Abstract
The current work seeks to ascertain whether rulings on dismissal cases issued by incumbent judges in Spanish labour courts are influenced by whether they are acting alone in their own court or sharing duties with other judges such as replacement judges, support judges or incumbent judges from other courts. We consider that a court is treated when more than one judge rules in it. Then, an analysis is conducted so as to determine the effect of such a treatment on the percentage of cases ruled in favour of the dismissed worker. The data used in the research are taken from the information recorded at court level provided by the statistics kept by the General Council of the Spanish Judiciary. A total of 2888 observations were available, corresponding to the period spanning 2004 to 2012. As regards the findings, it may be concluded that there is a significant positive impact on the number of dismissal cases ruled in favour of workers when incumbent judges are not acting alone in their court, particularly when the incumbent judge solves cases together with another professional judge.
This is a preview of subscription content, log in via an institution to check access.
Source: Authors’ own based on Frick et al (2012)
Source: Authors’ own based on CGPJ data
Source: Authors’ own
Source Authors’ own
Similar content being viewed by others
Notes
As well as those mentioned, it could be added that the work of Martín-Román et al. (2013) reports that incumbent judges and replacement judges in labour courts behave differently.
There is a third reason for separating our database into these two groups of non-incumbent judges, although it is strictly statistical: the structure and size of our database recommends such a distinction (the following section deals with this issue in greater detail).
In the former work, the author models the determinants of severance pay for dismissal in Spain through negotiation prior to any trial. In the second, the main interest lies in modelling severance payments in cases of collective dismissals.
This finding concurs with that of Donohue and Siegelman (1991), who claim that workers tend to make greater use of the legal system during periods of economic downturn if their job is at risk.
The author justifies this finding drawing on experimental evidence from the work of Farber and Bazerman (1986). For the case of the United States, these authors find that in arbitration proceedings concerning wage rises, the arbitrator tends towards the position of the firm when the latter is in a precarious financial situation.
Two studies exploring other effects of the business cycle on labour courts in Spain are those of García-Martínez and Malo (2007) and Frick et al. (2012). The former examines how the business cycle affects companies’ strategic use of individual dismissals compared to collective dismissals in an effort to adjust the workforce. In the latter, macro panel regressions are performed with the 17 Spanish regions and the 11 German states as units. The main finding to emerge is that when the business cycle is at a low point workers are more prone to use the labour courts when dismissed (as well as in cases concerning salary disputes). In a similar vein, Berger and Neugart (2011) also report a positive link between the legal activity of German labour courts and unemployment.
In order to delve more deeply into the definition of “peer effects”, three important references are Manski (1993, 2000) and Dietz (2002), who explain the various types of social effects or proximity effect that might exist. Strictly speaking, the real “peer effects” correspond to the endogenous effects in Manski’s classification. This corresponds to the “emulation effect” defined in Martín-Román et al. (2015). Correlated and exogenous effects are also social effects although their rationale differs somewhat. A detailed explanation of how the three types of proximity effects can operate in Spanish labour courts can also be found in Martín-Román et al. (2015).
This process is compulsory and it is regulated in articles 63 to 68 of Labor Procedure Law.
Royal Decree-Law 10/2010, of 16 June, applying urgent measures for labour market reform and Law 35/2010, of 17 September applying urgent measures for labour market reform.
Royal Decree Law 3/2012, of 10 February, applying urgent measures for labour market reform and Law 3/2012, of 6 July, applying urgent measures for labour market reform.
It is sufficient for a firm’s revenue or sales to be less than for the same quarter of the previous year.
Although information is available for before 2004, it cannot be used since it is not broken down into type of judge.
By “extreme situations” we mean that when there are few cases it is easier to find extreme values in the courts’ resolutions (i.e. observations with 0% or 100%). Put it in other words, the higher is the number of rulings, the closer we get to the mean value of the variable, according to the law of large numbers. In any case, we have checked whether our results were sensitive to this decision. We have carried out different econometric regressions including those observations that were removed originally and our outcomes are barely affected. These results are available upon request.
For those interested, the results of these tests are available upon request from the authors.
We tried different specifications, removing these variables from the regression, and the result was that the treatment effect rose. Therefore, we deem that those covariates are relevant for the analysis and should remain in the basic specification.
There is another indicator, used in Italian Judicial System, which can be obtained as the sum of pending cases at the beginning and at the end of the period divided by the incoming cases plus the number solved cases and then multiplied by 365. We have tested both indicators and the results we got were very similar.
As already mentioned in the descriptive section, all the courts in which fewer than ten dismissal cases were handled were removed from the analysis to correct part of this effect.
This effect associated to the annual dummy variables is what might be leading to the unemployment rate’s lack of significance.
Except in the case of module I where there was no control for heteroscedasticity and, therefore, the effect is identical on treated and non-treated.
The inclusion of this variable in the regression together with the case duration reduces the statistical significance of both variables without yielding major consequences in the treatment effect. For this reason, we use the case duration as the instrumental variable in our baseline model within the main text and utilize the congestion rate as an alternative instrumental variable in the robustness analysis within the appendix.
The analysis was also repeated with 500 iterations and the results were identical.
References
Berger, H., & Neugart, M. (2011). Labor courts, nomination bias, and unemployment in Germany. European Journal of Political Economy, 27, 659–673.
Berger, H., & Neugart, M. (2012). How German labor courts decide: An econometric case study. German Economic Review, 13(1), 56–70.
Booth, J. F., & Freyens, B. P. (2014). A study of political activism in labour courts. Economics Letters, 123(3), 370–373.
Brennan, T., Epstein, L., & Staudt, N. (2009a). The political economy of judging. Minnesota Law Review, 93, 1503–1524.
Brennan, T., Epstein, L., & Staudt, N. (2009b). Economic trends and judicial outcomes: A macrotheory of the court. Duke Law Journal, 58, 1191–1230.
Burgess, S., Propper, C., & Wilson, D. (2001). Explaining the growth in the number of applications to industrial tribunals, 1972–1997. Employment Relations Research Series. Londres: Departamento de Industria y Comercio.
Cerulli, G. (2014). ivtreatreg: A command for fitting binary treatment models with heterogeneous response to treatment and unobservable selection. Stata Journal, 14(3), 453–480.
Cox, D. R. (1958). Planning of experiments. New York: Wiley.
Danziger, S., Levav, J., & Avnaim-Pesso, L. (2011). Extraneous factors in judicial decisions. Proceedings of the National Academy of Sciences, 108(17), 6889–6892.
Dietz, R. D. (2002). The estimation of neighborhood effects in the social sciences: An interdisciplinary approach. Social Science Research, 31(4), 539–575.
Donohue, J. J., & Siegelman, P. (1991). The changing nature of employment discrimination legislation. Stanford Law Review, 43, 983–1033.
Farber, H. S., & Bazerman, M. H. (1986). The general basis of arbitrator behavior: An empirical analysis of conventional and final-offer arbitration. Econometrica, 54, 1503–1528.
Fraisse, H. (2010). Labour disputes and the game of legal representation. CESIFO Working Paper No. 3084.
Frick, B., Malo, M. A., García-Martínez, P., & Schneider, M. (2012). The demand for individual grievance procedures in Germany and Spain: Labour law changes versus business cycle. Estudios de Economía Aplicada, 30, 283–310.
Galdón-Sanchez, J. E., & Güell, M. (2000). Let’s go to court! Firing costs and dismissal conflicts (No. 823). Princeton University, Department of Economics, Industrial Relations Section.
Galdón-Sánchez, J. E., & Güell, M. (2003). Dismissal conflicts and unemployment. European Economic Review, 47(2), 323–335.
García-Martínez, P., & Malo, M. A. (2007). The strategic use of dismissal legislation: An empirical analysis using Spanish data. European Journal of Law and Economics, 23(2), 151–167.
García-Rubio, M., & Rosales-López, V. (2010). Justicia y economía: evaluando la eficiencia judicial e Andalucía. Revista para el Análisis del Derecho, 4, 1–26.
Ichino, A., Polo, M., & Rettore, E. (2003). Are judges biased by labor market conditions? European Economic Review, 47(5), 913–944.
Jimeno, J. F., Martínez-Matute, M., & Mora-Sanguinetti, J. S. (2015). Employment protection legislation and labor court activity in Spain. Documentos de Trabajo Banco de España, 1507.
Malo, M. A. (2000). A simple model of severance pay determination: The case of individual dismissals in Spain. Labour, 14, 269–290.
Malo, M. A. (2001). European labour law and severance pay determination in collective redundancies. European Journal of Law and Economics, 12(1), 73–90.
Malo, M. A., & Pérez, J. (2003). Individual dismissals in Europe and the United States: A model on the influence of the legal framework on firing costs. European Journal of Law and Economics, 15, 47–63.
Manski, C. F. (1993). Identification of endogenous social effects: The reflection problem. Review of Economic Studies, 60(3), 531–542.
Manski, C. F. (2000). Economic analysis of social interactions. Journal of Economic Perspectives, 14(3), 115–136.
Marinescu, I. (2003). L’application du droit du travail est-elle biaisée par les conditions économiques. In Papers for the ADRES conference, Paris.
Marinescu, I. (2011). Are judges sensitive to economic conditions? Evidence from UK employment tribunals. Industrial and Labor Relations Review, 64, 673–698.
Martín-Román, Á., Moral, A., & Martínez-Matute, M. (2013). Tipo de juez y estimación de los casos de despido: un análisis de los Juzgados de lo Social en España. Cuadernos de Economía-Spanish Journal of Economics and Finance, 36(102), 142–154.
Martín-Román, Á., Moral, A., & Martínez-Matute, M. (2015). Peer effects in judicial decisions: Evidence from Spanish labour courts. International Review of Law and Economics, 42, 20–37.
Mora, J. J. (2005). El justo despido y el mercado laboral: Una aplicación para España. Semestre Económico, 8(16), 119–127.
Mora, J. J. (2006). La dinámica de los despidos en España. Estudios Gerenciales, 98, 131–138.
Pastor, S., & Manspons, L. (2004). Cifrar y descifrar: indicadores judiciales de las Américas (Vol. II). Chile: Centro de Justicia de las Américas.
Posner, R. A. (2008). How judges think. Cambridge, MA: Harvard University Press.
Rosales-López, V. (2008). Economics of court performance: An empirical analysis. European Journal of Law and Economics, 25(3), 231–251.
Rubin, D. B. (1978). Bayesian inference for causal effects: The role of randomization. The Annals of Statistics, 6, 34–58.
Viscusi, W. K. (1999). How do judges think about risk? American Law and Economics Review, 1(1), 26–62.
Acknowledgements
The Ángel Martín-Román was partially supported by the Spanish Ministry of Economy and Competitiveness under project ECO2014-52343-P, co-financed by FEDER funds. The Alfonso Moral has been partially supported by Ministry of Economy and Competitiveness under project CSO2015-69439-R.
Author information
Authors and Affiliations
Corresponding author
Appendices
Appendix 1
See Table 4.
Appendix 2: Robustness analysis
The final part of the econometric analysis involves using a robustness analysis to check the sensitivity of the results obtained. With this goal in mind, three different approaches are followed: first a bootstrap analysis is conducted to determine whether the decomposition of the effects of treatment is significant and whether said effects change for random sub-samples. Secondly, the estimations are repeated (applying the more complete model) using maximum likelihood techniques to gauge the effect of treatment. Finally, we carry out additional regressions taking the most completed specification as the benchmark model and substituting the case duration by the court’s workload as the instrumental variable.Footnote 21
Table 5 shows the results of bootstrap analysis on the effects of the two treatments, both on treated as well as non-treated. The analysis shown in the table has been carried out using 100 iterations.Footnote 22As can be seen, there are no major changes in the values observed in Tables 2 and 3, thus indicating the results are robust. Overall, it can be seen that the decomposition in the more comprehensive models is significant for the two treatments and that the effects are noticeably greater in cases when judges other than replacement judges are acting in the same court. It is not clear whether the effect is greater on those treated or on those not treated, although the size of the aggregate effect remains the same in the case of both treatments.
The second part of the robustness analysis repeats the estimations carried out using maximum likelihood methods of a linear model augmented with a specification of treatment which enables endogeneity to be corrected. The estimations performed are displayed in Table 6 which presents two results for each treatment. In all cases, the more complete specification method is used (the one corresponding to models III and IV) with the following consideration. Only the second specification in each treatment uses interaction variables. These variables are assumed to have a different effect on the group of treated and non-treated. Thus, in the second specification of each treatment, a distinction emerges between the mean effect of treatment and the effect of treatment on those treated.
The results obtained are very similar to those presented in Tables 2 and 3, again bearing out the robustness of our results. In all cases, the Wald test rejects the null hypothesis of non-correlation between the errors of treatment and those of the outcome variable. Again, there is a greater positive and significant effect of treatment when those sharing the court are not replacement judges. The data concerning most of the variables are also repeated. The only different results to be seen is that the case duration has a positive and significant effect on the possibility that there might be replacement judges acting in the same court. The conclusions with regard to the effects of treatment are also similar although their size increases somewhat. Sharing a court with a replacement judge increases the percentage of cases ruled in favour of workers by 6.6 points (close to a 10% increase on the observed mean). However, if the court is shared with professional judges, the increase rises to 11% points (a 15% increase).
The third part of the robustness analysis repeats the estimates of models III and IV using the methodology proposed by Cerulli (2014) for the two types of treatment but with a different instrument. Instead of using the case duration indicator, we utilize the so-called congestion rate (Pastor and Manspons 2004), a well-known index in this sort of literature. That rate is calculated as a quotient with the sum of pending cases at the beginning of the period and incoming cases in the numerator and the number of resolved cases in the denominator. The estimations results are displayed in Table 7. The new instrument yields similar outcomes in terms both of the sign and the magnitude of the coefficient. The percentage of cases ruled in favour of the workers is increased by about 5% points when the court is shared with replacement judges and about 10% points when the court is shared with other type of judges.
Rights and permissions
About this article
Cite this article
Malo, M.Á., Martín-Román, Á. & Moral, A. “Peer effects” or “quasi-peer effects” in Spanish labour court rulings. Eur J Law Econ 45, 497–525 (2018). https://doi.org/10.1007/s10657-018-9576-9
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10657-018-9576-9