Justifications of repression in autocracies: an empirical analysis of Morocco and Tunisia, 2000–2010

ABSTRACT How do autocrats communicate about repression? Previous studies have analysed how autocratic officials justify the repression of large-scale protests to avoid backlash effects. However, we know much less about how everyday repression against dissidents and ordinary citizens is communicated and justified under authoritarianism. This paper is the first to systematically investigate how officials in autocracies justify, conceal, or deny repression employed by different state actors. It studies the communication of repression in two North African autocracies by analysing the novel Justifications of Repressive Incidents in Morocco and Tunisia Dataset (JuRI). The event dataset contains 439 instances of repression between 2000 and 2010 and disaggregates various dimensions of repression and its communication. The empirical analysis shows how the chosen forms of repression influence ensuing patterns of communication and justification. Studying the communication of repression helps us better understand the nexus of legitimation, judicial repression and political violence in autocracies.


Introduction
Authoritarian leaders do not safeguard their rule by using repression alone.Instead, they employ sophisticated techniques of legitimation and are concerned with their image (Dukalskis, 2021).When autocrats repress mass protests, indiscriminate violence perceived as unjust may backfire, leading to public outrage, the mobilisation of dissent, as well as increased support for the opposition (Hess & Martin, 2006;Mason & Krane, 1989, p. 192).Therefore, state officials often publicly communicate why violence is 'necessary' and right (Edel & Josua, 2018).Existing research has shown that discursive justifications are used in the context of highly visible acts of large-scale repression that have many witnesses (Dukalskis & Patane, 2019).
Going beyond this established focus on extraordinary acts of repression, the paper studies how autocratic officials communicate about everyday repressive incidents outside times of crisis.This contribution analyzes justifications for repression in when dealing with internal oppositions, adding insight into their state-society relations and under-researched causes for their endurance.
The article first presents a conceptual framework of communication and justifications of repression.It then introduces the JuRI dataset on which the analysis is based and sketches the empirical background of the chosen cases Morocco and Tunisia.I then analyse the data and present the main findings and arguments, comparing patterns of communication in the two countries relating to the extent and types of justification, the justifying agents, and the targets of repression.This gives us systematic insight into which justification patterns follow different forms of repression, furthering our understanding of political communication in autocracies.Beyond the focus on protest policing, this article highlights the relevance of judicial repression and everyday repression more broadly as underestimated factors contributing to authoritarian endurance.

Conceptual framework: communicating repression
This section introduces relevant definitions and elaborates on how important concepts relate to one another before delving into why, how and whether repression is justified or otherwise communicated under authoritarianism, with a special focus on judicial repression.I then give an overview of existing literature on the communication of repression in the MENA and outline my own conceptualisation as well as the analytical steps taken in the empirical section.

Definitions
Autocrats seek to ensure regime survival as the ultimate goal of their rule.To this end, they use both legitimation and repression as overarching strategies, which have the functions to garner support for their policies, and to deter dissent, respectively (Gerschewski, 2013).Following the classical definition, repression 'consists of government action which grossly discriminates against persons or organisations viewed as presenting a fundamental challenge to existing power relationships or key governmental policies' (Goldstein, 1978, xxviii).Just as Tolstoy suggested that each unhappy family is unhappy in its own way, each authoritarian regime uses repression as it sees fitwith a particular mix of forms, agents and targets hereof.
Repression may target rogue elites, political opponents, societal groups such as minorities, or the population at large.Autocracies vary in the levels of repression they apply, although they all restrict civil and political rights to some degree.With regard to violations of physical integrity rights, they may use more or less lethal forms.Agents involved in repression vary, as different parts of the security apparatus become active, and it can be applied in a more or less systematic way.E.g. some states may rely more on intelligence services to surveil the population, whereas others may use preventive policing, jail members of certain groups, or employ harsh measures against individual dissidents to deter others.
Repression is a costly strategy for authoritarian elites, one potential cost being the risk of losing legitimacy with audiences who disapprove of coercion, and thus endangering autocratic survival.Authoritarianism research has come to a consensus that even autocratic rulers strive for legitimation with at least certain parts of their populations to ensure some degree of support (Dukalskis & Patane, 2019).Legitimation means the efforts to be regarded as justly exerting rule.Autocrats use legitimation to create support and to 'reduce the need to rely on sheer repression as a mechanism of control' (Wedeen, 1999, p. 26).Beyond rhetorical strategies, legitimation also includes material policies, such as laws and other regulatory outputs, to form a coherent package of discourse and performance.
The concept of justification is related to legitimation, but they are not identical.Justifications are important for the image a state wants to convey, since '[j]ustification is about communicating reasons' (Abulof & Kornprobst, 2017, p. 127).These reasons aim at 'explaining or defending state policies against potential or actual criticism' (Josua, 2021).Presenting justifications is always part of a legitimation strategy, but legitimation as understood above encompasses more than justification, which happens merely on the rhetorical level.
Why do autocratic state officials flank repression with justifications?When autocrats use repression, their aim is to deter dissent and demobilise opponents.However, repression may backfire and lead to the mobilisation of dissent (Hess & Martin, 2006;Mason & Krane, 1989, p. 192).This can happen in cases of indiscriminate violence that the population perceives of as unjustified.Legitimating messages are used to avoid backfire effects, as extant research demonstrates (Edel, 2019).
When autocrats engage in repression and talk about it this might have different effects for different audiences, such as signalling deterrence or garnering applause for the unequal treatment of their citizens from their supporters.This is because all political communication is situated in a relational societal context.Under certain circumstances this may lead to counterintuitive results: '[R]epression can legitimize a regime by signalling to supporters a commitment to hold a hard line against their political rivals' (Lachapelle, 2022, p. 696).Studying justifications of repression bridges the conventional divide between legitimation and repression under authoritarianism (Gerschewski, 2013) and illuminates the complex entanglements of these strategies to ensure regime survival.

How repression is communicated and justified
When devising justifications, state officials draw on the norms that are prevalent in a given society and on the global level.The potential variety of norms allows for even contradictory justificatory arguments.As actors within autocratic regimes have divergent interests, it may be possible that different state officials follow different communication strategies at the same time.Heller and Kahl argue that '[b]y using justifications, actors principally accept their responsibility for action that is usually assumed to be "wrong," but deny the validity of the behavioural norm in the case at stake ' (2013, p. 419).Autocratic elites use rhetoric in two main ways: negative framing delegitimises the opposition while positive framing turns policies that are undesirable for certain parts of the population into a legitimation story (Li, 2022).
Existing research has shown that discursive justifications are used in the context of highly visible acts of large-scale repression (Dukalskis & Patane, 2019).These and other overt forms of repression, e.g.legal restrictions, are so visible that they usually require justification.Other forms of repression are covert, such as forced disappearances or torture (Aguilar & Kovras, 2019).Since they have no legal basis, they are more likely to be denied than justified (Earl, 2003).The same holds true for lethal repression.
Justifications only make up one part of communication, so some repression will be acknowledged, while the bulk of repressive incidents may be denied, ignored, or concealed (Dukalskis & Patane, 2019, p. 474).Even when the public knows about repressive events, autocratic elites might still deny or ignore them.Denial refers to public statements admitting something happened to the target of repression but rejecting any intentionality and the state's role in hurting the victim.In some cases, denial is advantageous because 'if successful, cover-up reduces or even eliminates the need to demonstrate legitimacy' (Hess & Martin, 2006, p. 252).
Diversion involves blaming the victim or trumping up related or unrelated charges against them to deflect from the state's agency.Denial and diversion both alleviate the backfire effect of state violence and are analysed together as non-justifications in this article.So far they have received even less scholarly attention than justifications, with the exception of Edel (2019).

Judicial repression
Besides the forms of repression that involve policing or various types of state violence, judicial repression takes an ever more important role (Shen-Bayh, 2018).Judicial repression occurs when individuals or groups perceived as posing a threat to existing power relations are summoned to court, a prosecutor initiates proceedings, or a full court trial takes place against them.Autocrats use trials against dissidents as a form of repression in itself.A politically motivated court trial is a way of humiliating the defendant and deterring them from future activism.
The advantage of using courts as sites of repression in autocracies is that through the legalistic appearance, repression is more easily justified in terms that are universally acknowledged.Courts have the aura of being impartial institutions and play an important role in maintaining a semblance of legitimacy for an autocratic regime, as examples from around the world show (Moustafa, 2014, 286-287).Leaving the repression of opponents to the judicial system is thus more elegant and appears more legitimate than the exertion of physical violence, although the effects on dissent may be similar (Pereira, 2005).
In the context of judicial repression, the autocratic logic of regime survival still reigns supreme.Autocratic courts do not necessarily adhere to the principles of due process and are thus not instruments of the rule of law.The subversion of law in autocracies has been termed 'rule by law' (Ginsburg & Moustafa, 2008).'Rule by law' refers to the arbitrary implementation of laws, using them or overlooking them according to the needs of authoritarian elites rather than for the primacy of the legal order.In trials, judges often at least implicitly approve of extralegal repression when defendants complain about being exposed to excessive violence, turning a blind eye towards mistreatment by security forces.
A consensus has emerged that autocratic judiciaries are neither mere instruments of the executive without any agency nor quasi-independent, but situated somewhere in between on this spectrum, as most autocratic regimes constrain the independence of their courts to some degree, depending on the relations between institutions (Moustafa, 2014;Pereira, 2005).Given its significance in autocracies, judicial repression is considered separately in the empirical analysis.

Communicating repression in the MENA and my conceptualisation
Existing research on the political communication of repression in the MENA region has mostly dealt with severe forms of violence, where many witnesses observe and document repression.Examples include the massacres of Muslim Brothers in Syria in 1982 (Wedeen, 1999) and in Egypt in 2013 (Edel & Josua, 2018;Grimm, 2022;Lachapelle, 2022;Pratt & Rezk, 2019).Regarding the Syrian civil war, Scartozzi (2015) uncovered how the changing strategic narrative influenced popular support for the regime.Selvik (2018) traced how the regime discourse on 'insiders' and 'outsiders' enabled the repression of the Green Movement protesters in Iran.We thus have some knowledge on the communication of the repression of large-scale protests in the MENA.Most existing works look at speech acts in single case studies.Apart from Edel (2019), who investigated obfuscation as well as rhetorical and procedural justification as strategies altering the perception of repression in Tunisia, the growing literature on judicial repression in the region has not tackled the question of justifications or legitimation.
My argument is that justifications are not only used in situations of high-scale repression, but also in the context of 'everyday repression'.Although a certain baseline level of repression is characteristic of autocracies, in addition to more targeted measures against dissidents, we know little about how it is communicated.To explore to what extent and in what ways autocrats communicate about everyday repression I have chosen the years 2000-2010 in North Africa, a period prior to the crisis mode of the Arab uprisings, although some locally confined protests occurred.This offers a glimpse into the spectrum of justifications available in autocracies below the threshold of large-scale protests and provides us with useful background information when studying more contentious times.
Political communication depends on the prevalent state-society relations and patterns of repression.For analysing their communication, the actual beliefs of state officials are irrelevant; the focus is on what they choose to say.The main finding of this analysis is that the variation in justifications offered is mediated by the different forms of repression that autocrats choose to use.Depending on those forms of repression, divergent patterns of communication ensue regarding the volume and content of the shared justifications.In particular, my arguments refer to regime characteristics in that autocratic officials in monarchies use more justifications than those in republics.The chosen forms of repression are influenced by who the targets are, as regime outsiders experience more severe forms of repression and less justifications than insiders.In terms of overt and covert forms of repression, officials justify visible repression while denying or concealing obviously illegal repression, such as torture, forced disappearance and killings.Finally, the actors involved in communication differ depending on the form of repression.They may include members of the legislative devising repressive laws, but mostly officials in the executive, such as ministers, heads of state and security officials of various branches.Finally, the judiciary assumes a role in the prosecution of dissenters.Based on their specific roles, state officials use different frames.Figure 1 below summarises how I conceptualise the variables relevant to the communication of repression.
The regime type influences to some extent who the agents and targets of repression are and what form of repression is chosen.Based on these factors, the justifying actors either communicate about repression or not.When they communicate, they may choose denial, diversion, or justification.The type of justification is then the final outcome.These multiple theoretical aspects can be mapped for all repressive incidents, although the importance of single factors is hard to pin down.
The analysis of the empirical material illuminates relationships between some of the variables in this figure in more detail.I use the dataset to answer the following questions: how does the communication of repression vary across different regime subtypes?Which types of justification were used in Morocco and Tunisia?Approaching the centre of the figure, what forms of repression were used in the two countries, and which of them were accompanied by justifications, which were denied, and when was there silence?More specifically, how much judicial vs. non-judicial repression was used in Morocco and Tunisia, and did justifications vary accordingly?Did the extent and type of justifications vary depending on the targets of repression?Finally, which justifying actors used which patterns of communication and types of justification?The following section explains the case selection and empirical background of the study.

Case selection and empirical background
Among autocratic regime subtypes, monarchies and republics vary in their use of repression.This variation can partly be explained by the different legitimation strategies at the disposal of monarchs and presidents, where kings have the advantage of claiming traditional legitimacy, while some also strive for a liberal image and avoid too harsh repression (Dukalskis & Patane, 2019).
The MENA is home to many typical autocracies, including monarchies, from which findings can be generalised to other world regions (Flyvbjerg, 2006, pp. 232-233).Also, MENA monarchies have used overall less repression than republics (Yom, 2014).Republics in the region tend to employ not only more, but also more severe forms of repression.In contrast, while monarchies restrict freedom of assembly and speech, they violate physical integrity less (Møller & Skaaning, 2013).To capture this variation, I compare Morocco and Tunisia to elucidate the specific patterns of repression and ensuing communication across a monarchy and a republic.
Until 2010, Morocco and Tunisia exhibited some notable similarities besides their differences in regime subtype.Both the Moroccan king and the former Tunisian president were neopatrimonial rulers in resource-poor countries, concentrating decision-making power in their own hands.Indices on the macro level showed similar degrees of repression during the studied time period, with Political Terror Scale values (ranging from 1 to 5) mostly settling at 3 with occasional fluctuations to 2. Morocco and Tunisia both belonged to those autocracies where the police but not the military were involved in domestic repression (Berman, 2021).Comparable structural factors offer the chance to examine the effect of particular patterns of repression on its subsequent communication.
To outside observers, Ben Ali presented Tunisia as a secular, neoliberal posterchild flanked by a 'rhetorical commitment to democratic procedures and human rights' (Cavatorta & Haugbølle, 2012, p. 187).However, this secularism masked an exclusionary regime that repressed Islamist movements of all shades and leftists.The economic modernisation process benefitted crony capitalists and corruption at the top.The smaller social base existing in such exclusive regimes reduced available alternatives to repression (Rørbaek & Knudsen, 2017).The Tunisian police apparatus was very close to Ben Ali and his preferred instrument of coercion.Ben Ali also used the judiciary as a political tool to get rid of opponents, outlawing the entirety of the opposition.A specific feature of repression in Tunisia was that it was more widespread and broadly targeted against different groups than in other autocracies (Nugent, 2020).
Morocco was a more inclusive society that relied on the king's legitimacy, although significant parts of the population in some Berber regions and the Western Sahara were disenfranchised.While state violence is path-dependent, after the royal succession in 1999 King Muhammad VI established a Truth Commission investigating previous human rights violations in the 'years of lead' as a measure of authoritarian upgrading (Vairel, 2004).However, following terror attacks in Casablanca in 2003 heavy-handed repression returned.Overall, repression was more widespread in Tunisia than it was in Morocco.The analysis shows the ensuing variation in the respective patterns of communication and justification.The period of investigation stops before the beginning of the Arab uprisings in Tunisia in December 2010, as ending the authoritarian 'normalcy' hitherto.

Methodology and data
The article contributes to theory-building of an understudied topic and establishes novel data.The units of analysis are 'repressive incidents', defined as repressive acts targeting individuals or organisations, often involving more than one actor and form of repression at the same timefor example, from arresting somebody, to mistreating them in custody, to bringing them to court.Some theoretical arguments are based on findings from existing literatures, for which repressive incidents drawn from two different countries function as plausibility probes (Levy, 2008: 3).As the limited number of data points in the dataset does not allow for a large-N test of hypotheses, in the analysis I will adopt an outcomecentred understanding in the sense of configurational methods and study where the dependent variable of justification is present to explore under which conditions it takes on different values.
While the sensitive issue of repression is notoriously secretive, the benefit of focusing on justifications is that they are by nature publicised.This alleviates potential ethical pitfalls when researching repression.For mapping the communication of repression in Morocco and Tunisia from 2000 until 2010, I created a dataset on repressive incidents from publicly available documents (Josua, 2022).Together with my research assistants, I systematically screened qualitative data on repressive incidents as obtained from pertinent annual reports on human rights by Amnesty International (AI), Human Rights Watch (HRW) and the U.S. Department of State (USDS).
In the Justifications of Repressive Incidents in Morocco and Tunisia (JuRI) Dataset, we disaggregated and coded various dimensions of the repressive incidents.The codebook with definitions of categories and codes can be found in the Appendix.The data include a description of the incident, geographical data, the supposed reason(s) for repression (which might deviate from official statements), the name and affiliation of the target of repression, the forms thereof used, the agents of repression, the justification and justifying actors.
Data from the main sources, accounting for 348 incidents, were supplemented by more specific reports from local organisations, NGOs working on behalf of journalists, as well as from media reports by newspapers and blogs covering the offered justifications (should they contain additional relevant information).These reports resulted in 91 additional repressive incidents.For coding the justifications, we proceeded in part deductively based on the justificatory arguments found in the literature (e.g.violence, criminal behaviour, public disorder, or terrorism-related justifications).In part, we added further codes inductively that emerged from the data by using in-vivo coding via the software MaxQDA, such as membership in illegal organisations or administrative infractions.Over various iterations we aggregated recurrent justifications to 10 different codes, which can be found in the Appendix.Furthermore, I consulted scholarly works on repression in the two countries to support the interpretation of the data.I cleaned all the data and coded the justifications myself, as here consistent interpretation is paramount.Depending on the primary language of the human rights or news reports, most sources we worked with were in English, French and Arabicwith the latter mainly drawn on to complement the information that was obtained from the English-language reports.While current developments move towards machine learning (Cordell et al., 2022), for identifying fine-grained disaggregation and the nuances of justification our hand-coded approach proved to be useful, especially given the diversity of the sources' languages.
The primary interest behind collecting this data was not quantitative, as the data selection underlies certain limitations (see below).However, the descriptive data give some hints regarding the extent of justified repression as compared to non-justified instances thereof.The dataset contains in total of 439 repressive incidents: namely, 280 for Tunisia and 159 for Morocco over the 11 years under investigation. 1  The reports upon which the JuRI dataset is based recount repressive incidents against activists and organisations, but also excessive violence against citizens by state actors.Some activists, journalists and politicians were targets of repressive measures in multiple years.These incidents were recorded separately.When a repressive incident spanned more than one year, for example, someone was arrested in a certain year while a court trial occurred in a later one, the earliest recorded repressive act counts.Given the fact that our focus is on the justifications and not merely the quantification of repressive incidents, the dataset condenses such cases and takes together repetitions of similar acts against the same person over the span of a year that might otherwise be counted multiple times.That is the case, for example, when a journalist reports being repeatedly followed or otherwise harassed.In some instances, the dataset aggregates lawsuits against various defendants who are tried together in a mass trial.
Among the limitations of the data, most obvious is the reporting bias, as human rights organisations do not provide a full universe of incidents but report on a variety of human rights violations.This coincides with our research interest of investigating the diversity of justifications.One important shortcoming of the human rights organisations' reports is that they can only address instances of repression they actually learn about.Thus, repressive incidents that target well-connected activists are more likely to be documented.This is especially true for incidents entailing low-level harassment, regarding which the 'average' or even marginalised citizen would hardly ever turn to international organisations.Local or diaspora organisations may be connected to certain groups and therefore emphasise some targets of repression more than others.Another bias stems from the availability of reports themselves.While the three main reports (AI, HRW, USDS) are published on an annual basis, those from certain local or more specialised organisations were not available for every year.As a result of these limitations on availability, the data are not easily comparable over the years.
While I identify patterns regarding how repressive incidents are distributed, these numbers should not be considered exact figures.In the analysis, I adopt an outcomecentred understanding in the sense of configurational methods and study under what conditions the dependent variable, namely justification, takes on different values.I am interested in the patterns of different forms of repression and ensuing justifications.Working in our favour is the fact that the under-reporting of incidents affects unjustified repression more than justified repression.The following section presents the main results of the descriptive analysis.

Empirical analysis
The following sections analyse the JuRI dataset containing 159 repressive incidents in Morocco and 280 in Tunisia from 2000 to 2010.I will first present the influence of regime type on the volume of justifications (5.1), then outline the types of justifications found in both countries (5.2) as well as how they relate to the forms of repression (5.3 and 5.4), in particular judicial repression (5.5).Finally, I investigate how forms of repression and ensuing justifications are influenced by certain targets (5.6) and actors (5.7).The tables report a selection of variables that are relevant to the pertinent question introduced in the conceptual part.Not all aspects are disaggregated for the two countries, as sometimes this would lead to very small absolute numbers where results are driven by single incidents.

Regime type and volume of communication
The first step of analysis looks into how the communication of repression varies across different regime subtypes.Variation in the forms and extent of repression used is shaped by regime characteristics.Research has shown that in the MENA, monarchies have used less repression than republics.As republics tend to employ more severe forms of repression, which are less likely to be communicated and justified, we expect and find that autocratic officials in monarchies use justifications more than those in republics do.
This becomes clear when looking at the question of to what extent officials communicate about repression, and, if they do, whether they offer justifications or denial and diversion.Denial or diversion are both treated as non-justifications, although they mean different things practically, with denial rejecting the repressive act and diversion involving gas-lighting or victim-blaming.
For the 439 repressive incidents, in 324 cases we found information about communication (see Figure 2).Tunisia used relatively fewer justifications than Morocco.No communication at all was reported for 17 cases in Morocco and 98 in Tunisia.In 50 instances, what was communicated was not an actual justification but rather diversion or denial (10 in Morocco, 40 in Tunisia).The relatively small share of non-justifications implies that it is advantageous not to mention repression at all when there is no justification at hand.Regarding the argument relating to regime type, this first look at the data shows that Tunisia used relatively less justifications than Morocco.This is in line with the assumption that officials in monarchies use more justifications than those in republics.

Communication and justifications
The second step in the analysis addresses the question: which types of justification were used in Morocco and Tunisia?For investigating these differences, I first present the direct comparison of findings for the two countries before showing the respective results separately.
Reiterating the contrast between how much communication was reported for repressive incidents in Morocco and Tunisia, Table 1 below shows Tunisia to offer lower levels of justification than Morocco.In Morocco, the defamation-insults nexus tops the list of justifications; in Tunisia, the most frequent form of communication is non-justificationthat is, denial or diversion, followed by terrorism-related arguments.Tunisia also stands out regarding administrative control and other legalistic arguments.In Morocco, targets of repression are more frequently blamed for using violence and for protesting.Morocco also uses slightly more moral and religious arguments, and many more national interest or security justifications, where the image of the monarchy comes into play.Figures 3  and 4 below illustrate the domestic rankings of such justifications even clearer, with 100% corresponding to repressive instances with communication only.

Forms of repression
To approach the question of what forms of repression are accompanied by justifications, which are denied, and when there is silence, we first look at the relative share of different forms in the repressive 'mix' that is employed in the two countries.Table 2 below lists the absolute numbers of repressive incidents certain forms were reported for, their share in all incidents, and percentages as relating to Morocco and Tunisia respectively.
One first observation is the importance of arrests and detention, which is connected to the human rights reports informing the dataset calling for action to be taken in these  situations.Harassment and physical violence are also frequent, with higher occurrences in Tunisia.Judicial forms of repressionevident in court trials and convictionswere substantial in many incidents, especially in Morocco.Further common forms of repression are intimidation and torture.Under 'other illegal activities' I subsume manifold tactics such as forced confessions, forced disappearances, sexual assault, arson, evictions, confiscation and theft.Movement restrictions pertain to travel bans as well as administrative control measures in Tunisia, where former prisoners must report to police stations daily or even multiple times a day, making normal life impossible.Financial restrictions were mostly fines imposed in court trials in Morocco, an elegant way of bankrupting journalists and human rights activists.Authorisation regarding media outlets, political parties and organisations was often denied in Tunisia.Killingsmostly resulting from medical  negligence in prisons, less often during protestswere reported in 25 instances.Work-or education-related punishment refers to job loss, disbarment of lawyers, the forced transfer of judges to remote areas, and bans on attending university.Censorship was imposed on newspapers, magazines and radio stations.Finally, denial of medical care was mostly reported for prisons in Tunisia, as were measures of degradation such as slander.

Justifications of different forms of repression
The core claim of this paper is that communication and justifications are mediated by the forms of repression that autocrats choose to employ.Turning to how different forms of repression were communicated in Morocco and Tunisia, the expectations based on the literature regarding overt and covert repression are that autocratic officials justify visible repression while denying or concealing obviously illegal repression.My findings confirm that officials justify visible repression such as court trials, while denying or concealing obviously illegal repression like harassment, physical violence, forced disappearance, the denial of medical care, and killings.Restrictions of personal and civil rights (freedom of movement, of association) are imposed in an arbitrary manner with few justifications given.
Figure 5 lists the share of different forms of repression that were flanked by some form of communication, and whether it was denial/diversion or justification.This leads to a 'net' share of justifications for various forms of repression.Most incidents included more than one form of repression, whereas justifications rarely refer to every form of repression that was used.Looking at the 'net justification,' meaning repressive incidents where substantive arguments were brought forward, for court trials and convictions around 95% of all cases were accompanied by justifications.Likewise, nearly all financial restrictions were imposed in court and thus flanked by justifications.The net rate of justifications for arrests or detention dropped to 74%, while this category comprises both well-reported prison sentences as well as short-term detention and arrests that serve to interrogate and harass activists.Many forms of repression range in the middle zone of 40-70% of net justifications.This is true for forced disappearances, which are subsumed under illegal activities and have a comparatively high share of non-justifications, meaning denial or diversion, as is the case for physical violence.Killings unsurprisingly take the extreme position, with most denials and a distinctly low share of justifications.
Non-justifications were most salient in two larger clusters of the various forms of repression.As expected, one cluster pertained to the use of violence, such as killings, physical repression, various illegal acts including sexual violence and the withholding of medical care mostly of prisoners.The other cluster contains harassment, intimidation, censorship, the arbitrary denial of authorisations, and work-or education-related punishment.In such cases, there is often no paper trail and no hard proof of what agents of the state did or whether they ordered these acts or not.These forms of repression are characterised by (im)plausible deniabilityespecially when the perpetrators are unknown.
When we dig deeper into the various forms of repression, Table A2 in the Appendix shows that when people were killed in Morocco, in 7 out of 10 cases preceding violence on their part was invoked as a justification, or diversion.The argument that the victims were violent is also the most prevalent justification offered when they endured physical violence, followed by diversions and denials.The most important justification in harassment cases in Morocco was arguments about morals/religion.Many cases of harassment and physical violence occurred without any public statement being made.Court trials and convictions mostly used charges of defamation, insults, or spreading false information in Morocco.Tactics of intimidation were connected with missing authorisation and membership of illegal organisations, where repressed journalists and activists defied the bans and restrictions on their work.Lastly, torture was mostly used when defendants were accused of terrorism or other criminal offenses.In these cases, the mentioned justifications were the basis for arrests or prison sentences of the defendants, while torture as an additional form of repression they were subjected to was not justified as such.
In Tunisia, the picture is more extreme.Table A3 in the Appendix demonstrates even fewer justifications in cases of harassment, intimidation, physical violence and killings, and a substantial number of arrests without communication.Most justifications were terrorism-related, and in many of these cases, the defendants were also tortured.A report by Tunisian human rights organisations on counterterrorism trials shows that evidence of torture was often ignored or even denied by judges (ROJ, 2016).Among 92 reported terrorism trials, 70 lasted for less than one hour, demonstrating how judicial repression unfolded (ROJ, 2016, p. 12).Courts often failed to inform defendants of their right to a lawyer or did not designate them a defence lawyer (ROJ, 2016, pp. 14-15, 20).
Regarding other charges, judicial processes varied greatly.The charges brought against political activists were often based on laws curtailing freedom of speech, although in addition there were trumped-up charges (Hudáková, 2021).Frequent justifications included law and order narratives, charges of criminal offenses, as well as defamation, insults, or spreading false information in connection with court trials and convictions.The form of repression that was communicated about the least was denial of medical care, which often occurred in prison.In addition, movement restrictions in terms of administrative-control measures were only rarely brought up publicly.Finally, denials of authorisation were mostly left without communication.

Justifications for judicial vs. extrajudicial repression
As judicial repression turned out to be important in many repressive incidents, I have analysed this aspect separately to uncover the differences in communicating judicial vs. extrajudicial repression.Since legal charges count as justifications, judicial repression includes a high share of justifications.In contrast, non-judicial repression is most likely to be denied.When there are no grounds for judicial action, justifications often point to morals, religion, or the national interest.
Judicial repression covered nearly half of all reported incidents, but was distributed unequally across the two countries.As Figure 6 below shows, in Morocco judicial repression made up two-thirds of all incidents, whereas in Tunisia almost 60% of all repression took place extrajudicially.
Table 3 below gives an overview of the content of the justifications used in judicial repression and contrasts them with the share among cases of non-judicial repression.
In non-judicial repression, diversion and denial make up around 40% of communication.This means that there are justifications for only 60% of all non-judicial repression as opposed to for 97.6% of judicial repression.The only justifications that are used more in extrajudicial than in judicial repression are moral and religious arguments, ones of national interest or security, and administrative infractions, which are the domain of the police force.With relation to the criminal offenses that the defendants in judicial repression were accused of, the most prevalent justification was defamation etc., followed by various other crimes.This shows a marked difference in the justification of judicial versus non-judicial repression.

Target-related differences
The question who the targets of repression are influences the forms of repression employed against them.Given that legitimation and repression strategies address different societal groups, people who are defined as outsiders and who do not enjoy equality in the eyes of the majority population experience more severe forms of repression.This relationship has even been demonstrated with regard to the support for torture of outgroup members in democracies (Conrad et al., 2018).In terms of the volume of justifications and communication, I found for Morocco that the repression of regime outsiders is less often justified.
While it is often hard to disentangle attitudes in heterogeneous societies and the data for Tunisia are not fine-grained enough to test this assumption there, this pattern can be illustrated using Morocco's territorial conflict with Western Sahara. 3We analysed the data for Moroccan and Sahrawi targets of repression separately, identifying 39 Sahrawi and 120 non-Sahrawi targets.Sahrawi individuals and activists made up 24.5% of all targets of repression, however, their share among all victims of torture and physical violence amounted to 38%; and they represented 30% of all those killed.Sahrawis were thus over-represented as victims of severe repression.
The justifications that were used against Sahrawi versus non-Sahrawi targets show some distinct differences.Even in absolute terms there was more denial and diversion with regard to Sahrawi victims than there was vis-à-vis Moroccans.Combined with the reduced communication shares, this means that substantial justifications were only given for 69.2% of repressive incidents against Sahrawi targets, as opposed to 86.6% for Moroccan targets.
A comparison of justifications in Figure 7 shows that Sahrawis were disproportionately more often accused of 'inciting violent protests,' mostly in connection with pro-independence activism.Protests were more frequent in Western Sahara than on Moroccan territory.The defamation-insults nexus and criminal charges that characterise a great deal of judicial repression in Morocco were less present in Western Sahara.
The different types of justification were related to the more severe forms of repression employed by Moroccan security forces.The authorities seemed to be less concerned with justifying them, as the separatists were victims of 'othering' and largely remained under the radar of national and international audiences.The high level of justifications was reserved for Moroccans, while Sahrawis were treated as second-class victims of repression.

Justifying actors
The final question to consider is how the contents of justification are influenced by the actors involved in communication.Depending on the chosen form of repression, different justifying actors communicate about it.They may include legislators but more often officials in the executive, such as ministers, heads of state, security officials and members of the judiciary. 4Based on their specific roles, autocratic officials use different frames.I found that security sector officials justify repression by referring to national interests and security as well as administrative or other legal infractions.Government officials often deny repression; when they do justify it, they most often refer to morals or religion.Judges and prosecutors often use justifications centred on law and order or criminal offenses, and raise charges related to defamation, insults, spreading misinformation, or terrorism.
We identified 329 justifying actors for the 324 incidents where the dataset reported communication, as in some instances different persons or institutions were involved in one incident.The bulk of identified actors included courts, judges, or prosecutors, underlining the important role of the judiciary in retrospectively justifying and thus enabling repression (Shen-Bayh, 2018).Police and law-enforcement personnel engaged in communication in 13% of these incidents. 5Government officials and parliamentarians were active in 21.5% of communication.Cabinet members merely figured as justifying actors in cases involving prominent journalists or the like.In only two cases in Tunisia did the head of state communicate about certain incidents, while this was not reported for Morocco.As the focus was on justifications by state officials, we recorded pro-regime media and other actors close to but outside the government in only 2.5% of justifications.
Table A4 in the Appendix shows what forms of repression were communicated by which justifying actors.Strikingly, governmental officials communicated most about incidents of killing and censorship.Denials of authorisation and work-or educated-related punishments were largely communicated by them too.Judicial actors justified the fines they imposed, initiated court trials and convicted defendants.Pro-regime media engaged in degradation measures damaging the moral reputation of dissidents.Finally, police and law-enforcement personnel communicated in very few incidents relating to harassment, physical violence, killings and movement restrictions entailing administrative-control measures.In Table 4 below, I analysed which actors picked what justifications most frequently.
The results show that judicial actors mostly use justifications in their 'natural' domains, such as criminal offenses, law and order, the defamation nexus, and terrorism-related charges.Again, government officials overwhelmingly used non-justifications.This pattern was evident in many instances where reports noted that the 'authorities denied' state involvement or any negative outcomes for the victims.Given that government officials often communicated about killings, this is hardly surprising.Their classic strategy was promising investigations of repressive acts without ever following up on them.Police and law enforcement personnel mostly referred to administrative infractions and other legal issues, in addition to national interest and security.Arguments pertaining to morals and religion were mostly used by pro-regime media, although the number of cases is small.

Conclusion and areas for further research
This study presented a first systematic analysis of how repression is communicated and justified in autocracies.The novel event dataset offered a more disaggregated view on repression and its communication than large-N studies, while unravelling more systematic patterns than a single case study could do.Although only representing two cases, even this narrow empirical basis of data supported various theoretical arguments on how the chosen forms of repression influence the related communication.The distinct regime types of Morocco and Tunisia led to different dynamics, as monarchies communicate and justify repression more than republics do.Regime outsiders face more severe forms of repression and encounter fewer justifications.Regarding chosen forms, autocratic officials justify visible repression such as court trials, while denying or concealing harassment, physical violence, forced disappearances, the denial of medical care and killings.Restrictions of personal and civil rights were imposed in an arbitrary manner, with few justifications given.
Regarding the actors involved and contents of justifications, security officials justify repression by referring to national interests/security as well as administrative or other legal infractions.Government officials often deny repression; when they do justify it, they most often refer to morals or religion.Finally, judges often use justifications centred on law and order or criminal offenses and raise charges related to defamation, insults, spreading misinformation, or terrorism.A central finding highlights the importance of judicial repression, with non-judicial instances more likely to be denied.When there are no grounds for judicial action, justifications often focus on morals, religion, or the national interest.The extensive use of judicial repression and legal justifications begs for greater scholarly attention going forward.
The main differences between the patterns of communication in the two countries were mediated by the distinct forms of repression that were used.The choice of form (s) in turn was influenced by regime-specific characteristics, with decidedly more severe repression and less communication in the Republic of Tunisia than in the Kingdom of Morocco.Some justificatory arguments in Morocco were tied to the king's claim to legitimacy.Thus, insulting him or the monarchy was a big issue.
In general, the style of repression was more sophisticated in Morocco.This was evident in the stronger reliance on judicial repression and ties in with the general model of authoritarianism in Morocco, which is more liberalised and smarter with regard to the cooptation of political elites, not least in the 'alternance' rotation in government.This contrasts with the exclusionary governance in Tunisia that was reflected also in a larger share of extrajudicial repression.Repression under Ben Ali was more in the shadows, with pervasive online surveillance, which unfortunately could not be traced in my data.Also leadership seemed to matter, as in the 2000s with the new king, 'Morocco has renounced certain repertoires of severe repression' (Berman, 2021, p. 738).In contrast, in Tunisia the honeymoon phase after Ben Ali's ascent to power had long passed.That being said, both states refrained from employing large-scale lethal repression given other means of repression and strong ties to Western countries, vis-à-vis which they maintained a façade of benevolent authoritarianism.
The two cases were in line with conceptual considerations and corroborated previous findings by the literature, and thus represent typical cases of how repression is communicated in the MENA.Future research should study a larger number of cases also from other world regions to test the theorised relations with additional data and to allow for more generalisations.
The study faced various limitations in terms of scope.In autocracies, the media operate under limited freedom and repressive incidents are often concealed.Unless judicial action is taken, government officials are reluctant to mention repression while reliable sources reporting them are hard to find.Data from independent media sources on certain instances was limited for the years studied.Such under-reporting could be alleviated in future studies by taking into consideration social media usage, which took off with the Arab uprisings.
Going forward, one promising line of investigation would be to focus on frame resonance, analysing which kinds of justifications target which audiences and whether their contents varied based on who the respective addressees were.Given the nature of event data used, this was not feasible here.However, audiences are a key element for understanding which justifications stick, are accepted, and thus further routinised, or which of them backfire because of their lack of credibility.This is a vital for taking the relational nature of state-society interactions seriously.
This article has contributed to our understanding of autocratic survival and linked the literatures on state repression and judicial repression.While we have found which actors tend to communicate about certain forms of repression, what remains a black box is how autocratic officials decide on whether and what to communicate, and whether to route repression through the judicial system or not.Learning more about framing strategies more generally could add an important element to the study of authoritarianism in the future.A takeaway message for research on political violence is the benefit of a holistic and disaggregated perspective on different forms of repression, the multifaceted actors involved, as well as on the interconnections with legitimation.Beyond policing, a focus on everyday repression and judicial repression adds to explaining the endurance of autocracies.In this sense, disentangling the communicative and potentially legitimating aspects of repression can be fruitful for authoritarianism scholarship., 2019, p. 47).The Tunisian Human Rights League recorded 1028 complaints concerning human rights violations during the first decade of the new millennium, including economic and social grievances (Avocats Sans Frontières, 2012).2. For example, torture as such was hardly ever justified, instead the dataset recorded the justification in the legal charges that victims of torture faced.3. The dispute between Morocco and the Polisario organization was not evaluated systematically in the dataset since at the time the USDS recorded human rights violations in the Western Sahara separately.However, incidents mentioned in the AI and HRW reports are included in the dataset, and Sahrawis were also targets of repression on Moroccan territory.4. The overwhelming majority are civil courts; only in 11 repressive instances were military courts explicitly mentioned.5.The police and other security forces were named as the perpetrators of repression in half of the reported incidents.

Figure 1 .
Figure 1.The communication of repression in autocracies.

Figure 3 .
Figure 3. Justifications in Morocco as a share of all communicated repressive instances.

Figure 4 .
Figure 4. Justifications in Tunisia as a share of all communicated repressive instances. 2

Figure 5 .
Figure 5.The share of communication and denial/diversion for different forms of repression.

Figure 6 .
Figure 6.Judicial repression in Morocco and Tunisia.

Notes 1 .
To put those numbers into perspective, Tunisia's Truth and Dignity Commission (Instance Vérité et Dignité, IVD) received 62,720 complaints covering the period between 1955 and 2013 (IVD

Table 1 .
Comparing the shares of disaggregated justifications.

Table 2 .
Forms of repression.

Table 3 .
Communication on Judicial versus non-judicial repression.

Table 4 .
Justifications that the different actors use the most.

Table A4 .
What forms of repression were communicated by which justifying