victims. These indirect victims include the citizenry who were negatively influenced by the corrupt policies of the autocratic regime. For example, a citizen who was prevented from voting in public elections as part of a general stream also qualifies as a victim under this broad definition. However, it should be noted that victims themselves do not all share the same preferences: some of them are for, some are against and some are neutral towards the TJ issue. Although the idea of a victim who is against justice – known as the ‘reluctant victim’ – might seem counterintuitive, it is common and can happen for different reasons. One example is the Argentinian children of the disappeared adopted by junta families, who refused all justice efforts to investigate their origins and reconnect with their families (Mégret 2018). This sub-categorization means there could be various ways for new rulers to be representative of, constitutive of or connected to different victim groups, which could then influence the understanding of the power dynamics and information gap between them and the victims. For this reason, the upcoming scenarios work on the assumption that the critical mass of the victims, which initiated and fuelled the revolution, are pro-TJ, while victims against TJ are considered as part of either the past-regime or the transitional rulers’ supporters, in which case the last are against TJ as well, and the neutral victims are mere bystanders who do not influence the power dynamics.

The TJ map

Mapping the post-revolution transitional phase depends not only on each case’s circumstances, but also on the subject under consideration. For example, a simplified map showing the constitutional transformations could involve the steps of changing the constitutional rules made from the start of the revolution until the adoption of the new constitutional order. However, in the case of explaining the map of TJ, the steps are much less clear. First, the steps genuinely differ in nature and sequence, depending on the type of transition after the revolution: was it a regime collapse that involved a clean break from the old regime, or a negotiated transition with that regime still involved? Second, unlike the other constitutional arrangements, TJ does not end with the adoption of the new constitution. On the one hand, the new constitutional order itself is an institutional reform that falls under TJ; on the other hand, the policies of TJ could still function after the new constitution as they usually take a long time to undergo adoption and application. Third, TJ policies do not take place within one decisive moment that marks the end of the transitional phase: many TJ policies are adopted and/or applied gradually through different points of the transition. All these considerations make it hard to introduce one explanation for the structure of the post-revolution transitional phase concerning TJ. However, for the sake of clarity, Figure 1 presents a simplified theoretical map of scenarios of the post-revolution phase regarding TJ.

Figure 1: Map of TJ stages in post-revolution transitional phase

The overthrow of a past-autocratic regime can generally happen according to two basic arrangements (Olsen, Payne and Reiter 2010b: 155). ¹⁸ The first is when the regime collapses after the revolution, which usually means it loses its legitimacy and does not have staunch supporters who could spoil the transition. In this case, the bargaining over TJ usually starts after the past regime’s collapse, and it excludes the representatives of that regime. Consequently, the parties to the negotiations are the transitional rulers and the representatives of any other groups in society who are powerful enough to influence the transition. Romania is a clear example of a regime collapse. The violent revolution in 1989 ended the 42 years of communist rule and publicly executed the longtime Romanian Communist Party Partidul Comunist Român (PCR) General Secretary Nicolae Ceaușescu and his wife Elena (Nelson 2014). Since then, the political and economic system of the country has changed significantly, despite the persistence of many challenges inherited from the past-communist era (Stan and Vancea 2015). In the second setup, the negotiated transition, the past-regime is too weak to continue, but it also still holds some power, which makes it costly to exclude it from the transition. In this case, the bargaining starts before leaving authority because this move would not happen at a reasonable cost without the past-regime’s approval, and the negotiations continue after the transitional rulers take over because the past-regime remains a party to the process. ¹⁹ An example of a negotiated transition is what followed the Tunisian revolution in 2010–11. After almost a month of protests, ex-president Ben Ali was forced to flee the country and was then granted protection by the Saudi government (BBC News 2011). Although the head of the regime and his family left the political scene, and despite the steps taken afterwards to achieve a democratic transition, the past-regime continued to be a major party to the post-revolution phase through Nidaa Tunis Party and its personnel, who co-designed that phase, including how the transitional justice process was undermined (Abderrahmen 2018). For example, the ex-president, El Sebssi, was one of the well-known figures of Ben Ali’s regime, an heir to the Dostourian Party formed by Bourqiba, even though the Muslim Brotherhood, an ex-opposition political power, was simultaneously in government (Mohsen-Finan 2018).

However, even in a general mapping that disregards the specific types of past and post-transition regimes, things are more detailed and complicated than this. In this research, I go a step further and divide the transitions on a second level according to the preferences of the three active actors indicated earlier. In this categorization, I refer to the victims as the principals because they are supposed to be the original actors who delegated the power of reaching their goals to their agents – that is, the transitional rulers ²⁰ – drawing on the theory in constitutional economics that sees citizens as the principals who delegate their powers and try to control their agents – that is, the government. ²¹

Applying the probabilities theory to a situation that involves three actors, where each actor could have one of two possible sets of preferences – whether to support the TJ goals or not – means we should have six possible scenarios. However, testing these scenarios against real examples, even theoretically, shows the impossibility of four of them. The four impossible scenarios would involve two where the past regime members would desire that the revolution achieved its goals and two scenarios where the principals would wish that the revolution failed to reach its aspirations. Obviously, these sets of preferences are non-applicable. While this calculation should leave us with two possible scenarios – that is, where the agents and principals’ preferences regarding the revolution’s goals either unite or diverge – both history and rational choice theory show the prevalence of a third scenario (Geddes, Wright and Frantz 2014; Tullock 2005). This scenario is when agents share the principals’ preferences to overthrow the past regime and punish it, but replace it with another authoritarian regime rather than transforming it into a democratic system.

The first scenario is when the victims, referred to as the principals, share the same preferences with the transitional rulers, referred to as the agents, in getting rid of the past regime, holding the regime accountable and transforming into democracy. This scenario is the classic setup of the transitional phase, upon which normative assumptions of the traditional legal scholars and international guidelines were built. The principals’ and the past-regime preferences are the same in the second and third scenarios; only the agents’ preferences (transitional rulers) change from one scenario to another. In the second scenario, the preferences of the principals and the agents diverge. While the principals seek democratization and accountability for the past-regime, the agents seek to turn around the revolution and find a safe exit for the past-regime. The third scenario involves the divergence of the preferences of the three actors: principals desire TJ policies, and agents seek the recreation of an autocratic regime. However, in this case the agents want to punish the past-regime and exclude it. To make it simpler and easier to follow, I refer to the three scenarios as follows:

$$ Scenario\;1\;\left( PR+ AG\right)\;X\;(EX) $$

$$ Scenario\;2\;(PR)\;X\;\left( AG+ EX\right) $$

$$ Scenario\;3\;\left( PR\left)\;X\right( AG\right)\;X\;(EX) $$

Where $ PR $ is principals/victims, $ AG $ is agents/transitional rulers, and $ EX $ is the past-regime. ²²

Scenario 1

The first scenario might look a bit rosy, and that is why it has a few historical examples, which could represent merely a temporary fragile phase that afterwards stabilizes in the form of one of the other two scenarios. One recent example is the transitional regime in Burkina Faso after the 2014 uprising and up until the coup in 2022 (The Guardian 2014; BBC News 2022). The preliminary consequences of ex-president Blaise Compaoré fleeing the country in response to the protests involved military rule (BBC News Mundo, 2014). However, a transitional map was shortly decided along with civil national, regional and international actors, transferring the power to a coalition civilian government (BBC News 2014). Moreover, measures were taken to ban politicians allied to the ex-president from running for the presidency (BBC News, 2018). ²³ On Polity IV, a measure indicator of autocracy characteristics, ²⁴ Burkina Faso jumped from a long history of score 0 in 2014, which listed it as an anocracy, to a consequent score of 6 from 2015 to 2018, which qualifies it as a democracy (INSCR Data Page 2019). The American Revolution (1765–91) is another excellent example of a revolution that witnessed unity in preferences between both principals and agents against the past-regime. However, the American Revolution’s aftermath did not involve a TJ process, unless we consider the American Constitution the primary institutional reform that could act as what was referred to earlier as transformative constitutionalism. ²⁵

The parties’ behaviour in Scenario 1 can be explained through both rational and behavioural analysis. In the example above, and also in other examples where we can find an application for this setup, it can be noticed that no one party holds power that outweighs the other parties’ powers. Looking at the Burkinabe military, religious leaders, civilian politicians and civil society (International News 24/7 2014b; Mackey 2014; Ouoba 2016), each of them was too strong to be ignored; otherwise they could spoil the transition. Along with the regional and international pressure (International News 24/7 2014a; Reuters 2014; UN News 2014), it was too costly early in the transition for one party to oppress the others, so they had to apply cooperative measures to reach a compromise. This compromise aimed to find a cooperative equilibrium, something also reflected in their adopted TJ mechanisms until the 2022 coup, which did not include large-scale mechanisms but focused further on the institutional reforms to preclude a repetition of the past. When the power dynamics changed later, the whole setup also changed.

Politicians are still humans, so although it may not maximize their subjective expected utility in the neo-classical economics meaning, altruistic behaviour that is motivated by bounded self-interest can also be expected from them (Simon 1995) – especially in such historic moments. These rare behavioural attitudes explain such an exceptional setup involving the similarity of preference between principals and agents.

This perspective suggests another explanation of the positive empirical results in terms of democracy and human rights, correlated with the balanced approaches of TJ that refuses both maximalist and minimalist TJ policies (Olsen, Payne and Reiter 2010b). TJ’s maximalist approaches tend to favour the ‘justice’ or retribution considerations, which lead to trials and isolations of the past-regime. ²⁶ Minimalist approaches favour the ‘political constraints’ or the transition concerns, which lead to amnesties and reconciliation. ²⁷ The current explanations attribute the relative success of the justice-balance approach to its balancing nature between the competing considerations of achieving accountability and politics of transition. However, this may be a form of reversed causality; it could be that witnessing higher rates of achieving TJ goals in these cases is due to the balancing nature and cooperative behaviour of the involved parties, which reflects on all their choices in the transitional phase, not because of the type of TJ mechanisms they choose.

Scenario 2

In the second scenario, although the autocratic regime could not prevent a revolution, it still holds more power than any other actor. Usually, in this case, the transitional rulers – who are part of the past-regime itself – will try to absorb the revolutionary wave through making a change only in the prominent face of the regime or partial constitutional changes. The transitional rulers are a reproduction of the past-regime, and consequently they are not expected to adopt a systematic TJ process that emerges from democratic consultations. Instead, they may tend to give away partial reparations or pardons to the victims while punishing some figures of the past-regime as scapegoats, especially in the immediate aftermath of the revolution. With the past-regime still controlling the state, and time passing in the crisis, the transitional rulers can finally reach the point where the principals lose any weight in the bargaining process and then could easily be overcome.

An example of this scenario is the Algerian uprising in 2019 (El Gantri 2019). Although the ‘Hirak’ succeeded in removing the ex-president Boutaflika, the military, which is the most powerful authority in the state, was the only one holding power afterwards (Rabahi 2020). The newly elected president is one of the main figures of the past-regime (MCD 2019). Some prominent members of the past-regime were presented at trial – not including Boutaflika (Chikhi and Ahmed 2019a, 2019b; CNN 2019) – along with promises of constitutional amendments and freeing of the demonstrations’ leaders (Baum and Ahmed 2020). However, Hirak still refuses to accept all these movements and considers them to be a mere facade to carry on with the reproduction of Boutaflika’s regime (France 24 2019; France 24 2020a; Sky News 2019). Nevertheless, with its power fading and more victims moving to the inactive parties’ side, the costs it can impose on the other actors to force its demands decrease.

The second scenario is a perfect manifestation of the asymmetric information problem in the context of TJ as a constitutional arrangement. The transitional rulers do not just have more information about the governing laws, crimes of the past-regime and the ongoing negotiations; they also have the intention of manipulating the principals. Giving temporary little-valued measures to falsely signal to the victims that they are committed to achieving the goals of the revolution and TJ results in an adverse selection problem. The principals tend to accept a welfare-decreasing situation that involves an absent or malfunctioning TJ process because they do not know how this process is planned to end by the transitional rulers (agents). The agents tend to give their scapegoat measures in an autocratic form instead of democratic procedures to keep the information low on the principals’ side. They also count on the victims’ preferences changing over time due to the prolonged economic and political pressure, ²⁸ which should diminish their expected utility from a long-term TJ process compared with the utility of the short-sighted ‘stability’. In the meantime, through their partial reform, they try to mitigate the negative outcomes of the past-regime violations.

Scenario 3

In the third scenario, the transitional rulers’ preferences diverge from the preferences of both the victims and the past-regime. The agents ultimately want to end the past-regime, punish it and prevent a possibility of its resurrection. However, in the meantime they do not aim to produce a regime that achieves the revolutionary goals or initiates a TJ process. Instead, they seek to establish a radically different autocratic regime of their own.

An example of this scenario is the Iranian revolution in 1979, which resulted in transferring the country from the autocracy of the Shah’s empire to the Islamic Republic’s theocracy (Afary 2020). Although both student and socialist movements participated along with the Islamists led by El-Khomeini in the protests, the religious powers were able to seize the authority after the Shah fled the country (Brumberg 2004). There were trials held against the past-regime that witnessed severe punishments under the label of revolutionary courts, resulting in thousands of executions against different Shah regime members of various levels (Abrahamian 2008: 181). However, these trials were not held after either a democratic process or any due process considerations (Bakhash 1984: 61). The mechanisms used after the regime change could be considered more like revenge than a transitional justice process, which did not just exclude supporters of the past-regime but even most of the victims themselves. ²⁹

In this scenario, the transitional rulers free-ride over the principals to achieve opposing pay-offs. It could be that the agents alone could not expel the past-regime, so they used the victims to achieve their goals and then started the rent-seeking activity. It could also be that the transitional rulers did not participate in the revolution at all, but had tools that enabled them to transform their political power into votes, which gave them control over the authorities. In this latter case, the risk-averse agents free-ride over the revolutionary activity caused by the risk-taking principals, who end up with no return on their investment in the revolution. It is a scenario that will be costly to achieve in the case where the victims are also organized in political parties with qualifications to oppose the new autocratic regime.

In all these scenarios, although TJ deals with past violations, it is an ex ante constitutional arrangement in the phase in between the past and the new regime. The political actors of this phase decide, when in uncertain times, ³⁰ a tradeoff between their expected political revenue and the expected risk. Given the emotional baggage involved and the critical momentum of the transition, parties usually view TJ policies as a tradeoff between conflicting pay-offs. For example, providing amnesty policies for past regime members entails limiting the pay-offs for the victims who supported the revolution and wanted to punish those who committed human rights violations against them. Similarly, seizing all the political pay-offs for the non-members of the past-regime through reform of electoral laws involves losing their expected utility as human capital. This view gives the sense that TJ is a zero-sum game. However, TJ’s balanced approach tries to escape this view by explaining all TJ mechanisms as advantageous to all the parties involved. Accordingly, amnesties conditional on truth revealing and confessions are not only profitable for the past-regime members but also for the victims, who will be able to reach the truth about those accountable for the violations committed against them and receive the proper moral reparation for their suffering. Usually, though, TJ mechanisms inevitably involve a tradeoff between the different parties to reach a compromise. TJ should be a structure of incentives to not only prevent the repetition of the autocratic past but also internalize whatever disutility was caused by that past.

The complexities of TJ, then, can be found on two levels: the first is in resorting to TJ in the first place, which was explained in this scenarios’ landscape; the second works on the assumption that the preferences of both the victims and transitional rulers are the same – that is, for TJ to operate towards its original ends. Still, on this last assumption, TJ as a concept is much easier said by lawyers than done by politicians.

IV. The TJ dilemma

To internalize the consequences of the past negative externalities as an ex post action, while simultaneously preventing the repetition of the harmful activity as an ex ante arrangement, there should be mechanisms that incentivize the agents to achieve the principals’ preferences. As explained in the last section, the agents’ preferences differ depending on the scenario, while the principals’ preferences are always to achieve the revolution’s goals. This can be problematic in two ways: first, regarding the TJ concept generally; and second, concerning the costs of each mechanism specifically. The second part is beyond the scope of this article. So, for the first part, how can reaching a point that incentivizes the agents to perform up to the principals’ standards be problematic?

Asymmetric information problems

Information asymmetrically distributed in the market between the exchanging parties is one of the market failures which leads to undesired results. ³¹ In the case of post-revolution transitional justice, two groups of actors have better knowledge and understanding of the legal institutions, the facts of the crimes and the political bargaining. These groups are the past-regime members and the transitional rulers. Those in the third group of actors – the victims –usually do not have the same advantages.

Two-direction incentives influence why this asymmetry usually continues throughout the TJ process. On the one hand, for most of the victims, the cost of collecting information on the TJ process is high. On the other hand, the other two actors have strong incentives to keep the victims imperfectly informed, providing them only with the information that would keep the process on the track they designed. This manifests a principal–agent problem (PAP).

The principal–agent problem (PAP)

The principal–agent problem results from information being differently distributed between the disparate parties of the transaction concerned. The principal’s benefit is derived from assigning specific responsibilities to the agent, who is then expected to carry them out in a particular desirable manner for the principal. In other words, the agent has the informational advantage, and their actions impact upon the principals’ pay-offs. However, the problem is that the agent’s incentives could be contradictory to this performance because their rational choice will be to seek their own benefits, not those of the principal. Monitoring the agent is costly both in terms of collecting the necessary information and sanctioning the agent if they fail in realizing the principal’s benefits (Cooter and Ulen 2012: 283–84). ³²

In this typical meaning, PAP is a problem in the second and third scenarios, where the victims’ and transitional rulers’ preferences diverge. In this case, the victims assign the transitional rulers the responsibility of achieving TJ goals. However, the problem is that the victims do not have the necessary information about the applicable laws, the possible efficient laws, the facts of the cases that can be under consideration, or the incentives and behaviour of the relevant ‘agents’. Additionally, sanctioning the agents for their rent-seeking behaviour could be precluded by the typical collective action dilemmas, in a phase that does not yet have established democratic authorities. Subsequently, monitoring the agents’ performance is highly costly for the principals. On the other side, the transitional rulers may have opposing incentives – for example, they may want to secure a safe exit for the past-regime due to close connections with it.

Take, for example, the Libyan transition after toppling the ex-president, El Gaddafi, in 2011. It was argued by analysts that the anticipated democratic transition was reduced only to formal reforms that would not achieve the original aspirations of the revolts (Giblin 2021). A significant number of the National Transitional Council (NTC) members were defectors from the past-regime. These members pushed towards quick elections that would guarantee that they would be elected because no space was given for public deliberations or the formation of solid political parties. On the other hand, the revolutionary activists demanded a two-year gap to give them the opportunity to create effective constituencies within the Libyan tribal society that had a severe constitutional vacuum. The vision of the transitional rulers prevailed, and the result was a parliament formed from ex-regime members supported by their tribes and foreign forces, which failed to achieve any transitional goals. In this case, it was not just the access of the revolutionary activists to the public deliberations and decision-making that was severely limited. Even when they had the necessary information regarding the rent-seeking behaviour of the transitional rulers, they could not stop it because, unlike the transitional rulers, they did not have the necessary resources or mechanisms to do so. The Libyan transitional process involved many other factors and mistakes that could explain its failure; however, the conflicting preferences in light of limited power and access available to the reformists not only contributed to stalling the whole process, including transitional justice efforts, but also led to a civil war in the country. ³³

However, how does PAP interplay in the context of the first scenario, where the principals’ and agents’ preferences are the same: to achieve TJ goals? In this case, the asymmetric information leads to divergence in the desired means, even if the desired ends are the same. For example, concerning trials against the past-regime members, agents may try to avoid costly judicial procedures because they prefer economic gains or political stability over the other pay-offs of ‘achieving justice’, which may be more important to the victims than to the politicians. Assume that a change in the international political landscape leads to finally toppling the Al Assad regime in Syria. Assume also that the new government desires to achieve the long-awaited democratic transition and bears the hundreds of thousands of victims’ sacrifices in mind. However, how is it possible to design a transitional justice process for a state that was torn by excessively entangled war for over a decade after a peaceful revolution over a horrific longtime autocratic regime where the regime victimized part of the population while the other part was violated by armed militias combatting that regime? ³⁴ The victims’ reactions to the current trials under the universal jurisdiction in Europe signal how victims would not let go of severe punishment for any person involved in these violations (Schaer 2022). However, such a maximalist approach can be worrying for politicians, who might be eager to reconcile the society at the cost of punishing the leading criminals only. It can also be a possible scenario that Al Assad would agree at some point to leave power in exchange for a safe exit, which can save the country and the Syrian victims more tragedies. However, would all this information necessarily be available to the Syrian people? ³⁵

The traditional solution to the agency problems is to design a contract that incentivizes the agent to perform according to the principal’s preferences. So why is this unable to serve as a solution in the TJ context? First, as explained earlier, TJ is a constitutional arrangement. This means that TJ policies themselves cannot work as a commitment device between the different actors because there is no third party who will enforce this ‘law’ or ‘contract’. The only guarantee that these policies will operate is that they are self-enforcing, which means they reflect an equilibrium point between the actors’ pay-offs. This enforcement difficulty is a typical dilemma of constitutional law that has been analysed by different scholars (Voigt 1997: 29–30). Second, the TJ agency model deviates from the typical agency problems where the desired end is always for the agent to perform in the way the principal wishes. In other terms, the bias in the traditional agency models is expected only from the agent’s side. So is this also the case in the TJ model?

Behavioural biases in the TJ agency model

The PAP in the public choice literature focuses on the intergovernmental relations and bureaucracies besides the relations between the rulers and the public as agents and principals. ³⁶ Miller (2005: 220–23) explored the possibility of the principal’s moral hazard. He also discussed the behavioural biases that could disrupt the rational assumptions of PAP in the political realm and the empirical results that motivated them. In the original PAP, the problem comes from the agent side because there is an assumption that whatever the principal’s preferences are, they are legitimate and natural. However, if we take the overall efficiency of the principal–agent relationship, it could be that what principals seek is self-destructive.

The TJ PAP departs from the classic canonical assumptions of PAP ³⁷ in a number of points. There is still an asymmetry of information, and the agent still impacts upon the principal’s pay-off. However, the preferences’ asymmetry is not valid in all the cases, as the principal does not always have the privilege to take the initiative; the principal does not hold the possibility of ultimatum bargaining; and, lastly, the principal does not have common knowledge of the agent’s preferences, which eliminates the backward induction possibility. Considering these deviations, how does the behavioural bias interplay within the TJ PAP model?

Behavioural biases from the principals’ side in the TJ context

Due to asymmetric information, principals can have unrealistic expectations about how TJ works, how long it takes and what it requires because they have no reference point or sufficient technical knowledge. They also tend to be over-optimistic regarding its results. This bias stems from uncertainty about the future. For instance, due to their lack of proper legal information, the victims could estimate the range of criminal sanctions in relation to specific crimes and start calculating the probabilities of the sanctions that members of the past-regime might receive. However, if for some reason related merely to the applied laws, the accused do not receive the sanctions the victims expected, they will then tend to think that the judges were biased, or that the transitional rulers are allied with the past-regime. This complicates the crisis or incentivizes the people to engage in further protests, and thus the unrest continues. Their lack of information about the economic costs of trials or investigations can also lead them to blindly support these mechanisms, despite the fact that one of the revolution’s original strongest motives is usually economic hardship. Instead, the over-optimism bias could lead them to believe that the exhaustion of the economic resources is not because of their TJ decisions – at least partially; instead, they will tend to blame the agents for their poor economic policies. ³⁸

Principals in the TJ model could also have a high-time preference, so they prefer short-term over long-term pay-offs, which could result in errors in decision-making, negatively impacting social welfare. They may also select options that bear implied costs that outweigh the direct benefits. For instance, victims could prefer arbitrary tribunals that would lead to the execution of every person who was ever a member of the past-regime party, even if this meant they gave the new system the tool of arbitrary tribunals that did not conform to international standards of fair trials. This may not be in their interest in respect of their personal welfare in the long run because this new system would then be able to use this same tool against them. The reasons behind this could be that, although they have information about the gains and losses of their choice, they have bounded willpower, or they are being myopic, or the benefits of the efficient solutions are not apparent to them. It could also be that the alternative policies desired by agents do not align with the principals’ self-serving conceptions of fairness, which they desire to provoke. Finally, the possible harm of the wrong choices may not be important to them; combined with over-optimism, that may lead to substantial underestimation of the risks of their choices.

Imagine, for example, the previously discussed case of Syria. Assuming that a holistic balanced approach of TJ would be the best possible scenario for a successful transition in Syria, would the communication of the cost and benefits of each TJ approach to the public be sufficient to bring the victims and their families into the deal? Despite the understanding of the need for a proportionate approach to criminal sanctions imposed on the Al Assad regime defectors who helped the justice process, which incentivizes other defectors to cooperate, some families of the victims would still think that these criminals do not deserve a fair and humane trial, given that this was not given to their own beloved ones (Alamir 2020).

These considerations combine to complicate the PAP further. In the cases explained, the problem is not only that the agents’ incentives are not in the same direction as the principals’ preferences, but also that the latter’s preferences are against social welfare. This could lead policy-makers or advisers to consider paternalistic policies, which brings us to a discussion of biases from the agents’ side.

Behavioural biases from the agents’ side in the TJ context

Consideration of paternalistic behaviour in the TJ context would be motivated by the assumption that not only are the end goals of both principals and agents the same – that is, achieving the TJ goals – but also the behavioural bias is only imagined on the principals’ side. These assumptions are obviously rarely met. Consequently, relaxing the first assumption only, which leads to the possibility of rent-seeking activity by the agents, leads to a rejection of the paternalistic approach. However, there is more to this. Even on the first assumption, relaxing the second assumption leads to the possibility of having the same expected behavioural biases from the principals but from the agents’ side. For example, transitional rulers could miscalculate the long-term risks of TJ, which leads to sub-optimal decisions.

For example, the policy of silencing and forgiving was part of the democratization process in Spain after Franco’s death (Escudero 2014: 131–33). For decades, it was thought that Spain could achieve a successful transition without any form of TJ. However, this policy came back to haunt Spain with what was known as the ‘memory boom’ in the 2000s (Kolon 2021: 1). It is not necessary that the Spanish politicians did not desire a democratic transformation; in fact, to a significant extent the political transformations that followed would defy this claim. However, it could be that transitional rulers at the time underestimated the long-term consequences of burying the past alive. This possibility leaves us with the anti-paternalist conclusion; paternalistic solutions to the TJ dilemma are also rejected, but for different reasons than unconditionally supporting the principals’ preferences. ³⁹ Consequently, paternalistic policies cannot be the answer to the TJ dilemmas in either of the cases.

Finally, given that TJ in the post-revolution context is different from the other public policies that could benefit from paternalism. In a very tense emergency situation, the different parties are asked to choose their movements, while they do not have similar personal experiences to which to anchor their perceptions; there is no credit of trust or reciprocity between them because the past-autocratic regime did not give them the space to interact on a political level. Moreover, there are no established democratic institutions to which TJ policies can be delegated with the guarantee of a checks-and-balances system to prevent a possible rent-seeking activity by these institutions. All we have here is a group of transitional rulers who monopolize the state’s power in this phase – angry victims who are not yet represented through elections and the past-regime trying to spoil the transition. This situation is what I refer to as the ‘TJ momentum’, which precludes resorting to typical solutions under law and public policy.

V. Concluding remarks

During their attempt to move from autocracy to democracy after revolutions, nations witness transitional phases that are full of doubts, compromises and experiments. Like any other kind of transition, democratization is a process that can be anything but short, clearcut or elegant. It is a concept with many aspects. However, usually the legal and political aspects are the first on the agenda of any revolution. One of the major tools to address these is TJ.

To explain TJ as a constitutional arrangement after revolutions over autocratic regimes, I first explain the setup and possible scenarios of the post-revolution phase through real historical examples. In most of the TJ scenarios, victims (principals) desire to achieve TJ goals, while transitional rulers (agents) seek to seize power and either offer a safe exit for the past-regime or eliminate that regime to create their new autocracy.

Consequently, achieving the original goals of TJ – that is, internalizing the negative externalities caused by the consequences of the past regimes’ human rights violations, preventing their repetition and precluding the past-regimes from spoiling the transition – is usually easier said than done.

The asymmetric information between the victims and the transitional leaders creates a principal–agent problem. The constitutional nature of TJ added to the extra-classical dynamics of the principal–agent relationship between the victims and the policy-makers, behavioural biases, especially on the side of the victims, and the nature of TJ momentum in a post-revolution phase in the absence of trusted democratic institutions that could play neutral roles all complicate the work of the traditional solutions to these problems.

This conclusion re-emphasizes the exceptionalism of TJ in relation to ordinary justice in opposition to a tendency in the literature that rejects this exceptionalism ( Bell 2008: 12–13; Posner and Vermeule 2003). It is necessary to realize the roots of TJ’s exceptionalism in order to avoid over-estimating its performance or suggesting solutions to its challenges that simply would not work.

Some scholars have suggested that civil society organizations can be the key to the TJ dilemmas. However, it is important to note that there is no third-party enforcer in such an emergent and constitutional moment. Nevertheless, there could be a third-party informer. This informed party could use the available empirical results, past-nations’ expertise, cost-benefit analysis of the possible options, and legal knowledge to supply the necessary information in the market to both principals and agents. This informed actor should not be a direct party to the bargaining. A national or international civil society could play this role – that is, non-governmental organizations (NGOs), or independent international organizations such as the United Nations, the International Center for Transitional Justice (ICTJ) or the International Institute for Democracy and Electoral Assistance (IDEA).

However, this suggestion is not perfect from a constitutional perspective. First, although civil society may not be a direct party to the bargaining, it is still a party to the general setup, which has preferences that may diverge from or unite with one or more of the other parties. Considering civil society as ‘the communicator, the facilitator, the arbiter’ (Arthur and Yakinthou 2018: 262) ignores that civil society is itself a party to the conflict. It is associated with political parties, international actors and even victims. All it can do is provide more information to the market. This information will probably be biased for the reasons explained earlier; however, the more academic institutions involved, the better the competition between civil society organizations, and the more variable the information distributed in the market will be. This will not create a perfect informed setup, but a better-informed setup. Civil society is then not a veto player, and not an enforcer; rather, it is an informer.

Second, it is worth noting that fixing the information problem does not fix the irrational behaviour problem. Consequently, on the assumption of the unity of preferences between agents and principals, the sound solution may be to observe the democratic procedures in designing these policies, as far as possible, to ensure all conflicting parties and views are included. In this case, the ‘checks and balances’ that they will offer each against the other may balance the potential inefficient behaviour by any of them. Further, both interdisciplinary theoretical and empirical research will be needed to test how far this could be achievable in reality.

These conclusions support the new attempt to reconsider the mainstream assumptions about the role of civil society in transitional justice as a secular, unified, neutral and virtuous player. Although our conclusions sound as if they are pessimistic regarding the role that civil society can play in fostering an efficient TJ process, they support the tendency in the TJ literature and practice to focus on the social transformation aspect of TJ away from the justice/reconciliation dichotomy (Fletcher, Weinstein and Rowen 2009: 218). Civil society in all its forms cannot hold trials or truth commissions, or execute an institutional reform. However, it is the party that can perform the role of distributing information that not only supports the achievement of TJ mechanisms but also spreads awareness about the roots and dynamics of the oppressive regime in a way that addresses the public communities themselves. This role plays a massive part in achieving the social transformation at the constituencies level, where the popular revolutions – and supposedly, transformations – start.

Finally, although not a perfect solution, promoting the role of civil society organizations as informers is the best available solution to foster a self-enforcing TJ process. Merely designing constitutional rules where TJ procedures have clear deadlines, sanctions and authorities to observe are insufficient under the momentum effect. Where constitutional democracy is not yet consolidated, clear constitutional sanctions can be ignored or misinterpreted using the excuse of popular sovereignty, exceptional circumstances, national security or lack of resources. The Arab Spring states, including Tunisia, are a clear example of this. The change that civil society organizations can make to this dilemma, so the TJ policies become more self-enforcing, is that when the principals (the victims) are better informed about the TJ process, this makes the costs borne by the agents (transitional rulers), if they deviate from achieving the TJ goals, relatively higher.