Barely expressed in the NYT’s long story about our use of paramilitary strikes in places we’re not officially at war is a conflict between three choices. The NYT piece describes the first two–a covert war run by CIA and briefed to Congress, or a covert war run by JSOC subject less oversight–as the choice the Administration is currently debating.

The Yemen operation has raised a broader question: who should be running the shadow war? White House officials are debating whether the C.I.A. should take over the Yemen campaign as a “covert action,” which would allow the United States to carry out operations even without the approval of Yemen’s government. By law, covert action programs require presidential authorization and formal notification to the Congressional intelligence committees. No such requirements apply to the military’s so-called Special Access Programs, like the Yemen strikes.

Implicit in the choice, is the question of whether or not we want to partner with the Yemeni government as we launch attacks on extremist groups in the country.

In part, the spotty record of the Yemen airstrikes may derive from another unavoidable risk of the new shadow war: the need to depend on local proxies who may be unreliable or corrupt, or whose agendas differ from that of the United States.

American officials have a troubled history with Mr. Saleh, a wily political survivor who cultivates radical clerics at election time and has a history of making deals with jihadists. Until recently, taking on Al Qaeda had not been a priority for his government, which has been fighting an intermittent armed rebellion since 2004.

And for all Mr. Saleh’s power — his portraits hang everywhere in the Yemeni capital — his government is deeply unpopular in the remote provinces where the militants have sought sanctuary. The tribes there tend to regularly switch sides, making it difficult to depend on them for information about Al Qaeda. “My state is anyone who fills my pocket with money,” goes one old tribal motto.

The Yemeni security services are similarly unreliable and have collaborated with jihadists at times. The United States has trained elite counterterrorism teams there in recent years, but the military still suffers from corruption and poor discipline.

So we are partnering with forces with occasional ties to our enemies, but the Administration fights fully briefing this stuff to Congress for fear it will be leaked?

Partnering with local governments also make them a target for al Qaeda retaliation, effectively setting off a contest between the government and al Qaeda about who does more damage. It seems to me this creates a need for a counterinsurgency strategy–but with a governmental partner that (like the corrupt Hamid Karzai) we don’t particularly want to partner with.

Meanwhile, this expanded secret war always seems to be expanding into places were the absence of real government and civil society creates a haven for extremists.

Which is presumably why the former Ambassador to Yemen suggests we need to do far more to develop government and civil society.

Edmund J. Hull, the United States ambassador to Yemen from 2001 to 2004, cautioned that American policy must not be limited to using force against Al Qaeda.

“I think it’s both understandable and defensible for the Obama administration to pursue aggressive counterterrorism operations,” Mr. Hull said. But he added: “I’m concerned that counterterrorism is defined as an intelligence and military program. To be successful in the long run, we have to take a far broader approach that emphasizes political, social and economic forces.”

Obama Administration defenders say they are–but they emphasize training troops, not investment in things that would lead to civil society.

Obama administration officials say that is exactly what they are doing — sharply increasing the foreign aid budget for Yemen and offering both money and advice to address the country’s crippling problems. They emphasized that the core of the American effort was not the strikes but training for elite Yemeni units, providing equipment and sharing intelligence to support Yemeni sweeps against Al Qaeda.

As I read the article and thought about these issues, I kept thinking back to one of the better responses to the flap over the Time cover showing the mutilated Afghan girl. Richard Sanchez argues that the problem for women in Afghanistan–indeed for all Afghans–arose from a US policy that served to strengthen warlords. He argues that the solution must aim to eliminate the divisions that empower warlords.

The answer to the warlords — and more importantly to what Afghans, especially the young, call “warlordism” — is the economic strengthening of the popular base. Secretary of State Hillary Clinton recently unveiled a program to lure the economic Taliban, that is, fighters who fight mainly for the wage, away from the insurgency with the lure of jobs. In this she has heeded the words of Karl Eikenberry, now the U.S. ambassador to Afghanistan, but formerly the commander of U.S. forces there, who told the House Armed Services Committee in 2007: “Much of the enemy force is drawn from the ranks of unemployed men looking for wages to support their families.”

But Clinton’s proposal threatens to fail by not going far enough. If jobs, preferably involving the construction of basic infrastructure, are politicized and given only to those who quit the Taliban, then those ex-combatants and their families will become targets for retaliation. This would add yet further impetus for violence.

A decade after it was a central topic of debate in the Presidential election, we still haven’t figured out how to “nation build,” how to eliminate the vacuums of power that al Qaeda’s affiliates exploit. And we seem to have little imagination of how to do so outside of the context of militarization which tends to polarize communities in dangerous ways.

The Administration seems focused on whether to conduct such polarizing strikes with or without a discredited partner. But both options, it seem to me, serve to undermine the most powerful alternative to al Qaeda, the development of a credible alternative.

Now, I presume the COIN folks would say that’s precisely what they’re trying to do. But so long as we’re dropping cluster bombs, so long as we’re choosing one corrupt leader over other corrupt alternatives, how seriously can we be trying?