The debate on the pros and cons of Afghanistan is raging inside the Beltway. And it is a bit unsettling.

On the one side are those who say no, America has no national interests in Afghanistan — and yes, it’s a war of choice: let’s leave the hellhole and get out asap. On the other side are those who say yes, our security is on the line and al-Qaeda must be defeated in Afghanistan — and no, it’s a war of necessity

: let’s do it seriously and pour in more troops and money. Until it’s fixed, like Iraq.

You’ve seen it. Shrill and loud, some of the contributions. The other side is brandished as “foolish” and “not serious.” Both sides make up straw-men and then mow them down. And don’t look at the reader comments. All that is even more disturbing if you consider that people are dying in this business.

So what should we make of it? As often in verbal fistfights, both sides have a couple of valid points. Let’s block out the shouting and try listening to some of the nuances. I would venture to say that most experts should be able to agree on ten assumptions — some of them are just statements of fact.

(1) Islamic radicalism and jihadi activity will continue to be a pain in the world’s back for the foreseeable future. The necessary ingredients will be provided for some time to come: ideology, frustrated young men, technology, platforms to organize and train, money, second-rate leaders with malicious intent, even failed or weak states as sanctuaries. Ask any Israeli or European intelligence analyst.

(2) Afghanistan and Pakistan influence the global extremism problem. Whatever the U.S. and NATO do there, it will have an effect on radicals. The war may alleviate the problem — yes, there can be local successes against extremists; al-Anbar province is one example — or it may make the problem worse.

(3) It follows that the problem cannot be solved in one region. No serious analyst will explicitly say that Islamic fundamentalism of the vicious kind, or al-Qaeda, can be “defeated” in one region alone. Some might imply it, but they seem to be very careful not to say it.

(4) What happens in Afghanistan and Pakistan is connected to terrorist activity in Europe and elsewhere. Fighters with training and know-how from Afghan camps came to Europe in the past, and some will likely do so in the future. Limited evidence even shows that some attacks were ordered or coordinated from Afghanistan or Pakistan. — But the influence goes both ways, therefore:

(5) A number of potential sympathizers are motivated to go down the path of radicalization and eventually fight in their jihad because NATO and America and Britain are fighting in Afghanistan. This might be an unpleasant truth, but it has to be part of the analysis.

(6) Western intelligence and law-enforcement agencies have become better at counterterrorism, at containing and managing the global threat. Some NATO armies have become better at counterinsurgency, at containing and managing the local threat.

(I was thinking of including the assumption that jihadis have also become better — but is that so?)

(7) Afghanistan will not be a Switzerland, even Bangladesh will be tough. On some level — perhaps a lower level than now, perhaps a higher level than now — corruption, poppies, poor governance, sectarianism, a weak state, and probably extremism will be part of Afghanistan for the foreseeable future.

(8) Nuclear-armed Pakistan in combination with an unstable government and extremists is highly dangerous. But there is no simple causal relation between chaos or stability in Afghanistan and chaos and stability in Pakistan.

(9) Time is limited. Elections loom, and eventually the public’s patience in all NATO countries will run out.

(10) The U.S. and NATO have options in Afghanistan: troop-heavy, population-centric counterinsurgency with high political ambitions — or a lighter approach, less population-centric, and less politically optimistic.

The question is what follows from these assumptions? A complex cost-benefit calculus, I would say. A nasty one. There are costs and benefits attached to trying harder — and there are costs and benefits attached to stop trying harder. And at some point somebody will have to make the decision to start a withdrawal. That much is safe to say. The question is: when do the net costs of trying to solve the problem outweigh the net benefits of trying to solve the problem? The answer cannot be a yes or no. Only a: then. Of course that doesn’t mean that a timetable should be communicated publicly — but the calculation should be clear, or at least clearer. Words like “winning” and “victory” have no place in this debate, even if the street is shouting for it.

What makes this calculus more complicated is that past losses in blood and treasure tend to make actors more irrational. Like on stock markets. The effect is twofold: first the costs of failure are constantly overrated. Naturally those with a vested interest look left and right for reasons to bolster their cause politically. They have to. Second the benefits of success are also overrated — for essentially the same reasons.

Or, to use the imperfect analogy: if I’d given my money to an activist investor (that would be the previous administration) who then forced a previously rock-solid company (that would be the U.S. military) to go after a product that doesn’t seem to work as advertised (that would be the war on terror) — I’d consider selling. Certainly the investor and his former advisers would have reduced credibility today, but also the current executives of the troubled company. Or at least they would have to put forward careful and considerate arguments for waiting out the crisis — in a calm, civilized, and sober way. If they’d overpromise or shout them in my face, I’d think something’s wrong.

Dr. Thomas Rid is the Calouste Gulbenkian Fellow at the Center for Transatlantic Relations. This essay was first published at Kings of War.