Download PDF The most fateful American choice in the rise of ISIS was also the oldest one: the 2003 decision to invade Iraq, followed by the mismanagement of the occupation.

Barring some catastrophic policy blunder by the United States, the Islamic State, also known as ISIS or ISIL, will eventually be defeated. The US-led international coalition that has assembled to fight the most formidable terrorist organisation of modern times overmatches ISIS on every relevant dimension – manpower, lethality, financial resources, global reach. As such, the defeat of ISIS, at least in its current form, is only a matter of time. But the group’s defeat will not resolve all of the questions that have been raised by its emergence. Looking forward, US policymakers will have to decide what to do next in America’s ongoing ‘global war on terror’.1

It is also worth looking back to ask whether ISIS’s rise could have been averted in the first place. Was it inevitable that the defeat of ISIS’s predecessor, al-Qaeda in Iraq (AQI), would simply pave the way for the emergence of a new and more deadly terrorist organisation? Or might different US policy choices have forestalled ISIS’s rise, or at least prevented it from becoming so powerful and destructive? These questions are of more than academic interest. If ISIS’s rise was predestined, then perhaps the United States and its counter-terrorism partners are simply stuck in a Sisyphean cycle of endless conflict. If, alternatively, different choices might have led to a different outcome, then perhaps learning from this episode can help America avoid similar setbacks in the future.

This question is important, moreover, because even though ISIS’s defeat is inevitable, that defeat will come at an extremely high price. Among other things, ISIS’s rise has directly threatened the survival of Syria, Iraq and Libya, and at its height posed an ominous threat to Jordan and even Saudi Arabia.2 It has generated waves of foreign fighters that, upon their return home, have posed ongoing domestic security threats in a way that members of previous terrorist groups did not.3 It has inspired lone-wolf and copycat attacks that have inflicted significant costs in lives and proven devilishly difficult to forestall.4 It has served as an incubator for innovative techniques of recruitment, mobilisation and organisation that future terrorist groups may exploit.5 It has exacerbated what the UN human-rights chief has called ‘the worst man-made disaster since World War II’, the Syrian civil war, generating additional waves of refugees and thereby threatening the political unity and stability of Europe.6 It has resulted in enormous destruction, including of priceless cultural landmarks in the Middle East.7 It has imposed the serious cost of strategic distraction on the United States, dragging the Obama administration into yet another war in the Middle East.8 It has resulted in the loss of tens of thousands (if not more) of innocent lives in Iraq, Syria and far beyond.9 And given the extent to which the threat posed by ISIS figured in America’s 2016 presidential campaign, the group’s rise may have helped decide that election.10 Determining whether these severe costs – for the Middle East, neighbouring regions such as Europe, and even the United States – could have been avoided promises to be painful, but necessary.

ISIS’s defeat will come at an extremely high price

Answers to the question of whether different choices by US policymakers could have averted some of these downsides, or allowed the problem of ISIS to be managed at a lower cost, have so far been both impressionistic and polarised. The Obama administration and its defenders have argued that no plausible policy choices from 2009 onward could have produced a better outcome. Other observers have argued that the Obama administration missed multiple opportunities to derail ISIS’s rise, while still others have argued that the best and most obvious way to avoid ISIS’s emergence would have been not to invade Iraq in 2003.11 One thing that all these assessments have in common, however, is that they tend to be based more on assertion than on evidence or analysis, and they are often coloured with a partisan or ideological hue.

Here, we seek to provide a more systematic and balanced assessment of this question, by revisiting four key inflection points in US policy and the rise of ISIS: 1) the original decision to invade Iraq in 2003, with all the mismanagement of the occupation that followed; 2) the US political disengagement from and military drawdown in Iraq in 2010–11; 3) the decision not to intervene more robustly in the Syrian civil war between 2011 and 2013; and 4) the decision in late 2013–early 2014 not to strangle ISIS in its cradle by taking military action before it conquered much of western Iraq and swooped down upon Mosul. For each moment, we briefly recount what US policymakers did and why, and then consider what might have happened had they taken a different – but still plausible – course. In essence, this article is an exercise in counterfactual history – an attempt to examine various roads not taken, and to engage in informed speculation about whether those roads might have led to a better destination.

This analysis yields three principal conclusions. First and foremost, the rise of ISIS was indeed an avertable tragedy. Had US policymakers made different but nevertheless plausible choices at one or more of several key junctures, ISIS probably would not have emerged as the full-blown threat it ultimately became. Had the United States not invaded Iraq in 2003, had it not effectively disengaged from Iraqi politics in 2010 and failed to secure a modest stay-behind force in that country after 2011, or had it moved to strike ISIS militarily in the crucial months leading to the fall of Mosul in June 2014, ISIS might not have emerged at all – or, at a minimum, the military threat it posed and the strategic havoc it wrought would have been significantly reduced, and the United States and its partners would have been better placed to respond. To be clear, ISIS’s rise owed principally to deep historical forces, as well as to complex local and regional dynamics over which US policymakers had little direct control. But the United States was hardly powerless to constrain or mitigate the effects of that rise, and there were multiple occasions on which US policymakers might have done so.

Secondly, however, virtually all of the counterfactuals explored here are less clear-cut than they might initially appear, either because different policies would have produced only moderately better results, or because different policies would have introduced other costs and difficulties that somewhat cloud the ultimate cost–benefit analysis. We find, for instance, that limited intervention in Syria in 2011–13 might have had benefits, but it probably would not have shifted the course of the conflict so fundamentally as to head off ISIS’s rise. Likewise, not invading Iraq in 2003 would have left the United States saddled with the costs of continuing to contain that country, whereas striking ISIS militarily in late 2013 or early 2014 might have weakened that organisation militarily while exacerbating the political conditions that were fuelling its rise. Intervening more heavily in Iraqi politics in 2010 in order to bring about a less sectarian government than that which ultimately emerged, and leaving a stay-behind force in Iraq after 2011, represent a fairly compelling counterfactual in the sense that such policies could have had numerous constructive effects. But even here, choosing a different path from the one actually taken would have meant courting non-trivial costs, liabilities, uncertainties and limitations.

Thirdly, and in light of the foregoing two points, the debate over why ISIS arose, and what US policymakers might have done to forestall that eventuality, needs to shift. In particular, it needs to move away from polemical and polarised assessments focused on assigning blame, and toward more granular, balanced analysis based on a fairer-minded view of what went wrong, and what the possibilities for better outcomes truly were. As we argue, there were numerous decisions that contributed to ISIS’s rise, and both the George W. Bush and Barack Obama administrations arguably missed opportunities to head off or mitigate that rise. There were also numerous ways – under both administrations – in which different policies might simply have forced the United States to contend with different problems. More broadly, the proper purpose of counterfactual analysis is not to indict policymakers before the court of history, but to explore feasible alternative pathways and thereby enable better decision-making in the future. And that requires putting partisan point-scoring aside and grappling with what happened – and what might have happened – in its full complexity.

Counterfactuals and the rise of ISIS

ISIS exploded into the US and international consciousness in June 2014, when its fighters captured Mosul, Iraq’s second-largest city, before approaching the outskirts of Baghdad and Erbil. In the previous months, dating back to 2013, it had already captured significant swathes of Iraq’s Anbar and Nineva provinces, in addition to large areas of northern and eastern Syria. By this point, ISIS could – and did – credibly claim to have established an Islamist caliphate in the heart of the Middle East. It had exerted control over perhaps 90,000 square kilometres of territory and millions of individuals, and had begun to attract tens of thousands of foreign fighters and to create the necessary infrastructure to carry out significant external terrorist attacks.12

By mid-2014, in other words, ISIS had become the strongest, best-resourced and most ideologically potent terrorist quasi-state of the post-9/11 era. Many of the costs associated with its rise were unavoidable from that point on; any struggle to defeat ISIS was bound to be long and difficult. Perhaps Washington and its coalition partners could have marginally accelerated that struggle by taking a less incremental approach to counter-ISIS military operations from June 2014 onward, but by then ISIS’s rise was a fact to be dealt with rather than an eventuality to be averted.13 Our counterfactual analysis therefore addresses decisions taken prior to the fall of Mosul, the point at which ISIS reached the pinnacle of its power.

Counterfactual analysis has a bad name among many academic historians, who associate it with parlour games and the ‘alternative histories’ favoured by popularisers.14 Yet political scientists, and some historians, have long understood that counterfactual analysis is integral to causal claims. As James Fearon has argued, every causal claim is simultaneously a counterfactual claim: proving that X caused Y is also proving that not-X would produce not-Y.15 Policymakers also continually engage in counterfactual reasoning, whether they realise it or not. If a policymaker claims that some initiative will produce the best possible outcome, they are, implicitly or explicitly, making the counterfactual claim that the alternative choices will fare worse. Moreover, counterfactual arguments have been central to public and political debates over key policy issues for decades. The idea that the Second World War might have been avoided had the British and French not appeased Hitler at Munich, that North Korea might not have invaded South Korea had Dean Acheson not placed the latter country outside the US defence perimeter, or that the history of the twentieth century might have looked very different had the United States joined the League of Nations after the First World War all represent counterfactual arguments that have figured prominently in such debates. For better or worse, then, counterfactual analysis is integral to conducting, and thinking about, foreign policy and international relations.

For counterfactuals to be useful, however, analysts should follow four basic rules of thumb. Firstly, the most powerful counterfactuals are those that are at least fairly proximate to the event one is trying to explain or alter. It is true but not particularly helpful to point out that, if not for the Sykes–Picot Agreement of 1917, or the dispute over the Prophet Muhammad’s succession in the seventh century, ISIS might never have arisen. Far better to examine moments at which prudent policymakers could reasonably have looked down the road a bit and anticipated something like what ultimately happened. Moreover, as demonstrated by even the relatively proximate counterfactuals assessed here, the further back in time one goes, the more any single policy change produces increasingly diverse downstream changes in the subsequent course of events – all of which makes for counterfactuals that are enormously interesting, but also analytically messy.

Secondly, the counterfactual in question must have been plausible at the time, given what policymakers knew and were prepared to contemplate. It is fair to ask what might have happened if the United States had left 10,000 troops in Iraq after 2011, for instance; it is not fair to ask what might have happened had it left 150,000 troops in Iraq. It is useful to ask what might have happened had the United States armed the Syrian opposition more aggressively in 2012, or undertaken limited airstrikes against the Assad regime; it is not useful to ask what might have happened had Obama sent in the 82nd and 101st Airborne Divisions to forcibly overthrow Assad and secure Damascus. It is not that the go-huge options would have violated the laws of physics, but that they were so far beyond the boundary of what was politically possible that no serious participant in the policy debate proposed them. Counterfactuals provide a way of constructively assessing what might have been only if they deal with reasonable, or at least conceivable, roads not taken. Thus, every counterfactual we analyse here was proposed, at the time, by serious policy professionals either inside or outside (and, often, both) the sitting administration, thereby passing the ‘reasonable threshold’ test.

Thirdly, and for largely the same reasons, the most useful counterfactuals are those that change as few critical variables at the initial decision point as possible. Political scientist Jack Levy, following Philip Tetlock and Aaron Belkin, has referred to this as the ‘minimal-rewrite rule’.16 This is because it is reasonable to ask if policymakers might have done one thing, or perhaps a very small number of things, differently or better in a given context; it is less reasonable, given the inherent challenges of policymaking, to ask if they might have done ten things differently or better in that same context. John Lewis Gaddis illustrates this point, as well as the previous rule of thumb, in a discussion of counterfactuals and the road to Pearl Harbor. He writes:

It’s perfectly appropriate to ask what might have happened had the United States not imposed the oil embargo on Japan after the takeover of French Indochina. It’s not appropriate to ask what might have happened if the Roosevelt administration had combined that decision with an offer to transport Free French forces to that part of the world, together with a massive buildup of American forces in the Philippines, together with an effort to settle the Soviet Union’s war with Nazi Germany so that Stalin could shift his forces east and also intimidate the Japanese.17

Finally, and notwithstanding the foregoing, good counterfactual analysis is alert to the fact that changing a given policy decision or historical event invariably results in multiple changes to the subsequent course of history. Had Lee Harvey Oswald’s bullet missed John F. Kennedy, for instance, there might have been no US escalation in Vietnam in 1964–65; there might also have been no landmark civil-rights or voting-rights legislation during that same period.18 Honest counterfactuals must therefore account for both good and bad variations.

Counterfactual 1: No invasion, or a better invasion, of Iraq

Critics have long contended that the invasion and botched occupation of Iraq in 2003 was the principal ‘deep cause’ leading to the rise of ISIS.19 By toppling Saddam Hussein, and then failing promptly to establish a viable post-Saddam political order, the United States, according to this argument, created a political vacuum in Iraq. That vacuum, in turn, fostered catastrophic internal violence, unleashed long-repressed sectarian tensions, and stoked the sense of Sunni displacement and grievance that fuelled the rise of AQI and later ISIS. In Richard Engel’s pithy formulation: ‘No Iraq war, no ISIS.’20

According to the best currently available information and analysis, the United States invaded Iraq in 2003 for a relatively simple reason: the 9/11 attacks changed the Bush administration’s risk calculus about the relative dangers of action versus inaction. Prior to 9/11, there was a bipartisan consensus that Iraq was a serious national-security problem, as a result of Saddam’s continuing intransigence and suspected stockpiles of weapons of mass destruction; that this problem would only be resolved by removing Saddam from power; that there was no plausible way of removing Saddam short of major military intervention, which was likely to be difficult and costly; and that there were higher-priority concerns than Iraq on the national agenda. Therefore, it was agreed that the Iraq problem could and should simply be kicked down the road. The 9/11 attacks convinced Bush and his senior team that the dangers of kicking such problems down the road could in fact be appalling, and so they determined to confront Iraq as soon as the initial invasion of Afghanistan had apparently succeeded.21 When this confrontation did not produce the administration’s desired results – conclusive proof of Saddam’s disarmament, perhaps accompanied by his and his family’s departure from Iraq – the United States went to war.

Were there plausible alternative courses the United States might have taken in 2003? Answering this question is made somewhat more difficult by the massive intelligence failure that convinced senior policymakers that Iraq was a greater and more immediate danger than it actually was.22 Yet even so, there were alternatives to an invasion of Iraq that were mooted by serious observers at the time.

The United States could have settled for using its enhanced diplomatic and military leverage after 9/11 and the invasion of Afghanistan to persuade Saddam (and the international community) to accept a revitalised, long-term sanctions and inspections regime. It might have re-evaluated its assumptions about Saddam and his weapons programmes when inspectors who had been readmitted to Iraq as a result of US military pressure failed to uncover evidence of the suspected weapons stockpiles. Alternatively, the Bush administration could have chosen to emphasise different aspects of the counter-terrorism problem set by placing greater stress on clearing out ungoverned spaces in countries such as Somalia, Yemen or Pakistan, or going all-in on finding a diplomatic solution to the Israeli–Palestinian conflict. Or it might have chosen to confront a different rogue state with extensive ties to terrorism and access to weapons of mass destruction – Iran, for instance, or North Korea. The plausible alternatives, in this case, were numerous, but the point is that Bush’s decision clearly could have gone another way.23

The Iraqi political order probably would not have collapsed

Had the president chosen not to invade Iraq, the subsequent course of events would have been vastly different. In the near term, the Iraqi political order probably would not have collapsed and created a void that non-state or quasi-state actors could fill. The Sunni–Shia cleavage that has made Iraq so difficult to govern still would have been present, but without the violence, political chaos and Sunni marginalisation of the post-invasion period, that cleavage would have remained in a less combustible state, and terrorist groups such as AQI and ISIS would not have found such fertile ground for recruiting. The broader wave of sectarianism and Shia ascendancy that convulsed the entire region after 2003, thereby setting the stage for the Sunni backlash of which ISIS was a part, would presumably not have occurred, or at least would have been more muted.24 Similarly, without the toppling of Saddam’s regime, displaced Ba’athist military and intelligence officials would not have been available to form the mid-level cadre leadership of both AQI and later ISIS. Iraq still would have been awash with weapons, but those weapons would have been largely in the hands of organised units answering to the Iraqi state.

Clearly, then, insofar as the invasion of Iraq created the conditions that ISIS would later exploit, not invading, and not creating those conditions, would have inhibited ISIS’s rise. It is possible, of course, that Saddam might have been toppled by an internal rival or an Arab Spring-like uprising, or simply died of natural causes, thus catapulting the country into a political crisis absent an invasion. But there would have been a long and winding path from that crisis to the ISIS threat in 2014, whereas the direct effects of the Iraq invasion and post-invasion difficulties are obvious and relatively easy to measure. In this sense, the ‘no Iraq war, no ISIS’ proposition is almost undoubtedly true.25

The CPA could have announced it would rebuild the security forces

What gives this counterfactual some complexity, however, is that it is nonetheless unclear how much better off America would have been in this scenario. The United States still would have faced all the problems associated with Iraq prior to March 2003: an aggressive rogue state in a crucial part of the world; a sanctions and inspections regime that had proven difficult to keep in place over time; the military, political, economic and humanitarian costs of containing Saddam’s regime; and so on. Washington would have had to keep American forces in Saudi Arabia to keep Saddam bottled up – a presence that had served as one of al-Qaeda’s primary grievances and most effective recruiting tools.26 Moreover, forgoing the invasion would have deprived the United States of one the great, ironic, benefits of the Iraq War – the opportunity to kill so many of the terrorist fighters that flocked to Iraq after the US invasion, but then were decimated by the Anbar Awakening of Sunni tribes and the nearly concurrent surge of American forces and reinvigorated counter-insurgency strategy between 2007 and 2009.27 And, of course, had the United States pursued another risky policy – confronting Iran or North Korea, or placing more pressure on Pakistan in an effort to mop up the remnants of al-Qaeda and the Taliban – in place of invading Iraq, it would have had to deal with all of the fallout from that initiative as well. It seems likely that these difficulties would not have equalled or exceeded the problems engendered by the Iraq War itself, but they do render the net assessment somewhat less clear-cut than it might initially seem.

What about a separate but related counterfactual – that of managing the invasion and particularly the occupation of Iraq in a more competent way? Here, there are countless ‘what ifs’ that loom large today.

The United States could have used airstrikes to target Abu Musab al-Zarqawi, the leader of AQI, in 2002; without Zarqawi’s ruthless and charismatic leadership, AQI might not have emerged as the lethal force it eventually became.28 The United States could have delayed the invasion by a few weeks to give the Erdogan government in Turkey more time to secure permission for the northern invasion route, which would have brought the 4th Infantry Division into Baghdad more promptly, and presumably after destroying many of the Fedayeen and other units in Anbar province that formed the initial insurgency. The United States could have gone in heavier with more troops, or at least more permissive rules of engagement directed against the looters, which would have provided for greater security in the critical early days of the occupation. The Coalition Provisional Authority (CPA) could have announced from the outset that it would rebuild and continue to pay the former Iraqi security forces, thereby avoiding the alienation of the crucial ‘swing group’ in post-Saddam Iraq and preventing a vast number of trained military personnel from joining the nascent insurgency. The CPA could have opted for a more generous de-Ba’athification programme that might have given more Sunnis a stake in the new order. The coalition could have chosen wiser detention policies that would have made it harder for AQI and ISIS to incubate inside the prisons. All of these policies were suggested, or were at least within the realm of plausibility, at the time. Had all or most of these roads been taken, the occupation might well have proceeded more smoothly, and the conditions that eventually produced ISIS might have been averted or at least significantly mitigated.29

There are, however, two crucial caveats here. Firstly, at least some of these decisions would have unleashed other negative consequences that could have produced the violent sectarianism that AQI and ISIS exploited. Preserving the old Iraqi military and pursuing a less stringent de-Ba’athification programme, for instance, might have precipitated a Shia revolt and thereby sparked civil war from a different direction. A heavier force presence, combined with harsher measures against looters and other purveyors of disorder, might have lessened the post-invasion chaos but also inflamed Iraqi nationalist sensitivities more quickly. Secondly, because the invasion, and particularly the occupation, of Iraq was such an inherently fraught proposition, averting the conditions that produced ISIS might well have required handling not just one but a number of these issues in the way dictated by perfect hindsight. It might not have been enough to use more troops, for example; the United States might also have had to pursue a more enlightened de-Ba’athification policy, a smarter approach to the Iraqi military, and so on. And as discussed previously, the more changes in initial conditions that one requires to make a counterfactual work, the less plausible the counterfactual becomes as a whole. In view of all this, it still seems highly probable, as Daniel Byman has argued, that the occupation could have been handled in ways that increased the chances for success and decreased the likelihood of an ISIS-like force later emerging. But it seems improbable that the United States could have invaded Iraq without seriously risking at least some of the dynamics ISIS later exploited.30

There is at least somewhat greater complexity here, therefore, than is sometimes assumed. Adding to the complexity is that Iraq was actually on the upswing both politically and militarily by 2007–08, and that ISIS’s rise thus occurred only after those positive trends had been reversed by later events.

Counterfactual 2: Less disengagement from Iraqi politics and security in 2010–11

If entering Iraq in 2003 constitutes one key inflection point in the story of US policy and the rise of ISIS, leaving the country in 2010–11 constitutes another. In retrospect, two key aspects of US policy during this period loom large in debates about whether the rise of ISIS was inevitable. The first was the US approach to government formation after the 2010 Iraqi elections produced a political stalemate. The US strategy during this period can best be described as an arms-length approach, one that consisted of taking no position on whether Nuri al-Maliki or his primary political rival, Ayad Allawi, should be given the first chance to form a government after Allawi’s party won the most seats in the parliamentary elections; quickly acquiescing in a second term for Maliki despite widespread concerns about his sectarian impulses and the fact that he had failed to win an electoral majority; and declining to fully apply US leverage to force Maliki to accept a meaningful power-sharing arrangement with Allawi, or make other concessions that might have resulted in a more pluralistic, less sectarian Iraqi government. The second fateful aspect of US policy was the failure to secure a status-of-forces agreement (SOFA) with the Iraqi government during 2011, which led to the withdrawal of virtually all US troops from the country at year’s end.

Together, these developments constituted a significant attenuation of US efforts to shape Iraq’s politico-military trajectory, coming just as events were taking a dark turn within that country. Iraq had been on an unmistakably positive slope in politico-military terms from 2007 through 2010. AQI had been brought to the verge of military defeat by the US surge and the concurrent Anbar Awakening, which inflicted devastating blows on an organisation that had seemed ascendant in the years immediately following the US invasion.31 Overall violence declined sharply, sectarian tensions receded (though hardly disappeared), and a more democratic and inclusive political process slowly began to take hold. After 2010–11, however, the pendulum swung back in a more ominous direction. The dominant trends were increasing political polarisation, hyper-sectarian governance, increasing violence and terrorism, and a dramatic erosion of the capabilities and professionalism of the Iraqi Security Forces (ISF).32 All of these trends, in turn, created the context in which ISIS exploded onto the scene.

Critics have subsequently alleged that different US policy decisions, such as prodding Maliki to step aside or to accept a power-sharing arrangement with Allawi in 2010, or leaving a residual US force of perhaps 10,000–20,000 troops after 2011, would have averted or mitigated these problems, and prevented ISIS’s emergence.33 The administration and its defenders, on the other hand, have argued that better choices were not available – because there was no plausible political alternative to Maliki, because the United States had insufficient leverage to shape Iraqi politics in more constructive ways, or because the United States could not obtain a SOFA that provided adequate legal protections for US troops – and that Iraq’s problems were so deeply rooted that US policy changes would not meaningfully have altered the course of events.34 Thus, there is still significant debate about whether there were plausible alternatives available that might have resulted in different outcomes. Because the 2010–11 episode has given rise to the most hotly contested counterfactuals regarding US policy and the rise of ISIS, we explore these questions at some length here.

The Iraqi government was insisting on a poison pill

To be fair to the Obama administration, the decisions made in 2010–11 had a credible rationale. The decision in 2010 not to exert greater pressure on Maliki, and ultimately to support him for a second term as prime minister, was not taken lightly, coming after it became clear that Allawi, whose party had won the most seats, would encounter great difficulty in attracting the broader support necessary to form a government of his own, and that other likely candidates for prime minister had significant liabilities of their own. It was also taken in the knowledge that any meaningful power-sharing arrangement would be difficult and painstaking to broker given the intricacies of Iraqi politics and the meddling of outside actors such as Iran; and that giving Maliki – who enjoyed Iranian support – a relatively free hand was thus the best near-term option for ending Iraq’s political paralysis. Additionally, the administration did seek a post-2011 military presence in Iraq (although Obama himself does not appear to have been particularly invested in that effort), and desisted only once it became clear that the Iraqi government was insisting on what administration lawyers unanimously considered to be a poison pill: a refusal to offer US troops immunity from prosecution backed up by an act of parliament. Instead, immunity was to be guaranteed only by executive agreement and an exchange of diplomatic notes.35

Moreover, top US officials seem to have believed, despite fairly stark warnings from State and Defense department officials in Iraq, that the country was more politically and militarily stable than later became clear. Indeed, it is hard to imagine that Obama and vice-president Joe Biden would have repeatedly touted the drawing down of US involvement in Iraq as a great success for the administration – as they repeatedly did during 2011 and 2012 – had they not been fairly confident about the country’s prospects.36 Finally, all of these decisions were influenced by Obama’s guiding belief that he had been elected, in large part, to wind down rather than extend America’s deep entanglement in Middle Eastern politics, and that the best way to do so was to force the Iraqis to take greater responsibility for their own political and military fortunes.37

But might the administration nonetheless have done things differently, and would those things have altered Iraq’s trajectory sufficiently to prevent or significantly mitigate ISIS’s rise? In retrospect, the idea that the United States could have induced Maliki to step aside, to share real power with Allawi, or simply to accept some other political arrangement than the one that eventually emerged – in which power was shared only nominally and Maliki was not simply permitted but incentivised to govern on an increasingly sectarian agenda – does seem somewhat more plausible than Obama and administration defenders later claimed. After all, the administration did recognise Maliki’s liabilities as a leader in 2010, as shown by the fact that numerous officials explored the idea of – or simply advocated – supporting an alternative candidate as prime minister. As vice-president Biden, himself a supporter of sticking with Maliki, acknowledged at the time, ‘Maliki hates the goddamn Sunnis.’38 Emma Sky, an adviser to US and coalition forces, later recounted that conversations with other Iraqi political leaders revealed that ‘nobody wanted a second Maliki premiership’.39

At this point, moreover, the United States still possessed significant leverage in Iraqi politics. It had influence with nearly all the major Iraqi political factions, because the roughly 50,000 US troops still in the country represented the primary guarantee of its security, and the primary hedge against a resurgence of the out-of-control sectarian violence that had nearly torn Iraq apart in earlier years.40 The United States also had significant leverage vis-à-vis Maliki, who was in a weakened political position after coming in second in the 2010 elections, and was thus vulnerable to the threat that Washington might back another candidate for prime minister or simply attenuate its support for a Maliki-led government. Efforts to shape Iraqi politics in 2010 ‘would be extremely difficult’, noted US embassy official Ali Khedery, one of the American advisers most opposed to Maliki, ‘but with 50,000 troops still on the ground, the United States remained a powerful player’.41 Indeed, the United States had used its military presence and the resulting influence in Iraq to good political effect between 2007 and 2009, to promote – gradually and with great difficulty – moderating trends in Iraqi politics.42 More broadly, in previous conflicts, such as the counter-insurgency in El Salvador in the 1980s, the United States had gradually and laboriously succeeded in translating lesser commitments of resources into a meaningful ability to shape a partner country’s political trajectory.43

In these circumstances, the United States might plausibly have invoked its still-considerable leverage in several ways. It might have insisted that Allawi be given first chance to form a coalition, perhaps on the understanding that Allawi himself would not serve as prime minister but would essentially act as kingmaker for a government that would be acceptable to Iran but nonetheless broader and more inclusive than a Maliki-led government.44 It might subsequently have withheld backing for Maliki – by threatening to break with him politically, to support another candidate for prime minister, or even to curtail US economic and security assistance to Iraq – unless Maliki made significant concessions to the Sunni political blocs and formed a government that was authentically inclusive. At the very least, the United States might have made a greater effort to avert the situation that ultimately materialised, in which Iranian influence predominated and Maliki formed a coalition via an alliance with Moqtada al-Sadr, a fiery Shia politician who was prone to stoking sectarian impulses and who stridently opposed a longer-term US presence in Iraq. US officials clearly understood the liabilities of this arrangement at the time. As then-deputy secretary of state James Steinberg commented, ‘We need to be very careful about the Sadrists and we need commitments from Maliki that he will contain them.’45 Had the United States been willing to make full use of the influence at its disposal, it might thus have had a reasonable shot – indeed, several reasonable shots – at achieving more advantageous political arrangements and reducing the autocratic, sectarian tendencies that quickly re-emerged after 2010. As it was, one close observer of Iraqi politics concluded that US policy signalled that ‘it would not continue to enforce the new, democratic rules of the road’, a signal that Maliki understood and would exploit, with disastrous effect, in the subsequent years.46

To be clear, this course of action would have entailed disadvantages, costs and uncertainties. It would have required deep, invasive meddling in Iraqi politics, which would have contradicted the drive to push Iraq to take greater responsibility for its own security and political outcomes. In the near term, this approach might have produced greater political instability, at a time when the administration was increasingly focused on creating maximum near-term stability to allow for the continued drawdown of US troops. US efforts to exert pressure on Maliki might have been contested by Sadr or outside actors such as Iran. Such efforts might simply have failed, because Iraq faced no existential crisis in 2010 (in contrast to 2014), and most of Iraq’s political leaders therefore felt safe enough to engage in partisan jockeying. Finally, had the United States tried and failed to render Maliki more pliable, it might have precipitated a breakdown of the diplomatic relationship, similar to what had happened when the United States unsuccessfully tried to find an alternative to Afghanistan’s Hamid Karzai during and after that country’s contested elections in 2009.47

The US still possessed significant influence

For all these reasons, the desirability of this counterfactual is perhaps less obvious than some critics of the Obama administration’s policy have argued.48 Nevertheless, given how much the United States had invested up to this point in creating a pluralistic, non-sectarian Iraq; given that the liabilities of essentially allowing Maliki a free hand were widely understood both within and outside the administration; and given that the United States still possessed significant political influence in 2010, one might still argue that a more involved approach to Iraqi politics, one that insisted on different leadership or simply stricter conditionality in US dealings with Maliki at a time when he was politically exposed, could have mitigated the political conditions that would ultimately give rise to ISIS.49 At the very least, in light of the costs of the policy actually pursued, it is hardly unreasonable to argue that the administration could and should have tried harder to find another course.

The US inability to achieve a better governing outcome in 2010 helped set the stage for the second failure: the unsuccessful negotiations for a stay-behind force in Iraq after 2011. The Obama administration subsequently insisted that no stay-behind force was possible due to the diplomatic stalemate over legal protections for US troops. Yet as the journalist Dexter Filkins has written, the leaders of Iraq’s political parties ‘had privately told American commanders that they wanted several thousand military personnel to remain, to train Iraqi forces and to help track down insurgents’.50 Maliki reportedly told Iraqi political leaders that he might accept a force as large as 20,000 US troops.51 This being the case, it is hardly unreasonable to believe that different negotiating tactics might have yielded a different result.52

President Obama could, for instance, have kept up with the kind of leader-to-leader diplomacy that had characterised US relations with Maliki under Bush, rather than downgrading the relationship – symbolically, if nothing else – by handing it off to Biden. The president could have personally engaged Maliki earlier, when the negotiations stalemated in 2011, rather than hanging back from the process as it ground to a halt. The Obama team also could have made a stay-behind force more attractive to the Iraqis. As it was, the team was very slow to settle on a realistic post-2011 force posture to propose as its opening bid in the negotiations, deferring the start of the effort until well after the US midterm elections in November 2010, and continuing to debate well into 2011 whether a stay-behind force was even desirable.53 With respect to the size of the force, the Pentagon initially proposed numbers in the range of 20,000 troops, and the White House was initially willing to seek a force of some 10,000 troops – a contingent sizeable enough to contribute significantly to Iraqi security. The White House soon began to retreat from that position, however, and subsequently insisted on extremely low numbers – ultimately, around 3,500 troops – that undermined the security benefit to the Iraqi government without materially lowering the political cost of permitting them to stay.54

Additionally, when General David Petraeus and Ambassador Ryan Crocker, a country team that worked exceptionally well together, cycled out in 2008, Obama replaced them with a team that did not work well together, comprising General Raymond Odierno and Ambassador Christopher Hill, and the political team in particular quickly lost influence with its counterparts.55 Moreover, the administration’s decision to back Maliki unconditionally after the 2010 parliamentary elections had the unintended secondary effect of making the most extreme Shia wing, led by Sadr, a veto player inside Maliki’s governing coalition. US negotiators thus found themselves trying to strike a bargain with a coalition that included the one faction that was unalterably opposed to a US presence. Finally, the president apparently refused to consider creative workarounds to the legal-protections issue, such as those involved in the 2008 Status of Forces Agreement, or to accept the higher risk that came with having those protections guaranteed only by diplomatic agreement rather than by an act of Iraqi parliament.56 Had the administration altered some or all of these negotiating tactics, it is altogether plausible that an agreement might have been reached, and a stay-behind force of 10,000–20,000 US troops might have been possible.

Washington lost its eyes and ears

Had such an agreement been reached, what good might such a stay-behind force have done? President Obama later argued that keeping US troops in Iraq would have achieved nothing (which makes one wonder why he pursued this approach in the first place); critics have contended that it would have significantly improved both the political and military situation.57 Our view is somewhere in between, but closer to the consequential end of the spectrum. We think that a plausible stay-behind force of 10,000–20,000 troops could have altered ISIS’s rise substantially, or at a minimum could have positioned the United States to respond more effectively at an earlier point, by providing seven key benefits.

Firstly, such a force would have provided far greater situational awareness inside Iraq. When US forces were reduced to the bare minimum needed to protect the US embassy, Washington lost its eyes and ears throughout multiple levels of the Iraqi government and security establishment, as well as the military footprint within which other intelligence agencies can more readily operate. The Obama administration did not go blind, of course, retaining a robust diplomatic presence in Iraq (although that presence was largely confined to Baghdad, another result of not having a stay-behind force). But it did lose both breadth and granularity of perspective that had helped policymakers in the previous decade, with respect to both emerging threats such as ISIS and the underappreciated deterioration of the ISF. Maintaining that perspective might have provided more advanced warning, so that policymakers would not have been so surprised in 2014, when a supposedly ‘jayvee’ (junior varsity) group made short work of the ISF.58

Secondly, such a force would have enabled a more robust training and advisory mission, leading to a less fragile ISF. Indeed, the principal line of effort for US forces from 2012 onward probably would have involved training, advising and assisting the ISF, in an effort to perpetuate their professionalism and combat effectiveness.59 Of course, by the end of 2011, the United States had already invested great time, effort and money in building up the ISF, and a few more years at a reduced level of effort likely would not have been a game-changer. But a similarly sized presence from 2014 onward did significantly improve Iraqi capabilities at the margins, enabling Iraqi forces to more effectively defend against ISIS attacks and to successfully retake the offensive. It also helped re-instil a modicum of the professionalism and morale that had been decimated by Maliki’s politicisation of the force.60 At the very least, then, a similar effort from 2011 onward probably would have mitigated – if far from fully prevented – the erosion of ISF effectiveness leading up to 2014; and it would, at a bare minimum, have provided greater insight into the state of Iraqi forces.61

Thirdly, a stay-behind force would have enabled greater engagement with the elite Iraqi Counter-Terrorism Service (CTS). CTS, the most capable of Iraq’s special-operations forces, was particularly compromised during the 2012–14 period. CTS was a credible counter-terrorism partner to US forces by 2010–11, but in the ensuing years its military effectiveness eroded as Maliki used that force as his own praetorian guard. More extensive US engagement with CTS would have helped sustain its professionalism and effectiveness, and (as discussed subsequently) allowed more aggressive counter-terrorism operations against AQI and ISIS as they began to re-emerge in Anbar and Nineva. Indeed, the fact that CTS has once again become an effective counter-terrorism and counter-ISIS force since US re-engagement in 2014 makes it seem all the more likely that continued US engagement after 2011 would have had a constructive effect.62

Fourthly, a stay-behind force would have allowed improved logistics and sustainment, thereby bolstering ISF effectiveness. US logistics are unrivalled, and the same capacity needed to keep an ongoing military presence from 2012 onwards would have been available to help the Iraqi forces respond more nimbly to the ISIS invasion. Even if US forces were not providing logistics and sustainment on a regular basis, they would have been able to provide such support in extremis in late 2013 and early 2014. As it was, a near-total breakdown of Iraqi logistics resulted in key units having insufficient supplies such as water, food and ammunition, badly depressing morale and almost ensuring that such units would simply collapse when confronted by ISIS.63

Maliki antagonised Iraqi Sunnis

Fifthly, an ongoing, visible US presence might have shifted the psychology of Iraqi politics in helpful ways. One of the crucial contributions of the 2007 Iraq surge was the way the redoubled American commitment adjusted the psychology of each of the Iraqi political factions, by creating a climate of security and reassurance in which political compromise could occur. According to Stephen Biddle, US forces during the late surge period operated essentially as peacekeepers, keeping warring factions separate and providing a guarantor for deals that were otherwise not self-reinforcing.64 As thoughtful observers such as Kenneth Pollack have noted, the US departure at the end of 2011 had the reverse effect: it removed any sense that there was a relatively neutral outside actor to mediate disputes and support the nascent democratic rules of the road. It is difficult to say with any certainty whether a modest residual presence might have avoided this backsliding, but it is not implausible that such a presence might have continued to act somewhat like a shock absorber, creating a sense of security to mitigate zero-sum politics.65 It might also have taken the hard edge off of Maliki’s sectarianism, by reassuring him that he would not be violently deposed by rivals, and thereby reducing his incentives to revert to winner-take-all politics – just as the American commitment had done during 2007–08.

Sixthly, the United States might have enjoyed marginally greater leverage to restrain Maliki from politicising the ISF, reneging on commitments to Sunni tribes, and steering Iraqi politics in a sectarian and authoritarian direction. Virtually all American observers agree that Maliki’s increasingly sectarian and dictatorial policies from 2012 onward had the effect of antagonising Iraqi Sunnis and crippling the ISF. (ISIS took double advantage of this, by expanding operations in an increasingly permissive environment within Anbar province, and by directly challenging and defeating sizeable ISF units that were combat ineffective.66) It also seems clear that the United States lost nearly all leverage to restrain this behaviour by withdrawing in late 2011 – indeed, Maliki began moving harshly against his political opponents just as the US withdrawal concluded.67 Although the modest stay-behind force that would have been plausible in 2012 likely would not have been large enough to enforce the kinds of strict conditionality that would have been necessary to change Maliki’s behaviour fundamentally, it might have provided some greater degree of leverage to push back on the most damaging policies, such as the repression of Sunni protest movements and persecution of Sunni politicians.68 ‘We used to restrain Maliki all the time’, the US deputy commander in Iraq, Lieutenant-General Michael Barbero, remarked in 2014; a stay-behind force could have preserved at least some of this effect.69

Finally, a crucial benefit of a residual presence would have been the ability to conduct ongoing counter-terrorism operations against ISIS when it was still in its formative stage. Had US troops remained in Iraq in 2012 and after, they likely would have combined training and advisory duties with direct-action missions against radical extremists (as was the case with a roughly similar-sized US presence in Afghanistan after 2001–02). These operations could have further suppressed an already decimated AQI, and might have had an effect in attriting ISIS units when they were in the formative stage. At the very least, ISIS would have been constrained from the increasingly brazen activities and operations it undertook from late 2013 onward, when it was operating openly and in fairly large formations. Moreover, the United States would have possessed significantly greater insight into its tactics and operations.70 Across a variety of dimensions, then, a modest and plausible stay-behind force could have had a constructive effect.

In fairness to the Obama administration, there also would have been limitations and downsides to this approach. A more constructive approach to Iraq would not necessarily have improved the situation in Syria, where ISIS truly took hold between 2011 and 2013, and so the United States and its partners might still have confronted a more geographically confined, militarily and ideologically weaker version of the organisation – something along the lines of an ‘ISIS lite’. Even within Iraq, it is possible that some of the benefits discussed previously would not have materialised, or materialised only to a relatively small degree, and that the overall effect would therefore have been too small to constrain ISIS’s rise. Some scholars have argued, for instance, that a stay-behind force of perhaps 10,000 personnel was just too small to achieve much in the way of augmenting ISF capabilities, and that what was needed was not additional trainers but a more robust peacekeeping force of considerably larger size.71 Similarly, it took the emergence of ISIS as an existential threat to the Iraqi state in 2014 to compel what very limited political accommodation has occurred since then; a modest US residual presence might not have been enough to dissuade Maliki and other Iraqi politicians from proceeding down their disastrous sectarian course.

A residual force might have been targeted by Shia militias

Moreover, a residual US force might have been targeted by Shia militias who viewed Washington as a competitor for power in Baghdad, resulting in American casualties and forcing the United States either to devote increased resources to force protection or to become increasingly reliant on the ISF for such protection. In that scenario, even if the presence of a small stay-behind force would have given the United States leverage over Maliki, it would also have given Maliki some leverage over the United States, since he could have threatened to stop defending US personnel from Shia militia or Iranian paramilitary operators.72 Had Maliki accepted a residual force with greater legal protections only reluctantly, the United States might have found itself in the position of being condemned by Iraqi political leaders for continuing to ‘occupy’ the country. And had US forces stayed behind but not produced enough of the positive benefits outlined here, the result might simply have been to make Sunni factions view the United States as complicit in Maliki’s sectarian policies, thus stoking Sunni anger even further – and directing it, increasingly, at Washington. Finally, had ISIS nonetheless emerged in its full-blown form in 2014, there would have been thousands of American troops in harm’s way.

It is thus far too simplistic to contend that a modest stay-behind force would have acted as a sort of silver bullet with respect to ISIS’s rise. Rather, the effects of that force would have been limited in proportion to its limited size, and the United States would have had to accept the costs – specifically, the potential for casualties – that would have accompanied any residual deployment.73 That said, we believe this counterfactual remains fairly powerful, because it is possible to identify several concrete mechanisms by which a stay-behind force might have either inhibited ISIS’s rise, or at the very least put the United States and the Iraqi government in a stronger position to react more effectively.74 ISIS might still have been a significant problem in this scenario, in other words, but it probably would not have enjoyed the extremely permissive environment that allowed it to reap such dramatic gains by mid-2014. Moreover, a stay-behind force would likely have set in motion multiple downstream effects – improved awareness and early assessment of the ISIS threat, improved understanding of Iraqi vulnerabilities, better intelligence in the crucial 2014 phase – that could have markedly improved the options at Obama’s disposal later. At the price of a relatively modest military presence, the administration might have gained a substantially better footing against ISIS before it reached its global apex in 2014–15.

Counterfactual 3: More robust intervention in Syria in 2011–13

If events in Iraq contributed powerfully to ISIS’s rise, it was nonetheless in Syria where ISIS came into its own. Although the group’s leadership may have formed out of the remnants of AQI members languishing in Iraqi prisons, it was the ability to seize and hold land in Syria that transformed ISIS into the threat it ultimately became. In Syria, ISIS won a safe haven and forged alliances with other terrorist groups. It became strong enough to defy and ultimately break away from core al-Qaeda, and refined the tactics it would later use to great effect in Iraq. It leveraged the internet and social media to achieve enormous recruiting successes, and established a two-way foreign-fighter transmission belt that brought radicalised individuals to the Middle East and sent them back to host countries around the world. In all of these ways, the Syrian civil war was the crucible in which ISIS was hardened and forged.75

ISIS’s rise in Syria paralleled the decline of political order there, and the civil war – particularly Assad’s repression of Sunnis – was a godsend for an organisation whose rise depended on access to ungoverned spaces and the weaponisation of Sunni resentment. Even at the time, in fact, fears that the Syrian conflict might breed instability and extremism throughout the region occasioned vigorous internal debates on how Washington should respond. Despite some prominent officials arguing to the contrary, the Obama administration initially chose to take a fairly hands-off approach to the conflict in 2011 and 2012, providing only some non-lethal assistance, along with diplomatic and moral support, to moderate opposition groups. It then took a go-slow, incremental approach to arming and training moderate forces from 2013 onward, while still avoiding more aggressive options – such as grounding Assad’s air force and enforcing a no-fly zone, striking Assad’s regime in response to the massive use of chemical weapons outside Damascus in 2013, or providing more robust support for the opposition – that experts inside and outside the administration argued were necessary to bring the civil war to a close, and to forestall the destabilisation and radicalisation that conflict was causing.76 In effect, the United States staked out a maximalist goal in Syria by declaring, in 2011, that ‘Assad must go’, but adopted only modest measures in pursuit of that goal.

Obama had good reasons for limiting US involvement in Syria. He was committed to not launching another potentially long and costly Middle East war, a conviction he brought to his presidency and one that was, ironically, only strengthened by the unhappy aftermath of the administration’s earlier violation of that conviction: the intervention in Libya in 2011. Until the summer of 2013, moreover, many knowledgeable analysts believed that Assad’s fall was only a matter of time, and so a policy that blended rhetorical support for the opposition (‘Assad must go’) with minimal concrete involvement seemed a prudent way to get on the right side of events without assuming significant responsibility for the aftermath.77 It also seemed a reasonable way of letting American antagonists that were deeply involved in the Syrian civil war – Iran and Hizbullah on one side, and al-Qaeda on the other – bleed each other while the United States stood safely on the sidelines.78

Finally, although the question of whether military or covert options could succeed in toppling Assad (or force him to make major political concessions that would lead to a negotiated settlement) was hotly debated inside and outside of the administration, Obama was consistently sceptical. He worried that options such as no-fly zones would be expensive to implement and carried no guarantee of decisively influencing the course of the civil war; that covert programmes to support the moderate opposition could inadvertently end up empowering radical jihadis who were intermingled with other anti-Assad forces; and that absent UN Security Council authorisation – something Russian support for Assad precluded – any meaningful intervention would rest on tenuous legal foundations. He also worried that toppling Assad would create even more catastrophic instability, allowing al-Qaeda-linked radicals to claim vast swathes of Syria and precipitating massive revenge killings against Syria’s Alawite minority. At virtually every key decision point between 2011 and 2013, these concerns were seconded by other key actors within the administration and inhibited a more robust response.79

Obama was consistently sceptical

So what were the plausible alternative options, and might they have inhibited ISIS’s rise, either directly or by bringing the civil war to an end? As noted above, in 2011–13 Obama’s advisers broached a number of measures that were seriously considered within the administration: arming rebel factions (or arming them more aggressively), carving out safe zones within which to build a moderate anti-Assad force, establishing no-fly or even no-drive zones that would constrain the regime’s military, cratering Syrian runways to prevent Assad’s air force from flying, targeting key military units or even threatening the safety of the Syrian leadership as a way of exerting coercive pressure on the regime, among others.80 Some of Obama’s military and national-security advisers favoured at least some of these options, and argued that they could be executed at a tolerable cost. After he left the administration, former vice-chairman of the Joint Chiefs of Staff General James Cartwright argued, for instance, that an effective no-fly zone could be achieved over key areas of Syria primarily by relying on ground-based air-defence systems in neighbouring countries such as Turkey or Jordan, and at a tiny fraction of the cost of a traditional no-fly zone, which would have required first striking Syrian air defences and then patrolling Syria’s skies indefinitely.81 In 2012, moreover, many of Obama’s top aides supported a plan to begin arming the Syrian opposition in hopes of creating a more viable, moderate alternative to Assad.82

What might these approaches have achieved? At a minimum, they would have alleviated some civilian suffering, particularly among the Sunni population that increasingly served as ISIS’s base, and provided much better intelligence into the Syrian civil war and ISIS’s rise. More speculatively, these policies might have slowed that rise by strengthening more moderate opposition forces such as the Free Syrian Army, and allowing them to push back against the extremists at an earlier stage in the conflict.83 A more empowered opposition that enjoyed more visible US support might have appeared to be a more viable alternative both to the Assad regime and to extremist groups such as ISIS and the Al-Nusra Front (now Jabhat Fateh al-Sham), thus inhibiting the swing of popular support and military strength to the latter groups as time went on. Even if a more empowered opposition had not been capable of defeating Assad militarily, earlier US support for relatively moderate groups in eastern and northern Syria might have permitted them to check ISIS’s subsequent rise in those areas.84 Most optimistically of all, these developments – and particularly direct US military intervention against Assad – might have constrained and pressured the regime to the point where it was compelled to surrender power or make significant political concessions to bring the war to an end.

These more optimistic scenarios were part of the argument advanced by advocates of intervention, both within and outside the Obama administration. In retrospect, however, there are several reasons to doubt that limited intervention of the sort contemplated by the administration would have changed the dynamics of the Syrian civil war thoroughly enough to avert ISIS’s rise.

Firstly, it seems likely that any sort of increased US support for the anti-Assad opposition simply would have been matched and exceeded by Assad’s foremost patrons, Russia and Iran – which is, in fact, precisely what happened when increased US support eventually helped anti-Assad opposition groups gain the upper hand in mid-2015.85 Secondly, when the United States did undertake a more robust effort to arm and train the Syrian opposition in 2015 (this time in a counter-ISIS, rather than counter-Assad, context), some of that assistance did fall into the hands of extremists, thereby vindicating Obama’s earlier concerns about this approach.86 Thirdly, some of the most commonly advocated forms of direct military intervention, such as cratering Syrian runways or establishing a no-fly zone, would have constrained the regime and sheltered the civilian population only marginally, due to the fact that most regime operations were not critically dependent on air superiority in the early stages of the conflict.87

Finally, and perhaps more consequentially, it seems clear that advocates of greater US intervention in Syria fundamentally underestimated Assad’s resolve and determination to either retain power or die trying, something that has become eminently clear over the past several years.88 As a result, the level of coercion required to force Assad from power, or even to compel him to conclude an equitable peace settlement, was probably considerably higher than many advocates of intervention believed – and considerably higher than what the limited military options they advocated would have produced. Fundamentally changing the trajectory of events in Syria, in other words, likely would have required far more robust intervention than was contemplated in 2011–13, with all the accompanying dangers and costs.89 Even had limited US intervention exerted the level of pressure necessary to force Assad from power – for instance, by convincing other Alawite elites that they had to break with Assad before they lost all hope of influencing post-war outcomes – there is no guarantee that his departure would have been followed by political stabilisation and curbs on radicalisation. Rather, just as Obama and other officials feared, the outcome might just as plausibly have been a power vacuum in which significant areas of Syria went ungoverned and radical elements flourished.90 The experience of the United States in Iraq in 2003, and in Libya in 2011, lends particular weight to this possibility.

Such analysis is not conclusive, of course. After all, many of the downsides of US intervention that worried Obama – empowerment of extremists, creation of ungoverned spaces or other safe havens in which malign actors could operate, increased tensions with Russia and other backers of Assad – arose anyway in the context of US non-intervention. And it remains possible that the limited options contemplated by the administration might have had a more constructive effect. Perhaps a determined but carefully calibrated programme of US military intervention in 2011–12, combined with aggressive arming of the moderate opposition, would have shocked Assad and changed his calculus at a time when the course of events remained more malleable, or at least had that effect on other power brokers within the Alawite regime. Perhaps a signal of greater US resolve in Syria would have deterred greater Russian and Iranian intervention by preventing the strategic vacuum that was created by US inaction (one that Moscow eventually filled in 2015) and raising the costs of their own involvement.91 Perhaps Assad might have been pushed from power without destroying the state institutions necessary to preserve stability and prevent extremist groups from running wild in a post-Assad Syria. But hindsight makes such ‘what ifs’ seem more dubious, given that both Russia and Iran – particularly the latter – had a great deal riding on the survival of the Syrian regime; given the tenacity with which Assad has clung to power; and given recent US experience with regime change in the Middle East.

The downsides of intervention arose anyway

On balance, then, limited US intervention might have bolstered moderate Syrian factions vis-à-vis other malign actors in that conflict – whether Assad, Russia, Iran or jihadi terrorists – and thereby helped create partners with which the West could have worked more comfortably down the road. Moreover, it certainly would have improved the intelligence picture with respect to ISIS, no small benefit in light of subsequent underestimation of ISIS’s capabilities by US policymakers. But on its own, such intervention probably would not have affected the course of the Syrian civil war sufficiently to head off ISIS’s rise. Even if successful, it could have replaced one type of catastrophic instability with another.

This judgement does not conclusively exonerate Obama’s decision, because there are other desiderata beyond the counter-ISIS fight that a more robust policy in Syria might have achieved – for instance, strengthening the US hand in negotiations to end the civil war on more favourable terms, enabling better pushback against the spread of Iranian influence in the region, and averting the massive refugee flows, regional destabilisation and other adverse outcomes to which the Syrian civil war powerfully contributed. Because these dimensions lie outside the scope of our present analysis, we do not analyse them in enough detail to make a confident judgement one way or another. In the aggregate, however, it does seem that Obama’s Syria choices were considerably less consequential for ISIS’s rise than other decisions he took.

Counterfactual 4: Strikes on ISIS before the fall of Mosul

The final opportunity to block the full-blown emergence of ISIS would have been in late 2013 and early 2014, in the months prior to ISIS’s dramatic seizure of Mosul. By this point, ISIS’s rise was becoming increasingly apparent; the organisation was morphing from a shadowy insurgent network into a quasi-state that held territory, collected taxes and conducted large-scale military operations.92 As this transformation occurred, the Iraqi government requested greater US assistance, including American airstrikes, in confronting an organisation that was now threatening much of western and northern Iraq, and that was clearly preparing for some sort of major assault on Mosul.93 Even some US officials warned that the danger was increasing, with then-deputy assistant secretary of state Brett McGurk publicly testifying as early as November 2013 that the organisation was exploiting a ‘permissive operating environment’ in Anbar and Nineva provinces and saying that ‘time may not be on our side’.94

The Obama administration’s response was mixed but, on the whole, fairly tepid. The administration did expedite certain arms deliveries to Baghdad in late 2013 and early 2014, and provided additional intelligence support in the form of occasional unarmed drone flights.95 Yet it declined to intervene militarily to thwart ISIS’s advance, acting only after Mosul had fallen, ISIS had taken control of perhaps one-third of Iraq, and Baghdad and Erbil, with their significant populations of US diplomats and civilians, were under grave and imminent threat. As late as June 2014, in fact, administration spokespersons insisted that ‘ultimately, this is for the Iraqi security forces, and the Iraqi government to deal with’.96

The Obama administration’s response was fairly tepid

The reason Obama did not intervene during this period was not, ironically, because he believed that it was too late to alter the course of events in Iraq, but because he had misplaced confidence that such intervention was unnecessary. As captured in his infamous remark dismissing ISIS as a ‘jayvee team’, the president still viewed ISIS as a local problem for Iraq and Syria, and not the full-fledged regional and even international security challenge it would soon become.97 The Obama administration, whose insight into events in Iraq was limited by the extremely light US footprint there, also believed that the ISF, which on paper was still a robust, sizeable, force, could blunt ISIS’s advance.98

Moreover, although the Maliki government was requesting greater US assistance on ever more urgent terms, Obama judged – correctly, in our view – that Maliki was as much a part of the problem as part of the solution.99 Maliki’s hyper-sectarian policies were providing the ideological fuel that powered ISIS; coming to Maliki’s aid without changing those policies could potentially have entrenched him in power and worsened the long-term political situation in Iraq, and perhaps even made the United States complicit in a bloody, heavy-handed effort to punish Sunni communities that had welcomed ISIS into Anbar province. Instead, up to and even after the fall of Mosul, Obama withheld aid as a way of exerting leverage on Maliki, pushing him to reform – or failing that, pushing Iraqi political leaders to jettison Maliki in favour of a less sectarian alternative. As Obama said in June 2014, ‘The United States is not simply going [to] involve itself in a military action in the absence of a political plan by the Iraqis that gives us some assurance that they are prepared to work together.’100 Finally, in late 2013 and early 2014, the attention of top-level policymakers was being consumed by the eruption of a political, and then military, crisis in Ukraine, which further reduced Obama’s incentives to take on increased commitments in Iraq.101

What might have happened had Obama taken a different course and intervened militarily against ISIS in late 2013 and early 2014? In a narrow operational sense, an intervention directed against ISIS during this period would have been eminently feasible, even without changing earlier decisions. There were sufficient resources within range in the Persian Gulf and the broader CENTCOM area of operations to execute sustained airstrikes against ISIS; if necessary, the United States quickly could have augmented available resources by deploying an extra aircraft carrier to the region or surging additional aircraft to bases in the Gulf or perhaps Jordan. Moreover, at this stage, ISIS, which was using large columns of vehicles and other quasi-conventional tactics in an effort to gobble up additional territory, would have been highly vulnerable to even a rapidly assembled air campaign.102 The United States might have also bolstered the struggling ISF by providing logistical support and advisers, which could have proven very useful in addressing the crippling logistical shortfalls and lack of confidence and professionalism that ultimately doomed the defence of Anbar and Mosul.103

It therefore seems likely that a US intervention consisting of airstrikes, advisers and logistical support could have prevented ISIS from effectively massing to invade Mosul, even had that intervention come as late as the end of May 2014. Had the intervention come earlier, in late 2013 or the winter of 2014, it also probably could have enabled a more robust defence of Fallujah and other key areas of Anbar province.104 More broadly, although the rot within the ISF was well advanced by this point, the presence of US advisers and support might have mitigated the near-total collapse that occurred in Nineva province by June 2014, thereby allowing ISIS to turn what had initially been conceived as an operational raid into Mosul into a blitzkrieg-style advance that culminated just an hour’s drive from Baghdad.

The strategic benefits of all this would have been considerable. Allowing ISIS to consolidate control of Mosul and much of Anbar dramatically raised the costs and lengthened the timeline of any campaign to defeat that organisation, as operations to liberate those areas from 2015 onward clearly demonstrated.105 Denying or even limiting ISIS’s territorial advances thus would have been a significant improvement over what ultimately transpired. Similarly, ISIS’s lightning advance in the first half of 2014 was what allowed it (credibly) to declare a caliphate, to excite ideological fellow-travellers across the Muslim world, and to attract hordes of foreign fighters to the cause; limiting that advance might have counteracted these damaging dynamics somewhat.106 Additionally, had ISIS not taken Mosul and threatened Baghdad, the Iraqi government would not have become as dependent on the mobilisation of sectarian Shia militias, often supported by Iran, to prevent the outright collapse of the state.107

The US might well have been stuck with Maliki

Yet there also would have been limits to any gains produced by US intervention. For starters, ISIS might have adjusted its tactics to mitigate the impact of US airstrikes, as the group did in its later attacks on Ramadi in May 2015.108 In that case, ISIS took advantage of a sandstorm, which negated the impact of US airpower, using large vehicle-borne improvised explosive devices to capture a nominally well-defended city; it is possible that earlier US intervention might simply have caused an earlier shift to such tactics. More significantly, by late 2013 and early 2014, ISIS was already a formidable threat, controlling significant territory in Syria and attracting foreign fighters to the region. This dimension of the problem would have remained even had the United States prevented such dramatic gains on the Iraqi side of the border.

Finally, and most problematic of all, had the United States intervened militarily at this point, it might well have been stuck with Maliki, at least for a while. At the outset, US forces would have been especially dependent on the Maliki-led ISF for intelligence and targeting information, simply because it takes time to develop the intelligence required to make airstrikes effective. US forces might therefore have become a de facto ‘Shia air force’, undertaking strikes that stymied ISIS but also entrenched Maliki in power and allowed him to perpetuate his sectarian agenda, at precisely the time when many Iraqi Sunnis had concluded that ISIS was a better alternative to the Iraqi government.109 It is distinctly possible that US intervention under these terms might have beaten back ISIS militarily, but perpetuated or even exacerbated the political and sectarian problems that facilitated its rise.

In theory, the United States might have threaded the needle here by making military intervention against ISIS strictly contingent on Maliki’s commitment to decentralise power, cease repression of peaceful domestic opponents and pursue greater political inclusion of Sunnis. (This was, in fact, precisely the bargain the United States made with Iraqi leaders in August 2014, when Erbil and potentially Baghdad seemed threatened.110) But it remains unclear whether, in practice, such an approach would have produced the desired results prior to the fall of Mosul. After all, Maliki was willing to risk the loss of vast swathes of Iraqi territory to ISIS throughout early 2014 without making any significant, tangible steps toward reform, remaining obstinate until ISIS was barely 100km from Baghdad. During this same period, and in the face of these same setbacks, Maliki was able to maintain his grip on his political coalition and so was not forced to make concessions. If it took such a grave disaster to shift the Iraqi political equilibrium, and then only marginally, it seems doubtful that even strict US conditionality in early 2014 would have produced major political changes.

On the whole, then, intervening in late 2013 or early 2014 probably would have produced a marked improvement over actual policy in terms of checking ISIS’s military onslaught. But the improvement would not have affected ISIS’s redoubt in Syria, and the military gains in Iraq would have been offset to some degree by the longer-term political costs of becoming Maliki’s air force. To change the trajectory of events in a more unambiguously positive manner, this counterfactual would likely have needed to be paired with other changes earlier in the chain of events – for instance, a modest stay-behind force to improve targeting and leverage over Maliki, thus making an intervention in 2014 more effective and less politically costly.

* * *

The emergence of ISIS as a fully formed, major danger to regional and international security was not inevitable. To be sure, the group’s ascendancy can be traced in part to deep historical forces that took generations to develop and would have been hard to shift on the timeline contemplated in this assessment. And it would be a mistake to underplay the powerful local and regional dynamics, from sectarianism to state failure, that created such a conducive context for its emergence. But the rise of ISIS – or more precisely, the failure to thwart the rise of ISIS – also can be traced to policy decisions made by US administrations that enjoyed more freedom of choice, and more ability to affect the trajectory of that organisation, than they are generally inclined to acknowledge.

In retrospect, the most fateful choice was also the oldest one: the decision to invade Iraq in 2003, followed by the mismanagement of the occupation. That choice either unleashed or exacerbated many of the forces that drove ISIS’s ascendancy. A different course of events would have left Iraq as a major problem (though one that was perceived to be larger than it actually was), but not the type of problem that was likely to lead to the rise of an ISIS-like threat within a decade.

It is therefore correct to assert that the policies of 2003 loom large in any assessment of ISIS’s rise. But it is not correct to claim that the invasion of Iraq set in motion forces that led ineluctably to the problems that the United States has faced since mid-2014. On the contrary, we identified three other pivot points in the story where a different but plausible policy choice might have reduced the ISIS threat by a greater or lesser extent, albeit at some cost.

One of these decisions was Obama’s strategy for dealing with Iraq in 2010–11, which involved, firstly, a reluctance to apply US leverage to shape the formation of the Iraqi government after the 2010 election, and, secondly, the failure to secure a significant stay-behind force in Iraq. Had the Obama administration taken a different approach in these episodes, it is reasonable to believe that the situation in Iraq might not subsequently have deteriorated so dramatically, and that Washington would, at a minimum, have had better options for blocking or mitigating ISIS’s rise at lower risk before that organisation reached its terrifying peak in late 2014. Even in 2010–11, earlier decisions had fateful repercussions for later decisions. For instance, the failure to achieve a power-sharing agreement or otherwise impose conditionality on Maliki after the 2010 election significantly complicated the SOFA negotiations a year later. There would have been important limitations and drawbacks inherent in pursuing different policies in 2010–11, of course, but in our view the balance sheet on this counterfactual remains a net positive.

By contrast, the implications of the non-intervention in the Syrian civil war are murkier, and may well have had only a marginal effect on the rise of ISIS. Based on our assessment, it is hard to argue that the kinds of Syrian options that were viable in 2011–13 would have decisively tilted the balance against ISIS. And it is easy to identify ways in which US actions might have fallen prey to unintended consequences or been neutralised by countermoves by other actors. One could still argue that Obama should have intervened more robustly in Syria to accomplish other ends, but in our judgement his non-intervention probably had only a limited effect on the ISIS story.

The decision to let ISIS take Mosul in 2014 falls somewhere in between those two in terms of its net effect on the scope of the ISIS challenge we face today. Obama had an opportunity to thwart ISIS’s advance in Iraq at the invitation of the Iraqi government and, had he done so, the group would have been a much less formidable military threat. Moreover, the military costs of subsequently defeating it might have been considerably lower. Yet the political price of that intervention could well have been to prolong Maliki’s grip on power in Iraq, and given Maliki’s pernicious role in creating the conditions that ISIS exploited, this would have been a very high price to pay. As we have argued, then, the United States did indeed miss, under both Bush and Obama, opportunities to thwart ISIS’s ascendancy, but all of the counterfactuals here are messier than they first appear, and at least some of them are far less obviously superior to what actually happened than they may initially seem.

Indeed, regardless of how one assesses the various counterfactuals presented here, it is important to keep in mind that analyses such as these inevitably fall somewhat short of definitive proof of what would have happened if policymakers had pursued a different course of action. Had any of these decisions gone differently, they might have produced second- and third-order effects that are difficult to assess even with the benefit of hindsight. The enemy also gets a vote, and so had the United States acted differently in 2003 or 2011 or 2014, it is possible that other actors would also have adjusted their behaviour in significant ways. Saddam Hussein might have aggressively ramped up his weapons programmes had the United States not invaded in 2003; Iran might have moved aggressively to undermine a less sectarian Iraqi coalition or a more conditional US approach to Maliki in 2010; Iranian-backed Shia militias might have attacked US stay-behind forces after 2011; ISIS might have ramped up suicide attacks inside Baghdad or embraced other unconventional tactics to counter US airstrikes on the outskirts of Mosul, and so on. We believe that the counterfactual assessments offered here are supported by the available evidence and a plausible reading of events, but we acknowledge that there is a limit to how fully anyone can game out how the past might have been. Counterfactual analysis, like foreign policy itself, eventually comes down to making informed judgements under conditions of at least some irreducible uncertainty.

This relates to a final point, which is that balance and a degree of empathy are vital in assessing counterfactuals such as these. Counterfactual analysis should not be used solely as a way of assigning blame or indicting policymakers for their failures. Every policy decision considered here was understandable in the context of what was known at the time, and reasonable people inside and outside the Bush and Obama administrations debated those decisions using many of the arguments we have explored here. And as we have noted, certain counterfactuals that are treated as near-certainties in hindsight – for instance, the idea that the United States might have blocked ISIS’s rise by intervening more heavily in Syria in 2011–13 – are far less clear-cut than one might initially assess, while even the ‘cleaner’ counterfactuals, such as those involving Iraq in 2010–11, would not have been unambiguous in their effects. Understanding the complexity and messiness of counterfactual analysis is thus essential to accurately judging whether ISIS’s rise might indeed have been averted, and at what cost.

It is also essential for getting future policy judgements right. If ISIS’s rise was inevitable, then perhaps there is nothing the United States can do to prevent the rise of other ISIS-like movements in the future. If, however, the counterfactual analysis that we offer here is correct in asserting that ISIS’s rise was not inevitable, then it follows that there may be policy steps the United States can take to ensure that the defeat of one powerful terrorist quasi-state is not simply followed by the rise of another – steps that are likely to involve trade-offs and costs of their own. Unfortunately, the history related here does not yield an exact formula for determining those policies and assessing their costs and trade-offs. History is, at best, a general rather than a precise guide to statecraft. Moreover, assessing the best course of action for US policymakers in the next phase of the ongoing struggle against terrorists inspired by militant Islamism should hinge on many factors beyond the scope of this article. But the policy debates that precede any such decisions must begin with the insight we offer here: that good policy hinges on strategic judgements that are themselves counterfactual assessments, and that a nuanced and intellectually honest approach to making such assessments is therefore vital.


1 For a discussion of the options, see Peter Feaver and Hal Brands, ‘Trump and Terrorism: U.S. Strategy after ISIS’, Foreign Affairs, vol. 96, no. 2, March–April 2017, pp. 28–36; and Daniel Byman, ‘How To Fight Terrorism in the Donald Trump Era’, National Interest, January–February 2017,

2 See Kristin Roberts, ‘ISIS Sets Its Sights on Saudi Arabia, and That’s Bad News for Washington’, Defense One, 29 May 2015; Andrew Spath, ‘Jordan: Between Stability and Spillover’, Foreign Policy Research Institute E-Note, 1 August 2014,; and Ken Dilanian, ‘DIA Chief: Iraq and Syria May Not Survive as States’, Military Times, 10 September 2015.

3 See Eric Schmitt, ‘Caliphate in Peril, More ISIS Fighters May Take Mayhem to Europe’, New York Times, 17 September 2016; and The Soufan Group, ‘An Updated Assessment of the Flow of Foreign Fighters into Syria and Iraq’, December 2015,; International Centre for Counter-Terrorism, ‘The Foreign Fighters Phenomenon in the European Union: Profiles, Threats, and Policies’, April 2016,

4 See, for instance, Harleen Gambhir, ‘ISIS’s Campaign in Europe: March 2016’, Institute for the Study of War, March 2016,; Rukmini Callimachi, ‘Not “Lone Wolves’” After All: How ISIS Guides World’s Terror Plots from Afar’, New York Times, 4 February 2017; and Barak Mendelsohn, ‘ISIS’ Lone-Wolf Strategy’, Foreign Affairs, 25 August 2016.

5 See, for example, Brendan Koerner, ‘Why ISIS is Winning the Social Media War’, Wired, March 2016,; and Alberto M. Fernandez, ‘Here To Stay and Growing: Combating ISIS Propaganda Networks’, US–Islamic World Forum Papers 2015, Brookings Institution, October 2015,

6 ‘UN Rights Chief Says Syria “Worst Man-Made Disaster Since WWII”’, RadioFreeEurope/Radio Liberty, 14 March 2017,

7 See Margaret Hartmann, ‘How ISIS is Destroying Ancient Art in Iraq and Syria’, New York Magazine, 6 March 2015,; and Camila Domonoske, ‘ISIS Destroys Ancient Theater, Tetrapylon in Palmyra, Syria Says’,, 20 January 2017,

8 On the relationship between ISIS and Obama’s broader grand strategy, see Hal Brands, ‘Barack Obama and the Dilemmas of American Grand Strategy’, Washington Quarterly, vol. 39, no. 4, 2017, especially pp. 113–14.

9 See Julia Glum, ‘How Many People Has ISIS Killed?’, International Business Times, 10 August 2016; and Khushbu Shah and Tim Hume, ‘Nearly 19,000 Civilians Killed in Iraq in 21-Month Period, Report Says’, CNN, 19 January 2016.

10 According to exit polls, Donald Trump beat Hillary Clinton by 17 points among the nearly one-in-five voters who identified terrorism as their top concern. See the Washington Post summary of the exit polls here: See also Michael Crowley, ‘Will Terror Help Trump?’, Politico, 22 March 2016.

11 For examples of these various perspectives, see Obama’s comments in Jeffrey Goldberg, ‘The Obama Doctrine’, Atlantic, April 2016,; Edward Morrisey, ‘Obama’s ISIS Failure’, The Week, 17 November 2015,; Bradford Richardson, ‘McCain: Obama’s Failed Leadership Led to Rise of ISIS’, The Hill, 15 November 2015,; Dexter Filkins, ‘Did George W. Bush Create ISIS?’, New Yorker, 15 May 2015,; and John Cassidy, ‘The Iraq Mess: Place Blame Where It Is Deserved’, New Yorker, 13 June 2014,

12 For a description of the territory controlled by ISIS at its peak, see George Packer, ‘The Common Enemy’, New Yorker, 25 August 2014, For a discussion of the rise of ISIS more broadly, see William McCants, ISIS Apocalypse: The History, Strategy, and Doomsday Vision of the Islamic State (New York: Picador, 2015); and James Fromson and Steven Simon, ‘ISIS: The Dubious Paradise of Apocalypse Now’, Survival, vol. 57, no. 3, May–June 2015, pp. 5–57.

13 For advocacy of more aggressive strategies, see Steven Metz, ‘The Case for a Punitive Expedition Against the Islamic State’, World Politics Review, 6 February 2015; and Max Boot, ‘Defeating ISIS’, Policy Innovation Memorandum No. 51, November 2014,

14 For a sceptical view, see Edward H. Carr, What is History? (New York: Knopf, 1962).

15 See James Fearon, ‘Counterfactuals and Hypothesis Testing in Political Science’, World Politics, vol. 43, no. 2, 1991, pp. 169–95; John Lewis Gaddis, The Landscape of History: How Historians Map the Past (New York: Oxford University Press, 2004); Niall Ferguson (ed.), Virtual History: Alternatives and Counterfactuals (New York: Basic Books, 2011); and Jack S. Levy, ‘Counterfactuals, Causal Inference, and Historical Analysis’, Security Studies, vol. 24, no. 2, 2015, pp. 378–402.

16 See Levy, ‘Counterfactuals, Causal Inference, and Historical Analysis’, pp. 389–94; and Philip Tetlock and Aaron Belkin, ‘Counterfactual Thought Experiments in World Politics: Logistical, Methodological, and Psychological Explanations’, in Philip Tetlock and Aaron Belkin (eds), Counterfactual Thought Experiments in World Politics (Princeton, NJ: Princeton University Press, 1996), pp. 23–5.

17 Gaddis, Landscape of History, p. 101.

18 On the Vietnam counterfactual, see Fredrik Logevall, Choosing War: The Lost Chance for Peace and the Escalation of War in Vietnam (Berkeley, CA: University of California Press, 1999).

19 As Fawaz Gerges argues, however, the roots of ISIS run much deeper than the 2003 Iraq War, and can be traced to the decades-long process that gave rise to al-Qaeda itself. Fawaz A. Gerges, ISIS: A History (Princeton, NJ: Princeton University Press, 2016).

20 Richard Engel, And Then All Hell Broke Loose: Two Decades in the Middle East (New York: Simon & Schuster, 2016), p. 107. See also David Kilcullen, Blood Year: The Unraveling of Western Counterterrorism (New York: Oxford University Press, 2015), especially pp. 15–36.

21 For balanced treatments of the decision-making concerning Iraq, see Melvyn Leffler, ‘The Foreign Policies of the George W. Bush Administration: Memoirs, History, Legacy’, Diplomatic History, vol. 37, no. 2, 2013, pp. 190–216; Robert Jervis, ‘Reports, Politics, and Intelligence Failures: The Case of Iraq’, Journal of Strategic Studies, vol. 29, no. 1, 2006, pp. 3–52; and Hal Brands, What Good is Grand Strategy: Power and Purpose in American Statecraft from Harry S. Truman to George W. Bush (Ithaca, NY: Cornell University Press, 2014), Chapter 4. For an interesting contrarian take, see Frank Harvey, Explaining the Iraq War: Counterfactual Theory, Logic and Evidence (Cambridge: Cambridge University Press, 2012). On the 1990s-era debate over Iraq, see also Derek Chollet and James Goldgeier, America between the Wars: From 11/9 to 9/11: The Misunderstood Years between the Fall of the Berlin Wall and the Start of the War on Terror (New York: PublicAffairs, 2008), Chapter 7.

22 On the intelligence failure, see Jervis, ‘Reports, Politics, and Intelligence Failures’.

23 See the discussion of Bush’s overall strategy in Condoleezza Rice, No Higher Honor: A Memoir of My Years in Washington (New York: Crown, 2011), pp. 148–81.

24 For a discussion of this phenomenon, see Vali Nasr, The Shia Revival: How Conflicts within Islam Will Shape the Future (New York: W.W. Norton, 2006).

25 The fallout from the invasion is described in Kilcullen, Blood Year, pp. 15–36; Dana Priest, ‘Iraq New Terror Breeding Ground’, Washington Post, 14 January, 2005; Ahmed Hashim, Insurgency and Counter-Insurgency in Iraq (Ithaca, NY: Cornell University Press, 2006); and Isabel Cole and Ned Parker, ‘How Saddam’s Fighters Help Islamic State Rule’, Reuters, 11 December 2015. On the idea that an Arab Spring-like event could have occurred even without the US invasion of Iraq, see Juan Cole, ‘Top Ten Myths about the Arab Spring of 2011’, Informed Comment, 30 December 2011,

26 These points are well covered in Harvey, Explaining the Iraq War; and Douglas Feith, War and Decision: In the Pentagon at the Dawn of the War on Terrorism (New York: HarperCollins, 2008), especially pp. 179–228.

27 Peter Bergen, The Longest War: The Enduring Conflict between America and al-Qaeda (New York: Simon & Schuster, 2011), especially pp. 266–96.

28 On this particular decision, see Micah Zenko, ‘Foregoing Limited Force: The George W. Bush Administration’s Decision Not to Attack Ansar al-Islam’, Journal of Strategic Studies, vol. 32, no. 4, 2009, pp. 615–49.

29 See Daniel Byman, ‘An Autopsy of the Iraq Debacle: Policy Failure or Bridge Too Far?’, Security Studies, vol. 17, no. 4, 2008, pp. 599–643; James Fallows, Blind into Baghdad: America’s War in Iraq (New York: Random House, 2006); James Dobbins et al., Occupying Iraq: A History of the Coalition Provisional Authority (Santa Monica, CA: RAND Corporation, 2009); and James Pfiffner, ‘U.S. Blunders in Iraq: De-Baathification and Disbanding the Army’, Intelligence and National Security, vol. 25, no. 1, February 2010, pp. 76–85.

30 Byman, ‘An Autopsy of the Iraq Debacle’.

31 On declining violence and the blows inflicted on AQI, see Stephen Biddle, Jeffrey Friedman and Jacob Shapiro, ‘Testing the Surge: Why Did Violence Decline in Iraq in 2007?’, International Security, vol. 37, no. 1, Summer 2012, pp. 7–40; and Bergen, Longest War, pp. 266–96.

32 See Kenneth Pollack, ‘Reading Machiavelli in Iraq’, National Interest, November–December 2012, pp. 8–19; and Dexter Filkins, ‘What We Left Behind’, New Yorker, 28 April 2014,

33 We chose the 10,000–20,000 range because it is bounded, on the low end, by the size of the stay-behind force that Obama himself was initially willing to seek, and, on the high end, by the size of the force that many US military commanders thought appropriate. The various proposals for a stay-behind force are covered in Michael Gordon and Bernard Trainor, The Endgame: The Inside Story of the Struggle for Iraq, from George W. Bush to Barack Obama (New York: Pantheon, 2012), pp. 651–71.

34 On these competing claims, see, for example, Ali Khedery, ‘Why We Stuck with Maliki – and Lost Iraq’, Washington Post, 3 July 2014; Rick Brennan, ‘Withdrawal Symptoms: The Bungling of the Iraq Exit’, Foreign Affairs, vol. 93, no. 6, 2014, pp. 25–34; Emma Sky, The Unraveling: High Hopes and Missed Opportunities in Iraq (New York: PublicAffairs, 2015), pp. 330–59; Mark Landler, ‘U.S. Troops to Leave Iraq by Year’s End, Obama Says’, New York Times, 21 October 2011; Christopher Hill, Outpost (New York: Simon & Schuster, 2014); and Yochi Dreazen, ‘U.S. Troops Are Leaving Because Iraq Doesn’t Want Them There’, Atlantic, 21 October 2011,

35 Katie Sanders, ‘Obama Wanted to Keep 10,000 Troops in Iraq, ABC’s Raddatz Claims’, PolitiFact, 24 August 2014,; and Gordon and Trainor, The Endgame, pp. 651–71.

36 For examples of such statements, see Caren Bohan, ‘Obama Touts Foreign Policy Successes in Iraq, Libya’, Reuters, 22 October 2011; and Eli Lake, ‘Why the White House Ignored All Those Warnings About ISIS’, Daily Beast, 14 July 2014.

37 This point was made to us by multiple former Obama administration officials.

38 Gordon and Trainor, The Endgame, p. 643.

39 Emma Sky, ‘How Obama Abandoned Democracy in Iraq’, Politico, 7 April 2015.

40 This was the level of US troops toward the end of summer 2010; the US presence had been even greater when the electoral processes started months earlier.

41 Khedery, ‘Why We Stuck with Maliki – and Lost Iraq’.

42 Peter Mansoor, Surge: My Journey with General David Petraeus and the Remaking of the Iraq War (New Haven, CT: Yale University Press, 2013), pp. 209–32; and Pollack, ‘Reading Machiavelli in Baghdad’, p. 9.

43 The El Salvador case is discussed in Hal Brands, Making the Unipolar Moment: U.S. Foreign Policy and the Rise of the Post-Cold War Order (Ithaca, NY: Cornell University Press, 2016), pp. 132–7. For a more sceptical take (but one that still concedes that US involvement had certain constructive political effects), see Benjamin Schwartz, American Counterinsurgency Doctrine and El Salvador: The Frustrations of Reform and the Illusions of Nation Building (Santa Monica, CA: RAND Corporation, 1992).

44 For a discussion of the idea that the United States should have insisted that Allawi be allowed to try to assemble a government, see Sky, ‘How Obama Abandoned Democracy in Iraq’.

45 Gordon and Trainor, The Endgame, p. 644.

46 Pollack, ‘Reading Machiavelli in Baghdad’, p. 10.

47 On these various issues, see Ahmed Rashid, ‘How Obama Lost Karzai’, Foreign Policy, 21 February 2011; and Hill, Outpost, p. 372. See also David Rohde, ‘The Rise and Fall of Our Man in Baghdad’, Defense One, 1 July 2014; and Colin Kahl, ‘No Obama Didn’t Lose Iraq’, Politico, 15 June 2014,

48 See Khedery, ‘Why We Stuck with Maliki’; and Sky, ‘How Obama Abandoned Democracy in Iraq’.

49 To be fair, the administration did try, relatively late in the game, to induce Maliki to accept an arrangement in which Allawi would become president of Iraq and other political concessions would be made. Yet this effort was undermined by the fact that the United States had already essentially acquiesced in Maliki’s premiership, as well as by the fact that it required the Kurds, who controlled the Iraqi presidency, to sacrifice their own power for the sake of Sunni–Shia harmony. In the event, little ultimately came of this effort. See Gordon and Trainor, Endgame, pp. 628–50.

50 Filkins, ‘What We Left Behind’.

51 Brennan, ‘Withdrawal Symptoms’, p. 33.

52 For an alternative view, albeit one that does not evaluate any of the counter-arguments presented here, see Juan Cole, ‘The Fall of Mosul and the False Promises of Modern History’, Informed Comment, 11 June 2014, Gregory Gause makes a similar argument in ‘Arab Politics Is Not All About Us’, Foreign Policy, 22 October 2012.

53 As Colin Kahl writes in a piece that otherwise defends administration policy, ‘Through the first half of 2011, there was a vigorous debate within the administration about whether U.S. forces should remain in Iraq beyond December, and if so, in what numbers and with what missions.’ Kahl, ‘No Obama Didn’t Lose Iraq’, emphasis added.

54 Defenders of Obama’s choices point out that Maliki was not keen on accepting the higher numbers proposed by the Pentagon and so welcomed some reduction, but it is doubtful he saw the final offer of 3,500 as worth the political cost. On these issues, see Gordon and Trainor, Endgame, pp. 651–72; Josh Rogin, ‘How the Obama Administration Bungled the Iraq Withdrawal Negotiations’, Foreign Policy, 21 October 2011; Michael Gordon, ‘In U.S. Exit from Iraq, Failed Efforts and Challenges’, New York Times, 22 September 2012; Michael Schmidt and Tim Arango, ‘Despite Difficult Talks, U.S. and Iraq Had Expected Some American Troops to Stay’, New York Times, 21 October 2011; and Leon Panetta, Worthy Fights: A Memoir of Leadership in War and Peace (New York: Penguin, 2014), pp. 392–3.

55 ‘End of Petraeus, Crocker Era Dawning’, CBS News, 29 July 2008,; Al Kamen, ‘In the Loop: Are Gen. Odierno and Ambassador Hill Not Getting Along?’, Washington Post, 30 September 2009; and Sky, ‘How Obama Abandoned Democracy in Iraq’.

56 On these previous workarounds, see Gordon and Trainor, Endgame, pp. 549–50.

57 See Brennnan, ‘Withdrawal Symptoms’; and Thomas L. Friedman, ‘Obama on the World’, New York Times, 8 August 2014.

58 For the ‘jayvee’ comment, see David Remnick, ‘Going the Distance’, New Yorker, 27 January 2014, For a discussion of the underestimation of ISIS, see Peter Baker and Eric Schmitt, ‘Many Missteps in Assessment of ISIS Threat’, New York Times, 29 September 2014. For more on the importance of a military footprint for better intelligence, see Keren Fraiman, Austin Long and Caitlin Talmadge, ‘Why the Iraqi Army Collapsed (and What Can Be Done About It)’, in ‘Iraq between Maliki and the Islamic State’, Project on Middle East Political Science, 9 July 2014, p. 14,

59 Gordon and Trainor, Endgame, pp. 651–60.

60 See, for instance, Ben Kesling and Matt Bradley, ‘Victory Marks Turnaround for Iraq Army’, Wall Street Journal, 29 December 2015.

61 Panetta, Worthy Fights, pp. 393–4. On the erosion of the ISF, see Eric Schmitt and Michael Gordon, ‘The Iraqi Army Was Crumbling Long Before Its Collapse’, New York Times, 12 June 2014; and Tom Bowman, ‘U.S. Faces Challenges in Shoring Up Iraq’s Crumbling Military’, National Public Radio, 24 June 2014,

62 On the trajectory of the CTS, see Loveday Morris, ‘The Force Leading the Iraqi Army’s Fight against ISIS Went from “Dirty Division” to Golden Boys’, Washington Post, 26 July 2016.

63 On logistical shortfalls and the collapse of ISF, see Ned Parker, Isabel Coles and Raheem Salman, ‘How Mosul Fell: A General’s Story’, Reuters, Special Report 1, 14 October 2014, For a discussion of these and other issues pertaining to US support for the ISF, see also Brennan, ‘Withdrawal Symptoms’.

64 Stephen Biddle, ‘Iraq After the Surge’, testimony before the House Committee on Armed Services, Second Session, 110th Congress, 23 January 2008.

65 See Pollack, ‘Reading Machiavelli in Baghdad’; and Pollack, ‘The Fall and Rise and Fall of Iraq’, Brookings Institution, July 2013, See also Sky, The Unraveling, p. 311; and Filkins, ‘What We Left Behind’.

66 See, for instance, Michael Flynn, ‘DIA Annual Threat Assessment 2014’, statement before the Senate Armed Services Committee, 11 February 2014,

67 Filkins, ‘What We Left Behind’.

68 The dynamics and limits of conditionality are explored in Stephen Biddle, ‘Evaluating U.S. Options for Iraq’, Statement Before the Committee on Armed Services of the United States House of Representatives, 29 July 2014.

69 Quoted in Filkins, ‘What We Left Behind’.

70 On ISIS’s tactics prior to Mosul, see David Kilcullen, ‘We’re Losing the War Against ISIS in Iraq’, National Interest, 15 September 2015,

71 Stephen Biddle, Ryan Baker and Julia Macdonald, ‘Small Footprint, Small Payoff: The Military Effectiveness of Security Force Assistance’, manuscript in draft.

72 This prospect has been a continual preoccupation of US officials since the start of the counter-ISIS campaign. See, for instance, ‘Will PMU Attack U.S. Troops in Iraq?’, Al-Monitor, 3 October 2016,

73 One question that arises is whether a smaller residual force along the lines of the final US offer to Iraq (around 5,000 troops in total) would have provided the benefits outlined here. Although space constraints preclude addressing this sub-counterfactual in any detail, we believe that a force of 5,000 troops would have been markedly better than the near-zero residual presence that ultimately eventuated, and would have at least positioned the United States to respond more effectively to ISIS’s rise. We do acknowledge, however, that the benefits of a smaller force would have been proportionally smaller to those of a larger force.

74 Our assessment thus largely tracks with that of other close observers, such as Peter Beinart, who summed up the situation by saying: ‘sooner or later, honest liberals will have to admit that Obama’s Iraq policy has been a disaster’. Peter Beinart, ‘Obama’s Disastrous Iraq Policy: An Autopsy’, Atlantic, 23 June 2014,

75 Gerges, ISIS: A History, especially Chapter 6.

76 Obama’s deliberations and decisions on Syria are covered in Derek Chollet, The Long Game: How Obama Defied Washington and Redefined America’s Role in the World (New York: PublicAffairs, 2016), pp. 127–56; Mark Mazzetti, Robert Worth and Michael Gordon, ‘Obama’s Uncertain Path Amid Syria Bloodshed’, New York Times, 22 October 2013; and Mark Hosenball, ‘Obama Authorizes Secret Support for Syrian Rebels’, Reuters, 1 August 2012.

77 For a discussion of such expectations, see Helene Cooper, ‘Washington Begins to Plan for Collapse of Syrian Government’, New York Times, 18 July 2012.

78 This rationale was related to us by former members of the Obama administration.

79 See, for instance, the comments of Leon Panetta in the PBS Frontline Documentary, ‘The Rise of ISIS’; Elisabeth Bumiller, ‘Military Points to Risks of a Syrian Intervention’, New York Times, 11 March 2012; Martin Dempsey, ‘General Dempsey’s Letter to Senator Levin on the U.S. Military and the Syrian Conflict’, 19 July 2013,; and Chollet, Long Game, pp. 127–56.

80 See, for instance, Anthony Cordesman, ‘U.S. Options in Syria: The Dempsey Letter’, Center for Strategic and International Studies, 23 July 2013; Kenneth Pollack, ‘Breaking the Stalemate: The Military Dynamics of the Syrian Civil War and Options for Limited U.S. Intervention’, Brookings Institution, August 2013; David Sorenson, ‘Dilemmas for U.S. Strategy: U.S. Options in Syria’, Parameters, vol. 43, no. 3, Autumn 2013, pp. 5–15; Chollet, Long Game; and Tara McKelvey, ‘Arming Syrian Rebels: Where the US Went Wrong’, BBC News, 10 October 2015.

81 David Willman, ‘How a $2.7 Billion Air-Defense System Became a “Zombie” Program’, Los Angeles Times, 24 September 2015,; and private correspondence with General James Cartwright, 23 February 2017.

82 McKelvey, ‘Arming Syrian Rebels’.

83 Some of these scenarios are explored in Derek Chollet, ‘Inside Obama’s Syria Choices (A Guide for Dissenting Diplomats)’, Defense One, 20 June 2016,; Kimberly Kagan, ‘The Smart and Right Thing in Syria’, Strategika, April 2013, pp. 11–14; and Charles Dunne, ‘The Syrian Crisis: A Case for Greater U.S. Involvement’, Freedom House Policy Brief, 14 March 2013,

84 From 2014 onward, in fact, the Syrian Kurds became Washington’s primary local partner in rolling back ISIS gains in Syria. Rukmini Callimachi, ‘Inside Syria: Kurds Roll Back ISIS, but Alliances are Strained’, New York Times, 10 August 2015.

85 See Eyal Zisser, ‘Russia’s War in Syria’, Strategic Assessment, vol. 19, no. 1, 2016, pp. 41–9.

86 See Michael Shear, Helene Cooper and Eric Schmitt, ‘Obama Administration Ends Effort to Train Syrians to Combat ISIS’, New York Times, 9 October 2015.

87 The limits of airpower options are discussed in Karl Mueller, Jeffrey Martini and Thomas Hamilton, Airpower Options for Syria: Assessing Objectives and Missions for Aerial Intervention (Santa Monica, CA: RAND Corporation, 2013).

88 See, for instance, Joby Warrick and Anne Gearan, ‘Syrian Dictator Bashar al-Assad Digs in for a Fight to the Death’, Washington Post, 31 July 2012.

89 As one perceptive analysis from 2012 noted, ‘At the end of the day … removing Asad [sic] may not be doable at a price the United States is willing to pay.’ See Daniel Byman, Michael Doran, Kenneth Pollack and Salman Shaikh, ‘Saving Syria: Assessing Options for Regime Change’, Brookings Institution, Middle East Memo #21, March 2012, pp. 1–2.

90 On this danger, see, for instance, Erika Soloman, ‘Insight: Syria Rebels See Future Fight with Foreign Radicals’, Reuters, 8 August 2012; and Tom A. Peter, ‘Syria After Assad: Will Radicals Rule?’, USA Today, 4 January 2013.

91 The counterfactual here is that the United States might have presented Russia and Iran with some of the same dilemmas that their intervention in Syria later presented to Washington. On some of those dilemmas, see David Sanger, ‘Russian Intervention in Syrian War Has Sharply Reduced U.S. Options’, New York Times, 10 February 2016.

92 See McCants, ISIS Apocalypse; Institute for the Study of War, ‘ISIS Annual Reports Reveal a Metrics-Driven Military Command’, 22 May 2014,; and Michael Weiss and Hassan Hassan, ISIS: Inside the Army of Terror (New York: Regan, 2016), pp. 216–58.

93 Michael Gordon and Eric Schmitt, ‘U.S. Said to Rebuff Iraqi Request to Strike Militants’, New York Times, 11 June 2014; and Adam Entous, Julian Barnes and Siobhan Gorman, ‘Secret U.S. Plan to Aid Iraq Fizzled Amid Mutual Distrust’, Wall Street Journal, 21 June 2014.

94 ‘Testimony of Deputy Assistant Secretary Brett McGurk, House Foreign Affairs Subcommittee on the Middle East and North Africa Hearing: Iraq’, 13 November 2013; See also Baker and Schmitt, ‘Many Missteps in Assessment of ISIS Threat’.

95 See Michael Gordon and Eric Schmitt, ‘U.S. Sends Arms to Aid Iraq Fight with Extremists’, New York Times, 25 December 2013; and Entous, Barnes and Gorman, ‘Secret U.S. Plan to Aid Iraq Fizzled Amid Mutual Distrust’.

96 See Gordon and Schmitt, ‘U.S. Said to Rebuff Iraqi Request to Strike Militants’; and Entous, Barnes and Gorman, ‘Secret U.S. Plan’.

97 Remnick, ‘Going the Distance’.

98 On the status of the ISF as of June 2014, see Anthony H. Cordesman and Sam Khazai, ‘Shaping Iraq’s Security Forces’, Center for Strategic and International Studies, 12 June 2014,

99 US scepticism of Maliki is discussed in Karen DeYoung, ‘U.S. Seeking Alternatives to Maliki’, Washington Post, 19 June 2014.

100 Priyanka Boghani, ‘Obama Says Iraq Needs to Solve Its Own Problems, but the U.S. Will Do Its Part’,, 10 June 2014,

101 This point was related to us in conversations with former Obama administration officials.

102 Kilcullen, ‘We’re Losing the War Against ISIS in Iraq’. See also Judit Neurink, ‘ISIS in Iraq: The Fall of Mosul to the Jihadists Was Less of a Surprise to Baghdad Than Many Were Led to Believe’, Independent, 25 February 2016.

103 On these issues and the fall of Mosul, see Parker, Coles and Salman, ‘How Mosul Fell’.

104 On the fall of Fallujah, see Liz Sly, ‘Al-Qaeda Force Captures Fallujah Amid Rise in Violence’, Washington Post, 3 January 2014.

105 See, for example, Loveday Morris and Mustafa Salim, ‘After a Slow and Bloody Fight against ISIS, Iraqi Forces Pick up the Pace in Mosul’, Washington Post, 14 January 2017.

106 See, for instance, Nancy Youssef, ‘3,000 More Foreign Jihadists Join ISIS’, Daily Beast, 30 April 2015.

107 On the role of these militias, see Liz Sly, ‘Pro-Iran Militias’ Success in Iraq Could Undermine US’, Washington Post, 15 February 2015.

108 Eric Schmitt and Helene Cooper, ‘ISIS Fighters Seized Advantage in Iraq by Striking During Sandstorm’, New York Times, 18 May 2015.

109 Current and former officials often publicly discussed this problem at the time. See the comments of former CIA director General David Petraeus in Loveday Morris, ‘Iraqi Army Increasingly Bolstered by Shiite Militias as ISIS Advances’, Washington Post, 20 June 2014. See also Martin Chulov and Spencer Ackerman, ‘Iraq Requests U.S. Air Strikes as ISIS Insurgents Tighten Grip on Oil Refinery’, Guardian, 18 June 2014.

110 See Jay Solomon and Carol Lee, ‘U.S. Signals Iraq’s Maliki Should Go’, Wall Street Journal, 19 June 2014; and Jonathan Ferziger and Nadeem Hamid, ‘Maliki Resigns as Prime Minister to Make Way for Abadi’, Bloomberg, 15 August 2014.

Hal Brands is the Henry A. Kissinger Distinguished Professor of Global Affairs at the Johns Hopkins School of Advanced International Studies and Senior Fellow at the Center for Strategic and Budgetary Assessments. His most recent book is Making the Unipolar Moment: U.S. Foreign Policy and the Rise of the Post-Cold War Order (Cornell University Press, 2016).

Peter Feaver is Professor of Political Science and Public Policy at Duke University, where he directs the American Grand Strategy Program and the Triangle Institute for Security Studies.

Back to content list

Survival: Global Politics and Strategy

June-July 2017

Also available in Kindle and iPad format:

Kindle UK > 

Kindle US >

iBookstore UK >

iBookstore US >

Table of Contents

Available to download as a PDF >