National Congress of Brazil
Seminar: Intelligence Activities in Brazil: contributions to sovereignty and democracy
Panel: Intelligence, Public Control and Democracy.
Introduction
In the past thirty years throughout Europe, the Americas and more sporadically elsewhere the issue of how to institute some democratic control over security intelligence agencies has steadily permeated the political agenda. There have been two main reasons for this change. In what might be described as the ‘old’ democracies (North America, Western Europe, Australia and New Zealand) the main impetus for change was scandal involving abuses of power and rights by the agencies. Typically, these gave rise to legislative or judicial enquiries that resulted in new legal and oversight structures for the agencies, some of these achieved by statutes, others by executive orders. The best known examples of these are the U.S. congressional enquiries during 1975-76 (chaired by Senator Church and Representative Pike), Justice McDonald’s enquiry into the RCMP Security Service in Canada (1977-81) and Justice Hope’s into the Australian Security Intelligence Organisation (1976-77, 1984-85).
Elsewhere, this shift has been a central, and sometimes painful, aspect of the democratisation of formerly authoritarian regimes, both civilian and military. For example, the death of Franco in 1976 precipitated democratisation in Spain that included the de-militarisation of intelligence (Giménez-Salinas, 2002). Military rule ended in Brazil in 1985 though the military dominated National Intelligence Service (SNI) was not replaced until 1990 as part of a continuing process of de-militarisation (Cepik & Antunes, 2001). During 1993-94 a more rapid transformation of formerly repressive security agencies was attempted in South Africa (Joffe, 1999). The other major examples of this transition since 1989 are the countries of the former Soviet bloc where no agency has been immune from the changes although the amount of real as opposed to nominal reform varies widely (for example, Rzeplinski, 2002; Szikinger, 2002).
Whether scandal or the democratization of former authoritarian regimes (and sometimes both together) have been the main impetus for change, the main emphasis of reforms has been on increasing the legality and propriety of security intelligence operations. Although in some cases attention was paid also to the issue of obtaining effective security intelligence (e.g. Macdonald, 1981), the overall direction of change was to the better control and accountability of agencies whose past activities had been dominated by the surveillance of political opponents rather than genuine security threats.
But since the September 11 2001 attacks in New York and Washington DC, the debates around security intelligence have shifted to the contemplation of ‘intelligence failure’ and how future threats can be averted. This is most obviously the case in the US itself but the impact of the global ‘war on terror’ has been much more general. This repeats the historical pattern in which concern regarding propriety has increased following scandals while intelligence ‘failures’ such as 911 give rise to increased concern with efficacy. In this atmosphere it is easy to see how the democratic gains of the last thirty years might be swept away in the naïve belief that agencies ‘unhampered’ by oversight requirements might somehow be more efficient and effective.
What is the objective of ‘public control’?
For the future the objective must be to avoid lurching between the alternative poles of ‘efficacy’ and ‘propriety’. Rather, the aim within democratic states must be to secure intelligence services that are both effective and operate within the boundaries of law and ethics. Historically, intelligence services have enjoyed a high level of autonomy from democratic control that has allowed them to be both ineffective and abusive of human rights. Measuring the effectiveness of security intelligence agencies is extremely difficult since their successes are likely to be seen primarily in ‘events’ that do not take place. Insiders may be able to point to security actions that prevented planned and predicted events (for example, riots, bombings, theft of secrets) but will be loath to publicise these for fear of exposing sources and methods of intelligence. More generally, measuring the performance of agencies is problematic: the production of intelligence cannot simply be reduced to counting up the numbers of reports disseminated.
It is a mistake to view efficacy and propriety as being in a zero (constant) sum relationship such that gains in one are outweighed by losses in the other. Rather, they should be viewed as being in a non-zero (variable) sum relationship such that both can be improved. This is not to say that there is no tension between the two: it is quite easy to see how, in the short run, the ability to conduct surveillance of an individual or group may be reduced by the requirement to follow procedures that seek to protect privacy but, in the longer term, such procedures are required if a state is to be entitled to call itself democratic. Such procedures should be designed in order that, even in the short term, the invasion of privacy is proportionate to the alleged threat but also to prevent it being directed at the wrong person or conducted in such a way as to amount to intimidation. Thus legal rules themselves may contribute to efficacy as much as to propriety.
But in the search for better public control of intelligence, improved legal rules alone will be insufficient. The task of democratisation and search for efficacy/propriety includes shifting both the legal contexts for intelligence work and the culture of the agencies. Although the process of achieving legislative change can itself be difficult and require considerable political will, there is a danger that, once it is achieved, it will be assumed that real change in the agencies and their behaviour will result. This is a dangerous assumption: new laws themselves may only achieve symbolic change (Edelman, 1964) so that people can be reassured that problems have been dealt with. If they are not matched by even greater effort in implementing those laws then little that is real may change. Beneath the surface of new laws, what the agencies actually do and how they do it might remain essentially unchanged. Achieving cultural change in agencies that may have long histories of complete autonomy from outside control or influence is a long term project that may require even greater political will than achieving initial legal reform. This issue is not addressed in detail in this paper since a number of other contributions to the seminar did so (see papers by Del’Isola e Diniz, Cepik, Bruneau). It should be noted, however, that while the ‘professionalisation’ of intelligence is an important development – and certainly a great improvement over the crude ‘politicisation’ characteristic of authoritarian states – it does not obviate the need for energetic external oversight or review of intelligence activities.
What are the major problems in establishing public control?
It is important to consider at the outset what ‘public control’ means. In the literature and laws of security intelligence, different terms are used and confusion can arise from their use to describe different things. This is not a semantic argument: it is not just that the words themselves matter; also it is that there may be a confusion as to functions and the division of responsibilities and powers between agencies and those controlling them. It is important that those involved within any particular control system understand the significance of the different functions even though actual practices will vary between countries in line with their differing constitutional and governmental structures. In particular there is a need to distinguish ‘control’, ‘oversight’ and ‘review’.
‘Control’
For the purposes of this paper ‘control’ is taken to refer to the management or direction of an agency that is normally achieved by means of administrative regulations or orders (cf. Caparini, 2002).
Security intelligence is prone to two forms of abuse, each being the ‘mirror-image’ of the other. In the older democracies the tendency was for there to be no control by elected ministers so the danger was that agencies pursued their own security agendas. For example, it was this characteristic in the United States that led Senator Frank Church to describe the Central Intelligence Agency (CIA) as a ‘rogue elephant’. But the problem was not restricted to the United States; more generally it can be summarised in the term ‘plausible deniability’. It was commonly accepted by state authorities that security intelligence agencies broke the law: this was clearly the case abroad since they were spying on other countries but it was also the case domestically where their surveillance activities routinely broke laws relating to burglary, theft etc. Since ministers could not publicly support such actions, it was important that they could ‘plausibly deny’ any knowledge of them and the easiest way to achieve this was if they really did not know! Therefore both ministers and agencies shared an interest in the ignorance of the former.
Plausible deniability is still maintained by many nations with respect to covert operations by foreign intelligence agencies except where this capacity is part of an attempt to convince an opponent of one’s overwhelming power. But democratisation has seen the reduction of plausible deniability in the domestic sphere: the legislation in the older democracies has normally included procedures for authorising the deployment of ‘special’ powers such as entry on property and theft of records. It has also incorporated the recognition that democratic control requires public preferences to be reflected in security policy. Therefore it is now normal for some provision to permit or compel ministers to issue directions to agencies with respect to both targets and methods. Directives have replaced ‘Deniability’.
Of course, ministerial control does throw up the opposite danger that the agency may be used by governments to infringe on the rights of opponents or minorities. There are examples of this form of abuse in the older democracies but in several of the transition states of Eastern Europe it is simply assumed that governments use the agencies as a means of spying on and, sometimes disrupting, the activities of competing political parties (for example, Rzeplinski, 2002; Szikinger, 2002). Thus, in any oversight or review structure it will be important to build-in protections against this kind of abuse, for example, for ‘whistle-blowers’.
‘Oversight’
This term is taken to mean supervision of an agency including its ongoing activities. The most familiar example of this in practice is the US Congress where the Intelligence Committees determine agency budgets and must receive prior notification of covert action. Through diligent use of large committee staffs (cf. those in most other legislative assemblies) and their extensive access to information, the US seems to provide a model of energetic democratic oversight. However, there are two central factors that require consideration. First, the US is a presidential system and its structures may not be appropriate for parliamentary regimes. Much has certainly changed in the institutional landscape of US politics since the Founding Fathers envisaged the Congress as the central governing institution at the end of the eighteenth century but enough remains of their legislative and budgetary powers that they still make key decisions regarding the size and distribution of resources within the agencies (e.g. Holt, 2000, 196). This happens rarely, if ever, in parliamentary systems.
Second, this unparalleled contribution to ‘control’ comes at a cost to the ability of committees to investigate and criticise. In the case of the prior notification of covert or other sensitive intelligence operations, if the Committee personnel agree that the agency may proceed then if there is some subsequent scandal or political embarrassment how can the Committee really criticise the agency? More generally, their own budgetary and resource-allocation decisions may themselves contribute to the agencies’ ineffectiveness or intelligence ‘failures’ and the Committees’ ability to criticise the agencies will thereby be compromised. For example, it has been suggested that, since the Committees’ ‘own lax oversight of the CIA and other spy agencies contributed to the nation’s vulnerability’ in September 2001, it is inappropriate that those agencies investigate the 911 ‘intelligence failure’ (New York Times, August 10, 2002).
‘Review’
By contrast with ‘oversight’ this is taken to mean survey after the event. In the recent history of intelligence control these two terms have often been used interchangeably but it can be useful to distinguish them. The Canadian Security Intelligence Service Act 1984 implemented many but not all of the proposals of the McDonald Commission, notably the creation of the new civilian Canadian Security Intelligence Service to replace the security service that had existed within the Royal Canadian Mounted Police (RCMP). In the early days of the new intelligence structure in Canada a lively debate ensued between those arguing for a more activist role for the new Security Intelligence Review Committee (SIRC-CSARS) and those criticising it for seeking a broader role than that envisaged by Parliament. There was certainly some substance to this debate, especially the extent to which the initial Committee rejected the idea that it should limit itself to a historical review of CSIS but it was also complicated by the broader meaning of the term surveiller in the French version of the Act compared with review in the English versions.
Other things being equal, ‘review’ committees are more likely to be found in parliamentary systems where government control of agendas through the majority party (or coalition) may be stronger, certainly, than in the US Congress. But, to the extent that review bodies are not implicated in making central budgetary or resource decisions, their ability to criticise failures is unimpaired. Their inability to comment on ongoing activities may reduce their influence on the agencies but they will not be compromised by involvement in ‘control’ decisions. The other problem that may inhibit review committees is that they may enjoy less than full access to information – this is not inevitable, but it may be the by-product of the greater governmental control of official information often associated with parliamentary systems.
Secrecy
This leads directly into consideration of the second major problem facing those seeking greater public control of intelligence. Security intelligence policies and agencies are distinguishable from other state activities by two factors. The first is that they are concerned with the very existence of the state itself in a way that policies regarding education, health, housing and so on are not and the second is that the policy field is characterised by high levels of secrecy. For many scholars this is the defining characteristic of intelligence work. For example, all government agencies seek to gather information and analyse it in order to develop policies; what distinguishes intelligence agencies is that they will also use covert methods to obtain information that is simply not available openly. Similarly, some level of secrecy will apply to all information in the hands of government. But even where freedom of information legislation makes it possible for citizens to obtain access to much state information, there will always be exemptions with respect to information relating to, for example, ‘national security’ and the ‘prevention and detection of crime’.
Secrecy has two main dimensions that are relevant to intelligence: the first seeks to ensure that state officials will only have access to information if they have been cleared by security vetting for access at the appropriate level of classification. Normally, the higher an official is promoted or the nearer she is working to military or security matters, the higher the clearance she will need – for example, from ‘confidential’ to ‘secret’ to ‘top secret’. Within the security intelligence sector the second dimension is compartmentalisation. Even though officials may be cleared to the highest level, it is still believed that the circulation of knowledge with respect to particular techniques, operations or targets should be minimised in the interests of security. Therefore individuals only have access to the information that they ‘need to know’.
Now, these dimensions of secrecy have many implications. For example, they may hinder the efficacy of intelligence by reducing the flow of information both within agencies and, even more, between them. The failure of agencies to share information through some combination of proper concerns for security and petty bureaucratic jealousies is a common feature of intelligence ‘systems’ but there is insufficient space to consider this fully here (Center for the Study of Intelligence, 1977). Clearly, secrecy presents a major hurdle to be surmounted if public control is to be achieved. The ability of outside bodies to oversee or review intelligence agencies depends on their ability to obtain relevant information; if the agencies themselves will not provide it then those bodies are stymied because there will be little information available that is independent and useful. In most areas of state policy there is a broader ‘policy community’ of research organisations, ‘think-tanks’, lobbying groups, journalists and academics that can provide a source of information and ideas independent of the state but in the area of security intelligence it is only small.
Intelligence Networks
A third difficulty inherent in the attempting to establish public control is the very scale and diversity of contemporary intelligence networks that cover both state and private sectors. Taking ‘intelligence’ in its broadest sense of governments seeking to develop knowledge in order to inform policies then these networks are as extensive as the state itself and, indeed, beyond into the private sphere where corporations in all sectors of the economy seek competitive advantage through obtaining and using knowledge. But even if we limit ‘intelligence’ to those activities characterised by secrecy in pursuit of safety and security, we find that contemporary networks are both extensive and spreading. Within the state structure the networks cover both civilian and military departments. The former is not just limited to specialised security intelligence agencies, however, because as ‘security threats’ are broadened to include, for example, organised crime and environmental degradation, so the number of government agencies deploying covert forms of information gathering also increases. In many places, the field of military intelligence has tended to be seen as separate from civilian or security intelligence departments but this is not helpful in analysing the recent past of South America or Eastern Europe where military intelligence was a key component in the maintenance of internal security by authoritarian regimes.
There are several reasons why attention must be paid to overseeing this entire sector. Because agencies do form a ‘network’ there are increasing numbers of information and personnel exchanges between them. Many of these take place in the interests of improving efficacy and there are strong arguments for seeking to improve such exchanges as an antidote to bureaucratic fragmentation and ‘turf wars’ between agencies. However, there is also a darker side to the network: it operates secretly and there is a danger that those agencies subject to control may ‘sub-contract’ illegal activities or operations to those that are not (either in the public or private sector).
The possibility of such problems occurring is enhanced by certain recent developments. For example, the democratisation of a large number of previously authoritarian states in South America and Eastern Europe has led, to a greater or lesser extent, to turnovers of personnel, sometimes associated with ‘lustration’ – a process of seeking to exclude those guilty of rights abuses from employment in the new regime (for example, Rzeplinski, 2002; Szikinger, 2002). But if this process is successful questions still have to be asked about where these former security employees go. We know, for example, that in Russia, many took advantage of the rapid privatisation of state assets to enrich themselves and/or to establish security companies with close links to criminal organisations (Voronin, 1997). This has not happened so extensively elsewhere but care must be taken that former security intelligence officers do not move into ‘security-industrial’ complexes that may provide fertile ground for the existence of some kind of ‘parallel power’ to that of the state such as that exercised by SNI in Brazil between 1964-85 (Cepik and Antunes, 2001, 6)
Some principles of control and oversight
Much can be gained from the comparative study of security intelligence (Hastedt, 1991): the use of security intelligence by states displays certain common features regardless of the precise form of the state concerned, for example, secrecy, a tendency to confuse ‘security threats’ with ‘political opposition’ and the use of ‘extra-legal’ methods to obtain information and disrupt opponents. Also, it is possible to see the development of cross-national intelligence ‘communities’ so that the differences between national agencies may be less than might be assumed. To be sure, this tendency is clearest within coalitions of nations, for example, the UKUSA pact of anglo-saxon countries, especially their SIGINT agencies or the Warsaw Pact between what were ‘counterintelligence states’ in Eastern Europe. Elsewhere, and sometimes even within coalitions, there are fierce ‘intelligence wars’ between agencies but there are clear signs now of convergence between agencies in the context of the globalised ‘war on terror’ led by a hegemonic United States.
Still, the actual structuring of any particular state’s security intelligence agencies and the appropriate forms of control, oversight or review will be determined finally by the particular political culture and traditions of that state. Therefore, it is idle to suggest that states might simply pick and choose from institutions operating elsewhere; political institutions cannot simply be transplanted from one political system to another. But studying institutions elsewhere may well help to prevent a state ‘reinventing the wheel’: states can learn from each other. So, in this paper, the object is not to lay down some set of hard and fast rules for effective public control; rather, it is to suggest that there are certain fundamental questions that have to be answered and certain basic principles that can be enumerated based on the study of intelligence reform in several countries.
The Figure ‘Control and Oversight of Security Intelligence Agencies’ summarises these. The horizontal axis is based on the proposition that ‘states’ are not single entities: they operate at three main levels, the demarcation between them often defined in terms of secrecy barriers. First, there is the most secret level occupied by security and military intelligence agencies; second, the executive branch (or government) and, third, the broader array of state institutions including elected assemblies, judiciaries and bureaucracies. Since we are concerned with the issue of public control, we must also include a fourth – non-state – level in our analysis, representing citizens, groups and social movements.
The vertical axis seeks to summarise, first, the different institutions and forms of control that need to exist at each level and, second, the complementary institutions of oversight or review. Forms of control become more specific the closer the level is to the agencies. The manifestos generated by political parties or social movements are not strictly-speaking a form of ‘control’ because they may have no impact on agencies but they will provide a general set of demands that might, at a later date, inform the generation of more specific statutes or court actions. Some Parliaments pass more detailed legislation than others; but in either case ministers are likely to provide yet more detailed directions for agencies. Some legislation actually requires ministers to provide directions, for example, the CSIS Act. The most detailed rules or ‘guidelines’ will be those developed with the agencies and are normally unpublished.
Clearly, the central institutions of control identified in the Figure also play a role in oversight. Indeed, in some parliamentary systems prior to intelligence reform, it was claimed that it was inherent in the constitutional process that there could be no independent oversight of security intelligence and that both control and oversight was provided by a single institution, usually a minister. This was certainly the case in the United Kingdom based on the doctrine of ‘ministerial responsibility’ to Parliament. Even though the inadequacy of this doctrine has now been acknowledged, we can see that agency directors, ministers, Parliaments and some Judges will exercise both functions. This is inevitable but only becomes a problem if there are no additional institutions of oversight with their own organisational basis.
Thus oversight institutions must also exist at each level, must report to those responsible for control at that level and will normally be located there (for example, agency, ministry, assembly). This location within agencies or ministries raises concerns as to the real extent of their independence but the danger of their being compromised can be reduced by securing their right to communicate with oversight bodies at other levels (see further below). Regarding ‘level’ one, it may seem odd to talk of oversight functions within agencies themselves but if oversight is only an external function then it becomes easier for agencies to see it as something troublesome that should be resisted. Instead, ideas of propriety must be internalised within the culture of agencies. However, although internal oversight is a necessary condition for public control, it is not sufficient: it must be backed up by external oversight at ‘levels’ two and three.
Oversight bodies are usually quite small with limited resources and their effectiveness can be enhanced in several ways. One way of seeking to protect their independence is to require them to copy reports to the oversight body at the next level. Depending on the precise institutional arrangements, this may be subject to some secrecy constraints but it will help to reduce the dependence of oversight on the agencies themselves. So, for example, if an internal agency body such as an ‘Office of Professional Responsibility’ reports to the Agency Director on some matter the report should also be made available to whatever oversight institution exists within the ministry, for example, an inspector general. Similarly, reports from inspectors general to the minister should be made available to the review committee at ‘level’ three, whether it is a joint parliamentary committee such as in Brazil or the UK or a non-parliamentary body such as SIRC in Canada. If reports cross the secrecy barriers existing between the different levels of the state then how is appropriate security of information to be maintained? Ultimately this has to rely on consultation and trust between institutions at different levels and the discretion exercised by those involved. This is particularly the case for those working at ‘level three’ who, elected or not, must provide some accounting to citizens. Clearly these people cannot simply reveal all they know to the public (hence the diagonal ‘secrecy’ line in the Figure) but they must be prepared to challenge the fetish of secrecy and reveal what they discover unless it would clearly damage the security of the nation or the rights of individuals.
In general, it is most important that oversight institutions at different levels co-operate and help each other; this will not be without difficulties since the primary organisational loyalties of agency staff, inspectors general and parliamentarians are very different but without such co-operation, oversight will be fragmented and consequently less effective. If control of intelligence networks is to be effective then there must also be an oversight network.
The basis for organising oversight bodies should be functional rather than institutional. The reason for this is simply the point made earlier regarding the rapid extension of intelligence networks in both the state and private sectors. If oversight is restricted to just one or two specific agencies there is always the risk that sub-contracting will enable agencies to avoid the gaze of the overseers. However, it has to be acknowledged that this raises profound practical challenges, especially as regards the oversight of private sector intelligence. States may not be prepared to take on corporate interests in the political battles that would be required to provide some regulation of private sector intelligence operations. Nor may they be anxious to provide the resources required for such comprehensive oversight but it must be remembered that the extent of private sector surveillance (including ‘dataveillance’) far outreaches that carried out by the state in advanced technological societies (Whitaker, 1999).
Inevitably oversight bodies will tend to be small – certainly minute by comparison with the agencies they are monitoring – so careful attention must be paid to their resources. Just two issues are considered here: access to information and staff. As discussed earlier, secrecy is a defining characteristic of security intelligence agencies and therefore a central problem for those with responsibilities for oversight or review. A key test of the real, rather than symbolic, significance of introducing oversight is the degree of access that oversight bodies are granted to the staff and files of the agencies themselves. If access is granted to everything then, at least, there is the potential for full oversight but if access is limited in some way then doubts must arise as to the genuineness of reforms. Naturally, the agencies themselves will be nervous at granting access to compartmented and classified information to outside bodies and special arrangements will have to be made for meetings and viewing documents so that trust can be created.
Officials employed in oversight departments within the agencies or their parent ministries (‘levels’ one and two in the Figure) will normally be full-time employees but in parliamentary or other external committees the members will often have duties other than intelligence oversight. However energetic parliamentarians or part-time appointees to committees such as SIRC may be, the real work of investigation that is at the heart of effective oversight requires full-time staff. As with research in general, the first issue is to decide what the questions are. Intelligence officers are very good at answering precisely the question that is asked of them and no more, so the expertise of staff will be crucial.
Other pre-conditions for effective control and oversight
The ‘principles’ of control and oversight discussed in the previous section are concerned mainly with some basic legal and structural questions that need to be answered in order to develop an effective system. But laws and structures can only provide the potential for effective control and oversight of security intelligence agencies. In this most secret area of state operations there is a real risk that the impressive-looking architectures of oversight that may be constructed are actually empty shells devoid of substantial content. This would reflect not just a failure of public control but might actually make the situation worse. The appearance of control and oversight can make it easier for agencies to defend dubious practices. If abuses or failures of intelligence power occur and the overseers have no real ability or intention to hold the agencies to account then public control is actually weakened.
Therefore, however elegant the architecture of control and oversight, it will work only if there is sufficient political will on the part of those charged with its implementation. In the face of inevitable secrecy barriers and agencies long used to autonomy from external control, those responsible for oversight must display sustained commitment and use much energy in order to surmount the considerable political challenges they will face. As was noted at the beginning of the paper, the objectives of public control are to increase both the propriety and efficacy of security intelligence. Those responsible for oversight must be careful to maintain this dual focus to their work. For example, while they must look for inefficiencies, they must resist being drawn into simply becoming unpaid management consultants.
In broad terms oversight committees will face choices as to whether they become the adversary of the agencies or their advocates (Caparini, 2002). Again, the committee should resist becoming simply one or the other. The agencies themselves and their parent ministries will hope that the oversight committee will become an advocate for the agency and its work; this might be an appropriate role for the committee if it knows that specific unjustified allegations have been made against the agency but is not the real objective of oversight. If the committee comes to be perceived as simply the mouthpiece for the agency then it will have no public credibility. Conversely, if the committee adopts an entirely adversarial stance to the agencies then it is likely to find it impossible to achieve even minimal objectives. Even if it has full access to information, it will be faced by concerted obstruction from the agencies and an associated propaganda offensive from government that its activities are ‘harming national security’.
Oversight committees must also seek to steer a delicate path between other extremes: the very fact that we are considering security intelligence accounts for the reluctance of agencies and governing executives to see it discussed in the arena of democratic politics. Yet this is the natural location for parliamentarians. While there are certainly dangers if the discussion of intelligence becomes merely a means by which political opponents can criticise, harass or discredit their opponents, it is vital that security intelligence is seen as an appropriate subject for democratic debate. If the issue is seen as being beyond the bounds for political debate then the danger is that agencies may pursue their own agendas, rather than those of elected governments, and may ultimately come to constitute the kind of ‘parallel power’ discussed earlier. The ‘politicisation’ of intelligence is necessary to its democratisation but overseers need to beware the perils of excessive partisanship that might mirror too closely the crude ‘politicisation’ of authoritarian regimes in which political opponents are defined as ‘security threats’.
Conclusion
Intelligence is the most challenging of all areas of democratic control. Its importance as a state function – ultimately seeking to protect the state itself - is matched by the extent to which it can be abused in the self-serving pursuit of power and wealth at the cost of the right, liberties and lives of citizens. The history of the use and abuse of intelligence power in the maintenance of corrupt and authoritarian regimes accounts for the degree of public cynicism and suspicion with which it is viewed. During the process of democratisation these public attitudes will change only slowly and the success or otherwise of new oversight procedures will be crucial in determining the pace and direction of that change. In the short term, those who are elected and serve within oversight committees need to bear in mind the following sobering thought passed to me during research some years ago:
‘It is not possible for politicians involved in oversight to win votes on security intelligence matters, but they can certainly lose them.’
References
Caparini M. (2002) ‘Challenges of Control and Oversight of Intelligence Services in a Liberal Democracy,’ Paper presented to Workshop on Democratic and Parliamentary Oversight of Intelligence Services, Centre for the Democratic Control of Armed Force, Geneva, October.
Center for the Study of Intelligence (1977) ‘Critique of the Codeword Compartment in the CIA,’ Central Intelligence Agency, March, accessed at www.fas.org/sgp/othergov/codeword.html
Cepik A. and Antunes P. (2001) ‘The New Brazilian Intelligence Law: an institutional assessment,’ Paper given to Center for Hemispheric Defense Studies, Washington DC, May.
Edelman M. (1964) The Symbolic Uses of Politics, Illinois: University of Illinois Press.
Gill P. (1994) Policing Politics: security intelligence and the liberal democratic state, London: Cass.
Giménez-Salinas A. (2002) ‘The Spanish Intelligence Services,’ in J.-P. Brodeur et al (eds.) Democracy, Law and Security: internal security services in contemporary Europe, Alershot: Ashgate.
Hastedt G.P. (1991) ‘Towards the Comparative Study of Intelligence,’ Conflict Quarterly, XI:3, 55-72.
Holt P.M. (2000) ‘Who’s Watching the Store? Executive-Branch and Congressional Surveillance,’ in C. Eisendrath (ed.) National Insecurity: US Intelligence after the Cold War, Philadelphia: Temple University Press.
Joffe A.H. (1999) ‘Dismantling Intelligence Agencies,’ Crime, Law & Social Change 32: 325-46.
McDonald D. (1981) Commission of Enquiry Concerning Certain Activities of the RCMP, Three Reports (First published 1979), Ottawa: Minister of Supply and Services
Rzeplinski A. (2002) ‘Security Services in Poland and their Oversight,’ in J.-P. Brodeur et al (eds.) Democracy, Law and Security: internal security services in contemporary Europe, Alershot: Ashgate.
Szikinger I. (2002) ‘National Security in Hungary,’ in J.-P. Brodeur et al (eds.) Democracy, Law and Security: internal security services in contemporary Europe, Alershot: Ashgate.
Voronin Y.A. (1997) ‘The Emerging Criminal State: economic and political aspects of organized crime in Russia,’ in P. Williams (eds.) Russian Organized Crime: the new threat? London: Cass.
Whitaker R. (1999) The End of Privacy: how total surveillance is becoming a reality, New York: New Press.
Peter Gill
Reader in Politics and Security
School of Social Science
Liverpool John Moores University
Liverpool, UK, L3 2ET
November 2002