2021 - Serbia Judical Functional Review
      
2021 - Serbia Judical Functional Review

Quality of Justice Services Delivered

Key findings

  1. The Serbian judicial system continues to struggle to fully comply with ECHR requirements, as evidenced by the large caseloads in Strasbourg. Non-compliance tends to be found in a significant number of case types, highlighting specific problems relating to non-enforcement of the final decisions, length of proceedings, protection of property, and lack of effective investigation. In addition, there are challenges in the enforcement of ECtHR judgments, and further actions are needed to establish organized coordination between all various state bodies.
  2. Overall, judges and prosecutors think that judicial quality has improved since 2013, but lawyers see less improvement. Unreliable data quality and availability, inconsistency in jurisprudence, and fragmented administrative systems are overarching challenges in addressing court system quality. On the positive side, members of the public who have been involved in court cases are generally satisfied with court quality.
  3. Citizens and the business sector are highly satisfied with the quality of notary work, while there has been a decrease in public satisfaction with court administrative services. While most members of the public remain satisfied with the quality of court administrative services, the downward trend in satisfaction should be compared with positive public opinion about notaries. Part of the courts’ administrative responsibilities was transferred to notaries in 2014, and public satisfaction suggests that the reform was successful.
  4. There are some concerns about impartiality. These include lawyers’ perceptions of selective enforcement of laws. Prosecutors have complained the police do not cooperate with them during investigations. Conversely, lawyers complain that they do not have access to all the information that prosecutors and judges have. Further, there is a concern that wealthier people may obtain deferred prosecution by making monetary donations to good causes, and those decisions to drop prosecutions are sometimes politically motivated.
  5. Because of gaps and ambiguities in legislation, laws are not applied consistently, and unwarranted appeals are filed, and, conversely, lower court decisions are reversed on appeal. Two related issues are the clarity of legislation and its application in the judicial systems. Regarding the first, about 40percent of judges, 37percent of prosecutors, and 46percent of lawyers believe that laws are ambiguous and inconsistent to a great extent or to some extent. While lawyers’ perceptions have improved over time, there has been uneven progress in judges’ and prosecutors’ perceptions.
  6. The proliferation of new legislation continues, often without analysis of the impact on or harmonization with existing laws. Ad hoc working groups are convened to consider and draft each new law, but there is sometimes an inadequate representation of stakeholders, working group members report inadequate guidance, and proposals are not necessarily subjected to formal analysis. Legislation continues to be routinely passed by the National Assembly under emergency procedures and without sufficient transparency.
  7. 84 percent of judges cited that less frequent changes in laws could contribute to a better quality of justice services. Criminal prosecution provides an example of the impact of frequently changed legislation and the quality of judicial services. The Criminal Code was amended 10 times over the last 15 years. During this period of change, offenses can be charged as both criminal and misdemeanor offenses - or as both criminal and commercial offenses. The same incident burdens the courts twice: once for the misdemeanor offense, with its procedure and legal remedies, and again for a criminal offense with its procedure and legal remedies.
  8. Following the enactment of new legislation, there have been challenges in implementation. These include limited outreach and training. A primary example is low awareness of the availability of free legal aid (see Access Chapter).
  9. Inconsistent interpretation of laws and inconsistent jurisprudence remain challenges for the Serbian judiciary. 70 percent of judges and prosecutors and 90 percent of lawyers stated that inconsistent interpretation of laws and inconsistent jurisprudence happen at least from time to time, if not often. More than 80 percent of lawyers reported that selective implementation of laws and non-enforcement of laws occurs frequently, but only about one-third of judges and prosecutors shared this view. Judges’ and prosecutors’ perceptions have been slightly improved since 2013, but lawyers’ perceptions have worsened over the time, especially in the area of selective enforcement of laws.
  10. The judicial system still lacks a standardized approach to routine aspects of case processing. The quality of case processing has not improved significantly since the 2014 Judicial Functional Review. There are no checklists, standardized forms, or templates for routine aspects of case processing, nor is there a consistent approach to drafting routine documents, such as legal submissions, orders, or judgments.
  11. There are few examples of specialized case processing for the types of cases that often warrant a tailored approach. The law on the prevention of family violence is an example of the potential for improved coordination in case processing. It envisages the establishment of a group for coordination and cooperation (Article 25) that consists of representatives of public prosecutors, police, center for social work, and, if there is a need representatives of other institutions (educational, employment services, etc.).
  12. Lawyers who represent criminal defendants in particular point to shortcomings in information and communication technology. For instance, some databases are available only to judges and prosecutors. There is no comprehensive countrywide system to process and interlink cases across courts and prosecutorial networks.
  13. There is a continuing lack of data about the reasons for dismissals by prosecutors. Since 2013, Serbian law has allowed the filing of complaints about the dismissal of criminal complaints, and Serbians have made extensive use of this process.
  14. The number of cases concluded by plea bargaining decreased by eight percent in 2019 due to a 17 percent drop in plea bargains in the Belgrade appellate region.
  15. The implementing legislation for deferred prosecution is incomplete and imprecise, prosecutors’ decisions are not uniform, and guidelines and criteria for its use are missing. There is a lack of consideration for the interests of the victims of the crimes involved. The conditions imposed in deferred prosecution measures seldom benefit the community at large through rehabilitation programs or community service. The most frequent condition is a cash donation to humanitarian causes. This can give the impression that defendants have bought their way out of the criminal justice system.
  16. The lack of official guidelines and political will for cooperation between police and prosecutors continue to impede the effective investigation of criminal cases. Prosecutors have no practical means for compelling police to follow their directions. Prosecutors reported this problem arose particularly in cases that might have political implications. In addition, when police submit both misdemeanor and criminal charges for the same incident, they often do not inform the prosecutor, which leads to duplication in court proceedings, as noted above.
  17. Serbia’s prosecutorial system also remains highly hierarchical, with higher-instance Public Prosecutors authorized to control the work of lower-instance ones. Higher-instance prosecutors can take over any matter from a lower-instance Public Prosecutor within his or her jurisdiction and issue mandatory instructions to those lower-instance Public Prosecutors. On the one hand, such oversight could be useful in promoting consistent practices. On the other, it may allow selectivity in prosecution.
  18. Standardized forms and templates used by PPOs are not being updated on a system-wide and regular basis, despite amendments to the criminal code. The use of up-to-date templates and standardized forms would facilitate consistency in routine prosecutorial tasks, reduce mistakes, and fast-track daily actions.
  19. The 2014 Functional Review found the appeals system is at the heart of Serbia’s problems in terms of quality of decision-making and remains high but has declined. The rate of appeals filed and the rate of reversals on appeal, are relevant to legislative quality, judicial quality, and public trust. A high rate of reversals can indicate that lower courts are struggling to interpret ambiguous laws. Lack of uniformity in the application of laws can encourage parties to hope for a more favorable result on appeal.
  20. Trust in the appellate system among court users in Serbia has decreased in the past decade. However, Court users who received an unfavorable judgment filed an appeal in 84 percent of the cases if they considered the decision unfair, an increase by 21 percentage points over the 2014 Functional Review.
  21. Serbian Basic PPOs appealed in 12 percent of cases in 2019 and were successful in only 21 percent of their appeals, indicating that prosecutors may be pursuing appeals that were not justified. Appellate success rates varied significantly among PPO types, among individual PPOs, and over the years. There were no written policies or guidelines governing the selection of cases to appeal. Appeal rates varied considerably among Basic PPOs, including those of similar size.
  22. While appeal rates vary markedly across court types, case types, and court locations, the data management system is not adequate to compare performance. It is not possible to generate a report on lodged appeals or dismissed appeals. It is not possible to distinguish between cases appealed from Basic Courts and those appealed from Higher Courts, which are entered in the same registry.
  23. It is possible that appeal and reversal rates will decline as the quality of judges’ decisions improves. The clarity in written decisions may help the parties, and the reviewing courts better understand the reasoning of the first instance courts. Existing judicial training has improved the clarity of written decisions. The Supreme Court of Cassation has organized round tables to discuss criminal judgments and identify shortcomings and good practices in judgment writing.
  24. As well as improving quality, specialization can result in more efficient use of limited resources. For example, the courts are burdened with many repetitive cases that derive from the same underlying issue. An example is over 56,000 military reservists’ claims. Serbia has not adopted the practice used in some countries of consolidating cases to resolve similar or identical factual and legal claims.

Introduction ↩︎

  1. This chapter assesses the ability of the Serbian judicial system to deliver quality services to citizens and its progress since the 2014 Judicial Functional Review. Quality of justice services was assessed through a range of dimensions, including the uniform application of the law, user satisfaction with the justice services received, consistency with ECHR standards, and perceptions of integrity.247
  2. The quality of the justice system is a significant part of effective justice, underpinning business confidence, job creation, and economic growth and providing protection from violations. However, according to the World Bank 2020 Regional Judicial Survey, more than 40percent of citizens and business representatives in Serbia believe that the quality of the judicial system has not changed over the course of the past three years, although many measures were implemented with the aim of improving e the quality of work.
  3. In comparison to general perceptions about the quality of judicial services, experience with court cases has a positive influence on citizens’ assessment. Citizens with recent personal experience are noticeably more positive about court work quality in their own case (69percent) than is the general public (45percent). The outcome of the case does not seem to play a role in the perception of court work quality. At the same time, business representatives with recent experience in court cases are the most satisfied with court work quality. When court users do perceive low quality, they see bad laws, followed by poor work by the judge, as the main reasons for the low quality.

Quality of Laws and Law-Making ↩︎

  1. The need to have good quality laws is stipulated in the jurisprudence of the ECtHR. Therefore, the legislatures of Member States need to respect the principles of the rule of law and the minimum requirements of good law-making. This aspect includes accessibility to information about laws and policies and foreseeability about how they are applied.248 Otherwise, there can be concerns about arbitrary interference by the public authorities.249
  2. Clearly, the quality of justice depends on the quality of laws and the performance of the law-making system.250 This section looks at three dimensions of the quality of laws: perceptions of the quality of existing laws, the law-making process, and the rollout of recent law reforms.

Perceptions about the Quality of Existing laws ↩︎

  1. Justice system professionals are concerned about whether laws are clear and consistent. Among judges, 40percent believe that laws are unclear and ambiguous to a great extent or to some extent. Among prosecutors, 37percent share that view. Among lawyers, 46percent have that concern, although lawyers’ perceptions have improved over time. Judges’ perceptions of clarity of the laws fell between 2009 and 2013, then improved in 2019, but only back to the 2009 level. Prosecutors’ perceptions fell between 2009 and 2013, then improved in 2019, but are still below the level in 2009. (See Figure 82)

Figure 82: Extent to which Serbian Laws are Clear and Unambiguous, as Expressed by Judges, Lawyers and Prosecutors, 2009, 2013 and 2020251

Chart, bar chart Description automatically generated

  1. Further, professionals expressed reservations about the fairness of Serbia’s laws. Only 13 percent of judges and prosecutors considered the laws to be generally fair and objective, although these perceptions are an improvement in comparison to 2009. Again, most professionals reported somewhere in the middle.
  2. The survey also highlights how imprecise and unclear laws can impact the quality of justice services. Lack of clarity and precision of the laws has a greater impact on the work of less experienced judges. Compared to their older peers, 55 percent of whom raise this issue, 69 percent of judges whose working experience does not exceed five years point out the need for greater precision of the laws. 19 percent of lawyers and 9 percent of prosecutors cited unclear laws as the main reason why the quality of judicial work is not higher.252
  3. Improvement in the law-making process and less frequent changes in legislation could enhance quality. 84 percent of judges stated that less frequent changes in laws could contribute to a better quality of justice services. In addition, 84 percent of judges see a better quality of drafting legislation as a measure that would improve quality.
  4. In interviews, stakeholders noted that overlapping and conflicting laws cause problems for the courts. Several stakeholders highlighted the need for greater harmonization of existing laws, as well as the need to consider existing laws when drafting new ones. Other stakeholders noted that there are gaps in the law, and that judges struggle to deal with these cases in the absence of clear guidance. Stakeholders in prosecution offices highlighted challenges in the application of environmental protection legislation and the use of ambiguous terms regarding wage laws, which causes problems in interpretation.
Environmental pollution is addressed in Article 260 of the Criminal Code. The basic form of the law applies when air, water or land are polluted to a greater extent or in a wider area, but it remains unclear what is considered a larger extent or wider area. The previous practice has interpreted these terms quite extensively and unevenly, which causes additional challenges. A more severe form of crime exists when large-scale destruction of animal or plant life has occurred or it takes a long time or large costs to eliminate the consequences. The meaning of the terms "large-scale", "long time" and "big costs" have to be interpreted in practice. Apart from uneven results, the interpretations are still not sufficient to crystallize these concepts.

Quality of the Law-Making Process ↩︎

  1. Unfortunately, the quality of the law-making process is still problematic in Serbia, despite the adoption of rules for the preparation and adoption of laws. There are several problems that lead to the adoption of laws of low quality. These include very frequent use of urgent procedures for the adoption of laws, which stifles democratic debate and lowers the quality of legislation; lack of transparent and genuine debate; lack of strict rules on the membership in working groups; and transposition of rules from other systems without adequate assessment of conditions and their implementation in Serbia.253 Furthermore, the National Assembly does not exercise its supervisory function, and changes in laws are not based on an assessment of the impact on the practice or pre-existing laws.
  2. Several stakeholders identified poor drafting practices in recent years as contributing to unclear or ambiguous new laws, which have led to uncertainty about the application of laws by the courts. In addition, some changes to legislation were introduced to improve practice, but without assessment of the impact of previous laws and practice. For example, to address the risk of corruption in the public procurement area, a special crime was introduced in the Criminal Code – abuse in the public procurement procedure. However, an insignificant number of cases have been prosecuted under this law because public prosecutors have reported that is more difficult to collect evidence for this crime than for abuse of office.
  3. Organizational methods within working groups and representation in working groups have not always been clear. Stakeholders who are members of various groups expressed frustration that working groups often are not given clear direction about the goals to be achieved by the law and the specific mandate and methods for their work. Some working groups are guided by prior analytic studies, but others simply debate their views. Official working groups do not always include representatives from the populations or entities with the most expertise or those most directly affected by the legislation.
  4. Although there is a requirement to assess the financial implications of proposed laws and institutional capacities to deliver reforms, working groups do not always conduct such analysis in detail. Lack of robust assessment of the financial implications of the 2011 Criminal Procedure law, which entered into force in 2013, led to significant financial arrears in public prosecutor’s offices.254
  5. Although there are have been improvements in the regulation of consultation processes and public debates, there are still shortcomings. Amendments to the Law on Public Administration255 from 2018 brought some improvements in the rules on public debate, such as the possibility of opening a public hearing in the early stages of preparation of an act (article 77), prescription of information that must be published before a public hearing, and the obligation for public consultations during the preparation of laws. The Government’s Rule of Procedure stipulates the obligation to prepare a report on the public debate and publish it on a webpage.256 In research on public debates held in 2019, Transparency Serbia found that state administrative bodies did not act the same way in similar situations and did not comply with the provision of the Law on Public Administration and the Government’s Rules of Procedure.257

The Rollout of New Laws ↩︎

  1. Stakeholders still highlight concerns regarding the successive and continual reforms in the law over the last decade. Legislation is amended often without adequate awareness-raising campaigns among practitioners and users. For example, the Criminal Code was amended 10 times over the last 15 years, which could cause confusion among practitioners and challenges in practice. All that could lead to lack of trust and legal certainty, making it difficult for potential court users to follow all those amendments and to know what the law is.
  2. There should be a greater focus on the dissemination and popularization of new laws, particularly given the pace of the reforms, the limited consultation, and the emergency passage of laws. Awareness of new laws is low among the public, court users, and even among legal professionals (see Access to Justice Chapter and discussion of awareness on Law on Free Legal Aid). Yet, they are the subjects and actors in the new laws, and their understanding is needed for laws to be implemented effectively.

Quality of Administrative Services within the Courts ↩︎

  1. The level of satisfaction with administrative court services is important from the perspective of court users because they directly rely on such services to conduct their everyday business. Administrative services to citizens and businesses comprise 24percent of all administrative tasks within the court. Basic Courts provide administrative services and issue certificates.258 Pending the appointment of notaries for some municipalities, some courts continue to verify signatures, manuscripts, and transcripts, including in probate proceedings.
  2. According to the 2020 Regional Justice Survey, court users assess the overall quality of administrative services to be good259 (see Figure 83). Court users from the general population and the business sector which had to complete administrative tasks related to their court cases were more satisfied with the quality of the administrative services than with the quality of the court work related to their case.

Figure 83: Perceptions of Users of Court Administrative Service of the Quality of Work in that Specific Administrative Case, 2009, 2013 and 2020260

Chart, timeline Description automatically generated

  1. The image of the conduct and competence of service providers is worsening over time. Most users of administrative services are satisfied with the knowledge, efficiency, and pleasantness of staff. However, the number of dissatisfied users has increased over the last seven years. Satisfaction with court administrative services should be compared with satisfaction with the work of notaries to whom part of courts’ administrative competencies were transferred in 2014. Citizens and the business sector are highly satisfied with the quality of notary work; 81 percent of citizens and 97 percent of businesses reported being satisfied with the quality of notary work in their specific case. Such high satisfaction confirms the success of that reform.

Figure 84: Court User Perceptions of Efficiency, Pleasantness, and Knowledge of Administrative Service Staff261

Chart, bar chart Description automatically generated

Quality in Case Processing ↩︎

  1. The quality of case processing has not improved significantly since the 2014 Judicial Functional Review. This section reviews several indicators and European benchmarks relating to the quality of case processing, including standardized forms, consistency in the implementation of laws, use of specialized case processing for particular case types, and coordination in case processing.

Use of Standardized Forms, Templates, and Checklists ↩︎

  1. Consistency in case processing is still undermined by the absence of a consistent approach to routine documentation. There is no uniformity in the online availability of relevant templates that could support users’ communication with the court and court administration.262 There is no common approach, nor have any changes been made by the Appellate Courts, SCC, or HJC.
  2. The RPPO took some measures in the direction of standardization to facilitate the application of the Criminal Procedure Code. The RPPO provided standardized forms and templates in an electronic format aligned with the new CPC in October 2013, but they would benefit from a system-wide update now, after five years of application. Prosecutors have altered some of the RPPO templates themselves already. In addition, the OSCE issued guidelines for different types of the prosecution to support prosecutors and provide interpretation of provisions.263 However, to ensure unified practice, it would be useful to issue a Guide by the RPPO as a mandatory general instruction.

Consistency in the Implementation of Law and Perceptions of the Quality of Judicial Work ↩︎

  1. Inconsistent interpretation of laws and inconsistent jurisprudence remain challenges for the Serbian judiciary. In the 2020 Regional Justice Survey, 70 percent of judges and prosecutors and 90 percent of lawyers stated that inconsistent interpretation of laws and inconsistent jurisprudence happen at least from time to time, if not often. More than 80 percent of lawyers reported that selective implementation of laws and non-enforcement of laws occur frequently. However, only about one-third of judges and prosecutors shared this view (see Figure 85). Judges’ and prosecutors’ perceptions have slightly improved since 2013, but lawyers’ perceptions have worsened over time, especially in the area of selective enforcement.

Figure 85: Share of Judges, Prosecutors, and Lawyers who Estimate that Listed Problems Occur from Time to Time or Frequently in the Enforcement of Laws, 2013 and 2020264

Chart, bar chart Description automatically generated

  1. Despite improvement, lawyers are still mostly dissatisfied with the quality of work of judges.265 By contrast, 73 percent of prosecutors rated the quality of work of judges as high or very high in 2020, compared to 54 percent in 2014, and 67 percent in 2009.266 87 percent of judges rated the quality of judges as high or very high in 2019, compared to 50 percent in 2014, and 61 percent of judges in 2009.
  2. The evaluation of improved quality over time is substantially higher among judges and prosecutors (53percent of judges and 48percent of prosecutors), than among lawyers (only 17percent). 42percent of lawyers actually report that the quality has worsened over time, compared to only 7percent of judges and 15percent of prosecutors (see Figure 86 below). Lawyers’ opinions are influenced by personal experience with the shortcomings of the existing system, such as the lack of information and communication technology systems, the absence of a comprehensive countrywide system to process and link cases across courts and prosecutorial networks, and limits to some databases, which are available only to judges and prosecutors.

Figure 86: Quality of work over time267

Chart, bar chart Description automatically generated

Use of Specialized Case Processing for Particular Case Types ↩︎

  1. There are few examples of specialization in case processing in the Serbian judiciary. Commercial Courts have specialized their case processing somewhat. Misdemeanor Courts are a type of specialized court, but within their jurisdiction is a broad range of cases, from customs and tax offenses to traffic infringements, yet few mechanisms exist to tailor case processing to these very different types of cases. The Administrative Court has similar challenges, with a broad range of cases ranging from competition cases to cases related to election legislation.
  2. Lack of specialization prevents prosecutors from developing special competencies and thus resolving cases with greater success. In addition to specialized PPOs and four specialized departments for corruption cases, only the larger PPOs have established specialized departments. The First Belgrade PPO has departments for commercial offenses and domestic violence, and the Belgrade Higher PPO has a department for combating high-tech crime. On the other hand, there is a specialization of case processing for juvenile cases in courts and prosecutor offices as required by law268.

Coordination in Case Processing ↩︎

  1. Coordination in case processing still presents challenges for the Serbian judiciary. Overlapping criminal and misdemeanor offenses still exist in the Serbian legal system. Elements of specific offenses can be charged as both criminal and misdemeanor offenses - or as both criminal and commercial offenses.269 In Serbia, police often submit both misdemeanor and criminal charges for the same incident and do not inform the prosecutor of the duplication.
  2. Overlapping offenses also cause inefficiency within the court system. The same incident burdens both the prosecution and the courts - once for the misdemeanor offense, with its procedures and legal remedies, and again for the criminal offense, with its procedures and legal remedies.
  3. There are examples of the roll-out of good practices in coordination of case processing across all courts. A positive experience from inter-sectoral coordination in family violence cases from Zrenjanin has been incorporated in legislation. The law on the prevention of family violence270 envisages the establishment of group for coordination and cooperation (Article 25) that consists of representatives of public prosecutors, police, the center for social work, and representatives of other institutions (educational, employment services, etc.) if needed. The group is obliged to meet once every two weeks.
  4. The great majority of prosecutors strongly believe that cooperation with other investigative bodies contributes to the quality of their institution, while judges and lawyers have more moderate opinions on this issue. It seems that the view of prosecutors is more accurate, and it is recommended that Serbia’s political leaders implement an effective, no-tolerance policy for the unwillingness of police to follow prosecutors’ instructions during all investigative phases of a case.271 Otherwise, it will be impossible to produce consistent improvements in the quality and timing of case resolutions or increase public confidence in the judicial system.
  5. Lack of political will, accompanied by the lack of official guidelines, generally impedes the effective investigation of cases.272 On the other hand, there are some positive trends of cooperation that improve the quality of work. For example, a cooperation agreement between Eurojust and Serbia entered into force in December 2019, and in 2020 Serbia took part in three joint investigation teams.273 Also, the cooperation between the War Crimes Prosecutor's Office and the War Crime Investigation Service has been improved by forming joint investigation teams and introducing a new methodology.274 These positive examples of good cooperation resulted in greater optimism among prosecutors who took part in the Regional Justice Survey.

Quality of Decision-Making in Cases ↩︎

Use of Standardized Judgment Writing Tools ↩︎

  1. Although there is no template or a common approach to judgment writing, some initiatives have been undertaken by the Supreme Court of Cassation and professional associations. The Supreme Court of Cassation has organized round tables to discuss criminal judgments and to identify shortcomings and good practices in judgment writing. The Judicial Academy, in cooperation with the USAID ROL project, has organized training for judicial assistants on judgment writing technics.
  2. Judicial training, both initial and continuous, includes a judgment writing module. A judgment-writing component was included in the Judicial Academy’s continuing training program for 2014, but the training is general and does not teach a standardized approach. As part of the initial training at the Judicial Academy, trainees receive compulsory training on the writing of various types of judgments and other court decisions in civil, non-litigious, enforcement, and criminal cases; in their final evaluation as trainees, they are evaluated on judgment-writing skills by their mentor judges.275

Consistency of Decision-Making with the ECHR ↩︎

  1. The statistics of the ECtHR in Strasbourg suggest that the Serbian justice system is still struggling to comply fully with the standards of the ECHR.276 Between 2017 and 2020, there has been an increasing number of cases where Serbia has been found in violation of the standards for non-enforcement and length of proceedings. Compared to 2010-2013, the number of violations related to the length of proceedings significantly increased, from 10 percent to 28 percent of overall violations, while violations concerning failures to enforce final court decisions and non-enforcement remain high.277 Out of a total of 63 judgments in 2017-2020 in which the ECtHR found Serbia in breach of the ECHR, 28 percent of violations related to an excessive length of proceedings and 27 percent of violations concerned failures to enforce final court and administrative decisions. Other violations were found for the right to protection of property and right to a fair trial. Serbia also has been cited for lack of effective investigation, inhuman or degrading treatment, and the right to respect for family and private life.

Figure87: ECtHR Judgments against Serbia by Case Type (2017-2020)278

Chart, pie chart Description automatically generated

  1. There also has been an increase in the overall number of Serbian cases pending before the ECtHR. Serbia is still among the countries with a significant number of pending cases at the ECtHR (2.8 percent of the pending applications at the end of 2020).279 This is only surpassed by far larger countries, such as Russia (22 percent), Turkey (19 percent), and Ukraine (16.8 percent). Almost 97 percent of the Serbian cases heard by the ECtHR have been declared inadmissible or stricken.280
  2. Of applications decided by a judgment, a significant number have found at least one violation of articles of the ECHR (22 out of 24 in 2019; 4 out of 5 in 2020). Among these, it is common for the ECtHR also to find a violation of the length of proceedings and non-enforcement.
  3. There has been a recent noticeable increase in the number of friendly settlements, an effective way in which Serbian authorities can resolve matters without the need for cases to go to hearings.281. In 2017, there were 32 friendly settlements; by 2019, the number of settlements had risen to 103, but this is still significantly lower than 679 friendly settlements in 2013. The negotiation of friendly settlements is likely to be a useful litigation strategy for the state, given that awards for non-pecuniary damages can be quite high. Friendly settlements also are good for applicants because they prevent further delay in resolving their case and receiving compensation.
  4. The Serbian authorities are taking measures, both legislative and non-legislative, to enforce ECtHR judgments, but certain challenges remain. Cooperation among different state authorities is the biggest challenge because enforcement of an ECtHR judgment may include the adoption of legislation and change of court practices and case law to come into line with the rulings of the ECtHR, as well as having budgetary implications.282 Therefore, it is important to establish organized coordination between all relevant state bodies.283

Effectiveness of the Appeal System in Ensuring Quality of Decision-Making ↩︎

  1. The appeal system in Serbia remains one of the judicial system’s impediments, with high appeal rates and deteriorating public perception of trust. The system still provides only unprecise data on lodged appeals, which hinders precise analysis and required the FR team to use estimated figures. Rates varied noticeably across court types, case types, and court locations. High appeal rates prolong the overall duration of cases and increase caseloads. On a more positive note, the reversal rates have declined and have been partially substituted by increased amendments, most likely due to the legislative obligation of the appellate court to decide on its own on the second appeal.
  2. Ambiguity in laws and lack of uniformity in their application may contribute to high rates of appeals and reversals. Ambiguity may cause lower-court judges to make reversible errors, while lack of consistency in lower courts may encourage parties to hope for a more favorable result on appeal. Other factors also may have encouraged parties to lodge appeals, such as the attorneys’ interest in charging for more actions taken in a case and/or dilatory tactics to postpone enforcement in adverse decisions.

Using Data on Appeals to Evaluate the Quality of Judgments and of the Appeals System ↩︎

  1. Due to the lingering lack of more appropriate data, this FR, like the one from 2014, relies on an estimate of lodged appeals and appeal rates.284 That is, present-day reports still do not provide information on lodged appeals but only on decided appeals, which does not necessarily equate to appealed lower instance decisions made in the same reporting period. Also, as found in the FR 2014, Appellate Court statistics still do not distinguish between cases received from Basic Courts and cases received from Higher Courts. Instead, cases deriving from both Basic and Higher Courts are entered into the same registries.285
  2. To calculate appeal rates286, the FR team used the number of resolved appeals adjusted by clearance rates of higher instance courts. Since the clearance rates of all higher instance courts examined here were close to 100 percent, the number of resolved appeals should be reasonably similar to the number of lodged appeals.287 This calculation is rather straightforward for all court types except for Basic Courts, for which the team needed to include an additional estimate to distinguish the appeals disposed of by the Higher Courts from the ones disposed of by the Appellate Courts.
  3. Serbian data on resolved appeals lacks one more dimension – dismissed cases. The SCC’s reports disaggregate resolved appeals by the following categories; confirmed, remanded288, amended, and partially amended or remanded. Dismissed appeals are left out, although they should be reported as a separate category. Therefore, the FR team could not include dismissals in its estimates. If dismissals were included, the appeal rates would have been somewhat higher than estimated.
  4. Confirmation and reversal rates, without the appeal rate, do not mean much individually, but they mean a lot combined. High appeal and high confirmation rates in combination indicate stalling or other abusive tactics by parties. High appeal and low confirmation rates indicate quality and case law harmonization problems. The ideal situation would be a low appeal rate and a 50 percent confirmation rate, suggesting that only cases where even the judge may be uncertain of the right outcome go to higher instances.
  5. Appeals are crucial not only as an indication of quality in decision-making but also as a factor in efficiency and timeliness. High appeal rates prolong the overall duration of cases and increase caseloads. Reversal causes a case re-opening in the lower instance court, after which the same case probably will be appealed again. This could happen several times in a single legal matter. However, procedural reforms have removed some of these procedural loopholes. For instance, appellate tiers are required to substitute the reversed decisions by their own judgments on the second appeal.

Appeals by Court Type and Case Type ↩︎

  1. There are four court types with appellate jurisdiction in Serbia: the Higher Courts, the Appellate Courts, the Appellate Commercial Court, and the Appellate Misdemeanor Court, in addition, appeals may be made to the Supreme Court of Cassation (see Figure 88, below). The ensuing analysis tracks appeals by court type and offers views on the current state of the appellate system in Serbia.

Figure 88: Court Appellate Jurisdiction in Serbia

Diagram Description automatically generated

Source: WB Team Illustration

Appeals from Basic Court Decisions

  1. As a rule, appeals from Basic Courts go directly to the Appellate Courts for review (so-called ‘big appellation’), but some, usually concerning simple matters, are reviewed by Higher Courts (so-called ‘small appellation’).289 According to the data collected by the SCC for this FR, on a national level, appeals against Basic Court decisions comprised 62 percent of the Appellate Courts’ caseload in 2019, and 66 percent in the first half of 2020. Table 14 below displays ratios of received cases from Basic and Higher Courts per Appellate Court in 2019.

Table 14: Received Cases in Appellate Courts from Basic and Higher Courts in 2019

Appellate Court From Basic Courts percent From Higher Courts percent Total in Appellate Court(s)
Belgrade 11,010 54percent 9,558 46percent 20,568
Kragujevac 8,764 66percent 4,541 34percent 13,305
Nis 8,313 67percent 4,092 33percent 12,405
Novi Sad 9,772 68percent 4,704 32percent 14,476
TOTAL 37,859 62percent 22,895 38percent 60,754

Source: SCC Data and WB Calculations

  1. The cumulative appeal rate of Basic Courts in 2019 was nine percent – one percent higher than in 2013. This appeal rate was calculated by dividing 105,464 resolved appeals in Higher and Appellate Courts by 1,110,393 resolved cases in Basic Courts. 72.40 percent of the decisions or 76,356 were confirmed, 13.64 percent or 14,381 decisions were remanded, 8.18 percent or 8.622 decisions were amended, and 5.79 percent or 6,105 decisions were partially amended or remanded.
  2. Appeals against Basic Court decisions in civil litigious cases290 in 2019 were high, with an appeal rate of 30 percent or approximately seven percentage points more than in FR 2014.291 Almost three-quarters of the appeals pertain to general civil litigation (‘P’ registry) where the appeal rate in 2019 was estimated to be around 33 percent and the confirmation rate 72.64 percent. Labor civil litigious cases (‘P1’ registry) occupied just under one-quarter of civil litigious appeals, with an appeal rate of 43 percent and a confirmation rate of 71.72 percent. Family civil litigious cases (‘P2’ registry) comprised three percent of the civil litigation appeals, with an appeal rate of six percent and a confirmation rate of 59.10 percent.
  3. Appeals rates against Basic Court decisions in criminal matters were also high at 24 percent.292 Interestingly, the appeal rate in criminal matters appears to be stable over time, as it was only one percent higher in 2019 than in 2013, as reported by FR 2014. Of appeal decisions made in 2019, 68.79 percent were confirmed, 18.27 percent were remanded, 10.50 percent were amended, and 2.45 percent were partially amended or remanded. There were another two noteworthy categories of criminal cases, so-called criminal panels (‘KV’ registry)293 and parole cases (‘KUO’ cases), with appeal rates of 17 and 10 percent, respectively. In both categories, the confirmation rates were almost 100 percent: 97.78 percent in criminal panels cases and 98.58 percent in parole cases.
  4. Appeals against civil non-litigious cases remained low, under five percent, and varied significantly among case types. This is because non-litigious cases essentially do not involve a dispute between the parties and because not all non-litigious decisions of the Basic Courts can be appealed. On one side of the spectrum, the appeal rate in probate cases in 2019 was 0.3 percent, and the confirmation rate was 99.85 percent. On the other side, a 38 percent appeal rate was reported in cases concerning requests for monetary compensation for immaterial (non-pecuniary) damages due to violation of the right to a trial within a reasonable time, while 94.49 percent of these cases were confirmed.
  5. Appeals against enforcement decisions stayed very low at approximately three percent.294 Out of this percentage, 69.93 percent were confirmed, 12.64 percent were remanded to the lower court, 11.98 percent were amended, and 5.45 percent were partially amended or remanded. Compared to 2014 data, the number of amended and partially amended or remanded decisions increased by multiple times, meaning that the higher instance courts now opt for resolving the case by themselves more, rather than returning the cases for retrial and prolonging their duration.295

Appeals from Higher Court Decisions to the Appellate Court

  1. In Higher Courts, a total of six percent of all decided cases were appealed in 2019, approximately as many as in 2013.296 77.33 percent were confirmed, 11.21 percent remanded, 8.42 percent amended, and 3.03 partially amended or remanded. In comparison to 2013 figures, the confirmations have increased by 11 percentage points, the remands have decreased by 3.5 percentage points, while the amendments and the partial amendments or remands varied only slightly, up to one percentage point.
  2. Appeal rates among major case types in Higher Courts varied significantly, primarily due to the ease of appeal. In the first instance civil litigious cases297, the estimated appeal rate was 20 percent,298 and the confirmation rate was 72 percent. Conversely, in the second instance civil cases299, where there are very few legislative options for appeal, the appeal rate was three percent300, and the confirmation rate was 95.80 percent. By contrast, in criminal cases,301 the parties appealed in 14 percent302 of the decided cases, and the confirmation rate was 61.82 percent.
  3. In other case types in Higher Courts, appeal and confirmation rates varied considerably. The lowest individual appeal rate was 0.1 percent in cases concerning measures to ensure the presence of the accused in the preliminary proceedings. The confirmation rate for the same case type was 100 percent. In criminal panels cases, the appeal rate was 28 percent, while the confirmation rate was 85.86 percent.

Appeals of Commercial Court Decisions

  1. In 2019, the Commercial Courts, aggregate appeal rates and confirmation rates were both moderate. Of the total of 140,082 Commercial Court decisions made, 15,242 were appealed to the Appellate Commercial Court, representing around 11 percent of the Commercial Courts’ decisions for that year. 75.37 percent of the appealed decisions were confirmed, 11.30 percent were remanded to the lower court, 9.24 percent were amended, and 4.09 percent were partially amended or remanded.
  2. The Appellate Commercial Court displayed a greater inclination to substitute the lower court decisions with its own, i.e., the remanded decisions decreased, while the amended and partially amended or remanded decisions increased. In 2013 19.5 percent of the decisions were remanded, which is 8.2 percentage points more than in 2019. Conversely, the amendments increased by 3.44 percentage points and the partial amendments or remand by 3.39 percentage points.
  3. Civil litigious cases in Commercial Courts reported very high appeal rates of 39 percent, while their corresponding confirmation rate was 73.71 percent. Out of 14,483 resolved cases, 5,721 were decided in the Appellate Commercial Court, and 4,217 were confirmed. Similarly, high appeal rates were reported in Commercial Courts in some case types involving bankruptcy proceedings (reorganization plans) and in enforcement proceedings regarding the right to a trial within a reasonable time.

Appeals of Administrative Court Decisions

  1. In the Administrative Court, appeal and remand rates remained low. In 2019, Administrative Court decisions were appealed to the SCC in 1.5 percent of all Administrative Court decisions for that year. This was a reduction by 2.3 percentage points compared to FR 2014. Of the 329 appeals decided by the SCC in 2019, 91.08 percent of the decisions were confirmed. This is almost exactly the same as in 2013, according to the FR 2014 data, when the estimated confirmation rate was 91.11 percent. The latest data confirm the previous FR 2014 finding that there is a higher level of uniformity and consistency in administrative law than in other fields and that a large number of appeals are lodged without merit. However, this analysis is not able to distinguish if the appeal rates were low, and the confirmation rates high because the parties find it hopeless to go against the state in administrative matters.
  2. Among individual case types in the Administrative Court, the appeal rate is high (38 percent) only in cases concerning the right to a trial within a reasonable time. Even in those cases, the confirmation rate is also high, at 91.08 percent.

Appeals of Misdemeanor Court Decisions

  1. In the Misdemeanor Courts, the aggregate appeal rate in 2019 was low, at four percent, while the remand and the amendment rates were fairly high. In 2019, out of 614,246 decided cases, 25,539 decisions of the Misdemeanor Court were appealed to the Appellate Misdemeanor Court. Of these, 58.78 percent were confirmed, 19.50 percent were amended, 21.48 percent were remanded to the lower court, and 0.24 percent were partially amended or remanded.
  2. In comparison to the 2013 data analyzed in FR 2014, the Appellate Misdemeanour Court doubled the number of amendments and reduced the number of remands by roughly one-third. Almost eight percentage points fewer decisions were remanded in 2019 than in 2013 (when 27.73 percent of decisions were remanded to the lower court). At the same time, the percentage of amendments in 2013 (9.66 percent) doubled to about 20 percent in 2019. Other categories are roughly comparable to the 2013 data.
  3. The increase in the amendments and the decrease in remands is an improvement in line with the FR 2014 recommendations. The FR 2014 argued that misdemeanor cases should be relatively straightforward and the Appellate Misdemeanor Court would be well placed to amend the decision and save the parties and the Misdemeanor Courts the necessity of a retrial. The latest data indicate that the Misdemeanor Courts complied successfully with the given recommendation.

Appeals by Location ↩︎

  1. Outcomes of appeals varied among Serbian Basic Courts without any clear pattern. The average reported confirmation rate was 68 percent. The Basic Court in Subotica reported the highest confirmation rate in 2019 of 88.87 percent, while the lowest rate, of 34.18 percent, was reported in the Basic Court in Vrsac.303 Simultaneously, the Basic Court in Vrsac also reported an unusually high percentage of amended decisions – 57.02, primarily due to a very high number of amendments of civil litigious cases registered under ‘P’. Amendments were also high in Basic Courts in Sremska Mitrovica and Backa Palanka, at 20.63 percent and 19.86 percent, respectively. By contrast, the highest remand rate of 27.49 percent was reported in the Basic Court in Bor, followed by the Basic Court in Velika Plana, Trstenik, and Senta, in which over one-quarter of decisions were remanded.

Figure89: Appeals Outcomes in Selected304 Basic Courts in 2019

Source: SCC Data

  1. Appeal outcomes varied also among Higher Courts but to a lesser extent than in Basic Courts. The majority of Higher Courts remained close to the average confirmation rate of 75 percent. The only two true outliers were the Higher Court in Kraljevo, with a confirmation rate of 34.64 percent, and the Higher Court in Prokuplje, with a confirmation rate of 41.24 percent. In Kraljevo, the low confirmation rate was caused directly by 59.75 percent of amendments of 406 labor civil litigious cases, most probably identical or very similar disputes that could have been resolved uniformly. In Prokuplje, 36.60 percent of the decisions were remanded due to 47.90 percent or 57 remanded civil litigious decisions.

Figure 90:Appeals Outcomes in Higher Courts in 2019

Source: SCC Data

User Perceptions of Appeals ↩︎

  1. Trust in the appellate system among court users in Serbia decreased from 2013 to 2020, and fell below 2009 values, as demonstrated in Figure 91. In 2020, under one-half (41 percent) of the citizens with recent experience in court cases stated that they trust the appellate system. Meanwhile, a slightly higher percentage (47 percent) of business sector representatives with court case experience stated that they trust the appellate system. Interestingly, both the number of people who stated that they trust the appellate system and the number of people who responded that they do not trust the appellate system decreased in comparison to both 2009 and 2013. The number of indecisive respondents grew by seven to twenty times. It is unclear whether this lack of trust in combination with indecisiveness, encourages or discourages court users from lodging appeals.

Figure 91: Perceptions of Trust in the Appellate System, as Reported by Court Users, 2009, 2013, and 2020305

Chart, bar chart Description automatically generated

  1. The decision of a party to file an appeal remains strongly related to the party’s perception of the fairness of the first-instance trial, even more so in 2020 than in 2013. Court users who received a judgment that was not in their favor filed an appeal in 84 percent of the cases if they considered the decision to be not fully fair, an increase by 21 percentage points over 2013. In contrast, court users who received a judgment that was not in their favor but who considered the decision to be fair appealed in only 10 percent of cases, an increase by two percentage points over 2013.

Figure 92: Relationship between Perceived Fairness and Decision to Lodge Appeal among Court Users who Received a Judgment Not in their Favor, 2013, 2020306

Chart, bar chart Description automatically generated

  1. In 2020, in over fifty percent of the cases, court users reported that the higher instance court upheld the judgment. However, in 20 percent of cases involving the public, the judgment was overturned and a retrial was ordered. In 17 percent of cases involving the business sector, the judgment was overturned and a retrial was ordered. Simultaneously, the number of amendments increased from 2013 to 2020 in both citizens’ and businesses’ cases by six and four percentage points, respectively.

Figure 93: Outcome of Appeals as Reported by Court Users, 2009 and 2013307

Chart Description automatically generated

High Appeal Rates and High Variation in Appeals ↩︎

  1. High appeal rates in Serbia, particularly in specific case types, and deteriorated perception of the appeal system suggest the systemic presence of quality-related difficulties. Lack of uniformity in the application of law may encourage parties to hope for a more favorable result on appeal. Furthermore, it has been frequently reported that attorneys also may play a role in driving up appeals, since their expenses are predominantly calculated per each action they take in a case. They also may instruct their clients about the likelihood of success on appeal and the tactical advantages appeals may offer, such as the delayed enforcement of an adverse judgment.
  2. Reasons for geographic variations remain inexplicable in any way other than lack of uniformity in the application of the law. Although one-time effects of specific circumstances in courts, such as remands of many uniform labor cases, may cause sudden variations, it is not likely that this could be the case in all of Serbia, especially since the variations are persistent in 2013 (2014 Functional Review findings) as well.
  3. The FR 2014 found the appeals system is at the heart of Serbia’s problems in terms of quality of decision-making. Appeal rates were found very high on average, as were reversal rates. Rates also varied markedly across court types, case types, and court locations. Appeals were poorly monitored. The perceived unfairness of the system, combined with its lack of uniformity and consistency, encouraged court users to appeal. Attorney incentives were also identified as one of the factors driving up appeals. At the same time, levels of trust in the appellate system among court users were low. Procedural amendments to reduce successive appeals (known as the ‘recycling’ of cases) were found effective. Nonetheless, appellate judges (notwithstanding their lighter caseloads) continued to remand cases back to the lower jurisdiction for retrial more often than they were required to.
  4. On a more positive note, the higher instance courts more often replaced the lower instance decisions with their own, as supported by data earlier in this section. In FR 2014, only a small percentage of cases were higher instance courts amending the decisions of lower courts, although the benefits of such amendments are numerous. They save the parties the trouble of re-litigating in the lower instance court, ease the workloads of judges and courts, shorten the trial, and increase uniformity in the application of the law over time. Legislative changes obliged the higher instance court to replace the decisions with their own on the second appeal. Higher instance judges should work toward, whenever possible, replacing the decision of the lower court in instances other than a second appeal in the same matter. Higher instance decisions in which a reversal is issued should contain precise reasoning and instructions to be followed by the lower court in subsequent proceedings.
  5. A range of other measures are available to improve the quality of decision-making. Some of them were already suggested in the FR 2014. These comprise education of judges, better use of existing case law harmonization tools, and implementation of new ones (e.g., meetings of judges in the same department to discuss legal issues).

Quality of prosecution ↩︎

Introduction

  1. This chapter builds on the analysis of Serbia’s prosecutorial system (Prosecutorial FR) by the World Bank and the Multi-Donor Trust Fund for Justice Support in Serbia308, by examining data from 2017 through 2019. The Governance and Management Chapter of this report covers those functions for the prosecution as well as for the courts, but readers should consult the Prosecutorial FR for more details about the structure and hierarchical nature of the prosecutorial system overall. The Prosecutorial FR also covered the functions of the SPC, the RRPO, and the different types and jurisdictions of PPOs.
  2. The Prosecutorial FR, formally published in January 2019, focused primarily on 2014 through 2016, when Serbia’s prosecutors were adjusting to extensive changes in the nation’s Criminal Procedure Code (CPC), adopted in 2013. These changes included the introduction of adversarial proceedings, which challenged many prosecutors as they adjusted to their new, more active roles.
  3. At the same time, leaders of Serbia’s political and judicial systems were under continuous pressure to make major additional structural changes to the governance and management of the country’s prosecutorial functions as part of Serbia’s planned accession to the EU. That pressure has continued to the publication of the present FR, as discussed in the various EU reports related to the EU’s Enlargement Policy.309
  4. The official data on which many of the statistics in this Chapter rely were not always consistent and could not always be reconciled. Data in this study came from statistics in RPPO Annual Reports from 2014 through 2019 and from the SPC, but other data was derived from interviews and published analytical reports such as those produced by CEPEJ.
  5. The available data for prosecutorial services still was far less extensive than it was for courts, and the data that was reported was of limited use because of the collection methods and formats. There was no unified electronic case management system for the prosecutorial system in place by the end of 2019. The available RPPO Annual Reports were published in a format310 that was not suitable for computer processing. Preparation of those reports depended highly on manual data collection and individual interpretation, which made the reports prone to inconsistencies and inaccuracies.

Quality in Case Processing ↩︎

  1. There were no significant advancements in modernizing performance measuring for prosecutors or PPOs. Prosecutors still lacked support in measuring their performance and how to use this information to their advantage to improve case management, support funding requests, foster public support, and respond to criticism clearly and precisely. The prosecution system would undoubtedly benefit from such modernization.
  2. Regular system-wide updates of the standardized forms and templates provided by the RPPO in 2013 were still required, especially since there had been several amendments to the CPC since 2013. Individual prosecutors reportedly altered some of the RPPO templates for their own use, but there was no centralized revision of the official forms. As noted in the Prosecutorial FR, the use of templates and standardized forms facilitates a consistent approach to routine prosecutorial tasks, reduces the number of mistakes in documents, and fast-tracks regular daily actions.

Conviction rates ↩︎

  1. As important as they are for assessing the quality of prosecution, conviction rates alone do not provide a complete picture of how well any prosecutorial system has performed.311 In addition to conviction rates, factors such as the timing and reasons for dismissals and deferred prosecutions can be used to examine both the quality of decision-making and the skills of professionals within a judicial system.
  2. The types of cases included in Serbia’s conviction rate statistics did not change between the publication of the Prosecutorial Functional Review and 2019, so they included only cases in which a court entered a decision of guilty or innocent. As a result, the statistics about convictions included cases concluded through plea bargains312 but did not include deferred prosecutions, which involve the dismissal of charges by a prosecutor without a court finding of guilt or innocence, while imposing a sanction on the defendant.
  3. The overall conviction rates of Basic and Higher PPOs held steady or improved during 2018 and 2019, compared to the rates for 2015-2017. The average for Basic Courts was 91 percent for 2018 and 2019, compared to 90 percent for 2015-2017, while the average for Higher Courts increased from 86 percent in 2015 and 2016 to 89 percent in 2017 and 2018, and 91 percent in 2019.
  4. However, there were wide variations in conviction rates among even PPOs of the same size and jurisdictional levels and sometimes by year within the same PPO from 2015 through 2019. See Figure 94 and Figure 95 below. The FR team was not able to obtain any analyses of the reasons for the variations that may have been completed by the SPC or the RRPO. The Basic PPO in Petrovac on Mlava had a 100 percent conviction rate from 2015 to 2017 and a 99 percent conviction rate in 2018 and 2019. Similarly, high conviction rates throughout the period were reported in Basic PPOs in Jagodina and Smederevo. The Basic PPOs in Belgrade had rates that were roughly at the national average. Conversely, similarly-sized Basic PPOs in Lebane and Mionica reported much lower conviction rates and/or higher variations.313 . The medium-sized Basic PPOs in Vranje had lower conviction rates than the average, ranging from 74 percent in 2015 to 83 percent in 2017 and 2019. The Basic PPO in Nis improved its conviction rate significantly in 2018 to 94 percent, which was approximately 14 percentage points higher than in previous years, but the rate fell again in 2019 to 86 percent. In 2019, the lowest conviction rate (61 percent) and a drop of 27 percentage points compared to the previous year was reported in the Basic PPO in Prokuplje. This sudden drop in Basic PPO in Prokuplje was caused primarily by a 400 percent increase in the number of acquittals.314

Figure 94: Convictions for Selected315 Basic PPO from 2015 to 2019

Chart, bar chart Description automatically generated

Source: RPPO Annual Reports 2015-2019

  1. The conviction rates of Higher PPOs also had similar variations, but the FR team could not locate any analyses of the possible causes of the variations. For instance, the conviction rate for the Higher PPO in Vranje dropped from 84 percent in 2017 to 58 percent in 2018. The rate for the Higher PPO in Belgrade increased each year by a total of 18 percentage points from 2015 to 2019. The three Higher PPOs with the lowest rates in 2019 were Vranje (72 percent), Jagodina (81 percent), and Kraljevo (82 percent). In contrast, the rate for the Higher PPO in Leskovac rose from 64 to 97 percent in 2019. Only two Higher PPOs, in Novi Pazar and Uzice, managed to maintain conviction rates of 90 percent or higher over the five observed years(there was no data available for Uzice for 2017).

Figure 95: Convictions per Higher PPO from 2015 to 2019316

Table Description automatically generated

Source: RPPO Annual Reports 2015-2019

  1. Conviction rates of Specialized PPOs and specialized departments in select Higher PPOs also varied significantly. The Special Prosecutor’s Office for Organized Crime reported conviction rates from 75 to 91 percent from 2015 to 2019317 while conviction rates of the Special Prosecutor’s Office for War Crimes ranged from zero to 83 percent.318 Conviction rates for the Special Prosecution Office for High Tech Crime within the Belgrade Higher PPO improved from 37 percent in 2015 to 91 percent in 2019.319 The specialized departments to combat corruption, established in 2018 in four Higher PPOs, produced a 100 percent conviction rate that year and a rate of 96 percent in 2019.320

Quality in Decision-Making ↩︎

Control Mechanisms and Coordination in Case Processing ↩︎

  1. There were no changes to the highly hierarchical structure of Serbia’s prosecutorial system between the publication of the Prosecutorial FR and the end of 2019. Higher-instance Public Prosecutors still had the right to control the work of lower-instance ones; the higher-instance prosecutors could take over any matters of lower-instance Public Prosecutors within his or her jurisdiction and issue mandatory instructions to those lower-instance Public Prosecutors.
  2. By the end of 2019, there still were no effective means prosecutors could use to force police to follow their instructions.. Prosecutors interviewed for the FR reported it still was common for police to ignore or to vary from prosecutorial instructions about steps to be taken during the investigations. Prosecutors reported this problem arose particularly in cases that might have political implications because of political issues or the political roles of persons involved in a case, which also was true when the Prosecutorial FR was published in 2018.
  3. To ensure better quality and control of prosecutors’ work, starting in 2013, the CPC has allowed the filing of complaints about the dismissal, suspension, or abolition of a criminal complaint, and Serbians have made extensive use of this process.321 An alleged victim or the person who submitted a criminal complaint may request that a higher-instance PPO reconsider a dismissal.
  4. The complaints mechanism was applied in eight percent of the dismissals in Basic PPOs (4,749 complaints) and 43 percent in the Higher PPOs (2,122 complaints) in 2019. In the absence of any data to explain the difference of 35 percentage points, the FR team presumes persons affected by the more serious crimes handled by Higher PPOs were more apt to feel they had significant interests in the outcome of those cases.
  5. In 2019, 12 percent of the 4,749 complaints against Basic PPO dismissals were adopted by the Higher PPOs while 75 percent were rejected, and 12 percent were carried over to 2020. Since the time limits for this type of complaint against any dismissal are quite strict, presumably, the carried-over cases were those received at the end of the year.
  6. In 2019, eight percent of the 2,122 complaints against Higher PPO dismissals or 166 were adopted by the applicable Appellate PPOs, while 1,739 were rejected, and 217 were carried over.
  7. In both Basic and Higher PPOs, the figures related to injured person complaints varied to some extent from 2014 through 2019 but without a distinct pattern. Approximately 3,500 to 4,500 complaints were submitted against Basic PPOs dismissals, and 1,000 to 2,500 against Higher PPOs dismissals. Roughly 75 to 85 percent of complaints against both Higher and Basic PPOs were not accepted, with the exception of 2017, when 93 percent of the complaints submitted against Higher PPOs were rejected.

Dismissals ↩︎

  1. Due to the continuing lack of detail collected about the reasons for dismissals, the system was still missing a significant amount of critical data about the quality of this process. Dismissals in the RRPO’s Annual Reports were broken down into only five categories, namely ‘insignificant offenses’,322 cases dismissed for lack of evidence, deferred prosecutions, unfinished deferred prosecutions and “other”. It was not even possible to determine, for instance, how many of the “other” cases had to be dismissed against adult defendants because the applicable statute of limitations had expired. In 2019, 40 percent of the dismissals of criminal complaints against adult defendants were handled by Basic PPO, as shown by, and Figure 96, 87 percent of the same set of cases handled by Higher PPOs fell in the “other” category.

Figure 96: Dismissals in Basic PPOs by Type in 2019323

Chart, pie chart Description automatically generated

Source: RPPO Annual Report 2019

  1. From 2014 to 2019, Basic PPOs dismissed a higher share of cases than Higher PPOs due to the number of cases for which there was insufficient evidence. In 2019, only five percent of dismissals in Higher PPOs were for lack of evidence, compared to 13 percent of the dismissals of the same type in Basic PPOs. See Figure 97.

Figure 97: Dismissals in Higher PPOs by Type in 2019 (Adult Perpetrators)

Chart, pie chart Description automatically generated

Source: RPPO Annual Report 2019

  1. Comparison of earlier figures to the share of cases shown as “discontinued” in the CEPEJ 2020 report,324 which was based on 2018 data, should be done with caution. Serbia’s statistics show the share of dismissals in the number of disposed of cases ranged from 47 to 58 percent of the cases disposed of for Basic PPOs and 39 to 47 percent for Higher PPOs.325 The CEPEJ 2020 report found that Serbian prosecutors discontinued 3.28 cases per 100 inhabitants, which was higher than the averages for the EU (1.91) and EU11326 (1.10), and the Western Balkans327 (1.33) average. While these numbers could be read to show a tripling of the cases discontinued by Serbian prosecutors compared to the previous CEPEJ report based on 2016 data, between the two reports Serbia changed its definition of what counted as a case against adult defendants processed by a public prosecutor.328 Therefore, the differences from the earlier numbers probably are not as high as they appear in the CEPEJ report. See Figure 98.

Figure 98: Percentage of Dismissed Cases in Basic and Higher PPOs from 2014 to 2019

Chart, line chart Description automatically generated

Source: RPPO Annual Reports 2014-2019

  1. Also in 2019, of the dismissals of juvenile justice complaints handled by Higher PPOs, 44 percent were dismissed because the defendant was too young to be prosecuted for the crime charged and deferred prosecution was applied in 23 percent of the dismissals. The bases for the remaining 33 percent of dismissals were not specified; for instance, dismissal statistics for juveniles did not contain separate categories for lack of evidence or insignificant offenses.

Deferred Prosecution ↩︎ 329

  1. According to the CEPEJ 2020 report (2018 data), with a score of 0.27 Serbia reduced the number of cases ‘concluded by a penalty or a measure imposed or negotiated by the prosecutor’, including deferred prosecutions, for the second time in a row. As opposed to the previous two evaluation cycles, when the reported ratio was one of the highest among covered countries,330 the latest figures are more modest and well under the CoE average of 0.45. However, they still were more than double than the EU11 average of 0.10. In the CEPEJ 2016 report (2014 data), Serbia’s reported ratio was 0.53, and in the 2018 report (2016 data) 0.36. Many of these differences probably were affected by Serbia’s decision to expand its definition of cases against adult defendants processed by prosecutors between 2016 and 2018.
  2. Compared to 2018, in 2019 Basic PPOs increased the number of deferred prosecutions by 6 percent, while the Higher PPOs reduced the number of deferred prosecutions by 74 percent.331 The reasons behind the decline among the Higher PPOs were not documented; presumably, at least part of the decline was due to concerns about possible overuse of the procedure, as discussed at page 41 of the Prosecutorial FR.332
  3. Deferred prosecution was applied in 707 juvenile criminal cases in Higher PPOs in 2019, which represented 23 percent of the dismissals of juvenile criminal cases. In preceding years, deferred prosecution attained a similar share in dismissals of juvenile criminal complaints. There were 619 deferred prosecutions (22 percent) in 2014, 689 (26 percent) in 2015, 875 (28 percent) in 2016, 834 (24 percent) in 2017, and 468 (14 percent) in 2018.
  4. Some deferred prosecutions were classified as ‘unfinished’ because the defendant still had time to meet his or her obligations. Although there still were no statistics available by the end of 2019 for the number of cases in which the deferred prosecution was revoked because the conditions of deferral had not been met, the Prosecutorial FR estimated defendants had failed to comply with the terms of the deferred prosecution in up to 10 percent of the deferred prosecution cases.

Table 15: Deferred Prosecution per PPO Type from 2014 to 2019 (Adult Perpetrators)

    2014 2015 2016 2017 2018 2019
Basic PPOs Deferred Prosecution 17,447 21,074 20,083 16,706 17,802 18,858
Deferred Prosecution (Unfinished) 15,706 14,216 9,011 8,787 7,874 7,320
Higher PPOs Deferred Prosecution 132 159 161 229 371 97
Deferred Prosecution (Unfinished) 48 er173 46 66 309 50
TOTAL 33,333 35,622 29,301 25,788 26,356 26,325

Source: RPPO Annual Reports 2014 – 2019

  1. Deferred prosecution sanctions that could benefit the community at large, such as rehabilitation programs or community service, were still of only limited availability by the end of 2019, since the legislative or regulatory measures needed to implement those types of programs still were not in place. There were no official figures available about the number of defendants participating in rehabilitation programs or community service. The most frequently reported sanction imposed on deferred prosecution defendants remained a cash donation to humanitarian causes, a sanction that can be relatively easily monitored. As the Prosecutorial FR noted, the widespread use of cash donations can give the impression that defendants have bought their way out of the criminal justice system.
  2. By the end of 2019, Serbia also had failed to take any steps to resolve other issues about deferred prosecutions noted in the Prosecutorial FR. These issues included concerns that (1) implementing legislation was incomplete and imprecise; (2) there was a lack of clear guidelines or criteria for the use of deferred prosecution and inconsistent use of it by prosecutors, and (3) there was a lack of consideration for the interests of the victims of the crimes involved, as prosecutors decided whether deferred prosecution should be offered and/or what its terms should be.

Plea Bargaining ↩︎ 333

The use of plea agreements needs to be carefully balanced against the need for justice being seen to be done and avoiding any impression of impunity in cases of serious and organized crime.

Source: EU 2020 Report334

  1. The most common types of plea bargains in 2019 resulted in suspended sentences (about one-half) and imprisonment (about two-fifths). Other measures such as fines, precautions (e.g., restraining orders and home detentions), and community service made up the rest. The RPPO Annual Report for 2019 indicated the sanctions awarded totaled more than 3,208 years of imprisonment and almost RSD 85 million335 in fines.
  2. After five years of consistent increases, the number of cases concluded by plea bargaining in Basic and Higher Courts decreased by eight percent in 2019, due to a 17 percent drop in the Belgrade appellate region. However, even with the 2019 decrease, there was a 190 percent increase in the number of plea bargains throughout Serbia in 2019 compared to 2014. Of the 5,971336 plea bargains proposed to courts in 2019, courts accepted 90 percent (5,363) were accepted. This percentage was very close to the percentages accepted in 2017 and 2018, and nine percent higher than the number of proposals accepted in 2014. Of the other 608 proposed plea bargains in 2019, the RPPO reported that more than 169 were not accepted by the courts, while roughly 783 requests were still pending in the courts. The RRPO Annual Report also indicated that defendants initiated around one-half of the proposed agreements, and prosecutors were more likely to initiate an agreement during the pre-indictment period than later in the case; of 4,612 plea bargains initiated before indictment, prosecutors initiated 2,680 of them.
  3. The highest combined number of plea bargains for Basic and Higher PPOs in 2019 occurred in the jurisdiction of the Appellate PPO in Belgrade, with a total of 2,490. This was 204 percent more than the 820 plea bargains for the same PPOs in 2014, and, as mentioned, 17 percent fewer than the 2,994 plea bargains in the previous year.

Effectiveness of Appeals in Promoting Quality ↩︎

  1. The decision to appeal for every case lies with the prosecutor handling a case and the higher-instance PPO. Prosecutors may appeal when the court has acquitted the defendant or when the prosecutor is not satisfied with the defendant’s sentence.
  2. On average, Serbian Basic PPOs appealed in 12 percent of cases in 2019 and were successful in only 21 percent of their appeals,337 indicating prosecutors should have realized at the outset that many of the appeals would not succeed.338 In the preceding year, Basic PPOs appealed at the same rate of 12 percent of the cases but had a slighter higher success rate of 27 percent.339 However, since the resolution of an appeal often occurs one or more years after the appeal is filed, the success rates reported in a certain year generally covered appeals filed at least a year earlier.
  3. Appeal rates varied considerably among Basic PPOs, including those of similar size. The lowest appeal rate of two percent was recorded in Basic PPOs in Pancevo, Ruma, Senta, Pozega and Subotica, while the highest appeal rates appeared in Basic PPO in Kragujevac (41 percent), Mionica (35 percent), Leskovac (33 percent), and Lazarevac and Arandjelovac (both 31 percent). The success rates varied even more, from zero to 72 percent. Zero percent was recorded in Basic PPOs in Kraljevo, Bor, Backa Palanka, and Novi Pazar, while 72 percent was recorded in Subotica. The wide variation in appeal rates for 2019 is displayed in Figure 99 below.

Figure 99: Ratio of Submitted Appeals and Successful Appeals in Selected340 Basic PPOs in 2019

Chart, bar chart Description automatically generated

Source: RPPO Annual Reports 2019

  1. There were no clear correlations between the Basic PPO appeal rate and success rates throughout the observed period. For example, in 2019, the Basic PPO in Mionica reported an appeal rate of 35 percent and succeeded in only two per cent of its appeals. In contrast, for the PPO in Leskovac, the appeal rate for 2019 was 33 percent and 23 percent of its appeals were successful. (As noted above, most if not all of the cases reported in the success rates probably were cases in which the appeals had been filed at least a year earlier.)
  2. Over time, appeal rates for most Basic PPOs were essentially stable, even when success rates varied widely. For instance, the Basic PPO in Sabac had an appeal rate of 23 percent in 2015 and 2016, 20 percent in 2017, 21 percent in 2018, and 22 percent in 2019, while its success rates ranged from three to 28 percent. There was a similar pattern for the Basic PPOs of Kragujevac and Nis.
  3. Compared to Basic PPOs, Higher PPOs filed appeals more frequently and succeeded in more cases. The average appeal rate of Higher PPOs was 49 percent in 2015, 50 percent in 2016, 39 percent in 2017, 36 percent in 2018, and 41 percent in 2019. The success rates were 30 percent, 26 percent, 30 percent, 29 percent, and 25 percent, respectively.
  4. However, very few Higher PPOs had consistently high or low appeal or success rates from 2015 to 2019. The Higher PPO in Belgrade has some of the most consistent showings; it appeal rate varied between 64 and 82 percent, while the success rate ranged between 12 and 18 percent. Similarly, in Zrenjanin, the appeal rates were between 59 and 82 percent, while the success rates varied between four and 16 percent. In contrast, the Higher PPO prosecution office in Prokuplje had appeal rates from 117 to 164 percent341 and success rates from 44 to 95 percent.
  5. The appeal and success rates for specialized PPOs were in line with the rates for other Higher PPO offices. The Special Prosecutor’s Office for Organized Crime appealed in 63 percent of the cases in 2019 with a success rate of 16 percent, and the Special Prosecutor’s Office for War Crimes appealed in 67 percent of the cases in 2019 with a success rate of 25 percent. The Special Prosecution Office for High Tech Crime within the Belgrade Higher PPO appealed in 33 percent of the cases and succeeded in 28 percent. During 2018, which was their first year of operation, the four anti-corruption Higher PPO specialized departments had an average appeal rate of 13 percent and their reported average success rate was nine percent (note, however, that the successful cases probably were ones that started by other offices, because the anti-corruption departments opened in 2018). In 2019 the average rate of appeal for the four PPOs was seven percent of their concluded cases and the average success rate was 27 percent.

Recommendations and Next Steps ↩︎

The 2014 Functional Review provided seven detailed recommendations and next steps for improving the quality of court services in Serbia. Although some recommendations have been implemented over time, with more or less success, some are still unattended.

Recommendation 1: Improve the clarity and consistency of legislation.

  • Develop consistent standards for representation of stakeholders in working groups considering new legislation. Provide guidance as to the tasks expected of such groups. Ensure that they have access to factual and analytical resources, including information on existing laws and relevant statistics. (MOJ, HJC, SPC – short-term)
  • Encourage legislative groups to think ahead to implementation of new legislation, including which stakeholders might take the lead in implementation, what are the budgetary limitations, etc. (MOJ, HJC, SPC – short-term)
  • Subject all proposed legislation to a review of consistency with existing laws. Develop a procedure to conform to older and new laws. (MOJ, HJC, SPC – short-term)
  • Adopt clear standards and limit circumstances in which emergency procedures are used for enacting legislation. (Parliament, MOJ – medium-term)
  • Conduct a legislative review to determine whether certain offenses should be uniformly charged as misdemeanors, criminal cases, or commercial offenses, and which should remain subject to the discretion of prosecutors’ offices. (RPPO – medium-term)

Recommendation 2: Improve the consistency of application of laws by courts.

  • -tandardize training in judicial writing (JA, HJC – short-term).
  • Adopt templates for drafting routine documents, such as legal submissions, orders, or judgments. Adopt system-wide procedures for routinely updating all such documents (SCC, HJC, MOJ – short- term).
  • Evaluate judicial quality along with both quantitative metrics (such as remand rates) and qualitative criteria (such as quality of writing in judicial decisions) (SCC, HJC – short-term).
  • Establish regular exchange of data on human resources and quality of decision-making between the SCC and the HJC to inform both bodies in performing duties. (SCC, HJC – medium-term).
  • Harmonize judges’ interpretation of statutes and case law using tools such as departmental meetings, issuing legal opinions, establishing case law departments in higher instance courts, and developing an easily searchable case law database. (SCC – continuous)
  • Regularly monitor and analyze reports, and discuss potential improvements in workshops, meetings, and colloquia. (SCC, court presidents, HJC, MOJ – continuous)

Recommendation 3: Unify and streamline court practices.

  • Adopt checklists and standardized forms for both routine and specialized cases. Adopt system- wide procedures for routinely updating all such documents. (SCC, HJC, MOJ – short-term)
  • Implement a standardized approach to routine aspects of case processing. (SCC, HJC – short-term)

Recommendation 4: Improvement of work of public prosecutors’ offices.

  • Monitor the reasons for dismissals of cases by prosecutors. (RPPO – short-term)
  • Develop uniform standards for the conditions associated with the deferred prosecution (principle of opportunity). (RPPO – short-term)
  • Develop uniform standards for prosecutors’ decisions to dismiss criminal complaints, appeal decisions and impose sanctions, and enter into plea bargains. (RPPO – medium- term)
  • Develop standardized guidelines for the decision of whether to charge an offense as criminal, misdemeanor, or commercial. Require police to inform prosecutors of the nature of charges. (RPPO medium-term)
  • Develop uniform standards for police-prosecutor cooperation. (MOJ, RPPO, MOI – medium-term)
  • Develop standards for prosecutors to decide which cases to appeal. (RPPO – medium-term)

Recommendation 5: Improve the functioning of the appeals system.

  • Set up a permanent body (working group or unit) in the SCC accountable for monitoring quality indicators in courts, i.e., confirmation rates, remand rates, and amendment rates. Monitor the quality of lower-instance courts' decisions and the appellate judgments to identify whether the appellate courts are appropriately using the possibility of amending first-instance decisions. (SCC - short-term/continuous)
  • Separately record Appellate Court statistics for cases received from Basic Courts and cases received from Higher Courts. (SCC – short-term)
  • Align statistical data on appeals of Basic Courts decisions to enable tracking of so-called ‘small appellation’ and ‘big appellation’. (SCC – short-term)
  • Enable tracking of lodged (not only resolved) appeals through the existing case management systems. (SCC, MOJ – medium-term)
  • Statistically monitor dismissed appeals as a separate category. (SCC, MOJ – medium-term)
  • Adopt policies that higher-instance judges should avoid reversals and replace the lower court's decision with their own. Ensure that remands contain precise reasoning and instructions to be followed by the lower court in subsequent proceedings. (SCC, HJC – medium-term)
  • Design and develop appropriate aggregated and disaggregated reports for monitoring appeals and corresponding higher-instance decisions (including information on confirmations, amendments, and remands). Reports should include data on court type, court, and case type, to enable evaluations to identify court types, individual courts and/or case types with adverse quality indicators (e.g., high remand rates) and identify reasons for poor results. (SCC – medium- term/continuous)
  • Identify causes of appeals (case law harmonization problems, loopholes in procedural laws, dilatory tactics, or other abuses by court users). (SCC, MOJ - medium-term)
  • Analyze the extent of appeals abuses in the Serbian judicial system, particularly in those court types and case types with the highest appeal rates. (SCC, MOJ – medium-term)
  • Develop possible sanctions for the abusing parties in line with COE recommendations, amend procedural laws, and issue instructions to stakeholders as appropriate. (SCC, MOJ – medium-term)
  • Develop standards for prosecutors to decide which cases to appeal. (SCC, MOJ – medium-term)

Recommendation 6: Regarding the ECtHR judgments, coordinate various state bodies to improve investigations, protection of property, length of proceedings, and enforcement of final decisions. (MOJ – medium-term)

Recommendation 7: Increase the use of specialized courts and case processing systems.

  • Analyze options for using specialized case processing systems in cases of general and specialized jurisdiction, with specific emphasis on Misdemeanor Courts and Administrative Courts. (SCC, MOJ – medium-term)