Liberalism, Intervention and Regime Change: A Problematic Record

This column by Michael W. Doyle and Camille Strauss-Kahn is part of Global Policy’s e-book, ‘Lessons from Intervention in the 21st Century: Legality, Legitimacy and Feasibility', edited by David Held and Kyle McNally. Contributions from academics and practitioners will be serialised on Global Policy until the e-book’s release in the summer of 2014. Find out more here or join the debate on Twitter #GPintervention.

Whether force is an admissible or desirable means to spread liberal values abroad has long been central to the political thought of classical liberals such as Kant and J.S. Mill and their modern interlocutors such as Rawls and Walzer. At the risk of oversimplification, one can identify broadly two groups of liberal scholars in the contemporary normative debate on military intervention and regime change: cosmopolitans, on the one hand, and communitarians, on the other. Both schools of liberals typically reject the power, resources and income-enhancing justifications that have been acceptable grounds for intervention for some realists.

Cosmopolitans typically assert that anyone who has the ability to intervene militarily in the face of massive human rights violations also has a moral duty to do so, subject to criteria of effectiveness and/or proportionality. But if the claimable cosmopolitan rights are limited, few grounds for intervention exist. Kant, for example, favored international nonintervention as a matter of principle. The only right that could be claimed by individuals on a cosmopolitan basis was the right of refuge. He also thought nonintervention necessary in the eighteenth century world of global cultural pluralism in order to stabilize international relations and ensure that each political community could freely determine its own way of life. For most modern cosmopolitans, however, the list of claimable human rights is much more extensive. For them, if a state is tyrannical and systematically oppresses its own population, it “forfeits any respect for its independence.” By implication, “international [military] intervention to displace the government and, if necessary, place the country under international trusteeship” is always prima facie morally justified and indeed required. But even these more extensive claims must, they rightly insist, be subject to proportionality considerations (the intervention must do more good than harm) and prudential considerations that often counsel against coercive regime change and make cosmopolitans in practice much closer to the next group, the communitarians.

Communitarians, for their part, have tended to recognize greater value in state sovereignty, not only because it protects small states from the imperial projects of the major powers but also because it reflects the inherently local character of legitimacy. Mazzini and Mill were willing to justify military intervention in the pursuit of a few clearly delimited objectives, such as to end protracted civil wars and save helpless populations from systematic slaughter; yet they vigorously opposed the use of force for the purpose of promoting liberty and democracy more generally. Mill famously argues that much more likely than a genuine liberation was renewed civil war, another autocracy or imperial rule. Both sensed that unless tyranny was defeated domestically ─ perhaps with economic and diplomatic assistance from the outside but crucially without foreign military intervention ─ any liberty achieved would remain exceedingly fragile and could hardly be sustained.(1)

Contemporary liberals of both sorts typically justify humanitarian military intervention as a last resort, suggesting that state sovereignty can be overcome or “disregarded” in the presence of gross violations of a population’s rights to survival, such as state-sponsored slaughter or genocide. Cosmopolitans and communitarians differ in lesser humanitarian crises, as illustrated in the famous debate between Walzer and Luban over Nicaragua, where the more cosmopolitan Luban favored intervention to reduce the casualties of civil war and the more communitarian Walzer favored nonintervention to permit the Sandinistas to broaden their base of support.

Both cosmopolitans and communitarians insist that military intervention ought to be multilaterally authorized and overseen, if it is to be legitimate. This generally implies authorization and oversight by the UN Security Council; but most liberal internationalists concede that when the Council is unable to act, for instance because of a (threatened and unjustified) veto by one of its five permanent members, regional multilateral organizations such as NATO or the African Union (AU) might usefully step in to carry out a surrogate legitimization function. The underlying assumption is that collective oversight and related institutional checks, however weak at the international level, reduce the risk of unwarranted interventions and usurpation by powerful states. Multilateral deliberation also increases the likelihood that the preferences of the target population, as well as broader consequentialist concerns about international order, will be taken into account.

When we look at the empirical record, results are also mixed. Many scholars – John Montgomery, John Owen, Minxin Pei, and David Edelstein among them – have done valuable work on this topic. These scholars do not all agree, but they are building a coherent picture. They note the many failures to impose a stable domestic regime through foreign occupation. We need only think of the U.S. intervention in Cuba in 1898 and again in 1907; in the Philippines from 1898; in Nicaragua in 1912 and Haiti in 1915. Democracy promotion and strategic dominance were dual aims in these cases. In all those cases there was a failure to establish a democratic government. In the 1920-1932, the U.K. failed in Iraq, in Palestine and in Egypt (from 1882 to 1954) to leave behind the rule of law and semi-democratic government. In the postwar period, the Soviet Union failed in Eastern Europe to leave behind stable, self-sustainable Communist governments.

On the other hand, the postwar occupations by the U.S., U.K., and France in Germany and western Austria; and the U.S. in Japan were all instances of successful democratic transplant. How and why?

First of all, there was a complete defeat. In no case was there just a liberation of one group that was then freed to rule in its own interests. A complete defeat offered a fresh slate for transformation.

Second, the occupiers were able to draw upon indigenous traditions of liberal capitalism and representative rule. The occupation had a restorative aspect to it.

Third, a good strategy was adopted, including an equalizing strategy in most of these cases, offering new opportunities for hitherto subordinated classes.

Fourth, there was an assured departure. That is, they drew a public distinction between occupation and imperial rule. The occupiers were known to be temporary.

Fifth, they were well prepared. As David Edelstein has noted, as early as 1943 the U.S. set up schools at the University of Virginia and at Yale to train future administrators of Germany and Japan. In 1943, it was not clear the Allies were going to win the war. Nonetheless, in 1943 the U.S. began to develop adequate language and other civil administration skills and undertake long-term planning. (2) This contrasts strikingly with a story about the 2003 occupation of Iraq in The New York Times in which a senior U.S. staff officer of the 3rd Infantry Division mentioned that, after successfully taking Baghdad, his division had no further orders whatsoever. That is, they had no instructions on how to occupy or govern, or on what was to happen next: a striking and, we now know, consequential difference. Add that to the weakness of democratic traditions in Iraq, the incomplete defeat of the insurgents (to put it mildly) and the very slow pace of reconstruction, and the challenges of a successful occupation in Iraq become clear.

But are these patterns generalizable to a wider range of democracy-inducing (by intention) interventions? Analyzing a list of 30 “major US intervention(s)” from 1898 (Philippines) to 2003 (Iraq) compiled by Michael McFaul, one can see that only seven (23% including the four post WWII occupations of Italy, Germany, Austria and Japan; and Grenada, Panama, and Kosovo) produced democracies ten years afterward. Two produced partial democracies (Bosnia and Afghanistan). The remaining 21 interventions (70%) all became autocracies. His similar list of 57 “covert military interventions” from Greece (1947) to Haiti (2004) yielded 15 (26%) democracies ten years afterwards; two partial democracies and 40 (70%) autocracies.

A wider examination still underway by the authors attempts a comprehensive empirical assessment of the effects of all armed, overt interventions. The investigation is still in its very early stages. We have identified 322 major, overt interventions since 1815.(3) An intervention for these purposes is an armed attack by one state in the territory of another state that is designed to intervene (come between) that second state and its population in order to change or protect its political regime, to liberate or restrain a rebellious population. Of such interventions, 126 were by liberal countries, and 195 were by non-liberal countries. Only 211 were successful, in the limited sense that they militarily succeeded in invading rather than being repulsed and defeated or having no effect. In these narrow terms, only 66 percent succeeded, and 34 percent failed altogether. We assume mixed motives in all cases; and so do not try to distinguish interventions of a specifically reformist character (though we acknowledge that such discrimination would be warranted).

What were their effects on the target state? Among the successful 211 interventions: 52 led to a new or renewed civil war within two years; 66 led to an deepened autocracy; and 136 led to empire – that is, the interveners stayed on to rule.(4) Only 25 produced a government no worse in measurements of democratic and civil liberties than the preceding government. That is, only 12 percent were potentially successful in advancing the cause of liberty and democracy. (Ironically, seven of those 25 were produced by interventions by non-liberal states, for which we can presume that producing a constitutional liberal government was not their major purpose.) In short, liberals succeeded in not worsening liberalism only 18 times – in nine percent of militarily successful interventions.

All this seems to confirm the warnings of the liberal communitarians (and Mill’s warnings specifically). One out of five are not great odds for reform by intervention, risking lives and national treasure – unless there is no other alternative to an ongoing or looming massacre.

Michael Doyle is the director of the Columbia Global Initiative and Harold Brown Professor of International Affairs, Law and Political Science at Columbia University. Camille Strauss-Kahn is a Ph.D. student at Columbia University. We thank David Held and Alicia Evangelides for useful suggestions.

Notes

(1) Mill, however, excepts huge swaths of “noncivilized” Asia and Africa from the benefits of nonintervention.

(2) For a thorough analysis of establishing strategically friendly regimes, see Edelstein “Occupational Hazards,” International Security 29, 1, (Summer) 49-91. Edelstein is defining success differently, not in terms of democracy or self-determination, but in terms of U.S. security interests. For a discussion of democracy promotion, see Carnegie Council on Ethics and International Affairs, Multilateral Strategies to Promote Democracy (New York: Carnegie Council, 2004).

(3) Inevitably, we do not have sufficient comparable information on covert interventions.

(4) The negative outcomes add up to more than 211; some cases produced more than one of these harms.