Empirical Analysis of Firing Disputes

Empirical Analysis of Firing Disputes Are Common Perceptions on Termination Cases in Mexico Supported by Empirical Data Abstract Using a random sample of cases from the Federal Labor Court in Mexico, this research analyzes firing disputes in Mexico from an empirical perspective. In particular, it focuses on presenting evidence in regard to the following three common perceptions on termination cases: (i) that they are often terminated by an adjudicated decision; (ii) that they are extremely delayed; and (iii) that the final payments obtained by the workers are extremely high. The statistical evidence suggests that an emphasis in the literature and media on a particular subsample of cases with extreme characteristics (e.g. adjudicated or appealed processes) produces these conceptions. The study concludes with a brief comment regarding the importance of empirical studies as a way of better understanding legal phenomena in Mexico. Table of contents I. Federal labor tribunals in Mexico II. Methodology and Data III. Results and Discussion A. Termination payments are extremely high B. Termination cases are extremely delayed C. Termination cases normally go to trial IV. Conclusion Although the analysis of the law in Mexico has changed slightly in the last years, Pounds vision effectively reflects the atmosphere of pure law of the Mexican mainstream. The analysis of labor law in Mexico—the general topic of this research—reflects this analytical path. In fact, the overwhelming majority of the literature related to labor justice has been doctrinal and normative. These studies often ignore factual evidence of how the legal framework affects the legal reality; and if some evidence is presented, it is merely indicative or anecdotal. This doctrinal research figures prominently in public policy debates. This study departs completely from the Mexican mainstream insofar as it is an empirical study based on case file analysis. The research results are based on a random sample of 3,203 files selected from the Junta Federal de Conciliacià ³n y Arbitraje (JFCA) archive. These files correspond to termination cases presented before the Tribunal 15 of the JFCA (Tribunal 15), a labor tribunal (Junta Especial) whose jurisdiction includes the pharmaceutical, chemical, automotive, and paper industries; and the Tribunal 6 and 8 of the JFCA (Tribunal 6 8), the labor tribunals whose jurisdiction includes the textile industry. The period studied extends from 1991 to 1998. Note that case file analysis has been used extensively in other countries with fruitful results. The objective of this research is to characterize empirically the termination disputes in Mexico in the context of the tribunals analyzed. In particular, it focuses on presenting evidence in regard to the following three common perceptions on termination cases: (i) that they are often terminated by an adjudicated decision; (ii) that they are extremely delayed; and (iii) that the final payments obtained by the workers are extremely high. This article is divided into three sections. The first section explains the operation of the federal labor tribunal in Mexico. The second section describes the data and the methodology. The third section examines the above mentioned perceptions on termination cases. I. Federal labor tribunals in Mexico The JFCA is an administrative court that belongs to the executive branch. Because of this dependency, Mexican legal authors have questioned its degree of judicial independence. The JFCA consists of 61 labor tribunals, called Juntas Especiales. Sixteen tribunals are located in Mexico City, and the remainders are dispersed throughout Mexico. Labor tribunals have the legal power to conciliate and adjudicate. In fact, the LFT mandates at least one conciliation hearing before trial. Among the labor tribunals, jurisdiction is determined by industry. Although the labor law openly promotes settlement of disputes, it does not permit the settlement terms to remain confidential. The relevant tribunal must ratify the settlement for it to become binding. An employee cannot credibly forfeit the right to sue his employer unless and until the court approves his settlement. Employers and workers very often submit a settlement jointly to the labor court simply to obtain ratification, which then makes the agreement binding. The approval of settlements primarily serves as a mechanism for confirming that the worker has not renounced some legal benefit and for making the agreement binding at law. Tribunals must record details about the settlement, such as the date of the settlement and the amount paid. Once a lawsuit is filed, the tribunal with jurisdiction over the dispute must schedule at least one conciliatory hearing. At the conclusion of that hearing, the tribunal schedules subsequent hearings for the presentation of evidence and for trial. At any point during the process, however, the suit can be terminated by a settlement. The tribunal must approve this settlement. In fact, the parties can agree to defer hearings if they think they need more time to negotiate the terms of the agreement. If no agreement is reached, a tripartite commission composed of an employer representative, an employee representative, and a government representative may issue an award deciding the dispute. Note that the employee may also drop the complaint at any point during the lawsuit. As noted in the tables below, dropped lawsuits are not uncommon. An appeal mechanism, the juicio de amparo, is available to challenge the resolutions of the tribunal. Note that the parties cannot only challenge the final decision issued by the tribunal, but also other decisions that could potentially affect the outcome of the case, for instance, the fact that the tribunal notified the employer about the workers complaint in an incorrect manner. It is fair to state that the scope of review of the juicio de amparo in labor cases is limited to protect the due process rights of the parties. II. Methodology and Data Clermont Eisenberg divide legal studies based on statistical research into three groups depending on the way in which the data is assembled. Studies of published judicial decisions, studies based on data produced by experimental work or by archival research, and studies involving analyses of publicly available, and usually governmental, databases. The present study fits into the second category, legal research based on archival data. The logic behind analyzing archival cases is simple: random samples of large dockets can provide us with useful information to analyze the patterns of litigation in a particular tribunal. This is particularly relevant when there is no information available—or only very incomplete information—on how certain type of legal dispute is resolved. In particular, information regarding labor disputes is extremely general and imprecise. The data used for this research was obtained through a public information request under the Ley Federal de Transparencia y Acceso a la Informacion Publica Gubernamental, the law that regulates the access to public governmental information. While some of the variables used in this study are considered public information under this law, other variables are not public information, and have been obtained under a confidentiality agreement. The confidentiality agreement was negotiated with the authorities of the JFCA, who only authorized to review those cases that complied with the following characteristics: (1) closed cases that cannot be reopened by the parties anymore, (2) and cases that had been initiated in 1998 at most. Given these constraints, I was able to assemble a random sample of cases initiated between 1991 and 1998 in Tribunal 15 and Tribunal 6 8. Note that these years refer to the dates in which the cases were presented before the labor tribunals, and not to the dates in wh ich the cases were actually solved. There are two main types of cases observed in the sample: filed settlements and litigated cases. In filed settlements, the employer and employee file a jointly agreement, which is merely ratified by the tribunal. In litigated cases, the parties usually present and produce several documents, such as the employees initial petition, the employers response (if any), the terms of agreement reached if the case is settled, the final award issued by the tribunal if the case is not settled, and the appeals carried out by the parties. Note that litigated cases have three possible outcomes: dropped suits, settlements, and trials leading to a final decision issued by a tripartite commission. As mentioned before, the final decision is known as an award (laudo). Using the information from the documents mentioned in this paragraph, I calculated the duration and the final payment of the cases handled by Tribunal 15 and Tribunal 6 8. These calculations will be presented in further tables. A comment should be made on the validity and usefulness of the results coming from the data analyzed. First, the LFT has not been substantially modified since its enactment in 1970. Second, and based on five interviews with representatives of the government from five federal labor courts that handle similar disputes, it is reasonable to state that the solving pattern of the termination disputes here portrayed is accurate and reflects the current situation. III. Results and Discussion A. Termination payments are extremely high Some Mexican authors have analyzed the LFT from an economic perspective. Basically, their argument is that the LFT should be modified to eliminate all those provisions that create restrictions on hiring, mobility, and termination of employees. These legal restrictions, originally designed to protect the worker, generate a higher social cost than benefit. In particular, these authors claim that the LFT raises termination costs, diminishing the quantity of labor that an employer hires below the socially efficient level. This argument assumes that most of the termination payments are paid according to the LFT and are thus extremely high. No empirical evidence concerning this assumption has been produced. Before presenting the empirical evidence, the rules regarding termination payments should be explained. These rules, provided in the LFT, mainly refer to the provision of fringe benefits, overtime, and the mechanics of termination. Fringe benefits are mainly composed of vacation pay and year end bonuses. Each employee is entitled to a certain number of days of paid vacation depending on his or her tenure at the firm. Also, every employee has the right to receive an end of the year bonus equal to at least 15 days wages. A normal workweek cannot exceed 48 hours. If an employee works more than 48 hours in a single week, he is entitled to overtime pay. The law mandates double pay for up to 9 hours of overtime, and triple pay for any hours above 57 per week. Terminations are classified under the law as justified or unjustified. Justified termination is limited to wrongdoing on the part of the worker, such as three continuous unjustified absences. Termination for any other reason, such as low employee productivity, is considered unjustified and it often implies higher termination cost. For either type of termination, the firm must cover all payments owed to the worker up to the termination date, including overtime and the prorated proportion of fringe benefits until the termination date. Additionally, the worker is entitled to severance pay equivalent to 12 days wages for each year worked, with the wage rate capped at twice the minimum wage. At the time of termination the firm must issue the worker a written document explaining the exact cause of termination as defined by the LFT. Workers have the right to challenge the grounds for termination. According to the LFT, the firm carries the burden of proving that it fired the worker for just cause. Workers terminated in an unjust manner have the right to receive two additional payments. Back pay from the termination date through the date when judgment is issued (including not only the salary, but also the fringe benefits), plus three months salary with benefits. Also, those workers who are denied reinstatement are enti tled to 20 days wages plus benefits for each year worked, without any cap on the wage rate. Note that a worker who proves he was fired without justification can request to be reinstated in his job. The firm may refuse reinstatement for certain categories of workers, such as temporary workers, those with less than one years tenure, and those considered to be at will employees. A firm may also avoid having to reinstate workers it fires in the case of justified layoffs. Table 1 presents the average final payments obtained by workers depending on the form in which the cases concluded. Both absolute and relative amounts are shown. Relative amounts correspond to the amount that workers win as a percentage of the amount they claimed. The data demonstrate that fired workers who initiated a lawsuit usually claimed considerable amounts of money that seldom obtained. Also, note that workers who resolved the case through a filed settlement or a settlement reached during the lawsuit had better final payments. Interestingly, in cases solved by an award, the decision of the parties to appeal increased the amount received by the worker. As the tables portray, most of the termination cases do not receive an extremely high final payment, which contradicts the widely held assumption of the Mexican media and Mexican academic literature. Although this section only presents a descriptive analysis of the data, these results deserve much more attention and analysis in future research. B. Termination cases are extremely delayed This section presents empirical evidence that supports the idea that the duration of termination cases is not necessarily extremely delayed, but it actually depends on characteristics of the dispute like the termination mode, the decision of the parties to challenge the tribunal decisions, or to defer the case. Many Mexican legal scholars have criticized the fact that labor procedures are extremely formal and complicated. According to this view, procedure is the main source of delay in labor disputes. Proposals for diminishing delays in labor disputes include: improving the conciliation system , hiring more legal clerks , replacing the old technology in the JFCA (i.e. principally computers) , and, of course, changing the law to eliminate procedural complexities. The LFT procedures do not limit the time period for case resolution; rather they establish a term in which each action in the process should be completed. The goal of establishing such terms is to create a fast labor procedure (article 685). The labor statute requires different types of procedures depending on the case at issue. The procedure by which termination cases are solved is known as ordinary procedure (article 871). The ordinary procedure starts when the employee files a claim before the JFCA (article 871). After this claim is admitted, the parties are scheduled for a hearing (article 873). The hearing has three stages: the conciliation stage; the claims and exceptions state; and finally, the stage of submission and admission of the evidence (article 875). Each one of these stages has particular time and form characteristics according to which it operates. Article 876 governs the conciliation stage. The goal is for the parties to reach a conciliation agreement through the intermediation of the tribunal. If the parties reach an agreement, the disputed is terminated and the tribunal endorses the corresponding agreement. This agreement, like an award, is enforceable. If the parties do not reach a conciliation agreement, the hearing proceeds to the claims and exceptions stage. Article 878 also governs this stage. The worker presents his claim, and the employer answer him. After this stage, the submission and admission of the evidence stage occurs, which article 880 governs. Next, the evidence is presented before the tribunal in another hearing governed by Article 884. When all these stages have been completed, a legal clerk creates a draft of the final ruling. This draft has to be approved by members of the tripartite panel (the representative of the employees, the employers, and the government). This process follows the rules established in articles 887, 888, and 889. After approval, and if no corrections are proposed, the draft is turned into the award. According to the terms established in the statute for each one of these proceedings, an ordinary procedure should be resolved in 75 effective court days. As table 2 portrays, the average duration of an ordinary procedure (i.e. case solved by an award without being appealed or being deferred) in Tribunal 15 is 440.76 effective court days, and in Tribunal 6 8 is 419.89 effective court days. Several things should be clarified with regard to the description of the ordinary procedure. First, although the labor statute speaks of one hearing divided into different stages; in fact, there are different hearings in which all these stages are completed. Therefore, if the parties cannot reach an agreement after a conciliation hearing, the conciliation continues in the next hearing, and the subsequent stage is postponed and so on. Second, the statute permits the parties to defer the hearings if both sides agree. In other words, if the parties are in the process of negotiating a conciliation agreement, they can ask the labor tribunal for more time. According to table 2, this is not unusual. Third, although the statute specifies that the conciliation stage should be performed at the beginning of the case, the parties can reach a conciliation agreement at any time during the ordinary procedure, even after the conciliation stage has passed. Fourth, the description portrayed for the or dinary procedure does not account for whether the parties (either the employee or the employer) decide to appeal the award or not. If the parties decide to appeal the award issued by the judge, the procedure should be extended by no more than 45 days until the appeal sentence is rendered. Table 2 portrays that the duration of the case increases considerably when the parties decided to appeal the rulings of the labor tribunal. While cases concluding through an award that was not appealed took 450.45 effective court days to be solved in Tribunal 15 and 442.69 effective court days in Tribunal 6 8, cases concluding through an award that was appeal took longer: 904.79 effective court days to be solved in Tribunal 15 and 790.28 effective court days in Tribunal 6 8. The ordinary procedure requirements suggest the excessive formality of the process for resolving termination cases. The real question is how many cases actually pass through all these complicated stages. Table 2 shows the average duration of cases in court effective days depending on certain decisions of the parties, particularly, their decision to appeal the tribunals rulings and to defer the case. Note that 56.97% of the total cases from Tribunal 15 and 23.48% of the total cases from Tribunal 6 8 were filed settlements. These cases were resolved without passing through any of the above mentioned stages of an ordinary procedure and went to the court as a mere formality. The situation is different for litigated cases. Litigation cases are divided between those cases in which the parties deferred the case at least once and those in which the parties never deferred the case. As noted in the tables, the average duration is regularly higher when the parties decided to defer the case tha n when they chose not to defer it. Observe that the difference in duration of a lawsuit also depends on the parties decision to settle, drop, or go to trial seeking an adjudicated award. Finally, the duration of the case depends on whether the parties chose to appeal the rulings of the tribunal or not (including the award). Therefore, it is clear that when an award is issued to resolve a labor dispute or when the parties challenged the tribunals decision, the case was more delayed. However, the real question is how frequently this situation actually occurred. According to the evidence presented here, termination cases were rarely solved through an award (6.09% in Tribunal 15 and 9.76% in Tribunal 6 8), and even more rarely through an award issued after an appeal (1.88% in Tribunal 15 and 2.55% in Tribunal 6 8). Adjudicated awards and appeals appear to be more the exception than the rule. Short settled cases (i.e. filed settlements and settled lawsuits without appeal) instead of long disputes seem to be the typical termination case. Finally, and given the results presented in table 2, it is reasonable to believe that the substantial length of the adjudication process could be itself a contributing factor to the settlement of most cases. Although a profound analysis of such idea goes beyond the scope of this work, it should definitely be considered in further empirical studies. C. Termination cases normally go to trial The media and legal scholars in Mexico have portrayed Mexican labor justice as complicated, corrupted, delayed, and costly. This vision has deeply shaped the perception of the labor justice system. Cases reflecting the view of delayed justice and low final payments appear in the random sample of cases analyzed in this study. Although these cases exist in the data set, they do not at all reflect the typical case. First, case 1/1991 , which began in 1991. In this case, a 26 year old man filed a claim of unjustified termination. He worked in the state of Oaxaca, in a paper factory. He had worked for almost five years receiving a daily wage of $78 pesos and claimed $191,900 pesos. A public defender represented him. In 1995, after four years, an award issued by the judge determined that the termination had been lawful. Thus, the claim was dismissed without any payment to the worker. Second, case 2/1994, which began in 1994. In this lawsuit, a 32 year old man also claimed an unjustified termination. He worked in the state of Chihuahua for a pharmaceutical firm. He had worked for six years and received a daily wage of $307 pesos. A private lawyer represented him, and the amount at stake was $922,429 pesos. The lawyer appealed the first award, which had favored the employer. The final decision, rendered near the end of 1997, favored the worker, but only granted him $7,894 pesos. The case took slightly more than three years of effective court time; and interestingly, the parties decided to defer the case once during the process. At the end of the day, the worker waited three years and received only a small fraction of his original claim (less than 10%). Although the cases portrayed above match the descriptions provided in the news for termination cases that reach the trial stage, it is not clear that these cases are representative of the typical termination case with which the tribunal commonly deals. Other cases in the sample reflect a completely different result. In other words, high payments and little court time. For example, case 3/1994 began in 1994 when a 50 year old man from Mexico City claimed an unjustified dismissal after more than thirteen years of work at a well known pharmaceutical company. The case did not reach judicial evaluation of the merits, and was settled only three months after filing. Interestingly, the parties decided to defer the hearings twice, and a private lawyer represented the employee. The employee, who claimed $267,115 pesos, received $214,959, 80% of his original claim. In another termination case, 4/1997, a 57 year old man filed a lawsuit against a pharmaceutical company asking for $1,364,927 pesos. The employee had a 22 year tenure and a daily wage of $2,472 pesos daily salary. The case was settled after ten months, and the employee obtained $1,001,167 pesos, 73% of his original petition. Again, a private lawyer took the case, and the parties deferred the hearings twice. Note two relevant commonalities in the two cases presented: both settled before reaching the trial stage, and the parties decided to defer the hearings at some point in the process. This study finds that settled cases more accurately reflect the resolution of Mexican labor disputes than the cases highlighted in the media, which typically cover cases that go to trial and terminate with an adjudicated award. In other words, the media emphasize cases that usually take a long time to resolve; and in particular, cases in which workers obtain outlying results (very low or high payments). Although these cases eventually come up in labor tribunals, they are rare. A bias for highlighting cases that go to trial has been identified several times in the literature. Laurence Ross and Alfred Conrad showed, in 1970 and 1964 respectively, that only a very small fraction of disputes go to trial and an even smaller fraction are appealed. Consequently, an analysis based only on those cases that go to trial is not representative of the influence of legal rules on social affairs. The conclusions of doctrinal studies based on trials or appealed cases disclose very little about how legal rules affect the behavior of those subject to them or affect the generation of legal disputes themselves. Mexican literature concerning labor law in general, and particularly termination cases, has followed this pattern. The methodological problem in studying only final decisions, like appeals or awards, is that the: simple extrapolation from a sample of litigated cases to the population of all cases is valid only if the sample is random. If cases are not randomly selected for litigation, then a researcher who uses litigated cases is necessarily studying both the selection mechanism and the underlying population of cases simultaneously. The particular set of data gathered for this research avoids this problem to a great extent. First, as explained earlier, the labor statute requires parties to ratify terminations before the labor tribunal for the termination to become legally binding. If an employer fires an employee and wants to have a legal document verifying that the employee was fired without any violation of the provisions of the law, he needs to ratify his dismissal agreement before the labor tribunal. Therefore, termination cases (the underlying dispute) should pass by the court, at least for ratification. Second, cases sampled not only include filed settlements, but also cases that starte d as lawsuits and were later settled, dropped, or decided by an adjudicated award. In short, the random sample of cases studied accurately represents the solving pattern of termination cases because the cases analyzed are a random sample of the underlying disputes, not a random sample from a particular biased sub sample. Table 1 and 2 portray the solving pattern of termination cases in Mexico. Statistics from these tables suggest that the typical termination case is settled in some fashion, and that awards are issued occasionally. A brief survey of the Mexican literature shows that one of the main concerns for labor reformers in Mexico is the long delay in solving disputes. Table 1 and 2 show that this perspective is imprecise. Most of the disputes went to court merely seeking a ratification of the parties agreement and were solved the same day they were presented. On the other hand, if the dispute was not a mere ratification but litigation, duration varied greatly according to the way of solving the dispute and to the parties decision to appeal the tribunals decision (particularly the award). Therefore, it appears that when studies and news refer to delay in labor justice, they actually refer to a sub sample of cases with certain characteristics, in particular, to those cases that finish with an adj udicated award, are appealed, or both. Many analyses of Mexican labor justice in general, and of termination cases in particular, have not seen the whole picture, but only a small fraction of cases with extreme characteristics. These works, though somewhat useful, tend to be biased in their conclusions, and they definitely do not appear to be the best guide for implementing public policy measures or legislative changes. Interestingly, according to the tables, a final judgment (i.e. award) was issued only in 6.09% of the total cases sampled in Tribunal 15 and in 9.76% of the total cases sampled in Tribunal 6 8. Alexander reports that fewer than 5% of litigated cases are tried to judgment in the U.S. Trubek et al. state that approximately 8% of civil suits filed in state or federal court went to trial. Also, Resnik notes that 85% to 90% of all federal civil suits end up settling. Note that 86.14% of the sampled cases were settled in some manner in Tribunal 15 and 77.80% in Tribunal 6 8. Therefore, the results of the present study are consistent in supporting the conclusion that cases that go to trial provide only a distorted window into peoples behavior in response to legal rules. IV. Conclusions This article seeks to analyze termination using empirical evidence. Using a random sample of cases from three Mexican labor tribunals, this work has analyzed three widely held perceptions on termination cases, which are commonly supported by the Mexican media and some Mexican legal scholars. The conclusion is straightforward: the perception of termination cases in private industry has been biased by a sub sample of cases with extreme characteristics, cases that go to a final judgment or are appealed. Interestingly, this biased has not been detected by research on labor law produced by Mexican scholars who study the law from a doctrinal perspective, proving that limiting the study of the legal system to doctrinal analyses of legal norms could lead to erroneous conclusions. Approaching research questions from an empirical perspective is necessary, and it will eventually reveal new features of the Mexican legal system.