France's Charles de Gaulle aircraft carrier will be deployed to support operations against Islamic State in Syria and Iraq, November 18, 2015.
Jean-Paul Pelissier / Reuters

As France intensifies its airstrikes against the Islamic State (or ISIS) in Syria after the November 13 terrorist attacks in Paris, the question of when to intervene—that is, to forcibly come between another government and its subjects—is once again pertinent. We’ve dealt with this question before. Think of Russia and the United States as they began to arm their allies and intervene in Syria; or France, the United Kingdom, the United States, and other NATO allies in Libya in 2011; or, the failure to intervene in a timely or sufficient manner in Rwanda in 1994 or in Bosnia before 1995.

But where you stand on intervention depends in part on your approach.

If you believe that there is a humanitarian commitment to save the lives of people around the world, you are likely an interventionist. If you assume that there is a right to national self-determination and sovereignty, you are likely to be a noninterventionist. And, if you regard national security as a responsibility that no government can fully cede to an international organization, you will want to intervene when it is necessary for your own national security, but not otherwise.

But what if you think that all three of these approaches should influence a decision to intervene? Then you are in the very good company of the great, nineteenth century philosopher, John Stuart Mill, who made a landmark attempt to reconcile these values with his 1859 essay, “A Few Words on Nonintervention.”  

In my recent book, The Question of Intervention, I analyze Mill’s arguments, defending some, rejecting others, and refining a few. His classic treatment of nonintervention helps us begin to unravel two fundamental puzzles that arise when determining when to intervene and when not to.


The first puzzle is reconciling human rights and democracy with nonintervention, which Mill argues should be the default norm. Why not enforce basic human rights, democratic government, and beneficent administration abroad?

Mill’s view of mankind regards humans as sentient beings who are fundamentally similar, capable of experiencing pleasure and pain, and recognizing good and bad. Compassion should thus drive us in making decisions that maximize the greatest good for the greatest number of people.

This philosophy leads, in turn, to two guiding principles when it comes to governance. The first is to maximize equal liberty: Permit all individuals to live without coercion and develop their potential freely, as long as they do not interfere with the rights of others. Second, support representative government: Collective decisions should indirectly reflect the will of the majority through elected representatives.

One might think that these principles, when applied to an international framework, would resemble a global version of the U.S. Constitution’s “Guarantee Clause,” which promises a representative, republican form of government to all states.

John Stuart Mill, 1872-1873.
Instead, Mill believes that forcefully imposing a free democratic government upon another country violates the very principles of liberty and representative government. Authentic freedom means we, and our neighbors, choose how that freedom is defined. For example, consider the United Kingdom and the United States, two card-carrying liberal democracies. The former has retained its monarchy and recognizes an established church; the latter has an elected president and prohibits the establishment of a state religion.

Moreover, Mill rejects the idea of intervention to impose liberty or democracy, writing that “to go to war for an idea, if the war is aggressive, not defensive, is as criminal as to go to war for territory or revenue.” He goes on to say that nonintervention is counterproductive in promoting freedom and democracy: Intervention will not do any real good and the war that accompanies intervention always does harm.

It is only through what Mill calls an “arduous struggle” (or in the extreme, a violent revolution), do a people come to define their freedom, and in the process, develop the capacity to exercise self-government—obey laws, fight for their country, and pay taxes.

In fact, a “knapsack regime,” one established through foreign armed intervention, is likely to be harmful and often brings about three undesired outcomes. The first is that the knapsack regime will collapse as soon as the interveners leave since the liberals ruling the regime have not earned political support from the population. The result then is another civil war. A second scenario is that the knapsack liberals, given their thin domestic support, turn to despotism in order to hold on to power. The third outcome is that the interveners never leave. They stay in power in order to prop up the knapsack regime, which becomes, in a sense, a client that loses state self-determination.

Mill rejects the idea of intervention to impose liberty or democracy, writing that “to go to war for an idea, if the war is aggressive, not defensive, is as criminal as to go to war for territory or revenue.”

I, along with Camille Strauss-Kahn, a doctoral candidate at Columbia University, evaluated the soundness of Mill’s intuitions by looking at historical data. We examined every intervention since 1815, limiting our analysis to overt armed interventions by foreign troops that come between a government and its people. We found that of the 334 interventions, 221 were militarily successful. Of those, 56 led to new or renewed civil war; 68 produced deeper autocracy; and 146 led to imperial rule. (These bad outcomes total more than 221 because some unfortunate countries experienced multiple harms.)  Only 26 interventions—or 12 percent—produced a free, independent, more rights respecting or participatory government.

The data does corroborate John Stuart Mill’s views that nonintervention should be the default norm. A 12 percent success rate is far too small to warrant optimism about intervention. Indeed, every justifiable intervention needs to address the possibility of one of these three harmful outcomes and seek to avoid them.


The second puzzle is then, why, if Mill is so committed to nonintervention, he argues that it is sometimes permissible to intervene? Here, Mill offers a number of cases where external humanitarian concerns or national security needs can overridethe right to self-determination or where countries can have their rights to national self-determination disregarded because they are not effective and unified nation states.

Mill, unfortunately and unconvincingly, argues that benign imperialism, a form of paternal authority, is acceptable in India and other countries that he shortsightedly portrayed as incapable of ruling themselves.

Mill is more persuasive, however, when he draws readers’ attention to national liberation or secessionist movements, in which the oppressed minority makes an “arduous struggle” against an oppressor that cannot be defeated. He argues that indigenous struggles often need help in lifting the oppressive (and often foreign) yoke. For example, Britain would have been justified in assisting Hungary break from Austria from 1848–49. But Britain chose not to intervene. Even here, Mill’s theory is incomplete. National liberation for every ethnic group is a recipe for permanent rebellion. We need to develop principles of and procedures for self-administration that are designed to satisfy, where they can, demands for legitimate autonomy.

Mill offers a number of other examples where it is acceptable to override the principle of nonintervention, such as when national security is under imminent threat or when, following a war of self-defense, the victorious defender need not stop at the border, but can intervene in order to eliminate a “standing menace” to peace. Mill had in mind the exile of Napoleon Bonaparte in 1815. Modern examples would include the post World War II occupations of Germany and Japan. Mill also discusses the humanitarian crises that result from protracted civil wars. Waiting for the “arduous struggle” to determine a legitimate sovereign can be a prescription for unending massacre. Sometimes external mediation can produce a coalition regime capable of ruling the country on a stable and more legitimate basis. Mill’s example, Spain and the United Kingdom’s 1846 intervention in Portugal, is relevant when considering modern debates about peacekeeping.

With the advantage of a hundred plus years of hindsight, we now know that humanitarian interventions were frequently exploited to further imperial projects. The British and the French decried barbarism in Africa and Asia, as Mill did in India, and then imposed imperial rule. In 1898, the United States rescued Cubans and Filipinos from Spanish oppression, but stayed on, indirectly or directly controlling domestic politics for half a century. So how can this be curbed?

Today, we have the Responsibility to Protect (RtoP) doctrine, adopted unanimously by the United Nations in 2005 to protect populations from genocide, war crimes, crimes against humanity, and ethnic cleansing. It also has bipartisan support from the United States and was implemented in Libya in 2011 by NATO members and the administration of President Barack Obama—but with very mixed results. RtoP is not law, but it is a revolutionary international commitment. It sets a new norm for legitimate action that is designed to provide humanitarian protection while preventing imperial exploitation.

French President Francois Hollande attends a ceremony in the courtyard of the Hotel des Invalides in Paris, France, November 19, 2015.
Philippe Wojazer / Reuters

The lessons of the 2011 intervention in Libya are that we need more responsibility while protecting: a better understanding of the facts on the ground as strategy is developed, more accountability in the implementation of Security Council mandates, and more attention to responsible state building in the aftermath of intervention.

Even with these reforms, we need to uphold the strong presumption against intervention, as Mill argued. We should keep two essential lessons firmly in mind: First, don’t intervene unless there are very strong reasons to override or disregard nonintervention. Today, this requires RtoP authorization from the Security Council whenever possible. Second, if one does intervene, it must result in a self-sustaining peace. This means a peacebuilding strategy that involves building the capacity for sovereign rule—or else the intervener will be morally responsible for the resulting civil war, despotism, or empire.

The complexities of Syria today defy easy prescriptions. At the same time, we see an international air war waged by France, the United States, and Russia against a non-state actor (ISIS); a messy civil war in Syria (between Assad and the various rebels) but with unilateral interventions by multiple outside powers; a UN-endorsed convention against chemical weapons; and ongoing civilian atrocities (perpetrated primarily by Assad’s regime and ISIS). In 2011, a “Mill-ian” would have kept the outside players at bay and left the struggle to the Syrians. In 2015, the best and least that should be done is to assist the displaced and refugees, defeat ISIS, and put concerted international pressure on Assad and the Syrian rebels to lay down their weapons and build a Syrian future. Admittedly, all easier said than done.

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now
  • MICHAEL W. DOYLE is University Professor and Director of the Global Policy Initiative at Columbia University and former Chair of the United Nations Democracy Fund Advisory Board. He is the author of The Question of Intervention: J.S. Mill and the Responsibility to Protect (Yale University Press, 2015).
  • More By Michael Doyle