Taming intervention is harder than it seems
Bernard Finel has an interesting response to my most recent post, in which he argue that interventions with limited goals get an unfair rap:
Opponents of intervention in Syria argue, essentially, that since we can’t solve the problem, we shouldn’t try. But is that the right standard? Why isn’t mitigation, or punishment, or norm-reinforcement a perfectly valid goal? Indeed, I would argue that having modest goals and limited expectations is the sine qua non of a plausible intervention policy.
We shouldn’t be asking “can we fix it?” but rather “can we make things better?” I get that this is can become too loose a standard. But that is a bridge we can cross when we get to it. If we ever get to the point where our objectives strike me as too modest, I’ll be happy to weigh in and try to get the pendulum moving back in the right direction. But at this juncture, we’re dealing with the opposite problem.
I am actually sympathetic to this point, even if I might disagree with some of the specific cases under which it’s been employed. Limited intervention does have value, and I don’t think I’ve ever made an argument claiming that an intervention must solve all a country’s problems, only that the new problems the intervention might create must be more tolerable than the old ones, and furthermore, that we be willing to walk away from them. Additionally, a limited intervention must still keep the means of its ways and means in balance with the end it is seeking to achieve. The problem, of course, is that seemingly limited objectives often turn out to require disproportionately expensive or risky uses of force relative to what they achieve.
In terms of mitigation, exactly what problem can the United States mitigate for a worthwhile cost? Safe zones or “no-kill zones” have already been exposed as a fallacy that would necessitate either the deployment of significant numbers of ground troops alongside a major air campaign, or else suffer the same fate as Srebrenica or Tuzla. A regime decapitation might also accelerate rather than mitigate the bloodbath, as a wider sectarian war might indeed claim more casualties than government oppression which concentrates killing power on discrete cities in sequence. But might intervention in Syria still have value as punishment or norm-enforcement? Larison has some good points here:
Punishment and norm-enforcement may or may not be valid goals, but they would almost certainly come at the price of an escalated conflict and a higher death toll than would otherwise be the case (to say nothing of the costs that the U.S. and Syria’s neighbors would have to pay). It wouldn’t be much consolation if far more Syrian civilians perished in the course of punishing Assad for his earlier crimes. If “punishment” ends up meaning the collective punishment of entire pro-regime communities, that hardly seems justifiable. Punishing Assad might make sense in theory if there were any reason to believe that such punishment would make other authoritarian states more reluctant to engage in such abuses, but Assad’s crackdown is Exhibit A that intervening against one abusive regime for the sake of deterring abusive regimes elsewhere doesn’t work at all.
Precisely. In proportion with what we get, the potentially large amount of military power we might need to commit for an intervention simply doesn’t appear justifiable. Furthermore, the Syrian reaction to a high-end military intervention, which is very often needed to achieve even limited political goals that do not involve regime removal or permanent pacification.
But, I’m not sure Finel is interested in arguing the specifics of Syria so much as the mindset in general – a point to which, as I’ve said, I’m sympathetic. There are indeed times when limited intervention is a useful tool.
The problem is that limited intervention often entails conditions which most interventionists, whether liberal or conservative in flavor, often reject. Take, for example, the Banana Wars of the late 19th and early 20th centuries. The Banana Wars provide an interesting model for limited intervention because they were the product of an inherently limiting policy – the late 19th century and early 20th century interpretation of the Monroe Doctrine – and employed to advance concrete U.S. interests under that policy. In doing so, they took advantage of favorable conditions for power projection and a norm-enforcement system that privileged the reproduction or manipulation, rather than destruction or production, of existing political orders. Reintroducing these controlling criteria for when an intervention should even be contemplated is a necessary antecedent of any defining of objectives.
The Banana Wars and Limited Intervention
The Banana Wars in Latin America were the product of a geographically limiting and ideologically, relatively neutral policy controlling intervention – that the U.S. should secure a sphere of influence excluding foreign encroachment and protecting U.S. economic and political interests. The United States rarely employed the “fixing” attitude towards governance when it acted in Latin America during this time period, perhaps with the exception of the 1915-1934 occupation of Haiti. Indeed, what is notable is that even in the many interventions into Latin American domestic politics, the United States basically maneuvered within the confines of local political standards.
For example, during the United States occupation of the Dominican Republic, the U.S. was essentially picking from a bevy of competing political figures rather than attempting to unseat the regime – that is, not just the leadership but the constitution, institutions and norms which support it. In fact, the United States engaged in regime consolidation. The intervention began when the abdication president of the Dominican Republic, Juan Isidro Jimenez threatened to result in the Secretary of War, Desiderio Arias, taking control of the country. Accordingly, the United States Navy under Rear Admiral William Caperton threatened Santo Domingo with bombardment if Arias did not withdraw. Arias retreated and within days the Marine Corps was onshore, and within months the Marines and Navy were essentially running the Dominican Republic. The U.S. declared martial law but left most laws and institutions continue as normal, although U.S. officers operated in the cabinet and controls on media and speech were imposed.
The campaign of counterinsurgency they imposed in the east of the country was brutal by modern standards – after all, this was a U.S. military whose Lieber Code was essentially ratified into international humanitarian law by the Hague Regulations of 1907. Under these rules, the U.S. could kill POWs by firing squad for reprisals, or conduct summary executions of irregular forces. The occupying force could also imprison or relocate populations aiding guerrillas, or even take hostages. The Lieber Code and the associated counterinsurgency standards conceived of the relation between occupying force and population in a Hobbesian, even Schmittian sense - protego ergo obligo. Rather than overthrowing the old regime (as opposed to its potential or actual leading individuals), the U.S. would ultimately attempt to extend the longevity of the established political order by creating gendarmeries and other security institutions to give governments a more solid power base than militias and warlordism.
Another case of limited intervention in Latin America was the 1914 occupation of Veracruz, which was ostensibly to prevent the shipment of German arms to the forces of Victoriano Huerta during Mexico’s civil war. Of course, the arms were actually made in the U.S. and being transshipped through Germany to skirt the U.S. embargo, but the conflict does point out another common theme of limited interventions – the mitigation of extra-regional involvement by U.S. rivals within the Western Hemisphere, rather than the outright resolution of conflicts themselves. There were also, of course, many punitive interventions designed to influence the conduct or depose certain figures within Latin American regimes, or prevent certain policy changes that harmed U.S. interests. Also extremely common during the Banana Wars and associated conflicts was the use of U.S. military assets to protect U.S. citizens and property, and norm-enforcement and perceptions management tied to protecting the reputation of the U.S. to prevent its citizens and commercial interests from being imperiled at will. While doubtless a modern schema of intervention would be quite different, it is notable that the ultimate limiting factors in the objectives and scope of U.S. interventions were the limited scope of U.S. interests themselves.
Searching for Limiting Factors
All of these interventions have some important implications for when a limited intervention is an appropriate use of U.S. force, or likely to actually remain limited. Firstly, a voluntary intervention should generally occur under circumstances where the United States has impunity of power projection. Despite all the talk about drones and other capabilities giving the U.S. the ability to intervene with impunity, a U.S. warship in the early 20th century had enormous degrees of latitude to act against a minor state. When Caperton sailed his ships to Santo Domingo and threatened the city with bombardment, his ships were virtually immune to the force of Dominican arms. In other words, a war against a target which possesses the capability to employ an anti-access/area denial strategy, or otherwise strike and harm infrastructure or assets involved in power projection, cannot be guaranteed to remain limited and should generally be reserved for cases where vital interests at stake. Even the presence of a robust air defense network, such as Syria’s, should give the U.S. pause and force it to consider whether or not the limited objectives at stake are worth a massive aerial and naval campaign and the potential escalation of the conflict to involve Iranian covert counter-actions against the United States.
Another limiting consideration should be the degree to which an intervention requires the alteration of an existing order or its reconstruction. While Finel is correct to show that missions such as Afghanistan and Iraq show that the United States has the capability to conduct regime removal, regime removal must be considered in the context of what problems it leaves in its wake. Ultimately, even successful regime removal can still leave problems for U.S. policy too costly to justify the removal, and perhaps even worse than the regime itself as far as U.S. interests are concerned. As for the actual mitigation of humanitarian problems and the protection of civilians, as Adam Elkus has noted, the circumstances under which this can actually remain a limited endeavor are quite rare, precisely because cases where we can act without the high-end objective of destroying the rival political order and mitigating the negative consequences it previously produced are so rare:
There is actually a formula that already exists for using force to protect civilians in vulnerable situations and guaranteeing human security. The model, dubbed “first-generation peacekeeping” was an immediate product of the first generation United Nations. The purpose of peacekeeping was not to resolve the underlying conditions that created the conflict, or necessarily peace enforcement, but to keep a peace that already existed.
… if the peacekeeping paradigm was not a challenge to an order but a production of it, effective protection of civilians is eminently possible. Depoliticizing intervention could produce human security with a minimum of “mission creep” and without substantial risk to peacekeepers.
… it had one cardinal strength: it was rooted in the reinforcement of an existing political order rather than the creation of a new one. It also understood something that Alice Hills, a theorist of post-conflict policing, found in her latest research—security is produced by order, but security doesn’t necessarily produce lasting order.
An intervention which seeks to create security through the destruction and production of order, rather than the reproduction or consolidation of an existing order, is far less likely to remain limited, and therefore far less likely to have a benign effect on U.S. interests.
Ultimately, though, for genuinely limited intervention to occur, we need policymakers with a conservative outlook – not in the sense of their domestic political affiliations, but their international expectations and willingness to tolerate problems that arise in the pursuit of legitimate and higher-prioritized U.S. interests. The reason that U.S. policies premised on regime removal are often so pernicious is because the regime is a stand-in for a host of other things – not just the actual persona and policies pursued by the leadership being removed, but its nature (authoritarian, theocratic, “failed,” etc), its tools of governance (violence, intimidation, corruption, etc), the effect that rule has on the civilian population and outlook (chaos, poverty, anti-Western attitudes, etc), and the consequences for the region (destabilizing, antagonistic, etc). The problem is that intervention can make some or all of these problems worse rather than better, even if the specific policies of the regime just removed are gone. This demands a conception of interests that is prioritized and can be translated into concrete end state objectives. Otherwise, it will be too tempting to ameliorate the failures or excesses from an intervention with yet another mitigating attempt. The possibility for such a sequence of frustrated interventions is evident in U.S. foreign policy leading up to the overthrow of Saddam Hussein.
Opportunistic Mitigation and the Unraveling of Limits
The U.S. containment of Iraq during the 1990s is actually a strong example of how the inability to tolerate the mitigation of a problem resulted in a prolonged series of limited actions, ultimately culminating with frustration and resort to an extreme solution. The U.S. had already neutralized the threat of Iraq to its neighbors and ensured, for the most part, the non-functionality of the Iraqi WMD program. But the problems the limited interventions failed to solve, or the problems they created soon became intolerable to policymakers who eventually put regime change into U.S. policy. In other words, limited goals and expectations are useless if policymakers can’t treat the counteractions of the enemy with an extreme degree of tolerance. The first U.S. war against Iraq had the limited political aims of defending Saudi Arabia, liberating Kuwait, and reducing Saddam’s offensive capacity. However, it created the problem of an extant anti-U.S. state engaged in the bloody repression of its own people. The U.S., believing another limited intervention could improve the situation, employed a mix of economic, diplomatic, and military measures to mitigate these problems. These, in turn, ultimately required the mitigating invasion of 2003, and the problems that created invited the much longer counterinsurgency campaign to follow.
Ultimately, interventions tend to undermine their own constraints without policies and policymakers willing to reject the sort of opportunism that a model of more and more limited interventions implies. Precisely because the results are never final, policymakers confronted with an unexpected insurgency in the wake of the fall of Saddam asked Finel’s question, “can we make things better?” They decided they could. As pernicious of a question “can we fix it?” is, “can we make things better?” is often just as bad. If “can we fix it” leads us to pursue maximalist ends from the beginning, “can we make things better” still makes it extremely difficult to determine when the use of force is no longer a good idea.
When we ask whether or not U.S. action can “help,” the relevant question is not whether it helps Americans. Too often, the national interest component of interventions is couched in upholding a system of norms or improving American perceptions abroad, but neither of these are really viable objectives of military interventions. Norm-enforcement can be worthwhile outcome of military endeavor, but only if it is produced through the pursuit of a legitimate U.S. national interest. How do we know when a norm has been adequately enforced? How do we know when if the regime will be adequately punished? The enemy, through his counter-actions, could undermine these perceptions through the force of arms. If, say, Assad in Syria were to simply disperse his arms, sacrifice a proportion of his military forces to the U.S. and then resume the task of brutal counterinsurgency later, will we feel our norm has been adequately enforced, or that Assad has been adequately punished?
When U.S. interests in a country are low, the chances that mitigation of a problem will result in a worse problem or an amelioration not worth the expenditure of resources involved increases. Rather than tolerating the unfixed problems in the wake of an intervention, a more useful approach would be to tolerate the existence of insoluble problems before the U.S. even launches another intervention. Rather than asking “can we make things better?” we need to be asking “do we need to make things better?” and “can we tolerate what we can’t make better, or make worse?” All of these questions have to be considered in line with the potential cost on the high end, particularly because what goals will seem appropriate and what inefficiencies we will be willing to tolerate with low end estimates will seem much less appropriate and tolerable if casualties, costs, and consequences are higher than expected. Ultimately, though, the same kind of mindset that limits intervention goals effectively – the acceptance of insoluble problems – should also make us less willing to use intervention generally. The conditions which make limited and safe interventions realistic – the existence of a permissive environment, the reproduction of order rather than its destruction or production, and the presence of concrete or more immediate U.S. interests at stake – are not necessarily forthcoming, in cases where they are not, we should be asking whether or not such interventions are necessary before we ask whether they are possible.