 My name is Dan Bannick and I have the pleasure of chairing what I think is going to be the most exciting panel in The whole conference and that says quite a lot because this conference has been so wonderful We are I'm from the University of Oslo and we have a fantastic panel we have Two presenters here one discussant physical physically present and one online The the topic today is looking at state capacity how state capacity and the nature of social contracts determine the evolution of taxation capacity taxation states and Sort of a brief I'm a political scientist and you know when we think about capacity or state capacity we often think a little bit about simply the ability to get things done and We think about many sort of variables like it could be organizational cultures. It could be Resources the ability of the state to actually hire civil servants based on merit to insulate them from social pressures The legitimacy of the state in the eyes of the citizens all of these things are important In terms of understanding the ability to get things done And there's also quite a lot of literature about you know the distinction between the scope of states and the strength of States right so scope as in the range of international and domestic functions that the state is supposed to perform and Strength as in the the effective Implementation capacity and sometimes you have states that are very extensive in scope But damagingly weak in terms of strength particularly when there's a lot of corruption and patronage Now there's so many wonderful sessions here on state capacity on fiscal capacity and we have three Really really good papers that that further sort of nuanced this perspective between looking at state capacity and the ability to generate More revenue so one of the papers and I'm not going to mention which papers are trouble I was just thinking highlighting three aspects one has to do with trying to identify the institutional factors That determine state capacity and here of course the equal sort of allocation of resources if citizens receive Resources and they perceive this to be fair then state capacity increases correspondingly if there's corruption then you don't really want to pay taxes But in in this sort of new cycle of doom and gloom There's actually a lot of optimism out there and in one of the papers You will the presentations you will hear that since the 1990s the capacity to To tax or taxation capacity has actually increased in many low-income countries, which is really a positive thing the second point here has to do with looking at a Related aspect of cost of states capacity, but that is often termed as information Capacity so the ability of the state to know if obviously if the state has good accurate information About its citizens about the economy About different types of economic activities than the ability to tax of course increases on a side note I just we received my wife and I yesterday a letter from the Norwegian Department of Motor Vehicles our oldest son is turning 18 So and I was a little worried because we got two letters my wife and I from the same authority I thought I hadn't paid any you know some tax or something wrote tax. It turned out the state knows too much about me and my family Information capacity. It was a very nice message saying your son is eligible for a driver's license You know, you're welcome to this little meeting and we can help you and provide you with more information About how a good a driver a safe driver. He can be I Was a bit blown away with this It's like I feel that the Norwegian state knows too much about me in any way for information capacity is Important and finally of course we think about state capacity as in you know The Weberian state the role of the bureaucracy effective Implementation recruitment all of that it turns out social networks are also very important and in one of the presentations You will hear about how the state can use dense social networks for good purposes, but also for very bad purposes So state capacity linked to to social networks. So we have this fantastic panel We have somebody online at 2 o'clock in the morning. I think in Boston Leander Eldering from Northwestern University. Good morning, Leander. So I'm going to be very kind and let you start first But let me just also introduce Matias Vom Hau from the Barcelona Institute of International Studies Oliver Morrissey from the University of Nottingham And then we have you any widers very own Rachel Gisselquist who's going to be the the Discussant I have Informed the panel 15 minutes each around 15 and then around 12 minutes for Rachel And then we'll take questions from the online audience and then people in the room Please use the hashtag on Twitter rev for dev Rev for dev if you want to tweet and otherwise Welcome to the session and Leander. It's over to you great So thanks to the organizers to you for giving me the opportunity to do this online And I want to talk to you today About Rwanda, and I'm going to present a brief version of this paper that I wrote with my colleague James Robinson on State capacity in Rwanda, and I'm going to try and go a little bit quick because we have 15 minutes. So It's not a surprise to anyone in this conference that there are some consensus around The fact that some of the problems of development of fertility in Africa are due to the weak state It's one of the reasons why we have this conference and This consensus is bolstered with many disparate pieces of evidence across the social sciences so whichever science you're in I'm sure you have some Strand of evidence in your field of study that tells us something like this Here's a here's a picture. I like this is the PK market in Bangui and sent in the Central African Republic and Louise Lombard a very interesting ethnography of the Central African Republican state says that this is effectively where the state stops This is 12 kilometers away from the presidential palace just as an illustration now The sort of the intellectual tradition behind of this behind this is that We all start from waivers to a distinction between the rational legal and the patrimonial state We also know all of this so the African state is thought to be patrimonial So the way to do better is to move from being patrimonial to being rational legal now what we do in our paper is to question this consensus and We have three claims that I want to bring across today so our first claim is that in Rwanda, which is the focus of our study that The state looks quite weak on the barian criteria. So just to give an example Taxes raised as a percentage of GDP are very low But in fact, it has a great deal of observed capacity Now we talk in the paper quite a bit about the terrible genocide that happened there where it was very perversely evident That the state implemented this immense plan Which killed around 800,000 people in three months which had not been possible had it not had tremendous control over society Much more positively. We also documented in the growth miracle Subsequent growth miracle that Rwanda is going through now. The state is very intimately involved in many aspects of Economic life in sectors that are central to this miracle most centrally agriculture and We ask the question how So our second claim that we make is that the key to understand this is that the Rwandan state Basically as at his edges blurs into society So the state is very densely connected to society at a local level Bureaucrats are not really bureaucrats. They have jobs. They're unpaid But they're part of the state hierarchy They work for the government a couple hours a week a couple hours a day So you don't see them on the budgets. You don't see them on the books, but they are very clearly Advancing state objectives Now such sort of networks between state and society have of course been widely pointed out by Africanist political scientist sociologist economist. They have typically been interpreted as purely Patrimonial and redistributed. So these are avenues for people to extract favors from the state We argue conversely that this has also been an avenue for the state to implement policy and This can be very bad like in the genocide or it can be very good like during the recent growth Miracle now we have a long section in the paper and I think this is part of the part of the interest of the paper and it Licks up with some of the sessions in the conference that this capacity is his historically constituted So i'm going to skip over that now in the in in the interest of time um But the one point I will make about it is that because this type of capacity is historically constituted It is not sort of the general purpose Ability to get things done because you're tapping into social networks. You're Tying into society The scope of the things you can get done Through sort of networked state capacity are more limited than you can get done through Vibrarian or sort of more formal state capacity So this limits what the Rwandan state can do today and you know, I'd love to talk to about that With anyone who's had a lot of thoughts on that But what I would like to do is talk a little bit about like real data and real empirics because we're social scientists after all I want you to keep in mind two Samples two empirical samples one is a sample of communes 142 communes in Rwandan What we're going to try and do is relate the Vibrarian state To a measure of the networked state and what i'm going to try and do is show that these are uncorrelated The second thing i'm going to try and do is show you a sample of Rwandans in a lab in the field experiment. I implemented some years ago In 21 villages close to bottom out your side of an expanding of the border of an expanding historical state So the idea is that networked state capacity being historically constituted Historically constituted varies with exposure to the historical state So this is how we're going to measure the network capacity and what we're going to try and do Is show you that first of all this Capacity is an orthogonal to Vibrarian capacity And then that it does correlate with outcomes. We care about So let me do the first thing. This is the expansion of the historical state in Rwanda So darker shaded areas of this map Are areas where the state expanded into earlier And what we're going to try and do is relate this variation to outcomes. We care about This is the second sample the right part of this map The gray outlines are also in injected into the left map in black outlines The blue line is a River where the expansion of the state stopped for some time The little black outlines are villages in which we implemented the lab in the field exercise with 20 persons per Village, I'll tell you exactly what we did So how we're going to measure The Vibrarian state so at the country wide commune level we have various regular measures of taxes revenue Several measures of presence of the infrastructural capacity schools hospitals, etc In the field work villages We did a more detailed survey so we know whether there's a government office at police station We know corruption. We know bureaucratic responsiveness So and what we're going to do in the commune samples, we're going to regress the Vibrarian state capacity on network informal capacity In the sample from our field work exercise, we're going to Focus on whether a participant Is inclined to follow an Unenforced rule And so this is a very important part of the modern state that we sort of get buying Of individuals even when we don't stand next to them Controlling every move so I measure that in a field work exercise And what we're going to try and do is regress this on both a measure of network capacity and modern Vibrarian capacity in a horse race So this is a dense table But the way it's structured is that there is four panels And each panel has three or four columns and are different categories of the presence of the Vibrarian state So we measure all of this just before the genus small so not to have that contaminate our outcomes So the first panel is about taxes. So if taxes receive per capita trade taxes other taxes then the second panel is expenditure Then we have various panels that measure sort of infrastructural presence like various public services energy Water then is cooling at the bottom We have hospitals Social centers other things like that and then Agricultural cooperative markets trades commercial centers, etc The point here is that across all of these measures There is basically no association With what here is in the table is called state presence Which is our measure of the density of penetration of these networked relationships between state and society now Just as an interlude For some measures, you can push these results back in time. So you can measure schools in 1916 and missionary schools in 2435 and hospitals going back in time And there's no relationship at all between those outcomes So the Vibrarian state outcomes and the presence of the state historically Now we're switching gears into the field work Sample so here the unit of observation is no longer commune, but it's an individual And on the left hand side of the dependent variable is a measure of compliance with the rule And the first row you see here is our measure of networked state capacity And then columns include different measures of the Vibrarian state And so what we see is that people's tendency to comply with the rules today correlates with how strongly historically Their communities were connected with the state and not so much whether the state today sets up an office Or is more or less corrupt? Okay, so I'm going very quickly But these results are consistent with the idea that state capacity is multi-dimensional So it's no not just the case that being stronger connected to network for the capacity of the Vibrarian state No, this is an orthogonal dimension of the more broader idea of the capacity of the state to get things done So Is it the case that the network stage is uniquely associated with bad outcomes? No, absolutely not Let me let me show you The center graph here is from some earlier work. I did where we're on the x-axis We have the same measure network state capacity y-axis of violence in the genocide During the genocide the network state is positively correlated With violence in the genocide before and after these are the other two graphs the opposite relationship is true So the network state is associated with less violence So it really depends on the policies that the state are pursuing how this type of capacity plays out on outcomes So I think I may be moving To the end of my time. Let me say one one last thing which is that we have a large section in the Paper where we then push the argument to the modern Rwandan growth miracle So it's very clear that the Rwandan state tries to in the main micro level target and push around economic activity by ordinary Rwandans and it does this despite not Formerly employing that many people and despite still having tax revenues relative to GDP That are below what the IMF calls a viable level for running a Government and we argue again that this is done by people that are informally employed By the government but are in a network structure With the central state in Kigali Okay, so we make three claims in this paper first Even though the Rwandan the Rwandan state looks weak on observed barbarian criteria It has a great deal of capacity The second is that the key to understand this Is then dense links between the state and society the boundaries blurred The third claim is that you this is not a general purpose technology Like the barbarian state capacity might be because you socially interact with people in society what you can get done Is circumscribed by a social Contract so I didn't get much time to talk about that but perhaps in the discussion we'll have time to Talk Thanks very much