This week we’re looking at the impact of war on rates of infant mortality. How big is it? And can it be mitigated?
It seems obvious that war harms civilian populations, not least children. But research can reveal much more about the nature and scale of those harms and perhaps also about what can be done about them.
This week we’re focusing on a new study of the impact of war upon rates of infant mortality. The study is by Rod Abouharb, Associate Professor of International Relations here in the UCL Department of Political Science.
Mentioned in this episode:
SUMMARY KEYWORDS
conflicts, war, civil wars, infant mortality rates, effects, infant mortality, data, interstate, research, geneva conventions, question, civilians, immediate effects, bombed, mitigate, lingering, defence, review, lingering effects, important
SPEAKERS
Alan Renwick, Rod Abouharb
Alan Renwick 00:05
Hello, this is UCL Uncovering Politics, and this week we're looking at the impact of war on rates of infant mortality. How big is it and can it be mitigated?
Hello, my name is Alan Renwick, and welcome to UCL Uncovering Politics, the podcast of the School of Public Policy and Department of Political Science at University College London.
Our television screens have been filled over the past year by images of the war in Ukraine – images of destruction, death, injury and hardship. It seems obvious that war harms civilian populations, not least children. But research can reveal much more about the nature and scale of those harms and perhaps also about what can be done about them.
And so this week we're focusing on a new study of the impact of war upon rates of infant mortality. The study is by Rod Abouharb, Associate Professor of International Relations here in the UCL Department of Political Science, and I'm delighted to say that Rod joins me now.
Rod, welcome to UCL Uncovering Politics. And I guess the first question relates to the point that I was just beginning to hint at there. We're probably safe to assume that war damages children's health and increases rates of infant mortality, so what more did you feel we needed to know that motivated you to conduct this study?
Rod Abouharb 01:36
Well, I just want to thank you, Alan, for having me on the podcast. So I'm delighted to be here today.
So I think I would say we actually know much less than we think we do when it comes to this. So even though it should seem like kind of an obvious question that there is this systematic link between war and infant mortality rates, the research up to now actually has only focused on sort of particular conflicts or particular regions.
And I think one of the things that really struck me about what we're sort of missing in our understanding is that the large-N systematic work on this has looked at adults and even children, but not infants. And that might sound pedantic but it's actually quite important. So in the sort of public health research on this, it looks at children up to the age of five. But infants who are up to the age of one are really entirely missed out as a separate category.
And I think that's important for a number of reasons. So one is that we're thinking about really, sort of, the most vulnerable populations. And actually the populations who have, in many ways, the least voice. And so those are the things that really struck me about it.
And then in trying to understand the link between war and infant mortality rates, I think the other thing that was sort of missing in our understanding of this is: do different types of wars have similar or different effects? And then how do, kind of, the immediate effects of war differ from the kind of lingering effects of conflict? And again, in some ways, the previous research has been quite descriptive, but not particularly analytical.
And I'm really interested in understanding, so you know, why might these wars have different effects? And are there ways that we should be trying... You know, if that's true, then I think that also points to, you know, how we can think about kind of mitigating the negative effects of these conflicts, which you sort of mentioned in the intro.
Alan Renwick 03:53
So you mentioned the two different distinctions: the distinction between different types of war and the distinction between different kinds of effects – immediate and lingering effects. Do you want just to unpack those a little bit further – what different types of war do you have in mind, first of all?
Rod Abouharb 04:07
So in sort of international relations the kind of two big categories of conflicts that we tend to think about are, one, what I would describe as civil wars. So these are conflicts within national boundaries. And then the second is sort of interstate wars. So these are kind of wars between states.
And the research tends to look at either one or the other. And of course, the thing is, if we think that both of these types of conflicts have a public health impact (and in my research, I'm looking at infants in particular), the next year, if you only look at one type of war, then you're missing the effects of the other. So that's, I think, an important reason to have both in any of our empirical models – to understand, kind of, the overall impact of conflict.
And then the second one is between what we talked about – kind of the immediate and then the sort of lingering effects. So the immediate effects in some ways are, you know... If you look at the coverage of the conflict in Ukraine at the moment, you know, you're thinking about hospitals being bombed, you're thinking about populations being bombed, about the electricity networks and the sewage networks and all these other things – the infrastructure that keeps populations healthy or is being attacked.
And these types of immediate effects are really, you know, what happens if you're a victim of the conflict itself, right. So you are in the wrong place at the wrong time and the bomb hits you; or you get caught in crossfire; or, as we know, actually, in a number of wars, parties to these conflicts will target healthcare facilities, will target other types of civilian infrastructure, which they shouldn't be doing. And even in the early weeks of the war in Ukraine, children and infants were being killed as part of that.
And then sort of moving onto the lingering effects aspect. This tries to get at the idea of what happens if you are in sort of a war-torn society. So we know that, you know, kind of the progressive consequences of these wars destroy some of these key infrastructures, destroy many of the key sort of support processes that keep populations healthy. And, you know, it makes it sort of very difficult to go to the doctors, right. So things that you and I would take for granted in peacetime (going to the GP (well maybe not UK right now), but going to the GP; accessing and going to hospital; getting medical care when we need it) – that is very, very difficult in war zones, and very, very dangerous. And so these things become much more limited for the civilian population.
And then the other types of infrastructure which are typically destroyed during conflicts are things like water treatment works. Even things that you wouldn't directly link to mortality rates like, for example, the road network, right? So if you destroy the road network it's then very difficult to get supplies in, it's very difficult to get access to medical care, access to foodstuffs and the like.
And then the lingering component of that is we know it takes a long time for societies to recover from that. So it could be that, you know, the hospital was destroyed three or four years ago, but actually it takes a long time to rebuild that.
And we know from previous research that these lingering effects continue even after the conflict stop, right, because that hospital hasn't been rebuilt, right; that bridge hasn't been rebuilt that, you know, links the key arteries across a particular country. So it's really trying to get a sense of, you know, what do these sort of longer-term consequences of conflict look like on mortality rates?
Alan Renwick 08:17
Right. So we have kind of direct and indirect effects of war. And then the direct effects might drop away once a war stops, but those indirect effects can linger for a long time afterwards.
And so what sorts of expectations did you bring to the analysis here? And were you expecting to find a difference, for example, between civil wars and interstate wars?
Rod Abouharb 08:43
So I think... And I suppose I have kind of a background interest here in how the international system post-World War Two kind of organised itself. So if like, in the background here is the role of the Geneva Conventions. And I think, you know, we can talk about how we might sort of assess that directly later on.
But in the background here is the Geneva Conventions. And then what that obliges states to do is that it really... You know, so even wars have rules, right? And it obliges states to protect civilians from harm in these conflicts. And in particular, those Conventions were written with a focus on sort of interstate wars. Now, there's an argument about the extent to which they apply in civil wars. But certainly with interstate wars, I was expecting that the effects of these wars would be smaller in comparison to civil wars, precisely because of this legal obligation that states have to protect civilians from harm.
And also we know that there's been some really interesting research on how civil wars are fought. And in particular NGOs, for example in Syria, in Sudan, in Central African Republic were all showing how actually the state had incentives to bomb its own healthcare infrastructure.
So we saw this in Syria where, you know, kind of in opposition-held areas of the country the hospitals there were repeatedly bombed by joint Syrian and Russian air forces. And of course, that's against the laws of war, right? So they shouldn't be doing this. And in that case, we would either call that a civil war or an internationalised civil war.
But the argument would be that there are probably different incentives there for states. So they may actually have fewer incentives to follow these rules, given that the conflict is within its own sovereign boundaries. So based on that I was expecting the civil wars to have maybe a worsened effect in comparison to interstate wars.
Alan Renwick 11:07
Okay, interesting.
Rod Abouharb 11:12
That was sort of the basis for the idea. And then if you want to think about what, kind of, the immediate or the lingering effect was, the literature is pretty overwhelming on this – that the lingering effects tend to be worse overall. So I was suspecting that these kind of longer-term effects of these conflicts, you would see that the highest increase in infant mortality rates.
Alan Renwick 11:37
So we will get onto what you actually found, but before we do that, we should explore the methodology through which you're finding these things out a little bit. And I guess in essence it's quite simple: you get the data on the incidence of war, you get data on infant mortality, you look at the relationship between the two, with some control variables in there as well. But of course, it's always more complicated than that.
So I guess two questions, maybe. Firstly, are there aspects of the methodology and the design of this research that you'd particularly like to highlight as being important for our listeners? Second, there was one aspect that I found particularly interesting, which was just simply around the measurement of infant mortality. And I suppose my hunch would be that, in the context of war, data on infant mortality are going to become very unreliable. So you know, if the exercise here is about actually measuring these effects as precisely as we can, is that actually a thing that is possible for us to do?
Rod Abouharb 12:39
So that's a great question. You sound like my 'reviewer two'. For listeners who aren't aware of the 'reviewer two' joke, this is kind of in academic papers, we always get one horrendous reviewer who asks all the really difficult questions, and invariably it's the second reviewer on the panel.
No, no, but seriously, these are really important questions, actually. And I think the literature to now has looked at, I think I mentioned before, sort of under-five mortality rather than infant mortality. And that may seem, you know, a bit technical, you know – isn't it all pretty similar? And actually, one of the things that you find when you start to kind of delve into the data is that we're missing a lot of cases when it comes to under-five mortality.
So the period that I examined for the paper was 1950s to 2007. So over that period, just to give you a comparison, I had sort of 8,207 cases of infant mortality rate data available. Now, if I had been doing an under-five study, which is what most of the research looks at, I would have only had 2,605 cases. So almost sort of four times or, no, three times the amount of data available. So I think there's the breadth of data that's available.
And I think the other thing that we know is often data is missing in the countries that are in the worst circumstances, right? So either have the worst conflicts or maybe have the worst registration systems. And those may well be the cases where you would think that the mortality rates would worsen the most in these types of conflicts. So the IMR data is much better for that purpose: we have much broader coverage; we're missing very few cases; and I think it's sort of a better baseline to begin trying to study this stuff.
Now, your question about the quality of the data during periods of conflict, I think, is a really important one. We know that the coverage is going to be more spotty, even with the IMR data – we still have gaps in that. Now, I think there's, you know... The International Committee has done a number of things to try to sort of mitigate that. So the WHO, since its inception in 1949, has placed an emphasis on sort of good quality, comparable data including, in particular, infant mortality. The UN has, kind of, a sort of complementary system where they're also trying to generate sort of comparable IMR data over time.
So I think the data quality definitely dips and it's certainly more spotty during periods of conflict. But I think if anything, it probably means that we underestimate the effects of these conflicts. Nevertheless, I think I have a lot more confidence in this data than if I was trying to make the case using sort of under-five mortality data as a comparison.
Alan Renwick 15:59
Okay. And so you're looking, as you said, at 1,000s of kind of country years, I guess, across the globe from 1950 to 2007. And you have data on levels of infant mortality, you have data on the incidence of war. We know whether that's civil war or interstate war.
One thing that struck me was you don't look at the scale of conflict – I would have guessed that the scale of conflict would make a difference to infant mortality rates. Was there a reason for that?
Rod Abouharb 16:33
So that's a great question. The interstate war data tends to have a threshold of 1,000 battle deaths. The civil war data you can get at lower thresholds. And in earlier versions of the work both are in there.
I think it came out for a variety of reasons, but one is to think about, sort of, comparability. So if I could have, let's say, a lower threshold for interstate wars, then I could be comparing different types of conflict but at similar levels of hostility. And I think that sort of data limitation at the moment is one thing to think about.
The 1,000-death threshold also makes it comparable to previous research on this. The most systematic previous work on this was by, sort of, Li and Wen, and they were interested in the impact on adult mortality. So I can sort of better compare my results to their earlier work, too. So there are some sort of data constraints and also, I think, making sure that I'm not comparing apples and oranges but maybe, you know, two flavours of apples in the research, in order to kind of understand if they have similar or different effects.
Alan Renwick 17:56
Okay, so we have a threshold of 1,000 combat deaths per year. So we're talking here about significant conflicts, but not necessarily really large-scale conflict.
Rod Abouharb 18:07
So I mean I think, if you like, it is another problem with the correlates of war data, right? And to be fair, this is the best data out there, right? So this is the data that everyone uses.
I think when they created the data they weren't necessarily focusing on what the consequences of these conflicts were for civilians and other groups. They were particularly interested in trying to understand what leads to these conflicts in the first place, right?
So if you like, I've sort of flipped around what is typically the dependent variable in a lot of IR research and it becomes my independent variable, my sort of thing that I think has the effect. And then I'm looking at something which, you know, is of interest to some political scientists, but there's a lot more public health researchers who look at this than sort of political scientists as a whole.
So in many ways, I'm trying to use data that wasn't exactly designed for this purpose and then thinking out, right, so if this is the best data we have at the moment, then how can we sort of reasonably use it to try to understand the link between, you know, different types of wars and mortality rates?
Alan Renwick 19:25
Interesting. So we know why you wanted to do this research. We know how you did the research. The most important question is: what did you find?
Rod Abouharb 19:34
So the world in part works the way I think it does, but not in all aspects. So we certainly find a significant link between both types of conflicts. So when civil wars erupt to this level of severity, they have a significant impact, worsening infant mortality rates. The same is true of sort of interstate wars.
If you like, kind of the point predictions around these were sort of the opposite of what I expected. Now, I think you have to be very careful with these point predictions because, you know, we have what we call sort of confidence intervals around them which say: look, in reality, if you want to be sort of 95% sure, it could be a bit worse, it could be a bit better than this. And those confidence intervals overlap between the civil wars and the interstate war. So the point prediction, you know, civil wars flare up and you have just over a 5% increase in infant mortality rates. You get to the stage of a major interstate war, and we're looking at a 10.5% increase in IMRs on average.
But the predictions here overlap. So it could be that these interstate wars are, in reality, still worse, but we have to be somewhat cautious about that. The thing is, when you're looking at these cases, we don't have that many cases, which is why these confidence intervals are relatively wide, you might think. So that's the cautionary note with them.
Having said that, it does seem to go against what I thought. And I think that's interesting, right? So I'm not going to entirely go against, you know, discount that because it goes against what I thought. And, you know, it may well be that we need to think more about the types of munitions that are being used – really kind of how destructive, you know, national armies are, national air forces are, in comparison to rebel groups, right? So rebel groups typically don't have access to the kinds of weaponry that the Syrian Air Force or the Russian Air Force was using on civilians in in the Syrian civil war. So there may be something about that, that we're kind of under estimating at the moment.
Alan Renwick 21:58
So in some sense, interstate wars may just be bigger, on average, than civil wars, is kind of what you're suggesting there?
Rod Abouharb 22:05
Yeah, I think so. And it could be that how we think about the threshold data of this, sort of, 1,000 battle deaths in many ways muddies the severity of these conflicts, right? If you have, sort of, fleets and fleets of bomber planes going over another country's boundaries, that is going to do a lot of damage.
Alan Renwick 22:29
Yeah. And so one of the mechanisms that you mentioned there in terms of why you were expecting interstate wars to have lower rates of increase in infant mortality was to do with the Geneva Conventions. And I think you were able to explore further the impact of Geneva Conventions as well.
Rod Abouharb 22:47
I mean, so at the moment the Geneva Conventions are in the background here because effectively all states are parties to that after World War Two. And I think in order to understand, kind of, directly if the Conventions have kind of constrained effects of these conflicts, it would be ideal to take this research back into the 19th century.
So I think this is something you can do. We have civil war data; we have interstate war data; we even have infant mortality rate data that goes back to 1816. So this is something that I was part of when I was a PhD student, staying up far too late encoding this stuff.
So there is a, I think, not unreasonable level of scepticism when you go back with this historical data: it tends to be bit more unreliable; we have to be even more careful than we are, you know, trying to understand the links between sort of conflict and these outcomes when you go back that far.
Alan Renwick 23:50
Okay, interesting. And you also had that expectation that the lingering effects of conflict would be worse than the immediate effects. What did you find on that?
Rod Abouharb 24:00
So there's a good deal of evidence to support that. You know, if we look at the sort of core results, the core results kind of suggest if you have spent most of your time as a state involved in a major interstate war, that this has the largest negative effect on infant mortality rates. So again, sort of going against what I expected of the interstate versus civil wars, but certainly in line with previous arguments that the sort of lingering effects tend to be the worst overall.
Alan Renwick 24:41
So what do we learn from this study about whether it's possible to do anything in order to mitigate the effects of war? I guess through these patterns you can start to see what factors seem to be having a bigger effect, what factors may be having a lesser effect – so you can begin to get a sense of where we should be focusing attention and trying to mitigate these effects.
Rod Abouharb 25:04
Yeah, I think it's a really good question. And I think some of the robustness tests, you know, flagged up some potential avenues that I think we need to better understand that may sort of mitigate the effects of conflict. So, you know, a number of NGOs have written about, kind of, the usefulness of kind of humanitarian interventions on trying to mitigate the health effects of these conflicts.
But so far, I haven't seen any kind of systematic work on it. And I think in many ways that will be sort of the next stage to understand – do these interventions have, kind of, a sizeable impact on that? I think that's one route.
And then I think the other one is actually... And it's a really interesting topic, so for example, in 2022, the US Department of Defence has now launched a review into what it calls the Civilian Harm Mitigation and Response Action Plan. The Secretary of Defence has tasked the armed services to precisely think about how they can limit the effects of these conflicts on civilians. So it's clearly of importance to the US Department of Defence at the moment.
As a comparison, there was no sort of similar review taking place in the UK Ministry of Defence. There was, kind of, a group of civil society organisations that have been pushing them to open up a review. So I think the other route is to, you know, actually have, kind of, the interested parties – whose obligations are actually to meet the sort of the Geneva Conventions – to actually start to understand what, you know, are the actual effects of what we're doing, right? And can we change that?
So I guess, this type of research, says: okay, the effects of what you're doing, you know, seems to be bad, here's some actual evidence that maybe you can think about. Alright, so what did we do in these conflicts; you know, what kinds of things happened; and what kinds of things should we be, sort of, changing?
Alan Renwick 27:21
So where do you plan to go next in this research? You've mentioned some of the questions that a broader review should be taking on. But what are your own research plans in this area?
Rod Abouharb 27:34
So I think it would be kind of really interesting to go back as far as you possibly can with this data. And then that will allow us to ask the big question about, you know, what's the role of the Geneva Conventions here?
So again, there's a lot of qualitative work on this. Public health specialists who you know... So for example, there's the violations data centre that has been capturing, you know, how civilians died in Syria, in an amazing level of detail. But the question that they can't answer is: would this be any different if the Geneva Conventions were not in place? And I think I feel like that for me is one of the key questions that I still think we don't know very much about.
And then asking these questions about do humanitarian interventions matter? Like, do they, sort of, mitigate the consequences of these conflicts?
So I think those would be the two sort of biggest avenues for future research that I would be thinking about. And then probably sending various emails to various government departments, you know, seeing if they would be interested in what I've been up to and if they think it's of, kind of, any value in terms of how they go about, you know, making policy in the future.
Alan Renwick 28:55
Well let us hope so because this is clearly incredibly important stuff.
Thank you so much, Rod. It's been really very interesting to discuss this with you on, as I say, such a very important topic as well. So thank you so much.
Rod Abouharb 29:09
Thank you very much for inviting me on today. It was a lot of fun.
Alan Renwick 29:12
And the study that we've been discussing is called War and Infant Mortality Rates by Rod Abouharb. It's available for online early view free of charge in the Journal of Human Rights, and we'll put the full details in the show notes for this episode.
Next week, we'll be looking at a new proposal for how we should think about the role of the state in our society. It's called the 'precautionary state' and it's going to make for a fascinating discussion.
Remember to make sure you don't miss out on that or other future episodes of UCL Uncovering Politics, all you need to do is subscribe. You can do so on Apple, Google Podcasts or whatever podcast provider you use. And while you're there, we'd love it if you could take a moment of your time to rate or review us.
I'm Alan Reddick. This episode was produced by Conor Kelly and Eleanor Kingwell-Banham. Our theme music is written and performed by John Mann.
This has been UCL Uncovering Politics. Thank you for listening.
Literature mentioned in episode:
Abouharb (2022): https://www.tandfonline.com/doi/full/10.1080/14754835.2022.2122786