from china with love
homeparlor maidchina spyingcounterintelligence
interview: t. van magers
photo of van magers

Magers was a special agent in the FBI from 1969 to 2002, during which time he specialized in Chinese counterintelligence and got to know both J.J. Smith and Bill Cleveland. In this interview, he describes the evolution of Chinese espionage over this period and recounts some of the FBI's major investigations into China's espionage program, including his own role in the Larry Wu-Tai Chin case. Magers tells FRONTLINE that among the lessons learned from the Parlor Maid scandal is that counterintelligence agents need to be careful in how they handle assets and authenticate the information they've provided. "You have to evaluate how much you're willing to compromise in order to get the information you need," he warns. This interview was conducted on June 4, 2003.

[Take me] into the world of Chinese counterintelligence in the FBI at the time you joined up with that crowd. Who were the people, how big is the unit, what is the orientation, etc.?

I began working Chinese counterintelligence in 1971 before the People's Republic of China was admitted to the United Nations, when the FBI's China program consisted primary of interviewing ship jumpers or Chinese aliens. … [T]here really wasn't much of a program until "ping pong diplomacy" in 1971, the vote in the United Nations in that fall, and the arrival of the first delegation to the U.N. in October/November, 1971.

We don't have many cases in which we have convictions. You just don't in counterespionage period, and particularly in Chinese...

From that day on, the focus changed. We had a small group of people working in those offices with China, a presence with the Chinese communities. After that, we focused on persons from the People's Republic of China -- diplomatic personnel, visitors from China after normalization in 1979, Chinese students, delegation members and the like. The whole purpose was to identify Chinese intelligence activities in the United States, and identify Chinese efforts to influence the ethnic communities in the United States.

Influence, meaning?

I'll do it by analogy. [We talked regularly to] students from the People's Republic who came to the U.S. as exchange students. … We would identify people based on perhaps family, perhaps their subject area. Perhaps other students on campus told us that they were doing something that they found suspicious.

We told them that we're FBI. We're here to protect the national security of the United States. While you're here you have certain rights, and if you feel that someone from the Chinese consulate or the embassy is trying to abridge those rights, you can come to us, and we'll do our best to prevent it. We'll protect your identity if necessary. It was such a user-friendly situation that we later heard from several sources that the Chinese government told their students before they came to the United States that they would probably be interviewed by the FBI but don't worry about it, because we're not like the KGB. We won't hurt them, we'll just talk to them, and you can talk without fear, without concern. …

Now in dealing with persons who grew up in a Chinese background, you don't expect the first visit to be anything more than sipping tea and getting to know each other. You probably don't expect the first dozen [meetings] or so. But you build up, over a period of time, at least a respect for each other, and when something happens, you hope that you'll be called.

After Tiananmen in 1989, we had calls from literally more than hundreds -- I don't know the number, and I don't know that I could use it if I did -- but maybe a few thousand people we had talked to who called back to the FBI, or friends of those people, who said, "Something happened, and I'm worried about it." "A Chinese diplomat tried to get me to stop speaking out against the Tiananmen crackdown." "They wanted me to disassociate myself from the students for democracy." Those persons would not have talked to us if we had not developed a level of respect and understanding with them.

Actually, an awful lot of what we were doing was trying to sell ourselves over a long period of time, so people would trust us.

This does not sound at all like FBI standard operating procedure, at least in the movies. This is not what one would expect the behavior of FBI agents to be.

And it isn't. … It was a group of iconoclasts. One, you had to be willing to do your FBI career in something other than kicking in doors and arresting bad guys. We have not had a lot of arrests in this field. We don't have a lot of arrests in foreign counterintelligence anyway, because often that's not the best way to resolve an issue. You have to go through a period of time and accept that you're probably not going to be involved in an arrest. If you are, you may have one or two in a full career.

You have to know that many of your cases will reach unsuccessful conclusions, from the perspective that you may think you have enough information, but you can't prove a case sufficiently for the Internal Security section of the Department of Justice to [be] willing to prosecute. They want an absolutely successful case before you start, before they're willing to take something to prosecution, because you don't want the precedent of the government bringing cases that aren't successful. We've had one case that we've lost that we've ever brought to prosecution, to my recollection.

It has to be a committed group of people. It's a culture that's different from what most of us grew up in, a Judeo-Christian background. You have to understand a little bit about the Chinese people, Asian people. You have to be willing to define your own successes. That small group of people stayed together for a number of years, and made us successful at it.

… Help me understand where Chinese counterintelligence is on [the FBI's organizational] chart.

… It's obvious that the function that the national security division, the intelligence division, whatever, performs is different from general criminal matters, and even from the international terrorism division. And it's different because the goals are somewhat different. We're trying to stop intelligence activities, which doesn't necessarily mean prosecute bad guys.

You have to be prepared to prosecute. You have to operate your case so that you can if you need to. But in your heart, you know that you're probably not going to, because we so seldom have the opportunity when everything comes together -- information that's available to declassify to use in a prosecutable case, foreign policy considerations. There are just so many people who have a different part, a different perspective of what we're doing, that it's difficult to have everyone agree that we should prosecute.

Because there aren't the prosecutions, I think at times, foreign counterintelligence is perceived by those people who are criminal investigators as Sleepy Hollow, and the unwanted stepchild of that is Chinese, because Chinese is different from everything else. …

When I talked to I.C. Smith, he was very good about how he measured the success of an agent by the agent's sources -- the quality of the agent's sources, how many one had.

Not everyone develops sources, and you understand that. You have to use your resources well. When I was a supervisor, I didn't want to have a specific [rule that] you have to have a certain number of sources. There was no specific rule about that. I did it the other way. If you had three or more sources, I looked at them during file reviews. If you had just two, I might ask you to submit me a monthly, a memo on what you were doing to enhance your asset base. If you had one, it might be every two weeks. If you had none, I wanted a weekly explanation of what you were doing.

FBI agents are like anyone else. If you make it easier for them to do something one way, they'll do it that way. So it was easier to have the sources than to have to write the memos explaining why you don't have the sources. That sometimes creates sources for the sake of sources. But I remember a fine young agent who was responsible for keeping track of travel of persons we were interested in, and he developed assets with airline companies. In an inspection, we had an inspector who was a friend of mine who said, "Why, this is no good. You don't have good sources." This person actually had a couple of good sources, and we had the airline people. And that inspector, when he went back to his field office, had a suspected intelligence officer visiting. They lost him, and one of the agents on his squad, [who] had heard his comments about the worthless airline sources, called us. We had one of the sources go into the database and found out where their missing spy was.

So sometimes what you do just to meet numerical goals or guidelines maybe, not quotas, can help you in the long run, because you have those people in place that you don't have to develop each time a case arises. If you have somebody you can pick up the phone and talk to and say, "I need this," and they know you, they trust you, and they know you're not going to compromise them or their identity, then they're able to tell you right [away] or call you right back with an answer.

It's absolutely necessary to develop sources to be a good agent. The problem that you have [is] in developing a source, you have to give a certain amount of information away in order to get information. In that case I was talking about, we had to give the name away, and obviously this person knows that the FBI is interested in so and so from China. Well, one could argue that that fact is classified, because an FBI-instigated interest in a specific person would indicate knowledge of that person's activities, and the fact that that person may be engaged in illegal activities. But if you don't give the name of the person you're interested in, you're not going to get the information back.

Operating a source always is going to be a value judgment of how much you give the informant in order to get your results from the informant. You try to make sure you give less than you get. You hope for complete control from the source, where the source does exactly what you ask them to do -- nothing more, nothing less -- and gets the information you want, brings it back to you, and tells no one else. My experience is that … no source runs perfectly. Even those, even if you ran your brother-in-law as a source, there would be some issues involved. I don't care who it is; there are going to be issues with a source. Your effort is to get as much information as you can and minimize those problems.

How do you turn a source? How do you make somebody a source?

You have to give them a reason to do what they're doing. What you want is to give them a reason that one, you buy, and two, that they buy. You don't come up to an official for a foreign government and say, "I want you to betray your country and give the United States information. We'll pay you for it, you dirty so-and-so." It doesn't work that way.

You have to say, and if you don't believe it -- I hope you believe it, but if you don't believe it, you have to at least be convincing about it -- say, "I need your information. The United States needs your information. The United States needs to deal with your country. We need to deal effectively, and the only way we can do it is if we know what your country thinks, if we can develop a foreign policy that works well with your country, if we have your cooperation in learning what your government plans to do."

Now obviously, there are other times. If you're dealing with intelligence activities, you say, "We know about these activities going on. We need you to help us, so that those activities don't get in the way with the United States' relations with your country."…

With the Chinese, for example, one of the issues is that it takes a long time to develop that relationship to the level of trust where you're going to be able to do that. I know cases where we were convinced the persons were intelligence officers, but they never told us. But they gave us good information, and we never asked the handling agent to confirm whether or not the person was with a particular group of Chinese intelligence. It just wasn't important, as long as they were giving us the information. Later on in the relationship, we were able to ask that question and get the answers.

Who is a Chinese spy? Who are you looking for when you're in Chinese counterintelligence? Are they different than, for example, Russian spies?

Absolutely, and that's one of the difficulties in dealing with Chinese counterintelligence. Against most countries, you identify intelligence officers, you determine the agency for whom they work, you know what sort of information they will seek based on their employment, and you go from there, developing a technique to counter them.

With Chinese, you have known intelligence officers who don't engage in intelligence activities; you have known intelligence officers who do. You have persons who have no connection with Chinese intelligence -- I mean, not even a peripheral one -- who engage in intelligence collection.

You have intelligence officers in the country who are here to make money. The Chinese are bureaucratic, as I guess most governments can be, and there is a level of approval for expenditures that's really relatively low. You have to go quite high to approve spending in even the most unimportant area.

However, if you have someone overseas who's making money, then the level of approval of that operation overseas making money can approve spending the money that you make. So you immediately increase your budget substantially if you have moneymaking operations abroad.

So we see intelligence personnel abroad, and all we see them doing is making money. At first we couldn't understand that. Later, we realized they were making money and that was the goal, so that other personnel would have money to spend on operations.

Our standard statement is Chinese is different, and it is. Because it is different, it's made it a challenge for us to be successful. Ultimately we weren't successful in identifying a lot of their activities and a lot of their targets, but still, because of the difference, it's often difficult to explain and justify budgets for continuing operations against the Chinese.

In the 1970s, 1980s and 1990s, did what they were looking for in the United States of America change?

Sure. Initially, when the first group got to the United Nations -- and this is one person's perspective -- they were interested in leading the non-aligned nations group in cultivating ethnic Chinese to win the hearts and minds of the Chinese people vis-à-vis Taiwan, and to get general information.

I remember when the Chinese first came to the mission to the U.N. Even before the liaison office in Washington, they obtained the entire printed output, every document that NASA put out each year; how much it was to upgrade their system, or to know what our system was, I don't know. But we knew what they got in just public source material.

An awful lot of what they did then was public source material, where they would meet with U.S. politicians or U.S. reporters, wine them, dine them, and talk to them about what was happening. This was major information. They considered it the most sensitive information again, and classified it very highly and sent back. It's the stuff that you read in the Op-Ed page of the Washington Post or The New York Times each day. You could read political columnists. If it were today's environment, you would watch Fox News, or CNN, or PBS, and you would listen to the talking heads. That's what the Chinese got, and that was important to them.

For us, we wouldn't have turned around for it. But they reported it, and were given high marks by their senior officials in Beijing for getting that information.

Then the arc of information gathering -- take me to the 1990s. What are they after by the 1990s?

When I left the field in the early 1990s, there were so many different groups of people in the United States from China. There were still political people who were interested in that. There were persons in the embassy and the consulates who were responsible for keeping track of ethnic Chinese in the United States. There were those responsible for students, and they wanted to make sure that they toed the party line, particularly in the aftermath of Tiananmen. There were commercial officials in the U.S. running companies, some legitimate, some engaging in business activities but with other purposes, such as to make money for intelligence; some, it seems, as loss leaders, just for raising the Chinese profile. …

One of the things about countering Chinese intelligence activities that is such a challenge is that again, the persons who collect intelligence, including classified information, often are not traditionally trained spies. If you have someone from an institute in Shenyang that makes cruise missiles, and that person comes to the United States and develops information that his factory might use, an institute might sponsor a student to come to the United States. The purpose is to get more information to come back and to enhance that facility's capacity. …

[Tell me about the Larry Chin case.]

There was information that the U.S. government had been penetrated. …

We didn't have much information about [the spy], except that he might have attended a party on a certain night in China, had been given some sort of award, and that he might have come over on a Pan Am flight on a certain date. The problem was that Pan Am didn't fly that day, and there were all sorts of difficulties in taking that little information and trying to pair it with someone in the intelligence community. It was traditional, standard, very hard investigating work, where the two guys primarily on the case looked through airline records, and came up with the theory that CAAC, a China international carrier at the time, used Pan Am facilities in New York, so maybe Pan Am used CAAC facilities in China. Perhaps it was the other airline.

Of course if it's CAAC, you can't exactly go to a Chinese flag carrier and ask for travel information about their passengers to investigate Chinese espionage activities. So ultimately we had to work backwards, look at customs declarations. Obviously it's not going to be immigration if it's a U.S. government employee who has top-secret clearance.

After looking through microfiche files of those records for a period of weeks, we came up with one flight that would have fit, and on it was Larry Chin. …

The problem is that you have the name; you then have to prove the case.

There was some superb investigation in the Washington field office of the case, where court-ordered electronic surveillances came out of court. There were various special techniques. We watched him for a long period of time, and we really wanted to do this right. It was after [Edward Lee] Howard. This was the first prosecution for espionage against the United States from someone representing China, and it was the longest case of active espionage against the United States in history, to the best of my knowledge. We had to do it right.

We had a Chinese-speaking agent who had handled a lot of the monitoring of some of the electronic surveillances, who represented Larry Chin in some interviews ahead of time, where we tried to go over techniques that we would use. After the interview of Larry Chin, one of the guys said, "This was easy. The agent was a whole lot harder to interview than Larry Chin was." But even then there are things that could have caused it to go wrong.

During the course of the interview, Chin said, "I wonder if I should talk to my attorney." One of the persons said, "I'm an attorney, you can talk to me." But another guy who was actually managing the program then for the FBI said, "You ought to stop and think about it. If you want to talk to an attorney, you ought to stop talking now. You can talk to an attorney, and the first thing the attorney is going to tell you is not to talk to us. So if you think you want to talk to an attorney, you ought to stop talking to us now." He said, "But when you do that, just remember if you were the attorney general, would you prefer to cooperate with somebody who has been cooperative with the FBI, or somebody who wouldn't tell the FBI anything?"

Amazingly, in the suppression hearing, Chin admitted exactly what had been said. As long as he told the truth, we were OK, and he did tell the truth in that. So the confession was admitted. …

Chin was a bit of a pack rat. He kept diaries. He kept records of money paid. He kept even a key for one of the hotel rooms he used in Beijing. The information obtained through the search warrant, I think, would have led to a conviction, even if he hadn't, even if the confession had been suppressed, which sure made it easier. It made it a five-day trial. …

How important was it to get the conviction of Larry Chin?

No question at all that that was a seminal event in Chinese counterintelligence, not just in the United States, but throughout the English-speaking world. We periodically met with our counterparts in Canada, Australia, New Zealand, the United Kingdom, to discuss similar experiences. We were able to use that case to justify our existence through Tiananmen in 1989. So it was good for at least four years in justifying budgetary expenses. …

[Did the conviction] really depend on that interview working out, and him kind of giving it up?

We might have had authority to arrest him after the search warrant had been executed, if he hadn't volunteered. But to make it a slam dunk, it had to be a successful interview. I suppose that's the case with any investigation, but certainly with counterintelligence. It's hard to have forensic evidence of espionage, unless you have somebody who clears, drops, or you get from a source abroad the person's fingerprints on material that's passed, or something like that. You can look at information that was lost and see who had access to that information, and have a good idea that Joe Jones is guilty, because Joe was the only one who had all this information that went over. But it's still not a provable case.

Statutorily, espionage is a hard offense to prove, and it's meant to be, because it's a very serious crime. It shouldn't be something that's bandied about the way some other countries do. The Chinese have prosecuted people for espionage for giving crime statistics to foreign reporters. That's not what our country is about. We make it a serious crime intentionally. But that means if you're a counterintelligence officer, it's hard to assemble the elements of the crime sufficiently that you can be confident beforehand that there is no doubt you're going to get a conviction. …

Tell me about Bill Cleveland.

I should of course tell you about Bill. Bill is one of the heroes of the Chinese counterintelligence program. I talked about a number of cadres who had been there from the beginning, working with the Chinese targets. To be successful, it helps to have an interest in China. It helps to have some language proficiency. Bill had both. It helps to be a student of intelligence, to be willing to do something more than just nine to five. You have to care about the work when you go home at night. Bill always was that. Bill cared about what he did.

I think if you talked to every person who has worked with the Chinese for any length of time for the last 25 years and asked them to list five people who were important to the program, Bill would be on almost every list. He was that respected. I consider him a close personal friend.

What was he good at?

He was good at interviewing. He was good at developing relationships with people. Often in the bureau, you have people who develop assets and people who make cases. They're often not the same people, because some of the requirements are different. Bill was gifted at both, and he also handled things well administratively. So dealing with someone like that, when you had a representation that there was a good case, you knew it was a good case. I just don't have enough good words to say about Bill Cleveland, whom I did respect, and whom I do respect. …

One of the things that [Cleveland] does is he gets [suspected spy Gwo-Bao Min] into the famous interview moment, where he essentially has to get the guy to give it up.

Yes. And if you don't give it up, it's often hard to make the case. From the perspective of prosecution, the case was a failure, because we didn't prosecute. But if you use that as the standard in Chinese, we're almost always failures. If you use instead the perspective that a person who, at the very least, has misused classified information -- and I think all of the reports would agree with that -- that person loses his clearance, and then is denied access to classified information. That's a success. If you look from the perspective that you learn how sensitive U.S. facilities are targeted in this one case, and to the extent that you can extrapolate from the case, you learn how to try to prevent that from happening in the future, then it's a success.

By just about every standard but prosecution, it was a success. I know Bill was disappointed, because I know Bill's work ethic and his expectation of success for him and for the FBI, and it didn't meet that level. But by a reasonable level, it was a success.

What did [the FBI suspect] Gwo-Bao Min want?

I don't know how much of that is unclassified. He worked in, he had access to weapons information at Lawrence Livermore. …

Then there were other cases following that where we identified National Laboratories as being a potential target of Chinese intelligence collection efforts; not necessarily intelligence service collection efforts, though some were, but intelligence collection efforts.

So this is a new territory being mined by the Chinese in the 1980s?

Well, it was new to us. How new it was to them? It's hard to say. I remember when I first joined, [I] attended Chinese seminars in the early 1970s. A friend of mine -- I still consider him a friend -- was in the unit at the time. He said, "The FBI is the world's greatest counterintelligence agency, and we've never detected Chinese intelligence activity in the United States. Therefore, they don't exist."

Well, unfortunately at that time, Larry Chin had been working for the Chinese for what, 25 years. You know, it's not easy to prove a negative. What I'm saying is that we saw that. Tiger Trap represented the first case that I recall where we could definitively say there was targeting of National Laboratories. Probably it happened before, perhaps not to that extent, certainly not to the extent that we were able ever to prove it. But at least we saw it then. …

Also at some time around [the Tiger Trap investigation], the Peter Lee case gets moving down the runway as well. Is Peter Lee in your mind more of the same?

The active part of Peter Lee came after I left the program. … But yes, it's the same basic thing. Certainly I have read about it since then. Again, it's a case where we would have liked to have had a more meaningful sentence, but we got a conviction. As I said earlier, we don't have many cases in which we have convictions. You just don't in counterespionage period, and particularly in Chinese, it's hard to make a case.

In dealing with the Chinese, we've had any number of persons the Chinese have dealt with who we've later talked to, many of whom have been in positions with great access. The Chinese tend not to give specific instructions, in most cases, about what information should be gleaned. We have a person who's in the FBI. If I were Foreign Service, I'd want to know who their assets are targeted against China. I'd want to know what cases are in the works. The Chinese response would probably be, "You give us what you think we need."

Well, try proving that in court as a case of Chinese espionage. They tend to be non-specific, and I guess if you accept their position as being willing to send a lot of sources out to get small bits of information, and then to reassemble it back in China, it works well. If you have a relatively large number of sources, and you're interested in running them for a long time, and you're not willing to compromise them for short-term gain, then you do it the way the Chinese do it.

If, on the other hand, you want immediate gratification, you want the hard facts, you want them now, you want specific information, you want it sooner rather than later, then you develop the sort of facts that the FBI can prosecute, or the FBI can develop information about so the Department of Justice can prosecute them. And the Chinese don't work that way. As a result, countering them is so much more difficult, because you don't have those hard-edged intelligence officers who look guilty on the stand.

I remember one case in which a person engaging in intelligence activities had his kids in the room with him when he was meeting with someone. You can't imagine how unattractive it would be to present in court espionage in the background and two kids watching TV in the foreground. You're not going to make a case that's going to be binding. This is the bad guy you want to send away for life? It's a hard target, because of the way they do what they do -- the fact that it's so different from the others, and the fact that there isn't, for the most part, the central direction and control that leads to the specificity you need to get the elements of the client. …

Tell me about J.J. [Smith].

… J.J. was certainly one of the two or three really important persons in Los Angeles. He and another agent who later went to Albany handled most of the important cases in Los Angeles for a number of years. J.J. was competent, conscientious. Certainly in retrospect, you would assume that he's not maybe the most administratively pure agent, but he always tried hard to get the job done.

Before this, I considered J.J.'s career to have been an outstanding career that brought all sorts of successes to the program. In reading the affidavit, it would appear that there is nothing in the affidavit that indicates that J.J. was guilty of a sin of commission, that there was guilty intent in anything that was done, and I hope that's correct, absolutely. …

So J.J. sits out there, kind of a different kind of a guy, working Chinese CI in the LA office, and he develops Katrina Leung. Do you know about her from your perch in Washington? Do you know her specifically? Did you know her as Parlor Maid? Did you know her as anything?

I certainly knew her as Parlor Maid. I made it a habit not to learn true names. It was too easy in my position if I knew true names of a lot of the assets to say them. You just don't want to run the risk of compromising an investigation that is going to hurt a person, much less field operations. I knew who Parlor Maid was in general. …

How important was Parlor Maid?

Parlor Maid was important. Parlor Maid provided information that we considered good information.

Like what?

I remember one time Parlor Maid met an official who said, "When you get back, give this information to the president of the United States." Parlor Maid said, "I don't meet with the president of the United States, I meet with the FBI." And the official said, "Give it to the FBI, they'll introduce you to the president." My response at headquarters was, "When she's introduced, can I go along? I'd like to meet the president, too."

I'm not sure the Chinese understand how the U.S. government operates. One, she doesn't go nearly as far in the U.S. as in China. She had [guanxi], and that was important. She had developed relationships with a lot of officials, and reported on those relationships.

Important and vital information, or?

Again, I'm not sure how much I can say. But I considered Parlor Maid to be an important resource of the FBI.

One of the best?

Sure. You know, define "is." Certainly she was probably the best in Los Angeles. …

How do you, in your capacity at headquarters, know whether somebody like Parlor Maid is a double agent? Or do you just automatically assume she probably is working both sides of the street?

… You have got to give information in order to get information. Your ideal is to give next to no information to get a lot. Obviously, the worst thing is when you're giving a lot and get none. Generally you're somewhere toward the middle, with the emphasis on getting more than you give.

According to the presidential executive order, if you have information that reveals methods and sources, it's classified. The fact that we have an investigative interest in Joe Wong, pick a name, could be classified. … But if you're asking an asset to provide information about Joe Wong, then the asset knows that Joe is of investigative interest, so you have classified information. You just try to limit it.

Controllability is an issue. How much are you convinced that you're able to know everything the asset does, and be sure that the asset does exactly what you say -- no more, no less? Not many assets come under that. Parlor Maid certainly didn't.

Certainly didn't?

Did not.

In what sense?

When she meets as many people as she does, there is no way you could direct and control whom she meets, what she asks them, what she gets from them. She's developing guanxi. She's developing her network of friends. That certainly doesn't hurt her business, but it's an advantage to the FBI as well. …

So in this relationship, J.J. to Katrina -- I'm not talking about the sexual relationship, I'm talking about the source/agent relationship -- what is the process? How is he working her? How is he supposed to work her? How does he actually work her?

...Obviously, the court case will determine how it actually happened. You identify a person like Parlor Maid who has access to persons of interest, as in Chinese officials, Chinese diplomats, Chinese business persons. You know that that person is going to have information. You try to develop a relationship with that asset, so that the asset will give you the information you want.

You tell them, "The XYZ company is coming to Los Angeles next month. See what you can find out about them." "So-and-so from the consulate has been nosing around certain students, or certain U.S. businesses. See if you can get to know him and give us a read on whether he's legitimate or not."

What you want to do is use the person's access, perhaps expand it if possible, to get you the information to identify and neutralize the spies. That's sort of a puerile statement, but in general, that's your organizational message. That's what you're trying to do. You try to use the asset in order to do it.

How much did a good source like Parlor Maid make somebody like J.J. bulletproof in the FBI?

Depending on the level of information that a source has, depending on the value of the case and our periodic inspection, you're more likely to overlook administrative imperfections if the end results are there. If you have the world's greatest asset in a certain field, you're not going to say, "Yes, but you were two weeks late in the renewal process." Well, of course you were two weeks late. You were processing all of this information that the asset provided. "Bulletproof" is too strong a word. …

Is the J.J./Katrina story unique in your experience, the love affair, the whole thing?

… Fortunately, it's not something that happens often, but one time is too many. You develop a vested interest in your sources, because sources are so important to successful operations. You want your source to be good. You know, maybe that's an incentive not to look too closely. I would hope that we do, but that's one reason why you're supposed to have independent people do the evaluation. You don't evaluate your source yourself, because you can't look at it in an unprejudiced manner. You have to have somebody look who doesn't have a bias, and say, "Look, we've got some issues here. I read the reports that you've provided of what the source has given us. I read the approval for what you've given the source, the passage information," if it's that. "It doesn't seem to me that we're getting value received for information spent."

But that was never the case with--

Not ever, not with Parlor Maid. It's difficult not to have a relationship with your source. Having it go into a sexual relationship seldom happens. More often than not, it would be a situation where it's financial where you're interested. You want to help the source. …

It's a pretty hard-hearted person who deals as closely with a source as you have to have a successful source, particularly in Chinese, who doesn't develop a caring relationship, and as a result, be at least somewhat susceptible to being manipulated if that's not reciprocated. If I ran organized crime, I might find it a little bit easier not to identify with my sources. But in Chinese work, where there is no quick response to a source, where you have a long, developing relationship before it's successful, you're going to care about the person. Because the success of the source is directly related to your own success, you're going to hope that the source is successful. …

[How do you think the aftermath of the Wen Ho Lee and Parlor Maid investigations will affect the FBI's China counterintelligence program?]

The one thing we could have done is more easily justify the [Wen Ho Lee] investigation. It would have been easy to point out that it was not a singling out of a particular ethnic group. That's one of the problems we had with the Chinese. The Chinese intelligence services, and those Chinese collectors who were not members of the intelligence services, tend to focus on ethnic Chinese. [The] exception is military intelligence, because they're dealing with other U.S. military, or foreign military. … So dealing with the Chinese community in the aftermath of Wen Ho Lee, in the aftermath of something like Parlor Maid, it's going to be difficult.

I think, at one time, the FBI had the best program in the world. Obviously, that's a personal perspective, and it's from my FBI viewpoint, where in another agency I might not think that. But I did, and I thought we had successes that were unique in the world. We downsized our mandatory retirement at 57, cycled the number of people out. Others retired under other circumstances.

In the aftermath of Chinese campaign contributions, Wen Ho Lee, this case, I'm concerned that we may be sowing the ground with enough salt that nothing will ever grow there again. Never say never, but it's going to be much more difficult for those who follow this than it was for me in 1972, going in basically unplowed fields, and trying to be successful.

And the implications of that?

Well, if as some people suggest, China is going to be the successor to the Soviet Union as the major geopolitical opponent -- perhaps [opponent] is too strong a word, or opposition, I don't know, maybe that's the same word -- if it's going to be U.S. versus China in the world by 2050 as some people suggest, then we will have lost a lot of our resources to be able to track their activities. …

Is there a feeling, though, that sometimes as good a job as you can do on the ground on these things, there is another level of writers, politics, there are diplomatic realities--

Sure. With China, there is always the diplomatic reality. You've heard it before, but our relations with China -- and this goes back from 1971 on -- are always in either two stages; either they're good and we shouldn't do anything to mess them up, or they're bad and we shouldn't do anything to make them worse; which means, basically, we shouldn't do anything. Except that if you don't have the information, then you've really not done anything, and you've compromised the security of the United States. For years, when I was a street agent, I heard that, and we still went about our business. You do it because you care. You do it because you have a commitment to something [and] you think that is a righteous end. …

So often you see old retired guys talking about how much better it was "when I was young." I don't want to make that mistake. But the people in the field now have to overcome the perception -- in most cases, I think the misperception -- of how the U.S. intelligence community has operated in the ethnic Chinese environment.

The Chinese traditionally have an ethnically focused operation. For us to respond in kind is simply not possible in the United States. You can't assume because of a person's particular ethnic group that they're going to be more susceptible to recruitment. I think the vast majority of Chinese-Americans are just as loyal as any other ethnic group. Other countries have worked in their own ethnic communities, just none to the extent as the Chinese, and none with the exclusivity of the Chinese.

So how does Parlor Maid affect that?

Again, it looks like we're picking on the Chinese. It's certainly not to the extent of the complaints about ethnic focus during Wen Ho Lee or during Chinese campaign contributions. But added to that, it reinforces a misperception, and I do believe it's a misperception. …

One of the things we always read in the paper is that all of the cases for 30 years, your life's work, [are] now under evaluation because of Parlor Maid. Who knows what's been compromised? Has everything been compromised? What's your assessment of that?

Two things. One, the bureau has to make that assumption, just as we did after the Hanssen case, just as the agency did after [Aldrich] Ames, or just as we did after Ames. When you have someone who may have had access and may have provided information, you try to develop a list of things to which the person would have had access, and to see if anything happened in the case that would [indicate] it was compromised. You might be unwilling to send persons back in harm's way. In other words, you don't want to send assets back who might be arrested, unless they're under a diplomatic passport or something like that.

You probably don't want to send Chinese permanent resident aliens who have been cooperating with us if you think they've compromised with that, because they're not U.S. citizens and don't have rights to protection in China that a U.S. citizen would. We'd be foolish not to make every effort to find out if we can determine if there has been any compromise.

On the other hand, do I think there has been? You know what? I don't know what the information is to say that. I don't know what the bureau has. If we have information that indicates that she hemorrhaged all the stuff she was supposed to have had and a whole bunch more, then yes, we've got stuff to worry about. You'd have to determine what Los Angeles knew specifically, and what in general might have been learned in seminars.

But to get all of that information, you have to assume that there was someone who was a willing participant along the way. I certainly don't know whether J.J. was. I hope he wasn't.

It's one of those things that, even if nothing was actually compromised, the fear of the compromise has set the program back.

Absolutely, and the fact that you have to be careful about certain sources -- for example, if you think they might have been compromised -- about sending them to China, about dealing with the Chinese.

So it's a setback?

Absolutely. There is no way a case like this could not be. It's part of that sowing the fields with salt thing. You've made it difficult to grow a crop in the future.

When you heard of the charges and allegations against J.J., what did you think?

Well, I had been interviewed ahead of time, of course, based on my previous position.

So you knew it was coming?

I knew that there was an investigation. I was surprised when there was an indictment, because again, interviewing somebody in a case like this, you try not to give away too much. The questions were couched in such a way that I saw no reason to believe that there had been significant compromise. Obviously, the fact that the incidents went forward would indicate that someone doesn't share that opinion. And someone who has access to more information than I do doesn't share that opinion.

I read the affidavits and heard about Bill. I characterize Bill as one of the heroes in our program. I was pleased to read throughout the thing that Bill cooperated fully throughout. I take that to mean [he] was probably afforded polygraph tests and passed them, because we wouldn't have made the statement unless that was the case. Then that confirms my opinion of Bill.

The problem is that someone whose career was so exceptional, someone who did as much as he did for the United States government and for the FBI, is likely to be remembered by those people who don't know the true story as a person involved in an espionage case. That hurts, because Bill's contribution over the years has been so great; not just to the FBI, but to the United States as well.

What is the meaning of all of this? Where do we stand now? You've talked about salting the ground and all of that. What are the lessons to be drawn from it?

Part of the lesson is that you have to have good assets in order to be successful. Assets can bring you success, but more often than not they bring you trouble as well. You have to be very careful in handling assets. You have to be able to have some method of objectivity in evaluating [them]. You have to be convinced, or be able to convince someone that your long-range plan -- with the Chinese it has to be long-range -- has the potential, at least, for getting more information than you give up.

With Chinese, you have to be able to run them more than just on a quarterly, or annual, or biannual basis. You have to be in it for the long run. You remember the statement when Chou En-lai was asked what the significance was of the French Revolution, he said, "I don't know, it's too early to tell."

You have to look at it from a different perspective, and we don't share the same cultural background. You have to be willing to understand the background of those with whom you're dealing. You have to accept that, in guanxi, you're going to have persons that may not do things the way you would want them to. But that's the same thing with a criminal informant who, if he's never engaged in any criminal activity, probably isn't going to be a very good criminal informer. You just have to make sure you don't wind up with another Whitey Bulger-type situation. …

In both cases, you have to evaluate how much you're willing to compromise in order to get the information you need. Compromise may mean not prosecuting a criminal informant for some minor thing. But when it comes down to the guy committed murder, hey, I'm sorry. You passed it. You're going to get the full equal justice under law to the ability that we have to make sure he gets it from the Chinese perspective.

You have to understand that a person who does well with Chinese has to have some sort of relationship with his Chinese handler, or with someone else, and that may make it not a black/white issue, but a very gray issue.

So why did we hammer J.J. and Katrina, and, in the end, ruin Bill?

Well, my presumption is that we have more information than has been released. I don't know. I don't know that, because I don't know all the things about the case. I don't need to know, and I guess the public doesn't either. …

The affidavit said that there is, plus other information not included. If the affidavit represents the sum total of the information the FBI has, then I don't know, unless it's the decision of someone senior, let's make an example, and why here, why this case, why now? I don't know.

The sex?

Maybe, maybe. It certainly makes it appeal to prurient interests. It grabs headlines. … Sex with assets has not exactly been a roadway to administrative or career success. Heck, it has not even been a roadway to staying out of jail. This is just not a good idea, and there are other ways to get that message across, I would hope. It has to be more than that. I can't believe that this is all that it is.

Do you think she turned him, or he turned her?

It presupposes that she was directed by anyone. I think either he turned her for operating for the United States, I don't know. One equally possible evaluation, based on the information available, is that she was operating for herself; that she was providing the FBI information she thought that the FBI wanted, that she was providing the Chinese the information she thought the Chinese wanted; that she, in the end, was in for Katrina Leung, for purposes of helping her career, her business.

It is true she was making a lot of money?

Yes. Yet the $1.7 million, or whatever it is, that [is] said [to be] from the FBI is not really representative, because most of that was for expenses. But when she had people over, when she traveled to China, all of those things were done to our benefit -- perhaps to the Chinese as well -- but to her own business's benefit. But if that's ancillary to what's good for us, that's value received for money spent. …


home + introduction + the "parlor maid" case + how china spies + inside fbi counterintelligence
interviews + join the discussion + producer's chat
tapes & transcripts + press reaction + credits + privacy policy
FRONTLINE + wgbh + pbsi

posted january 15, 2004

FRONTLINE is a registered trademark of wgbh educational foundation.
web site copyright 1995-2014 WGBH educational foundation