 explaining why countries have births of growth and other countries don't have births of growth and then why some countries have growth that persists like Korea and Singapore and other countries like Brazil have births of growth. Brazil was growing as fast as Korea was from 65 to 80 but then Brazil had zero growth from 80 to 2000 whereas Korea continued on. So it's the dynamics of growth that are the interesting puzzle. The economics doesn't yet have a good explanation for it. We have a pretty good explanation for why countries, some countries are rich and some countries are poor but we don't have a lot of explanation for how the dynamics of that work out. Now part of my solution or proposed solution is that economists think a lot in terms of rules. What are the, they often use the phrase what are the rules of the game? Well, my sort of basic metaphor with a deals world versus a rules world is rules happen when you have a referee who enforces the rules. What you really have and what in some sense we mean by weak institutions is that people don't follow the rules. That what is available say to one firm in Ghana is not available to every firm in Ghana but it's a specific deal cut with a specific firm that may happen because of personal connections between the politicians or regulators involved in the specific firm and hence I have empirical evidence basically showing that the difference in regulatory compliance costs of firms in the same country is wider than the difference in regulatory costs on average across countries. So it's not really where you are that matters, it's who you are. And once you're in a deals world you have a completely different dynamic so often times in rules world countries we think of the private sector and the government and we think there's a dynamic in which the private sector wants better rules for the private sector and the government might have regulatory interests and the politics are between the private sector and the government. But in a world of weak institutions businesses actually have vertically integrated industry meaning you have politicians and businessmen in bed together and the dynamic isn't between private sector and government. It's a dynamic between insiders in private sector and government and outsiders to private sector or government which is outsiders in the private sector but not in government and outsiders of people not in government and it's that insider outside that is what happens in the deals world and in that world the dynamics of capitalism are completely different because it's all about who is inside versus who's outside rather than there being a neutral transparent good set of rules for all. Resources of aid need to be more directed to the ecosystem and less to the plants. So to pick up on metaphor that Michael Wilcock was using about planning a sunflower versus planning an oak tree. The difficulty is that in the projectized world of aid too often you focus on which seed should I plant whereas what really the aid business should be in it is creating the ecosystem in which the plants that will flourish can flourish. Now to some extent this is why the what works versus what could work. I think in some sense there's been too much emphasis on what works with the assumption that if we have what works in one place in one time in one sector the what could work, what's scalable, what's transferable are all easy. We just take what does work in one environment and we replicate it around the world. So it's the over replication of innovations that's in part the problem. Now partly what we're working on is what we call what's the scope for positive deviation. Let's say you have an idea of something that might work in your context in Mozambique, in Vietnam, in Ghana, in South Sudan, in some place and it doesn't look like what has worked elsewhere. Right now I think the aid environment is doing harm by assuming that what could work has to look a lot like what does work in other places and doesn't create enough space for innovation. So in order to create rules to provide strict accountability to taxpayers from the aid giving countries we create tight conditions about how that aid should be used that are actually too tight and we choke off the space for innovation. So rather than asking the question what could work and be scalable here we skip from what works to what's transferable and we assume that the scale one could work are all solved. So what I would work on most is creating an ecosystem we were comfortable with into which we could rely on positive deviation. Now one of the way that ecosystem would work is we would be more rigorous about performance targets and less rigorous about process targets. So right now a huge amount of expenditure of aid is into process compliance. Did you procure the thing? And the way that the procedure specified should you procure the thing? Not did the thing work? Not was it the cheapest thing? But we have faith in the process requirements leading to the performance requirements that's unjustified by the evidence. So if we could set up the performance requirements then we could let up on the process requirements because we'd say we don't care how you procured it if it works for the purposes that we've agreed upon and it works. So a lot of the metrics on performance is where the aid industry could have been much more helpful with its expertise and yet hasn't been. And so I would put a lot more emphasis on to what extent are the efforts of the donor community help creating an ecosystem for positive deviation, meaning positive innovation that could be tested then scaled and replicable in the countries that it's in, which is basically coming to some sort of set of performance. Are we satisfied that the performance metrics against which innovations are judged are right? If those are right then we could unleash the power of innovation into that ecosystem. And specifically on the issue of capability traps, the first thing it could do is do no harm. And it's not yet meeting that standard because one of the things that we point out that helps reinforce capability traps which are organizations that persist and persist with budget and legitimacy in spite of lack of performance. They do so by adopting what we call, a term we borrow from sociologists of organization called isomorphic mimicry, which is they pretend to look like something and they get by with the pretense. So when I teach this in class as I say, okay, how many of you have ever been to a country where the police are corrupt? Everyone raises their hand. Everybody's been in a country where policemen are corrupt. Then I say, how many of you have ever been in a country where the policemen don't wear uniforms? And no one raises their hand. Well, policemen, they wear uniforms. That's what policemen do. And then you realize, well, that's isomorphic mimicry. These police forces, often which are corrupt, not in fact enforcing the law, if anything, they're not just institutional mimicry, but a term I just heard the other day, institutional mockery. They make a mockery of justice, but yet they still wear uniforms. Why? Well, because just the wearing of the uniforms and having the organogram makes them the police force. They are, in fact, a police force. And so in some ways, they fill the institutional ecosystem niche of law enforcement without actually being functional. Now, I think in a lot of domains, the donor community has fallen for isomorphic mimicry and in fact reinforced it by continuing to channel money into organizations on the premise that those organizations continually look like they're developing, they've managed to prevent in some sense space for positive deviation from emerging in which those ineffective practices are challenged. And that can lead to a capability trap where you have low capability. And in the interest of building higher capability, you reinforce the trap of lower capability. So I think, and again, just their example, upper example where the development business is not new. People have been at the idea of building the capability of these newly emergent states, particularly in the countries that were decolonialized in the post-World War II period. There were ideas and notions about how one would build the capability of the state that you might have forgiven us for having those ideas in 1962, but it's 2012 will be in a week or two. That's a long time to have had the same set of ideas that aren't working. So if in 2012 countries that have been independent since 1950, since 1960 still haven't acquired capability, we have to step back and say, something must have been deeply wronged with how we were thinking about our theory of change or else we would have had more success. So I think on this score, being willing to step back, revisit and admit you might have been deeply wrong about your theory of change, can maybe get us off the donor community being part of the actively doing harm and perpetuating capability traps. It's much harder to say when we look at isomorphic mimicry, we call that a technique of successful failure. So at this stage, we're much better at saying, what are the techniques that people are using to perpetuate this than necessarily describing the deep causes? Any social science researcher in these days of you use the word reasons, we're very nervous because you're applying, we understand the whole causal model, right? I think there are lots of reasons in which the configuration of politics and power have led the state to want to appear to be making progress on issues where they don't for either internal or intrinsic reasons have a deep interest in performing and isomorphic mimicry has been a technique they have used to achieve their objectives. So they haven't in fact had the objective of making sure every child is learning. They've had the objective of making sure every child was in school, which is a very different thing. That they have managed to succeed in their objectives at hiring teachers and building buildings without pursuing education by creating an environment in which isomorphic mimicry can be used that look, we have a building, kids are wearing uniforms, the desks look like desks and no one's asking the question, is there learning going on inside this building? So the reasons why they've wanted to do that is a complex political story. We've been so far more focused on the technique. How is it they did this, right? It's really amazing. How is it that you managed to build schools for 50 years and never once measured learning? How did you do that? Well, that's what we're explaining first of all. Now, the objective going beyond that is what I call isomorphic mimicry is a type of camouflage. It's looking like one thing when you're really not there. So the next stage of our research is how to sabotage the camouflage. How do you unmask and delegitimize essentially isomorphic mimicry as an organizational survival strategy in favor of forcing them to adopt a performance strategy results well to some extent because I do have research going back. So I think there are two basic explanations for it. One, which there are some new results by researchers, Eric Hineschek and Ludger Wosman from Germany that suggests when they try and decompose the effect between schooling, the number of years kids sat in a building and learning, they actually find very strong growth effects of learning, but they find that there's very little correlation between schooling and learning because the variance across countries and the amount of learning per year of schooling just varies enormously. So like micro research I've been doing, I've been looking at what I call learning profiles, which is if you just look at what children know, what you find, just very recent research that I've done on a variety of data sets from India, I find that if you take some very simple question like can they do an arithmetic operation of division, only about 9% of kids learn how to do that in an entire year of schooling. So in second grade, 20% of kids can do it and let's round it to 10 because it makes it easy. In third grade, only 30% of kids can do it. In fourth grade, only 40% of kids can do it. So the pace at which they're acquiring this skill you would think well division, you should learn that in second grade so it should go from 20% to 100% know how to do it in third grade. Well that's not what's really happening. What's really happening is that there are these very flat learning profiles such that even by seventh and eighth grade, only about half the kids, well of course in this example, where would I have been, but anyway, by eighth grade, it's still only a small fraction master evenly very rudimentary what we think of as second or third grade skills. So that happens in some countries. In other countries, there's actually very rapid learning progression. So if you compare two countries, eighth grade, but then put them on an equivalent learning basis, actually the eighth grade schooling is only three years of education. Whereas in this country, eight years of schooling is eight years of education. So the first thing, unambiguously, that's gotta be important. It's gotta be important not just years of schooling and Hanyu Shagin was, and I think they're starting to put some reasonably good numbers on this, although it's very difficult, because in the interest of isomorphic mimicry, most schooling systems of the world have consistently refused to measure their performance. So it's actually hard to put numbers on learning, whereas it's easy to put numbers on what they've been doing for the mimicry side, which is building schools and hiring teachers and progressing kids through grades. Very, very hard to put numbers on learning because they were too smart to allow themselves to be tested. I mean, so now that is part of the explanation. I can't, that isn't all of the explanation. I think the second half of the explanation is the labor market into which the newly educated children went. In many countries, what happened was that the far and away most attractive thing from a private point of view to do with an education was to go into the government. Now, that's not a terrible thing because what you find is the best and the brightest emerging from the system of education in Taiwan and Korea, Japan went into the government. But the best and brightest went into the government in Nigeria, and the best and the brightest went into the government in Cote d'Ivoire. Matter of fact, in the 1990s, 80% of every individual with a junior high school degree or more was working in the government in Cote d'Ivoire. So then you have to ask yourself the question, how productive do I think the government in Cote d'Ivoire was? If it was productive, like it was in Taiwan and Korea, that wasn't such a terrible thing because they provided needed public goods. If, on the other hand, you educated a huge population, most of them went to work in a government and that government was corrupt and dysfunctional, then it's not that surprising that what you found was that when Cote d'Ivoire could sort of afford the government that it had through commodity earnings, it all worked well. Kids got educated, the government hired them, paid them nice wages. Once the commodity shock happened in Cote d'Ivoire, Cote d'Ivoire had no, completely lack of robustness of its economy, so it started going downhill and it just basically had no way of stop going downhill. So what you find is lots of countries in the world where Cote d'Ivoire, Venezuela, Jamaica, lots of countries where education stocks go up, growth is going up, growth tapers off and starts to fall and education stocks continue. So there's this divergence. That can't be explained by learning because it's not like kids stopped learning less and less. That is explained by the structure of the labor market and you've got a mismatch. Then at Broadway in Finland, so examples close to home, where learning isn't the issue, prior to the end of the Soviet empire, the countries of Eastern Europe actually had higher education stocks and higher learning stocks than Southern Europe. So for instance, at one point in the early 1990s, Czechoslovakia had much higher levels of schooling and much higher levels of measured learning performance in Spain, but had half the GDP per capita because they were working in an unproductive system. So the second answer to the solution is if you educate children, put them into a system in which their education is devoted to unproductive work, you're gonna eventually, it will be revealed that if there was no tight correlation between schooling and sort of productivity gaps. One is how to create this organizational and systemic learning environment. I actually have a group of students that's founded a brand new consulting firm, just the four of them, but on providing organizations feedback loops. So a lot of what's happening out there is organizations want to do things and they want to do things better, but they don't actually know how to measure their performance, measure what the space of what they can do differently is and feed that back in their performance. So for instance, an example of an organization that does this well is Profim in India, it's an NGO that's been working on education issues. And they have a very tight feedback loop. They go out, they try things at scale, but they build learning into what they're doing and then they have a pretty sophisticated notion of what they can do differently. Do we do the training here, do we do it there, do we do it these people, do we do it those people, do we do literacy camps in the summer, do we do literacy program in the school year? And they're always doing multiple things in parallel, but they have performance feedback loops so that then when they say gee, it worked here but it didn't work there, then they think through why. So I think this notion of we're in an active search for relevant solutions and we need to use data feedback into that, how you do that in a way that organizations actually learn from it is a very exciting research area. The second which has nothing to do with anything I'm doing I think is the question of entrepreneurship and how you facilitate entrepreneurship, how you promote entrepreneurship. I think one of the sort of global crises is in some sense the global crisis of the legitimacy of capitalism and part of the crisis is the sense that the game is rigged and that we're really not living in opportunity societies, we're really living in deal societies where the few have access to preferred deals and the rest of us are just, the rest are just sort of at the mercy of the deals and trying to fill in the niches as best we can. Now the reality would be a real opportunity society would be in a society in which well the metaphor I use from learning Latin America is every child in every neighborhood of Brazil believes that if they're the best soccer player they'll play on the Brazilian national team and when they're out playing the game in the worst favela or the most remote rural area when they score a goal they imagine that I could be on the Brazilian national team. The question is do they feel the same way about Brazilian capitalism? When they look around do they think if I work hard, if I have new ideas, if I do my best do I have the same opportunity to be at the top of the economics in the way that they have faith in the sports and the answer is probably not because that's not in fact the way capitalism is currently structured. So there's a lack of faith in entrepreneurship because we don't have mechanisms of reaching out and making entrepreneurship really translate into an opportunity society. And that's a combination of how regulatory happens is a combination of how firms are able to form and expand. It's a combination of financing. It's a matter of sort of education and what people see as education instrumental too. On all those dimensions I don't think we have anything like the right answers of how you really create a society that both is and appears to be a society that really unleashes the energy of everybody through entrepreneurship. So I think that's a really interesting research area.