After 16 days of negotiations, the U.S. Special Representative for Afghanistan Reconciliation, Amb. Zalmay Khalilzad, tweeted that "peace requires agreement on four issues: counter-terrorism assurances, troop withdrawal, intra-Afghan dialogue, and a comprehensive ceasefire … We’re now ‘agreed in draft’ on the first two." This is a positive development, as previously the two sides had been only "agreed in principle." The next step is for both sides to consult with decision-makers and seek their approval of the draft agreement so as to finalize it. This would presumably pave the way for direct talks between the Taliban and the Afghan government to address the remaining two issues.

While there have been numerous attempts at negotiations throughout the war in Afghanistan, these talks have made by far the most progress to date. They have also engendered the most handwringing, concern, and pushback. Some of the worries surrounding these talks stem from legitimate concerns, such as the rights of women and minorities — concerns that the Taliban have occasionally sought to allay. However, the group’s statement at the recent Moscow discussions was ambiguous at best on the topic of women’s rights, which has left some to interpret it as indicative of a continued harsh stance by the group on this issue. Indeed, recent CNN reporting from Taliban-held areas confirmed some of these critics’ worst fears.

Other concerns have arisen from the fact that so far these talks have not included the government of Afghanistan, and it is clear that President Ashraf Ghani’s administration is feeling excluded and suspicious. For example, Ghani’s National Security Advisor Hamdullah Mohib recently remarked to a set of journalists that he fears the United States is "selling out" Afghanistan and that Khalilzad is negotiating with intent to put himself in power as the leader of a new government. The U.S. State Department subsequently asked Mohib privately to apologize for his comments, which he apparently refused to do — leading the Trump administration to cease communications with him.

Still others are concerned that the United States is giving away too much, too early in these negotiations. For example, some have argued that Washington has squandered leverage by agreeing to address the Taliban’s demand of a troop withdrawal timeline prior to its engagement with the Afghan government. Or worse, that the United States is simply repeating its history from Vietnam by looking for a "decent interval" between the end of the war and the inevitable collapse of the government in Kabul.

There are many reasons for all of these concerns — some are legitimate, stemming from real, personal fears of people likely to be directly (and perhaps repressively) affected by the results of these talks. Others stem from fears of loss of power, prestige, and/or streams of wealth that are assured to a select group that benefits from a wartime economy — a group that is not confined to the borders of Afghanistan. Still others stem from sunk cost bias, ignorance, ahistorical beliefs, and outright paranoia.

One aspect of these talks that seems to be missing is a discussion of the continued costs of war versus the costs of peace. In other words, a dispassionate rendering of what it means for these talks to fail, versus what it might mean for them to succeed. While such discussion is inevitably fraught with challenges since there is no way to calculate such costs with absolute accuracy, let’s nonetheless press ahead and see what we can learn.

Let’s start with the costs of war, and specifically the two most tangible ones — human and economic. The number of personnel from the United States and its coalition partners being killed in the war these days is relatively small (e.g., the total number of U.S. military and Defense Department civilians killed since 2014 is 64). Much higher are the human costs among Afghans. For example, the United Nations Assistance Mission in Afghanistan has been tallying civilian casualties of the war over the past decade. Since 2014, the number of civilians killed each year has varied slightly around an average of about 3,600. There is little reason to think that, if the war continues apace, the total for 2019 would be any less than this. In terms of Afghan security force casualties, neither the United States nor the Afghan government releases official tallies any more, but estimates of these figures can nonetheless be made. For example, projections of the New York Times’ weekly casualty data (which it has been collecting since October of last year) suggest that we might expect at least 4,000 deaths of pro-Afghan government forces in 2019. However, there is reason to believe — based on recent statements by Ghani and other Afghan government statements — that the number could be expected to rise as high as 11,000 or more.

The Taliban also do not release tallies of their casualties, but a recent press release in the wake of a Ranger battalion deployment cited the unit’s claim of over 1,900 Taliban killed or captured during its 4-month deployment. Let’s assume this claim is accurate and that the vast majority of these Taliban were killed (say, 90 percent, or about 1,700). Let’s further assume a consistent operational tempo over the next year by units that follow. This would give a rough order of magnitude estimate of about 5,000 Taliban to be killed by the U.S. counter-terrorism mission alone over the next year. If we assume that at least as many would be killed by the entire Afghan Army, police, and special operations forces, we get a very rough estimate of about 10,000 Taliban who might die over the course of the next year.

Taking all of these estimates together, one might anticipate total deaths stemming from the war in 2019 to be in the range of 17,000 to 25,000 people. This would in effect add another 12 to 17 percent to the war’s total number of deaths (roughly 147,000 according to Brown University’s "Costs of War" project). And none of these numbers say anything about those who would be wounded or displaced by the war, the number of which would likely total multiples of the number killed.

In terms of economic costs, a senior Pentagon official publicly cited the cost of the Afghanistan war to the United States as $45 billion last year — which includes about $5 billion for Afghanistan’s security forces, $13 billion in operating costs for the U.S. military, $750 million in economic aid, and a host of other categories of spending such as logistics and sustainment. This does not include the costs borne by the rest of the international coalition in Afghanistan, nor by the Afghan government — which totals several billion dollars more. And these are only direct costs, which fail to capture the opportunity costs to the economies of Afghanistan, the region, and all of the countries involved in the war.

What would be the costs of peace? These are, of course, even harder to calculate, since peace remains a hypothetical scenario for Afghanistan and therefore empirical data are not available. But let’s make an attempt. While it would be nice to assume that war casualties would go to zero, as I discussed previously, there would likely still be a requirement for a counter-terrorism mission against the Islamic State and al-Qaeda in Afghanistan (whether conducted by the Americans or Afghans alone), and so it’s likely there would still be some hundreds of deaths from that mission each year until the threat from these groups is diminished. Similarly, it would be nice to assume that costs would go to zero, but here again there are residual requirements — perhaps $2 billion or more per year for continued support to Afghanistan’s security forces, and sizeable additional amounts in continued development aid, refugee resettlement programs, and programs to demobilize and reintegrate insurgent fighters.

There are, of course, many intangible costs associated with both war and peace. For the former, these include the drag that Afghanistan has become on America’s aims to refocus its strategic and operational attention on state adversaries such as China and Russia. For the latter, they might include some degree of the rights and freedoms that Afghan women and minorities have enjoyed for nearly the past 20 years. They might also include a degree of the power and wealth that some members of Afghanistan’s elites have obtained as a result of the war economy.

So what can we learn from this admittedly rough comparison of the costs of war and peace in Afghanistan? Three points come to mind. First, based on the calculations above, the continued costs of not getting to the second two issues cited by Khalilzad — intra-Afghan dialog and a comprehensive ceasefire – are sizeable, in both human terms (roughly 45 to 70 people killed every day) and economic ones (about $135 million per day in war expenditures). Anyone questioning the urgency of the negotiations should bear these figures in mind.

Second, the costs of the war are highly bifurcated: While Washington is paying the vast majority of the financial costs, Afghans (whether aligned with the government or not) are paying the vast majority of the human ones. This is important to recognize because it helps clarify why the U.S. government is willing to meet the Taliban demand of direct talks first and intra-Afghan talks second. Such an approach allows the United States to address its costs first — to minimize its expenditures in Afghanistan (via troop withdrawal) while also minimizing the risk of another 9/11 (via a verifiable "no safe haven" guarantee by the Taliban). The sequencing of these issues is such that the Afghan-borne costs — both the current human ones and those that would accrue from a peace deal — are to be addressed second. This is clearly not lost on many Afghans and is creating a sense of hostility among the likes of Mohib for the American approach. But it’s worth noting that this approach is consistent with both the United States’ strategic interests in Afghanistan and this administration’s "America First" view of foreign policy and is therefore unlikely to change.

Third, while reaching a peace deal would mean sizeable cost savings for the United States, the Afghan government would instead accrue significant near-term cost increases. Some of these would stem from external patrons reducing their overall contributions to Afghanistan, some from a requirement to find gainful employment for demobilized Taliban fighters and Afghan security forces, and others from the likely return of hundreds of thousands if not millions of refugees from Pakistan, Iran, and elsewhere. As discussed above, there will be continued tangible costs even in a post-peace deal setting. Which means that Afghanistan’s patrons — and especially the United States — should temper expectations of "peace dividends" from Afghanistan and instead be thinking of ways to help the Afghan government mitigate these costs as an insurance policy against the possibility of resurgent wartime costs in the future.

At the same event in which Mohib questioned Khalilzad’s motivations, he said "It is not for us [Afghans] to prescribe American national interests. What we don’t accept is to be thrown under the bus." While it is easy to understand this kind of frustration, an examination of the costs of both war and peace make it clear that the United States is not throwing the government of Afghanistan under the bus. Rather, it is pursuing its own national interests and reduction of its own costs first, with a stated intent to address the costs to Afghans second. Rather than railing against this situation — one over which it has no control and little influence — the Afghan government would be better off preparing to expeditiously address its own interests and costs in the second phase of negotiations. Given the continued human and economic costs of the war, one can only pray that this phase will begin soon.

This article originally appeared in War on the Rocks