The Intelligence War against the IRA

by Thomas Leahy

  Despite this, the statistics for the 1990s suggest that the North and Mid-Armagh units became more active. The IRA killed eleven intended targets in these areas in 1990 and a further six in 1991. Whilst killings declined to between one and three ‘intended targets’ per year between 1992 and 1994,85 the North and Mid-Armagh IRA posed a threat to surrounding towns and security-force barracks. Lurgan town centre suffered significant damage after an IRA bombing in 1992.86 In February and September 1993, another two large devices were detonated in Portadown and Armagh City, causing extensive financial damage.87 As with other more sparsely populated areas, what made the North and Mid-Armagh IRA particularly threatening was that they could inflict numerous deaths in a single attack. Examples include the IRA’s killing of three RUC officers in a landmine attack in October 1982, another three RUC officers in a similar attack in Armagh City in July 1990, and two UDR officers following a mortar attack in Armagh City in March 1991.88 The evidence of these attacks supports Michael Dewar’s view that the North and Mid-Armagh IRA remained ‘particularly challenging’ for the security forces up to 1994.89

  The North and Mid-Armagh IRA escaped debilitating infiltration for similar reasons to those that will shortly be described for many rural units. But allegations of collusion between the security forces and loyalists, and an alleged ‘shoot-to-kill’ policy in the early 1980s, seemed to maintain republican support in north and mid-Armagh. Throughout the conflict, there were accusations of collusion between elements of the north and mid-Armagh security forces and loyalist groups, including the killings of Catholic civilians by the ‘Glenanne gang’ operating out of a farm in the mid-Armagh area in the 1970s.90 Accusations of collusion were not isolated to single incidents. In March 1990, republican Sam Marshall was shot dead by loyalists ‘as he was walking to his home at Donnelly Gardens in the Kilwilkie estate, Lurgan, after signing bail conditions’ with two other republicans at Lurgan RUC barracks. Lost Lives explains that: ‘Republicans claimed police collusion in the attack, saying that [only] the men, their solicitors and the RUC knew what time they had to sign on for bail.’ Republican suspicions are also based on reports of British intelligence and Army members conducting surveillance in the location where Marshall was shot on that day. Sinn Féin and the Marshall family have called for further investigations into the killing to explore the possibility of collusion. They dismissed a Historical Enquiries Team report of 2012, which rejected the possibility of collusion, and want a review. The persistence of IRA activity in the area shows that Marshall’s killing certainly did not reduce levels of militant republicanism there.91

  There were also the ‘shoot-to-kill’ allegations made against the security forces in the early 1980s. These incidents did not deter militant republicans and their supporters from increasing IRA activity in north and mid-Armagh thereafter. Similar to south Armagh, controversial incidents involving intelligence and special forces arguably encouraged further paramilitary activity. Whilst the SDLP won local council elections to Craigavon before 1998, Sinn Féin consistently held one or two seats.92 There was a core republican support base that remained loyal to the republican movement in spite of intelligence operations and security-force pressure. The fact that An Phoblacht continues to raise cases of suspected British intelligence and loyalist collusion during the Troubles in these areas partly demonstrates the anger that these incidents still cause for local republicans. Republican suspicions of collusion in the area were further heightened shortly after the Good Friday Agreement. Rosemary Nelson’s death at the hands of loyalists in March 1999 was also suspected by nationalists and republicans to be a collusive act. She had defended prominent republicans in court. Nelson had also called for an inquest into the Sam Marshall killing and publicly accused the RUC of colluding with loyalists, which the RUC denied. According to Lost Lives, following her death ‘200 people carrying black flags and placards call[ed] for the RUC’s disbandment’ and they ‘staged a largely silent protest’ outside Lurgan RUC barracks.93

  Tyrone

  Before the 1990s, the East Tyrone IRA and other units in Tyrone sustained a persistent low-intensity campaign. The figures for intended killings in Tyrone in the late 1970s stand at four in 1976, eleven in 1977, five in 1978, twelve in 1979 and a decline to one in 1980. Only the South Armagh and Belfast IRA killed more security-force members in these years. Between 1981 and 1985, the IRA killed thirty-three further ‘intended targets’ in Tyrone. Between 1986 and 1990, this average had only slightly declined to twenty-nine.94 What made Tyrone units particularly alarming for the security forces was their ability to inflict numerous casualties in single attacks in the open space of the countryside. Here are some examples: the IRA killed four British soldiers via a landmine near Dungannon in December 1979; later, in July 1983, four UDR officers died after their vehicle struck an IRA landmine near Ballygawley; in December 1985, the IRA destroyed Ballygawley RUC barracks and shot two RUC officers; and in August 1988, the IRA killed eight British soldiers and wounded others when a landmine exploded underneath an Army bus travelling between Omagh and Ballygawley.95

  Nonetheless, between 1991 and 1994 Tyrone units killed thirteen ‘intended targets’, marking a decline overall.96 With a few exceptions, such as in January 1992 when the Tyrone units killed eight contractors working at a British Army base travelling on a bus in Teebane Cross, attacks that killed members of the security forces, or those working with them, declined.97 Evidence does suggest that the erosion of the East Tyrone Brigade’s armed capacity was a key reason for this decline. The British state area review in July 1980 commented, ‘East Tyrone PIRA is … particularly secure [from infiltration]’, partly because the local population was staunchly nationalist.98 It appears that intelligence increased thereafter. At the Loughgall police barracks on 8 May 1987, for example, the SAS killed eight East Tyrone volunteers when they came to detonate a bomb.99 Brendan Hughes and Tommy McKearney both suspect that informers were involved.100 In contrast, Ingram claims that technical surveillance led to the Loughgall ambush.101 Neither Ian Phoenix, who helped organise the Loughgall ambush for southern Tasking and Co-Ordinating Groups (TCGs), nor William Matchett, an RUC Special Branch member at the time in the southern region of Northern Ireland, specify exactly where the intelligence came from.102 Matchett is more specific when he suggests that whilst ‘the intelligence for Loughgall was compelling it was not definitive’, and that the security forces did not know the precise details of who was involved and the size of the bomb to be used.103

  Certainly, in the area review for Tyrone in 1980, British officials mentioned ‘[t]he removal of BLANK and a small number of his close associates from the terrorist organization would have a major beneficial effect’. At the time, the British lamented the ‘shortage of usable intelligence’, but hinted that in the coming years, ‘[i]t is essential to attack the structure of East Tyrone PIRA with more efforts to place human sources of intelligence within the organization’.104 Since the East Tyrone IRA were ‘attacked’, it could be that ‘sources of intelligence’ were acquired, allowing for a number of SAS ambushes to take place. Following the Ballygawley bus bombing, the southern TCG struck back. They killed three IRA volunteers in August 1988 when the IRA attempted to kill an off-duty UDR member. Phoenix claims that he knew the identity and plans of the three volunteers in advance through ‘intelligence’.105 In June 1991, the SAS ambushed another three Tyrone volunteers during an operation in Coagh. And in February 1992, the SAS targeted four more Tyrone volunteers after the latter had attacked Coalisland barracks.106

  Martyn Frampton and Ed Moloney suggest that the erosion of the East Tyrone Brigade was an important factor influencing the IRA into a ceasefire, since it had been one of the IRA’s strongest units.107 Matchett believes the Coalisland barracks attack shows how cautious the East Tyrone IRA had become: ‘Unlike Loughgall, a village supportive of the police, Coalisland was a town hostile to the police … East Tyrone … had been reduced to conducting attacks within the geographic boundary of its small-
support base and even then it failed miserably’. Overall, Matchett argues that Loughgall was ‘the beginning of the end’ for the East Tyrone IRA, whom he believes the security and intelligence services had ‘decimated’ by the early 1990s. Matchett argues that the IRA in places such as south Armagh suffered following successful attacks against the East Tyrone IRA because security-force resources could now be used elsewhere. Whilst potentially accurate on east Tyrone, Matchett’s argument is not convincing for south Armagh, based on the evidence previously discussed.108

  It certainly could be the case that agents and informers were at work in Tyrone. With the ambushes, the SAS knew exactly where the IRA were going to strike. Shortly after the Clonoe ambush in August 1992, the IRA killed Robin Hill, a former member from Tyrone, who they claimed was an informer. The family deny the allegation. It is not clear whether the IRA believed Hill was responsible for the ambush, although the fact that they were looking for informers suggests that they suspected infiltration had contributed to the ambushes.109 On the other hand, Tommy McKearney, a former Tyrone Provisional, believes that senior informers set up SAS ambushes in Tyrone:

  SAS operations were removing personnel at a time when the IRA leadership was seriously considering … a ceasefire … it was directed against the Tyrone IRA to a large extent … [so Tyrone] was not capable of creating political opposition to a ceasefire; and in consequence … did not offer any significant opposition to a ceasefire … Tyrone … had the political views and numbers to oppose the Belfast republicans. For example, Fermanagh–South Tyrone elected Bobby Sands … [and] had elected an independent MP in the late 1970s … In my view, there was surgical sight here to remove not only the military but the political opposition to any emerging ceasefire negotiations.

  McKearney added that ‘agents of influence’ such as Donaldson could have identified those who opposed the peace process for the intelligence services.110 The trouble with analysing this view is that there is not enough evidence available to correspond with McKearney’s view. Donaldson’s role will be detailed elsewhere, but it seems that he was primarily accessing political intelligence. Otherwise, it would be expected that Donaldson or Stakeknife would have set up frequent SAS ambushes against the South Armagh and Fermanagh IRA too. Since IRA attacks occurred in rural areas on a regular basis, it seems that Donaldson and Stakeknife had, at best, limited access to many rural units.

  Alternative explanations could account for the erosion of the East Tyrone IRA’s campaign. Laurence McKeown suggests that perhaps IRA casualties were unusually high there because of ‘an element … rushing into things and not giving the attention that you had’ in places such as south Armagh.111 Matchett does imply that the East Tyrone IRA felt overconfident following the killings of security-force members and their ability to evade arrest.112 Brendan Hughes also believes that whilst informers were involved at Loughgall: ‘that operation went ahead without the proper intelligence, without proper organization and without proper training’. In hindsight, Hughes alleges that senior republicans ‘threw caution to the wind’ at the time by deliberately damaging strong IRA units to prevent them from opposing a ceasefire.113 There is not enough evidence available to test his view.

  These attacks should not overshadow the political support that republicans maintained in Tyrone up to August 1994. Although Sinn Féin no longer held the Fermanagh–South Tyrone constituency, they consistently performed well in council elections. In Omagh, Sinn Féin won the majority of nationalist seats in 1985 and 1993. Republicans beat the SDLP in Cookstown in 1985, although the SDLP narrowly gained more seats there in 1989 and 1993. The SDLP held a narrow lead over Sinn Féin in Strabane and Dungannon district council elections too, between 1985 and 1993.114 Sinn Féin support was sizeable across Tyrone during this period. Other parties could not ignore the republican movement in local politics. This electoral support suggests that the ambushes had not significantly reduced republican support for Sinn Féin before 1998 either. Similar to the case of the IRA across Armagh, it appears that special-forces operations did not encourage a long-term reduction of support away from republicanism in Tyrone. There remained a core support base on which the East Tyrone IRA could have increased the intensity of its campaign in the late 1990s if it had wanted and if it had managed to locate where intelligence leaks were coming from. The IRA ceasefires means that unlike North and Mid-Armagh units, who did regenerate their campaign after earlier intelligence setbacks, it is impossible to say whether the East Tyrone group would have been able to return to the persistent level of attacks that they had maintained before the 1990s.

  Newry and South Down

  The other main geographical area where the IRA struggled was in Newry and south Down. As in areas such as south Derry and north and mid-Armagh, there needs to be more research on the IRA’s campaign here. The available evidence suggests that agents and informers certainly helped disrupt their activity. The South Down Brigade, for instance, only killed a few security-force members between 1976 and 1986, before a decline to zero in the majority of years leading up to the 1994 ceasefire.115 Eamon Collins, who joined the South Down Brigade in the early 1980s, admits that ‘no more than 50 per cent of our plans worked out as intended’, and that ‘betrayal by informers’ was a key reason for this.116 The fact that Stakeknife is mentioned by Collins and Kevin Fulton (a British Army agent from Newry) as being involved in internal security in Newry suggests that damaging infiltration partly arose from outside interference.117 Fulton says that it was the IRA’s internal security unit that allowed him to join the IRA in Newry and later Belfast, despite previously serving in the British Army. He went on to disrupt various operations in Newry, including a planned attack on Newry courthouse and a plan to kill a worker at the RUC barracks, which did not eventuate as the worker retired before it could be carried out.118 Yet there are other explanations for the low level of IRA activity in Newry. Collins puts some failures down to ‘sheer incompetence … and … bad luck’. On the theme of ‘incompetence’, for instance, he recalls an incident when the IRA killed a boy with a bomb in Banbridge town centre in the early 1980s because the timer was too short.119 Collins’s view cannot be verified, but it does support McKeown’s point about leadership being crucial to the rural IRA’s ability to conduct an effective campaign.

  Overall Assessment of Rural Units’ Influence on the IRA’s Campaign

  Whilst the activities of the East Tyrone and South Down IRA had reduced by the 1990s, this was not the case with other rural units. As a veteran Troubles commentator suggested: ‘[w]ith the possible exception of east Tyrone … in … other [rural] places … [i]t was a continuing battle’.120 In many rural areas and small towns the IRA continued disrupting local political, social and economic life until 1994. In south Armagh, north and mid-Armagh, and Fermanagh they posed a persistent threat to the security forces. The implication of this evidence is that many rural units were not a sideshow. As Danny Morrison argues: ‘[r]ural units … were extremely important in the fight against the British presence’.121 Despite suggesting that the IRA was in decline in rural areas by 1994, Matchett agrees that rural units were fundamental to IRA efforts from the mid-1970s. He even labels the rural and cross-border IRA as ‘the greatest threat’ for the security forces by the mid-1970s.122

  The importance of rural units to the republican movement can be broken down into three key elements. First, they provided the IRA with opportunities to kill a greater number of security-force members in single attacks in wider terrain without risking numerous civilian casualties, which could impact negatively on Sinn Féin’s electoral support.123 Lost Lives statistics, for example, show that the greatest number of full-and-part-time RUC, RUC Reserve and UDR deaths occurred in Armagh and Tyrone, with Down and Fermanagh also high on that list.124 The second reason is that high-profile bombings in England provided another outlet for the IRA to inflict damage on the British state. These attacks are said to have involved, in the 1990s, the South Armagh IRA, who were generally resistant to in
filtration. Lastly, rural units stored and ferried supplies across the border for other IRA units.125 The benefits of open landscapes for rural volunteers compared to the environment of those operating within compact cities was a major advantage for rural units.126 Ultimately, without rural units the IRA would have struggled to continue a persistent, low-intensity conflict between 1976 and 1994. Such units were so important that Operation Banner has reflected: ‘[w]ith hindsight … the Border area was critical to the conduct of PIRA operations and therefore should have been the geographical focus of the campaign’.127

  Why were many rural IRA units generally resistant to significant infiltration? As discussed earlier, many border nationalist communities were historically opposed to the British state for forcing their area into unionist Northern Ireland in the 1920s. Chapter 2 also described how discrimination against majority nationalist populations by local unionist councils before 1969 had created hostility towards the state in some areas west of the River Bann. But the evidence presented in Chapters 2 and 5 also indicates that short-term factors also made rural nationalists reluctant to assist the state. The contentious killings by the SAS, HMSU, the Army and the RUC of suspected and actual IRA members in areas such as south and north Armagh provided additional impetus and motivation for the conflict to continue. These intelligence and security operations may have achieved a short-term reduction in IRA activities. But they did not succeed in reducing (and maybe even cemented) the republican support base in these areas. Frank Foley suggests: