In the past week or so, we’ve seen the emergence of what must be the most curious — okay bizarre — argument yet about why the United States should remain in Afghanistan.  According to Jari Lindholm and Joshua Foust, both normally thoughtful regional experts, the war in Afghanistan is crucial to prevent a nuclear war between India and Pakistan.

Foust writes:

The big danger, as it has been since 1999, is that insurgents, bored or underutilized in Afghanistan, will spark another confrontation between India and Pakistan, and that that confrontation will spillover into nuclear conflict. That is worth blood and treasure to prevent.

Lindholm argues:

Without access to the training grounds of Afghanistan, Pakistan’s intelligence service would’ve found it difficult to build the militant armies it sent to invade Indian-controlled Kashmir in the early 90s; and without Kashmir in flames, the nuclear close-calls of 1999 and 2001 never would have happened.

As it turns out, these arguments are wrong… but not just wrong in details, but wrong in strategic logic and structure as well.  I mean, they are essentially as wrong as wrong can be.  They are in a way, the antithesis of what a reasonable debate about national strategy ought be.

First, let’s consider the basic structure of the argument.  In order to prevent a nuclear war between India and Pakistan, the United States has to defeat an Islamist insurgency in Afghanistan.  Hmmm.

Prevent is an awfully strong word, isn’t it?  The case is stronger if you cast it as an either-or situation.  Either we remain in Afghanistan OR a nuclear war occurs in South Asia.  Now, as a matter of fact, if that were the choice, one could make a compelling case for an international response.  One might even make a compelling case that the United States ought to lead this response.  But as a matter of principle, is it in America’s interest to spend blood and treasure to prevent a war — nuclear or otherwise — between two independent nations?  What is the principled reason for this obligation?  And how far does it extend?  Must we also prevent large scale conventional wars?

So, in the very strongest version of the argument, the case is flawed, but not impossible to make.  But the strongest version of the argument is also the most dishonest.  Realistically, we’re talking about “reducing” (at best) the risk of nuclear war rather than preventing it.  And now, the argument becomes less sustainable.  How many American lives is it worth to reduce the already low risk of nuclear war between India and Pakistan?  Lindholm and Foust may think this is an argument that would resonate domestically.  I think only a maniac would try to make it.

I mean, think about it.  Is this a humanitarian argument?  That the loss of life in South Asia would be so great as to justify the investment?  We can’t build a consensus on Darfur, and most Americans regret the Somalia intervention which saved hundreds of thousands at the cost of 19 American lives.  Is it an environmental issue?  Concern over potential fallout and maybe broader climactic considerations?  We can’t even build a consensus on a carbon bill to address the certainty of climate change when even the most aggressive bill would cost a fraction of the cost of the Afghan war.  The notion that pitching the Afghan conflict as a response to the threat of war between India and Pakistan would cement its public support is a fantasy.  And in this case, the fact is that the public would have a better sense of costs and benefits than Foust and Lindholm.

So structurally, it is an argument which relies on appeals to unlikely risks and vague principles to justify the very real loss of American lives and the expenditure of hundreds of billions of dollars of taxpayer dollars.

Second, the argument relies on incoherent strategic logic.  While Lindholm and Foust talk about preventing nuclear war, and I’ve suggested that at best we’d be reducing its risk, as a matter of strategic logic, “fixing” Afghanistan would have no effect on the risk of nuclear war and might indeed increase it by some tiny fraction.

Lindholm and Foust seem to believe that Islamist radicals, if allowed to return to Afghanistan, would create pressure for the Pakistani government to behave more aggressively toward India or independently provoke a conflict between the two.  The first is almost certainly the reverse of what would occur.  The second unlikely but also unrelated to Afghanistan.

Does radicalism in Afghanistan pressure Pakistan into more extreme behavior?  Clearly not.  Some indiscreet Pakistani strategists consider Afghanistan as providing strategic depth in the case of a conflict with India.  The concept is fuzzy frankly, because unless the Pakistani army plans to retreat over the border — with no infrastructure or supplies — to avoid Indian advances, there is no compelling conventional strategic depth argument.   But, on the other hand, having as a neighbor a state that would unquestionably take Pakistan’s side in a conflict would provide some opportunity to retreat strategic assets, leadership, and potentially provide a base for guerrilla resistance against an Indian incursion.  In short, what a Taliban controlled Afghanistan might provide Pakistani leaders is a small measure of reassurance that in a conflict, the Pakistani army would not find itself crushed between an Indian advance and a closed Afghan border.  But while in a constrained and unlikely case, the “strategic depth” provided by Afghanistan could provide Pakistan with some options in the case of an Indian attack, there is no way in which the “strategic depth” provided by Afghanistan could aid in the development of an offensive military option against India.

In short, strategic logic suggests that a Taliban-controlled Afghanistan would reduce the risk of conflict by reducing insecurity among Pakistani elites.  Because it would not provide Pakistan with any additional offensive capabilities, it ought not increase insecurity among Indian decision makers.  Lower levels of insecurity usually result in lower levels of risk taking and less pressure for military pre-emption in times of crisis.  An Islamist — and likely anti-Indian — regime in Afghanistan almost certainly eases the security dilemma in South Asia rather than increasing it.

Is there a possibility that bored jihadists, having conquered Afghanistan would now turn their attentions to India?  Maybe, but that ignores the particular nature of the Taliban, which is not an undifferentiated radical Islamist movement, but is in fact deeply grounded in local conditions.  Indeed, some analysts go so far as to consider it more of a Pashtun insurgency than anything else.  That almost certainly overstates the case.  But the notion that Taliban forces, having conquered Afghanistan could suddenly pack up and spearhead an anti-Indian crusade is fanciful.

We tend to both underestimate and overestimate the Taliban.  We underestimate its strength in Afghanistan, but overestimate its ability to destabilize Pakistan — much less take on India.  It is a mistake to see the Taliban as either all powerful or impotent.  Indeed, some of the very reasons it is a force in Afghanistan — its existence as a Pashtun insurgency — are precisely what also limits its reach.

But what about the strategic political dynamics?  Is it possible that Islamist radicals always need an outlet, and that having won in Afghanistan some Islamists both within Afghanistan and in the Pakistani government would seek the next step?  Yes, that is a source of some concern.  Though, in the final analysis what that suggests as a strategy is not a massive U.S. commitment to Afghanistan to defeat the Taliban, but rather a minimal U.S. commitment that ensures the war there never ends.  If you really believe it is important to have an outlet to keep the jihadists busy, then both a Taliban victory and an unequivocal Taliban defeat are problematic.  This sort of amoral theorizing is fun, but of limited utility.  In the final analysis, the steam valve theory assumes a lemming-like enemy rather than one with strategic insight.

So, in terms of strategic logic, we’re left with the risk that radical Islamist groups might provoke a war with India on their own.  And here, we run into an empirical problem.

While Kashmir led to frightening tensions in 1999, what I consider the greater risk is 9/11-style shocks such as the Indian Parliament attack in December 2001 and the Mumbai attacks of 2008 — because of the risk of a visceral reaction poses a greater threat than premeditated clashes between militants, paramilitaries, and soldiers in Kashmir.  But neither the 2001 nor the 2008 attacks (nor indeed 9/11 for that matter) required a secure territorial base.  These were operations launched by a few dozen individuals.  And it is a fantasy to believe that anything we can do in Afghanistan could prevent a group like Lashar-e-Taiba from launching such attacks in the future… especially since Lashkar is based in Pakistan, not Afghanistan.

The thing about Kashmir is that it tends to confirm the controversial logic of nuclear deterrence proposed by Kenneth Waltz in his seminal essay,“The Spread of Nuclear Weapons: More May Better.” I am not actually a convert on this argument, but we do have to admit that after three major conflicts in 1947, 1965, and 1971, the Indian-Pakistan relationship settled into a Cold War-style competition following the 1974 Indian nuclear test and Pakistani hints at a nuclear capacity starting in the mid-1980s.  Indeed, low-intensity conflicts like Kashmir in 1999 and the mobilization of 2002 reflect the existence of a “stability-instability paradox.” This paradox suggests that while states will shy away from direct military confrontations, they will engage in irregular conflicts and brinksmanship precisely because they know full-blown war is unlikely.

As a matter of empirics then, we have to acknowledge that (a) the kinds of attacks that might trigger a dangerous crisis are unrelated to Afghanistan, and (b) that the low-grade conflicts and posturing that have defined Pakistani-Indian relations since the 1980s might reflect the dynamics  of nuclear deterrence rather than opening a window onto the possibility of uncontrolled escalation.

There are pithier ways of attacking the Lindholm/Foust thesis.  I’ve tried to understanding the argument, but I just don’t see how it holds up to careful scrutiny.