British Journal of Industrial Relations
40:4 December 2002 0007-1080 pp. 753-777
Abstract
The study examines all cases ( 1990-7) in Israel in which employers petitioned the labour court to issue an injunction against striking workers, and identifies how judges use their discretion in deciding the petition. The findings indicate that, judicial rhetoric to the contrary, the labour court limits its considerations almost solely to the parties' formal compliance with the legal rule. The implications of these findings for the relationship of the legal and industrial relations systems are discussed, particularly with reference to the dual task of labour law: to govern the industrial relations system, and to facilitate its autonomy.
1. Introduction
Strikes are complex phenomena with a clearly dialectic quality. They are adversarial and militant, and financially harm employers, the economy, the public, and even the employees. Yet strikes are an integral part of the freedom of association, and therefore are an important component of a democratic regime that respects human rights. They can also provide economic gains in the long run, given their role in voicing the workers' interests. Consequently, all democratic regimes respect the right to strike, but within prescribed limits (Ben Israel 1988; Blanpain and Ben-Israel 1994). These limits are established by law and may be found in legislation, secondary regulations, executive orders and judicial precedents. Strike activity must therefore be studied through observation of the impact of the legal background on the strategic choices made by the relevant parties, and their interaction.
The relationship between the legal rule and the study of strike activity is frequently based on observing legislation and leading judicial precedents. These are the visible frontiers of law. For example, the US Supreme Court decision to sanction the permanent replacement of striking workers has been
Guy Mundlak is at Tel Aviv University. Itzhak Harpaz is at the University of Haifa.
© Blackwell Publishers Ltd/London School of Economics 2002. Published by Blackwell Publishers Ltd, 108 Cowley Road, Oxford OX4 IJF, UK and 350 Main Street, Malden, MA 02148, USA.
754 British Journal of Industrial Relations
noted for its adverse effect on workers' capacity to strike (Weiler 1990). Similarly, changes in the British legislation during the 1980s were shown to have an adverse affect on strike activity (Wedderburn 1995). These analyses take a considerable step towards explaining the relationship between the legal background and the strategic choices made by employers and workers. Even so, the focus on the formal statement of statutory law and highvisibility precedents is not sufficient, and may even be misleading.
The need to study 'law in action' dates back to the development of legal realism in the United States and in Europe (Pound 1910; Kairys 1990). The underlying assumption is that a gap lies between the formal statements of the law and what the legal agents actually do. This approach requires scrutiny of the legal process rather than just of the legal rule. This shift in the focus of study may call for examination of the obstacles placed in the way of access to justice, the process of dispute resolution, and the outcomes of the process itself.
The objective of this study is to move away from the formal law of strikes, and to focus instead on one of the more important components of its implementation-in-fact, namely the use of judicial discretion in cases where employers seek an injunction against striking workers. The law of strikes in Israel is deliberately general and highly discretionary. Hence the study is based on the assumption that the law of strikes is what judges make it to be. It examines the construction and implementation of the law by the Israeli labour court between 1990 and 1997. More particularly, the study tests two competing views on the role of labour law, and of labour courts as a derivative, in the resolution of collective labour disputes. One view limits the law to a set of procedural rules, which it is the court's task merely to administer. The other view holds labour law as a means to actively facilitating collective dispute resolution, granting judges broad, flexible, and discretionary standards, which can be applied in the way that best fits the needs of the parties to the dispute. Currently both views appear in the formal statement of law, but it is the day-to-day adjudication, rather than the high-visibility precedents, that shapes the court's agenda and its task. Identifying what courts do in fact can contribute to the understanding of the role law and its agents play in the industrial relations system (Dunlop 1993).
The Law of Strikes in Israel
The law of strikes is an important component of the autonomous collective labour law in Israel. Like the law of contracts, which governs the individual employment relationship, autonomous labour law seeks to establish procedural rules that govern relationships on the basis of agreement. Both bodies of law are different from mandatory regulation, because the law seeks only to define the 'rules of the game' rather than actual standards (the 'game's outcome'). While statutory standards implicate the state as a regulatory agent, collective labour law and the law of contracts seek to
© Blackwell Publishers Ltd/London School of Economics 2002.
Israeli Judicial Discretion 755
facilitate systems that are separate from the legal system (Rogowski and Wilthagen 1994). The law of contracts is necessary to ensure the operations of the market, while autonomous labour law is necessary to facilitate the industrial relations system. Autonomous labour law is therefore a mode of reflexive law (Teubner 1993). Its scope is defined by the need to ensure the autonomous functioning of the industrial relations system and its need for self-regulation.
The reflexive nature of autonomous labour law suggests two objectives for labour law: to govern the agents of the industrial relations system, and to facilitate their autonomy. This requires not just leaving the parties alone, but stabilizing the system and taking part in developing norms of cooperation and legitimacy for peaceful dispute resolution. The law of strikes therefore seeks to distance itself from the system (by focusing only on technical procedures) and to draw closer to the system (by identifying its needs and facilitating them).
Israeli law of strikes is complex. For the most part it does not appear in legislation. The recognition that the right to strike is a human right appears in case law, and the statutes supplement this recognition with a number of immunities dispersed among various pieces of legislation. Beyond the immunities, the few rules that appear in statutes deal with technical procedural requirements, such as the duty to provide advance notice before a strike, and the allocation of responsibility to fulfil this duty to the central organs of the trade union.
Unlike the statutory provisions, case law on strikes is extremely rich.
Generally, while the legislature's contribution to the law of strikes adheres to the formal procedural objective of the labour law, the court's development of broad legal standards is more reflective of labour law's facilitative objective (Kennedy 1976; Mautner 1988). The labour court provides a broad definition of a 'strike' (to include all partial strikes and slowdowns), but also limits the legitimacy of strikes not aimed at the employer (political strikes) or aimed to pressure the employer on issues unrelated to industrial relations or employment conditions. The court further limits legitimate strike activity to economic disputes alone (rather than legal disputes). It enforces relative industrial peace clauses in collective agreements when strikes are initiated at the time the agreement is in force and addresses issues clearly regulated by the collective agreement. The court further requires the striking workers to act in good faith and to show reasonable respect for the employer's property.
When one or more of the legal rules has been violated, the court holds the strike to be illegitimate. An illegitimate strike is not illegal in the criminal sense, nor does it deny the striking workers their immunities. On the basis of its general prerogative to provide remedies, the court views the injunction as the most appropriate means of legal intervention in strike activity. The injunction orders the workers back to work, and can be administered in a very short time (occasionally even a few hours after the employer's petition to the court). It does not seek to penalize the striking workers, but merely to
© Blackwell Publishers Ltd/London School of Economics 2002.
756 British Journal of Industrial Relations
eliminates the illegitimacy of the strike. It therefore accords the industrial relations system more autonomy in administering its own sanctions, rather than drawing on civil legal sanctions.
A few leading precedents have elaborated on the considerations for issuing an injunction. These reveal that the court is sensitive to the tension inherent in the function of law in strike activity. On the one hand, the court warned that 'it will not issue an injunction as a routine response to a breach of collective agreement. The court has broad discretion in issuing an injunction, whether temporary or permanent, and it must assess a broad range of considerations in its decisions' (NLC 4-28/89, The Ports Authority v The Histadrut). On the other hand, the court recognized that the legal rule must seek to protect employers, consumers and the general public from harmful strike activity, even if the strikers comply with the minimal procedural rules prescribed by law. Relying on a 'balance of convenience' test, the court noted that it has discretion to issue an injunction even when a strike is legitimate.
The complex body of law that has evolved reflects the dialectic nature of strikes, as well as the dual function of the legal system. Law therefore intervenes in and withdraws from the strike at the same time. Rather than setting a rigid balance between labour law's dual objectives, the law is highly discretionary and has the potential to accommodate almost any ruling. The labour court in this respect provides the most important venue for the study of law-in-action. The court writes substantial segments of the law, giving itself a high level of discretion, and then implements its own laws. It is this implementation that is instructive for understanding the balancing of the law's conflicting function with regard to the industrial relations system.
The Labour Court
Labour courts or tribunals are common in many countries and can be distinguished according to their level of integration in the general legal system (Rogowski 1994). At one end of the continuum are administrative tribunals such as the NLRB in the United States or the Conseils de Prud'hommes in France. Other states have established fully fledged judicial tribunals with powers matching those of the civil judicial system. The Israeli labour court system falls in the latter category (Bar-Niv 1974; Ben-Israel 1977). As in some other countries (Porcher 1994), in Israel the agents of the industrial relations system have played an important role in establishing and staffing the labour court (Zamir 1974). The labour court system was originally conceived in a collective bargaining agreement in which the social partners agreed to lobby the legislature to establish a labour court that would respond to their specialized needs in dispute resolution.
Two major functions justify the establishment of a separate court for the resolution of legal labour disputes. The first is expertise. It is assumed that adjudicating labour disputes requires specialized knowledge of labour law's polycentric body of norms, as well as knowledge of extra-legal disciplines
© Blackwell Publishers Ltd/London School of Economics 2002.
Israeli Judicial Discretion 757
that are required to understand the administration of law in organizations. Second, it is assumed that labour courts should provide a different kind of adjudication, which is more closely related to alternative dispute resolution. Such adjudication is necessary to sustain the autonomy of markets and, more notably, of the industrial relations system. The two justifications therefore reflect the legal system's dual objectives: to govern, and to facilitate the autonomy of the industrial relations system.
In its jurisprudence, the court repeatedly makes claims that reflect its dual role. First, case law emphasizes its adjudicative role, which requires the judges to determine disputes on the basis of strict implementation of the law. Here it conceives itself as a sub-system of the general court system. This view, designated hereinafter the formalistic approach, perceives the labour judge as law's gatekeeper; the court's role is to ensure compliance with the procedural requirements of the law, while disregarding the extra-legal nature of the dispute. Second, the court emphasizes its role as an industrial mediator, seeking to move the parties to a constructive settlement. This role brings the labour court closer to the industrial relations system. This view, designated hereinafter the strategic approach, holds that the labour judge ought to draw on the discretionary components of the law in order to identify solutions constructive to the resolution of the labour dispute in the long run, rather than strictly applying the procedural requirements. This function is emphasized in particular with regard to collective disputes, where dispute resolution seeks to sustain the constructive ongoing relationship between the parties, unlike individual disputes, which are usually brought to the court on the termination of the employment relationship. Note that both views are well grounded in the legal norms themselves, and are emblematic of law's dual function with regard to the industrial relations system.
This study examines the court's choice out of the two views. It does not purport to assess the impact of judicial decision-making on strike activity directly. Instead, it holds judicial decision-making as the dependent variable and hence the subject of inquiry. However, the concluding section remarks on the implications of the study of judicial discretion for the impact of law, as implemented in fact, on the industrial relations system, and on how this study promotes understanding of the relationship between the law and the partners to collective bargaining.
Previous Studies
No systematic data are available on how the Israeli labour court implements its discretion. Previous research has mainly been limited to a descriptive depiction of judicial outcomes (Galin and Harel 1978; Weiss 1989). Concerning the level of the court's responsiveness to employers' petitions, Galin and Shirom (1978) found it to be only 21 per cent in the early 1970s, while Weiss (1989) reported a higher rate (29 per cent) during the 1980s. This difference can be explained by Galin and Shirom's account that, during the first years after the establishment of the labour court (in 1969), judges
© Blackwell Publishers Ltd/London School of Economics 2002.
758 British Journal of Industrial Relations
sought to mediate cases rather than simply adjudicate them, in comparison with Weiss (1989), who found an increasing trend of formalistic adjudication, rather than informal mediation. Prima facie, this shift suggests the growing relevance of the formalistic approach at the expense of the strategic approach, and hence a preference for the governing objective of labour law over its facilitative objective. However, while these studies, as well as other extensive studies of strike activity in Israel (Chermesh 1983), observe judicial outcomes, they do not study the construction of judicial discretion as such. They therefore preserve the image of law as an unopened 'black box' (and hence an independent variable) rather than an output of a dynamic and strategic process (and hence a dependent variable).
Few attempts have been made to study empirically the role of injunctions in other countries, even though these provide only a partial understanding of how judges make their decisions (Gall and McKay 1996). Other studies that provide a rigorous empirical analysis of judicial discretion do not study injunctions as such (Cooke et al. 1995). Rigorous studies on the use of judicial discretion with regard to individual employment disputes (Dickens 1994) are less instructive on how the legal and industrial relations systems interrelate and on labour law's dual objectives, because individual disputes are often removed from the industrial relations system at the outset. Consequently, aside from the fact that most Israeli studies are already somewhat outdated, the present study aims to provide a piece that is still missing in the overall theoretical puzzle.
Hypotheses
Based on previous studies, judicial precedents and statutes that provide the formal content of the Israeli law on strikes, and the theoretical approach outlined here, it was possible to generate three hypotheses regarding the determinants of judicial discretion in cases involving petitions against striking workers. The hypotheses distinguish three categories of variables that are assumed to affect the judges' construction of legal discretion. The first of these concerns variables that correspond to the procedural rules determined by law; the second concerns extra-legal considerations, which the judges are permitted to consider, given the high level of discretion accorded to them by law; and the third concerns variables related to the legal process itself.
The hypotheses were tested in two ways. The first was to test them only on cases in which judges issued a decision (i.e. to issue an injunction as requested, or to deny the petition for injunction). The second was to test the hypotheses on all cases, a distinction being made between those in which the judge adjudicated and those in which the parties reached a settlement after the petition for an injunction was filed in court. As will be explained, in most of these cases judges either were in the shadow of the settlement or actively promoted it, so these settlements are as instructive as judges' formal decisions for the study of what judges do.
© Blackwell Publishers Ltd/London School of Economics 2002.
Israeli Judicial Discretion 759
Hypothesis 1: When the strike is not legitimate the court always issues an injunction.
This hypothesis draws on the governing function of law, aimed to ensure compliance with the legal rules of the game. It can be collapsed into a number of sub-hypotheses as follows. The court will issue an injunction whenever workers do not give prior notice as prescribed by law (15 days prior to the strike, and by the authorized organ of the union); whenever the strike is over a legal dispute, rather than an economic dispute; whenever there is an industrial peace clause in effect; whenever the strike is quasipolitical, limiting the strike to a short period of time; whenever the strike is not about wages or working conditions; and when the strike is unprotected. In contrast, the court will tend to refrain from issuing an injunction when the cause of the strike is an economic dispute during negotiations over a new collective bargaining agreement. When the cause of the strike is different, the court will tend to issue an injunction.
The first hypothesis assumes that the court strictly implements the legal rules of the game. The particular sub-hypotheses match the formal legal rules as described in the Appendix.
Hypothesis 2: Judges will consider extra-legal characteristics of the strike. The second hypothesis addresses extra-legal characteristics of a particular strike. Optimally, it would have been best to investigate directly the nature of the parties' positions in the dispute (e.g. the wage increase that the workers demand, or the number of workers they demand not be dismissed), but such an inquiry would have required observing immeasurable and incommensurable factors. Consequently, the following variables are hypothesized to proxy the judges' willingness to consider factors other than the procedural rules of the game.
First, it is hypothesized that the greater the hardship inflicted by the strike on the employer or the public, the greater is the willingness of judges to issue an injunction. Such hardship is evident when strikes are in essential services and in the public sector, when work stoppage is complete (rather than partial), when all workers are striking (rather than some), and when the strike extends beyond one organization or sector (rather than being confined only to one organization). These variables are not legal in nature, because the law of injunctions does not explicitly distinguish strikes on the basis of these variables (e.g. no difference exists between injunctions in the public and the private sector). However, the judges are allowed to consider them under the open-ended 'balance of convenience' test that was developed by the court.
Another support for the second hypothesis can be found if judges consider the parties' autonomous efforts to resolve the dispute prior to the involvement of the court. In the time period studied here, labour law did not mandate the parties to negotiate, but negotiations were considered appropriate and necessary by the industrial relations system. It is therefore expected, as part of the second hypothesis, that, if the parties have made no
© Blackwell Publishers Ltd/London School of Economics 2002.
760 British Journal of Industrial Relations
effort to reconcile their differences outside the court before the employer petitions the court for an injunction, the court will refuse to issue the injunction. Similarly, if the employer seeks an injunction before a strike materializes, the court will refuse to issue an injunction. As the court itself has noted, it will not give employers an insurance policy against strikes.
Hypothesis 3: The legal process itself affects the use of judicial discretion. Previous studies have recognized that characteristics of the legal process have an impact on its outcomes. Two important factors were therefore integrated into this study, although they cannot be directly classified as legal characteristics, which are dealt with in the first hypothesis, or as extra-legal considerations, which are dealt with in the second.
One variable, recognized in previous studies as making an impact on judicial discretion, is the employer's petitioning for ex parte hearings (Weiss 1989). In ex parte hearings, the court decides the petition with only the petitioning party (i.e. the employer) present. These findings were limited to the Tel Aviv district labour court during a one-year period. To test this hypothesis, the present study distinguished two stages of the decision: whether the employer petitioned the court to conduct the hearings ex parte, and whether the court consented to the petition.
A second variable tested under this hypothesis is the identity of the judge.
Jurimetric studies in Israel, as elsewhere, have indicated that different judges have different judicial temperaments and beliefs, which are reflected in their decisions. It was therefore hypothesized that, when testing judicial use of discretion in issuing injunctions, different judges will display different levels of responsiveness, and will not act in a monolithic fashion. One limitation in testing this hypothesis must be noted. Given that dozens of judges have decided on petitions for injunctions, some of them sitting in as many as twenty cases and others in only one, it is impossible to observe each judge independently. Aggregating the labour judges in each labour court could therefore serve as an indication regarding variations among judges. The results regarding this hypothesis can therefore be instructive only if courts are found to vary. If no variations exist between courts, there may still be immeasurable variations among judges in each and every court, which simply cancel themselves out in a comparison of the court's judges in aggregate.
2. Method
Data and Measurement Instrument
The database included 324 portfolios, which constituted most cases in which an employer petitioned for an injunction against striking workers, in all labour courts in Israel (five regional courts and the national labour court, both as first instance and as court of appeal) from 1990 to 1997. Limiting the study to cases from 1990 ensured that most court files were still
© Blackwell Publishers Ltd/London School of Economics 2002.
Israeli Judicial Discretion 761
accessible in the courts' archives (except for cases decided in the Jerusalem District labour court in 1992, which were missing and could not be located in the archives). The study ended at the end of 1997, when the president of the National Labour Court resigned and was replaced by a new president. The incumbent president initiated many changes by means of precedents. Given the aim of this study – to examine judicial use of legal discretion – it was conducted only on cases handled under the same legal assumptions. Otherwise stated, the selected time-frame of the study ensured against variations caused by change in the formal statement of the law.1 The concluding section of the paper further develops the significance of the time limitations adopted.
Out of the 324 portfolios, 183 reached a final court decision (56 per cent), while 141 (44 per cent) were not decided by the court because a settlement was reached by the parties themselves. Of the cases reaching a final court decision, in 162 cases the court issued an injunction against striking workers, whereas in 21 cases employers' petitions were rejected, and the strike was not terminated. The variables included in the data were the following.
(a) The dependent variable
As noted, the hypotheses were tested twice. In the first test all compromises were omitted, so the sole dependent variable was whether the judge issued the injunction (1) or refused to issue the injunction (0). In the second test the sole dependent variable was whether the case was settled by the court through adjudication (1) or settled by the parties after the employer petitioned the court for an injunction (0).
( b) The independent variables
The information collected through the database could be divided into three sets of independent variables, corresponding to the three hypotheses outlined earlier:
Variables related to the observation of the legal rule: (a) the strike was legitimate; (b) the strike was protected; ( c) 15 days prior notice was given; (d) notice was given by the authorized representative entity; (e) an industrial peace clause was in effect at the time of the strike.
In addition, different types of strike, classified according to legal distinctions, were used, and each was treated as a distinct variable, as follows: (f) an economic dispute during negotiations on a new agreement; (g) a legal dispute regarding the enforcement or interpretation of an existing collective agreement; (h) disputes over new claims when a collective agreement existed; (i) demands for more influence in managerial decision-making (issues not related directly to pay and employment conditions); (j) a quasi-political strike (aimed at influencing the sovereign and not the employer).
Variables related to non-legal characteristics of the strike which, according to case law, may be considered: (k) whether the strike was in the public or the
© Blackwell Publishers Ltd/London School of Economics 2002.
762 British Journal of Industrial Relations
private sector ('public' being broadly defined to include all organizations directly, fully or partially, supported by the state); (1) whether the employer provided an essential service ('highly essential' being services that can affect human life or security, such as an extensive strike by all physicians employed by the state; 'medium-level' being services that provide important services or the production of commodities that are in great need, such as a strike in the National Insurance offices; and 'low-level' being the residual category, such as a strike in a textile factory); (m) whether there was a full work stoppage (rather than, say, a slowdown); (n) whether the strike was comprehensive (all workers were on strike); (o) whether the strike covered more than one sector of the economy; (p) whether a strike was already in effect when the employer petitioned for an injunction; (q) the parties had made efforts to settle the dispute before strike began.
Variables related to the legal process: (r) the labour court; (s) whether the employer requested ex parte hearings; (t) whether the court approved ex parte hearings.
Data Analysis
Given that most variables, whether dependent or independent, were dichotomous, the hypotheses were examined by chi-square analysis and logistic regression analysis. The latter was applied in order to scrutinize which of the independent variables described above best predicted the dependent variable. The logistic procedure suits linear logistic regression models for both binary or ordinal response data, by the method of maximum likelihood estimates. Coefficients in the logistic regression analysis describe the log odds, or likelihood ratio, of 'success'. The hypothesis that a coefficient is not different from zero is tested by the Wald statistic (Hosmer and Lemeshow 1989). Both forms of analysis provide data regarding the relationship of each independent variable to the dependent variable (outcomes of judicial decisions).
3. Results
Frequency distributions of the data's characteristics are presented in Table 1. As can be observed, 88 per cent of the strikes were not legitimate, in 75 per cent no prior notice was given, and 74 per cent of cases were not authorized by the proper organ of the union (as determined by the union's bylaws). With regard to the cause of the strike, 63 per cent were legal disputes, 6 per cent quasi-political, and 18 per cent economic disputes. In most cases (76 per cent) a partial or full industrial peace clause was in effect; in 81 per cent of cases a partial or full work stoppage was in effect. Existence of a strike prior to petition was recorded in 68 per cent of cases, and in 38 per cent of cases no previous efforts had been made to settle the dispute. Finally, ex parte
© Blackwell Publishers Ltd/London School of Economics 2002.
Israeli Judicial Discretion
763
TABLE 1
Frequency Distributions of Data's Characteristics (N = 324) |
|||
Variable |
Response |
N |
% |
(a) Dependent variable |
|||
An injunction was issued |
154 |
47.5 |
|
No injunction was issued |
29 |
9 |
|
Parties compromised (out of court) |
141 |
43.5 |
|
(b) Independent variables |
|||
Legitimate strike |
Yes |
39 |
12 |
No |
285 |
88 |
|
Prior notice |
Yes |
81 |
25 |
No |
242 |
75 |
|
Union authorized |
Yes |
85 |
26 |
No |
238 |
74 |
|
Legal dispute |
Yes |
121 |
37 |
No |
203 |
63 |
|
Industrial peace clause: |
|||
1. No labour contract |
6 |
2 |
|
2. Partial industrial peace |
230 |
73 |
|
3. Full industrial peace (labour contract) |
10 |
3 |
|
4. Contract about to be signed |
70 |
22 |
|
Quasi-political strike |
Yes |
19 |
6 |
No |
305 |
94 |
|
Protected strike |
Yes |
23 |
7 |
No |
194 |
60 |
|
(not applicable) |
107 |
33 |
|
Economic dispute |
Yes |
57 |
18 |
No |
267 |
82 |
|
Essential service |
High |
98 |
30 |
Medium |
192 |
59 |
|
Low |
34 |
11 |
|
Work stoppage |
Partial |
171 |
53 |
Full |
49 |
15 |
|
Threat of partial |
29 |
9 |
|
Threat of full |
24 |
7 |
|
(not applicable) |
51 |
16 |
|
Existence of strike prior to petition |
Yes |
222 |
68 |
No |
102 |
32 |
|
Previous efforts to settle a dispute |
Yes |
198 |
62 |
No |
120 |
38 |
|
Request ex parte hearings |
Yes |
140 |
44 |
No |
181 |
56 |
|
Court has approved ex parte hearings |
Yes |
42 |
13 |
No |
98 |
30 |
|
(not applicable) |
184 |
57 |
|
Public sector |
Yes |
217 |
67 |
No |
107 |
33 |
Frequency Distributions of Data's Characteristics (N = 324)
hearings were requested in 44 per cent of cases, while in only 13 per cent of cases were actual ex parte hearings conducted.
Results of hypotheses, when tested only on cases in which the labour court fully adjudicated, are presented in Table 2. As shown in the table, the following hypotheses were confirmed. The court tended to issue an injunction (a) when the strike was not legitimate (z2 = 61.550, p = 0.001); (b) when the strike was unprotected (z2 = 18.609, p = 0.001); (c) when 15 days' prior
© Blackwell Publishers Ltd/London School of Economics 2002.
764 British Journal of Industrial Relations
notice of the strike was not given (z2 = 32.583, p = 0.001 ); ( d) when notice was not given by the authorized organ of the union (z2 = 30.483, p = 0.002); (e) when an industrial peace clause was in effect (z2 = 16.481, p = 0.001); (f) whenever the strike concerned a legal dispute over the enforcement or interpretation of an existing agreement (z2 = 8.133, p = 0.004); (g) when no previous efforts had been made to settle the dispute (z2 = 5.950, p = 0.015); (h) contingent on the court in which the petition was litigated (z2 = 17.553, p = 0.004); and (i) when petitioner requested that the case be heard ex parte (x2 = 4.177, p = 0.041). By contrast, the court tended to refrain from issuing an injunction when the cause of the strike was an economic dispute during negotiations over a new collective bargaining agreement (z2 = 16.204, p = 0.001).
Results of hypotheses tested on all cases, comparing those that were fully adjudicated with those in which the parties settled, are presented in Table 3. This table shows that the courts tended to adjudicate more (a) when a strike was legitimate (z2 = 4.230, p = 0.040); (b) when the employer requested ex parte hearings (z2 = 13.406, p = 0.037); ( c) when the court granted ex parte hearings (z2 = 18.754, p = 0.001); and (d) when the rate of adjudication was
TABLE 2
Hypotheses Testing of the Relationship between the Outcome of the Court's Decision to Issue an Injunction and the Independent Variables
Variable |
P-value |
Df |
Conformity with the legal rule Non-legitimate strike Unprotected strike Prior notice (15 days) Union-authorized Industrial peace clause in effect The strike's cause Economic dispute during negotiations of a new agreement Disputes regarding the enforcement or interpretation of an existing collective agreement Disputes regarding new claims despite an existing collective agreement Demands for influence on managerial decision-making Quasi-political strike Non-legal characteristics of the strike Public or private sector Essential service Full work stoppage Comprehensive coverage of all employees Strike covers more than one sector of the economy Existence of strike prior to petition Previous effort to settle dispute |
Characteristics related to the legal process The labour court Request ex parte hearings Court has approved ex parte hearings |
61.550 18.609 32.583 30.483 16.481 |
16.204 8.133 |
3.222 |
0.157 |
1.127 1.282 6.621 1.164 3.976 3.655 5.950 |
17.553 4.177 0.254 |
1 1 1 1 3 |
0.001 0.001 0.001 0.002 0.001 |
0.001 0.004 |
NS |
NS |
1 2 3 3 2 1 1 |
NS NS NS NS NS NS 0.015 |
5 1 1 |
0.004 0.041 NS |
© Blackwell Publishers Ltd/London School of Economics 2002.
Israeli Judicial Discretion 765
TABLE 3
Hypotheses Testing of the Relationship between the Nature of the Petition's Resolution (Adjudication or Compromise) and the Independent Variables
Variable |
P-value |
Df |
Conformity with the legal rule Non-legitimate strike Unprotected strike Prior notice (15 days) Union authorized Industrial peace clause in effect The strike's cause Economic dispute during negotiations of a new agreement Disputes regarding the enforcement or interpretation of an existing collective agreement Disputes regarding new claims despite an existing collective agreement Demands for influence on managerial decision-making Quasi-political strike Non-legal characteristics of the strike Public or private sector Essential service Full work stoppage Comprehensive coverage of all employees Strike covers more than one sector of the economy Existence of strike prior to petition Previous effort to settle dispute Characteristics related to the legal process The labour court Request ex parte hearings Court approved ex parte hearings |
4.230 2.118 0.298 0.221 5.816 |
1.254 2.159 |
1.259 |
1.921 0.682 |
0.373 3.080 3.117 0.172 6.732 0.009 0.000 |
13.406 4.362 18.754 |
1 1 1 1 3 |
0.040 NS NS NS NS |
NS NS |
NS |
NS NS |
1 2 3 3 3 1 1 |
NS NS NS NS NS NS NS |
5 1 1 |
0.020 0.037 0.001 |
contingent on the labour court at which the petition was filed (z2 = 13.406, p = 0.020). All other variables were found to be insignificant.
The regression analysis revealed only two variables as significant predictors of a court's decision to issue an injunction: an illegitimate strike, and a petitioner's request that the case be heard ex parte. Four variables proved significant predictors of a court's strike endorsement: a legitimate strike; an economic dispute during negotiations over a new collective bargaining agreement; a valid collective agreement being in effect; and the strike lasting less than one day.
4. Discussion
Descriptive Findings
At the outset, one of the more notable findings emerging from the study is the high rate of non-legitimate strikes (i.e. strikes not conforming to the legal rules) that reached the court. Although this finding is not directly
© Blackwell Publishers Ltd/London School of Economics 2002.
766 British Journal of Industrial Relations
related to the hypotheses, it has not been documented empirically before, and it affected the testing of the hypotheses. One explanation of this phenomenon could be that the legal rule does not especially influence workers' behaviour. However, such an account clearly disregards the fact that strikes that reach the labour courts are not a representative sample of all strikes in Israel. To consider the proportion of strikes that reach the labour court, and whether they are distinct from the general pattern of strikes in Israel, the findings of the present study may be compared with the database compiled by the Ministry of Labour and Welfare, which documents most strikes in the country.i The database has few variables that indicate judicial intervention in the course of the strike. Out of a total of 1008 strikes in the time period studied here where such an indication existed, the labour court was involved in 239 (23.5 per cent).3 Using 'prior notification by the appropriate organ of the union' as the best proxy for the legitimacy of a strike, approximately 52 per cent of all strikes in Israel were illegitimate; however, this study found that 88 per cent of the strikes that reached the court were illegitimate.
The Decision to Issue an Injunction
The hypotheses consisted of three aspects that could affect the final decision of the labour court to issue an injunction: compliance with the legal rules of the game, the extra-legal characteristics of the strike, and the characteristics of the judicial process. These hypotheses were tested at first only on those cases that ended in a judicial decision (i.e. issuing or refusing to issue an injunction). The predominant factors affecting the labour court's use of discretion proved to be limited to compliance with the legal rule of the game, and to a lesser extent to the nature of the judicial process. Other characteristics of the strike were not found to be significant.
The court was very systematic in its behaviour regarding issuing injunctions. It adopted a highly formalistic approach, namely a strong tendency to issue injunctions when the legal rules of the game were contravened. The evidence in Table 2 supporting this conclusion is unequivocal. Generally, when a strike was not legitimate (i.e. when at least one of the legal rules had not been complied with), the court was most likely to issue an injunction. Furthermore, the logistic regression analysis revealed that the non-legitimacy of the strike was the best (and practically the sole) predictor of a decision to issue an injunction (parameter estimate -3.661; Wald Chi square = 32.078; p = 0.0001). In addition, when out-of-court settlements of the disputes were omitted, only 12 (out of 183) cases did not follow the general practice. In seven cases the court issued an injunction even though the strike was found to be legitimate. In five cases the court decided not to issue an injunction, despite the non-legitimacy of the strike.
The legal rules try to channel the use of strikes to periodic negotiations over the renewal of collective agreements. Although Israeli law is not as strict about the timing of strikes as is the law in other countries (e.g. Canada),
© Blackwell Publishers Ltd/London School of Economics 2002.
Israeli Judicial Discretion 767
and permits strikes during interim periods when agreements are in force, these strikes tend to conflict with industrial peace clauses. The data indeed support this preference for economic strikes during negotiations over a new agreement. Logistic regression analysis indicated that the economic nature of the dispute strongly predicted that the court would endorse the strike (parameter estimate -2.862; Wald chi square = 12.358; p = 0.0004).
In sum, the data document the court's adherence to formal legal considerations, and indicate that these outweigh all others. The remarkably strong relationship between the workers' adherence to the rules of the game and the outcome of adjudication indicates that the courts do not get intensely involved in the details of workers' or employers' substantive demands in negotiations, or in the extra-legal characteristics of industrial action taken. Unlike an arbitrator, mediator, or facilitator, who usually seeks to advance the parties' relationship, the labour court adopts a more formalistic approach, characteristic of courts in general, according to which a dispute must be resolved – and can be resolved – by mere reliance on procedural rules. It is worth emphasizing again that, on the basis of its own jurisprudence, the labour court is not compelled to issue an injunction upon finding that one of the formal requirements was not complied with.
Consistent with the analysis thus far, the second category of variables, which are related to the particular characteristics of a strike (essential services, full work stoppage, existence of a strike prior to petition), proved insignificant in explaining the court's use of discretion. By contrast, in cases involving parties that had made no effort to reconcile their differences before the employer petitioned for an injunction, the judges tended to issue injunctions. However, the exceptional significance of this variable can be explained by the formal legalistic orientation of the court as well. The requirement to negotiate in good faith, which was introduced into the Law of Contracts (General Part 1973), has been gradually elevated to become a supreme principle governing contractual and quasi-contractual relationships, including the collective labour-management relationship (Yadlin 1999). One of the duties deriving from the principle of good faith is to negotiate 'open-heartedly', although there is no obligation to reach an agreed outcome. This doctrine, which developed throughout the 1980s, can account for the marked judicial tendency to issue injunctions in the 1990s, compared with findings reported for the earlier years of the labour courts. Although negotiations were declared mandatory by law only after the period of time studied here, their significance was nevertheless prevalent in previous years.
It is noteworthy, however, that unlike illegitimate strikes, in which the workers are responsible for non-compliance with the legal rule, the absence of prior negotiations is generally an outcome of one or both parties' behaviour. In issuing injunctions in these cases, the court may motivate employees to negotiate. At the same time, the incentive for the employer to negotiate diminishes once the force of injunction has removed the threat of the strike. The tendency to issue more injunctions as a response to the
© Blackwell Publishers Ltd/London School of Economics 2002.
768 British Journal of Industrial Relations
absence of previous efforts to negotiate places the burden on the employees, and, given the denial of their right to strike, it may be detrimental to their bargaining position.
Finally, the characteristics related to the legal process itself were found to affect the outcomes of adjudication. The evidence indicates significant differences among the various courts, hence also significant differences among judges. To interpret this result, it is necessary to recognize that all labour judges are appointed by a centralized appointments committee on a nonpolitical basis, and also are occasionally transferred from one labour court to another. Therefore, no reason exists to assume inherent regional differences among courts. At the same time, the evidence indicates a lesser inclination in the national labour court to issue injunctions than in the regional courts. The labour court hears some of these petitions on appeal, and more rarely at first instance. The differences between the regional courts and the national labour court are therefore difficult to generalize and may stem from various causes. However, they do support the observation that there is a discrepancy between the judge-made law formulated at the national labour court and its routine implementation in the district courts.
Second, when a request was made for ex parte hearings, judges were found to evince a greater tendency to issue injunctions, in contrast to cases where such a request was not made. Yet the actual approval of ex parte hearings was not a significant variable in the court's final decision to issue or deny an injunction, although the reason for this finding is mostly methodological. Because the regression analysis was conducted only on the cases in which ex parte hearings were requested, and in which the marked tendency to issue an injunction was already established, the second stage – whether or not the request was approved- was found to be insignificant. Note however that in each and every case, without exception, where ex parte hearings were approved, and an out-of-court settlement was not reached, an injunction was issued.
Two explanations can be provided concerning the significance of petitions for ex parte hearings. The first is that the court approves ex parte hearings when the facts are lucid, and the costs incurred by the strike are so significant and imminent, that the outcome is virtually considered to be known in advance. The second explanation is that lawyers add to the petition, offhandedly, a request for ex parte hearings. The lawyers' assumption is that the mere request for ex parte hearings creates some kind of a judicial disposition in favour of issuing an injunction.
Clearly, the first explanation must account for at least some of the cases.
However, the data indicate that the second explanation cannot be dismissed altogether. First, if ex parte hearings are granted when the outcome is quite clear, we will expect an even higher share of cases in which ex parte hearings are permitted. This is simply because most strikes that reach the court are not legitimate, and, as is clearly demonstrated by the evidence, the judges' willingness to issue an injunction in these cases is almost automatic. Yet ex parte hearings require more than certainty regarding the outcome. It is
© Blackwell Publishers Ltd/London School of Economics 2002.
Israeli Judicial Discretion 769
necessary to demonstrate some sense of urgency that justifies violating procedural justice. It would hardly seem reasonable to think that 44 per cent of the cases encompass such a sense of urgency. In fact, the data do not reveal that ex parte hearings were requested, approved and conducted only when strikes were in sectors classified as essential. The findings therefore lend some support to the employers' effective use of petitions for ex parte hearings, given that the mere request is sufficient to predict the outcome of the case.
The Decision to Adjudicate
The three hypotheses were tested again, distinguishing between cases that were adjudicated and those that were terminated upon a settlement reached by the parties themselves. The study found that 44 per cent of the cases were not fully adjudicated because the parties reached a settlement. The significant number of cases in which judges were not required to decide is puzzling. The judicial process in this context is expedited. Once a petition is filed, it is assigned to a judge, who immediately views the petition and schedules the initial hearings on the basis of the petition's urgency. The initial hearings are often conducted within hours or at most a few days after the petition is filed. Interim decisions to issue an injunction can therefore be expected within a very short time. Given the nature of these disputes, the interim decision usually also ends the role of the court. In theory, permanent injunctions are given after a full hearing of the circumstances, which may last for a few months. Neither strikers nor employers tend to be patient and to pursue the dispute to the stage of a permanent injunction. The relevant time frame is therefore that of the interim injunction.
This short time frame calls for an explanation of the employer's willingness to settle with workers after petitioning the court for an injunction. Given the short time lapse between the petition and the interim decision, it is assumed that only in a negligible number of cases is the settlement a result of the strike's termination for reasons wholly unrelated to the employer's petition. In other cases the legal process serves as a catalyst for reaching an agreement. Sometimes the mere petition enables the parties to advance their negotiations and reach a settlement independently. In most cases, however, the settlement is conducted with the active participation of the judge. This may take place at the initial screening of the petition, when the judge discusses the petition with the parties. Often the judge postpones the decision for a certain length of time, ordering the parties to resume negotiations, and to update the court on progress. In these situations the judge occasionally suggests various propositions to aid the parties. Sometimes these are suggestive in nature; at other times they are proposed as an 'offer one cannot refuse'. Whether the judge imposes a compromise or mediates between the parties towards a compromise is a difficult matter that rests on values and facts. The cases ending in compromises are generally closed with a handwritten decision by the judge holding that the petition is cancelled.
© Blackwell Publishers Ltd/London School of Economics 2002.
770 British Journal of Industrial Relations
The laconic decision does not reveal the process that brought it about, and the process itself is not recorded in the court's protocols. Beyond the factual question, there is also a value-laden issue of how to interpret these outcomes. When a judge 'warmly suggests' that the parties cease industrial action and resume negotiations, a party's rejection of the offer may be detrimental to its cause if the judicial process continues. Therefore, the difference between mediation and coercion seems to be a matter of degree.
In testing the three hypotheses, when the dependent variable was whether or not the court adjudicated the case, the findings generally matched those of the previous testing. The most significant finding was that, when a strike was legitimate, there was a greater likelihood of adjudication, rather than an independent settlement by the parties. These were the cases in which judges more frequently denied the employer's petition for injunction, and in which workers had greater leverage. The workers were therefore reluctant to compromise. Similarly, requests for and approval of ex parte hearings decreased the number of out-of-court settlements. The explanation for this finding is the flip side of the previous explanation. Employers who are well aware of the high probability that they will be granted an injunction if they request ex parte hearings – and, moreover, if their request is approved – are not likely to settle the case without the relatively predictable injunction they are likely to obtain. Consistent with the first testing of the hypotheses, it was found again that the judges' identity was a significant determinant, indicating that different judges (proxied by labour courts) tend to balance their adjudicating and mediating roles differently. Finally, similar to the findings in the first testing of the hypotheses, the extra-legal variables were altogether insignificant.
Role of Labour Judges in Industrial Disputes
The findings confirm that paramount in a court's decision to issue injunctions are considerations related to compliance with the legal rule. Characteristics related to the legal process itself were also found to affect the final outcome, unlike extra-legal characteristics of the strike itself, which carried no implications for the implementation of the legal rule in fact. The same determinants were similarly found to affect the decision of the parties to settle the case, rather than to reach the final outcome of adjudication.
At first glance these findings may seem obvious: the labour court implements the rule oflaw. However, this is an over-simplification oflabour law's objective, and accordingly of the labour courts' task. The findings indicate that the labour court functions like a regular civil court, but resists an alternative vision of a tribunal, which was in fact the underlying reason for its establishment. Moreover, even the formal statement of the law, mostly encoded in the courts' case law, does not assume that the court should be merely administering technical rules. The findings are hardly a statement of the obvious.
What do these results indicate regarding the role of the Israeli labour judges in industrial disputes? First, the dominance of the legal variables as
© Blackwell Publishers Ltd/London School of Economics 2002.
Israeli Judicial Discretion 771
opposed to the insignificant effect of the non-legal variables indicates that the labour court judges adhere to the formalistic law approach, and are concerned with simple maintenance of compliance with the legal rule. The findings are consistent with earlier studies asserting that the labour court has shifted the perception of its own task from that of an industrial mediator in the 1970s to that of a formalistic adjudicator in the 1980s (Galin and Shirom 1978; Weiss 1989). In balancing the competing roles ascribed to the labour judges, the judges themselves prefer the role of adjudication.
This strict adherence to the legal rule requires an account for the significant share of cases ( 43 per cent) that are settled without a judicial decision. However, compared with previous years, as well as with individual labour disputes, this is a relatively small proportion of the total number of cases. Moreover, the findings indicate that an important determinant for settlement is the expected outcome of litigation, given the consistent behaviour of the courts. The judges' mediating role is therefore partially passive. It is not so much that the judge succeeds in creating an alternative form of communications that does not rely on the legal discourse of rights and obligations. Rather, the judges, or even the threat of judicial intervention, signal the expected outcome of adjudication to the parties, so that an out-of-court settlement is for the most part a catalyst for clearing the case from the court's docket.
Several explanations can be provided for the judicial preference of the formalistic approach. Generally, this approach seems to be favoured by judges since it enables them to make a lucid and straightforward decision, devoid of value judgements. Judges apply a straightforward yardstick: compliance with the legal rule. Consequently, the judge succeeds in achieving 'judicial peace', but is less concerned with the court's potential role in inducing 'industrial peace'. It seems that the judge seeks legitimacy of the court's decisions on the basis of the mode of communications of the legal system, not of the industrial relations system.
The inclination towards the formalistic judicial task can also be explained by the nature of the judicial process in these cases. Employers submit petitions to the court in times of an industrial crisis. Given that most strikes take place in the public sector, there is also the occasional pressure created by those who will be harmed by industrial action in public services, backed by intensive press or electronic media coverage. Hearings are scheduled within a matter of hours, or at most a few days. Occasionally a decision must be given before the morning news, in order to inform workers and users of public services (e.g. schoolchildren's parents) whether a strike will be in effect. The process is rather informal, and decisions tend to be very short. The judge has a lot to lose if efforts at reconciliation do not succeed. It seems that, when the courtroom environment is so intense, clinging to the formal rules is a relatively safe haven that can protect the judge from criticism.
Finally, despite the interpretation of the findings given here, whereby judges prefer to act as gatekeepers of the legal rule, a competing explanation
© Blackwell Publishers Ltd/London School of Economics 2002.
772 British Journal of Industrial Relations
holds that judges may actually be hostile to strikes. The descriptive findings indicate that a petition seeking an injunction is almost a secure avenue to a consequence favourable primarily to employers, because in 47.5 per cent of the cases the court issues an injunction, and in almost 44 per cent of the cases the parties reach an out-of-court compromise that involves the cessation of the strike. However, given that most strikes that reach the court are not legitimate, judicial hostility can neither be confirmed nor ruled out.
Judicial Discretion in Context: The Impact of Formalistic Adjudication on the Industrial Relations System
The relationship between the legal and the industrial relations systems is complex. Labour law seeks to accomplish two objectives that are inherently discordant. On the one hand, its seeks to dominate the industrial relations system; on the other, and at the same time, it seeks to facilitate its autonomous decision-making capability. The latter function indicates not the need to leave the industrial relations system alone, but rather the need to recognize its objective – the promotion of shared regulation of the employment relationship – and to provide the rules and means that will best serve this objective. The labour courts are entrusted with carrying out this dual task. The law of strikes provides the judges with a wide leeway to balance the dual objectives of the law, and the judges' decision-making-infact constructs the precise substance of the law. This is the process of translating the formal statement of the law into law-in-action.
The findings outlined thus far shed light on the institutional position of the labour court vis-a-vis the industrial relations system. The establishment of the labour court was motivated by the industrial relations agents' desire for a legal forum to accommodate their particular needs. These were defined earlier as the need for expertise in the area of labour, and the need for a dispute resolution process distinct from that provided by the regular civil court system. The formal application of the procedural rules does not deliver either of these two objectives. Counting the number of days of 'advance notice' does not require any particular expertise, nor does it provide an alternative to standard adjudication. The judges' quest for legal peace (rather than industrial peace) was explained in the previous section, but it risks the alienation of the industrial relations agents from what was to have been 'their own courtroom'.
Arguably, formalistic adjudication has the potential to promote the autonomy of the industrial relations system. The judges' sole concern with the parties' compliance with procedural rules may be interpreted as according respect to the autonomous parties, and delegating to the industrial relations system sole responsibility for administering sanctions that are sensitive to all extra-legal issues. For example, repeated strikes in the public sector should be penalized non-legally by the withdrawal of public support for labour or, alternatively, remedied by a redesign of the public employment system. These are deemed not legal issues, but issues of policy and autonomous law-making by means of collective bargaining. However, the
© Blackwell Publishers Ltd/London School of Economics 2002.
Israeli Judicial Discretion 773
withdrawal of the courts from all of these issues reflects a very particular sense of granting autonomy to the parties. The court seeks to 'leave the parties alone' to achieve whatever outcomes they can feasibly negotiate. If they cannot do so, it is their 'autonomous choice'. The reasons for this 'choice' are deemed outside the law's interest and reach. There is however an alternative conception of autonomy and choice, according to which the law should facilitate autonomy in the sense of aiding the parties to reach a constructive and sustained resolution of disputes. This requires the court not simply to leave the parties alone, but to aid them in regaining a sense of stability and shared norms that can keep the industrial relations system intact. The evidence produced here does not support such a view. Formalistic adjudication is therefore merely a mode of legal governance, not of legal facilitation.
During the last decade the Israeli industrial relations system has undergone tremendous change, and what was once an almost paradigmatic corporatist system has adopted the characteristics of liberal Anglo-American systems (Cohen et al. 2003). Since 1998 the labour court, headed by a new president, has tried to develop labour law in order to halt the decline of collective bargaining and to compensate for the weakness of organized labour. This has been accomplished through dramatic changes in the formal statement of the law, and an effort to reclaim the facilitative role of the labour courts. In a small number of high-visibility precedents, the president of the labour court has emphasized the role of the labour court as an industrial mediator, while explicitly holding that this task requires rejection of a formalistic implementation of procedural law, as observed in this study. The immediate response of employers has been to withdraw from the courtroom and to resort to self-help methods (such as lockouts). This process is indicated by a drop of approximately 50 per cent in the number of petitions filed by employers for injunctions against striking workers (although the overall number of strikes has not declined). Employers therefore demonstrate their lack of interest in the labour court as a venue for third-party intervention in collective labour disputes. The labour court is currently facing a serious legitimacy crisis.
The labour court's present legitimacy crisis sheds light on the findings of the present study and can advance two competing accounts. One account would suggest that formalistic adjudication was important to sustain the court's legitimacy in the industrial relations system. The second account suggests that years of formalistic adjudication, at the price of a greater involvement in facilitating the constructive resolution of labour disputes, has made the labour court a distant and unimportant agent in the industrial relations system. Both accounts are possible and in fact need not be mutually exclusive. The outcome is one and the same: the labour court no longer holds the position it did in the 1970s. Whether this is a result of the dramatic change in the industrial relations system, which no longer seeks alternative forms of adjudication, or an outcome of its own approach to adjudication, which, as depicted here, distanced itself from its facilitative task, is a matter for further inquiry.
© Blackwell Publishers Ltd/London School of Economics 2002.
774 British Journal of Industrial Relations
Appendix: Summary of the Law on Strikes in Israel
The right to strike is a fundamental right in Israeli law. A 'strike' is defined as any concerted refusal to work.
When is an Industrial Action not a 'Strike'?
- When it is an individual employee's refusal to provide work.
- When it is a concerted refusal to provide work when the employees' demands are not about the conditions of work or industrial relations.
- When it is a concerted refusal to provide work that is directed towards an entity other than the employer (a political strike).
- When it is a quasi-political strike, defined as a strike directed at the sovereign, on matters of wages or working conditions (permissible for only a short duration).
- When workers cease work, but their action does not fall under the definition of a strike (in which case the Court must issue an injunction holding them to go back to work).
- A strike can be called only in the course of an economic dispute, not in a legal dispute. In this latter case, the parties must adjudicate their dispute in courts or through alternative dispute resolution mechanisms.
- If the parties are bound by a collective bargaining agreement, the strike must not violate existing industrial peace obligations (absolute or relative). It is assumed that collective agreements always contain, even implicitly, a partial industrial peace clause.
- According to the law of collective bargaining, when the agreement's contractual period expires, the agreement continues to apply. The court does not however enforce industrial peace clauses at that stage.
- Parties to a conflict must give 15 days' prior notice before beginning industrial action.
- Only the union's organs authorized by its constitution are allowed to declare a strike.
- Strikers must avoid violating employers' property rights and damaging physical property.
- If all the above rules are observed, a strike is legitimate. Otherwise it is not legitimate, and the court may issue an injunction against the striking workers. If the strike is legitimate, the court still has the discretion to issue an injunction based on general principles, such as the balance-ofinterests test.
- All strikes (legitimate and non-legitimate) provide workers with immunity against dismissals and a partial immunity from being sued in
- When a strike is unprotected, the employees no longer enjoy the common protection accorded to striking workers. The employer can therefore sue the workers for a tort committed, or dismiss the striking worker. Such cases are extremely rare in Israeli law.
- The courts will be reluctant to issue an injunction, given the importance of the right to strike. Issuing an injunction is the final resort, and the court will seek more optimal means to resolve the dispute.
- The injunction is not a substitute for collective bargaining, and should be viewed as an aid to bring the parties back to the bargaining table.
The court tends to issue an injunction when the workers have not tried to resolve their dispute by other means. The court avoids issuing an injunction when an employer refuses to bargain with the workers. - The court refuses to issue an injunction as a carte blanche against all future industrial action.
- The court will not issue an injunction ex parte except in rare circumstances.A Legitimate Strike is a Strike that Observes the Rules of the GameAn Unprotected Strike
© Blackwell Publishers Ltd/London School of Economics 2002.
Israeli Judicial Discretion 775
torts. The term 'unprotected strike' is reserved for some strikes in the public services, as defined in s.37 A of the Law of Industrial Disputes Resolution.
General Considerations Guiding the Court (based on precedents)
Final version accepted 26 March 2002.
Acknowledgements
This study was carried out with the aid of grants from the Israel Science Foundation and the Israel Foundation's Trustees.
Notes
- The only exception to the stability of the formal law on injunctions against striking workers is the Supreme Court's recognition of the quasi-political strike in 1992, allowing it for a short period of time (which was established in later implementations of the law as ranging between 1 and 8 hours). Given the very small number of quasi-political strikes, this change in the law has a negligible effect on the study.
- The database is compiled on the basis of a formal questionnaire submitted to employers by the Ministry of Labour and Welfare, after it has become known that they had a strike in their establishment. The employer's reports are complemented by data generated by the Histadrut and from the media. The collection method is incomplete, and not all strikes that were brought to the labour court were identified in the database. This, however, is the formal database the Israeli government uses, and on which reports to the International Labour Organisation are based.
- This figure requires two explanations. First, judicial intervention in a strike may be required in a context other than a petition for an injunction (e.g. in a union's petition for an injunction against an employer in the event of a lockout). On the basis of a general reading of the case law, these situations are negligible. Second, there is a discrepancy between the 239 strikes indicated by the Ministry's database and the number of cases studied here (324). This discrepancy can be explained almost in full by the fact that 102 petitions for injunctions were made before the strike commenced (see Table I), while the Ministry's database includes only strikes, and not those industrial disputes in which a strike was threatened but did not materialize.
© Blackwell Publishers Ltd/London School of Economics 2002.
776 British Journal of Industrial Relations
References
Bar-Niv, Z. (1974). 'The labour courts and their place in the legal and labour systems of Israel'. Israel Law Review, 9: 558.
Ben-Israel, R. (1977). The Industrial Tribunal in Israel, report on industrial tribunals in Israel presented to the ILO. Geneva: ILO.
— (1988). International Labour Standards: The Freedom to Strike. Deventer:
Kluwer.
Blanpain, R. and Ben-Israel, R. (eds.) (1994). 'Strikes and lockouts in industrialized market economies'. In Bulletin of Comparative Labour Law, Vol. 29.
Chermesh, R. (1983). Strikes in Israel as a Social Problem (in Hebrew). Tel Aviv:
Institute for Productivity of Work.
Cohen, Y., Haberfeld, Y., Mundlak, G. and Saporta, I. (2003). 'Unpacking union density: union membership and coverage in the transformation of the Israeli industrial relations system', Industrial Relations, forthcoming.
Cooke, W. et al. (1995). 'The determinants of NLRB decision-making revisited'.
Industrial and Labour Relations Review, 48: 237-57.
Dickens, L. (1994). 'Deregulation and employment rights in Great Britain'. In R. Rogowski and T. Wilthagen (eds.), Reflexive Labour Law. Deventer: Kluwer, pp. 225-48.
Dunlop, J. (1993). Industrial Relations Systems, 2nd rev. edn. Cambridge, Mass.:
Harvard Business School Press.
Galin, A. and Harel, A. (1978). Developments and Change in the Industrial Relations System in Israel (in Hebrew). Tel Aviv: Masada.
— and Shirom, A. (1978). Assessment of the Labour Courts' Impact on Collective Disputes (in Hebrew). Tel Aviv: Faculty of Business.
Gall, G. and McKay, S. (1996). 'Injunctions as a legal weapon in industrial disputes'.
British Journal of Industrial Relations, 34: 567-82.
Hosmer, D.W. and Lemeshow, S. (1989). Applied Logistic Regression. New York:
John Wiley.
Israel Labour Court (1998). Annual Report to the Department of Labour and Welfare. Unpublished.
Kairys, D. (ed.) (1990). The Politics of Law: A Progressive Critique. New York: Basic Books.
Kennedy, D. (1976). 'Form and substance in contract law adjudication'. Harvard Law Review, 89: 1685-778.
Mautner, M. (1988). 'Rules and standards in the new civil legislation: a question of legislative doctrine' (in Hebrew). Law, 17: 321-52.
© Blackwell Publishers Ltd/London School of Economics 2002.
Israeli Judicial Discretion 777
Porcher, P. (1994). 'The changing role of the French labour courts in employment relations'. In R. Rogowski and T. Wilthagen (eds.), Reflexive Labour Law. Deventer: Kluwer, pp. 249-62.
Pound, R. (1910). 'Law in books and law in action'. American Law Review, 44: 12-47.
Rogowski, R. (1994). 'Industrial relations, labour conflict resolution and reflexive labour law'. In R. Rogowski and T. Wilthagen (eds.), Reflexive Labour Law. Deventer: Kluwer, pp. 53-94.
— and T. Wilthagen (1994). 'Introduction'. In R. Rogowski and T. Wilthagen
(eds.), Reflexive Labour Law. Deventer: Kluwer, pp. 1-20.
Teubner, G. (1993). Law as an Autopoietic System. Oxford: Blackwell. Wedderburn, Lord (1995). Labour Law and Freedom. London: Lawrence & Wishart. Weiler, P. (1990). Governing the Workplace. Cambridge, Mass.: Harvard University
Press.
Weiss, R. (1989). 'Injunctions in the labour courts' (in Hebrew). Unpublished Masters' thesis, Tel Aviv University.
Yadlin, 0. (1999). 'Good faith in employment relations: from class to individual' (in Hebrew). Iunei Mishpat, 22: 867-900.
Zamir, Y. (1974). 'The law of labour disputes'. Israel Law Review, 9: 548-57.
© Blackwell Publishers Ltd/London School of Economics 2002.