Predicting Dangerous Acts: Specialists Find It a Hazardous Pursuit

Your neighbor gives you a funny smile as she clips her hedge.

You look up from your Big Mac and notice a fellow toting a pistol, a shotgun and a semiautomatic rifle.

In each of those encounters, you might catch yourself wondering whether the person in question is dangerous. If you’re lucky, someone may already have made that determination. If you’re very lucky, the person doing the determining did a good job.

Predicting violent behavior is becoming increasingly important and increasingly controversial as society becomes increasingly violent, according to specialists on psychology and the law.

About 60 psychiatrists, psychologists, marriage counselors, nurses, attorneys and other professionals gathered Saturday at UC Irvine to hear two panel discussions of “Dangerousness: Legal and Clinical Aspects.” The group got a detailed look at its members’ complex legal and professional responsibilities and the clear message that how they and their colleagues handle those responsibilities concerns everyone.

As attorney Byron Chell, former chief counsel to the state Department of Mental Health, explained, the legal and ethical quagmire that therapists face dates back to a 1976 California Supreme Court ruling. Known simply as “Tarasoff,” that ruling offers insight not only into the dilemma of therapists but also into society’s growing concern with what had traditionally been deemed the sacred domain of client-therapist confidentiality, speakers explained.

According to various accounts of that case, in 1969 Prosenjit Poddar, who had voluntarily sought outpatient therapy at UC Berkeley’s Cowell Memorial Hospital, told a psychologist about violent fantasies he had concerning Tatiana Tarasoff, a young woman with whom he had become obsessed. Poddar’s therapist consulted with a psychiatrist on the hospital staff, and they decided that Poddar should be hospitalized under California’s new civil commitment statute.

The psychologist wrote to the university police, and officers interviewed Poddar at his apartment. But instead of committing him, they left him with a warning to steer clear of Tarasoff. Two months later, Poddar stabbed the young woman to death.

“You can imagine how the family felt,” Chell told the group. “Suppose this is your daughter. . . . (Poddar) went to see therapists and let them know he was going to kill your daughter . . . then the police let him go.”

The family reacted with a lawsuit charging the therapists and the campus police with negligence. University attorneys, in turn, claimed that neither party in the case had a duty to protect or warn Tarasoff. The Supreme Court disagreed and established, for the first time, that therapists had a legal duty to warn an intended victim.

Panelist Robert Sullivan, counsel to the California State Psychological Assn., said he understood the pressure that therapists sometimes feel. “My father was a psychiatrist, and I remember he’d have white knuckles and beads of sweat at night when he’d get calls from people threatening suicide or threatening to kill someone else.”

The pressure on therapists increased with the Tarasoff decision, and a series of related cases have put them in more of a bind, Sullivan said. In one of those decisions--the Hedlund case--therapists’ legal duty was broadened to include liability not only for the person actually threatened but for other people injured in an attack on the threatened person. In another ruling--the Jablonski case--therapists were found to have a duty to predict or diagnose violent behavior, even though no specific threat was made. In the spring, 1984 issue of California Health Law News, Chell wrote: “It can be argued that Jablonski has quickly taken the ‘duty to warn’ from the land of therapeutic reality to the wonderland of clairvoyance.”

Relief Expressed

According to Sullivan, though, a recent statute may signal a change in that tide of decisions. Many of the therapists gathered expressed relief that their responsibilities seem to have been more clearly defined by California Assemblyman Alister McAlister’s Assembly Bill 1133--which was passed last year after strong lobbying by the California Psychiatric Assn. But that law, which requires a therapist to “make reasonable efforts to communicate the threat to the victim or victims and to a law enforcement agency” and provides that a therapist is immune from liability “except where the patient has communicated to the psychotherapist a serious threat of violence against a reasonably identifiable victim,” is open to broad legal interpretation, panelists said.

Professionals at the daylong seminar ranged from a Long Beach radio psychologist concerned about his liability for advice given over the air, to sheriff’s deputies, psychiatric social workers and a Riverside attorney who believed her client had been falsely committed for voicing a vague and meaningless fantasy to his therapist.

The group’s general mood of self-sympathy was reflected in its reaction to a story told at the afternoon seminar by John Monahan, a psychologist and author whom moderator Stephen Wells introduced as “by far the foremost psychologist in America studying questions of the prediction of dangerousness.”

Monahan recalled the day that Dr. John J. Hopper--who had been John W. Hinckley’s psychiatrist in Colorado--was called to the stand and asked why he hadn’t predicted his patient’s attempt to assassinate President Reagan.

Heartfelt ‘Awwww’

“The next day a Boston newspaper had as its full-page headline, with a picture of Dr. Hopper walking to court with his head bowed down: ‘Hinck’s Shrink Stinks,’ “Monahan said. The audience did not giggle or titter but instead let out a sustained and clearly heartfelt “Awwww.”

“I’d hate to be the next young, depressed male going in to see Dr. Hopper,” Monahan said.

Monahan argued that clinicians should let the courts and lawmakers decide what to do about criminal violence whenever possible. He also noted, though, that for a variety of reasons, society and the courts are “pressuring” therapists to predict violent behavior--and the consequences of such predictions range from denial of bail for the person in question, to 72-hour civil commitment, to a decision to execute someone in murder cases. (The legality of using a prediction of continued violence in determining a capital sentence was confirmed in the 1983 Supreme Court case of Barefoot vs. Estelle.)

With that in mind, Monahan said that even the question of exactly what is dangerous is often open to debate. For example, he cited a survey he conducted of jurors’ attitudes while he was at UCI. “Liberal people in general rated pollution-type offenses as more dangerous than did conservative people. Conservative people rated pornography-type offenses as more dangerous . . . and residents of Orange County found life in general to be more dangerous” than people in other parts of the country, Monahan said.

Another important question is just how accurately a professional can foresee violence. Explaining that any prediction of a person’s dangerousness is going to be subjective, Monahan said recent surveys have shown that clinical predictions of violence are at best accurate only a third of the time. According to one test, psychologists and psychiatrists who are highly dogmatic are more likely to predict dangerousness; other studies have shown that some psychologists and psychiatrists have “absolutely zero ability” for accurately making such predictions, he said.

Despite such unpromising evidence, all sorts of variables are used to predict violence--many of which raise serious ethical questions, Monahan said.

Multitude of Predictors

For instance, according to some of the demographic factors Monahan cited, a poor, nonwhite, unmarried male in his late teens or early 20s with a history of drug or alcohol abuse, a low IQ, a poor education, changes his address frequently and has employment problems would not be off to a good start in an evaluation of dangerousness.

A multitude of other factors have been suggested as predictors. For example, Monahan recalled that an old survey showed that boys in the third and eighth grade who preferred the television show “Have Gun Will Travel” to “The Lawrence Welk Show” were three times more violent. (“Any third-grader who watched Lawrence Welk would have to be mentally ill,” one psychologist in the audience muttered.)

Using such statistical devices, however, has “engendered lots of litigation” as well as other reactions, Monahan said. For example, when word got out that the Michigan Department of Corrections was using demographics to predict dangerousness at parole hearings, “there was a rash of prisoner marriages.”

But despite problems with attempting to predict violence (the American Civil Liberties Union has flatly condemned such attempts, according to an article Monahan wrote for the January, 1984, issue of the American Journal of Psychiatry) Monahan said he believes such predictions are necessary and will be made more frequently.

‘Bending Over Backwards’

“At least in the courts, and in many legislatures as well, in the past five years, there has been a bending over backwards in terms of allowing public policy to be predicated on predictions of violence. . . . The question isn’t how well can we predict violence in some absolute sense but rather if we don’t base public policy on a prediction of violence, what else are we going to base it on?

“I’d personally be satisfied with a lower acceptable level of predictability” in cases where 72-hour involuntary observation is at stake than in cases where the person might go to the gas chamber.

Monahan explained that he sometimes consults for the U.S. Secret Service, which keeps a detailed record of anyone who has made any sort of threat against the President or his family. That list now includes 200 people who have made “explicit threats,” he said. “And (the Secret Service) will take action even on a very low level of predictability.”

Monahan noted that “of the 13 people who have attempted to assassinate a President, all but one have had a history of mental illness.” So when the President’s motorcade is passing through a given town, Secret Service agents may visit a potentially dangerous citizen and say, “Here’s a newspaper; here’s what’s playing at the movies and we’re going . . . ,” he said.

“That’s an intervention that seems to me isn’t really bothersome, given the enormity of the consequences if there’s a false negative prediction (of dangerousness).”

Such solutions, however, are not generally acceptable for a therapist, the afternoon panelists seemed to agree.

Panelist Dr. Mark J. Mills, a psychiatrist and attorney who is chief of psychiatry service at the Veterans Administration Medical Center in Brentwood, offered a case study that, he said, suggests how important appropriate diagnosis and action can be.

“Not too long ago at the VA we had a guy drop by for an evaluation,” Mills said. In the course of the interview, the man mentioned that he was planning on going to the Hollywood Bowl that evening. The therapists nodded and said, “That’s nice,” Mills continued. But as the interview progressed, the man added a few details about his itinerary. He said, “Well, he did have a couple AK-47s, 5,000 rounds of ammunition, and he was planning to sit up in the hills and blast away till somebody blasted him . . . because he felt he didn’t have the courage to commit suicide.

Man Was Hospitalized

“Needless to say, we did hospitalize him . . . ,” Mills said, later adding, “In general, when asked to evaluate a patient, I would try to err on the side of overcomprehensiveness.

“But let me remind you very clearly that violence is not a psychiatric diagnosis,” he continued. “And . . . often there is no single best treatment for violence.”

In fact, Mills’ prognosis for the treatment of violence in general was rather bleak.

Once a diagnosis is made and organic problems, such as the sort of brain damage that makes a person chronically violent, are ruled out, there are several options, he said. One strategy is to treat the mental illness, Mills said. He warned, though: “Sometimes treating the illness will actually breed violence. And very often I believe it will make no difference whatsoever.

Among his other suggestions were “remove the stress” and “teach new coping strategies.” Psychotherapy also “is great stuff,” he said, but “mighty slow stuff.”

Then there are the “psychopharmacological strategies,” which raise “very subtle and complicated questions,” Mills said.

“Clearly, any medication that gets the patient comatose treats the violent behavior,” he said. He added, though, that there are obvious problems with such an approach.

Therapy Effective

Less severe drug therapy with medications such as Haldol, however, has been shown to control violence in many types of patients, he said.

Mills also pointed out that “there is a very clear association between testosterone and violence” and suggested, “The other thing you can do, of course, is you can castrate folks. . . . There are some chemical ways you can do that.” Dryly he reminded the audience, “But in most states that’s a high-risk, experimental approach,and it’s not something you want to do routinely.

Extreme Remedy

“Another, even more extreme, remedy is psychosurgery . . . but in this country, we generally--and I would say this is generally appropriate--recoil at that notion. Somehow the brain is pretty special stuff, and to have some gloved surgeon mucking around in there is not a very appealing fancy.”

Mills concluded, though, that after a therapist has effectively diagnosed dangerousness, warned any intended victims and decided whether to request involuntary commitment, there’s not a whole lot of good news about what comes next--for the therapist or the patient or society.

“What can you do?” Mills asked rhetorically.

He answered his own question: “Lord knows.