A suicide bomb killed at least 20 in Kabul earlier this month, just as the United Nations reported a marked rise in the number of children dying in Afghanistan’s protracted conflict during 2016.

Mark Bowden, the United Nations humanitarian aid coordinator in the country, said the level of civilian casualties in Afghanistan was unlike anywhere else in the world outside Syria.

Speaking with International Peace Institute Senior Adviser Els Debuf, he said there was a need for “far better rules of engagement in the conflict and far more thought being given to how you can minimize the impact on the civilian population.”

This conversation has been edited for clarity and length.

You’ve been the UN humanitarian coordinator in Afghanistan for the past four years. How would you describe the situation today, and how do you see it evolving in 2017?

We have a more serious humanitarian situation now than we had four years ago because the nature of the conflict has changed. When I first came, there were considerable civilian casualties, but over the years we’ve seen the continuing increase in the numbers of civilian casualties, with a disproportionate impact on women and children.

The other concerns that I have about the humanitarian situation is with a declining economy the ability of the population to withstand economic shocks and natural disasters has also declined. So vulnerability has increased, and will increase further this year because of the influx of a number of returnees who remain displaced. In general, we see a changing situation with regard to displacement, where people are being displaced for longer periods of time, and we see more casualties from the protection of civilians in a rather weaker economic environment.

What would you say are the main causes of the massive internal displacements in Afghanistan, which affect most if not all of its provinces, and how can they be addressed? And, in light of the protracted nature of displacement, how have you been able to concretely alleviate immediate suffering at the same time as investing in more sustainable solutions?

Internal displacement has followed many different patterns across Afghanistan. I would say the displacement in the north has been, up until now, rather different than the displacement in the south, and the northeast has another pattern of its own. In the north the war is being fought more through local militias, and the targets of the main displaced populations tend to be from district capitals—lower-scale displacement with essentially the families of militia members and others.

In the south, the displacement continues from what is essentially large-scale war, with more and more people going into Lashkargah and the capitals. There’s also a changing pattern of displacement due to the way in which roads have been cut and transport and trade routes have been affected, which has led to, again, more displacement in those areas. Places like Kunduz are seeing far more conflict, and this is a pattern that worries me a lot more because we’re now seeing cases of prolonged displacement.

The problem we face is that our own response mechanisms are geared toward just providing support on new displacement, and then there’s no secondary support for the prolonged displacements. So this year’s UN funding appeal is actually trying to address that issue and the increased vulnerability that goes with it. Another impact of this is that when returnees come back a lot of them become displaced because they can’t go back to their [exact] places of origin. So, paradoxically, the refugee return is another push in displacement, mainly to major urban centers.

So you’ve got a variety of displacement taking place, but one underlying feature is that we are stretching coping mechanisms and support mechanisms because most support comes from host communities and families, not just from international assistance. Secondly, we have, unwittingly in some cases, created potential competition between displaced and returnees by having different programs of support. So where the refugee has been getting quite a high level of support they have to some extent increased the cost of housing also being occupied by displaced people. I think one of the important humanitarian elements of this is to get a better harmonization of response.

One of my main responsibilities as humanitarian coordinator is to ensure such harmonization of response, including between the humanitarian and development responses. On the latter, as more areas are contested or come under direct Taliban control, I think that both the development agencies and donors are beginning to recognize—and the bi-national donors—that development assistance is reaching less and less people, and that we also need to look at our service delivery models.

I’ve always been very keen to say that humanitarian assistance should be trying to maintain agreed models of service delivery—national models of service delivery in areas that are otherwise inaccessible, which are covered by international humanitarian law (IHL). So a lot of the design of the humanitarian program and appeal is to continue to ensure the delivery of services to a lot of the population. And I think donors are now beginning to look at how they need to engage with the humanitarian organizations and with humanitarian strategies to ensure that we can keep the basic social infrastructure through this prolonged period of conflict. That’s, in a sense, how we’re developing the humanitarian development relationship.

There hasn’t been a lot of discussion about resilience in Afghanistan, but I think that’s now beginning to come on to the development agenda rather more. In the last month or so I’ve seen a lot more engagement from the donor community, where they’re drawing up their five-year strategic plans to look at how this humanitarian development relationship can be strengthened.

As you said, apart from significant internal displacement dynamics, we have seen more recently an increase in Afghan refugees returning from Iran, from Pakistan, and from elsewhere, where they’ve spent years and decades fleeing the conflict. What are the specific challenges returnees face when they come home to a country that is no less fragile or less conflict-affected than when they left?

There are lots of issues of social integration, because a number of the returnees are second or third generation. And some of them said to me, “We’re treated not as Afghans returning; we’re treated as Pakistanis coming into the country.” And their dress patterns are very Pakistani, their speech patterns are Pakistani, so there are issues of integration that will be a problem.

The second issue is a very flat economy—which Afghanistan has at the moment, with only 1% to 2% growth in a year—and a very young population. This means that 400,000 youths are coming on to the job market each year. Returnees also put more stress on the economy and particularly on the job market. So, while some of them will have jobs and skills, which will contribute to the economic growth of Afghanistan, a number of them are daily laborers, and that’s going to create another problem in terms of competition. So that’s why I think we’ll see a lot more vulnerability.

To what extent are these massive returns voluntary? You spoke about social integration and that people might be more perceived as foreigners, so do those people feel like they’re Afghani, and do they want to come back to Afghanistan?

I think it’s quite mixed. I think that some of them do feel Afghani and will tell you they’re coming back for reasons of patriotism, but a lot of them also have told stories—and particularly undocumented Afghans—about how they’ve suffered from night raids in Pakistan, how they’re not welcome. I met with NGO communities in Pakistan who would provide support to the Afghans, who confirmed that there has been, particularly in Khyber province, a lot of hostility, and the government explains some of their actions by local hostility. I think many Afghans in Pakistan recognize that there’s a very difficult environment for them to live in.

Do you feel that humanitarian organizations within the UN system or in parallel are able to reach all people in need and to deliver an adequate response? And, if not, what would you say are the main obstacles to full humanitarian access, and how can they be overcome?

When we launched the appeal, we said that we could reach 5.7 million out of the 9.3 million Afghanis in need, so there is substantial access, but there are still key areas where access is more difficult. It is very much a localized issue, because I think both parties to the conflict have recognized some need for humanitarian access, and through discussion, local negotiation, you may have both parties willing to find a way through. So it’s not the most difficult and intransigent situation; we don’t have anything like Syria or anything else where there’s siege, but it’s something that needs constant effort and work.

I think we’ve improved access over time, but bear in mind it is a large-scale conflict and it remains difficult. It also varies very much by province to province. The biggest challenge has been the ISIS-controlled areas where there is no access, and it’s unfortunate that these areas are also areas where you have endemic wild polio virus. We put a lot of effort in trying to get into those areas, but it does pose a broader risk by not being able to do that. Other areas it sort of comes and goes. Places like Farah, where there’s been a lot of fighting, it’s been difficult.

Interestingly, in places like Kunduz I think the humanitarian organizations working there have been allowed to continue their support across most of the districts, and there have been mechanisms to do that. And whether it’s the UN or NGOs, luckily we’re in an environment where there are quite strong partnerships. The UN is not, I think, seen as a hostile political entity, which also helps. We’ve put a lot of emphasis on the UN side in trying to maintain and to be seen as impartial, and that, fortunately, is understood.

As you are approaching the end of your mandate in Afghanistan this coming March, can you share a story that has stayed with you and is illustrative of your meeting the people of Afghanistan?

The ones that mark me have been when I’ve been to the Emergency [an Afghan NGO] hospital and seen some of the child casualties there—victims of bombings that have taken place. I remember a kid I saw there who had lost his eye and a hand and was just bewildered by what had happened, and I think that’s the tragedy: It’s a conflict where the collateral damage on civilians is dramatic and not understood, and you begin to see it in casualty figures; we’re talking about levels of casualties, civilian casualties, that aren’t experienced anywhere else except in Syria. You know, far more than Yemen, far more than Somalia and other places, and it’s actually seeing that impact in trauma centers that I think has really affected me.

You mentioned earlier on also that you’re worried about the civilian casualties and the respect of IHL. How much do you feel is really collateral damage and how much of it is actually a targeting of the civilian population? In terms of IHL, is enough attention paid to precaution, or is it calling into question the actual respect of very basic norms?

Well there are two different aspects to this. The actual number of suicide bombers has gone down, but the big events that have really affected the civilian population have been the truck bombs where we’ve had massive casualties. We’ve witnessed some really horrific things which are clearly targeting civilians. The attack on Hazara in March, where you had IEDs placed in ice cream carts, can’t be seen as anything but an attack on the civilian population. When you have a truck bomb with huge amounts of explosives, you can’t see that as anything but an attack on the civilian population. So that, I think, is of course a concern. On the other side, we’re now seeing far more use of airstrikes, use of helicopter gunships, and it’s impossible to imagine them without affecting the population. I think there’s a need for far better rules of engagement in the conflict and far more thought being given to how you can minimize the impact on the civilian population.

Is that a dialogue that you are able to have with all the parties in the conflict?

Yeah, it is. We’ve had some success with the Taliban in that there’s recognition that IEDs are indiscriminate, and that we’ve seen a noticeable decline in their use—particularly no more pressure plate IEDs which tend to blow up buses, so I think that’s one area of success. I’m working with the chief executive to set up a high-level policy group, which will develop strategies to specifically look at mitigating the impact of the conflict on the civilian population. It will look at issues like the rules of engagement for the Afghan army and how to better protect health facilities. Although there’s clear guidance and training to respect international humanitarian law on the Afghan side, there are clearly breaches on this. So we are looking together to see how we can improve that and have a positive impact to ensure that the casualty figures go down.