The latest Western “victories” over the Islamic State—the U.S. killing of the number two man in the heinous group and the Russian supported retaking of the ancient town of Palmyra, Syria by the Syrian government—should not mask the long-term difficulties of eradicating the Islamist insurgency. Examining these two “achievements” may, however, tell us what is likely to work and what is not.

The two recent developments continue a series of setbacks for the Islamic State—the fall of the towns of Tikrit and Ramadi in Iraq and al-Shadadi in Syria. The U.S. assassination of Haji Iman, the organization’s day-to-day commander and finance minister, will probably have less of an effect than is widely believed. Even U.S. Secretary of Defense Ashton Carter admitted that the senior leadership of terrorist groups can be replaced. And since war is an evolutionary hothouse, an often more ruthless leader, who was tough enough to survive, replaces the slain leader. Experts on counterinsurgency are skeptical that any great power can kill its way out of an insurgency.

The fall of Palmyra seems to belie the U.S. government’s line that President Vladimir Putin of Russia was only trying to weaken pro-U.S. anti-Assad rebels at the expense of taking on the Islamic State. Indeed, the heavy use of Russian air power in support of the Syrian Army seemed to be decisive in recapturing the town. This accomplishment should also make the United States chary about deposing Bashar al-Assad’s Alawite (an offshoot of Shi’ism) government in Syria. The Russians have successfully shored up Assad’s control of the Shi’ite part of Syria. Assad may have a very poor human rights record, but he is better than the Islamic State getting control of all of Syria, something the U.S. government cannot seem to understand.

Yet, even if the Islamic state is defeated in most towns and cities—such as Mosul in Iraq and its capital of Raqqa in Syria—the group likely will then fight strictly as a guerrilla force, requiring years to completely eradicate. And since the Islamic State was motivated to develop its ability to strike in the West—thus creating an international threat out of what had been merely a regional one—only after the U.S.-led Western coalition began bombing it in Iraq and Syria, it is easy to anticipate that the group’s overseas effort would continue, and maybe even accelerate, after the Islamic State loses the cities in its caliphate.

Now that a cease fire exists in the rest of Syria, the U.S. government needs to face the hard realities. Assad needs to stay and rule a rump Shi’ite (Alawite) Syria. To get the Sunnis in Iraq and Syria to drop support for the Islamic State, they need to fear such Shi’ite central governments in Syria and Iraq less than the Islamic State group. Allowing the two countries to break up into autonomous regions, with self-rule by various ethno-sectarian groups, would immediately give the Sunnis in both countries a big incentive to throw out the Islamic State fighters and more importantly would dissolve support for the group among such populations. This option is much better than fighting a bloody long-term counterinsurgency, with the brutal Islamic State group continuing to lash out at Western, including possibly U.S., targets. The decentralization of Iraq and Syria would drain local public support from the insurgency, instead of fueling it— as the current Russian and U.S.-led allied bombing is doing by accidentally or purposefully killing civilians. The international community—stocked full of countries with restive minorities—has always been reluctant to set the precedent of breaking up countries, but to restore stability in this volatile region, the United States needs to insist on this natural solution.