Troika ’ s Portuguese Ministry of Justice Experiment , Part II : Continued Positive Results for Civil Enforcement Actions in Troika ' s Aftermath

This article constitutes part II of a series of studies on the impact of measures implemented by the Portuguese Ministry of Justice. Particularly, it addresses the results at the level of civil enforcement actions resulting from goals inscribed in the MoU signed between Portugal and the so called Troika (IMF/EC/ECB). The empirical study is now expanded to encompass the quantitative analysis of results achieved not only during the Troika’s period but during the post-Troika period as well. Results show a continued positive effect for civil enforcement actions in Troika's aftermath.


Introduction
In the 21 st century, the rule of law model has struggled to assimilate and accompany the vertiginous, simultaneous evolution of societies, states, companies and citizens all over the world.Judicial systems appear to have become victims of an important economic problem pointed out, several decades ago, by intellectual heavyweights like Barnard and Hayek: the inability to rapidly adapt to changing circumstances 2 .millennium.This state of affairs led to high levels of financial obligation defaults and, in turn, to big increases in litigation associated with debt collection, including massive surges in the numbers of civil enforcement actions processed by the judicial system 5 .
The judicial system composition was noticeably complex and, many times, even counterproductive and burdensome.It was not equipped to deal with such an increase in demand, particularly of civil enforcement actions.It needed reform.
The reform of the Portuguese justice system took as one of its main priorities the substantial reduction of the civil enforcement actions backlog.To achieve that goal, a series of measures 6 ensued, including: • Improvements to the role of enforcement agents; • Enhancement of the supervision and disciplinary powers of the regulatory body; • Judicial map restructuring; • Improvements to the alternative dispute resolution system; • Improvements and expediting of civil procedures 7 (comprising the elimination of useless formalisms).
Progressively, after an implementation period that spanned approximately one year and six months, a decline in the number of pending civil enforcement actions became apparent.These results were regarded as particularly meaningful as they brought to an end more than 20 years of continual pending increases.
More than two years after the end of the financial and economic assistance programme and after overcoming the most critical phase, one question remains: has the Portuguese judicial system taken advantage of the rare contextual, political and social, window of opportunity in order sustain the improvements obtained during the Troika's years?
The approach used in this paper closely follows the one presented by Correia and Videira (2015).Despite the mounting, noteworthy literature on the subject no particular theoretical perspective was favoured by the authors in order to encourage readers to abstain from any type of theoretical framework 8 lure.
Therefore, it is important to stress that the purpose of this article is to circulate and make accessible a theoretically neutral and statistically sound analysis of the Troika's Portuguese Ministry of Justice experiment aftermath (giving particular emphasis to the period subsequent to the adjustment program).By doing so, future theoretical and empirical work can be inspired in an unbiased manner.

Methodology
Empirically, this study analyses civil enforcement actions in Portugal's first-instance courts from the perspective of behavioural temporal evolution of this case type, based on a 111-month sample spanning from January 2007 to March 2016 (a 33-month, 42.3% increase from the 78-month period used by Correia and Videira (2015)).53 of these 111 months, from January 2007 to May 2011, precede Troika's arrival in Portugal (referred to from now on as pre-Troika's period); the next 34 months, from June 2011 to March 2014, postdate Troika's arrival and predate Troika's departure (referred to from now on as Troika's period); the remaining 24 months, from April 2014 to March 2016, postdate Troika's departure from the country (referred to from now on as post-Troika's period).The numbers of new, completed 9 and pending 10 enforcement actions in Portuguese first-instance courts comprise the raw data used in this research paper 11 .
5 Directorate General for Justice Policy (2015Policy ( , 2016)). 6For a detailed account and description of the measures, see Correia and Videira (2015). 7The new Civil Procedure Code (CPC) entered into force on September 2013 (Portugal, 2013b).It is relevant to highlight a particular legislative measures, the Decree-Law 4/2013, of 11 January (Portugal, 2013a), that approved a set of provisional actions to fight the backlog pendency and was later absorbed into the new Civil Procedure Code. 8Such as the new public management perspective (Lane 2000;Gomes, 2007;Frederickson et al., 2012;Hill and Hupe, 2014), the theories of judicial governance perspective (Frederickson et al., 2012;Guimarães et al., 2015), the permanent notion of crisis in the judicial systems perspective (Campbell, 2013), the erosion of judicial legitimacy and concern for power separation perspective (Stephenson, 2004;Langbroek, 2008), the human resources productivity and demand pressure perspective (Walsh, 2008), or the macro-model approach of the judicial system perspective (Bell, 2006;Ambach and Rackwitz, 2013). 9A precise definition of completed case if provided by the Directorate General for Justice Policy in several of its official data releases (see, for instance, Directorate General for Justice Policy (2014)).For an accurate translation of the concept, see Correia and Videira (2015). 10A precise definition of pending case if provided by the Directorate General for Justice Policy in several of its official data releases (see, for instance, Directorate General for Justice Policy (2015Policy ( , 2016))).For an accurate translation of the concept, see Correia and Videira (2015).
This data (available at http://www.siej.dgpj.mj.pt) is released to the public by the Directorate General for Justice Policy, the government agency responsible for statistical data for the justice sector.
The three original variables were complemented by three additional compounded indicators: procedural balance12 , clearance rate13 and disposition time14 .The calculation formulas of these indicators can be found, for instance, in Correia and Videira (2015): In the absence of Gaussian distributions for the monthly data in each category (pre-Troika's, Troika's and post-Trioka's groups of months) the application of the ANOVA parametric test was abandoned15 in favour of the non-parametric test of Kruskal-Wallis16 to determine whether: H0: The pre-Troika's period, Troika's period and post-Troika's period datasets have equal medians.H1: The pre-Troika's period, Troika's period and post-Troika's period datasets do not have equal medians17 .

Results
This section is divided into two distinct sub-sections.Sub-section (a) presents a descriptive statistical account that establishes a robust body of evidence for the prevalence of positive results not only in the Troika's period but also in the post-Troika's period.Sub-section (b), in turn, makes use of statistical tests to obtain unbiased confirmation of the results suggested by the descriptive statistics analysis.Due to the customary judicial vacation period, with particular emphasis on the months of August, seasonality is an inherent feature of the data presented on figure 1.This characteristic seasonality, more intense at the level of the resolved cases, is particularly visible in the 2007-2012 period.Source: prepared by the authors.

Descriptive Statistics -Evidence Accumulation Continues
In order to reduce these effects and obtain a less biased reading, the data was adjusted to compensate for the seasonality effects 18 .Figure 2   The seasonality adjusted procedural balance is, for the period in question, plotted in figure 3. The change in trends reported by Correia and Videira (2015), initiated about 1,5 years after the start of the adjustment program, continues to be well visible in the post-Troika's period.Of the 111 months considered in the analysis, 49 presented favorable procedural balances (number of completed cases higher than the number of incoming cases, resulting in a decrease in the pendency equivalent to that same balance).Of those 49 months, 7 (or 14.3%) were recorded in the pre-Troika's period, 19 (or 38.8%) were recorded in the Troika's period and 23 (or 46.9%) were recorded in the post-Troika's period.More noteworthy, however, is the fact that in the 53 months comprising the pre-Troika's period only 13.2% (7 months) presented favorable procedural balances, in the 34 months covering the Troika's period 55.9% (19 months) presented favorable procedural balances and in the 24 months spanning the post-Troika's period, a staggering 95.8% (23 months) presented favorable procedural balances.The seasonality adjusted clearance rate is, in turn and for the same period, shown in figure 4  Source: prepared by the authors.White bars represent favorable clearance rates (above 100%) and dark bars represent unfavorable clearance rates (bellow 100%).
In turn, for the period under analysis, figure 5 displays the seasonally-adjusted disposition time.As for the previously shown indicators, a change in trend is observable particularly from January 2013 forward.Of the 111 months considered in the analysis, 39 presented disposition times under 1,500 days22 .Of those 39 months, 8 (or 20.5%) were recorded in the pre-Troika's period, 18 (or 46,2%) were recorded in the Troika's period and 13 (or 33.3%) were recorded in the post-Troika's period.However, more noteworthy is the fact that in the 53 months comprising the pre-Troika's period only 15.1% (8 months) presented disposition times under 1,500 days, in the 34 months covering the Troika's period 52.9% (18 months) presented disposition times under 1,500 days and in the 24 months spanning the post-Troika's period 54.2% (13 months) presented disposition times under 1,500 days.Figure 6 presents the evolution of the number of pending civil enforcement actions, from January 2007 to March 2016, which is a consequence of the previously presented results.A detailed analysis of figures 2 through 6 suggests that the stabilization and initial decline of the pending cases during the Troika's period, and its subsequent continued decline in the post-Troika's period, were not achieved by chance or by random fluctuations, but rather by Portugal's measured implementation in the context of the MoU objectives set by the adjustment program.If one considers the maximum reached in August 2012 as a reference point, by March 2016 a reduction of more than 360,000 pending cases had been achieved (including a reduction of nearly 160,000 pending cases in the first 24 months of the post-Troika's period).The measures, even after Troika's departure from Portugal, appear to continue to be effective.Source: prepared by the authors.
The data poses an interesting and relevant question: do the values of the above examined indicators (incoming, completed and pending cases; procedural balance; clearance rate; and disposition time) for the three considered periods (pre-Troika's, Troika's and post-Troika's), have different statistical properties?If so, those differences cannot be merely attributed to the common fluctuations of the studied phenomena and should be regarded as consequences of the Portuguese judicial system's continued resolve to tackle the challenges identified in the 2011 MoU.

Statistical Evidence -Once More
Due to the absence of Gaussian distributions for the various data groups, the Kruskal and Wallis test (1952) was applied in order to determine whether the three sets of data come from the same population (H0) or if, alternatively, they originate from distinct populations (H1).Table 1 presents the results of the Kruskal-Wallis test for the six variables included in this study.In all the considered variables H0 is rejected (p-values=0.000<0.05),and H1 is validated, i.e., the three data sets do not come from the same population.
The logical question arising from the Kruskal-Wallis test results is: if the three data sets do not come from the same population, can their outcomes (for each of the six indicators) all be considered statistically different or, instead, can they be statistically grouped?
In order to answer that question, table 2 presents a stepwise comparison23 for statistical similarity of medians for the pre-Troika's period, the Troika's period and the post-Troika's period24 .Understanding the contents of table 2 complements and enhances the interesting outcomes reported in previous studies 25 .On the one hand, regarding the median of incoming actions adjusted for seasonality, it is possible to conclude that there is no statistically significant difference over the pre-Troika's and Troika's periods, but the difference of roughly 7000 incoming cases between these two periods and the post-Troika's period is statically significant and signals a decrease in society's demand for the resolution of this type of cases (median of 22028 incoming cases per month before the arrival of Troika to Portugal, of 22314 incoming cases during Troika's stay in Portugal and of 15037 incoming cases per month after the departure of Troika from Portugal).
such as procedural balance, clearance rate or disposition time; and the number of pending civil enforcement actions had sharply declined 29 .
The next logical and essential step in this process was, to understand if the implemented public policies would produce only short term results or, alternatively, if those positive results would endure.
After Troika's departure, in the period comprised from March 2014 to March 2016, the statistical answer is clear: society's demand for civil enforcement actions has decreased (probably as a reflex of the improving socio-economic and financial environment); the judicial system's supply for such actions has sustained most of the improvements achieved during the Troikas's period; and the number of pending civil enforcement actions continues to sharply decline.All these results converge on improved values for indicators such as procedural balance, clearance rate or disposition time (note that those indicators improved not only when compared to the pre-Troika's period but when compared to the Troika's period as well).
Therefore, it is acceptable to conclude, on the basis of the empirical study hereby presented, that the noteworthy Portuguese civil enforcement actions public policy implementation success story, reported by Correia and Videira (2015), was not confined to the Troika's period but extends well beyond it, with statistically significant results observable 24 months after Troika's departure from the country.
We leave to others the task of analyzing to what measure and extent these results constitute evidence of a successful IMF, EC and ECB strategy toward countries that beneficiated from financial assistance 30 .
Future studies should consider following two distinct paths.The first consists of carrying out similar empirical studies on other types of actions specifically targeted in the MoU (for instance, bankruptcy, insolvency and corporate recovery actions, or special revitalization procedures).This will permit a more comprehensive assessment of the tangible accomplishments of the adjustment program for the Portuguese judicial system.The second is to continue the empirical monitoring studies of the results at the level of civil enforcement actions resulting from goals inscribed in the MoU.This will allow for the creation of a rich detailed history that will make it possible to determine whether the successful story resulting from the experiment carried out in the justice sector in Portugal is limited to a well-defined time span or, instead, if its fruitful results will be carried out into the mid-term and long term future.

Figure 1
displays the evolution, between January 2007 and March 2016, of the number of civil enforcement actions (incoming and completed) in the Portuguese courts.

Figure 1 -
Figure 1 -Incoming and completed civil enforcement actions, between January 2007 and March 2016 presents the monthly numbers of incoming and completed civil enforcement actions, between January 2007 and March 2016, adjusted to compensate for seasonality.The calculations of the procedural balance, clearance rate and disposition time (figures 3 to 5) are based on the values of incoming and completed cases displayed in figure 2 19 .

Figure 2 -
Figure 2 -Incoming and completed civil enforcement actions, adjusted according to the seasonality, between January 2007 and March 2016

Figure 3 -
Figure 3 -Procedural balance for the civil enforcement actions, adjusted according to seasonality, between January 2007 and March 2016

Figure 5 -
Figure 5 -Civil enforcement actions disposition time, adjusted according to the seasonality, January 2007 to March 2016

Figure
Figure 6 -Pending civil enforcement actions, January 2007 to March 2016

Civil enforcement actions clearance rate, adjusted according to the seasonality, January 2007 to March 2016
20.As for the seasonality adjusted procedural balance, starting in mid-2012, a trend change is noticeable becoming unequivocal from 2013 onward and extending well beyond the end of the adjustment program.Of the 111 months considered in the analysis, 49 presented favorable clearance rates (i.e., above 100%, resulting in a decrease in pendency).Of those 49 months, 7 (or 14.3%) were recorded in the pre-Troika's period, 19 (or 38.8%) were recorded in the Troika's period and 23 (or 4.,9%) were recorded in the post-Troika's period.Similarly, more noteworthy is the fact that in the 53 months comprising the pre-Troika's period only 13.2% (7 months) presented favorable clearance rates, in the 34 months covering the Troika's period 55.9% (19 months) presented favorable clearance rates and in the 24 months spanning the post-Troika's period, a staggering 95.8% (23 months) presented favorable clearance rates21.Figure 4 -