On Tuesday this week, a high-powered drug war summit convened in Mexico City with Secretary of State Hillary Clinton, Defense Secretary Robert Gates, and Department of Homeland Security Secretary Janet Napolitano in attendance. The good news is that a more effective Mexican anti-drug strategy, supported by U.S. assistance, is finally underway. But the bad news is indeed awful.

Since President Calderon came to office in 2006 and deployed over 45,000 soldiers to Mexico’s streets, violence has continued to increase. More than 18,000 people have been killed over the past three years, with 2009 being the bloodiest year yet. In Ciudad Juarez alone, the city most engulfed by violence, 2,600 people were killed last year.

The narcos’ murders are brazen, brutal, and often senseless. They no longer bother to confirm the identity of their victims before their hits, meaning innocent people die. There have been deep repercussions: many inhabitants have moved out and businesses fearing extortion have closed down.

Other communities are suffering too: Tamaulipas, where violence has spiked in a new turf war between Los Zetas and other drug trafficking organizations; Tijuana, where a truce among the cartels appears to be unraveling; and Michoacán, where the cult-like La Familia is far from eliminated. Other forms of crime, including kidnapping and extortion, have also soared, and the drug-trafficking organizations continue to take over other illegal enterprises and informal economies, posing an even greater challenge to the state.

But despite these discouraging developments–indeed because of them–the strategy of the Mexican government is becoming better organized. Facing an outraged public, the Mexican government can no longer dismiss the violence as bad guys killing bad guys.

The government’s previous policy of “decapitation” focused on eliminating high-value targets–the heads of the cartels–and deploying military patrols to the streets. But it failed at its most basic goal: to boost public safety. Despite the unending parade of captured or killed narcos, the policy actually generated new turf wars among the cartels over territory and corruption networks who tried to capitalize on recently “decapitated” rivals.

Crucially, the Mexican government has realized that it needs a multifaceted response to serious organized crime. It has begun replacing the decapitation strategy with one that focuses more on entire networks, not just their heads. Such a policy requires law-enforcement and intelligence apparatuses that have a robust investigative capacity and are reasonably free of corruption. Mexico currently has neither, but American assistance can help.

Tijuana, for example, provides a model on how to deploy the military in trouble spots. Instead of soldiers, who often lack a mandate or the ability to conduct systematic investigations, newly vetted police units can take the lead in providing safety on the streets and conducting investigations. The military can be deployed in the background to supplement the police if they are subject to intense violence by the narcos.

The state will need to build links to civil society and persuade the population that it can provide them with public goods and social services better than the narcos can. Such bonds between the community and the state are what at the end of the day will allow the state to prevail over the cartels. To this effect, President Felipe Calderón has unveiled a host of social programs oriented toward bringing jobs, education, and public spaces to Ciudad Juarez. This overall strategy needs to be extended throughout Mexico.

Of course, the most effective way for Mexico to combat drug violence is to create jobs that will give people an adequate alternative to the cartels. Given the structural deficiencies of Mexico’s economy, generating sustainable jobs will be hard. The socioeconomic programs cannot be construed simply as limited handouts and buy-offs, but as a systematic strategy to bring marginalized urban communities into prosperity and legality–complex and long-term urban planning.

The new strategy will need commitment and perseverance. Many a good strategy dies in poor implementation. Politically it will be very difficult to concentrate newly trained police units and military backup in particular locales, instead of deploying them thinly throughout all the hot areas. If the latter takes place–driven by the politics of Mexico’s upcoming elections and a natural desire to respond to any area where violence flares up–competent law-enforcement officers will be spread too thin to effectively tackle any problems, and Mexico will simply continue putting out fires without making a systematic difference.

Most importantly, although the state needs to reach out and cooperate with the community; it cannot push civil society to take on the narcos on its own and prematurely. Achieving a strong bond between the society and the state is critical, but if the state asks the society to act and then itself fails to deliver on public safety and protection, community efforts will fizzle. It will be very difficult for the state to mobilize society second time around, and the narcos will score a great victory.