On September 16, General Raymond Odierno will succeed General David Petraeus as commander of U.S. and coalition forces in Iraq. The surge strategy Petraeus and Odierno developed and executed in 2007 achieved its objectives: reducing violence in Iraq enough to allow political processes to restart, economic development to move forward, and reconciliation to begin. Violence has remained at historic lows even after the withdrawal of all surge forces and the handover of many areas to Iraqi control. Accordingly, President Bush has approved the withdrawal of 8,000 additional troops by February 2009.

With Barack Obama's recent declaration that the surge in Iraq has succeeded, it should now be possible to move beyond that debate and squarely address the current situation in Iraq and the future. Reductions in violence permitting political change were the goal of the surge, but they are not the sole measure of success in Iraq.

The United States seeks a free, stable, independent Iraq, with a legitimately elected representative government that can govern and defend its territory, is at peace with its neighbors, and is an ally of the United States in the war on terror. The Iraqi leadership has made important strides toward developing a new and inclusive political system that addresses the concerns of all Iraq's ethnic and sectarian groups. But it has also taken steps in the wrong direction. An understandable desire to seize on the reduction in violence to justify overly hasty force reductions and premature transfer of authority to Iraqis puts the hard-won gains of 2007 and 2008 at risk. Thus, the president's announcement of new troop withdrawals has come before we even know when Iraq's provincial elections will occur.

Reducing our troop strength solely on the basis of trends in violence also misses the critical point that the mission of American forces in Iraq is shifting rapidly from counterinsurgency to peace enforcement. The counter-insurgency fight that characterized 2007 continues mainly in areas of northern Iraq. The ability of organized enemy groups, either Sunni or Shia, to conduct large-scale military or terrorist operations and to threaten the existence of the Iraqi government is gone for now. No area of Iraq today requires the massive, violent, and dangerous military operations that American and Iraqi forces had to conduct over the last 18 months in order to pacify various places or restore them to government control. Although enemy networks and organizations have survived and are regrouping, they will likely need considerable time to rebuild their capabilities to levels that pose more than a local challenge--and intelligent political, economic, military, and police efforts can prevent them from rebuilding at all.

American troops continue to conduct counterterrorism operations against Al Qaeda in Iraq, which has not given up, and against Iranian-backed Special Groups, which are also reconstituting. U.S. forces support Iraqi forces conducting counterinsurgency operations in the handful of areas where any significant insurgent capability remains. But mostly our troops are enforcing the peace.

In ethnically mixed areas, American troops are seen as impartial arbiters and mediators. In predominantly Shia or Sunni areas, they are seen as guarantors of continued safety, destroying the justification for illegal militias. American brigades also play critical roles in economic reconstruction, not by spending American money but by helping Iraqis spend their own money. American staffs help local Iraqi leaders develop prioritized lists of their needs, budgets to match those priorities, and plans for executing those budgets. American troops support the Provincial Reconstruction Teams that mentor Iraqi provincial leaders and help local communities communicate their needs to the central government. American soldiers provide essential support to Iraqi soldiers and police working hard to develop their ability to function on their own.

Indeed, American combat brigades have become the principal enablers of economic and political development in Iraq. When an American brigade is withdrawn from an area, there is nothing to take its place--all of these functions go unperformed. Clearly, then, the number of brigades needed in Iraq should be tied not to the level of violence but to the roles the Americans perform and the importance of those roles to the further development of Iraq as a stable and peaceful state.

But American brigades do more than that. They also give us leverage at every level to restrain malign actors within the Iraqi government and to insist that Iraqi leaders make concessions and take political risks they would rather avoid. The notion, popular in some American political discussions, that withdrawing our forces increases our leverage is nonsensical. The presence of 140,000 American troops on the ground in Iraq requires the Iraqi leadership to pay attention to America's suggestions in a way that nothing else can. Every brigade that leaves reduces our leverage just when we need it most.

For all the progress made to date, the next president will face significant challenges in Iraq. In recent testimony, Secretary of Defense Robert Gates enumerated them: "the prospect of violence in the lead-up to elections, worrisome reports about sectarian efforts to slow the assimilation of the Sons of Iraq into the Iraqi security forces, Iranian influence, the very real threat that al Qaeda continues to pose, and the possibility that Jaysh al-Mahdi could return."

The existence of malign sectarian actors in the Iraqi parliament and in the prime minister's inner circle is not news. Nor is it news that Iraqi politicians, elected under a closed-list system that emphasized ethnosectarian identity at the expense of political interest, have weak electoral bases and much reason to fear the results of open and honest elections. It is similarly well known that Iran seeks to drive the United States out of Iraq and has been putting tremendous pressure on Iraq's leaders to obey Tehran and reject Washington. These three factors help explain the development of significant negative trends in Iraq in recent months: the downward spiral of negotiations over the Strategic Framework Agreement, delays in the passage of an electoral law, escalating tensions along the Arab-Kurd border, and Iraqi government attacks on certain Sons of Iraq groups in and around Baghdad.

American errors have contributed to these developments. At the outset of negotiations over the Strategic Framework Agreement, for instance, we should have offered Iraq a security guarantee. Iraq's signing a Strategic Framework Agreement would have openly and publicly committed themselves to the United States--and against Iran, in the zero-sum thinking of Tehran. It was only reasonable that Maliki and others in the Iraqi government should have expected an American commitment to match their own, and we should have given it to them. But American domestic politics made that impossible.

Leading congressmen and senators insisted that a security guarantee would raise the Strategic Framework Agreement to the level of a treaty requiring Senate ratification--which is true. They also made clear that no such ratification would be forthcoming if the document bound the next administration. The Bush administration therefore had to tell Baghdad at the outset that America would not match the commitment we were asking the Iraqis to make with an equal commitment of our own. American domestic politics also prevented the administration from placing the security agreement in the larger context of a U.S.-Iraqi strategic partnership, since that concept was ridiculed by those who refused to accept the possibility of success in Iraq.

The Iranians sensed an opportunity and responded with a massive public information campaign in Iraq and a virulent private campaign to put pressure on Iraq's leaders. America's refusal to offer a long-term security guarantee gave weight to the constant Iranian refrain that Iran will always be there, while America will ultimately leave Iraq to its fate. Shrewdly refusing to admit the degree of direct Iranian pressure, Maliki and his associates used the cloak of "Iraqi sovereignty" to conceal their uneasiness at taking responsibility for making a deal with the United States--uneasiness not before their own people, but before Tehran. As a result, the negotiations have dragged on, Iraqi demands have increased, and it is possible that Maliki will now wait until after the American election to see who wins--all because domestic political constraints prevented the Bush administration from making the necessary opening bid.

Maliki has been using "Iraqi sovereignty" to do more than delay those negotiations, however. He has also used it to insist on the accelerated transfer of Iraq's cities, especially Baghdad, to Iraqi control and the withdrawal of American forces from those cities. As a result, the problems that premature transition can cause are on display in the city of Baquba, the capital of Diyala Province northeast of Baghdad.

Diyala has always been one of the most challenging provinces in Iraq because of its swirling mix of Kurds with Sunni and Shia Arabs and its proximity to Baghdad. It served in the past as a staging area for Shia militias and al Qaeda terrorists launching attacks in Baghdad. It was pacified in 2007 with a great deal of hard fighting that resulted in the defeat of illegal Shia militias and the capitulation of the local Sunni insurgent groups, many of whom joined the Sons of Iraq, volunteer security forces organized and initially paid by the United States. More remained to be done in Diyala as the surge ended, however. Surge operations had cleared Baquba and areas further east, but not the rim of the province from Khanaqin along the Iranian border and then through Balad Ruz toward Baghdad. The end of the surge meant the withdrawal of significant American forces from Diyala, so U.S. troops largely turned responsibility for the city of Baquba over to the Iraqis and moved out to clear the peripheral areas of the province.

Rumors began circulating that the Iraqi government believed it would have to re-clear Baquba, even though violence remained low and American leaders did not agree. In August 2008, the Iraqi security forces, with limited support from American troops, did re-clear the city--but their targets were primarily leaders in the Sons of Iraq movement and members of the local government and community that had supported them. This action--which could not have taken place if American forces had continued to patrol the city--was part of a larger effort by Maliki to weaken the urban Sons of Iraq. It appears that the current Iraqi leadership has recognized that it must allow the Sunni tribal movements, particularly in Anbar, to organize and gain power in their own communities, but it sees the urban Sons of Iraq movements as political threats to its power.

The return of the Sunni Iraq Islamic party (IIP) to the government appears to have created an unholy alliance between Maliki and IIP leader (and Iraqi vice president) Tariq al-Hashimi aimed at weakening grassroots Sunni political movements in and around Baghdad and ensuring that the unpopular and unrepresentative IIP continues to wield power after provincial elections. A similar alliance is operating in Ninewa Province, where Kurdish leaders appear to have joined with the IIP to ensure that they will continue to have influence in the largely Arab province when provincial elections eliminate the current disproportionate Kurdish sway in the provincial government. This Kurdish-IIP alliance helps explain why there are virtually no Sons of Iraq in Ninewa. The extremely limited American presence in Ninewa, as in Baquba, has enabled these developments, which may call into question the legitimacy of the upcoming provincial elections in some areas.

Maliki's actions may reflect the continued powerful influence of malign sectarian actors among his advisers, or it may reflect the determination of a temporarily strong political leader confronting elections that are likely to weaken his base. The specter of Iranian power combines with the enormous question mark hanging over the future of American support to make Maliki look to his own resources to stabilize his position. Again, contrary to conventional wisdom, the threat of American withdrawal and America's refusal to guarantee the security of Iraq and its constitutional processes presses Iraq's leaders to make bad decisions, not good ones.

Whatever Maliki's motivations, however, the bottom line is clear. Although a dramatic increase in violence or the rebirth of a large-scale Sunni insurgency in the next six months is unlikely, it is possible that American policies are combining with Iraqi mistakes to undermine the long-term prospects for success. These trends can be reversed, with care, over the coming months if the United States can summon some strategic patience.

There is no question that we should be able to start withdrawing significant numbers of American forces from Iraq in 2009 and accelerating our withdrawal in 2010. Assuming that Iraqi provincial elections in 2008 or early 2009, and parliamentary elections in 2009 or 2010, are accepted as legitimate by the Iraqi people and the international community, it is also highly likely that we can continue to withdraw from Iraq's cities, including Baghdad, and move from a patrolling role to an advisory and support role in the same period. But the timing of force reductions and withdrawals from urban areas is critical, and the current pace is too fast.

It appears from media reports that General Petraeus initially proposed no reduction in the number of U.S. brigades below the pre-surge levels, and that was certainly the right recommendation. Current force levels may, in fact, already be too low. At all events, we must see Iraq through the upcoming two elections, pressing the government to conduct them fairly and inclusively as well as ensuring that enemy groups do not disrupt them with violence. Doing so requires a significant American presence on the ground in Iraq's population centers, where, in addition to all the other key non-combat roles they play, American soldiers are the canaries in the mine shaft. They know before anyone else when Iraqi leaders at any level are starting to play games that can undermine mission success.

We should therefore not withdraw any brigades from Iraq before the provincial elections have occurred and the results have been certified and accepted. We should not accept timelines for the departure of American troops from Iraq's cities, particularly Baghdad, before the parliamentary elections of 2009. We should continually press the Iraqi government not simply to pay the Sons of Iraq (as it has announced it will do beginning in October), but to bring most of them into the political process. Some of the Sons of Iraq were leaders of the insurgency and should have no place in Iraqi politics, but in its Baquba operation, the Iraqi government was not sufficiently discriminating in whom it sought to exclude (much less detain). We must also support the Iraqi government in its efforts to push Kurdish militias out of Diyala and Ninewa provinces.

This is not a matter of Iraqi sovereignty. American troops will not stay anywhere in Iraq if ordered by the Iraqi government to leave. We are not going to depose Maliki or retake control of Baghdad. We are not going to force the Iraqis to do anything. And, above all, we are not going to maintain a large military presence in Iraq indefinitely. But we are engaged in continual negotiations with the Iraqi government about what our forces will do and what Iraqi forces will do, and we have tremendous leverage in those negotiations.

For too long, we have allowed domestic American political considerations to reduce our leverage and weaken our bargaining position, and we have refused to recognize the critical role the presence of our combat forces plays in keeping us in the game at all. When America provides combat forces to maintain internal or external security in a foreign state, it acquires the right to bargain hard for what it thinks is best for the common interest, even when the host state's government does not agree. We have engaged in such hard bargaining in South Korea and in Europe, and it is a normal part of alliance relationships. We must bargain harder in Iraq and give ourselves the tools and leverage we need to succeed.

Above all, we must recognize that there is never a glide path in war. As long as the outcome remains in doubt, we must never imagine that the situation is under control and we can put it on autopilot and ignore it. The relief of getting Iraqi violence under control and American casualties down turns naturally into a desire to declare victory and withdraw. That is a danger to be avoided at all costs. This administration must ensure that it hands its successor not only a relatively peaceful Iraq, but an Iraq that is headed in the right direction.