 Well, as Dirk Willem alluded, this is a very different sort of a paper, and a very modest paper in a way. A lot of the recent writing on industrial policy, largely led by Danny's work, but also summarized in the very good survey by Andreas Rodriguez-Claire and Ann Harrison, leads us to a conclusion that there's probably less controversy about the what of industrial policy today than there is about the how. How do you actually implement industrial policy as opposed to what should be the major themes for interventions or lack of interventions? And what I wanted to do today was really to just talk about one unintended experiment in the implementation of industrial policy. And the reason why I say it was unintended is that it's actually the result of the odd couple tour of Africa by Horst Kohler and Jim Wolfenstein back in 2001. For those of you who aren't as old as I am, you may not recall why they were such an odd couple, but believe me, they were. If for no other reason than the conclusion of this tour resulted in a major recommendation that governments in Africa, presidents in Africa, should create something called a Presidential Investors Advisory Council. The reason why I say this was already evidence of the odd couple is that Jim Wolfenstein was heard to grumble in the halls of the World Bank where I dwelled in those days that what the hell is Horst Kohler talking about industrial policy for? After all, the IMF should confine themselves to macroeconomic management. So it was a tour that was somewhat fraught with conflict, but also resulted in this one widely shared vision of both the bank and the fund. The councils were essentially on the face of a good idea. I think both of the principles came away with the notion that business-government relations in Africa were often difficult, that there was a good deal of lack of communication between business and government, and that getting a group of corporate champions together with the head of state bringing experienced and successful business leaders to bear on this question of what should you do to expand the private sector was a fairly good idea. It was also intended, at least in the view of the staffs of the bank and the fund to do one other thing, and that quotation is important, to identify constraints to foreign and domestic investment, to generate recommendations for concrete action, and to reinforce and accelerate policy reforms. So the councils were created in 2002 by the hosts, Ghana, Tanzania, Senegal, Mali, and Uganda followed in 2004, as we heard from Lilly yesterday. Following that the IFC basically took over the program and there are now similar kinds of mechanisms across a broad range of countries, both in Africa and in Asia. There are at least 10 more councils in the cases of Asian countries. Ethiopia, which is an interesting case to which we'll return, launched a public-private consultative forum, more or less modeled on these presidential advisory councils in 2010. The unintended consequence of this, I would argue, is when the management of the World Bank, and especially the World Bank, got involved in the outcome of the deliberation of the councils. Remember, the councils were originally conceived by their originators as a chat shop. Put the heads of corporations together with the president, let them build trust, let them engage. But the World Bank management immediately moved in and linked progress in the councils to their own private sector development programs. Now keep in mind, again, that donors in Africa are a very important actor in the management of the economy and that the outcomes of the councils were suddenly to have operational consequences for the programs of the World Bank. In addition, both the Bank and the Fund wanted to link the process of the councils and the outcomes of the councils deliberations to the poverty reduction strategy papers. These, as you'll recall, and being one of the guilty parties in having originated the poverty reduction strategy papers, were intended to demonstrate domestic ownership of the reform agenda. What happened, I would argue, is that these then went from being public-private deliberation mechanisms, chat shops, if you will, to coordination mechanisms. Suddenly, the public and private sector were supposed to meet together to identify obstacles to investment, to come up with actionable solutions to the problems, and to embody those in policy. And there was real money to be had if one did that in a way that was acceptable to the Bank and the IMF. So suddenly the bar was raised quite substantially in terms of the expectations of the outcomes of the councils. So the question that I want to ask is a fairly straightforward one. To what extent have these councils succeeded as business government coordination mechanisms? It draws on case studies that I convinced my colleagues in the African Development Bank to do of Ethiopia, Senegal, Tanzania, and Uganda in 2012. And I use a very primitive framework, certainly much more primitive than the one we've just heard about, reflecting the Asian experience of coordination mechanisms. Keep in mind that these are an East Asian import into Africa. To say, well, how well did they seem to work on a number of dimensions of good performance of coordination mechanisms? The first possible mistake, instead of seeking the views of the countries as to what their coordination mechanism should look like, the bank and fund staff in their wisdom came with a one-size-fits-all approach. Every council was to look the same. It was to consist of 15 corporate champions, five from the domestic private sector, five from foreign investors present in the country, and five from potential foreign investors. It was intended to identify obstacles to investment, focus on a limited number of issues, generate concrete recommendations, et cetera, et cetera, and so forth. To do the heavy lifting, a secretariat sponsored by the World Bank, funded by the World Bank, was to do the basic analysis. Maybe that's the pocket of excellence that they were looking for. What do we know about what the councils did? Keep in mind we're dealing now with almost 14 years of experience. Two evaluations were conducted, as Lily told you yesterday, one in 2005 and 2009. The results were to say the least decidedly mixed. By 2009, the council in Ghana had completely disappeared. The president lost interest. He failed to hold a meeting for two and a half years, after which the private sector quite realistically, and I think justifiably said, this is not the way in which we're going to engage with the government on issues of policy affecting the private sector. Let's try to do something else. To come back to the politics of the story, the complication also arose that the preceding government in Ghana was the sponsor of the council. So when it was replaced in the elections, there wasn't much of a political champion behind it to look into the council. Uganda was judged to be far and away the most successful, and we'll talk a little bit about why that was the case. Senegal and Tanzania fell between the stools. In terms of general findings from the experience up to 2009, the first point that even the bank evaluators noticed is that the councils were much better at accelerating an agenda of previously identified reforms. Then they were at generating new ideas for reform or new ideas for other kinds of actions. So not too many public actions coming out that hadn't already been on someone's, and usually some donor's laundry list of things that needed to be done. The reason the bank attributed this to the councils was that it created, quote, an atmosphere of discipline and pressure for action in the face of government inertia. So already you can see there's a certain amount of capture going on here. It's our agenda, and the government is the inertial force, and the council is the tool by which we hammer the government to move in the direction we want them to go. But then they discovered that there was a limited amount of implementation of council decisions due they ascribe to capacity constraints in the public sector. Where are they now? Even greater diversity. Uganda remains the best performer. By far and away judged either in terms of the range of activities it's undertaken, the seriousness with which its recommendations have been implemented, or the durability of the actions of the councils. Tanzania is the laggard, mainly because even decisions that have been taken jointly by the public and private sector have not been implemented. Senegal is kind of muddling through, and Ethiopia really is too new to tell, but as I'll tell you, begins to show some signs of trouble. The councils have been best still at implementing a donor driven regulatory reform agenda. And my least favorite World Bank product of the entire 50 years of its, 60 years of its existence, unfortunately became the agenda for virtually all the councils. And that was the doing business laundry list of reforms. They've been somewhat useful, and this is important, in building public-private consensus around a number of important initiatives. Two that I'll just mention, Kalimokwanza in Tanzania, which is the agriculture first set of policies, substantial amount of private sector buy-in needed there, and some identification by the private sector of problems in pursuing the Kalimokwanza agenda. Also, ICT backbones in Uganda and in Tanzania have gotten substantial public sector, private sector partnership buy-in in those areas. They have been extremely poor at identifying constraints to firm performance or implementing solutions, finding broken wheels, as John Sutton would put it. And there's, of course, no quantitative evidence that they've succeeded in boosting investment, much less economic growth. So what can we do to evaluate it? Well, what I wanted to do was to go back to a literature that looks at the coordination process in Asia. Basically comes up with two themes that we see running through the new literature on the implementation of industrial policy. And the idea is that you need to have a process of consultation and coordination, a close relationship with the private sector first of all to identify the binding constraints. You need to know what's really the broken wheel, as opposed to perhaps what people in the basement in Washington think is the broken wheel. And secondly, you need feedback. And this is a very important thing I think we oftentimes miss, but it's a point that Danny's made in a number of his writings. You need to know whether the intervention that you've designed is actually working. This idea is very similar to the coordination mechanisms we've seen in Asia, ranging from Japan to Vietnam. It has different institutional expressions. It has different champions, but it's very, very much a theme that runs through it. I'll mention there are two exceptions. Thailand and Hong Kong are quite different in terms of these things, but otherwise you can find a thread that runs through coordination mechanisms. And in my simple-minded way I came to four elements that I thought were really the key drivers of success. The first one is a high level of commitment of senior government officials. And for that, I mean very senior. The one clear insight that Jim and Horst had, which was correct, is that the head of state has to be engaged. He has to be perceived to be engaged, and the private sector needs to know that he's following up. But beyond that you also need, and we can put different labels on it, senior members of the political administration and the civil service to be on board. And without that it simply doesn't work. The second one I think of is focus, and it's a simple idea, but it comes back to the broken wheel notion. Focus on a limited number of actionable issues actually deliver an actionable program and see that it gets done. And that may be the hardest thing in the implementation experience, not only of African countries, but Asian countries as well. A willingness to experiment, and this is something I think the Asians get too little credit for in a way. And it's partly our biases economists trying to interpret Asia's experience in terms of neoclassical economics. I once had the privilege of meeting with a man who had been the career vice minister for planning in Korea. And the first thing he told me, I was with Nancy Burtsall and Dick Sabbath at the time. The first thing he told me was, you know what we're proudest of? We didn't have a single economist running the Korean economy during their rapid growth period. Well, how do you make public policy without economists? Well, you observe, you inquire, you find out where the broken wheel is, you try an intervention, and you find out whether it works. Closest science I can think of to this is not economics, it's pharmacology, right? Bark off a tree reduces cancer. If it doesn't kill too many patients at the time you treat them, you put it out to the general population. 35 years later, someone wins the Nobel Prize in medicine for telling you what the biochemistry of it was that worked. So experimentation is important, but to experiment then you need to observe, to experiment, and to give feedback, and careful attention to feedback. So how did these councils perform on those dimensions? Commitment. Well, that's what happens when a president gets bored, right? Loses interest, council loses traction, private sector walks away. Not good for the World Bank's reputation, but you notice that they don't talk about that when they report on these things. Uganda, that's what happens when a president follows up. The important difference between Ethiopia, Senegal, and Tanzania, and Uganda, is that Museveni for all his faults, one, conducts the meetings twice a year. That's as rapidly and as frequently as anybody does it in Africa. Two, he actually asks his civil servants and his minister what went on at the meeting. Three, he says, have you followed up on what went on in the meeting? And he actually has a reputation within the private sector for demanding reports. So a president is actually showing interest. Ethiopia, Senegal, Tanzania, I would argue it's tepid interest. It's keeping the donors happy, it may be keeping the business community happy. Melis is an interesting case. Melis agreed to establish the council. Melis also has a reputation for dealing directly at the sectoral level with industries such as cut flowers, with exporters, with metal mechanical industries and plastics. He was far more enthusiastic and there was far more follow through at that level than there was at the level of the council where before he died a council was now actually held. So this is why I say it may be in trouble. Focus. Here I think the biggest issue was having the agenda of the council hijacked by the donor community. Let me spread the blame a little bit more broadly than just to the World Bank. But the point is the very easily quantifiable box tick regulatory agenda in doing business became not only something the World Bank was very proud of but also got embedded in the country like Tanzania into the budget support program. Which meant that it was a high stakes game for the government to make sure that there was follow through on the doing business agenda in order to stay in the favor of the donor community. What that does I argue is fundamentally contracts the rationale for having a public-private coordination mechanism in the first place. You don't want your agenda written in Washington. You want your agenda written in Dar es Salaam. By this process of having the private sector identify the actual constraints to investment and firm performance. There was certainly an attempt to address too broad a range of issues even if we think about the successes in terms of Kalimo Kwanzer ICT. A bit of a kind of dilution I think of the idea of a coordination mechanism and very very little evidence of homegrown analysis and action on sector industry specific constraints. Working groups were the idea for the generators of ideas. This was supposed to be where the experimentation took place. Here the issue has become largely membership of the councils. Because of the original design even though it's evolved small scale of this min size industries business associations are largely excluded from the council process so they don't get engaged in creating new ideas. It was also of course limited by the donor agenda itself. Outside of regulatory reforms we've seen a few attempts at innovation and they're quite interesting but they don't add up to very much. In terms of feedback this is probably where all of the councils get a failing grade even the one in Uganda. There hasn't been a systematic process of follow through call it secretariates were supposed to do that. Except using the doing business scorecard which I would argue is the wrong scorecard. There hasn't been an effort to quantitatively determine the outcome of council decisions. And it's been uniformly poor in terms of other initiatives to follow up so follow through feedback remains perhaps the most fundamental problem. Some lessons learned. This is kind of circular statement but it makes sense. Commitment depends on getting things done which in turn depends on commitment. Think of the two bookends. One of the reasons why before lost interest in Ghana was because the council didn't get anything done. Because the council didn't get anything done before lost interest. And of course the council didn't get anything done because everybody knew the president didn't care about it. So you have to have that sort of follow through of the leadership focusing solving and then getting the feedback. The donor dominance the agenda was unhelpful. The public administration in all cases takes its cue from the top that's why it works better in Uganda than it works in Senegal or Tanzania. And I would argue still Philanene blank you like for the country unlike Calvin Coolidge where the business of America was business. The business of any country in Africa of the political leadership the top leadership still is not business. This is part of the reason why there's lack of attention. I do not still believe that the kind of notion that it's the middle level firm that's going to drive the transformation of economy has yet really filtered through to the people who are at the very top. That I think is important. The donor driven agenda stifles focus and experimentation so a lesson here for the donor community if you want to create something and have it be a homegrown solution to coordination problems then back away and let them determine the agenda. They need to be more flexible and these coordination mechanisms need to be more flexible in terms of representativeness and improve an agenda setting. They certainly need to dump doing business. And use the councils that they were intended to identify binding constraints to investments. And remember it's not just these four that we're talking about. We're talking about 16 of these across Africa. So there is an important change here in the way in which I think the major actors who support them exercise self restraint finally focus on a few fewer things. Follow through get feedback and try to implement what you do in a way that creates that circularity. Higher commitment because you've actually followed through which in turn generates higher commitment. So with that thank you all very much. And a shameless ad for Finn and my project. Please visit our websites at Brookings and wider for learning to compete.