"In the wake of such general conflicts as the Napoleonic wars or World Wars I and II, two conditions prevail that are virtually essential to the fact, or the illusion, of collective security: (1) The victorious powers are momentarily in concert; this provides the basis for equating their provisional coalition to a disinterested concern for universal world order. (2) The defeated powers, are by consensus of the victors, clearly labeled the 'aggressors.'"
--Earl Ravenal, "An Autopsy of Collective Security," Political Science Quarterly (Winter 1975/6).

Earl Ravenal's rule applies equally in the wake of the Cold War. Buoyed by the collapse of the Soviet Union, proponents of collective security have been writing up a storm, producing literature full of hope for the future. Central to this new vision of collective security are new roles for two familiar Cold War international organizations: the United Nations, now freed from a paralyzing standoff between two superpowers with vetoes in the Security Council; and NATO, the victor over the Soviet Union and the Warsaw Pact, now reaching to embrace new friends and missions.

Much of the hope placed on the UN in a new era is grounded in the belief that it can finally provide for collective security. The hopeful cite the collective expulsion of Iraqi forces from Kuwait as evidence of this. As a function of its only unique virtue, all-encompassing membership, the UN becomes the natural instrument for the wishful tinkerer in trying to implement collective security.

More surprisingly, perhaps, with war in the Balkans and evidence of instability elsewhere in Central and Eastern Europe, collective security also figures increasingly in the context of developing a new role for NATO. As Josef Joffe writes: "[Collective security] has been the idŽe clef in the debate on post-Cold War security in Europe. Only the label has been modernized. Instead of C[ollective] S[ecurity], advocates use terms such as 'overarching security system,' 'pan-European security,' 'co-operative security' or 'expanded CSCE.'"

While the Conference on Security and Cooperation in Europe (now officially the Organization for Security and Cooperation in Europe, the OSCE) and West European Union (as the security organ of a post-Maastricht EU) have gotten some attention in this discussion, most hopes have fallen upon NATO as the prime agent for Europe's collective security, and the alliance's Partnership for Peace has been seen as an instrument for defining such a mission.

Before one can determine whether the UN, NATO, or the OSCE is an appropriate vessel for pursuing collective security, it is necessary to understand precisely what is meant by the term. And that involves clarifying not only what "collective security" is, but what "security" itself means. For, unfortunately, both terms are now in danger of being stretched into unrecognizable shapes and established distinctions ignored, to accommodate the agenda of some of their users. After a recent Freedom House symposium on the expansion of NATO, for example, I asked one of the participants, Michael Mandelbaum, whether his use of "collective security" to include preparation to meet threats from outside the membership of the alliance fit the traditional meaning of the term. Mandelbaum dismissed my query as a semantic quibble. A lot more than a quibble is at stake.

A Taxonomy

Richard Betts has noted the propensity of many to apply the label "collective security" to institutions and processes which do not fulfill the unique, defining criteria that were set out and championed by Woodrow Wilson, the century's leading proponent of the concept:

[I]n the generations after Wilson many felt the need to endorse collective security while defining it in ways that overlapped significantly with traditional arrangements. . . If a collective institution is really to function as a security system rather than a slogan, the elements that are conceptually unique rather than those which are shared with other constructs should set the standard for assessing the idea.

What are these unique elements? Perhaps the best way to define collective security is in terms of a three-part typology.

First there is "collective security" as it has been conventionally understood until recently. It reflects a security arrangement whereby a clearly delineated group of nations agrees that if one nation in their membership encroaches upon the sovereignty of another, the rest of the members will take collective action against the aggressive member. This internal focus is the distinguishing mark of collective security, and the basis for action against aggression is, hence, universal group consensus minus the aggressor(s). Thus, collective security is and has always been a commonly agreed upon arrangement to protect the status quo within a given group from change brought about by warfare, without identification of coalition partners or threatening revisionist powers beforehand. The League of Nations is the archetypical and indeed original example of collective security in practice. Japan and Italy made a mockery of it when their expansionist designs in Manchuria and Abyssinia went unpunished. Its abject failure has much more to do with its premises than an insular America's failure to join it.

A second distinct security arrangement may be labeled "unspecified collective defense." Collective defense is a formula whereby group members are concerned with threats from without. But subscribers to a collective defense arrangement, established to deal with outside threats, need not know or declare at the time of entering into the arrangement who the outside threat might be. Collective defense of this kind is concerned with creating a pledge for united action against one or more unspecified external aggressors; it is limited to outside aggressors, but unlimited as far as they are concerned. An example is the "Little Entente" of the 1920s and 1930s, the alliance formed in 1920 by the three successor states of the Austro-Hungarian dual monarchy, Czechoslovakia, Yugoslavia, and Rumania. Its purpose was to guard against the twin dangers of Hungarian revisionism and a Habsburg restoration. The threats to its members were diverse and diffuse--including Germany's threat to Czechoslovakia, Italy's to Yugoslavia, the Soviet Union's to Rumania. And when the League of Nations sat dormant as Italy assaulted Abyssinia and Germany remilitarized the Rhineland, these partners turned to France to guarantee their unspecified collective defense arrangement.

Third, and complementing the above, there is "specified collective defense," an alliance in the strictest and most limited sense. It represents preparations made against outside threats which have already been identified. The expansive literature on alliances is concerned overwhelmingly with this articulated collective defense. The North Atlantic Treaty Organization, as it existed from 1949 until the end of the Cold War is an archetypical example, and a remarkably successful one.

Two observations about this typology. First, the widest conceptual divide lies between collective security and the two types of collective defense. This is to say, the difference between an arrangement aimed at coping with an aggressor arising within the membership and one aimed at external threats, whether clearly identified or not, is greater than the difference between an arrangement to cope with an unknown (or unacknowledged) threat and one aimed at dealing with a known enemy. Second, other things being equal, the likely success of these arrangements is in ascending order. Historically, an alliance against a specified threat has worked and will work better than an unspecified collective defense; unspecific collective defense, in turn, has worked and will continue to work better than collective security.

The late 1940s illustrate this hierarchy of efficacy. The United Nations was created to provide collective security. It would fail to keep the peace over the next forty years, as the two superpowers exercised their Security Council vetoes. Just as this UN gridlock was becoming apparent, the 1948 Treaty of Brussels--involving the Benelux countries, Britain and France--was signed. At this early moment in postwar history, not only the Soviet Union but Germany was seen as a potential threat, and in order to cover both, the treaty provided for unspecified collective defense. The following year the North Atlantic Treaty, based on American leadership, was signed--clearly a specified collective defense agreement, with the Soviet Union and its satellites as the identified enemy. The fact that the UN, the Brussels Treaty Organization, and NATO were formed in succession reflects the relative efficacy of the three types of organizations they represent.
A fundamental conceptual failing goes a long way toward explaining why collective security is at the bottom of this list in terms of credibility, and why its historical record is dismal: the argument that sustains it involves circular logic. Joffe puts it well: "[Collective security] presupposes, rather than creates, its own necessary condition: a common commitment to the peace. If so, institutions do not matter. When all are agreed, there will be stability no matter what--with or without collective security. Synonymous with a sturdy status quo, collective security turns outs to be a state of affairs, not a mechanism that brings it about."

What to Make of NATO?

Those who envision NATO as an instrument of collective security exhibit this circular reasoning. Collective security would surely work as a new goal for NATO, but it would do so precisely because it is unnecessary among its present members, who are not likely to go to war with one another. NATO's members are democracies; democratic nations very rarely fight one another.

NATO's very success has prompted the search for a new role. As Betts observes, "[C]ollective security generates interest more by default than by its own merits: unilateralism seems ineffective or illegitimate, and alliance without adversary seems anachronistic and empty."

Obviously the alliance has lost its original utility as an instrument of specific collective defense, since the Warsaw Pact and Soviet Union no longer exist. But the Partnership for Peace not only seeks to move away from specific collective defense, but from any kind of collective defense altogether. Henry Kissinger has identified the essential flaws: "The Atlantic Alliance assumed that its threats came from outside the treaty area, while the Partnership seeks to pacify threats within its area by affirmations of good intentions."

There is some concrete good to be won from these good intentions, and much to be said for the idea of regular, systematic contact between the democratizing nations and the established democracies in NATO. In particular, the Visegrad nations could be "socialized" in the ways of democracy, and particularly in the vital norm of civilian control of the military instrument of policy. Arguably, NATO in its earlier Cold War existence already performed this valuable function with respect to West Germany, Portugal, and Turkey; but in earlier days these benefits were only by-products of alliance, not its central purpose.

If collective security presupposes what it pretends to create, and NATO has lost its enemy, is there a realistic future role for the alliance? Conceptually, what I have termed unspecified collective defense remains an option. It is closer in purpose to NATO's original collective defense mission than is collective security, and an unspecified collective defense arrangement can easily and quickly be transformed into a specified one.

It is possible to see the glimmerings of such a future in the alliance's role in the Desert Shield and Storm operations. Though the Gulf War fed hopes for collective security, the American-led coalition was not in essence a collective security arrangement under the auspices of the United Nations, as many observers now assert. The Bush administration cobbled together a coalition using the UN as a forum, but basing substantive membership upon NATO, along with regional powers such as Saudi Arabia. In this light, the role of the NATO core within the Gulf War coalition could be a useful model. In contrast to its utter failure in the Balkans--caught between incongruent roles as interposed neutral and anti-Serb partisan--NATO was quite successful in the Gulf War. It led a campaign against an opponent, Iraq, which was not previously "specified" as a threat to NATO members' interests, but which proved to be just that. Accordingly, if anything, unspecified collective defense should be NATO's new mission.

Security is Security

A fellow sufferer along with "collective security" in this new world of stretched concepts and mangled language is the notion of "security" itself. Simply put, and for the sake of clarity and precision, international organizations created to sustain the security of their members should continue to define "security" as the protection of the territorial integrity of those nations against external attack. It should not be defined in terms of internal conflict within nations, economic development, social development of the "nation-building" variety, or any broader processes and phenomena vaguely related to nations' interests.
This point needs to be made because the strict definition of "security" is under energetic assault. For example, the secretary-general of the United Nations, Boutros Boutros-Ghali, ties all these phenomena together in one undifferentiated mess: "The concept of security, which has traditionally been defined in strictly military terms, has evolved to encompass the economic, social and environmental problems that threaten national and international security. We have seen how problems emanating from poverty, social unrest and humanitarian tragedies in just one state can--if left unchecked--reach a magnitude that disrupts an entire region. That is why I believe that there can not really be peace without development."

Americans, too, are particularly susceptible to this line of reasoning. Twenty years ago, Robert Packenham of Stanford University identified the cultural tendency of American government officials and academics to assume "all good things go together," so that there is no need to distinguish clearly between them, to confront the task of ordering priorities, to consider whether some are incompatible with others, and that choosing is unavoidable. Now there is a corollary trend: all bad things go together. The Clinton administration has created a position of undersecretary of state for global affairs, occupied by Tim Wirth, which deals with a grab bag of issues from human rights to sustainable development to aids to population control to environmental protection and so on--a collection that has nothing in common except that it involves trying to cope with "bad" things.

Lumping these issues together packs excess baggage into the new mission statements of organizations devoted to security. Thus, under President Clinton, the financially strapped Pentagon has been ordered to divert spending toward projects which are not defense, such as defense conversion, medical research and environmental clean-up; Jeane Kirkpatrick has called this "reinventing defense." This seeing all phenomena as connected, and the consequent inability to see any one problem clearly, is reminiscent of the kind of thinking that elsewhere gave us a crime bill in which midnight basketball leagues became part of crime fighting. But defense is defense just as crime fighting is crime fighting.

All-inclusive thinkers of this kind fill many of the senior posts in the Clinton administration. Ashton Carter, when he was chairman of the executive board of the Harvard University journal, International Security, proposed expanding that distinguished journal's focus after the Cold War: "[T]he Cold War period was in a sense anomalous: the "normal" complexity and range of international security concerns and behaviors were unnaturally foreshortened, dominated by a single conflict with a rather narrow set of characteristics--heavily military, bipolar, static, and with stakes that by the end were almost abstract. Now international security issues will exhibit themselves in all their variety once again--issues of markets, resources, technology, ethnic animosities, political philosophy, and different conceptions of world order, as well as armies and nuclear weapons. There will be a lot to understand and thus a lot for the [International Security] community to do."

Holding these views, it was more or less inevitable that Carter should end up as assistant secretary of defense for international security policy in a Clinton administration.

In a similar spirit, Jessica Tuchman Matthews, who served in the early days of the Clinton administration, has written: "The 1990s will demand a redefinition of what constitutes national security. In the 1970s the concept was expanded to include international economics as it became clear that the U.S. economy was no longer the independent force it had once been, but was powerfully affected by economic policies in dozens of other countries. Global developments now suggest the need for another analogous, broadening definition of national security to include resource, environmental, and demographic issues."

While all of the issues cited by Carter and Tuchman are worth considering in policymaking, their relevance to security is highly questionable. What they are certainly not relevant to are international organizations charged with keeping the peace, and, on occasion, fighting wars. This omnibus view of social problems around the globe is a pernicious force threatening to divert organizations like NATO and the UN from their primary purpose.

"Keeping the peace"--maintaining sovereign nations' territorial integrity from other aggressive nations--is too often permitted to slide into ill-conceived military and social interventions in the affairs of nations, glibly labeled "peacekeeping."

 And these interventions do not even meet the narrow criteria of interpositional peacekeeping. This slide will occur all the more easily with the expanding impact of the "chaos industry" inside and outside of government today.

Of course, in the end there is more at stake than mangled terms and stretched concepts: these international security organizations are meant to deter and, if necessary, fight wars. Perhaps it is well to keep in mind Earl Ravenal's sardonic coda to his observation about the perennial enthusiasms for collective security: "But of course this situation does not last."