PROJECT SPOTLIGHT: REPORT NO. C-06-QSC-1

INTRODUCTION

Project Spotlight is a research and analysis programme conducted by World Fathers Union, a Canadian-based, all-volunteer NGO serving the interests of children and fathers in all parts of the world. The intent of the programme is to validate (or disprove, as appropriate) anecdotal reports of bias in the decisions rendered by specific family-court judges in child custody cases brought before them.

While it is generally accepted that there is a strong but unspoken bias in the family courts against fathers who seek custody of their children, and while it is evident from existing government statistics that custody is granted to mothers in the great majority of cases, it is less clear as to whether the stories of individual judges being especially biased against fathers have any basis in fact, or whether such allegations are simply the result of frustration felt by men who believe themselves and their children denied a fair hearing. The Project Spotlight programme was conceived to answer this question in a rigorous manner, using only facts drawn from the public record as a basis for the studies.

HISTORY AND DISCUSSION

Placing the Family Court in Perspective

Family court judges in virtually all democratic nations operate in a judicial environment which is free of the restrictions placed upon judges by the checks and balances of the criminal rules of evidence and procedure. Whether through accumulated jurisprudence or specific statute, family court judges have become endowed over time with an almost unlimited amount of what is known as 'discretionary authority'--the latitude to decide for themselves not only what testimony they wish to hear, but also what kind of material or testimony constitutes evidence.

Most importantly, they are free to interpret any or all of this material in the manner they see fit, and are not, in any practical sense, bound by statute to give weight or credence to any particular testimony. As all students of law know, equally intelligent and conscientious justices often come down on opposite sides of the same question and justify their differing opinions by a different selection or application of supporting citations. The differing arguments advanced by honest jurists merely reflect the fact that our systems of law are themselves a reflection of the imperfect societies and people who have created them. Law is not a science; it is an art--and it is clear that one artist may paint an entirely differerent image of a model than another, though both use the same colours and brushes and canvas. Each artist sees things through his own set of personal 'filters' developed as a result of his preferences, education, and life experiences.

Thus it is with judges. Whilst in theory all judges are impartial and render decisions free of any biases their personal beliefs may bring to the questions placed before them, in reality they are but human and as subject to human frailty as any man. The best of them recognise this and try valiantly to compensate for the good of Justice; the worst of them profit by the awful power entrusted to them and pursue their personal agendas as a way to modify the world to their liking. It is for this reason strict limits on trial judges in criminal cases are an essential element of the legal systems of democracies. It is far easier to legislate impartially in the abstract than it is to judge impartially in the face of real, live, human agony. Thus, in the criminal venue, the legislator constrains the jurist to rule based on fact and not on opinion or supposition, and to demand that all allegations be demonstrated beyond a reasonable doubt.

Family courts, however, are a world apart. Relying on the argument (or statutory directive, in some jurisdictions [1]) that all decisions by a family court judge must be taken solely in consideration of the best interests of the child, these judges find themselves given virtual carte blanche, because the determination of what constitutes any particular child's best interests rests primarily with them. In that the view of the child as innocent victim of his parents' errors is difficult to challenge, the concept of resolving the inherent conflict of rights amongst the parties becomes moot in any practical sense and only the very hardiest of legal theorists dare to enter upon such dangerous terrain.

Yet even without addressing this problem, it can be seen that such a vague and undefinable concept as the 'best interests of the child' may easily lend itself to abuse. If a judge were to believe that all boy children should be raised by their fathers and all girl children by their mothers, given his discretionary authority to interpret testimony he could choose to lend credence to only such testimony as would support his point of view. Any testimony tending to contradict his bias could be ignored as insufficiently credible, and any difficulties encountered in so doing could be justified by claiming the necessity of using a 'conservative approach' to protecting the child's best interests.

Examining existing avenues of redress

What deepens the problem is the manner in which justice systems are structured. Democratic nations pride themselves on having systems of justice with safeguards against error built into them. Thus, in theory, an unjust or unreasonable decision by a court may be appealed to a higher court, which has the authority to reverse or nullify the decision. In practise, however, two impediments often stand in the way of such a course of action.

The first impediment is the very prosaïc and mundane matter of finances.

Prosecuting an appeal is an extremely costly venture. Lawyers truly competent to advance the appelant's cause do not often work on a pro-bono basis, and the additional costs of the transcripts and other expenses inherent in preparing and filing the appeal are substantial. The initial costs for appealing a custody decision from a provincial family court in Canada are rarely less than $15,000, and can easily exceed $25,000. To carry an appeal all the way to the Supreme Court places the upper limit of costs in the high six-figures bracket. This places such a proceedure out of reach of all but the very poor and the very well-to-do portion of the upper middle class. The poor who rely on legal aid attorneys rarely find themselves represented by the best legal talent available; the upper-middle-class appelants frequently find it necessary to sell or re-mortgage homes, empty life savings accounts, and beg money from their own parents and siblings in order to fuel the voracious legal process.

Thus, a very small percentage of cases are ever appealed, because the would-be appellant simply cannot afford to do so. It is therefore impossible to draw any conclusions as to the soundness of judgements from the courts of the first instance by looking solely at the numbers of cases appealed.

The second impediment to appeal is the technical question of grounds.

The fact that one party is disappointed by the outcome does not mean that a decision is improper. In resolving a contested question, it is rare indeed that a court or other tribunal can do so in a manner which leaves both parties to the conflict satisfied. The law therefore sets practical limits as to what sort of errors may give rise to an appeal. These are called grounds.

In general terms, only errors of law or errors of fact are considered grounds for an appeal. Errors of law are those in which a judge misapplied or ignored statute. Errors of fact are those wherein he is shown to have made a factual mistake, such as stating in his decision that the respondent committed an act which the record clearly shows the respondent did not commit. The category of errors of fact generally does not include questions as to a judge's interpretation of testimony (which, in family court cases, consists primarily of opinion and is thus subject to interpretation in a way that forensic evidence is not).

Appelate courts are extremely reluctant to comply when asked to review the testimony heard by the judge of the first instance. The standard usually set is that of 'patent unreasonableness'--which is much stricter than the more familiar standard used in criminal law of 'beyond a reasonable doubt.' Unless an appeals court finds that the original judge was patently or grossly unreasonable in his interpretation of the testimony presented to him, it will not review that testimony as to its credibility or lack thereof. The decision of the judge of the first instance will therefore be let stand, even if there is a reasonable doubt that he may have interpreted the testimony in a questionable or incorrect manner.

Again, but for different reasons, we see that no conclusions can be drawn as to the overall soundness of family court decisions based solely on the numbers of appeals filed.

In that the mechanism contemplated by the legislator for redress of error in judicial rulings is not accessible for an unknown number of litigants who may have a legitimate need for such redress, another mechanism--that of public scrutiny and comment--must take its place in those instances.

In conclusion, we find that both a necessity and justification exist for conducting the research programme.

RESEARCH PHILOSOPHY AND METHODOLOGY

A research philosophy of complete transparency has been chosen. Studies to be undertaken are identified by public comment, data is collected from public sources, and decisions by individual judges are compared against decisions of other judges serving in the same arena. Researchers remain anonymous so that no pressure can be brought to bear upon them. Duplication of this and other studies is encouraged.

Study Selection: Individual studies in the Project Spotlight programme are commissioned following receipt of anecdotal reports of family court judges who are alleged to have an overriding and evident bias against fathers seeking custody of their children. Following receipt of several similar complaints about a particular judge, an investigative (as opposed to analytic) overview of a number of that judge's decisions is undertaken. If this overview tends to confirm that the allegations may have some substance to them beyond the personal dissatisfaction of the attorneys or fathers making the complaints, a full study is commissioned and a team of volunteer researchers is assembled.

Parameters of Study: The basis for all the studies is a comparison of an individual judge's record of custody awards to the overall record of custody awards rendered by the court in which he sits. The subject judge's bias or lack thereof is thus determined as a function of his or her own performance viewed against that of brother and sister judges, all of whom are operating under the same rules and statutes, within the same social context, and having reference to the same 'local standards' as the subject. This is an important distinction: Judges are evaluated for bias only within their own judicial and social environment. While it might be useful in another context to compare Canadian or British judges' records to those of New Zealand or the United States, such a comparison would not say anything about the bias held by any particular judge. Only a comparison to judges working under the same conditions can testify as to that.

Data Selection: The list of decisions to be tabulated is derived from a key-word search performed upon the public data-base for the court in question. Two keywords are used to return the data studies. They are: the name of the subject judge; the word 'custody' [2]. The ranking of pertinence assigned by the court or justice department's own search engine is accepted as unbiased. Elimination of cases not applicable the study is made by examination.[3]

Study Protocol: Each member of the research team is randomly assigned a number of decisions to read and is provided with a spread-sheet document in which to enter data. Researchers record the following information:

These data are extracted from the public record of the judges' decisions and entered by the researchers into the spreadsheet for tabulation.

Once all the data are collected, they are then tabulated and sorted alphabetically by the names of the presiding judges. The keyword search contains the name of the subject judge but not those of any other judges on the court; thus, failing an anomolous distribution of divorce cases within the court roster for the period studied, a much larger sub-sample (as much as an order of magnitude in some instances) is generally returned for the subject judge than for any other individual judge. To control against any such case distribution anomolies, note is taken of any other individual judge represented in the total sample by an excessive number of cases and, if necessary to avoid unduly weighting the results for the whole court, an appropriate number of that judge's cases would be eliminated by random selection[5].

Researchers are provided with a field to enter comments on the issues raised in each case, but this information is not used to perform the analysis, only to validate the premise that child custody cases constitute a class, or type, amenable to binary (won/lost) analysis. For custody cases, it is found that a short list of common issues is present in virtually all of them, and that a predictable selection from another short list of issues is present in most. Thus the individual cases in any sample generated by an unbiased search-engine from the available data base will be similar to all other custody cases in the data base when examined together.

Finally, straight percentages as to the four types of custody awards possible (sole custody to the mother; sole custody to the father; joint custody as demanded by the mother; joint custody as demanded by the father) are then calculated separately for both the subject judge and for the entire court. Figures for the whole court are calculated in two ways: including the decisions of the subject judge, and excluding those decisions. The second calculation is done so that the subject judge's rulings, if biased, would not introduce that bias into the results for the whole court.

RESULTS

In the present study, 742 cases were returned by the keyword search.[6] Of these, the 200 most pertinent (as ranked by Justice Quebec's own search engine) were examined. After elimination of civil cases not applicable to this study[7], 112 decisions were read by the researchers[8]. The subject judge's decisions comprised 13.4% of the total sample before reading[9]. The period studied was approximately five years, with the earliest decisions in the sample being from 2001, and the most recent from 2006.

There were 59 judges in the total sample after elimination of non-applicable civil cases by examination. Two were eliminated after reading as explained in note 7. The cases were heard in all regions of the province; a regional distribution was not drawn from the data as it would not inform the question. In the province of Quebec, many Superior Court judges sit at more than one courthouse, and any biases they might have would thus be portable and not regionally dependent.

For the court as a whole

It was found that the court as a whole ruled in favour of fathers in 38.0% of cases (n=100), including sole custody awards to the father as well as joint custody awards at the father's demand. This result includes the subject judge's decisions as part of the total sample. When his decisions are removed from the sample, the court ruled in favour of fathers in 40.7% of the cases (n=91), a difference of 2.7 percentage points.

When broken down by sole custody and joint custody, the figures for the entire court showed that fathers were granted sole custody in 24.0% of the cases and were granted joint custody at their demand in 14.0% of the cases. With the subject judge's decisions removed from the sample, the awards to fathers for sole custody represented 25.3% and awards of joint custody at the father's demand represented 15.4% of the sample.

Awards of custody in favour of the mother were much higher, as was expected. In 58.0% of cases in the total sample, sole custody was awarded by the court as a whole to the mother. Another 4.0% of the decisions granted joint custody to the mother at her request. With the subject judge's decisions removed from the sample, mothers received sole custody in 54.9% of the cases decided by the 56 other judges, and 4.4% of the decisions awarded the mother joint custody at her request. Overall, mothers received custody (either sole or joint) in 62.0% of the total cases, and in 59.3% of the cases when the subject judge's decisions were removed from the calculation.

For the Subject Judge

The subject judge awarded custody (sole) to the father in only 11.1% of cases studied. He awarded sole custody to the mother in 88.9% of the cases.

As for joint custody demands, the subject judge awarded joint custody in 0.0% of the sample cases he heard.

CONCLUSIONS

Based on the results of this study, the research team finds that the anecdotal allegations of evident and overriding bias against fathers in the rulings of the subject judge are amply justified by the facts shown in the public record. In a court which renders decisions in favour of fathers in almost forty percent of the cases it hears, a judge whose record reveals that he does the same in only eleven percent of the cases identifies him as a statistical outlier, performing radically off the performance curve established by his peers working under identical conditions. There is thus sufficient cause to warrant a full investigation into this judge's conduct by the appropriate authorities; in this case, the Minister of Justice of Quebec.

The subject judge in this study is Justice André Wery; he is currently serving as Associate Chief Justice of the Quebec Superior Court. The names of the other judges whose rulings constituted the sample are given in Appendix A. All rulings were from the Superior Court of Quebec, Family Division.

RECOMMENDATIONS

It must be stated that in the reading of the material necessary for the completion of this study, most of the researchers remarked that Judge Wery delivered long and intricately reasoned decisions. It was also noted that his points of law were cited with some regularity by other judges to support their own reasoning (even when they were ordering an award of sole or joint custody to a father). The scholarly nature of Judge Wery's legal efforts is therefore not called into question.

What is needed is an examination into why this judge consistently rules in a way contrary to the usual practise of his peers, and that, unfortunately, is beyond the mandate and resources of the research team. Psychological profiling of the subject based upon the public biographical information about him might suggest some answers, but lacking hard evidence to support it, any such surmise would be non-rigorous, superficial, and prejudicial to the subject. The research team therefore refrains from advancing any theories as to the subject's reasons.

Once surmise is set aside, what remains are the facts as determined and presented by this study, and the evident and immediate need for this judge to be re-assigned off child custody cases unless or until a valid explanation for his abnormal performance can be found.

The recommendation is therefore made that, in the interests of preserving and protecting the good administration of justice, the subject judge should recuse himself from all present and future family court proceedings. Failing his willingness to do that, it is recommended that the Minister of Justice remove him from such duties immediately, and notify litigants who have appeared before Judge Wery in custody disputes that an application for rehearing would be looked upon favourably if properly constituted and well-founded.

A further recommendation is made that the resources of the Ministry of Justice be used to investigate the causes for this judge's aberrant custody decision record, so that a reasoned decision as to his future in the justice system can be made. The research team does not presume to comment upon Judge Wery's competence in other judicial venues, and does not recommend ipso facto his removal from the bench. That decision is best left to the competent authorities after they have completed their investigation.

Published, for the public good and without prejudice,

November 22, 2006

WORLD FATHERS UNION, Yarmouth, Nova Scotia (CAN)

Notes

[1] See, for instance, Magill v Magill [2006] HCA 51 at ¶246

[2] In the present study, the majority of the decisions are rendered in French, thus the word 'garde' was also included.

[3] In addition to divorce and custody cases, the initial search returns a case list which generally contains a number of ordinary civil cases (not from the Family Court) which are not applicable to the study. Most of these are returned by the search engine because the keywords are found in the text in a non-family context ("custody", for instance, can refer to posession of a chattel), or because the subject judge's name appeared in footnotes or citations. These cases are eliminated by cursory examination prior to reading.

[4] Joint custody awards are attributed as Joint-Mother or Joint-Father, according to which party demands this type of custody. A cross-demand for joint custody by the father is frequently opposed to a demand for sole custody by the mother. It is less frequent that the mother makes a cross-demand for joint custody. In rare cases, joint custody is ordered by the Court in the face of opposing demands for sole custody; this is treated as a dual decision (see note 8).

[5] In the present study, it was not found necessary to perform this control procedure as no individual judge was represented by an excessive number of cases. One judge was represented by 5 cases. As a check, calculations were performed for that judge's decisions, and it was found that the percentage of awards granted fathers or mothers by this judge concorded closely with the percentages shown for the court as a whole. A second check was made by re-calculating the whole-court percentages without those five decisions, and an even closer agreement with the original figures was noted. The distribution for multiple cases per judge in the validated sample was (No. of Cases=No. of Judges): 5=1; 4=2; 3=5; 2=15. All other judges (33) were represented in the sample by one case. (The subject judge's decisions are not included in the distribution.)

[6] The keyword search was performed in April of 2006. Because the total number of cases in the data base increases regularly as new decisions are rendered and older ones are digitised, an identical keyword search peformed at a later date would not necessarily turn up the same number of results or exactly the same case list.

[7] A number of applicable (i.e.: divorce) cases (2) were eliminated after initial examination because upon reading they were found to be procedural hearings or corrected judgements wherein no new decision on custody was rendered. An additional three divorce cases were eliminated because although custody was attributed in the order, it was not adjudicated. Seven more divorce cases were eliminated because custody was not being contested by the parties. The resulting case list is designated the validated sample.

[8] The number of individual cases read was 107. Of those, five cases contained two different custody awards within the same decision for various reasons (different children, dates, conditions, etc.). Each separate custody award required individual attribution in the tabulation, thus those cases were entered twice in the table and one of the two awards was attributed to each entry. This resulted in a count of 112 decisions in 107 cases.

[9] After reading, a number of the subject judge's cases were found to fall into one of the categories given in note [7], and they were removed from the case list. Validated cases for the subject judge made up 9% of the validated sample for the entire court.

APPENDIX A

The judges whose decisions were read in conducting this research are listed below in alphabetical order.