As a policy to curtail the drug supply, "Plan Colombia" has been a total failure. Cocaine and heroin are as cheap and pure as ever in the United States, and coca production in Colombia has been holding steady over the past two years, despite the fact that the government has been dumping herbicides anywhere it can reach. Coca growers, who depend on the crop for their livelihood, have become more innovative in response to the aerial spraying—cultivating smaller, harder-to-detect plots; developing new strains that grow more quickly; and planting in the shade.

And so long as the growers have no other options—alternative development programs are under-funded and reach only a small fraction of rural Colombians—they'll keep on innovating. Not to mention the fact that all that constant spraying makes growers more sympathetic to guerilla groups. Seeing as how the United States is currently in the middle of following a similar strategy to fight opium production in Afghanistan, these seem like lessons worth learning. Doubtful it will happen, though.

Now in 2002, the Bush administration expanded military aid to the Bogota government to help it fight FARC, one of the two leftist insurgency groups. Colombia has seen a few crucial security improvements over the past few years—under President Uribe, kidnappings have dropped by 57 percent, massacres by 71 percent, and murders by 31 percent (at least according to "official" figures). On the other hand, this can't count as an American success; most U.S. military and police aid simply doesn't go towards protecting civilians, so it's hard to credit "Plan Colombia" here. (Isaacson instead credits President Uribe's decisions both to redeploy troops in population centers and to induce the right-leaning paramilitaries to agree to a ceasefire.) And then there's the downside to all that military aid:

Since Plan Colombia's inception, Colombia's attorney general has demonstrated markedly less will to prosecute cases of human rights abuse by the military. The State Department's last human rights certification memo named only thirty-one military personnel… currently under indictment for human rights abuses or support of paramilitaries. This impunity undercuts much of the well-publicized improvements that the Colombian armed forces have made to their human rights training. If a soldier knows he stands almost no chance of punishment for committing an abuse, will the mere knowledge that the crime is wrong consistently prevent him from committing it.

Some numbers suggest that the Colombian government really is acting with more impunity. The share of abuses committed by the Colombian security forces rose from 5 percent in the late 1990s to 7.8 percent in 2003. Killings and disappearances of human rights activists rose from 29 in 2001 and 2002 to 33 in 2003. 340 people were tortured in 2002-03, up from 242 in the previous twelve months.

Nor has Plan Colombia improved the prospects for peace talks; indeed, there's some evidence that the specter of U.S. military aid may have helped scuttle the moderately promising negotiations in 2000 between the government and FARC and ELN by fostering mistrust among the guerrilla groups. And the counterinsurgency campaign isn't proving very effective either: in part because over 80 percent of the U.S. aid is of a military nature, the Colombian government has had the same trouble that coalition forces have had in Iraq—they can seize rebel-held territories, but they can't hold them once they withdraw. Not only that, but should the conflict in Colombia start escalating rapidly, the U.S. could find itself committing more and more resources to a major war in its backyard. Already some American policymakers have been linking FARC to Hugo Chavez in Venezuela, which should make for fun times.

Here's one comprehensive suggestion for a change in strategy in Colombia, which would involve: Less military aid; more support for institutions that bolster human rights. An end to crop spraying; more support for alternative development strategies in neglected rural areas. At home, drug treatment programs would be a much cheaper and actually effective means of reducing the demand for drugs; but this fraction of the $12 billion that the federal government spends each year on the "war on drugs" hasn't changed for a long while. Budget cuts this year focused on hacking up health care for the poor rather than taking even the slightest look at a billion dollar foreign policy adventure that doesn't seem to be achieving much of anything.