A Review of the September 2005 Shooting Incident Involving
the Federal Bureau of Investigation and Filiberto Ojeda Ríos

August 2006
Office of the Inspector General

Chapter Seven:
OIG Assessment of the FBI’S Decisions
Regarding Entry of the Residence

In this Chapter, the OIG assesses the FBI’s decision not to enter the Ojeda residence until 12:35 p.m. on September 24, over 18 hours after the agents at the scene reported that they believed Ojeda had been shot. This decision has been criticized because Ojeda bled to death from his bullet wound. Some individuals have alleged that the FBI’s failure to provide medical assistance to Ojeda during this 18-hour period reflected negligence, poor planning, or an intent on the part of the FBI to allow Ojeda to die. The controversy may have been further fueled by the FBI’s failure to make any public statements regarding the operation until 7:00 a.m. on September 24, even though reports that Ojeda had been shot and wounded or killed had begun to circulate widely in the media during the evening of September 23.

The OIG found that the decision by the FBI to delay entry of the residence until September 24 was motivated by considerations of agent safety and not by an intent to let Ojeda die. In the hours after 6:08 p.m. on September 23, when the agents at the scene reported that Ojeda had likely been wounded, the primary concern of the FBI commanders in Puerto Rico was that Ojeda might not have been disabled and that there might be an additional armed subject inside the residence posing a continuing deadly threat to agents who might enter the house. Accordingly, the agents in Puerto Rico took deliberate and reasonable steps in preparation for a “clear” of the residence after dark, when the agents believed they would have a significant tactical advantage. If control of the operation had remained in Puerto Rico, the clear would likely have begun shortly after 8:09 p.m., when electric power to the residence was cut off.

At approximately 8:05 p.m., however, Counterterrorism Division (CTD) AD Hulon in FBI Headquarters informed Fraticelli that any deliberate entry of the residence would require CTD approval. AD Hulon, in consultation with EAD Bald, ultimately decided that no entry would take place until relief units from HRT in Quantico arrived at the scene the next morning. Again, we found that the decision was motivated by concerns for agent safety because of the unsuccessful assault earlier on September 23 and because of uncertainty regarding the continuing threat posed by Ojeda and any other occupants of the house.

We did not find CTD’s decision to assume control over the entry decision improper and believe the decision to postpone the entry until the next day reflected a good-faith balancing of the available information. We also found, on the basis of expert forensic analysis, that by the time CTD assumed control and made the decision not to enter the residence that evening, Ojeda had very likely already died. Therefore, the decisions made by CTD after assuming control at 8:05 p.m. likely had no impact on the outcome of the operation.

However, we did find troubling aspects in the entry decisions regarding the clarity in the chain of command in Puerto Rico and the accuracy of information Bald and Hulon relied upon in making their decisions. We believe the FBI can learn important lessons from this incident that will benefit future operations.

I.   Assessment of FBI Entry Decisions in Puerto Rico from Approximately 6:08 p.m. to 8:05 p.m.

From the time that Ojeda was shot until 8:05 p.m., SAC Fraticelli and HRT Deputy Commander Steve were responsible for decisions relating to whether to enter the residence. We concluded that their decisions during this period were reasonable under the circumstances.

At approximately 6:13 p.m., immediately after Ojeda was hit, Red Squad Supervisor Doug requested permission to conduct a “clear” of the residence. Steve rejected this request because of the uncertainty about Ojeda's condition and whether he was still a viable threat to the HRT agents. In addition, the reporting from the scene following the exchange of gunfire between Ojeda and HRT suggested the possibility that a second armed subject was in the residence. Moreover, Ojeda's wife, Rosado, refused to answer FBI questions when she surrendered about the number of people inside the residence, as did Ojeda when asked by San Juan FBI agents Rodger (the negotiator) and Ron.

Steve, with Fraticelli's apparent consent, elected to proceed slowly and methodically. As described in Chapter Three, this strategy involved making a limited breach of the gated door on the residence to try seeing inside, calling out to Ojeda for a response, assessing alternative entry points to the residence, and assembling an entry team and ensuring it was properly equipped. The strategy was also designed to provide HRT the tactical advantage of operating under cover of darkness by waiting for nighttime and shutting off the electricity to the residence.

The OIG’s experts agreed that a cautious, deliberate approach was the appropriate one, even considering the likelihood that Ojeda was wounded and needed medical care. In their judgment, the agents were not required, legally or ethically, to place themselves at risk by immediately entering a residence to provide medical aid to a person who still could present a threat and who had demonstrated a propensity for violence that day, and in the past. They said that the incremental steps taken by HRT to assess the situation and ensure agent safety were entirely appropriate for the circumstances.97

II.   Differing Accounts Regarding the Chain of Command from 6:08 until 8:05 p.m.

Before assessing the decisions made by CTD after it assumed control of the situation, we comment here on the differing accounts that the OIG received regarding the chain of command during the period from 6:08 p.m., when Ojeda was wounded, until 8:05 p.m., when Hulon told Fraticelli that CTD approval would be required for any entry into the residence.

Our conclusion that San Juan FBI and HRT approached the entry decision cautiously but reasonably credited Fraticelli’s and Steve’s accounts that they were in fact planning to enter Ojeda's residence that night. As we set out in Chapter Five, however, their accounts directly conflicted with HRT Commander Craig's statement to us that he ordered Steve not to enter the residence after the limited breach, which took place at approximately 7:00 p.m. In contrast, Steve told us Craig did not give him any orders at all during their conversations and Fraticelli told us that he thought Steve’s conversations with Craig were limited to providing situation reports. We did not find evidence of any communications between Steve and Fraticelli or between the TOC and the scene suggesting otherwise.98 In effect, what Craig considered at the time, and asserted to us was an order, evidently was a non-factor in San Juan FBI and HRT’s entry decisions between 6:08 and 8:05 p.m.

The conflict between the versions of events provided by Craig and Steve is troubling for several reasons. Apparently, Steve misinterpreted or ignored what his superior considered an order, indicating a significant lapse in either communication or command. Alternatively, Craig did not communicate the order to Steve at all. Just as troubling is the fact that Craig, although Steve's superior, was not the tactical advisor to the SAC for this operation. We do not believe Craig had authority to make tactical decisions at that time. Thus, whatever Craig intended to convey in his discussions with Steve, its proper status was that of a recommendation. Assuming it was in fact made to Steve, the recommendation should have been passed along to Fraticelli for his consideration, and it apparently was not.

We do not believe this circumstance occurred because the FBI's chain of command policies lacked clarity. Chain of command responsibilities were set forth in the FBI’s Crisis Management Program guidelines and were clearly defined in the CONOP for this operation, which provided that Fraticelli was at the top of the chain of command, followed by San Juan FBI ASAC Leslie, HRT Deputy Commander Steve, and Red Squad Supervisor Doug. Therefore, to avoid a repeat of this situation in future operations, we recommend that CIRG and HRT management review and discuss the miscommunication described here and the discrepancy in the chain of command that Craig’s order demonstrates.

III.   Whether the FBI Ignored Bloodstains on the Front Doorstep

One allegation made in the Puerto Rico press was that the FBI knew Ojeda was seriously injured and bleeding because a large amount of blood came out under the front door, making a large stain on the front step and the balcony porch door. Several Puerto Rico newspapers published photographs similar to Figure 11 as support for this allegation. Some individuals suggested that because the agents at the scene would have seen the blood stain on September 23, the FBI’s delay in entering the house must have reflected an intention to let Ojeda bleed to death.

The OIG found no evidence to support this inference. To begin with, the dramatic bloodstain photographs published in the newspapers reflected the appearance of the doorstep on or after the afternoon of September 24, by which time Ojeda’s body had been turned over and pulled out onto the porch to check for explosive devices. (See Figure 11.) Before the body was turned over, the appearance of the doorstep was very different. As shown in Figure 10, there was a thin stream of blood at one corner of the doorstep. A substantial amount of blood gradually accumulated on the floor of the balcony porch.

None of the agents at the scene said they saw a blood stain on the doorstep before the FBI entered the house on September 24. We found the agents’ statements that they did not see a stain to be credible for several reasons. First, as noted above, before Ojeda’s body was moved the stain was not nearly as dramatic as suggested in the newspaper photographs.

Second, many of the agents covering the Ojeda residence after Ojeda was shot were positioned at locations below the elevation of the balcony porch and therefore could not have seen the doorstep or the floor. The agents’ view of the porch also was significantly obscured by heavy foliage. The newspaper photographs were taken after the foliage was removed, giving the impression that the doorstep was much more visible than it in fact was during the operation. Figure 8, a photograph taken from a position in the front yard before the foliage was removed, gives a more accurate illustration of how the porch would have looked to many of the agents on September 23.

The agents positioned on the hillside behind the retaining wall (Brian and Ken) also may not have had a clear view of the area of the porch where the bloodstain formed. As shown in Figure 9 – a view of the residence from the approximate location where Brian was positioned when he fired at Ojeda – the view of the doorstep and the porch floor near the front door was obscured by the balcony railing.

Third, the blood on the porch floor would have accumulated gradually as blood flowed from Ojeda’s body under the door. Ojeda was shot at 6:08 p.m. Sunset on that day was around 6:30 p.m. There was no light on the porch. Therefore, it likely was dark before blood would have been present in an amount visible to the agents from their positions of cover, even assuming that their views had not otherwise been obscured.

Finally, we found it significant that most of the agents who were present at the scene on September 23 told us that they believed that Ojeda was wounded on the basis of hearing him cry out and fall, even without seeing any blood under the door. Most notably, the two agents in the best position to view the blood stain – Don and Scott, who breached the gated door at 6:49 p.m. and remained in the alcove for some time after – strongly advocated entering the residence and therefore would not have had any reason to disregard the significance of the blood stain. To the contrary, the presence of the stain would have made their argument for entering the residence that night more compelling.

IV.   Assessment of FBI Entry Decisions at FBI Headquarters

In this section, we assess separately two aspects of the entry decisions made at FBI Headquarters: first, the reasons for requiring CTD approval for any entry, and second, the explanation for not entering the residence the night of September 23. We also address how removing the entry decision from the San Juan FBI Command Post affected the accuracy of the information available to the CTD decision-makers.

  1. Decision to Require Counterterrorism Division Approval of any Plan to Enter the Residence
  2. CTD’s authority to require, as it did at 8:05 p.m. on September 23, that Hulon approve any plan to enter the residence is undisputed. The Macheteros investigation, which included the arrest operation for Ojeda, was classified as a terrorism case and therefore fell under Hulon’s authority as the Assistant Director for CTD. In addition, following the September 11 terrorist attacks, the FBI mandated that FBI Headquarters, instead of the field division with primary investigative responsibility, would be the lead in all FBI counterterrorism cases and their related operations.

    The issue for the OIG, therefore, was to assess the reasons CTD exercised its authority in the manner it did in this particular case. We interviewed the FBI Headquarters officials most closely involved in the matter – EAD Bald, AD Hulon, and DAD Lewis. Based on these interviews, we found that Bald decided to require CTD approval for any entry plan because he thought that Fraticelli was “overwhelmed” and lacked confidence in his own ability to manage the situation, a perception Bald felt was confirmed by what he believed were indications of command confusion between San Juan FBI and HRT. The decision was made in the context of a shared belief among all three Headquarters officials that FBI Headquarters’ involvement was needed to provide balance and perspective, and to ensure sound decision-making.

    For the reasons explained below, although we did not find this explanation entirely satisfying, we did not conclude that the decision was improper.

    1. Perception that the SAC lacked confidence
    2. Bald’s perception that Fraticelli felt overwhelmed by the situation and lacked the confidence to manage it was based on what he was told about Fraticelli’s demeanor in calls with Hulon, and the significance of what Bald considered Fraticelli’s premature request for the help of another SAC. The fact that Fraticelli requested a specific SAC with whom he had previously worked and that he made “repeated” follow-up inquiries about the status of the request reinforced Bald’s perception.

      It is difficult for us to assess Bald’s perception that Fraticelli lacked confidence. Bald had years of experience overseeing and supervising crisis situations that informed his response to Fraticelli’s request. In addition, Hulon – who as the person actually talking to Fraticelli was in a position to assess his demeanor – told us that he thought it was fair for Bald to form the perception he did based on what Hulon related about his conversations with Fraticelli. Hulon told us that Fraticelli seemed excitable and stressed during their calls.

      However, the agents we interviewed who were at the scene, the TOC, and the Command Post did not tell us anything that supported the perception that Fraticelli lacked confidence or was overwhelmed during the period after 6:08 p.m. when the perimeter shots were fired. In addition, we did not find unreasonable Fraticelli’s explanation for requesting that another SAC be sent to Puerto Rico. Fraticelli told us that while he did not know how long the situation was going to last, he wanted to ensure that he had a counterpart available who could speak Spanish and had experience working in Puerto Rico.

    3. Indications of confusion between San Juan FBI and HRT regarding who was in charge
    4. Bald also told us that he felt his perception of Fraticelli’s lack of confidence to manage the situation was confirmed when Hulon told him at approximately 7:00 p.m. that an HRT agent had gotten close enough to the residence to look inside through a window. This suggested to Bald that HRT was making decisions independent of Fraticelli because the agent’s alleged activity was inconsistent with the earlier guidance Hulon had given Fraticelli to hold the perimeter.

      This aspect of Bald’s decision is also problematic. Hulon told the OIG that he does not believe he could have been the person who reported this information to Bald because he did not recall hearing that any agent got close to the window. Fraticelli, who was Hulon’s primary contact in Puerto Rico, also told us that he did not recall anything about an agent getting close to a window. And we concluded, on the basis of our interviews with the HRT agents who were at the scene, that no one got close to the window at or near the time Bald recalled it being reported to him. Therefore, the particular activity that Bald said confirmed his perception of Fraticelli’s lack of confidence never actually occurred.99

      Nevertheless, it was apparent from Bald’s contemporaneous notes that some event prompted Bald to question who was in charge of the operation. Some of the concern apparently resulted from a statement made by Craig, the HRT Commander in Quantico. Hulon called Craig – we believe at about 7:25 p.m. – to discuss the situation. Craig told Hulon that HRT had the lead at the scene.100 When this statement reached Bald, he consulted EAD Ashley, who agreed with Bald that CTD, through the SAC, had the lead. This resolved in Bald’s mind that he needed to act to ensure there was a proper chain of command in place before any further deliberate action was taken.

      Yet, here too, the information Bald relied on did not accurately portray the situation at the scene. Craig, who Bald incorrectly believed was in Puerto Rico, was not at that time in the chain of command. Moreover, whatever authority Craig believed he was exercising apparently was not registering in Puerto Rico, where Steve and HRT were preparing for a nighttime entry under Fraticelli’s authority. Therefore, the chain of command problem in Puerto Rico that Bald believed was indicated by Craig’s comment about who had the lead did not actually exist.

      We recognize that Bald had to rely on the information he was given, and we do not believe he was required to verify everything he was told. However, we also observed that Bald did not consult with Hulon regarding the chain of command implications of the agent reportedly looking in the window or Craig’s statement that HRT had the lead. In fact, while Hulon told us that he agreed with Bald’s ultimate decision because he felt Fraticelli needed support, he said he was not personally concerned with the management of the operation up to that point.

    5. Belief that FBI Headquarters would provide balance and perspective, and ensure sound decision-making
    6. The context of Bald’s decision to require CTD approval for any entry plan was his view of the appropriate role for FBI Headquarters in circumstances like those in Puerto Rico after the emergency assault failed: to force the managers and agents in the field to step back from the action, carefully review the situation, and methodically assess the next possible courses of action. The goal, as Bald told us, was to avoid making a bad situation worse.

      Hulon and Lewis shared Bald’s view of FBI Headquarters’ role. Hulon told us that while he considers HRT an exceptionally well-trained unit capable of taking dynamic action upon command, he believes there are circumstances where it is appropriate and even necessary for FBI Headquarters to provide an additional, broader-based perspective. Similarly, Lewis told us that in his experience SWAT agents, like those on HRT, are action-oriented and sometimes inclined to be aggressive in barricaded subject scenarios.101 He said that while there are limited circumstances when quick action is required, he felt the situation in Puerto Rico called for stepping back from the earlier violent encounter. Lewis also explained that FBI Headquarters gets involved in situations such as the one in Puerto Rico because ultimately Headquarters is responsible for the outcome.

      These officials’ views have significant implications for many FBI operations, and it was beyond the scope of this review to analyze their effect and ramifications on FBI operations generally. We believe, however, that after Ojeda was shot the FBI commanders in Puerto Rico were doing precisely what Headquarters would expect of them: cautiously and methodically planning the next course of action.

      On the other hand, as we discussed in Chapter Six, we found that HRT had earlier persuaded the SAC to approve an extremely aggressive and risky daylight assault by helicopter, in which the element of surprise was foreseeably absent. The failure of the assault to that point would have been a reasonable explanation for Bald’s concern that further aggressive action be planned carefully and justified.

      However, Bald told us that his decision to require CTD approval for any entry did not reflect a judgment about the earlier daylight assault because he did not have sufficient information at that point to assess the action. Rather, he said the fact that HRT agents had been shot indicated the severity of the situation and impressed on him the importance of having someone in firm control of the operation. Based on the considerations we discussed above, Bald said he concluded Fraticelli was not the right person.

    7. Conclusion regarding the decision to require CTD approval for any entry
    8. We were troubled by Bald’s explanation for requiring CTD approval for entry because we determined that two of the purported factual predicates for the decision – an agent’s activity near a window and the command confusion suggested by Craig’s statement that HRT had the lead – did not reflect what was actually happening at the scene. In addition, we believe that at the time CTD assumed control over the decision, the commanders in Puerto Rico were exercising caution and deliberation in preparing for a nighttime entry to be conducted after the power was cut.

      The most important reason for CTD’s decision to assume control was Bald’s perception that Fraticelli was overwhelmed and lacked the confidence to manage the situation. Bald’s perception was largely based on Hulon’s characterization of Fraticelli’s demeanor during their phone conversations, and it is difficult for us to assess this subjective judgment after the fact. While we found Fraticelli’s explanation for requesting assistance from another SAC to be objectively reasonable, Hulon believed, based on his conversations with Fraticelli, that Fraticelli was stressed and that the request for another SAC was an unusual request under the circumstances. Bald believed the request indicated a lack of confidence. While we are not completely persuaded by the evidence supporting these judgments, we cannot say that Bald’s decision, based primarily on Fraticelli’s perceived demeanor, was improper.

      Finally, there was an additional aspect of the requirement that CTD approve any entry decision that warrants comment. When Hulon called Fraticelli at 8:05 p.m. to tell him that CTD must approve any entry, Hulon initially left open the possibility of a nighttime entry and told Fraticelli and Steve that the proposal should be put in writing and sent to Headquarters for review.

      The FBI’s Manual of Investigative Operations and Guidelines states: “It is recommended that whenever possible, written arrest plans be prepared prior to conducting law enforcement activities that may result in the arrest of a potentially dangerous subject.” It also provides that: “Certain situations may necessitate an oral briefing in lieu of a written plan in exigent circumstances.”

      In our view, Hulon’s requirement that the plan be in writing was unnecessary and unduly bureaucratic, particularly in light of the fact that Hulon had been told that Ojeda was believed to be wounded and on the floor. Hulon told us that he asked for the plan in writing because he considered it standard procedure under the circumstances and because it would allow FBI Headquarters to carefully review the proposed action. However, Bald told us that although written plans should be done if time permits, there would have been nothing wrong in this case with HRT presenting the plan orally to the SAC and then to Hulon. We concluded that the fact that Ojeda was believed to be wounded and on the ground created an “exigent circumstance” that justified an oral briefing in lieu of a written plan. We recognize that Hulon ultimately rejected the plan for a nighttime entry before it was submitted in writing, so that the writing requirement had a limited practical impact on the outcome. While we believe Hulon still would have rejected the plan, allowing Fraticelli or Steve to present it orally might have accelerated the decision and avoided the confusion between Fraticelli and Steve and the agents at the scene, who were preparing until 11:33 p.m. that evening to enter the residence, long after Hulon had decided that the entry would be delayed until the next morning.

  3. Decision Not to Enter the Residence the Night of September 23
  4. We found that CTD’s decision not to approve entry of the residence on the night of September 23 was motivated by a concern for agent safety and a desire to avoid additional casualties. We found no evidence that Bald, Hulon, or Lewis was motivated by a desire to withhold medical assistance from Ojeda. Rather, the decision reflected a good-faith balancing of the available information regarding circumstances at the scene.

    As we discussed in detail in Chapter Three, CTD ultimately rejected a nighttime entry based on its belief that there might be a second shooter in the residence; the lack of certainty regarding Ojeda’s condition and whether he was still a threat; the belief that Ojeda’s intimate knowledge of the interior of the residence mitigated HRT’s nighttime tactical advantage; and the fatigue of the HRT agents at the scene. Hulon also told us that he believed there was a risk based on the history of Ojeda and the Macheteros that the residence contained improvised explosive devices. In Hulon’s judgment, these considerations outweighed concerns that Ojeda might need medical treatment.

    Hulon recognized that, in addition to Ojeda’s possible need for medical attention, there were other considerations that favored resolving the situation on September 23. For example, delaying the entry until the next day meant that the already fatigued HRT agents would have to hold the perimeter overnight. Hulon told us that he felt the San Juan FBI SWAT agents and POPR officers would be available to assist on the perimeter, which in fact is what happened (San Juan FBI SWAT on the inner perimeter, POPR on the outer perimeter). Hulon was also aware of the potential security risk posed by the crowds at the scene but believed this, too, could be managed by San Juan FBI and POPR personnel. With the assistance of a rainstorm that induced much of the crowd to disperse, San Juan FBI and POPR personnel were in fact able to control the crowd.

    The theme that the FBI Headquarters officials stressed in our interviews was that there was no compelling reason to rush the entry and that the information available indicated a sufficient level of risk to warrant delay. We did not find that this consideration reflected indifference for Ojeda’s medical condition. While there was reporting in the SIOC log indicating that Ojeda was “on the floor” and “possibly injured,” the officials told us that they did not know for certain Ojeda had been shot or what his condition was, although the perception of the agents at the scene was that Ojeda had been killed or was seriously wounded, as we discuss in the next subsection.

    Although the CTD officials took into account Ojeda’s possible need for medical care, they concluded that the continuing potential threat to the agents posed by Ojeda or others inside the house was a more important consideration. We do not believe the FBI was required to enter a residence under circumstances that it had reason to believe constituted a continuing threat to its agents.

    We recognize a legitimate argument can be made that the FBI should have entered the residence that evening, as some of the HRT agents at the scene advocated. This argument relies on HRT’s significant tactical advantage in nighttime operations; the near certainty – based on his screaming “ay, ay, ay,” followed by sounds from the residence suggesting a fall – that Ojeda had been seriously wounded and was “on the floor;” the deteriorating security environment on the outer perimeter; the possibility that delay would give Ojeda and any other occupant time to fortify their positions; and the interest in rendering medical care to Ojeda. In retrospect, given our current knowledge regarding Ojeda’s condition at the time, it would have reduced (but surely not eliminated) the criticism of the FBI if the agents had entered the house on September 23, as Fraticelli and Steve initially recommended. However, Hulon did not have this knowledge at the time and knew only that Ojeda might have been hit but that his precise condition was unknown. We found that the decision Hulon made with the information then available to him was not improper.

  5. The Informational Consequences of CTD Assuming Control of the Entry Decision
  6. We described above how Bald’s decision to require CTD approval for any entry was based in part on inaccurate information regarding what was occurring at the scene. We saw a similar problem regarding the decision not to enter the residence on the night of September 23. In our view, the information available to Hulon and Bald was somewhat inaccurate because of their distance from the scene, their lack of continuous communication with the San Juan FBI Command Post, and the filtering of information through at least two levels of reporting. As a result, the CTD officials’ perception of the threat to the agents entering the residence was different in important respects from HRT agents’ perception of that threat. The agents at the scene who were closest to the action said they were virtually certain, based on what they saw and heard, that Ojeda had been killed or very seriously wounded by the perimeter shots at 6:08 p.m. The agents’ certainty on this point, and their increasing confidence in the belief that there were no other subjects in the house, was reinforced by the lack of any response to the partial breach and the absence for several hours of any noise from inside the house except for the sound of a radio playing.

    In contrast, the CTD officials were operating throughout the evening of September 23 under the assumption that there may have been a second armed subject in the house and the belief that Ojeda may have been wounded but that his condition was essentially unknown. We believe that this difference is one reason (but not the only reason) the agents at the scene were so much more eager to conduct an entry that night than CTD.

    Fraticelli told us that he was not aware of the HRT agents’ growing certainty as time passed that the threat from the residence was minimal. In our view, this was because the HRT agents anticipated that the nighttime entry plan would be approved by CTD and therefore did not report their evolving perception of the threat. Further, Fraticelli did not solicit the information because he understood, based on his earlier call with Hulon, that CTD had made a decision to conduct the entry the next day with the relief HRT team from Quantico. In fact, in his later call to Lewis to persuade him that a nighttime entry was the better course of action, Fraticelli said he told Lewis that Ojeda might still be alive.

    The evident disconnect in the assumptions underlying the entry deliberations troubled us. CTD was making decisions based on assumptions about the threat from inside the residence that were not entirely supported by the observations of the HRT agents at the scene. Yet, because the agents at the scene believed it was only a matter of time before the entry would be approved, they did not report their evolving perception of the threat to the TOC. And because Fraticelli believed that CTD already had made its decision relatively early in the evening – for reasons that he told us he found reasonable – he did not solicit from HRT its assessment of the threat as time passed and then report this information to Hulon.

    We asked Hulon whether a different decision might have been made if CTD had been informed during the night of September 23 that the HRT agents at the scene were virtually certain Ojeda had been seriously wounded, if not killed, and were increasingly confident there were no other subjects in the house. Hulon told us that this information would have prompted additional discussion with Bald, and possibly input from Lewis and Craig as well, but that he could not state whether this would have changed his decision about delaying entry. Hulon said they likely would have assessed the basis for the agents’ judgment and considered that together with the other relevant factors, such as agent fatigue and the possibility of explosives in the residence.

    We concluded that the distance and multiple layers of reporting between the agents at the scene and the FBI Headquarters officials in Washington D.C. affected the information that was the basis of the decision not to enter the house during the evening of September 23. However, we cannot say with certainty that the decision in this case would have changed if the more immediate assessment of conditions at the scene had been available to the CTD officials.

    The OIG’s experts were critical of the command and control structure where tactical decisions are made from a location other than the scene of the operation. As one expert explained, circumstances in crisis incidents like that in Puerto Rico are frequently changing and the tactical decision-maker must be at the scene to keep pace with what is happening and to best appreciate the relevant risk factors. In the expert’s view, remote management that relies on intelligence being accurately reported through multiple layers is problematic.

    In contrast, the FBI Headquarters officials we interviewed told us there are circumstances where Headquarters’ involvement in a crisis incident is important. In this case, for example, Hulon felt Bald’s decision to require CTD approval for any entry plan was appropriate because it eliminated any command confusion and provided additional support to the operation. Hulon also told us that the involvement of senior managers from Headquarters brings significant operational experience to crisis incidents and provides perspective regarding how decisions in one incident might impact FBI operations more broadly. Headquarters’ involvement can also bring a measure of deliberation and caution to crisis incidents where, as Lewis told us, stepping back from a violent encounter is a preferable strategy to the quick, aggressive tactics sometimes favored by action-oriented agents such as those on HRT.

    It is beyond the scope of our review to determine generally the relative merits of our experts’ opinion (that any command and control structure where FBI Headquarters makes tactical decisions is a poor model for crisis incidents) versus the desire to vest ultimate decision-making authority in a senior manager who while removed from the scene has a broader perspective. However, this case demonstrates how an information disconnect can affect the decisions of remote managers. In considering when, and how, to vest decision-making authority for crisis situations in FBI Headquarters, we believe that the FBI should take this consequence into account in assessing the conditions under which Headquarters will assume control over a crisis incident. And when FBI Headquarters assumes such operational control, the FBI should ensure that adequate information flows to the Headquarters officials who must approve operational decisions and that Headquarters officials adequately communicate with on-scene commanders.

V.   The Tension Between the FBI’s Aggressive Strategy Before the Gunfire Exchange and its
       Cautious Strategy after Ojeda Was Wounded

There was an obvious change in the degree of aggressiveness in the FBI’s strategy during the course of the operation. When the sniper-observers reported that they had been compromised, Fraticelli approved the extremely aggressive immediate assault strategy recommended by Steve. After three HRT agents were shot and one was seriously wounded, the agents at the scene still wanted to immediately enter the residence, but the FBI adopted a far more cautious approach. CTD’s decision not to approve a nighttime entry after Ojeda was wounded clearly reflected a greater emphasis on caution and safety than informed the selection of the emergency assault plan.

This change in emphasis subjected the FBI to harsh criticism, ranging from the suggestion that the FBI was not prepared to finish what it had started to the accusation that the FBI’s decisions were part of an intentional plan to assassinate a symbol of the Puerto Rican independence movement.

We determined that the change in emphasis toward a more cautious approach was not the result of malicious intent, but rather FBI Headquarters’ adjustment from the initial strategy that had failed.102 As we discussed in Chapter Six, we concluded that HRT’s decision to conduct an emergency daylight assault was extremely risky in comparison to available alternatives. The selection of a more cautious approach after Ojeda was shot was, in significant part, an understandable reaction to the results of the more aggressive approach, which had resulted in three agents being shot and one seriously wounded. Thus, in our judgment, the contrast in approaches was the result of an overly risky assault as opposed to an overly cautious entry afterwards.

In addition, the circumstances had changed by the time the nighttime entry decision had to be made. The emergency assault, in theory, required rapid action so that Ojeda did not have time to prepare any resistance. After the assault failed and Ojeda was believed shot, there was a wholly different scenario to contend with: the apprehension of a barricaded, potentially wounded subject who had shot at law enforcement, with the possibility of a second shooter in the residence. The OIG’s experts concluded, and we agreed, that this scenario supported the FBI’s cautious approach.

VI.   Consequences for the Outcome

It is critical to note that, although the decision to put off entry into the residence until the next day was one of the most controversial aspects of the operation, we found that it likely had no impact on Ojeda’s death. As we discussed in Chapter Four, the Forensic Pathologist who performed the autopsy told us that based on the size of the wound and reasonable assumptions about Ojeda's heart rate and blood pressure, he estimated that Ojeda died from loss of blood approximately 15 to 30 minutes after being shot, which would place the time of death at between 6:23 and 6:38. In addition, when the OIG asked the Office of the Armed Forces Medical Examiner whether Ojeda could have survived longer than two hours (the period during which Fraticelli was in charge of the entry decision), the Medical Examiner responded that it was unlikely.

If these medical opinions are accurate, Ojeda died from blood loss before the agents at the scene conducted a partial breach of the gated doorway at 6:49 and well before CTD took control of the entry decision at 8:05 p.m. Indeed, if the estimates from the Forensic Institute are accurate, the only chance to save Ojeda’s life would have been for the agents to enter the residence some time before 6:38 p.m. and evacuate Ojeda immediately. At 6:38, the partial breach had not yet been conducted. Given the result of the earlier effort to enter the residence, the uncertainty regarding Ojeda’s condition, and the belief that more than one weapon had been fired from inside the house during the gunfight, an immediate effort to enter the house would have been a risky action and we cannot fault the FBI for taking a more cautious approach.

VII.   Conclusions Regarding the FBI’s Entry Decisions

In this Chapter, we considered the entry decisions made by the FBI commanders in Puerto Rico and the officials at FBI Headquarters. We concluded that the decisions were motivated by considerations of agent safety and not by any desire to withhold medical treatment from Ojeda. The FBI’s primary concern was the threat posed to agents entering the house by the uncertain condition of Ojeda and the possibility of an additional armed subject in the residence. Early in the evening of September 23, this concern justified the cautious and deliberate approach taken to preparations for entering the residence after dark. Later in the evening, after CTD assumed control over the entry decision, this concern was the reason for CTD’s decision that no entry would take place until the relief team of HRT agents from Quantico arrived the next day, on September 24. Additional reasons for CTD’s decision included agent fatigue and the possibility of explosives in the residence. We concluded that CTD’s decision reflected a good-faith balancing of the information known at the time, and we found no evidence that the decisions made by the FBI commanders in Puerto Rico or the CTD officials were based on any desire to withhold medical attention from Ojeda.

We also assessed in this Chapter the decision to require that SAC Fraticelli obtain CTD approval before entering the residence. This decision was primarily based on the perception at FBI Headquarters that Fraticelli appeared overwhelmed and lacked the confidence to handle the situation. Although we found that CTD’s decision to assume control was based in part on a misperception of events at the scene, we could not conclude the decision was improper.

We concluded that there were troubling aspects to the entry decisions in terms of the clarity in the chain of command in Puerto Rico and the accuracy of information used by the CTD officials in making their decisions. A clear chain of command is critical in operations where HRT is deployed in support of an FBI field division. In this case, Fraticelli was in charge of the operation and Steve served as his tactical advisor and was the ranking HRT manager in the CONOP’s chain of command. Yet, according to Craig, he gave Steve a tactical order not to enter the residence after the limited breach of the gated door. In our view, Craig’s order was troubling because we do not believe he had authority to make tactical decisions at that time, and because Steve – who told us Craig did not give him any advice or recommendations at all – either misinterpreted or ignored what Craig considered an order. Craig also incorrectly told Hulon that HRT was in charge, which caused Bald to question Fraticelli’s leadership. In fact, Fraticelli was in charge, contrary to Craig’s statement. This apparent discrepancy in the chain of command should not occur, and we recommend that HRT management take measures to ensure it does not happen in the future.

There also was an evident lack of clarity between Fraticelli and Steve regarding CTD’s decision not to enter Ojeda’s residence on September 23. Based on the statements of Fraticelli and Hulon, CTD’s final decision was conveyed to Fraticelli sometime before 9:00 p.m. However, the decision apparently was not conveyed to Steve because HRT continued to draft a nighttime entry plan, which was finally faxed to FBI Headquarters for approval at 11:25 p.m. By that time Hulon had already left for the night. According to Steve, he did not learn CTD had made a final decision until he and Fraticelli called Lewis to try to persuade him – in the hope that he could then persuade Hulon – that HRT should conduct a nighttime entry. Although the recollections of those involved were imperfect as to the exact timing of the calls, we believe this call and Lewis’s return call occurred after 11:15 p.m., when Hulon and Lewis had already left FBI Headquarters for the night. When Fraticelli told Steve that Lewis said there would be no entry, the decision was conveyed for the first time to the HRT agents at the scene. According to the TOC Log, this occurred at 11:33 p.m. While we could not determine precisely why Steve and HRT believed a nighttime entry was a possibility nearly three hours after CTD decided against that course of action, this demonstrated a troubling lack of communication between Fraticelli and Steve concerning the most significant tactical decision left to be made.

Finally, in this Chapter we also highlighted the significant informational consequences of CTD’s decision to assume control of the entry decision. A disconnect existed between CTD’s assumptions regarding the threat from inside the residence and the HRT agents’ assessment of that threat as time passed. Although we could not conclude that the decision in this case would have changed if the more immediate assessment of conditions at the scene had been available to CTD, we believe that the FBI should take into account the potential adverse informational consequences when assessing the conditions under which Headquarters will assume control over a crisis incident and in ensuring that adequate information flows to and from Headquarters officials who must approve operational decisions.

  1. We identified two issues relevant to San Juan FBI’s and HRT’s decisions that relate to the level of preparedness and include them here for the FBI’s consideration in future operations. First, one of the OIG’s experts questioned whether HRT had equipment available that could have been used to see inside the residence, such as fiber optic video equipment. He thought this should be standard equipment for an operation such as this and considered it a mistake that it was not brought. None of the experts believed that having the equipment would have changed the FBI's obligation to enter the residence, but using it might have provided intelligence that could have hastened the decision to enter.

    Second, regarding the significant delay shutting off the electricity to the residence, the experts questioned whether more thorough intelligence gathering before the operation began could have hastened this process. Of course, one possible explanation for this not being done was that neither San Juan FBI nor HRT anticipated a barricaded subject scenario and therefore did not consider in advance how the utilities could be shut off. However, we found in Chapter Six that there should have been contingency planning for this scenario. As we discuss in Chapter Eight, the FBI’s failure to anticipate a barricaded subject scenario also affected its negotiations preparedness.

  2. In fact, on the subject of Craig sending HRT reinforcements to Puerto Rico, Steve said he told Craig that they would not be needed if the situation was resolved that night (September 23), suggesting that entry that evening was still considered a viable option.

  3. We believe it likely that the incident described by Bald was actually the limited breach of the residence. Agents at the point of the breach reported that they were able to look into the hallway behind the wooden gate but could not see Ojeda.

  4. Craig told the OIG that when Hulon asked him who was in charge of the scene, he replied HRT was unless Hulon ordered otherwise. In fact, that was incorrect – according to the CONOP and FBI policies, SAC Fraticelli was in charge.

  5. The OIG’s experts likewise told us the dynamic approach designed to overwhelm a subject through confrontation and force is very popular in law enforcement training today and that, as a result, HRT-type units are most comfortable in environments requiring that approach.

  6. The OIG experts also told us that, based on their review, HRT did not send enough agents to Puerto Rico to conduct the operation. It is possible that, with an additional shift of agents available the fatigue factor might not have been as persuasive a reason for CTD to reject a nighttime entry. In this sense, it could be said that the change to a cautious approach reflected inadequate preparation. But in light of the outcome of the emergency assault, we believe that CTD would have selected a more cautious approach even if more HRT agents had been available at the scene.

« Previous Table of Contents Next »