 So, we start from the idea that arm group governance is extremely common. By the way, I should say this is not about civilian protection or civilian resistance or anything related to this panel. I apologize. It's an important caveat before I launch in here. Yeah, I guess this is probably pretty obvious that it bears no relationship to this panel. So we start from the idea that a variety of arm groups engage in governance. Whether we're talking about rebels, militias, pro-state paramilitary groups, criminal groups, all exercise some form of governance in the territories that they control. Here we have a bunch of different examples. I won't read them out, but basically everything from resolving disputes to arbitrating and mediating problems between neighbors to providing health care education, etc. These groups do this governance in radically different ways, okay? Some displace, destroy, compete with local pre-existing mechanisms for governance, while others cooperate pretty freely with them and actually provide some coercive capacity for those local communal institutions. Others impose these really illiberal policies that seek to restrict freedom of movement, freedom of worship, consumption of drugs, etc. While others adopt kind of rules for participatory engagement that help shape how arm groups are going to govern in territories that they control. So we think that this variation in the ways that arm groups govern are likely to produce long-term consequences for the ways that civilians participate in politics, okay? For a variety of reasons, I'm not going to have a ton of time to talk through the theoretical mechanisms, but here are just some ideas. When arm groups enter a territory and they destroy local mechanisms for governance, this could obviously crowd in state institutions if they decide to work through state institutions. This could crowd in state institutions, but generally we expect that this should shape the way that civilians think about what governors, would-be governors should do, how they should do it, and under what conditions. So we're going to argue in this paper that there are basically two factors we think that are going to affect how civilians engage in politics even after conflict has ended, okay? In particular, we're going to think about the arm group's position to the state. That is, it can either be aligned with the state or it can be against it, right? So here we're thinking principally about pro-state militias, paramilitary groups that are nominally aligned with the state, or rebel groups that are nominally against it and seek to replace the state. And we're going to talk about different governance ideologies, okay? So in particular, whether arm groups adopt shared governance ideologies that seek to work in concert with local communal institutions and other mechanisms for local governance, including civilians, bringing them into governance, or a centralized kind of more top-down rule in which arm groups are interested in displacing, eliminating, or kicking out pre-existing institutional forms, okay? So we're basically interested in explaining variation both in the kinds of formal engagement in politics that civilians undertake. Here we're thinking about state-provided channels. This is things like voting, contacting politicians, and so forth. I'll talk a little bit more about that in a minute. But also informal forms of engagement in politics, and this is things like joining civil society organizations, engaging in protest, and so forth. And we think that this is important, well, because we think it's intrinsically interesting, but also because it has some policy consequences for the health of post-conflict democracies, right? So if arm group governance, for example, radically depresses engagement in democratic avenues, then this is of particular concern for post-conflict stability. And so I'll just very quickly kind of walk through some theoretical expectations, big alarm bells, or at least a warning that we don't find support for all of these hypotheses, and we'll talk about why that is in a minute. But basically, if you think about a 2x2 between whether an arm group is anti-state or pro-state, and if they adopt a shared governance ideology or a centralized one, we've got four boxes. So the first one is an anti-state arm group. This is roughly a rebel group with a shared governance ideology that's really interested in working with local mechanisms for collective action and governance at the local level. Those remain in place. And they're socializing, these arm groups are socializing civilians to reject the legitimacy of formal state institutions. And as a result, we should see something like this, which is protest marches, engagement in non-state forms of collective action. When we, on the other hand, have anti-state arm groups with a centralized governance ideology, we've got this same socialization to reject the legitimacy of state institutions, but the mechanisms for collective action are going to be destroyed, will be replaced. And as a result, you'll see this dude sitting on a couch, and there should be kind of deactivation. We shouldn't expect to see either participation in formal state institutions or in formal state institutions. When we have a state-aligned arm group, these are the paramilitaries and militias with a shared governance ideology. They leave in place mechanisms for engagement in local politics. We have the legitimacy of state formal institutions are reinforced. We should see increased engagement in state, in pro-state mobilization, like voting and contacting politicians and so forth. But we have a capacity for collective action that also remains in place, and so we might see additional informal mobilization as well. And then finally, just for sake of completeness, state-aligned arm groups, militias and paramilitaries with a centralized governance ideology that destroy local mechanisms for governance. We should see this increase in formal participation and no effect or decreased effect on informal participation. And we're going to study this in the context of Colombia, not because not only because that's where I live, but because it's a really interesting place in which, you know, in 2016, the Far Que Pe after more than 50 years of conflict decides to demobilize and leaves behind, you know, really interesting variation in exposure to rebel governance. Paramilitaries had governed. I'll talk about that in a second as well. And you have the creation of what are called pedets, which are areas that are heavily prioritized for implementation in which citizens are given a really muscular role in determining local priorities for peace agreement implementation. Okay, I thought you were going to tell me that I have like two minutes left and I panicked. Okay. What did arm group governance look like in Colombia? This is obviously a radical simplification, but basically insurgents were mostly known for letting pre-existing governance institutions flourish, and they collaborated with those local institutions. This is what Andrés Madagascar calls polycentric governance, what we call shared governance, but basically there was a symbiotic relationship between the Juntas Axon Comunal, these local communal institutions that provide FARC local legitimacy, FARC lends the hacks coercive capacity at the local level. And there's some interesting stuff that we can talk about in the Q&A maybe, but basically this is an anti-state arm group that allows for this polycentric governance, the shared governance to flourish at the local level. Paramilitaries on the other hand, basically sought to eradicate these pre-existing forms of governance at the local level, and they are state aligned. And so we should expect basically that pro-state forms of political participation in the post-conflict period in these areas will continue, will be strengthened, and probably communal forms of organization, participations in NGOs and so forth will be curtailed. Let's talk about the survey. So Abby and I and a bunch of other people ran a really, really big survey with 12,000 people, and we actually have a panel, so it's two waves. We're not going to talk about the panel today, we're only going to talk about the 2019 data, but basically this survey is representative at the level of each Pevets, which is each subregion prioritized for implementation of the peace agreement, and here what we're going to do is we're going to measure self-reported past exposure to arm group governance, whether by the FARC, whether by the ELN, whether by paramilitaries, and we're going to connect this to self-reported political participation, okay, to test some of our hypotheses. This just shows the distribution of respondents in our sample. This shows the distribution of responses to arm group governance in our sample. So basically what can we see here? Almost half report actually no arm group governance in their community. We can measure this in a lot of different ways. We do measure this in a lot of different ways. Here it's simply did the FARC, ELN, or paramilitaries ever rule this community? A question for each of those, obviously. We have roughly 38% saying paramilitaries ruled in their community, FARC in about 8.5%. We also ask questions about tax collection, about dispute resolution, and a bunch of other things. I'm just going to use this question for the purposes of this presentation today. Here, this just kind of shows you the geographic distribution of those, of arm group governance in Colombia according to our survey. We don't have a lot of time to talk about the details, but you can see that you have places where the FARC. This is more than 50% of respondents in each given place, responding that the FARC was responsible for arm group governance for governance in their community. This is the ELN, which is a smaller rebel group, and these are the paramilitaries. You can see very clear places where the paramilitaries are overrepresented basically as identified governors of these communities. This just shows you kind of, you may worry that there isn't significant consistency among respondents. If you ask each individual respondent in a given community who ruled you, maybe there isn't a lot of coherence among their responses. All that I'm showing you here, I don't know why that's bouncing up and down. That is frightening. That's not just me, right? It is bouncing. All right. Good. We can just see. What I mean to show you here is that with the FARC, you can very clearly see tremendous consistency within communities and responses to whether FARC governed that community or not. Same goes for paramilitaries. These results are a little bit more suspect. Here we disaggregate by urban areas and rural areas and the ELN. In urban areas, there's quite a bit of disagreement within particular communities about who was ruling. Okay. I see you're getting a number ready, so I should keep moving. This just shows the distribution of reports of kinds of political participation. We've got lots of different ways to measure this, but here's just an index of different forms of political participation, contacting local leaders, politicians, voting in the peace referendum, voting in legislative elections. There's quite a bit of formal participation, quite a bit less informal participation, and what we're going to try to do is connect these self-reported experiences with armed group governance with variation in this kind of political behavior in the post-conflict period. Many, many years later in many cases. We do a bunch of stuff econometrically to try to identify the causal effect of armed group governance. This is going to be really hard. We have no quasi-random exposure to armed group governance. There's nothing of the sort, so we're going to try to soak up some unmeasured heterogeneity in a variety of ways, both at the individual level with some controls, as well as at the community level with municipality fixed effects, and we're going to use population weights to correct for some imbalances in the survey when we're thinking about population size and so forth. Let me just show you kind of what we find. We find increases in informal participation, let's start with informal participation, that can be traced to both insurgent and paramilitary governance. This is surprising, for paramilitary in particular, because we thought that the destruction of local communal institutions at the local level by paramilitaries would depress informal engagement in politics, things like participating in marches, participating in protests, participating in NGOs, and that actually isn't the case, and I'll try to get to in a minute why that might be the case. We do see the expected increase in informal participation for the FARC, this is interesting in places where all groups were reported to have governed together, these are places of chaos, habesian kind of disorder, and we see a radical decrease in informal participation in these places. In terms of formal participation, another surprise, FARC governance actually increased formal participation, I'll get to that in a minute why that is, and we see the expected increase in paramilitary governance associated with increase in formal participation. How do we account for these results which are puzzling? We ask ourselves, have we mischaracterized how armed groups govern? We're not sure, and I'll talk about that in just a second. Is this driven by underlying preferences of communities where armed groups choose to govern? That is, there's a selection process here, in which armed groups select into particular kinds of communities, and this helps explain our results, we don't think that's true, like the existing literature says that armed groups choose what kinds of communities they enter, not due to the underlying political preferences of communities necessarily, but for factors that are kind of orthogonal to that, that have to do with proximity to state resources and natural resources and other things, so we don't think that's what's going on here, we can show econometrically that this actually isn't driven by IDPs or migration because we limit our sample only to those who are from a community, who have not moved into a community or moved out of a community, so these survive those kinds of tests, we know that armed group governance isn't simply measuring victimization because we actually have victimization data that allows us to say this isn't being driven by victimization, and so the question is like why, right? Why do we see FARC governance increasing formal participation? So one interesting thing is that the FARC actually taught civics lessons in places that it governed, and actually it would constantly engage in lessons about how the Colombian government wasn't living up to the Colombian Constitution, and so it could be the case that in the process of socializing local communities into these kinds of civics lessons, they've actually increased expectations about what the state can do and therefore increased engagement in state provided mechanisms for participation. On this one, this is a little bit more puzzling, I think, why does paramilitary governance increase informal participation? One potential mechanism here is that paramilitaries were far more decentralized than the FARC was, and so there could have been just simply more geographic variation in how mid-level commanders decided to either destroy informal institutions or work with them, the existing narrative is that that's not the way it worked, but it could be the case and I'm interested to hear of the folks who really know Colombia in this room if you buy that. So I'm going to kind of wind down here. One important thing to say is armed group governance happens at a very micro level, and normally we don't have data at a really geographically disaggregated level. So at the municipality level, if we measure armed group governance at the municipality level, there's a lot of interesting variation that's being obscured, so that's why we think surveys are actually a really interesting opportunity to do that, but we face all sorts of identification problems of the sort that I've kind of been talking through. I think that the general approach here is to say that we think that armed group governance has an effect not just on preferences but on behavior, political outcomes that survive the end of conflict. So we're talking about it could be a space of 30 years between the armed group governance experienced and when we're measuring these outcomes in the post-conflict period, and interestingly in the places where state authority was most challenged, was most at risk, most at threat, we see both more formal and informal participation. This is actually like a feel-good story in a weird sort of way. If we're interested in engagement in democratic politics. We've got a couple of things that we want to do next. We want to look at how these kind of legacies interact with variation in peace agreement implementation across space. We actually have, through the survey, a lot of data on how the agreement is being implemented, perceptions of that implementation, and then finally what's the relationship between these kinds of legacies in post-conflict stability? So all sorts of outcomes related to crime, violence, armed group governance today. These are interesting questions I think that we haven't fully grappled with. So with that, thanks so much.