Dr. Wilhelm Agrell
Lund University, Sweden
Although a formally non-aligned country with strong economic and security links to the Western powers, Sweden nevertheless developed an expansive national intelligence system during the Cold War. After the tumultuous shift of European security policy between 1989-91, Sweden realized immediate benefits in the area of national security; it went from the exposed position of a front-line state in the Baltic to an embedded position behind a new Cordon Sanitaire to the east. As other small European countries, Sweden in the 1990s was thus faced with the task of aligning its national intelligence system with new international premises and a broadened, but largely unknown, future security agenda.
The attempts to reform its system offer insights into the process involved in changing intelligence agendas and institutions, and into the problems facing national intelligence policy caused by globalization and European integration.
The Challenge of Reform
The rapid, dramatic changes in European security and politics in the late 1980s and early 90s had the effect of redefining national intelligence, its goals, and its institutions. The Cold War was over, and gone with it were the military structures and day-to-day activity that Western intelligence systems had been designed to survey and assess. If the Western defense establishments and defense industries faced situations where much of their raison d'etre was gone, the same was true for the intelligence communities with no Cold War Order of Battle as a given priority. But as traditional tasks lost priority, new tasks took their place. The post-Cold War world was not less complicated, and certainly not less hazardous1.
The challenge of intelligence reform, however, was nothing new for the 1990s. National intelligence systems in most Western countries became increasingly outmoded in the 1970s and 80s as new problem areas became apparent, from the oil crisis to control over technology transfer to growing refugee and illegal migration.
The intelligence institutions were also faced with the double dilemmas of maintaining democratic control while managing the effects of technical intelligence collection. Once the peak of the Cold War had passed in the1960s, the legitimacy of closed state-in-state intelligence organizations was put in question. Occasional intelligence "scandals" continued to poison the political climate and the democratic credibility of national intelligence in countries such as Britain, West Germany, Norway, and Sweden2. Cold War methods and Cold War ethics, especially in domestic intelligence, continued to bring discredit to all aspects of intelligence. The stigma of illegality and oppression only started to crumble in the 1990s. One cause of the change in climate was the increasing practice of companies, banks, and public institutions to euphemize their intelligence efforts by using more positive or neutral designations3.
If ethics, legality, and political control were the overt problems, the imbalance between intelligence collection and intelligence analysis, or finding questions for all the answers, was the predominant internal problem facing the intelligence communities and also the decision-makers dependent on its output. Already in the1970s, the intelligence communities felt the impact of the information explosion that society was to experience in the 1990s. Few effective countermeasures to this phenomenon were developed; in addition, the increasing international tension and the arms race in the late 1970s and early 80s further stressed the need for more capable and effective technical means for collecting intelligence.
At the end of the 1990s, the Western countries, as well as the former east and central European countries in the process of transition, faced the tasks of reassessing their security priorities and redefining their intelligence goals, both designed to accommodate the new ways of thinking about national security.
The Swedish Intelligence System in the Cold War period
Sweden, being a formally non-aligned country conducting a policy of neutrality between East and West, from the late 1940s informally established close ties with a number of key NATO countries: the Nordic neighbors Denmark and Norway, but first with Great Britain and the United States. In 1960 Sweden became a major recipient of US military technology and far-reaching security guarantees that de facto gave Sweden a position similar to the NATO allies.
Day-to-day cooperation developed mainly in the areas of advanced defense technology and intelligence collection. Sweden was well situated for the monitoring of activities in the Baltic and Western USSR through radar surveillance, signal intelligence, and sea/air reconnaissance. From the late 1950s, Swedish intelligence (along with Norway) received significant support from the US.
The Swedish national intelligence system, established during the Second World War, remained basically unchanged throughout the Cold War4. While the larger powers concentrated their intelligence assets in organizations with a broad spectre of operations, the Swedish intelligence community remained highly diversified. The core of the system consisted of two centers for analysis and national intelligence estimates: the Defense Staff Intelligence Branch (later renamed MUST) and the Political Department of the Ministry of Foreign Affairs. This relation was at best politely distanced, at worst suspicious or hostile. Both tried to have the last say in the assessment of any threat to the country. The bulk of intelligence collection was carried out by two independent or semi-independent institutions, the Defense Radio Agency, which was responsible for all signal intelligence short of domestic illegal radio communication (a task for the Security Police), and a secret military intelligence bureau, which engaged in domestic and foreign intelligence collection; the latter's existence was unknown to the public (and most politicians) until 1973.
As the Cold War wore on, intelligence liaison became increasingly important. Swedish intelligence collection in the Baltics and from platforms to the east could supply neighbors and Western powers with pieces of intelligence that they could not readily collect themselves5. In exchange, Swedish intelligence acquired advanced intelligence technology as well as parts of the intelligence flow distributor among the intelligence services of the NATO countries.
Cold War intelligence was focused on external and internal aspects of the perceived threat from the Soviet Union and its allies. The State Security Police was the major counter-intelligence agency responsible for monitoring and investigating crimes against state security (espionage, infiltration, subversion, and sabotage). In the early 1970s, the efforts of the Security Police began to focus, with growing emphasis, on counter-terrorism and the monitoring of armed political groups of foreign origin (among others, the German Rote Armee Fraktion, Palestinian PFLP, Kurdish PKK, and Croatian nationalists).
A comparatively high priority during the Cold War was given to counter-subversion, almost exclusively directed against the Moscow-oriented communist party and its various genuine or suspected front organizations. As in Norway, the ruling Social Democratic Party in the late 1940s and early 1950s established an impressive nation-wide intelligence network to counter communist infiltration in the trade unions6.
The public revelation of secret domestic and foreign intelligence collection in 1973-74 resulted in the first comprehensive efforts to evaluate and reform the Swedish intelligence system. The parliamentary intelligence commission of 1974 concluded that the Swedish intelligence system lacked guidance, effective political control, and analytic efficiency, shortcomings that were shared by many Western intelligence systems in the 1970s7. The commission of 1974 formed a parliamentary board to monitor the intelligence services. However, the role of the intelligence board was limited by budgetary concerns, and operational control remained within the services.
The commission of 1974 discussed a new intelligence structure, but in the end it suggested only marginal changes. All attempts to change the existing intelligence institutions turned out to be a slow and complicated process. Institutions could be assigned new tasks and new instructions, but with staffing and internal culture left in place, little change occurred in the 1970s and 80s. The Security Police successfully resisted repeated attempts to establish political control. The parliamentary board appointed to supervise the vast register of Swedish citizens appears to have had very limited insight, and no impact on conduct.
The role of intelligence in national decision-making
While large segments of the Swedish intelligence system were engaged primarily in counter-intelligence and counter-subversion, its main task was to supply the military and political decision-makers with short- and medium-term early warning. The role of the intelligence system in the conduct of long-term policy and the day-to-day implementation of this policy was utterly limited; the system generated a plethora of "nice-to-know" intelligence, but it did not necessarily fulfill the intelligence needs of top- or medium-level decision making.
The commission of 1974 also examined the function of this early warning system. After reviewing the raw intelligence and intelligence assessments from a number of international crises in the 1960s, a commission concluded that the system really had been effective8. The study revealed, among other findings, that Swedish intelligence had monitored in detail the Warsaw Pact mobilization and exercises prior to the 1968 invasion of Czechoslovakia. However, it misinterpreted the political developments and thus was unable to predict the timing of the invasion.
Although the intelligence system had ample resources for intelligence collection, the analytic function as well as contact with national decision-makers remained a bottleneck. The more complex the tasks grew, the more obvious the limitations became. Being mainly targeted against the perception of a military threat, the intelligence system possessed only limited assets in the areas of economy, trade, infrastructure, and social indicators. An office for East European economic intelligence was established in the 1950s by the secret intelligence bureau and the Swedish Industry Association, but the focus was on the defense-relevant Soviet and WP economy; trading with the Eastern bloc remained a marginal business for Swedish industry.
From the 1970s, the focus of business and national decision-making was shifting away from the Cold War agenda, but intelligence support in new fields was at best incidental. The consequence was often no or inadequate intelligence support for vital decisions and no early warning regarding crises outside the East-West framework. One example was the multibillion losses suffered by the Swedish government due to over-optimistic export credits supplied to Swedish companies in Iran prior to and during the fall of the Shah's regime in 1978. This was probably neither a case of inadequate intelligence information nor a lack of competent analysis. The relevant questions simply were not asked9.
A changing security agenda
After the turmoil caused by the shift in European Security policy in 1989-91, Sweden realized immediate benefits in national security, moving from the exposed position of a front-line state in the Baltic to an embedded position behind a new Cordon Sanitaire. The potential geo-strategic threat from the Soviet Union disappeared, Soviet bases in the now independent Baltic states were closed, and the once powerful Soviet armed forces were rapidly crumbling. In the late 1990s, the Swedish military intelligence (MUST) concluded that Russia no longer possessed the operational weapons systems, trained units, or transportation to sustain an offensive operation against Scandinavia. The threat was gone for the foreseeable future; and internally, the Signal intelligence complained that the level of routine Russian activity in the Baltic was so low that the personnel manning the electronic intelligence stations were not receiving adequate training.
Although the traditional threat was gone, the intelligence agenda was not empty. Just the opposite. Sweden in the 1990s was confronted by a mix of intelligence needs, but without any self-evident priority. Since a Russian invasion appeared remote, defense was cut down and essentially mothballed. Instead of on rapid mobilization, Swedish defense was based on timely early warning (one to two years warning of drastic change in the European security environment). The intelligence system was relieved of most of its routine monitoring tasks and assigned a new role.
Bosnia and Kosovo not only revealed that European security in the early 1990s was unstable; in addition, the demand for collective crisis management pushed Swedish political and military decision-makers into situations where they had to evaluate the consequences of Swedish military participation in multi-national forces, such as UNPROFOR, IFOR/SFOR and KFOR. Suddenly, detailed intelligence about remote conflicts was crucial; for example, the government in 1993 decided not to deploy Swedish UN forces in the Srebrenica enclave.
Other new threats came as side effects of the political and social changes in the East and Southeast. As most other EU countries, Sweden by the mid 90s became increasingly restrictive toward spontaneous non-union migration and took strong action to prevent and block asylum- or employment-seekers from reaching the union area. A more hostile threat came from the rapid increase in organized crime, facilitated by increased freedom of movement and international business.
However, the most profound changes in the intelligence agenda concerned matters outside traditional security policy and new transnational threats. Globalization, the de-regulation of the flow of international capital, and Swedish membership in the European Union in 1995 transformed the core of national security, changing it to participation in a process of political, social, and economic integration. The new short- medium- and long-term intelligence assessments required for the conduct of national policy were radically different from the old, traditional - or extended - threat assessments.
Attempts and limitations of reform
In the 1990s, an array of "new" state intelligence services was in place to offset the "new threats." Prompted by the free movement within the Union, the custom authorities shifted their emphasis from traditional border crossing control to an active campaign against organized smuggling. The role of customs service intelligence rapidly changed; as a result, the customs criminal investigation branch added a new arm.
In a similar fashion, the police created a criminal intelligence service (CIS) designed initially to combat organized crime, but also to fulfill the role of Swedish participation in the Europe intelligence cooperation. Thus the national police suddenly had two intelligence institutions with overlapping functions - the Security Police and the CIS - the former responsible for operations against armed (domestic and/or foreign) political groups, the latter for operations against similar (often identical) groups with criminal intention. As expected, the two services' areas of responsibility often overlapped. For example, in the late 1990s, Nazi extremists were engaged in a mix of "ordinary" criminal acts and in acts of political and racial terrorism. The security police, however, strongly resisted suggestions to combine the two institutions.
In 1996, the Swedish government finally appointed a committee to survey "the tasks, guidance, and outline of the intelligence service." The committee's task was to evaluate the impact of the many changes in the international environment, the impact of the new security agenda, and resolve the problems inherited from the 1974 commission. The committee, chaired by a leading jurist, included several senior intelligence officials but no representatives from the private sector or the universities. The committee's findings were published in March 1999 in a lengthy document that was fully declassified, in itself an important concession to the necessity for increased openness in intelligence matters10.
The committee's findings concerned the nature of the changes in national security and the problems involved in redirecting and coordinating national intelligence efforts. National intelligence tasks had, the committee found, become increasingly long-term, geographically remote, and more demanding of analytic capability. Any changes, therefore, should result in a more flexible national intelligence structure with full competence in areas previously subordinated. But first, achieving coordination and guidance was a primary objective.
Lack of coordination in the Swedish intelligence system had been debated internally since the end of the Second World War, but little had been decided and still less implemented. The major reason was the strong inter-institutional rivalry between the intelligence branches and the structures that they served. The services fought for "their" intelligence services; so did the joint defense staff, the ministries of foreign affairs and defense, and the state police. Although there were compelling arguments for a centralized intelligence agency, the politicians and senior military and civilian officials feared the concentration of power in such an organization and the subsequent lack of external control. Also, the problem of inadequate political and legal control over the powerful Security Police was still ongoing.
Although the 1996 committee was aware of the changes in the international environment and thus the need for comprehensive intelligence efforts in new fields, it failed to suggest any corresponding institutional reform. On the contrary, the committee declared itself satisfied with the existing order (influenced, perhaps, by an internal clash between the Ministry of Foreign Affairs and the Military Intelligence Service (MUST) over the crucial task of assessing future military threats). Instead, the intelligence services should first recruit analysts with broad competence; then they should compensate them so that they stay in the service. But with business intelligence offering higher salaries, it is probably a vain hope.
Perhaps most surprising, the 1996 committee dealt only marginally with the topic of most concern to intelligence analysts; that is, the future relationship between national intelligence assets and goals and the growing number of serious non-state intelligence entities. Much of the analytic capacity lacking in the national intelligence services also existed in the intelligence departments of the large corporations and in the consulting firms employed by them and by the major financial factors.
The outcome of the Swedish reform process in the late 1990s was not impressive. Sweden adopted an intelligence coordination board similar to that which existed in Britain for decades and in the US for most of the Cold War. The intelligence community, therefore, consisted of the same institutions created during the Second World War. Meeting the challenges of complex international developments will thus continue to be an issue of division of the labor, rather than one of integral analysis.
Is a national intelligence system an illusion?
The attempts to reform national intelligence systems in Sweden and the other small and medium-sized European countries are based on the implicit assumption that creating a new, more appropriate system is possible. This may be so, at least in theory. National intelligence systems were once created to counter the extreme circumstances of the Second World War and the effects of the Cold War. National security and national decision-making in this period was fairly one-dimensional and coherent, even in countries with an export-oriented market economy.
Advanced business intelligence is still predominantly a component in large corporations; there, the level of professional skill is about even with and, in some instances, surpasses the traditional national intelligence institutions. Beyond doubt, the analytic capability and the skill to plan and execute mission-oriented intelligence efforts are higher in large corporations competing in the world markets. If national intelligence institutions are to reform, they must learn from and adapt to the experiences and methods of professional business intelligence.
But with long-term policy becoming less a restricted national matter, so is national intelligence. Today, the European Union is establishing a structure for crisis management that includes the military capability to intervene in ancillary conflicts not directly threatening member states. I suspect this effort will fail if it does not include - or rather is precluded by - the establishment of an integrated European intelligence system for early warning and evaluation of threats from local conflicts. Without such a system, Union members must rely on the United States for intelligence collecting, for intelligence analysis, and also for policy recommendations. European attempts to intervene in the successive Balkan crises of the 1990s are replete with examples of the hazards that accompany intelligence voids and an over-reliance on US estimates.
Is a national intelligence system feasible in a world where international structures, security goals, and intelligence agendas are inconstant and fluid? The answer must be no, if national intelligence systems remain the type of omnipotent, closed intelligence bureaucracies of the Cold War. The answer could be yes if national intelligence is based more on cooperation and networks, more on national goals that are supported by information networks and by the analytic competence of an internationalized economy and society. True, the key functions of national intelligence will still rely on specific institutions for coordination and intelligence collection, especially in the field of international crime prevention, counter-terrorism, and the monitoring of incipient military threats. But to achieve the ideal - insightful perception of the goings-on on the international stage - still demands a surgical change in intelligence priorities and attitude.
1 The debate on the changing intelligence agendas has been dominated by an American perspective; see Roy Godson et. al. US Intelligence at the Crossroads, Brasseys 1995. It is notable that a number of books published in Sweden have focused more on strategic intelligence for private enterprise than on national intelligence.
2 See Stephen Dorril, The Silent Conspiracy, Mandarin 1995 that deals with the problems of legality and political control in British counter-intelligence and counter-terrorism in the 1970s and 80s.
3 The leading Swedish business bank Skandinaviska banken introduced an intelligence department in the 1990s. The department was not new; it originated from one established in the 1920s, but the name Internal Statistics was chosen as a euphemism.
4 Swedish intelligence archives from the Cold War have only recently been available for research. No major overview similar to the ones published on Norwegian and Finnish intelligence yet exists. The Second World War is dealt with in Wilhelm Carlgren, Svensk underrättelsetjänst 1939-1944, Liber 1985. An insight into the secret intelligence service in the first two post-war decades is given in the posthumously published memoirs of the director 1946-1965: Thede Palm, Nagra studier till T-kontorets historia, The War Archives 1999.
5 The first known photograph of the Soviet Tu 126 Backfire bomber was, for example, taken in the mid 1970s by a pair of Swedish fighters that intercepted the bomber on a trial run over the eastern Baltics.
6 The former director of the secret intelligence service IB, Birger Elmér, in 1993 informed a government investigation commission that the Social Democratic Party could draw on a network of 20,000 informers that covered virtually every state and private enterprise in the country. This vast network was co-employed by IB.
7 Den militära underrättelsetjänsten. Betänkande av 1974 ars underrättelseutredning. Sveriges Offentliga Utredningar 1976:19. This is the declassified version of the report; the full version is still classified. Some of the annexes were only recently released (see below).
8 The historian, Stig Ekman, was not able to publish his investigation; classified as top secret, it took him more than twenty years to get a declassified version released with large deletions. This declassified version was eventually published under the title Den militära underrätelsetjänsten. Fem kriser under det kalla kriget (The Military Intelligence Service. Five crises during the Cold War), Carlssons, 2000.
9 The US intelligence community experienced a similar failure, but with more devastating consequences for national security. See Iran: Evaluation of US Intelligence Performance Prior to November 1978. Permanent Select Committee of Intelligence. US House of Representatives 1979.
10 Underrättelsetjänsten - en översyn. Betänkande av 1996 ars underrättelsekommitté. Statens Offentliga Utredningar 1999:37.