The Intelligence War against the IRA

by Thomas Leahy

  8

  The Intelligence War against the IRA in Belfast and Derry City, January 1976 to August 1994

  Chapters 8 and 9 consider whether the intelligence war reduced IRA activities to such an extent that this influenced the republican leadership to call a ceasefire in August 1994. This argument is suggested in various parts of the available secondary and primary literature.1 I present the counterargument that the intelligence campaign against the IRA did not force the organisation into the peace process in 1994. This chapter explains why the Belfast and Derry City IRA persisted with a low-level campaign, in spite of the intelligence methods deployed against them. I explain that electoral defeats against the SDLP in Derry City, and the political necessity of avoiding if possible multiple civilian casualties in Belfast after 1981, best explains the decline in IRA killings in the cities. It is vital to compare the trajectory of the IRA’s campaign in the 1980s and 1990s to levels of IRA activity directly after 1975. The introduction of the cells, the reduction in active IRA volunteers and the ‘Long War’ declaration implied that the IRA in Belfast and Derry City would no longer be able to reach the activity levels managed in the early 1970s. Chapters 8 and 9 therefore evaluate whether the IRA remained capable of persisting in its attacks against some of its ‘intended’ targets. Targets included (on- and off-duty) RUC officers, the UDR, British soldiers, contractors working for the security forces and retired security-force members. The IRA felt that by targeting these groups and individuals it could influence British policy. The impact of civilian killings on the IRA’s campaign are considered in Chapter 10.

  The Intelligence Structure, 1976–1998

  Before evaluating the impact of the intelligence campaign on the IRA, it is important to explain how the intelligence system functioned. In Northern Ireland, the RUC Special Branch led the intelligence matrix and gathered intelligence through ‘surveillance and … human agents’.2 The RUC’s Special Branch was divided into regional branches. Within each region there were agent-handlers. The recent inquiry into the killing of lawyer Rosemary Nelson in 1999 records that RUC Special Branch had ‘three regions, Belfast, North and South, [which] enjoyed a considerable degree of autonomy’. In terms of human intelligence, an agent-handler for RUC Special Branch would provide an agent or informer with a reference number and code name if the ‘source’ was perceived as reliable after a few meetings. A local RUC source unit led by a detective chief inspector would debrief handlers after they met their agents or informers. An SB50 form was then created. According to de Silva, a SB50 contained: ‘brief summaries of the intelligence received from an agent’. If RUC Special Branch chose to share the intelligence with other agencies, such as the FRU, the information on the SB50 form would be copied to an RIRAC form. The RUC’s Special Branch copied intelligence centrally and stored it at their headquarters in Knock.3 They could determine what – if any – information other agencies could receive. Special Branch also maintained its own surveillance and rapid response units. The unit E4A carried out surveillance. Each region (Belfast, North and South) would have an E4A surveillance unit attached. Unit E4B, also known as Headquarters Mobile Support Units (HMSUs), carried out special operations. Other E departments focused on administration, specific regions and training. William Matchett describes the surveillance and rapid response units as consisting of eighty-four officers, whose duties included surveillance, conducting search operations and protecting other undercover units.4

  British Army intelligence had altered by the early 1980s. The major innovation was the creation of the FRU as its central human-intelligence-gathering agency. Before the 1980s, individual brigade intelligence units collected information from agents and informers. The FRU centralised and coordinated human intelligence for the Army. Its structure consisted of a headquarters and four regional units covering the north, east, south and west of Northern Ireland. Each detachment had an officer commanding (OC) and agent-handlers. The OC would report intelligence to the FRU’s Commanding Officer (CO). From there, the FRU CO would provide reports to the CLF and GOC. The FRU handlers created two types of intelligence reports. The ‘Contact Form’ noted ‘detail about a meeting between a handler and an agent and were circulated internally within Army structures’. In contrast, a Military Intelligence Source Report ‘was a summary of the intelligence received that was transmitted to [RUC Special Branch] and, when appropriate, to [MI5] personnel’.5 The 14 Intelligence Company (also referred to as the DET or Special Reconnaissance Unit or Regiment) emerged out of the MRF. But 14 Intelligence Company focused on undercover surveillance and assistance in arrests when required by the RUC Special Branch and British Army. The British government also deployed the SAS in south Armagh in 1976. The SAS were required thereafter when arrests or active engagements of IRA members were needed in ‘hard’ areas, often rural republican areas. Urban estimates that the 14 Intelligence Company and SAS units in Northern Ireland were made up of approximately 150 operators altogether.6

  The primary role of MI5 was to provide technical surveillance. The RUC Special Branch and FRU directed MI5 where to install electronic listening devices. However, MI5 did run a small selection of agents and informers. Their role later expanded because they took over the lead role against the IRA in Britain from the Metropolitan Police Special Branch in late 1992.7 Also, MI5 had liaison officers at Army headquarters in Lisburn and RUC headquarters at Knock. But they had to gain permission from RUC Special Branch if they wanted to recruit an agent or informer.8

  The focus of GCHQ was on complementing the electronic and technical side of the intelligence war against the IRA. From the 1970s, they engaged in technical warfare against IRA bombers and engineers. Aldrich writes: ‘GCHQ came up with special equipment that inundated the Province with random radio signals on the bomb command frequencies. This caused a number of bombs to detonate while they were being constructed and tested.’ The IRA eventually realised this, and began to design new signals and methods of detonating their devices.9 Additionally, GCHQ had two sophisticated listening and intercept towers in the mountains of south Armagh and in Capenhurst in England. The ‘150-foot-high concrete tower built in 1989 within a secure compound at Capenhurst … was on a direct line between the British Telecom Medium Wave Tower at Holyhead in Anglesey and another tower at Sutton Common near Macclesfield’. This tower enabled GCHQ to monitor phone calls between the UK and Ireland. Aldrich adds: ‘[a] similar station on Croslieve Mountain in South Armagh is thought to have taken [telephone] traffic between Belfast and Dublin’.10

  Another important contributor to the intelligence network was the Director and Coordinator of Intelligence Northern Ireland (DCI). The DCI delivered ‘high-level policy direction and advice relating to intelligence activity in Northern Ireland … to the Secretary of State for Northern Ireland … the Chief Constable of the RUC and the GOC’ (General Officer Commanding). The DCI met regularly with the RUC Chief Constable and GOC to ensure that the Secretary of State’s priorities were in the minds of the security-force commanders. There was a direct link between the intelligence effort and government policy during the conflict. Nonetheless, government ministers and officials in Whitehall would not demand that specific operations be carried out, such as to recruit a particular individual as an informer. Instead, Westminster and Whitehall would provide overall objectives for political and security policy as guidance for the intelligence services.11 Former agent-handlers from the RUC Special Branch and FRU certainly felt that government objectives influenced their work. Whilst complaining that the British government did not provide guidelines for the handling of agents, a former FRU handler told de Silva that the government would be ‘calling for counter terrorist measures’ at particular times.12 The former head of RUC Special Branch was told by Margaret Thatcher that Special Branch were ‘working well’. But when it came to asking for guidance on the boundaries of agents’ activities, the advice was simply ‘don’t get caught’. Matchett identifies the long-term legal difficulties this approach caused the RUC
Special Branch and FRU.13

  The Tasking and Co-Ordinating Groups (TCGs) created in the late 1970s were also crucial in the intelligence network. To avoid conflict between the various agencies involved in undercover work, TCGs were comprised of representatives from RUC Special Branch, MI5, FRU, 14 Intelligence Company and the SAS. Led by a regional RUC Special Branch superintendent, they planned and implemented operations based on information gathered from all agencies in a regional TCG. There were three TCGs: TCG Belfast; TCG South, based in Gough Barracks and covering Armagh and the southern areas; and TCG North, covering the Derry area.14 Matchett estimates that there were 1,500 security-force personnel involved in intelligence and surveillance activities.15

  In practice, this system had some success in improving cooperation between the intelligence agencies after 1975. At Loughgall, the SAS killed eight East Tyrone IRA volunteers whilst attacking a police barracks in May 1987. Ian Phoenix, a former Special Branch leader of TCG South, claims that his TCG set up the ambush.16 Nevertheless, a former colonel who worked at British Army headquarters told the Nelson inquiry: ‘[o]nly … crumbs from [RUC] Special Branch’s table came to us’.17 Rob Lewis, the pseudonym for a former FRU agent-handler who worked in Fermanagh during the 1980s, agreed: ‘[RUC] Special Branch did not always play ball with us’.18 Phoenix recorded in his diaries that parts of his own organisation saw ‘people [build up] little fiefdoms’ that did not share information. Yet Phoenix also found that MI5 increasingly interfered with RUC Special Branch’s work after the end of the Cold War.19 Former Special Branch employees have also stressed they felt that the Army and FRU sought immediate results during their short tours of duty, explaining RUC Special Branch’s reluctance to always share intelligence.20 Matchett argues: ‘[w]hilst the FRU contribution is not insignificant, it was small and not effective as the [14 Intelligence Company] or the SAS’. ‘Just by existing’, he adds, ‘the FRU made it hard for [RUC] SB’ to coordinate intelligence. For Matchett, FRU’s existence was about ‘power’, which meant that the British Army did not have to rely on RUC Special Branch in the fight against the IRA and could continue to play a central role.21

  Conflicting evidence makes it difficult to attribute blame to one specific agency for the lack of cooperation sometimes experienced. The crucial point is that ‘institutional rivalries’ did create a ‘strained relationship’, particularly between the FRU and RUC Special Branch.22 There is no doubt that these flaws within the intelligence system made it more difficult for the British state to bring about an ‘acceptable level’ of IRA violence.23 De Silva discovered government documentation from the late 1980s detailing meetings between British government ministers and the security forces demanding better coordination, particularly following the IRA killing of eight soldiers at Ballygawley in 1988.24 These divisions within the state security and intelligence services were significant. The challenges they presented in terms of a successful intelligence campaign against the IRA can be seen with the restructuring of the intelligence network in Northern Ireland after 1998. Restructuring included the disbanding of the RUC Special Branch. The Rosemary Nelson Inquiry report states that in 2007: ‘the Security Service [MI5] took over from the PSNI the lead responsibility for national security intelligence work in Northern Ireland. The arrangement is now similar to that operating in other parts of the UK.’25 The experience of intelligence-sharing and conflict in Northern Ireland contributed to this decision. The intelligence effort during the Troubles faced some difficulties due to institutional rivalries.26

  The Intelligence Campaign against the Belfast IRA, 1976 to August 1994

  In Future Terrorist Trends in 1978, Brigadier Glover argued: ‘the PIRA leadership appreciate that their campaign will be won or lost in Belfast’.27 Much of the secondary literature agrees with his view for three reasons: Belfast was Northern Ireland’s capital city; it was where the Provisionals emerged in 1969; and leading figures in the republican movement such as Gerry Adams came from Belfast.28 Tommy McKearney also suggests that Belfast accounted for roughly 50 per cent of the IRA’s strength even after the introduction of the cell structure, because there was a larger concentration of republicans there to recruit.29

  By the 1990s, Bew and Frampton believe that the Belfast Brigade’s campaign was ‘being brought to a standstill’. To support their argument, they cite the case of Stakeknife and the fact that the IRA did not kill any British soldiers in the city after 1992. They also quote Phoenix, who suggests that eight out of ten Belfast IRA operations were being prevented by the 1990s, and note that the last commercial bombing in Belfast took place in 1993. Bew et al. argue that the ‘decline’ in the vitality of the IRA’s campaign in Belfast ‘[prompted] a growing realization within [the IRA] leadership that the “armed struggle” had reached a point of deadlock’ by 1994.30 Whilst there was at times significant infiltration of the Belfast Brigade between 1976 and 1994, this chapter disagrees with the argument that the Belfast IRA was facing terminal decline by the 1990s because of the intelligence war.

  After the collapse of the 1975 ceasefire, IRA killings of alleged agents and informers recommenced in Belfast. Seamus O’Brien was killed on 17 January 1976 in west Belfast. According to Lost Lives, the IRA killed him for allegedly informing and colluding with loyalist paramilitaries to attack republican pubs. O’Brien was supposedly ‘named to the IRA as an informer by Vincent Heatherington’. O’Brien’s mother and journalist Martin Dillon both dispute these claims.31 The most unusual case of alleged IRA infiltration during the late 1970s involved Vincent Heatherington and Myles McGrogan. Both were young men from west Belfast. Details surrounding their activities have emerged from various sources including Lost Lives, the former Belfast IRA commander Brendan Hughes and journalists Ed Moloney and Martin Dillon. In early May 1974, two RUC officers were shot in Belfast. The security forces arrested Heatherington and McGrogan. When the prisoners entered Crumlin Road jail they asked to be placed in the IRA’s wings. Neither of them were active republicans. According to Hughes, following questioning by the IRA in prison, Heatherington admitted ‘he had been working with … Brits’. Heatherington allegedly gave the IRA the names of other agents and informers. The IRA began interrogating prisoners in Crumlin Road and Long Kesh. Heatherington’s next ‘confession’ spread absolute panic in IRA ranks: his mission was to poison leading IRA figures including Hughes. After poison was allegedly discovered in the loyalist wing, Hughes recalls ‘mad hysteria’. The IRA monitored all food and water. The plot thickened. Heatherington withdrew the original names he had provided and created a new list. Hughes was in disbelief: ‘he was playing with me. It was basic counter-intelligence disinformation that they were spreading’. Hughes added: ‘we were in total disarray … You didn’t know who was a tout, or who was going to poison you’. McGrogan and Heatherington were acquitted and left prison. The IRA abducted Heatherington and killed him in Belfast on 6 July 1976. The IRA killed Myles McGrogan later in the Lisburn area in April 1977. A month after McGrogan’s killing, the IRA shot James Green. Green was a former member of the British Army and worked with taxi firms in Belfast. Unconfirmed reports say that Heatherington named him as a British agent. There is no proof that any of those killed were agents and informers. The RUC and Army have not confirmed any allegations.32 But Ingram admits that it certainly was a tactic of British intelligence to divert IRA energies to spy hunting in order to ‘sow alarm, despondency and paranoia’ within the organisation.33

  Killings of alleged agents and informers by the IRA continued. For example, on 8 April 1978 a masked gang entered Brendan Megraw’s flat in west Belfast and abducted him. He was never seen again. For years, the IRA denied involvement. His widow denied that Megraw was an informer. In April 1999, the IRA provided the following statement: ‘Brendan Megraw, a civilian from Belfast … in 1978 … admitted to being a British army agent’. This claim cannot be verified. His body was found in late 2014 in County Meath. The reason why he was ‘disappeared’ remains unclear.34

  The
IRA’s killings of ‘intended targets’ between 1976 and 1979, however, do not indicate any particular difficulties in Belfast with agents and informers. Ten deaths of IRA ‘intended targets’ occurred in 1976, eighteen in 1977 and six in 1978, before a rise again to eighteen in 1979.35 Of course, these figures represent a significant decline from 1972. But this decrease resulted partly from the IRA having fewer active volunteers by the late 1970s. As McKeown explains:

  If you are trying to build something long-term … we cannot be running about on the off-chance that there might be a soldier down the road. We would lose weapons and volunteers doing that … This would inevitably reduce the number of operations that you are carrying out.36

  The British security area review for Belfast in October 1980 concurs that a decline in random IRA shootings at security forces was ‘probably a reflection of … increased discipline of the terrorist organization, which is no longer willing to expose men in … unproductive attacks’. The review also commented: ‘[t]he threat posed by [Belfast] PIRA is now based on a limited, secure and refined organization mounting relatively few attacks’.37

  In contrast, the Belfast IRA did experience a decline in the 1980s. The IRA struggled to maintain its killing levels at those of the period between 1976 and 1980. A particular period of concern for the IRA was between 1982 and 1986, during which time there was a dramatic decline to low single figures each year, although they did temporarily rise thereafter.38 Nevertheless, the statistics do correspond somewhat with alleged and actual infiltration of the Belfast Brigade in the 1980s. Furthermore, Operation Banner argues: ‘[t]he quality of intelligence became very good indeed – by the end of the 1980s PIRA was unable to mount a bombing operation in Belfast for about two years’.39 The decline continued into the 1990s. Five intended targets were killed in 1990, rising temporarily to nine in 1991, before a decline to three in 1992, one in 1993 and two in 1994.40