Behavioral Science and the Secret Service: Toward the Prevention of Assassination (1981)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

PREDICTING VIOLENT BEHAVIOR: A REVIEW AND CRITIQUE OF CLINICAL PREDICTION STUDIES John Monahan, Ph.D. Professor of Law and Professor of Psychology University of Virginia Law School Charlottesville, Virginia In looking at the state of the art in the area of predicting violent behavior, I will briefly review the clinical prediction studies. To my knowledge, there are only six studies which test the validity of predictions made by mental health professionals (clinicians) in the area of violent behavior. Five of them are reviewed in the monograph that has been distributed to all of you* (and are summarized on page 48), and one has been published since the monograph went to press.** The level of accuracy in the predictions found in the existing studies ranges from 14 to 41 percent, with many different factors affecting the specific figure, most notably the base rate for the violent behavior in the population being studied. The study that reported the highest percentage of "true positives" (41 percent) also had the highest percentage of "false negatives" (31 percent). ("True positives" are persons predicted to be violent who did, in fact, subsequently engage in the violent behavior. "False negatives" are persons predicted to be safe who turned out to be violent.) So there was a 10 percent differential in favor of the individuals predicted to be violent. A fair summary statement of the existing literature on the prediction of violent behavior would be that mental health professionals are accurate at best in one out of three predictions of violent behavior that they make. Many reasons can be offered for this figure, perhaps the one most often cited being the low base rate problem—the fact that in many cases the violent behavior of interest *J. Monahan, The Clinical Prediction of Violent Behavior, U.S. Department of Health and Human Services, National Institute of Mental Health (DHHS Publication No. [ADM] 81-921), 1981. **E. S. Rofman, C. Askinazi, and E. Fant, "The Prediction of Dangerous Behavior in Emergency Civil Commitment," American Journal of Psychiatry 137 (September 1980): 1061-1064. 129

occurs so infrequently in the population under study that the error rate will of necessity be high. There are also many psychological factors that have been mentioned as accounting for the preponderance of "false positive" predictions generally (that is, predicting that an individual will engage in a behavior, when in fact he does not engage in it—in this case, a type of violent behavior), most important among them that the "costs" of the two types of mistakes—false positives and false negatives—are drastically different for the people doing the predicting. The costs of false negatives are generally much higher. There are several observations that have to be made regarding the existing body of literature on prediction of violent behavior. One is that all the studies, with the exception of the one that was just recently published, involve long-term predictions. The persons predicted to be violent had all been in institutions for at least several months, and in some cases many, many years—15 years on the average in one study. They were followed up in the community for between three and five years. So these are not studies of violence that is about to take place imminently. They are studies of persons who have generally been off the streets for a long time—predictions about how they will behave when they are back on the streets for a long time. Second, all the persons predicted to be violent in these studies were diagnosed as mentally ill. To what extent these findings can be generalized to people not diagnosed as mentally ill is not known. Third, all the persons predicted to be violent in these studies had a history of violent behavior in their background—often an extensive history of violent behavior—which meant that the base rates for violent behavior among this population were already higher than expected in the population at large. To what extent one would get predictive accuracy rates approaching one out of three in populations that did not have such a high base rate for violent behavior is unknown. The existing research can be criticized on many grounds. The principal, and I think most serious, criticism is that the criterion that is used—arrest for violent crime, civil commitment for dangerous behavior, or, in some cases, an aggressive act noted on a hospital record—may in fact underestimate the actual occurrence of violent behavior. Thus, violent behavior may be occurring; it simply has not been found. So the question is, How many of the false positives are really true positives in disguise? This is a question that could be debated at length. My conclusion is that the existing research provides reasonable estimates of the accuracy of clinical predictions of violent behavior. It is likely, I think, that much of the unreported or unsolved violence is actually committed by those people who are arrested or civilly committed for other violent acts. 130

Thus it is not so much that the false positives are really true positives in disguise, but rather that the true positives are actually truer than we might imagine. Not only might they have done what they have been accused of doing or what they admit to doing, but much other violence as well. In the monograph that Saleem Shah of NIMH commissioned me to write, I tried to systematize the clinical prediction process by providing a list of questions that clinicians might ask themselves in terms of how to structure an examination to determine the violence- potential of an individual. I would simply emphasize one point here: an examination of what situations or contacts appear to have elicited violent behavior in the past for any person and how likely is it that he or she will confront the same kinds of contacts and situations in the future might, indeed, be a very important factor to analyze in assessing that person's potential for violent behavior in the future. While mental illness in general, for example, does not, from the research, appear to be significantly related to violent behavior, in any given person case, a certain form of mental illness may be part of a pattern that sets the stage for violent ways of coping with life stress. In another individual, the presence of mental illness might be part of a pattern that leads the person not to be violent. I think, then, that the question is not so much whether various factors in general are related to violent behavior; but, rather, whether in specific cases, patterns can be found where a situation that the person is in tends to elicit violent ways of coping with the stress presented by that situation. In my opinion, it is important for Secret Service agents to be aware of the literature on the clinical prediction of violent behavior. There may be clues in that literature relevant to the assessment of violence against protected persons. Certainly the methodological issues are similar—concerning true and false positives and base rates, for example—but, as I mentioned in my memorandum distributed to you, the predictors of political violence—of violence against protected persons—may not be at all synonymous with the predictors of street violence, and it has been street violence that has been studied in the literature: murder, rape, robbery, aggravated assault. I am really struck by the fact that some factors that seem to be strongly implicated in street violence seem not at all to be implicated in political violence. Race is the most obvious example. Every account of street violence heavily implicates race as a factor predisposing toward violent crime; yet, I do not know of any black presidential assassins, for example. Race seems either not to be involved in political violence, or to be involved in a manner opposite to the way it is involved in street violence. Perhaps what is really called for with regard to Secret Service assessment procedures is two assessments. Stage one assessment 131

might be an evaluation of violence potential, without regard for the identification of the possible victim. For this assessment, the existing prediction research is relevant. Stage two assessment might be called victim assessment, in which the "direction of interest" of the individual being assessed receives special attention. Perhaps it is at this second stage, where things like political ideology, membership in groups wishing to overthrow the government, and attribution of the sources of the stress in one's life to political figures are most relevant, rather than in the first stage of assessment for violence potential. 132

ON INTERVIEWING POTENTIALLY DANGEROUS PERSONS Shervert H. Frazier, M.D. Psychiatrist-in-Chief McLean Hospital Belmont, Massachusetts I, too, have thought about the various questions that have been raised here because most of our experience is based on assessing individuals who have crossed the personal assault line—who have killed someone, who have demonstrated dangerous behavior, and who are also mentally ill and in institutions for the mentally ill. We could start out by noting that the Secret Service has various types of persons to protect, and not just the president who is now in office. In fact, most of the agency and staff energy are devoted to protecting a large number of other persons. The list of protected individuals has been expanded gradually over the years, so that what may be appropriate for use in examining or assessing and interviewing a subject about a president now in office may be different from what is appropriate for a protected person who is out of office, or one about to lose his or her protected status. There are, also, some conflicts from my point of view about what the Secret Service agents have to accomplish. They have to obtain a lot of demographic data in the initial interview— citizenship status, residence, changes of residence, employment, military service, aliases, and so forth—which they have to report back and which they must keep updated. Obviously they also need to find out about any mental or criminal history, education, training, skills, weapons knowledge, addictions, sexual-affectional preferences, family history and, more importantly, the specific ideation the individual has about the protected person. Obviously there are volumes and volumes about interviewing techniques, interviewing skills, and interviewing methodologies, written for general interviewing, psychiatric interviewing, and so forth. Probably the most pertinent here is Merton's work on focused interviewing.1 Although the book is old, some parts are relevant to Secret Service assessments of potentially dangerous subjects, as the subjective experience of a particular person in a given situation is discussed. Merton also discusses the dyadic and triadic interview situation, and the fact that the number of persons present during the interview is an important factor. The book is thus relevant to Secret Service considerations as to whether interviewing should be 133

done one-on-one or by pairs of agents with a subject. Then, of course, the subsequent writing of Kahn and Cannell^ at Michigan describe the communication process—the individuals receiving the messages—and the problems involved in these communications, and the various judgments which go into that dyadic or triadic situation. There is also the MacKinnon and Michels book on the psychiatric interview,3 which essentially takes each particular diagnostic category and describes the psychopathology and then proceeds to show how to open the interview, how to manage it, what to expect, and what kinds of questions to ask in an enlightened fashion, so that the greatest amount of information is elicited in the shortest period of time. When it comes to what I do when I sit down with someone who is said to be dangerous—said to have a threat in mind—I want to know what is on the person's mind. I work in a male hospital for the criminally insane, but which is staffed with a large number of women who are very important to the process of determining what actions and interactions take place. We always have a woman in the room during the interview; she may be a nurse, a social worker, a psychologist, or a psychiatrist. I start out by trying to settle the person down and inform him of the who, why, where, and what is going on. This has to do with being appropriately direct and at the same time not at all misleading the patient. When we communicate with Secret Service people, we note they are very direct about what they are doing, and that is important. I don't know how they came by it, but intuitively they have learned to keep aloof in interview situations—not to be buddy-buddy and not to get too close to the very disturbed, the very paranoid person. I think this is an essential part of that kind of process. They certainly have been doing a good job with the mentally ill, at least; and sometimes I think the question of feeling inadequate about what one does is probably related to the factors which I will talk about in a little while. I always state the obvious: what I am doing there and what I expect to find. I assume, also, that there are no secrets. That is the position I put myself in so that I can try to find out any information and be open to ask any question that I think is necessary. I am also wary of prejudgments—fixed ideas on my part of what this client is up to or what I think is in the offing for him. This helps avoid stereotypes about what one anticipates from the person one is interviewing. I believe this is very important. I don't always assume that a subject's interest in a protected person (the president, for instance) is menacing. The interest may simply reflect the individual's need at a specific time. Sometimes the nature of that interest can best be extracted by eliciting the strengths of that person in a very straightforward way. That is, not 134

everyone who is concerned about the president or the president's wife (as in the case described last night) has sinister intent. A good many people are concerned about them and have very long and detailed fantasies about them as authority figures, as individuals who represent persons out of their past, and I think that is an Important part of what does happen. The role of the Secret Service agent who is doing the interview may be quite different from the role of the physician who does an interview. In my view, the identity or role assumed by the agent is important. Do you, a Secret Service representative, think you are part of the criminal justice system, or do you think you are part of the mental health system when you are asking questions of persons? I know that you, also, encourage and participate in long-term care of persons who have made threats. The question is, What is your role identity in this? It is pretty important to get that straight so that you know which system you represent. One can use dignity and respect in either role, but certainly a police or criminal justice role and a mental health role are different. It is a special task which you have, to be sure. Since your statistics show that the majority of the persons you see are mentally ill, you are obviously into the business of evaluating the mentally ill and taking care of them or recommending some kind of care. It is important in any interview situation that the interviewer pay attention to the following questions: Am I tuned in or am I being misled or getting confused? Is the situation not clear? The response to each is a visceral reaction and also a mental reaction to what is going on. That kind of awareness is essential to being tuned in and recognizing all the time that one must not enter into a persons's delusions or unrealistic expectations. It is also essential to be pragmatic, to keep one's mind on what is real. Individuals being interviewed utilize symbolic references, symbolic ideas. They distort ideas. They avoid some things. They block in certain instances. And they often have a private logic. All of these things require careful determination. I like to elaborate on the thinking processes of the person being interviewed. I like to know how concrete he is in his ideas. The degree of concreteness has something to do with the ability to carry out such ideas. I do this by putting a temporal frame around an event when a person is suspicious about something and I ask him, "Where were you when you first had that thought?" That essentially helps him come back to a place in space and he can begin to concretize. Then I do what is called a "micro interview," going into minute details to obtain a sample of an idea that was conceived and elaborated on during a given period of time. This leads to the very often unwitting demonstration of feelings about the situation being 135

discussed. It also helps to build the data base on which to make a judgment. I am interested, also, in persons with memory defects about these particular events because that, also, brings in the whole question of drugs and alcohol and amnesia and other kinds of repression which go on in a number of persons, especially paranoid individuals. I do not push early in the interview for data when the person cannot deliver the data. I reserve that for later or for a subsequent interview. I know that sometimes the Secret Service agent must be concerned with the time element. This may be the only opportunity to meet that particular person. I have watched a number of people interview, and those who pace themselves to be insistent and persistent with the interviewee are able to induce stress without relying on an overt stress type of interview. I am aware that there are gradations of stress in any dyadic or triadic situation. The introduction of a self-recognized persistence which leads to a certain amount of stress has something to do with the kinds of data which come forth. I also confront the person's dangerousness. When someone is dangerous I think it is often wise to let him know that we consider him dangerous, why he is dangerous, .that we have concerns about him, and that we want to stay in touch with him. This has to do with correcting reality. For very many people this approach rearranges their perception of what is going on and their impact upon other persons. I have found this tactic useful. Obviously, when there are other psychological barriers, such as language or race barriers in the interview, there is no question that specialized interviewers from the particular language or racial/ethnic group are essential. This also applies to female clients, and the need for female interviewers, as we have outlined in our paper. Knowledge of the psychopathology of various kinds of persons is helpful in pinpointing what to track down and what their specific defects are. For instance, if you are looking at schizophrenic persons, you question whether they have problems of personal identity, problems of dependency, problems with assertiveness or aggression or a struggle for power and control, and other defects usually present in this disorder. This helps in determining the kinds of questions that are important to ask. 136

The persistence of ideational patterns and the distortions of perception of the here and now are important in identifying the difficulty. I also note the recurrence of fixed delusional patterns in assessing behavior. We recently found at Bridgewater an extremely high number of persons who, in manic states, have the potential of being very dangerous. We also assess suicidal risk, because in the last analysis the suicide-homocide axis is very often intertwined. I would like to caution Secret Service agents about investing too much expertise on the part of professionals. I think that very often you resort to physicians and psychologists and mental health workers in a hospital setting and are relieved that you have the assailant under wraps. I want you to know that your relief is false—that you probably know more about dangerous people than those mental health professionals in that particular situation. Notes 1. R. K. Merton, M. Fiske, and P. L. Kendall, The Focused Interview (Glencoe, I11.: Free Press, 1956). 2. R. L. Kahn and C. F. Cannell, The Dynamics of Interviewing (New York: Wiley, 1957). 3. R. A. MacKinnon and R. Michels, The Psychiatric Interview in Clinical Practice (Philadelphia: Saunders, 1971). 137

PROBLEMS AND METHODS IN PREDICTING PAROLE BEHAVIOR: UTILITY OF PREDICTION INSTRUMENTS Don M. Gottfredson, Ph.D. Dean, School of Criminal Justice Rutgers University Newark, New Jersey John Monahan already mentioned that problems and methods of prediction in areas such as parole are no different generally from problems of prediction in other areas. There is a long line of studies in the prediction of parole behavior, going back at least 50 years. Quite a lot has been learned in those studies. I shall mention about 10 or 12 things that may have some relevance to the concerns we are discussing. I assure you that there is good evidence for some of what I will have to say, but there is quite a bit of opinion mixed in, too. (I am not going to have time to tell you which is whichl) Some fairly stable predictors of parole behavior have been identified. These include such variables as the type of offense of conviction, age, indices of prior criminality or conviction, and drug abuse. These tend to be found repeatedly in many studies and to be fairly stable over time. At the same time, items of data that we might think would be useful in prediction turn out not to be. That is, there is simply a great deal of information on offender attributes that is not predictive. It has been found that valid parole prediction devices can be developed. It has also been found that quite a lot of modesty in describing the degree of validity that we are talking about is in order. Also, despite our knowledge of how to devise predictive instruments (that is, statistical prediction devices) , corrections agencies, probation departments, and similar agencies often fail to use widely accepted procedures in constructing them. Fundamental errors of method are made repeatedly, concerning such issues as representative sampling or the need to test validity in new samples before application or operational use of the instrument. In the area of parole behavior, the evidence about clinical versus statistical prediction is much the same as in other fields. There has not been a lot of study in this particular area, but generally the statistical methods have fared better when pitted against clinical predictions. One thing that has been learned more recently, from much debate and discussion about particular statistical methods for combining 139

information into some kind of instrument, is that the more sophisticated statistical methods (and even more appropriate statistical methods with respect to the statistical theory involved) do not seem to work any better than much simpler methods. It should be mentioned that the statistical literature on parole shows little or no ability to predict person offenses as distinct from property offenses. Although something is known about transitions from one offense to another when there is recidivistic behavior, one cannot predict the nature of the next offense very well from the prior offense. It can be claimed that there are three areas of demonstrable utility for predictive instruments of the sort mentioned above. I will mention them in what I think are the rank order of confidence that I would have in the appropriateness and utility of their use. It might be a different order than you would suspect. The main utility, I think, for instruments of this sort is as a research tool in program evaluation studies in which there is a lack of appropriate experimental controls, such as randomization. Prediction methods provide some ability to correct for demonstrable bias at the start of the comparison. This may make possible a useful quasi-experimental design; and that is, in my opinion, the best use of this kind of instrument. A second use of this kind of instrument is in program planning. A third would be for placement decisions in cases where this kind of instrument is believed useful. One of the things that has been learned from this line of investigation is that there is a need—if there is to be a serious attempt to investigate the potential usefulness of such a tool—for a research capability within the organization of the agency that will use it. There are a variety of reasons for this. One relates to the quality of the data that will be available. It is necessary to monitor the reliability of the data items. Another concerns the need to assess validity repeatedly, to see whether the empirically demonstrated relationships remain constant with the passage of time and with social change. Thus, the agency must have an ongoing, continuous evaluation program or function. Finally, as discussed in John Monahan's monograph, it might be in order to move away from comparing the usefulness of clinical versus statistical prediction to examining the utility of statistical prediction via clinical prediction. Based on this literature I would say that the most relevant issues to discuss in this conference are strategies for dealing with the base rate problem, the need for systematic data collection for this kind of research effort, and the idea that the development and continuing program should be located within the agency. 140

Bibliography 1. Gottfredson, D. M. Assessment and Prediction Methods in Crime and Delinquency. Appendix K to the Task Force Report: Juvenile Delinquency and Youth Crime of the President's Commission on Law Enforcement and Administration of Justice, 1967. 2. Gottfredson, S. D., and Gottfredson, D. M. " Screening for Risk, A Comparison of Methods." Criminal Justice and Behavior 7 (1980): 315-330. 141

SUICIDE AND SUICIDE PREVENTION RESEARCH James H. Billings, Ph.D., M.P.H. Director, Institute of Epidemiology and Behavioral Medicine San Francisco, California Research problems presented by the low base rate for assassination have been raised several times in our discussion thus far. Some of the issues in suicide and suicide research provide interesting and informative parallels which may be relevant here. For example, suicidologists have been involved in attempting to identify the suicidal patient who ultimately kills himself or herself since Durkheim's Le suicide established a beginning for systematic investigation of suicidal behavior in 1897. The suicide death rate in the general population is approximately 10 per 100,000, or .0001 percent. The rate of suicide among the highest risk group, previous suicide attemptors, is roughly 1,000 per 100,000 per year, or one percent. While one percent is a more researchable base rate figure, the reality continues to be that there are 99 chances out of 100 of being wrong in attempting to predict who will suicide even among the highest risk population. Prediction of suicide continues to be a complicated and controversial issue. There are also similarities between suicidal patients and those exhibiting "dangerous behavior" in terms of management issues. What does one do legally, ethically, and therapeutically once the high risk person has been identified? One of the most creative approaches to the prediction and management problems raised by the suicidal patient is exhibited by a study conducted in San Francisco by Dr. Jerome A. Motto.* This prospective study (1968-1974) followed 3,006 high-risk patients who were hospitalized at nine psychiatric inpatient facilities in San Francisco for a depressive disorder, suicidal ideation, or a suicide attempt. Each patient was evaluated by research associates, and 175 variables were coded on a research protocol for each case. All patients were followed up to five years after discharge to determine suicide or non-suicide status. *J. A. Motto, "The Psychopathology of Suicide: A Clinical Model Approach," American Journal of Psychiatry 136 (April 1979): 516-20. 143

There were basically two purposes in conducting this study. One was to generate statistical models whereby it would be possible, with greater sensitivity, to identify the persons among the high risk group who were most likely to kill themselves. Second, the study was developed to test a suicide prevention technique designed for those patients who had declined to participate in mental health treatment following discharge. The findings of this study provide some encouraging perspectives on the problem at hand. Dr. Motto has been able to generate two relatively successful predictive models, and others are in the process of being created. The ultimate value of this approach can emerge only through repeated prospective application. This experimental suicide prevention program involved unobtrusive, non-demanding contact maintained with a random sample of the patients who had refused treatment after discharge. It consisted of a telephone call or letter which communicated an interest in how they were doing and expressed the hope that things were going well. The contact was maintained at regularly scheduled intervals for five years following discharge. It is my impression that there is a statistically significant difference between the contact and no-contact groups in terms of suicidal outcome. The study described above suggests that there may be therapeutic value in using contact maintenance procedures for persons identified as "dangerous" by the Secret Service. That is, there is likely to be significant preventive value in providing the "dangerous person" with consistent and basically non-threatening contact with an organization, even the Secret Service. 144

ASSESSING THE CREDIBILITY OF NUCLEAR THREAT MESSAGES Brian M. Jenkins Director, Security and Subnational Conflict Program Rand Corporation Santa Monica, California One of several research projects on terrorism that we at Rand have been involved in over the years deals with assessing the credibility of nuclear threat messages. During the past 10 years in the United States there have been more than 50 communications—most of them written, a couple of them verbal—involving nuclear threats against American cities and including some sort of demand or message. The problem in reacting to these messages is that the two traditional modes of assessing them are declining in utility. One was to dismiss them because we had confidence in the integrity of our safeguard system for nuclear materials. If no one can get the nuclear material, no one can make a nuclear device. However, there is an increasing amount of material in that system that is unaccounted for. More important, there are more and more nations which have their own nuclear programs. So there is a lot of nuclear material out in the world that is beyond the control of our inventory system, and therefore that method of dismissing a threat is less reliable. The second criterion of assessing such messages had to do with the technical quality of any diagrams, jargon, descriptions of devices included with the nuclear threat. As I am sure you have all seen in the newspapers, it has become fashionable for college students to design atomic bombs instead of writing term papers. Given public discussion of nuclear matters, those things which were once closely guarded secrets are now available to a much larger population; and we see more sophisticated messages which use nuclear terms correctly. Therefore, it was considered necessary to develop a systematic capability for assessing the credibility of those messages. What is important here—and I think interesting about our study in the context of this conference—is that we are talking about assessing the dangerousness of an individual. We are talking about the credibility of a threat made by an individual to carry out a very specific act. That takes into account a broad range of facts and attributes about that person—not just his mind set, but also evidence of certain capabilities, and so on. We were asked at Rand to develop the behavioral component of this threat assessment capability. (Others handle the technical and operational aspects.) We look at the message and try to provide information about the individual himself, about his mind set. We try to address the issue of credibility. I should mention here that for 145

the majority of the threateners we look at—the authors of these threats—it appears that the message is not being used as a vehicle of communication, but rather is the "event" itself. As concerns the Secret Service, that raises a question: Does the assassin population come from a different group than the population of threateners? There are some similarities between our problems and those of the Secret Service, especially as concerns methodology. For instance, we, too, are dealing with a low base rate phenomenon. Out of all the threats we have looked at, we have judged only two as having credibility. These two did not involve nuclear bombs. One was a non-nuclear threat. The other involved a threat to disperse radioactive material. In fact, a person was apprehended, and did have the stuff in his possession. One important difference between our assessment and that of the Secret Service is that ours involves a rather elaborate evaluation process. There are different teams involved in different parts of the country, all linked by computer, and each team has a fairly broad variety of people in it. We do it very few times. We are talking about 50 or 60 of these nuclear events in the last 10 years. In contrast, the Secret Service has 20 or 30 investigations opened per day. And we are dealing primarily with written threats and have no opportunity to interview the threateners. That is because the author of the message rarely identifies himself—at least not directly—so that one could communicate with him further. 146

COPING WITH VIOLENCE W. Walter Menninger, M.D. Senior Staff Psychiatrist Division of Law and Psychiatry Menninger Foundation Topeka, Kansas I. Behavior of the Protected and Situations of Risk In this conference, we are addressing a specific kind of potential violence: assassination. Management techniques may, therefore, be directed toward modifying any of the key elements which are necessary for an assassination to occur. As noted by Shahl, the key elements or variables are the targets (protected persons), settings and situations, and assailants (potential assassins). While most of my remarks will address certain features of the assailants, I want to briefly touch some factors that should be considered with regard to the protected. Noting that some victims of aggression, violence, and accidents have unconsciously played a part in eliciting that violence, Rothstein^ has speculated about the unconscious factors in the psychology of those who become presidents or great leaders which might make them more prone to assassination. In his review of the behavior of some of these leaders, Rothstein found some indications in their observable behavior of what one might view as unconscious fantasies of omnipotence or counterphobic elements aimed at mastering man's mortality. As he puts it, "In one way, this could be viewed as evidence of an unrealistic belief in their own indestructibility which would allow them to take unwarranted risks." However, he goes on to acknowledge: It appears that a successful leader may have to be an individual who acts as if he is, and perhaps in some ways believes himself to be, incapable of failure, assured of success. This assuredness of success may have to be communicated to the public non-verbally by confident action. Thus, the same attitude which conveys confidence to the public may make safety more difficult.^ The implications of these observations by Rothstein are that protected persons need to be confronted with what they are doing when they present a pattern of repeated risk-taking behavior. With regard to excessive risk, the National Commission on the Causes and Prevention of Violence (1969) urged limiting the public 147

exposure of the president and presidential candidates and avoiding situations and settings which might increase the possibility of assassination. In its Final Report the Commission observed: There can be no perfect system for guarding the President short of isolating him, confining him to the White House and limiting his communication with the American public to television broadcasts and other media. This extreme solution is neither practicable nor desirable. For political reasons and for the sake of ceremonial traditions of the office, the American people expect the President to get out and "mingle with the people." (Among the eight [now ten] Presidents who have been assassination targets, all but Garfield and Truman were engaged in either ceremonial or political activities when they were attacked.) Still, a President can minimize the risk by carefully choosing speaking opportunities, public appearances, his means of travel to engagements, and the extent to which he gives advance notice of his movements. He can limit his public appearances to meeting places to which access is carefully controlled, especially by the use of electronic arms-detection equipment. Effective security can exist if a President permits.^ In order to achieve this reduction of risk, an effective alliance must be developed between the Secret Service and those it protects. In the case of the president, this means not only with the president himself, but with members- of the president's family and key members of the presidential staff whose opinions can influence the president's behavior. II. Factors in the Individual Which Contribute to Violence/ Assassination Before discussing principles of management of potential assailants, it is well to review factors which underlie or precipitate violent action. While understanding these factors may not be critical in order to deal with an acutely violent situation, understanding is important if one seeks to prevent violence from occurring. Violent behavior is commonly the end product of extreme anger (rage) or fear (panic), stemming from an actual or threatened injury or loss, which may be real or perceived. In assassination, the violence is directed toward a specific target—a significant governing leader who takes on a special symbolic meaning in the mind of the assailant. The motivating forces are usually multiple and 148

both conscious—such as achieving political change or personal revenge—and unconscious—seeking to destroy a hurtful father figure and at the same time achieving unique recognition or martyrdom. From his research of the violent interaction between individuals and authority figures (police officers), Toch made these observations about motives which underlie violent behavior: Our assumption is that if we want to explain why men are driven to acts of destruction, we must examine these acts, and we must understand the context in which they occur. We must know how destructive acts are initiated and developed, how they are conceived and perceived, and how they fit into the lives of their perpetrators. We must also assume that we cannot make sense of violent acts by viewing them as outsiders. Ultimately, violence arises because some person feels that he must resort to a physical act, that a problem he faces calls for a destructive solution. The problem a violent person perceives is rarely the situation as we see it, but rather some dilemma he feels he finds himself in. In order to understand a violent person's motives for violence, we must thus step into his shoes, and we must reconstruct his unique perspective, no matter how odd or strange it may be. We must recreate the world of the Violent Man, with all its fears and apprehensions, with its hopes and ambitions, with its strains and stresses.^ An additional perspective which may enhance the capacity to understand and deal with violent behavior is offered by Pinderhughes^ from his work with violent marital situations. He calls attention to the frequency with which ambivalent motives are associated with violence. Also he notes that it is not possible to do violence without projecting evil onto the person who is the object of violence. This occurs through a paranoid process whereby that person represents a renounced and projected part of the self which is perceived as totally evil. Then, at the moment of violence, that person is viewed as an enemy who is deserving of destruction. In the study of violent behavior in both clinical and non-clinical settings, various factors have been identified as predisposing, potentiating, and precipitating violent behavior. Some of the predisposing factors in the individual include a defect in the capacity for basic trust in relationships with others; exposure to violence in formative years by child abuse and family abuse; impaired ability to cope with frustration and anger through learning 149

disability, mental retardation, or organic brain damage; and a defective sense of personal identity, self-esteem and self-integrity. A number of environmental factors have been observed to increase the potential for aggressive behavior. As noted by O'Neal and McDonald', these are noise, heat, anonymity, diffusion of responsibility, audience approval, territoriality and crowding. With particular reference to an aspect of territoriality, Kinzel^ discovered a significant difference in the "body buffer zone" of violent offenders, in contrast to non-violent offenders. This zone is the area surrounding the individual within which anxiety is produced if another person enters. Kinzel found the average body buffer zones of the violent group were almost four times larger than the zones for the non-violent group, particularly in the rear of the individual. He concluded that the violent offenders appeared to be most threatened by emotional and physical closeness, and that this sensitivity was rooted in problems dating back to pre-adolescent years. In many of his offenders there was evidence of homosexual panic. The actual precipitation of violence is often triggered by a specific incident or an accumulation of stresses which prompts in the individual a feeling of desperation and a need to communicate that desperation or resolve the tension by violent action. At the point at which violence occurs, Toch^ notes the intersection of three factors in the individual. First, there is an assumption that the authority has acted unfairly. Second, the individual concludes that the unfairness has reached a point where it cannot be further tolerated. Finally, the individual must feel a sufficient sense of potency (or disregard of consequence) to initiate violence. With regard to the specific violence of assassination, Abrahamsen10 and RothsteinH>12 in testimony before the hearings on assassination and political violence, held by the National Commission on the Causes and Prevention of Violence, identified several characteristics of the potential assassin. Such a person commonly is more concerned about world events than his peers; he has exaggerated feelings of omnipotence, feeling he can do more to save the world than others; he is attention-seeking; and he often has significant resentment consciously felt toward the government for alleged mistreatment, either experienced personally or by persons or causes with which he deeply identifies and for which he feels a need for revenge. III. Some Techniques for Managing Potentially Violent Persons The management of the potential assassin by the Secret Service involves some direct activities—contact for assessment and surveillance—and some indirect activities—referral of individuals 150

to appropriate community agencies (either police or mental health) for the institution of external control and possible treatment. Clinically, violent persons may present any of several diagnostic pictures--paranoid, manic, antisocial, episodic dyscontrol, and so forth. Nevertheless, the experience of clinicians who have worked with violent persons suggests a number of principles to be considered in the direct contact with a potentially violent person. Most of these principles follow from an understanding of the psychology of the paranoid process and a sensitivity to approaches which respect the limited coping skills of the potentially violent person. Some of the elements of interviewing have already been reviewed by Frazier and colleagues.13 This subject is also discussed in a number of papers in the literature. Di Bella^ outlines three steps in educating staff to manage threatening paranoid persons. He feels the first step is helping the staff appreciate the diagnostic features of a hostile, paranoid person. Second, he emphasizes the awareness and mastery of one's own feelings when one is dealing with a paranoid person. While Di Bella directs his attention toward persons working in a clinical setting, the principles he outlines are no less applicable in a non-clinical operation. The objective is to be sensitive to feelings generated in oneself during an interaction with a threatening, paranoid person, and to be able to use these feelings both diagnostically and, in a general sense, therapeutically. The third step suggested by Di Bella addresses the practical techniques: The [intervenor] must take a scrupulously frank, honest, straightforward and structured approach to the threatening [person]. He or she must not be too warm and friendly, or attempt to get close quickly. Touching . . . must be curtailed. Similarly, staff . . . should not approach the paranoid (person) suddenly and rapidly . . . .^ Since the paranoid person tends to project responsibility for problems elsewhere, he or she will not often seek help and initiative will have to be taken in offering resources to that person. The family should be engaged as much as possible in the process. Since the paranoid person usually explodes after the period of mounting tension, the intervenor needs to be alert to evidence of that tension or turmoil. With regard to specific interventions, it is important to gauge the degree of failing psychic function on the part of the potential assailant and to arrange for intervention which will provide control, either physically or chemically. This may be through arranging referral for medication to reduce tension and activity, soliciting companionship by a significant person in the individual's life, or 151

arranging for protective custody in a hospital or detention facility. Where there is clear evidence of problems in judgment, it may be possible to refer the person to a setting where he can think and talk about his problems or concerns, and correct distorted perceptions and ideas by confrontation with the reality of the situation. If there appear to be some clear environmental factors contributing to the immediate crisis, it may be necessary to involve social work assistance or participation of a community agency that can attempt to alter the environmental stress.16 How can one change violent persons? Pinderhughes points out, "To change primitive thinking patterns, primitive methods are needed. When thinking patterns are unresponsibe to reason, pleasure or pain—in the form of reward or power—is required to change them."!' In his discussion of strategies for changing violent persons, Toch believes that the task is to reduce the needs and incentives for the violent behavior, or to furnish alternative expressions or courses of action. He notes that violence is more likely to occur in men "who are unsure of their status or identity," and preventing violence requires an awareness and sensitivity to that. In instances where the Secret Service identifies a person in whom they feel an emotional illness is a significant factor, they need to have access to mental health resources to which they can refer such persons. These agencies may then apply medical/psychiatric approaches to the evaluation and treatment of the potentially violent person. °»1° IV. Special Considerations for Law Enforcement Officials As noted previously, the potential for violent action may be potentiated by the way with which a potentially violent person is dealt. With particular reference to the role of law enforcement personnel in their contact with potentially violent people, Toch observes that "police officers sometimes unwittingly cooperate with self-defined champions by letting them play the role and crowding them into a duel." That is, the law enforcement officer may, by the way he approaches and deals with a potentially violent person, significantly potentiate and provoke a violent outcome. Further, the potential for violence may be influenced by those persons who gravitate to the law enforcement profession to satisfy unconscious wishes to express violent impulses as well as to reinforce inner controls for those impulses. Drawing on his work with police officers, Toch made these observations: Our research indicates that the ranks of law enforcement contain their share of Violent Men. The personalities, outlooks and actions of these officers are similar to those of the other men in our sample. They reflect the same fears and insecurities, the same fragile, self-centered perspectives. They display the same bluster 152

and bluff, panic and punitiveness, rancor and revenge as do our other respondents. To be sure, the destructiveness of these officers is circumscribed by social pressure and administrative rules; but it is also protected by a code of mutual support and strong esprit de corps. And whereas much police violence springs out of adaptations to police work rather than out of problems of infancy, the result, in practice, is almost the same.20 One may assume that the Secret Service selection processes are much more rigorous than those of the police departments studies by Toch. Nonetheless, the training of Secret Service agents should be addressed to the potential errors that can be made in the interaction with a potentially violent person which increase the likelihood of a violent outcome. Notes 1. S. A. Shah, "Problems in Assessing and Managing Dangerous Behavior: Some Comments and Reflections," paper prepared for the Workshop on Behavioral Research and the Secret Service: Problems in Assessing and Managing Dangerous Behavior, Washington, D. C., March 1981. 2. D. A. Rothstein, "The Assassin and the Assassinated—as Non-patient Subjects of Psychiatric Investigation," in Dynamics of Violence, ed. J. Fawcett (Chicago: American Medical Association, 1971), pp. 145-55. 3. Ibid., p. 152. 4. M. Eisenhower et al., To Establish Justice, To Insure Domestic Tranquility, Final Report of the National Commission on the Causes and Prevention of Violence (Washington, D. C.: Government Printing Office, 1969), p. 131. 5. H. H. Toch, Violent Men (Chicago: Aldine, 1969), p. 5. 6. C. A. Pinderhughes, "Managing Paranoia in Violent Relationships," in Perspectives on Violence, ed. G. Usdin (New York: Brunner/Mazel, 1972), pp. 111-39. 7. E. C. O'Neal and P. J. McDonald, "The Environmental Psychology of Aggression, " in Perspectives on Aggression, eds. R. G. Geen and E. C. O'Neal (New York: Academic Press, 1976), pp. 169-92. 8. A. F. Kinzel, "Violent Behavior in Prisons," in Dynamics of Violence, ed. J. Fawcett (Chicago: American Medical Association, 1971), pp. 157-64. 153

9. Toch, Violent Men, p. 49. 10. D. Abrahamsen, Testimony at Hearing of the National Commission on the Causes and Prevention of Violence, Washington, D. C., October 3, 1968. 11. D. A. Rothstein, Testimony at Hearing of the National Commission on the Causes and Prevention of Violence, Washington, D. C., October 3, 1968. 12. D. A. Rothstein, "Presidential Assassination Syndrome," Archives of General Psychiatry 11 (September 1964): 245-54. "" 13. S. H. Frazier and colleagues, "Problems in Assessing and Managing Dangerous Behavior, paper prepared for the Workshop on Behavioral Research and the Secret Service: Problems in Assessing and Managing Dangerous Behavior, Washington, D. C. , March 1981. Note especially "Conducting the Interview." 14. G. A. W. Di Bella, "Educating Staff to Manage Threatening Paranoid Patients," American Journal of Psychiatry 136 (March 1979): 333-35. 15. Ibid., p. 334. 16. R. W. Menninger and H. C. Modlin, "Individual Violence: Prevention in the Violence Threatening Patient, " in Dynamics of Violence, ed. J. Fawcett (Chicago: American Medical Association, 1971), pp. 71-78. 17. Pinderhughes, "Managing Paranoia in Violent Relationships," p. 123. 18. S. E. Edelman, "Managing the Violent Patient in a Community Mental Health Center," Hospital and Community Psychiatry 29 (July 1978): 460-62. 4 19. J. R. Lion, Evaluation and Management of the Violent Patient (Springfield, I11.: C. C. Thomas, 1972). 20. Toch, Violent Men, p. 240. 154

THE SECRET SERVICE AND THE MENTAL HEALTH DELIVERY SYSTEM: PROBLEMS AND PROSPECTS Joseph T. English, M.D. Director of Psychiatry St. Vincent's Hospital New York, New York First, let me say a word about how we became involved in assisting the Secret Service in connection with the 1980 Democratic National Convention. Over a period of time, one of the St. Vincent resident psychiatrists had two patients that were a matter of concern to the Secret Service, one who was potentially threatening to Pope John Paul, II, when he was visiting New York and another who was threatening Senator Kennedy. The psychiatrist worked with a Secret Service agent who came to look into the situation. We were enormously impressed with the professionalism with which we were approached. As a result of our contact with these two patients, the location of St. Vincent's on the Lower West Side of Manhattan, and the fact that for the last two Democratic conventions St. Vincent's has been the medical back-up and has had a medical installation on the convention floor, we were asked if we would be willing to assist the Secret Service. We also learned that the Secret Service field headquarters happens to be in our cachment area. Let me also share a personal experience that has influenced my response to the Secret Service request. Walter Menninger and I were working with the Peace Corps in Washington when the assassination of John F. Kennedy occurred in Dallas. I had the unfortunate experience of having to tell two immediate members of President Kennedy's family as to what had been learned in the Dallas emergency room. Subsequently, I spent four days in the White House helping Mr. Shriver to organize the funeral arrangements. During that period of time, one had a sense of the enormous impact of an event like this, not just on the people who came from all over the world, but as evidenced by the largest assemblage of heads of state in the history of the world attending the funeral. I had a chance to talk to some of the Secret Service agents, and it was very clear that they too were profoundly affected. Physicians have all had the experience of "losing" a patient and are trained to expect that as a part of their professional life. When the Secret Service loses a president, it is just as tough on them. I think the memory we all have of such events affects our willingness to help the Secret Service. However, not all my colleagues were sympathetic. In a department like ours, we discuss this sort of thing before we get into it, and despite my mind set, a good many thorny issues 155

were raised. First, there were the legal issues, the ethical issues, the confidentiality issues, and the liability issues that hospitals worry about these days. Second, there were questions as to just what we had to offer the Secret Service, recognizing the limitations of our predictive ability. We had long discussions about these questions. Third, someone raised the concern: "What is this going to do to the image of our department? We will be considered the Lower West Side extention of the Gulag Archipelago; the Secret Service today and the FBI and the CIA tomorrow." Someone else said, "Listen, we are trying to provide treatment here. What is the effect of this involvement going to be on our patient care environment?" Yet another colleague queried, "Where are we going to get the financial support? Are these patients going to be covered, and how long are we going to have them after the Secret Service and the convention leave town? We know you are trying to kill the president, but do you have Blue Cross?" Finally, I think we were concerned that we might somehow raise some false expectations about what we could do in relationship to this enormous responsibility that the Secret Service was going to take on. Despite all of that we said, yes, we would do it. Somewhat in the Peace Corps spirit, we decided if we were going to do it, we would avoid the way mental health professionals often do these things, i.e., in total isolation and without any sense of the context of what the Secret Service is up against. We met with the agents that were going to be involved in this task on their turf, in their headquarters. In their conference room, we met a bright, interesting and obviously concerned group of agents. I took our clinical director, our chief resident, and some of the others who would actually be involved at St. Vincent's. The first thing our clinical director did was to get up and say, "Look, what worries us most is that we are going to create expectations just by being here that we cannot fulfill. We are going to start by telling you what we cannot do." He proceeded to explain the commitment laws of the State of New York, the fact that we could not be involved in preventive detention, that essentially where we could help was if someone was clearly mentally ill and in our judgment represented a danger to himself or someone else. We tried to give a feel for what the ground rules are in our world, the world of the psychiatric hospital in New York City. It became apparent that something we could do would be to expedite the evaluation of people that the Secret Service agents would want to have evaluated. It became obvious in the discussion that one of their problems is it can take a lot of time to get an evaluation done in a large medical center. Setting up arrangements that would provide a fast-track evaluation would be useful to limit the time agents would be away from their primary area of activity, i.e., the convention. 156

When we met in advance with the Secret Service, we arranged to have a debriefing, within one week or two after the convention, to see what we might have learned together. By that time, we had also received word from David Hamburg about this meeting, and we said, "Perhaps there will be something coming out of the convention collaboration that might be relevant to a meeting such as this." I must say they were as enthusiastic about that as we were. Our actual experience was really anticlimactic. We were asked to evaluate only two patients who were brought to the emergency room. One we hospitalized, clearly mentally ill, a chronic paranoid schizophrenic who was up in the Waldorf looking for Senator Kennedy who "owed him a lot of money"; he himself was "a billionaire," had "23 Nobel prizes", and all the rest, and was very angry about all this money that Senator Kennedy owed him. When he came to the emergency room, the question of mental illness was clear, but the dangerousness determination was made by the clinical director together with our resident's perception of the agents' expertise in this regard. The agents' judgment was that this person was imminently dangerous and their opinion was a critical factor in influencing that admission. In the other instance, there was evidence of a personality disorder, but certainly no major mental illness requiring treatment, no interest in treatment; and the person was discharged. From our experience, let me offer several observations. First, it was enormously important to have the prior association to the event that was developed between our hospital and the Secret Service itself. We wonder whether every field station of the Secret Service—certainly in major cities where protected persons are coming in and out—might not establish such a relationship, develop rapport and mutual understanding so that whatever help a medical center could give could be facilitated. Second, though we got a clear perception of what the agents' task was, it was very clear that they had great compassion for some of these people that they would see who might or might not be dangerous but who clearly were in need of help from some mental health professional. It was obvious that what we were able to offer was helpful, i.e., some evaluation and an attempt to link that person to some kind of mental health care which would not only be good for the individual but, hopefully, would also reduce the possibility of what the Secret Service is principally concerned with preventing. The issue of training fascinated us as well. After our initial meeting with the agents before the convention, they started asking us about how they could improve their ability to assess dangerousness. However, sitting there with them, we developed enormous respect for their experience and expertise in this regard. To be quite frank, we 157

were very worried about contaminating that expertise, so we said, "We have nothing at this point to say to you in that regard. We really wonder about our own profession's ability. We would suggest that you continue doing whatever you have done. At some point we may have something to contribute. Right now, when you think mental illness is a major factor and you are wondering about its impact on the problem with which you are concerned, we will be glad to be a part of the picture; but you may be much more adept at this than we are." We think that statement may have reinforced whatever sense of adequacy they had and maybe that was, in itself, valuable. Our concern about the training issue would come down to this: On the basis of the discussions that we have had with them, it is clear that they are supporting a lot of chronic schizophrenics, but we don't think they do this essentially to perform a mental health function. We think that is another way they have of getting the kinds of information that they need to do their job, and I cannot think of a way that we could more misserve them than to try to turn them into amateur mental health professionals. We would, however, like to experiment with ways in which we could reinforce each other's adequacies in dealing with problems of great mutual interest. 158

LEGAL AND ETHICAL IMPLICATIONS OF SECRET SERVICE INTERVENTION Charles R. Halpern, J.D. Associate Professor and Director Institute for Public Representation Georgetown University Law Center Washington, D. C. In general, I will limit my comments to what we might call involuntary or unwelcome intervention. If the Secret Service adopts as a strategy getting a dangerous person out of Los Angeles by sending him to Hawaii for a couple of weeks while the president is visiting, that is the kind of intervention that I think does not pose legal or ethical problems, but as the discussion last night and today has suggested, intervention can sometimes come in less benign guises. It can involve triggering a civil commitment process. It can involve a more or less intrusive kind of surveillance which might involve interviews or discussions with the employer of the person or members of his family or circle. Those kinds of interventions could have extremely negative consequences for the person who is the subject of those interventions. I should think that being identified by the Secret Service as a threat to the president or another protected person could have really quite devastating effects on the life of the person into whose world this intervention comes. An argument for humility in intervention grows out of the statistics that John Monahan gave us—that mental health professionals have, at best, a one-in-three success rate in predicting dangerous behavior. Nothing I have heard today suggests that the present prediction rate of the Secret Service would be any higher than that now or in the foreseeable future. In thinking about the legal and ethical implications of intervention, it is appropriate to take a moment to step back and look at some of the other values and legal considerations that are implicated when such an intervention is made. First and foremost there is the presumption of innocence. We are, as American citizens, presumed innocent until proven guilty. By hypothesis, most of the interventions will be into the lives of citizens who have committed no crime and been convicted of no crime. It is true some of them may have violated the threat statute, but they have not been adjudicated in any event. Many others have not even arguably committed any crime. So there is that highly prized, constitutionally-based value that must be taken into account. 159

Second, there is what might be called the right to privacy or, more generically, the right to be left alone. We all have the right to be left alone unless we commit an act which has been defined as criminal. As for the mentally ill or arguably mentally ill suspect, each state has a commitment law which defines mental illness and defines what kinds of consequences can flow from that characterization. In thinking about Secret Service intervention, it is important to consider not only the mandate of the Service to protect the person of the president and certain other enumerated people, but also to consider the structure of constitutional principles and laws which limit the Secret Service as well as all other policing agencies. It is important to remember that, as special as the Secret Service mandate is, it is not unique. The Federal Bureau of Investigation, for example, has responsibilities which are no less awesome than those of the Secret Service. Ted Kennedy, after all, is protected one day by the Secret Service and the next day by the FBI. Nancy Reagan is a protected by the Secret Service, but the problems of protecting her are probably no more awesome than the problems of guarding against the kind of nuclear disaster that Brian Jenkins was discussing this morning. So, it is important to think sympathetically and constructively about the Secret Service's situation, but not to make the mistake of thinking that this is some kind of unique problem in the world of American policing that is discontinuous from all other problems of that dimension. This should be a cautionary note, if we start thinking about amending federal laws to help deal with threats to protected persons. The "protected" category is an arbitrarily defined creature of a couple of legislative spasms. It does not define the category of those in greatest jeopardy or those whom our country can least afford to lose. Moving on to the subject of the interface between the mental health professionals and the Secret Service, I think many of the suggestions that came up in my workshop group and that I have heard in the plenary make a great deal of sense for facilitating an effective relationship; but again, my suggestion is that psychiatrists and other mental health professionals should not "throw away the book" when it comes to dealing with the Secret Service and the persons it protects. Many of you are probably familiar with the Tarasoff decision in California about the obligations of a psychiatrist when one of his 160

patients makes a credible threat against another person.* That seems to me to be the appropriate kind of analysis for mental health professionals to make in deciding what their obligations to their patients are, how the confidentiality obligation should be interpreted, and the like. The fact that a patient threatens Senator Kennedy before he withdraws from the presidential race rather than afterwards, which is the litmus paper test for Secret Service protection, does not resolve the question for the therapist of when to break confidentiality. This leads me to the conclusion, at least until I am persuaded otherwise, that there is no need for special federal commitment legislation or other legislation to deal with those who threaten persons protected by the Secret Service. A better assessment of state laws, smoother coordination, and anticipatory coordination between the Secret Service and state facilities all seem quite appropriate, but I do not believe that this is a situation in which special laws should be contemplated or passed. Finally, let me just note that the problem of intervention is much larger than the interface between mental health problems and the Secret Service mandate, which has tended to dominate our discussion today. A significant number of people who are of concern to the Secret Service are not even arguably mentally ill. The Puerto Rican nationalists who shot up the House of Representatives were not, to the best of my knowledge, even arguably mentally ill. James Earl Ray was not arguably mentally ill. I understand that some significant number of the 300 "Quarterly Investigation" subjects are not even arguably mentally ill. The problem with intervention in that situation is still more acute. We do, at least, have some laws and some intellectual constructs for justifying intervention to prevent dangerous behavior by people who are arguably mentally ill; but if one is talking about people who are not, then from the standpoint of law and ethics one has an even more complicated problem. We have not yet gotten into the question of the intervention strategies that the Secret Service uses in those contexts. It would perhaps be an interesting subject for inquiry, and from the legal standpoint an even more difficult one. Let me end by expressing my sympathy for the enormous task that falls to the Secret Service and my gratitude to these people who seem to discharge this responsibility with so much sensitivity and so much loyalty to their duties. *Tarasoff v. Regents of the University of California, 529 P.2d 553 (Cal. 1974), vac., reheard in bank, and aff'd. 551 P.2d 334 (Cal. 1976). 161

I think, however, that there are legal and constitutional limitations as to what kinds of interventions are justifiable to control the behavior of dangerous people. They are not matters of legal technicality or changing a particular test in a jurisdiction to the left or to the right. They are fundamental questions about how we have ordered our affairs as a society. This leads me back to the point that Dave Hamburg has raised in this conference, namely to think about other ways in which we can facilitate the Secret Service in the discharge of its responsibilites through altering the behavior of protected persons. 162

ETHICAL AND MEDICAL IMPLICATIONS OF SECRET SERVICE INTERVENTION Robert Michels, M.D. Barklee McKee Professor of Psychiatry Cornell University Medical College Psychiatrist-in-Chief The New York Hospital New York, New York One of the themes of the afternoon has been the potentially different ethical standards which we might apply to the behavior of different classes of persons. For example, consider citizens in general. The ethical standards are those that would apply to a used car dealer or .soap salesman. Government agents might be different. I can imagine things that one might find acceptable in a private citizen, but not acceptable in a government agent. Finally, a member of a profession might be judged by still a third set of standards. There might be knowledge that we would consider appropriate for a government agent to use, but not appropriate for a professional to reveal. We must consider the difference between these roles and, also, the significance of the blurring among them that might occur with certain types of arrangements between the government and the professions. The word "co-opt" was.used earlier, and some might even say that everyone in this room has been co-opted. Shifting to another issue, we are talking about a population of persons at least a significant portion of whom demonstrate behaviors that would suggest that mental health professionals might have relevant expertise regarding their management. However, these persons are not identified as patients and do not have doctor-patient or caretaker-patient relationships with mental health professionals. We are asking, "What are the rules under those circumstances?" The answer is that we do not have clear rules under those circumstances, and we tend to get into difficulty as a result. There are two different conflicts that we have been discussing. One involves conflicting values that are respected by the Service. The mission of the Secret Service is to protect its "protectees." At the same time, like everyone else in our society, Secret Service personnel recogize and are constrained by the rights of citizens, and they have to deal with the tensions between the mission and those rights. We heard in the case presentations last night how they have to balance these two values. 163

A third value that is clearly lower in priority, but recurrent in the discussion, is the desire of Secret Service agents to help people in need. It is not part of their official mission, nor is being helped by the Secret Service a right of the people whom they approach. Yet that theme emerges repeatedly. It is interesting in part because it may offer a way of resolving some of the conflicts between the first two values. It might be that the best way to guarantee the safety of protected persons, and at the same time to respect the rights of citizens in whom the Service becomes interested, is to assume a helping role—as exemplified, for instance, by sending a "dangerous" subject on vacation for the week that the president is in town. The other tension that we have discussed is the tension between two basic frames of reference in ethical discourse, the consequentialist and the deontologic. As applied to the Secret Service, the former might refer to what would result if the Service followed a certain course of action, while the latter would involve consideration of what basic principles of human value should determine how the Service is to function. This came up clearly in one of the cases presented last night, in which we heard the tension between the official justification for an act which included concern for the basic rights of the subject, and the more urgent consequential analysis in the statement, "We got the gun away from him but I am not sure I want to tell you how." This is an inevitable tension in front line work, whether it be in a psychiatric emergency room where there is a document to be signed that says "voluntary consent for admission" (but we often prefer not to know how the signature got there), or in a police station where a subject negociates with an officer concerning whether he can go home or must spend the night in the station. One of the dangers in this kind of tension between avowed principles and immediate consequences is that it creates the possibility of a crisis if it should become clear that the principle was violated in practice, with a backlash that can have very dramatic consequences. Some problems are related to the types of interventions that are used. Charles Halpern has emphasized that the very act of investigating involves potential intrusions into privacy, a value held dearly in our society. In addition, at times there is deceit, or at least the failure to disclose the person's individual rights to him. For example, it was not clear that the two persons who were brought by Secret Service agents to St. Vincent's Hospital [New York City] were told that they had no obligation to go there.1 We heard the examining physician say that he returned the patient to the *See presentation by Joseph English, page 155. 164

"custody" of the agent, although the agent never had custody in the first place. The physician's confusion would certainly raise questions concerning whether the patient knew that he was not in custody. Still other problems arise in response to the use of overt coercion. Finally, the least noxious type of intervention leads to the most interesting problems: the dependency that develops in response to long-term supportive contact. For example, in one of the cases we learned about, the Secret Service may have become the most meaningful relationship in the subject's life. There is an old Chinese custom that if you save a man's life you are responsible to care for him forever. One might argue that the Secret Service develops a caretaking responsibility as a result of its intervention because it becomes the next of kin to some of its subjects. One way of solving some of these problems is to provide attractive interventions rather than noxious ones, so that subjects will embrace the interventions rather than shrink from them. For subjects who are mentally ill, facilitating good treatment might be a socially acceptable, attractive, and effective strategy for diminishing risks without infringing on rights. Charles Halpern says that he does not think that we should change our medical ethics because the person being threatened is or is not protected by the Secret Service. I agree, but the Service may want to change medical decisions rather than medical ethics. You can change decisions without changing principles by using the Service's resources to change the cost/benefit ratio of different courses of action. For example, if the Service made treatment resources available for subjects, those treatments might be prescribed more often by those caring for them, and if those treatments happened to contribute to the safety of the Service's protected persons, that might be to everyone's benefit. Most of the intervention strategies we have heard discussed have been acute strategies. Most of the psychopathology we have heard described has been chronic psychopathology. The current health delivery system in psychiatry does not provide adequate chronic care for most patients. Instead, what the Secret Service can expect from the mental health care system is a series of acute interventions for people with chronic diseases who have as a recurring symptom threats to protected persons. This is not a rational treatment strategy; it simply reflects the general irrationality of the treatment offered the chronically mentally ill. If the Service is going to offer rational intervention systems, it will have to build its own because no others are available. This would be a formidable task, but it might be the cheapest, most humane, and most libertarian way to manage 300 subjects who cause the greatest concern. 165

PERSONAL LIBERTY AND THE SECRET SERVICE PROTECTIVE FUNCTION Charles H. Whitebread, L. L. B. Visiting Professor of Law University of Southern California Law Center Los Angeles, California I concur with Charles Halpern that no issue has for so long tested our legal system as devising lawful means in our free society for preventing criminality. We are correctly skeptical of restraining liberty unless that restraint is based upon demonstrable conduct. Removing people from the streets—putting them in even temporary restraint—on mere hints and suspicions, will not square with the constitutional ideal. Of course, part of our concern stems from lack of confidence in the dangerousness assessment, but even if we were far more sure of our predictions than the present state of the art in behavioral science permits, I would not condone or advocate restriction of constitutional liberty interests of citizens without demonstrable anti-social conduct. This preference for liberty and personal privacy has certainly made law enforcement in general and the Secret Service protective function in particular more difficult. In other countries the counterparts of the Secret Service can sweep the streets clear without the need even for reasonable suspicion. We have made the social policy judgment that such police conduct may damage society more than the harm it seeks to prevent. This choice of a free society in favor of personal liberty has made the job of the Secret Service more difficult, but there are at least three concepts of constitutional criminal procedure which may serve to help the Service in fulfilling its mission. First, there is a significant line of cases permitting police officials to stop and briefly detain any citizen reasonably suspected of involvement in crime. So long as the officer's suspicion is reasonable and articulable, detention is authorized even if no crime has been committed. Officers may detain for investigation those they reasonably suspect may commit crimes. These short-term investigative detentions on the standard of reasonable suspicion provide law enforcement officers a significant weapon in the crime prevention effort. Second, as some of the Secret Service representatives pointed out in the planning session for this conference, the Secret Service surveillance and investigation may in itself have deterrent value. For example, if the agent investigating Mr. Smith during a visit by a 167

protected person to Smith's town says, "During the visit, I will be assigned to you. You can be sure I will be right with you all during this time," such surveillance deters. There may be some hypothetical legal cause of action by the person under surveillance, but so long as the decision to undertake the surveillance is reasonable, I suspect courts will err on the side of approving the investigative conduct, especially in light of the minimal nature of the intrusion compared to the significance of the Secret Service protective function. Third, many state law provisions for emergency civil commitment may permit constitutional restraint of the acutely mentally ill at least for a period sufficient to permit the protected person to finish his visit and leave town. While we have heard of the inadequacy of some state mental health laws during this conference, I do not think amendment of federal law to permit easier commitment of those assessed dangerous is advisable. The Secret Service has an awesome responsibility to protect our leaders; nevertheless, that function is neither unique nor considerably more important than the duties of other police agencies such as the FBI, airport police, harbor patrol, customs and immigration authorities and the like. Tampering with legal bars to involuntary loss of liberty is dangerous. We must remember the fundamental principle that once any citizen's liberty is unfairly or unconstitutionally lost, even in the name of a noble cause, all of us have lost some part of our hard-won liberty. This libertarian rhetoric, while not new, needs no apology as an appropriate restraint on hasty proposals for easing the way to involuntary losses of liberty. As to the other legal issues—which often trail or precede moral and ethical questions as well—I concur in the view of the planning committee that they are best saved for full discussion at some later conference and are not best served with short shrift here. 168

SOME THOUGHTS ON "THREATS" AND FREE SPEECH, AND CASE MANAGEMENT R. Kirkland Gable, Ed.D., J.D. Associate Professor Department of Psychology California Lutheran College Thousand Oaks, California I would like briefly to discuss some of the things that go into weighing competing interests. The Supreme Court in the Watts case^, set out some issues in viewing 18 USC 871, the so-called "threat statute," with which we are in some measure concerned here today. The Court said that the statute under which the petitioner was convicted was constitutional on its face. The nation undoubtedly has a valid, even an overwhelming, interest in the protection of the safety of the Chief Executive, and in allowing him to perform his duties without interference or threats of physical violence. Nevertheless, a statute such as this one, which makes criminal a form of pure speech, must be interpreted with the commands of the First Amendment clearly in mind. What is a threat must be distinguished from what is constitutionally protected speech. In these types of cases, there has been something of a separation between an absolute threat and a conditional threat. As an example of an absolute threat, let me cite an old case, Ragansky^, in which the fellow said something like, "We ought to make the biggest bomb in the world and take it down to the White House and put it on the dome and blow up President Wilson and all the rest of the crooks." That is a kind of statement for which the fellow was convicted under the threat statute. In contrast, here is an example of a conditional threat, which looks like political hyperbole: "Now, I have already received my draft classification as A-l, and I have got to report for my physical this Monday. I am not going. If they ever make me carry a rifle, the first man I want to get in my sights is LBJ." If a statement is conditional, there is no conviction. When we begin to look at the intent of the person, we balance the likelihood of particular risks. We are looking at the purpose of free speech, and so on. My guess is that there should perhaps be a clarification of what is legally required for conviction and what is permissible. Such clarification might not involve changes in statutes, but rather issuance of guidelines as to what is permitted versus what constitutes a threat, and so on. There is, perhaps, something useful as a guideline in Tarasoff3, regardless of how one may feel about that case. 169

Turning to an issue which was illustrated last night in the case presentations, I am wondering whether we might want to think about increasing the use of conditional release from hospitals for subjects we know are about to be discharged. Perhaps the release would be conditional upon medication maintenance of some kind, non-possession of certain kinds of weapons, or specified limitations in travel. We might begin by examining legally and clinically relevant conditions that might be placed on travel. The use of case managers to monitor such subjects upon their conditional release might be an appropriate strategy. I am inclined to mention a device used at one time by the late Dr. Schwitzgebel^S^ who was a close friend of mine—an electronic monitoring system for tracking the locations of persons released from institutions. This equipment allowed the monitoring of a person every 30 seconds as he or she travelled in an urban area, and was required as a condition of release. The person was free at any time to return to the mental hospital or prison, if he or she so desired. (Very few wished to do that.) I think there are probably many such intervention strategies which could be devised and used effectively. The issues are not going to be so clearly civil libertarian versus police. What we really need is a more thoughtful consideration of the legitimate competing interests and a clarification of what happens in practice—not what happens on paper and who signs admission certificates, and so on. We need a really serious and honest look at the processing of these people; and where critical legal issues are raised, we need to begin to develop guidelines for handling such persons. Notes 1. Watts v. United States, 394 U.S. 705 (1969). 2. Ragansky v. United States, 253 F. 643 (CA7, 1918). 3. Tarasoff v. Regents of the University of California, 529 P.2d 533 (Cal. 1974), vac., reheard in bank, and aff'd. 551 P.2d 334 (Cal. 1976). 4. R. Schwitzgebel, "Development of an Electronic Rehabilitation System for Parolees," Law and Computer Technology 2 (1969): 9-12. 5. Anthropotelemetry: Dr. Schwitzgebel's Machine, Harvard Law Review 80 (1966): 403-21. 170