 One of the things that we find in a lot of the attempts at improving governance in developing countries and to improve anti-corruption is that a lot of these programs or approaches fail because we don't ask what are the incentives and the capabilities of insiders to engage in this peer group monitoring and to provide the information, not only the information on which enforcement can be based but also have the capacity and the voice to engage in some of these enforcement themselves. In other words, unless information and enforcement are both at play, governance improvements don't happen. So this is a very fundamental observation of our whole approach to anti-corruption. In each case, the problem is that for the governance to work, you need to have a feedback loop which generates information about what's going wrong but also creates incentives to act on it by actors who have the incentive and the capacity to act. As a system, the top-down enforcement which you take for granted often in advanced countries where when you identify wrongdoing or you identify evidence which says something should happen, you can rely on enforcement mechanisms and systems which will enforce certain rules. This is absolutely not the case in developing countries to varying extents, ranging from very fragile countries and Jonathan will talk a little bit about that to even the relatively well-working developing countries, that loop is very weak. So unless you also ask yourself who is going to enforce this policy or use this information, then your attempts at governance or anti-corruption don't work. When the COVID-19 pandemic hit us last year, we started looking at some of the health systems in the countries we work in. The fear was that they would have catastrophic failures and the fear was that they would not be able to respond to this crisis at the scale and with the efficiency that was required. And indeed in many cases, the outcomes were extremely poor. But the really interesting thing is that the outcomes in many other cases were not as poor as many of us feared. And so we started looking at what was working and why and we found that there were some patterns in the bits of the system that were working better. Those patterns have some implications which go beyond the pandemic response and might help us to rethink how we think about systems organization for improved governance in developing countries. So this research is at a very early stage. A lot of information still has to be collected, but we wanted to share with you some approaches at this very initial phase. So we know that health systems are very centralized and for good reason health systems are centralized because you want to have coordination between different parts of the health system and a lot of these bits are interdependent. And so in every country, you find that the bulk of the health, I'm talking about developing countries, is provided by very centralized public sector health delivery systems with the private sector and so on operating side by side with that. Particularly in context of poor rule of law and poor governance, the monitoring of flows from the top down into this vast pyramid of activities which is the health system has always been very weak. And in most developing countries that has been organized with a lot of informal transactions which you can call corruption and sometimes it's just not corruption but informality, a lot of problems are solved informally. So when you start pumping massive amounts of resources into this really fragile and informally organized badly controlled system, you are likely to get catastrophic outcomes. The interesting thing is that the response was not linear. A bad and fragile system did not completely collapse in every case. There were many counter movements which emerged spontaneously with innovative health sector individuals, entrepreneurs, NGOs, concerned bureaucrats and politicians popping up spontaneously because they had strong incentives to find solutions. This did result in well known problems of coordination and corruption did spike but it also had a positive effect. We found that by allowing these multiple networks to mobilize, a number of things happened. It generated information about what was working and to what extent but it also created checks and balances between these parallel systems because there was a kind of friendly competition between them. And those, that peer group monitoring that actually we can deliver this cheaper than the state or the state's public saying actually it turns out that we can do it better, that check and balance had a remarkable effect in not only generating information but acting upon it. And this is the really missing thing often in developing country governance but there were a few conditions which made this work when it did work. The first condition was that you have to have more than one potential delivery agent with the capability to deliver. So there has to be a minimum, you know, multiplicity of capabilities. So this is not the case in every context even within the same country. So it only works where that condition is met. Secondly, there has to be some degree of coordination and feedback so that there is a coordinating body which is looking at these multiple deliveries and saying actually this is not working and that is working. And that coordinating body must have the incentive to do that. Now in a pandemic, there's a very strong incentive to do that because a prime minister or different levels of the health authority are under pressure to deliver and so that coordination happens. And this is a really strong learning that we have to identify the appropriate level of coordination where these different mechanisms can be assessed and then scaled up or scaled down. The third observation is that there was redundancy built in by definition but this redundancy, that is you are trying different things at the same time, actually was less costly than the alternative. And to make it less costly than the alternative, obviously the redundancy had to begin on a small scale. If you try to do the whole system in multiple ways, obviously that would not work. That kind of redundancy would cripple any fiscal system. And finally I want to end with a couple of observations before handing over to Jonathan. The first observation is that the least cost and most effective delivery system is not defined by who can procure at the lowest price and who has the lowest unit cost in terms of salaries. The least cost and most effective delivery system, as we all know, depends fundamentally on the internal governance of that system. In other words, if the governance is good and you are managing your delivery properly, then that is the least cost system. The second observation is that the implicit competition between these multiple delivery systems was extremely beneficial in triggering socially desirable behavior. Now, we often think of how to constrain public actors by removing or reducing their discretion because we are always worried that if they have discretion, they will do some bad things and often they do. But this is an example where by creating discretion but controlling it with this lateral competition, you actually generated a lot of pro-society or pro-social behavior. And finally, the final point is corruption did not go down by any means. There were many factors pushing corruption up during this pandemic but the improvements of governance meant that the corruption was contained and you had better outcomes in sectors where this multiple adaptive system could work. There is a different route forward perhaps which is to have coordination at the center of the public delivery system and have multiple adaptive systems on a small scale experimenting and competing amongst themselves to show effectiveness and then you scale that up. But you don't scale that up or scale it down permanently. It should be a continuous process of adaptation and experimentation. And I think this is the big learning from the pandemic. And I think it's incredibly important at this time. I think these learnings and the ability to formalize them as you're doing now is extremely important for health responses to pandemics but also other health crises. From a top line perspective, the most important contribution that this framework makes is it provides a set of anti-corruption solutions but in a way that doesn't inhibit but in fact actually advances some of the other key operating imperatives that you need in a health response. Namely, it promotes a resilient approach, an adaptive approach as Moustap mentioned, and also a scaling approach. You see a lot of inter-corruption work trying to limit and control and become very rigid, reduce discretion in performance but that comes at a great cost of then limiting experimentation and flexibility which is so important in emergency responses and longer-term engagements and factor basis. What's fascinating for me is that a lot of all of these different contributions of the framework, they all hinge on this notion of redundant capacities. The first one is this inter-corruption outcome about weeding out which is allowing through the process of invitation it to become clear who the high performers are, who are actually delivering on health outcomes. And in terms of lower performers on one level, they're probably low performers because they have very limited capacities, assets, human resources but there's another essence here of they could also be low performance because they're not really in the business of high performance. They're actually coming to this with a sort of a rent seeking mindset, this framework to sort of pet contracts, fine leakages, etc. And this framework, this redundant capacities, this competition as Mushna referred to it, kind of weeds that out. The other pieces I call sort of a disciplining measure or function here which is to say that you probably have some really high performing or high capacity to partners but who are doing corruption, who are opting for operations in the health sector for private gain. And when they enter this framework, this horizontal network as you've called it, they see pretty quickly that this is a competitive piece and that they need to move above board to actually deliver on these efficiencies and then produce effective outcomes. And so they're disciplined in this way. And the last piece is perhaps the most interesting for me, the last measure of anti-corruption that this model introduces, which I call the enabling function. And I think one of the most important contributions of ACE's work more broadly in the interruption space is to recognize that there are firms and other kinds of implementing partners that would, under other circumstances, choose not to be doing corruption. They actually want to be working in legitimate ways but there are such barriers to markets or particular sectors like health that they really have to use corrupt measures to kind of pay their way into functionality or survivorship. When these types of partners are allowed to enter the horizontal framework with its fast track mechanisms for bureaucratic permit licensing more streamlined approaches for procurement, which we would imagine in a kind of a highly politically supported type of framework or network, they then become able or enabled to act with more effectiveness and efficiency. And what I want to say now is just how I think this model enables as opposed to tracks from more adaptive resilient and scalable solutions as well because you can't have anti-corruption measures again that undercut other key imperatives for a health response. So the first one, as Moushtak alluded to as well, is this adaptive component. So it's very hard to know at the outset what the right solution is in a place. You have a technical set of solutions that you know you need to do, contact tracing, for example, lockdowns, etc. But how you do that in very different political, economic, and social contexts is not apparent from the outset. And we take the technical approach and align it to the political, social, and economic configuration that you need to actually make it work. For men in the adaptive community, something that they often see as a kind of a strange bedfellow with anti-corruption measures, which really try to have an ex-anti-pre-conceived solution and then create linearity and predictability so that they can control every step of the process. So what this model does is it provides a way to be adaptive, to find out what's working and how to build that up or scale it, but also in a way that advances anti-corruption. Anti-corruption isn't really talked about in the health sector, in the way it is, say, in the economic sector. So this brings into focus a way in which you can do anti-corruption in a way that promotes resilience as well. And as you know, Palafian particularly fond of the argument that Moustak articulated that sometimes when we talk about resilience, we talk about as creating excess capacity. And that seems to be at loggerheads with efficiency. But in fact, your arguments are that you need resilience in the system to promote efficiency. There's a very antiquated notion of scaling right now in the aid industry that sets up this notion that you have to do a pilot project, often a very small one somewhere. And if that works, then you scale it. And we're increasingly trying to move away from this and towards this idea of outcomes-based scaling, the idea that you're promoting an outcome like health, like anti-corruption, like resilience. And you're allowing multiple scaling partners to work in this kind of scaling consortium or horizontal network, as you called it, in a way that allows you to see variation, what's working where, who's working and where. And from there, scale up what's really working and scale down what's not working. The next part of the discussion, which is really exploring, in a sense, how can we bottle up these learnings from this framework on crisis response, the broader learnings, the broader intuition of this framework of scaling, which is really about scaling for revealed competencies, building resilience, and in the process emerges an efficient, relatively non-corrupt outcome. And one of the sectors that the anti-corruption evidence program in SOAS has done a lot of work on is the electricity sector. And it actually got us thinking, hang on a minute, are there aspects, as I said, within the broader intuition of the framework, are there aspects of which that can actually be transferable to the electricity sector? And obviously, at first here, this jump to the electricity sector might sound like a huge stretch. There are some broad mechanisms which might actually work. And just to give a brief headline of how the electricity sector works across many developing countries, the enforcement of rules in centralized electricity delivery systems is weak and can result in poor outcomes. So as in the health sector, the focus tends to be solely on procurement rules or it tends to be on let's just enforce the contracts, never mind whether there is the capacity to enforce or the incentive indeed to stick to the rules. And obviously, it doesn't work very well in the context where the rule of law is weak. Specifically, in the context of Lebanon and Nigeria, we also see that it's politically captured to the extent that any reform of the national grid within the short to medium term is actually not possible. And we're also looking at almost levels of crises, certainly in Lebanon to a lesser extent, but also through Nigeria or under supply to the extent that those countries productive capabilities are actually hugely constrained. So what we are suggesting is taking a step back and scaling down from the grid down to local networks. Now, the recent literature on decentralization has suggested that decentralization might lead to greater amounts of rent seeking. And that's probably true in the sense in the case of administrative decentralization. But like with health, it might just be the opposite in the case of electricity generation and supply. So downscaling to local networks using disaggregated decentralized mechanisms, sometimes involving efficient but informal arrangements is possibly the way forward in the short term. It's a way of thinking very differently of responding to an electricity supply challenge, which can't have a more sort of immediate response like we do in health, but it's certainly, you know, on a more achievable scale than looking at the entire grid and saying how do we improve transmission, then distribution, and all the political nature of capital that takes place from end to end. So essentially what we are talking about here is not the physical delivery mechanism. That's not the rapid scaling that we are talking about. What we are talking about are different de-risking strategies, different contractual mechanisms across segments of users within a country that can be scaled as required. In Nigeria, we worked with small and medium enterprises in what used to be the beating heart of Nigerian manufacturing in the Aba region, in the southeast of Nigeria. And it actually turns out that there is what we call willingness to pay. Consumers who were once forced to be corrupt, once they realize that there is a technologically efficient mechanism which provides them electricity at an optimal tariff for them, no longer want to be corrupt. This, you know, creates a virtuous cycle wherein you get an investor who sees the willingness to pay as proof for investing in that locality. And these investors tend to be more nimble, more technologically capable because they're coming in with either solar. And in the case of Nigeria, because it's really more a gas economy than an oil economy, we're looking at CNG. In sectors like power, investments are lumpy, and checks and balances have to be created by designing different delivery mechanisms. So you test, this is working in a particular cluster, a particular cluster might need a different kind of a supplier, a different kind of financing mechanism and what's very critical in countries like Lebanon and Nigeria. And this speaks to what Jonathan was talking about, the community aspect, you know, the community bind has to be very high. We worked in Zahle in Lebanon, which has its own sort of sectarian complexities in Nigeria with the SMEs, they have their own political economy complexities. Community bind is very important. But once you get that community bind, once you get that willingness to pay, you can actually get in less politically connected investors. And this by itself, becomes a sort of anti-corruption program, not sort of becomes an anti-corruption strategy. Because when you're improving development outcomes, which is you are ensuring that more electricity supply, rent capture isn't taking place and people are able to improve productivity by definition, by increasing efficiency, you are reducing corruption. This is not scale on the lines that Mushtaq talked about for the health sector. But as he said in the beginning, these are early days, and we are doing so much interesting research, we've been having conversations with Jonathan and his team of colleagues. And there are very interesting ways in which we can join this thinking up.