When assessing progress in Iraq, most Americans focus principally on the counterinsurgency campaign and related political violence. This is understandable, given the number of American and Iraqi casualties the conflict continues to cause. But it should not be our paramount worry in Iraq today.

Coalition and Iraqi security forces will ultimately defeat the rejectionist remnants of the Ba’ath Party, as well as foreign terrorists who have entered the country. These dead-enders are few in number and have little ability to inspire a broader following among the Iraqi people. The American people are not likely to waver in their willingness to tolerate U.S. casualties in the process of defeating the remnants of Saddam Hussein’s regime as well as affiliates of al Qaeda so long as the resistance appears geographically confined and numerically limited.

The bigger and harder question concerns the transition to Iraqi sovereignty this summer. The current U.S. plan for creating an interim Iraqi government now seems untenable. That plan would first lead to a provisional assembly chosen through an elaborate caucus system. The assembly would then select a provisional government to run the country until a new constitution can be written and then general elections held in 2005.

But this plan seems increasingly unpopular in Iraq. The influential Shi’ite cleric, Ayatollah al-Sistani, appears strongly opposed to the caucus approach. Largely as a result, as many as 100,000 Shi’ites have recently protested the plan in demonstrations in Baghdad.

Any U.S. plan for creating an interim government that is unacceptable to a majority of Iraqis could lead to disaster. If we lose Iraqi hearts and minds in this way, the insurgency could broaden. Then American public opinion could turn against the operation, leading to its premature termination and its failure.

The United States and its partners need to find a way to create a provisional Iraqi government perceived as legitimate, authentic and representative. The Iraqi population has two prerequisites for their future government. First, most Iraqis, especially among the Shi’ite majority, want direct elections. Second, many Sunnis and Kurds want protection from the “tyranny of the majority”—in particular, the Shi’ite majority.

These goals are not unattainable or mutually exclusive, as some U.S. officials claim. Direct elections are feasible by the end of June, which is the U.S.-imposed deadline for transferring power. There is a distinct possibility that Kofi Annan’s U.N. delegation to Iraq will soon reach that very conclusion, making the standing U.S. caucus plan even more unacceptable to Iraqis than it is right now.

Even though proper voter rolls in Iraq do not exist at present, they are not absolutely needed, as the first post-apartheid elections in South Africa proved some years ago. Individuals can be prevented from voting twice by being stamped with indelible ink after their first trip to the ballot box. They can also be verbally tested for fluency in Arabic to ensure they are not Iranian; a few foreigners may still get into the voting booth this way, but not enough to measurably affect the outcome. Ditto for the possibility of a few underage voters. And as for the security risk incurred by holding elections in a strife ridden society, this danger has been successfully addressed in other volatile election environments from Sri Lanka to El Salvador to Kashmir to Cambodia.

As for preventing the tyranny of the majority, three straightforward measures can limit the risk. First, a supermajority vote of about 65 percent of the constituent assembly could be required to make decisions on big issues facing Iraq. Those would include the secular vs. religious character of the state as well as the federal vs. unitary nature of the new government. This provision could be negotiated into the interim constitution or “basic law” being developed among various Iraqi parties this month.

Second, voting in Iraq should not be conducted on the basis of a simple nationwide ballot, but province by province. And each province should be granted equal representation in the provisional assembly. Of Iraq’s 18 provinces, seven are dominated by Kurds and/or Sunnis (and two more are quite mixed). That means that the Shi’ite delegation in a future transitional government—even if fully unified itself, which seems unlikely—would not be able to generate a supermajority vote without at least some support from other groups. This approach, combined perhaps with a three-person presidency including representatives from each major ethno-religious group, could prevent any tyranny of the majority from becoming a major problem in Iraq.

Third, the basic law that the U.S.-led coalition is currently developing with Iraqis could also include various forms of protections for the rights of individuals and provinces. In particular, each could be guaranteed a certain minimum fraction of Iraq’s oil, the Kurdish region could retain its autonomy, and all individuals could be ensured freedom of speech and religion as well as due process under the law.

Much is going well in Iraq today. But our mission there could still fail because of centrifugal ethnic and religious forces. Anti-Americanism could also become the rallying ideology for a much broader guerrilla movement than we have witnessed to date, if Washington insists upon a political transition plan that enjoys limited support and inspires little confidence among the very Iraqi population it is designed to serve.