There is ample evidence, however, that the Pakistani intelligence relationship with the Taliban does go far beyond mere shelter. Credible sources from Pakistan’s Federally Administered Tribal Areas told me that the leadership of the Haqqani network was indeed removed from North Waziristan after the Pakistani Army reconquered that territory in 2014—but they moved to the Tirah Valley in Khyber Agency, from there to Parachinar and then into Afghanistan. If true, this move could not possibly have taken place without the help, or at the very least acquiescence, of Pakistani intelligence.

In private, many Pakistani soldiers and officials explain that the terms set out by pragmatic Taliban interlocutors for peace are the most that the Taliban could ever be brought to accept and, therefore, that Pakistan will urge upon them. Senior officials say that at the very start of the peace process Islamabad laid down the following conditions, which they claim U.S. officials accepted: that Pakistan would influence the Taliban to come to the negotiating table, but would not try to force it to accept any given settlement; that such a settlement was for the Afghan government to propose; and that Pakistan would take no responsibility for the actual implementation of any agreement.

WHAT CAN Washington and Kabul do to change Islamabad’s approach? Kabul could take concrete measures to restrict Indian influence, for example by closing Indian consulates. This course was rejected by almost everyone with whom I spoke in Kabul, whether on the grounds of national pride and independence or because powerful forces in Afghanistan do in fact regard India as an essential future backer. Kabul could join the international community and finally recognize the Durand Line, the frontier drawn by the British Empire between Afghanistan and what is now Pakistan, as an international frontier (thereby also falling into line with the almost-universal international tradition of recognizing the borders left behind by colonialism). This was also rejected by almost everyone with whom I spoke in Kabul and Jalalabad, whether out of nationalist, irredentist passion (“the sacred soil of Afghanistan”) or because they feared that it would lead to insuperable fragmentation in the Afghan political system and provide a vehicle for mass protests that former president Karzai could use to overthrow the present power-sharing government and return to power.

Karzai has been using this issue assiduously in his rhetoric in recent months, and when the United States (which does recognize the frontier) suggested this to Karzai in 2007, his response was so furious that Washington dropped the idea and has never revived it. As for the prospect of the United States distancing itself from India and reinstating Pakistan as its chief regional ally, Islamabad’s spotty record and New Delhi’s key place in Washington’s strategy of containing Beijing make this scenario highly unlikely.

America’s ability to put economic pressure on Pakistan is severely restricted by China’s aid. The $46 billion promised in the China-Pakistan Economic Corridor (CPEC) and associated projects is more than three times greater than all U.S. nonmilitary aid since 2001. An expansion of the American drone campaign against the Taliban leadership raises the question of just how far the United States is prepared to go. If the leaders withdraw into major Pakistani urban centers, firing drones at them would bring about a collapse of the U.S.-Pakistan relationship (including crucially important cooperation against terrorist attacks aimed at the U.S. homeland) and further radicalize Pakistani society. The killing of Mullah Mansour may seem an American victory, but the best that U.S. officials with whom I have spoken have been able to come up with as justification is that “it hasn’t made things any worse than they were already.” Even that is not true, since the killing has increased the power of the Haqqanis.

Unfortunately, it seems likely that these differences cannot by their very nature be decided by negotiation, but only by “life itself” (as the Soviets used to say). Above all, this means changing the calculations of the Pakistani establishment and the Taliban leadership concerning the durability of the Afghan state and of U.S. support for it. Such a change could come about if four years from now the Taliban has still not captured and held a single provincial center; if the Afghan National Army remains resilient and American military support and European economic support hold firm; and if the present Afghan government, having survived until 2019, holds more-or-less legitimate elections leading to a relatively legitimate new government. At that point, Pakistani analysis will have been proved wrong, more Taliban fighters will be willing to surrender on terms and a peace along the lines envisaged by Kabul and Washington may be a possibility. If, conversely, Afghanistan faces recurrent internal political crises, elections are seen as illegitimate, U.S. support wanes and the Afghan army loses significant ground (or collapses altogether), Kabul and Washington will have to seriously reevaluate the terms put forward by Taliban pragmatists and by Pakistan.

Anatol Lieven is a professor at Georgetown University in Qatar and a visiting professor in the War Studies Department of King’s College London. In the 1980s he was a British journalist in India, Pakistan and Afghanistan. His most recent book is Pakistan: A Hard Country (2011). Between 2013 and 2016 he conducted five trips to Afghanistan for research.

Image: “070719-A-6849A-365 - Two Afghan National Army Soldiers talk with a local Afghan during Operation Saray Has July 19 near Forward Operating Base Naray, Afghanistan. The ANA worked with Headquarters and Headquarters Troop, 1st Squadron, 91st Cavalry Regiment (Airborne), during Operation Saray Has. www.army.mil”