Small Wars Journal

The Success of the Surge

Tue, 12/25/2007 - 5:13am
The Success of the Surge

By David Glasner

The real world, even under normal conditions, is a complicated and confusing place. In war, complications and confusion increase exponentially. Ever since January, when President Bush announced what he called a new way forward (and others called "the surge") in Iraq and selected General David Petraeus, whose ideas for counterinsurgency warfare, codified in the Army's new Counterinsurgency Manual which he co-authored, provided the rationale and blueprint for the new strategy, to take command of American forces in Iraq, arguments about whether the new strategy was really working have been going back and forth. After months of confusion, the picture has become unmistakably clear. The surge worked.

At first, critics of the surge refused to acknowledge that anything new was being tried other than to send another 25,000 troops into what had become a hopeless situation. In fact, the surge was the first (or second if one counts the belated sacking of Secretary Rumsfeld) serious, albeit tacit, acknowledgement by the Bush administration that it was facing a real insurgency in Iraq and that any strategy for success (as opposed to a classic but irrelevant concept of military victory) had to aim at changing the conditions on the ground that allowed the insurgency to flourish and gather strength.

The most critical condition fostering the insurgency was the lack of security for the local population. During the first four years after the invasion, the provision of security to the local population was at best a subsidiary part of the military mission that American forces were supposed to accomplish in Iraq. Secretary Rumsfeld and Undersecretary Wolfowitz made that very clear from the outset, when they sacked General Eric Shinseki for daring to tell Congress that a post-invasion force of several hundred thousand troops would be required in Iraq to provide security and was emphatically punctuated with an explanation point by Secretary Rumsfeld's infamous "stuff happens" comment when the looting started in Baghdad in April 2003. The provision of security was a mission for Iraqis not Americans to discharge. If Iraqi forces were unable to provide security, the Iraqis would just have to live without security until American forces trained and equipped enough Iraqi police and troops that could provide it. Aside from training Iraqi forces, the primary American mission was to wage a "war on terror" by killing "terrorists" wherever they could find them.

With understaffed, under-equipped, poorly trained and often disloyal Iraqi security forces unable to provide basic security to the local population, with an undersized and disengaged American force, largely alienated from the local population and mainly confined to large base areas, there was literally nothing to check the growth of an insurgency (actually multiple insurgencies) with unrestricted access to the huge stockpiles of arms and ammunition left behind by Saddam Hussein which American forces were too few in number or too distracted to bother to secure. So, for almost four years, security conditions in Iraq were, almost as a matter of administration policy, left to spiral downward into utter chaos.

The main task of the surge was therefore to bring the American force in Iraq for the first time to a minimally adequate size, and, more important, to deploy that force in close and regular proximity to the Iraqi people to provide them with a security environment in which insurgencies could no longer thrive and sustain themselves. When the added American troops began arriving in Iraq in February 2007, the initial indicators of their effects were far from favorable. The new strategy was predicated on sending American troops out of the relative (but illusory) security of their large bases, supposedly insulated from the violence that was overwhelming the rest of the country, to the least secure areas of Baghdad and Anbar province. The new strategy, at least initially, exposed Americans troops to an increased danger of hostile attack by putting them in neighborhoods in which insurgents and hostile militias of various religious and ideological stripes were doing battle with the Iraqi security forces, with each other, and with any American forces in sight, while preying upon and terrorizing the local populations. As more and more Americans were sent into such dangerous territories and neighborhoods, American casualties steadily mounted to some of the highest levels of the war.

The linked figure provides a convenient way of tracking American fatalities resulting from hostile engagements (a slightly different number from the more frequently cited total fatalities, which includes fatalities not resulting from hostile engagement). It plots a moving sum of American hostile fatalities on each day from the start of the invasion and the 15 days preceding and following that day. Each point on the graph represents the total number of hostile fatalities in the 31-day period centered on the corresponding day. For example, on February 1, two Americans died from hostile fire, while in the 15 preceding (subsequent) days 57 (40) died, for a total of 99. The surge began at the peak of an especially violent period in late January and early February (25 hostile deaths on January 20 and 11 on February 7). US hostile fatalities thus abated slightly in the first few weeks of the surge, but the downturn was short-lived as increasing numbers of American troops entered unsecured areas and took exposed positions from which they were obliged to engage insurgent forces to bring those areas under their control.

The combination of additional American troops in Iraq and their deployment in exposed positions in vulnerable areas that had to be wrested from insurgent control caused the trend in the number of hostile fatalities to begin rising steeply in March, a trend that continued into April and May. Figure 1 shows an almost continuous rise in the 31-day moving sum of hostile casualties from the dip at the end of February until mid to late May. From May 11 to May 28, the daily moving sum of American fatalities was not less than 115, and for 41 consecutive days, from May 4 to June 14, the 31-day moving sum was over 100. One way to put that number into perspective is to consider that in the 1736 days since the invasion of Iraq (March 21, 2003 to December 20, 2007), the 31-day moving sum of hostile American fatalities has exceeded 100 on only 143 days, of which 14 came in the early days of the invasion. Before 2007, the longest stretches in which the 31-day moving sum of hostile fatalities exceeded 100 were from March 27 to April 24, 2004 (28 days) and from November 1 to November 27, 2004 (27 days).

The progress achieved by the surge did not come cheaply. The period from early May to mid-June was the longest period of intense fighting by American forces of the entire war. It saw the loss of 143 Americans troops between May 4 and June 14. But unlike the previous four years in which American losses were squandered by a feckless leadership with no coherent strategy, the sacrifices of this spring were made to implement a strategy that was reversing the trajectory of the conflict. The spikes in violence in the spring and autumn of 2004 were initiated by Iraqi insurgents, and the resulting heavy losses were taken by an American force on the strategic defensive. In the spring, however, it was American forces that, by taking control of unsecured neighborhoods in Baghdad and elsewhere in Iraq and providing security to the people in those neighborhoods, were finally on the strategic offensive. The insurgents did not relinquish control over those neighborhoods without a fight. But the superior firepower, training, and determination of the American forces overwhelmed the insurgents, while the decency and reliability of the American troops, finally coming into meaningful and regular contact with the local population, contrasted so starkly with the wanton brutality of the insurgents that the local populations soon realized that an ongoing American presence in their neighborhoods was their best hope for restoring a tolerable way of life.

Nor can one emphasize too strongly that the primary strategic goal in these engagements was not to kill terrorists but to secure the population. Killing terrorists without providing security to the population is a prescription for defeat in fighting against an insurgency.

Just as proponents of a counterinsurgency strategy had predicted, the American forces, by establishing an ongoing (and hence reliable) presence in an area, and providing security to the local population against insurgent attacks, began to elicit the cooperation and intelligence from the local population necessary to anticipate and frustrate insurgent attacks on both the population and themselves. Before such cooperation and intelligence could be obtained, the local population had to be convinced that the security forces would indeed stay as long as needed to protect the population from insurgent violence and reprisals against those cooperating with the security forces. The increasing cooperation between the local population and the counterinsurgents and the increasing quantity and improving quality of the intelligence received tends to marginalize the insurgency as they suffer attrition and lose active and tacit, —and unwilling, support that allows them to conduct their operations. The success of the counterinsurgency strategy, if it is sustained, triggers a virtuous cycle that gradually saps the vitality of the insurgency, forcing them to move further and further from population centers and increasingly isolates them from the population and from each other. The best indicator of a successful counterinsurgency is, therefore, not the insurgent body count, but a decline in the overall level of violence.

So even as American troops emerged from their seemingly secure bases (actually isolated enclaves from which troops could venture only at high risk) into vulnerable positions among the Iraqi population, they were degrading the capacity of insurgents to inflict damage on the Iraqi population and thereby on themselves. For example, the IEDs that took such a large toll of American forces, are now a much diminished threat thanks largely to cooperation and intelligence obtained from the local population. The best way to reduce the number of deaths by IEDs was not to add more armor to Humvees (as if insurgents could not add more explosives to the IEDs), but to gain more and better intelligence about who was making them and where they were being placed. This is just what General Petraeus meant when he wrote in the Counterinsurgency Manual "sometimes the more you protect your forces the less secure you are."

Similarly, it is an improved security situation on the ground, not diplomacy or threats to neighboring countries, that is reducing the flow of fighters and supplies into Iraq from the outside. That flow is demand-driven by conditions in Iraq, not supply-driven by external sources. External sources may, for a variety of reasons, be happy to cooperate with, and provide material support, to the insurgents. But those external sources are mainly responding to the demands of insurgency. If the ability of the insurgency to sustain itself is degraded, support from the outside cannot keep it going. As long as security conditions inside Iraq continue to improve, the flow of material from outside to support it will dwindle correspondingly.

What is remarkable in retrospect is how quickly improvements in the security environment in areas where American troops established an ongoing presence reduced hostile fatalities even as newly arriving American troops were entering other areas that had been effectively conceded to insurgent control. American fatalities peaked in the third and fourth weeks of May, about four weeks before the full complement of additional troops under the surge had been deployed. At the peak of the fighting, from May 16 to May 23, the 31-day moving sum of hostile fatalities was at least 119. The 31-day moving sum then began to decline almost without interruption, reaching 119 again only on June 2 and dropping below 100 on June 14, just as the final reinforcements under the surge were arriving in Iraq.

The first clear indicators of an improved security environment became visible in July. US forces, having reached full strength under the surge, began a multi-pronged offensive centered on Diyala province northeast of Baghdad. Diyala had become the center of Al-Qaeda activity in Iraq after the Anbar awakening earlier in the year had, with the help of a Marine battalion, largely ousted Al-Qaeda from that formerly hospitable province. But in contrast to previous offensives against Al-Qaeda, this one was also simultaneously directed at escape routes and potential safe havens to the north and south. Although the Diyala offensive occasioned some intense engagements between American and Al-Qaeda forces, the battles, surprisingly, did not reverse the downward trend in US casualties that had started a month earlier.

Critics of the surge at first tried to discount the reduction in American casualties in July, despite the additional American troops that had taken increasingly exposed positions among the Iraqi population, by attributing the reduction to seasonal factors. July is indeed the hottest month in the brutally hot Iraqi summer, and in each previous year of the war, the number of hostile fatalities in July had been relatively low. Although one could not absolutely dismiss the notion that it was the July heat rather than an improving security environment that was causing a drop in hostile fatalities, it was inherently implausible that an average temperature in July two or three degrees higher than that in June would be responsible for the continuing decline in hostile fatalities that had begun after the rate of hostile fatalities peaked in late May.

The July theory proved increasingly untenable as the Diyala offensive continued through July and early August. There was a short pause in the downward trend of hostile fatalities in the second half of July, but the trend resumed in August. By the time General Petraeus reported to Congress in September, he was able to cite a substantial reduction in American casualties as one indicator of an improving military situation. Other, less precise, measures of the security situation, such as the number of Iraqi civilian casualties, could already be seen to be following a downward trend broadly similar to that evident in the number of hostile US fatalities.

Since General Petraeus testified before Congress, the rate of hostile American fatalities, mirroring other measures of violence including Iraqi civilian casualties, continued to decline until it has reached the lowest levels since the first year of the war. In both October and November only 29 Americans were killed as a result of hostile fire. The 31-day moving sum of hostile American casualties has been less than 40 for 75 consecutive days since September 21 with no sign of an uptick. Indeed, the 31-day moving sum on December 4 was just 14, a level not seen for any 31-day period since February 24, 2004 when it was 13. Since the winter of 2003-04, the 31-day moving sum of hostile fatalities has not remained under 40 for more than 31 days in the early spring of 2005 (March 9 -- April 8) and the late winter of 2006 (February 22 -- March 24).

The unexpectedly rapid improvement in the security environment is, of course, in large part a testament not just to the successful implementation of a classic counterinsurgency strategy by General Petraeus, but to the singular ineptitude, combined with the unequaled brutality, of Al- Qaeda. Having exploited the incompetence of the American leadership for almost four years, Al-Qaeda returned the favor by so alienating the broad mainstream of Sunni society that most Sunnis came to perceive the American forces as their best hope for protection against Al-Qaeda fanaticism, brutality, and venality on the one hand and Shia retribution on the other.

With the refutation of the July theory of reduced hostile fatalities, critics of the surge seized on yet another alternative explanation for declining violence in Iraq, namely, that the ethnic cleansing of mixed neighborhoods that has been going on for the past two or three years has finally run its course. The idea once again is that the surge changed nothing, it just happened to coincide with other trends that were working independently. But even if ethnic cleansing was a factor affecting the overall level of violence, though its importance was probably overstated by critics of the surge, it could hardly have accounted for more than a fraction of the total number of Iraqi civilian casualties and, especially, of the total number of American hostile fatalities. Moreover, having been deployed once to explain away the rapid and sizeable reduction in violence in the summer, the supposed completion of ethnic cleansing is no longer available to explain the rapid and sizeable reduction in violence that has continued since the summer. The process of ethnic cleansing can come to an end only once.

Critics say that we have heard about improved security before and that the earlier optimism was unfounded. But previous improvements were brief and episodic (usually lulls after a severe outbreak of violence) and followed by a ratcheting up of violence to an even higher level. We have now witnessed an almost unbroken trend of reduced violence and improved security that has lasted for seven months, far longer than any previous hiatus in the violence. It is sometimes said that there was never a doubt that American troops would achieve any military objective that was asked of them, so that the improved security environment achieved by the surge is of little or no consequence. However, the problem was never that American troops could not accomplish the objectives that were asked of them. The problem was that their civilian and military commanders had no conception of what objectives had to be accomplished. With the departure of Secretary Rumsfeld and the arrival General Petraeus, that is no longer the case.

Critics also complain about the lack of political progress. Certainly, political accommodation among hostile elements of the Iraqi population is vital to achieving any acceptable outcome in Iraq, and their failure so far to arrive at any mutual understanding is a source of continuing frustration and remains the chief danger to a successful outcome in Iraq. However, critics are proving to be as short-sighted and narrow-minded, in their own way, about conditions in Iraq as the administration was before finally adopting General Petraeus's counterinsurgency strategy.

For nearly four years, the administration's attention was riveted on formal (and largely meaningless) political achievements and symbolic military milestones: the toppling of Sadam Hussein, the killing of Uday and Qusay Hussein, the capture of Sadam, the transfer of power from the provisional authority to the interim government, the ratification of a constitution, and the election of a national assembly. All of these events were hailed as harbingers of a reduction of violence, the end of the insurgency, and the creation of a democratic government. In fact none of these discrete events had any lasting effect on the underlying security situation, which continued to deteriorate after each milestone.

By focusing on specific discrete political benchmarks that so far have not been achieved, critics of the surge are similarly ignoring the new reality emerging in Iraq, a reality more important than meeting externally imposed benchmarks. That new reality is a gradual grass-roots accommodation between the local population and the American forces, which, in turn, is fostering cooperation between the disparate elements of Iraqi society. An improving security environment is a necessary, but not sufficient condition for such cooperation and for an ultimate accommodation between the main ethnic and religious groups contending for power in Iraq. Iraqis will have to learn to live along side each other in peace before they can reach a political accommodation about the kind of political arrangements they want to live under.

In its arrogance and ignorance, this administration actually thought that it could dictate the structure and composition of a sustainable democratic regime in Iraq. That was a foolish and puerile conceit, for democracy requires habits of conduct that must be deeply ingrained in the culture of a people, before any set of formal institutions delineating a democratic form of government can be sustained over time. Such habits have to be cultivated with care and dedication until those habits become part of the instinctive repertoire of behavior of the vast majority of those who participate in the public life of the people.

Those habits cannot be cultivated in the war of all against all into which Iraq had descended and from which it is only now beginning to emerge. For critics of the surge to expect that the power blocks that emerged from the political process imposed on Iraq by the American occupying force to reach an agreement that each block may rationally view as devastating to its own vital interests is no less fanciful than the administration's notion that it could simply prescribe a set of democratic institutions for a non-functioning Iraqi polity that was certainly unready and unable, and probably unwilling, to nurture and sustain those institutions.

The undoubtedly correct notion that a successful counterinsurgency strategy requires a political as well as a military component does not mean that the political component -- the reconciliation of hostile and mistrustful ethnic and religious factions and an understanding that their differences will be resolved peacefully through a political process that all sides regard as legitimate -- can be imposed, on a predetermined timetable, from the outside. Rather, the military component must be oriented towards providing the essential security preconditions that must be achieved before the main factions of society can arrive at their own political accommodations based on their own judgments of what compromises are tolerable and what agreements are credible. The chances for an enduring reconciliation are not enhanced if the compromises and agreements designed to achieve reconciliation are dictated and imposed by an outside power. The reconciliation that we all fervently hope for will be far more likely to endure if it results from compromises and agreements that are worked out by the contending factions themselves, not under duress exerted by an outside power. Especially after the horrific factional violence of the past three years, the pace of reconciliation cannot easily be determined from the outside to suit the political exigencies of the outside power that is seeking to reconcile the hostile factions. In these very difficult circumstances, patience is almost surely a virtue.

The security environment in Iraq is now better than it has been at any time since the inception of the insurgency in late 2003 and early 2004. This means that after three lost years in which the situation in Iraq was allowed to spin almost irretrievably out of control, the combination of an inspired and inspiring new commander and an incompetent and overreaching enemy has brought us to the cusp of a moderately tolerable outcome in Iraq. It would be unbearably tragic if the unwillingness on the part of this chief executive and his first secretary of defense to acknowledge the reality that we were failing in Iraq and that a coherent strategy was needed to avoid defeat would be followed in turn by a corresponding unwillingness on the part of opponents of this administration to acknowledge the extraordinary progress that has been made on the ground in Iraq in 2007 and that success may now be well within our grasp if we can only muster the patience and the fortitude to allow the strategy that has brought us to a place that a year ago seemed hopelessly out of reach to bear its ultimate fruit.

David Glasner, is an economist in Washington DC and the author of Free Banking and Monetary Reform, Politics Prices and Petroleum.


1. Surprisingly, but actually perhaps not so surprisingly, the acknowledgment drew a notable demurral from the Vice-President who, imprisoned in his own version of reality, publicly dissented from the decision to sack his one-time mentor and patron.

2. Any process of withdrawing American troops from Iraq must, therefore, take into careful consideration the delicate psychological balance necessary to ensure that the confidence of the population that they will remain secure is not undermined as a result of withdrawing American troops. A hasty withdrawal that induces the public to lose confidence that they will remain secure can trigger the unwinding of the virtuous cycle of increasing security and increased cooperation between the public and the security forces, thereby providing the insurgency with an opportunity to reestablish itself.

3. The glimmerings of the kind of halting steps toward mutual reconciliation that I am referring to were highlighted by David Ignatius in a recent (December 19, 2007) column "Skirting the Abyss in Iraq." His concluding paragraph echoes my point. "Crocker and other U.S. officials don't talk about reconciliation as an end state but as a process. As security improves, they say, so do the local economy and the government's ability to provide services. They hope to see an upward spiral, with an increasing return to stability and order. Just as no Iraqi wanted to be the last to abandon what appeared to be a sinking ship, neither will any want to be the last to clamber back aboard."

4. A good example of how the improving security environment is causing a change in how Iraqis are relating to each other on a persona level is provided a report in the Financial Times 16 December 2007 "Baghdadis Enjoy the Moment by Steve Negus in Cairo and an Iraqi correspondent in Baghdad.

"Militia members themselves apparently feel compelled to go on a charm offensive, portraying themselves as Iraqi nationalists rather than sectarian killers.

"Hassan", a 34-year-old member of the Shia Mahdi Army, says he is now trying to woo Sunni refugees to return home.

He joined the militia after his younger brother, also a militiaman, was killed by gunmen, and for much of last year patrolled the neighbourhood stalking Sunni raiders and ejecting Sunni from houses to make way for incoming Shia refugees.

But now he says he wants the Sunni to come back: "Everything has changed. We are now for reconciliation." His Mahdi Army cell recently pitched in to repaint a Sunni mosque that they had previously bombarded with rocket-propelled grenades.

David Glasner, is an economist in Washington DC and the author of Free Banking and Monetary Reform, Politics Prices and Petroleum.


Discuss at Small Wars Council



Tue, 12/25/2007 - 10:48am

Yes, while not attempting to speak for Dr. Glasner and as a matter of what lessons not to learn, it is necessary as some of the neo-con crowd who couldn't punch their way out of a Phase IV paper bag are still out there - revising history via the op-ed pages of mainstream news media and books.

Your wishing away the serious missteps of the administration in the planning and execution of the early and mid stages of OIF is, quite frankly, ludicrous.

Moreover, some of the same who would have been fired - say, during WW II - and received our countrys highest civilian honor - the Medal of Freedom - are in the process of being "rehabilitated" with "administration" consultant positions. So in this case - digging is not only appropriate - it is patriotic and should serve as a harbinger for and on those that might like to follow in the arrogant footsteps of their predecessors.

Is it really necessary to dig out those tired old rants about 'this administration' and 'this chief executive' before admitting the success of its/his latest initiative in Iraq? Perhaps you wish it had happened earlier, but it's at least arguable that more American troops on the ground would simply have resulted in higher casualties, given that in 2004 the Iraqis had not yet suffered the consequences of their adoption of a terrorist insurgency, whether native or imported. No doubt you live in a community (as I do) where 'this chief executive' is rarely called such, but by a bullying semi-expletive of which 'idiot' is one of the kinder versions. Still, adopting that mindset does not inspire great confidence in your conclusions. Blue skies! -- Dan Ford

Here's another cut at using averages:

As we know, statistics are very sensitive to start-stop points and so on. So from the data we have, how about a comparison by half-year? Start with either Jan. '03 or March '03.
6 months ending:
Calendar | From Onset
2003/06 134 | 2003/08 178
2003/12 189 | 2004/02 196
2004/06 311 | 2004/08 359
2004/12 408 | 2005/02 405
2005/06 309 | 2005/08 333
2005/12 367 | 2006/02 335
2006/06 293 | 2006/08 301
2006/12 410 | 2007/02 463
2007/06 529 | 2007/08 502
2007/12 239 | 2008/02 117 (to date)
Total: 3189 3189
6-mo. Ave: 318.9

Ist half of year averages:
315.2 | 334.6

2nd half of year averages:
326.5 | 307.1
6-mo averages excluding '07:
~303 | ~321

1st half excluding '07:
~262 | ~293

2nd half excluding '07/'08
~344 | ~350

So: the 1st half of calendar '07 exceeded the previous average by ~267. The second half is, so far, below the average by ~105; there is the rest of December to make up the difference. Any shortfall will represent a 6-mo total below the average of the 2nd halves of the year. Compared to the previous overall 6 mo. average, the 6-mo calender 2nd half of '07 is 79 lower.

For the 6-month periods since onset, there is longer to make up for the final period, but the 1st half of '07/08 is ~209 above the previous averages. The second half is 233 below the previous average, with under 3 months to make up the difference. If the current total (last 3 months+) doubles, the total will be 234, 116 below the previous 2nd half average. If the ~29/month number from Oct/Nov holds, the total will be 187, or 163 below the previous average, and as of this writing it looks like December will be about 20.

So two alternate predictions come out of this.

If you believe this is a fluke, not a trend, you need 105 more fatalities in December, and 233 before the end of February to bring the current 6-month averages up to snuff.

If you believe the numbers show a trend, you'd expect about 3 more fatalities by the end of December, and perhaps 30 more total by the end of February.

So the differences in predictions are: anti-liberationists need 105 combat deaths by the end of December; liberationists expect no more than 3, a difference of 102.

Anti-liberationists need 233 US combat deaths by the end of February to bring things back on track, and liberationists expect no more than 30, for a difference of 203.

Place yer bets, ladies and gentlebongs!