 I was extremely pleased to be invited and I'm deeply honored and I was also extremely happy that the first keynote speech was by my good friend Martin Reveillon and I must say that after hearing Martin I remembered when I first started teaching development economics this was decades ago and I would start with Weber and Marx and then I would finish essentially with Herod Domar and I would always say it's going from the sublime to the ridiculous and in some respects I'm afraid that I some of the things I'm going to say because I'm going to look at the anatomy of growth and institutional structure in a very nitty-gritty way so I hope that what I have to say is not going to sound too ridiculous so let me get started and again a quick digression this morning as I was walking over I thought about the serendipity of creativity and that a person of my age simply cannot be creative I mean they say that mathematician reach the the age beyond which they cannot be creative at about 22 for physicists at 25 for most of the population it may be 10 there's nothing more creative than the mind of a young child and lately I've completed reading a book on physics by Letterman and one of the examples that he gives is Elvin Guth theory of expansion of the universe Guth was a young particle physicist who knew nothing about cosmology and he was working on a very different problem and suddenly he realized that he had all of the pieces together to explain something that until then had not been explained namely how come the universe was constantly expanding and in a very modest way he said really my contribution was ready to be achieved I just put all of these pieces together nothing to be very proud of and Wolfgang Pauli who was one of the Wunderkinder of physicists in his time when he heard about it said I know too much to be creative well this sounds arrogant but there is some truth to it over time you accumulate a baggage and it's difficult to see with entirely new eyes but I will try to make some contributions I think there are a few things that I have to say which you might be considered considering if not creative a new way of looking at things so let me very briefly go over the content of my presentation the first thing I want to do is to say something about inclusive growth in sub-Saharan Africa and perhaps how it could differ from other developing countries then I will go into a tale of two worlds two very very different ways of looking not only at the degree of inclusiveness in growth but also of the kind of strategy that is called for to address inclusive growth then I will go into structural transformation and the kind of structural transformation that is needed to generate growth then one of the topics that I have been working on for the last ten years and some of you have heard me talk about it already I'll say something about the interrelationships linking growth inequality and poverty and two different strategy the pro-pro growth as well as the pro-growth poverty reduction then I'll say something about what Martin was mentioning yesterday how is it or why is it or is it really so that a high initial poverty can dampen growth can be a break on growth and is this really true within the context of Africa then the second part of my presentation and again because of time constraints I'm afraid I won't be able to say too much here is on the architecture of institutions conducive to inclusive growth in a different paper that I did for Jaika I looked at institutions that could be conducive to greater growth so again depending on time I will say something about it so before it is too late and I've been caught short many times over a long lifetime let me give you the the main messages first of all what I will do is emphasize the reverse causality from poverty reduction to growth which I think is a very important link I will claim and it's a tentative claim I think much more research is needed that the recent structural transformation in many of the African countries is more inclusive than it has been in the past and I'm going to raise a question as to whether there is something different in the anatomy of growth in Sub-Saharan Africa compared to other developing countries at one time Eselin and Levin were talking about the African dummy so in a way it's an attempt at if not elucidating this dummy trying to understand it better one of the points that I'm going to make and I hope Martin is here because it bears directly on his work it is not clear that high initial poverty we don't retard subsequent growth in Sub-Saharan Africa I've done some additional work using his own data set and it shows that the results for the African sub-sample tend to be different than for the the whole sample then my recommendation is going to be to follow a strategy of pro-growth poverty reduction in the selection of inclusive institutions and again I will claim that such institutions exist and can be designed to conform to the conditions which exist in Africa so very quickly what are the the features of inclusive growth within the context of Africa which perhaps are somewhat different than within the context of other developing countries the in the case of Asia and I just completed a paper for the Asian Development Bank on inclusive growth it is not clear that growth has led to greater inequality it has led to greater inequality in some countries particularly in China but overall there is one cannot necessarily generalize in a number of countries inequality has not increased so in the case of Asia and perhaps less so Latin America to adopt the relative definition of pro-pro-growth does make a little bit less sense than within the case of Africa but I think within the case of Africa it's quite clear that it is important to have a reduction in inequality as an objective a second I think difference and again it's maybe a minor difference is that the inequality of opportunities is probably greater in Africa and it means that in terms of benefits derived by the the poor and the near poor in terms of education in terms of health are really significantly less than the benefits derived by individuals and households in the higher income brackets so if one really wants to if one really wants to change the income distribution and facilitate a structural transformation and eventually lead it to a social transformation where the poor have greater voice I think it's extremely important to emphasize social protection schemes that provide more education better health to the to the poor and then employment employment is absolutely crucial within the context of Africa so now the tale of two worlds and I think and this is a caricature and take it as a caricature there are two economic camps I think and you can distinguish these camps are going to two different distinctions the first distinction is that everybody agrees that growth is a necessary condition is essential for development to take place but some economists feel that a proactive government is crucial others feel that we ought to have a minimalist government leave it very much to free enterprise and in this connection the New York Times recently had a wonderful article where they confronted a marcher sense view with that of Baguatti and Sen was quoted as saying this Indian administration thinks that the only thing that works is business that's a disastrous position to take but what he on the other hand said sense position is mistaken and dangerous since money spent on government programs is largely wasted India's myriad problems have less to do with poor health and literacy than a poor investment climate so again for the sake of simplification let's call sense position growth with government and Baguatti growth is enough GG and GE the second distinction has to do with how the perception by different economists how of how inclusive the present growth spell is in sub-Saharan Africa and here the glass is seen as half full and again it's a caricature by people like so like Martin young IMF on the other hand it's seen as half empty by certainly AEC a large number of publications the project on the interaction between growth and poverty and many other collaborative projects for Sue the African Development Bank and I could count myself in this camp so there is I think a large overlap between the half empty and government proactive government role and between the half full and growth is enough for more proponents however and I found this very surprising I went back to some of the recent IMF document there appears to be much more convergence between the HG GG world and the HF GE world in the context of Africa than other developing regions even the IMF recognizes a an important role for the for the government and the need for social protection and labor schemes and this I think is is somewhat encouraging okay quickly some facts to judge the degree of inclusiveness in Africa first since 19 to 2000 remarkable growth something like 3% per capita GDP per annum compared to 0% between 1960 and 2000 headcount poverty went down from 58% to 48% between 96 and 2010 Ravi Kambo at one time made the I think the very incisive comment that really one shouldn't just look at the headcount ratio one should look at the number of poor there the situation is not quite as good there has been an increase in the absolute number of poor income inequality is still very high only 28% of the labor force has stable jobs 63% have vulnerable jobs the the McKinsey report that I would recommend has changed the definition we used to talk in terms of formal as opposed to informal employment they talk in terms of stable versus vulnerable jobs and then I think it's fair to say that the the distribution of social benefits tends to be very uneven in Africa so let me now move to the what I call the anatomy of growth and particularly the the structural transformation the world development report of 2008 which was on agriculture and development made the point that until the early 2000s in sub-Saharan Africa workers tended to be pushed out of agriculture rather than pulled into other sectors into more productive sectors and this you could characterize as a migration of my of misery these people essentially by being pushed out were forced into unproductive jobs in the informal sector and they contrasted this with a very successful structural transformation in most of the Asian countries now the I don't know how much of this you can see but the the point here is that the flawed taking the wrong one a flawed structural transformation would lead to a vertical time arrow what it would mean is that the share of the labor force in agriculture declines with no growth in income on the horizontal axis you have the the log of per capita income so for many African countries until the early 2000 the structural transformation was flawed in the case of most of the Asian countries you see that it's followed the more normal pattern so what I did I was curious to see well here's again the same graph the for Africa and you see in many countries this almost vertical line so what what I did I'm getting I updated the WDR 2000 a data set and it's very difficult to find data on labor shares for more than two sometimes three periods but I was able to find 14 countries and the structural transformation appears much more normal for most of these countries the only really outlier was Mali there were a few countries Botswana Ghana Liberia that reveals a rising share of agricultural employment together with growth and the only possible explanation is that maybe because of the changing composition of agricultural output agricultural workers were doing better in these in the production of these new products so this is this is what we were able to do and you can see it may not be obvious but you can see that the structural transformation over the the last decade I should say between 2000 and 2011 tends to be following much more the normal pattern than it did in the in the past so tentative conclusion the structural transformation appears to be relatively more inclusive what does it reflect I would say it's reflected a more orderly and productive migration process the policy environment has changed the exploitation of agriculture is less now than it used to be yet much more needs to be done so let me now move to the interrelationship among growth inequality and poverty and here basically two forces are influencing this interrelationship the first force is globalization and that's essentially exogenous from the standpoint of a given country there's very little they can do about it and the other is the development strategy that is followed by a country which at least partially is a endogenous so the these two forces globalization the development strategy that is adopted will affect this nexus and each of the links first the link from openness to growth the link from openness to distribution the link from growth to distribution which is the old Kuznetz curve the link from inequality to growth as well as a link from growth to poverty and the link from distribution to inequality will have to be looked at very carefully to understand this the dynamics of growth so very very quickly let me go into some of these links if you take the the blue the light blue arrow it's very clear that a country like Kenya relying on export of horticultural products that the impact of this growth will be very different than oil exports from Nigeria in one case it's a labor intensive product in the other case it's almost a completely capital intensive product the pace of growth may be very high in Nigeria but the pattern of growth is hardly likely to be inclusive and vice versa the link the green arrow linking the development strategy and globalization to distribution here looking at at a whole sample of African countries I found as many countries in which inequality had increased over the last ten years as countries in which it had gone down but again if you if you want to come up with a composite index and you weigh it according to population then you can say that the genie coefficient has gone up not very much I think it went up from something like forty four point four to something like point four seven the the Kuznetz curve has been largely dethroned for the simple reason that there is no iron law that necessarily leads to a an increase in inequality at an early stage of development followed by a reduction it depends at least partially on the policies that are followed by the by the countries and then of course the the other two from growth to poverty from distribution to poverty you can say that income distribution acts as a filter on the impact of growth on poverty now later on what I will do is emphasize very strongly the reverse causality starting from poverty but this will come in just a few minutes so in terms of the impact of inequality on growth there are two conflicting theories the old neoclassical Kaldor theory claimed that since the marginal propensity to save of the rich was higher than that of the poor a an unequal income distribution was almost a free condition for economic growth the new political economy of development people like Roderick claim on the contrary that greater inequality leads to social conflict political conflicts and can dampen growth so again depending on which one of these two theories you believe in you're going to have very different policy prescriptions I am I rather strongly in favor of the the new political economy of development story okay the now we come to the reverse causality from poverty to growth and Martin in a I think a very important article in the AR finds that high initial poverty rates have sizable negative impacts on the growth rate and the what we already knew the growth elasticity of poverty reduction tends to be lower in countries with a higher initial level of poverty so if if I mean in a way you can have two different strategies I mean they they very much overlapping if you want to but one strategy which is the well-known pro pro growth strategy grows from growth to lower inequality to poverty reduction but the other one is poverty reduction will lower inequality and by lowering inequality and also for other reasons that are not entirely clear is going to contribute to growth and I in my message I would strongly favor looking at the feasibility of this strategy okay yesterday Martin talked this was at a probably a meeting where not all of you were about poverty traps well we know that there are many many poverty traps and I list some of them here I don't want to go into them and then the question is can some of these poverty traps be changed these are vicious circles can be can they be converted in virtuous circles so what I would propose here is a development strategy that by having institutions and having policies that reduce poverty directly can also lead to greater productivity by changing very often the behavior of the poor think of a social protection scheme a la a la Progresso a la oportunidades a la Bolsa Familia in Brazil or in the case of Africa the Ethiopian public works program by reducing poverty it frees the recipients from being unable to take certain risks there's plenty of evidence that with the money that they get or that they save by essentially having not to pay for fees for the children free lunches and so on they can start businesses and it can be productive so you could in a way think of the intervention reducing poverty as being productive and leading to a virtuous spiral okay the some of you may say well what's the difference between pro gross poverty reduction and pro poor growth strategy I think the the main difference is at the the trigger point in the former is directly on poverty reducing poverty directly which it isn't in the in the second case now coming back to Martin his point that tie initial poverty rates have sizable negative impacts on growth I confronted it with some evidence well the first point I think that I should make here is that when you first encounter this statement you say well of course the countries that have high poverty will have low growth but his point is a much more subtle one he checks for the endogeneity and he claims that there is a causal relationship between high initial poverty and subsequent growth now I confronted the this hypothesis with micro evidence from 15 Ethiopian villages and something like 31 district and what I found was exactly the opposite that in both cases high initial poverty led to high growth so let me first give you the evidence this is a case of Ethiopian villages on the vertical axis you have average for capital consumption growth you see that countries with high initial incidents of poverty have higher growth in the case of again Ethiopian villages and looking now at the average rate of growth in poverty reduction poverty the average rate of growth of poverty reduction was higher in the villages where the initial poverty was high same thing for the Rwandan district so how does we how can we resolve this paradox well one way of resolving this paradox is to hypothesize that the Ethiopian and Rwandan governments allocated anti-poverty and social protection funds proportionately to the incidence of poverty and I checked with the same as who arguably is one of the most knowledgeable Ethiopian economists and he said this clearly or there clearly was and still is a conscious effort in in that direction so then my curiosity led me to try to redo retest the rebellion equations using the African sub-sample Martin's sample consistent something like 97 countries the African sub-sample 32 so we estimated the regressions and surprisingly the results suggest no statistically significant correlation between initial consumption expenditure and per capita consumption growth but low and behold significant and positive correlation between initial poverty and consumption growth and here you have a diagram that shows this again on the horizontal axis you have the initial log head count ratio and on the vertical one you have the annual per capita consumption growth and this was true for those countries that had two growth spells as well as for those countries that had three growth spells what and this is highly speculative what would be a speculative explanation one could be that the poorest African countries in the past underwent the greatest improvements in governance and vice versa for diva would be an example of vice versa suffered more from ethnic and other conflicts and shocks that were at least partially resolved and and again it it raises the issue of the African dummy and I raise this as a big question mark at and as an area that really ought to be researched further so now and and do I have about five minutes five five minutes so very very briefly the institutional architecture for inclusive growth and in I think an extremely incisive book that I recommend to you by Asimoglu and Robinson why nations fail one of the main conclusion is that growth and development can only be sustained if anchored on inclusive political and economic institutions and then they essentially say that the tragedy of Africa is that highly extractive colonial institutions created an opening for unscrupulous leaders after independence and I think it's it's not unfair to say that in many African countries there is a vacuum of national inclusive institutions now the the I think the perhaps the main reason for this is that in the case of many African countries the the loyalty of an individual is not to the state it's not to the nation but it is to the clan or the ethnic group so to come up with national inclusive institutions is is a very difficult process because it means working out all of the necessary compromises and different values and objectives of different groups now there is again this is a claim on my part there is strong and convince convincing evidence that social protection and labor schemes exist that are both poverty reducing and productive so that means that there is no necessary trade-off between equity and efficiency in the old days it was taken for granted that any attempt at reducing poverty would have an opportunity cost in terms of efficiency so appropriate poverty reduction institutions could lead to higher growth now I try to think of the the major pillars of the architecture of inclusive institutions and it seems to me that they are in the case of Africa there are four of them greater centralization secondly they have to be broad-based given the large number of poverty traps that exist you cannot really operate on each of these poverty traps independently you really need a broad-based strategies the schemes already have to be there you cannot wait until a shock it's too late I mean Indonesia paid a very high price in 97 by not having in place such safety nets and again it's very obvious that there is no one size fits all the specificity the context specificity is very high so they have to be tailored to the the culture institutional capability governance of a of a given country so what I did and I probably won't have time to to we go into it but I tried to come up with not only features but also issues and recommendations for specific institutions within three areas very interrelated small-scale agriculture infrastructure and social protection schemes I will say that Africa can learn a great deal from the experience in East Africa one of the issues is that since so many of these countries are small you need supranational institutions in terms of infrastructure as well as in terms of research which which means greater regional integration the present structural transformation is inclusive but it needs much more in order to create a social transformation and this brought up a thought that I had I know the transformation transformation is a key concept in the new program of research of wider and I think one area that you could look at is how structural transformation by changing the income distribution and also by providing the poor greater benefits in terms of education health and so on how this can lead to greater voice for the poor greater representation and ultimately social transformation so the key here is how do you go from structural transformation to social transformation so I get the message let me conclude what are the concluding messages structural transformation in the present gross spell is more normal than it used to be which is a good news the relationship between initial poverty incidents and subsequent growths needs to be clarified in Africa the new pattern of growths even though more inclusive than in the last millennium requires major institutional changes I believe that a pro-growth poverty reduction strategy is both appropriate and feasible there's a strong case for the productive role of social protection protection and labor scheme and the the present growth spell in Africa what people call the African Renaissance is probably the right time to do something it's a lot easier to come up with institutional changes during a period of growth than it is during a period of stagnation so thank you very much mr. chairman