Arbitrary Decision-making and the Rule of Law: The Role of the Jurisdiction in Migrants’ Detention Proceedings – Between Discretion and Arbitrariness

Many studies have highlighted a substantial "bureaucracy domination" in procedures relating to migrants’ access to territory. This form of domination is marked by highly discretionary and arbitrary practices, enacted by the administrative authorities of the state. Only minor attention, however, has been devoted to the arbitrariness of judicial decisions and to the judicial role in general in the numerous proceedings that increasingly affect the path of migrants. This path is the main object of this paper. The study focuses on Italian case law in expulsion and detention proceedings of irregular third country national citizens and asylum seekers and presents qualitative empirical research on decisions issued by the competent national authorities. The results have been analysed using a selection of theoretical tools, all referable to the general concept of the rule of law. The judicial decisions on pre-removal detention proceedings in two case studies are examined: the jurisprudence on detention of irregular migrants, in different offices of the Justice of the Peace in Italy; and the case law on detention of asylum seekers in the Ordinary Tribunal of Rome. The assumption underlying the research is that various conceptions of the rule of law may have different explanatory power when it comes to explaining the empirical results. To verify this hypothesis, the study proposes an overview of the main rule of law doctrines in the Western tradition of political and legal thought and applies the method of historical-conceptual analysis. As a result, the explanatory power of six theoretical models of the rule of law was verified against the data with the view to highlight the virtues and vices of the respective explanatory frameworks. 
     This article reaches a two-fold conclusion. First, as far as the explanatory frameworks are concerned: the results of the two case studies cannot be fully explained by any of the models considered in this study. This fact alone casts doubts on the explanatory power of these theories and calls for further research on judicial decision-making more generally. Secondly, a key finding of the study regarding the notions of discretion and arbitrariness is that the judicial approach which assures the highest protection of rights is also the one that is most easily influenced by arbitrariness. The author argues that this paradox can be easily dissolved by paying attention to the plural dimensions of arbitrariness. If we consider arbitrariness from a legal point of view, i.e. as an illegal decision, it is unsurprising that the authority that most uses its discretionary powers is also the one most at risk of abusing these discretionary powers and hence of exercising arbitrary power. However, if we consider arbitrariness from the point of view of philosophical-political theory, i.e. as a form of domination characterised by the absence of sufficient justification, it is unsurprising that the judicial approach which assures the highest protection of rights is also the one that takes its own role as guarantor of these rights and of the constitutional democratic legal order as such most seriously. This judicial approach thus most often risks exercising its power in criticisable ways, as compared to an authority much more in line with the requirements of law enforcement agencies. 
Keywords: migration, discretion, justice, arbitrariness, civil rights, Rechtsstaat, expulsion, mixed constitution

arbitrariness. The author argues that this paradox can be easily dissolved by paying attention to the plural dimensions of arbitrariness. If we consider arbitrariness from a legal point of view, i.e. as an illegal decision, it is unsurprising that the authority that most uses its discretionary powers is also the one most at risk of abusing these discretionary powers and hence of exercising arbitrary power. However, if we consider arbitrariness from the point of view of philosophical-political theory, i.e. as a form of domination characterised by the absence of sufficient justification, it is unsurprising that the judicial approach which assures the highest protection of rights is also the one that takes its own role as guarantor of these rights and of the constitutional democratic legal order as such most seriously. This judicial approach thus most often risks exercising its power in criticisable ways, as compared to an authority much more in line with the requirements of law enforcement agencies.
Keywords: migration, discretion, justice, arbitrariness, civil rights, Rechtsstaat, expulsion, mixed constitution Many studies have highlighted a substantial "bureaucracy domination" (Sager 2017) in procedures relating to migrants' access to territory (Landman 2002;Heyman 2009;Alpes and Spire 2014). This form of domination is marked by highly discretionary and arbitrary practices, enacted by the administrative authorities of the state (see Lipsky 1980). Only minor attention, however, has been dedicated to the arbitrariness of judicial decisions and to the judicial role in general in the numerous proceedings 1 that increasingly affect the path of migrants (Ramji-Nogales, Schoenholtz Schrag 2007). This path is the main object of this article. 2

Proceedings on Expulsion of Migrants Before the Rule of Law: An Empirical Research Study
The study focuses on Italian case law in expulsion and detention proceedings of irregular third country national citizens and asylum seekers 3 and presents qualitative empirical research on decisions issued by the competent national authorities. The results have been analysed using a selection of theoretical tools, all referable to the general concept of the rule of law. In particular, I have examined the judicial decisions on pre-removal detention proceedings in two case studies: the jurisprudence on detention of irregular migrants, in different offices of the Justice of the Peace 4 in Italy; 5 and the case law on detention of asylum seekers in the Ordinary Tribunal of Rome. 6 The Italian legal system requires double jurisdictional protection in matters of expulsion of non-nationals: an elective judicial remedy regarding the substance of the expulsion act, and a mandatory proceeding on the lawfulness of the expulsion's executive measure in cases of forced deportation. 7 In particular, the data collected and analysed involve the detention hearings of validation/rejection in the centres for deportation (CPR). 8 The Italian legal framework on the detention of nonnationals has been subject to numerous modifications, most recently with the socalled "Salvini Decree", 9 now conversion law no. 132/2018. The legal regime concerning the detention of irregular migrants and asylum seekers differs to some degree, 10 nonetheless, without entering into the details of legal analysis, it is important to note that in both cases detention hearings are ruled by the same summary procedure. This discipline is characterised by a reduction of formalities in the trial, which leaves more discretionary power to the judge. 11 The judgement expressed on the validation of detention and other executive measures concerning the expulsion falls within the competence of the civil jurisdiction, even if hearings are held according to a typical criminal law model, such as the application of precautionary measures affecting personal liberty. This "anomaly" is due to the coercive feature of these measures, included among the restrictions to personal liberty regulated according to Article 13 of the Italian Constitution. 12 Therefore, the judge has a controlling function over the activity of the administrative authority and operates as a limit to state power in its ability to expel and detain nonnationals. 13 As the empirical study shows, the relationship of the judicial authority to that of law enforcement is a key factor, which is able to explain the main differences in the approach of the judges. Moreover, the relationship between the branches of government constitutes a core point in the current debate on the rule of law, which consequently offers the paradigm of reference for reading the data.
Over the last few years, the rule of law has become the core of a rich scientific debate, and it is considered by several scholars to be a model affected by crisis in both its principal elements, i.e. constitution and legislation (Ferrajoli 1995;Chevallier 1998;Ferrarese 2000;Santoro 2008). With the introduction of rigid constitutions and the implementation of European and international laws, 14 the traditional-continental conception of judicial power has changed, and the role of judges in society is now called into question. 15 In this context, it is even more evident that the idea of the rule of law has a polymorphic character; therefore it is necessary to start with a preliminary analysis of the different meanings of this notion in relation to its various historical embodiments. The assumption underlying the research is that various conceptions of the rule of law may have different explanatory power when it comes to explaining the empirical results. To verify this hypothesis, the study proposes an overview of the main rule of law doctrines in the Western tradition of political and legal thought and applies the method of historical-conceptual analysis (Begriffsgeschichte). As a result, the explanatory power of six theoretical models of the rule of law was verified against the data the research has collected. 16 Although sometimes these models appear outdated (e.g. the Rechtsstaat), it is still interesting to note the partial explanatory power that these "outdated" ideas about the rule of law have when tested against the case studies.

Rule of Law and Arbitrary Power: The Pure Arbitrary Justice Paradigm
Arbitrariness of power has always been an open question in the Western philosophical tradition, 17 and the rule of law formula in all its various meaningsstarting from the famous distinction between the "rule of law" and the "rule of man" 18 -has offered the principal answer to the challenge of arbitrariness. The idea of the history of the rule of law is marked by a continuous dialectic between the need to ground judicial decisions in the law, running the risk of formalism, and the necessity to situate justice with regard to the individual case, running the risk of arbitrary and unequal decisions (Taruffo 2002: 177-178) 19 . The problem naturally concerns the role of the judge in adjudication, one of the most common subjects in the current legal-philosophical debate. Already in the 19 th century, Jeremy Bentham observed that the exercise of interpretative powers by the judge violated the principles of the rule of law (Bentham 1988;Endicott 2014). The doctrine of literal interpretation of the law offered a possible way out, but in adapting to the intention of the lawgiver, the lawmaker may also act arbitrarily. In fact, as Timothy Endicott states, there is no specific reason to consider the creative power of judges to be more arbitrary than that of the legislator: there are simply different types of law, some made by judges, some by the lawgiver (Endicott 2014: 54-67). 20 So far, arbitrariness of verdicts or decisions made by judges appears to be a particularly complex subject, involving other important issues, such as the interpretation of law and the limits to the discretionary powers of judges. 21 A good starting point for addressing the problem of arbitrariness in judicial decisions may be the notion of discretion in the legal context. Let us start with the definition given by Aharon Barak, who describes discretion as the power awarded to a person with authority to assess between two or more alternatives, specifying that all of those are lawful (Barak 1989: 17-20). 22 Often discretion is defined in a "negative way", through the preliminary discussion of its limits, distinguishing discretion from arbitrariness. 23 Barak indicates two kinds of limits: procedural and substantive limits. Procedural limits address forms of decision-making and can be summed up by the term fairness; substantive limits concern the relevant elements for judicial assessment and can be summarised by the notion of rationality (Barak 1989: 27-30). 24 However, the distinction between discretion and arbitrariness is not very useful for explaining judicial decisions in the two case studies. As I will explain later -except for a few examples where an act is "illegal", or unlawful because of the absence of legal reasoning in the motivation 25 or because it is unlawful in substance -most of the time the decisions appear to be legitimate and within the discretionary powers attributed to the judge, who selects from among a range of possibilities, and conform to the procedural and substantial requirements of the law. This is a first relevant element emerging from the research, since the issue of the administrative detention of migrants has been defined by some scholars as an example of "special law in a state of exception" (Agamben 1995;2003). 26 The observations in this study, however, do not seem to support this being the case, because the institution of detention of migrants is nowadays well established in the constitutional framework and also regulated by a rather thick legal regime (Rigo 2017). Therefore, it seems more useful to read the judicial decisions in light of the notion of discretion rather than the notion of arbitrariness. This position is strengthened by the results of applying a pure arbitrary model of justice to the empirical data, well exemplified in the "Justice of the Khadi", 27 . This model of justice is well-described by Max Weber (Weber 1978) with reference to materially and formally irrational forms of law, where judicial decisions are made on the basis of ethical or emotional considerations in the first case, or founded on divine interventions and on the authority of religious figures in the second case (Marra 2005). In both cases, no formal or substantial limits to adjudications are in place, except to the "hallowed" laws, based on obeisance to a specific authority (Weber 1978: 809-815). The historical reference is to all theocratic and patriarchalauthoritative systems, but also to some democracies, such as the popular justice in Attic democracy, or more modern forms like the approach of the English Justice of the Peace (Weber 1978: 811-814). In a system of this kind, we would expect to find a great variety of unpredictable judicial decisions. This is not, however, what happens in the case of the Italian Justice of the Peace. In fact, one of the most interesting elements arising from the empirical research is the uniformity of the case law made by the Justice of the Peace. On the other hand, the Ordinary Tribunal of Rome shows a higher variability of results in adjudications that are less predictable, but that generally present a good level of legal reasoning, an element that we would not expect to find in an arbitrary and irrational form of justice.

The Formalism of the Rechtsstaat: A Bureaucratic Justice
The second model considered is the nineteenth century European-continental conception of the rule of law, developed in the doctrine of the so-called Rechtsstaat, 28 which also spread to France in the theory of the Etat de Droit. 29 Fundamental to this model is the principle of conferral of power by law and the exercise of power according to the procedures and forms established (Ferrajoli 2002: 349). The model is described by Max Weber as rational and formal, and functional for the development of capitalist society (Weber 1978: 652-659). The main values of this paradigm are legal certainty and foreseeability of judicial decisions, which are also safeguards for the citizenry against the arbitrariness of sovereign power. 30 Although individual rights are always subject to the state's will, a fundamental assumption of Germans jurists is in fact the idea of the "state's voluntary restraint" 31 in granting rights. Rights, in this view, are therefore conditional and not absolute. In this context, jurisdiction is the expression of a rational-legal power, dominated by the formalism of impersonal bureaucrats: the "bureaucrat" exercises his work "without passions", regardless of who stands before him; the bureaucrat uses "specialized knowledge" as a primary tool. According to this perspective, judicial activity is not too far removed from bureaucracy, which gives jurisdiction and administration a similar character (Weber 1978: 644-645). As far as this model of the rule of law is concerned, adjudication appears to reproduce the lawmaker's intention in the enforcement of the law (Di Donato 2010: 463), and it would therefore be more precise to speak of "rule by law" rather than "rule of law". The same principles are partially replicated in the French version of the Etat de Droit, where judicial power and the judge are seen as the bouche de la loi. The more extreme expression of this doctrine is the judicial syllogism and the "mechanical" application of the law. 32 When tested against the backdrop of the approach by the Justice of the Peace in the empirical study, the explanatory power of this theory appears to be quite high. If the bureaucratic model of justice were thoroughly applied, we would expect to see little or no legal reasoning, uniformity in the case law and strong limits in the discretionary powers of judges. This is quite similar to the results emerging from the analysis of the detention proceedings by the Justice of the Peace. The general scenario that arises from analysing the hearing minutes reflects an extremely summary and superficial proceeding, which can rightly be defined as "bureaucratic". Hearings last on average between 5 and 10 minutes, hearing reports are most often reduced to a few lines written on an outlined form, where each part of the process has a preset section. 33 Finally, the hearing report has a check box reserved for the judge with two options, "validation" or "rejection", and a specific section for legal reasoning. This last is very often limited to a "ritual formula" such as "the judge considers the detention requirements to be fulfilled", or "the judge considers the instances advanced by the law enforcement agency (Questura 34 ) to be well founded." 35 This modus operandi necessarily impacts the outcomes of adjudications. In 80-90% of the cases the detention of the migrant is validated. The most extreme example is the Office of the Justice of Peace in Turin, where the validation of detention occurs in 98% of the cases. 36 Furthermore, as a general rule, validation and extension hearings are held inside detention centres instead of in the offices of the judges, a practice which highlights how influential are practicalorganizational reasons put forward by the law enforcement agencies. In fact, the agencies have the effective power to establish the location of the hearings. 37 By contrast, the adjudication practices of the Ordinary Tribunal reflect a different landscape: the outcomes are equally balanced between validation and rejection and they vary to a great extent. Hearings are characterised by a more indepth inquiry, and decisions present a detailed legal reasoning. The results of the case studies are therefore not really intelligible according to this theoretical model, which thus only has partial explanatory power.

The English Rule of Law: The Protection of Civil Rights Throughout the Courts
The third theoretical framework considered here is the rule of law in the English tradition, an example of a "thick" or substantial idea of the rule of law, where public powers are regulated by the law not only formally but also in substance. This model requires the existence of an unavailable sphere of competence for state action -also called iurisdictio -38 that ensures the protection of fundamental rights. 39 It is a more ancient tradition than the Rechtsstaat and finds its roots in medieval constitutionalism, 40 and therefore it presents some peculiar aspects compared to the European-continental rule of law model. 41 Jurisdiction evolves in the English context as an autonomous power: courts assume a leading role in claiming and strengthening a series of rights and liberties, thanks to their interpretation of the common law -the customary laws of the land. 42 The case law system becomes complementary to Parliamentary sovereignty: even if the latter has wide discretionary power to adopt any act, the interpretation of the law is assigned to the judges who apply it in compliance with the common law. In this perspective, the judiciary conception confers broad discretionary powers to the judges, which are not considered arbitrary, because they play the role of guarantor against political power. Judges therefore enjoy an autonomous "law-making authority" geared to protect individual rights, thanks to the "jurist's artificial reason". 43 This type of rationality differs from the "reason" of Enlightenment thought: it is a discursive reason that understands law as "a legal reasoning language", a special lexicon used by jurists to justify adjudication (Santoro 2008: 110).
According to this paradigm, justice should be characterised by consistent legal reasoning, a contextual variability in adjudication and should exercise the role of controlling the activity of the administration, with the aim to protect individual rights. This model does not capture the policy of the Justices of the Peace. Indeed, judges provide rather shallow legal arguments and do not play an independent role in relation to administrative instances, to the detriment of the rights of migrants. Instead, this model seems more appropriate to capture the jurisprudence of the Ordinary Tribunal: even if procedural rules are the same, detention proceedings before the Ordinary Tribunal of Rome appear to be different from the practices of the Justice of the Peace. Tribunal hearings are not located inside detention centres, but in judicial offices. The general landscape emerging from the analysis of the minutes of the hearings is quite a detailed and rather complex process as shown in the precise development of legal arguments. 44 Frequently, the judge conducts a proper interview of the asylum seeker, asking about relevant elements such as residential and working conditions, his or her personal story, with the aim to evaluate if the request for international protection is merely instrumental to avoid or delay deportation. Judges sitting on the Tribunal use their discretionary powers entirely to conduct hearings and to apply the law; the legal reasoning is fully developed in the motivations offered in judges' decisions, 45 with consideration of each individual case. As a result, the outcomes of the adjudication are more balanced: the detention of migrants is not a foregone conclusion, since in approximately 50% of the cases, judges decide to reject the detention request. 46 Thanks to more fully exercising their discretionary prerogatives, the magistrates working in the Ordinary Tribunal are able to elaborate judicial decisions and guidance that take into consideration the protection of fundamental rights. This approach is well exemplified by the jurisprudence of the Ordinary Tribunal in Rome concerning the recognition of international protection to female Nigerian asylum seekers. 47 In fact, the policy has a direct influence even on detention proceedings that involve mostly women, in particular from Nigeria. 48 Most of the time, Nigerian detainees were released. In their arguments judges sometimes specifically refer to this jurisprudence: "Since in the case law of this Tribunal [...] asylum request of women from Nigeria is granted in the form of subsidiary protection, because of the high risk in the country for women to become victims of human trafficking" (Tribunale di Roma 2018 n.r.g. 9190). The decision-making criteria adopted by the judges shows the attention paid to individual rights and the assumption of an independent role in relation to the administration. One case illustrates this perspective in a particularly meaningful way. Here, the judge tacitly reveals his attitude favouring a more guarantee-centred approach and defines the interpretative guidelines for detention legal regimes; those norms need a restrictive reading since they affect personal liberty according to Article 13 of the Italian Constitution. 49 In conclusion, this theoretical model has significant explanatory power as far as the case study is concerned, nonetheless on some points the jurisprudence of the Tribunal contains inconsistencies and critical aspects: noteworthy are the differences in the quality of decisions taken by the judges and in the legal arguments. The variations do not seem to merely be the result of the concrete circumstances of the case in point, as I will clarify below.

The Mixed Constitution Model and the Social Conception of Justice: Risk of Arbitrariness
Another theoretical model considered in this article can be defined as "the social equality of the mixed constitution". There is no specific historical reference here, but different relevant conceptions refer to the balance of power between competing social forces, starting with the ancient idea of the mixed constitution. 50 In this classical model, distribution of power was understood to consist essentially in a relationship between the social classes, expressed in inter-institutional terms. 51 This idea was widely debated by scholars in a variety of historical settings, crossing Western philosophical thought from ancient times 52 to the Middle Ages; 53 it is well represented by the example of the English monarchy. 54 Montesquieu mentions this institutional model while analysing the French system of "intermediate powers" (Goldoni 2013) in his De l'esprit des lois, a manifesto of the separation of powers doctrine. 55 With the introduction of the theoretical tool of sovereignty in the modern theory of the state, the tradition of the mixed constitution was abandoned (Portinaro 2007: 162), even if it can be traced to some contemporary conceptions of the so-called welfare state. 56 Indeed, the recognition of a range of social rights on the constitutional level has reopened the debate on balance between the social classes, a question now fully integrated into the judicial point of view. However, the discussion on the effectiveness of social rights is clearly a different and peculiar perspective, since it is inserted into the constitutional framework. In a model of this kind, without additional elements, judicial power can be seen as an autonomous institution aimed at preserving the equilibrium among the social forces. Adjudication would be based mostly on social or extra-legal considerations and therefore might be arbitrary. This theory does not explain the case law of the Justice of the Peace. In fact, the role of the judge is here reduced to complying with law enforcement agencies, as the legal arguments per relationem show. This term described the routine of the Justice of the Peace to introduce an explicit mention of the arguments advanced by the law enforcement agency "Questura" in the legal reasoning in the motivation, without adding any other element to justify the validity of detaining the migrant.
Instead, considering the adjudication practice in the Tribunal, there are some cases where judges based their assessment on social or extra-legal factors, rather than legal requirements. As an example, in two decisions (Tribunale di Roma 2018 n.r.g. 8738; 8740), the judge links the evaluation of the instrumental character of the asylum request with the absence of a "condition of exploitation" of the female applicant. 57 Other similar decisions date back to 2015: in those cases, social reasons, with reference to the vulnerability of female applicants, are left explicitly to prevail in the legal argument offered by the judges concerning on detention. Decisions appear to be paradoxical and arbitrary because the judge states the validity of detention to be "in the interest" of the detainee, even if the asylum request is considered well founded: "Detention is considered necessary to facilitate the recognition of international protection, in order to guarantee the presence of the detainee before the Commission and avoid her involvement in the sex and drug trade" (Tribunale di Roma 2015 n.r.g. 14207). 58 The explanatory power of this model, however, is quite low: indeed, except in some isolated cases, social evaluations are not very relevant, at least not directly, in the decisions made by the judges. Many judges argue on the basis of legal reasons or renounce their role of independence in relation to the administrative authority, as the Justice of Peace case study shows.

Modern Constitutionalism: A Debated Model of Justice
Finally, the explanatory power of the constitutional model, i.e. the actual legal framework within which the Italian judges operate, is assessed and considered in its prescriptive and theoretical dimension. This category is much discussed in literature. There is no single definition of it, yet it is possible to identify some elements common to various conceptions of the constitutional model (Bongiovanni 2013: 85-86). First, in the constitutional model, power is expressed and limited by the law; both "rule by law" and "rule of law" are relevant; 59 the constitution usually includes a bill of rights and general principles of law, with a judicial authority set to guard these. 60 An essential theoretical assumption behind twentieth-century constitutionalism is the idea, elaborated by Hans Kelsen, according to which the legal system is characterised by a hierarchical structure. 61 The introduction at the constitutional level of fundamental rights and principles modifies the relationship between the judge and the law: judicial power can no longer be expected to apply law automatically, since it is now required to interpret legislation according to constitutional standards (Bobbio 1992;Ferrajoli 2002: 354-355;Ferrajoli 2010). This is a very widely discussed issue in the legal-philosophical debate, since different ideas of judicial power 62 rely on different conceptions of the constitution. This can be understood as a tool to limit power and the reference framework in the ongoing political conflict, or as a prescriptive declaration of values in society, that provides certain policies (Pozzolo 2001: 338-39).
According to many scholars, the continental nineteenth-century paradigm of jurisdiction cast as a technical and neutral power still affects the role of the judge in the constitutional state (Pozzolo 2001: 91-92). 63 Nonetheless, from a normative point of view, a coherent model of justice in the constitutional state should be firmly independent from the other branches of government, with the aim to protect constitutional rights (Celano 2013: 411-421). This aspect is comparable to the idea associated with the English rule of law of judges in society, with the difference that constitutionalism upholds legal certainty as a value and generally does not involve discretionary powers attributed to the judicial branch in the form of equity, for example. 64 What one ought to expect as results from the case studies, under the assumption that the constitutional model would have a high explanatory power, would be judges' full exercise of their discretionary powers, expressed through welldeveloped legal reasoning as far as the motivation of the decision is considered, with the overall aim to justify the decisions taken. However, the approach of the Justice of the Peace, as illustrated in my empirical study, is not intelligible on the basis of the aforementioned expected results. Because judges do not make full use of their prerogatives, little work is being done as far as the legal reasoning is concerned, and consequently decisions are swiftly made and adjudication is extremely summary.
However, the case law of the Ordinary Tribunal is much more in line with the constitutionalism model: judges use their discretionary powers to safeguard rights and develop relevant jurisprudence for this purpose. Here judicial authority therefore has the role of controlling the activity of the administration. However, the high variability of the outcomes -which is not always the result of an objective evaluation related to the concrete case, but rather seems to be affected by subjective judgements -cannot be easily framed within the framework of the constitutional model. The case study I have conducted on the Ordinary Tribunal of Rome shows that heterogeneity is high in the decisions taken, both in the way hearings are conducted and in the quality of the arguments put forward to justify the decision taken. 65 According to the law, detention of asylum seekers can be authorised in the event that migrants pose a risk of absconding, danger to society, or in the event that the asylum request is merely instrumental to avoid deportation. Notably, the most relevant divergences between judges concern whether or not the asylum request was made instrumentally. 66 Magistrates use different "criteria" of interpretation. 67 Yet the most recurrent criterion is "the prognostic judgement" (giudizio prognostico), which can be defined as a forecast or prevision (explicit or implicit) concerning the outcome of the asylum proceeding yet to be started. In sum, judges are considering the probability that the asylum request will be granted by the competent administrative authority; magistrates frequently give arguments on the credibility of asylum seekers for instance, or they express themselves on the substantial legitimacy of the asylum request instead of expressing themselves, as they would be held to do, merely on whether the person may be lawfully detained or not. Indeed, according to the law in the cases I considered in this study, the judge cannot rule on whether or not the asylum request is in itself justified; the judge is only supposed to rule on the "reasonable grounds" that could make the asylum request instrumental merely as a way to avoid deportation. 68 However, one aspect that ought to be mentioned here are relevant differences between the different phases of conducting the study: in the sample taken during 2015, this "forecast judgement" rarely includes an assessment (of the rather objective criteria) of the socio-political situation in the applicant's country of origin. 69 In most cases, the declarations made by the asylum seekers during the hearing are highly influential: these declarations are frequently used directly in the motivation of the decision taken concerning the groundlessness or instrumentality of the asylum request. The main ground for the decision taken was simply the "story" told by the migrant and his or her credibility in doing so. The adjudication that was supposed to consider only the grounds for detention is thus turned into a decision on the justification of the asylum request, which ultimately rests on the personal story of the migrant. The decision concerning whether or not it is justified to detain is hereby transformed into a kind of "personal judgement". Similar considerations can be made regarding the decisions issued in 2018, even if in this latter case, the nationality of the asylum seekers is a more central argument. In the most recent case law of the Tribunal, in fact, releasing Nigerian women is a guideline. Yet, it should be said that some judicial provisions are simply inconsistent 70 and the judgement on the instrumentality of the asylum request produces opposite outcomes, even when the person's country of origin is the same. For example, in the two cases mentioned above that involved two asylum seekers of Nigerian nationality, 71 the judge did not consider the nationality of the asylum seekers and decided on the basis of social elements outside the legal system (e.g. the condition of "exploitation of women").

The Legal Realist Perspective and the Exercise of Discretionary Powers
To sum things up, none of the models reviewed above can fully explain the results of the two case studies. This fact alone is sufficient to cast doubts on the supposed explanatory power of these theories. Yet, what remains to be tested is whether the legal realist perspective offers a better chance of coherently explaining the collected data.
Realists generally hold that a judge's decision-making rests on elements that do not exclusively amount to legal reasons regulating the case at hand. Let me start by fleshing out two versions of this claim. One, perhaps naïve, version of this claim is associated with the "Frankified" version of American legal realism, as Brian Leiter famously called it in Naturalizing Jurisprudence. Frankification holds that the factors impacting judges' decision-making are essentially personal or ideological in nature. This view is closely associated with what has been called the "digestive theory of justice", 72 according to which justice varies depending on what the judge had for breakfast. Another version of the aforementioned basic claim of realism concerning adjudication is that of Alf Ross. As once observed by Alf Ross, a prominent figure of the Scandinavian realist tradition, 73 the "normative ideology" of judges is a key factor guiding magistrates in the decision-making process. According to Ross, the mental process by which the judge decides to base his decision on one rule rather than another is not a capricious and arbitrary matter, varying from one judge to another, but a process determined by attitudes and concepts, a common normative ideology, present and active in the minds of judges when they act in their capacity as judges (Ross 1959: 75). The normative ideology can be detected only by analysing the magistrates' actual behaviour and consists of directives concerning the way in which judges proceed in order to select the "decisive directives" regarding a specific question (Ross 1959: 76). Ross combines two levels in his analysis: on one side are the personal attitudes and ideological convictions of judges that he calls the "material or moral" legal consciousness, and on the other is the "formal or institutional" legal consciousness. This latter form of consciousness consists of the particular practices enacted by the judicial operators, and "it is directed toward the institutions in recognition of their 'validity' as such, irrespective of whether the demands in which they manifest themselves can be approved as morally right or just" (Ross 1959: 54). 74 Ross' view of how both juridical and extra-juridical factors may play a role in the decision-making process offers a more articulated and complex approach to the influences on the judge when compared to the more radical versions common in American legal realism. The crucial factors explaining the decision-making of the judge in the more radical versions are reduced to extra-juridical elements, often of a merely subjective character. 75 Therefore, I find Ross' framework more articulate and perhaps more useful for exploring whether it is possible to explain the approach of the judge in my two case studies. My hypothesis is that Ross' idea has a greater explanatory power than the legal realist view attributable to Jerome Frank, since both personal and/or ideological factors -as well as juridical factors, depending on how the legal order works -seem to affect the magistrate's decision.
However, another point needs to be made before presenting the compatibility of the data with the explanatory frameworks inspired by legal realism. It is clear that the exercise of discretionary powers by the judge may lead to different outcomes, producing judgements that can either be evidence-based and respectful of a complex social reality, or simply hasty and summary. What remains unclear is the nature of the elements that affect the judge's decision-making. This point is important in order to assess the explanatory power of the available theories. Is it the case, for instance, that the elements influencing the judges' decision-making are essentially of an extra-juridical nature, as most radical realist theories would claim? Or is it the case that the discretionary choices made by the judge are determined by intra-juridical considerations, such as the institutional role of the judiciary in the legal system? Were we to find that no intra-systemic, law-related or juridical considerations matter, we would not be in a position to exclude the possibility that the "digestive theory of justice" -according to which justice varies with what the judge had for breakfast -might best explain what is going on in the case studies. This would be worrisome for several reasons, yet it would enable us to distinguish the theory with strongest explanatory power in relation to the data I have presented. However, were we to find that intra-juridical considerations matter for the way in which judges choose to exercise their discretionary powers, we could show that the explanatory power of at least some of the most radical versions of legal realism would fall short of explaining the data that emerged in the case studies.
The relevance of personal beliefs and elements depending on social evaluations of an extra-juridical nature is clear in the case law of the Ordinary Tribunal of Rome. A pattern is discernible in the way the magistrates conduct the hearings and in the decisions they make. Indeed, we can often explain the decisions taken by judges on the basis of the following explanation: Judge x usually does not carry out a thorough inquiry during the interview with the asylum seeker, provides shallow legal arguments, and most of the time confirms the detention of the applicant if he or she comes from a given country, or alternatively will not validate the detention order when the applicants come from other countries. By contrast, Judge y always investigates in depth during the hearing and is more inclined to reject the detention request. Some decisions show this pattern clearly: for instance, in the hearings held on 24 January 2018, the same judge decided on four proceedings, three of which involved Moroccan women; in all three of these cases the detention was validated. In these three proceedings (Tribunale di Roma 2018 n.r.g. 1191; 1192; 1193) 76 the hearing minutes were almost nonexistent; the declarations made by the asylum seekers, according to which they had fallen victim to harassment and violence in Libya, were not questioned any further. 77 In the fourth proceeding (Tribunale di Roma 2018 n.r.g. 1191) regarding a Nigerian woman, a similar conduct of the hearing leads to a quite different outcome (i.e. invalidation of the detention) with reference to the vulnerability of this young woman who could become a victim of trafficking. The possibility that we might be dealing with personal preferences or other extrajuridical factors in the determination of the outcome is strengthened by the fact that another judge has quite a different approach to Moroccan women travelling through Libya: in these cases (Tribunale di Roma 2018 n.r.g. 1613; 1614), 78 the story told by the applicant is given ample coverage in the hearing minutes; the abuses suffered in Libya are considered indicative of vulnerability and the decision is made to release the detainee.
Finally, analysing the legal reasoning of the judges sitting on the Ordinary Tribunal, a more complex landscape emerges, which cannot be made intelligible within the framework of classic or "Frankified" legal realism. It seems that explaining the variation in outcome by personal ideologies is unwarranted or does not offer a complete explanation of the data we have. In fact, in some cases the judges mention the case law of the Tribunal regarding the issue at hand or provide an explanation for their choice of interpretive criteria. 79 This kind of argument can be better understood in reference to the theory of Alf Ross than to the theory of Jerome Frank. We may argue that this kind of argument introduces the "formal or institutional" legal consciousness that Alf Ross spoke of as the judges' normative ideology, which is to be categorized as a legal element influencing the judicial decisions, not merely as an extra-juridical element of personal predilection. Specifically, in the abovementioned cases, this normative ideology, as Alf Ross calls it, can be identified first in reference to the consolidated guidelines of the competent Tribunal's division on that issue and in the interpretation that the judges make of the requirements of the law regulating detention for asylum seekers. As we have previously pointed out, the common or shared opinion of jurists is an important parameter in judges exercising their discretionary powers when choosing between alternative legal options, which echoes the point of view expressed for instance by Aharon Barak. 80 This criterion of shared opinion of lawyers is not a very accurate criterion and it is necessarily uncertain, since a split may occur in the community of judges over a specific matter of law. Nonetheless, shared opinion may enable the judge to distinguish his personal thoughts on the matter from his ideas on what the law requires.
I have found that the importance of the intra-systemic legal elements in the approach of judges is even more evident in the jurisprudence of the Justice of the Peace than in that adopted by the Ordinary Tribunal. In the case law of the lower instance, the intra-systemic legal elements primarily consist of the institutional framework within which judicial authority is exercised. As already mentioned, the way the hearings are conducted and the significant difference in outcome found in the adjudication of cases show that the policies on detaining foreigners, as these emerge in the two competent judicial authorities, differ quite a lot. This gap in policy cannot easily be explained away merely as a consequence of somewhat different legal norms apply to the detention of irregular migrants and asylum seekers. Since the differences in the laws applying to these two categories of migrants cannot provide the explanation we are looking for, we need to look further. I suggest that a better explanation for the divergence in policy that I have been able to observe in the two authorities -the object of the two case studies -has to do with the institutional role attributed to the adjudicative organ in the two different settings. Note that this amounts to an intra-systemic, i.e. entirely lawrelated, attribute that thus shares little or nothing with the extra-systemic factors that the Frankified view of legal realism considered responsible for explaining the difference in outcomes. Now, the institutional shaping of the two authorities within the legal system -the Ordinary Tribunal on the one hand and the Justice of the Peace on the other -has to be considered an entirely legal element, that has nothing to do with the individual preferences that the judges may have. In particular, regarding the Justice of the Peace, the institutional role of the authority in the legal system needs to be considered a crucial element to account for. The Justice of the Peace is an honorary judge, introduced into the legal system in 1991 with law 374/1991. 81 The Justice of the Peace appears to be a somewhat ambivalent and controversial figure (Scamuzzi 1896;Picardi 1984). This is so especially because of two factors: the temporary affiliation of the Justice of the Peace to the judiciary and the piecework remuneration. The fact that the Justice of the Peace is not technically speaking a professional judge and is paid for the output in proportion to the number of cases examined, has led some scholars to define the institution of the Justice of the Peace in terms of a "precarious" judge 82 and calls into question the requirement of independence of judicial power in relation to this authority.
The relationship between branches of government appears to be a crucial factor affecting the adjudications: the two competent judicial authorities -the Justice of the Peace and the Ordinary Tribunal of Rome -diverge in their jurisprudence on detaining migrants. These two authorities are also differently framed within the legal system. The Justice of the Peace is merely an honorary judge and, as I have pointed out above, there are some controversial elements in its institutional configuration. These elements relate to its independence from executive power -a necessary element for the judge to play the role of guarantor in the proceedingsand to how effective this independence may be said to be. On the other side, the Ordinary Tribunal of Rome -usually composed of professional judges who are generally more competent in civil and criminal matters -has a different attitude towards the administrative authority; the Tribunal, but not the Justice of the Peace, is able to elaborate case law autonomously, and its overall jurisprudence is aimed at protecting individual rights.
The exercise of discretionary powers is a crucial element that distinguishes the case law of the Justice of the Peace from the jurisprudence of the Ordinary Tribunal of Rome: while the first authority does not use its prerogative in the hearings, judges on the Tribunal use the discretionary powers invested in them to play a central role when conducting the hearings. There is generally a very low level of discretion in the adjudications by the Justice of the Peace, in the sense that their discretionary powers are not put to use, and we can generally say that the exercise of these discretionary powers does not exceed their place and become arbitrary. The decisions taken by the judges on the Ordinary Tribunal, by contrast, are the result of fully exercising their discretionary powers. In most cases, the use of these discretionary powers is meant to guarantee individual rights, but sometimes their use is not properly justified or may seem to lead to arbitrary decisions.
To sum up what we have found, the results of the two case studies cannot be fully explained by the available models in legal realism, either. Realists generally hold that there are elements that explain the decision-making of the judge which do not exclusively amount to reasons provided by the law regulating the case at hand. I found that one version of this claim most commonly associated with the "Frankified" version of American legal realism -the belief that the factors that influence judges' decision-making are essentially personal predilections -did partially succeed in explaining variations in outcome among judges' decisions. Indeed, it would be possible to track some positions on key factors (e.g. country of origin combined with genre) that individual judges hold in order to foresee the outcome of the case, which is in line with what the mainstream version of American legal realism held in its infamous "predictive theory". However, this version of realism had trouble explaining variation across institutions. Why does the case law of the Justice of the Peace differ so significantly from that of the Tribunal? One possible explanation for this divergence was presented and discarded: the fact that the authorities are dealing with somewhat different legal norms, since one concerns the detention of irregular migrants and the other concerns the detention of asylum seekers. This strictly formal explanation contained some logic but not enough to explain the scale of the divergence in the case of the two authorities. Once this formal explanation had been discarded, I could assess the virtues of the other version of the aforementioned basic claim of realism concerning adjudication, associated with Alf Ross and the Scandinavian realist tradition that ascribes weight to "normative ideology" of judges in explaining their decision-making. I found that this version of legal realism was a much better fit to explain some data, in particular concerning the divergence of the case law between the two authorities. In particular, the normative ideology thesis advanced by Alf Ross helps to explain the fact that the judge on the Tribunal made a much broader use of discretionary powers than the Justice of Peace. This fact could in turn be accounted for by the different institutional features (ways of working and being paid, professionalization, etc.) and ultimately by the role each authority plays with regard to the separation of powers within the legal system. This explanation, rooted in Ross' notion of normative ideology, was found to offer the best framework for accounting for the divergence between the case law of the two institutions, although other models have been shown to have different virtues relating to other features of the data.

Concluding Remarks
The impetus for this article came from observing the lack of scholarly attention to the arbitrariness of judicial decisions involving migrants. Whereas many researchers have highlighted the discretionary and arbitrary practices surrounding the "bureaucracy domination" in procedures concerning access to territory by nonnationals, and many have studied what the police do and how the government acts, few have looked at the role of the judiciary in this "bureaucracy domination". This study thus fills a void in the literature on migrant issues. In a qualitative empirical research study, I have studied the Italian case law concerning expulsion and detention proceedings of irregular third country national citizens and asylum seekers. I have looked at decisions issued by the competent authorities and how they have been carried out. The results were analysed using a selection of theoretical tools, all referable to the general concept of the rule of law in one way or another.
The assumption underlying the research is that various conceptions of the rule of law may have different explanatory power when it comes to explaining the empirical results. To verify this hypothesis, the study proposes an overview of the main rule of law doctrines in the Western tradition of political and legal thought and applies the method of historical-conceptual analysis. As a result, the explanatory power of six theoretical models of the rule of law was verified against the data with the view to highlight the virtues and vices of the respective explanatory frameworks. The explanatory frameworks or models of the rule of law that I have worked with are the following: (1) The Pure Arbitrary Justice Paradigm; (2) Rechtsstaat; (3) English Rule of Law tradition; (4) Mixed Constitution Model; (5) Modern Constitutionalism; (6) Legal Realism. The last is divided into two versions: (6') one centred on personal predilections in explaining the decision-making of the judge, associated with Jerome Frank and the movement of American legal realism; and (6'') the 'normative ideology' thesis advanced by Alf Ross and favoured by members of the movement of Scandinavian legal realism, more centred on institutional features of the legal system.
These frameworks consist of four prescriptive theories about how judges ought to decide, a descriptive legal theory about how judges do decide, and the hypothesis of complete absence of the rule of law, in which asking whether or not judges act as they should or even if their decisions follow a pattern would not make sense. These explanatory frameworks do not offer an exhaustive list but still cover a wide range of theories. The fact that some of these explanatory frameworks are normative in kind need not worry us since we are not asking whether the model of the rule of law is good or not -an inquiry that would require a normative methodological outlook that is not employed here -but merely whether or not the judge lives up to the requirements of a given prescriptive model of the rule of law and -if he does not -how far off the mark the decision-making of the judge really is.
The conclusion this article reaches is two-fold: first, concerning the explanatory frameworks and second, concerning the notions of discretion and arbitrariness. Let me start with the first. The results of the two case studies cannot be fully explained by any of the models that I have reviewed above. This fact alone casts doubts on the explanatory power that these theories may be said to have and calls for further research on judicial decision-making more generally.
In the Pure Arbitrary Justice Paradigm and the Mixed Constitution Model, the explanatory power was so low as to make it possible to simply discard these models as unfit to explain what is going on. As far as (1) the Pure Arbitrary Justice Paradigm is concerned, we would expect to find a great variety of unpredictable judicial decisions, but this clashes with great uniformity of the case law made by the Justice of the Peace. The Ordinary Tribunal of Rome applies a good level of legal reasoning, an element that we would not expect to find in an arbitrary and irrational form of justice. Therefore, we must discard (1) as having particularly low explanatory power. The fourth model of the rule of law -the Mixed Constitution Model -had quite a low explanatory power. Except in some isolated cases, social evaluations are not deemed directly relevant in the decisions made by judges.
However, in other cases, the explanatory power was greater and succeeded somewhat in explaining what was going on, yet not to a sufficient degree or with a sufficient scope to fully cover the data. Thus, we may conclude that they provide partial explicative force in terms of the empirical data of these two case studies. As far as (2) the Rechtsstaat is concerned, the explanatory power appears strong when related to the data of one case study (that of the Justice of the Peace), but poor when related to the other case study focusing on the Ordinary Tribunal. When (2) the Rechtsstaat is tested against the backdrop of the approach adopted by the Justice of the Peace, the explanatory power of this theory appears to be quite high. This bureaucratic model of justice, thoroughly applied, would yield little or no legal reasoning, uniformity in the case law or strong limits to the discretionary power of judges. This is quite similar to the results that emerged from the analysis of the detention proceedings of the Justice of the Peace. However, when (2) the Rechtsstaat is tested against the backdrop of the approach adopted by the Ordinary Tribunal, the explanatory power drops significantly. Here it clashes with the results. Therefore, we can say that (2) will partially explain the results, but it is not yet satisfactory as a framework for explaining what is going on. Yet, what I found is in itself of interest: it is often claimed that the Rechtsstaat model is outdated. Yet it looks rather alive when compared to the data from the case law of the Justice of the Peace. In sum, model (2) only offers a partial explanation of the data. As far as the third model of the rule of law is concerned, it can be said to have a similar level of explanatory power as the Rechtsstaat, albeit in reverse order. In fact, the (3) English rule of law tradition does very little to explain what goes on in the case law of the Justice of the Peace, but it is partially useful for explaining what goes on in the case of the Ordinary Tribunal in Rome. According to this third model, justice should be characterised by consistent legal reasoning, a contextual variability in adjudication, and judges should play the role of controlling the activity of the administration, with the aim to protect individual rights. This model does not capture the case law of the Justices of the Peace, who provide rather shallow legal arguments and do not play an independent role in relation to the instances of the law enforcement agencies, often to the detriment of migrant rights. Instead, this model seems more appropriate to explain the jurisprudence of the Ordinary Tribunal. The general landscape emerging from the analysis of the hearing minutes is quite a detailed and rather complex process. The precise development of legal arguments also illustrates this. All in all, we may say that the English rule of law tradition still carries a partial ability to explain the data. As far as (5) Modern Constitutionalism is concerned, the explanatory power may only be said to be partial. The results one ought to expect from the case studies, under the assumption that the constitutional model would have a high explanatory power, would be that judges would fully exercise their judicial discretion, expressed through well-developed legal reasoning regarding the motivation of the decision, with the overall aim to justify the decisions taken. However, the approach of the Justice of the Peace is not intelligible in this way: judges do not make full use of their prerogatives, little work on legal reasoning is being done and consequently decisions are swiftly made and adjudication is extremely summary. However, more coherent with the constitutionalism model is the case law of the Ordinary Tribunal: judges use their discretionary powers to safeguard rights, and the judicial authority plays the role of controlling administrative activity.
Finally, the results of the two case studies cannot be fully explained by the available models in legal realism either, although realists generally hold that there are elements that explain judicial decision-making that do not exclusively amount to reasons provided by the law regulating the case at hand. I found that one version of this claim most commonly associated with the "Frankified" version of American legal realism -the belief that the factors that influence judges' decision-making are essentially personal predilections -did partially succeed in explaining variations in outcome among judges' decisions. Positions on key factors (e.g. country of origin combined with genre) made individual judges recognizable: indeed, something of a pattern of personal choices could be detected. More to the point, however, I found that the "normative ideology" thesis advanced by Alf Ross succeeds in explaining the decision-making of judges. I found that this version of legal realism was much better suited to explain the divergence of the case law between the two authorities. This divergence could be related to different institutional features (ways of working and being paid, professionalization, etc.) and ultimately to the role each authority plays with regard to the separation of powers within the legal system. This finding is also in itself of interest since it opens the possibility of an explanatory framework for legal decision-making that develops from the realism understanding that there is life outside the law, yet that gives particular weight to procedural legal aspects.
Let me now move on to findings of the study as far as discretion and arbitrariness are concerned.
I started my investigation of the concept of arbitrariness by introducing a definition linked to a legal perspective. I understood arbitrariness as the opposite of discretion: an arbitrary act is one that falls beyond the scope of discretion, where discretion involves the possibility of choosing between different equally legitimate solutions. In this sense, an arbitrary decision can be defined as illegal, to the extent it does not comply with the law's procedural and substantial requirements. Nevertheless, if we limit our definition of arbitrariness to 'an illegal act', the overall result could be quite paradoxical: the judicial approach which assures the highest protection of rights -adopted by the Ordinary Tribunal -is also the one that is most easily affected by arbitrariness. While the Justice of Peace's jurisprudence is an extremely shallow and summary process, most of the time it is lawful and hence not arbitrary. This paradox can be easily resolved by paying attention to the multiple meanings that can be attributed to the notion of arbitrariness. If we consider the notion of arbitrariness from a purely legal point of view, i.e. as an illegal decision, it is unsurprising that the authority that most uses its discretionary powers is also the one most at risk of abusing these powers, and hence of exercising arbitrary power. However, if we consider the notion of arbitrariness from the point of view of political theory, or even of so-called critical legal thinking, the result could be different.
The discussion on the concept of arbitrariness within the political-philosophical debate is quite extensive and beyond the scope of this article. Nonetheless, I will clarify my use of the term 'arbitrariness' for the purpose of this study and look forward to exploring the subject in further research. I decided to make reference to one conception of arbitrariness within the political-philosophical theory related to the republican debate on domination. 83 In this sense arbitrariness is understood as the illegitimate rule of an individual by another or by an authority, and an arbitrary power is one that is not limited by any obligations or duty to justify its action (Mindus 2020: 3). The idea of 'justification' in this context is a more comprehensive concept compared to the definition of 'legal reasoning' in the judicial lexicon; however, these two aspects are strictly connected when it comes to studying adjudication. Indeed, legal reasoning is essential to the judiciary function, and providing a satisfactory legal argument is a means of monitoring the judicial authority by society.
The legal reasoning should provide good reasons for the choice of the judicial authority: the discussion of what should be counted as a 'good reason' in order to justify adjudication is still open, but an essential list might include consistency and rationality of decisions (Taruffo 2002: 209-216, 220). However, the justification of decisions raises the question of a problem of legitimacy, beyond the mere lawfulness argument. Occasionally an 'unjustified' decision can be illegal, but that is not always the case, as the jurisprudence of the Justice of the Peace seems to attest.
Generally, the decisions of the Justice of the Peace are characterised by extremely shallow legal arguments and are the result of a summary process. However, if we try to qualify them as illegal, only a few of them would be considered unlawful decisions, although the empirical results show several critical aspects to the detriment of the individual's rights. I would relate those problems to a lack of justification of judges' decisions.
Indeed, if we include in the analysis the definition of arbitrariness as a power that does not provide any kind of justification for its actions, it is unsurprising that the judicial approach which assures the highest protection of rights is also the one that takes its own role as guarantor of these rights and of the constitutional democratic legal order as such most seriously. This judicial approach hence most often risks exercising its power in criticisable ways, as compared to an authority much more in line with the requirements of the law enforcement agencies. Were we to adopt the view according to which an arbitrary power can be defined as a form of domination, characterised by the absence of sufficient justifications in its exercise, a good example of such an arbitrary power would be represented by the sovereign prerogative of the state to exclude aliens from access to territory without giving reasons (Mindus 2016). In this sense, the policy of the Justice of the Peace, where in the vast majority of cases the judge authorises the detention with the view to deport the applicant, is strongly aligned with such an arbitrary power to exclude, exercised by the executive branch of government and ultimately by the law enforcement agencies (Campesi 2014). The state's reasons to exclude aliens from its territory and to limit migration have been the subject of wide debate, and in some cases were considered arbitrary (Finnis 1992;Cole 2000;Carens 2000;Bosniak 2008;Shachar 2009). Similarly, the "reasons to detain" are actually unclear and not well justified by the sovereign authority (Rigo 2017: 482-483). It is claimed that migrants are detained in order to be deported or returned to their home country, yet in most European countries this goal is never achieved: numerous studies have shown that the number of deportations is very low compared to the number of detainees in detention centres (Beilfuss 2017;Könönen 2017). Since the goal cannot in practice be deportation, scholars have started to stress the disciplinary nature of detaining migrants, or have in general claimed that the practice of detaining migrants serves the purpose of managing and controlling the phenomenon of migration (Guild 2009;Campesi 2013: 66). The purpose of detaining migrants is therefore controversial: it is unclear what purpose the practice serves and whether the practice serves the purpose well or not. Given the controversy surrounding the practice of detaining migrants, it is crucial for the judiciary to exercise its role of controlling the lawfulness of the practice well. The judiciary may act as a limit to the power of the state and fill in the lack of justification that the state offers for its practice, as the Ordinary Tribunal of Rome case study generally shows. Otherwise, the judicial authority may add lack of (legal) reason(ing) to this (already unreasonable) form of restriction on personal liberty, as the Justice of the Peace case study seems to prove, in a perverse form of exchanging legitimacy between the executive and judicial branches of government.
2 The selected literature of this study refers mostly to non-anglophone scholars because the case study took place in Italy and the paper addresses different juridical aspects of the Italian legal system linked to the topic, therefore a consistent part of the relevant literature was Italian. 3 The competent authorities for this kind of proceeding are the Justice of the Peace for irregular migrants and the Ordinary Tribunal for asylum seekers. 4 The Justice of the Peace is an honorary judge, introduced into the legal system in 1991 with law 374/1991, and has jurisdiction over minor civil and criminal matters. 5 This study takes its springboard from the results of a wide research project at national level: The Observatory on the judicial review on migrants' rights (http://lexilium.it), with a unit located in Rome. The project, which involved different University departments, started in 2013 under the direction of the Legal Clinic of the University Roma Tre. It focused on migration and asylum, and collected and analysed a very large sample of judicial decisions on the expulsion and detention proceedings of migrants in Italy. In the last phase of the research between 2015 and 2016, 1220 judicial decisions had been collected. The monitoring centre also collected case law from the Justice of the Peace of Bari, Bologna, Prato, Rome and Turin, including judicial reviews of removal orders, immediate removal orders, pre-removal detention orders and alternative measures to detention. In order to consider the "jurisprudence" of the Justice of the Peace, I used the global results of the Observatory, and specifically the executive reports of each research unit. I also accessed the data collected by the research unit located in Rome. 6 Data for this case study was collected independently in a research study limited to the detention of asylum seekers in the deportation centre of Ponte Galeria, situated close to Rome. This part of the study was organised in two phases, occurring in 2015 and in 2018 respectively, and collected a total of 188 decisions by judges. 7 §5bis of Art. 13 of the D.Lgs. 286/1998 -the immigration consolidated lawestablishes the following: the public authority of security can order the immediate deportation to the border of the foreigner, which in most cases is impossible; then it can impose a coercive measure such as administrative detention or alternative measures to detention on the basis of a discretionary evaluation made by the law enforcement agency, the Questura. 8 CPR stands for Centri per il Rimpatrio, literally centres for repatriation. 9 D.L. stands for Decreto Legge, literally law decree. According to Art. § 77 of the Italian Constitution, a legal decree is an act with legal force, adopted by the Government, if emergency and necessity conditions occur. The Decreto Legge is framed as an exceptional law-making instrument, and has to be converted into law by the Parliament within 60 days to not expire. In recent years, the Italian governments made extensive use of this instrument, and adopted legal decrees also when Constitution requirements were absent. This is especially the case in migration matters. 10 As an example, the time limit of detention is 180 days for irregular migrants, and 1 year, globally, for asylum seekers in the meantime of the asylum adjudication. Furthermore, a fundamental legal requirement to detain third country nationals is the "risk of absconding", while in the case of the asylum seekers, the more common argument for the decision to detain is the exploitative nature of the asylum request. 11 This civil procedure is called "rito camerale" literally "chamber procedure" (Carratta 2009;2012). 12 Art. 13 §2 of the Italian Constitution establishes the following: "In exceptional circumstances and under such conditions of necessity and urgency as shall conclusively be defined by the law, the police may take provisional measures that shall be referred within 48 hours to the Judiciary for validation and which, in default of such validation in the following 48 hours, shall be revoked and considered null and void". A fundamental decision in the issue is the Constitutional Court judgement no. 105/2001, on the basis of which the Court stated subsequent legitimacy decisions on the matter. 13 For a reconstruction of the genesis of administrative detention see Campesi (2013). 14 On the current crisis of the traditional conception of Etat de droit in France with regard to the impact of EU law, see Heuschling (2010). 15 This topic has been the object of many theoretical debates, from the "neoconstitutionalist" approach to the debate on access to justice. In the latter context, there are different positions: some scholars consider judicial power to be excessively present in social life, while others see emancipatory potential in implementing the competences of judges. See Cappelletti (1984); Garapon (1996); Sassen (2000). 16 More precisely, the theoretical models that have been identified are four prescriptive legal theories, a descriptive one with reference to legal realism, and the hypothesis of complete absence of the rule of law. This is not an exhaustive list, since it would be possible to find other useful models connected to the idea of the rule of law. The models are intended to represent ideas that would be more or less "true" in themselves but only more or less capable of explicating the empirical results. 17 Starting with the Ancient Greek thought of Plato and Aristotle. Contemporary legal thinking has sometimes overlooked analysing the notion of arbitrary power. Cuono (2013); Mindus and Cuono (2018). 18 More specifically the question is whether the rule of "good" law or the rule of "good" men is preferable. See Tamanaha (2004: 105-115). 19 See also Calamandrei (1954: 49-50). 20 The analysis is proposed as a reply to the critics of Bentham. 21 The exercise of discretionary powers by the judge in the trial is a fundamental feature of the judicial function. Calamandrei (1954: 64); Ferrajoli (1996: 556-557); Villa (2012). 22 This conception is quite different from that of Ronald Dworkin, for example, which instead affirms the existence of one right solution to every judicial case, see Dworkin (1977: 81-130). 23 In this context, the term "arbitrariness" is used from a legal and not philosophical point of view. 24 In the same direction note the contribution of Michele Taruffo, with his definition of judgement as a "decision-making method", see Cardozo (1921); Taruffo (2002: 157-179). 25 Especially in front of the Justice of the Peace, the legal reasoning is completely absent in a certain number of decisions . 26 For a critical view on migrant detention in the Agamben thesis, see Rajaram and Grundy-Warr (2004). 27 The term "Khadi" comes from the Arab and means "the one who decides" (a question, a judicial dispute, etc.), as a technical description of the magistrate who, in the name of the sovereign, administers justice according to the sharī‛ah, or Muslim law. For Islamic public law, the administration of justice is a sovereign prerogative, therefore the sovereign is the supreme judge and can replace the Khadi anytime or limit his competence. See the entry "Khadi" ad vocem in Enciclopedia Treccani. Retrieved 31 August 2019 from http://www.treccani.it/ enciclopedia/cadi_%28Enciclopedia-Italiana%29/ 28 The notion of Rechtsstaat was first introduced by the German jurist Robert von Mohl around 1830, but was implemented in Germany only in the second part of the nineteenth century, in the restoration after the Revolutions of 1848. In the doctrine of the Rechtsstaat, evolution can be seen in two core issues: the passage from natural law tradition to legal positivism, and the opposition between conservative and liberal perspectives, expressed respectively by Robert von Mohl and Friedrich Julius Stahl. See Mohl (1872); Gozzi (2002); Ridolfi (2017). 29 In the French legal tradition, the limitation of power was the subject of much reasoning, starting from the times of the ancien régime. However, in France Jean Bodin first introduced the influential notion of sovereignty: indeed, even if the Revolution had produced a rupture with the absolutist model, the traditional notion of sovereignty as a unitary power was still very influential. In particular, even after the restoration, the idea of a single subject entitled to power stood. The sovereign was identified in the representative bodies and in the law, and was considered an expression of the "volonté générale". Only in the French Third Republic, in the 19 th and 20 th Centuries, was the doctrine of the Etat de droit actually elaborated in France, taken from the German model. See Rousseau (1839); Bodin (1986); Laquièze (2002). 30 A tradition that dates back to the thought of Thomas Hobbes and Jeremy Bentham, Bentham in particular identifies the fundamental values of the legal system in the state security and in the unconditional predictability of adjudications. See Bentham (1996); Hobbes (2006). 31 A theory first developed by Rudolf von Jhering and thereafter completed by Georg Jellinek with his notion of "subjective public rights". Jellinek (1892); Jhering (1997). 32 This doctrine, first elaborated in the context of the "French exegetic school" in the years of the codification by Napoléon, enhances the role of the law and that of the state authority in the legal system, stating the necessity to apply law as close to its literal meaning as possible. See Bobbio (1996); Santoro (2008). 33 The space consists of some fixed items regarding whether the migrant is in possession of his or her documents and pertinent to the migrant's residential condition and income. Often, the defence activity is also inadequate: in various cases the lawyer only filed a formal opposition to the validation of detention, without adding any objections in the interest of his or her client. 34 This is the name of the law enforcement agency responsible for the urban area. 35 As mentioned before, legal reasoning is totally absent in a relevant number of cases. These amount to 12% of the cases in the offices of the Justice of the Peace in Rome, 30% of the cases in Bari, while in Turin 50% of the decisions consist in pure ritual formula. 36 The number of validations in the other judicial office sites of a detention centre are: 76% in Rome and 86% in Bari. 37 This praxis is allowed by law (Cf. D.Lgs. 286/1998, Art. 13 comma 5ter). The norm provides the possibility to hold the hearings in the locations of the urban law enforcement agency questura. 38 The reference is to the medieval constitutional distinction between gubernaculum (government) and iurisdicitio (jurisdiction, norms, law), where the latter is intended as a negative limit set to the crown's discretionary power. This conception was developed in the historical context of 14 th Century England. McIlwain (1940); see Palombella (2009: 36). 39 This framework is generally compared to that of a constitutional state, but indeed, the English legal tradition constitutes a variation on this model, because at least until recently it was not created from a written constitution, yet it states a series of intangible rights as part of the "law of the land", beginning from the 1215 Magna Carta. See Dicey (1889: 273); Santoro (in Santoro Zolo and Costa 2002: 174); Ferrajoli (2002: 349). 40 The formula 'Rule of law' is attributed to William Edward Hearn (1867) but was later elaborated by Albert Venn Dicey (1889). 41 The idea of the uniqueness of sovereign power is not widespread in the English context, which instead affirms the model of the "King in Parliament": a polycentric distribution of power. Furthermore, English law evolves for the first time as a judicial claim -or remedy -through the so called "writs". This way the link between the statement of rights and their remedies in front of the Court are still very strong. See D'Avack (2000); Santoro (in Santoro Zolo and Costa 2002: 174). 42 The common law and the legal English customary tradition are handed down and preserved by ordinary judges; even if courts do not formally have the competence to control the acts of Parliament, the relevance of the previous decisions entails the necessity to apply the law according to a consolidated jurisprudence, which has therefore supported the autonomy of judicial authority. Santoro (2008: 110). 43 The expression was introduced by Edward Coke, see Fassò (1999). 44 The cases with particularly poor legal reasoning are few, especially in the second sample of the data referred to 2018. 45 In the majority of cases the decision is issued a few hours after the end of the hearing and is recorded in a separate document, different from the praxis of including the motivation in the hearing minutes, as the Justice of the Peace usually does. 46 Specifically, the percentage of validations observed in 2015 was around 50%, and in the 2018 sample were more validations with 50 rejections out of 88 analysed proceedings. 47 The Ordinary Tribunal of Rome has a settled case law of granting subsidiary protection to female asylum seekers from Nigeria, because of the severe social context of the country, especially for women who are often victims of violence and abuses. This jurisprudence is directly mentioned in some decisions. See: Tribunale di Roma 2018 n.r.g. 9190. 48 In fact, from December 2015, the male area of the Ponte Galeria detention centre closed following a riot initiated by some migrants. For that reason, at the moment of the last research survey (June 2018), detention proceedings involved only female asylum seekers, and in 2018 around half of the detainees were Nigerian (37 of 88 total). 49 Tribunale di Roma 2018 n.r.g. 2572. According to this parameter the judge considers, in this case as in others, the rootedness in the territory to be more relevant than the risk of absconding even if Police records exist. 50 The ideal of the mixed constitution derives from the Roman tradition as instrumental to maintaining the political order, thanks to a compromise between the interests of different social factions. This concept is closely linked to the government of law issue, because the distribution of powers within social parties is functional for its limitation, therefore as Pietro Costa observes, it represents the "prehistory" of the rule of law. Costa (in Santoro Zolo and Costa 2002: 104); see Portinaro (2007). 51 This model seems to constitute a settlement between the "rulers" and the "ruled": as Norberto Bobbio states, it assures stability both from the ruler's point of view and from that of the population, since it offers more guarantees of freedom, thanks to the reciprocal control of powers. This last element makes the mixed constitution tradition similar to constitutionalism. Passerin D' Entreves (1967: 155-157); see Bobbio (1983). 52 This refers to Aristotle, for whom the mixed constitution represents the merging of two corrupted forms of government: oligarchy and democracy. In the tradition of ancient Rome this idea was first elaborated by Polibio, who intended the equilibrium between the social classes as concrete powers in the legal system; and by Cicero who defines the aequabilitas, as equal distribution of rights and powers, with precise legal lexicon. See Passerin D'Entreves (1967: 167-168); Portinaro (2007: 154-155). 53 In the medieval context the mixed government theory was also enriched by spiritual elements, of which one of the best known examples is Thomas Aquinas's conception of regimen commixtum, which provides the temperament of the monarchy through the participation of the people in legislating. Passerin D'Entreves (1967: 170); Portinaro (2007: 154-155). 54 Already in the 1465 John Fortescue affirmed the superiority of the regimen politicum et regale compared to the pure regnum regale, but it is in 1642 with the release of the Carlo I "Answer" to the Nineteen propositions raised by the Parliament, that the English system assumes a stable configuration as a mixed government. D' Avack (2000: 236). 55 In the sense of a rigid separation between the functions of the state. 56 This is, for example, the perspective of the "Access to Justice movement" that developed in Italy during the 1970-1980s, with the aim to investigate the duties of the modern welfare state in the light of the difference between classical liberalism on the one hand, marked by the ideal of negative liberty, and contemporary democracy on the other hand, where the principle of positive liberty could find full expression instead. Cappelletti (1979: 54) 57 In this case, the "exploitation" element is considered relevant for evaluating the asylum seeker's credibility, even if there isn't any reference to this sense in the law.
Here an excerpt of the judge's decision: "[...] the applicant arrived in Italy more than one year ago which has not resulted in being in an exploitative situation, considering she has had access to a free service in a Camp in Milan since her arrival ". Tribunale di Roma 2018 n.r.g. 8738. 58 This refers to two decisions taken by the same judge involving Nigerian women. The other extra-legal factor considered by the judge is the absence of a reception