The following is a lightly edited conversation between two people working in research and policy. We felt the value of including this kind of conversation is as a form of meta-commentary on research and policy work itself – the questions, doubts, emotions, and thinking process that sit alongside the strict subject matter but rarely make it into publications.
We find from our own experience that reports do usually drive some discussion and commentary by those working in the field. Many reports are published by many organizations, and many have been in years past, so what makes a given publication worth putting time and mental energy aside for an overworked staffer, busy senior official, development organization heads responsible for multiple countries, or experts working on a range of projects, to go through and digest fully? We can point to the important subject matter, the impact it would have if implemented, the transformative impact that policies based on these findings could have, or the stakes for those affected by the status quo – but a lot of other work also deals with important topics that can have large impacts and have high stakes for those affected by the status quo.
The question isn't about whether the topic is important. What we want to reflect on here are the questions and doubts, the emotions and worries, the balance of hope and despair or motivation and burnout. And for junior staffers, researchers, students, or interested members of the public, we also aim to show insights into the thinking process around research and policy work that hopefully give some useful perspective.
Transcript
[A]: When you spend time going through research on people who've experienced severe harm – reading about systemic failures, about the gap between what the law says should happen and what actually happens – it's difficult not to feel a kind of despair about the whole system. You read about people whose legal protections were violated repeatedly with no consequences, whose lives were spent being failed by the institutions that were supposed to help them. Honestly, it has me feeling a bit like "the system is broken, nothing works." Sometimes it even makes me feel like, what's the point of these reports and publications? If people will still keep on suffering because what we do in our offices don't meet the scale of the problem?
[B]: I've spent a lot of time in that exact place emotionally over my career. There are a lot of times where learning about the circumstances experienced by people out in the world makes me feel so much hopelessness and despair. I think that's actually a trap. Because when you care about a topic, when you're reading about horrific situations and systemic failure, you can swing between righteous anger and cynical hopelessness. Both feel like the right response. And in a way, they are! The way things are demands righteous anger. The trap is how righteous anger can be emotionally draining and then turn into cynical hopelessness. And when you feel just hopeless about it all, that fatalism can sometimes be paralyzing. So I have to regularly tell myself to keep focus. The people who shared their experiences didn't do it so we could feel appropriately devastated. They did it hoping something could change. So I have to keep asking myself: am I engaging with what's actually fixable here, or am I just cataloguing injustice in a way that feels morally engaged but is actually useless?
And I think there's something I have seen in and appreciate about you – I've seen it across all the time we've known each other as colleagues in this field – where the passion for justice, which comes from exactly the right place, can sometimes tip into a kind of fatalism. When you're so affected by reading about horrific situations that you start wondering whether institutions or workers in government could be trusted at all given the pain that keeps being suffered, or asking what good anyone was doing if this keeps perpetuating. It's a natural response. I think it's actually the right response, because this should be driving us to want to dig in and defend people from the status quo that causes them such harm. I think the risk is where it can also make problems seem less solvable, make systems seem more intentionally hostile than they are.
Sometimes systems perpetuate oppressive status quos not because the people in them are hostile, but because the bureaucracy itself is badly organized, understaffed, using manual processes that should be automated, staffed by people who learned on the job and don't have the expertise to improve their own workflows. Those look identical from the outside – cases don't get handled, orders don't get enforced – but they require completely different interventions. Training someone who simply doesn't recognize a problem because nobody ever taught them what it looks like is different from confronting someone who actively doesn't care. Both are problems, but they're different problems.
I mean, think about the people we meet at all these conferences and events, people we work with from across agencies and organizations. They care and they want things to get better. In general, I think a lot of people, a majority of people, want things to be better. Including for the institutions they're part of to be better. And the thing is, they have made things better! There has been real change. There are real people out there whose lives are better because of the work done by those people we know and admire in institutions that we work with. From institutional interviews across different research, we've seen examples where some enforcement mechanisms do work. And from reviews of literature and previous publications, we see that past recommendations have become policies that affected real lives. There's still a long way to go. But we are going. Like when you're staring at the moon, it doesn't look like it's moving, but it is.
When I first started this work, I had to learn this distinction myself. It's easy to look at outcomes and infer intent. But a lot of what looks like malicious neglect is just institutions that are bad at improving themselves. Regular staff know their subject matter but they're stretched just keeping things running. They don't have the expertise or bandwidth to step back and redesign systems and policies. I've seen courts that can't track compliance systematically because their data systems are manual – monthly statistics generated by staff from different departments sending individual Excel sheets that someone manually compiles. They can't easily tell someone their case status without looking it up individually. These are fixable problems if someone actually helps them fix them. So you have to not lose hope about how we can impact things. I'm not saying this as cope – I've worked in this and have seen firsthand how much of what a lot of people blame on intent is just poor capacity or outdated barely working systems staffed by people who understandably lack the specific domain knowledge or expertise to change the systems and have only learned on-the-job to continue doing what they do daily.
[A]: Here's another concern I have. I'll be honest. A lot of research in this space comes with very long lists of recommendations and I feel like that can be a liability – if trying to do everything means you end up doing nothing well. If you had to pick five that would actually move the needle, which would they be? Because unless there's some specific reason, a list that long starts to look like you're thinking too big without prioritizing. Is there a reason to go that path? I feel like there's a risk of trying to do everything and potentially doing nothing well.
[B]: That's a good question and it is something worth considering carefully. It can feel wrong or at least unorthodox to have a very long recommendations list without having a good reason for it. But the thing is, thinking about what the problem actually looks like – in a lot of the work we do, someone's trajectory through vulnerability might touch education, the court system, economic structures, housing, police response, enforcement of obligations. Intervening at one point doesn't address the compound nature. It's a "death by a thousand cuts" kind of situation.
A long list of recommendations doesn't have to be a sequential plan that one actor has to execute. They can be a wide net of narrow recommendations directed at different actors. NGOs for awareness sessions, professional associations for sensitization of their members, courts for procedural reforms, local councils for community initiatives, corporations for CSR possibilities, development organizations for aid or expertise, and so on. Different organizations doing what they're actually equipped and able to do, in parallel. And you're asking each stakeholder for relatively narrow, specific commitments within their existing capacity. You're not asking a teachers' association to reform the court system – you're asking them to help their members recognize warning signs. That's bounded, achievable. A lot of this work is coordination across actors rather than sequencing.
[A]: That's another thing. Some recommendations are relatively big and do address root causes, yes. But there's also a lot of those narrow recommendations you mentioned. Isn't that just addressing the symptoms without the disease? Do you worry about whether including narrow recommendations risks blunting the more crucial ones? Or feel like maybe it's encouraging a technocratic way of approaching these issues that will prevent the needed root cause changes?
[B]: That last question, I do worry about that, yeah. You have to emphasize that narrow recommendations aren't replacements for the big root causes. In any given context there are massive structural factors – social norms, economic insecurity, geographic disparities, housing – that create the conditions for harm. This shouldn't be approached as a purely technocratic problem. But I think including even the narrower recommendations is the right way. Partly because of the wide range of stakeholder commitments aspect – you want to include things that many stakeholders across society can all participate in in some small way, for problems to be addressed quicker while also having figures across society feel a sense of ownership or investment over these outcomes and see the value in achieving them as a society. Broad points alone wouldn't have any menu items for a lot of the stakeholders you do want to participate.
The other reason is, it goes back to the "death by a thousand cuts" thing too. So there's a lot of interventions that each address one cut. Some of them seem too small or narrow, maybe they even seem unambitious like they aren't actually going to address the root causes at all. And the root causes still need to be addressed. But in the meantime, in parallel, there's still these small narrow interventions that do matter. They might be small but even one person is one human being where some intervention could alter the entire trajectory of their lives, be the difference between the context of an entire life. So I do believe that also trying to pick the "low-hanging fruit" of smaller changes isn't somehow inconsequential or abandoning the bigger issue.
There are these still specific, identifiable things that could have been different. A lot of those specific things are symptoms of a bigger root cause or problem, yes, but there are still things where the root cause may have created that initial status quo but it's not going to regenerate it if we close it. Some of these symptoms yes, it's whack-a-mole where closing one will just have the issue pop up elsewhere. Like, misogyny within a society will find ways to manifest itself because it's this corrupting sickness in society and it will infect the fruits of society as long as we can't achieve a healthy society. But we can prevent some of the harms still, right? It's like, if we carry this metaphor of a sickness or poison within society, some of it will be pumping out these toxins into the air, and for those, even if we keep air filters in our homes, our air will never actually be clean and the toxins will flood in and keep making us sick the moment we stop until we can actually plug those pipes. But for those poisoned fruits, if we can provide people with other food sources so they never get that poison in their bodies, even while we go looking for ways to get rid of it from the trees. Changing the root causes and changing society as a whole is a necessity, but we also want to do what we can when we can for now. A lot of vulnerabilities identified in research do have very specific, identifiable points of harm. That exact point of harm alone, even by itself, does real damage. So when you hear of these narrow and specific issues, cutting off those little sharp thorns still matters. Looking at every individual example – if you'd prevented that exact point of harm and even if that's all you changed, that changes their entire lives, and that's still important. The people who share their experiences identify specific things that would have made a difference – knowledge they wish they'd had, support they wish had existed, enforcement they wish had happened.
And I do think that the distinction between symptoms and root causes are also somewhat intertwined. One of the biggest root causes you keep finding in this work is the way that childhood context affects people all their lives in ways that manifest in society ten, twenty, forty years later. People bring what they see as children into the rest of their lives. And the intergenerational cycle of trauma and violence also keeps churning along every day, the wheel turns just a little every day. So you want to include recommendations that can cut out some specific vulnerability fast, to prevent some of its transmission to the children growing up now whose outcomes depend on whether we get this right – that's the urgency. Every year we don't address these issues, more children grow up in unstable homes, witness violence, live in financial precarity. The intergenerational transmission continues until something interrupts it. Even case workers emphasize that harms are cyclical. Sometimes someone recognizes that cycle and breaks it for future generations in their families, and the support we provide can hopefully create an environment where they are better equipped for recovery and thriving in a supportive society with effective systems and institutions around them. Sometimes, some kind of harm begins a cycle, and hopefully every change that gets implemented can prevent harms that might have begun a generational cycle.
[A]: One thing that's stood out to me across various research is how information interventions can sometimes matter more than you'd expect. The general policy literature on this is pretty discouraging – you can tell people things but it doesn't usually change behaviour much because behaviour is driven by incentives and constraints, not just knowledge. Budgeting class isn't going to transform financial decision-making, because people aren't spending because they don't know how to make a budget, but because their circumstances affect their spending needs, or they don't have the time and mental bandwidth to keep track. So I'd normally be sceptical that you could move the needle just by telling people things. But there are cases where it's different – where people genuinely don't have the concepts or categories to name what's happening to them. That's not a "people know they should save for retirement but don't" type of problem.
[B]: Right, and I've had my own scepticism on that challenged. I expected information and education components to be relatively minor – something you'd include because stakeholders wanted it, but not something that would actually move the needle. You can't just tell people about behaviour and expect change, because usually there are specific incentives and underlying factors for people's suboptimal behavioural patterns. But what I've found in my work is that sometimes you encounter something different from the usual "knowing but not doing" problem. These are genuine information gaps – people who literally don't have the concepts to name what's happening to them.
In vulnerability research, for example, you find women who didn't conceptualize sexual coercion within marriage as abuse at the time it was happening. They only understood what had been done to them years later, after learning about their rights. That's not a behaviour change problem. It's a category problem. If nobody has ever told you that something is wrong and is illegal, you don't have the framework to name what's happening to you, let alone seek help for it. Similarly with psychological abuse – if your working definition of abuse is limited to physical violence, you can experience constant criticism, isolation, threats, humiliation, financial control, and not categorize any of it as abuse because it doesn't fit your mental model.
And when you find that people who went through harmful experiences learned through painful experience things they could have been told beforehand – and when multiple people explicitly say that if only someone had told them certain information earlier, it would have made a big difference in their lives – that's when information interventions stop being symbolic and start being substantive. The standard policy model of people having concepts and knowledge but facing barriers to acting on them doesn't apply in those cases. You're not trying to overcome behavioural inertia or compete with present bias. You're introducing entirely new conceptual frameworks. And you can reach people young, before patterns are established.
[A]: That makes sense – the layered approach of early education planting the framework and reinforcement at key decision points. What about the pushback question, though? Cultural or religious concerns about what you're teaching?
[B]: The pushback question is real, especially given the cyclical nature of a lot of these harms. If sessions help young people recognize which behaviours are harmful, some of them are going to recognize those patterns in their own homes, with their own parents. For some, that recognition is actually the intervention – a kid who grows up watching harmful dynamics and develops the framework to think "this is wrong" is on a different trajectory than one who internalizes it as normal. It might affect their future choices, how they respond if they experience harm themselves, whether they intervene for their own children someday. The intergenerational transmission has to be interrupted somewhere. But for others, there might be resistance – it's hard to accept that your parent might be an abuser. They could dig in, there could be defensive reactions, parents could get angry, community backlash. But that tension exists with any education about healthy norms. You can't not teach these things for fear that some people will recognize their own situations. The discomfort of processing that is preferable to us being okay with a status quo of perpetuating the cycle.
[A]: Let me ask about a different kind of challenge – geographic and resource constraints. In a place like the Maldives, or really anywhere with dispersed populations, services are concentrated in one urban centre while most people live far from it. You can't put professionals everywhere, so approaches like training community members come up. But that's asking a lot of unpaid or minimally-supported people, and it creates its own risks around confidentiality, burnout, and conflicts of interest in small communities.
[B]: The geographic concentration issue is massive – in the Maldives specifically, arguably the two issues at the root of almost every major social issue are geographic disparity and housing availability/affordability in the capital, and even those two are closely linked. But this applies to a lot of development contexts. Decentralization is such a massive institutional factor that even if you recommend it, you also need to work around it with the likelihood that sufficient decentralization to really solve the disparities overall won't happen. You have to target what CAN be done with what limited resources exist locally. Build on what's already there.
A community-based approach is meant to work within those constraints. The intent isn't to train volunteers to do the work of professional case workers on a day-to-day basis. That would be completely unrealistic. Instead, it's to teach people some key community-centric elements toward helping in little ways – looking out for vulnerable people, identifying signs of harm, creating a more participatory supportive community through concrete activities. The selection challenge is real, though. In close-knit communities, your volunteer could easily be connected to the person causing harm. You need people with respected roles and some level of independence – teachers, health workers – who are likelier to be familiar with duty-of-care concepts and who have some protection through their professional standing. And you have to accept that no single approach works for everyone. You don't need every touchpoint to work for every person – you need enough touchpoints that most people have at least one that works for them.
[A]: What about the children dimension? Across a lot of this research, children seem like the primary victims – not just witnesses during conflict, but bearing the worst consequences of family instability. Children who don't have anyone to turn to, who lose support systems, who blame themselves. Attachment research showing lifelong impacts. The pathways into crime that come from children seeking belonging outside a broken family.
[B]: That's exactly why I think framing matters – "Child and Family Well-being" rather than just protection or response. Those response framings aren't wrong, but the children who've witnessed years of conflict, been through contentious custody situations, lived with financial instability – that trauma is already internalized.
And it does seem like family breakdown is a real risk factor, but when you look deeper, it clicks that breakdown itself isn't the problem – it's how it's so often accompanied with very upsetting circumstances. Conflict, instability, financial inability, children being involved directly in adult disputes when they're unequipped to process having that role thrust upon them at a stage where they want their parents to have a full and unconditional role in their lives. I don't believe that the rate of family breakdown is an indicator to be targeting, especially if it incentivizes policies that make it harder to leave unhappy or harmful situations.
[A]: Enforcement of obligations like child support feels like it could be really high-leverage if you can crack it. Right now in a lot of contexts, the incentive structure means people can refuse to meet their obligations because they don't feel that commensurate consequences will come. If you could shift that calculation meaningfully, even imperfectly, it changes the whole dynamic.
[B]: One major challenge is that we often lack the detailed data to know where the real problems are. When we have better data, it can completely change our understanding of what needs fixing. This applies broadly to every area of policy. Without sufficient data, it's hard to know how exactly to address problems. Let's say for example that there's some very prominent or even viral instances of some kind of significant policy error on social media, and that what it shows is a massive problem. The ways we'd want to target that problem would be affected by data. Like, are these cases representative in which case a major overhaul is needed, is it revealing a specific design error in which case some targeted fix is needed, or are these rare examples in which case a fallback option to prevent harms for what falls through the cracks is needed. Three very different policy responses that would be counterproductive without info. In that example, a fund that covers expenses for the fail cases would be perfect for option 3, but would be an impossible budget drain if it had been option 1, right? Or announcing a targeted fix that might be perfect for option 2 would keep an infuriating problem ongoing if the issue had been option 1.
[A]: Institutions say they're committed – courts, police, government agencies all express buy-in during consultations. The question is whether that sustains beyond the initial momentum. Implementing many recommendations across a dozen stakeholders over years, with political changes and staff turnover and competing priorities – that's hard. Easy to commit in a meeting, harder to follow through when it's year two and nobody's paying attention anymore.
[B]: But the roadmap is there. Research documents what's wrong and what's needed. Recommendations can be specific and achievable. Stakeholder structures create some accountability. It's not inevitable that things will change – but it's also not inevitable that they won't. The problems are real, the gaps are documented, the pathways to address them exist. What happens next is up to the people with the power to act. And the people who keep pushing them to act.