Annual Report 2003 - 2004

ANNEX 3 - INQUIRY INTO CONCERNS RAISED ABOUT DIO BY LT COL LANCE COLLINS

Background

1. Mr Collins, a Lieutenant Colonel in the Australian Defence Force, wrote to the Minister for Defence on 6 December 2000, expressing concerns about the Australian defence intelligence system. A copy of his letter is at Annex 1.[8]

2. The minister wrote to me on 20 December 2000, asking me to investigate, report and make recommendations in respect of the issues raised by Mr Collins that fell within the scope of the Inspector‑General of Intelligence and Security Act.

3. These, as expressed in Mr Collins's letter and drawn out in later discussion with him, were his belief that:

Inquiry process

4. The inquiry involved:

Facts and conclusions

Quashing of assessment ‑ summary of events

5. In July 1998 Mr Collins was the primary author of a document entitled Intelligence Estimate on East Timor (Draft).

6. He prepared this document in his capacity as Deputy J2 (Intelligence) at Headquarters Australian Theatre (HQAST). In his letter to the minister Mr Collins says that he provided it to Major General Connolly the Commander Australian Theatre (COMAST), who instructed him to forward it to DIO for comment.

7. The DIO officer who received it considered that it contained comment and recommendations on policy matters. He therefore sought comments from the International Policy Division of the Department of Defence (IP).

8. The DIO officer also spoke with the officer at HQAST who had directed its preparation. He told him that it would not be possible to respond appropriately within the time requested. He also expressed concern about the policy content and told the officer that he had referred the assessment to IP. It would appear from his subsequent correspondence with the officer that he told him DIO would not be preparing a formal review and would require further tasking from HQAST if that was still required.

9. During the conversation, however, the two officers identified four questions about which HQAST required information and comment that DIO was prepared to provide.

10. The DIO officer forwarded these responses to HQAST on 14 July 1998 under cover of a memorandum (see Annex 2). Mr Collins seems not to have been aware of this because he told me at interview that there was never a formal comment back from DIO.

11. Shortly after this, according to Mr Collins, he visited DIO and an officer there made clear that he was unhappy with the draft. Mr Collins believes that this was because it went against a tendency in DIO to see things the Indonesian way.

12. On 20 July IP wrote to Mr Collins commenting on the draft. The comments (at Annex 3) were heavily critical of it on two grounds: first, that it focussed at the strategic level and raised issues outside the scope of an operational level assessment; and secondly, that it failed "to understand some of the fundamental drivers behind our strategic relationship with Indonesia and contains several assessments about Australian Foreign and Defence policy regarding the political status of East Timor".

13. The IP document went on to express the author's belief that an intelligence estimate is not the place to air differences with government policy. An attachment cited paragraphs that were of particular concern.

14. Handwritten comments, on the DIO copy of this document, apparently from a senior DIO officer, reflect agreement with its conclusions and concern about the draft assessment's discussion of policy matters.

15. HQAST did not seek further comments on the draft and DIO did not prepare any.

Quashing of assessment ‑ conclusion

16. I do not consider there was anything improper in DIO referring the document to IP. It clearly contained significant observations on policy. Mr Collins appears to have acknowledged this in an e-mail of February 1999 to a colleague in which he said that he got into strife (water off a duck's back) for making uncomplimentary statements about our polic[ies]. DIO was both entitled and, arguably, obliged to refer it to IP. IP's decision to correspond direct with HQAST was its own and appropriate in the circumstances.

17. Similarly, DIO's interaction with HQAST was proper in the circumstances. DIO expressed its concerns but also, in discussion, identified particular intelligence issues about which HQAST needed DIO input. It provided that input in a timely manner and made clear that it would respond to further formal tasking. Mr Collins has not pointed to any adverse consequences flowing from this approach and I am unaware of any. The assessment was criticised, not quashed.

18. More generally, Mr Collins has claimed that DIO attempts to suppress dissenting opinions. He bases this on DIO's criticism of his assessment and on its requests for him to desist from publishing, via e-mail, sometimes without informing DIO, various critiques of DIO product and alternative assessments.

19. DIO's justification for the latter requests was that publication of such material, which received wide distribution including transmission not only to colleagues of Mr Collins but also to senior politicians and Australian posts around the world, could lead to confusion with key national and strategic customers in Australia and also with allies.

20. This rationale, given the distribution of responsibilities described at paragraph 25 below, seems entirely logical and cannot, in my view, be accurately characterised as Mr Collins seeks to do. I note also that DIO invited Mr Collins to provide his views for consideration in preparation of its own assessments.

East Timor assessments ‑ Mr Collins's views

21. The critique advanced by Mr Collins alleges that there was a long history at the bureaucratic and political level of a pro-Jakarta line, deriving from support for a policy approach which saw Indonesian political and economic stability as vitally important to Australian interests.

22. This, according to Mr Collins, led to DIO assessments during 1998 and 1999 being overly optimistic about Indonesian, and particularly Indonesian military, intentions and policies in relation to East Timor. They failed, he believes, to give adequate weight to the fact that the military were opposed to independence for the province and would do whatever they could to counter it.

23. Secondly, Mr Collins contends that DIO methodology was flawed in that it relied primarily on reporting events, or statements made by key figures, rather than analysing all sources of available intelligence and seeking to predict likely outcomes.

24. Mr Collins also argues that there needs to be review of intelligence assessments by experts from outside the government, so as to provide a greater measure of quality assurance.

DIO assessment role

25. DIO has described its role as being responsible for the production of top-level, strategic assessments on defence and security issues, to support intelligence customers ranging from the government to the Chief of Defence Force and strategic level command to the operational and tactical level commanders. Within the framework provided by those assessments intelligence staff at the operational and tactical levels provide more detailed and focussed intelligence of direct relevance to their commanders.

26. DIO has explained that this process is designed to work as a cascading framework, with assessments produced at the operational and tactical levels drawing on and complementing the strategic level assessments. Staff at these levels are also encouraged to provide their views and input to the higher level assessments. The purpose of this is to ensure effectiveness and efficiency, and to avoid unnecessary duplication or wasted effort. DIO does not have any authority over the HQAST or Deployable Joint Force Headquarters intelligence staff, as they work within the operational chain of command directly to the Chief of the Defence Force.

DIO assessments ‑ official appraisals

27. In the period under review (mid-1998 to late 1999) DIO produced [a large number of] assessments covering East Timor. Predominantly but not exclusively, these were in the form of Current Intelligence Briefs.

28. In September 1999 all these assessments were collated and provided to the Prime Minister's office so that it could undertake a retrospective evaluation of the intelligence analysis produced during the period. The Director, DIO has informed me that the Prime Minister's office was satisfied with the quality of the DIO product.

29. DIO product is also subject to review by ONA, which is required to provide an annual performance report to the government on Australia's foreign intelligence effort.

30. The ONA report for 1998-1999 noted (at page 9) that DIO was generally accurate in charting the course of political, institutional and economic change in Indonesia and that it accelerated and expanded its reporting as its appreciation of the deteriorating circumstances in Indonesia and East Timor grew. [Sentence deleted]

31. [Sentence deleted] One of Mr Collins's criticisms of DIO was his assertion that it believed every Indonesian statement [words deleted]. The ONA report suggests that the risk was identified already.

32. ONA's 1999‑2000 report said (at page 4) that DIO continued to provide accurate and timely intelligence assessments, well‑received by its ADF, Defence and wider customers.

33. The report also said (at page 9) that among judgments on a range of issues, DIO maintained comprehensive coverage of the security situation in Indonesia and of TNI, including TNI involvement with militias in East Timor.

34. The report contained a separate section on East Timor [words deleted]. DIO also, the report said, staffed an East Timor Crisis Action Team and received good feedback from its customers for its intensive coverage in support of policy and planning and the ADF in East Timor.

DIO assessments ‑ views of Mr Collins's colleagues

35. As noted at paragraph 4.h, Mr Collins nominated a number of colleagues who, he said, shared his criticisms of DIO reporting. Five of these responded to my written invitation to comment. In fairness to Mr Collins, it seems appropriate to report their comments here in some detail even though the overall conclusion of this report is not supportive of his contentions.

36. One of his colleagues said that during the period in question he had only infrequent and indirect contact with DIO staff but was well aware of the DIO role in providing assessments. He said that he could not comment on institutional bias in DIO but from his perspective, despite what appeared to be a deteriorating situation and ADF planning for a worst case scenario, DIO reporting could best be characterised as optimistic, right up until the final crisis occurred.

37. Another colleague said that he had access during this period to reporting from all Australian intelligence agencies. He alluded to the history of TNI involvement with and support for militias and suggested that, despite this, DIO assessments were based in the main on what the Indonesian government and TNI were saying rather than what they were doing. The conclusions that can be drawn from this, he said, are that there was a failure to consider the problem comprehensively or that the people involved did not have the required depth of knowledge and understanding of the history of East Timor and Indonesia,

38. On the question of institutional bias this correspondent said he felt unqualified to comment but he was prepared to say that he and others found it very hard to place much confidence in DIO assessments based on what they themselves could see happening in Indonesia and East Timor.

39. A third colleague considered that the concerns raised by Mr Collins suggest a broader pathology whereby DIO was simply reflecting the long term strategy of successive Australian governments and departments when engaging with Indonesia. This correspondent said, however, that he could not offer specific examples to substantiate Mr Collins's perceptions.

40. A fourth recalled that the assessments of DIO and the HQAST staff differed in important aspects. In particular, this person recalled, DIO seemed more inclined to take at face value statements from the government of Indonesia which indicated that:

41. The fifth correspondent, agreeing that there seemed to be institutional bias, believed that the basic thrust of assessments was overly weighted towards positive public assurances by Indonesian officials and politicians, and contrary intelligence was somewhat ignored.

42. This person said he was satisfied, however, that the perceived bias was counteracted adequately by the existing Defence/ADF system and that accurate assessments were prepared and presented at the operational and tactical level to ADF planners and commanders. This the correspondent attributed to the distributed nature of the intelligence system allowing DIO to extract the broader‑grain intelligence required by strategic level customers in Canberra, and lower levels in the system to extract the finer‑grain information required to plan and execute ADF operations.

DIO assessments ‑ conclusions

43. Though claiming no special expertise on Indonesia and East Timor, this inquiry has reviewed all the DIO assessments referred to at paragraph 28. For purposes of completeness they are included as Annex 4 to this report.

44. I think it fair to say that, if one were of Mr Collins's view, ie that the Indonesian government and military would do whatever they could to prevent the loss of East Timor and, if that failed, would seek revenge on the East Timorese population, then any assessments that did not have that as their underlying premise would probably appear inadequate.

45. The DIO assessments, clearly, do not all have that as their underlying premise. The overall picture throughout the period is of conscientious attempts to analyse what was going on and come to rational conclusions about where that might lead.

46. [Sentence deleted].

47. Similarly, while there are assessments that an observer might say were optimistic about Indonesian government or military intentions, or possible eventual outcomes in East Timor, there are others that display scepticism and concern.

48. One cannot, therefore, discern from the totality of the written assessments the institutional bias that Mr Collins and others have identified.

49. It is also reasonable to point out that events did not go entirely as, presumably, Mr Collins would have predicted. In an unlikely development the civilian government of Indonesia decided that there should be a referendum and, albeit under considerable pressure, Indonesia subsequently withdrew from East Timor and accepted its independence. (This was described in Mr Collins's assessment (see paragraph 5) as the least likely of four possible scenarios for East Timor).

50. Mr Collins has argued that there needs to be a system of evaluation of DIO's assessment effort that is independent of the Australian intelligence community.

51. Paragraph 5(1)(d) of the Office of National Assessments Act requires ONA to: keep under review the activities connected with international intelligence that are engaged in by Australia and to bring to the notice of relevant Departments and Commonwealth authorities any inadequacies in the nature, the extent or the arrangements for co‑ordination, of those activities…

52. As noted at paragraph 29 above, ONA has reported on the foreign intelligence performance in annual reports to government. Those reports have covered assessment but recently have had a relatively greater focus on collection.

53. In the ordinary course there may not be a great deal to be gained from qualitative analysis of assessment performance, since intelligence assessments, while valuable inputs to policy as part of a range of information and opinion available to government, will not be critical.

54. Very occasionally, however, as with East Timor and the terrorism environment in Indonesia preceding the Bali terrorist attack, intelligence and intelligence assessments will be the principal aid to government policy and decision‑making. Potentially in such cases intelligence success or failure may mean the difference between policy success and failure.

55. In those rare instances, arguably, ONA may not be best placed to advise government on the quality of the assessment effort, because it is in the position of analysing both its own and peer agencies' products.

56. The government might, therefore, wish to consider establishing special arrangements, utilising resources outside the intelligence community, when it wishes to have the performance of the community in relation to a particular issue critically and objectively reviewed to identify strengths, weaknesses and possible changes to methodology.

57. Because the circumstances giving rise to the need for such reviews would, one hopes, be few and far between it would be premature to suggest the precise nature of any review mechanism that might be established. It could well vary depending on the nature of the intelligence issue that gave rise to it.

[Word deleted] access [to the database] ‑ background

58. The Joint Intelligence Support System (JISS) is a top secret computer network maintained by DIO and accessible to some other Defence entities [words deleted].

59. JISS hosts [a highly classified intelligence database].

60. [Sentences deleted].

61. Mr Collins had access to [the intelligence database].

62. As the inquiry proceeded it became apparent that DIO considered that Mr Collins's access to [the intelligence database] was not properly authorised. While this might have justified the deliberate withdrawal of access had it occurred, it was not relevant in light of the conclusion in paragraph 70.

Loss of [word deleted] access to [the intelligence database] ‑ circumstances

63. On 20 December 1999 [the database] became unavailable to [words deleted]. Mr Collins sent an e-mail message (copy at Annex 5) to DIO saying that inquiries by engineering staff had revealed that [the database] had been turned off on the orders of the Director, DIO. The message said that, if true, this was of the utmost concern. It asked for further details so that Mr Collins could brief the Commander [words deleted] on the following day.

64. The next morning Mr Collins received an e‑mail message (copy at Annex 5) from the JISS Project Office to the effect that the loss of [the database] was not a technical matter but a DIO policy decision. The Deputy Director, DIO annotated this with a suggestion to the Director that "When we get the security aspects fixed I believe we should turn [the database] back on, even though it's not normal for DJISS, and even though they're using it partly to do strategic assessment".

65. The Director, DIO instructed the recipient of the message referred to at paragraph 63 (in a handwritten note) to telephone Mr Collins and explain that he had not instructed that [the database] be turned off but had asked that it not be turned back on pending the telephone discussion, in which the Director, DIO wanted certain messages conveyed to Mr Collins. The feed was then restored.

66. On the following day the Director, DIO wrote to Mr Collins, with a copy to the Commander, [word deleted], a note (copy at Annex 5) saying that he had not directed withdrawal of the database feed and explaining the reasons for directing the delay in restoring it. These related to security and need‑to‑know issues. The note also dealt with DIO's reservations about Mr Collins's criticisms of DIO product and the appropriate demarcation between DIO's and other intelligence functions.

Loss of [word deleted] access to [the intelligence database] ‑ inquiry and conclusions

67. The records extant at the time of this inquiry were insufficient to determine conclusively the reasons for the loss of [the database] feed.

68. The message cited in the first sentence of paragraph 64 above appeared to support Mr Collins's concern that the withdrawal of the feed was a deliberate decision on the part of DIO management. Despite exhaustive searching over a period of months, however, including reconstituting e‑mail records from archived material, it was not possible to substantiate this. The material that could be obtained tended to support the claim that technical factors were responsible for the problem. It was evident, for example, that the [the system], [words deleted], had experienced a number of technical problems in the weeks immediately beforehand, as did [word deleted], and [the database] access was lost on other occasions in November-December 1999.

69. In view of the absence of conclusive evidence I decided that it would be appropriate to seek written accounts of their recollections from the DIO personnel involved. These were provided as statutory declarations, including declarations from the Director, DIO and the Deputy Director, DIO. Universally they were to the effect that there was no decision to withhold [the database] access and that technical problems were to blame. Copies of the declarations are at Annex 6.

70. My conclusion, therefore, is that there was no policy decision to withdraw [the database] access.

Summary

71. I have no doubt that Mr Collins's concerns were and are sincerely held. Nevertheless, his allegations do not stand up to objective scrutiny.

72. What Mr Collins interpreted as an attempt to quash contrary views appear to be legitimate expressions of concern about parts of the content of his assessment and about his wide distribution of assessments and comments.

73. DIO assessments during the period in question did not uniformly, or even predominantly, adopt a pro‑Jakarta line although there were instances where that interpretation might be available.

74. It might be desirable to consider, in special cases where it is warranted, establishing ad hoc external reviews of intelligence performance.

75. To the extent that evidence is available, it supports the DIO contention that the loss of [the database] access resulted from technical problems rather than a deliberate decision.

76. There is, therefore, no need for formal recommendations arising from this inquiry.

Inspector‑General of Intelligence and Security

May 2000

 

[8] To declassify the report for public release, a small number of deletions were made as marked. The Annexes were not released

previous

content

next