 of this final session. So our conference is almost over, but it's, there's one very important element still left, and that's the final keynote lecture. And so our keynote speaker today is Dr. John Page, who's the senior fellow at the Brookings Institution in Washington DC, and he's also affiliated provider as a non-resident senior fellow. Before joining Brookings, John had a very long distinguished career at the World Bank, where he served, among other things, as the chief economist for the African region and the director of the development policy unit. He has also a number of other affiliations, including the National Graduate Institute for Policy Studies in Tokyo and the International Growth Center in London. He's a author of numerous books and articles, and these have been published in, by, by leading outlets just as the, like the Oxford University Press or the Journal of Development Economics or Review of Economics and Statistics, just to name a few. He received his doctorate from Oxford University, where he was a road scholar. So, obviously, we are extremely happy to have Dr. Page here with us today, and his title is How to spend it financing Africa's infrastructure gap. John, please, the floor is yours. Well, thank you, Yuka, and I'm both surprised and pleased to see that anyone would applaud after two very rich and busy days, another talk. But let me try to talk a little bit about the expenditure side of the public economic story. When I heard that McKean was coming to give the first keynote, I said, well, that solves the question of how to get it because his specialty, of course, as you know, is tax policy and tax administration. And I didn't want to give a talk about, you know, the sort of usual public economic stuff of public expenditure. So, I thought, well, let me pick a specific topic, and the topic is infrastructure. And let me start with a little story to motivate this a bit. On Friday and Saturday last week, I was in Nairobi, and I was at another conference, and I was sitting, and the speaker was about to speak. And he said, you know, the following table will throw light on the answer to this question. And at that point, the power shut off completely. It's an experience which anyone who has lived and worked in Africa has felt on many, many occasions. And in fact, we find ourselves in a world in which Africa faces a genuine crisis in power. The World Bank estimates that roughly 600 million Africans have no electricity connection whatsoever. About 30 of the 53 countries across the continent face regular power shortages. A great number of those countries also pay very high prices for emergency power. And in every survey of enterprises that have been conducted, both by the World Bank and the World Competitiveness Survey, lack of reliable, efficient, and affordable electrical power turns out to be identified as one of the major impediments to competitiveness by firms across the continent. The map here I chose because it gives you a sense of the varying geography, if you will, of access to electrical power, the darkest shaded countries of the ones with the least. The north and south, as you can see, are the best access. And in the middle we have a large number of countries where between half and maybe three quarters of people have genuine access to electrical power. But it's not just electricity, unfortunately. It's all infrastructure measures. These are standardized infrastructure measures that were done by the World Bank for its 2009 African Infrastructure Finance Report. If you look down the two columns, what you'll see is that in every category, paved roads, mainline density of telecommunications, generation capacity, electricity, water and sanitation, Africa significantly lags other low-income countries in all of these characteristics. It is small wonder, I would argue, that the African delegations to the United Nations pressed very hard to make infrastructure one of the new sustainable development goals, resilient infrastructure, but infrastructure nonetheless. You will recall that the millennium development goals were completely silent on the subject of infrastructure. But just to say that we don't have it doesn't mean we should spend money on it. And this, I think, to be perfectly frank, was one of the mistakes of the millennium development goals. We set ourselves up with a set of merit goods, obviously meritorious, poverty reduction, health education. But we really never did the hard work of asking whether those were the priorities in terms of economic development. In the case of infrastructure, the infrastructure, the gap is growing and it's growing because we have starved it for funding over the last 15 to 20 years. Paved roads, telephones, power generation, they're all growing much more slowly than in other developing countries. It's one of these well-known statistics a little bit like the comparisons of Ghana and Korea in the 1960s to make the point that in the 1960s Africa had more power generation per person than South Asia. Today that is no longer true. Investment in power simply hasn't matched population growth nor economic growth. About 80 percent of the road network remains unpaved and the road stock is actually contracting as we speak. But the question is really why spend on infrastructure as opposed to something else? And the answer, I think, comes at least from a preoccupation of mine, which is that I've spent the last six to eight years studying the question of how to make Africa more competitive. And that involves raising the productivity of existing firms and attracting new firms. And the focus on infrastructure I think holds the key to both of those. Both in terms of the productivity of firms, there's now fairly persuasive evidence, recent work by World Bank colleagues that indicate that power infrastructure is strongly correlated with lower firm level productivity. Just to kind of pull statistic, if you will, the production loss in terms of sales from African firms that have power outages is about eight percent. That's more than double the three and a half percent that it is in East Asia. Uganda, not too far distant from us at the moment, on average firms lose about two percent of domestic sales and about one percent of export sales due to delays in transport services. In most landlocked countries, transport costs are 15 to 20 percent higher than they are in coastal countries. And this one really caught my attention because it's close to home. It's more expensive to transport cargo within Mozambique than to ship it to a different continent. So this is quite clearly a major impediment to competitiveness and to trade. There's also fairly robust evidence that infrastructure development has a positive impact on long-run growth and somewhat surprisingly reduces income inequality. It was surprising to me until I reflected on an article that Ricardo Hausmann wrote some years ago, in which he pointed out that one of the major drivers of inequality in most of the countries he knew was access to infrastructure and the fact that infrastructure is not uniformly distributed geographically in any country that we know. By one estimate, and this is an estimate due to the African Development Bank, the African Union, and NEPAD, all of this adds up to a loss of about two percentage points per year in GDP growth. What it all sums up, I think, is that in short, economic and social returns to, I'll put this in italics because it's very important, appropriate investments in infrastructure are high. But then we have to sort of sort out the question of what do we mean by appropriate investments and what sorts of investments and rules should we use? How big is the financing gap? It's big. The same World Bank study that came up with the measurement of the gap in terms of physical infrastructure on standardized measure made a very good faith effort to try to establish the volume of the financing gap. And they finally came to the conclusion that over the next 15 to 20 years, the continent will need about $93 billion a year in new funding to meet infrastructure needs in just the following energy, telecommunications, transport, water supply, and sanitation. 40% of that spending should be dedicated to power alone. That's the size of the power problem at the moment. And the topic we'll come back to, and something which is oftentimes neglected, about a third of that spending is needed for maintenance. Countries across the region spend about $51 billion a year, according to the IMF and its most recent report. That gives us a financing shortfall of about $42 billion a year. It's a big number by anyone's estimation. But the financing needs also vary across countries and across types of countries. Fragile states, conflict-affected countries have infrastructure spending needs that are on the order of 35% of GDP or more. Non-fragile states, somewhere around 20 to 25% of their GDP to build and sustain basic infrastructure. And the punchline is that magnitude of the gap is most likely underestimated. In 2009, the World Bank's estimates left aside completely the issue of urban infrastructure. And as we know in a rapidly urbanizing continent, which we face today, urban infrastructure, anyone who's experienced the traffic in Maputo, which is relatively mild, or I would argue Dora Salama Nairobi, has a sense of how seriously urban infrastructure is underdeveloped in a rapidly urbanizing Africa. The other thing is that the way in which we measure external financing by commitments rather than by disbursements is not entirely compatible with annual budget spending. So when the bank says we need 93 billion dollars a year, that's 93 billion dollars of budgetary allocations and spending. But commitments actually materialize in a naive and fashion both from development banks and from private investors in the course of the implementation of projects. So we're quite likely to see quite a bit of variability in the availability of foreign finance as we move forward. Bottom line, if infrastructure is important because it has high social returns, the size of the financing gap is such that we run a serious risk of being unable to realize those social returns without some major changes in the way in which we do business. Something that, again, as I begin to work on this topic struck me, and I suppose it's partly the bias of having worked for a long time for an aid agency, is that it is really domestic financing that drives infrastructure spending. A statistic that Mick didn't give the other day, and I find quite impressive, is that African countries generate about 500 billion dollars annually in domestic taxes. That's a fairly large revenue effort, and Yukon and I were talking on the bus on the way over about the fact that when you look at the numbers, African countries are actually expending quite a bit of tax effort in terms of revenues per dollar of GDP, but because their poor economy is the absolute size of the funding pool, it's quite a bit smaller than in richer economies exerting the same amount of tax effort. It comes to about 70%, 73% to be precise, of the total funding is done by domestic expenditures, of which about 10% comes from the IFIs, the African Bank, the World Bank, other international financial institutions, and to a lesser degree from bilaterals, so we get down to about 60% of funding of infrastructure, which is due in a sense to domestic tax effort. That's still much too little to close the infrastructure gap, but to increase revenues, countries will have to either address the set of problems that we discussed yesterday in terms of tax policy and tax administration with its attendant set of downside risks as well, or, and this is where I think there could be a free lunch, for the increasing number of countries who are becoming resource rich economies, seriously following through with effective and appropriate revenue policies for natural resources offers an untapped pool of funds. This has not been an easy task in Africa. One of the aspects of the export or the natural sources curse across the continent has been that countries with large endowments of natural resources have had very poor relationships with the natural resource extractive sector and very poor mechanisms for extracting natural resource rents. This is changing. There are new initiatives. The Natural Resources Charter, which is a product of work at Oxford by Paul Collier and his colleagues, offers a series of highly specific and very practical recommendations about how governments can follow through on taxing natural resources, but just in the last two weeks, as I've kind of been traveling around the East Africa region, I was reminded that it remains the case, for example, that people have been very unhappy with the tax regime in Zambia, in the copper sector, which has generated extremely little, especially during the copper boom. Tanzania has now introduced three bills into parliament that have been passed, as I understand it, to reform the regime of natural resources taxation in the mining sector. Nigeria remains confused and fraught in terms of its relationships and its natural resources taxation. So a large number of countries and an increasing number of countries here in Mozambique natural gas in Tanzania natural gas in Kenyan newly discovered reserves of petroleum in Uganda reserves of petroleum have to address this issue and this could become part of the solution to the natural resources curse. But money isn't going to solve the problem alone and here I want to turn to the public economics part of this, at least for a bit. I would argue there are at least three major domestic policy goals governments need to pursue in parallel with attempts to increase the funding for infrastructure. The first is increasing the efficiency of infrastructure services. The second is improving the quality of public spending and the third is aligning investment effort to investment outcomes. I'll talk about each of these for a bit. In terms of increasing efficiency it's interesting because the results are mixed. In some cases infrastructure and infrastructure services are costly because costs are high. In other cases infrastructure and infrastructure services are costly because of high profitability of service providers. So that it's not a kind of one size fits all single solution to the problem. Power is probably the clearest example of cases where infrastructure costs are higher in Africa than they are elsewhere. In part that's because Africa is a collection of small countries which means that they are small power markets which in many cases also means that they fall below the famous 500 megawatt line the engineers like to draw for high unit costs of delivery of power and many of these rely on small scale diesel and sometimes natural gas generation. What is perhaps disappointing here is that despite the calls for regional integration commitment to the African Union to regional integration and the growth of regional economic communities we have very very few cases of integrated power grids in the region and almost no cases of power trade. This should be the first thing that regional economic communities are looking to not a secondary or tertiary item to be put on the list after we've discussed trade integration and whether or not we're going to have a common currency but simply not happening. In the case of telephones and internet services it's a mixture. For those countries with access to submarine cables and significant competition in the sector cost to be relatively low and we've seen a number of countries Kenya and Senegal being among the most noteworthy that have suddenly become providers of back office services because of low telecommunications and information costs but countries without access to submarine cables continue to depend on satellite and that's an extremely high cost way of moving data and information around the globe. Those countries like Senegal that have moved from having competition to having a monopoly with respect to access to the cable have also found themselves in situations with very high costs of access to the global information highway simply because of high profitability on the part of a monopoly provider. Perhaps the poster child for high profitability in Africa is the trucking industry. High road freight tariffs in Africa have a lot more to do with profit margins than they have to do with road costs where it turns out some fairly careful work that's been done by people like Gail Ravallar and others tell us that the actual costs of operating the fleet in Africa are not that much in excess of those that we find in other parts of the world but because of regulations because of cartels and because of other non-competition arrangements among trucking services providers the cost of transport services are extremely high. In some regions it would probably be more cost effective to actually promote competition than to go through the process of rebuilding the highway network or improving the highway network. The bottom line is that again my former colleagues at the bank tell us that if you could reduce these inefficiencies if you could make changes to the market organization of the transport sector for example the financing gap would fall by about 17 billion dollars a year so that's a fairly significant slice of the 41 but there are some cautionary notes. The first one is that relatively few countries in the region have a modern regulatory framework for infrastructure. Telecommunications is most advanced and in many countries now telecommunications regulation is actually moving toward global best practice but transport frankly not making much progress at all. Add to that the fact that cross border institutional failures are widespread and we come back here again to the question of regional integration and regional communities. Transnational infrastructure not just power which I talked about already but rail and road links are underdeveloped and are characteristically difficult to implement because countries are unwilling to cede sovereignty to a transnational authority. In eastern and southern Africa where we are costs are oftentimes affected by the cost of delays at border crossings, at way bridges, at ports, long customs procedures, the dwell time for trucks along some of the major corridors in this part of Africa sometimes go into the four to ten day range. So if these delays again could be significantly reduced on average countries get about 20,000 more kilometers out of their truck fleets than they do today with a substantial potential for reduction in costs provided that there's more competition. Second point is to improve the quality of public spending. Something that's sometimes missed is that a key to the appropriate quantity of public spending is the quality of that spending and public investment efficiency in Africa to some degree as you can see from the right hand chart here lags at least according to the IMF public management index that of other developing regions. What is quite interesting is there's a huge amount of variability across countries including some state of the art world class, public investment efficiency or public investment activities in individual countries but on average across the region things are not working all that well. What's needed from the context of infrastructure are better capabilities to design, select, implement and evaluate public projects and so I want to talk about a particular pet of mine which is Better Project Appraisal. When I first broke into the economics profession applied well for economics and cost-benefit analysis were the hottest topic of the day. We had the UNIDO manual which was done by Steve Marglin, Marcia Sen and others. We had the famous Little and Merleys volume and subsequent studies on social cost-benefit analysis. The World Bank was rapidly adopting social cost-benefit analysis as a tool for project evaluation and many of us somewhat naively believed that that was going to be a new tool for better public decision making. Unfortunately much of that has disappeared and I say unfortunately because project selection and implementation need to be protected from the influence of political patronage and prestige and one of the tools in the toolkit of public economics is systematic use of project appraisal. In fact if you get a public commitment to systematic use of project appraisal as they had in Chile for many years and as they had in Botswana and still have in Botswana it's a kind of commitment technology to improve the efficiency of investment but what it also requires is actually building a group of economists with training in project appraisal making them responsible both in the line ministries and in the finance ministry for project appraisal in preparation and getting the incentives right making their career advancement, their promotion, their prestige dependent upon the ability to prepare sound appraisals, recognize weak appraisals and validating that by rejecting at the level of government those appraisals which do not conform to the criteria they've been established. The bottom line is a particular disappointment for me. The eight community here could assist quite substantially in this if indeed they did project appraisal themselves but the World Bank, most of the regional development banks and virtually all bilateral agencies have avoided systematic project appraisal precisely because it raises embarrassing questions of value. The viability of projects. There are also some cautionary notes here. One of the ones particularly in this region that may be the most dramatic is the movement of governments toward growth corridors or other very large geographically based investments combining infrastructure with changes in institutions in the hope of crowding in private investment. If you fail to consider the impact of infrastructure on private investment and the externalities that are generated by the infrastructure think the US transcontinental railway if you like for a really big picture example of this. Then you seriously underestimate or at least run the risk of seriously underestimating the benefits of the project. But if you give too much credit for these kinds of externalities then you sort of devolve into wishful thinking and into the realm of politics and into precisely the set of politically and prestige motivated investments that are taking place in some of the countries today. How you deal with this is an extremely complex problem. It's certainly not an excuse for not doing the basics. For not looking at the without project scenarios for not trying to get some idea of what the external benefits might be. But as Jagdish Bhagwati once said to me back in the days when I was working on these kinds of things and he was a bit of a skeptic he said John have you ever seen an externality walking down the street? There are a large number of people who are skeptical of the impact of externalities. And I would argue here that another depressing thought is that we as economists have simply failed to meet our social responsibility in terms of developing techniques to deal with. I'm always very impressed when I come to a conference like this. Everyone is sort of on top of randomized trials, differences and differences, privacy score matching. We're enthusiasts for the things that are fashionable today. But this is a huge public policy question. The commitment in terms of the volume of investment of these growth corridors is likely to completely dwarf any of the social interventions that we've been talking about today. Yet we have nothing in our toolbox to offer policy makers in order to help them. So for any of you who are interested in these kinds of things please get excited about it. It's something that we really owe as a profession to society. The last point is budgeting recurrent costs. We know that timely attention to maintenance reduces the expenditure needed sustained infrastructure. Hilda Johnson, the former Norwegian Minister of Development, who was born and grew up in Kenya, used to be fond of pointing out that in her lifetime, and she was a very young woman at the time that she was the minister, the Mombasa Road had been rebuilt seven times. And the reason why it had been rebuilt seven times was because there was no maintenance that was done to the road. The recurrent costs of maintenance are almost systematically underfunded. Half the countries that the bank sampled for its 2009 report weren't devoting adequate resources to maintain the infrastructure stock. The road stock in fact, as I said, is actually shrinking as a result of insufficient maintenance. And part of the problem here is that capital spending is oftentimes delinked from ONM expenditures. So the capital budgets on sort of one part of the house in the Ministry of Finance, the ONM budgets on another part of the house and the two don't seem to talk to each other. Botswana is perhaps the poster child here for worrying about operations and maintenance. It started with a rule of thumb. It said let's put 18% of the capital cost of a project into the operation of any asset in terms of operations and maintenance. And in fact, when they found out that this was a little bit too small, and it's something that's quite unusual, they actually reduced the size of the public investment program so that it could move the funding to the maintenance of the existing stock. I don't know actually of any other country in Africa that has tried to do that. There are some questions here. The first one is that the pro-investment bias is partly due, again, I'm afraid to say, to my colleagues in the aid industry. We do have a bias toward new construction. There's no question. I hope the Finnish ambassador isn't here because we know that any development minister or any ambassador is delighted to be seen cutting a ribbon to a new school, a new road, a new bridge. It ain't so pretty to be seen with a bunch of guys for the steamroller repairing potholes. So new construction is what we like. It's how we count development impact. The other problem, of course, comes from our colleagues in the IMF. Any device that you use to lock into expenditure, road funds, other commitment devices such as these, often run afoul of the IMF's theology about needing integrated budgets. And there's a good solid public economics rationale for that. We should have marginal equivalencies across all categories of public expenditure. But if we think about the politics of this and the institution, the economics of this, road funds and other kinds of maintenance funds, the 18% rule are commitment devices, just like cost benefit analysis. So these are things that might well be used to lock in a reasonable allocation to the maintenance of existing and future infrastructure assets. Finally, investment effort and investment outcomes. This is, of course, I think if I'm not wrong, due originally to, of all people, Larry Summers, who wrote a rather interesting paper on the relationship between investment outcomes and investment effort back in the 1980s. But in the context of Africa, it really comes down to the question of what goes on in the construction sector. The ability of the construction sector to respond to an increase in demand in some way sets the ability of the economy to respond in terms of investment outcomes to investment effort as construction prices go up. You get less physical output for the same amount of financial commitment to infrastructure construction. That suggests that maybe we need a set of government interventions or government policies to shift out and flatten out the supply curve for construction. A topic that Paul Collier, with his talent for titles, has called investing to invest. So let me talk just a minute about investing to invest. In some countries, and that's Uganda on the right hand side, I tried to pick an east African country, large increases in investment, particularly in the infrastructure sector, have resulted in very, very rapid increases in construction prices. This is not uniformly true. Tanzania, next door, has a relatively flat supply curve of construction. The reason for that is because virtually all large scale construction services are imported. It's not uniformly true, but in those countries where one starts to see rapidly rising construction costs relative to the overall price level, there is an argument for trying to deal with this investing to invest problem. One of the other unfortunate things that turns up in the literature is that the unit cost of construction in all other regions of the world are significantly lower than they are in Sub-Saharan Africa, so we're already in a world in which we're beginning from high costs and moving on about 50% relative to east Asia and we're about 20% higher relative to Latin America. Indeed, this is all beginning to be reflected in cost overruns on multilateral development bank projects, which are increasing in number, they tell us over time. Some areas for action. The first thing is look at the construction sector, which is not something most economists who are doing public economics think a lot about, but in fact it's quite important. What are the regulations and institutions of slow construction projects? In most places I know, especially in urban areas, land rights, zoning, planning permissions, all these kinds of rules, slow construction. In some cases it's residential construction, in other cases it's for projects. Reduce the cost and economize on the use of high cost inputs. A lot of inputs that could be produced or imported at a much lower cost oftentimes are protected and are part of the problem. Construction skills is it's quite remarkable that lack of plumbers and lack of stone masons which are complementary skillsets to much higher skilled construction workers, including architects and engineers, actually can constrain the ability of the construction sector to respond. So you need to kind of think about well what do we do to get people in in these trades? And then finally, and I think probably most interestingly, at least in the context of both Sadek and the East African community, what about temporary location of services providers? Why not open up the migration channels for people at skilled trades and fight that fight rather than fighting the fight over architects and lawyers, which is where it seems to be centered today? Finally, an obvious but not often observed point, it makes a great deal of sense to adjust the pace of public investment to changes in construction costs. Why should you be spending more and more effort to get less and less output? Perhaps slowing down public investment is one of the appropriate responses to this. Some cautionary notes on this. The first one and perhaps the most disturbing one is that there's really a lack of effective competition for civil works contracts. They're in general across Africa small number of bidders for these contracts and very wide price spreads across the bids. It is unfortunately the sector with highest incidence of bribes, both to officials and to other firms, and this is especially true in public works contracts. A sort of contrary example of that is the fact that the World Bank and bilateral donors who are often very sharp about scrutinizing these kinds of contracts for various sorts of illegal transactions are associated with lower unit costs than with government financing of similar public works projects. So one has to keep in mind that this is an industry which will not be easily reformed, but one where I think some attention is due. Let me finally then turn to external financing because at the end of the day, even with these improvements in the performance of the public sector, more efficient infrastructure, better project selection and better project appraisal and budgeting for operations and maintenance, reducing the construction cost problem, we're still going to face a funding gap. The question is where is it going to come from? If you don't get it from domestic savings, there's only one other place to get it from, that's from foreign savings. ODA has happily begun to recover its commitment to infrastructure. There was a very long period of about 30 years where economic infrastructure is a share of all ODA decline continuously. Beginning about the time of the Glen Eagle Summit and the time that people began to discuss moving on to the social, the sustainable development goals, we began to see some renewed interest by the official development community in infrastructure finance. It's about 35% of external financing, but perhaps what's most important about it is that it goes to sectors which the private sector does not find attractive. So it's the primary source of funding for water, sanitation, even transport. China is a cipher. What we think we know about Chinese commitments to infrastructure investment in Africa comes from very skilled detective work by people like Debra Broutigan and others who try to kind of put all the numbers together. But if you boil it down, we don't really have a very good understanding of the strategy the Chinese are taking to employ or their priorities. We do know that their instruments are oftentimes quite different and are associated with lending, so we'll come back to the question of sovereign debt in a bit, but quite clearly they've moved to become a major player in the region. So between, say, 2007-2012, which is about the last time anybody's managed consistent numbers together, they were committing about $5 billion a year, which was bigger than any other donor, bilateral, multilateral, you name it. So they've become huge player in African infrastructure, and as Finna and I were chatting just before the session, these are the guys who are building the bridge, and I suspect it's both Chinese construction and Chinese finance. Private finance, though, has become the largest financing source since about the turn of the century. It's more than 50% of commitments of external financing for infrastructure, but private finance comes with a few downsides. Private investment is basically quite concentrated in terms both of sectors and in terms of countries. In terms of sectors, mobile telephones are quite clearly the top of the list, then thermal power plants, then container terminals. Transport and water supply face highly significant obstacles to attracting private investment. We'll come back to a little bit about this concessioning and why that may be so. And countries other than South Africa and Nigeria haven't effectively been able to attract significant PPI, private provision and infrastructure outside of the telecom sector. About five countries in the region have attracted about 50% of private commitments to infrastructure development, and Nigeria and South Africa alone are about 30%. So it's a very small club of people who are attracted to private investors. It usually comes against a concession contract. That's one of the reasons why sectors are very carefully chosen, and that's just because you can then assign future project revenues to the provider of the service. This further constrains the kinds of activities that private providers are interested in. I'm happy to have a telecom, I'm happy to have a toll road, I'm happy to have a container terminal. I'm not very happy to think about power distribution, road networks, if I can't do a toll road or water supply and sanitation. So the private sector who I think in the naive days of the 1990s, certainly the international development banks thought we're going to be the salvation of infrastructure finance, not just globally but in Africa as well, has its role, but its role is limited and we need some new thinking about how to get engaged in Africa. So that brings me to the final part of this talk, and I've named it very sort of thoughtfully as fatal attraction. The reason why I call it fatal attraction is that some of us live through the period of the heavily indebted poor countries debt initiative, HIPIC. And sovereign debt at that point was a serious problem. We reset the continent and now I look at the right hand graph you see over here and I feel very worried and I'm very disturbed. Before about 2006 in the aftermath of HIPIC, South Africa was really the only sub-Saharan African country that issued a foreign currency denominated sovereign bond, starting in about 2007 and working its way through to now. We've got 13 countries that are engaged in sovereign borrowing, around $15 billion in international sovereign bonds. Much of this, at least the legislatures have been told, is for the purpose of infrastructure finance. In a couple of countries, Ghana and Senegal and Zambia, so a trio of countries I guess, this is now about 40% of outstanding public debt. So we seem to be moving very quickly into a world in which governments and their legislatures frustrated with the growing need for infrastructure, unable to tap the official development community, unable to tap the private sector, find themselves in the world in which sovereign borrowing becomes the most attractive option. But sovereign borrowing comes with a few significant downsides, the biggest of them being high costs and short maturities. And it raises the ugly question of debt sustainability, which as we know depends on how fast you grill and the management of savings and investment rates. And to come back to HIPIC this in a sense turns then on how to get it. It turns on tax policy. Because some of this when we first started out thinking about HIPIC and worrying about HIPIC were a bit confused. Because we said, look, this just doesn't add up. There are all these micro projects that we've done. Each intervention has been assessed as being successful. It's generated a social surplus. And yet these countries are all heavily indebted and can't pay. Was it the projects that were wrong? No. It was tax policy. It was the inability to appropriate the social benefits that the private sector had appropriated by more efficient tax policies. So it comes back again, how to spend it depends on how to get it. The point is one doesn't go willy nilly into sovereign borrowing without thinking about the implications from the point of view of debt sustainability. And I'm afraid that not every one of the countries certainly probably not gone to Senegal and Zambia have devoted quite as much time to that as they might have. So how do we avoid the fatal attraction? Well, the first one is the donors have to step up. The aid industry really needs to change its game here. The first thing, which is a very simple thing to do, and I'm surprised, it comes up about every three years and nothing gets done about it, which is what surprises me, is that my alma mater is constantly complaining about having excess headroom in its non-concessional borrowing, IBRD, while at the same time not allowing credit worthy countries to borrow IBRD for certain kinds of specific activities. The terms are hugely favorable, 1% over liborins to the 5% to 7%, much longer grace and repayment periods. And they could use blend financing, they could combine concessional with non-concessional financing in these kinds of things. In addition, they bring some virtues, they bring project management experience and they bring organizational skill. And that in and of itself is useful as a complement to the financing side of the package. But the bank has never shown real energy in trying to move in that direction, and I'm not quite sure if I can understand why, except the internal dynamics of bureaucracies. The finance vice presidency has as its primary objective in life to maintain the solid AAA credit rating of the institution, the development side of the organization wants to use the money for developmental purposes. Second thing that could certainly be done is to subsidize political risk insurance. I mean, we do have our own, the World Bank political risk arm, Miga. But it's quite interesting that today the excess profits that IFC makes are used as a subsidy to our concessional loan window, they're not used to subsidize political risk insurance. There's no reason why you couldn't let IFC run an internal cross subsidy between its profits and subsidizing political risks. And then finally, and this may be the greatest challenge of all, at some point the Chinese, the Koreans, and the rest of the Dave Dahlman community have got to get together. And maybe the way in which one can get them together is through the IFC, using the IFC as an instrument to get coordination in terms of infrastructure investment, because as we've discovered, certainly your whole set of work on native effectiveness, multiple and conflicting interlocket tours are never put in place. This is something where I think, again, the donors can step forward. Another couple of things, which I think are actually quite interesting. The first one is the tap-domestic markets, and Kenya, I think, has had the most success of this. They've actually generated over a billion dollars worth of local currency infrastructure bonds since about 2009, 2010. One trick that they've used is to allow banks to use these for collateral bank loads, and then calculate reserves. A fairly simple institution will change. Not being a specialist in bank supervision and regulation, there may be some downs there to that, but honestly just thinking about it in terms of first-person, if I could think of one. How do you make African infrastructure to an institutional investor something desirable when you're out there in the international bidding contest competing with East Asia? The first problem is that most countries, particularly in Africa, and most low-income countries in general, have tended to use public-private partnerships or private provision of infrastructure for project finance. Most of these kinds of projects are simply too small for large, long-term institutional investors. They don't have the staff to evaluate the quality of the projects, so they can't put a portfolio of them together. As a result, they're really only comfortable holding ideally guaranteed debt instruments backed by infrastructure assets. So what you need to do is think about how you produce an asset-backed financial instrument, which is attractive, and shares some of the risk characteristics that are desirable to institutional investors. Well, here's another job for the IFIs. The development banks can actually begin to take up this task. They can facilitate the development of larger-scale projects or larger-scale financial instruments, and they can also help to mitigate risk. But it's quite interesting. This is not how they behave today. IFC, which is the private sector arm of the World Bank, basically sits and waits for private sector proposals to be generated. Then it kind of does its due diligence and then decides whether it's going to invest or not. That's not really a developmental role. That's the role that any investment bank would take, and in fact, there are probably some more aggressive investment banks around the world who would be more active than IFC. But this has been the traditional mode of behavior in terms of financing. To change that behavior, you actually have to convert a two-way partnership to a four-way partnership. The two-way partnerships today are essentially a public agency, think the power authority, and a private developer and operator. The four-way partnership is the public agency, the developer operator, a development bank, and a set of institutional investors. This means that the development banks, or at least their private sector arms, and each one of them does have a private sector-oriented subsidiary, becomes the project initiator and the project coordinator. They can create then pools of infrastructure assets that are the basis for asset-backed securities, and then they can sell those on to investors. It's a way of reducing risk. It's a way of bringing an implicit guarantee that the multilaterals are reducing political risk and payment risk, and it's a way of solving the problem of matching the size of the investment to the preferences of the institutional investment community. I hope I've convinced you there's an infrastructure problem. I wish the lights would go out right now. That would be helpful, but there is, and it's growing, and it's not going to go away. Most of the finance, most of the heavy lifting to do this will come from domestic tax effort, but I don't think money is going to solve the problem, and that's where we get to the public economics of increasing the efficiency of the services, improving the quality of the public spending, aligning investment effort to outcomes, and provisioning for the maintenance of the existing stock of infrastructure. ODA, PPI, and China has filled part, but it will never fill all of this financing gap, and it's certainly led to a growing number of countries who have issued sovereign debt, and I suspect in the future a growing number of countries who will issue sovereign debt. To avoid my nightmare of a second hyppig, I think at least three things are needed. Back to the public economics, sound domestic fiscal policy, and recovering the economic benefits that accrue to private agents, some creative thinking by the IFIs, which will bring us to a new set of instruments that can be used to finance infrastructure projects by tapping the capital of the non-concessional portions of these institutions, and finally by making African infrastructure more attractive to institutional investors, and that is my sales pitch for infrastructure for today. Thanks very much.