57988 USES AND USERS OF JUSTICE IN AFRICA THE CASE OF ETHIOPIA'S FEDERAL COURTS THE WORLD BANK Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts Table of Contents Acronyms...................................................................................................................................................xi Acknowledgments ......................................................................................................................... xiii Executive Summary Introduction ..............................................................................................................xv Background on the Current Study: Ethiopia.................................................xvi Trends in Demand for and Supply of Federal Court Services ............... xvii Users of Justice ­ Who Uses the Courts? ..................................................... xviii Uses of Justice ­ For What Are the Courts Used? ........................................xix How Do the Parties Fare? .....................................................................................xxi Delays and Their Causes ......................................................................................xxii Enforcement/Execution of Judgments ........................................................ xxiv Conclusions ........................................................................................................... xxiv CHAPTER I: Introduction to Study and Explanation of Objectives Background .................................................................................................................1 Objectives and Organization of the Present Study .......................................2 Two Notes on Usages Adopted Here..................................................................5 CHAPTER II: The Ethiopian Justice System Country Overview .....................................................................................................7 Governance Structures ............................................................................................8 The Justice Sector Organization...........................................................................9 Some Additional Details about Case Processing, Especially as Related to the Further Analysis ......................................... 17 iii Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts CHAPTER III: Overview of Recent Trends in Demand for and Supply of Federal Court Services Introduction ............................................................................................................. 19 Growth in Demand (New Filings) and Distribution by Major Types of Cases (Criminal, Civil and Labor Benches) ................. 20 Trends in Supply of Court Services: Dimension One, Changes in Size of Federal Court System ................................................................... 27 Trends in Supply of Services: Dimension 2, Disposition, Clearance and Congestion Rates ................................................................ 30 Conclusions .............................................................................................................. 34 CHAPTER IV: Users Of Justice ­ Who Uses the Courts ? Introduction ............................................................................................................. 35 Gender ........................................................................................................................ 37 Individual versus Organizational Users .......................................................... 39 Repeat Organizational Users .............................................................................. 41 Litigation Matrices ­ Who Sues Whom? ......................................................... 42 Conclusions .............................................................................................................. 47 CHAPTER V: Uses of Justice ­ For What Are the Courts Used? Introduction ............................................................................................................. 51 Types of disputes .................................................................................................... 53 Monetary Claims ..................................................................................................... 56 Appeals ...................................................................................................................... 57 Conclusions .............................................................................................................. 61 TABLE OF CONTENTS CHAPTER VI: How Do the Parties Fare? Introduction ............................................................................................................. 63 Winners and Losers in Three Major Areas (Criminal, Civil, Labor) by Litigation Status (a pro-plaintiff or pro-defendant bias?) ................................................................................. 63 Winners and Losers in Cases Pitting Organizations against Individuals ........................................................................................... 65 Closed and Reopened Cases .............................................................................. 65 Other Factors Affecting Outcomes: Amounts at Stake and Legal Representation ............................................................................. 69 iv Table of Contents Government Litigation ­ Does the Government Always Win? .............. 70 Conclusions .............................................................................................................. 71 CHAPTER VII: Delays and their Causes Introduction ............................................................................................................. 73 Trends in Times to Case Disposition ................................................................ 75 Trends in Factors Influencing Delay: Number and Length of Adjournments ............................................................................... 78 Trends in Other Factors Influencing Delay: Appeals Rates and Appeal Outcomes....................................................... 81 Conclusions .............................................................................................................. 83 CHAPTER VIII: Enforcement of Judgments Introduction ............................................................................................................. 85 Enforcement of Judgments in Comparative Perspective ........................ 85 Execution of Judgments in Ethiopia ................................................................ 86 Conclusions .............................................................................................................. 88 CHAPTER IX: Conclusions Using Ethiopia as a Case Study .......................................................................... 89 Lessons of Wider Utility ........................................................................................ 92 References .............................................................................................................................................. 95 List of Tables TABLE OF CONTENTS Table 1: FSC Appeal Vs Cassation ......................................................................... 23 Table 2: Average New Caseload Per Judge ....................................................... 24 Table 3: Number of Court "units" (First Instance, High Court, and Supreme Court, courts and benches from 2005 to 2009) ........... 28 Table 4: Number of Federal Judges (by type of court) from 2005 to 2009 ..................................................................................... 28 Table 5: Individual Plaintiffs/Appellants in Civil Cases, by Gender, Instance, and Year ...................................................................................... 37 v Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts Table 7: Individual Plaintiffs/Appellants in Labor Cases, by Gender, Instance, and Year ............................................................... 38 Table 6: Individual Defendants in Civil Cases, by Gender, Instance, and Year ...................................................................................... 38 Table 8: Individual Defendants in Labor Cases, by Gender, Instance, and Year ...................................................................................... 39 Table 9: Identity (individual, private organization, public organization) of plaintiffs and defendants to all civil cases filed between 2005 and 2009, by Court Level ................................. 39 Table 11: Identity (individual, private organization, public organization) of plaintiffs and defendants to all criminal cases filed between 2005 and present, by Court Level................ 40 Table 10: Identity (individual, private organization, public organization) of plaintiffs and defendants to all labor cases filed between 2005 and 2009, by Court Level ..................... 40 Table 12: Repeat Users in Federal Courts, cumulative for 2005­2009 ....... 41 Table 13: Major Civil Issues, FSC, by Year .............................................................. 53 Table 14: Major Civil Issues, FHC, by Year ............................................................. 53 Table 15: Major Civil Issues, FFIC, By Year ............................................................. 53 Table 16: Major Criminal Issues, FSC, by Year...................................................... 54 Table 17: Major Criminal Issues, FHC, by Year ..................................................... 54 Table 18: Major Criminal Issues, FFIC, by Year (Lideta and Arada only) ..... 54 Table 19: Major Labor Issues, FSC, By Year ........................................................... 55 Table 20: Major Labor Issues, FHC, by Year .......................................................... 55 Table 21: Major Labor Issues, FFIC, by Year .......................................................... 55 Table 22: Mean Monetary Claims for Filed and Disposed Civil Cases, TABLE OF CONTENTS by Year, FSC, FHC, FFIC (Lideta and Arada Only) ............................. 56 Table 23: Appeals Rates (Applications) by Year for Major Types of Civil Cases, FHC to FSC ........................................................... 60 Table 24: Civil Cases by year and instances (FICC and HC, original jurisdiction only) ........................................................................ 64 Table 25: Labor Cases by Year and Instance (FFIC Only as all labor cases go there in the first instance) ......................................... 64 Table 26: Criminal Cases by Year and Instance (FFIC and FHC, Original Jurisdiction only)....................................................................... 64 Table 27: Outcome of all civil cases (trial stage) for involving individual vs organization ...................................................................... 65 vi Table of Contents Table 27: Percentage of Reopened Cases against Filed .................................. 67 Table 28: Reopened cases in FHC for Major types of Civil Cases ................. 68 Table 29: Reopened cases in FHC for Major types of Criminal Cases ......... 68 Table 30: Reopened cases in FHC for Major types of Labor Cases .............. 68 Table 31: Reopened cases in FIC for Major types of Civil Cases ................... 68 Table 32: Reopened cases in FIC for Major types of Criminal Cases ........... 69 Table 33: Reopened cases in FIC for Major types of Labor Cases ................ 69 Table 34: Government Litigation ............................................................................ 70 Table 35: Mean Times to Disposition for Criminal, Civil and Labor Cases, FFIC (for civil and criminal, Arada only, for labor, Yeka only) ................................................................................... 76 Table 36: Mean Times to Disposition for Major Types of Civil Cases, FFIC (Arada Only) ................................................................ 78 Table 37: Average Duration (in Months) of Appeals, FSC and FHC, by Major Types of Cases and Years ...................................................... 83 Table 38: Plan or Judgment Execution, Federal Courts, 2006­2009, Targets and Achievements ..................................................................... 87 List of FIgures Figure 1: New Filings (Original Jurisdiction Only) from 2005 to 2009; First Instance and High Court ............................................................. 22 Figure 2: New Filings (Total= Original Jurisdiction + Appeal + Cassation) ................................................................................ 22 Figure 3: Average New Filings per Judge (Original Jurisdiction) 2005 to 2009 .............................................................................................. 23 TABLE OF CONTENTS Figure 4: Litigation Rates: Number of New Filings per 100,000 Population by Year .................................................................................. 24 Figure 5: Federal High Court Appeal Rates, Appeals Requested by Major Legal Area ................................................................................ 25 Figure 6: Federal Supreme Court Appeal Rates, Appeals Requested by Major Legal Area ................................................................................ 26 Figure 7: Number of Dispositions by Level of Court and Material from 2005­2009 ....................................................................................... 32 Figure 8: Clearance (Disposition) Rates by Court, Material, and Year...... 33 Figure 9: Congestion Rates by Level of Court, Material and Year ............. 34 vii Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts Figure 10: Litigation Matrix, Civil Cases, FSC ...................................................... 43 Figure 11: Litigation Matrix, Civil Cases FHC....................................................... 43 Figure 12: Litigation Matrix, Civil Cases, FFIC (Lideta and Arada Courts only) .................................................................................. 44 Figure 13: Litigation Matrix Labor Cases, FSC .................................................... 45 Figure 14: Litigation Matrix Labor Cases, FHC.................................................... 45 Figure 15: Litigation Matrix Labor Cases, FFIC (Old Airport and Yeka, only) .................................................................................................. 45 Figure 16: Litigation Matrix Criminal Cases, FSC ............................................... 46 Figure 17: Litigation Matrix Criminal Cases, FHC .............................................. 46 Figure 18: Litigation Matrix Criminal Cases, FFIC (Lideta and Arada, only) ........................................................................ 47 Figure 19: Appeals Rates (Application to Appeal) from FFIC to FHC, by Legal Area and Year........................................................................... 58 Figure 20: Appeals Rates (Admitted Cases) from FFIC to FHC, by Legal Area and Year........................................................................... 58 Figure 21: Appeals Rates (Application for Appeal) from FHC to FSC, by Year....................................................................... 59 Figure 22: Appeals Rates (Appeals granted) from FHC to FSC, by Year..... 59 Figure 23: Number of Reopened Civil Cases ­ FFIC, FHC, and FSC by year ......................................................................................................... 66 Figure 24: Number of Reopened Criminal Cases, FFIC, FHC, and FSC by Year ........................................................................................ 66 Figure 25: Number of Reopened Labor Cases, FFIC, FHC, and FSC, by Year ....................................................................................... 67 Figure 26: Mean Times to Disposition for Criminal, Civil, TABLE OF CONTENTS and Labor Cases, FSC.............................................................................. 76 Figure 27: Mean Times to Disposition for Criminal, Civil, and Labor Cases, FHC ............................................................................. 76 Figure 28: Mean Times to Disposition for Major Types of Civil Cases, FSC ........................................................................ 77 Figure 29: Mean Times to Disposition for Major Types of Civil Cases, FHC ....................................................................... 77 Figure 30: Mean Number of Adjournments for Major Types of Civil Cases (FSC) .................................................................................. 79 Figure 31: Mean Length of Adjournments for Major Types of Civil Cases (FSC) ...................................................................... 80 viii Table of Contents Figure 32: Mean Number of Adjournments for Most Common Civil Cases (FHC) ....................................................................................... 80 Figure 33: Mean Length of Adjournments for Major Types of Civil Cases (FHC) ..................................................................... 81 Figure 34: Appeals and Reversal rates for all cases, FHC to FSC, by Year ......................................................................................................... 82 Figure 35: Appeals and Reversal Rates, all cases, FFIC to FHC by Year ......................................................................................................... 82 TABLE OF CONTENTS ix Acronyms BPR Business Process Reengineering CPC Civil Procedures Code ECA Eastern Europe and Central Asia EPRDF Ethiopian People's Revolutionary Democratic Front FFIC (FIC) Federal First Instance Court FHC (HC) Federal High Court FSC (SC) Federal Supreme Court ICT Information and Communication Technology JAC Judicial Administrative Commission MIS Management Information System PPP Purchasing Power Parity PSCAP Public Sector Capacity Building Program xi Acknowledgments This report was prepared by Linn Hammergren, World Bank Senior Rule of Law Advisor, and Sintayehu Mitiku, ICT Advisor for the Federal Supreme Court of Ethiopia, under the supervi- sion of Lisa L. Bhansali, Task Team Leader of the World Bank in the Africa Region. Funding for the research was provided by a grant from the Norwegian Ministry of Foreign Affairs to support rule of law work and enhance the Bank's effectiveness in facilitating reform. The team is especially grateful for the assistance provided by Justice Menberetsehai Tadesse, Vice President of the Ethiopian Federal Supreme Court, who served as Special Advisor to the project, offering insights on interpretation of the judicial database as well as on the details of Ethiopian law and legal practice. Work was also facilitated by the World Bank Country Office in Ethiopia through its sponsorship of two prior projects, including a study on the justice sector and a review of its Public Sector Capacity Building Program (PSCAP). The two principal drafters, Ms. Hammergren and Mr. Mitiku, as well as Ms. Bhansali would like to thank the other parties for their support without which this project could not have been started or completed, in particu- lar Ms. Nancy Chaarani Meza. Finallly, the conclusions and opinions expressed in this study are those of the principal authors, who take full responsibility for the report and its contents. xiii Executive Summary Introduction This report presents the findings of one of a series of studies sponsored by the World Bank on the "uses and users" of courts in the regions where it participates in justice reform projects. These studies typically use aggregate statistics and random samples of cases files to analyze court performance as regards: Changes in demand for court services and the creation of any supply-demand gap The effects of reform measures directed at reducing the gap or at correcting other reputed problems Patterns and changes in court uses, both as regards the identity of the parties and the types of disputes they take to court Patterns in case outcomes--who wins and who loses--and their relationship to such char- acteristics as party identity, the issues under dispute, or the amounts at stake The incidence of delay, where it occurs, and its causes Whether judgments once final are enforced. Exploration of other common beliefs (conventional wisdom) about the administration of justice in the given country The results of these studies are useful to countries and to the Bank in separating real from imagined problems, identifying their causes, planning reform programs, and tracking their results over time. Although in Africa court use is typically restricted to a small portion of the population, most donor funding goes to the formal court system and most countries are interested in expanding access to the latter. Hence, the studies can help evaluate those invest- ments and point them in directions where they are likely to do the most good in advancing objectives like increased and more equitable access, delay reduction, and satisfactory resolu- tion of common disputes. xv Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts Background on the Current Study: Ethiopia Ethiopia is one of the poorest countries in Africa and thus in the world. As in the rest of Africa, 1 traditional dispute resolution mechanisms and a series of "hybrid institutions" serve a majority of the population. However, its judicial system (incorporating federal, regional-state, and some municipal courts) is unusual in several aspects. First, since 1994, the government has been actively engaged in expanding access to the formal system and improving the professionalism of its judges. Thus, while still low as compared to other regions, Ethiopia's current 4.0 judges per 100,000 inhabitants is high for Africa, and judges, who before the early 1990s often did not 2 have law degrees are increasingly fully accredited lawyers. Second, at both the federal and sub-national levels, governments have been actively extending the reach of the court system, both by adding courts and judges and by using a variety of additional mechanisms (including circuit riding, mobile units, and modern ICT equipment). Third, and most important for this study, for the past ten years, the federal judiciary has been promoting the development of a computerized case tracking system to monitor performance. (It has also been computerizing other court activities, but this is a slow process owing to resource and logistical constraints.) The system's quality is currently high enough to allow this study to be based on its contents rather than on the usual random sample of case files. Only the federal courts' statistics were 3 used (and for the more detailed analysis of FFIC cases, only drawn from four of the nine courts ), but the systems now installed in several regions would have allowed comparable work there. Following the Federal Supreme Court's lead, Ethiopia's federal, regional, and municipal judi- ciaries have been using the system to identify shortcomings in service delivery, reduce delay, and resolve other prioritized problems. The full set of statistics is available on court websites and is used to generate over 100 reports on aspects of case flow. As Ethiopia's procedures are based on common-law practices, one particular concern has been to limit the number and length of adjournments of hearings, a major source of delay in largely "oral" systems. Judges' disposition rates and overall productivity are also tracked, and more recently the Federal EXECUTIVE SUMMARY Supreme Court (FSC) has begun looking into problems caused by the high appeals rate and contested enforcement of judgments. Thus, in addition to investigating the normal ques- tions for these studies, this one also examines the courts' success in using their database as a 1 In Ethiopia, these include the executive-dependent "social courts," which mix customary and formal law and use lay judges, and popular militias. Both were introduced in the 1970s and have been pre- served in modified form under the post 1991 government. 2 At the federal level, all judges now have law degrees and some have advanced degrees. Many regional judiciaries have introduced training programs to provide law degrees to judges with only a certificate (6 months of legal studies after High School) or a diploma (2 years). 3 Because of how the data are organized it proved too difficult to do some of the analysis across the entire FFIC. For civil and criminal cases, the two courts with the highest workload (Lideta and Arada) were used; for labor cases, we used two other courts (Old Airport and Yeka) which receive relatively more labor disputes. xvi Executive Summary lever for reform. It bears noting that Ethiopia's case tracking system is an exemplary model of this approach and generates information on court performance of much higher quality than in many regions that have invested far more money over much longer time in automating their courts. Trends in Demand for and Supply of Federal Court Services The Federal Courts can receive cases originating anywhere in the country, but most of their business at the trial level (First Instance and High Court, FFIC and FHC) comes from the two cities where they are physically located, Addis Ababa, the capital, and Dire Dawa. These cities hold a special status under Ethiopian law, and although both consequently have municipal courts, many disputes occurring within them go to the federal jurisdiction. The Federal High and Supreme Court (FSC) also function as courts of appeal and cassation (FSC only). Cassation cases come not only from the federal jurisdiction but from regional and municipal judiciaries and from the "social courts." Two additional features are important in shaping demand: court fees are very low and legal representation is not required (as there are currently fewer lawyers in private practice than judges). This increases accessibility of all courts to all citizens al- though admittedly an unrepresented peasant taking a case in cassation to the FSC may have problems shaping the legal issues. Cases are divided into three major legal areas--criminal, civil, and labor--with benches and judges specializing accordingly. In absolute numbers, criminal cases are most common, fol- lowed by civil (which also includes administrative disputes) and then labor cases. For the two (of 9 to 124) first instance courts (FFIC) surveyed in this study, criminal case filings reached a high of about 23,000 in 2007­08 while civil cases topped at about 17,000 in 2008­09. For the two courts receiving most labor cases, new filings reached only slightly over 2,000. EXECUTIVE SUMMARY Over the period covered by this study (mid-2005 to mid-2009), new filings in the federal juris- diction have nearly doubled, and despite the addition of judges, average caseloads have gone up accordingly. The rise in caseloads is significant, and in the FFIC is nearing what might be considered a maximum feasible workload. However despite this growth in demand, courts have increased their disposition rates accordingly, keeping their clearance levels in the 90­115 range (100 would indicate dispositions exactly equal to new filings) and reduced their congestion rates 4 Numbers of "courts" (separate physical facilities housing numerous judges and benches) vary be- cause some of them have been closed recently, owing to their poor condition. xvii Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts Average New Caseload Per Judge (Total= Original Jurisdiction + Appeals + Cassation, as applicable) Includes all FFIC filings and judges Year 2005­06 2006­07 2007­08 2008­09 Court Filings Per Judge Filings Per Judge Filings Per Judge Filings Per Judge FSC 5195 325 5746 319 7415 221 8265 344 FHC 7641 170 9108 175 9108 166 13237 241 FFIC 48115 891 51536 859 65452 747 80461 1018 from highs of 2.0 to under 1.5 and in some cases close to 1.0. At 1.0 no further backlog would be accumulating. Given the substantial increases in demand, and relatively lesser increases in the number of judges, this is a significant accomplishment, but one the courts are still work- ing to improve. Users of Justice ­ Who Uses the Courts? Since we know that most Ethiopians do not use the courts, and even fewer, naturally, take their disputes to the federal jurisdiction, we were interested in the extent to which court use was skewed toward certain, probably elite groups. Earlier studies in other countries, for example, identified a series of repeat users in civil cases even at the first instance level, largely banks, in- surance companies and similar organizations who use the courts to collect small debts (World Bank, 2002; Gonzales et al, 2002). We were also interested in knowing the extent to which women accessed the courts as it is often argued that they face exceptional obstacles. It should be noted that statistics on FFIC are drawn from only four (of the 9­12) courts for this and later analysis, as the organization of the database made it difficult to do this universally. EXECUTIVE SUMMARY Answers to both these questions suggested less elitism (and thus broader access) than might be supposed. First, in civil cases entering the FFIC, women were slightly more likely to appear as plaintiffs than men. Their representation declined in the FHC and FSC, but never to less than half. As original jurisdiction cases are a minority in the FHC and virtually nonexistent in the FSC, this may also indicate that they usually won their cases in the first instance (thus did not appeal). Women were significantly less frequent users for labor cases at all instances, but not knowing more about employment patterns in Ethiopia it is impossible to say whether they were "under-represented." Second, individuals were the most frequent plaintiffs in both civil and labor first instance cases, and in the civil jurisdiction most cases were between individuals. The anticipated strong showing of organizational plaintiff versus individual defendant in first instance civil cases did xviii Executive Summary not hold out. In labor, most FFIC cases featured individual plaintiffs against organizations, as was expected. These patterns also held at the higher levels, although in the civil area, organi- zation vs. organization cases, second place in the FFIC, became less significant in the FHC and FSC. This finding is worth further exploration but like many of the questions emerging from the initial analysis could not be pursued for lack of time. Repeat users in civil cases were identified, but they appeared much less frequently than might be expected, and were equally likely to be plaintiffs or defendants at the FSC and FHC. (The database did not allow their identification in the FFIC, and there was little point in including criminal cases as the state is always a party by definition). Litigation matrixes (iden- tifying parties as individuals or organizations) were developed for all three legal areas, but were less interesting for criminal cases, as they nearly always pit the state prosecutor against an individual defendant. The greater frequency of individuals as appellants in criminal cases at the higher levels has two explanations: they are appealing a guilty verdict (or the non- granting of bail) and possibly a reputed disinclination of prosecutors to appeal when they lose. Uses of Justice ­ For What Are the Courts Used? As frequency of labor, civil and criminal cases was tapped in earlier sections, here we looked at the most common disputes within each category and instance. Again a few additional details are required to understand the outcomes. First, the FFIC receives many civil cases that are not disputes but only require a "declaratory judgment" or judicial recognition of a legal situation. Second, while land in Ethiopia, as in much of Africa, increasingly gives rise to disputes, all land is owned by the state. Thus, conflicts center on use-rights or ownership of immovable prop- erty (houses, other buildings). Third, most civil cases have the initial claim recorded, allowing comparison with the actual award. The amount claimed is one of the factors affecting whether original jurisdiction cases go to the FFIC or FHC. EXECUTIVE SUMMARY Our review of what gets to court leaves many of the initial hypotheses untested. However, a few findings as relates to them and to other issues do merit mention. First, while the most common civil disputes sent to the FHC (original or appellate jurisdiction) and FSC (appellate or cassation) were very similar, only two of these categories (execution and inheritance) also appeared in the FFIC top list. Contract disputes were important in the FHC and FSC, but in the FFIC civil area, the comparable "diverse monetary claims" are not as prevalent as has been found elsewhere (where they may represent up to 70 percent of first instance civil cases). This may have more to do with exogenous factors--patterns in mak- ing loans and issuing credit--in Ethiopia than in any shortcomings of the courts. While one does not want the courts to become a debt collection agency for banks and retailers, as more xix Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts credit is issued, contract enforcement especially for smaller loans is likely to become more important. Second, issues regarding property, both movable and immovable, are important at the higher instances. Especially in urban areas (where the federal courts largely operate) ownership of buildings, mortgages and foreclosures, and right to construct are increasingly controversial, and not surprisingly accompanied by charges of considerable administrative corruption, ex- plaining in some part why cases are then taken to court. That they do not appear as frequently in the FFIC is doubtless because of their higher value. Not surprisingly, average monetary claims for civil cases were highest at the FHC for original jurisdiction cases. However, here there was also a significant difference between the claims filed and the final awards--suggesting only that claimants went for broke and that judges reduced the awards accordingly. Claims for appeals, cassation and FFIC cases tended to be relatively low (largely because those for appeals and cassation are based on the awards made at the lower level court). Mean claims at the FFIC averaged slightly less than US$2,000, still high in relation to what has been found elsewhere (World Bank 2002 for Mexico) and thus suggesting that poorer clients or those with minimal claims find other ways to resolve them. This may simply mean they go to municipal courts, but there the acknowledged problem is lesser citizen confi- dence (possibly warranted). Third, the high percentage of declaratory judgment cases in the FFIC is seen in part in the rising importance of adoption cases. Ethiopia is currently one of the three most popular countries for foreign adoptions and this is clearly generating business for the FFIC. We did not separate out other declaratory cases, but it is estimated that they account for 28 percent of the workload in the two courts included in the sample. Fourth, the large number of execution cases recorded in the FFIC and their substantial rep- EXECUTIVE SUMMARY resentation at the other instances is a growing concern for the courts. These are cases where winners request assistance in enforcement or where losers protest the form enforcement takes, and in either situation are regarded as arising in efforts to resist payment of claims. Finally, and related to all the above comments, appeals rates (number of appeals over num- ber of judgments) in Ethiopian courts remain very high (up to 50 percent for admitted cases and still higher for applications), and despite court screening of applications, are still not under control. By law, the right to a second appeal has been restricted, but this has still not brought rates to what might be considered a reasonable level, one closer to 10 percent. One obvious solution, to charge higher fees for appeals and cassation (where no fee is charged), is probably not advisable in Ethiopia because of its impact on access to justice. However, as the courts realize, something needs to be done as many appeals appear to be filed for purely dilatory purposes. xx Executive Summary How Do the Parties Fare? Owing to problems with the database, most of the findings had to be based only on the last year covered, 2008­09. Moreover, for the FFIC outcomes only two courts were used for each category--Lideta and Arada for civil and criminal cases, and Old Airport and Yeka for labor cases. Labor cases are only heard in the first instance in the FFIC for which reason there is no entry for the FHC. First instance cases entering at the Supreme Court are too rare to be worth noting. Original Jurisdiction Case Outcomes by Instance, 2008­09. For FFIC, data only from Lideta and Arada (Civil and Criminal) and Old Airport and Yeda (Labor) 2008­09 Civil Criminal Labor Litigants FIC HC FIC HC FIC Plaintiff wins 6107 81 4917 775 784 Defendant wins 3886 163 3129 1549 499 Plaintiff partial win 1200 27 894 258 143 Case concluded without judgement 9111 415 15,663 3174 1300 Three observations merit attention. First is the tendency for the plaintiff to win in the FFIC, but not in the FHC. We have no explanation for the FHC outcomes, but for the FFIC this can be regarded as positive, meaning that plaintiffs generally are filing cases with merit. There may be other less positive explanations, but in a well-functioning and efficient system, we would expect plaintiffs to file and courts to accept only cases involving a real, justiciable dispute Second, is the relatively poorer showing of criminal as opposed to civil plaintiffs. The criminal plaintiff is of course the state, meaning that the state is losing many cases, and that prosecu- EXECUTIVE SUMMARY tors, as many critics observe, too often proceed with a case without sufficient evidence and preparation--as one government commission characterized this, "charging to fail". Since most defendants are not represented by an attorney (are likely to be so only in the FHC), this is still 5 less understandable. Third, is the high number of cases concluded without judgment, at the FFIC about 40 percent for civil, 50 percent for labor and 60 percent for criminal. For civil cases, the outcomes are not as poor as they look, as the numerous "declaratory judgments" are located in the concluded 5 Ethiopia does have public defenders, but they are few in number and thus only provided to indi- gent defendants in more serious cases, those most likely to be seen in the FHC. xxi Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts without judgment category. However, for criminal cases, when combined with the state's higher loss rate, there is clearly a problem--and its explanation lies not only in the prosecutors' probably insufficient preparation, but also in their inability to get the witnesses, and sometimes the defendant to court. Since it is often expected that organizations will have an advantage over individuals, we also looked at outcomes of cases involving both types of parties, with the following results. Outcome of all civil cases (trial stage) involving individual vs organization 2008­09 FIC HC Litigants N = 3156 N = 647 Individual plaintiff wins over ORG defendant 1991 129 Individual plaintiff wins partially over ORG defendant 164 0 Here N represents the total number of disposed cases in this category (individual plaintiff ver- sus organizational defendant) in the FHC and for two FFIC courts (Lideta and Arada) in 2008­09. In the FFIC, individuals win wholly or partly in over two-thirds of the cases. Their showing is less impressive in the FHC, but still any notion that individuals do not win over organizations is clearly untrue. Closed cases can be reopened, depending on the reasons for closure, and we did find this occurred, but in a minor proportion. This was highest for the FFIC, where between 2005 and 2009, reopened labor cases constituted about 14 percent of the total workload, and reopened civil cases between 11 and 22 percent. However reopened criminal cases were less than 4 percent of the workload, suggesting that prosecutors are not trying to make up for EXECUTIVE SUMMARY their poor showing on the first round. There are many explanations for the prosecutors' prob- lems, but the one most often offered by their agency, that they are too few in number, seems doubtful. In Ethiopia the number of prosecutors is equal to and sometimes more than that of judges in all jurisdictions. Delays and their Causes Delay is a universal complaint about judicial performance, and in Ethiopia reducing its incidence has been a high priority goal of the courts. However, not all delay is directly under judicial control. Parties and their lawyers may work to create delay--to avoid justice--and various logistical problems, common to developing countries, may also increase its inci- dence. xxii Executive Summary As a result of their programs, and the use of the database, the courts not only keep track of disposition times, but also have succeeded in reducing them for the most part, both in original jurisdiction cases and at the appellate levels. Average Duration (in Months) of Cases Disposed by First Instance Courts (Arada for Civil and Criminal and Yeka for Labor) Case Type 2005­06 2006­07 2007­08 2008­09 Civil 10.03 10.11 5.32 5.52 Criminal 7.42 6.00 4.28 2.27 Labor 5.33 6.44 7.73 4.03 As can be seen, the efforts to reduce disposition times in the FFIC have been very successful, with only a little backsliding for labor cases in the middle years. As shown below, patterns in the FHC and FSC are slightly more erratic although generally on a downward trend as well. We were told that the FSC civil bench was the most problematic, and the data do support that claim. Average Duration (in Months) of Appeals, FSC and FHC, by Major Types of Cases and Years Case 2005­06 2006­07 2007­08 2008­09 Category SC HC SC HC SC HC SC HC Civil 3.0 16.5 5.4 11.6 7 6.9 5.7 7.5 Criminal 4.7 12.9 6.0 7.0 2 2.2 2.0 2.7 Labor 3.0 2.9 3.6 2.1 2 2.5 1.3 1.8 While times for appeals have generally gone down, and rates of overturns on appeal are low, EXECUTIVE SUMMARY less than ten percent, appeals rates remain high. It thus appears that appeals are still being used, if less successfully than before, to augment delay. The judiciary is attempting to find ways to resolve this, so far unsuccessfully. Where it has had more luck is in reducing the frequency and length of adjournments, setting targets for the mean number of adjournments per case and applying the following rules: No judge initiated adjournments are allowed If the plaintiff or both parties do not appear at a hearing the case is closed If the defendant does not appear, a default judgment is given While success in limiting the number of adjournments has been erratic, and varies by type of case, they are generally lower now for the FSC and FHC, and moreover the mean length of xxiii Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts adjournments for civil cases in the FHC has declined fairly steadily. One concern raised by this policy is that judges may be closing cases too quickly, and while cases closed without judg- ment can be reopened, as we have seen this is not that frequent. Enforcement/Execution of Judgments This is an area attracting more judicial attention in Ethiopia and universally. A favorable judg- ment is of little worth if it cannot be enforced, and even in countries that have not made much effort to reduce pre-judgment delays, parties often resort to additional dilatory tactics at the enforcement stage. One further problem is that many courts lack much information on enforce- ment, only hearing about it when something goes wrong. Ethiopia has some data, especially after its introduction of enforcement benches, but still not enough to develop adequate remedies. The available data do, however demonstrate, through the increasing importance of execution cases at all instances, that the problem is growing. Ethiopian scofflaws have been especially inventive in avoiding execution of civil judgments, as the courts are well aware. In addition to the usual tricks of claiming insolvency, hiding assets, or protesting seizures of certain goods, they have also begun to undermine enforcement by en- couraging the last minute (post-judgment) appearance of third parties who claim an interest in seized assets or collateral. If done successfully (often with backdated deeds, and other falsified records), this can cause the nullification of a judgment and the reopening of a case to recog- nize the new interest. The FSC has developed a targeted work plan to accelerate enforcement of judgments, but it clearly will need other tools, as against a wily debtor, reliance on judicial incentives alone will not be sufficient. Conclusions EXECUTIVE SUMMARY As compared to other countries in Africa, and in many other regions where the World Bank works, Ethiopia has made significant progress in developing a management information system and applying it to improve court performance. The analysis it currently does and could do with its database is possible in few countries, and in none of the five Latin American nations (Argentina, Brazil, Ecuador, Mexico and Peru) where "uses and user" studies were done earlier. The Ethiopian database facilitated the current study (although the study also revealed areas where it needs improvement, not that the FSC was not aware of most of this). It also offers the potential for far more exhaustive analysis of many of the questions and issues raised here. In short, whether or not Ethiopia has achieved all its goals in improving judicial performance, it is clearly on the right track as regards its approach and is a lesson to other countries that have used automation primarily for other purposes--to speed up drafting of standard documents, to allow internet submission of massive filings, or to tell judges when a hearing is about to xxiv Executive Summary start. These alternative uses are helpful, and Ethiopia is pursuing some of them now, but its experience demonstrates that the biggest improvements in court performance are necessarily based on accurate information on how cases are progressing within the courts and on an initial investment in automation for this purpose. Since few other countries, and even fewer in Africa, have performance statistics that allow com- parison, it is hard to say how Ethiopia is doing in relative terms as regards its goals of increas- 6 ing production, productivity, access, and the quality of results. However, it does appear that Ethiopia has at the very least stopped increasing congestion and diminished the average times for resolution of cases in its Federal Courts (and most likely in some of the regional systems as well). Federal Court caseloads, especially in the FFIC, are quite high by international standards, making the clearance rates still more impressive. The large number of cases "closed without judgment" is a concern and a possible consequence of the tracking of times to resolution (since a closure without judgment is also a resolution). However, this should be a learning process, and as with the paradox of high appeals and low overturn on appeal rates, the courts' response may eventually get court users to change their behavior. In a country like Ethiopia which has reduced some of the usual barriers to access (court fees and the need for legal representation), it is harder to get these messages out. A good lawyer should make sure his/her client appears in court along with the witnesses, but with many first-time, and probably unrepresented parties, there is a wide margin for mistakes and a good deal of room for still greater effort to help them understand the rules. In short, and despite whatever shortcomings in the reform program, Ethiopia offers an impor- tant example for other African countries. A first lesson is that rather limited automation can still produce a good management information system and that this probably should be a first priority as automation begins. A second is that this type of system can and should be devel- oped gradually--trying to introduce a multi-functional system as a first step is a near guarantee for failure. Third, court leadership needs to understand and utilize the system as a means of improving performance--it does no good to have information no one uses. And fourth, the EXECUTIVE SUMMARY court requires support in this effort. Users may provide some pressure for change, but it is also important that the other branches of government recognize and support the importance of making these improvements as they have in Ethiopia. Donors, like the World Bank, also need to absorb these lessons. One frequent question asked in discussions of automation is "what 6 One of the report's authors has done considerable work in Latin America and can verify that no country there, with the possible exception of Chile, could provide the type of performance statistics Ethiopia routinely collects and publishes ­ for example clearance, congestion, and appeals rates, rates of overturns on appeal, and certainly not mean frequency and length of adjournments. The aggregate statistics collected elsewhere in Africa (we did not include South Africa and trust it is the exception to the rule) were generally of dubious quality and definitely would not permit the type of analysis done here. xxv Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts system should we recommend or adopt?" This question has no answer--because systems as developed as that of Ethiopia, or those even further developed in OECD countries, are not the model for a country that is starting out. What is most important, and what apparently hap- pened in Ethiopia, is that the interested parties observe more advanced systems and under- stand where they want to go, but that they begin more modestly, with what is achievable and practical in their current situation. This was not meant to be a discourse on court automation, but as most of the other conclu- sions have been mentioned earlier, it is logical to add these thoughts here. Most donor-spon- sored court automation projects have failed because they tried to do too much and started with the wrong thing. This is also true of country financed projects--the application that does everything except prepare the judge's coffee, but at the same time does not provide basic case flow statistics. The bells and whistles (and Ethiopia has some of them--cell phone connections to determine the status of one's case in the FSC) can win friends, but they are less important as regards improving performance and that is what reform in the end is about. EXECUTIVE SUMMARY xxvi CHAPTER 1 Introduction to Study and Explanation of Objectives Background During the early years of the past decade, the World Bank sponsored a series of research projects in five Latin American countries as input to improvement of its judicial reform efforts in the region. The studies used a methodology that had been developed in the U.S and was based on the analysis of random samples of court case files to determine how, by whom, and to what ends litigation was undertaken and how users fared once they initiated or were drawn into legal action. Given the expense of the methodology, the samples were inevitably small and covered only a few courts in each country. Nonetheless the results were interesting enough to encour- age the initiation of similar studies in other regions, most notably in ECA.7 In both regions, the absence of good statistics and other empirical data on "court use and users" had forced reform- ers to design their judicial or justice reform programs on the basis of not much more than intuition, conventional wisdom and anecdotal evidence, As the research demonstrated, much of what "everyone knew" about the courts proved either inaccurate or completely wrong. This con- sequently gave rise to reform measures that could be equally misguided--attempting to resolve nonexistent problems or providing solutions for real ones that attacked the wrong causes.8 On the basis of the positive reactions to these first studies, the World Bank decided to under- take similar research in Africa, a region where good empirical information on court perfor- mance tends to be still scarcer. This is certainly true for the first country selected, Ghana, where judicial statistics and methods for tracking individual cases are still in their infancy. Interestingly, the second country, Ethiopia, is an exception to this general rule. Ethiopia not only has good performance statistics for its federal and many of its regional courts, but also uses them to identify and resolve problems--and moreover makes its database available on the judicial websites. As a consequence, the study undertaken in Ethiopia could take a 7 Many of these studies have not been released publicly. However, an overview of the Latin Ameri- can work is found in Hammergren (2002). Two published examples are World Bank (2002 and 2003) 8 The presumed excessive judicial workload as a problem in its own right and as the major cause of delay is a good example of both phenomena. In most instances, the studies found that judges did not have an unreasonable workload and that even those handling higher numbers of cases ap- peared able to keep up with most of them. What got left behind remained there for other reasons, meaning that adding more judges was not a satisfactory solution. As documented in World Bank (2002), other remedies adopted in Mexico to speed up case processing seemed to make no differ- ence, largely because they did not target the most critical contributing factors. 1 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts methodological short cut, eliminating the costly phase of creating a database from random sampling, and moreover focusing on the universe of cases filed in the federal courts since roughly 2005.9 Although Ethiopia does its own analysis of its database (some of which is used here), it was determined that additional work, using categories developed for cross-country analysis, would still be of interest, both to Ethiopians and to those concerned with African court performance in general. Moreover, the results of Ethiopia's data-based reforms could also be usefully contrasted with different approaches taken in other countries. There are obvious strengths, but also some potential shortcomings of the Ethiopian strategy and one of the ad- ditional aims of this research and of studies underway or planned in other African countries is to explore them. From the start, however, it should be recognized that the development and use of a court information system, both to track individual cases and caseloads and to identify and resolve systemic performance weaknesses is a very positive step and thus that Ethiopia's experience in this area merits serious study, not only by other African judiciaries, but by court systems in other regions. None of the five Latin American countries covered in the first round of research had anything comparable despite their substantial investments in court automation over often much longer periods of time.10 Thus, while the purpose of the present study is not to evaluate or showcase Ethiopia's automation process and automated database, the researchers would be remiss if they did not call attention to this significant accomplishment. Objectives and Organization of the Present Study Like the prior and on-going studies in Latin America, Eastern Europe and Central Asia, and Africa, the current research looks only at cases handled in the formal court system. In light of our growing understanding that courts are only one of a series of institutions involved in the resolution of disputes, strengthening of the normative framework and imposition of sanctions on norm violators, this is already problematic, but throughout Africa, where up to 85 percent of the population may use other mechanisms (e.g. religious or traditional bodies, other community authorities or notables, some kind of more modern alternative mecha- 9 The database for the federal courts actually goes back further, but the 2005 start date was used for two reasons: first, to make it comparable with the period covered in the Ghana study and second, because the database has been under development since its inception, and not all variables of inter- est for this study could be tracked for the at least 10 years of its existence. CHAPTER 1 10 This is largely because Latin American countries have invested heavily in giving everyone a computer and introducing a sort of "judicial Word," a program allowing staff to draft standard docu- ments rapidly. Ethiopia, despite having insufficient funds to automate fully, prioritized a system that tracked case movements, thus producing a complete set of caseload statistics which has been used to guide its subsequent reforms. 2 Introduction to Study and Explanation of Objectives nism) the value of such an approach could be questioned from the start. However, there are several arguments in its favor. First, nearly all donor support goes to formal institutions, and thus the studies should help direct these contributions toward resolving real problems. Second, in most countries, governments are attempting to expand access to formal institu- tions and thus understanding the latter's strengths and weaknesses should help organize that effort. Third, the studies do capture how various court users (including the poor) fare under the current system and also indicate what kinds of cases are or are not taken to the courts. This information will be valuable in designing access programs from a variety of standpoints--eliminating obstacles to effective access; redesigning laws and procedures so that cases presumed to be of more interest to the poor are recognized in the formal jurisdic- tion and so on. Finally, the studies may provide both donors and governments with a better idea of how much to invest in strengthening the formal system, as opposed to trying to work with alternative mechanisms. As both the title ("Uses and Users") and the above explanation suggest, these studies focus on a number of issues by collecting information on what gets to court and how it is handled once it arrives. The case file (or in the case of Ethiopia, statistical) analysis is usually complemented by a review of aggregate data on trends in overall demand and supply of court services, the extent of the latter being dependent on the quality and coverage of existing statistical systems. Issues and questions explored in the research are as follows: Tendencies as regards changes in the overall demand for and supply of court services. This is typically captured through aggregate statistics and reviews such issues as changes in numbers of filings, numbers of dispositions, and growth of the judicial infrastructure (number of courts and judges) to respond to new levels and types of demand. Where ex- isting statistics allow, these changes can also be tracked at the level of major legal areas (civil, criminal and administrative cases for example) and for sub-types of cases within them. The creation of any demand-supply gap (courts' inability to keep up with growing demand as evidenced by increases in congestion or "cases being left behind") and the identification of the courts, proceedings, and substantive areas in which it is most prominent. The effects of any reform measures on reducing the demand-supply gap or on any other problems identified by key stakeholders. Patterns and changes in court use--both as regards the identity of the parties and the types of cases filed. This is especially important in countries or regions where access to courts is limited, but even in those where it theoretically is not, it addresses concerns that certain groups are monopolizing court use for a very limited set of cases, while a vast ma- CHAPTER 1 jority of citizens are excluded for a variety of reasons. Patterns in outcomes of cases as related to the issue under dispute, the identity of the par- ties, or a variety of other characteristics--essentially who wins and who loses, and whether there are any distinguishable patterns here. These questions address a variety of concerns, 3 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts ranging from inherent biases and corruption to the argument that "repeat users" or those with access to private counsel (or any counsel at all) may have an automatic advantage. A review of the incidence of delay, where (in what cases and in which courts) it occurs, and the reasons for its occurrence. Delay may be a less serious problem than bias or corruption, but in the end its impact tends to be unequal. Generally, those with deep pockets can withstand its impact better, while for a party destined to lose, provoking delay can be a means of postponing or entirely avoiding "justice." Delay of course is a relative term,11 but where some cases take notably longer than others, or where presum- ably simple cases take years to resolve, it is worth investigating why. In addition to delay, non-resolution (a case that never reaches judgment, at least not within the time period covered) is also important--as it represents a possible, if often overlooked, miscarriage of justice as well.12 Questions about enforcement of judgments--the emphasis on reducing delay often stops with the delivery of a firm judgment (after all appeals have been exhausted). How- ever, an unenforced judgment usually has little value to the victor, and thus the problem of enforcement has recently begun to receive more attention.13 Unfortunately, this is one of the most difficult areas to investigate, as even in more developed judicial systems, the courts may not keep a record, unless they are directly involved in the enforcement process. These are the questions reviewed in the current research, in the order indicated above. Before getting into the analysis, a preliminary chapter describing the Ethiopian justice system is pro- vided for those not familiar with it. A final concluding chapter summarizes the most important findings and looks ahead to their implications for further justice reform in that country. Even with a very good statistical system as a basis for this analysis some questions simply could not be answered. Data on enforcement and, surprisingly, on legal representation, were not suf- 11 The definition of delay hinges on legal or practical standards being set as to how long cases should take to resolve. Because such standards do not exist in many court systems, it may be more accurate to speak of "duration" or time to judgment, but delay is more easily understood. 12 This was first noticed in an early World Bank study (2002) in Mexico which revealed that of 100 debt collection cases filed in the Federal District courts, only 20 reached judgment. The rest were ei- ther withdrawn or abandoned by the parties, and while it can be expected that some of these were settled out of court, it is nonetheless a concern that no one really knows what happened to them. This finding also suggests that judicial workloads are often lower than the statistics would indicate, as many of these cases cease moving forward very early in the process, although remaining in the courtrooms for considerably more time. CHAPTER 1 13 See World Bank (2002), Henderson et al (2004). We would note that in some cases, the winner does gain some benefit even from an unenforced judgment. This was noted both in Peru (Gonzales et al. 2002) and Colombia, where banks use the unenforced judgment to reduce their tax burden by counting it as a loss. In Colombia, banks play this outcome two ways ­ an unenforced judgment reduces their taxes, but a favorable judgment (with no mention of enforcement) increases their overall assets. 4 Introduction to Study and Explanation of Objectives ficient to allow the analysis the authors might have wished to do,14 but all and all, the database provides a rich resource for this and other kinds of studies. Two Notes on Usages Adopted Here Translating local conventions into a document to be read by non-Ethiopians posed a few quan- daries, which have been resolved, for better or worse, as explained below: Ethiopian calendar, judicial, and fiscal years--the years cited in the text correspond to the "Western" calendar, but it is important to recognize that the Ethiopian calendar (and fiscal) years do not follow "Western" practices. (For that matter neither do the months of which there are 13, or the dates which are 7 years different from the Western system--thus 2009 in the Western system is 2002 in Ethiopia). Moreover, the Ethiopian calendar year begins in early September while the fiscal year runs from the European July 1 to June 30. The judicial year co- incides with the fiscal year, meaning that, as Western dates are used in this report, each judicial year covers the respective halves of two Western years--e.g. 2008/2009. Names and bibliographic references--in Ethiopia, individuals are usually identified by their first names (taken from the second names of their father). This convention has been respected in the bibliographic references. Thus the entry for Andargatchew Tesfaye is entered under An- dargatchew and not Tesfaye. This convention was also used for Ethiopians residing outside the country who, in all probability, have adopted Western usage. 14 There were also problems, as the FSC was already aware, with the self-correcting function built into the database. This was convenient for the Court's immediate purposes, but created problems in reconstructing the trajectory of cases because, as one example, when a judgment was annulled, its existence disappeared entirely from the statistical record. The FSC Vice President who oversees the database continually, first noted this in finding cases that had lasted for years ­ something he thought had been resolved already. The problem was that since the initial, now annulled judgment, CHAPTER 1 had disappeared, the database made it appear that the case had simply been proceeding through the normal process at an extremely slow rate. Similarly, the number of judgments made by any court might suddenly be reduced for the same reason. The Court is taking actions to correct this problem as it interferes with its own performance monitoring exercise. To some extent, has also affected the statistics used in the present study. 5 CHAPTER 2 The Ethiopian Justice System Country Overview Ethiopia is arguably the oldest African nation and the only Sub-Saharan country not to have been colonized. Its history goes back millennia, although its current size is the result of the more recent (post 16th century) expansion of a historical empire formerly centered in the pres- ent-day regional states of Amhara and Tigray. Until the end of the last century, its governments were autocratic, making no pretense of democratic practices. The most famous of them were the 44 year reign of the Emperor Haile Salassie (1930­1974), and the military regime that over- threw him. The latter, termed the Derg (Committee), remained in power from 1974 to 1991, and was noteworthy for its extremely repressive policies. Under the Derg, many Ethiopia profession- als fled the country (or were imprisoned and/or killed). The overthrow of the Derg also resulted in the creation of the independent Eritrean state in what was formerly the Northeastern corner of Ethiopia (and the only part of the country to have been colonized, by the Italians from the late 19th century until the end of World War II). Despite its illustrious history, Ethiopia is one of the poorest countries in Africa and thus in the world with a per capita income of roughly US$420 ($779 when converted to PPP). Situated in the Horn of Africa and bordered by Eritrea, Sudan, Djibouti, Somalia and Kenya, its cur- 15 rent population is estimated at 85 million, divided among some 80 different ethnic groups. By population, the largest groups are the Oromo (about 32.1 percent of the population), the Amhara (30 percent), the Tigrayans (6.2) and the Somalis (5.9). Amharic is the official language for the federal government, and the most widely understood and spoken, but at least 84 distinct languages are used in the country, and some are making a resurgence in regions and sub-regions dominated by other ethnic groups. The 1994 Constitution recognizes the rights of different cultural groups to operate in their own languages and to continue with many of their other customary practices, so long as they do not violate the basic constitutional tenets. Inter- estingly, the Constitution also recognizes the right to secession from the federation by "every Nation, Nationality and People in Ethiopia" (Article 39:1) or the creation of their own separate regional-state. 15 Most of the figures cited here are very rough estimates and some are subject to considerable controversy ­ for political reasons as well as the absence of accurate information. When in doubt, fig- ures from US, Central Intelligence Agency (2009) have been used on the assumption that as regards Ethiopian controversies it is more neutral than many other sources. 7 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts The economy is largely agricultural, accounting for 80 percent of the economically active popu- lation, many of whom are engaged in subsistence farming. Literacy rates are low (only about 36­42 percent, depending on the source) as are many of the other basic social indicators. Trans- portation and communication are problematic for many areas of the country, and frequent droughts pose further problems, not only for food security but also for energy (85 percent of which is hydroelectric). Although rates of urbanization are increasing, over 80 percent of the population still lives in rural areas. Governance Structures Until the early 1990s, Ethiopia was a unitary country. The 1994 Constitution created a federal government, composed of nine regional states. Early on two city states (Addis Ababa, the capi- tal, and Dire Dawa) were accorded special status, partially within the constitutions (for Addis, Article 49) and by federal proclamation (law) for both. Over time various regions have also give a similar status to some larger municipalities. Inter alia, this allows them to create municipal courts, prosecutors, and police. There are the usual three branches of government, and the pat- tern is repeated in the regions, each of which has its own constitution, executive, assembly and judiciary. To hold the country together, the government has adopted a unique policy of "ethnic federalism" (Turton, 2005) under which to the extent possible, different ethnic groups have their own region, or within the region their own zonal (district) government. Regions are divided, in descending order, into zones, woredas, and kebeles. Woredas and kebeles have their own execu- tive and assembly; although the largest units, zones generally do not. However, judiciaries are either federal or regional, except for those connected to a few of the larger municipalities, and regional courts generally operate at the central (capital city), zonal and woreda levels. The government formed after overthrowing the Derg has remained in power since that date. The current Prime Minister Meles Zenawi, is finishing his third term with elections to be held again in 2010. The official party, the EPRDF (Ethiopian People's Revolutionary Democratic Front) operates in alliances with a variety of regional parties, and elections are contested by a number of opposition parties. In 2005, for example, the EPRDF lost control of the Addis Ababa govern- ment to an opposition coalition. The current political situation is only relevant to the present study as it affects judicial and broader justice sector autonomy. While the constitution guarantees judicial independence, and there have certainly been improvements in that area, some external and internal critics contend that political interference in judicial appointments and decision-making is a continu- CHAPTER 2 ing problem. Interestingly, among the judges interviewed in connection with this and an earlier report (Hammergren, 2009), those still on the bench (and some who had left it) reported that they had never been pressured, although some noted that this could be a problem for oth- ers--either as regards actual decisions, or more frequently, a judge's willingness to oversee a 8 The Ethiopian Justice System politically charged case. Nonetheless, the federal and regional governments have made serious efforts to expand access to formal institutions and to improve the quality of sector profession- als, through training, higher salaries, more transparent appointment systems, and better equip- ment and infrastructure. The Justice Sector Organization Overview: As is typical throughout Africa, Ethiopia has a relatively small formal justice sector, and it is estimated that up to 85 percent of the population rely on alternative, largely traditional mechanisms, to resolve disputes and provide security (Alula and Getachew, 2998). In effect, the country has several parallel justice systems: the formal federal and regional state structures, religious (Islamic) courts which at the federal level and in regions where they are recognized are overseen by the respective Judicial Administrative Commission; traditional or customary justice (associated with different ethnic groups or communities); and a series of more recent hybrid innovations, the most prominent of which are the social courts (operating at the kebele level and using a combination of formal and customary law) and the militias (a sort of community "police," also operating at the kebele level and largely in rural areas). As regards sheer quantity of units and members, these hybrid structures far outnumber the formal ones, and thus over the past few years, many regional judiciaries and police have made attempts to build linkages with them. Cases originating in the social courts can be appealed to the regional judiciary and, in cassation, can be reviewed by both the Regional and the Federal Supreme Court Although Ethiopia's number of judges (4.0 per 100,000 inhabitants) and prosecutors (at least as many and possibly more) is proportionately higher than in much of the rest of Africa (where the ratios are closer to 1 per 100,000) it is still far short of sufficient for covering the country's needs. Moreover, it is unlikely that the ratios can be raised to what might be considered an ad- equate level (more in the 10­12 per 100,000 range) for some time to come. This is not unusual 16 for Africa, unlike other regions (e.g. Latin America and Eastern Europe ) where state finances permit the placement of far larger numbers of police, prosecutors, judges and public defend- ers. Elsewhere citizens may use non-state mechanisms out of preference; in much of Africa it is 17 often a matter of a lack of alternatives. Preference of course also figures in Africa, and where state structures are located conveniently, citizens may still choose one of the alternatives out of 16 It bears noting that the judge to population ratio in Eastern Europe is somewhat higher (20 to 100,000 inhabitants) than in Western Europe. CEPEJ, 2005. It is not, however, evident that the larger CHAPTER 2 numbers are needed. 17 Because of Ethiopia's higher, but still inadequate judge-to-population ratio, preference may be a greater factor there. It was noted by one FSC source for example, that despite efforts to place more judges in the Somali and Afar regional states (arguably least developed in the country), citizens there still preferred their traditional mechanisms. 9 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts distrust of the state system, disagreement with or lack of understanding of formal procedures, or because they think the outcome there may be less favorable (Meron, 2009). This study will only focus on the state structures and principally the courts, but it is well to keep the preceding discussion in mind as a reference point. The Ethiopian government has made vast strides in expanding the reach of the formal system since 1991, but economic limitations have been a major constraint. It also bears mentioning that the post 1991 efforts to recreate the federal and create a regional justice structure have faced other impediments, most notably the lack of qualified personnel. Prior to the 1960s, most judges did not have a law degree but rather were "members of the clergy or persons well versed in the canon law of the Ethiopian Orthodox Church" (Andargatchew, 2004; 105). The same was true of prosecutors and there were no public defenders. Ethiopia's first law school (in the University of Addis Ababa) was founded in 1963, and despite the very recent proliferation of law schools and faculties (now numbering 23) the country still has a shortage of trained attorneys. This fact, plus the flight of many professionals under the Derg, and after its overthrow, meant that efforts to create the new federalized structure inevita- bly required the placement of many individuals with little or no legal training. Over the ensuing years, the government has made every effort to enhance the quality of personnel. Several regions now have judicial institutes that offer training to aspiring or seated judges and prosecutors with only a certificate (6 months of post High School legal training) or a diploma (two years) and over time the percentage of persons holding these positions without any significant formal training has gradually decreased. At present, all federal judges have at least a basic law degree; however, federal (and regional) prosecutors are not doing as well. A further problem is that the state has absorbed so many of the lawyers that it is estimated that judges (currently about 3,400 at all lev- els) outnumber the attorneys in private practice (estimated at 1,800 in 2005, but probably at least 500 more by now). This means, inter alia, that while citizens now have a better chance of airing their cases before a qualified judge, they may have to do so without an attorney to help them. Legal Tradition and Framework: Because of its parallel structures, Ethiopia can be considered to have a hybrid legal system, but it is also hybrid in the more traditional sense. With the re- drafting of its basic codes in the 1960s the country adopted a civil law system for its substantive laws, but based its procedures on common law models. However, since that time, it appears to have relied on local experts for any further revision (except for those cases where donors provided foreign experts to draft or redraft more specialized legislation). Ethiopia has not been impervious to extra-national trends in the further legal modifications, but it also has been de- cidedly eclectic as to what it decides to adopt. In any event, until the late 1990s, it was too busy with other issues to spend much time modernizing its basic legal framework. It was only in 2005 that it adopted two major pieces of legislation--a new Criminal Code and a Family Law. It CHAPTER 2 currently has drafts of a Vital Events Law, an Administrative Procedures Law, a Law of Evidence, and a Criminal Procedures Code nearly ready for adoption and there are some 15 other major laws under consideration. It should be noted that all of these laws are federal in nature, but that regions can, and in some cases must adopt them, with usually minor modifications. 10 The Ethiopian Justice System The two existing federal procedural codes (Civil and Criminal) and their regional variations, while based on common law (using Indian and British models respectively) practices, are clearly outdated and have not kept up with recent common law changes. This is especially problemat- ic for the Criminal Procedures Code which was drafted during a period when the English relied on police prosecution. As a result one of its major flaws is a reduced role for the prosecutors, 18 who may not even receive notice of a police investigation until it is completed. This is one of the reasons for an extremely low conviction rate, an enormous backlog in the prosecutorial of- fices, and quite probably, a certain amount of police corruption (as police can be paid to ignore a crime, overlook evidence against the major suspect, or alternatively to fabricate a case for 19 pay). Ethiopia arguably has more than enough prosecutors (is perhaps the only country where they often outnumber judges) but they are extremely inefficient, possibly in large part because of their minimal involvement in the police investigation. Whether the new Criminal and Family Laws constitute improvements remains debatable. Critics charge for example that the Family Law is written with an urban bias and thus ignores the dif- ferent situation (especially as regards the significance of divorce and the chances of remarriage) in traditional rural areas (Meron, 2009). However, it should be noted that the Family Law is a federal law, and thus was intentionally drafted for urban populations; regions adopting their 20 own variation on the model may thus be the ones who overlooked the bias. Whether or not this is true, it does appear that current drafting is overly dominated by lawyers and thus tends to ignore the context in which laws will be applied or the capacity of the relevant institutions to apply them. For example the draft Vital Events Law, while accompanied by a proclamation to establish a central registry does appear to ignore how the transition to new types and forms of registry (of births, deaths, marriages, etc) would be made at the local level. Drafting also incor- porates other errors--two consultants working on a SEC law noted that the new Criminal Code apparently made all sales of stock illegal as it omitted the term "fraudulent" in the description of what was outlawed. Two comments are merited here. First, the legal framework does need updating, but a bit more carefully. Second, careless updating is not uncommon elsewhere. Much of the criticism of the World Bank and other donors' efforts to reshape the legal frameworks in Eastern Europe and the former Soviet Union rests on the observation that absent appropriate institutions and 18 The Criminal Procedures Code does include language suggesting that prosecutors can guide the police investigation and requires judicial warrants for searches, seizures and arrests, but in practice prosecutors generally are not involved in most investigation while the requirement for warrants is frequently honored in the breach. CHAPTER 2 19 For an effort to tap corruption in Ethiopia's justice sector, see Hammergren (2009). 20 The Criminal Code has national effects, and thus regions have less flexibility in introducing their own modifications, which usually involve adding more crimes and infractions. Two examples given were Oromia's criminalization (as a misdemeanor) of not using seatbelts in cars and Tigray's similar inclusion of the infraction of using a cell phone while driving. 11 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts incentives, a new law is just a piece of paper (Gupta et al, 2002). In Bolivia in the 1990s when debtors' prison was eliminated, the drafters of the new Criminal Code simply forgot to substi- tute another sanction for those crimes to which it had applied. And then there was El Salvador which initially proposed that an 11 person Supreme Court be renewed by one-third every three years; fortunately the arithmetic error was corrected (by increasing the number of justices to 15, divisible by three) before the bill was sent to Congress. Structure: As noted Ethiopia's justice systems follow federal outlines, meaning that each regional state has its own constitution, secondary laws, courts, prosecutors, police, and public defenders (although this last category is in extremely short supply). Addis Ababa and Dire Dawa, while not having their own constitution have or can introduce the other actors. The judiciaries are formally independent, a third branch of government, overseen by Judicial Administrative Commissions 21 (JACs), composed of members selected by the judiciary and legislature (and sometimes with representatives from the private bar). The police and prosecutors belong to the executive, the police housed in the Ministry or Bureau (at the sub national level) of Public Security, and the pros- ecutors in the Ministry or Bureau of Justice. Prisons are also under executive control, at the federal level under a separate Prison Administration Authority, and at the regional level, most often as part of a Bureau of Public Security. The militias, where they exist, are also formally overseen by the respective Bureau of Public Security, but until recently have had no more direct formal links with the police. Social courts are regarded as executive bodies, a status which is itself a source of criticism. Their "judges" are lay persons, although in some regions they are expected to use proce- dures dictated by the Civil Procedures Code despite their lack of formal training in the latter. At both the federal and regional levels the courts (and the prosecutors' offices) follow a three instance pattern--headed by a Supreme Court (SC), with High Courts (HC, with both appellate 22 and original jurisdiction) and first instance courts (FIC) under them. Personnel management and other administrative matters are handled by the respective Judicial Administration Com- mission the composition of which varies from region to region. However, each Commission is usually presided by the respective Supreme Court President and has a majority of judicial members. This has not eliminated some complaints that they are dominated by the minority of "political members"--i.e. those chosen by the assembly. Jurisdictional division of labor between the federal and regional judiciaries is set by federal law (Federal Courts Proclamation 25/1996 and subsequent amendments). However, because the 21 While the executive does not have a formal right to name members, legislative appointees may CHAPTER 2 come from the executive, a consequence some observers hold to be a function of the parliamentary government system (Canada, National Justice Institute, 2008). 22 Addis Ababa and Dire Dawa's municipal judiciaries have only two tiers--a first instance and appel- late jurisdiction. Both have also created a cassation division within their respective appellate courts to review complaints about fundamental errors of law. 12 The Ethiopian Justice System federal courts until recently only functioned in Addis Ababa and Dire Dawa, some of the cases they would normally see were delegated to the regional courts. (On the other hand, in Dire Dawa and Addis Ababa the federal jurisdiction receives many cases that would normally go 23 to a regional court. ) Five new federal High Courts were created in 2003. Because of the small workload, their cases are still covered by judges located in Addis Ababa and Dire Dawa who periodically "ride circuit." Over time the organization of both the federal and regional courts has become more complex, with separate benches for at least civil, criminal and at the first instance, labor and more recently family issues. In Addis Ababa, the federal courts have also cre- ated first-instance bail and enforcement benches to facilitate and standardize these operations and "real time" benches to handle minor felonies. Over the past decade, most of the growth of the judiciaries has occurred in the regions, with the number of federal judges remaining fairly stable (at about 130) or even declining in some years. Most of the original and appellate (as opposed to cassation) caseload of the federal courts in fact originates in Addis Ababa and Dire Dawa. At least in the former city, the local (municipal) courts do not enjoy a high reputation for efficiency or professionalism and thus even the cases they might see are often directed by the parties to the federal courts. It bears mentioning that Ethiopia's judiciary does not enjoy constitutional review powers which are 24 instead exercised, at the Federal level, by the House of Federation. Thus, the Federal Supreme Court operates largely as an appeals court (for cases originating in the Federal High Court) or a 25 court of cassation (for those originating in both the federal and regional courts ), The House's constitutional review role remains controversial (Mamo, 2008), but as the Canadian National Justice Institute (2008) notes this type of arrangement if unusual is hardly unique. Ethiopia's prosecution service and Ministry (or Bureau in the regions) of Justice (where it is now located) have had a still more complex history. Under Haile Selassie, the Ministry had consider- able powers over the judiciary. In 1942 when it lost some of its powers (but not that for disci- plining judges), the Office of Public Prosecution was created within the Ministry. The Minister appointed all prosecutors except the Advocate General and his Deputy; both were appointed by the Emperor. The first Advocates General were foreigners; an Ethiopian did not hold this position until 1961. Until the late 1960s, most prosecutors were police officers because of the shortage of trained lawyers. The office of the Procurator (Advocate) General, while remaining in the Ministry of Justice, assumed more powers under the Derg, including extensive supervisory 23 This is foreseen in Proclamation 25/1996 (Articles 14.2 and 15.2), is also a function of the limited jurisdiction of the respective municipal courts, and can be the choice of the parties. CHAPTER 2 24 Technically a part of the legislative branch, the House of Federation does not have law-making powers, but rather exercises other, special functions, including that of constitutional review. 25 Regional (and municipal) courts also have cassation benches to review errors of law related to state or municipal matters. However, their decisions may be submitted for final review by the FSC cassation bench. 13 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts control over other agencies involved in law enforcement (prisons, the police investigators, and "the legality of the administration of justice:" Andargatchew, 2004; 70­71). The courts became less important (because of the creation of special tribunals; Aberra, 2002; 514) and their organi- zation was simplified. Proclamation 53 of 1975 instituted a Judicial Commission for the appoint- ment of judges, prosecutors and registrars, except for the Presidents of the Supreme Court, the High courts, and the Attorney General (all appointed by the Head of State pursuant to consulta- tions with the Minister of Justice). Perhaps because of the Derg policies, both the Ministry and the Office of Prosecution seem to have suffered a certain amount of neglect post-1991, both by the government and by donors. Ethiopia has an unusually large number of prosecutors relative to its complement of judges-- sometimes equaling or even exceeding the latter. However, the prosecutors tend to be less qualified (in part because the lower age limit is only 18 years as opposed to 25 for judges) and more poorly paid and equipped. Automation, which has firmly taken hold in the courts, is only beginning in the prosecutorial agencies, and the latter still lack an information system to track caseloads and performance. This makes it difficult to assess their productivity, but the statistics that are available (for the most part collected manually) indicate a very low conviction rate (33 percent for the federal level) and a very large backlog. In the past couple of years, the federal and regional ministry/bureaus have been attempting to address their shortcomings through a business process reengineering (BPR) exercise, the introduction of an "integration model" to improve coordination with the police, and the adoption of some of the administrative innova- tions introduced by the courts. However, despite all this, most of the agency heads interviewed continued to argue that they needed substantially (up to triple the current amount) more personnel to get the job done. Police forces follow the same federal, regional, and occasionally municipal organization. Nationwide, numbers of police were impossible to ascertain, but it appears total staffing is less than 100 per 100,000 inhabitants, far below international benchmarks (between 250 and 300). The federal forces tend to focus on crimes of "national interest" (involving national secu- rity, foreigners, terrorism, organized crime and so on) while the regional and municipal forces handle more mundane issues. (One exception is the Addis Ababa police force, which unlike its courts, deals with more serious threats to national security in addition to ordinary crime and such minor matters as traffic violations). The regional and municipal, but not the federal police have also begun to adopt community policing policies although at the moment it is not clear what this means beyond trying to place at least one police officer in every kebele (or even every community). As noted above, under the existing Criminal Procedures Code, the police conduct criminal investigations with little if any direction from the prosecutors, a CHAPTER 2 situation regarded as problematic, but without a single solution as yet. Either because of cor- ruption, incompetence or simple work overload, police failures to bring witnesses or even the defendant to court lead to many criminal cases being temporarily closed (and as discussed in later sections, possibly never reopened). 14 The Ethiopian Justice System Public defense remains nearly non-existent in Ethiopia, largely because of financial constraints. Only those accused of crimes subject to serious (including capital) penalties are given subsi- dized assistance by the state. Some NGOs provide additional assistance both for criminal and civil cases, but a majority of the parties to all types of cases continue to work without an at- torney. Many defendants in criminal cases must thus organize their defense on their own, often while in pre-trial detention. Given resource constraints and other priorities, it is unlikely this situation will be improved for some time. 26 Procedures: As noted the procedural codes are based on the British version of the com- mon law system (for the Civil Procedures Code, as translated through the Indian CPC), but adopted that model in the 1960s without keeping up with later changes. A mentioned, this is most apparent in criminal law; the 1961 procedures code accords an independence to police investigation no longer enjoyed under the English system (which introduced a crown prosecution office in the 1980s and gave it the ability to direct the police actions). In civil law, pleadings are fact rather than code or notice based (a trait shared with continental Europe as well, and now being recommended for the U.S. whose federal and some state courts adopt- 27 ed notice pleading in the middle of the past century ). The Civil Procedures Codes covers all non-criminal matters (thus labor, administrative, commercial, family and other civil issues). Despite the English "adversarial," influence, both codes accord the judge certain "inquisitorial powers" especially as regards the ability to proceed with a civil case even when the defen- dant accepts (or at least does not oppose) the facts in the plaintiff's claim, and to take a more active role in the pre-trial and trial stages than is usually associated with the common law model. Proceedings in both criminal and civil matters feature oral hearings at both the pre-trial and trial stages. There are certain expedited procedures (summary and accelerated proceedings 28 in civil cases, and the practices adopted in the new real time criminal courts ), but in legal theory even ordinary cases should be relatively simple. In civil disputes, there is a first prepa- ratory written stage involving the exchange of pleadings (including the plaintiff's statement of claim, the statement of defense, and the submission of lists of witnesses and other docu- ments by both); a first hearing to frame the issues and decide on preliminary objections; a second hearing or trial stage, and a judgment. All FFIC and all FHC non-appellate civil cases 26 Much of the material summarized here is based on Menberetsehai (2009). 27 However, the earlier US fact and writ-based system is not being recommended, but rather a mod- ern, simplified version. See Love Kourlis et al (2009). In theory this is what Ethiopia has, and thus may CHAPTER 2 represent an improvement over the system in place in England in the 1960s. However, as critics note (Menberetsehai, 2009), judicial interpretations of the procedural rules may well undermine some of the presumed benefits. 28 Although real time courts are gradually taking on a larger proportion of criminal cases, the expe- dited civil proceedings are apparently infrequently used. (Menberetsehai, 2009) 15 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts are heard by individual judges, although most of the preparatory matters may be attended by administrative staff. For criminal cases, the judicial part of the process (once the police investigation is completed and the prosecutor has prepared his/her case) should be still simpler--a preparatory stage in which the charges are filed, the defendant is served (if not already detained), and witnesses are summoned; a trial stage, and a judgment. The criminal trial is intended to be (as defined in the code) a continuous and primarily oral event. According to a local expert (Menberetsehai, 2009), both civil and criminal proceedings deviate considerably from the legally defined practices, with far greater reliance on written submissions, far less continuity, and far more intermediate steps than the codes anticipate. This, it is argued, is one principal cause of delay and is attrib- uted to judicial practice and judges' unwillingness to pressure the parties into a less drawn out mode. Interlocutory appeals are not allowed, but either party can appeal the initial judgment or sentence, and if not satisfied with the outcome, can also request a cassation review (for funda- mental errors of law) by the Federal Supreme Court. (Regions may have cassation divisions in their Supreme Courts, and their decisions can also go to the FSC cassation bench.) Appellate courts may refuse a full hearing, reviewing all factual and legal issues, to an appellant but they must accord him/her an audience before deciding whether to go further. This also takes time and may encourage the courts to simply proceed to the next step. Cases can be submitted for a second appeal (to a higher level) only when the first appeal overturns the initial judgment. Appeals are not supposed to postpone enforcement of civil awards, but temporary injunctions may be requested. Moreover, enforcement in increasingly a problem, and one further defense mentioned by court sources was the sudden appearance of a third party protesting a proprie- tary interest in assets to be seized for sale or turned over to the winner. When this occurs, a firm judgment may be nullified and the case reopened. There are a number of other factors that may encourage parties to pursue litigation despite slight chances of receiving a favorable judgment. Court fees are relatively low for first instance cases and even lower for appeals (50 percent of the initial fee for the first appeal and half of that for a second). There is no fee for cassation. While lawyers' fees can be high, legal representation is not required for any party, even in cases going to the Federal Supreme Court. Judges can award costs to punish an abusive litigant, but it appears they are usually awarded as incurred (i.e. each party pays her or her own costs). In criminal cases, while prosecutors by law can decide not to pursue an indictment, this requires a written justification (and can be contested by the victim or the victim's representative). This encourages prosecution of hopeless cases CHAPTER 2 (without sufficient evidence or possibly not meriting attention for other reasons) as the easier route. In a country where so few citizens use the courts, the absence of more stringent filtering mechanisms and other disincentives may still be regarded as positive, but it does threaten to swamp the judiciary with cases with which it can do little. 16 The Ethiopian Justice System Some Additional Details about Case Processing, Especially as Related to the Further Analysis Some of these points will be repeated in the relevant sections below, but they are consoli- dated here for easy reference. Most of these observations are the result of an initial analysis of the data provided by the court and some apparent anomalies in the contents. Some of the latter may represent errors in data entry, but most are a result of certain practices peculiar to Ethiopia. Land cases--in Ethiopia all land is owned by the state which means that land cases do not involve ownership but rather, in rural areas, right to till, and in urban areas, right to other types of uses. Land cases are nonetheless important, and many of the rural cases make it as far as the Federal Supreme Court (where they are often dismissed for lack of merit inasmuch as the parties rarely have a lawyer to advise them on the legal issues). Other property cases, especially in urban areas, most often involve ownership of houses and other buildings. In rural areas land cases often originate in the social courts, but there is a move to shift them to the formal system. In urban regions like Addis Ababa, they may be first handled by special administrative tribunals with decisions appealable to the formal system. Both extra-judicial arrangements have been criticized for possible political interference and corruption. Labor cases--these are the only non-criminal cases in which the amount at stake is usually not included in the database, largely because the plaintiff commonly makes a series of demands, many of which will not be admitted. Despite this, it is contended by informed observers that labor judges, especially in first instance courts, tend to favor the worker in their decisions. This might explain why although plaintiffs at the first instance are nearly solely individuals, those appealing are usually organizations (the losers on the first round). Administrative cases--cases to which the government is a party go to the civil courts, although in some instances, only after they have been reviewed by administrative tribunals. In those cases, issues seen by civil courts are limited and usually do not enter into the substan- tive content of the administrative decision. Representation of the state is usually handled by specialized attorneys attached to the affected ministries or state agencies. Corruption cases are heard by criminal courts and are usually prosecuted by lawyers working for anti-corruption commissions at the federal or state level. Legal representation--while many parties to litigation operate without a lawyer, some who CHAPTER 2 do have legal representation do not have this registered with the court. The explanation given was that lawyers prefer not to be registered for tax purposes. They prepare the brief or plead- ing but have the client present it under his or her own name so as not to have the payment for services officially recognized. 17 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts "Split decisions"--a number of the cases with judgments indicated a "split decision"--i.e. neither the plaintiff nor the defendant won entirely. This apparently is a consequence of their having had the amounts at stake registered at the beginning. In short, the judge might rule in favor of the plaintiff, but would not give them everything they were requesting. Quite prob- ably some change in the registration of outcomes is required so that it would indicate that the plaintiff's demand was recognized but simply not at full value or that in labor cases only a part of the claimant's pretensions were met. Declaratory Judgments--a significant portion of the civil caseload (estimated at 28 percent) of first instance courts comprises non-controversial cases requiring judicial recognition of a le- gal situation or agreement reached by the parties. The result is termed a "declaratory judgment" and while it closes a case, is not counted a "judgment" within the database. Dismissed or closed cases--these represent a large category in the database, a subgroup of cases that are temporarily or permanently dismissed. What is included in this group differs by instance. At the FSC level, most "dismissed" (or "discarded") cases are not admitted for lack of merit. However, at lower levels many of these cases are temporarily closed, most often be- cause of excessive adjournments. These cases may be reopened, and one question that will be explored later is how many of them actually are since for someone attempting to "avoid justice," this may represent an interesting way to end movement forward. A good part of the backlog accumulated by prosecutorial agencies apparently is composed of cases closed temporarily be- cause the witnesses for the prosecution or the defendant did not appear for a hearing. Closing cases temporarily allows judges to increase their disposition rate, but it clearly can also obstruct justice. Enforcement or Execution cases--enforcement of judgments is increasingly recognized as a universal problem inasmuch as a party derives little satisfaction from a favorable judgment the other party does not honor. The Ethiopian federal courts have opened enforcement benches to handle cases not executed spontaneously or which, in the case of assets seized for sale, require judicial supervision. As enforcement cases handled by the special or ordinary benches repre- sent a substantial percentage of the workload in all instances, they to some extent constitute a sort of double counting--cases which have been counted once before judgment, and now are counted again afterwards. However, given the real work this does represent, the extra count is justified, and furthermore, unlike many countries that ignore the issue, the Ethiopian courts are now working on remedies as discussed in a later chapter. CHAPTER 2 18 CHAPTER 3 Overview of Recent Trends in Demand for and Supply of Federal Court Services Introduction With this chapter we begin our analysis, looking first at overall trends in case filings (demand for services), by aggregate number and by category, and in case dispositions and other aspects of supply. Data come only from federal courts, but in both Addis Ababa and Dire Dawa, the federal judiciary sees many "normal cases" because of the restricted jurisdiction of the munici- pal courts (and most probably, citizens' greater confidence in the efficiency and efficacy of the federal system). The Federal High Court (FHC) receives some original jurisdiction cases from outside the two cities and also receives a few regional cases on appeal. The Federal Supreme Court's cassation bench takes cases from both the federal and regional jurisdictions. Thus, while not representative of all cases filed nationally, the federal courts still cover a reasonable selec- tion of the types of cases seen by all the country's court systems. Here we are particularly interested in several factors: changes in the level of demand for court services, changes in the composition of demand, and changes in the courts' ability to respond to both. Generally, we would expect demand for services to increase over time, thereby putting pressure on the judiciary's ability to respond. This seems to be a universal tendency, even in countries whose court systems are not regarded with great confidence. As public confidence and ease of access increase, demand tends to grow still more rapidly. This is the downside of improving judicial performance and a good example of the conflicts inherent among the usual objectives of judicial reform programs--greater access, efficiency, and quality. There are, of course, many other factors accounting for increased demand--economic development, eco- nomic downturns, societal change and especially the weakening of traditional means of norm enforcement, and so on. The differential impacts of these exogenous factors should also be reflected in varying growth rates among types of cases. Where, as in Ethiopia, courts have spe- cialized jurisdictions, this poses an additional challenge of ensuring a balance between types of demand and types of supply. In many court systems worldwide, these tendencies have led to a demand-supply gap and increases in both backlogs and delays, either across the board or in different levels or types of courts. In Ethiopia, the negative consequences appear to have been avoided in large part. The statistics demonstrate that the federal courts have been able to increase their production (and 19 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts productivity), thus keeping up with the higher demand for their services, and even reducing accumulated backlog. Although the number of judges has been increased, the success in re- sponding to higher demand appears to be largely a result of the Supreme Court's ability, thanks to its database, to track judicial output, delays, and factors like number and length of adjourn- ments. The FSC has set targets for all these areas and monitors judicial compliance--especially as regards clearance and congestion rates, average age of disposed cases, and number of adjournments. Targets have gradually been adjusted to more demanding levels--for example where once an average of five adjournments per case was considered reasonable, the target is now two. The emphasis on productivity and speed may have a downside, as explored in later chapters, but it clearly has had a major positive impact on this dimension of court performance. Although not covered here, it bears mentioning that the federal courts' methods for increasing efficiency and thereby responding to growing demand are gradually being adopted by the re- gional and municipal judiciaries and that several of them have sufficiently developed databases 29 to allow this same type of analysis there. The increase in caseload is still more dramatic in the regions, and as a result their judiciaries have grown both in number of judges and numbers of courts and benches. The intent is not only to keep up with rising demand, but also to increase access, and as at the federal level, the approach includes several mechanisms beyond track- ing productivity--use of video-conferencing and hearings, introduction of information desks, and a move toward internet filing. Some of these mechanisms only facilitate access for normal court users, since they require familiarity with and access to modern equipment (internet, cell 30 phones ) most of the population may not have. However, efforts to spread court presence by constructing new units or using mobile facilities are aimed at drawing in those who would not access the formal system otherwise. Growth in Demand (New Filings) and Distribution by Major Types of Cases (Criminal, Civil and Labor Benches) A supply-demand gap with the accompanying problems of delay and congestion can be the re- sult of years of slightly deficient production and the slow accumulation of a substantial backlog, or the consequence of rapid increases in demand and the courts' inability to adjust to it quickly. For Ethiopia's federal (and regional) courts it appears that both factors have been at work. Thus, as we will see in the following sections, remedial actions have had to take two tracks, raising production and productivity to respond to higher levels of demand, but at the same time, taking CHAPTER 3 29 The authors had initially considered adding a regional court, but this proved too complicated logistically. 30 Despite the well publicized cell phone revolution throughout much of Africa, cell phone use in Ethiopia is very limited--estimated at less than 4 percent of the population. Costs and infrastructure are two major limitations. 20 Overview of Recent Trends in Demand for and Supply of Federal Court Services additional actions to reduce the backlog that had accumulated from earlier years when less at- tention was being paid to keeping courts up to date. To a greater or lesser extent, the two tracks are usually needed, as backlog reduction only deals with historical, not new cases (and many of the former may have gone inactive years before, thus not affecting current operations) while delay reduction typically only affects new cases, leaving the mass of pending files untouched. The simplest way of measuring demand is to track the number of new filings entered each year. There are nonetheless some caveats as regards this technique. First, new filings may be heard only at one instance or they may generate applications for appeals and cassation. Counting their entrance at each level of the judicial hierarchy thus risks double or triple counting of cases, but may be a more accurate reflection of real work (and real demand, which logically would include requests for review of first instance decisions). In order not to under-represent the workload of the High and Supreme Courts we have thus tracked "filings" at each level, but also provided additional graphs to distinguish between those received as original jurisdiction (start- ing at that level) and those undergoing review on appeal or in cassation. Second, especially when reviewing caseload per judge, the fact that some cases are reviewed by panels has led to objections (in other countries) that it would be fairer to look at caseload per panel. This makes it very difficult to compare across judiciaries, for which reason we have simply calculated workload per judge irrespective of how judges are organized to decide each case. Since at the High Court level, most cases are heard only by one judge in Ethiopia, this is less a problem there, and in favor of the technique it is a better representation of use of human resources. Whether or not judges review cases in panels, individually, or sequentially (different judges handling different stages of a case at a single level--for example, in Ethiopia the desig- nation of a specialized bench to hear questions of bail or enforcement) the fundamental issue is 31 how many judges it takes to hear X number of cases, not how they are organized to do it. Third, workload also depends on the complexity of proceedings something over which the judi- ciary has some but not total control. We have noticed in other countries that complaints about an excessive litigation rate often seem exaggerated when compared with international statistics. Aside from a failure to do such comparisons (which would demonstrate that 4,000 cases per 100,000 inhabitants is, for example, relatively low), another explanation may be a focus not on the number of entering cases (usually calculated as cases per 100,000 inhabitants), but on what 31 In reviewing criminal cases in Latin American countries, this is also an issue as the new criminal CHAPTER 3 procedures codes assign different stages of a first instance case to different judges--a "guarantee judge" to hear pre-trial motions, a trial judge to conduct the trial, and possibly an enforcement (or execution) judge to oversee issues regarding those sentenced to prison. In Latin America, this has arguably led to an inefficient proliferation of judgeships especially because the number of cases reaching the judiciary has been decreased due to the prosecutor's exercise of the "principle of op- portunity," (ability to decide not to pursue an investigation or indictment). 21 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts happens to them after admission--whether complexity expands the work required for process- 32 ing each of them, or whether they can be handled quickly and directly. Finally, this measure of "demand" does not take into account backlog or pending cases from prior years. This omission is in line with conventional practice, and the common definition of demand. It will be considered, however, in later sections on "supply" as backlog reduction is Figure 1: New Filings (Original Jurisdiction Only) from 2005 to 2009; First Instance and High Court* 90000 80000 70000 60000 50000 40000 30000 20000 10000 0 2005/06 2006/07 2007/08 2008/09 FHC FFIC * The Federal Supreme Court is omitted because its original jurisdiction caseload is virtually nil. Figure 2: New Filings (Total = Original Jurisdiction + Appeal + Cassation)* 90000 80000 70000 60000 50000 40000 30000 20000 10000 0 2005/06 2006/07 2007/08 2008/09 FSC FHC FFIC CHAPTER 3 * Cassation is only heard by the FSC and for it is more common than appeals. For the FHC both original jurisdiction and appeals are included. 32Throughout Latin America, real litigation rates are generally low to moderate by international standards. However, even simple cases may be strung out unnecessarily, as part of what one author (Uzelac, 2008) calls the Mediterranean Model. 22 Overview of Recent Trends in Demand for and Supply of Federal Court Services a part of that dimension of judicial performance. In later chapters, we will include "reopened cases." This category is especially important in Ethiopia because of the high number of cases initially closed without judgment. In the accompanying charts we have shown both original jurisdiction cases and where relevant, appeals and cassation. Thus for the FFIC (Federal First Instance Courts) all cases are original jurisdiction; for the FHC, appeals are also added, and for the FSC, original jurisdiction cases do not appear (as they are extremely rare and average less than one a year). For both the FHC and the FSC, appeals constiute a large share of the caseload, as is evident in the change in FHC workload when they are added. Another interesting facet of the FSC work- load is that admitted cassation cases far exceed those seen on appeal. A principal explanation is that whereas appeals come largely from the federal courts, cassation cases come from the regions (and even the social courts) as well. Table 1. FSC Appeal Vs Cassation (includes only cases admitted) Year FSC 2005­06 2006­07 2007­08 2008­09 Appeal 1240 1480 1955 2301 Cassation 3955 4266 5460 5991 As the above charts make apparent, even over the short time period covered here there has been a substantial increase (a near doubling) in "demand" for all three instances. This has also produced an increase in average caseloads per judge as shown in the figure and table below. Figure 3: Average New Filings per Judge (Original Jurisdiction) 2005 to 2009 1200 1000 800 600 400 CHAPTER 3 300 0 2005/06 2006/07 2007/08 2008/09 FHC FFIC 23 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts Table 2. Average New Caseload Per Judge (Total= Original Jurisdiction + Appeals + Cassation, as applicable) 2005­06 2006­07 2007­08 2008­09 Court Filings Per judge Filings Per judge Filings Per judge Filings Per judge FSC 5195 325 5746 319 7415 221 8265 344 FHC 7641 170 9108 175 9108 166 13237 241 FFIC 48115 891 51536 859 65452 747 80461 1018 The increase in the caseload of the FSC has been least dramatic. That for the High Court judges has been proportionally greatest, but the current 241 new entries per year hardly seem exces- sive, while the 891 and now 1018 entries per FFIC judge are reaching appreciable amounts. Of course caseload numbers do not tell the whole story as much depends on the complexity of what is being seen and the normal trajectory for its processing. In any event, the bottom line is that the federal judiciary's aggregate workload has increased considerably and thus that judges will have to make adjustments to keep up with it. For curiosity's sake we have also calculated the litigation rate for the Federal Courts, based on the number of new filings per 100,000 inhabitants. The increase is also substantial, although as our base is Ethiopia's total population, the rate appears relatively low. Given that most cases for the first instance and high courts originate in Addis Ababa and Dire Dawa it would probably be fairer to use the combined population of those two cities rather than the national population as the base. Finally, we have separated out appeals to the High and Supreme Court as a subcategory of demand, looking here at appeals rates rather than absolute numbers. Appeals rates are also Figure 4: Litigation Rates: Number of New Filings per 100,000 Population by Year 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 CHAPTER 3 0.1 0.0 2005­06 2006­07 2007­08 2008­09 FSC FHC FFIC 24 Overview of Recent Trends in Demand for and Supply of Federal Court Services an indicator of "litigiousness" in the second sense discussed above--not people's tendency to go to court but rather their "argumentativeness" once they get there. Here we only consider "applications to appeal" as a sign of demand. In a later chapter we recalculate appeals rates in a second fashion--based on the appeals actually admitted. Appeals rates are difficult to calculate, and the database does not do so. Consequently, we have approximated the rates by taking the number of appeals over the number of judgments in the lower level court. This is far from ideal, but an accurate calcuation would require linking appeals to individual cases since one cannot assume that the base year for judgments captures all the cases from that year that are appealed. Because of these difficulties, comparative appeals rates are hard to come by. The studies done in Latin America (Mexico and Brazil) found relativ- iely high rates for Mexico (30 percent for judgments on simple debt collection cases) and very high rates (over 100 percent) for the same types of cases in Brazil. Rates over 100 percent are possible there not only because both parties may appeal, but also because each may attempt several types of appeals. The few Western European countries on which data were available suggested appeals rates for all civil cases of about 10 percent. In Western Europe (and the United States, Canada, and Austrailia) there is a tendency to restrict appeals--by being very rig- orous as to the requirements for admission (or granting "leave" to appeal.) In most developing countries such policies have yet to be adopted meaning that the only effective restrictions on appeals are the costs of taking the case further. It is widely acknowledged that while appeals are important to correct procedural errors, they are most commonly used as a dilatory measure. A threat of an appeal can also be used to negotiate further with the winning party--as occurs in what is called post-judgment bargaining in the United States., a process whereby the winner may agree to accept less than the award made by the judge so as to avoid further delays (and costs) before collecting. Figure 5: Federal High Court Appeal Rates, Appeals Requested (percentages based on lower court judgments) by Major Legal Area 60 50 40 30 20 10 CHAPTER 3 0 2005­06 2006­07 2007­08 2008­09 Civil Criminal Labor 25 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts Figure 6: Federal Supreme Court Appeal Rates, Appeals Requested (percentages based on lower court judgments) by Major Legal Area 50 45 40 35 30 25 20 15 10 5 0 2005­06 2006­07 2007­08 2008­09 Civil Criminal Labor Several observations can be made as regards appeals rates. First, the rates (for applications) have tended to increase over the period covered, nearly doubling in a few cases. Although hardly unique to Ethiopia, one factor adding to the high rates is the potential for co-defendants (or co-plaintiffs) to file separately. Second, they are high to extremely high for the most part, the only exception being labor cases appealed to the FSC from the High Court. (Since these would necessarily be second appeals, the prohibitions on the latter--only possible when the first appeal overrides the initial ruling--explain the low rate.) The judiciary regards both trends as problematic, especially given (as discussed in a later chapter) the very low reversal rates (which is to say that the higher court upholds the lower court's judgment, a situation which precludes the possibility of a second appeal, but not of cassation). However, it is still seeking ways to bring the rates down since in its interpretation of events, appeals are predominately a delaying 33 tactic. Third, the rates vary among the legal materials between the two courts. Since rates are percentages this has nothing to do with the quantity of judgments that might be appealed. It probably has more to do with the types of cases seen within each instance, the identity of the parties, and the issues at stake. We will explore these issues more in later chapters, but one hy- pothetical explanation for the very high rates of applications to appeal for labor cases coming from the FFIC is a purported tendency for judges to decide in favor of the plaintiff (the worker) and thus for the loser (the employer) to protest. Since a second appeal (to the FSC) could only be requested if the FHC judgment reversed the FFIC ruling, this would account for the very low rate of requests to the Supreme Court. However, to flesh out these speculations it would be important to know who registers the appeal and that information is largely lacking. CHAPTER 3 33One wonders however to what extent this is really the case. Given the low cost of entering an ap- peal (and the null cost for cassation) it may simply be that people are willing to take their chances, either because they do not know how low they are or because they figure they might be among the 5 percent who earn a reversal. 26 Overview of Recent Trends in Demand for and Supply of Federal Court Services In summary, as the above tables indicate in a variety of manners, even over the short period covered here demand for services of the federal courts has increased, nearly doubling in all three instances. While, as discussed below, the numbers of judges have been augmented, the average caseload per judge has also increased, if less dramatically. Somewhat surprisingly, caseload per judge has remained lowest at the High Court Level. At the first instance (FFIC) it is clearly approaching what might be considered a maximum feasible amount. For simple cases, 1,000 new filings per judge is not necessarily unreasonable, but does require good case management and an exercise of control over parties and counsel to ensure their compliance with deadlines and discourage their efforts to complicate processing needlessly. Courts in other countries have been able to handle even substantially higher workloads, but only with the help of techniques and technology (high levels of automation; batch processing of similar cases; delegation of many routine matters to highly qualified courtroom staff, and so on) that the Ethiopian courts do not enjoy. Automation began in the federal courts, but was first used largely for recording court actions, and while the situation is improving, many staff members 34 and even some judges still do not have their own computers to do other work. Given that workloads for the FFIC in particular were already high in 2005, the rapid growth since then has thus posed a challenge for the courts and motivated a series of measures to help address it by working on the supply side of the equation. Trends in Supply of Court Services: Dimension One, Changes in the Size of the Federal Court System The "supply" of court services is determined by two factors--number (and location) of court units and judges and their efficiency in processing cases. This section only reviews the first factor; the next section looks at efficiency or productivity. Courts attempting to respond to increased demand can work on either of the two dimensions although in many court systems there seems to be an expectation that efficiency cannot be increased and thus that supply can only be augmented by adding judges. Ethiopia appears to be an exception to this rule, having emphasized both dimensions, and at the federal level, putting still more emphasis on efficiency. Here we first look at the number of court "units"--divided into courts and benches. In Ethio- pia, the term "court" refers to a physical facility in which a number of benches and judges are located. In the FFIC, the number of benches coincides exactly with the number of judges, but they are located in 9 to 12 physically separate court facilities. For the High Courts and Supreme CHAPTER 3 Court, judges are grouped into benches (or divisions, a term often used in translation) special- 34At present, all FSC judges have computers as do 70 percent of the FFIC judges. At the High Court level, coverage is lower ­ only 8 percent. Another problem, outside the judiciary's control, is occa- sional blackouts, meaning that courts cannot afford to be completely automation-dependent. 27 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts Table 3. Number of Court "units" (First Instance, High Court, and Supreme Court, courts and benches from 2005 to 2009) 2005­06 2006­07 2007­08 2008­09 Court Court Benches Court Benches Court Benches Court Benches FSC 1 3 1 3 1 4 1 3 FHC 2 34 2 34 2 31 2 31 FFIC 12 54 11 60 10 69 9 79 izing in one material or another, although most cases decided by the High Courts are seen by only one judge. As is apparent from the above table, the number of courts has stayed fairly stable or even declined, while the number of benches has only increased at the First Instance (FFIC). In light of increases in demand, this might appear nonsensical, but there are some simple explanations. First, at the FSC and FHC level, the number of "courts" is fixed--one Supreme Court for the entire country and one High Court for each region where the federal system operates directly. In reality, although not shown here, there are now an additional 5 High Courts, created in five more regions as of 2003. Until then, the Federal High Courts only functioned in Addis Ababa and Dire Dawa (with most of the Federal First Instance Courts also physically located in the former). To date these new High Courts still use judges riding circuit from Addis Ababa and Dire Dawa--they simply do not receive enough cases to merit making separate appointments. There apparently has been no perceived need to create federal first instances courts in these regions as their own court systems adequately cover any demand. Second, the reduction in the number of First Instance Courts and of High Court benches has a more mundane explanation--problems with infrastructure (demolition or poor condition of buildings which forced their closure). Meanwhile, the addition of benches (in the FSC and FFIC) is a more direct response to policies to handle demand. The temporary creation of an extra Su- preme Court bench in 2007­08 was a measure to reduce backlog on the civil bench. Once the aim was accomplished the additional bench was closed. At the FFIC level, benches (and thus judges) have been added to deal with rising demand and to implement a policy of introducing specialized jurisdictions (creation of a bail and an execution bench, and of real time, tax, family, and sex offender benches among others). Table 4. Number of Federal Judges (by type of court) from 2005 to 2009 CHAPTER 3 Court 2005­06 2006­07 2007­08 2008­09 FSC 16 18 26 24 FHC 45 52 55 55 FFIC 54 60 69 79 28 Overview of Recent Trends in Demand for and Supply of Federal Court Services Addition of benches has also meant addition of judges, although most notably at the FFIC level where each judge constitutes a bench. As with the prior table, the yearly cut-offs do not reflect the changes perfectly--especially where reductions in numbers are caused by the retirement or resignation of judges and delays in replacing them. For example at the end of 2009, there were only 20 justices on the FSC and the number of High Court judges has also fallen. In the latter case this is because of resignations caused by the impending enactment of a law that would require a two year waiting period before judges could resume private litigation activities. Since a successful lawyer can earn far more than a judge, many Ethiopian judges have traditionally regarded their stay on the bench as a first step to private practice, allowing them to get experience, make contacts, and not incidentally, become accredited as a lawyer without passing a bar 35 exam. This reverses a pattern seen in other countries--private practice first, and then, in later years, appointment to the bench. The pattern in Ethiopia is not only unusual; it is also regarded as problematic because ex-judges may use their contacts with friends on the bench to enhance their chances of winning a case. This is the principal motivation for the new law, and as the exit pattern suggests, one which may have interfered with some tradi- tional expectations. Aside from these coincidental influences, the changes in the number of judges over the 4-year period represent a response to differences in the patterns of demand. This is also vis- ible above in the charts on the average workload per judge over the same period. At both the FSC and FFIC levels, the number of judges has been augmented by about 50 percent, producing a fairly stable per judge workload in the FSC and an increase of about 20 percent in the FFIC. For the High Courts, where the per-judge workload started and remains relative- ly low, the increase in judges has been less dramatic, about 20 percent over the same period. The Supreme Court and the Judicial Council can reach these decisions on their own, subject only to budgetary constraints. They can create new courts, benches, and judgeships without formal approval from the other branches of government (although they of course need the additional funds to do so). The addition of judges and the emphasis on productivity have kept the average workload and disposition rate per judge more or less at a constant level, suggesting inter alia, that the Court has been making its decisions with these objectives very much in mind. CHAPTER 3 35 Once an individual earns a law degree (and formerly without even that step), he or she can be ac- credited as a lawyer on the basis of five years of experience or passing an examination administered by the Ministry of Justice. For whatever reason, most law students have traditionally preferred the five years of practice to the examination. Menberetsehai (2009) provides figures showing that of the 1,800 or so accredited attorneys registered with the Ministry in the early 2000s, a minority did not have a law degree and a few lacked any legal training. 29 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts Trends in Supply of Services: Dimension 2, Disposition, Clearance and Congestion Rates Here we look at the second dimension of supply, increases in productivity or efficiency. There are several common ways of measuring this--average number of dispositions per judge; clear- ance rates (number of dispositions as percentage of the number of new and reopened filings 36 for any given year ); average time to disposition; and congestion rate. The latter is calculated in a number of ways, but that used in Ethiopia is as follows: Congestion Rate = [(number of pending cases plus new and reopened filings) divided by annual dispositions] minus 1 Here we will only use clearance and congestion rates. Both can be expressed as percentages or ratios. Following Ethiopian practice, we use percentages for the former and ratios for the latter. For clearance rates a percentage of 100 or above is good as it at least ensures judges are not falling behind (or enlarging the supply-demand gap). For congestion rates calculated in this manner, 0.0 is the lowest (best, and nearly impossible) number, indicating that courts are clearing their desks each year. Ratios over 0.0 signify that cases are being retained although numbers between 0.0 and 1.0 probably mean that some 37 backlog reduction is occurring, while those over 1.0 signify that backlog is accumulating. Ide- ally, once a backlog reduction program has begun, the congestion ratio should decrease over time, meaning that we are less interested in a single number than in trends. As we have already examined workload per judge and will not look at delays until later, these calculations are not reviewed here. As many other countries have discovered to their dismay, simply increasing the number of judges may not increase overall production (total number of cases processed) and almost certainly will not increase productivity (cases processed per judge) unless additional measures are taken. The limited or null impact on productivity is easiest to explain. There is no reason to believe that having more people do the work will make each of them work any faster or harder. Production is a more difficult question, but in the context of the judiciary as well as other organizations, it has been observed that individual "workers" often have their own "set points" for output--frequently based on how much of their average workload they will CHAPTER 3 36 Because of the practice of reopening temporarily closed cases, these have to be added for the Ethiopian calculation. Normally clearance or disposition rates are simply: number of dispositions/ number of new filings for any given year. 37 It is difficult to be precise here as the proportion of backlog to new filings would determine whether backlog is being reduced--where initial backlog is less than new filings, even a 1.0 or lower ratio could mean backlog is accumulating. 30 Overview of Recent Trends in Demand for and Supply of Federal Court Services 38 process in a given time. If the workload decreases (as would happen if the same amount of work is divided among more workers), but they continue to process the same percentage, this would lead to a lower absolute output per worker, and thus a stable or even lower level of production for the entire group. This sort of reaction by federal court judges in Mexico was reported by researchers there (Magaloni and Negrete, n.d.). If the set-point is based on the absolute amount of work, production would go up but only as a result of the greater number of "production units." Similar reactions have been observed with the introduction of theoreti- cally "time-saving" technology--tasks are accomplished more quickly, but the saved time is 39 not used to accomplish more tasks. Although most judicial informants worldwide usually assert that increases in productivity are impossible--they are working at the maximum--and thus that the only way to deal with increased demand is to add more judges and staff, a comparison of output figures among and even within countries, suggests this is not the case. Within any single country, output per judge usually varies considerably, both absolutely and as a percentage of incoming work. Perhaps not surprisingly, the least burdened judges, often appear to produce less on both counts (on Colombia, see Corporación Excelencia, 2008). There clearly are limits to what any single individual can produce, but they are arguably more than the 200­300 annual judg- ments many court systems in less developed countries seem to believe reasonable. Much of course depends on the composition of those 200­300 cases, and as European commentators have noted, courts with good filtering systems (eliminating the "garbage cases" or channel- ing others to alternative forums) produce less than some systems without them (Blank, et al, 2004; 28). A case in point on the latter end is Brazil, where justices in the Constitutional Court now each write 10,000 opinions annually, and the average annual dispositions per judge are well over 1,000 (World Bank, 2004). However, these "surrealistic numbers" (in the words of a Brazilian judge) are possible because (in the words of another judge) those 10,000 opinions really involve only 125 issues (Jobim, 2003). For most courts, lacking either stringent filtering systems or Brazil's high proportion of redundant (and probably unnecessary) cases, reason- able workloads also fall between the two extremes, over 300 but less than 1,000 dispositions 40 a year. 38 This is an entirely reasonable reaction to a situation where there is always more work to do and seems to be based on workers' experience and the reaction of those monitoring them. Judges may not like being compared to ordinary workers, but as regards the seemingly endless nature of their task, the situations are not dissimilar. 39 This kind of reaction was also seen in Mexico by one of the authors. When the federal courts intro- CHAPTER 3 duced automation to speed processing of document production, court workers reported that before automation they had produced an average of 5 documents daily, and afterwards were producing the same number. 40 It also bears mentioning that in systems where out-of-court settlements are promoted (as in the U.S.) filings may be far in excess of this amount, but most cases are disposed (by settlement or with- drawal) before requiring much work by the judge. 31 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts Because we are talking about Ethiopia, there is another factor that merits mention--disposi- tions (closure) do not mean issuance of a judgment. They simply mean a case is terminated, not necessarily with prejudice (meaning it could be reopened). In fact, in Ethiopia, many of the dis- positions are closures without judgment, and as discussed in a later chapter, this could be prob- lematic. However, in countries like the U.S. where closure without judgment is frequent, this is not interpreted as a system failing because, it is assumed, cases terminated in this fashion have either been decided by the parties out of court (settled) or have been abandoned because one or both parties see little point in continuing. There is a tendency, in virtually every judicial sys- tem in the world, for parties to file cases for a number of reasons having little to do with a just cause--to punish or harass an opponent, to encourage the other party to enter into a negotia- tion, because they simply want to make a statement, because their lawyers misrepresent the ease of winning and so on. Where filing and later abandoning a case reflects the parties' ability to reach a satisfactory agreement, the courts can be considered as playing a positive role in the latter. However, where no agreement is reached and one or both parties remain unsatisfied, the courts cannot be viewed as having made much contribution and the result could be immediate injustice or further conflicts. We will not pursue these issues here, for lack of information on the outcomes, but we will return to some speculative conclusions later in this report. The following two figures examine the first efficiency measure for the federal courts, tracking changes in the number of disposition and disposition rates for civil, criminal and labor cases by year and by instance Figure 7 simply demonstrates that in response to increasing caseload (and partly as a result of adding more judges, especially in the FFIC), the absolute number of dispositions (production) in each instance and in each material has increased fairly steadily over the years. Figure 7: Number of Dispositions by Level of Court and Material from 2005­2009 70000 60000 50000 40000 30000 20000 10000 CHAPTER 3 0 Labor Civil Criminal Labor Civil Criminal FSC Labor Civil FHC Criminal FFIC 2005­06 2006­07 2007­08 2008­09 32 Overview of Recent Trends in Demand for and Supply of Federal Court Services Figure 8: Clearance (Disposition) Rates by Court, Material, and Year 250 200 150 100 50 0 Civil Criminal Labor Civil Criminal Labor Civil Criminal Labor Civil Criminal Labor 2005­06 2006­07 2007­08 2008­09 FSC FHC FFIC As Figure 8 demonstrates, with a few exceptions clearance rates (one measure of productivity) are in the 90 to 115 percent range, which can be considered satisfactory, especially if there is no large backlog to eliminate. Rates over 100 percent would indicate an attack on backlog, as is most visible in the FHC's treatment of criminal and labor cases in 2005­06. It is, however a concern, that clearance rates seem to have stabilized, and even slightly decreased during this period--suggesting either that more incentives are needed to raise them, or that a reliance on judicial incentives may not be sufficient. For countries with much higher levels of demand, the recommendation might be to control it. For Ethiopia, however, solutions are more likely to involve a more aggressive program for accelerating case processing. This is where the observa- tions discussed above about the deviations from the simple trajectories seemingly stipulated by the codes may merit more attention. As explained, the second efficiency measure reviewed here is congestion, which looks at how well the courts are doing in reducing backlog. Congestion and disposition rates have some relationship to each other, although much depends on how much backlog existed at the begin- ning of the period. Generally, the trends are for a steady reduction in congestion rates, most visible for the FFIC and for the criminal benches of the FSC and FHC, although the latter's congestion rate remains relatively high, and distressingly show some tendency to increase or remain stable for the other two areas. It should be remembered that this is being done in the context of an increase in new CHAPTER 3 filings at all levels, meaning that courts are either having to resolve many old cases or attack both new cases and backlog. When the rate drops below 1.0, they are clearly doing both, and it is significant that over the entire 5-year period, rates were at 2.0 only in the first two years, and only for criminal cases in both the FHC and FFIC. It is also significant that the courts can do these calculations, something we found in no other countries where similar studies have been 33 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts Figure 9: Congestion Rates by Level of Court, Material and Year 2.5 2.0 1.5 1.0 0.5 0.0 Civil Criminal Labor Civil Criminal Labor Civil Criminal Labor FSC FHC FFIC 2005­06 2006­07 2007­08 2008­09 done. That does not get them out of the woods, or into a solid sub-1.0 range, but it is clearly a first and very important step and they are making progress. Conclusions Over the period covered by the statistics (mid-2005 to mid-2009) the numbers validate the fed- eral courts' efforts to increase production and productivity and so keep up with an increasing caseload. The increases in the caseload and the overall litigation rate suggest a rising demand for services, in roughly the same proportions for all three instances and of course, in absolute terms, highest in the FFIC. Appeals (and cassation) cases are also increasing and are a source of concern to the FSC which is monitoring the entire process. In an era where the congestion rates of many other courts are steadily increasing, Ethiopia's ability to put them on a downward trend is significant. The High Court, with relatively few original jurisdiction cases, a lower caseload per judge than the other two instances, and an absolute number of appeals which in recent years has been lower than that of the FSC's combined appeal and cassation levels, is a continuing concern. We will look more into these problems in later sections. Dispositions, however, are not everything and as we will see in following sections, the drive to keep abreast of the increases may also have some perverse results--the early closure of cases because of excessive adjournments and the temporary closure of many criminal cases because CHAPTER 3 the witness or even the defendant does not appear for the hearing. Temporarily closed cases may be reopened, but if a majority is not, this does give defendants in particular an additional means of avoiding justice. 34 CHAPTER 4 Users of Justice -- Who Uses the Courts? Introduction We already know that a majority of Ethiopia's population does not use the formal court system, and quite naturally even fewer use the federal courts. However especially in Addis Ababa and Dire Dawa, the federal courts receive, because of party preference and jurisdictional limits on the respective municipal courts, a good share of the litigation there and thus are not necessar- ily hearing only cases initiated by well-heeled parties or those exclusively reserved for the fed- eral jurisdiction. Their caseload may not be representative of the national situation, but it still is indicative of who uses the courts and for what ends. Thus in this section we address the first question--who are the court users, especially as regards gender, individual versus organiza- tional parties, and public versus private organizations. Although "user" might be most narrowly understood to include only the plaintiff--the defendant being an involuntary party--we will look at both. The analysis has been guided largely by hypotheses and patterns originating in other countries or in the general literature. Aside from the (well-founded) assumption that only a minority of citizens uses the formal system to resolve disputes, and that even many criminal offenses are handled in alternative forums, there does not appear to be much speculation in Ethiopia as to the identity of court users. Fortunately, the court database captures many of the categories re- quired for this analysis although the use of some of them is relatively novel. The major missing variables--relating to socio-economic status of individual users--are difficult to capture and are not found in even the most sophisticated statistical systems. Hypotheses based on patterns found elsewhere and to be explored here include the follow- ing: Women are minority court users either as plaintiffs or defendants, a reflection of various cultural and legal obstacles. This generally signifies a lesser protection of their rights by the courts. In civil cases, individual versus individual disputes predominate. We would be surprised to find anything different in Ethiopia as worldwide most disputes seem to be between individual parties even if the most economically significant ones are assumed to involve organizations. 35 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts Although organizational users are often a minority, there are certain types of or even 41 specific organizations which use the courts frequently (are repeat users ). Repeat users are significant for several reasons: their familiarity with court practices is presumed to give them an advantage in litigation; to the extent services are subsidized, they derive consider- 42 able benefits; and their greater dependence on these services gives them a larger stake and thus interest in how courts perform. Repeat users thus may be important stakehold- ers in discussions of reform, but their pro- or con- positions and what they do want can be 43 expected to stem from their specific interests in court actions. In labor cases, individual plaintiffs against organizational defendants predominate. Again this is simple common sense as labor disputes usually involve a worker suing another individual or an organization, and since the organization has deeper pockets, most legal 44 actions focus there. In civil cases, mixed litigation (organization versus individual) are the second most com- mon (after individual versus individual) pattern with organizations usually operating as plaintiffs. The predominant thesis here is that civil litigation becomes a debt collection ser- vice for private organizations (banks and others making small consumer loans). Individuals may have many complaints against organizations, but the costs of litigation often deter their taking them to court. Individual or private organization plaintiffs suing public organizations are unlikely in less democratic systems, and even when they do occur, plaintiffs are unlikely to prevail. 41 Attention to this phenomenon is usually attributed to Marc Galanter (1999) who first made these arguments in regard to litigation in the U.S. Very few developing countries seem to pay attention to it although when they do they often turn up some surprising findings ­ for example the discov- ery (Poder Judiciário, 2004) that a large proportion of the civil cases seen by Rio de Janerio's first instance courts were generated by consumer complaints against a few public enterprises. In both Mexico (World Bank, 2002) and Peru (Gonzales et al, 2002) banks were found to be frequent users of the lowest tier of civil courts for the purpose of collecting small debts. 42 Generally, unless fees are very high, court use is always subsidized as the fixed costs (judicial salaries, infrastructure, etc) are covered by the state. This fact has lead to a suggestion that repeat us- ers be charged higher fees, not only to cover these costs but also to reduce the incentives for using courts for services they might themselves provide by other means. This is especially true as regards debt collection cases where it has been argued that low fees and an efficient judicial service might discourage creditors from doing adequate checks on those to whom they make loans. 43 Predicting whether and which reforms repeat users will support is hardly only intuitive. It is im- portant to take a look at what they are doing and how they are faring now. Repeat users who take advantage of judicial corruption, can, for example, be significant opponents of reforms aimed at curbing it. However, they can also be advocates of change, especially when corruption begins to be CHAPTER 4 used successfully by all comers. 44 Although many, sometimes a majority of labor relationships involve individuals hiring individu- als, there are several reasons for assuming these will not be a majority of complaints. Commonly such relationships lack legal formality, existing legislation may exempt the employers from certain responsibilities, amounts at stake may not be worth litigation, and personal ties may impede taking the issues to court. 36 Users of Justice -- Who Uses the Courts ? As we will see, not all of these patterns hold up in Ethiopia. The reasons why they do not in part have to do with the legal framework and how the jurisdictions are defined, but they also arise in some apparently positive developments as regards people's trust in the courts and willing- ness to use them to resolve disputes. The low cost of litigation is doubtless another contribut- ing factor. It may encourage some abuses, and may leave many users disappointed with the outcomes, but it clearly facilitates access for many citizens who in other countries would never even consider taking a case to court. One further explanation is needed here as regards data on Federal First Instance Courts (FFIC). Because of the way the database is organized, it proved extremely difficult to track these vari- ables over all FFIC. Thus here, and in the following chapters, unless otherwise indicated, data on first instance criminal and civil cases are taken from two federal first instance courts (Lideta and Arada) and those on labor cases from Old Airport and Yeka. All four courts are located in Addis Ababa. Lideta and Arada were chosen as the busiest courts, and Old Airport and Yeka because they receive proportionately more labor cases. Additionally, given that we are covering first instance cases as well as appeals and cassation, we have simply used the terms "plaintiff" and "defendant" to identify the actor initiating the action at each level--thus "plaintiff" should be understood to include the "appellant" in cases being heard on appeal. We apologize for the liberties taken with technical terminology, but the alternative would have made for a much more complex set of tables and graphs. Gender Because of a universal assumption that women do not use formal justice, for a variety of reasons (culture, bias, finances, or even the legal framework) this variable is of interest as an indication of structural or attitudinal obstacles. Here we examine it from two perspectives--frequency of women as plaintiffs or defendants in any case. In this and the following charts, we are not interested in the identity of the other party (defen- dant or plaintiff ) but only in the gender of the party being studied. This is mentioned to explain why numbers of plaintiffs and defendants are not the same (because, often times the other Table 5. Individual Plaintiffs in Civil Cases, by Gender, Instance, and Year Year CHAPTER 4 2005­06 2006­07 2007­08 2008­09 User SC HC FIC SC HC FIC SC HC FIC SC HC FIC Male 2477 1939 5020 2924 2172 5282 3240 2455 8191 4079 2361 8455 Female 1255 1162 5436 1618 1387 5376 1782 1650 8016 2188 1449 9149 37 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts party is not an individual, and because sometimes there are multiple plaintiffs or defendants). We look at the role and incidence of organizational defendants in another section. As regards the gender of plaintiffs, two observations should be made. First, women are not underrepresented in the FFIC (probably because they are frequent plaintiffs in family cases). Second, at the higher levels, their incidence decreases--meaning that first (for the High Court alone), they are less likely to initiate cases of higher monetary value, and second (for both the High and Supreme Court) they apparently are less likely to enter appeals or cassation. It should also be remembered that for the FSC both appeals and cassation (but predominately the latter) may also originate in cases first seen in the regional or social courts. The FHC also receives some appeals from the regions, but this is fairly rare. Table 6. Individual Defendants in Civil Cases, by Gender, Instance, and Year Year 2005­06 2006­07 2007­08 2008­09 User SC HC FIC SC HC FIC SC HC FIC SC HC FIC Male 2466 2358 5568 2932 2443 5376 3285 3004 7407 3691 2514 8245 Female 1425 1359 2365 1666 1384 2519 1941 1911 3258 2192 1471 3457 As defendants, women are underrepresented at all levels, although in the FSC and FHC some- what less so than as plaintiffs. For the FFIC, their lower incidence as defendants is again a prob- able consequence of their more likely involvement in family cases as plaintiffs. Assuming they usually win those cases, this could also explain the slightly higher representation of woman as defendants in the higher courts. In Ethiopia, as opposed to other countries reviewed (especially those in Latin America where appeal-itus seems to be a typical trait) the winner of a case gener- ally does not appeal, and thus if (as is suggested in a later chapter) courts tend to find for the plaintiff more frequently in civil and labor cases, at the appellate level, the plaintiff and defen- dant will reverse identities (i.e. a losing defendant will become the "plaintiff" or appellant). This holds even more for criminal cases, where if a judgment is rendered, it usually is a conviction. (The bigger problem is that most criminal cases do not come to judgment). Thus the observa- Table 7. Individual Plaintiffs/Appellants in Labor Cases, by Gender, Instance, and Year CHAPTER 4 Year 2005­06 2006­07 2007­08 2008­09 User SC HC FIC SC HC FIC SC HC FIC SC HC FIC Male 372 366 1324 339 231 1278 762 352 1230 955 245 1905 Female 59 42 134 72 38 221 187 81 246 210 49 392 38 Users of Justice -- Who Uses the Courts ? tion that "the prosecutors never appeal" may be logical assuming they usually win if the case gets that far. However, as discussed in a later chapter, this assumption is not necessarily valid. Absent good data on the percentage of women in the workforce, it is hard to tell whether their very low representation as plaintiffs in labor cases is a bad or good sign. However, they are also (see below) very much underrepresented as defendants. It should be remembered that all labor cases enter at the FFIC level and thus that FSC and FHC labor cases are appeals or cassation (only for the FSC). Table 8. Individual Defendants in Labor Cases, by Gender, Instance, and Year Year 2005­06 2006­07 2007­08 2008­09 User SC HC FIC SC HC FIC SC HC FIC SC HC FIC Male 438 382 56 352 175 26 668 241 49 795 201 132 Female 105 58 22 57 28 8 286 28 9 316 34 18 Individual versus Organizational users In many countries it is believed that private (and public) sector organizations are more fre- quent users of justice than are individuals. However, as noted, prior research using samples and statistics suggests a predominance of individual disputes, or of those where at least one party is an individual. Other work, while admitting the greater frequency of individual conflicts, finds that certain organizations use the courts very frequently--are repeat users. Here we examine the competing hypotheses in the Ethiopian context, and also review the types of organization, public or private. The numbers given below are cumulative for all four years, although once again only two First Instance Courts are the source for each table. Table 9. Identity (individual, private organization, public organization) of plaintiffs and defendants to all civil cases filed between 2005 and 2009, by Court Level Status and Instance Individual Private Org Public Org Plaintiff FIC N = 64036 (Lideta and Arada Only) 43104 7701 6445 CHAPTER 4 Plaintiff HC N = 16564 13015 1317 1597 Plaintiff SC N = 18139 15473 841 1616 Defendant FIC N = 45145 39461 2131 2516 Defendant HC N = 23851 12882 1320 1565 Defendant SC N = 20843 14383 888 2505 39 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts The predominance of individuals as both plaintiffs and defendants at all levels comes as no surprise. What is surprising is that public organizations figure more frequently than private ones, at all levels and also as both plaintiffs and defendants. On the one hand, this doubtless is a logical consequence of the larger role of the state in many economic activities, but on the other, it also suggests a greater inclination than might be expected for private individuals and organizations to contest its actions. (There are of course instances where public organiza- tions engage each other in disputes, but they are not frequent enough to account for these patterns). Table 10. Identity (individual, private organization, public organization) of plaintiffs and defendants to all labor cases filed between 2005 and 2009, by Court Level Status and Instance Individual Private Org Public Org Plaintiff FIC N = 8157 (Old Airport and Yeka only) 7932 62 136 Plaintiff HC N = 6220 3258 1444 730 Plaintiff SC N =2943 1440 606 687 Defendant FIC N = 6881 1001 4916 964 Defendant HC N =5 462 3334 1533 595 Defendant SC N = 2731 1595 471 665 It should be mentioned as regards the above and later tables, that we suspect some errors of data entry as regards the identity of the parties. However, in the above table the small number of organizational plaintiffs in first instance labor cases is apparently not one of them. There are labor cases initiated by organizations, even against individuals. For the FHC and FSC all cases are appeals or cassation, and thus we can generally assume that the "plaintiff" lost at the first instance. It should also be remembered that FFIC data come from only two courts, thus explain- ing the lower N's for the cases surveyed. Table 11. Identity (individual, private organization, public organization) of plaintiffs and defendants to all criminal cases filed between 2005 and present, by Court Level Status and Instance Individual Private Org Public Org Plaintiff FIC N = 75794 (Lideta and Arada only) 1169 0 74625 Plaintiff HC N = 18762 3677 0 15060 CHAPTER 4 Plaintiff SC N = 5539 4928 0 592 Defendant FIC N = 75091 75091 0 0 Defendant HC N = 18723 15327 0 3396 Defendant SC N = 5482 733 0 4749 40 Users of Justice -- Who Uses the Courts ? In theory, criminal trials should always involve the state (public organization) as plaintiff. However, two exceptions explain the appearance of individual plaintiffs at the FFIC. First, while Ethiopia does not allow "private prosecution" (most commonly elsewhere for "crimes against honor"), this type of complaint has to be initiated by the private party although it is taken forward by a public prosecutor. Thus some of these cases may have been entered with private plaintiffs. Second, and perhaps more importantly, when individual defendants file applications for bail, this action is recorded as another case--possibly inflating the total number of criminal cases, but there was no easy way of separating this out. The small numbers of public organiza- tions as plaintiffs and of individual defendants in the FSC (all appeals or cassation) do suggest that either 1) the state wins all its criminal cases or 2) it doesn't appeal what it loses. In effect, both explanations are likely in Ethiopia with the latter, however, accounting for the larger role. Repeat Organizational Users Given the importance assigned to repeat users in the general literature, we also did a search over the four-year period to determine their role here. The following table it the result. The way data are kept only allowed tracking of repeat users at the FSC and FHC. Clearly, as the topic is of potential interest, it would be well for the appropriate additions to be made to the database for the FFIC. Entries correspond to the number of cases in which these parties appear over the 2005­2009 period. The "repeat users" are the expected ones, but the incidence of their repeat appearance is not very high. Unless the low incidence is a function of errors of data entry, repeat users do not appear to play an unusually large role in Ethiopia. However, it is also possible that many more cases are filed at the first instance and simply do not go to appeal Table 12. Repeat Users in Federal Courts, cumulative for 2005­2009 FSC FHC Identity Filed Disposed Filed Disposed Plaintiff Banks 394 344 478 547 Ethiopian Telecommunication Corporation 96 95 80 93 Ethiopian Customs Authority 76 76 84 117 Public Housing Agency 160 195 248 356 Insurance Companies 131 111 234 255 Defen- Banks 502 471 341 423 CHAPTER 4 dant Ethiopian Telecommunication Corporation 78 77 5 106 Ethiopian Customs Authority 62 61 107 115 Public Housing Agency 276 277 367 458 Insurance Companies 119 99 188 231 41 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts or cassation. One additional note is that the higher numbers of disposed than filed cases for several categories reflects attention to backlogged cases from earlier years. This is simply an indication that the courts are catching up (but also may show a greater willingness to address cases with more powerful parties). Litigation Matrices ­ Who Sues Whom? To this point we have only looked at the frequency of court use (as plaintiff or defendant) by gender and individual or organizational status. The analysis allowed us to investigate several common hypotheses about who uses the courts, but said very little about the types of disputes, especially as regards the identity of both parties. In this section we explore one dimension of this issue, looking at the incidence of conflicts between individuals, between individuals and organizations, and between organizations. For this purpose, we graph what we have termed a "litigation matrix" identifying parties in this fashion. While only relevant for individual-orga- nizational conflicts, the first party mentioned is the plaintiff. For criminal and labor cases, this is largely self evident at the first instance or trial level (whether in the FFIC, FHC, or for the few criminal cases seen in the FSC). In criminal cases, "organizations" (that is the state) are nearly al- ways the plaintiff, and for labor cases, the plaintiff is usually an individual. However, at the appeal stage (some High Court and Supreme Court cases), this rule does not hold, and it is therefore of interest to know whether individuals or organizations are most likely to appeal (appearing as the plaintiff no matter what their initial status). The point here is not only to determine patterns of use within Ethiopia but also to see how they stand up against common assumptions or trends detected in other countries, for example: A predominance of individual versus individual disputes in civil cases, with individual versus organization or organization versus individual coming in second. Organization versus organizational cases are assumed to be less common, everywhere, although (to be explored in later chapters) they usually involve larger amounts A predominance of individual appellants in criminal cases--because where the defendant loses s/he will normally appeal and because (in Ethiopia) it is commonly believed that the state, even when it loses, will not appeal because of prosecutors' overwork. In other coun- tries (especially in Latin America) the pattern is reversed because prosecutors are either required to appeal (in both criminal and administrative cases) or do so because of a fear they will be accused of dereliction of duty if they refrain. A greater proportion of organizational than individual plaintiffs at the appellant stage in labor cases because 1) workers are favored by judges and so win at the first instance and 2) whether CHAPTER 4 or not they are satisfied with the results, workers lack the deep pockets required to appeal. As the matrices are by year, but show absolute numbers, rather than percentages, it is the rela- tive size of the boxes that count in interpreting changes over time. Again in all the following 42 Users of Justice -- Who Uses the Courts ? charts, statistics from the FFIC represent only two courts, and not the entire universe. We pres- ent and discuss the tables by materials, for all instances. As regards the civil litigation matrices, there are a few surprises. First while individual versus individual cases predominate at all levels, cases pitting organizations against each other unex- pectedly hold a second place in the first instance courts, and seem to be increasingly frequent. As civil cases also include administrative issues, this may well involve disputes in that area. Second, the greater incidence of individuals as plaintiffs in cases involving organizations at all three levels is also surprising. On the one hand (and especially for the FFIC cases) this puts Figure 10: Litigation Matrix, Civil Cases, FSC 4000 3500 3000 2500 2000 1500 1000 500 0 Cases individual Cases individual Cases organization Cases organization against individual against organization against individual against organization 2005­06 2006­07 2007­08 2008­09 Figure 11: Litigation Matrix, Civil Cases FHC 3500 3000 2500 2000 1500 1000 500 0 CHAPTER 4 Cases individual Cases individual Cases organization Cases organization against individual against organization against individual against organization 2005­06 2006­07 2007­08 2008­09 43 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts Figure 12: Litigation Matrix, Civil Cases, FFIC (Lideta and Arada Courts only) 10000 9000 8000 7000 6000 5000 4000 3000 2000 1000 0 Cases individual Cases individual Cases organization Cases organization against individual against organization against individual against organization 2005­06 2006­07 2007­08 2008­09 into question another assumption founded on experience elsewhere--that banks and other (largely private, but also public) organizations are among the most frequent users of courts to collect debts. Instead the numbers indicate that individuals are increasingly willing to take on organizations, something not exactly anticipated in present-day (or traditional) Ethiopia. On the other, and in the higher courts, it suggests that even if they lose on the first round, individu- als are not dissuaded from going to appeal or cassation. Third, the continued strong showing of individual versus individual cases in the higher courts, and especially in the Supreme Court, where they are proportionately more frequent than at the lower levels, is also surprising. Fourth, the decreasing importance of organization versus organi- zation cases at the higher levels, and especially in the FSC, is also a surprise. It should be remembered, as regards all four points, that at the FSC level appeal and especially cassation cases are drawn from both federal and regional courts, and even from social courts. Moreover, the numbers shown here also include cases that will be dismissed for lack of merit after an initial brief review. According to Supreme Court sources, there are a good number of these involving land cases which began in the social courts. By the time they get to the Su- preme Court, the usual lack of legal representation often leads to their early dismissal--not because legal representation is required but because it is more necessary in preparing the ar- guments at that level. Even if this is the explanation, it does demonstrate a willingness of even fairly humble parties to use the formal system. However, if they continue not to receive a full CHAPTER 4 hearing, the trend could easily be reversed. We next use the same framework for reviewing labor cases, which it should be remembered always initiate in the FFIC. 44 Users of Justice -- Who Uses the Courts ? Figure 13: Litigation Matrix Labor Cases, FSC 500 400 300 200 100 0 Cases individual Cases individual Cases organization Cases organization against individual against organization against individual against organization 2005­06 2006­07 2007­08 2008­09 Figure 14: Litigation Matrix Labor Cases, FHC 1200 1000 800 600 400 200 0 Cases individual Cases individual Cases organization Cases organization against individual against organization against individual against organization 2005­06 2006­07 2007­08 2008­09 Figure 15: Litigation Matrix Labor Cases, FFIC (Old Airport and Yeka, only) 2500 2000 1500 1000 500 0 CHAPTER 4 Cases individual Cases individual Cases organization Cases organization against individual against organization against individual against organization 2005­06 2006­07 2007­08 2008­09 45 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts In light of what we know about Ethiopia and about labor cases in general, the patterns shown above are not unexpected. Individuals are nearly the exclusive plaintiffs in the FFIC which not only by name (first instance) but also by legal limitations are the nearly exclusive original jurisdiction courts for labor cases. However as cases go to appeal and cassation, individuals and organizations are nearly equally likely to be plaintiffs suggesting either that individuals win labor cases less frequently than Ethiopian observers suppose or that they are not satisfied with a partial victory. Finally, we look at litigation patterns for criminal cases, for which the FFIC is the primary trial venue, but where some cases are initiated in the FHC. Figure 16: Litigation Matrix Criminal Cases, FSC 2000 1800 1600 1400 1200 1000 800 600 400 200 0 Cases individual against individual Cases individual against organization Cases organization against individual 2005­06 2006­07 2007­08 2008­09 Figure 17: Litigation Matrix Criminal Cases, FHC 5000 4500 4000 3500 3000 2500 2000 1500 1000 CHAPTER 4 500 0 Cases individual against individual Cases individual against organization Cases organization against individual 2005­06 2006­07 2007­08 2008­09 46 Users of Justice -- Who Uses the Courts ? Figure 18: Litigation Matrix Criminal Cases, FFIC (Lideta and Arada, only) 25000 20000 15000 10000 5000 0 Cases individual against individual Cases individual against organization Cases organization against individual 2005­06 2006­07 2007­08 2008­09 The nearly exclusive appearance of individuals as defendants against organizational plaintiffs in the FFIC is no surprise, since this is how criminal cases occur under Ethiopian law, the only exceptions being applications for bail or the occasional entry of a "crime against honor" as having a private plaintiff (although even here a public prosecutor must take the case forward). At the FHC, individuals' moderate incidence as plaintiffs corresponds to the appeals made against guilty verdicts in the lower level courts. Most of the FHC organization versus individual cases are those being judged for the first time at that level. A second observation in support of that conclusion is the predominance of individual plaintiffs (on appeal and cassation) at the FSC level, suggesting either that all judgments find the defendant guilty in the High Court or that, as conventional wisdom holds, Ethiopian prosecutors rarely appeal. It is interesting that the absolute number of individual appellants/plaintiffs is slightly higher at the FSC level than for the High Courts, despite the far greater number of criminal cases seen in the FFIC. There are several likely explanations for this which probably operate in conjunction: cases seen in the High Courts carry stiffer penalties; their defendants more often have legal representation (either by private attorney or by a state-subsidized lawyer), and they are more likely to have the means to pursue an appeal (if only because many of the crimes sent to first instance courts are the sort of things--petty robberies, assaults and so on--where the perpetrator is more often poor, uneducated and thus unlikely to attempt even a free appeal.) The FSC cases also of course include cassation, which may include cases from the FFIC, FHC, and the regional courts. CHAPTER 4 Conclusions The two most important conclusions arising from this analysis are that woman do indeed have an active role as plaintiffs in the first instance courts, thus indicating that at least in Addis Ababa and Dire Dawa, they do not face unusual obstacles, except perhaps in labor cases, and that indi- 47 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts viduals are the most common users of the courts, with most civil disputes in particular coming between them. This second finding is not unusual, and repeats patterns observed in most other countries. Indeed, had the result been different, it would have been a cause for concern, as it was in the Peruvian study which showed banks and insurance companies as majority users of small claims civil courts (Gonzales et al, 2002). The second place standing of organization versus organization cases in the first instance civil courts (FFIC) is somewhat surprising as is its failure to be replicated at the higher levels. By law where FFIC cases have a monetary value, it must involve amounts of less than 500,000 birr (cur- rently, roughly US$45,000) which means they are not necessarily negligible and thus could en- 45 compass inter-organizational disputes. However, the finding still runs counter to that of other studies (largely in Latin America) where the second (and in some cases first, as in Peru) use of the lowest level courts was for organizations collecting small debts against individuals or indi- viduals protesting actions of organizations. Unfortunately, our inability to identify repeat users at the FFIC level does not provide additional information to test this explanation. As regards the unexpected frequency of organization-to-organization cases these may also involve tax or land disputes, as these administrative cases are usually heard on appeal in the civil jurisdiction once they have received a first decision in the executive administrative tribunals. We will pursue this in the next chapter where we look more closely at types of cases within each of the broader legal areas. All of these preliminary findings suggest several tentative hypotheses about court use in Ethio- pia which we will explore further over the course of this report. They also, of course, suggest themes for further exploration which, for lack of time and sometimes of sufficient data, we have not been able to pursue here. First as regards the hypotheses: For a variety of reasons court use seems to be less restricted in Ethiopia than in many developing countries. (These we conclude from the high number of disputes involving in- dividuals as plaintiffs or as both parties and the significant representation of women). Low costs for filing and the ability to file without a lawyer probably explain a good deal here. Taking legal action without a lawyer may not be advisable if one wants to win a suit, but not needing one to file does enhance the first step in access The use of courts by organizations to collect debts owed by individuals is seemingly not as common in Ethiopia as in other countries. This may be more a consequence of economic practices (lesser frequency of small consumer and bank loans) than of the ease of using or the efficacy of the courts. However, as this use is often regarded as "abusive" (using the courts as a debt collection agency) it can be regarded as a plus. CHAPTER 4 45In theory civil cases with a value under 5,000 birr go to social courts, which are not in operation in Addis and Dire Dawa. They could however go to the municipal courts in each city, and probably do. 48 Users of Justice -- Who Uses the Courts ? While there are some "repeat users," they appear less frequently than in many other coun- tries, and interestingly seem as likely to appear as defendants as plaintiffs. This is a good sign, suggesting that private parties are willing to take them on and indicative of one or more positive developments: 1) increasing faith in the courts; 2) fewer barriers to access; and 3) lesser fear of reprisals. There is little information (see Chapter VI) as to how plaintiffs (or defendants) fare against the repeat users, but at the very least there is no suggestion that the latter have monopolized court use to forward their own agendas. The high caseloads in the FHC and the FSC suggest a lack of resistance to appealing a decision. We will look at this in more detail in later chapters, but while it is positive from the standpoint of access, it has its drawbacks as regards court congestion and efforts to avoid justice through the use of dilatory tactics. In any event, and as a bottom line to the topic, although the federal courts can be thought of as an "elite forum" they appear not to function in this fashion here. Of course to under- stand this better, we need to know more about specific types of cases, which is the topic of the next chapter. CHAPTER 4 49 CHAPTER 5 Uses of Justice ­ For What are the Courts Used? Introduction In this section, we look at the patterns in types of ligation, and their changes over the five years covered. Which are the cases most likely to be received by Ethiopia's federal courts and has there been any significant variation over this short period? As we have already looked at case distribution in the three main legal areas (civil, criminal, and labor) as defined in Ethiopia (remember for example that administrative cases when they go to the judiciary, enter in the civil jurisdiction), that is not our interest here. Instead we are concerned with the principal categories of disputes within each of these larger classifications as an indication of the kinds of conflicts that occur in Ethiopia and in turn are handed to the courts for resolution. In this sense, it would have been useful to have information from a second kind of research--interviewing individuals about the types of potentially justiciable problems they encounter. Doing this kind of study proved impossible for reasons of both time and costs, but at some later stage it is worth attempting. Similar studies done elsewhere within the region (Republic of Kenya, 2006; Cook et al. 2007) and outside it (Genn, 1999; Castro et al, 2007) are helpful in demonstrating not only what gets to court, but also the difference between what arrives and the conflicts people face most frequently. Typically such studies also ask people how they resolve their disputes (within what forums), why, and with what level of satisfaction. Especially in regions like Africa where court use is limited, this is a good means of determining access needs and how they might best be addressed. Still even a partial view--only of what does arrive in the courts with the most frequency--is useful, especially if it can be cross checked against parties (users) and outcomes. Thus, for example it would be important to know whether women using the courts do so for family disputes, and if so (or if not) what kinds of outcomes they receive. In the present study, we have only begun to tap into those questions, and thus in the following chapters in addition to findings we also suggest useful future lines of analysis. Some of these can be done now; others had to be eliminated for lack of time or because they would require slight adjustments to the database to permit their being conducted. As in prior chapters, most of the questions being addressed here are based on findings from other studies. "Conventional wisdom" about what cases reach courts in Ethiopia remains under- developed, which may be a good thing inasmuch as it will not provoke resolution of imaginary 51 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts problems or inattention to real ones, as it has elsewhere. Thus, drawing on beliefs and findings in other studies, the hypotheses for investigation are as follows: 46 In civil cases, debt collection (or contract enforcement ) will be the most common issues-- in other countries (e.g. Argentina, Colombia and Mexico) it often reaches 50­70 of civil filings. Family cases (divorce, child support, inheritance) will be a second common civil category The incidence of different types of original jurisdiction cases within higher and lower level courts should vary, as some issues automatically go to one or the other (e.g. labor cases always originate in the FFIC) while the naturally larger amounts (or stiffer penal- ties) for others will send them to the higher courts. (Many family disputes on the other hand will not because they either fall by definition with the lower courts or involve lower amounts). It thus follows that average amounts claimed will differ by instance, as well as by issue. Appealed civil cases will have different frequencies than those for the first instance--this is on the assumption that appeals are most likely to be entered when larger amounts or more severe penalties are at stake. Labor cases will either be for unfair dismissal or for a number of causes--frequently in la- bor cases the strategy of the plaintiff is to ask "for everything," in the hopes that something 47 will be granted. Depending on the local legal framework and other factors (like cost for using alterna- tive mechanism--e.g. notaries) court caseload may have a high incidence of what might be called "notarial" cases (resulting in Ethiopia in "declaratory judgments")--that is cases where a document or right is only registered by the judge, with no controversy being involved. It is often argued that these cases can unnecessarily inflate judges' workload and thus we are interested in knowing the extent to which they do here. This effect is most likely to occur in civil matters or in a separate family jurisdiction if one exists. Land cases are another common issue, especially in developing countries, but because all land "belongs to the state" in Ethiopia, cases involving ownership are not possible. However, land cases over rights other than ownership may be substituted--we simply have no pre-estab- lished expectation here. 46 As debt collection cases are not classified separately in Ethiopia, we will include "contract cases" CHAPTER 5 instead, but in Ethiopia as elsewhere, much of contract enforcement is about collecting money due. For the FFIC, the comparable category is "diverse monetary claims." 47 Although this might seem more likely for an unrepresented client, this was common in other countries, even when the employee had legal representation. Lawyers apparently also adopted the tactic of asking for anything and everything on the assumption that some of the claims would be awarded (World Bank, 2004 on Brazil). 52 Uses of Justice ­ For What are the Courts Used? Types of Disputes In this section we look at the most common types of disputes within the three major areas-- criminal, civil, and labor--by level of court. This was done by taking the 5 or 6 more frequent categories listed within the database. We have used filings, rather than dispositions for these graphs as the emphasis is on demand. Once more, while statistics for the FHC and FSC are based on total caseload, those for the FFIC are drawn from two courts (Lideta and Arada) for civil and criminal cases, and from Old Airport and Yeka for labor. Table 13. Major Civil Issues (Filings), FSC, by Year No Major Civil Cases 2005­06 2006­07 2007­08 2008­09 1 Contracts 408 447 479 542 2 Immovable Properties (House) 394 491 819 1023 3 Movable Property 243 262 233 169 4 Inheritance 251 352 359 390 5 Execution 252 305 295 343 Table 14. Major Civil Issues, FHC, by Year No Major Civil Cases 2005­06 2006­07 2007­08 2008­09 1 Contracts 315 355 263 143 2 Immovable Properties (House) 428 524 624 514 3 Movable Property 172 131 110 68 4 Inheritance 304 357 503 478 5 Execution 512 583 615 570 Table 15. Major Civil Issues, FFIC, By Year No. Major Civil Cases 2005­06 2006­07 2007­08 2008­09 1 Execution 3012 3567 3852 3884 2 Adoption 2281 2932 3913 4177 3 Diverse Monetary Claims 1584 1352 3086 3812 4 Divorce 3342 3979 4593 5076 5 Inheritance 272 293 382 562 CHAPTER 5 Several observations can be made as regards civil cases. First family cases not only predominate at the FFIC level; they also, with the exception of inheritance, seem to be settled there. Adop- tion cases are largely declaratory, constituting a good portion of the approximately 28 percent 53 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts of the FFIC civil caseload in that category. Divorces are usually contested, as is inheritance. Second, and surprisingly "diverse monetary claims" (or contract enforcement) cases are only the fourth most important category the FFIC and are also, as contract cases, in that position for the FHC. However, execution (enforcement of judgment) cases are very frequent at the FFIC level and are also important for the other two instances. This is a growing concern for the courts as they believe this is becoming the second line of defense for avoiding payments of debts. Other patterns are hard to identify, except for the enormous growth in appeals and cassation cases involving immovable property (ownership of houses and mortgage disputes) in the FSC. "Land" cases in general are becoming more prominent in Ethiopia, especially in urban areas where they usually involve rights to immovable property. They are often heard first, by law, by admin- istrative tribunals, but apparent dissatisfaction with outcomes there (and with possible corrup- tion) quickly move them into the courts. Table 16. Major Criminal Issues, FSC, by Year No Major Criminal Cases 2005­06 2006­07 2007­08 2008­09 1 Murder 245 334 621 854 2 Application for Bail 126 75 96 41 3 Application for leave to appeal out of time 113 67 219 168 4 Fraud 106 57 68 112 5 Robbery 69 88 174 507 Table 17. Major Criminal Issues, FHC, by Year No Major Criminal Cases 2005­06 2006­07 2007­08 2008­09 1 Murder 576 708 839 499 2 Drugs 240 282 350 321 3 Fraud 747 1292 1268 1280 4 Robbery 513 1066 1184 1869 5 Bodily Harm 77 92 101 131 Table 18. Major Criminal Issues, FFIC, by Year (Lideta and Arada only) No Major Criminal Cases 2005­06 2006­07 2007­08 2008­09 1 Theft 3264 3765 5979 6308 CHAPTER 5 2 Bodily Harm 1906 2689 3727 3707 3 Rape 598 624 617 379 4 Defamation 561 1002 1508 1200 5 Breach of Trust 502 540 778 834 54 Uses of Justice ­ For What are the Courts Used? As regards criminal cases, a first important reminder is that virtually all cases heard by the FSC are on appeal or cassation. The Federal Supreme Court typically hears only one or two crimi- nal cases in original jurisdiction annually. Two other items seen here--application for bail and application for leave to appeal out of time--are peripheral issues attached to other cases. If one assumes a one-year time lag in appealing serious crimes (e.g. murder) from the High Court (where they are first heard) to the Supreme Court, it is apparent that a good portion of these are appealed. Ethiopia is not a country known for high levels of violent crime and this is reflected in the distribution of cases. The relatively high incidence of cases involving defama- tion and breach of trust in the FFIC are also indicative of that. Theft and lesser injuries are more important, but the numbers are not terribly high, especially considering that the two courts sampled are among the busiest. Table 19. Major Labor Issues, FSC, By Year No Major Labor Cases 2005­06 2006­07 2007­08 2008­09 1 Termination of Employment 238 331 147 40 2 Promotion 42 26 58 49 3 Application for leave to appeal out of time 17 12 10 17 4 Execution 30 17 20 24 5 Compensation 100 110 185 276 Table 20. Major Labor Issues, FHC, by Year No Major Types of Labor Cases 2005­06 2006­07 2007­08 2008­09 1 Termination of Employment 421 569 1143 1156 2 Promotion 48 65 81 46 3 Back Pay 237 15 172 117 4 Transfer 31 18 8 10 5 Compensation 41 21 69 41 Table 21. Major Labor Issues, FFIC, by Year No Major Labor Cases 2005­06 2006­07 2007­08 2008­09 1 Termination of Employment 353 469 237 401 2 Promotion 4 5 11 33 CHAPTER 5 3 Back Pay 148 180 344 612 4 Disciplinary Issues 68 10 3 2 5 Compensation 78 77 134 93 55 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts For labor issues, all of which start at the FFIC, the surprise is the number of appeals to the FHC, and especially those for termination of employment. Otherwise, and with the exception of the categories for execution and application to appeal out of time (late), the issues remain very similar and more or less in the same dimensions. The lower numbers at the FSC respond in part to the impossibility of a second appeal if the first appeal (at the FHC) supports the initial judgment. Some FSC cases (those for compensation perhaps) may arrive on cassation or from regional courts. Monetary Claims In the judicial database, monetary claims are only recorded for civil cases, but contrary to patterns seen elsewhere, in most civil cases a monetary amount is recorded with the claim. Exceptions include cases receiving declaratory judgments (e.g. adoption) and some issues like electoral dispute (which also go to the civil jurisdiction). Because the amount at stake is one of the determinants of where cases will start (at what instance) we would expect significant differences in the average amounts for cases entered at each level. However, this also suggests differences among the frequencies of types of civil cases entered at each level. We are interested in the amounts claimed (and awarded) for several reasons. They are an indica- tion of how the courts are used (and by whom)--to settle mundane minor issues or only for big stakes. They also, when tracked through the appeals and cassation level indicate something about party willingness to pursue cases, even for minor amounts. Because Ethiopia is a poor country, and because we know court use is limited, the underlying issue is how "elitist" court use Table 22: Mean Monetary Claims for Filed and Disposed Civil Cases, by Year, FSC, FHC, FFIC (Lideta and Arada Only) In Ethiopia Birr; roughly 11 birr to the US dollar Court Case Level 2005­06 2006­07 2007­08 2008­09 FSC Filings Appeal 214,564 329,633 196,251 149,546 Cassation 21,312 155,774 55,505 88,007 Disposal Appeal 107,017 196,096 242,142 233,266 Cassation 16,618 147,833 51,196 52,338 FHC Filings Appeal 39,855 47,873 38,676 80,579 First Instance 3,458,506 193,709 213,734 2,118,304 CHAPTER 5 Disposal Appeal 35,820 48,471 40,439 69,325 First Instance 1,654,305 255,935 147,824 154,233 FFIC Filings First Instance 20,083 21,040 14,913 16,914 Disposal First Instance 27,081 19,825 22,756 20,686 56 Uses of Justice ­ For What are the Courts Used? is. Fortunately the database also allows us to track awards as well as claims, and appeals as well as original jurisdiction, letting us see how use and tactics (claim more than you can hope to get) work out in practice. It should also be noted that we have used arithmetic averages rather than medians. This is not the preferred technique as a few very large (or very small) claims can distort the outcomes. Some distortion is also presented at the FFIC level because of the large propor- tion of cases (those receiving declaratory judgments) that do not have a value attached. The most noteworthy observation here is that the average amounts for filed and disposed cases in the FFIC are relatively low, and far under the 500,000 birr ceiling. This is not unexpected in light of findings elsewhere (World Bank, 2002 and 2003) and is positive in the sense that it indicates that people use the courts for smaller claims. However, taking into account the coun- try's poverty, the claims are not that small. Fortunately, the municipal courts in Addis Ababa and Dire Dawa may take up some of the slack. The generally higher average for disposed as opposed to filed cases has several possible explanations (time lags in dispositions, tendency for lower value cases to drop out, etc) but we lack sufficient information to select among them. The higher average claims for filed and disposed FHC cases (especially for the first and last 48 years) may simply be a function of a few large cases being added, in addition to the higher floor for submitting cases to the High Court. It is noteworthy that for all years, but one, the amounts for disposed cases drop considerably This clearly reflects a tendency for claims to be higher than awards finally made, and is thus also reflected in the lower mean amounts for the FSC, where the claim registered corresponds to the lower court's award. Finally, it is interesting that the amounts for appeals (from the FFIC to the FHC and from the FHC to the FSC) and for cassation (FSC only) are relatively modest. While partly a function of the lower awards made at the subordinate level, this also indicates a willingness of individuals with modest amounts at stake to take their chances with another review. As discussed in the next section, Ethiopia's high appeals rates are a concern to the judiciary, but in a poor country, one would not want to impose monetary floors on this privilege. Appeals The conventional wisdom prevailing in most developing regions is that everything is appealed. The question is whether this also holds in Ethiopia. The relative poverty of the population and the low level of legal representation might lead us to expect it would not. However, the low costs of filing appeals (and the lack of fees for cassation) provide counter arguments. CHAPTER 5 48 As we are using arithmetic means, this is one problem ­ the effect of very large or very low ex- tremes. However, it proved too difficult to use median amounts for the entire database. 57 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts There are two ways of calculating appeals rates. Both are based on the number of judgments from the lower court, with the difference being that the numerator can be either the number of applications for appeal or the number of appeals actually accepted. By using both rates, or calculating the percentage of appeals accepted over those for which an application is made, we can also get some sense not only of the parties' inclination to appeal but also the courts' ability to place some control on its exercise. There is also the question, as regards either formulation, as to whether the rate of appeal varies by type of case, type of party, or other characteristics, like the amount at stake for civil cases. It should be remembered that appeals are facilitated in the Ethiopian system because of 1) the lesser fees for filing them in civil cases and 2) the automatic right of appeal for all criminal Figure 19: Appeals Rates (Application to Appeal) from FFIC to FHC, by Legal Area and Year 60 50 40 30 20 10 0 2005­06 2006­07 2007­08 2008­09 Civil Criminal Labor Figure 20: Appeals Rates (Admitted Cases) from FFIC to FHC, by Legal Area and Year 20 18 16 14 12 10 8 6 4 2 CHAPTER 5 0 2005­06 2006­07 2007­08 2008­09 Civil Criminal Labor 58 Uses of Justice ­ For What are the Courts Used? convictions. The ability to appeal without using legal representation lowers the cost, although it probably also lowers the chances of success for those working without a lawyer. There are several interesting characteristics here. First is the difference in appeals rates by major areas in the two stages. In labor cases applications to appeal, which reach a high of nearly 60 percent from the FFIC to the FHC, only reach 7 percent from the FHC to the FSC. As these numbers represent percentages of judgments at the lower level, the quantity of judgments does not affect them. The most probable explanation for the lower FSC rate is that since all cases originate in the FFIC, only a second appeal would go to the FSC, and this is only allowed if the first appeal overturns the lower-level judgment. Apparently, it does not and thus second Figure 21: Appeals Rates (Application for Appeal) from FHC to FSC, by Year 50 45 40 35 30 25 20 15 10 5 0 2005­06 2006­07 2007­08 2008­09 Civil Criminal Labor Figure 22: Appeals Rates (Appeals granted Hearing- Both Parties Present) from FHC to FSC, by Year 45 40 35 30 25 20 15 10 CHAPTER 5 5 0 2005­06 2006­07 2007­08 2008­09 Civil Criminal Labor 59 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts appeals to the FSC are very rare. Meanwhile, civil and criminal cases which only reach roughly 31 and 28 percent respectively for applications from the FFIC to the FHC, reach highs of nearly 45 and 40 percent respectively for applications from the FHC to the FSC. A second observation is that with the exception of labor applications and the last two years of civil applications to the FSC, all rates are increasing considerably over the 4 year period. This is a cause of concern for the judiciary as a whole, because of its perception that appeals are largely used as a dilatory maneuver. The FSC does save some work by reviewing all appeals and only proceeding to a full hearing with those deemed with merit. Still the review takes the work of 3 justices and does involve a hearing for the appellant. Another solution might be to raise fees (or impose some of cassation), but an interest in maintaining or raising levels of access might argue against its adoption. Third, the FHC seems to be substantially cutting the rate of admitted cases in all areas (see Fig- ures 19 and 20, noting that the scaling has changed so that the 60 percent rate for applications in labor in 2008­09 is reduced to 10 percent on admission). The FSC has apparently not done as well, reducing applications to admissions by only a few points. In fact the reductions are so slight that if one does not consider the scaling, the tables appear identical. We next look at appeals rates (applications) for the most common types of civil cases. It was not possible to do this for cases appealed from the FFIC to the FHC, so only FHC to FSC cases are considered. Table 23: Appeals Rates (Applications) by Year for Major Types of Civil Cases, FHC to FSC Types of Civil Cases 2005­06 2006­07 2007­08 2008­09 Immovable Properties (House) 19 23 43 31 Contracts 62 51 65 42 Execution 45 46 40 52 a Inheritance 23 20 6 0 Movable Properties 18 25 26 42 a The figures for inheritance in the last two years appear to be problematic, possibly a result of errors in data entry. Although the highest rates were not reached in the last year considered, only inheritance cases seem on a significant decline. Contracts cases have gone up and down (and remain very high CHAPTER 5 even in the last and lowest year) while execution cases are increasing as are those involving movable properties. Since some of these cases may have originated in the FFIC and gone to the FSC on second appeal, the tendency of the FHC to uphold or overturn FFIC judgments is another factor in determining the number of applications going to the FSC. Of course, not all 60 Uses of Justice ­ For What are the Courts Used? applications will be admitted, but as we have seen the FSC seems to have a very loose filtering system, admitting most of what is requested on appeal. Conclusions Our review of what gets to court leaves many of the initial hypotheses untested. However, a few findings as relates to them and to other issues do merit mention. First, in the civil area, "contract enforcement" cases are not as prevalent as was believed, and as has been found elsewhere. This was already tentatively suggested in Chapter IV, and is substan- tiated here by the review of the most common civil proceedings. As noted earlier, this may have more to do with exogenous factors--patterns in making loans and issuing credit--in Ethiopia than in any shortcomings of the courts. While one does not want the courts to become a debt collection agency for banks and retailers, as more credit is issued, contract enforcement espe- cially for smaller loans is likely to become more important. Second, issues regarding property, both movable and immovable, are important at the higher instances and apparently are generating more controversy despite the fact that "land" is owned by the state and thus not subject to judicial decisions. Especially in urban areas (where the federal courts largely operate) ownership of buildings and right to construct are increasingly controversial, and not surprisingly accompanied by charges of considerable administrative corruption, explaining in some part why cases are then taken to court. Addis Ababa's adminis- trative land tribunals have been under investigation by the Federal Ethics and Anti-Corruption Agency, and some municipal judges have also been accused of illegally taking on "land" cases. The high percentage of declaratory judgment cases in the FFIC is seen in part in the rising importance of adoption cases. Ethiopia is currently one of the three most popular countries for foreign adoptions and this is clearly generating business for the First Instance Courts. We did not separate out other declaratory cases, but it is estimated that they account for 28 percent of the workload in the two courts included in the sample. The large number of execution cases recorded in the FFIC and their substantial representation at the other instances is an growing concern for the courts, and will be treated in more detail in Chapter VIII. These are cases where winners request assistance in enforcement or where losers protest the form enforcement takes, and in either instance are regarded as arising in efforts to resist payment of claims. CHAPTER 5 Finally, and related to all the above comments, appeals rates in Ethiopian courts remain very high, and despite court screening of applications, are still not under control. By law, the right to a second appeal has been restricted, but this has still not brought rates to what might be con- 61 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts sidered a reasonable level. One obvious solution, to charge high fees for appeals and cassation (where no fee is charged) is probably not advisable in Ethiopia because of its impact on access to justice. However, as the courts, realize something needs to be done as many appeals appear to be filed for purely dilatory purposes. We return to this issue in Chapter VII on delays and their causes. CHAPTER 5 62 CHAPTER 6 How Do the Parties Fare? Introduction Our interest here is in determining how the different types of parties fare in the major catego- ries of cases. Outcomes are important because just filing a case is one thing, and having it de- cided in one's favor is another. There are several hypotheses about outcomes that can be drawn from other studies and other countries. For criminal cases, prosecutors will only indict those who they have some chance of con- victing--thus conviction rates over indictments should be well over 50 percent Contract enforcement cases are usually decided in favor of the plaintiff, but often after very lengthy delays--in any open and shut case, the defendants lawyer will aim to postpone judgment Between private parties, organizations tend to win over individuals, largely because they can afford better representation (and also because they are often repeat users) There is a small bit of conventional wisdom in Ethiopia as regards these questions: No one wins against the government in administrative disputes FFIC judges favor employees over employers in labor cases Winners and Losers in three major areas (Criminal, Civil, Labor) by Litigation Status (a pro-plaintiff or pro-defendant bias?) Here the study ran into major limitations of the database. Winners and losers of cases were only recorded more or less consistently in the last year (2008­09) and even then we have some doubts about the reliability of the numbers. Thus the following should be taken as illustrative, but with the proverbial grain of salt. It nonetheless is suggestive of some interesting patterns for which reason we have included it despite our doubts. Two findings for civil cases, which also apply to the following tables, deserve attention. First is that at the FFIC level the plaintiff wins twice as often as the defendant, a pattern that is reversed 63 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts Table 24: Civil Cases by year and instances for the FHC original jurisdiction civil cases. Sec- (FICC and HC, original jurisdiction only) ond, nearly half the cases at the FFIC level and over half at the FHC are closed without judg- 2008­09 ment. However, for the FFIC, as "declaratory Litigants FIC HC judgments" figure in the "concluded without Plaintiff wins 6107 81 judgment" category, the results are not as Defendant wins 3886 163 grim as they appear. For other cases, conclu- Plaintiff partial win 1200 27 sion without judgment generally means that Case concluded without 9111 415 the cases are temporarily dismissed because judgement of excessive adjournments owing to the failure of the parties or their witnesses to appear in court. This is part of the judiciary's efforts to Table 25. Labor Cases by Year and Instance eliminate needless delay, but it also has its (FFIC only as all labor cases go there in the downside as regards justice possibly denied. first instance) However, cases can be reopened, and we 2008­09 return to this issue below. We will look at the Litigants FIC positive side of the coin--the speedier dispo- Plaintiff wins 784 sition of cases--in the next chapter. Defendant wins 499 Plaintiff partial win 143 The results for labor cases are similar to those Case concluded without judgement 1300 in the civil cases--the plaintiff wins more of- ten than not but fully half the cases are closed without judgment (for the same reasons). This finding also bears out the Ethiopian belief that Table 26. Criminal Cases by Year and employees win their labor cases in the FFIC, Instance (FFIC and FHC, Original Jurisdiction although whether this is because of "judicial only) favoritism" cannot be said. As all labor cases 2008­09 enter at the FFIC level, there is no column for Litigants FIC HC the FHC which only reviews labor cases on Plaintiff wins 4917 775 appeal. Defendant wins 3129 1549 The results for criminal cases are still more Plaintiff partial win 894 258 interesting, and very negative for the state. Case concluded without 15,663 3174 judgement At the FFIC level, the plaintiff (the state) wins more often than the defendant, but the dif- ference is not as great as above. Rather than 2:1 or 5:2, the state's win rate is only 3:2 (slightly more if partial wins are counted). Taking into CHAPTER 6 consideration that roughly two-thirds of the cases are terminated without judgment, this is a terrible record for the prosecutors. Moreover, at the FHC level, the state loses two thirds of its cases and over half the total are terminated without judgment. The good news is that the judges apparently are not on the side of the state, as if they were the prosecutors' win record 64 How Do the Parties Fare? would be far better and quite probably, the terminations without judgment (usually caused by the prosecutor's failure to get his/her witnesses there) would be less frequent. Winners and Losers in Cases Pitting Organizations against Individuals As one assumption, drawn from the broader literature, is that individuals are at a disadvantage when litigating against organizations, we used the database to test this hypothesis. The follow- ing table, still not definitive, was the result. Here N represents all disposed civil cases in which an individual plaintiff opposes an organiza- tional defendant in the FHC and in the two Table 27. Outcome of all civil cases FFIC courts (Lideta and Arada). (trial stage) for involving individual vs organization Although we have not captured other out- 2008­09 comes (closure without judgment, or organi- FIC HC zation wins) the results at the FFIC are particu- Litigants N=3156 N=647 larly significant, as individuals win these cases Individual plaintiff wins 1991 129 wholly or partially over two-thirds of the time. over ORG defendant Their showing in the FHC is less dramatic, but Individual plaintiff wins 164 0 does indicate that they can win even when the partially over ORG defendant stakes are higher. Closed and Reopened Cases One effect of limits on adjournments is the tendency for a large proportion of cases to be tem- porarily closed with the possibility of reopening them latter. While this policy decreases delays, it also may work against many cases ever reaching any resolution. It should be noted that at the FSC, cases are not closed without prejudice (or for excessive adjournments). Closures without judgment are generally cases dismissed for lack of merit. The very low numbers of reopened cases usually already have judgments but are reopened because new evidence or legal argu- ments provide the basis for reconsidering the case. We do this analysis by parts, first looking at the absolute numbers of cases reopened at each in- stance and by legal area and years. Given the great difference in numbers at the various levels, CHAPTER 6 it is nearly impossible to appreciate those at the FSC level. When one considers the number of cases filed at each instance, and the number of cases closed without judgment at the FHC and FFIC levels, it is already apparent that reopened cases are only a small portion of the caseload--or in other terms, that most cases closed without prejudice are 65 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts Figure 23: Number of Reopened Civil Cases ­ FFIC, FHC, and FSC by year 3000 2500 2000 1500 1000 500 0 2005­06 2006­07 2007­08 2008­09 FSC FHC FFIC Figure 24: Number of Reopened Criminal Cases, FFIC, FHC, and FSC by Year 450 400 350 300 250 200 150 100 50 0 2005­06 2006­07 2007­08 2008­09 FSC FHC FFIC Figure 25: Number of Reopened Labor Cases, FFIC, FHC, and FSC, by Year 350 300 250 200 150 100 50 CHAPTER 6 0 2005­06 2006­07 2007­08 2008­09 FSC FHC FFIC 66 How Do the Parties Fare? not reopened. The only significant increase is in labor cases at the FFIC level, where the numbers of reopened cases have gone steadily upwards. In all the other areas and instances, some initial increase seems now to be on the decline. To make this point still more apparent, we now compare reopened cases as a percentage of all annual filings at each level by year. Table 27. Percentage of Reopened Cases against Filed Court Case Type 2005­06 2006­06 2007­08 2008­09 FSC Civil 1.3% 1.2% 0.7% 0.4% Criminal 0.2% 0.5% 0.6% 0.0% Labor 1.0% 0.4% 0.7% 0.5% FHC Civil 9.3% 11.6% 8.8% 3.6% Criminal 4.4% 4.8% 3.5% 1.0% Labor 3.3% 4.6% 3.2% 1.2% FFIC Civil 22.0% 12.0% 12.0% 11.0% Criminal 3.0% 4.0% 2.3% 2.1% Labor 13.5% 14.0% 14.2% 12.6% As can be seen the highest percentages are reached in the FFIC, and although they have declined from earlier levels (even in labor), they still constitute over 10 percent of the caseload in civil and labor. Still when one considers the still higher percentage of these cases closed without judgment this means that at most a quarter of them are reopened. In criminal cases the results are still lower and again reflect on the prosecutors' apparent discinclination to pursue their cases. As noted by an FSC source, many criminal cases may in fact be reopened not by the prosecutor but by defendants after a higher level court quashes the initial verdict on ap- peal. What this means is that a poorly formulated criminal case, and one where the prosecutor cannot get the witnesses and defendant to trial is for all intents and purposes concluded. How much hardship this works on the defendant (who may be spending time in pre-trial detention) is another question, but in any event the problem is not with the courts but with the other ele- ments of the criminal chain. For curiousity's sake we also examine the number of reopened cases by type of dispute in the FHC and FFIC. The FSC is not included as the numbers are extremely low, the only exception being 40 cases for termination of employment reopened in 2008­09. CHAPTER 6 What stands out here is the large number of reopened civil cases, especially for execution in the FFIC and in the FHC, and the general increases in all civil categories for the former. What this implies is that even though the percentage of reopened cases is declining, in the most com- 67 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts Table 28. Reopened cases in FHC for Major types of Civil Cases No Major Civil Cases 2005­06 2006­07 2007­08 2008­09 1 Contracts 23 0 40 27 2 Immovable Properties (House) 36 0 42 63 3 Movable Property 23 0 15 8 4 Inheritance 24 0 31 31 5 Execution 73 0 123 87 Table 29. Reopened cases in FHC for Major types of Criminal Cases No Major Criminal Cases 2005­06 2006­07 2007­08 2008­09 1 Murder 20 15 14 9 2 Drugs 7 0 7 2 3 Fraud 11 49 50 28 4 Robbery 45 119 86 36 5 Bodily Harm 0 0 1 0 Table 30. Reopened cases in FHC for Major types of Labor Cases No Major Types of Labor Cases 2005­06 2006­07 2007­08 2008­09 1 Termination of Employment 14 24 37 13 2 Promotion 1 2 1 0 3 Back Pay 9 11 5 2 4 Transfer 1 2 0 0 5 Compensation 1 1 3 1 Table 31. Reopened cases in FIC for Major types of Civil Cases No Major Civil Cases 2005­06 2006­07 2007­08 2008­09 1 Execution 964 801 982 1028 CHAPTER 6 2 Adoption 82 98 59 361 3 Diverse Monetary Claims 396 134 347 657 4 Divorce 632 741 829 734 5 Inheritance 98 101 177 171 68 How Do the Parties Fare? Table 32. Reopened cases in FIC for Major types of Criminal Cases No Major Criminal Cases 2005­06 2006­07 2007­08 2008­09 1 Theft 105 82 74 70 2 Bodily Harm 48 67 97 128 3 Rape 27 26 16 11 4 Defamation 19 31 36 60 5 Breach of Trust 12 13 12 16 Table 33. Reopened cases in FIC for Major types of Labor Cases Major Labor Cases 2005­06 2006­07 2007­08 2008­09 1 Termination of Employment 52 51 63 64 2 Promotion 0 0 0 5 3 Back Pay 20 40 50 83 4 Disciplinary Issues 4 0 0 0 5 Compensation 10 11 15 11 mon disputes they are not declining as fast. The rather erratic patterns as regards reopening of criminal cases deserves more exploration, by the prosecutors themselves, to determine why certain cases tend to be reopened more often than others. Other Factors Affecting Outcomes: Amounts at Stake and Legal Representation We mention these factors as possible explanations of outcomes, but unfortunately cannot track them in the database. Legal representation we assume should have some effect, especially when a represented client goes up against with one operating without counsel. The problem here, as mentioned in the introductory section, is that the number of apparently unrepresented clients is probably highly inaccurate. While a majority of court users do not have counsel, for some who do, that fact is not registered. This is a maneuver used by lawyers to avoid register- CHAPTER 6 ing taxable income. In any event, it is an important question not only for Ethiopia but for other countries where pro se representation is allowed. We can surmise some difference in viewing the outcomes in criminal cases in the FFIC and FHC, as in the former defendants are less likely to have representation and they do lose more frequently there. However, otherwise this Ethio- pian peculiarity makes it impossible to investigate further. 69 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts As for the effects of the amounts at stake, it simply proved too difficult to explore this at pres- ent. However, the database should allow work to be done and it would be interesting to follow up with other studies. Another factor that may influence outcomes is notification, notoriously difficult in developing countries. However, again the court's database does not include suffi- cient information to allow us to explore this issue. Government Litigation ­ Does the Government Always Win? We have already seen that individuals operating as plaintiff can win over organizations, but have yet to test the Ethiopian assumption that the private party (organization or individual) never wins, either as defendant or plaintiff, against the state. Such "administrative cases" (those to which the government is a party) are heard by civil benches. Again, the database presented some difficulties, so what was done was to review the outcome of appellate cases in the FHC and FSC. The numbers in each box (the N) represent all appeals entered where a private party (individual or organization) is pitted against a government agency. It includes both administrative and criminal cases (although the latter are virtually limited to individual not organizational private parties). Four scenarios are considered depending on who enters the appeal and who the other party is. The percentages represent the outcomes favorable to the appellant, and thus over- turning a lower-level decision. A first general observation is that the government does not always win, although for the most part it seems to do considerably better in the FSC than in the FHC. Second, at both levels, private organizations seem to do better than individuals both as plaintiffs and defendants. This may in part reflect the fact that many cases involving individuals are criminal matters. Third, Table 34. Government Litigation 2005­06 2006­07 2007­08 2008­09 Outcome SC HC SC HC SC HC SC HC Individual appellant against 17% 78% 16% 73% 13% 52% 10% 67% government 1284 1215 1503 1069 2116 2263 2471 2152 Organizational appellant 32% 78% 23% 70% 83% 83% 7% 86% CHAPTER 6 against government 19 70 52 63 47 47 68 37 Governmental appellant 69% 3% 74.5% 3% 65% 2.2% 71% 2.1% against Individual 541 6642 514 5309 685 6509 696 6224 Governmental appellant 44% 18% 69% 15% 71% 4% 82% 6% against private organization 32 105 36 105 63 74 49 67 70 How Do the Parties Fare? appeals involving private organizations are relatively rare across the board although not so in- frequent as to preclude identifying some trends. They are also the most interesting as they are typically purely administrative cases, and thus do not mix in criminal appeals where a different set of incentives may be operating. In any event, as the table demonstrates, the government is not always the victor whether operating as an appellant or a respondent to an appeal. Some further investigation is needed in the FFIC and also to identify patterns as regards the type of cases government wins or loses, both on appeal and at the trial stage. Conclusions A first observation is that our analysis has only touched the surface as regards what could be done with the database. However, some of the findings are extremely important, and, since this is not an analysis usually done by the Ethiopian judiciary, may even offer some novelties for its members. Of the three hypotheses drawn from studies in other countries, only two could be partially investigated and to the extent we could explore them they did not hold up entirely. First, prosecutors' low conviction rates, the large percentage of criminal cases concluded without judgment, and the low percentage reopened do not support the general theory that prosecutors "indict to win." In fact, the reference, made by an Ethiopian commission charged with reviewing criminal justice problems, to prosecutors' "charging to fail" seems more appropriate here. While the commission did not intend to accuse the prosecutors of intentional failure, the phrase does capture some major problems in the criminal justice system. How these problems may harm defendants who in the end are not found guilty remains to be seen--as we have no way of telling whether the cases merited prosecution albeit with more preparation and better orchestration by the prosecutors. It is particularly telling that the prosecutors do even more poorly with cases initiated at the FHC which are by definition for more serious crimes. With the recent introduc- tion of "real time courts" where defendants may simply plead guilty in the face of overwhelming 49 evidence the conviction rate may go up, but this is not necessarily the best solution. Second, the notion that individuals cannot win against organizations appears to not hold up or at least not when the individual is the plaintiff. It is encouraging that many individuals are will- ing to take on organizations in the federal courts, and the outcomes are a sign that the courts are willing to give them a fair shake. The fact that individuals can also win against public sector organizations is likewise important and contradicts the Ethiopian expectation that the state always wins. Here it would be important to explore in what kinds of cases the individual wins and whether this is an increasing pattern or not. CHAPTER 6 49Conviction rates for these courts which in theory review in flagrante crimes is said to be about 95 percent, with 80 percent of the defendants pleading out at the beginning. 71 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts Third, we have not been able to explore fully the notion that contract enforcement cases (or at the FFIC level, "diverse monetary claims") are usually won by the plaintiff, but only for lack of time. The database would allow this investigation and it probably should be done. However, the increasing incidence of "execution" (enforcement of judgment cases) in original jurisdiction, on appeal, and as reopened cases is a counter tendency, suggesting that once a defendant in a debt or related cases loses, s/he will then proceed to contest the enforcement, fairly or not. The FSC is well aware of this problem and as discussed in a later chapter, is seeking ways to address it. Finally, while not a theme taken from other countries and studies, and not part of Ethiopians' conventional wisdom about their courts, the issue of reopened cases is important. As we saw, up to half (even more for criminal cases) of cases seen at the FFIC and FHC are terminated without judgment. Most of these cases could be reopened, but the amount that are is less than a quarter of the total and appears to be declining. Since these cases are usually closed because of parties' failures to appear at hearings or get their witnesses there, the general practice can be considered positive--the courts' effort to enforce discipline on the court users. However, in a country like Ethiopia where many parties have no legal representation and may not fully understand the rules, this is a concern and an area, one among many, where the courts may want to further investigate the reasons. As regards criminal cases, this is not their problem as it is evident that the failures lie with prosecution and police. However, for civil and labor cases, it might be well to determine whether the reasons are parties' failure to understand the rules, their inability to comply, or simply a lack of sufficient interest to go on with the case. CHAPTER 6 72 CHAPTER 7 Delays and Their Causes Introduction One of the problems confronted by justice systems universally is a tendency, especially in the face of rapidly growing demand, for cases to take "too long" to resolve. This is a common com- plaint of citizens, a blight on the judicial image, and thus a problem targeted by judicial and broader justice sector reforms worldwide. While the saying "justice delayed is justice denied" has its detractors (who commonly cite overly speedy trials as an equally injurious vice), it is evident that lengthy delays can defeat a legitimate plaintiff and also discourage people from taking cases to court. However, delay is a complex issue, if only because, absent legal standards (which themselves may be arbitrarily and unrealistically set) delay is very much a subjective appreciation. One person's delay may be another's sufficient time for judicious consideration of the facts and issues of a case. There are those who argue (Mohando, 2009) that a certain amount of delay is in fact necessary to discourage abusive use of the courts or to encourage disputants to reach extra-judicial agreements of minor conflicts. Aside from these legal, philo- sophical, and behavioral considerations, there are other issues surrounding the definition of delay and the identification of its causes: Absent fixed legal standards (which do exist internationally, but may be unsuitable for many countries) it may be more appropriate to speak in terms of "duration of cases" and to set goals for the reduction of average processing times or for the resolution of X percent- age of cases within defined periods. We will continue to speak of delay here, but it is well to keep this caution in mind. One problem with legal standards (and with popular appreciations) is that many of them do not take into consideration the different requirements for different types of cases. It is frequently observed that some common perverse impacts of setting broad standards (all civil or criminal cases to be terminated within X months) are that judges will tend to cherry pick (focus on the easy cases), that they will regard the time limit as license to draw out even easy cases to meet it, and that many complex cases will end up being terminated without judgment. Delay can be a result of overburdened judges or of judicial inefficiency and thus in deter- mining whether cases are taking too long to process it is also necessary to understand what judges are receiving and how they are processing their work. 73 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts 50 Delay is often party-initiated--one party (and sometimes both ) often benefits from delay and thus may use a variety of "dilatory tactics" to generate it. Delay can also be caused by attorneys independently of and sometimes contrary to the interests of the parties they are representing (World Bank, 2002). Attorneys may time their cases to meet their own schedules, or if payment agreements encourage it, draw them out or add extra steps to be able to charge more. Judges can combat party or attorney initiated delay, but this requires support from their superiors as they may otherwise be the victims of user-complaints. In criminal cases, delays may also result from those further up on the criminal justice chain--police and prosecutors who like judges may be overburdened, inefficient, or just not very dedicated to their jobs. Police or prison authorities responsible for getting defendants or pre-trial detainees to courts may also cause delay, sometimes as a result of corruption (refusing to transport prisoners unless paid or taking payment for not delivering the detainee or non-detainee to court). Delays may also be a consequence of the parties' incomplete understanding of the rules, especially but not exclusively when they are not represented by a lawyer. While the emphasis on "orality" (substituting hearings for written submissions) as a means of reducing delay has become very popular with donors and in some regions (e.g. Latin Amer- ica), absent strict control of party and witness appearance in hearings, orality opens up still more potential for generating delays. Moreover in countries with problems of transporta- tion and communication, or where courts are located in few areas, there may be legitimate reasons for non-appearance, especially for rural populations and for the poor in general. And finally, when cases take "too long" to process, the problem may be generalized (all phases take too long) or limited to only one or a few steps. For example, at the trial stage, the problem may be in slow notification, excessive adjournments of hearings, or problems in identifying assets to support payment in a civil claim. Or the initial judgment may occur rapidly, but the slowdown may come on appeal or in enforcement. As the above suggests, combating delay effectively requires first defining it and identifying where it occurs and then finding the reasons (and the remedies). All delay is not the same, and thus all the solutions are not similar. In fact, experience demonstrates that absent a good prior analysis of what is really happening, reforms may do nothing to improve or may even complicate the situation. In reviewing Ethiopia's experience we are thus interested in two main questions: how long it takes to resolve a case and when there are delays, what causes them? Two further sets of questions, inspired by the Ethiopian judiciary's efforts to increase productiv- ity and speed up case processing involve the visible effects of these efforts over the time period covered, and how this relates to some of the specific policies adopted (especially as regards CHAPTER 7 control of the number and length of adjournments). Finally, we will also consider some possibly 50 It is not unheard of for plaintiffs to draw out cases in the expectation of collecting more money in the form of interest and fines. 74 Delays and their Causes perverse consequences of these latter policies--as regards any tendency to close too many cases too rapidly and the chances that they will be reopened later. Trends in Times to Case Disposition Ethiopia does not have precise legal standards for defining delay, but rather a general goal 51 of shortening case processing times. This can be considered the courts' principal policy for reducing delay (followed by, as discussed below, an effort to limit the use of adjournments). However, the goals set are reasonable by international standards, and in fact somewhat ahead of those set for universal application. This is logical insofar as much of what gets to the federal courts is at least theoretically fairly simple. Moreover, most parties appear to lack legal repre- sentation meaning that the arguments advanced are less likely to involve complex interpreta- tions of existing law. Of course, that also means, as we will discuss later, that parties may be insufficiently protected or even unable to present a case that will be admitted to the higher courts in particular. In this section we will only look at trends in times to disposition of cases at the three court levels. In attempting to limit case processing time, the courts are relying on judges' incentives--judg- es will be evaluated on the basis of their ability to meet the goals, generally set in terms of the percentage of their workload that should be decided within certain periods (e.g. so many in one month, so many in under six months, and so on). This is a good policy as it allows for differ- ent time requirements and insofar as some of the time periods are short, also encourages judg- es to allocate the time needed to resolve each case--the easy ones will be decided quickly, and there is a margin for more complex cases taking more time. The policy does assume that the goals are feasible--that judges are not being asked to do the impossible--and thus will require support to judges who for some reason are faced with an insurmountable challenge, too many cases, too many complex cases, etc. However, since the benchmarks are based on experience, and are gradually moved as performance improves, they are likely to be reasonable. In addition, as we saw in Chapter III, the court is also adding judges as workloads become overly large. Here we look first at mean disposition times for criminal, civil and labor cases in the three court levels (again using only a few first instance courts for this exercise) and then at times for major types of civil cases, again in all three instances. It should be remembered that only the FFIC have original jurisdiction cases only and that the FHC has both these and appeals, while the FSC has nearly exclusively appeals (and cassation) cases. As a second, and very important note, disposition does not mean judgment--it also includes cases temporarily and permanently CHAPTER 7 closed without judgment. 51 The procedural codes do of course define time limits for filing a case and for other actions by the parties, but do not say how long a case should last. 75 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts Figure 26: Mean Times (Months) to Disposition for Criminal, Civil, and Labor Cases, FSC 18 16 14 12 10 8 6 4 2 0 2005­06 2006­07 2007­08 2008­09 Civil Criminal Labor Figure 27: Mean Times (Months) to Disposition for Criminal, Civil, and Labor Cases, FHC 60 50 40 30 20 10 0 2005­06 2006­07 2007­08 2008­09 Civil Criminal Labor Table 35. Mean Times (Months) to Disposition for Criminal, Civil and Labor Cases, FFIC (for civil and criminal, Arada only, for labor, Yeka only) Case Type 2005­06 2006­07 2007­08 2008­09 Civil 10.03 10.11 5.32 5.52 Criminal 7.42 6.00 4.28 2.27 Labor 5.33 6.44 7.73 4.03 CHAPTER 7 76 Delays and their Causes As the above makes clear, even over the short period covered, the average times for case disipositions have been declining in all areas and in all instances. The FSC is a slight exception, but after an increase in 2006­07 it has brought times back down to the initial levels. As we will see below, the FSC problems tend to lie in parts--its civil bench, hearing appeals, tends to be slower for a number of reason, and in fact, the brief addition of an extra civil bench in 2007­08 was intended to clear up its backlog. It should be noted that while the FHC shows a steadier reduction, the scale had to be changed to reflect the longer times it generally takes in relation to the two other instances. Having reviewed the overall picture we now look at times to dispositon in the most common types of cases, again by instance and by year. Figure 28: Mean Times (in months) to Disposition for Major Types of Civil Cases, FSC Moveable Properties Inheritance Execution Contracts Immovable Properties (House) 0 2 4 6 8 10 2008­09 2007­08 2006­07 2005­06 Figure 29: Mean Times (in months) to Disposition for Major Types of Civil Cases, FHC Moveable Properties Inheritance Execution Contracts Immovable Properties (House) CHAPTER 7 0 5 10 15 20 2008­09 2007­08 2006­07 2005­06 77 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts Table 36. Mean Times to Disposition (Months) for Major Types of Civil Cases, FFIC (Arada Only) Major Civil Cases 2005­06 2006­07 2007­08 2008­09 Execution 10.11 5.50 2.00 3.01 Adoption 3.67 5.40 6.00 3.45 Diverse Monetary Claims 14.00 19.00 10.00 7.00 Divorce 0.24 0.23 0.08 0.24 Inheritance 17.00 12.00 6.00 8.80 The decrease in mean times to disposition for civil cases is most dramatic in the FFIC, and espe- cially in those cases (execution, diverse monetary claims, and inheritance) that started with the highest averages. Here mean times have dropped by half or even more. Trends for the FSC and FHC are more erratic, with contract cases being the most problematic for both. However, the FSC's times are significantly lower than those for the FHC (note that the scale changes on the two graphs) with few FHC averages being as low as even the highest FSC numbers (8 months for inheritance and contracts). Moreover, it should be remembered that the average caseload in the FSC is significantly higher than that in the FHC. Trends in Factors Influencing Delay: Number and Length of Adjournments The second principal policy adopted by the courts to reduce delays is to limit the number and length of adjournments. In oral systems, adjournments are a frequent source of delay and moreover, are often used by the parties to that end. However, adjournments may also be judge initiated (an indisposed or simply unprepared judge who decides to postpone a hearing), by the lawyers (because of schedule conflicts or a similar lack of preparation) and by parties and witnesses who for some reason cannot or will not show up for the hearing. The current policy on adjournments, intended to limit their use, is as follows: No judge-initiated adjournments are allowed In civil cases, two adjournments are allowed, and these in effect are the "normal ones"--the setting of future dates for the pre-trial hearing and the trial. Where the plaintiff or both parties fail to appear, the case will be closed Where the defendant does not appear, the hearing will go on without him/her CHAPTER 7 The general aim is to have hearings continuous without any intermediate recesses. The policy is incentive based and only made possible by its inclusion in the database. Judges are evaluated as regards their performance on adjournments and thus have a strong reason to limit them and especially to curtail their own initiation of postponements. Attorneys and 78 Delays and their Causes parties, at least those interested in getting a judgment, also have an incentive to comply. As regards, parties and attorneys more interested in delay, no new incentives are added and enforcement depends on the judges' own motivation. The same applies to other actors (police, prosecutors, and prison authorities responsible for transporting detainees) although over the longer run, their agencies could also impose disincentives for non compliance. To some effect, this is already happening with prosecutors, if in a slightly perverse fashion. As judges will close cases when a party or key witness does not show up for a hearing to avoid an "excessive num- ber of adjournments" this has had a negative impact on prosecutors' conviction rates which are based on the ratio of convictions over indictments. Federal prosecutors have attempted to ar- 52 gue that this calculation (done by the Ministry of Justice) is unfair. Having lost that battle they are now having to ensure that once an indictment is registered, the case is completed without further delays (meaning that defendants, witnesses, and the prosecutor him or herself must show up for scheduled hearings and limit requests for further postponements). Currently the Court tracks both numbers and lengths of adjournments as both can contribute to delays. Thus, the following figures illustrate trends in both areas over time. If the data are accurate, the FSC is not doing all that well with its own policies. However as noted earlier, one problem is that the database self corrects, meaning that when a judgment is nullified, the case is recorded as though it never occurred. Thus some of the categories (e.g. inheritance and contracts) with excessive numbers of hearings may well include those carried over and re- opened. In any event, the civil bench of the FSC is regarded as most problematic in the area of de- Figure 30: Mean Number of Adjournments for Major Types of Civil Cases (FSC) Moveable Properties Inheritance Execution Contracts Immovable Properties (House) 0 1 2 3 4 5 6 2008­09 2007­08 2006­07 2005­06 CHAPTER 7 52 They argued that contrary to conventional practice, the conviction rate be calculated as convic- tions over judgments. This would raise the rate from the current 30 percent or so to at least 50 percent given that up to 60 percent of the indictments now end in closure without judgment.. 79 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts lay and clearly a too lenient policy toward adjournments is one reason. However, 5 adjournments was the earlier goal, and still much better than what has been observed in other countries. Figure 31: Mean Length (in days) of Adjournments for Major Types of Civil Cases (FSC) Moveable Properties Inheritance Execution Contracts Immovable Properties (House) 0 10 20 30 40 50 60 70 80 90 2008­09 2007­08 2006­07 2005­06 The FSC's success in limiting the length of adjourments is more positive, even in the areas like contracts and inheritance where the number of adjournments has increased. As the efficacy of the policy depends on both factors, progress is being made, and once again, it is extremely rare that any court is able to track and thus attempt to control these variables. We review the same factors for the FHC, mean number and mean length of adjournments for the most common civil cases. Figure 32: Mean Number of Adjournments for Most Common Civil Cases (FHC) Moveable Properties Inheritance Execution Contracts Immovable Properties (House) CHAPTER 7 0 1 2 3 4 5 6 7 2008­09 2007­08 2006­07 2005­06 80 Introduction to Study and Explanation of Objectives Figure 33: Mean Length of Adjournments (in days) for Major Types of Civil Cases (FHC) Moveable Properties Inheritance Execution Contracts Immovable Properties (House) 0 20 40 60 80 100 120 2008­09 2007­08 2006­07 2005­06 While the FHC has too many adjournments (up to seven in some areas), it is graudatlly reduc- ing both the number and the length, with, like the FSC, more luck reducing the length than the number. Contracts are problematic for both instances, but in the FHC, it is movable property that appears to generate more postponements. Still on a whole, the FSC policies apepar to be having a decsive impact. Trends in Other Factors Influencing Delay: Appeals Rates and Appeal Outcomes Another factor affecting delays for total case-processing time is appeals. Appeals are a neces- sary part of any judicial process, and a protection for the parties against judicial error, arbitrary or idiosyncratic decisions, and intentional miscarriages of justice. However, as is often observed they are also a common delay-creating tactic, and moreover, because they both create delays and often higher costs, inevitably work against the party with fewer resources. The threat of an appeal can also be used to leverage negotiations with the party that won on the first round; it is less clear that this is entirely negative, but it often comes as a shock to a one-time court user who now finds he or she will either face several years of delay (and an uncertain outcome) or the need to relinquish some of the initial reward. So-called post-verdict or post-judgment bargaining is apparently not a common practice in Ethiopia, but it is frequent in other countries (including in the US for civil suits). CHAPTER 1 Countries vary considerably as to the parties' tendency to apply for appeals, the circumstances under which an appeal can be requested and will be granted, and the time it takes to process both the application and the appeal itself. Whereas the right to an appeal, whether used or not, was once regarded as absolute, many judiciaries are finding the need to impose some limits. 81 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts This is a very controversial area and one with no easy answers. Apparently simple remedies like limiting appeals to awards over a certain amount can work hardship on many parties with just cause for complaint (and especially the poor). More complex formulas inevitably bring charges of subjective decision-making and other kinds of bias. Nonetheless in countries with high ap- peals rates, delay and the injustices it brings are a growing threat, and thus some solution is clearly in order. The Ethiopian judiciary has still not found a solution to what it regards as an excessive number of appeals and their use primarily as a dilatory practice, an appreciation in part based on the fact (also recorded in the database) that nearly all appeals uphold the lower court judgment. In the first set of tables we look at appeals and reversal rates to illustrate this situation. Figure 34: Appeals and Reversal rates for all cases, FHC to FSC, by Year 45 40 35 30 25 20 15 10 5 0 Appeal Reversal Appeal Reversal Appeal Reversal Appeal Reversal 2005­06 2006­07 2007­08 2008­09 Civil Criminal Labor Figure 35: Appeals and Reversal Rates, all cases, FFIC to FHC by Year 45 40 35 30 25 20 15 10 5 0 CHAPTER 7 Appeal Reversal Appeal Reversal Appeal Reversal Appeal Reversal 2005­06 2006­07 2007­08 2008­09 Civil Criminal Labor 82 Delays and their Causes Patterns for reversals are more erratic in the FHC than in the FSC, but except for labor cases are generally far lower than appeals rates. As noted above, an efficient appeals system should have a low appeals rate, but a 50 percent reversal rate, thereby indicating that only truly difficult decisions are being appealed. Patterns like those seen in Ethiopia suggest a use of appeals to create delay as the chances of reversal are relatively low. One partial solution to appeal generated delay is to speed up processing of appeals. Here the Ethiopians have had more success. As the following table demonstrates they have made con- siderable progress here over the past four years, although with a certain amount of backsliding. Improvements in the FHC are clearly the result both of the policy and the addition of more judges. Although as we saw in Chapter III, the caseload for individual FHC judges is now higher than it was four years before. Table 37. Average Duration (in Months) of Appeals, FSC and FHC, by Major Types of Cases and Years 2005­06 2006­07 2007­08 2008­09 Major Case Type SC HC SC HC SC HC SC HC Civil 3.0 16.5 5.4 11.6 7 6.9 5.7 7.5 Criminal 4.7 12.9 6.0 7.0 2.0 2.2 2.0 2.7 Labor 3.0 2.9 3.6 2.1 2.0 2.5 1.3 1.8 Conclusions Clearly the FSC's efforts to reduce delays in all the federal courts is paying off as shown in the decreases in average time to disposition. The two factors of most importance here have been tracking judges' disposition rates and times, and limiting the frequency and length of ad- journments. As regards the latter, the recent experience has been mixed, but generally shows positive trends. It should also be remembered that we are only reviewing a four-year period and that the bigger changes likely came earlier. At some point, further reducing the number and length of adjournments may only lend itself to very incremental improvements, as these are also to some extent necessary actions. The larger impact may be in limiting the reasons for their occurrence and outlawing judge-induced adjournments. While at the time of writing the Ghana study is still not complete, preliminary results indicated that adjournments requested by CHAPTER 7 parties were only a fraction of those occurring, suggesting that judge-induced adjournments may still play a major role there. We cannot conclude this section without a minor caveat. Raising judicial production and pro- ductivity are clearly important values, and the Ethiopian Federal Courts have made important 83 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts progress in improving both. However, as several observers have warned (Tshuma, 2000 and Mohando, 2009, for example), there are dangers in stressing these values to the exclusion of others. These include the following developments which have been observed in other coun- tries: Where productivity is tracked, judges may "cherry pick" selecting cases that can be resolved quickly Where a failure to admit, or early closure counts as a disposition, judges may exaggerate these outcomes so as to accumulate more dispositions There is no way to tell whether this is happening in Ethiopia, but the high percentage of "dis- missals" is a concern, especially as many of them represent "temporarily closed" cases which apparently are never reopened. Without backtracking on the progress already made, it may be time for Ethiopia's courts to look more closely at the reasons for closure without judgment and to take necessary steps where it appears this is fomenting its own injustices. On the criminal justice side, the ball is in the prosecutors' court, but for civil and labor cases, it is of some con- cern that an emphasis on speed may be prejudicing some court users. CHAPTER 7 84 CHAPTER 8 Enforcement of Judgments Introduction Winning is not everything. You also have to collect. This is one area where the Ethiopian statis- tics are not adequate, but we still can provide some information on the dimensions of the prob- lem and the steps being taken to remedy it. The Supreme Court is well aware of this problem and is already exploring alternative solutions. Problems with enforcement along with excessive appeals appear to be the Federal Court's principal priorities for its on-going reforms. Enforcement of Judgments in Comparative Perspective It is worth mentioning that enforcement and adequate information on its occurrence are 53 problems in many countries. Losers to a civil (or for that matter other kind of ) case often resist enforcement. They may simply refuse to pay, hide their assets, or should assets be identified protest their seizure and sale for a variety of reasons. Should goods go to auction, measures may be taken to prevent their sale or to "sell" them back to the debtor's representative at a low- er price.54 Where registries of movable or non-movable property do not exist or are poorly run, debtors may be aided in all these tactics. A World Bank (2002) study of debt collection cases in Mexico found that more than half stopped at the stage of identifying an asset to embargo (in Mexico done before the judgment) because no assets could be found. Where cases proceeded to judgment without this step, failure to identify assets later often prevented payment--lead- ing to a saying heard throughout Latin America, "gané pero no cobré" ( I won but I didn't collect). Moreover, even where assets were identified, if they were not cold, hard cash, the rest of the enforcement proceedings could take years, with the debtor raising issues at every step, from 53 Labor cases often face this problem as well. A study by Peru's Human Rights Ombudsman found that government and especially municipal authorities were among the worst scofflaws in this regard--claiming inadequate funds to make monetary payments, or their inability to reinstate a worker because "the position no longer exists." (Peru, Defensoria del Pueblo, 1999, Chapter III) A study on labor justice in Honduras (USAID, 2008) identified similar problems with private employ- ers. Interviews with labor lawyers in Brazil (World Bank, 2004) found that employers' resistance to compliance with the judgment commonly led to a post-judgment negotiation with the worker and a reduction of the terms of the award. 54 In Mexico in the economic crisis of the 1990s a revolt among home owners whose mortgages had been foreclosed often stymied efforts to sell off the properties. 85 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts the calculations of interest and fines, to the valorization of the asset, to the holding of a judicial auction. Of the 200 adjudicated cases analyzed in the Mexican study, not a one had gone through the entire process in the 4 intervening years. Although identification and seizure of assets may be less problematic in more developed countries, creditors also face problems there with insolvent debtors against whom a court can do little or with assets that cannot be easily liquidated. A related problem is the lack of adequate court records on enforcement. When an award is paid (or otherwise made) spontaneously, the wise payer has this recorded somewhere, but that may not necessarily be in the court. Even in the United States, judges rarely know whether their judgments have been enforced or whether, as commonly happens there, the awards have been reduced through the practice of post-judgment bargaining between the parties. The assump- tion there, and elsewhere, has traditionally been that if nothing has been heard, the payment was made, but we are increasingly realizing that this may not be what happened. Although increasing court involvement in enforcement has been one suggested means of resolving these problems, it increases judicial workload and often leads to new kinds of corruption, if not among the judges than among a host of other judicial employees whose services may be re- quired to carry out these actions. While these individuals may theoretically work under judicial supervision, there is often a large gap between theory and practice. In some European coun- tries, enforcement has been turned over to executive agents (e.g. Sweden) or to other officers of the court (e.g. the French huissiers). Both practices appear to work well in the countries cited as examples, but would probably not be recommended for less developed regions because of the enormous potential for abusive actions. Execution of Judgments in Ethiopia Turning to the Ethiopian case, as shown in Chapter V, execution or enforcement cases are among the most frequent issues seen by the federal courts. Their incidence appears to be rela- tively stable at the FHC level, but it is growing in the FFIC and the FSC. In the civil area, enforce- ment cases are those where court involvement is requested or required in order for an award to be paid. The winning party may request assistance where the loser is not complying or may require it when assets must be seized, evaluated, and sold. Most cases at the FHC and FSC level involve protests (appeals or cassation) of decisions reached in the FFIC. Because execution or enforcement cases are so common in the federal courts, they have also become the targets of the general goal setting for limiting delays. As shown in the figures and CHAPTER 8 tables in Chapter VII, this has had some positive effects. Mean times to disposition are being tracked, and have been reduced at the High Court level. While fairly stable in the FSC, they do not have the longest times to disposition among the most common proceedings in both the FSC and FHC, and both the number and lengths of adjournments are declining. However, in all 86 Enforcement of Judgments three instances they are among the most common types of cases seen, and this itself is a prob- lem since they put in jeopardy an apparently firm judgment. The problems mentioned above in the context of other countries also occur in Ethiopia, but Ethiopian debtors have been still more creative. According to FSC sources, the most frequent defense against payment, aside from simply displaying empty pockets, is to have a third party suddenly appear demonstrating a proprietary interest in whatever asset has been put up as col- lateral or seized to repay a debt when the debtor claims insolvency. This it is believed, frequent- ly involves fraud--the backdating of deeds, falsification of other proofs of ownership and so on, meaning that the creditor's pretensions go up in smoke, unless he wants to become involved in still another case in which the third party will also be represented. The situation does not appear as grim as that reported in Mexico, and Ethiopia's federal Su- preme Court has responded, not only by creating enforcement benches to specialize in these issues but also with the development of a plan to set targets for and monitor enforcement cases. The following table shows the plan for 2006 to 2009, largely in terms of the targets for carrying out specific actions. Table 38. Plan or Judgment Execution, Federal Courts, 2006­2009, Targets and Achievements 2006­07 2007­08 2008­09 Plan Achievement % Plan Achievement % Plan Achievement % Assessment 1400 657 46% 700 369 53% 700 249 35% of Monetary Value Movable 72 81 112% 72 95 132% 72 80 111% Property Valuation Trials 4000 4219 105% 4000 3816 95% 4000 3087 77% Auction 1200 235 19% 600 259 43% 300 228 76% Payment 1000 834 83% 500 1052 210% 500 613 122% Transfer 1000 365 36.5% 500 351 70% 500 308 62% New Filings 1000 1645 164% 1200 1676 140% 1200 1573 131% As should be apparent from the above, enforcement of judgments frequently constitutes a CHAPTER 8 second trial in Ethiopia. This is regarded as undesirable and the court is seeking ways to address the problem. Given the many factors outside its control--unreliable registries, lack of infor- mation on other assets of the debtor, possibility that he/she has bank accounts outside the country--this will not be easy. However, few developing countries are paying attention to this 87 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts problem and even fewer have the statistical data to inform their decisions. Clearly, the Federal Courts can use the data they already have and refine their database on enforcement cases to develop a better understanding of where the problems lie and which parties are most likely to create them. In some sense, the courts' success in speeding up the time to judgment has pushed those seeking to delay justice to focus on this further step--if you can't hold off judg- ment than you can prevent collection of the award levied against you. Conclusions The conclusions here are brief owing to the relative shortage of information. Execution (en- forcement) of judgments is an area where Ethiopia's federal courts have recognized a problem and are now seeking answers. The creation of specialized execution benches was a first step, but it is uncertain how much improvement this has brought. Enforcement cases have benefited from the general delay-reduction measures. Since times to disposition and adjournments are being tracked, both are declining. However, the underlying challenge is how to discourage the increasingly frequent appearance of these cases, and to combat the many questionable issues and tactics used by losing parties to prevent having to honor the initial judgment. It will also be important to track, as we have not been able to do here, the success rate of these efforts to avoid enforcement through fraudulent practices. CHAPTER 8 88 CHAPTER 9 Conclusions The purpose of this final section is not to repeat all the conclusions of the preceding chapters but rather to give an overview and introduce some ideas that were not sufficiently developed before. Two themes seem important here: the utility of this and similar studies for understand- ing the broader justice issues facing Africa, and the lessons from the Ethiopia research that are more widely applicable. Using Ethiopia as a Case Study The findings should be of use to Ethiopia, although given the amount of analysis the FSC already does, they probably have not added much to what it already knew. The larger question is the utility of this type of study for understanding and addressing justice sector problems in Africa writ large. It would not be impossible to conduct similar work in more nations, although as not- ed, few would provide a statistical database allowing this to be done without drawing samples. This second and more usual technique has been applied in Ghana and it is both more costly and less able to capture the breadth of cases included here. However, each study alone, and the two in conjunction will provide information so far unavailable to inform any African reform. What has not been emphasized previously, but deserves mention now, are the many common- alities across the region in terms of challenges faced. Here Ethiopia's situation is hardly unique; it is only its approach to addressing the problems that is unusual. We have mentioned some of these commonalities already--in terms of the limited use of courts by a majority of the popula- tion, the accompanying reliance on traditional dispute resolution and "hybrid mechanisms," the shortage of attorneys for work in the private or public sector; budgetary constraints on court operations, salaries, and further expansion; and outdated legal frameworks. There are numer- ous other details that could be included: Problems with communication, transportation, and energy. These affect court operations in a number of ways, making it difficult to ensure that the parties, their lawyers and wit- nesses get to court (or even receive notice that they should be there); interfering with the operation of and even the ability to install new ICT equipment; and complicating equip- ment maintenance, especially in areas outside the urban centers. When equipment breaks down, as it inevitably does, there may be no one available to service it. Photocopiers, print- ers and computers distributed to outlying courts frequently go unused for months because 89 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts of this, because basic supplies are not available locally, or because if they are, courts have no funds to buy them. Lack of knowledge about court proceedings among a majority of the population. This ac- counts in part for their unwillingness to use the courts, but also affects their participation when they do. Parties may not understand the need to appear at hearings and most prob- 55 ably are not getting the "messages" court actions are supposed to convey about likely outcomes (what happens if they fail to show up, what deadlines must be met and what the consequences are for not meeting them, what constitutes a justiciable dispute). Lack of understanding can generate frustration when the predictable occurs, and lead to impres- sions that they were somehow cheated. Poor quality of courtroom staff--even in countries trying to improve the professional qual- ity of judges, this remains a problem, and a likely source of delays and corruption. Some judiciaries do not give sufficient importance to the role staff plays, bur many others simply lack the funds (and the available pool of recruits) to attract and retain good people. Poor quality of infrastructure--as noted, Ethiopia has had to reduce the number of federal courts and benches because some buildings were no longer serviceable. Court installa- tions are not necessarily worse than those of other government and even private organi- zations, but their poor condition, and sometimes inconvenient locations can aggravate a series of other problems Inadequate performance of other sector actors--police, prosecutors, public defense and private attorneys. Courts can at best exercise indirect influence over these individuals, and they must do so with care or the lack of cooperation could become worse. Political and other interference with judicial decisions and operations. Whatever the level of constitutionally guaranteed independence enjoyed by the courts they operate in a po- litical environment and to a greater or lesser extent are affected by it. Most of these problems are not unique to Africa, but they often take on extreme dimensions there. Because of their individual and combined influence, it thus becomes important to know how this affects "normal" expectations as to what courts can do and how far improvements to their performance can advance. Those problems arising directly or indirectly in such factors as budgetary constraints and relative citizen ignorance as to formal operations will probably not be resolved soon. However, this does not necessarily mean condemnation to a low level of performance and may simply require more creativity in designing realistic remedies. The dearth of available information on what African courts do is thus the motivation behind this and the Ghanaian research. Two countries are hardly representative of all of Africa, but bet- ter information on them can sharpen our appreciation of what the real problems and advances CHAPTER 9 55 Authors like North (1990) have put enormous emphasis on the signaling function of courts--the notion that predictability of outcomes tells potential users what their results are likely to be and so helps them decide whether to take a case to court. 90 Conclusions are. Five studies in Latin America also barely scratched the surface, but did call attention to many unrecognized emerging problems such as low enforcement rates; the increasing impact of government litigation and the role government plays in augmenting court congestion; the prevalence of dilatory practices, including unnecessary appeals; certain types of corruption not normally noticed; and the often low workloads of presumably over-burdened judges. The Ethiopian study has illuminated some of these same traits in its courts--high appeals rates, problems with enforcement of judgments, and the other dilatory maneuvers activated by par- ties and their counsel to avoid justice. (We don't say discovered because the judiciary already recognized these problems, but it is important that the knowledge be shared with others which is one role this report can play). However, the study also has revealed some positive findings. Although the courts covered are federal, and are physically located in two cities which also have municipal court systems, they seem to attract a variety of ordinary cases, and not just disputes of concern to more elite par- ties. In fact, civil cases are largely between individuals and women appear as plaintiffs in more than half of the FFIC civil disputes. Average claims even at the FFIC level are somewhat high (about US$2,000) but this may be because smaller claims are sent to the municipal courts. This may also explain the relative scarcity of cases where organizations sue individuals. In systems surveyed in other regions, these were sometimes a majority of the civil caseload in the first in- stance courts, leading to the assumption that the latter were used largely to collect small debts. It was also significant and probably positive that in civil cases the plaintiff more often won than not. However, in criminal cases, the state (the usual plaintiff ) does very poorly--it wins only half of its cases and fully two-third are closed without judgment. This problem is well known in Ethiopia, and does not originate with the courts, although the judicial policy of closing cases for excessive adjournments probably contributes to it. Still if the prosecutor cannot arrange to get his witnesses or even the defendant to a hearing, there arguably is no reason to continue, and closure without prejudice may be better than a default judgment against the state. Extrapolated to Africa in general, the conclusions are still more important--the fact that judges can manage what many might consider overly high caseloads; their ability, even in this situa- tion, to keep up with growing demand; the potentially extensive use of courts by women; the use of civil courts primarily to address disputes between individuals; the relatively lower inci- dence of repeat users (who in Latin America often flood the court with cases either as plaintiffs or defendants); the low success rate of prosecutors (positive only in the sense of suggesting that the courts can and often do rule against the government); and the fact that low fees and the ability to litigate without legal counsel has broadened access to more citizens. (Lawyers can and have objected to the positive tilt put on the latter finding, but where counsel is scarce the CHAPTER 9 only alternative is, as in much of Latin America, to restrict court use still further). In short, even given its limited reach and the myriad other challenges, the formal system is playing a growing role in dispute resolution in Ethiopia, and arguably elsewhere in Africa. 91 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts However, in Ethiopia this role is in large part due to the proactive approach taken by judicial leadership, and here we turn to the lessons the study generates. Lessons of Wider Utility The first and most obvious point is to stress the importance of Ethiopia's advance in developing a judicial management information system (MIS) and using it to plan and monitor improve- ments to judicial performance. This is a nearly unique accomplishment in Africa. (As has been repeatedly stressed, we are aware that South Africa has good court statistics and suspect these are used to similar ends, but have seen no sign of similar efforts elsewhere.) It also is very un- usual, if not unique in the developing world where automation, including that financed by the World Bank, seems to focus on anything and everything except generating accurate informa- tion on caseflow. Ethiopia's level of court automation is far less than that of many other coun- tries, and even some in Africa, but the point is that it put its automation where it would have the most impact--in generating information on what is happening in courts and in using it to 56 plan reform programs. A second point, as regards the federal courts surveyed here, is that the demand for their services (as indicated by new cases filed) has increased dramatically even over the four years covered (mid 2005 to mid 2009). Preliminary data from Ghana suggest similar trends. Ethiopian caseloads have grown correspondingly despite the addition of some judges, and judges have had to increase their output to keep up. Generally they have been successful in doing this in response to the FSC's oversight. Clearance rates in Ethiopia's federal courts are in the 90 to 115 range and congestion rates have dropped from 2.0 to 1.50 or lower. Time taken to decide cases has also gone down, although the high rate of appeals (and cassation) is another source of concern. This is especially true because the rate of reversal on appeal is very low (less than 10 percent) meaning that court users either do not recognize this, or ignore it because their only interest is creating more delay. An emerging problem is that of contested judgment enforce- ment as seen in the high percentage of execution/enforcement cases in all instances. However, it again bears mentioning that very few courts anywhere maintain statistics to support these kinds of conclusions, let alone use them to improve performance. They may suspect they have similar problems, but suspicions can also be misfounded or completely in error and they are a poor basis on which to build a reform program. 56 Although most federal judges, if not all their staff, now have computers to do their other work, CHAPTER 9 coverage is much more sparse at the regional and municipal levels. What this means is that the first computer is used to record case movement data, usually entered by a single staff member. This has several advantages as opposed to the frequent plan to have data entered as part of all staff's normal work ­ it guarantees more consistency in data entry, and, important for countries like Ethiopia, when the lights go out, court work does not stop. As for the data entry person s/he simply does the entry when the electricity returns. 92 Conclusions Thus, a third lesson is that court use in Africa can be expanded to a greater variety of citizens even if more slowly than in other regions. However, here it is not only important to make courts more physically accessible but also to improve the ease of use and quality of the results. Low fees and the ability to litigate without counsel have already been mentioned as means to facili- tate access, but Ethiopia's courts have also adopted a number of low tech practices to increase quality--these including the introduction of information desks and common intake systems, the use of "color-coded" files to attack a universal problem of intentional file loss or misplacement, and the more active role occasionally played by judges to assist court users who may not under- stand the rules. Higher tech innovations like cell phone consultations and automated informa- tion kiosks are also important although they may mainly benefit more sophisticated court users. Over time, the goal is to give those with disputes a choice as to the forum they will use, and even if the number enjoying the opportunity to choose is still limited, it is gradually expanding. The positive lessons from Ethiopia can thus be summarized as follows. In Africa, formal justice institutions will remain constrained in their reach for some time to come, but their use and user satisfaction can be augmented with the right programs. Information is the key, both as regards its generation and its use to improve performance. Automation is critical here, but it must be used intelligently. A basic MIS that grows over time is more important than a ready-made application that promises too much and in the end does little. But even more important is the involvement of judicial leadership in using it to identify problems and resolve them. CHAPTER 9 93 REFERENCES References Abera Miriam. 2002. "APAP's Experience on Social Court Judges Training," unpublished report. Aberra Jembere. 2002. "Ethiopia." In Herbert M. Kritzer (ed.) Legal Systems of the World: A Political, Social, and Cultural Encyclopedia, Volume II; 510­515. Santa Barbara, Cal: ABC-CLIO. Alula Pankhurst and Getachew Assefa (eds.). 2008. Ethiopia at a Justice Crossroads: The Challenge of Customary Dispute Resolution. Document in draft, Centre Français des Etudes Ethiopiennes. Addis Ababa. Andargatchew Tesfaye. 2004. The Crime Problem and Its Correction, Volume II. Addis Ababa: Ad- dis Ababa University Press. Blank, Jos, Martin van der Ende, Bart van Hulst, and Rob Jagtenberg. 2004. "Bench Marking in an International Perspective: An International Comparison of the Mechanisms and Performance of the Judiciary System." Netherlands Council for the Judiciary, Division Labor and Social Policy, Rotterdam, May 3. Canada, National Justice Institute. 2008. "Judicial Independence, Transparency, and Account- ability: Assessment Report," Final Report prepared for the Canadian International Development Agency (CIDA) and the Federal Supreme Court of the Democratic Republic of Ethiopia, December. Canada, Office of the Commission for Federal Judicial Affairs. 2005. "Court Administration Re- form Project in Ethiopia, Final Report." Castro F. Mauricio, Mauricio Olivera G. and Daniel Monsalve O. 2008, "Resultados de la Encuesta de Necesidades Insatisfechas de Resolución de Conflictos (NIRC)," Final report produced by FEDESARROLLO for the World Bank, June 26. Center for International Legal Cooperation. 2005. "Comprehensive Justice System Reform Pro- gram: Final Report," Draft prepared for the Justice System Reform Program Office of the Federal Republic of Ethiopia. CEPEJ (European Commission for the Efficiency of Justice). 2008. European Judicial Systems 2006. Strasbourg: Council of Europe. 95 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts Crook, Richard with Simplice Affout, Daniel Hammond, Adja F. Vanga and Mark Owusu-Yeboah. 2007. The Law, Legal Institutions and the Protection of Land Rights in Ghana and Cote d'Ivoire: Developing a More Effective and Equitable System. Brighton: University of Sussex, Institute of Development Studies, Research Report 58. Corporación Excelencia en la Justicia (CEJ). 2008. "Línea de base de la congestión judicial en la Jurisdicción Ordinaria." Third Report, prepared for the Colombian Consejo Superior de la Judica- tura, April. Galanter, Marc. 1974. "Why the `Haves' Come Out Ahead: Speculations on the Limits of Legal Change." Law and Society Review 9: 95­160. Genn Hazel. 1999. Paths to Justice: What People Do and Think about Going to Law. Oxford: Hard Publishing Company. Gonzales Mantilla, Gorki, Jean Carlo Serván, Luciano López, and Hernando Burgos 2002. El sistema judicial en el Perú: Un enfoque analítico a partir de sus usos y de sus usuarios. Lima, Peru: draft report prepared for the World Bank. (On file in Bank) Gupta, Poonam, Rachel Kleinfeld, and Gonzalo Salinas. 2002. "Legal and Judicial Reform in Europe and Central Asia." Washington, D.C: The World Bank Operations Evaluation Department, Working Paper Series. Haile B. Zewdineh and Wendwesson Shewarega. 2007. "Evaluation Report on the National Court Reform Program," draft prepared for the Federal Supreme Court of the Federal Republic of Ethiopia, August 10. Hammergren, Linn. 2002. "Reforming Courts: The Role of Empirical Research," PREM Note 65. Washington, D.C.: The World Bank. ______. 2009. "A Preliminary Overview of Corruption in Justice Sector Institutions," draft paper prepared for World Bank Governance and Anti-Corruption Review in Ethiopia. Henderson, Keith, Angana Shah, Sandra Elena, and Violaine Autheman. 2004. Regional Best Practices: Enforcement of Judgments: Lessons from Latin America. Washington D.C. IFES Rule of Law Paper Series, April. REFERNCES Jobim, Nelson. 2003. "O processo do reforma sob a ótica do Judiciário." In Reforma do Judiciário: Problemas, Desafios, Perspectivas, ed., Armando Castelar Pinheiro, 13­40. São Paulo: IDESP 96 References Love Kourlis, Rebecca, Jordan M. Singer, and Paul C. Saunders. 2009. "New report established principles for improving the U.S. civil justice system," Judicature, 93:3, November-December; 121­123. Magaloni, Ana Laura and Layda Negrete. N.d. "El Poder Judicial Federal y su política de decidir sin resolver." Unpublished draft, Mexico, Centro de Investigación y Docencia Económicas. (On file with the author.) Mamo Mihretu. 2008. "Constitutional Adjudication in Ethiopia: An Ill-conceived Response to a "Countermajoritarian Difficulty?" December, 19; draft paper on file with author. Menberetsehai Tadesse. 2009. Judicial Reform in Ethiopia, dissertation to be presented to the University of Birmingham, United Kingdom, in partial fulfillment of a PhD. Meron Zeleke. 2009. "Ye Shakoch Madad (the court of the sheikhs): The Shrine of Tiru Sina and its Role in Conflict Resolution," Paper to be presented at the 17th International Conference of Ethiopian Studies. Mojando, Luis Felipe. 2009. "The World Bank Doing Business Ranking of Quality of Justice: Criti- cal Analysis," Juridica Internacional, XVII; 203­215. North, Douglass. 1990. Institutions, Institutional Change and Economic Performance. Cambridge: Cambridge University Press Peru, Defensoría del Pueblo. 1999. Debido Proceso y Administración Estatal. Lima.. Poder Judiciário, Tribunal da Justiça do Rio de Janeiro. 2004. "Perfil das Maiores Demandas Judi- ciais do TJERJ." Rio de Janeiro, August. Republic of Kenya, Governance, Justice, Law and Order Sector (GJLOS) Reform Programme. 2006. National Integrated Household Baseline Survey, September. Tshuma, Lawrence. 2000. "The Political Economy of the World Bank's Legal Framework." In Governance, Development and Globalization, Julio Faundez, Mary E. Footer, and Joseph J. Norton (eds,), London: Blackstone Press: 7­27. REFERNCES Turton, David (ed). 2006. Ethnic Federalism: The Ethiopian Experience in Comparative Perspec- tive. Addis Ababa: Addis Ababa University Press. United Nations Development Program. 2009. Human Development Report 2009. http://hdrstats. undp.org/en/countries/country_fact_sheets/cty_ETH.html 97 Uses and Users of Justice in Africa: The Case of Ethiopia's Federal Courts United States Agency for International Development (USAID). 2009. The Role of the Labor Sector in the Foreign Assistance Framework: Honduras Labor Assessment. Washington D.C: USAID United States, Central Intelligence Agency, The World Factbook, "Ethiopia" at https://www.cia. gov/library/publications/the-world-factbook/geos/et.html Uzelac, A. 2008. "Reforming Mediterranean Civil-Procedure: Is There a Need for Shock Therapy?" in Civil Justice between Efficiency and Quality: From Ius Commune to the CEPEJ, C. H. van Rhee y A. Uzelac (eds.). Intersentia, Antwerp; 71­99 World Bank, 2002. The Juicio Ejecutivo Mercantil in the Federal District Courts of Mexico: A Study of the Uses and Users of Justice and Their Implications for Judicial Reform. Washington, D.C, Report No. 22635-ME ______. 2003. An Analysis of Court Users and Uses in Two Latin American Countries. Washington D.C: The World Bank, Report No. 26966 ______. 2004. Making Justice Count: Measuring and Improving Judicial Performance in Brazil. Washington, D.C: The World Bank, Report No. 32789-BR ______. 2005. Doing Business in 2005: Removing Obstacles to Growth. Washington D.C: The World Bank, the International Finance Corporation, and Oxford University Press. ______, Legal Vice Presidency. 2004. Ethiopia: Legal and Judicial Sector Assessment. Washington D.C: The World Bank REFERNCES 98