Criminal Procedure And Public Prosecution Reform In Italy: A Flash Back

This paper addresses the struggle of the Italian justice system during the transition to implement the 1989 Code of Criminal Procedure. The Code was intended to solve some of the major problems of the Italian Courts, such as backlog and length of trials, by adopting special procedures and an adversarial model. Justice system officials assumed that the oral process would accelerate the pace of litigation and, thereby, increase the efficiency and the effectiveness of the administration of justice. In retrospect, the assumption was simply wrong. On the contrary, the Italian experience demonstrates how a significant reform in the criminal process and procedure involves not only amending the rules; in addition, it must consider the institutional and organizational contexts, the actors’ capacity to absorb change, and the risk that change may worsen the functioning of the criminal process. The paper also shows how the legal formalism that pervades the Italian justice system saddled the reform with rules that failed in practice to work.


INTRODUCTION
This paper focuses on select aspects of the Italian Code of Criminal Procedure in action following its implementation on 24 October 1989. I will use a judicial administration approach by reviewing the law in practice rather than the law in books, 1 attending more to caseflow management and organizational issues than to legal aspects.
In particular, I focus on select issues linked with the initial impact of the criminal process reform: the changing role of judges, prosecutors, and defense attorneys; the difficulties entailed in learning how to present oral argument and the related record; and the effort to introduce so-called special proceedings. I end the paper with conclusions based on the facts described.
The data and the information presented in this paper reflect extensive field research initially completed by the Research Institute on Judicial Systems of the Italian National Research Council and the Research Center of Judicial Studies at the University of Bologna when the Code entered into effect 2 and subsequently updated with the analysis of quantitative data and legislation.

A. LIMITED JURISDICTION FIRST-INSTANCE COURTS:
The first single-judge court of limited competence is held by the Justices of the Peace (Giudici di pace). 12 There are 849 Justices of the Peace offices nationwide which have limited jurisdiction in civil and criminal matters. Justices of the Peace do not receive a monthly salary but are paid on the basis of their actual work (i.e. they receive a certain amount of money for each decision, written judgment and hearing).

B. GENERAL JURISDICTION FIRST-INSTANCE COURTS:
The courts of first instance with general competence are called Tribunals (Tribunali). There are 165 of these all over the country, 13 as well as 222 detached offices 14 . These courts sit either as a single-judge or a three-judge panel depending on the type of case. Within these courts, there are specialized units such as the Judge for Preliminary Investigations (Giudice per le indagini preliminari) and the Judge for Preliminary Hearings (Giudice per l'udienza preliminare). The Judge for Preliminary Investigations checks the work of the public prosecutors in many ways. For example, the judge must authorize tapping a suspect's telephone, pre-trial detention, and suspension of the investigation upon the prosecutor's request. The Judge of the Preliminary Hearing is in charge of the formal indictment of the suspect as well as application of special proceedings introduced by the 1989 Code of Criminal Procedure. Of the 165 courts of first instance, there are 93 units called Assize Courts (Corte di Assise) which have criminal competence in serious crimes such as murder and kidnapping that can entail a sentence ranging from 24 years to life imprisonment. 15 Their composition includes two career judges and six citizen jurors. Of the 165 first-level courts of general competence, 26 are units called Revision Units (Tribunale della libertà o del riesame), which are in charge of revising court orders dealing with personal restraints within a judicial district. 16 In addition, a reform 17 has established a new special unit in only 12 of 165 courts of first instance and in 12 Courts of Appeal to deal with patents and intellectual property litigation.
The current court structure came into effect in July 1999. Previously, there was another court of first instance: the Pretura, a single-judge court with limited jurisdiction in civil and criminal matters. In 1999 the Preture were merged with the Tribunali (courts of first instance), creating just a single court of first instance of general competence.

C. GENERAL JURISDICTION FIRST-INSTANCE PROSECUTION OFFICES:
Attached to each of the 165 courts of general competence are 165 Public Prosecutor Offices.

D. "SPECIAL JURISDICTION" PROSECUTION BUREAUS FOR ORGANIZED CRIME:
In 26 of those Prosecutor Offices, there is a special antimafia team called the Antimafia District Bureau (Direzione distrettuale antimafia). Established in 1992, these bureaus oversee all mafia cases within a specific judicial district --the geographical area of each of the 26 Courts of Appeal. The work of these 26 special teams is coordinated by a central office in Rome called the Antimafia National Bureau (Direzione nazionale antimafia), formally attached to the General Prosecutor's Office of the Supreme Court. This National Bureau is directed by an Antimafia National Prosecutor and, among other responsibilities, it promotes mafia investigations through intelligence work. To preserve the independence and autonomy of each Public Prosecutor, the Antimafia National Prosecutor has only a single deputy and exercises no hierarchical authority over the regional offices. For the same reasons, within each Public Prosecutor Office, the Chief Prosecutor has very limited, if any, hierarchical authority over the deputies.

F. "SPECIAL JURISDICTION" PRISONER SURVEILLANCE COURTS:
A specialized competence over detainees and prisons has been granted to the so-called Surveillance Courts (Tribunale di Sorveglianza). They number 58 and are organizationally separate from the courts of first instance where judges (Magistrato di Sorveglianza) deal with all matters related to the treatment of detainees. The decisions of these courts are rendered by a single judge and may be appealed to a panel of four whose members include two career judges and two experts.

G. "SPECIAL JURISDICTION" JUVENILE COURTS:
Civil and criminal cases with defendants between the ages of fourteen and eighteen years are handled by 29 Juvenile Courts (Tribunale dei minorenni), to which specialized Prosecutor Offices are 12 The Justices of the Peace were introduced in the Italian justice system by statute n. 374, 21 November 1991, which came into effect in May 1995. The Justices of the Peace are temporary judges. They have to be qualified for the Bar to be appointed. 13 The Legislative Decree n. 491 of 3 December 1999 established two new courts of first instance with general competence (Tivoli and Giugliano). They were created nearby the courts of Rome and Naples, in order to decrease the huge caseload of these two courts. To date one of the two, the Giugliano court close to Naples, has never been in operation since it was established by statute in 1999. Therefore, theoretically, the courts of first instance with general competence are 166, but de facto they are 165 plus a "shadow court".
14 Only the hearings that require a single-judge court can be held in the detached offices of the "Tribunals". 15 There is also some exclusive competence for certain crimes which entail a prison sentence of at least ten years. 16 The "Revision Units" are not really separate organizational units within the courts; they are specific panels of three judges which decide on court orders dealing with personal restraints within a judicial district. 17  attached. Decisions are issued by panels of four whose members include two career judges and two experts, one male and one female, with professional expertise in social assistance, psychology, and related disciplined. Appeals from these courts are heard by a special unit of the Courts of Appeal.

H. FIRST-INSTANCE COURTS WITH "SPECIAL" SECOND-INSTANCE JURISDICTION:
The courts of first instance, sitting as a single-judge benches, also function as second-instance or appeal courts for the decisions made by the Justices of the Peace.

I. GENERAL JURISDICTION SECOND-INSTANCE OR INTERMEDIATE APPEALS COURTS:
The 26 Courts of Appeal (Corte di appello) -and three detached divisions -function as appeal courts for decisions rendered by the courts of first instance (Tribunali). Within these Courts of Appeal, there are several specialized units, including the Appeal Courts of Assize (Corti di assise d'appello) which handle appeals from the Court of Assize of first instance, and the specialized juvenile units, which hear appeals from the Juvenile Courts. These various units of the Courts of Appeal sit in panels of three judges, with the exception of the Corte di assise d'appello which sit in panels of two career judges and six citizens who serve as jurors.
The appeal process considers both factual or evidentiary matters and legal issues. In effect, new evidence can be submitted during the appeals process, a practice that has the potential to significantly protract the length of the criminal process on the appellate level.

J. GENERAL JURISDICTION SECOND-INSTANCE OR APPEALS LEVEL PROSECUTION OFFICES:
Attached to each Court of Appeal there is a Public Prosecutor Office which prosecutes cases on the appellate level.

K. GENERAL JURISDICTION SUPREME COURT OR COURT OF LAST INSTANCE:
Italy's highest court is the Supreme Court (Corte di cassazione), located in Rome. It deals with questions of law and conducts final appellate reviews of all provisional orders relating to personal restraints (art. 111 Cost.). The Supreme Court's jurisdiction extends to ensuring the uniformity of how the law of the state is interpreted and applied. 18 The Court is obligated to review all the pleadings that are appealed to it; it does not have a writ of certiorari discretion. This requirement generates a substantial caseload which, in turn, requires a disproportionately large court of final appeal. The Supreme Court currently has in excess of 300 judges, which is exceedingly high when compared with the terminal courts of appeals of other European judiciaries. 18 Notwithstanding Italy's status as a country in the civil law tradition in which the stare decisis doctrine does not apply, precedents have always played a major role in passing judgments. In particular, the judgments by the Supreme Court have always affected the decision-making process of the trial courts.

A. THE OLD CODE -LACK OF PROCEDURAL DUE PROCESS PROTECTIONS:
The structure of the first Code (known as codice Rocco, from the Minister of Justice who played a major role in its preparation) was based on a non-adversarial or inquisitorial model and hierarchical officialdom. 19 During a closed pretrial inquisitorial phase, evidence was gathered to determine whether a crime had been committed and, if so, by whom. Formally, a judge controlled the pretrial examination, performing the role of both judge and investigator. The defendant had no right to participate or even to be notified of the investigation. Without the presence of the defendant or defense counsel during the examination proceeding, interrogators could exert considerable pressure on witnesses who appeared before them. Following the examination phase, a trial proceeding was conducted for presentation of the evidence and structuring of a case on the basis of which the defendant could be convicted. The principles of orality and immediacy were abandoned; records and materials collected during the investigative phase became the basis for the verdict and sentence.

C. ANTICIPATED CONSEQUENCES OF IMPLEMENTING THE NEW CODE:
The new Code reflected an effort to transition from an inquisitorial to an adversarial model of criminal procedure, 24 one that has its distinctive origins in liberal ideology and is considered a more functional and transparent litigation model for liberal democratic states. 25 The reform effort also was motivated by the enormous backlog of pending cases and the archaic and time-consuming procedural requirements of the old code. The enormous investment of time and effort required by various court procedural rules, and its harmful impact on due process rights and the ability of the court system to deliver timely justice − an impact for which Italy has been repeatedly condemned by the The adversarial mode of proceeding takes its shape from a contest or a dispute. It unfolds as an engagement of two adversaries before a relatively passive decision maker whose principal duty is to reach a verdict. The non adversarial mode is structured as an official inquiry. Under the first system, the two adversaries take charge of most procedural action; under the second,

D. ASSUMPTIONS:
The decision to move from the old inquisitorial model to an adversarial one was based on two major assumptions: First, that the adversarial procedure is more compatible with democrat government; and second, that there was a pressing need to restore effectiveness and efficiency to the administration of criminal justice.

E. PRIMARY OBJECTIVES:
The two primary objectives of the reform were (i) to terminate eighty to eighty-five percent of new criminal cases prior to trial by employing primarily Italian plea-bargaining (applicazione della pena su richiesta della parti), and (ii) to accelerate the trial process through the use of other special procedures created by the new Code such as direct trial (giudizio direttissimo); abbreviated or summary trial (giudizio abbreviato); immediate trial (giudizio immediato); and penal decree (decreto penale).

F. ANALYSIS:
It is not clear how these legal reforms by themselves can make the judicial system more efficient and effective. Procedural law is just one of several factors that must be coordinated to effect a substantial change in the quality of justice; it may not by the most important one.
Changing from an inquisitorial to an accusatorial system, and from a written to an oral process, is complicated and difficult. These two models are substantially different. Table 1 summarizes some of the differences. Perhaps the most important and difficult element in the reform process is the initial implementation which can result in its success or failure. At this important juncture, Italy faced some enormous problems that, in hindsight, could have been be foreseen.

A. THE DIFFICULT RELATIONSHIP BETWEEN PROSECUTORS' OFFICES AND THE POLICE:
One problem with the former Code was the discretion it allowed the police with regard to how much time they could take to file criminal complaints. It was considered a source of possible abuse and a violation of the constitutional provision of "mandatory penal action." According to the Italian Constitution, 27 the prosecutor must file a criminal complaint if there are reasons to suppose that a crime has been committed. 27 Cost. art. 112.

Inquisitorial Accusatorial
Offense This principle of mandatory penal action or compelling initiative of criminal proceeding by prosecutors was a reaction to the Fascist period, during which prosecutors were under the control of the executive power and many abuses occurred. 28 This principle should have guaranteed the equal treatment of citizens. 29 "In fact, however, penal actions in Italy are largely discretionary. The magistrates to whom this task is assigned institute proceedings not only on request from outside (reports and accusations from the police, private citizens or public authorities) but also on their own initiative. In other words, it is quite legitimate for them to carry out, with the greatest independence, investigations of any kind, of any citizen, using the various police forces to verify whether the offenses they (more or less justifiably) assume to exist have actually been committed." 30 To reinforce the monopoly over initiating criminal proceedings, the 1989 Code of Criminal Procedure mandates that a complaint (or notice) that a crime was committed must be filed in the prosecutor's office by the police within forty-eight hours. 31 The major objectives of this rule were: (i) to eliminate the discretion of police agencies which are arms of the executive power, in determining the pace of the investigations; and (ii) to involve the prosecutor in the investigations as soon as possible in order to improve the administration of justice by advancing the process.
In practice, however, the rule did not fulfill the expectations. In the first two days after the Code went into effect, police agencies, in their efforts to comply with the new forty-eight hour rule, overwhelmed the prosecutors' offices − particularly those in the courts of limited jurisdiction which handle about 85% of the entire caseload − with crime reports that they had collected throughout the year. As a consequence, prosecutor offices were flooded with thousands of documents that, under the law, were supposed to be recorded "immediately" 32 by hand in an archaic docket registry book. 33 In effect, although crime reports were effectively filed in the prosecutors' offices by the police within the required forty-eight hours, most languished without action for six months or longer while the prosecutors struggled to deal with the backlog.
Moreover, even if the prosecutor's office could immediately record all the crime reports, it is hard to believe that they would be able to manage the huge caseload effectively. (In some cities, each prosecutor already has a caseload of more than 5,000 cases.) This was not the only problem inadvertently created with implementation of this new rule. Because the prosecutor is legally in charge of managing the investigations, the police determined it was unnecessary to conduct any investigation or assemble evidence prior to receiving the prosecutor's instructions. As a consequence, the police were losing or misplacing important pieces of information that should have been collected and secured immediately following the crime. In addition, conflicts between police officers and prosecutors had been increasing. Many police officers had little confidence in the ability and competence of prosecutors to lead criminal investigations, in part because the only training many of them received was on the job.
After three years of trouble, the government finally amended the forty-eight hour rule. 34 Now the police must transmit a crime report to the prosecutor's office "without any delay." This amendment helped to adjust the flow of reports, but it failed to address the fundamental challenge of how to respond effectively to the mandatory criminal action requirement which compels courts and prosecutors' offices to deal with an enormous paperwork burden to the detriment of the substantive investigation process. 35 The best intentions of the Italian Parliament notwithstanding to promote the principle that all crimes have to be prosecuted in a timely manner, the Italian Judiciary has tightened itself in a vicious loop. 34 Law n. 356, 7 August 1992. 35 Although reliable data are not available, it seems that about 95% of crime reports with unknown perpetrator submitted to the prosecutors' offices are dismissed by the judge without any real investigation. B. FORECASTING CASELOAD AND ANTICIPATING JUDICIAL NEEDS: Many prosecutors' offices were unable to manage the unanticipated increase in filings because their offices were insufficiently organized, staffed, and equipped. No legislative impact studies had been undertaken to assess what resources would be required to process the additional caseload. The Ministry of Justice does not have a functional system to determine judicial needs and to forecast judicial caseloads. Organizationally, the Italian Judiciary suffers from an inadequate strategic planning function, resulting in the failure to adequately orchestrate resources to deal successfully with the practical requirements of implementing the new Code. Indeed, a major judicial administration concern is the lack of reliable data on extent to which implementation of the new criminal code has been successful and its overall impact on the performance of the judicial system. Installation of automated caseflow management systems in recent years has led to modest increases in the reliability and availability of data. The primary cause of the modest increases appears to be not so much the unavailability of technology to handle data, but the judges and court officials have not yet come to recognize how important the recording of accurate and complete case information data can be in helping to manage institutional operations and in anticipating future requirements, In short, in many courts, accurate data collection and preservation appear to be viewed as a waste of time rather than an investment.
Pursuant to the Constitution, the Ministry of Justice is responsible for the organization and the functioning of the judicial and prosecutorial system. However, the Ministry lacks a qualified technostructure 36 that can be useful to analyze, design, plan, and implement progressive change in an organization. Almost all the executive positions in the Ministry of Justice (e.g. Director of the Prison Department, Director of Information Technology, Director of Statistics etc.) are held by magistrates who, with some exceptions, have neither sufficient training nor experience essential to their positions. Rather, they bring to them an archaic and formalistic approach which impedes their success and the adoption of flexible and progressive management principles and practices in the Ministry. This lack of organizational sensitivity and management competence give rise to problems in key areas of administration such as strategic planning, information technology, performance appraisal, finance and budget, and human resources management. The resulting dysfunctions, which undermine initiatives for dynamic reform, linked with the inadequate performance of the criminal justice system, were foreseeable had the legislators, legal scholars, and judicial officials who wrote the 1989 Code and who were in charge of the implementation process, adopted a more pragmatic, proactive, and goal-oriented approach.
Italian policy makers and criminal justice system executives continue to maintain that the function of criminal procedure rules is to yield a deductive and abstract intellectual effort rather than viewing them as a set of important organizational tools that facilitate the efficient administration of criminal justice. Therefore, very little attention has historically been paid to substantive application of the rules and to their organizational aspects.

A. DEFENSE ATTORNEYS
Another major problem that emerged with adoption of the new Code was the balance between the prosecutor and the defense attorneys, a problem that was exacerbated by the changing role of judges. In the adversarial model, criminal defense attorneys play a very active role. They conduct investigations to gather exculpatory evidence, and they must be prepared to mount a vigorous defense that includes examining their own and cross-examining prosecution witnesses. To collect exculpatory evidence, they have to acquire investigative skills and recruit professional private investigators to assist and support them in seeking evidence. 37 In Italy, prior to adoption of the new criminal code there were few such professional investigators because the existing inquisitorial model did not need them. As a consequence, defense attorneys had difficulty finding qualified investigators. That shortage has largely been eliminated, but for a number of years following implementation of the code, there were inadequate investigative resources available for the defense. Italy does not have a public defender system. Where an indigent person is charged with serious criminal behavior, privatesector defense attorneys are appointed by the court from a list prepared by the local bar association.

B. PUBLIC PROSECUTORS
Public prosecutors in Italy enjoy a privileged status among their Western European colleagues. Their status as judges provides them with status and benefits that include career protections, professional independence, immunity from liability, etc. In addition, the Constitutional principle of mandatory penal action authorizes them to exercise broad discretion in setting priorities and committing criminal justice system resources (e.g. wiretapping, number of police officers to involve in an investigation etc.) to the process of conducting a criminal investigation. Implementation of the new Code enhanced and strengthened their roles. For example, a prosecutor's decision to dismiss a case, to incarcerate a defendant for pretrial detention, or to authorize a wiretap on a telephone, must be reviewed by the preliminary investigation judge (giudice per le indagini preliminari). This judicial position − a creation of the 1989 Code that eliminated the traditional investigating judge of the inquisitorial tradition − is authorized to indict a defendant when the prosecutor has collected enough evidence to go to trial. These checks by a judge are considered necessary, because, as pointed out before, the prosecutor is compelled to initiate a criminal proceeding under the Italian Constitution; therefore, prosecutors cannot dismiss a case without judicial control. In practice, however, these judicial checks are often little more than legal formalities which are easy to overcome. Indeed, judicial controls are now exercised routinely and primarily only on the amount of evidence that the prosecutor intends to introduce.
The role of prosecutors has also been strengthen by what it has been called a countereform. This countereform, which reverted some of the oral innovations of the new Code back to written processes, was determined mainly by Constitutional Court decisions 38 and by a government law in the aftermath of the assassination of the General Manager of Criminal Affairs of the Ministry of Justice, Giovanni Falcone, and his colleague, Paolo Borsellino, who for many years led the investigations into mafia crimes in Palermo. 39 It is impossible to analyze in this brief overview all of the details attending the significant changes that occurred as a consequence of this tragedy. One of the most important provides that in organized crime trials, certain categories of evidence collected for other trials or during the preliminary investigations can be introduced in written form, primarily for the protection of potentially targeted witnesses who otherwise would have to appear in court and offer oral testimony. This was a substantial retreat to certain provisions of the inquisitorial model, and it is perceived as weakening the role of defense attorneys and strengthening that of prosecutors. Here it is important to point out how the full accusatorial and inquisitorial models are two ideal opposites on a continuum within which every criminal judicial system eventually locates the position which best suites its criminal justice system.
It also bears brief mention that an amendment to the Constitution (n. 2, 23 November 1999) modified article 111 of the Constitution by introducing a provision for the fair trial, similar to article 6 of the European Convention of Human Rights, to balance the impact of the inquisitorial countereform.

C. JUDGES
In the inquisitorial model, the public hearing is much less important than in the adversarial model. In the former, the judge already knows and understands the case prior to the trial. Documents and transcriptions of depositions taken before the investigating judge are part of the file. The public hearing is the end of a long process that transpires in successive stages. Essentially, the trial in the inquisitorial procedure is something quite different from the public accusatorial model trial. "The file of a case [was] the nerve center of the whole process, integrating various levels of decision making […] The trial was not concentrated, but organized in installments. After a matter has been considered at one session, new points can emerge to be the subject of another session, and so on, until that issue seems thoroughly clear". 41 In the adversarial model the trial, rather than the case file, is the focal point of the litigation process. "Attorneys on both sides must have good oral skills. The drama of a case is concentrated on few hearings and "requires that decisions be based largely on fresh impressions, including surprise, shock, the spell of superficial rhetoric, and, perhaps, even theatrics". 42 In sum, the litigators' performances have to be remarkable to allow the decision maker, the judge, to be informed of the details of the case through the parties' process of evidentiary presentation and argument.
The difference between these two opposing models is substantial. To trained observers who watch videotaped trials and various interviews, trial attorney performance in general and public prosecutor performance in particular requires several years of professional trial experience to build up to a level that is acceptable and appropriately competitive. Making the transition from a file-based process to an oral-based one is time consuming and requires commitment and effort. In the United States, only a minority of practicing lawyers learn or are trained to function as active trial-level litigators. Trial-level prosecutors are provided fairly significant training in how to present the case to court, to examine their own witnesses, to cross-examine defense witnesses, and to address a jury. The poor performance of Italian prosecutors in this transition to the adversarial model, linked with the civil law tradition where judges play a far more proactive role during the trial (they are responsible for developing the evidence, calling and questioning witnesses, etc.), generated another problem. Judges are supposed to remain neutral in an adversarial model, but in Italy, given the inertia of the old system, they frequently question witnesses directly, reverting back to the traditional inquisitorial model. 44 Here again, the criminal law reform required adaptation to and acquisition of new skills and patterns of actions 45 which cannot not be learned and integrated overnight.

THE CALENDAR AND THE TRIAL RECORD
Making the transition to the oral process from the written process anticipates a much greater need for adequate preparation and extemporaneous performance skills. To adequately organize and prepare, it is very important to design and implement an effective calendaring system as well as a verbatim trial recording system that is reliable, accurate, and complete.

A. CALENDAR MANAGEMENT
In several Italian courts, managing the calendar in a manner consistent with the new oral process remains an elusive objective for prosecutors, judges, police officers, counsel, and even witnesses. Effectively scheduling events for the investigation, pretrial hearing, and trial calendars is critical to effective management of the pace of litigation. It permits the prosecutor who led the investigation to easily track the case all the way to court.
In Italy, the case is initiated by the prosecutor's office. If not carefully managed, it may generate calendar definition problems. A review of the various stages of case processing clarifies what these problems are.
The prosecutor who has a major role in the investigation typically determines when to file the case in court. 46 The case is randomly assigned to a preliminary hearing judge who determines whether to indict the defendant or dismiss the case. If the judge decides to indict the defendant, the case will be assigned to a trial division for public hearing. Because the preliminary investigating judge, the trial division, and the prosecutor do not share an integrated calendaring and scheduling system and because there is no specialization by criminal case type or jurisdiction, coordinate their respective activities and scheduled is a difficult and time-consuming proposition, thus delaying prompt case processing. For example, five different cases assigned to the same prosecutor could be assigned to five different preliminary hearing judges, as well as to five different trial divisions, and all of them have independent calendars. As a consequence, it is not unusual for the same prosecutor to have two or more scheduled hearings in different courts at the same time or at overlapping times. Because he or she cannot simultaneously be present at the same time in different courtrooms, there incidence of adjournments is very high; moreover, where a substitute prosecutor is designated who has insufficient time to become familiar with the particulars of the case, his or her performance frequently is inadequate.

B. THE TRIAL RECORD
Also very important for the trial process in an adversarial system is the manner in which the oral record is taken and preserved. In the courts of the United States on both the state and federal levels, a variety of recording systems are deployed, ranging from the traditional stenotyping to the most advanced CAT (Computer-Aided Transcription) and CIC (Computer-Integrated Courtroom) systems to digital audio and videotape court reporting. The oral trial needs a recording system that is reliable, complete, accurate, and easy to use, 47 among other reasons for purposes of presenting a clear and complete record of the trial if the decision in the case is appealed.
Unlike the adversarial model, the inquisitorial procedure, which relies heavily on content of the file, can utilize short summaries of the oral discussion and witness testimony rendered in a hearing because most of the documentation is the dossier. 48 In most court systems that continue to rely on the inquisitorial model, the presiding judge will dictate short summaries of the proceeding to a court clerk who records them either in handwriting, on a typewriter, or on a PC. risk, of course, is that the judge's summarized description may contain biases or comprise an incomplete record of key witness testimony. For that reason, some inquisitorial-based systems are adopting verbatim recording systems to ensure the accuracy of the record.
The new oral trial model has forced the Italian courts to address the recording problem. The legislators, in part due to lack of useful information about state-of-the-art recording technology, selected stenotype as the standard and, for particular events, audio or video recording with a summary account of the interrogations. 49 The Ministry of Justice then recognized that there was an insufficient number of skilled stenotypists to staff the nation's courtrooms. The Ministry then rushed to procure and test audio-recording systems for the courts and to train courtroom staff in their use. Some courts took the initiative to locate and enter into contracts with free-lance reporters. A few months later, the Ministry sponsored pilot testing of videotaping systems in six courts. Following relatively successful results of the pilot test at the end of calendar year 1992, the Ministry of Justice purchased eighty of these video systems to deploy in the trial courts.
Judicial assessments of these different systems are as varied among judges in Italy as they are among judges in the United States. Generally speaking, most Italian judges who understand the role and function of the adversarial model prefer the videotaped record. They find it helpful in complex cases and particularly useful in organized crime or mafia trials where visual communication is extremely important. However, it is necessary to revise the law to make the videotape record really useful both at the trial and at appeal levels. The primary problem is the transcript which is still considered necessary even where video or audio recording is made. Judges, prosecutors, and attorneys have relied on paper for so long that transferring to an electronic recording involves too much change too quickly. Over time, each court appears to have migrated to a hybrid that responds to the needs of its judges, leaving justice system officials with a heterogeneous mix of systems that is costly to operate, administer, and maintain.

THE ITALIAN SPECIAL PROCEEDINGS
A major purpose of the new Code was to create new proceedings that would serve to help reduce enormous pending criminal case backlogs and to accelerate processing of the caseflow. The Code introduced alternative proceedings to the trial referred to as "special proceedings." 50 Their goal is to reduce the number of criminal cases that go to trial by using a variety of abbreviated procedures designed to encourage agreement among the parties. Table 2 summarizes the different special proceedings available and their main features. This paper cannot describe or analyze all of them; it focuses on the Italian version of plea-bargaining. 51 PLEA BARGAINING: Italian plea-bargaining is very different from the version of plea-bargaining generally utilized in the United States. Although the goal is supposed to be the same − settling cases before trial − the rules that govern the two systems and their respective effects are quite different. Italian plea-bargaining allows the defendant to opt for, at the most, a one-third reduction in the sentence he otherwise would receive based on the charges in the indictment. The reduced punishment cannot exceed five years; otherwise, the plea-bargaining is not applicable. 52 Unlike plea-bargaining in the Unites States, Italian prosecutors cannot bargain by reducing the number or kind of charges. In addition, the Italian model permits defendants to bargain pleas with the judge, even when the prosecutor does not consent. 53 TABLE 2 -Italian special proceedings: main features A very important aspect of plea bargaining has not been adequately considered in Italy. To effectively utilize pleabargaining to decrease the backlog of pending cases that are tried, it must be considered from the hurdles race perspective. In the United States, from initial filing of the charge through the preliminary hearing or indictment to the time of the trial, prosecutors may repeatedly approach the defendant with variations of a plea-bargain. Many prosecution offices as a matter of policy, make their best, most lenient offer early in this sequence. The longer the defendant waits, the greater the likelihood that the generosity of the original plea-bargain offered will be diminished. There is an inverse relationship between the amount of work the prosecutor invests in the case and how generous the offer is likely to be. There is little utility in bargaining when the trial is imminent − the prosecutor is prepared and all the paper work has been completed and filed with the court.
In Italy, plea-bargaining is not conducted in the context of a hurdles race. Defendants can plea-bargain or consent to the other special proceedings without penalty as late as the day of trial before the judge and either with or without the prosecutor's consent. 54 To that extent, no incentives exist to motivate defendants to opt for alternatives to trial, especially if they are not being held in pretrial confinement. Defense attorneys, unless they are court-appointed counsel, have no incentive to settle cases because they can earn more if the case is brought to trial. The only strong incentive to plea bargain is triggered when a defendant is caught red-handed in committing a crime because a plea bargain will result in reducing the punishment by one-third where the likelihood of being convicted is very high.
Plea bargaining, as well as the other special proceedings, have only slightly diminished the number of cases that went to trial in the first years of the implementation of the Code. 55 Things have changed during the years, particularly after the 2003 reform that enlarged the possibly to plea bargaining, with an average of about 35% of criminal cases brought before 54 See C.p.p. art. 446. 55 Although reliable data on the use of plea bargaining in all types and phases of proceedings were, and still are, scarce, from October 1989 to December 1993, the average nationwide percentage of plea-bargainings, on the total of the prosecutors' office filings to the courts of limited jurisdiction (preture) was less than 1.5% and less than 4% for the courts of general jurisdiction (tribunali). The other special proceedings also were not very successful. From October 1989 to December 1993, in the courts of limited jurisdiction there were 0,3% giudizi abbreviati, 1.9% direttissime, 25.1% decreti penali. In the courts of general jurisdiction, there were 1.5% giudizi abbreviati, 3% direttissime, 4.6% decreti penali, 2.6% giudizi immediati (Ministry of Justice, Monitoring Office, Criminal Division, 1994). Appeal is always allowed the judge that are disposed by special proceedings. 56 However, this percentage is still too low to give some relief to the overwhelmed criminal caseload, which still suffer of an outstanding number of pending cases, an extremely slow length of the procedure and a consequent huge percentage of criminal case that end because they go over the statute of limitation. 57 The integrity of the justice system is strongly diminished under these circumstances.

CONCLUSION
This paper has addressed important issues that have confronted and continue to confront Italian prosecutorial offices and courts since implementation of the 1989 Italian Code of Criminal Procedure. The Code was designed to address major problems with which the Italian Courts were wrestling, including enormous pending case backlogs and unnecessarily lengthy proceedings, by adopting special procedures and an adversarial model.
The Italian experience illustrates that assuming that the substitution of the oral for the paper process for purposes of accelerating the pace of litigation and improving the administration of justice was both wrong and naive. On the contrary, if criminal justice reform efforts are not adequately researched, planned, organized, and managed, the overall consequences can increase rather than diminish the difficulties with which the criminal justice system must deal. Substituting the oral process is just one of many important steps that may have to be taken to effect sustainable improvements in the administration of justice. A prime lesson learned in Italy is that it makes no sense to change the criminal procedures rules without simultaneously reforming the institutional roles of prosecutors, judges, and defense counsel.
The legal formalism that pervades the Italian justice system saddled the 1989 Code with rules changes that have been shown to not work in practice. The Constitutional principle of mandatory penal action requirement imposed on prosecutors generates enormous paperwork to the detriment of substantive action by law enforcement. Implementation of this Code provision has demonstrated that this ideal principle is impractical and does not work in practice. Conflicts between prosecutors and police agencies have been emerged as a result of the prosecutors' expanded role in managing investigations. Italian plea-bargaining and other delay reduction measures under the rubric of "special proceedings" have not fulfilled their initial expectations of reducing the number of cases that proceed to trial. The Italian experience also has show that cultural traditions and practices are often difficult impediments for lawyers and magistrates to overcome when attempting to integrate into an old system fundamentally new practices and procedures that may be perceived as counterintuitive and questionable. Judges have complained about prosecutors' poor performances during the oral trial process. Their substandard performance detracts from the quality of the public hearing and may impact the judges' ability to reach the correct decision. On the judges' side, they also had several difficult transitions to make in adapting to their new adversarial role.
"By definition, a reform is intended to cause change, and, in contrast with normal behavior, change conveys elements of risk and uncertainty that, themselves, tend to inhibit action [...] one must plan a reform with specific attention to normal behavior within the system to which the reform is to be applied". 58 In the Italian case the negative impact of these drawbacks might have been minimized had more attention been focused in developing the Code, planning for its implementation, and preparing and training the key players.