Written evidence submitted by Raphael Marshall (AFG0038)


Evidence for the House of Commons Foreign Affairs Select Committee’s Inquiry on Government Policy on Afghanistan




1.     From November 2018 until 21 September 2021, I worked at the Foreign, Commonwealth, and Development Office as a Desk Officer (equivalent to a HEO).


2.     From Saturday 21 August to Wednesday 25 August 2021 I worked on the Foreign Office’s Afghanistan Crisis Response.   Specifically, I worked on the  ‘Afghan Special Cases’ team. Our team’s responsibility was Afghans whose lives were at risk because of their association with the UK, or the West as a whole, who did not qualify for the Afghan Relocations and Assistance Policy (ARAP) scheme.  These people were eligible for evacuation under the UK’s Leave Outside the Rules (LOTR) scheme.  Applicants included Afghan soldiers, politicians, journalists, civil servants, feminists, aid workers, and judges.   The ARAP scheme’s criteria are strict, and therefore our scheme also included some Afghans who superficially appeared to qualify as ARAP cases, including guards and interpreters.


3.     I estimate between 75,000 and 150,000 people (including dependents) applied for evacuation under the LOTR scheme. The vast majority of these applicants feared their lives were at risk as a result of their connection to the UK and the West and were therefore eligible for evacuation.


4.     I estimate fewer than 5% of these people have received any assistance.  It is clear that some of those left behind have since been murdered by the Taliban.


5.     By 21 August, it was inevitable that the vast majority of Afghans appealing to the UK for evacuation would be left behind and that some would be killed by the Taliban.  This was not a result of problems with the UK’s evacuation effort but an inevitable consequence of the situation in Kabul and the broader decision to complete western withdrawal by 31 August.


6.     However I witnessed two important avoidable problems in the FCDO which significantly undermined the efficacy of our evacuation effort:


    1. In my opinion, the FCDO’s process for selecting which Afghan applicants to evacuate was not credible.  Given it was only possible to evacuate around 1-5% of applicants, it was very important that we selected this 1-5% of applicants rigorously and credibly to ensure we selected the individuals to whom the UK had the greatest obligations. In my opinion, these prioritisation decisions were therefore not made in a rational way. I accept there were extremely hard decisions to be made; my concern is that these decisions were made badly.


I judge the purpose of the evacuation was primarily to demonstrate the value of British promises by saving friends of the UK. Therefore accurately identifying the correct people to rescue was imperative as this objective would not be attained unless we selected the right people for evacuation.  Justifiable selection of evacuees was therefore at least as important as the number of evacuees.


    1. Avoidable bureaucratic problems significantly delayed the ‘calling forward’ of prioritised evacuees to the airport. Given the increasing difficulty of entering the airport, this significantly reduced the chance of evacuees being successfully air-lifted.


7.     In my opinion, an important underlying cause of these problems is that the FCDO’s Crisis Response structure is not capable of satisfactorily responding to a crisis of the urgency, importance, and complexity of the situation in Kabul in Afghanistan.  When I completed my Crisis Response training in 2019, the scenario we role-played concerned British tourists stranded in a Jamaican hotel in the aftermath of a hurricane. In my opinion, the Crisis Structure appears to be best suited to addressing situations of this type. In my view, this is concerning because the FCDO uses the Crisis structure as the primary mechanism for responding to difficult or dangerous international situations.


8.     I have divided flaws in our efforts to evacuate ‘Afghan Special Cases’ into 10 categories:


  1. Arbitrary and dysfunctional prioritisation process.
  2. Inadequate staffing, including lack of night shifts.
  3. Lack of expertise and arbitrary allocation of staff.
  4. Lack of urgency.
  5. Lack of coordination between the FCDO and MOD.
  6. Lack of coordination with our allies including the United States.
  7. Lack of buses for Afghan Special Cases.
  8. Assumption every visa-holder would reach the airport
  9. Ineffective cooperation between ministers and civil servants.  Difficulty of dealing with requests on 25 August.
  10. Lack of planning for after 31 August.


9.     In addition I will provide evidence about:

  1. The FCDO’s internal investigation in response to my report of breaches of the Civil Service Code.
  2. The transport of Nowzad’s staff and animals. Risk to British soldiers involved. 
  3. FCDO planning for the fall of Kabul.
  4. Benjamin Slater and Nomad Group.


10. Some of these flaws were present in the FCO’s spring 2020 Covid Crisis Response, which I also worked on. Most notably in spring 2020 we also struggled to manage the volume of emails received, including from MPs. Credible lessons learned exercises were conducted after the Covid Crisis and I expect they highlighted these problems.


11. On Tuesday 31 August, I wrote to the Permanent UnderSecretary to report that these problems constituted breaches of the Civil Service Code. The Permanent UnderSecretary met me the same day and I set out the flaws listed in this evidence in general terms. In response the Permanent UnderSecretary appointed a credible senior official to investigate. I spoke at length to the investigator in September and provided an early draft of this evidence.


12. The investigation concluded no breaches of the Civil Service Code had occurred but identified a number of lessons for the FCDO to learn. The Permanent UnderSecretary briefed me on this conclusion on Wednesday 29 September.


13. I respect the personal integrity of the investigator. However I am unconvinced by the finding that the Afghan Special Cases evacuation process was compatible with the Civil Service Code. I certainly do not believe that any individual person acted improperly or maliciously but I do believe that the FCDO’s institutional approach amounted to a failure to abide by the values set out in the Civil Service Code.


14. I did not expect the FCDO’s internal reviews would result in meaningful reform of the FCDO’s crisis response structure or on UK policy on the resettlement of Afghan friends of the UK.  As such I decided the right thing to do was to provide evidence to this Committee.


15. In my opinion, the specific flaws in the Afghan Special Cases process outlined in this evidence were likely representative of broader problems with the FCDO’s Afghanistan crisis structure.  I believe that the ARAP team and, to a lesser extent, the consular team suffered from similar problems.


16. I love the Foreign Office. I previously expected to work in the Foreign Office for many years.   I am very grateful to the Foreign Office for three largely very rewarding and interesting years. In particular, it was a privilege to work with many deeply inspiring colleagues especially in Eastern Europe and Central Asia Directorate.  I believe effective scrutiny by the Foreign Affairs Committee is strongly in the Foreign Office’s long-term interest.


17. I believe many of my colleagues working on the ‘Afghan Special Cases’ Team acted with great integrity.  I admired the person who acted as Crisis ‘Bronze’ on the afternoons of Saturday 21 August, Monday 23 August, and Tuesday 24 August.  I also admired the Senior Responsible Officer (SRO). In my opinion, the SRO was an honourable, professional, and compassionate public servant in an impossible position. The problems outlined in this evidence were not the SRO’s fault.  In normal circumstances I am sure they are an exemplary leader.


18.                         In my opinion, HMG should conduct a full external (non-FCDO) review of the FCDO’s Afghanistan Crisis Response and the FCDO’s Crisis Response structure. 



Flaws in the FCDO’s Afghanistan Crisis Response


1)      Arbitrary and dysfunctional prioritisation process.


19. The key task of our team was to process emails from Afghans at risk (or from their British friends) to the Afghan Special Cases Inbox.  There was a separate inbox for the ARAP scheme.  The inbox had received tens of thousands of emails since early August. Most emails contained a ‘principal’, the person in whose name the application was made, and a number of immediate dependents.


20. Many of these emails were not read.  Between Saturday 21 August and Wednesday 25 August, there were usually over 5000 unread emails in the inbox at any given moment, including many unread emails dating from early in August.  These emails were desperate and urgent.  I was struck by many titles including phrases such as ‘please save my children’.


21. Many of these emails also documented numerous recent grave human rights abuses by the Taliban, including murders, rapes, and the burning of homes.  The contrast between HMG’s statements about a changed Taliban and the large number of highly credible allegations of very grave human rights abuses HMG has received by email is striking.


22. Emails received an automatic response that the request for assistance had been ‘logged’. This was usually false. In thousands of cases emails were not even read.  In 100s if not 1000s of cases emails from MPs were also unread. 


23. Whenever I looked at the ARAP inbox, there were an even larger number of unread emails, including many unread emails from earlier in August. On the evening of Thursday 26 August, there were 4914 unread emails in the ARAP inbox. 


24. Hundreds if not thousands of emails from MPs of all parties were not read until the evacuation was concluded. There is no suggestion of any party political discrimination. 


25. The FCDO encountered the same problem during the Covid Crisis in spring 2020. In February 2020, I worked on processing a backlog of thousands of unread appeals for assistance from British tourists stranded in Peru.  In spring 2020, there were also a large number of unanswered emails from MPs, in some cases dating back weeks. Although lessons learned exercises were conducted after the Covid Crisis, clearly the FCDO failed to recognise and address this problem.


26. The role of our team was to judge who was eligible for assistance and then to prioritise who to evacuate. We were provided with three criteria to carry out this task: 


    1. Vulnerability.  In essence, does the applicant have a credible fear of being killed by the Taliban? This was clearly a very broad category and encompassed a sizeable majority of applicants. 


    1. Support for UK objectives in Afghanistan. Did the applicant work to support the UK’s objectives in Afghanistan or work directly with the UK? Given the UK’s broad objectives in Afghanistan, this criteria clearly encompassed a very significant percentage of Afghans and a sizable majority of applicants.


    1. Significance/Sensitivity. This criteria was not clearly defined. It was usually described as how prominent the individual was. However it could also include other factors such as whether an MP had lobbied on their behalf.  The words ‘Significance’ and ‘Sensitivity’ were in practice used interchangeably for this criteria.  At one point on Sunday 22 August or Monday 23 August, a colleague acting as ‘Bronze’ described significance as solely meaning an MP had lobbied on the applicant’s behalf.


27. These criteria were very unhelpful. The majority of those in the inbox clearly met the first two criteria.  The meaning of the third criteria was very unclear. It was not clear whether the intention was that applicants would be required to meet all three criteria to be eligible for evacuation but in practice the criteria of ‘significance’ was to a great extent disregarded because of the lack of clarity about its meaning.


28. We could only evacuate a very small proportion of those who met the criteria of ‘vulnerability’ and ‘support for UK objectives’.  There was no guidance on how to prioritise among the majority of the applicants who met at least these first two criteria.  What was the relative prioritisation between an Afghan judge, an Afghan National Army (ANA) commando, a human rights activist, or a former base guard?  In retrospect, it appears likely the criteria of significance was intended to act as the key criteria to distinguish between the numerous applicants eligible for the first two criteria.  However, in practice, due the lack of instructions as to how ‘significance’ was to be assessed, this did not happen. 


29. ‘Significance’ and ‘Sensitivity’ are both extremely subjective terms. This criteria was presented as though it was self-evident which characteristics would render particular Afghan people significant to the British Government. In my view this is in-fact ambiguous. Similarly ‘significance’ is also clearly a relative term, indicating an application for evacuation was more or less important than other applications we received.  It’s unclear how someone on the first day on the team could form a reasonable assessment of an applicant’s relative importance given they did not have an overview of the inbox.


30. In my opinion in practice it is likely that the criteria of vulnerability was often not applied in a rigorous way. Many emails simply stated as a matter of course that the Taliban had taken power in Afghanistan and therefore the applicant was at risk of murder. Some emails also included more specific evidence of targetting or detailed reasons to believe the applicant was at exceptional risk. In practice I expect insufficient distinction was observed between applicants who explained a specific risk, for example that they had received specific death threats or believed their work had made them especially prominent, and between applicants who merely referred to the general risk posed by the Taliban coming to power.


31. The criteria of ‘support for British objectives’ was significantly clearer.  At the same time, in retrospect this relied heavily on the team’s instincts as to British objectives in Afghanistan and their relative importance.


32. The criteria gave no formal preference for those who had worked closely with British soldiers or diplomats over the much broader group who had only supported broad western objectives in general terms. The SRO said to me on Sunday that we should favour those with strong UK ties, but I do not believe this was in practice reflected in our prioritisation. Only those who had worked directly for the UK, as opposed to merely very closely with the UK or for contractors hired by the UK, were eligible for the ARAP scheme.   Therefore guards who had protected the British Embassy and, also I believe, the guards of a National Crime Agency compound in Kabul were not prioritised for evacuation.


33. In my opinion given the extreme situation we should have accorded greater importance to strong ties with the UK. In essence this was primarily a political rather than a humanitarian evacuation.  Given the current situation in Afghanistan the number of people who would qualify for humanitarian evacuation overwhelmingly exceeded the number of evacuation places plausibly available, arguably including 4 million Hazaras and several million educated women.  Therefore in my view, the evacuation’s objective was clearly not humanitarian as such but primarily intended to demonstrate the UK’s credibility as a friend. For this reason, selecting the correct people to evacuate was at least as important as the number of people evacuated.  The Afghan Special Cases team did not receive instructions about the policy-objective of the evacuation beyond the three criteria.


34. On reflection, I believe the criteria were ambiguous because implicitly HMG wished to evacuate two separate cohorts of people.  Firstly, people who had worked closely with the UK but were ineligible for ARAP even if they were not ‘significant’ as such (guards, interpreters, and similar people) and secondly, Afghans of significance to the West as a whole even if they did not have close ties with the UK specifically.  The criteria included both these separate groups and as a result were excessively broad and did not provide a useful basis for prioritisation. 


35. I believe the Foreign Secretary had approved a submission including a list of categories of people to be evacuated (intelligence officers, journalists, judges etc).  However I believe the Foreign Secretary’s instructions did not address the key question of how to prioritise among the categories.  This was an important omission because the number of applicants in these categories considerably exceeded available evacuation slots.  These categories were not provided to the staff processing the emails, we only received the criteria.  It is not clear what the relationship between the categories and the criteria was intended to be.


36. Some claims in emails were evidenced by endorsements from British soldiers and diplomats, certificates, or other supporting documents. Many other emails contained no evidence. There was no instruction to distinguish between claims with and without evidence or to review evidence provided. Due to our team’s lack of expertise, I believe we will not have distinguished between claims in emails which were and were not credible. As a rule, all claims made in emails were accepted as true.


37. In practice any Afghan could likely therefore email the Afghan Special Cases inbox claiming to be eligible for evacuation.  It appears likely that such a person would not have a materially lower chance of evacuation to someone genuinely at great risk of being killed by the Taliban.  It is therefore possible that some evacuation spots were misallocated to people who misrepresented the level of risk they faced. Clearly in the circumstances application of onerous evidentiary standards would have been impractical but in my opinion greater efforts to distinguish between claims with and without evidence would have been appropriate.


38. An Afghan ‘Commando’ evacuated from Kabul by HMG was arrested in a raid by armed police in Manchester on 11 September. He may well have been prioritised for evacuation by the Afghan Special Cases team. I do not know the nature or comprehensiveness of the security checks performed by the Home Office. However it would not be reasonable to expect the Home Office to possess comprehensive databases.


39. No consideration was given to whether an evacuee was in Kabul. Although it is not clear how this could have been assessed, in consequence evacuation slots may have been allocated to people who could not plausibly reach Kabul Airport in the time between call-ups and the end of evacuations.


40. These emails were processed in different ways at different times.  Between Saturday 21 August and Wednesday 25 August we normally marked emails with Microsoft Outlook flags;  blue for eligible and red for ineligible. Earlier in August a wider selection of different coloured flags were used to denote different categories of evacuation groups.  I believe these flags classified emails by categories such as soldier, journalist, etc.


41. An initial group of priorities for evacuation had already been chosen during the week commencing Monday 16 August.  I judge the method used is likely to have been extremely similar to the process described here but I do not know for sure.


42. Eligible emails were usually entered into an initial spreadsheet of potential priorities.  On Saturday 21 August, I entered every eligible case I encountered into the spreadsheet.  I do not know who designed this system or on what basis.


43. Emails were frequently forwarded between the ARAP, Afghan Special Cases, and generic crisis mailboxes.  I believe there were long delays in processing emails in each of these inboxes. I judge the result of this confusion between email inboxes was likely that some applicants will not have been considered for evacuation because their email waited days for attention in one of the various inboxes before being forwarded on and waiting again for attention in another.


44. Arguably the most deserving cases for evacuation under the LOTR scheme were people who were almost entitled to evacuation under the ARAP scheme but not actually eligible due to technicalities (such as interpreters or guards employed through subcontractors). Logically, these cases are the most likely to have suffered from being forwarded between the ARAP and Afghan Special Cases mailbox. There was no mechanism for referring refused requests for ARAP to the Afghan Special Cases team for special consideration.


45. Earlier in August a large number of emails had been read but had not been processed (i.e. entered into a spreadsheet for potential evacuation).  These emails were marked with a black flag in addition to whatever other flags they were marked with.  We never returned to these emails due to lack of time. They were therefore de facto eliminated from the evacuation process. 


46. There were also a large number of emails which had been read earlier in August but not flagged in anyway. From Saturday 21 August, we operated on the assumption the people contained in these emails had already been considered for evacuation and therefore did not return to these emails. As set out in paragraph 53, this assumption appears to have been wrong in at least some cases. In at least some cases these emails appear to have been previously read but not considered for evacuation.   In my opinion, these problems likely arose due to frequent changes of who was managing the inbox leading to inconsistent approaches and an extremely brief institutional memory. 


47. The spreadsheet of potential priorities usually included short summaries of cases which were sometimes misleading as they had been written at speed by someone without meaningful Afghanistan expertise; for example, eliminating the distinction between someone who had once attended a gender empowerment workshop and a leading women’s rights activist.  These summaries were not written according to any consistent methodology; in some cases the entire email was copy and pasted, in other cases a summary was written, in other cases no summary or only a several word summary was provided.


48. The summaries were therefore inconsistent and likely often misleading . This was important because I believe subsequent prioritisation of the cases entered into the spreadsheet was performed primarily or exclusively on the basis of these summaries. I believe there was no time for the people performing the subsequent prioritisation of cases entered into the spreadsheet to return to the original emails in the vast majority of cases (or possibly in any cases at all). On the basis of having returned to some of the original emails sent by individuals subsequently prioritised for evacuation to look for passport details on the evening of Monday 23 August, I judge the subsequent prioritisation process would not have prioritised at least some of those selected evacuation if the people performing the prioritisation had seen the original emails rather than only summaries.


49. The spreadsheet also often did not include the applicant’s passport details.   In some cases this resulted in people being prioritised for evacuation who had not sent us the passport details necessary for security checks. This is discussed further in paragraph 61.


50. On the afternoon of Saturday 21 August, there were already around 150 rows of data on this spreadsheet.  Given a row typically included a ‘principal’, their partner, and a number of children, and some rows included institutional lists, this likely numbered over 1000 people already.  This may already have been more than the number of evacuation slots which remained available (although it was not clear how many evacuation slots were available).


51. The methodology for deciding who to enter into the spreadsheets appears to have varied. Judging by the cases included in the spreadsheet on Saturday 21 August, at the end of the week of 16 August primarily or exclusively cases put forward by MPs appear to have been entered into this spreadsheet at the expense of other cases.  At the same time many other emails from MPs remained entirely unread.


52. On the afternoon of Saturday 21 August, the Foreign Secretary’s Private Office contacted the Crisis Centre to say that Tom Tugendhat MP had complained that his emails to the Crisis Centre appear to not even have been read. Mr Tugendhat had sent roughly 10 cases to the Afghan Special Cases inbox (perhaps 80 people in total). I checked and none of these had been processed at all or entered into any spreadsheet.  In the case of Mr Tugendhat’s interpreter whom he had emailed about repeatedly, I could not find his details in our inbox and they must have been lost somewhere in the FCDO system.


53. Mr Tugendhat’s emails had been opened but not flagged and no action of any sort appeared to have been taken. This suggests that other read but unflagged emails from earlier in August may also not have been processed. In Mr Tugendhat’s case this problem was detected because Mr Tugendhat was in a position to contact the Foreign Secretary’s Private Office to complain. Clearly for the vast majority of people whose emails had been read but not processed or entered into any spreadsheet, this was not possible.


54. We emailed Mr Tugendhat including a list of all the names he had sent us, saying that we were processing them. This email was sent under my name but was effectively written by committee.  In fact, none of these cases had so far been processed at all. We subsequently entered these names in our initial spreadsheet as a result of Mr Tugendhat’s email.


55. On Sunday 22 August, a group of around six FCDO staff formerly in DFID volunteered to assist. It was hard to integrate them effectively because we could not share live documents or give them access to the inbox because the DFID and FCO IT systems are not yet integrated.  They were visibly appalled by our chaotic system.


56. On the morning of Monday 23 August, another civil servant, I believe of roughly my grade, decided which cases should be transferred from the initial prioritisation spreadsheet of eligible cases we had been compiling into another further prioritised spreadsheet. I do not know on what basis these decisions were made as no criteria for how to prioritise the thousands of eligible cases were ever provided. 


57. This second spreadsheet was then reviewed by the responsible senior civil servant on the evening of Monday 23 August. They carried out this task with integrity.  However, They were not in a position meaningfully to review the decisions made by their subordinates.  They did not have time to review the underlying emails on which decisions were made.  They only saw the cases flagged for prioritisation by their team, not those which had been rejected. In practice, there was therefore no effective review of these decisions. 


58. At this point two colleagues (I believe a D6/G7/Team Leader and a C4/HEO/Desk Officer respectively) further narrowed down the list in line with a senior civil servant’s instructions. These were committed people and  I am sure they carried out this task with integrity.  However it is unclear on what basis they made these decisions.  Both were clearly deeply drained by this task.  I judge at this stage they likely removed several hundred names or more from the evacuation list.


59. On the evening of Monday 23 August, I was also asked to review lists from a number of organisations and decide which of their staff should be evacuated; these organisations were the Nomad Group, the Afghan Women’s football team, the Afghan Paralympics team, and an Afghan Human Rights organisation. These lists totalled several hundred names. I had no basis on which to carry out this task beyond guessing the likely relative vulnerability of individuals based on their job titles.  I managed to prioritise the Human Rights organisation on this basis, but was told that the evacuation list was already full. Ultimately none of these hundreds of people made it onto the final list. I was very concerned that my failure to prioritise these names fast-enough may have been a factor in them not being added to the final list.


60. After our list of names was finalised we needed to send the spreadsheet to the Home Office for security checks.  However the spreadsheet often lacked the necessary details, a passport number and date of birth. We therefore went back to the original emails to extract this information. This took considerable time.  Some of the people who had been prioritised had not provided the necessary details in their original email. It was now too late to request these details, so these people were eliminated from the evacuation process.


61. On returning to the original emails to search for details, I judged that there appeared to be extremely limited grounds for preferring the emails prioritised for evacuation over many other similar ones.  Some emails prioritised were extremely brief and only included very unremarkable information, such as that someone was from a particular village and had attended women's rights seminars.  I believe that the imprecise summaries in our spreadsheet contributed to mis-prioritisation.


62. Most or all stages of this prioritisation were carried out without knowing how many evacuation slots we still had available. This obviously made effective prioritisation difficult because we were selecting an overall priority cohort rather than ranking priorities. To perform this task rigorously it would have been necessary to know the size of the cohort we were selecting for evacuation.


63. We did not have this information because, despite repeated attempts, it was not possible for us to find out the number of people we had already called-up. On Sunday 22 August, the person who was said to know this number appeared to be uncontactable. A colleague told me they were off-shift and I could not find who had replaced her.  They were also a former DFID employee and possibly working from home; given DFID and FCO Teams accounts are not fully integrated it was hard to be sure whether they were online. When I located the relevant team on the morning of Monday 23 August, they did not know how many people we had already prioritised for evacuation.  They were still counting and could not confirm when they would know. 


64. Around midnight on the night of Monday 23 August to Tuesday 24 August the list was sent to the Home Office for processing. After we sent names I believe it took roughly six hours for the first groups of names from this spreadsheet to be returned in the early hours of Tuesday 24 August.  Most did not return until somewhat later.  This was a predictable consequence of the FCDO sending a group of roughly 1000 names together.  


65. After the Home Office had carried out security checks on the names, they were placed on a final spreadsheet for ‘call-up’ to the airport. Although the order in which names returned from the Home Office was arbitrary, this became a de facto prioritisation. This is because the order of names on the spreadsheet determined the order in which they were ‘called-up’ for actual evacuation.


66. The system for the final stages of the prioritisation were designed as we went along on Monday by officials of my rank.  I arrived at 6:30 AM on Sunday 22 August determined to find out the process for bridging the gap between reading the emails and actually calling people forward for evacuation.  I found chaos, particularly as I understand an Afghan had been shot at while talking to British soldiers on the phone overnight. There was no clarity on what the next steps for prioritisation were.


67. A member of Lord Ahmad’s Private Office and I sent a separate shorter list of very high priority cases to the Home Office in parallel to this wider process on the night of Sunday 22 August.  This is because the overall list process seemed to be advancing dangerously slowly and we wanted to make progress on ensuring that at least some people were rescued.


68. I compiled this list primarily on the basis of deferring to credible experts who had been in contact with us. I am therefore confident that the choices made were at least as justifiable as the decision made by our overall process.  This included the MOD’s recommendations and two of Mr Tugendhat’s recommendations.


69. With permission from the SRO, the member of Lord Ahmad’s Private Office sent the list to a Home Office Private Office to be sent on for Security Checks. However unfortunately this list appears to have been lost somewhere within the Home Office.  When I tried to check whether this list had been approved on the night of Monday 23 August, the Home Office appeared not to have processed it.


70. Soldiers called up Afghan evacuees and issued them travel documentation.  I judge this was likely because there was a shortage of FCDO staff and because soldiers could be instructed to serve shifts, including night shifts, whilst FCDO staff could only be requested to do so.


71. On the morning of Tuesday 24 August, the soldiers performed a count of how many people had already been called forward to the airport and decided on this basis how many more people could be called for evacuation.  As the number of slots still available was considerably smaller than the number of prioritised people returned from Home Office after passing security checks, the Army therefore performed a further prioritisation by only calling up a limited number of people from their final spreadsheet. I believe that they did this by proceeding down the spreadsheet in the order names had arrived from the Home Office.


72. Although this was clearly a profoundly flawed approach to prioritisation, this was not the soldiers’ fault.  They could not have plausibly done anything else at this stage.  However clearly this final prioritisation should have been a policy decision rather than being made in this purely arbitrary way.  The FCDO had not made this policy decision because our team never managed to clarify how many evacuation slots we still had available.  Therefore circumstances required the soldiers to do it on an arbitrary basis.


73. I was impressed by the soldiers' professionalism.  However I believe that some of them were likely using Microsoft Excel or Microsoft Outlook for the first time in a professional context.  I understand that some administrative mistakes reflected this lack of experience, including sending 91 travel documents from the wrong email accounts which meant that we did not have a full record of them. Again, this was not the soldiers’ fault.  This was a predictable result of soldiers being asked to perform bureaucratic tasks for the first time in a high-stakes situation without any relevant experience or training.


74. Although I believe the soldiers held some form of MOD security clearance, this was not recognised by FCDO security. As such, they had to be escorted around the Foreign Office by FCDO staff to guard against potential espionage.  This wasted considerable time in a context where staffing was extremely constrained.


75. As of the morning of Tuesday 24 August, calls were made only in English. There was no capacity for making calls in Dari or Pashto.  It is possible the capacity to make calls in Dari or Pashto may have later been introduced on a limited scale later.  I understand many of the soldiers’ conversations with Afghans involved telling Afghans that only children under eighteen could travel with them and they would have to leave their older children in Kabul.


76. I was told the Dari text on the ‘call-up’ emails inviting Afghans for evacuation was inaccurate.  The Dari said that a printed version of the email was necessary to enter the airport but in fact a digital copy on a phone was fine, which PJHQ (Permanent Joint Headquarters, the part of the MOD which commands the UK’s overseas operations) confirmed to me.  Although less serious than other problems discussed, this reflects our lack of expertise as no one spoke Dari and therefore no one noticed this problem.


77. The soldiers calling up Afghan nationals for evacuation were issued a paper list of logins for the FCDO’s non-secure phone system. This phone system is not suitable for classified information of any type. Without these logins the phones do not work.  On the night of Monday 23 August the soldiers lost this paper list in the hand-over between shifts. This would have prevented them from calling any Afghan nationals to the airport. My colleagues and I obtained phone logins for them from my Fast Stream WhatsApp group, the British Embassies in Beijing and Tokyo (who were online), and other sources. 


78. I also emailed BE Washington to explain that we did not have phones for the soldiers and they needed to send phone logins to allow the evacuation to proceed. The British Embassy in Washington found the situation described so implausible that they reported my email to FCDO Security as clearly a Russian phishing attempt. On Wednesday 25 August, FCDO Security contacted me to say I had broken security rules and instructed me to email BE Washington to apologise for requesting they send phones. I was informed the correct course of action would have been to wait until the next morning and then request new logins from the relevant IT team. This would have wasted around 12 hours at a crucial moment to protect the integrity of an unsecure phone system. This would have delayed evacuation by around 12 hours and meaningfully reduced the chance of the circa 1000 people we were seeking to evacuate from getting to the airport by giving them much less time for multiple attempts often necessary to enter the airport. I sent an apology.


79. On the evening of Saturday 21 August, the soldiers were issued one FCDO computer for every two soldiers. These did not work because FCDO IT had not issued the passwords to unlock them. These computers were finally unlocked on the afternoon of Sunday 22 August. Until this, the soldiers worked with one computer shared between roughly eight people.  This obviously considerably reduced their efficiency and speed. I printed out A3 spreadsheets for the soldiers but this was no substitute for a computer.


80. The soldiers clearly needed computers to email travel documents to Afghans selected for evacuation. As noted in paragraphs 117-131, travel documents needed to be issued as soon as possible to increase the chance of the selected Afghans being successfully evacuated. The failure to issue soldiers with sufficient computers for over 12 hours clearly delayed the issuing of travel documents, it will therefore have reduced the chance of selected Afghans being evacuated, and consequently may directly result in the deaths of people unnecessarily left behind. 


81. For several hours on the night of Monday 23 August, we tasked groups of soldiers with reading and prioritising emails from Afghans appealing for evacuation.  I assume we did this because not enough FCDO staff were available.  I deeply respect everyone who serves in the British Army but clearly soldiers’ training is not intended to equip them for this type of task.  The soldiers were uncomfortable with this task.  One remarked that he belonged with his comrades in Kabul not processing emails in London.


82. Between Wednesday 25 August and Wednesday 1 September, I believe emails were processed by marking them with a flag once read but were not entered into a spreadsheet. I do not know what system has been adopted since 1 September.  I believe the purpose of this system was to allow the Prime Minister and the then Foreign Secretary to inform MPs that there were no unread emails. It is difficult to see what other purpose this method served because these emails will need to be reread and prioritised before any action can be taken.


83. After the Observer’s story on Saturday 29 August on the backlog of unread emails, the ARAP and Afghan Special Cases inboxes were locked.  The ARAP and Afghan Special Cases teams lost access to these emails for several hours. 


84. This could be interpreted in two different ways.  In my opinion, this was an admission that at this stage the FCDO’s method of processing the emails only served a public relations purpose.  On the other hand, if the FCDO believes that this method of processing the emails served any real policy purpose, then locking the inboxes undermined rescue efforts for public relations purposes.


85. Given the very limited number of evacuation spaces available, effective prioritisation was particularly important.  Given the system adopted, I believe this prioritisation was likely not performed effectively.  Some decisions made are likely impossible to justify.  For example, I understand that we evacuated the BBC’s Afghan cooking and cleaning staff. Although I wish these people the best, it is impossible to justify why they were prioritised above interpreters or others who were at much greater risk and had performed much greater services to the UK.  I believe this happened due to a misunderstanding about the BBC’s list of names.



86. This was an appalling task and mistakes were inevitable.  It would have been impossible to design an entirely satisfactory system for this task. However, in my opinion, it would surely have been possible to design a considerably better system than the one employed by the FCDO.


87. The Ministry of Defence used a prioritisation system for the LOTR scheme based on strong endorsements by a serving or former British officer of field-rank or above.  Although imperfect, in my personal opinion this system is likely to have been considerably more rigorous than the often arbitrary system that was used by the FCDO. An equivalent for the FCDO might have been asking former Ambassadors to Kabul, Heads of Station, and other senior diplomats to nominate their most important Afghan contacts for evacuation.


88. On Saturday 21 August, the MOD sent me a list of three priority individuals with a senior British sponsor and a brief justification for why they were priorities for evacuation. Two of these individuals were recommended by the Chief of Defence Staff. Without knowing the full details of the MOD’s prioritisation system, this suggests the MOD’s LOTR prioritisation system was considerably more organised and coherent than the FCDO’s. 


89. A senior official and I briefly discussed our prioritisation system overnight on the 21-22 August. I said I was not confident in our prioritisation system. The senior official said ‘I used to be confident that we had a system. I have lost confidence that we have a system.’ They did not say this lightly.  In my view it is to the official’s credit that they recognised the grave problems with our system.


90. On the night of 21-22 August, one of the other three team-members working overnight said they were ‘disgusted’, intended to resign, and hoped there would be a full inquiry so the people would be held accountable. 


2) Inadequate staffing, including lack of night shifts.


91. On the afternoon of Saturday 21 August, I was the only person monitoring and processing emails in the Afghan Special Cases inbox.  No emails from after early Friday afternoon had been read at that point. The number of unread emails was already in the high thousands, I believe above 5000, and increasing constantly.  Nonetheless, a colleague favourably compared this situation to the ARAP inbox; there were even more unread appeals for help from people who had worked directly for the UK (interpreters, guards etc) in the ARAP inbox.


92. My impression is that no one at all had been processing these emails on the morning of Saturday 21 August, although it is possible that a colleague may have spent some of their time on them. 


93. Two other people were working on the Afghan Special Cases team at this time; however both were primarily engaged on other tasks. One of these was not technically on shift but had decided they were morally obliged to stay.


94. Around four other people had been rostered to work on the Afghan Special Cases team but had not come on shift. I am not sure of the reason for this.  Possibly they had not seen the rostering email asking them to come on shift, or their names had been left off the mailing list, or they had been accidentally rostered on multiple teams.   In general, staff availability appears to have been lower over the weekend.


95. I had not originally been rostered on Saturday but had decided on Friday that I was morally obliged to put myself down because I saw the Special Cases team was not fully staffed.  If I had not, it is possible there would have been no one to process the emails at all.


96. From Sunday 22 August, staffing levels somewhat improved. There were normally roughly six people working on the Afghan Special Cases team at any one time. Given the volume of emails received, this was still inadequate and we consistently had a backlog of thousands of emails at any given point.  A group of former DFID colleagues volunteered to help which also somewhat mitigated the problem.


97. On the nights of Sunday 22 August to Monday 23 August and Monday 23 August to Tuesday 24 August, I believe no night shift was allocated at all.  If a night-shift of any type was allocated, it was extremely token. I believe it is likely this decision was taken by the rostering team because few people had volunteered for the night shift. I understand that there were also often no night shifts, or only extremely token night shifts, on the consular and other key teams. Even putting aside the enormous urgency of the situation, given that Afghanistan time was then 3.5 hours ahead of UK time this is extremely difficult to justify.


98. At midnight on Monday 23 August, I asked for help from my Fast Stream WhatsApp chat to transcribe passport details from photos of passports to allow cases to be sent to the Home Office for security checks.  This is because passports had not consistently been transcribed and included in the list by the teams processing the email.  At one point I also asked one of my siblings to re-transcribe a group of passports after my computer crashed. FCDO security subsequently treated this as a security breach and breach of GDPR.


99. Turnover of personnel was extremely high.  My impression is that on the morning of Sunday 22 August, everyone working on the Afghan Special Cases team except the SRO and I  were new to the team and to the Afghanistan Crisis.  This naturally resulted in several hours of chaos while people attempted to work out what was going on.  The result was that despite having only worked on the crisis for one day, I was a relative expert on the process and asked to integrate the former-DFID volunteers into the system and explain the criteria on a Teams call.  For the same reason I also became a point of contact for the MOD on 21-23 August and to a significant extent for Lord Ahmad’s Office on 23 August by virtue of having been on the team for three days in a row.


100.                     This turnover was likely the result of the Crisis rostering team reassigning those who had previously worked on Afghan Special Cases to new teams. This was likely a result of the problems with rostering set out in paragraphs 106-109. After Sunday 22 August the composition of the team stabilized and I judge functionality increased progressively.


101.                     In my opinion, staffing shortages were exacerbated by some staff working from home, which hampered communication.  This was on occasion significant in a context where policy was poorly defined and the situation unclear. On Sunday 22 August and Monday 23 August, the majority of team members were working from home (including the responsible Bronze/Team Leader during Sunday 22 August).


102.                     I judge that the FCDO mobilised a higher percentage of its staff to work on the Covid-19 crisis in Spring 2020 than on the Afghanistan crisis.


103.                     Despite the urgency of the situation, the default expectation remained that FCDO staff would only work eight hours a day, five days a week. FCDO employees were only asked to work shifts for which they volunteered.  This likely resulted in a lack of night shifts and limited cover over the weekend because these shifts were less popular.  It also resulted in frequent changes of personnel due to days off. FCDO employees were not asked to work over periods for which they had indicated they were unavailable, even if this led to serious shortages of capacity.


104.                     I believe this reflects a deliberate drive by the FCDO to prioritise ‘work-life balance’. FCDO employees, including senior leaders, are often told that working more than eight hours a day suggests that they are inefficient.  Working more than eight hours a day has also sometimes been portrayed as selfish, as it potentially pressures other employees to do so as well.  Senior leaders are sometimes encouraged to set a good example by not sending emails outside working hours.  Whilst a healthy ‘work-life balance’ is important in moderation, in my opinion the FCDO’s approach has undermined organisational effectiveness.


105.                     Staff who worked more than their designated eight hours in the Crisis Centre were often encouraged to leave by colleagues.  Many FCDO staff worked extremely hard, especially but not only senior leaders, but this was optional.  Staff were not discouraged from leaving after eight hours even if there was an extreme lack of capacity. 


106.                     Staff are allocated for teams within the FCDO crisis response by a rostering team which compiles a daily roster which is then sent out by email. I worked a shift on the rostering team on the evening of Thursday 19 August.  I was placed on the rostering team because I replied to an all-staff email sent at circa 2:30 PM asking urgently for volunteers for a task that evening.  I believe any member of FCDO staff (or possibly anyone from an extremely broad FCDO mailing list) who wished to could have replied to this email.  To the best of my knowledge, all staff working on rostering on the evening of Thursday 19 August lacked any HR experience or experience of Afghanistan.  On this basis I then made decisions on who should be allocated to which team within the crisis response, including allocating staff to the ARAP and Afghan Special Cases teams where they would make life and death decisions.


107.                     Staff were allocated to roles on the basis of FCDO volunteers’ one-line summaries of their previous crisis and Afghanistan experience.  The responsibilities of particular teams were unclear;  in many cases, ‘rosterers’ may not have been fully aware of what each particular team was working on when they made allocation decisions.


108.                     Compiling spreadsheets manually often led to errors, with people frequently rostered onto two or more teams for the same shift. Furthermore, the crisis rosters were sent out by manually compiling a mailing list of everyone on the roster.   This may have led to staff not receiving the roster - and thus not reporting for their shift - if their email address had been left off the list. 


109.                     FCDO Staff were strongly advised to work on the Afghan crisis but usually not expressly instructed to do so. I was on language training in August and received a general request to work on the Afghan crisis. If I had wished to, I could have continued learning Arabic. By contrast, during the Covid crisis in 2020, I believe all language students were instructed to work on the crisis response.


       3) Lack of expertise.


110.                     I believe no member of the Afghan Special Cases team had studied Afghanistan, worked on Afghanistan previously, or had a detailed knowledge of Afghanistan.  My own knowledge of Afghanistan is largely limited to reading Rory Stewart’s book.   I believe the Senior Reporting Officer had also not worked on Afghanistan before.


111.                     Members of the Afghan Special Cases team usually heard of an Afghan organisation for the first time when they were asked to decide whether its staff should be evacuated.   As discussed above at paragraph 39, I assessed the relative vulnerability of staff in different sections of an Afghan human rights organisation on the basis of their departments’ names.


112.                     To give one example, on the night of Saturday 21 August (I believe the only night with a night-shift), I individually briefed the two people who came on shift to prioritise the emails. One was clearly scared by being asked to make hundreds of life and death decisions about which they knew nothing.   They initially essentially declined to do the task.  I persuaded them that unless they accepted the task the emails would not be read at all which would be worse. Their pass indicated they did not hold Developed Vetting; in a Foreign Office context this is reasonably likely to indicate that they were an A2 administrative officer or B2 executive officer, although this is not certain.  People at these grades are not usually responsible for policy work.  I judged they were a relatively new entrant to the FCDO. The second person prioritising the emails overnight was a second year fast-streamer.


113.                     The person who was acting as ‘Bronze’, (i.e. team leader) on the Afghan Special Cases team on the morning shifts of Sunday 22 August and Wednesday 25 August did not know that the correct term for people from Afghanistan was Afghans and referred repeatedly to ‘Afghanis’.  Although I am sure they meant well, in my opinion this indicates they did not have sufficient expertise.


114.                     As noted previously, in practice the Senior Reporting Officer was not in a position to meaningfully scrutinise prioritisation decisions made by the team. This was structurally inevitable and not their fault.


115.                     There was no access to additional information about organisations or individuals beyond what could be found on Google. There was no ability to process applications in any language other than English. There was no access to a Foreign Office Research Analyst to provide expert advice.


116.                     On Thursday 26 August, I received several requests for information from the Afghan Special Cases team about individual Afghans. I did not know this information and it could not be found in our inbox.  I feared if this information could not be found then these Afghans might be left to die unnecessarily.  I therefore asked Mr Tugendhat because he clearly knew much more about Afghanistan than the FCDO staff dealing with the crisis.  This was of course a technical breach of confidentiality.  Mr Tugendhat was able to answer the questions but requesting potentially life-saving information from the Chair of a Select Committee on a case-by-case basis does not appear to be a sensible way for the FCDO to manage expertise.


       4) Lack of urgency.


117.                     Conditions around the airport were evidently hell on earth.


118.                     This was for three reasons:


Reason One: Crowds


119.                     The UK controlled entrance to the airport, at the Baron Hotel, was surrounded by a crowd of at least 25,000 trying to enter.   Anyone trying to get to the airport had to fight their way through this crowd. 


120.                     There were frequent stampedes and crushes.  Small children were at grave risk of being crushed to death. Getting through this crowd often appears to have taken around 12-13 hours if not more.



Reason Two: The Taliban


121.                     As 31 August approached, the Taliban increasingly attempted to prevent Afghans without Western passports from entering the airport. To prevent Afghans from approaching the airport, the Taliban beat people with sticks. If people insisted on trying to get to the airport, the Taliban sometimes shot at them.


122.                     The MOD informed me that to get past the Taliban, evacuees sometimes had to show their travel documents.  These documents were only issued to people at risk of murder by the Taliban, and therefore effectively identified the individual as a Taliban target.



Reason Three: Extreme difficulty in communicating with the British soldiers.


123.                     Given the immense crowds, eligible people often could not get close to the British soldiers.  It was often difficult or impossible for the soldiers to identify eligible people in the crowd.


124.                     Most people trying to get into the airport had no means of communicating with the British soldiers or the FCDO Rapid Deployment Team in Kabul unless they could get close enough to the soldiers to wave their documents. This was often very difficult or impossible due to the crowds.


125.                     Helping people into the airport often involved sending groups of soldiers into the crowd to grab eligible families and bring them in. A member of the FCDO Rapid Deployment team in Kabul informed me that, due to the limited number of soldiers, resources for this type of operation were extremely limited.  This lack of soldiers was a key constraint on our ability to bring evacuees into the airport.  It was very unusual for LOTR cases to receive this type of treatment.  It was not policy to give LOTR cases the contact number for an FCDO staff member to facilitate entry to the airport.


126.                     I believe British nationals were sometimes in contact with the soldiers via the FCDO Rapid Deployment Team and the consular teams.  This was extremely unusual for Afghan Special Cases and most had no means of contacting the Rapid Deployment Team and British soldiers at Kabul Airport. I do not know to what extent ARAP cases had means of communicating with the British soldiers at the airport; my expectation would be that most would not have had means of communicating. 


127.                     An ISIS-K terror attack was expected, the soldiers were constantly in close proximity to the Taliban, and had a responsibility to prevent the crowd from entering the airport. I admire the soldiers’ courage and professionalism in very difficult and dangerous circumstances.


128.                     At times the soldiers appear to have had orders not to admit anyone without a UK Passport even if the FCDO had called them up for evacuation. I repeatedly heard of evacuees without UK passports but with valid travel documents issued by the FCDO being refused admittance by British soldiers. I am not sure if this arose due to operational reasons, such as lack of capacity at the airport, or because of miscommunication between the MOD and FCDO in London as described in paragraphs 139-146.  The Committee may wish to investigate this point. Responsibility for problems of this type lies in London not with the soldiers in Kabul.



129.                     In two cases I was aware of, people invited for evacuation said they were grateful that the UK had thought of them but it was impossible for them to enter the airport.


130.                     Due to these conditions, entering the airport typically required multiple attempts; families needed to risk the crowd and the Taliban multiple times to increase their chance of entry.


131.                     The Afghans whose emails we processed from Saturday 21 August to Tuesday 24 August were approved for evacuation overnight (UK time) on the night of Monday 23 to Tuesday 24 August. If they were called up, they were called up on Tuesday 24 August to Wednesday 25 August (UK time).  Given the time difference and that it was very difficult or impossible to get into the airport overnight, this likely de-facto gave most only the day of Wednesday 25 August and the morning of Thursday 26 August to reach the airport before the gate was permanently closed ahead of the ISIS-K attack.  It was foreseeable that we would have to permanently close the Baron Hotel gate on or around Thursday 26 August as we withdrew from Kabul airport ahead of President Biden’s withdrawal deadline of 31 August.  Withdrawal from the airport was necessarily gradual to allow British and American soldiers to manage the security situation as they withdrew.


132.                     I was told that it could often take as much as 12-13 hours to get past Taliban checkpoints and the crowd to reach the Baron Hotel gate, so it is likely these people only had enough time to make one or at most two attempts to reach safety.


133.                     I believe it became progressively harder to enter the airport between Monday Saturday 21 August and the final closure of the Baron Hotel gate to the airport on Thursday 26 August.  Firstly, the crowd at the airport appears to have grown larger and more desperate over time. Secondly, the Taliban announced repeatedly that they would not allow Afghan nationals to enter the airport and appear to have become increasingly violent towards people trying to reach the airport.


134.                     Of the limited number of people I was in direct contact with, I am not aware of anyone successfully entering the airport after Tuesday 23 August. 


135.                     Why were people not called up to the airport faster?


    1. Delay in equipping soldiers with computers. The soldiers were tasked with emailing a travel document to people prioritised for evacuation to allow them to enter the airport.  However they were not equipped with computers at a rate of more than 1 between 8 until the afternoon of Sunday 22 August. Clearly an adequate number of computers was a prerequisite for issuing electronic travel documents at an acceptable rate. The process of issuing visas and calling people forward to the airport for evacuation will inevitably have been delayed, likely by around 12 hours, due to the lack of computers. 


This delay is because there was no tech support available overnight and because only one tech-specialist was available in the daytime over the weekend.


    1. Extremely limited capacity, including me being the only person processing emails from the inbox on the afternoon Saturday 21 August.  This prevented applications from being processed at any meaningful rate on the afternoon of Saturday 21 August.  I believe emails had been processed at the same slow rate or not at all on the morning of Saturday 21 August.


    1. Failure to allocate meaningful night shifts. For the same reason, failure to allocate meaningful night shifts wasted time unnecessarily.


    1. The decision to compile a single list of circa a thousand people for submission to the Home Office on the evening of Monday 23 August, rather than submitting multiple shorter lists to the Home Office between Saturday 21 August and Monday 23 August led to unnecessary delay. If we had submitted multiple shorter lists, the soldiers could then have called individuals forward on a rolling basis as soon as they passed Home Office’s security checks. I believe this also led to a bottleneck in the Home Office’s security checks system because around a thousand names arrived simultaneously.


    1. Inability to immediately call-up all evacuees after security checks on the morning of Tuesday 24 August.  When I left at circa 10:30 AM on Tuesday 24 August, the soldiers had paused calls because they needed to count how many people we had already called forward to the airport in order to determine how many slots were still available before making further call-forwards. 


    1. Failure to understand the situation at the airport or the broader FCDO system. Most people in the FCDO crisis centre had a poor understanding of the actual situation at Kabul airport and the consequent urgency of calling people up as soon as possible.  I judge that the unspoken assumption was that once an Afghan had been invited to the airport for evacuation, they would be able to enter with relative ease. I shared this assumption until Tuesday 24 August.


    1. This in turn perhaps due in part to Lack of Sitreps. Our Situation Report (‘Sit-Rep’) system does not appear to have functioned. It is possible some Sitreps were written but personally I never saw any Sit-Reps or heard anyone refer to them. I am sure that senior officials had accurate information about the situation in Kabul, but for others information about the situation in Kabul often appeared to take the form of rumours or to come solely from the media.


    1. Lack of awareness of the evacuation timeline. On the weekend of 21 to 22 August, I assumed that evacuations could continue until 31 August itself or only shortly before. I judge this assumption may have been widely shared and contributed to a lack of urgency. In practice it appears to have been largely impossible to enter the airport after the morning of 26 August.


    1. Fatalism. On Sunday 22 August a ‘Bronze’ (team leader) said on at least two occasions that it was unlikely we would now be able to evacuate anyone further due to the situation.  This and similar misunderstandings are likely to have reduced the urgency of work on Sunday and Monday.


    1. Pause on calling Afghans forward for evacuation on the afternoon of Sunday 22 August due to the FCDO’s failure to coordinate with the MOD (paragraphs 139-146).


136.                     One interpreter I was in WhatsApp contact with abandoned his second attempt to get into the airport because his 9 year old child was crushed and was not breathing.  He naturally wished to take his child to hospital to be resuscitated.    British soldiers had located him but it would take at least several hours for them to reach him. He said he would return the next day.


137.                     I knew the airport gate would likely be permanently closed the next day. I tried to persuade him over WhatsApp that he had no choice but to continue.  He took the child to hospital and was not evacuated. He had twice been refused entry to the airport by British soldiers the previous day despite having an FCDO issued travel document. I imagine he thought if he waited his son might die and he might also be turned back again. He had received his visa too late because of the process flaws set out above.


138.                     Another person I was in contact with asked me what UK policy would be if he managed to reach the airport but his wife and tiny children had to stay in Afghanistan.  He believed his children would likely be crushed trying to enter the airport. He wanted to know if they could follow him later.  UK policy was not clear and I was unable to reassure him to any meaningful extent.


       5) Lack of coordination between the FCDO and MOD.


139.                     In the week commencing Monday 16 August, the Foreign Office submitted to the Foreign Secretary proposing the evacuation of (I believe) 5000 Afghans under the LOTR scheme. This proposal had not been seen by or agreed with the Ministry of Defence.


140.                     PJHQ asked me about this evacuation group on the afternoon of Sunday 22 August. PJHQ and the MOD had not previously been informed of this proposal. PJHQ then requested the FCDO pause LOTR call-ups on Sunday 22 August because these numbers had not been agreed and factored into the MOD’s planning assumptions.


141.                     On the evening of Sunday 22 August, a colleague from the MOD called me to warn me that senior officials at the MOD believed that the FCDO was calling-up an unspecified number of Afghans to the airport without consulting or informing the MOD.   My colleague said that the FCDO’s failure to coordinate with the MOD threatened to undermine Operation Pitting.  They said that as a result senior MOD officials planned to ask for the cancellation of the whole LOTR scheme the next morning.


142.                     I promised that I would emphasise to my superiors at the FCDO the need for realism and proper communication with the MOD.  I promised the MOD that I would confirm first thing the next morning (Monday 23 August) the number of people the FCDO had already committed to evacuate and planned on evacuating over the remainder of the week.  However, when I contacted the relevant team the next morning, they did not know how many people we had already called-up or planned to call-up.  They were counting and could not confirm when they would have this figure.  I was therefore unable to update the MOD.


143.                     On Monday 23 August, the same MOD colleague requested that the FCDO agree to the MOD sending its much shorter list of names for LOTR to the Home Office for security checks. An FCDO colleague requested that I review the MOD’s list of names in order to recommend to him which names to remove. I was unsure on what basis the FCDO possessed authority to veto or amend the MOD’s list and did not judge I was qualified to do this.  This would likely take at least several hours. I believe MOD’s names had been relatively rigorously selected through the process set out in paragraph 87.  I also believe these names had already been cleared at a ministerial level at MOD.


144.                     My MOD colleague was frustrated. Time was extremely limited before our withdrawal from the airport. Even if the FCDO possessed authority to veto these names, it would delay call-ups and therefore likely cost lives. I suggested MOD send the list to the Home Office for security checks but also to me.  After symbolically reviewing the list, I would then recommend to my superiors that it was not necessary to veto any names.


145.                     In response to our threat of vetoing names from the MOD’s list, the MOD decided not to send us their list. Conversely the FCDO could not send the MOD the list of names we had already called-up or planned on calling up because, despite asking a large number of people on the evening of Monday 23 August, I could not find who had it. Given the FCDO and MOD were both seeking to evacuate Afghan officers, there was a risk of individuals appearing on both the FCDO and MOD lists and therefore wasting evacuation slots. 


146.                     Given the ineffective communication between the MOD and FCDO, it was the responsibility of the National Security Council (NSC) to coordinate the two departments.   As far as I can judge, the NSC did not meaningfully attempt to do this. 


       6) Lack of coordination with our allies, including the United States.


147.                     The UK, the US, and other western countries did not coordinate our evacuation efforts.  Given the UK’s evacuation priorities were similar to our partners, this may have led to duplication of visas. Using contact details supplied by BE Washington, I emailed a moderately senior official in the US State Department on the evening of Monday 23 August to suggest the UK and US compared our respective evacuation lists to prevent duplication.  I subsequently talked to the official on the phone and they were very receptive to the suggestion. Unfortunately this initiative failed because the UK’s evacuation lists were not available and therefore it was not possible to share them with the US. US evacuation lists were organised in databases which could not be transferred by email.


148.                     A senior official in Americas Directorate expressed concern that sharing this information with the US would be in breach of European Data Protection Law (GDPR) because it would violate Afghan nationals' data security rights. The same official later gave me ‘feedback’ that senior diplomats in Americas Directorate were ‘quite surprised’ by my email to the State Department and they considered my email ‘shrill’ because it implied that people might die unless the US and UK coordinated our approaches.


    7) Lack of buses for ‘Afghan Special Cases’.


149.                     I understand buses were provided for some ARAP cases and UK nationals to reach the airport. I believe this was on quite a limited scale. Buses clearly heavily mitigated the three problems set out in paragraphs 119-128. The committee may wish to investigate whether other western countries used buses on a larger scale than the UK.


150.                     With one exception on Thursday 26 August, I believe no buses were provided for Afghan nationals evacuated under the Afghan Special Cases scheme. As a result evacuees had to endure the appalling conditions described in paragraphs 119-128.


151.                     I do not know why this was. I am not sure whether this was the result of a conscious decision by the FCDO Crisis Centre.


152.                     If there were a shortage of available buses, it would have been justifiable to prioritise ARAP and British Nationals. There is no evidence that there was a shortage of available buses. Mr Slater, whose company ran a secure transport service to the airport, informed me he offered to charter his buses to the UK.


     8) Assumption every visa-holder would reach the airport.


153.                     I believe that the number of visas granted was equal to the number of evacuation spaces judged to be available on flights from Kabul airport.


154.                     However,  I believe that the majority of Afghans we offered visas to from Monday 23 August did not manage to enter Kabul airport. I hope this is wrong but anecdotally it appears to be accurate.  No one I was in contact with or aware of managed to enter the airport after Monday 23 August.  It was foreseeable that many of those we granted visas to would not succeed in reaching the airport.  As such, it is possible spare capacity was available on UK evacuation flights.


155.                     In my personal view, it would have been wise to grant more visas than spaces were available, on the reasonable assumption that many of those granted visas would not be able to reach the airport. Whilst this may have somewhat increased the number of people trying to seek entry at the Baron Hotel gate, this would have been immaterial relative to the size of the crowd already surrounding the gate.


9) Ineffective cooperation between Ministers and civil servants. 


156.                     I was informed by a colleague in a position to know on Sunday 22 August that the Foreign Secretary believed that he had already approved a list of people to be called up for evacuation under the LOTR scheme.  However this list did not yet exist and we were in the process of creating it. The Committee may wish to review how and why this misunderstanding arose.


157.                     I believe the Crisis Structure was reporting directly to the Foreign Secretary not indirectly through the Minister of State responsible for Afghanistan, Lord Ahmad.  The Committee may wish to consider whether this was the most effective arrangement. Lord Ahmad and his office were highly engaged with the evacuation effort but lacked a clear locus to provide direction. Lord Ahmad and his team appeared not to be in a position to instruct the Crisis Centre as such but only to make requests and suggestions. As such Lord Ahmad’s position was ambiguous.  In my judgement Lord Ahmad and his office understood the bureaucratic problems with our evacuation effort and tried to work to solve them.   In my opinion, given Lord Ahmad’s engagement, the system would have worked more effectively if he had been given a clearer mandate to identify and resolve problems within the crisis structure and particularly in the ARAP and Afghan Special Cases evacuation processes.


158.                     On Sunday 22 August, a member of Lord Ahmad’s Private Office worked on the Afghan Special Cases team in the Crisis Centre. This is likely moderately unusual.  One potential explanation for this is that Lord Ahmad did not feel he was being kept adequately informed by the Crisis Centre and sent someone he trusted to inspect our process more closely.  The member of Lord Ahmad’s Private Office and I talked through the flaws in our prioritisation process and how to mitigate them on the evening of 22 August.  On Monday 24 August Lord Ahmad’s office lent a further member of their office to the Afghan Special Cases team.


159.                     Lord Ahmad and his office received numerous requests for the evacuation of individuals. Lord Ahmad’s Office kept an office tracker of these. On Saturday 21 August (and possibly also on the week of 16 August) some of these individuals were offered evacuation as ‘Lord Ahmad Special Cases’.  However my impression is Lord Ahmad’s Private Office believed that Lord Ahmad’s did not have an adequate mechanism for coordinating the requests for evacuation he was receiving from contacts with the crisis structure and ensuring those requests were receiving fair consideration. 


160.                     On Monday 22 August Lord Ahmad’s Private Office referred several cases or groups of cases to me for the team’s consideration. I was trying to juggle too many other extremely urgent tasks. I am not confident that I ensured these cases were fairly considered and fear some may have been dropped with potentially fatal consequences.


161.                     Given the many flaws of the prioritisation process, in my opinion it would have improved our evacuation process if Lord Ahmad had been empowered to formally instruct us to prioritise specific cases. In my opinion individuals recommended by Lord Ahmad and his Private Office likely had a higher chance of being justifiable priorities than the individuals selected through our prioritisation process.


162.                     On Wednesday 25 August as it became clear we were approaching the end of evacuations, numerous departments and Secretaries of State contacted the Foreign Secretary’s Office and the Crisis Centre to ask why specific individuals or groups of individuals had not been selected for evacuation.  For example I believe the Secretary of State for Education contacted us about a group of Afghan academics.  The volume of senior requests for reconsideration received was so high that only those from a Secretary of State could be considered. I believe the Foreign Secretary’s Office compiled a tracker of all these requests.


163.                     The SRO had been receiving interventions of this type from across Whitehall throughout the evacuation process.  For example on Monday 23 August the Permanent Secretary of the Department of International Trade had highlighted a case.


164.                     These requests presented a dilemma.  Given we had already allocated our evacuation capacity on Monday 23 August, we likely did not have capacity to evacuate the vast majority of these additional people. In my opinion, there is reason to believe a significant number of people raised by Secretaries of State were likely to be more eligible and deserving of evacuation than many of the people we had previously selected for evacuation. However by Wednesday 25 August, given we could not rescind issued evacuation ‘call-forwards’ there was nothing to be done about this.


165.                     The responsible senior official was visibly drained by their responsibility to prevent individually extremely compelling requests from multiple Secretaries of State resulting in us calling forward an un-viably large number of additional people for evacuation.  It was their duty to prevent the extremely limited processing capacity at the Baron Hotel from being overwhelmed by additional people resulting in the breakdown of our process. This was clearly a deeply harrowing task for them. 


166.                     The British Embassy Kabul guards employed by GardaWorld had not been prioritised for evacuation through the Afghan special cases prioritisation process. I do not know why, but imagine this may have been the result of a failure to process GardaWorld’s email or similar oversight rather than a deliberate decision.


167.                     On 25 August, GardaWorld made a personal appeal to Sir Laurie Bristow, HMA Kabul, to evacuate 125 guards and their families.  Sir Laurie made a personal promise to GardaWorld that if the embassy guards could reach the airport on 26 August he would do what he could to secure their evacuation.  I understand, the guards attempted to enter the airport on 26 August but were prevented by ISIS-K’s suicide bombing.


168.                     The FCDO crisis centre in London was frustrated with Sir Laurie for making this promise because capacity at the airport was so limited. This was due to a concern that the 130 guards and their families would ‘swamp’ the limited available capacity at the Baron Hotel Gate. The Crisis Centre attempted to insist Sir Laurie reverse his promise due to limited capacity. I believe he refused to do this.


169.                     For the same reason of lack of capacity, British soldiers were instructed to forcibly remove ineligible people from the Baron Hotel. It was said that the soldiers in Kabul had become good at forcibly removing ineligible people from the Hotel because of the number of occasions on which it had been necessary. In my opinion, forcibly removing people ineligible for evacuation from the Hotel to free capacity was clearly necessary and justified.


170.                     On 25 August extreme measures were therefore being taken to preserve the extremely limited capacity at the airport. The measures taken include the Foreign Office requesting HMA Kabul not evacuate the embassy guards, British soldiers (entirely correctly) forcibly removing people from the Baron Hotel, and the Foreign Office refusing numerous requests from Secretaries of State to evacuate highly eligible people. In this context, we received an instruction from the Prime Minister to use considerable capacity to transport Nowzad’s animals. The Prime Minister’s instruction is discussed in paragraphs 205-221.


171.                     To address some of the numerous requests for the reconsideration of particular cases we received, the Crisis Centre adopted a system of submitting exceptional cases to the Foreign Secretary for approval. We therefore wrote notes or submissions on a series of individual cases which arose for the Foreign Secretary’s approval.


172.                     It took several hours for the Foreign Secretary to engage on any of these notes. In the circumstances, I am not sure why. The Foreign Secretary then replied through his Private Office to say that he could not decide on individual cases and he would need all the cases set out in a well-presented table to make decisions. I understand that he or his Private Office had commented that as a lawyer he could not take information without the full facts in a table. We therefore reformatted the table and sent it back to the Foreign Secretary.


173.                     In my opinion, for the Foreign Secretary to make this request suggests he did not fully understand the situation. Firstly there was very little time left for anyone to enter the airport, therefore  Foreign Secretary’s choice to cause a delay suggests he did not understand the desperate situation at Kabul Airport. Secondly, the Foreign Secretary’s reluctance to authorise any exception to our prioritisation process indicates he did not understand the, at best highly approximate, nature of our process.


174.                     If the Foreign Secretary felt he did not have sufficient information to make a decision, it would have been extremely reasonable for him to defer to the judgement of the Crisis Centre on this question by merely accepting the Crisis Centre’s recommendation that these people should be evacuated.  However in the circumstances it is hard to explain why he reserved the decision for himself but failed to make it immediately. In my opinion, it is likely that the decisions the Foreign Secretary initially declined to make were less ambiguous than decisions made by relatively junior FCDO employees.


175.                     On Wednesday 25 August, Lord Ahmad and the Minister for Armed Forces, James Heappey MP instructed us to evacuate a senior Afghan soldier, his adult children, and his small grandchildren.  The Afghan soldier had been prioritised for evacuation but his adult children had not, although they had separately applied for evacuation. The soldier judged that if his children were left behind they would certainly be murdered.  The MOD and Mr Heappey agreed that his children would be murdered. The Afghan officer said that as an honourable soldier and father he would not leave his adult children and infant grandchildren behind; he would rather remain and die with his family.


176.                     The Crisis Centre initially judged that Lord Ahmad’s and Mr Heappey’s instruction should not be implemented because the sons had not been prioritised by our overall evacuation process. I said that this process was very arbitrary; it is plausible that the applications submitted by the senior soldier’s adult children had simply not been read. We agreed to submit this case to the Foreign Secretary. As noted, the Foreign Secretary did not respond for several hours. I believe this family did not succeed in entering the airport. The bureaucratic delay may have been a factor.


177.                     A group of seven people, said to be women’s rights activists, had been promised evacuation by the Home Secretary. Their case was supported by senior executives at Sky News and were therefore referred to as the ‘Sky Seven’.  It was not clear on what basis the Home Secretary had agreed to this and we did not know the exact identify of these people.  We included them in our submission to the Foreign Secretary, I believe including noting their connection with Sky News.


178.                     On the evening of Wednesday 25 August, they were waiting by the Baron Hotel gate, which was about to close for the night.  The Foreign Secretary declined to make a decision on whether to admit these people without a properly formatted submission with a table setting out multiple cases.


179.                     In the best case scenario, this would have resulted in this group waiting in the midst of the 25,000 strong crowd without shelter overnight surrounded by Taliban fighters. The Crisis Centre therefore instructed Kabul to admit these people to the airport in anticipation of the Foreign Secretary’s instruction.


10) Failure to plan for after 31 August and lack of urgency granting further visas.


180.                     I believe that since Wednesday 25 August very few further visas have been granted, except to the Nowzad’s employees. 


181.                     I believe that, at least initially, little or no provision has been made by the FCDO or HMG to assist Afghans with UK visas to cross the Pakistani or Uzbek borders for onward travel to the UK.


182.                     An exception was made for Nowzad’s staff who received visas and assistance in crossing the border.  As set out in paragraphs 205, the MOD and the FCDO agreed that there was no meaningful risk to these people.  As set out in paragraph 218, this judgement has subsequently been proved correct.  As set out in paragraph 215-216, Nowzad’s staff were not eligible for evacuation under our criteria.


183.                     I believe assistance crossing the Pakistani Border was provided for Nowzad’s staff.  This demonstrates that HMG could have provided evacuees with assistance across the Pakistani border if it wished to but chose not to.  The Committee may wish to investigate why HMG made this choice.


Part Two: Additional Evidence


The FCDO’s internal investigation into breaches of the Civil Service Code during the Afghanistan Crisis Response



184.                     On Tuesday 31 September I emailed the Permanent UnderSecretary to report the problems detailed in this evidence as breaches of the Civil Service Code.


185.                     I met the Permanent UnderSecretary on the same day and set out the concerns listed in this evidence. The Permanent UnderSecretary subsequently appointed a credible senior official to investigate. The then Foreign Secretary Dominic Raab MP approved the terms of reference of this investigation. I spoke at length to the investigator.


186.                     On 29 September the Permanent UnderSecretary briefed me on the outcome of the investigation.  I thanked the Permanent UnderSecretary for appointing a credible investigator.


187.                     The Permanent UnderSecretary said the investigation had found no breaches of the Civil Service Code, but had  identified several lessons for the FCDO to learn from the Crisis.  These lessons would be duly fed into a broader internal review of the FCDO’s Afghanistan policy since President Biden’s announcement that the United States would withdraw from Afghanistan by 31 August 2021.


188.                     Overall the investigation had found that the FCDO’s Afghan Special cases evacuation sought ‘to deliver the right outcome with some success’. This strikes me as a notably qualified endorsement.


189.                     The Investigation had found that staffing was deliberately lower on the Afghan Special Cases team than the ARAP team or the Consular team. This is because the Leave Outside the Rules scheme was a less established evacuation route. On the wider resourcing of the crisis, the investigator had identified some lessons for the FCDO especially on staff turnover.


190.                     With regard to the prioritisation process, the investigation found that there was insufficient time to design a better process because the LOTR scheme was only approved six days before the end of evacuations, i.e. circa 20 August around 4-5 days after the fall of Kabul.


191.                     The investigation also found that everyone evacuated was eligible according to the three relevant criteria with one specific exception, where criteria had not been followed. This was the result of a ministerial decision taken elsewhere in Government. Clearly this was a reference to the Prime Minister's instruction to evacuate Nowzad’s staff.


192.                     I said I was unsurprised that the inquiry had found that FCDO staff were indeed seeking to implement HMG policy.  I said that my concern had not been that anyone had deliberately undermined the evacuation effort, but that the FCDO had not implemented HMG policy as effectively as the public had a right to expect. 


193.                     I asked whether in the FCDO’s understanding, the Civil Service Code was in-effect not concerned with efficacy but only with deliberate impropriety. The Permanent UnderSecretary said it would be fair to say that the Code was not primarily concerned with efficacy. The terms of reference of the investigation appear to have been relatively restricted. Although understandable, this may in-part explain the outcome of the investigation.


194.                     I respect the investigator’s integrity and I am grateful to the Permanent UnderSecretary for appointing a credible investigator. I have not read the investigator’s report.  However, in my opinion if the FCDO’s interpretation of the Civil Service Code is correct, this strongly calls into question the utility of the Code.  The FCDO’s interpretation of the code appears to imply that only explicit corruption or deliberate sabotage of government policy would breach the code. 


195.                     The Civil Service Code requires that civil servants ‘deal with the public and their affairs fairly, efficiently, promptly, effectively and sensitively, to the best of your ability’.


196.                     The evacuation of ‘Afghan Special Cases’ from Kabul was clearly a public affair of an urgent and unusual character. As I understand it, the FCDO’s investigation therefore found that the FCDO executed this task ‘fairly, efficiently, promptly, effectively and sensitively, to the best of your ability.’  In my opinion, this is not a reasonable finding.


197.                     I believe the investigation found that there was a functioning, if imperfect, prioritisation system in place. I believe that the investigation also found that I was not personally making life or death decisions but only advising on decisions as part of this system.  In my opinion, this is unfortunately not an accurate assessment of the ‘system’ employed.


198.                     With regard to the suggestion that, aside from Nowzad’s staff, the criteria were followed:


    1. There was no consensus to the meaning of the criteria of ‘significance’ and it was not applied consistently. Therefore it is clearly not accurate to say that all the cases were eligible according to the criteria of ‘significance’.


    1. It is correct that everyone called forward for evacuation was judged to be eligible according to the criteria of ‘vulnerability’ and ‘support for UK objectives’. However, given that a majority of people who emailed the Afghan Special Cases inbox were eligible for these criteria, this in no way indicates that we employed a functional prioritisation system.


The key task of the Afghan Special Cases team should have been to identify which of the 10,000s of eligible applicants the UK’s obligations and national interest dictated we should prioritise rescuing.  Merely stating that those selected for evacuation were eligible according to the criteria is therefore not a credible or substantive response to criticism of the prioritisation process employed.


199.                     In my opinion the finding that it was impossible to design a better system in six days is not credible.  I believe that within this timeframe it would have been possible for the FCDO to design a significantly better system than the one described in this evidence.  Available evidence strongly suggests that in the same timeframe the MOD created a prioritisation system of considerably greater rigour for the selection of the MOD’s LOTR cases.  Moreover:


    1. A large portion of this time was wasted due to understaffing, excessive rotation of staff, and lack of meaningful night-shifts.


    1. The Afghan Special Cases inbox was created in early August. Therefore the LOTR scheme must have been conceived in early August even if not formally approved until later. The FCDO could therefore have spent the time before the scheme was formally approved preparing lists for evacuation.


    1. By this account the LOTR scheme was created circa 19-20 August, 4-5 days after the fall of Kabul on 15 August. It is unclear how this delay can be satisfactorily explained.


    1. The Chief of Defence Staff told the Defence Committee on 9 November that the MOD’s planning for the Operation Pitting evacuation began in January 2020. It is therefore unclear why by contrast the civilian planning for the evacuation was seemingly not finalised until 4-5 days after the fall of Kabul.


200.                     It is justifiable that the Afghan Special Cases Team was a lower priority than the Consular and the ARAP teams. However:


    1. Staffing on the ARAP team was clearly also inadequate given the backlog of thousands of unread emails which accumulated. 


    1. If the FCDO believes staffing on the Afghan Special Cases team was adequate, it is necessary for the FCDO to explain why it was justifiable for only one member of staff (me) to be monitoring the Afghan Special Cases inbox on the afternoon of Saturday 21 August.


    1. The FCDO has circa 17,300 employees, most of whom are in policy-roles and continued their normal roles in August. Therefore it would not have been difficult for the FCDO to divert a sufficient number of other employees from other less important tasks.              


201.                     The investigation also found that there was no resistance to ministerial instructions and FCDO staff should receive training in how the FCDO relates to ministers in a crisis. I agree there was no resistance to ministerial instructions as such, but this does not address the more complex question of whether FCDO officials worked effectively with FCDO Ministers of State and the Foreign Secretary during the crisis. The implication of this second statement is that I had misunderstood the correct relationship between the FCDO and FCDO Ministers of State in a crisis. The Committee may wish to ask the FCDO to elaborate on the FCDO’s assessment of the correct relationship between the Crisis Structure and FCDO Ministers of State.


202.                     Although I am glad the FCDO recognises it has lessons to learn, in my opinion the problems identified were either highly foreseeable or ought to have been identified from the Covid Crisis.


Statements by the Prime Minister in the House of Commons on 6 September.


203.                     Following the Sunday 29 August Observer Story, the Afghan Special Cases Team was instructed to respond to every email ahead of the parliamentary debate of 6 September. On the basis of this instruction the Prime Minister informed the House of Commons on 6 September that all outstanding emails would be responded to by the end of that day.


204.                     On receiving this instruction on Sunday 29 August, my colleagues correctly remarked in the team's group chat that this was impossible given there were roughly 5-6 people on shift at any one time on the Afghan Special Cases team and tens of thousands of unanswered emails.


The transport of Nowzad’s staff and animals. Risk to British soldiers involved.


205.                     On Monday 23 August, we received a ministerial request to consider the evacuation of the animal NGO Nowzad.  I assumed this request referred to the NGO’s Afghan staff.  I believe this request came from the office of Wendy Morton MP, the Minister for Europe. Ministerial interventions to highlight particular cases were routine and perfectly appropriate. 


206.                     I was informed that Nowzad could supply its own plane and had therefore requested assistance outside the normal evacuation process.   Superficially I judged this to be reasonable.


207.                     I consulted PJHQ’s embed in the Crisis Centre, who noted that Nowzad's ability to supply its own plane was not relevant because the FCDO could charter any number of planes. The relevant constraint was the limited number of soldiers available to bring eligible people into the airport and limited capacity within the airport. PJHQ noted that at least seven people had already been crushed to death trying to get into the airport.


208.                     PJHQ pointed out correctly that there was no reason to believe the Taliban would target animal rights charities.  There was therefore no justification for concluding  that Nowzad’s staff were at significant risk.  By contrast many others would inevitably be left behind who were at risk of murder. 


209.                     Similarly the protection of domestic animals was not a UK war aim in Afghanistan. 


210.                     PJHQ therefore noted that even if Nowzad supplied its own plane, it would be unethical to prioritise Nowzad’s Afghan staff. This conversation concerned the evacuation of people rather than animals. My colleague replied to the relevant Private Office to this effect. 


211.                     My colleagues and I eliminated thousands of Afghan friends of the UK at risk of murder from the evacuation lists. We were instructed to do this due to lack of capacity to process people at the airport. On Wednesday 25 August, many people referred by Secretaries of State were rejected due to limited capacity.  This capacity was subsequently used to transport animals.


212.                     There was a direct trade-off between transporting Nowzad’s animals and evacuating British nationals and Afghans evacuees, including Afghans who had served with British soldiers.


213.                     This is because soldiers tasked with escorting the dogs through the crowd and into the airport would by definition have otherwise been deployed to support the evacuation of British nationals or Afghans prioritised for evacuation, notably by helping families out of the dangerous crowd into the airport.  As noted in paragraph 125, the limited number of British soldiers available to help UK visa holders and British citizens from the crowd into the airport was an important limiting factor on our ability to evacuate people.


214.                     Nowzad’s staff were subsequently granted visas for the UK and assisted to cross into Pakistan. This was an exception to our then policy.  At this point, we were not granting further visas to any other Afghans even if they had served the UK and were at direct risk of murder.


215.                     Our three criteria for evacuation were;

    1. Vulnerability.  As noted in paragraph 208, there were no grounds to conclude the staff of this charity were at higher risk from the Taliban than the average Afghan.
    2. Support for UK objectives in Afghanistan.  The protection of domestic animals was not a UK war aim in Afghanistan.
    3. Significance. It is not clear on what grounds the staff of this charity could be judged to be uniquely significant.


216.                     I wish the staff of Nowzad all the best in their new life in the UK but they were not eligible for evacuation let alone for exceptional assistance which prioritised them above British Army interpreters. The FCDO’s internal investigation also determined that the Nowzad staff were not eligible for evacuation according to these criteria.


217.                     I believe that British soldiers were put at risk in order to bring Nowzad’s animals into the airport.  It has been reported that British soldiers ventured into the crowd around Kabul airport to clear the way for Nowzad’s vehicles. At this point, the crowd was estimated to exceed 25,000 people, with many Taliban fighters present. There were many reports of people being crushed to death.  On 26 August, 13 US Marines and around 170 civilians were murdered by an ISIS-K suicide bomber in the same area of the airport.  In these circumstances, any task which involved British soldiers venturing into the crowd presented a meaningful risk to the soldiers.


218.                     I understand that an American animal charity is still operating in Kabul managed by an American. I understand the animals and staff of this charity have not been subject to any mistreatment by the Taliban.  I understand that far from being harassed by the Taliban, this charity has been able to hire members of the Taliban.   This vindicates the MOD and FCDO’s belief that neither Nowzad’s animals nor its staff were at risk from the Taliban. 


219.                     HMG therefore transported animals which were not at risk of harm at the direct expense of evacuating British nationals and people at risk of imminent murder, including interpreters who had served with the British Army.


220.                     I expect HMG may respond by noting that Nowzad supplied a crowd-funded plane for the animals. This is not relevant. The problem was not the lack of planes. The problem was lack of soldiers and lack of capacity at the airport.


221.                     I suggest the Committee requests answers to four questions from HMG:

    1. What was the policy justification for the decision to ‘evacuate’ Nowzad’s animals?
    2. What was the policy justification for the decision to put British soldiers at risk in-order to facilitate the ‘evacuation’ of Nowzad’s animals?
    3. Given Nowzad’s staff were not eligible for evacuation under the criteria of the LOTR scheme, what was the policy justification for prioritising their evacuation over tens of thousands of people who were eligible for evacuation under the criteria of the scheme?
    4. Nowzad’s staff received help travelling to Pakistan but not others with UK visas, including British Army interpreters, did not. What was the policy justification for assisting Nowzad’s staff in crossing to Pakistan but not British Army interpreters?


FCDO Planning for the Fall of Kabul


222.                     The majority of the UK’s Embassy to Afghanistan was evacuated in July. A colleague who served in the Embassy told me that this was because of fear that Kabul would fall in the short-term.


223.                     My colleague said that the UK had requested a promise of United States assistance evacuating the British Embassy to Afghanistan in the event of a Taliban advance on Kabul. The United States had declined their request. I was told that this was because if the United States promised the UK assistance it would also have to promise assistance to other NATO allies, which the United States believed would be an unmanageable obligation in the context.


224.                     Without United States assistance, the FCDO feared it would be impossible to safely evacuate the Embassy in the event of a Taliban advance on Kabul.


Benjamin Slater and Nomad Group


225.                     The Afghan Staff of the development and security company Nomad Group, owned by British veteran Benjamin Slater, were not prioritised for evacuation from the airport.  Mr Slater was formerly a Royal Military Policeman and a bodyguard to Lord Sedwill during Lord Sedwill’s service as NATO’s Senior Civilian Representative in Afghanistan.


226.                     As a British national, Mr Slater could have entered the airport at any time.  He made multiple trips to the airport to help his Afghan employees and contacts enter. Instead of entering the airport, he stayed in Kabul to try to save his staff.


227.                     Nomad’s staff were originally included in our second ‘priority’ spreadsheet on the evening of Monday 23 August.  When this spreadsheet was cleared by a senior official, I believe they instructed the team to prioritise the most vulnerable people on the Nomad list but remove others.  This is because the Nomad list included several hundred names and only a very limited number of evacuation slots were judged to be available.  It was the senior official’s duty to ensure institutional lists were correctly prioritised. Their instruction was proper and sensible.


228.                     At this point I was asked to prioritise Nomad, an Afghan Human Rights organisation, an Afghan Women’s Football Team and the Afghan Paralympics team.  I believe these collectively amounted to over 500 names.  I was also simultaneously trying to contact the United States.


229.                     I did not believe I could prioritise the names on the Nomad list in a justifiable manner as it did not contain the information I needed. I sent Mr Slater an email requesting that he prioritise the list further. Mr Slater replied to emphasise the extreme importance of his staff receiving visas.


230.                     Mr Slater’s email read ‘my staff are not animals’.


231.                     In parallel two of my colleagues further prioritised our list by removing further names.  I am sure they carried out this deeply distressing task with the greatest integrity and professionalism possible in the circumstances.  I believe that as part of this prioritisation exercise my colleagues entirely deleted Nomad from the evacuation list as not a priority. 


232.                     Our team had previously been confused by the Nomad Group.  The Nomad group combined several separate for-profit and not- for-profit organisations with a wide range of activities including both security, women’s empowerment, and development activities. As such it was hard to categorise. We had speculated that the Taliban might target Nomad less because it was a company.


233.                     I understand Mr Slater’s employees have been subject to systematic harassment and intimidation by the Taliban including threats of rape, ‘forced marriage’, and murder.


234.                     Lord Sedwill subsequently emailed me to endorse Mr Slater and ask that the merits of Mr Slater’s request for Nomad’s staff to be evacuated be fully considered. 


235.                     I left the FCDO at around 11:00 AM on Tuesday 24 August.  Before I left I forwarded Nomad’s case to the Bronze who was on shift and asked them to pick this up urgently. 


236.                     I later saw that they later replied to ask me to pick up Mr Slater’s email the next day, Wednesday 25 August.   This was because any reconsideration of Mr Slater’s case at this stage would have required an exception to our process because we had already sent our final list to the Home Office.


237.                     Mr Slater WhatsApped me on the morning of Tuesday 24 August at circa 11:30 AM as I was leaving the FCDO. 


238.                     He wrote ‘I desperately need to get my people on an evacuation flight…. I have to do my best for these people’.


239.                     I promised I had forwarded his email to a colleague and asked them to pick-it up.


240.                     Mr Slater said ‘I just hope my email gets looked at.’


241.                     Later on Tuesday 24 August Mr Slater WhatsApped me in the course of trying to escort some of his Afghan staff with Canadian visas to the Baron Hotel gate.


242.                     It was dark and Mr Slater was close to Taliban fighters. He had just failed to get some of his staff into the American-controlled section of the airport.  He said he had been ‘shot at’ by the Americans. Mr Slater later told me that he had been shot with a rubber bullet by an American soldier.


243.                     On Thursday 26 August or Friday 27 August, a British soldier informed Mr Slater by phone that HMG would not evacuate his staff. Mr Slater told me the soldier cried during this conversation because of the message he had to convey.


244.                     I understand from Mr Slater that HMG subsequently communicated that Mr Slater’s employees would be granted visas for the UK and allowed to cross to Pakistan if they travelled to the Afghan-Pakistan border. I understand from Mr Slater that if he had not received this assurance, he would not have put himself and his group in this exposed position by travelling to the border. However after the group had arrived at the border, HMG only granted a visa to Mr Slater himself and one other person. I believe this decision was made at a ministerial-level within the Home Office. Although the UK declined to assist, at this point Australia granted 25 of Nomad’s staff visas and assisted these 25 people in crossing to Pakistan.


245.                     In the Wednesday 1 September Foreign Affairs Committee, the Foreign Secretary said that it was impossible to help this group cross because further security checks were necessary. The Home Office informed me they had done security checks on 50 of Mr Slater’s employees on Thursday 26 August. Any further security checks could easily have been completed in the available time.


246.                     Mr Slater again refused to abandon his staff and declined to cross alone.


247.                     I believe Lord Sedwill and Mr Tugendhat both raised this case with senior decision-makers in Whitehall. I also mentioned this case to colleagues in the FCDO. I was politely asked to stop raising this case as it had ‘become political’.  On 31 August the Permanent UnderSecretary confirmed to me that he was aware of the Nomad Group case.


248.                     On the night of 1 September, the Taliban raided the hotel where the Nomad Group employees were staying. Mr Slater and six of his employees were then detained and imprisoned by the Taliban.  The detention of Mr Slater and his staff by the Taliban was arguably a foreseeable consequence of HMG’s decisions.


249.                     On 1 September, one of Mr Slater’s employees called me. She had managed to escape from the hotel and was hiding from the Taliban with a group of other employees.  She was 21 and terrified. She believed she would be murdered if she did not cross the border immediately.  The British High Commission in Islamabad could not help her because they had not been instructed to do so by London.


250.                     Mr Slater’s employee WhatsApped me to say:


‘I do not know where to go and what to do? I do not know why it is my fault for working for British people, now we are about to be killed but no one can really care.  Can you please talk to about only for my VISA i mean only one visa. I am asking for one visa.’


251.                     Mr Slater WhatsApped me from the Taliban prison where he was being held to say that HMG’s evacuation efforts were ‘disgusting’ and a ‘disaster’.


252.                     As has been reported in the media, Mr Slater was subsequently held captive for six weeks by the Taliban and tortured repeatedly. He was held with a substantial number of ISIS-fighters and witnessed a number of executions.  Whilst held captive, he continued attempting to coordinate his staff’s evacuation.


253.                     On 5 October Dr Martin Longden and Sir Simon Gass visited Kabul for talks with Taliban interlocutors.  The Taliban released Mr Slater and he returned to Doha with the delegation.  I believe Dr Longden and Sir Simon’s trip to Kabul on 5 October was the first official diplomatic visit to Kabul by any western country since 15 August.


254.                     The Committee may wish to investigate whether this visit represented an acceleration of HMG’s plans for engagement with the Taliban partly because, after HMG had arguably caused Mr Slater to be kidnapped and tortured, it was necessary for HMG to engage with the Taliban in return for his release.











November 2021