In Pursuit of Development

Corruption and political settlements — Mushtaq Khan

Episode Summary

Dan Banik speaks with Mushtaq Khan on whether definitions of corruption matter, the relationship between corruption and economic performance, and the impact of efforts to combat corruption in Bangladesh.

Episode Notes

One of the dominant explanations for elusive development in many parts of the world is the negative role played by corruption in the development process. And many national and local governments as well as international aid agencies have spent considerable time and resources trying to come up with plans to combat the corruption menace. But anti-corruption policy has often been difficult to implement and many well-intentioned efforts have had limited impact. 

Despite the challenges associated with researching the phenomenon, corruption has attracted considerable academic interest over the years. And one of the leading thinkers on anti-corruption, governance and economic development is my guest this week. Mushtaq Khan is a professor of economics at SOAS, University of London where he directs the Anti-Corruption Research Consortium (ACE).

Episode Transcription

(Prepared by Ingrid Ågren Høegh)


Khan                One reason why a lot of this kind of mid-level corruption is very difficult to solve in countries like Bangladesh is because when you go down to it, you find that everybody has something to hide. So when you go to a hospital and you say which doctors are not attending, you find that almost every doctor has broken their contract, and when you come to say, okay, let’s have a formal process of punishment or sanctions, no one will come forward to give evidence because everyone has something to hide. 


Theme music    You are listening to In Pursuit of Development with Dan Banik. 


Banik                One of the dominant explanations for elusive development in many parts of the world is the negative role played by corruption in the development process. And many national and local governments as well as international aid agencies have spent considerable time and resources trying to come up with plans to combat the corruption menace. But anti-corruption policy has often been difficult to implement and many well-intentioned efforts have had limited impact.


Despite the challenges associated with researching the phenomenon, corruption has attracted considerable academic interest over the years. And one of the leading thinkers on anti-corruption, governance and economic development is my guest this week. Mushtaq Khan is a professor of economics at SOAS, University of London where he directs the Anti-Corruption Research Consortium (ACE).

We discussed a range of issues, including definitions of corruption and whether they matter, the relationship between corruption and economic performance, the political settlements approach and efforts to combat corruption in Bangladesh.


I hope you enjoy our conversation.


Theme music


Banik                Mushtaq Khan, it’s such a pleasure to see you online today. I’ve long wanted to speak with you. Welcome to the show. 


Khan                Thank you, Dan. Pleasure to be here. 


Banik                I want to get us started by focusing on this term corruption, because in common parlance it is often used in a wide variety of different ways. Some claim that we should be talking about corruption in a very narrow legalistic manner, as in deviation from legal norms associated with public office, others would say we should be adopting much more of a broad understanding of the term where we see the entire set of societal norms. It’s not just bribes. It could be gift-giving, nepotism, favouritism, all of this, even theft. So let me begin by asking you whether you believe such definitional issues matter, because I’ve read somewhere that you feel that we shouldn’t really be talking so much about or discussing the definitional issue. So does it matter and how do you actually understand the term or phenomena that is corruption?


Khan                I think that’s a great question. And I think that sometimes it’s actually not useful to limit ourselves to terminology and ask ourselves a more fundamental question, which is what is the problem that we are trying to address? And I think the problem we are trying to address is the way in which, in societies, people use influence or power in ways that are harmful to other people, and harmful to the social common interest. And in a sense, you know, a lot of my early work was on this concept of rent-seeking and rent-seeking is the activity that people engage in to capture benefits that they didn’t have before. The broadest definition of a rent is, I have some benefit which I can get by some sort of activity of influencing, which is generally described as rent-seeking. And that activity of influencing is after all what politics and society is all about, except that sometimes this activity of influencing and getting laws changed and laws distorted and so on, is against the public purpose. So a better way of thinking about it is what kinds of social activities which are not strictly market activities, but are about politics, about influence, about connections, about networks, which harms the public purpose? But a lot of this actually doesn’t harm the public purpose. Indeed we would argue, it is actually integral to achieving the public purpose, right? So the discussion of the difference between value-enhancing or developmental rent-seeking and socially damaging rent-seeking which was really very controversial because people normally used to define rent-seeking as always damaging. And corruption, very simply, is all those activities of rent-seeking which are harmful but are also in some sense illegal or illegitimate. So we’re looking at the subset of damaging rent-seeking which is also illegal and illegitimate. But, not all damaging rent-seeking is either illegal or illegitimate. So, for example, lobbying by big companies to get special privileges for themselves might be completely legal but might be harmful to the public purpose. So, there are a whole range of issues here. One is that I think my real interest is in the harmful socially damaging rent-seeking. A subset of that is corruption or clientelism or nepotism. But I think we shouldn’t get too hung up about the distinction between what is nepotism and what is clientelism and what is corruption. If you are focused on the main issue which is when does collective activity or networking or influencing activity harm the public purpose and what we can do about it, then I think we can all agree about what is the phenomenon that we’re trying to understand? But in the context of developing countries, one of the big differences between developing and more advanced countries is a lot of the rent seeking is actually illegal because systems haven’t set up. These are not rule of law societies where everything is highly regulated. And I think too much of a focus on corruption might simply replace the damaging kind of illegal rent-seeking with damaging but legal rent-seeking and then we haven’t achieved very much. So while the focus on corruption is very important and interesting, the real focus is how to change the rules of the game so that the influencing activity which is inevitable and I think a normal part of any social activity is more aligned with the public purpose and less likely to be socially damaging.


Banik               I agree with you that there is sometimes too much focus. Obviously there’s a lot of focus on why it is bad for economic growth, it weakens political legitimacy and accountability, but others have argued that corrupt acts can actually also be beneficial for development. I just want to pursue this idea further, Mushtaq, whether it could be an important catalyst for growth in some ways, even though it is illegal and I’m thinking about this idea that corrupt illegal acts can actually be beneficial. Do you see any merit to that kind of argument?


Khan                I think that’s an argument which is very old. It goes back to Huntington who famously said and I’m paraphrasing, he said the only thing worse than a developing country with corruption is a developing country without corruption, because he was referring precisely to these kinds of red-tapism and obstacles to investment which corruption has been the traditional way of bypassing. When we talk about rule violating behaviour, in which some individuals or groups use money and influence to bypass rules. If you take that as a general definition of illegal corruption, there are many variants of that and the outcome varies greatly between the type of corruption that you’re looking at. So this type, where the problem was that the underlying laws are inappropriate. There’s too much red tape. This is a very specific type of corruption which in some of my work I refer to as market restriction driven corruption. So there’s a market restriction which has no social purpose. It’s just there because it’s put there either by bureaucrats or politicians to get some money in from the poor businessman trying to do business, or more likely because no one’s bothered to change it because the people who are doing business in these countries have found very easy and comfortable ways of going around it by paying a little bit of speed money here and there. This is one type of corruption, where actually the corruption doesn’t really harm development. But in fact, if you try to stop this corruption without changing the underlying rules, you would actually make things worse, but the solution is obviously not to allow this corruption to continue or not even to stop this corruption while keeping the rules in place, but obviously to change the rules. And so there is a very simple solution to this corruption which is to simplify the rules and make them easier to follow and that’s been happening in developing countries a lot in the last 15-20 years. So, you know this “doing business” kind of rankings of the World Bank is really about that. And a lot of countries have focused on simplifying these things so that those delays and the corruption associated goes down. But my point is that this is really not the most important type of corruption and Huntington made a mistake in focusing on that too much. There is a much more important type of corruption, which is connected to the rent-seeking processes that I was talking about earlier. The state of policy is trying to achieve something at least formally. So for example, the state might have a health policy or education policy or an industrial policy or an environmental policy on paper that looks like a good policy but then different parties who are affected by that policy, engage in corruption or influence buying or lobbying or clientelism, to have that law interpreted or applied in a way that benefits them. And the outcome of that is much more complicated and usually the outcome of that is socially damaging. In other words, here it is much less likely that the corruption is developmental. Even here there are some cases that you can find historically where there has been for example industrial policy and one of the countries I looked at in the very early stage of my research, 20-30 years ago, was South Korea. And South Korea in the 1960s was a country which had a very successful industrial policy and there was a lot of corruption, but uniquely or almost uniquely, the East Asian countries are countries where corruption around those critical policies did not distort the policies and did not harm the social outcome greatly because what happened in those countries, is that the power of the state, and this comes to another concept I have worked on, political settlement, which is a description of the distribution of power between the stakeholders engaged in this bargaining over rents, in the East Asian countries they have a very specific and different type of political settlement. They are different from each other but they’re different from most other developing countries. 


Banik               In relation to say corruption and economic performance, what I really like in some of your previous work is that you’ve argued that the economic forms of corruption and their effects are closely tied to the political forms of corruption. And what is particularly interesting to me is distinguishing between situations where corruption, as you were just saying, where corruption has impoverishing effects, from those where corruption actually allows rapid economic growth, which is what I suppose the South Korean example is very good in this context. So it seems that there are some forms of corruption or corruption scandals, irrespective of the amounts that are involved, that some of these undermine economic performance more than others. Is that how you see it? And that there are sometimes similar practices of corruption, but these have very different outcomes? And all the examples that you mentioned in relation to say both South Asia and also South Korea, is that some countries perhaps witness rapid economic growth and this is because corruption can be offset by what you call counter-prevailing factors. Have I understood you correctly?


Khan                That’s right. I think the argument is that corruption is a bargaining which includes lobbying which can include corruption which can include nepotism and clientelism, over policy resources that the state is trying to allocate to achieve some social outcome. So, take industrial policy. Industrial policy is a policy where a state has a policy which is targeting resources to companies, so that those companies invest these resources in building their capabilities, become more productive, and in the future, the social return in terms of extra jobs and taxes, more than outweighs the short-term support that they are receiving from the policy. So that’s the intention. Now what happens as soon as you have these resources going, companies, individuals, unions, bureaucrats, bankers, all kinds of groups become involved in bargaining, in getting hold of those subsidies without the hard work of delivering the productivity growth, which can then justify the subsidy, right? So that’s natural. That’s going to happen. The outcome will depend on the bargaining power interests and capabilities of the different players here. So what happened in China in the 1980s or in South Korea in the 1960s was a form of bargaining over these rents, which I’ve described as profit sharing corruption. So the state here had very high capabilities of enforcing its decisions and it had a long time horizon. So when it gave resources to companies it obviously would prefer if the companies shared some of the future profits from this investment back with the ruling coalition and they were unwilling to tolerate inefficient companies or banks or investors getting these resources and then not developing their productivity because that resulted in lower kickbacks to the ruling coalition over time. So, if you have a long enough time horizon, and if you have the capacity to implement things, then the ruling coalition doesn’t accept kickbacks which protect inefficiency. 


Theme music


Banik               In one of your works I read a very fascinating distinction that you made where you wrote about two determinants of the economic effect of corruption. The first is the economic effect of the bribe, that is the flow of resources from social actors to state officials. And this is important because the resources that are transferred in the bribe itself often leads to a reduction in the social value and there’s an economic cost to society but what is more interesting I suppose is the second aspect, the second effect of corruption, which is on the economic consequence of the new rights or relocations of rights that are brought about by state officials in lieu of the bribes that they’ve received. And I wonder whether this kind of a distinction explains why corruption, some forms of corruption, have perhaps been more beneficial in countries like South Korea than they have been in countries like Bangladesh or India or even Nigeria? So there may be similar practices of corruption, but with very different outcomes. 


Khan                I think that’s right. So that’s what I’ve been arguing for some time. And the point is not that some types of corruption are beneficial but some types of corruption, which look quite similar to other types of corruption, may be associated with rapidly growing countries, and another cases be associated with actually low productivity growth and in some cases even with economic collapse. And so that’s the puzzle that my work was trying to address for some time. And I think the missing part of this puzzle is the political settlement, the distribution of power between the different stakeholders who are negotiating over the allocation of policy resources. So this is this whole conversation is about what I call policy distorting corruption where the state has a policy, could be industrial policy, could be agriculture, could be health or education, where resources are being allocated through institutions to different organizations and they’re supposed to use those resources to achieve a public purpose which could be investing in the productivity of firms or delivering education or whatever. You can expect that there will be a lot of activity by interested parties to influence not only how those resources are allocated, but the terms on which they are allocated. And this is where the bargaining power of the different players, the time horizon, their capabilities, all of these things come in to play. So you can have two contrasting situations. Let’s say with industrial policy. One where the state is giving low interest credit or export subsidies to companies and the quid pro quo is that those companies have to generate productivity growth and become competitive and then pay taxes and employ people so that a public purpose is served. Now if the state has very short time horizon, very low capacity to implement things, if the businesses on the other side have low capabilities, if they don’t see a possibility of becoming really productive globally, then all of them might collude in capturing these resources quickly and sharing it between themselves so that you have infant industries that never grow up, you have a lot of industrial policy credit never being repaid, the banks start going bust. All of those things that we see in a lot of developing countries with industrial policy start to happen. And here the corruption is policy distorting in a very bad way and has a high social cost which is much bigger than the size of the bribes because on top of the loss of the resources that are being paid as bribes, which could have been invested or whatever, you now have the much bigger loss of social policy, industrial policy in this case, collapsing and not delivering a benefit. In contrast, if you look at the East Asian stories, and I’ve looked at South Korea, but a very slightly amended story holds in China in the 1980s, here you have states with a lot of implementation capacity, a long time horizon, you have businesses, which realized very quickly that their best strategy is to actually raise productivity because they can’t get the state to do short-term deals and make some quick money. Then what happens is that there is still informality in negotiation and corruption, but now what happens is that the bureaucrats and the state do not tolerate inefficiency. They do not support the companies which don’t generate productivity growth. They take resources away from them and they give them to companies which have a better chance of driving productivity growth, not because they’re nationalist or altruistic or not corrupt, but paradoxically because those companies can give them bigger kickbacks over time, because if you have a long time horizon, the more productive will generate more profits. And so what you have here is what I call profit-sharing corruption. So, the state becomes a kind of informal investor in the companies and informally they benefit from the profits that these companies make. The big difference is that the outcome is now, the productivity growth that is achieved, is so much higher as a result of these investments that the state is making, that net of all these costs, the outcome is still good. So I think it’s very important for listeners to understand that I’m not saying that the corruption drove growth, but that the way in which the bargaining over these resources was informally organized, allowed the policy, which was supporting productivity growth to happen, was successfully implemented and some of the extra returns from that were indeed kicked back to the state. So you have a kind of profit sharing corruption. And the point is that this is not very easy to organize because you can’t do that unless you have a very specific distribution of power between the players, what I have also described elsewhere as a political settlement. And the political settlement is not something that is a policy goal. You can’t change the distribution of power in a country. You have to take the distribution of power as a very complex historical outcome and you have to work within that. So, I’m kind of critical of people who use the South Korea and Chinese example to say that look, you know, correction can live with growth, why are we so concerned with it? I think we should be concerned with it, because most countries don’t have that configuration of power, the political settlement, that you find in some of these East Asian countries. They are different from each other but they’re much more different from the rest of the developing countries where the configuration of power is that the ruling coalition is much more fragmented, therefore they have very short-term interests. They usually lack a lot of implementation capacity, where businesses have lower capabilities. And so the interest between these competing fragmented and low-capability entities, is that those policy resources are then captured for immediate dissipation of the resource rather than investment and profit sharing. And so the outcome of industrial policy has been much worse. Very similar industrial policies with apparently very similar degrees of corruption have produced very different results across countries. So this is one of the areas that I’ve worked on and the policy implication of this is that if you’re in a Bangladesh or a Nigeria or an India, don’t try to achieve what the South Korean state, because you will not have that enforcement capacity or that time horizon. What you should be doing, what we should be doing, is redesigning our industrial policy so that the resources are now given in such a way that knowing what the configuration of power and capabilities are, they’re much more likely to achieve a good outcome. 


Banik               I think the political settlement framework does offer a very useful approach to understanding why certain very ambitious anti-corruption programs haven’t even worked. We’ll return to some of these countries that you mention like Bangladesh and India slightly later Mushtaq. I was reminded of the Chinese example, of course, when you were answering my previous question. I was reminded of a recent discussion I had with Yuen Yuen Ang, who has written this fascinating book on Chinas Gilded Age, and her major argument was of course, and she says that corruption is bad, but at a certain point of time, it is the access money that in many ways explains why China was able to grow as fast as it did and it is very similar to this political settlement argument that it was a bit of win-win for both influential politicians who used investments by certain corporate actors, the creation of jobs to further their political careers, while there was also all of these other things happening, benefits for the company, so access money was crucial. I was also reminded of this one article I read many years ago, suggesting that maybe certain types of bribes are more effective than others and I think he said something like the centralization of bribes is crucial, that say in Indonesia  it was centralized, that it was much more effective in terms of overall economic growth in the country. Whereas in a country like India, it was decentralised. It wasn’t very effective. So what you’re basically saying is that some types of corruption have more beneficial effects on growth, right, than others? And it depends on the political settlements and the nature of politics in these countries, right? 


Khan                I think people have observed for a long time that corruption seems to have different effects in different places and they have given different types of explanations for why that might be the case. And I think that these different explanations are slightly different from the other. So I think that what I’m saying is that you really need to take seriously the configuration of power which is holding power that is what is the bargaining power of the different parties negotiating over a rent, negotiating over a policy resource. And I think this is distinctive in my approach. So, I agree with Yuen Yuen Ang’s analysis of China as China, but you know what she calls access money is what I’m calling the productive use of policy rents. The point is that those policy rents have been subject to corruption in other countries where, if you like, the access money type of corruption has had extremely damaging effects. So, access money in India or in Bangladesh or Nigeria is where collusive businesses get engaged with politicians or bureaucrats and pay them to form policy or give them resources in ways that benefit them. So, it’s a form of access money, but the outcome is really damaging for society because they form effectively monopolies of different sorts and distort the market. Why that didn’t happen in China to the same extent and why the mix between productive and unproductive was much more positive in China has to be explained by the specificities of the Chinese political settlement, the organizational structure of the Chinese Communist Party, the way in which it controls in a pyramidical way, implementation capacity of the ruling political coalition is different from any other country of that size, or indeed any size. It’s the product of decades of very costly and bloody revolutionary struggle which created a kind of discipline political organization, which I think is unimaginable, in Bangladesh or Nigeria. I think that’s where you know people from those countries like Yuen Yuen Ang or others who I think are from South Korea often don’t see the real specificity of their contexts and how different it is from others. And the same access money in Bangladesh is usually extremely damaging. It’s not what drives development. It’s often what blocks development. I agree with that analysis for China, but I think the real challenge is what lesson do we learn from that in Bangladesh? And I think the lesson which I’m sure Yuen Yuen Ang doesn’t want to give, and I know that because I’ve discussed this with her recently, she doesn’t want to say that access money is always a good thing, indeed she’s saying it can be extremely damaging, except that in China, it was mostly not so damaging. So, the real question is why was it not so damaging in China? Why is it so much more damaging elsewhere? And in that elsewhere, what do we do then? And that’s where my work on designing the policy in such a way that the powerful players in that country can’t capture it in an unproductive way, becomes such an important thing. 


Banik               So Mushtaq, the first thing I’m thinking about now is of course the difference in political systems. You’re talking about democracies; however imperfect they are. We’re talking about lobbying, we’re talking about very many political actors, political parties, all competing, aligning themselves with corporate actors, getting campaign financing, all of these things happening. There are all these groups that are struggling to claim rights and to ensure that their needs receive political attention, which is very different in a country like China. So, is it the political system that we should be focusing on? Is it a democracy vs. non- democracy kind of issue? 


Khan                No, because of political system also works in a way that is underpinned by the configuration of power. I think that the most important variable is the underlying distribution of power between organizations because you could have a one-party state and we have had many in history where the actual configuration of power doesn’t give the ruling coalition either the implementation capacity or the long time horizon to do what China did, right? So the worst thing you can have is a one-party state where the ruling coalition is not confident of the future, faces a lot of internal resistance, the party internally is not disciplined and you get basically a Zimbabwe type outcome, right? One can name many other such outcomes. It’s not the political system. You could have a democracy where the configuration of power between the different players is such that the democracy can take a longer-term view and you can have much more developmental outcomes. In my work on political statements, I make that clear. So it’s not a question of democracy vs. dictatorship. Some dictatorships are much worse than even you know, the bad democracies. It’s not a question of the centralization of corruption and the decentralization of corruption, because you can have a centralized corruption which is absolutely kleptocratic and totally devastating and it might be better to actually break down that centralised corruption into some decentralized version. So the distinction I would make is between institutional explanations and organizational explanations. And I think you need to look at the two together and this is a distinction Douglas North made which is really interesting and which now also informs a lot of my work, although the philosophical motivation behind my work is different from North, I think there’s a lot of stuff that I’ve learned from him. The difference is that when you talk about a political system, you’re talking about institutions, you’re talking about rules, you know, how many times do you have an election? How do you elect people? Those are rules? So, a one-party state is an institution. How it operates depends on organizations. That is, it depends on who are the subunits within this set of institutions. How are they interacting with the institutions? So the political settlement, as I define it, is this equilibrium between a set of institutions and a set of organizations, where the organizations are trying to influence how the institutions are implemented and the institutions affect how the organizations behave and, through that interaction, there’s a distribution of rents and if that distribution of resources across these different players and that equilibrium is the political settlement. So you can have a set of institutions which look quite similar, like industrial policy, but one industrial policy will be very different from another industrial policy because the players, the organizations, the firms, the bureaucratic networks, the political networks, are differently constituted and have different power. In the same way, you can have an institution which is a one-party state or institution which is describing a democracy and they will behave very differently because the organizations are different. But I would not say that the type of corruption you have is simply a function of institutional rules about how policies are made and what the policies are, but also, and this is the complexity, about the organizations which can also be, to make it even more complex, informal organizations. That is networks. And how those networks are constituted, and their relative power, will determine how your institutions work. So, it is really dangerous to say that China works because it’s a one-party state, because the last thing we want is for developing countries to try and set up one-party states. They have done in the past and they have been terrible in their outcomes. And the last thing we want is to go back and reinvent the wheel. The reason why the Chinese one-party state works in the way it does, you have to look at it historically, at how the Communist Party was put together, and how that internal configuration happened through huge amounts of rather violent events, like the Great Proletarian Cultural Revolution, the Great Leap Forward, and the Revolution. I mean, it wasn’t a cost-less exercise. I mean, society was, if you like, pulverised into a different form of organization, which isn’t simply a question of: is it a one party state? But a question of how actually the constituent units of the society have been organized. So, it’s a much more difficult thing to try and imitate. You can imitate the one-party state but you can’t imitate the social organization of China without going through something quite similar, which no one would recommend as a policy solution. Having gone through that, China now has a structure of organizational power and institutions which works rather well for itself involving certain types of problems, but it might not solve other types of problems. And then the Chinese will have to reform the institutions and adapt and so on. So, I think it’s a complex story. And I think the danger is to draw simple messages from it, which result in quite bad outcomes if you try to replicate that.


Theme music 


Banik               Let’s move on to Bangladesh. It has in recent years been described as some sort of a paradox that Bangladesh has achieved sustained economic growth over the last couple of decades. It has reduced property. There’s been a return to democracy, more investment, confidence in the private sector, the country is no longer dependent on aid, and a lot of people associate Bangladesh with the textile and garment industry, millions of jobs. The sector has become globally competitive. There’s a lot of talk about innovative work being done by NGOs that have ensured that aid actually has had a positive effect on poverty and yet there are all of these governance and corruption related challenges. Even though there’s been a return to democracy. There have been elections. It is still a politically polarized society. And I find it particularly interesting and some of the work that your ACE-group has been doing, and also in your own work, is highlighting the fact that Bangladesh has actually had very strong anti-corruption laws. But like in many parts of the world, in many other developing countries, you could be very rich in policy but implementation poor. So the question is, is that really the correct picture? Is it the implementation of all of these laws that has been weak? Or is it just the laws themselves that have been flawed because I do read in some of your work that the big problem is perhaps that anti-corruption efforts have been highly politicized. So, what is it? Is it bad policies? Is it bad implementation? Is it both? Or is it just bad implementation? 


Khan                Great question. I think anticorruption is like any other law. It’s, if you like, a set of rules and how those rules are implemented depends on the distribution of power in society and who has the interest to actually enforce those rules will determine whether those rules are enforced. So, what is true for industrial policy is also true for anticorruption policy. One of the observations of the political settlements approach is that one difference between levels of development, of capabilities in society, is that in a society where capabilities are very broadly dispersed, you have actual real effective demand for a rule of law, because you have many organizations in society which have capacity to engage in complex productive activities and to interact with thousands of other organizations on an everyday basis. And when your society is structured like that, all these powerful organisations will demand a rule of law because without it, they can’t negotiate and organize all their contracts. This is what North, Wallis and Weingast in their Open-Access Order was also saying, using slightly different language, that you need to have quite a broad distribution of capabilities for what they call an open-access order and what I would call a rule of law society to emerge. Basically, a rule of law society means a society where not only do you have rules, not only are the rules enforced, but they’re also enforced on the powerful. And the reason why rules are enforced on the powerful is because the powerful are very broadly dispersed. There isn’t just one tiny group of powerful people. And when there’s a broad distribution of power, the checks and balances is that no powerful group will want some other powerful group to get away with it, and they will insist on the enforcement of rules. For example, with Trump when he was trying to break the rules, there were many other groups in America, which saw that as a threat to them in the long run and therefore they tried to stop it. And in this case, it was successful. If you look at a developing country, and Bangladesh is no exception, you have rules, the rules are enforced, but the distribution of power is still not that broad-based that it is enforced on everyone equally. And that basically means, in the language of political settlements, I would say they have a rule by law. They don’t have rule of law. Which laws are enforced and on whom, depends on the configuration of power between the players in that context. When you have a small number of players who are powerful, they actually don’t need a rule of law. They can survive quite well without rule of law. So, if you have a country with a small number of powerful firms, they know each other, they know the businesses, the politicians, the bureaucrats, so if I’m a corrupt business and you’re a corrupt business, and I know that you’re a thief and you know that I’m a thief, I can still do business with you. If you don’t pay me, I can get informal pressure from the police, from the law, courts, or just the mafia. I can hire some people to go on a motorbike and get my payment, and since you know that, you will actually make the payment. In other words, the enforcement becomes informal. This is no longer possible if I don’t know you because there are 10,000 people like you and I’m going to select the cheapest one. Then I need a rule of law, because then if you don’t pay me, I need some way of enforcing it, which I can’t use anymore because I am one of many powerful players. So, I think there is a simple explanation. Of course, it’s much more complicated and there are many more factors. But if you want to understand which laws are enforced in Bangladesh and, as you said, we have anticorruption laws. They are enforced. But they’re universally enforced on the opposition parties, right, and that’s because when you try to enforce this on the ruling party, of course, the distribution of power between the enforcement agencies and others is such that this cannot happen and the broad social distribution of power doesn’t demand a rule of law and cannot enforce a rule of law. This is not surprising at all. The more interesting question is: why does Bangladesh work at all? Why are there sectors which are working reasonably well? Why is it growing? Why is there investment? And to explain that, I go back to the political settlement, and the organization of power in Bangladesh. And, just like in China, there was a very traumatic event which at that time was very costly, but in the long run had unintended consequences, which are extremely good. In Bangladesh, the defining moment of Bangladesh was the 1971 Conflict, the War of Independence, Civil War, whatever you call it. That event mobilised huge sections of society in a way that they hadn’t been mobilized before. So if you look at it historically, East Bengal was the Boondocks of Bengal. So, when it was British India, most of the industry was in West Bengal, in what is now the Calcutta basin, and East Bengal was the peasant economy. When it became Pakistan after 1947, most of the industry was in West Pakistan, most of the administrative power was in West Pakistan. East Pakistan was exporting jute. It had very low organizational capabilities, almost no entrepreneurs, almost nothing, right. It was just peasants growing stuff. What happened in the War of Independence is that suddenly society became mobilized around injustice. Injustice that there was an election and the election wasn’t honoured and we all know that history. A lot of social actors came up from lower levels of society, which had previously not been organized, not been powerful, and that’s had a tremendous impact on the configuration of power in Bangladesh. 


Banik               One of the things that I was thinking about when you were mentioning the problems in Bangladesh, in terms of enforcing these rules, is this rather famous bridge that is under construction. And that is I suppose an illustrative example of attacking high-level corruption in that current political settlement of Bangladesh isn’t going to perhaps work because if I remember correctly there were some high-level politicians and ministers who are accused of corruption by international agencies, but not much happened to it. How do you see that? Is it a good example of why one should be perhaps pursuing other types of strategies? As you put it, more incremental strategies, perhaps even involving the local communities to monitor development activity? 


Khan                I think this is a really complicated question because the high-level corruption is very important to track and inform the public about and to have a discourse going in the country about what is going on. What I think is unrealistic to expect is that this will lead to immediate action prosecution or even a kind of ending of these activities. This is extremely unlikely because of the configuration of power in these countries and the weakness of a rule of law. What we have in these countries is rule by law, which is that the law is imposed by the powerful when it is aligned with their interest. So, if you are taking the very powerful people in your society at the very top to court, then in any developing country, we shouldn’t expect a very good outcome. So, there’s a difference between not having the investigation or the discussion. Yes, we should, and I think it’s important to keep pressure on the political players and the big businesses and so on engaging in collusive activities, but that’s not immediately implementable. And I think the danger for societies is that you mobilize a lot of your activity into doing something which is not going to be implemented, then you become really frustrated and you give up and you have hopelessness and you have political apathy. That’s the real danger. So what we say should happen, is we should focus on the corruption that is actually solvable because solving it is in line with the interest of a much more dispersed group of people who benefit from solving it and have the interest and the capacity to enforce it. And it’s not immediately taking on the really big and powerful people at the top of society because that kind of corruption which is the kind of feasible and implementable policy distorting corruption, which is stopping policy from working is what actually spreads the growth of capabilities and creates a broad-based society that in the end, in 20 or 30 years’ time, might create the demand for a true rule of law, where you will be able to get the really big players abiding by the rules. But I think most developing countries are still quite far away from that. So, I think we are not against that. It’s just that we go to it in a slightly more roundabout way with a greater attention to feasibility and what is achievable right now. 


Theme music


Banik               So rather than a top-down approach, would it be more effective to apply much more of a bottom-up approach to actively involve local communities to monitor activities? How about influential individuals who can use their power, their connections, their network to pressure officials and private-sector contractors? Is that more feasible? Is that what you’re saying? 


Khan                Very broadly but not precisely. I think again, it depends on very specific questions, perhaps if I give you a couple of quick examples from Bangladesh, it might be clear. So, one of the projects that we did recently, just completed, Bangladesh is spending billions of dollars on climate adaptation projects, building embankments on rivers, cyclone shelters and so on, because it’s a low lying country affected by climate change. Now, we know that these projects are subject to massive corruption. Something in the order of 30% of the project funds according to some estimates get diverted and many of them are not finished to a high-quality. We looked at a comparison of a number of big projects, recently completed, with our partner Transparency International Bangladesh. So, we got into a research partnership and we found that some of these projects had much lower levels of corruption than others. They were completed to a much higher quality than others. So, we did a large-scale survey of the surrounding areas, looking at four projects. And we surveyed around 2,000 people to figure out why was corruption low in one set of projects and high in another set of projects? And we found an answer that we hadn’t really expected which is that these climate change projects all have dual uses. Right? So, embankments can be used as roads, and the dual-use benefits are immediate. Whereas the climate adaptation effects are in the future and long-term. When you have very high levels of immediate benefits, if the design of the project is such that the road happens to be in a place or the community centre happens to be in a place which is of interest to the local community, then the more the dual-use benefits, the more the relatively more powerful people in the community become interested because they benefit most from dual- use benefits and these powerful people in rural Bangladesh might be in absolute sense quite poor, but they are networked to powerful people. So you then mobilize these informal networks. And what happens in anticorruption is that the formal rules don’t work, but the informal, when the local powerful, you know, landholders and business people, they want that road to be built properly, they sit in the tea shop and have a conversation with the contractor and the contractor knows that if they go too far wrong, these people will mobilize their networks in local politics and local administration. They will get into trouble. So by creating a check and balance between the powerful, you actually contain corruption much more successfully through this informal mechanism, than the formal mechanism of reporting accountability, town hall meetings, all of which happen, but they happen in a very tokenish way, as a tick box exercise. The real activity happening is when you can mobilise these networks. And so we have a very simple policy implication here, that at the project design stage you should seek to maximize the immediate dual-use benefits for the community because that is a very simple way in involving communities. And this is one of the strong points of Bangladesh, is that power is quite dispersed and at the lower levels, you need to create checks and balances between powerful groups at that level. So, if you can do that, you have a good anticorruption outcome, which is very feasible to achieve and doesn’t require you to enforce rules from above on anyone but the checks and balances are doing anticorruption. However, it doesn’t follow that, you know, everyone can do this because to check the powerful who are engaging in corruption, you need to have other powerful people in the local community interested in anti-corruption. If it is just, you know, poor peasants sitting in a committee and saying this is corruption, nothing will happen. Right? So the question of powerful is not, the powerful doesn’t mean the prime minister and the ministry, it means, in rural Bangladesh as you know, a powerful person is someone who might have half an acre of land. That’s powerful. Powerful is someone who owns a tea shop. Those people need to be interested in the monitoring of the project. If they are monitoring, then the other powerful, the local contractors, and so on, who would otherwise be giving kickbacks to the politicians and so on and getting the contract and making the money are then kept on check. This is the kind of evidence-based research that were doing, showing that there are feasible solutions. This is one type of answer. The other type of answer is a lot of corruption in developing countries happens, which is several types of corruption loaded on top of each other. Some people are corrupt. And now we’re not talking about the top-level, you know, big corruption, but everyday corruption. When you look at the corruption in hospitals or the corruption in schools or the corruption in these kinds of decentralized activities, you find that some people are corrupt because there is no other way of doing business unless you are corrupt. They don’t actually want to be corrupt but they’re forced to be corrupt because that’s the only way of getting any kind of return on their time and effort. But some people are corrupt because they could follow the rules, but they decide to be corrupt because they’re extractors. They want to steal and cheat. But that second group is usually a minority. One reason why a lot of this kind of mid-level corruption is very difficult to solve in countries like Bangladesh is because when you go down to it, you find that everybody has something to hide. So, you know, when you go to a hospital and you say, which doctors are not attending, you find that almost every doctor has broken their contract at some time. Some of them have done it because they have no option, some of them have done it because they have an option, but when you come to say, okay, let’s have a formal process of punishment or sanctions, no one will come forward to give evidence because everyone has something to hide. So, one of our strategies is that in these kinds of areas, you actually need to work to find out what is the problem that is causing the rule violations to happen and solving it to the extent that a large enough group of people who are actually not violators, not corrupt, can be rule following and that makes it possible to Identify the true free riders. Otherwise, you will never be able to identify them. A fascinating example of this: we were doing a project on skills training providers in the garment industry in Bangladesh. The industry says, we don’t have skilled workers. The government is spending billions of dollars from various banks, you know, like the Asian Development Bank, the World Bank, lending money to Bangladesh, which has been spent by the government in skills training programs, and we found that there is massive corruption in these skills training programs and the outcome is very poor. The government has, like many countries, incentivized these skills trainings, so that the training provider and the vast majority are in the private sector, the training provider is supposed to provide training and then a third of their money is only paid to them after their trainee gets a job. The idea is that you align incentives to outcomes. So, in theory, the training provider should work very hard to provide training that the market wants, because only when the trainee gets a job, they get the last transfer of their money. Fantastic in theory. What happens in practice is that there’s massive fraud in the reporting of employment. And there’s collusion among all kinds of players in doing this fraud so you can’t uncover who is doing the fraud and why. So we had a very complicated research program where we partnered up with one of the implementing agencies, which was paying the money out to a bunch of training providers to really understand the problem from inside. And that’s the starting point. These are actually quite difficult problems to crack, to understand what is going on. And, we found that some training providers were not doing any fraud whatsoever. And, we found that for some training providers, the people they were saying have a job, almost all of them had a job. They were not doing any fraud. In the case of other training providers with the same incentives, with the same monitoring, with the same overall political settlement, those guys were doing 60% or so fraud. In other words, most of their trainees were not engaged in employment, but they were reported as being in employment so they could collect the money. Why is this happening? And it wasn’t a simple question. It took a long time, a lot of data, to uncover. And the answer was something that we hadn’t expected. It turned out that the problem was not solely on the side of the training providers. The problem was also on the side of the employers. This garment industry has some high capability firms, which are extremely efficient inside, their production lines work very quickly, and those factories will employ the skilled worker because without a skilled worker, their production line slows down. However there is a long tail of garments factories, which are not so well organized and when you send them a skilled worker, it doesn’t raise the speed of their production line and the employer says, this is not a skilled worker, I’m not getting any faster production, and they prefer to hire people from the factory gate, who are slightly cheaper, and completely unskilled, because that doesn’t make any difference to their production line. So we found that, and this was a really fascinating result, that skills training providers who were close to clusters of high capability firms would send their trainees and they would immediately get a job, and those unfortunate training providers who were close to clusters of lower capability firms, would send their workers, but the firms would say they’re not really well-trained. You can’t see any difference in productivity. They would hire from the street. Now those guys, even if they didn’t want to engage in corruption, are forced to engage in corruption because they can’t place their workers, so they have to spend their resources in doing the training, they then do fraud, the workers know that they have been trained, so the training provider tells the worker, you know, when they phone you up and check, please say that you have a job, otherwise, we won’t get paid. So there is collusion between the workers and the training provider. 


Banik               My general observation, and I’ve done a bit of work on corruption in India and also in parts of sub-Saharan Africa, a general observation related to what you were just saying, is that almost everyone in society agrees that corruption is bad. Everybody publicly would never say anything good about corruption. They’ll always denounce corrupt acts, but in private, they may not be as critical. And, the example is, if you want to get a driver’s license or some sort of permission to do something, get any clearance, you may publicly say that’s really bad but in private you may be the first to offer a bribe. So, my question really is, is there something there in which you know how we as individuals perceive corruption? That we may be very critical of it if it involves some official but we’re not very critical if it benefits us?


Khan                Absolutely, but do you want me to finish what I was saying and then come to this? Because they’re actually quite closely connected. So, what we found in the end was that the training providers who were located near low capability firms, were engaging in corruption because there was no other way of them doing business. So, there’s a very simple policy conclusion that follows from here and that’s also based on experiences in Bangladesh, which is that the government or the development partners, should not just invest in skills training. They also have to invest in raising the capabilities of the firms which is actually a very simple exercise in advising them about how to set up their internal production lines and quality control and timekeeping so that the production line moves faster. As soon as you do that, important chunk of corruption by the skills training providers disappears and you have a real win-win. You have a win-win because now your skills training investment results in productivity growth, potentially higher salaries and more jobs, corruption goes down, and you have the employers having higher productivity because it’s a combination of skills training with capability development that unlocks this kind of corruption. And when you still have some residual training providers who are still corrupt, then the others will easily identify them and you will have that peer group pressure, which is the best kind of check and balance against corruption. This comes back to this observation that everybody is publicly against corruption, but in private almost everybody is doing some corruption, and this is really at the heart of the exercise that we are carrying out because that is actually true. For most people, the process of finding a rule-following way of getting things done is so difficult and time-consuming because the system has been set up in the wrong way, sometimes of course, the issue is that there aren’t enough resources and that you can’t do anything about, so you’re a poor country, you don’t have resources, so things will not work as rapidly, but there are areas where the system could be improved, but it is not, because it is assumed that anyone who wants to, can always find an informal way of working your way around that set of obstacles. The skills training is a very good example of that. You have a system that everybody, where everyone knows that however much effort I put into my skills training, it’s very likely my trainees won’t get jobs, but they also know that I can falsely report high levels of employment and everybody knows that this is being done by everybody. So, nothing will happen. The problem is that the unintended consequence of that, is that you are not identifying the real social problem here, which is the low capability firm, there is no policy response to that, and a lot of social resources are being wasted and this is another example of the social costs being much higher than the bribe, because the nature of this kind of policy distorting corruption, is that your policy is being distorted in a fundamental way. So, all of our solutions are incremental in the sense that sector-by-sector, activity by activity, we are looking for solutions where people who don’t want to engage in corruption, can find a non-corrupt way of doing their business, making their profit or meeting their doctor or whatever it is that they’re trying to do, or in the case of the skills training provider, getting paid after they have done their skills training work because once you do that, and only if you do that, will you have a bunch of people in that activity who say, actually we are following the rules and we are profitable and we can make money and we can get done what we need to do. And that is when the peer-to-peer, the horizontal pressure, for getting rid of the free riders, really begins to bite. And I think this is the real difference between anticorruption in more advanced countries and anti-corruption in less advanced or developing countries, because in the latter, the peer to peer pressure completely disappears. External enforcement only works if 90% or 95% of people are already following the rules. Then the 5% who are not following the rules will be identified by the 95%. The problem in developing countries is that 95% or 100% are not following the rules. They are breaking it for some reason or the other and so when they randomly pick up people and they systematically pick up people who are in the political opposition, and then the punishment is carried out on them. And this has no effect on behaviour in the activity. So, this is really what we are trying to change. We are saying the bulk of rule following enforcement is horizontal peer-to-peer monitoring.


Banik               Mushtaq, it was great fun to chat with you today. Thank you so much for coming on my show.


Khan                Great, Dan. You’re going to have a job editing all of this.


Theme music


Banik               If you enjoyed this podcast, please spread the news among your friends and share it on social media. The Twitter handle for this podcast is @GlobalDevPod.


Thank you for listening to In Pursuit of Development with Professor Dan Banik from the University of Oslo’s Centre for Development and the Environment. Please email your questions, comments and suggestions to