Reducing the Impact of Perjury In Domestic Violence Cases

by Jon Marin

Uncertain justice

Physical and sexual violence committed against children by adults in their household is a serious problem. So is actual violence committed by men and women against their intimate partners. The heightened official attention these problems have received in recent years has defused dangerous situations, assisted victims, and brought violent men and women to justice.

Unfortunately, official action taken upon false allegations of such criminal behavior have also caused great harm to many innocent children, women, and men.

The adjudication of accusations of domestic violence encounters systematic bias against the accused (perpetrator) that is not present, or less present, in ordinary criminal matters. The bias rarely if ever results from malice on the part of judges, but arises instead from the nature and context of the charges and the limited evidence often available in such cases. The key evidence is commonly limited to the testimony of the parties and other persons who may, and often do, have an undisclosed or under-disclosed interest in the outcome.

Asymmetric consequences of error

Humans are good at deception, poor at detecting deception.

The "he said /she said" nature of most domestic violence and abuse hearings causes the probability of erroneous outcomes to be unacceptably high. People begin to lie as little children, and continue to lie, more or less frequently, throughout their lives. They get very strong feedback on their proficiency - listeners tend to react unpleasantly when they catch someone lying. Though skill levels vary, most adults are pretty good liars when they have to be.

Humans are much less proficient at detecting deception.

People make subjective judgments of veracity all the time, of course, but the feedback is weak. Often, they simply never learn that they have been deceived. When they do, it is generally too late to reconstruct the details of the deceiver's voice, posture, and mannerisms at the time of the deception. Their memory of such details is no better when they learn they mistakenly judged someone to be deceptive.

This proficiency difference is reflected in the relative scarcity among attorneys of really good cross-examiners, the reverence in which they are held by their fellow attorneys, and the high fees they command. It is also reflected in the ongoing efforts to develop and prove novel lie detection methods, such as the analysis of stress in the voice, in minute short-lived contractions of the facial muscles, or in involuntary movements of the eyes.

The outcome of "he said/she said" cases is likely to be frequently wrong, even when the "playing field" is level in all respects. In the adjudication of domestic violence complaints, unfortunately, many factors combine to tilt the field severely.

The scales of justice are unbalanced

It is a basic principle of decision theory that a rational decision-maker should minimize the net cost of false positive and false negative errors. Application of this principle in domestic violence proceedings tilts them markedly in favor of the accuser. Judges are rational decision-makers by profession. They know that the real-world consequences of mistakenly finding in favor of a guilty man can be catastrophic, resulting in further acts of violence up to and including murder.

They know that such eventualities, though rare, attract the attention of the entire community when they occur. And they know that such a turn of events can damage their careers, cause painful lifelong self-recrimination, and further erode public confidence in the courts. Conversely, the harm from erroneously finding an innocent man or woman guilty - loss of income, reputation, and access to children - though substantial to the individual, is of modest or no consequence to the judicial system, and certainly less sensational. The likelihood of such an error coming to light is small. No one, including the judge, ever has reason to believe that a mistake was made. So wrongly-convicted, non- violent individuals go on their way, glue their life back together as best they can, and are seldom heard from again. The asymmetry creates a powerful incentive for judges, sitting as finder-of-fact, to find DV accusers to be truthful and render judgment accordingly. Though some, perhaps many, judges will have the strength to resist the incentive, none can be unaware of it. A heavy thumb is pressing down on one side of Lady Justice's scales.

Other factors accentuate the tilt:

Credibility: Motivation to lie

Judges, like everyone else, are naturally inclined to disbelieve witnesses who have a clear motive to lie. For the accused, the prospect of punishment constitutes a clear motive to lie. The accuser ("victim") takes the witness stand with a presumption of credibility which the accused ("perpetrator") must overcome. Even where the accused is able to do so, their motive to lie remains. Credibility becomes a wash.

Absence of effective appeal

The potential fallout facing a trial court for finding against an honest accuser pales in comparison to the fallout an appellate court could face for overturning a trial court's decision in favor of an honest accuser. Fortunately for appellate courts, their rules and traditions permit them to limit their review to errors in procedure and the law, but defer questions of fact to the finder of fact - the trial judge - enabling them to sidestep the danger. Applied to "he said / she said" DV cases, the limitation of review renders practically null the possibility of successful appeal. Where a trial judge has determined the accuser to be truthful, and found as a matter of fact that the accused committed or threatened acts of violence, little else matters. Unless a judge's errors have been absolutely egregious they will be dismissed on appeal as harmless error.

Unaccountable power and judicial immunity

Trial judges are human. They know that few DV findings are appealed, and that appeals rarely succeed. They know that their findings of fact will be decisive and that in their findings they are free to indulge not only the conflicts of interest due to the asymmetric consequences of error, but all manner of preconception, prejudice, and political inclination, all without fear of review. Some judges may have the strength to resist the temptations inherent in the situation; perhaps most do. There is no way to know who, or how many. What is certain is that the temptation is present, uncontrolled, and unchecked. Unaccountable power is abhorrent in a democracy. The risk is real, and reflects badly upon the entire judicial system.

Toward better justice

The article, He Said / She Said by Jonathan Marin proposes an approach that can sharply reduce the incidence and impact of perjured testimony in DV proceedings, with important benefits to society as a whole. The text on the web is a somewhat- extended edition of the version that appeared in Polygraph, Volume 29 Number 4 (2000) p. 299. The article describes how polygraph results when taken in combination in "he said/she said" situations, where one of a pair of opposed witnesses is almost certainly lying, can be reliably used to exclude untrustworthy testimony.

The underlying statistical concept is simple enough: If you roll one die, the chance of getting a six is 16.66%. Roll two dice, and the chance of getting two sixes is only 2.77%. The statistical gain applies to pairs of polygraph results viewed together. If the probability that either result alone will be wrong - whether false positive or false negative - is 14% (slightly less than that of getting a "six" on a single roll of a die) then outcome error rates of less than 4.0% can be confidently predicted. However, both of the polygraphists involved must use standardized, validated, techniques and, irrespective of other credentials, must have personally demonstrated an accuracy rate of at least 86% in a controlled protocol. Because the proposal advocates utilizing paired results solely to exclude untrustworthy testimony rather than admitting the results themselves into evidence, long standing precedent against the admissibility of polygraph results need not change.

The ideas are extended and detailed in a FAQ.

Applied to domestic violence and abuse cases, paired polygraph testing would: Reduce the number of groundless accusations. Reduce the incidence of meritorious suits stymied by perjury- dependent defenses. Reduce the load on the courts, thereby speeding justice for meritorious litigants. Reduce the incidence of tried cases that are decided incorrectly due to perjured testimony. Increase courts' willingness to penalize frivolous litigants and their attorneys. These benefits derive from the fact that litigants and their attorneys will understand that they have little hope of winning if their opponent's key witnesses will be allowed to testify unopposed about the dispositive facts in the case. The high costs of litigation provide a strong incentive against sustaining a case in the face of such odds.

Litigants who nevertheless persevered would risk being found frivolous by the court and burdened with their opponents' legal fees as well as their own. Rational accusers' ("victims") attorneys will advise their clients in their own best interest to abandon their case, and rational attorneys for the accused ("perpetrators") will advise them to acknowledge their guilt, when so indicated by the paired polygraph results.

Paired polygraph testing should not be applied to all testimony in dispute. It's use should be limited to witnesses where: The facts in dispute make it likely that the case will hinge on whom the finder of fact believes. The nature of the transaction makes it unlikely that either party could be honestly mistaken. The paired testing approach adapts well to the issue of post- adjudication relief. The accused would challenge the accuser to participate in paired testing, stipulating to incur substantial penalties if he failed and the accuser passed. They would both be examined. If the accused passed and the accuser failed, then appellate courts and other review authorities would be obliged to exclude that witness's testimony from the record on review - to treat the record as if that witness's testimony did not exist.

The Equal Justice Foundation can arrange paired polygraph testing in actual cases in conjunction with the Veritas Center, which maintains a roster of polygraphists who have rigorously established their personal proficiency at 86% accuracy or better. The Center assigns examinations randomly to polygraphists on its roster; a polygraphist who examines an attorney's client in one case may well be examining the opponent in a future one. The Center manages the flow of examination charts and videotapes, maintains an appeal procedure, and handles the reporting of results. To prevent even the appearance of impropriety, the Center collects fees and expenses before the polygraphists are assigned and later pays them; the polygraphists never have any financial contact with litigants or their attorneys.

Paired polygraph testing does not provide the absolute certainty that DNA often can. But it will help exonerate innocent persons accused of offenses that, by their nature, seldom produce forensic evidence. Exclusion based upon paired- test results, under the tight constraints proposed is a sensible and realistically achievable approach that will reduce the frequency of abuses and false allegations that so strain the process of adjudicating domestic violence.

* * *