They are surging more troops into troubled areas and pushing them off the remote fortress-like Forward Operating Bases and into neighborhoods where they conduct foot patrols, erect concrete barriers, and establish a street-level sense of security. The situation in Anbar province has improved substantially, and, while the areas around Baghdad remain deeply troubled, there are signs of progress in the capital itself. (Sectarian murders are down two-thirds since January, though deaths from spectacular suicide bombings remain high.)

Where such strategies have worked, the results were achieved in years, not months. The same is likely true of Iraq, so patience remains the order of the day. But while Petraeus has the fundamentals right, there are still reforms that could be implemented to improve the odds of success. During a recent two-week visit with U.S. forces in Iraq, I saw a number of problems that need fixing, starting with the inadequate size of the Iraqi army.

The army is the most effective and nonsectarian institution in Iraq. Although it has its share of woes, its combat performance has been improving, and it is less corrupt than the police. But it's too small. Saddam Hussein kept more

than 900,000 men under arms at the time of the 1991 Gulf war, a figure that had shrunk to fewer than 400,000 by the time of the U.S. invasion in 2003. Today the Iraqi army is only 136,000 strong. (There are another 194,000 police officers and 125,000 Facility Protection Service personnel, but many of them are useless or worse.) There is talk within the Iraqi Ministry of Defense and the American high command of expanding the Iraqi army by 35,000 to 40,000 soldiers a year for the next three years, but even this isn't enough. The army needs to be at least 300,000-400,000 strong. (The Afghan army likewise needs to be expanded, but that's another story.)

Recruiting hasn't been a problem, not when unemployment is 20 percent or more. But it still may make sense to introduce conscription--something that is alien to currently serving American soldiers, all of whom are volunteers, but that has a long history in Iraq and neighboring states. An army in a developing nation like Iraq isn't there merely to fight internal and external enemies. Its mission is also to inculcate a civic religion of nationalism and egalitarianism in its recruits. Germany, Japan, Turkey, and other newly created states in the 19th and early 20th centuries turned the army into a "schoolhouse of the nation." That requires exposing a large percentage of young men to army training and indoctrination, not just a handpicked few. Iraq could usefully emulate their example, even if it does run the risk that, as in those societies, the army could become the ultimate arbiter of political power.

Expanding the army will require more money, and the United States should probably increase its military-assistance spending, but the Iraqis are starting to help themselves as well. This year, for the first time, the Iraqi government is spending more on its security services (some $9 billion) than the amount of U.S. aid ($5 billion), and the Ministry of Defense has managed to spend 90 percent of its budget, which puts it well ahead of the curve in the semi-functional Maliki government. In any case, the cost of draftee soldiers would be less than that of volunteers, since they wouldn't have to be offered as much pay or benefits; they could also be employed with fewer restrictions on where and when they serve.
An expansion of the Iraqi army will also require an expansion of the number and quality of American advisers, which should not be that great a stretch, since even Democrats say they want to continue the advisory effort indefinitely. But while increasing support for the Iraqi security forces--and working hard to promote evenhanded, effective commanders--it is important to resist the temptation to impose our standards in all matters great and small. American advisers may unwittingly hold back the Iraqis in some instances by insisting they conform to the extraordinarily stringent standards of the U.S. armed forces--rules that, in terms of ethical conduct, are probably a good deal stricter than those previously employed by any army sent to quell any major insurgency in the long history of warfare.

The New York Times ran

a front-page dispatch by Alissa J. Rubin on April 22 ("3 Suspects Talk After Iraqi Soldiers Do Dirty Work") reporting that Iraqi troops managed to break up a major terrorist ring in the violent Baghdad neighborhood of Ghazaliya by beating up a captured insurgent. Such conduct is not tolerated in the American ranks, but the Iraqis are fighting for their lives against the most vicious terrorists on earth in a society that has never heard of the Warren Court. It's hardly surprising that they might resort to "third degree" techniques that were in widespread use by American police until a generation ago, and remain commonplace throughout much of the rest of the world. American advisers need to have the leeway to exercise their best judgment--to be able to turn a blind eye to minor abuses without risking court martial, while at the same time remaining vigilant against major abuses. Just as no counterinsurgency has ever been won employing the norms of modern Western peacetime policing, so too it is very rare to defeat insurgents by terrorizing the population into acquiescence--and in any case that is not a strategy that either the United States or its allies could employ.

In walking the line between excessive lenience and excessive brutality, the Iraqi government needs to build more prisons. The U.S. armed forces have been expanding capacity at their two main holding facilities, Camp Cropper in Baghdad and Camp Bucca in southern Iraq (which replaced Abu Ghraib after it was closed last year). The total prison population, which stood at less than 15,000 last year, is now 19,000, and the plan is to expand to 25,000-30,000 by the end of the summer, when the Baghdad security plan will be in full swing. The number of detainees held in Iraqi custody is unknown but estimated at perhaps 20,000. It's not enough. As military analysts Bing West and Eliot Cohen wrote not long ago, "One in 75 American males is in jail, compared to one in 450 Iraqi males." Since, as they note, "Iraq is not six times safer than the United States," the disparity needs to be addressed if Iraq is to become substantially more peaceful.

Part of the answer is to help the Iraqis build more prisons and appoint more judges. The coalition's Rule of Law Project is doing just that by constructing a facility within the Green Zone that will house not only 6,000 prisoners but also the judges who will try their cases. This is necessary to prevent intimidation of judges, which is said to be widespread and which results in "not guilty" verdicts in many cases in which U.S. troops are convinced the evidence is overwhelming.

But it is doubtful that any civilian legal system, much less one as anemic as Iraq's, could cope with the demands of wartime. The obvious answer is selective use of martial law to quell violence, giving authority to sentence insurgents to the same people who are risking their lives to catch them--Iraqi and American army officers. This would, of course, be controversial within Iraq. And since the United States is no longer an occupying power, we cannot impose martial law ourselves, but we could make it one of the points on which we lobby the Maliki government for results.

Next, create a readily accessible national identity database. This is an essential prerequisite for a successful counterinsurgency, yet it has never been implemented in Iraq, because successive American commanders have never thought they would be in the country long enough to pull off a project that might require a minimum of six to twelve months to implement. U.S. and Iraqi troops trying to identify potential insurgents have to rely either on food rationing cards, many of which are out of date and all of which lack biometric data such as fingerprints, or on their own haphazard surveys. Unfortunately, most of the data that U.S. troops amass during their tours of duty is lost when they rotate home. There is no uniform database to share population information among all security forces, American and Iraqi, current and future. Thus it's hard to know if someone stopped at a checkpoint belongs in the neighborhood or whether he is a wanted terrorist from another province. This is the kind of information that any U.S. cop would have available within seconds of a traffic stop, because he would run a check via a wireless computer terminal on the motorist's license plate and driver's license. Security forces need to have this same capability in Iraq. There is talk now within the Iraqi government of implementing such a system, but given the Maliki administration's lack of capacity, nothing meaningful will happen unless the Americans do it themselves.

Another necessity is to go more aggressively after foreign fighters. They comprise a relatively small percentage of the overall insurgency, but they account for a very high percentage of the most grotesque attacks--80 to 90 percent of all suicide bombings, according to General Petraeus's briefing with Pentagon reporters on April 26. These jihadists are of many nationalities, but most infiltrate from Syria. The Bush administration has repeatedly vowed that Syria would suffer unspecified consequences if it did not cut off this terrorist pipeline, but so far this has been an empty threat. The administration has refused to authorize Special Operations forces to hit terrorist safe houses and "rat lines" on the Syrian side of the border, even though international law recognizes the right of "hot pursuit" and holds states liable for letting their territory be used to stage attacks on neighbors. It's high time to unleash our covert operators--Delta Force, the SEALs, and other units in the Joint Special Operations Command--to take the fight to the enemy. They can stage low-profile raids with great precision, and Syrian president Bashar Assad would have scant ability to retaliate. We also need to apply greater pressure to Iran, which continues to support both Shiite and Sunni terrorist groups in Iraq, but that will be harder to do because Tehran is a more formidable adversary than Damascus.

There are some less urgent moves than those above, which still might significantly improve our effectiveness in Iraq. It would be helpful to streamline the U.S. command structure. At the moment there are a bewildering variety of senior headquarters in Iraq: Multi-National Force-Iraq (MNFI, the four-star command in charge of overall strategy), Multi-National Corps-Iraq (MNCI, the three-star command in charge of day-to-day operations), and Multi-National Security Transition Command-Iraq (MNSTCI, the three-star command in charge of training and equipping Iraqi security forces). In addition, many of the functions performed by these military staffs (e.g., economic aid, legal affairs, contracting, public affairs, liaison with the Iraqi government) are also carried out by diplomats at the world's largest American embassy. And that's to say nothing of the parallel, often Byzantine structure of the Iraqi government.

The senior American leaders in Iraq today--General David Petraeus of MNFI, Lieutenant General Ray Odierno of MNCI, Lieutenant General Martin Dempsey of MNSTCI, and Ambassador Ryan Crocker--seem to be working fairly well together, but that wasn't always the case among their predecessors. And in any case, no matter how much goodwill there is at the top, the overlap between staffs can cause needless duplication and confusion. One recently returned army officer who served as an adviser to Iraqi troops complained to me that he was never sure who he was supposed to report to: Both MNSTCI and MNCI had jurisdiction over him.

A more serious problem is that rebuilding projects undertaken by the U.S. Agency for International Development, the State Department, the Army Corps of Engineers, and their assorted contractors have often not been well coordinated with the pacification efforts of American combat troops. That's why so many projects have turned into white elephants--they were built in areas that didn't have a modicum of security. This problem is starting to be addressed by the embedding of the State Department's Provincial Reconstruction Teams within U.S. brigades, but greater efforts should be made to streamline and rationalize operations so as to further the essential principle of unity of command.

An Iraqi version of CORDS (Civil Operations and Rural Development Support) might help here. This was the agency created in 1967 under the leadership of "Blowtorch Bob" Komer (with a young Richard Holbrooke as his aide-de-camp). A veteran of the CIA and the National Security Council, Komer coordinated all civilian pacification efforts in Vietnam. He and his successor, William Colby, reported to the four-star commander of the U.S. Military Assistance Command, Vietnam, thus tying civil and military efforts closely together.

As part of a broader administrative overhaul, it would make sense to put more emphasis on "information operations" and to push these efforts down to lower levels of command. There is widespread agreement within the U.S. military that the war for hearts and minds is essential, and that so far al Qaeda and other jihadist groups have done a more effective job than the United States of competing in the "information battlespace." They are able to get their messages out more quickly and to make a bigger splash. Part of this is due to the natural disparity between a ruthless foe that can lie with impunity and intimidate the press and a democratic government that must tell the truth and not interfere with the free functioning of the media. But part of the disparity is also due to self-inflicted wounds on the part of the U.S. government.

I was stunned to learn in Iraq that leaflets and radio broadcasts need to be approved at the division level, and that press releases need to be approved one step higher, at the corps level. Even more amazing was the revelation that U.S. forces are forbidden to conduct information operations on the Internet--the jihadists' favorite venue--because of concerns at the highest levels of the U.S. government that American propaganda might inadvertently be seen by U.S. citizens browsing the web. Several junior officers told me that they have the authority to call in an airstrike that will kill dozens of people but not the authority to issue a press release. That's crazy. The authority to conduct public affairs and information operations needs to be pushed down to the level of the battalion and even the company, and American commanders at those levels and above need to be graded on their success in engaging in this all-important battleground.

Accountability should extend far beyond information operations, of course. There needs to be a much greater effort to promote good commanders and weed out bad ones. Imagine how poorly the Union would have fared in the Civil War if Lincoln had not cashiered McClellan, Pope, Hooker, Burnside, and numerous other ineffectual generals, while promoting Sherman, Grant, and Sheridan. President Bush has singularly failed to hold his commanders accountable. Lieutenant Colonel Paul Yingling, a veteran of two combat tours in Iraq, rightly complains in the new issue of Armed Forces Journal that, "As matters stand now, a private who loses a rifle suffers far greater consequences than a general who loses a war." Yingling isn't the only one upset by this. I've talked to many serving soldiers who are still fuming over the Medals of Freedom given to General Tommy Franks, Ambassador L. Paul Bremer III, and CIA director George Tenet--well-intentioned men all, but their medals were seen as a reward for failure. There was also a fair amount of grumbling within the ranks when the previous commander of U.S. forces in Iraq, General George Casey, was appointed Army chief of staff notwithstanding the deteriorating security situation on his watch.

Failed commanders ought to be fired or pushed aside, following the example of Major General Lloyd Fredendall, who was relieved after the debacle at Kasserine Pass in 1943 and replaced by George S. Patton Jr. Equally important, those who prove their mettle on the battlefield should be quickly promoted. At the moment, battalion and brigade commanders--the key combat leaders in this decentralized war--cycle through Iraq on their 6-7 month (Marine) or 12-15 month (Army) tours, and then proceed with the normal course of their careers. The successful ones may eventually be rewarded with promotion over their less successful colleagues, but the process will take years to play out. Given that we're at war, it would make sense to modify the peacetime personnel system and to resurrect the 19th-century practice of giving brevet ranks or field promotions to outstanding officers who have proven their merit in combat.

Of course, we could achieve an acceptable outcome in Iraq even without taking some of these steps. Conversely, we could lose even if we implement all of these recommendations. But the more of them we implement, the easier the job will become--and the greater the likelihood of success.

Max Boot is a senior fellow at the Council on Foreign Relations and a contributing editor to THE WEEKLY STANDARD. He is author most recently of War Made New: Technology, Warfare, and the Course of History, 1500 to Today.