Under-the-Covers Undercover Investigations: Some Reflections on the Stateís Use of Sex and Deception in Law Enforcement
1999 revision of article in Criminal Justice Ethics, vol. 11, no. 1, pp. 13-24, Spring 1992.

Back to Main Page

I am gratetul to John Kleinig for his critical comments. An earlier version of this article with extensive footnotes appeared in the Journal of Criminal Justice Ethics, Vol. 11/no. 1 Winter/Spring 1992.

By Gary T. Marx

I have no sympathy for those who are crybabies about the fact that police officers are selling to those who want to buy drugs. We use every legal means that we can. We want everybody to know that the next drug buy may be from a police officer.
--Mayor Marion Barry News Conference, 1988
Or it might be from a former girl friend working for police who has invited you to her hotel room. The arrest and trial of the former Mayor of Washington DC, Marion Barry, after he had purchased drugs from an ex-girl friend raises a variety of ethical and policy issues involving police deception. One of the most interesting involves friendship and undercover investigations. When, if ever, is it appropnate to use friendship and the lure of sex as part of an investigation?

State-sponsored deception, of course, raises all the ethical issues generally associated with deception.It also raises some issues that are unique to the state as the symbolic repository of societal values (for example, the need to avoid setting bad examples). But when friendship and sex are present, as in the Barry case, the situation becomes more complex. Manipulation, temptation and deception (whether involving motives and/or identity are joined in a potentially explosives mix.

In this article I will focus on the limited topic of sex and undercover investigations. The deceptive use of sex may magnify the basic issue of the violation of trust found within the broader topic of "false friend deception."

It might help our understanding of the larger topic if we focus on the narrower one. Here I ask (1) What is at stake and what is different when undercover operations have a sexual component? (2) What are the ways that sex is used in undercover investigations? (3) How should we judge this behavior?

There is a natural congruence between covert means and sex more generally and sex and undercover activities show parallels. Both are private activities. Those not directly involved are unlikely to know the details.

Secrecy and temptation may play important roles in each. Both romance and undercover activities can involve heightened efforts to create impressions, the keeping of secrets and intense bonding. Prostitution, like undercover work, may involve roleplaying and feigning emotions (and to judge from the classic scene in the film When Harry Meets Sally, faking it is not the exclusive preserve of professionals). Undercover activities, with their secret watching and audio and video recordings, have a voyeuristic quality. Terminology such as "deep penetration" and "access" have multiple referents. Targets of investigations frequentlv report being "screwed" by the agent after their arrest. Agents sometimes refer to the agency they work for as their "mistress." For some agents the excitement of undercover has a sexual parallel. As one highly experienced agent put it in an interview with a co-worker, "The best undercover is exalted in what he's doing; it's almost a sexual thrill."

In the case of the "Mata Hari" phenomenon, undercover and sexual roles may be professionally intertwined --although one recent account suggests that Mata Hari may have been framed. A possible link between homosexuality and spying, at least in the post-World War II British context, has often been noted.

As part of a larger study of undercover practices I have identified a number of ethical justifications for, and objections to undercover work. Among the objections is that of lack of respect for the sanctity of intimate relations. Restrictions on the use of spousal testimony reflect this concern. Unlike the impersonal and instrumental relationships of the marketplace, intimate relationships are valued as (and assumed to be) ends in themselves. They flourish to the extent that individuals feel free to express themselves without suspiciousness or fear of othersí ulterior motives.

In a larger sense, intimate relations can also have instrumental or functional consequences in positively linking the individual to others. Our sense of freedom, autonomy and well-being depend partly on our ability to control information about the self and on our being able to voluntarily enter into relationships with others free from both coercion and deception.

Undercover work exploits the cognitive and behavioral aspects of intimate relations by using them for purposes beyond the relationship itself. As Davis observes, intimate relations inherently involve and develop trust based on the revelation and toleration of a wider range of attitudes, inclinations, and behavior than is the case with more casual relations. This trust leads to the exchange of "confidences," some of which will clash with the public image the individual would like to project.

Covert operations, with their duplicity and betrayal, trade on the trust that is essential to, and defines, these primary relations. Anything that debases that trust must be viewed as undesirable (if not necessarily always indefensible). It is in this regard that seduction is the moral equivalent of rape because they both deny the dignity and freedom of the individual.

By contrasting deceptive situations with respect to whether sexual intimacy (or the hint or promise of it) is present with those in which (from the pointof view of the target of the investigation) psychological intimacv is, or is not, presumed to be present, we arrive at the four types of situation shown in Table 1. The gravity of betrayal increases as we move from type 2 to type 4. The deception in type 4 "seduction" (psychological and sexual intimacy) is more unethical and troubling than in type 3 "platonic" (psychological but not sexual intimacy). These in turn are more troubling than type 2 "prostitution" (sexual but-not psychological intimacy), although there are ethical objections to this as well.

Table 1: Types of Intimacy in Undercover Enforcement
Sexual Intimacy?
No Yes
1. 2. Decoys and prostitution
Presumed psychological intimacy on the part of a target? 3. Platonic friendships 4. Seduction

Decoys and Prostitution
I will firstconsider the Type 2 situation, where there is no psychological intimacy but there is implied or actual sexual intimacy. Other factors being equal, the use of undercover means is less troublesome here for three reasons (i) the violation itself involves sexual behavior (ii) such acts are coercive (attack on a decoy) or by definition, with prostitution commercial and instrumental, and (iii) the resulting arrests usually do not involve consummation.

The ethical objections (at least with respect to the sexual aspect) in using decoys as potential victims in response to a pattern of sexual assault and harassment are minimal. For example, after a series of rapes in a park, a female agent walks alone in the area with a backup team near by. After female patients complained of being assaulted by a dentist while they were under anesthesia, a policewoman poses as a patient and invites assault, with a backup team watching via a hidden camera poised to intervene. Departments in a number of cities have recently used decoys in an effort to combat violence against homosexuals. Some departments (with the cooperation of the victim) arrange for rape and sexual harassment victims wearing a wire to return to their assailants and talk to them in order to gather evidence. Since those cases are likely to involve responses to crimes that have already been committed, coercion on the part of the offender, and a relatively passive response by the agent they are acceptable. The fact that the bait is of a sexual rather than a monetary, or some other nature is irrelevant. Directing undercover tactics against sexual predators in response to a pattern of victimization is appropriate.

In response to concern over sex crimes involving minors located through the internet, agents visit internet chat rooms and pretend to be underage females. In one case a well known technology entrepreneur was arrested for the intent of having sex with a minor. Over a seven month period using the name of "hotseattle," he conversed over the internet with a "girl" who repeatedly said she was 13 years old. When the suspect invited "her" to meet him at a Los Angeles beach and promised to take her to his hotel room and show her computer pictures, he was arrested and his laptop computer was seized as evidence. Crimes involving intent rather than actual behavior are more difficult to prove. A key issue here is also whether the individual was targeted because of some prior reason for suspicion or was simply caught up in a trawling net. Was the crime only an artifact of a law enforcement created morality test? A jury concluded it was and found the man not guilty. The internet offers a new field for criminal and social control innovations. The assumption (although doubtful) of truly anonymous internet communication creates new opportunity structures for both.

What if those providing sexual services are the subjects of victimization unrelated to sex? Is it appropriate to use sex when it is only indirectly an issue? It is not uncommon for urban vice enterprises to be asked to pay organized crime for protection. In response to the extortion of pornographers and brothel owners, police occasionallv (as in Los Angeles and New York cases) act as pornographers and run brothels for a period of time. The goal is to gather evidence of extortion and racketeering. Again, the ethical objections we will note below do not apply.

Nor do these objections apply if those involved in the provision of sex are involved in other more serious crimes (rather than being victims themselves), and if agents facilitate, but are not directly involved, in providing sexual services. For example, in a Chicago investigation into links between organized crime and suburban prostitution, federal agents took over a credit card processing company (whose owner came to them complaining of an extortion threat).Over a four year period they processed $30 million in customer payments for sexual services. These were shown on credit card records as expenditures for food, beverages and office supplies which could be taken as business tax deductions. The FBI-run firm paid the sex clubs from its own bank account and then collected from the customerís credit card company. It also had to pay out $100,000 to local police in bribes to stay in business. The main targets of the investigation were not the actual deliverers of sexual services but organized crimes of extortion, corruption of public officials, and tax fraud.

With respect to prostitution and covert means, there is considerable disagreement about the appropriateness of criminalizing consenting behavior among adults. In enforcing such statutes the government engages in moral partisanship regarding the disputed relationship between sex and friendship/love. It hypocritically combats the separation it opposes through enforcement means that appear to further it. Enforcing these laws may also waste scarce police resources, demean the enforcers, and have a variety of undesirable and unintended consequences.

The use of covert means to enforce such laws (for example, an officer posing as a prostitute or the client of a prostitute, or as a buyer or seller of pornography) raises the problems that characterize any use of deception bv the state. Propriety and the symbolic importance of a pristine governmental image may militate against certain extreme activities (such as the New York City case in which the government financed the making of a pornographic film and had a policewoman direct on-camera sex acts to ensure that they were sufficently explicit to satisfy the legal definition of obscenity).

There also seems to be a threshold of reasonableness. Thus in some jurisdictions such as Seattle, Washington a man can be arrested for "loitering" for the purposes of soliciting a prostitute. This can happen even if there is no direct discussion of a sex act in exchange for money. In an area known for prostitution, simple questions (to an undercover police woman) such as, "are you looking for a date?" or "are you a policewoman?" are sufficient grounds for arrest. Beyond enforcing morality, police may believe that this is a way to combat drug problems, make the city look less "trashy" and even protect the client from being robbed.

Holding apart the morality around criminalization of prostitution, these laws are on the books, the substance of the violation (and its consensual and secret structure) practically requires a degree of participation by the state. There is a distinction between situations in which sex is used in the direct enforcement of laws regarding sexual conduct (for example, prostitution or pornography) and situations in which sex is merely a tool to some other end. There is an obvious justification for using sex (to some degree at least) in the enforcement of laws involving sexual conduct. In addition, targets may be protected by the entrapment defense. These factors are missing when sex is used as a resource against a person who faces unrelated, or no, charges at the time of the investigation.

Prostitution, of course, involves sexual intimacv but not emotional intimacy. In general prostitutes have no expectation of genuine emotional intimacy or commitment from their customers, nor do the customers expect it from them. Their role is strictly professional, and the operative principle is "Let the seller beware." Sex is separated from friendship. Indeed occupational folklore has it that when a prostitute starts enjoying sex with the customers, it's time to find a new line of work. With the publicity given to the use of policewomen posing as prostitutes, in an effort to arrest those who solicit them, the buyer must increasingly beware as well.

Prostitutes know their behavior is illegal and that police will use deception to arrest them. The prostitute chooses to convert sex into a marketable commodity. For law enforcement to use deception in the pretended, or real, purchase of sex is less morally questionable than to use deception in circumstances when it is not for sale, but instead, is voluntarily given out as part of what is presumed to be an intimate relationship.

To be sure, there are certainly questionable schemes to arrest prostitutes. Consider a newspaper advertisement in the Los Angeles Free Press which stated: "Sexy hostesses needed for gambling junket.
Entails foreign travel. Expenses paid." Applicants were interviewed, and some were then invited to a party in a plush hotel suite. Those invited were told the party was being held so they could meet some of the gamblers who were going on the junket. The message to the girls was, "This is a run-through for the forthcoming trip; it's to your advantage to be liked by the men at the party." Fifty-four of the women went too far in their efforts to be liked and were arrested for solicitation. The ad had been placed by the Los Angeles Police Department, and the gamblers were all undercover police officers.

There are undoubtedly better ways to use police resources, but given laws against prostitution, this is not
ethically very disturbing. It contrasts markedly with the ethical disaster that would exist if an undercover
officer formed a relationship with a woman, convinced her to become a prostitute, and then had her arrested for prostitution. I have not encountered such a case, but undercover officers do sometimes play the role of pimp. In a Chicago case an agent had trouble leaving the role once the investigation came to an end. He was suspended for operating a prostitution ring.

An experienced undercover agent offers the following practical advice: "Never arrest anyone you sleep with." Most prostitution-related arrests do not involve consummation of a sexual act. On that account at least, they are less objectionable than those in which the act is carried out. This is clearest when we consider the particularly egregious practice of an officerís consummating a sexual act and then arresting his partner.

Apart from the personal temptation involved, completion of the act may offer clearer evidence of intent
and permit higher charges to be brought. In such (presumably statistically rare) cases, the association
of the stateís power with deception gives a double meaning to the notion of being screwed. There is a theft of services and a terrible imbalance or lack of reciprocity that is not found with most other undercover encounters. Both parties participate in the criminal act, the male officer is paid by the state to have sex in the line of duty, and the participating female gets arrested.

Yet is the above situation any different from one in which a vice officer purchases drugs from a drug seller? The answer is "yes" because the former involves direct consumption (and to the benefit of the consumer, not the society on whose behalf he is authorized to act). However, if the agent consumes the drugs purchased the situations are more parallel. When the pursuit of private pleasures emerges as a new goal, the professional "disinterestedness" ideally associated with the role weakens or disappears.

Still, there is an elusive difference which further indicates what is at stake when sex is deceptively "taken" under state sponsorship. In the first case (involving sex rather than drugs) the active participation of the other party is required. In that sense sexual acts are "personal," even in the absence of psychological intimacy. Betrayal involving another's body adds an additional troubling element, beyond that occurring with the mere exchange of tangible objects or the failure to keep a promise. Such betrayal is less distant and mediated and hence, whether involving coercion or deception, it is more violative.

Another ethical issue concerns equity in assignments regarding gender and sexual preference with respect to the role the agent plays and the likelihood of his or her becoming a suspect/target.

Playing an undercover sexual role and consummating a sexual act within it seems to be disproportionately a phenomenon of male heterosexuals. Greater undercover activity is directed at heterosexual than homosexual prostitution, and male agents are more likely to play the role of the "John" than female agents are that of the prostitute. Some critics also see a double standard (or at least a sexist outcome) in that when policewomen are asked to play such roles, it is that of female prostitute, while policemen rarely pose as male prostitutes.

Paid female heterosexual prostitution is of course much more common than its male equivalent. This represents culturally, and perhaps physiologically, shaped demand and supply factors. Men have traditionally expressed greater demand and also have been more willing to offer the supply without cash and other inducements. In this sense men in principle might be much more vulnerable to sexual exploitation for undercover purposes than are women. However the double standards of our culture push in the opposite direction, and that helps to explain the empirical gender pattern.

It is extremely rare for the male officers who pose as homosexuals to actually have sex with those they arrest. In my research for the book Undercover I did encounter one case in Atlanta. But this involved a new recruit who told his sergeant, "I thought you were supposed to go all the way." With increased efforts by departments to recruit homosexuals, this might be less rare.

While in recent decades there has been a marked increase (partly as a result of court challenges) in the number of policewomen posing as prostitutes who arrest the men who proposition them, I have not found equivalent accounts of actual sexual involvement. However, sexual involvement may occur indirectly when prostitutes serve as informers although such involvement is not usually for offenses related to the act of prostitution (for example, the Georgia drug case discussed later).

The type 4 situation, labeled "seduction" in Table 1 (fake psychological intimacy and sexual intimacy), is the most troubling. I do not argue that such situations are always wrong as a law enforcement tool, although I think they can rarely be justified. In reviewing type 4 situations I have identified five contexts (based on goals) in which seduction is most commonly found: (1) blackmail, (2) stigmatization and/or disruption, (3) general intelligence collection, (4) evidence collection for an offense that is suspected but not yet committed, and (5) evidence collection for a prior offense.

With respect to blackmail a distinction can be made between covert surveillance that permits documentation of sexually embarrassing facts, and the creation -Ėmorally even more troubling--of opportunities for such documentation. The secrecy surrounding both of these makes it difficult to know how widespread they are.

Every few years a case becomes public in which a vice detective is arrested for creating and documenting enticing situations that are subsequentlv used for blackmail purposes. But this use seems more common in international spying than domestically.

The United States had its "honeypots" and the Soviet Union its "swallows," used for both blackmail and deceptive information collection, although the use of sex was probably not as widespread as those who write spy and detective novels would lead us to believe. Given the secrecy and the availability of the tactic to competing countries, it is sometimes difficult to decide just who was seduced and to determine who was the target and who was the agent. Sometimes the individuals involved are both. Thus former FBI agent Richard Miller, who had an affair with a Russian émigré, was accused of spying for the Soviet Union. He claimed that he was trying to use the affair as a counter-intelligence strategy to infiltrate the K.G.B. The U.S. government saw a Soviet plot and claimed that the émigré, who pleaded guilty to spying, was working for the K.G.B.

Domestically, the documentation of already existing embarrassing situations for purposes of blackmail is probably much more common than the effort to engineer them ab initio. The former involves an invasion of privacy but not a violation of intimacy. J. Edgar Hoover was very adept at this. His voluminous files on important people contained information on sexual and other improprieties. Such information was rarely made public or used to prosecute. Instead the implied threat of exposure was used as a political resource.

Information from local police intelligence units in Chicago, Los Angeles, Detroit, Seattle, and many other cities has been used in this fashion.

A British military scandal offers an illustration of the expansionary strategy of coercively "turning" (upward if possible) ever more people involved in violations. A communications specialist in Cyprus was allegedly lured to a party, provided drugs and photographed having sex with two males. He was then told that unless he delivered certain documents the photos would be made public. In a classic espionage technique, he was then pressured to recruit others in the military with whom he had been involved, under threat of exposing them. Seven servicemen were eventually charged with having passed important military secrets over a two-year period.

With the declining importance of ideology as a motive for spying, sex, along with money and drugs, may take on increased importance. But this may be undercut by greater societal tolerance for homosexuality and other sexual practices once condemned by Victorian morality. This tolerance underlies new 1988 Department of Defense regulations requiring employees and contractors with security clearances to report if they have engaged in activities such as adultery, spouse swapping, and "group sex orgies." Under the new rules acknowledged homosexuals have received security clearances because they are not seen as susceptible to blackmail.

In the context of stigmatization and/or disruption, the goal is to create difficulties for a group without resort to legal sanctions. This is a classic tactic directed against politically suspect groups. For example, an informer who infiltrated the Dallas CISPES (Committee in Solidarity with the People of El Salvador), a group which regularly met in a church, reports that his FBI contact encouraged him to seduce one of the nuns in the group and film it. The Church Committee Investigation of the FBIís COINTEL program cites a Klan informant who testified that he was instructed "to sleep with as many wives as I could" in an attempt to break up marriages and gain information.

In the context of intelligence collection, the official goal is preparedness and prevention. The infiltration of dissident political or criminal groups (or milieus) for monitoring purposes is a well-known tactic. In a Los Angeles example, an undercover agent who had infiltrated a local Maoist political organization became the boyfriend of one of the women he spied upon. Under oath he testified that, after consultation with his superiors, he regularly engaged in sexual intercourse with the woman. The relationship was used to help gather information about the woman and her associates and to establish the agent's credibility. He later testified that during the seventeen months he was undercover, he never heard anyone in the organization talk about committing crimes, nor did he see any weapons.

In the fifth context an undercover operation is directed against a particular group or individual there is reason to suspect of criminal activity. For example an officer who successfully infiltrated an international drug ring was greatly aided by having an affair with one of the groupís leaders.

In a large Oklahorna corruption case the govemment's chief witness admitted she had sexual relations with suspects in order to gather evidence. She stated: "At the times was working for the government and I said and did anything to open up and get the information for the government

A further example can be seen in the FBI's effort to locate fugitives from the Weather Underground faction of Students for a Democratic Society. A federal agent posing as a radical infiltrated a student milieu thought to be close to this faction. He developed a relationship with a political activist, and she became pregnant. After considerable indecision, and at the urging of the agent, she had an abortion. His efforts did not locate the fugitive. The agent's work then took him elsewhere, and he ended the relationship. The woman apparently never learned of his secret identity and true motives. The situation would have been more complicated had she decided to keep the child or died in childbirth or developed a sexually transmitted disease or become mentally unstable.

In an equivalent example, Kim Paris, a young female investigator, established a specious relationship which led to a successful murder prosecution. She contrived to meet a murder suspect and then dated him for two months. He proposed marriage to her. However, she told him she first wanted to know what it was that seemed to be troubling him. With Houston police listening through a transmitter in her purse, he confessed, was arrested and subsequently found guilty.

In a similar Greek case the mother of a murdered girl, in cooperation with police, posed as a prostitute in order to infiltrate a group believed to be responsible. According to a newspaper account she allegedly slept with the man believed to be the killer. She wore a wire and the evidence was turned over to the prosecutor who filed formal charges.

How should we assess these five contexts? The first two can be quickly dismissed. Blackmail and disruption are both illegal and immoral, and there is little to discuss in a societv in which conduct is governed by due process and political expression is protected. Of course controlling such conduct in organizations is another matter. But my concern here is with the ought rather than the actual.

The third context, the use of sex to gather general intelligence, should also be prohibited. Police resources are too scarce and the state-sponsored creation of specious intimate relations is too serious and potentially costly an undertaking to be used for fishing expeditions. This also violates the expectation that there should be specific grounds for suspicion before an investigation begins. In addition, in the post-Watergate age the infiltration of political groups may violate the law, or at least departmental guidelines.

In the other three cases, the mere fact that the goal is legal is not sufflcient to conclude that it is sound as public policy. Legality is a necessary but certainly not a sufficient condition for the use of sex in an investigation. Here it is appropriate to recall Justice Stewart Potterís observation that just because there is a legal right to do something, it does not mean that it is the right thing to do. After determining that the goal was legitimate we must next ask, How was sex used in the investigation?

Let us contrast two types of case. First, there is the case of "the beguiling serpent," in which sex is used as an inducement or incentive to encourage illegality on the part of the suspect. Sex is a resource offered in an exchange relationship.

In Florida a man was arrested after a female undercover agent offered him sexual favors if he would sell her marijuana.He at least had possession of the marijuana, which suggests a presumption of guilt, leaving to one side the conditions under which the evidence was collected. More troubling is an Atlanta case involving a former nude dancer addicted to cocaine who worked with the Georgia Bureau of Investigation. Eighteen men were arrested as a result of her efforts. In a sworn statement she said many of those arrested had not previously used or purchased drugs until she enticed them. She approached male acquaintances with an offer of sex if they would purchase cocaine for her, and she steered them toward making purchases from an undercover agent.

Such cases are troubling. But it is not immediately apparent why. The promise of casual sex certainly implied no serious emotional intimacy, so it canít be criticized on those grounds, and the purchase and use of cocaine is illegal.

A contrast of prostitution with drug use may help clarify why this example is troubling. The economic structure of the relationship is similar to that in prostitution. But there is a difference in that in the Georgia case the currency exchanged was an illegal substance, and the arrest that followed was unrelated to enforcing laws against prostitution.

The promise of sexual favors in return for procuring drugs (or participation in some other crime) is a different type of temptation from the promise of cash. We can imagine a continuum moving from more to less acceptable temptations. Temptations as public policy tools can be ranked by their ethical desirablity (or perhaps, more precisely, by their degree of undesirablity). One factor is of course how realistic or extreme the temptation is. Temptations should stay close to real-world settings. Also important is the degree of freedom of choice and absence of compulsion. In that regard using a physiological and/or psychological need such as sex seems less acceptable than the offer of cash. An extreme example is that of gaining the compliance of an addict experiencing withdrawal symptoms in an undercover scheme to buy or sell drugs. While the situation in which sex is employed generally do not go to this extreme, using it in this way seems sleazy and exploitative because it draws upon an individualís vulnerability.

The above cases contrast with cases of the facilitating serpent. Sexual involvement is not directly tied to a crime. It is not bait held out in reward for participation in illegal activities. Instead, it is used to help the individual fit into a criminal environment or setting. The sexual behavior adds credibility and may be a means of gaining information. It can also be a self-protective device since not to become involved (whether with sex or drugs) would cast suspicion on the agent. The offering of a woman can be a test and may create a dilemma for the agent. One agent notes: "Avoiding sexual compromise is no easy task when working undercover. It is, in fact, one of the job's greatest challenges ."

This second form certainly raises ethical issues, but they are ancillary to the criminal behavior. The violation of trust is a price paid to obtain, or at least not to block, obtaining some other end. This is not as abhorrent as using sex to tempt the target into illegality. It is also less morally unacceptable if the agent is a passive respondent rather than the initiator.

In contrast to the Georgia case, in neither the case of the failed romance to gain a confession nor the case of the search for the Weather Underground fugitives was sex used to further the commission of a crime. Nor were agents co-participants in a crime. Instead, a duplicitous relationship was established in order to solve a serious crime, and there was reason to choose the targets in question. These cases are also more acceptable than the infiltration of the Los Angeles Maoist group described above, which was undertaken for diffuse intelligence gathering rather than in response to a particular crime that had already occurred. In contrasting the last three cases, does the successful confession in the murder case mean that use of the tactic was more acceptable than in the Weather Underground case, where the fugitives were not found? If the Los Angeles case had discovered a plot to blow up City Hall, the agent would likely have been a hero. But he apparently found nothing. It is tempting to make the end result the criterion for judgment. Yet such a criterion conflicts with the important principle that means have a moral component, apart from ends.

If, in pursuing the broader social good, harm is done to individuals (for example, in the cases of the duped pregnant woman and the romantic partner in Los Angeles) should they be entitled to compensation as a result of being innocent third-party victims harmed by an undercover investigation?

Given the potential of intimate relationships for gathenng intelligence and evidence and our conventional assumptions about appropriate uses of sexuallty, should agents be permitted, expected, or required to use sex if it willl further an investigation? This involves the ethical issue of what can legitimately be asked of agents in situations where sexual deception may be justified. Given the unique character of such assignments they should certainly be voluntary. This is likely to be affected by gender, with men more likely to volunteer.

There is certainly resistance to this, but it is not universal. Thus in a novel drawing from her undercover experience, Kim Wozencraft recalls a conversation with her chief regarding evidence gathering about a reputed crime leader, "Iíll bet heíd give you anything you asked for." "Chief," I said, "you can get someone else to screw him, if thatís what youíre asking me to do. My job description says nothing about being a hooker." This contrasts with a more purely fictional account of a policewomenís eager infiltration of a radical group and the forming of a relationship with its leader.

Let us return to the Barry case with which we began. An elected official who is believed flagrantly to have violated the standards he is charged with enforcing communicates cynicism and hypocrisy. The behavior of leaders is not only instrumental; it is also educational and symbolic. But when law enforcement resorts to trickery and exploits intimate relations to make a misdmeanor arrest as in the Barry case, it in turn communicates the idea that the end justifies the means. This may hold in popular attitudes, but it is not the message of the law in its most exalted form. There is also communication when the suspect (as in this case) is from a minoritv group and the enforcers are not, and the case can be publicy perceived as a witch hunt. In the well-honored words of Justice Holmes, "For mv part I think it a less evil that some criminals should escape than that the Government should play an ignoble part."

To apply the criteria suggested here, the use of sex in the Barry case is ethically undesirable because it involves violation of psychological and sexual intimacy and was initiated by the state for an offense unrelated to statutes governing sex.

However, I do not suggest that there should be rigid rules against ever exploiting sex and friendship in an investigation.There is need for flexibility and appreciation of the great variety in law enforcement situations. There are clear costs when the state introduces mistrust and suspiciousness into intimate relationships. But never to do this may also have costs. I don't think categorical prohibition is appropriate, but neither is unlimited permission. Instead, each case must be examined separately in light of questions touching the key issues.

Local departments should follow federal agencies in establishing guidelines and restrictions on the instrumental use of intimate sexual relations. In extreme cases an individual's constitutional rights may be violated. The courts may eventually offer clearer standards. Circuit courts in the United States have refused to dismiss charges when an informant had a sexual relationship with a defendant. However a different standard may apply when a sworn agent is involved.

The U.S. Court of Appeals for the Second Circuit recently set forth criteria for establishing when sexual relations between a government agent and a defendant constitute "outrageous conduct that would require dismissal of an indictment. To rise to a level of a constitutional violation it must be shown that (1) the government "set out to use sex as a weapon in its investigatory arsenal" (2) a government agent initiated the sex or allowed it to continue to achieve a government goal, and (3) the sexual relationship occurred during or close to the period covered by the indictment and "was entwined with the events charged therein."

In considering the ethics of sex as a part of an undercover investigation, I would take the same approach as I took to the general question of whether undercover tactics were ethical.The answer depends on responses to a series of questions about the specific case. I argue that the greater the affirmative or "harm-avoiding" answers, the more justified undercover means are. The questions involve seriousness, alternatives, democratic decision making, consistency with the spirit of the law, prosecution as a goal, clarity of definition, crime occurrence, grounds for suspicion, prevention, autonomy, degree of deception, bad lessons, privacy and expression, collateral harm, equitable target selection, realism, and relevance of the charges. Our assessment should also be conditioned by the goal of the investigation, the cost of taking no action, the degree of intrusion and the nature of the betrayal, the skill of the operatives, and the likelihood that the deception will be publicly judged in court. But in general there should be a strong presumption against trading in the currency of intimate relations, no matter how noble the goal.

Back to Main Page  |  Top

You are visitor number  to this page since February 16, 2000.