 So, hello everyone, my name is Guillermo Creman and I'm a post-doctor researcher at Carlos de Ferreira University in Madrid. What I'm going to present today, it's a working paper entitled Encartion of Leviathan to Draw Control and Post-Conflict State Capacity Evidence from Peru. But before beginning, I would like to thank Marcela and Jesse for being the discussant chair of this panel and also to Patricia Justino for the organization of the conference. So, since the end of World War II, civil wars have become the most usual type of conflict worldwide. Actually, what we know is that around one-third of the countries have suffered any kind of process of internal armed conflict. And as you know, civil wars are a very destructive phenomenon. But there is one concrete feature that is fundamental to address. And that is the fact that more than 50% of the civil wars relapse into war in the next five years after the termination of the conflict. Therefore, we need to understand and to ask ourselves, how do states sustain peace in the aftermath of war? And in this project, what I do is to argue that state capacity plays a fundamental role in this. However, when we take a look at a literature on the relation between state capacity and conflict, it's a little bit unclear. On the one hand, we have the tillion, the bellousest approach that argues that war made the state. The state made war. But we also have the opportunity argument that is mainly developed by fearon and leitin that argues that weak state capacity is one of the core correlates of civil war onset. So, concretely, the research question that I address in the project is the following. Does war time to control affect post-conflict state-building measures? And the core finding from it is that, yes, it does. In the aftermath of war, the state will mainly focus its state-building measures to areas that remain contested or that were under insurgent control during the war. And the main rationale behind this is to reduce the threat of insurgent revival. So, one of the things that we know from the civil war literature is that internal armed conflicts have a direct impact on existing forms of order and structures of authority, but it also has the capacity of shaping a post-conflict attitude at the individual level. Therefore, we need to focus on the endogenous effects of war. In this project, what I argue is that the combination of exposure to violence, fragmented patterns of authority, and novel types of governance will provide fundamental changes both at the institutional and at the individual level. And therefore, the type of social order that emerges at the local level in the aftermath of conflict is going to be completely different from the type of social order existing before the onset of the war. So how does the state deal with this situation in the aftermath of conflict? What I argue is that the state will need to invest in the restoration of two core characteristics of state capacity. First of all, on the restoration of state control through the development of the political, military, and social order, and this will be mainly done through the deployment of state forces to these areas. And the second characteristic, it will be the reinvestment in state legitimacy. The legitimacy of the state during the war might have suffered from important levels of erosions due to patterns of state repression against civilians, as well as probably on the effectiveness of the rebel ruling at the local level. And therefore, to offset these potential losses, the state will need to distribute public goods and services. Turning now to the different categorization of territorial control, I distinguish between areas that were controlled by the state, that were controlled by the rebels, or that remained contested during the war. So let's go one by one in here. Where would the state in the aftermath of war mainly invest? Let's start with state-controlled areas. These areas remain under full and effective incumbent control during the conflict, and therefore, these are the areas that suffer the lowest level of erosion at the local level. So what I argue is that if the state is rational and wants to maximize the level of a state rich throughout the country, it will be rational for them to put higher levels of investment, both in state control and in state legitimacy, than in areas that were contested or that were controlled by the insurgents. Regarding contested areas, the core characteristics of these areas is that they remain under a vacuum of power or in a situation of dual power in which neither the rebels nor the incumbents or the state have the capacity of having full control of these territories. One of the things that we know from the book published by Statis Kalivas in 2006 is that contested areas are the ones that suffer the highest levels of violence. But there is a concrete type of violence that is fundamental for this project, and that is the selective violence engaged in these areas against state representatives. Therefore, what we have is a reduction during war time on the level of state control. However, what we don't have in here is the establishment by the rebels of their own governance institutions. Therefore, what I expect is that the level of state legitimacy is going to be a little bit higher than in areas that were controlled by the rebels and that they had the capacity of establishing their own institutions. So what I expect in contested areas is that there will be higher level of investment in state control in the deployment of state bureaucrats, but not in the distribution of public goods and services. And in insurgent controlled areas, what we see in here is that these are the areas that suffer the highest level of disruption of previous patterns of authority. In other words, what we see is that there has been the highest level of erosion of the social order during the war. The level of state power in these areas was completely minimum. And the main difference with the other ones is that in these areas rebels had the capacity of establishing their own structures. One of the things that is fundamental to understanding here is that both during conflict and in the aftermath of conflict, areas that were controlled by insurgents are considered insurgent strongholds. And therefore they can serve as a signalling device of potential future insurgent reemergence. On the basis of this, what I expect is that insurgent controlled areas will receive higher levels of investment, both in state control, but also in state legitimacy. So in order to test these hypotheses, I rely on the case of the Peruvian Civil War in which the main actors were the state and the Maoist insurgency of Sandero Luminoso, but other actors also took place, such as the Comites Autodefensas, also known as Rondas Campesinas, or a small urban insurgency called the MRTA. Around 70,000 people died in the conflict, and most of the conflict took place in the peripheral and rural areas. Due to time constraints, I'm just going to focus on the quantitative evidence of this paper, but it's a mixed methods approach. For the quantitative evidence, I use data from the Peruvian Civil War, which is the census between 1961 and 2007, using the district as the main unit of analysis. Regarding the variables, I have two proxies for state capacity, one for state control, which is the number of state bureaucrats per district, and regarding state legitimacy or the distribution of public goods and services, it's going to be access to public electric power. And as the main independent variable, I'm going to use a variable coming from a paper published by Luis de la Cádiz in 2017, which uses electoral boycott as a proxy of territorial control. So as you can see in here, this is the distribution of the number of districts on the basis of the different type of territorial control that they had during the conflict. And what we see is that around 10% of them were controlled by the rebels, 7% were contested, and around 83% remained under state control. And this is a geographic distribution of the territorial control during the conflict. Regarding the empirical strategy, I use a difference in difference design, in which both warrants of territorial control will be used as time and group treatment. I include distributed fixed effects, as well as a vector of a time-barriering control variables in order to control for potential confounding effects. Going fast to the results. These are the results of the difference in different models without the controls, and these are the results when the covariates are included. But in sum, what are we seeing here? In those territories that were controlled by Sendero-Luminoso, we see that the level of investment in state bureaucrats is around 75% higher than in those areas that were controlled by the state, and the level of investment in electricity is 8 percentage points higher than in areas that were controlled by the incumbents. Regarding contested territories, what we see is that the effect for the level of state bureaucrats goes down to 37%, but still it's positive and statistically significant, and for electricity what we see is a positive, but insignificant coefficient. In sum, what we see is that post-conflict state control is mainly targeted to areas that were controlled by the rebels and then remain contested. However, the distribution of public goods and services is mainly targeted to areas that were under and such in control. These results are rather sort of different, to the different in different assumptions as well as to a variety of robustness checks. So before concluding, I want to acknowledge some of the limitations of this paper. First of all, I want to be very clear that I'm using in perfect process, both for state capacity, but particularly for territorial control. The second problem is the usual reverse causality issue in here, the treatment, the type of territorial control is not exogenous. The third one is that there are important data constrained, for example, it would be great to have sub-national measures on rebel governance, not only on territorial control, or for example, patterns of displacement and resettlement during war and in the aftermath of it. And another limitation are the stringent scope conditions of the project and mainly focusing on irregular wars in which states have at least mid-levels of capacity as they have the capacity of distributing public goods to the population. I'm only focusing on wars in which the rebels are defeated, and fourth, I'm only focusing in insurgents that have the capacity of establishing their own governance structures. So these are conditions that must be acknowledged for understanding of the results. So to conclude, what I've shown is that we also need to focus on some of the potential side effects of civil war, as civil wars can be a good way of bringing order and development at the local level. What I've shown is that post-war state capacity at the local level seems to be informed by war-time dynamics and more concretely, not by violence, but by the spatial distribution of territorial control. And in here, the state is a strategic in the deployment of state building. It mainly targets those areas that had higher level of exposure to rebel influence. In other words, those areas that were controlled by the rebels, and also to those areas that remain contested during the war. Therefore, the outcome of the bellessist approach that war made state also seems to apply sub-nationally for civil wars. Thanks a lot.