International Development Committee - Minutes of EvidenceHC 334

Back to Report

Oral Evidence

Taken before the International Development Committee

on Tuesday 2 July 2013

Members present:

Sir Malcolm Bruce (Chair)

Hugh Bayley

Fiona Bruce

Jeremy Lefroy


Examination of Witnesses

Witnesses: Owen Barder, Senior Fellow and Director for Europe, Centre for Global Development, Andrew Rogerson, Senior Research Associate, Overseas Development Institute, and Peter Young, Director, Adam Smith International, gave evidence.

Q41 Chair: Good morning, gentlemen. Thank you very much for coming in to help us with our inquiry on the future of development co-operation. Could you introduce yourselves so that we have that on the record?

Owen Barder: I am Owen Barder from the Centre for Global Development.

Andrew Rogerson: I am Andrew Rogerson from the Overseas Development Institute.

Peter Young: I am Peter Young from Adam Smith International.

Q42 Chair: Thank you again for coming in. This is the second session we have had on this, and we are particularly looking at different financial models. As you know, the UK this year has announced that it will reach the 0.7% that we have been working towards for the last few years. I think everyone on the Committee wholly supports and welcomes that, but I think we also recognise that the background development situation is changing pretty rapidly, so we are looking forward to the next 10 or 15 years with regard to whether or not the model needs to change as the world changes. First of all, would you briefly like to share with us your thoughts on how you think it has changed and will change over the next, as I say, 10 to 15 years? What are the key factors you can see emerging now that will perhaps justify either a different approach or a different mix of approaches from where we are now?

Andrew Rogerson: May I start? Thank you very much for having us. You have heard quite a lot about how the world panorama is changing, particularly the location of poverty and the nature of global challenges. I do not think it is really that necessary to cover the whole landscape, but two particular points are worth drawing your attention to. Poverty will almost certainly be concentrated more in fragile states. They have particular needs, and the track record of development in those areas has been a tough one, and a not totally successful one. That is across the spectrum of instruments.

Secondly, we are in a world where we have to produce global public goods, including fighting climate change, and that is not just a matter of money but also a matter of smart international collective action. I will stop there, with the mention that for both of those things you will have to be more, rather than less, multilateral. That panorama has implications in terms of the nature of international development action, simply because you need to have scale and reach, you need to have the depth of expertise, you need to have convening power, and you need to have objectivity and neutrality, none of which are in the purview of any single state, and of course some states are better at it than others. I will stop there; we can talk about instruments later.

Chair: We will explore those in more detail.

Owen Barder: I agree with that, obviously, completely. I would add a couple of observations. In the changes we see, with a large number of new actors coming into development finance and the changing shape of development financial flows, and not only new donors but also private investment and remittances coming in, one question is what should happen to the British aid programme in the face of that. There is a tendency to say, "We should be doing more of what everybody else is doing. Everyone else is making more private loans, either hard loans or concessional loans; shouldn’t we be doing some of that? People are getting more engaged in infrastructure and private-sector growth; shouldn’t we be doing more of that?" I would argue that the proper response is to think about what we should be doing more of, in the light of what everyone else is doing. We should probably be sticking to the things that we already know.

We have learned lessons about how to deliver aid effectively over the last 50 years. That means we should continue to do that, and not follow the new donors into less effective ways of giving aid. The fact that there are more people giving hard and concessional loans should focus us on the areas where aid can make most difference in grant form. I think we should resist the tendency to be pulled into doing more of what everybody else is doing, because if aid does not focus on the very poorest, most marginalised people, who will? That is money that will reach people that those other sources of finance probably will not reach. I think we should resist the temptation to be like children playing football, where everybody follows the ball, and be more like adults playing football, where we hold our position on the pitch and do the things that each of us is best at.

If I may add to Andrew’s point about multilateralism, he is absolutely right that in the world we will see in 30 years’ time, there is a stronger case for multilateralism, but there is also a stronger case for multilateralism right now. It is not just that because the world is changing we should be doing more multilaterally, but there are strong biases towards bilateralism that I think are unhelpful, and we should be, in any case, looking to spend more money through the multilateral system. We can talk about that if the Committee wants.

Peter Young: I think the UK aid resources are increasing to the 0.7%, but still it is a relatively small drop in the bucket compared with the overall needs, and we should be trying to leverage those resources to maximum effect, trying to achieve the maximum amount of transformational change. I agree that the focus on the fragile states will continue, and the more difficult areas in which to operate. That means that having high-quality, fast, effective capability in our aid programme is particularly important. I would suggest that by and large the track record shows that multilateral institutions have not been able to demonstrate that-rather the opposite. While we may well need to put more money through multilateral institutions in order to address climate change and other major factors of that nature, and have them operate as development banks, in terms of fast, effective assistance, they are not very good.

Lastly, we need to do what we can to leverage the private sector, particularly in the poorest countries. I thought it was very encouraging that Bob Geldof’s latest initiative was to set up some kind of private equity fund. That kind of imaginative thinking is something we should be looking more into.

Q43 Chair: Thank you. I have a couple of supplementaries. You say that poverty, Andrew Rogerson, will be concentrated increasingly in the fragile states, yet what we are being told is that hopefully-because we are concentrating on them-there will be fewer fragile states, and in population terms they will not account for most of the poorest people. We are told, for example, if you talk about the bottom billion, that about 850 million of them will be living in middleincome states that are not designated as fragile. That raises the question: if we are going to concentrate only on grants, is it really sensible that one of the biggest donors in the world should simply say, "We are concentrating most of our aid on 15 to 20 states, and for the rest of the world we just hand it over to the World Bank, or whoever it is, and let them get on with it, and we have no further contribution to make"? Is that really the right ambition for UK development policy?

Andrew Rogerson: First, if I may offer a clarification, about 80% of the world’s poor currently live in-you can play around with acronyms-CHIPIN: China, India, Pakistan, Indonesia and Nigeria. The latter three at various stages were classed as fragile states, and China and India are reducing poverty quite dramatically, although admittedly at different speeds and from different starting points. It is not either/or. There will be fragile middleincome countries: think of Nigeria, think of Pakistan and others that are quite large and populous. There will also be some countries that are not middleincome that may become middleincome over time.

I think increasingly the issue of at what point you cross a statistical line called "middle income" will become irrelevant. The basic conditions-"Is the country well governed? Is it capable of making use of external resources?"-are the real driver.

Q44 Chair: The others can chip in. We are just ending the programme in India, really for the political reason that it is embarrassing for both the UK and India for the UK to be handing grant aid out to India. Yet India does have 400 million very poor people, and the argument is: should we not be working in a more co-operative way with India, using different instruments, to help them tackle poverty, not least if we are to achieve the elimination of absolute poverty and the post2015 Development Goals? It has to happen in countries like that. That is really what I am teasing out: can we just stick with the model we have now, or will we not potentially shut ourselves out of being players in significant parts of the world where we also have legitimate, strong, historical and indeed contemporary interests?

Andrew Rogerson: My take on that is that countries such as India have not yet reached a point where their fiscal capacity-in particular, their ability to tax their citizens-can automatically address all the abject poverty that they face.

Q45 Chair: So you would raise the bar on what you define as middleincome?

Andrew Rogerson: I offer you a hypothesis: should we use public money to assist the poorest in the United States? At some point you will have to decide that, for a number of reasons, they are fiscally and economically able, but perhaps socially these problems are harder to tackle, and then you have to ask, "What does the UK add to that mix, and is it capable of shifting those internal dynamics?"

Q46 Hugh Bayley: On Andrew’s particular point, when we did our study a few years ago in India, which concluded that one should not cut aid in a peremptory way but set a target date for ending the gift relationship of aid, we looked at taxation rates in India and found that, broadly speaking, middleclass people were paying a comparable, or at least a reasonable, level of their income in taxation. I also seem to recall that only 1% or 2% of public expenditure in India is contributed by donors, so although it may be the case that donor help is still needed, it is at the margin.

My question is this: there has been for decades this global ambition that rich countries should provide 0.7% of their GNI to aid. Would it make sense, in the post2015 Millennium Development Goals, to have a presumption that for states above a certain level-let us say at middleincome plus 50% levels-one would expect a certain proportion of the cost of public services to be met through taxation? Then you would have a benchmark of a good donor and a benchmark of a good co-operation partner.

Owen Barder: I think there has been some unhelpful conversation about the idea that India’s problem is essentially a distributional problem-that now it is a middleincome country and a rising economy, if they only sorted out their tax system properly, they would be able to deal with this. Martin Ravallion, then at the World Bank, did a very interesting study of what the tax rates would need to be in India in order to transfer money from the middle classes and the 61 billionaires that India has. He found that you would need average tax rates of over 100% to solve the poverty problem in India.

The average income per capita in India is still $10 per day. That is well above the $1.25 absolute poverty line, but it is still less than one third of the poverty line in the United States. You simply could not redistribute all the money from the wealthiest people in India and solve the poverty problem there. The fact is that the British aid programme in India, until it finishes in 2015, will be providing nutrition to 3.5 million kids. It will be sending 700,000 children to primary school; it will be sending another 700,000 children to secondary school. That will not make a huge difference to India’s growth prospects or to the performance of its private sector, or to its overall public spending numbers, but it is a very serious issue for the 3.5 million children who will get extra nutrition, who would otherwise be stunted for the rest of their lives.

Q47 Hugh Bayley: Could I pick at that a little bit? It would not make sense in the UK for us to say, "In order to finance Government programme x, we intend to increase taxation rates for high earners," because that is not how you would do it. When you were in Government, you did not say, "We have a particular regional development agency, and it will all be funded by higher levels of income tax on the top 25% of earners." You would look at a whole range of instruments for funding regional development, or whatever the project might be. I think the example you use-"Surely the middle class can pay 110% of their income in tax"-really is not a very helpful example.

Owen Barder: My point is you cannot do that.

Q48 Hugh Bayley: No, but do you understand what I am saying? Simply to say the only way the Government of India can raise money is by having everhigher rates of income tax on the better off is not a sensible way to manage public finances. The point I am making is a rather wider point about whether India is in a position for redistributive personal taxation to pay for one third, or two thirds, or 90% of its development spending. It is somewhere on that gradient. Shouldn’t we be saying that countries that have got beyond the DRC or Mozambique point should be generating their own resources, and that one should set a gradient? Somebody once gave us figures, didn’t they, which said that your internal resources will allow you to finance your own development when they are-I forget what the level was-$1,000 a year, or something like that, or a couple of thousand dollars per year. Should one not build that in as part of a development paradigm, so that the Indonesias and Indias recognise, as they already do, through that policy-

Owen Barder: Yes, and I was just looking for the reference. Development Initiatives is producing a paper in September on the sources of finance for poverty reduction, and what it shows is that, not just in India and China but across the developing world, Governments are spending much more of their own <?oasys [pc10p0] ?>resources on social services, health, education, and poverty reduction, including things like social safety nets. That is where by far the biggest increase in development finance is coming from: it is coming from developing countries themselves. That should be encouraged, supported and helped.

Q49 Hugh Bayley: Should that be codified into the post2015 framework in some way-some model of a good middleincome aid recipient?

Owen Barder: I do not think there is anybody, anywhere in the development system, who does not want to see, in the end, every country in the world able to finance its own social services, its own economic infrastructure and private-sector growth, and all those things. I think we have been somewhat seduced by having set a very, very low poverty line of $1.25 a day, and saying, "If people are getting across that poverty line, this country is now wealthy enough," or a very low status for middleincome status, which is $1,000 per person per year. That is not a good signal of whether a country is yet in a position to provide all its basic services for all its people and deal with pockets of poverty, and many of those countries will still need external financial support. Even if they are making great progress and spending more and more of their own resources, that is not a reason for us to walk away from them and say, "You do not need us anymore," because many of them still do.

Chair: That is a very interesting philosophy, but I think we need to get more of the detail.

Q50 Jeremy Lefroy: Good morning. I would like to ask a couple of questions about the type of financial instruments that may be required to meet the UK’s development objectives. It seems to me that in development you have two particular kinds of finance. One is obviously the more technical assistance, where you are helping the country to build up its ability to deal with a particular problem-to set up a social transfer programme, governance in the private sector, or whatever. Then you have actual funding for those programmes, and sometimes I think there is not enough analysis or discussion of whether those two things should always go together. It is often the case that you get both the technical assistance and the funding to go with it.

With the private sector increasingly being seen as incredibly important to development, in the sense that it generates 90% of new jobs in the developing world, so we are told, and therefore generates the tax revenues and helps to lift people out of poverty, and given that DFID is therefore concentrating more of its funding on the private sector, which ought to be able to return at least some of the funding it is given-as we have seen in some of the projects we have visited, whether in Pakistan, through bank guarantees, or in Sudan and elsewhere through the Africa Enterprise Challenge fund-we now have a kind of mix of funding in DFID where it is all grant but some of it effectively should be returnable capital. It is going to <?oasys [pc10p0] ?>the private sector, and they can easily afford to return it, but there is no mechanism for doing that. How should DFID respond to this? Let me ask that question first, and then come with some followups.

Peter Young: I think that the situation differs in different countries and regions. If you take somewhere like Nigeria, I do not think the problem is a shortage of money. There is lots of money sloshing around Nigeria-quite a lot of it is sloshing into this country too, and elsewhere of course-but for profitable ventures, you can find finance. There are more serious problems that prevent people from accessing that finance. For example, only about 3% of the population have titles for their land and property, so they cannot get bank loans because they do not have collateral. In that case it would be much more effective for DFID, as in fact it is doing, at least in some states, to try to tackle that issue and a range of other issues that prevent people other than the elite accessing finance.

As we suggested in the evidence paper we put in, one could try to encourage more direct UK private-sector investment into these countries by providing tax incentives for investment, either equity or debt investment, along the lines of the seed enterprise investment scheme, or the VCTs that we have in this country. That may well throw up all sorts of different initiatives that would, again, be more tailored to the particular circumstances of the countries concerned. A "one size fits all" setting up of some additional development bank, which would take a very long time to do and build up its own bureaucracy and whatever, would seem to me not to be the right way to go.

Q51 Jeremy Lefroy: Why do you think it will take a long time to do, and why would it necessarily build up a lot of bureaucracy? Would it necessarily have to be something separate? Could it not be from within the Department, as it is at the moment? After all, they are doing some of those things; they are just not acknowledging it, and not putting them on one side as a department of capital that is returnable, as opposed to that which is made by grant.

Peter Young: I guess there is nothing wrong with starting more of those types of programmes in some of the countries, where perhaps there is a particular need and there is no other way of fixing the problem. Usually these programmes, like the Africa Enterprise Challenge Fund, are run by a private-sector organisation-I think it is KPMG in that case-who have the expertise to implement the programme. They are not setting up some new huge organisation or department staffed by civil servants, who are, because it is public money, extremely nervous about everything that is spent, and everything basically goes towards a very riskaverse model, because there will be a row every time a load of money gets lost and so forth, and your Committee will investigate.

I am not sure it is the best means of providing such finance, but perhaps a more experimental approach might work. There already are a number of these development banks, none of which are noted for their particular speed and flexibility, it must be said, although some are better than others.

<?oasys [pc10p0] ?>Andrew Rogerson: May I add to that? I tend to share this view, but for another reason also: one of the success stories of DFID that is recognised is combining a very large portion of the UK’s external aid effort under one roof, and then being present incountry with that same sort of unified voice. We should be careful. If the answer to a problem is concessional loans or guarantees, then DFID already has the powers, as I understand it, to do those. They may need more capacity and more skills-we can talk about that-but it should be very careful not to throw away that advantage by creating something apart from DFID that then adds to the general fragmentation incountry and does not necessarily dovetail with how taxpayers’ money is being used in other forms.

Concessional loans potentially have their place, and the world will probably need more of them, including through IDA, the EIB and other people you will be talking to, but not necessarily as a freestanding, brandnew bank that one would magic up, with all the transaction costs that might involve.

Owen Barder: The 2002 Act does give DFID the powers it needs to do this kind of thing. The world absolutely does not need a new development finance institution.

Q52 Chair: That is not the view of the World Bank.

Owen Barder: It may not be the view of the World Bank. DFID is in a position to make sensible choices about the right mix of grant and concessional loan and hard project finance.

Q53 Jeremy Lefroy: Could I come back on that? You may not have the figures-we do not-but how much in the way of concessional loans has DFID made, say, in the last 10 years?

Owen Barder: Effectively none.

Andrew Rogerson: I think none.

Q54 Jeremy Lefroy: So it has the power, but it has not used it?

Owen Barder: It has not used it, but it could and should choose to use it. It is perfectly reasonable to say there are many circumstances in which that is something DFID should be doing.

Q55 Jeremy Lefroy: Could anyone say why they think DFID has not used this power, even though it has it?

Peter Young: I think the DFID funding can be often used to greater effect and create greater leverage if it is not simply provided in the form of a concessional loan. We do a lot of work in the PPP area, and everyone is enthusiastic about attracting private-sector funds into infrastructure, but the overwhelming problem is not the lack of finance, concessional or otherwise. It is the lack of bankable PPP projects, because not enough work has gone into creating those projects and addressing the underlying problems. That is a fairly expensive process, but not nearly as expensive as financing infrastructure itself, so I would say it would be better to put your money into creating <?oasys [pc10p0] ?>a flow of bankable projects than into attempting to bank the very few projects that are out there.

Q56 Jeremy Lefroy: That is exactly what PIDG-the Private Infrastructure Development Group-is supposed to do. It is precisely the point I made: if you do have or develop bankable projects, at some point they need to be paid for by the people who develop them on a commercial scale, which is what PIDG does, and therefore this returnable money is, in effect, what we are talking about. There are some things that DFID is doing at the moment along those lines, but it is not recognised as being returnable. It gets lost in the ether. I do not think anybody is particularly necessarily saying that the only answer is to formalise this into a development bank, but should DFID not do a lot more to identify these and really get to grips with them? At the moment we feel this money is often wasted. It disappears. I think there was the expression used earlier that money is lost. I think it would be a very interesting exercise to go through and say, "DFID has been giving several billion pounds a year over the last 10 years-so probably something in the order of £60 billion to £70 billion. How much of that has had some effect?" We do not know. I am sure even the Secretary of State could not tell us precise figures.

Peter Young: Giving money to what-generally?

Jeremy Lefroy: Yes, generally, whereas if you were doing something through concessional loans, the very fact that it is a 30year loan means you have a track record of what has happened to that project over the 30 years. In 30 years’ time when the loan comes to be repaid and the project does not exist because it is a white elephant, you have some accountability there, which you do not have in terms of having made a grant 30 years ago.

Andrew Rogerson: We should not forget that DFID can and does invest in the likes of IDA, which does essentially loanbased operations over 30 and 40 years, the EIB, from whom you will hear later, and others. That might be one reason, Sir Malcolm, why the World Bank is perhaps happy to have more friends around the block, because they can cofinance. They do not necessarily have to be the designer and the architect of every operation; they could come in and slipstream behind some of the other bigger fish. There is definitely a possibility of doing more on concessional lending. We are using the word "concessional" knowing that this is a very blurry story at the moment, but I use it to mean-as I think one of your other witnesses, Matthew Martin, said-well below 1%, something that cannot create unsustainable debt bubbles in the creditor countries.

Q57 Jeremy Lefroy: One final question, which is a bit more specific: we have read in evidence, I think from ASI’s Development Resources Trust, about development impact bonds, which are very interesting, and certainly need to be explored. Do you think there is a role for DFID to play in that, or should it just leave the market to get on with it?

Peter Young: In the case of the specific instruments that require UK tax incentives, you would need, I <?oasys [pc10p0] ?>think, DFID to lead the process of designing the appropriate intervention or series of interventions, and negotiate with the Treasury. I think it would be appropriate for the cost of those tax incentives to come off the DFID budget, but last time I looked, the cost of tax incentives for the venture capital trusts in the UK was £80 million, which is not an absolutely vast figure. A lot of useful economic activity is created through them.

You could either go down the route of creating a specific vehicle, such as the Development Resources Trust that we suggested, which would be private-sectorrun funds that would collect money from retail investors, have onboard expertise, and invest in particular projects-a sort of venture capital trust model-or you could simply provide a tax incentive for any investment, as we do through the Seed Enterprise Investment Scheme in the UK, where very small investment projects attract a high level of tax rebate. That would possibly lead 100 flowers to bloom and all sorts of new initiatives to arise, some of which then might be suitable for subsequent concessional finance, grant support or whatever, depending on what they were. I think it would be an interesting route to go down.

Owen Barder: If I may add to Peter’s point, the development impact bond proposal, which is modelled on the social impact bond proposal that you are familiar with-used, for example, in Peterborough prison-has private investors, social impact investors, putting in private capital to produce results, but those results are typically results that do not yield a financial flow of benefits. In the case of Peterborough prison, it is a reduction in recidivism, which is of benefit to society, the economy and the taxpayer, and in that case the savings come to the Ministry of Justice in lower prison costs, and the Ministry of Justice then remunerates the investor for the money they have put in.

In the case of development impact bonds, where very often a developing country Government would not have a financial benefit that would enable them to remunerate an investor for the results they have produced, if you have an investment that does not itself have a financial revenue stream, you might well look to a combination of the developing country Government and donors to remunerate the investors. There would be a clear role for DFID in creating this market by promising that it will repay the social impact investors for the investment they make in producing certain results in developing countries. DFID would need to figure out the legal and financial mechanisms, the procurement rules and so on, to enable them to do that.

There is a certain amount of work to be done to put them in a position to do that, but it is potentially a very powerful instrument that DFID could be using to pull through private investment into social results in developing countries.

Q58 Jeremy Lefroy: Are there any examples of such bonds in existence?

Owen Barder: There is an example currently being worked out, not yet committed to, to tackle sleeping sickness in Uganda, which would involve private investment to roll out the services, but it needs a donor-and I hope DFID could be one of those donors-to promise to repay the investors if the results are achieved.

Q59 Fiona Bruce: Good morning. We have heard on this Committee often that grant funding decisions are made in a very short time frame. I wonder whether, with reference to leveraging in more private sector investment, we need to look again at the returns from investment. Do we look too short term, and should we be engaging in the exercise of looking more medium and long term at the outcomes and the outputs of investments when we are looking at private sector partnering? How could we do that?

Peter Young: I think that we should, across the board, be looking more at how to measure the outcomes of the investment made by DFID. There is a tendency, when you cannot measure some of the harder-to-measure things, to only put money into the very easy-to-measure programmes. There is not nearly enough done in looking at proper cost-benefit analysis across the whole piece. Of course, if you are talking about some of the specific investments we are making alongside the private sector, it is perhaps in some senses easy to measure some of those, because the private sector is trying to measure the effect of the investment.

However, many of the things DFID would do would be more to create the enabling environment for private-sector investment. For example, we are working on a project at the moment to try to tackle the absence of electricity in Nigeria. There is a lot of private-sector investment coming in-indeed, some billions are lined up behind the current privatisation of the industry-but none of that will occur unless tariffs rise to an appropriate level, which in fact they now have, and unless assorted other technical issues to do with gas supply regulation and so forth are fixed, but the returns on that are absolutely gigantic, because an adequate supply of electricity will enable that country’s economy to grow much, much faster. No one has ever attempted to measure that, because we need to come to a view on what will be the increased economic growth that will result from these changes. It will be significant, but what will it be? No one has tried to measure it.

Andrew Rogerson: I think another part of your question is really the level of ambition that one should have in transforming the environment and the market infrastructure. The problem with being that ambitious is that you are also raising the risks. There are some good examples of donors putting in money for advertising; the publicising of improved wood stoves in Bangladesh, financed by USAID, comes to mind. That did not in itself sell a single wood stove, but then the private sector came in behind it and did what it can do well. It is not necessarily just an issue about the lending and creating financial constructs with the private sector; you are raising an issue about how return and risk are perceived, and whether Ministers and others are willing to make those investments, knowing that a lot of the benefits accrue to private individuals-maybe overseas, maybe in our own companies-but if it does not work, the risks come home to the taxpayer.

Q60 Fiona Bruce: That is the interesting dichotomy, isn’t it? The private sector is used to working in a riskbased environment. The public sector is not, and here we have the merger coming in international aid work in a way that perhaps we have not seen before. Do you think there is the capacity within DFID to engage on that basis?

Owen Barder: If I may say so, from my experience of working in DFID, I had the sense that as an organisation they understood very well the longterm nature of the risks they were taking. If anything, the dislocation working with the private sector was that the private sector, typically for reasons of market pressure, needed returns more quickly than the development sector was able to offer-more certain, and more quickly. If the implication of your question is, "Is DFID good enough at thinking about risk in the long term?" I think they are better at it than some of the private-sector partners with whom they have looked to partner in the past, and that has been an obstacle to successful co-operation.

My sense is that the political leadership-not always The Daily Mail and other commentators-and the civil service in DFID have a good sense of the longterm nature of the investments they are making, and the risks they are taking.

Q61 Fiona Bruce: Do you think we sufficiently measure that, both prospectively and retrospectively? Have we learned enough lessons?

Owen Barder: I completely agree with Peter. Though DFID has made good progress in recent years, we are a long way from properly measuring the impact of DFID programmes and making decisions based on that evidence.

Q62 Fiona Bruce: One of the concerns that this Committee has occasionally expressed is that DFID projects can tend to be too big financially, or their funding too big, for some of the smaller social enterprises who might want to engage-for example, for a group wanting to build a single health centre in Africa. Do you have a comment on how we could lever in more private sector funding and expertise, if DFID rethought how to engage with the smaller organisations?

Andrew Rogerson: I think others know more than I do about some of the specifics, but a number of organisations, not least the World Bank, the IFC and even CDC, have been able to delegate to incountrybased funds or regionalbased funds, so that the unit level of the investment, including in social enterprises now, is affordable, in terms of transaction costs, rather than being organised straight out of London or Washington or wherever it may be, which is prohibitively expensive when addressing the needs you are talking about. We are doing some work in ODI on trying to harvest some of the recent experiments on social enterprise funding by official donors; you will be hearing later from the Shell <?oasys [pc10p0] ?>Foundation. This is still in its infancy, and there are not that many surefire cases that we can put our hands on and say, "Here are small investments, deliberately and aggressively funded."

One final thing on this risk issue: committees such as yours, and other raters of the results for development, should have the patience to accept that in such portfolios there will be a lot of small or quiet failures, and some spectacular successes. That is the nature of venture capital. If one approaches it with a "Gotcha!" approach, if I may say so, you will probably discourage those few people who are quite keen to make more and more of these things-accepting that some will not work.

Q63 Fiona Bruce: A final question, unless you have any thoughts on how we can leverage more private sector funding, other than the ones we have discussed: how can DFID can ensure that private sector investment is consistent with its broader policy aims, such as gender equality and environmental sustainability?

Owen Barder: This is at the nub of the question of the respective roles of ODA and DFID funding and private sector investment. It is obviously fantastic news that we have much greater private investment in the developing world, both domestically and externally. The question for us should be, "How can we best use a very small additional resource to help leverage that, but also shape it?" We know, don’t we, that as countries grow, an enormous amount of their growth path depends on how unequal they are as a society, and how unequal that growth is. All growth is good, in some sense, but more equal, better-shared growth is better. It reduces poverty faster and it reaches more marginalised groups faster.

I think it is a proper role for DFID to figure out how it can use its very small resources, in the grand scheme of things, to interact with those other private investment projects and programmes, to ensure that people are not left behind in the growth process. Often that means doing some of the things in the background that the private sector will not reach, such as ensuring that projects and programmes reach women and children, and marginalised communities, in lots of different ways. It seems to me that finding ways of topping up the private investment to catalyse and shape that change is extremely important.

We talked earlier about development impact bonds, and they are one mechanism to do that, because they turn financially uninvestable propositions into financially investable propositions. You can bring the excitement and the capacity of the private sector to projects that they would not otherwise be able to invest in, because there would not be a sufficient financial return for them. It seems to me that is exactly the kind of role DFID should play. What we should avoid is a situation where we are just trying to do what everyone else is doing-where we are just throwing in our money with everybody else’s private sector investment and saying, "Here is our bit, because we think investment is good and we think growth is good."

We should be very strictly focusing on reaching particularly those people and those countries that private sector growth will not be reaching any time soon. There are 400 million or 500 million chronically poor people around the world who will simply not benefit from the creation of jobs and the private investment that is going on across the developing world, because they are too far away in lots of different respects from that economic growth. I hope eventually they will, but doing all that we can in the meantime to focus on them-to connect them to the economy, primarily by giving them health, education, access to water, housing and so on-is extremely important if they are to share in economic progress.

Peter Young: I certainly agree with that. I think that our relatively limited resources need to be concentrated on removing the barriers that prevent women and marginalised groups from participating in the growth that will come from the private sector investment. Let me give some specific examples. I mentioned land in Nigeria; we could enable women to participate in the land registration process, so it is not always done by the husband, even though it is the woman who owns the land, and stop inappropriate use of the tax system to harass small women traders, enabling them to build up their businesses more effectively. There is perhaps a role for even unlocking some concessional funds, which exist in the Nigerian Government, to enable some of those small women entrepreneurs to participate in the system.

All those sorts of interventions, which are not of high cost, will have a significant effect. That is the type of thing we should be doing. It is very careful, targeted, quality work.

Q64 Hugh Bayley: Could I, Owen, go back to what you were saying about development bonds? If you were talking about a development bond, shall we say, in Malawi or the DRC, whose credit rating would determine the cost of raising the finance: Malawi’s, or the DRC’s, or the country that agrees to pay the financing cost? If, to some extent, the low credit rating of the country issuing the bond would contaminate the bond, in the sense that it would require a higher cost, surely it would make more sense for the UK, the World Bank, or some more creditworthy institution to take the risk in its entirety and then provide finance through a different gift mechanism?

Owen Barder: The main risk for which the investors would require compensation is the delivery risk. It is the risk that they put money into a scheme to reduce sleeping sickness, and they do not manage to put in place the logistics network, the management network, to get the right vets to deliver the right injections to the right cattle at the right time. That is independent of whose finance is standing behind this. That is the majority of the risk that investors are taking, and because they are taking that risk, they are also putting their back into ensuring that that risk does not turn bad on them. They are putting their back into ensuring the right logistics are put in place so that the right cattle are vaccinated.

Then there is a risk that whoever has guaranteed to pay them at the end of it, in the event that the results are achieved, is not creditworthy or walks away from it. In that case, if it is a development impact bond financed by donors, it would be the UK Government, for example, whose creditworthiness would determine the value they would put on that risk. I would think in today’s market that would be at or very close to gilts, but by far the majority of the return to investors is the delivery risk.

Q65 Hugh Bayley: My observation, following many visits to subSaharan Africa-not to Asia, but to subSaharan Africa-is that there is a really serious problem with domestic capital accumulation. If you are a multinational company you can obviously finance your new Guinness brewery, or whatever it might be, and there is some effective development assistance, through micro-credit, at a very, very low level, but if you are a garment maker with a sewing machine who wants to employ four other sewers, or to move from being a market trader to opening a shop, or to move from selling paint to making paint, it is almost impossible to raise the capital. What should DFID do better to support the growth of an SME sector in developing countries, and indeed the use of SMEs to deliver development assistance programmes?

Owen Barder: To be honest, I would defer this question to when Chris West is giving evidence.

Andrew Rogerson: I think it might be part and parcel of what we are talking about: using the power of a small enterprise, and particularly enterprises that can deliver a social impact, whether that is by employing the people in the paint shop or running ambulance services, or whatever it might be.

Peter Young: As I was saying earlier, it will vary from country to country, but in many countries there are particular barriers that prevent that capital accumulation: the commercial court system does not work, or if it does work, it is for whoever bribes the judge the most; there is no proper land and property title; the regulatory requirements for particular activities are very heavy or it is difficult to get the appropriate licences; some particular aspect of infrastructure does not function; or there are predatory tax inspectors, or whatever. One should attempt to address those things, and that would have a significant effect on the availability of capital in those countries, as it would then become profitable for people to accumulate and lend capital to each other.

Q66 Hugh Bayley: I once asked a parliamentary question to find out what proportion of DFID’s bilateral spend in countries in Africa was spent on buying goods and services from local suppliers, and it was vanishingly small-a fraction of 1%. Whenever you go to a DFID or a World Bank office, it is an office that is wholly provisioned by Western, or maybe Chinese, paint, computers, air conditioning machines and furniture. Everything is shipped in from 5,000 miles away. Wouldn’t it make sense for DFID to set targets for local procurement, even if the paint it got was so terrible that you had to repaint your office every two years instead of every five years? Surely that in itself is generating economic activity and economic development in the country you are intending to help to develop.

Owen Barder: Absolutely. There is a very impressive Canadian organisation that used to be called the Peace Dividend Trust and is now called Building Markets. It was set up by a former Canadian diplomat who, having been responsible for administering Canadian development assistance, concluded that the only time it was any good was when they were procuring local goods and services, and at least helping provide demand for local businesses. He has set up a business connecting, initially, mainly aid agencies and aid donors-the multilaterals and the bilaterals-with local suppliers of everything from bottled water to cars and air conditioners. It is a fantastically inventive and effective intervention.

I was horrified to see the Permanent Secretary of DFID boasting the other day to the Whitehall and Industry Group that 90% of DFID’s centrally procured competitively tendered procurement went to British firms. If a developing country were making that boast, we would accuse them of corruption and ask them to undergo a process of procurement reform. You are absolutely right; it seems to me to be shocking that we buy so little and we have not worked harder to put in place systems to buy locally.

Andrew Rogerson: I think you also need to consider the scale, going forward. Somebody mentioned earlier that the minimum size of DFID operations is rising. The pressures that come, particularly from local staff, if you have to organise 400 construction sites, with artisans working in different districts of a country, are an enormous logistical headache. Even some extremely capable international NGOs find it difficult to meet the requirements of a £10 million, £15 million or £20 million operation for which they are screened. I know that, having sat on the boards of some of these. You would then have to go the other way, localising your action as against merely your contracts, and that is expensive in terms of administrative cost.

Peter Young: I think you need to be quite careful about this by laying down some rules and regulations. Firstly, a number of the things that one would purchase are probably not made locally, like cars, photocopiers and so forth.

Q67 Hugh Bayley: Maintenance could conceivably be sourced locally.

Peter Young: Oh, yes, sure. Those types of things, certainly: desks in the DFID offices could be sourced locally, though I understand they have fairly strict health and safety requirements for those sorts of things, and various inspections and whatever. You would not want, however, to lay down some list of rules and regulations that meant that, because it took another six months to procure something, a programme started six months late. General encouragement is a good thing, but some kind of prescriptive approach would probably get into some other type of labyrinthine bureaucracy, which would just cause a nuisance to everyone.

Q68 Jeremy Lefroy: Just to follow up a little bit on that before coming on to climate projects, Owen, you talked earlier about not wanting to follow the pack, as it were. I seem to remember that, in the mid-1990s and even later, things that are now taken for granted, such as the importance of agriculture, particularly smallscale agriculture, in nutrition and employment, and the importance of the private sector, were completely poohpoohed by the development community, with a few exceptions-in particular, there were honourable exceptions among NGOs-and certainly by the "official" development community. It has taken DFID and one or two others to take the lead, and that involved them moving into the kinds of instruments, such as the AECF and other kinds of returnable finance, that we are talking about.

I would slightly challenge the assumption that looking at instruments such as concessional finance, particularly if it is done in an innovative way, perhaps along the lines that you have been talking about, is simply following the pack. We tend to think that the current is normal, and therefore things like, as I say, the importance of agriculture, the importance of the private sector, have been taken for granted since the year dot. They have not. They have only been taken for granted in the last three to four years. There is a tendency to kid ourselves, and when we go back 15 years, the situation was completely different. I just wondered if you had any comments on that.

Owen Barder: You are absolutely right. These things tend to move in cycles. There was a long period in the 1960s during which infrastructure and economic growth were central, and people shifted away from that. Now people are shifting back to it, and that is excellent. My point about following the pack was less to do with that trend-I actually welcome the focus on economic growth-and more to do with focusing on the special contribution that aid can make that other forms of resource cannot and will not make. It is simply not the case that commercial investors will invest primarily for broader social benefits. Their private investment will deliver broader social benefits, and that is terrific, and there will be many people who get jobs, livelihoods, incomes and access to consumer goods as a result of private investment, and that should be welcomed.

However, there are some people that will not reach, and what the aid system needs to do is keep its eye on the prize of ensuring that everybody shares in that rising prosperity. We should not be sucked into saying, "Now that China and India are doing more infrastructure, we should do more infrastructure." If China-India perhaps is not as good as an example-is investing more in infrastructure, that could be a reason for us to invest less in infrastructure. It is great to see that infrastructure rising, but if they are, for whatever reason, less likely to invest in something else that is important, then we should be making a judgment about the role we can play to complement what other people are doing, while welcoming what other people are doing.

We should not feel that we all have to play the same thing, and British aid has a particularly good story to tell of focusing on improving the lives of the very poorest people. We will continue to need to do that while this economic growth is taking place. There will be several decades when there are very, very poor people who do not benefit from this economic growth, and we should not take our eye off the need to protect them while they remain vulnerable, while this growth is happening. We should not feel the need to be part of the growth part of that process if there is something that we could and should be doing that other sources of finance will not reach. That is the sense in which I feel we should be careful not to be sucked into whatever the latest trend is.

Jeremy Lefroy: I fully understand that. We do not have time now to explore that, but I think it is a very important issue.

Chair: The only point I was going to make was that unfortunately the UK Government has already made a decision that in India they will not do that, so the Committee’s reaction is: "If we are not going to do that, what are we going to do?"

Q69 Jeremy Lefroy: I think this is very important. I realise we are already running over time, so let me quickly ask a question about climate mitigation and adaptation. These projects tend to be riskier investments, because they are still at a very innovative stage. How do you think DFID can best invest its resources so it can get the private finance needed? Frankly, if the world is to meet the target of $100 billion per year in climate finance, there is absolutely no way it can be done through ODA. It needs to be done largely through private sector, or certainly nonpublic sector, finance.

Andrew Rogerson: I absolutely agree, and there is a whole range of financial tools. First of all, yes, on climate change particularly, transfers of technology will not happen without private sector engagement. There has to be a very careful study of what kind of terms can be offered, to the extent that there are sweeteners to these terms, such as the implicit subsidy involved in the World Bank using its main hard market terms window to finance climate change measures. They have to be looked at, as is always the case with public money going to support largescale private sector, but one has to be incredibly creative or we will not get there. The $100 billion Copenhagen target of additional flows covers the gamut from pure grants, which I think will be a minority, right through to market terms, with perhaps some support in accessing the market in some cases.

Owen Barder: If I may add a footnote to that, this is a good example of the complementarity issue. Almost all the private-sector money going into climate finance is going into mitigation efforts. There is almost no money going into adaptation, and yet for the poorest and most vulnerable countries, adaptation is the most urgent and pressing need. That seems to me a very good example of where concessional resources from the UK could play a role that the private sector does not at the moment seem willing to play.

Andrew Rogerson: Also, this adaptation/mitigation split is important, because if you chose to use public money massively to finance mitigation, that is to say grant or grantequivalent resources, there will be a huge sucking sound as UK money goes back to China, India, Brazil and possibly Russia, and away from the most fragile country contexts-fragile in every sense of the word.

Peter Young: Again, we need to do what we can to help the countries get the ground rules right, so that private sector investment can flow into these areas. Examples are solar power, which is more viable in Africa than it is in the North of England, and working with countries to make sure that there is an effective feedin tariff, and that the regulatory arrangements allow people to build these facilities and be remunerated from them. That sort of assistance can then create a private market that is financeable.

Owen Barder: Of course the best way to do this is to set a meaningful carbon price and then let the private sector figure out the investment needed to live within it.

Q70 Jeremy Lefroy: The UK is doing that. Do we know how many other countries have done that?

Owen Barder: We have the EU emissions trading scheme, which of course is not working very well and could use some serious overhaul. In the end, if you want the private sector to tackle this problem, there is no better way than setting the carbon price properly.

Q71 Jeremy Lefroy: Are there examples of developing countries that have done that, or middleincome countries setting carbon prices?

Owen Barder: China has an emissions trading scheme between its regions, designed to try to set a carbon price.

Q72 Hugh Bayley: I warmly and strongly support what you wrote on your blog about the relative merits of multilaterals visà-vis bilateral aid programmes, but the question I want to ask is: does the UK currently have the right balance between bilateral and multilateral aid, and are the BAR and the MAR effective tools to assess the relative strength of one multilateral versus another, or one country bilateral programme versus another, and to compare whether aid is more effectively spent through multilaterals in general, to set that split between multilateral and bilateral spending ceilings?

Owen Barder: To take your question about the BAR and the MAR first, I think those are tremendous innovations that DFID and the Government should be congratulated on. As Peter was saying, trying to make explicit what the impact is of different projects and programmes and then comparing them and making choices between them based on that seems to me absolutely the right thing to do. I hope this Committee will push DFID to continue to take that approach. It is quite similar to the way that foundations allocate resources. Do not start with how much you want to give each country; start by looking at the list of available projects and programmes and investments, and choose the best ones among them, and then the country allocations are whatever they are. I think that is a very reasonable process to take in the case of the bilateral aid review. Next time round, I would like to see a much more open and transparent process. I think all those offers that country programmes make should be publicly available for scrutiny and for people to comment on and crowdsource a bit before Ministers make decisions about what they want to invest in, so we can all see that that is properly scrutinised process.

Q73 Hugh Bayley: That presupposes that the BAR will be repeated alongside that.

Owen Barder: That is what I am saying. I hope it will be repeated and that, if it is, it is done in an even more open and scrutinisable way. That would be terrific. The multilateral aid review is a more difficult process. It is a big step forward from MOPAN, the previous way for the network to look at the effectiveness of multilaterals, but I think we are still, in that space, hampered by poor data and poor knowledge. It is extraordinary that we cannot answer more precisely the question of how much it typically costs the World Bank or the African Development Bank to build a mile of road, train 1,000 teachers, or buy 1,000 textbooks. It seems to me these are knowable facts. Of course there will be variations between them, just as there are between surgeons’ death rates, but we go ahead and publish those data anyway, and we use that to make informed judgments. We should be doing that in the case of multilaterals.

I hope that DFID will, as a major contributor to these organisations, continue to put a lot of pressure-as it has been doing, to some extent-on making those comparisons more transparent. Ultimately, that is what we mean by effectiveness. The MAR judges organisations on a whole number of things to do with their strategy, their overhead costs, their alignment with our values and so on, but actually what you want to know is how effective this thing is at delivering the results that we are contributing to.

Q74 Hugh Bayley: Briefly, because of a lack of time, it seems to me that there is a methodological gap, in terms of making a logical, transparent assessment about whether you have a 40:60 split or a 60:40 split in working out what proportion of your funding goes to multilateral or bilateral.

Owen Barder: You have preempted my next sentence, which is that the BAR/MAR process did not provide any rational basis for the split between multilateral and bilateral aid. I personally think that 40% is too little for the multilateral share. I would like to see that increased. I think the bilateral share should certainly be capped in cash terms. It should have been falling as a share of the aid programme as the aid programme rose. What I think is most demanding of scrutiny is the multibuy part of the aid programme. That is the part of the bilateral programme that is spent through multilateral organisations. That seems to me to be prima facie evidence that we are trying to spend too much bilaterally. What we should be doing is shifting most of that money into the core funding of multilateral organisations, with them using their governance and accountability systems for that money; we should not be scoring it as bilateral aid but putting it through the multilateral system. That is about 20% of our aid, and that seems to me far, far too much. It is well above the OECD average, and I strongly doubt whether that is an effective use of taxpayers’ money. I think we should be shifting that into the core funding of multilaterals.

Peter Young: The problem is that the governance and accountability systems of a number of these multilaterals are not sufficient for us to rely on to pump everlarger sums of money into them. The BAR and the MAR, while it is great that we are now looking at the effectiveness of organisations that we are providing these funds to, were in fact somewhat overly optimistic in a number of cases, perhaps because of the way this was structured and the questions asked. If you attempt to look into the details of the performance of some of these multilaterals, the first thing you find is that you cannot actually find the information on any coherent basis.

We do not have a bloody clue what the performance is, broadly speaking, of many UN organisations, many EU programmes and so forth, because there is not the information available. As we put in our note, in the case of the UN, for years and years people have been saying, "They do not do baseline studies so you cannot see where the improvement is and what the starting point was; they do not measure outcomes; they only talk about activities." You do not really have much of an idea what on earth is going on. I think a lot of the reason why we are pumping this extra money through the multilaterals is that it is administratively easier to do so, and DFID does not have sufficient administrative budget to handle it itself.

It does not make sense that you provide funds to an organisation that is administratively more expensive because you have some kind of cap. Also, if you are trying to spend your budget and you write a cheque to organisation x, you have spent the money, even if organisation x does not spend it for another five years, so there are incentives in the system that encourage the spending of money-particularly this extra 20% Owen is talking about-through these multilaterals. I think those incentives should be changed so that the money is counted as spent only when it is spent on the ground, not when it is transferred into the bank account of some multilateral. The administration costs of these multilaterals should be taken into account in the overall assessment of DFID’s administration spending, so the incentives operate properly.

Andrew Rogerson: The only part of this last statement that I agree with at all is this last point about more scrutiny on the 20%. That was a point that Owen brought out on multi-bi.. I declare an interest: I was on the external scrutiny panel of the bilateral aid review while I was writing the annual reports on multilaterals for the OECD’s Development Assistance Committee. I found that there was a very poor fit, in terms of the excellent work done at country level, between the work done at the level of the DFID bilateral programme, and the discussion at the international level about the quality of various institutions.

I think Owen will shortly wave one of his inhouse products, which deluges you with information and demonstrates-I think conclusively, in the eyes of most academics who look at it-that there are a number of leading multilaterals that perform at least as well as, or better than, DFID. DFID is an outlier; these multilaterals perform far better than the average bilateral. I think we ought to keep that in our head. The feeling that there is a long accountability chain is a real political issue, and I am sure the Committee wrestles with it, but it is part and parcel of an incentive structure where UK NGOs, UK businesses, and people who are interested in the UK bilateral activity are visited, quite rightly, and are seen in DFID’s programmes, but the multilaterals do not come into the same kind of contact.

As the UK is only one of 180 member countries-not the least, but not the only one-it is quite difficult to enforce or demand the kind of influence you would normally have on a bilateral programme. I think I would keep the kind of proportions that Owen raised. Of the 60% going through multilaterals-most of which, by the way, goes through AAArated multilaterals-only a small part goes to these rather bizarre normsetting smaller organisations. The 40% could go up safely within the governance structures-warts and all-of those organisations, and within the evaluation requirements they have, which are, in my view, tough.

The 20% deserves much closer examination. I do not think that at the country level, when people are deciding, "Okay, we will do this programme with agency x," they even check back on what the MAR says about agency x. You could ask that question when the case arises, but at one point, when Owen and I were both in DFID, we discovered quite by accident that DFID was giving more money to UNICEF India than DFID as a corporate entity was giving to UNICEF overall-obviously taking out the India proportion of it.

Peter Young: Yes, but Andrew, you participated in this MAR/BAR thing, which I had forgotten, and that rated the UNDP, for example, as "good or satisfactory". Anyone who attempts to look at the evaluation reports on that organisation-most of which it does itself-concludes that that is an entirely inappropriate statement. Basically, insofar as you can tell what they are doing, because they do not report on the results, there are all sorts of problems throughout the organisation, in terms of speed, bureaucracy and competence.

Q75 Chair: Unfortunately, I have to stop you. I think you have clearly made your point-a point of disagreement. Please feel free to add to it in writing, but I think we got the message. Our problem is that we have another set of witnesses, and we are going to run out of time. On a related issue, let us say that we do 40% bilateral, which you think should fall, and 40% multilateral, which you all think should rise, and the bit in the middle is mixed-is a bilateral purchase of multilateral services. I was interested to hear you imply that we do not have proper supervision of the 20% that we buy, or that it is rather arbitrarily done. What we are really looking at is how DFID could have a better relationship with multilateral delivery, especially in those countries were we do not have bilateral programmes. You are almost implying that we could get rid of DFID and have a small bilateral agency, and that the Treasury should just hand over the money to multilateral agencies and let them get on with it. There is a political problem here of accountability and British national interest. Do you really think The Daily Mail would be satisfied if we said: "We are delivering £8 billion a year to the World Bank, and we have a board member, but the details of what they do is up to them"?

Jeremy Lefroy: I do not think The Financial Times would be satisfied; it is not just The Daily Mail.

Chair: Indeed.

Owen Barder: I was in the position one year of being Britain’s negotiator for the IDA replenishment for the World Bank, where precisely this issue arose. One of the things that most struck me was that, around the table, my colleagues from donor nations were mainly from Treasuries. I was the only fulltime development professional representing a major donor nation, and it gives Britain an enormous advantage in those negotiations that that is the case, because the British negotiators in that process have a much better insight into the strengths and weaknesses of the World Bank, precisely because they are informed by DFID working on the ground in all these countries.

I was able to bring evidence and analysis from our country programmes about things the World Bank was doing well, and things the World Bank was not doing well and could improve. This combination of having a bilateral network on the ground, managing and delivering aid programmes, and a strong role in the multilateral system is extremely important.

Q76 Chair: That is really helpful. Can I just push another question, while you are answering? I think it would be helpful. In that case, would it be better to have, in addition to what we have, perhaps some slightly stronger regional offices that operate slightly more co-operatively? For example, we are going to Brazil because we have no real presence whatsoever in South America, yet we give money to be spent there. If we had a slightly stronger office that engaged bilaterally with the Bank and the InterAmerican Development Bank, would that help inform us and them, and produce a better result?

Owen Barder: No, I do not think the challenge that DFID faces is that it has insufficient intelligence on the ground about how the multilaterals are working. Again, the last thing the world needs is more spread of more bilateral agencies to more countries around the world. There has been a trend towards countries concentrating on fewer countries and fewer sectors to increase the efficiency of aid spending, as agreed in the Paris Declaration, and I think we should continue to keep our eyes focused on that.

Peter Young: I disagree with that. I think having an effective bilateral presence in a wide number of countries is a very good idea. It enables you to keep a handle on what is going on. This business of pulling out of countries-Burundi or whatever-is not a good idea. I do not think that, in fact, one has enough understanding of what is happening within these multilaterals in the performance of their programmes. You need some kind of forensic analysis, done independently, in order to determine that. You might, in the case of the EU, have the occasional report by the European Court of Auditors. I think they are reliable investigations, but they are very few and far between, and they are usually pretty negative, too. We really do not know, however, what is happening with the majority of EU aid, for example.

Owen Barder: If I may, I shall offer you two data points. If we look at such academic studies as there are-there is one by Easterly and Pfutze, one by Stephen Knack, and one by my colleagues at the Centre for Global Development-on the effectiveness of aid agencies, all three find that multilateral agencies are substantially more effective than bilateral agencies on the measures we have. The narrow point I agree with Peter on is that we definitely need more evidence. We definitely need more facts-those are knowable facts.

Peter Young: We need independent evidence.

Owen Barder: These are independent. These are three academic-

Peter Young: I am not talking about some vague academic studies.

Owen Barder: They are not vague academic studies.

Peter Young: I am talking about forensic analysis of what is happening on the ground.

Chair: Hang on.

Owen Barder: With respect, Peter, unlike you, we are not paid by DFID. We are independent. That is one data point; there is another, which is what the Finance Ministers of developing countries say. This year, they have again written to the major donors to say, "We would prefer you to put more money through the African Development Bank and the World Bank than to put that money through more and more bilateral aid programmes." I would have thought, Peter, that you would believe that asking customers what they want is a good start for understanding what is effective and what is not. If the multilateral agencies were wasting money on staff, overheads and administration, and none of that money was reaching developing countries, then developing countries would scarcely be putting pressure on the donors to put more of their money through the multilateral system and less of their money through the bilateral system. That is what they are asking us to do. They are asking for a good reason.

Q77 Chair: Andrew, you are literally in the middle.

Andrew Rogerson: Yes. On Sir Malcolm’s specific point, I think it is a very good idea to stay very much engaged in parts of the world where bilateral aid has withdrawn, if for no other reason than because many of them are success stories where aid has worked-or perhaps DFID was never there in the first place. As for the specific mechanism of having regional offices, I have my doubts on that, both because of the distance that still remains and, more interestingly, because once you are already inside the governance of organisations like the World Bank with the World Bank’s Executive Director or negotiators, as we have heard, the only other way of getting in touch with country offices, and really understanding what goes on, is if you are a partner in some way.

You need some kind of minimal financial stake. You need skin in the game to be credible with other donors and lenders, and you would have to fund this, which DFID did, to its credit, for a while in Latin America and other parts of the world, but then it decided it could no longer afford even regional programmes there, which were a matter of trying to get catalytically injected improvements in the country without an office at country level.

Chair: That is helpful. The Committee is really teasing out where things are going and what different things might work. As I say, we have another set of witnesses. In conclusion, I think you have all given different views. There is scope for loans; you place different emphases on what that is, but I think you are all fairly sceptical about the case for having a bank. I hope that is a fair summary. If you feel that, in addition to what you have already said to us, orally and in your submissions, you have anything more to add on how loans might work, what your preference would be, and anything negative-where it would not work and why-that would be helpful. We were going to explore that in a bit more detail, but we have run out of time. I think you have expressed your views and made your positions reasonably clear, but please feel free if, on reflection, you want to follow this up with further details. The Committee genuinely has a very open mind on this. We are teasing out all the options before we come to any definitive conclusion. We are really open to detail. Thank you very much. It has been a really interesting session, and it is always good to have lively disagreement.

Examination of Witnesses

Witnesses: Tamsyn Barton, Director General, European Investment Bank, Marc Engelhardt, Director of Development and Climate, KfW Development Bank, Dorothee Fiedler, Deputy Director General, German Federal Ministry for Economic Co-operation and Development, and Dr Chris West, Director, Shell Foundation, gave evidence.

Q78 Chair: Thank you very much for coming, and good morning. I am sorry you have been kept waiting, but I think you were listening to the previous debate, so you will have some idea of the context. First of all, may I ask you to introduce yourselves formally for the record? Then we can proceed to giving evidence.

Dr West: Chris West, Director of the Shell Foundation.

Dorothee Fiedler: I am Dorothee Fiedler. I am the Deputy Director General at the German Federal Ministry for Economic Co-operation and Development.

Marc Engelhardt: My name is Marc Engelhardt. I am the Head of the Department for Development and Climate at KfW Development Bank.

Tamsyn Barton: My name is Tamsyn Barton. I am Director General responsible for the European Investment Bank’s lending operations outside the EU.

Q79 Chair: Thank you very much indeed. As I say, you will have heard the flavour of the debate. Our previous witnesses were rather sceptical about the role of a development bank for DFID, but the Committee clearly is keen to explore how other banks operate, and to consider whether or not there is a model that may or may not be appropriate for the UK Government to consider. We very much appreciate that you are willing to share your views and experiences in that direction. We have certainly heard differing views. I think most people have said that some form of loan window is becoming an increasingly necessary part of the mix, and others have said, "Why not a development bank? DFID is a big player, and most other countries have it; why not?" You have heard counter-arguments already this morning.

We know that DFID is considering this situation, and I think it would be helpful perhaps to take the German experience first. You have, of course, a Ministry, and then you have a separate finance arm. How does that work? How do you interact? What are the advantages or disadvantages of being, if you like, in two separate compartments-or how separate are they?

Dorothee Fiedler: Let me thank you, first for inviting me, and then for founding KfW in 1948. We are very grateful to the British for founding this very important institute.

Q80 Chair: We seem to have similar responsibilities for the French bank as well.

Dorothee Fiedler: I think it is very important for me to mention that this is a development bank that has been active for very many years. You may be familiar with the German system: we have those two agencies, GIZ dealing with technical co-operation, and KfW dealing with financial co-operation. We think that with financial co-operation, we can of course broaden our scope of co-operating with third world countries. You may know that the Germans co-operate with about 70 countries all over, so we also include in our programmes the socalled emerging countries-no longer China, but Brazil, India or Mexico, for example. We think that financial co-operation is a very important instrument for co-operation with those countries, because what GIZ does is purely technical co-operation, mainly sending German experts out, but what KfW does is offer a broad range of grants and loans to those countries.

Furthermore, what is very important to us Germans is that the budget part of our grants and loans is more or less stable. It has increased a little, but as a bank with a triple-A rating that is in the ownership of the Federal State and the Federal Government, KfW is very well able to get money from the market. You see a huge increase in the financial co-operation between Germany and a group of developing countries. I think it is about a 14% increase every year for 10 years now. We more or less have the same amount of Government money, but there is a huge increase in the money that KfW has raised. I will leave it at that for the moment.

Marc Engelhardt: If I may add to that, since you asked, Mr Chairman, for the division of labour between the Government body and KfW, the Government clearly sets the political framework. All projects that KfW is financing, either with grants or with concessional loans, are agreed by the German Government with the respective partner Governments in the developing countries. KfW is always acting on behalf of the German Government. The structuring of the lending, the projects and the oversight of the implementation of the project lie in the hands of KfW, so there is a very clear and distinctive division of labour between these two layers.

Q81 Jeremy Lefroy: Thank you very much. Just on a point of fact, what is the total portfolio of KfW in developing countries at the moment, in euros, approximately?

Marc Engelhardt: The ongoing portfolio, you mean?

Jeremy Lefroy: Yes, the current outstanding portfolio.

Marc Engelhardt: It is roughly €20 billion.

Q82 Jeremy Lefroy: Thank you very much. I think in AFD it is something like €16 billion, so I just wanted to get some kind of comparison. We are looking at the possibility that has been raised of the UK establishing a development bank. One alternative that has been suggested is effectively a virtual bank, in the sense that DFID already does some form of what we would call returnable finance. At the moment, it does not get segregated; we personally think it gets lost in the system. It could put all that into one area and then perhaps also do some concessional lending, if it has the legal power to do so. We were told earlier this morning that it does, which I think was a bit of a surprise to some of us. Would you think that the idea of going one step towards a virtual bank within the existing set-up, which would mean that there was the ability to do concessional lending and that there were people there who were experienced in credit risk management and so on, was a sensible way forward, as opposed to setting up a completely new institution?

Dorothee Fiedler: I would not really dare to judge this, but let me just make the point again that in our view, what is really very attractive is that KfW uses more than 60% of its own money, which is an addup to Government money. We have not only KfW’s money but also the expertise of a bank. The development branch of KfW is just one branch. There is the 1948founded KfW with all its expertise in development banking-they started in Germany-so we can make very good use of all that. For us Germans, this is very important. A lot of the abilities of GIZ, but particularly that special expertise in banking, and of course the fact that they raise money on the financial markets, is something very special to KfW. That is why we think our system works quite well with a development bank, with its special abilities.

Of course I have to say that the people working at the development branch of KfW are not just banking people. They have different expertise and experience. Many of them are economists, sociologists, or people who have studied politics, but of course the people at the development bank branch of KfW can draw on the experience of their colleagues in the other parts of KfW, so we think that this is an especially good model.

Q83 Jeremy Lefroy: How much of German ODA goes directly to GIZ, and how much goes into KfW?

Dorothee Fiedler: In 2012, to KfW it was-

Marc Engelhardt: €1.5 billion.

Dorothee Fiedler: A little more: €1.8 billion, and for GIZ it should be around €1.3 billion.

Q84 Jeremy Lefroy: Together that makes a fairly small proportion of German ODA, so where does the rest go? German ODA is what-€12 billion to €15 billion in total? It is about 0.5% of GDP.

Dorothee Fiedler: No, it is 0.4% of GDP.

Jeremy Lefroy: 0.4% to 0.5%.

Dorothee Fiedler: Let me give you the precise figures later, please, because I do not have them here with me.

Q85 Jeremy Lefroy: That was one of the questions we were asking JICA in Japan last week, which has a huge development bank portfolio of well over $100 billion, and of course their gross ODA is enormous, but their net ODA is quite small, because of the repayments coming from the loans over many, many years. I wondered how much of an issue this was in Germany. If you have an established development bank that has been going for many years, and then some of those 20year loans on concessional terms are being repaid and count as negative ODA, how much of a political issue is that? Perhaps on the one side, as we suspect in some countries, raising money on the bond markets through development banks is being used as a way of trying to get towards 0.7% without it affecting the public budget. On the other hand, as we have seen for instance in 2006-07 in DFID, where a major surplus in the Commonwealth Development Corporation reduced the level of ODA substantially, that can become a problem. In other words, if longterm loans are being repaid and new loans are not being given out, there is a negative effect on ODA. I wondered how much of that is a consideration in Germany at this moment.

Dorothee Fiedler: I think this is a problem, yes, and of course with politicians being elected every four years, whoever is in power does not really think about what might happen in the middle or longer term. Of course we are striving very hard to reach the 0.7% goal. We can only congratulate you on already being there, but for us it is still very difficult. I do not think we have really solved that issue. Of course there are ongoing discussions at the OECD, because we think that a bigger part of those loans should be recognised as ODA; the money that can be made by KfW at the market should be accepted as ODA, which it is not right now. To be very frank, I do not have an answer to you on how politicians, in a couple of years, might deal with exactly that problem, which we know we are going to face-that the money will come back. Maybe Marc has another idea.

Marc Engelhardt: As a footnote to that, in the long run, this is clearly a problem if the system stays as it is. For the time being, the German ODA quota is rising, and KfW’s financed share of the ODA is rising. We had an ODA percentage of KfW’s lending activities and grant activities of 6% in 2005, and last year, we had a percentage of 20%, so the rise of German ODA from 0.3% to 0.4%-that is still not sufficient, but it has risen in the past years-was mainly due to the capacity of KfW to augment its share in concessional lending.

Q86 Jeremy Lefroy: Through raising money on the markets?

Marc Engelhardt: Correct.

Q87 Jeremy Lefroy: Which is what the French have done as well.

Marc Engelhardt: Correct.

Q88 Jeremy Lefroy: I have the figures here for German ODA in 2012: it was €10.2 billion. The total you told me for KfW and GIZ is €3.1 billion, so does that mean the German Treasury pays multilateral agencies directly?

Marc Engelhardt: No, because what Dorothee Fiedler was mentioning was the grant part of KfW’s share.

Q89 Jeremy Lefroy: €1.8 billion is grant?

Marc Engelhardt: It is mainly grants and a few soft loans from budget funds. In addition to that, there is another €3.1 billion from KfW’s funds from the capital market-the bonds KfW issues-and a large part of it, roughly 80% of that, merged with grant funds and/or covered by partial risk guarantees from the Government, is accountable as ODA, so the overall ODA that KfW is accountable for is well above €4 billion-€4.5 billion or so.

Q90 Jeremy Lefroy: Or nearly €5 billion, if there is another €3.1 billion on top.

Marc Engelhardt: Yes, as I said, it is not all accountable as ODA. We had total new commitments of €4.9 billion; we are talking about new commitments here, but the disbursements (which are the basis for the ODA calculation), of course, lag behind a bit.

Q91 Jeremy Lefroy: So approximately 20% of Germany’s ODA is through net concessional lending, over and above repayments?

Marc Engelhardt: Yes.

Q92 Fiona Bruce: How can public funding be best used to leverage in private investment? You have heard a lot of the discussion with the previous witnesses.

Marc Engelhardt: First of all, the way development banks-not only KfW, but also the French one and, to a certain degree, EIB and others-function is they issue bonds on the capital market, at least €3.1 billion last year, as I mentioned. Overall in the last 10 years, roughly €20 billion that KfW was spending in addition to Government funds in developing projects for developing countries was raised on the capital market, and so came from the private sector. Private investors who bought those bonds are automatically, if you want, contributing, in a very efficient manner, because they use the system KfW and the German Government have established, to funding development projects. This is one indirect but considerable effect of development banks that people often overlook.

Then of course we have direct ways and means of fostering private sector development. One example is SME funds, which KfW and others have set up in eastern Europe and now in Latin America and other regions. We structure the risks; we collect funds from Governments, from our own funds, and from private investors, to be mixed in one pot but structured so that the private sector takes a lower share of risk, and thereby is attracted in sectors and countries where it normally would not invest.

Q93 Fiona Bruce: So the private sector investment group, you think, is a good model?

Marc Engelhardt: That is also a good example of such ventures, correct.

Q94 Fiona Bruce: My next question is really to all witnesses: how do Governments ensure that private sector investment is consistent with our broader aims of gender equality and environmental sustainability?

Dorothee Fiedler: We have very clear rules and regulations, of course, in respect of KfW and GIZ, which are our executing agencies. We also do a similar thing with the multilateral institutions and banks, but mainly with GIZ and KfW we have those very clear rules and regulations. If the private sector is involved in any of our projects, they would have to cling to those same rules and regulations that we have, so we do not finance anything outside of those rules.

Fiona Bruce: Does anyone else want to come in?

Tamsyn Barton: Yes. I can explain a little bit about how it works in the European Investment Bank. We are, of course, owned by the member states of the EU, and every single one of our projects has to be checked by every directorate general in the European Commission, which you will be surprised to know we achieve within two months for each project. Nevertheless, that means in every area of EU standards that have been agreed by member states with the Commission, there are clear safeguards in relation to the areas you mention. On the environment side, I would say that probably we have the toughest standards. Sometimes we find we are not able to cofinance with other development financing institutions because our standards are that much higher, because of our commitment particularly in relation to action to tackle climate change.

As far as social safeguards go, in addition to the policies we have, we also have within the bank specialists, just as KfW does, so we have social and environmental specialists who obviously work closely with other international financing institutions to make sure that we use those standards. We use them partly in our ex-ante estimates of what the results of every project will be. We have a results framework: Peter Young, who was sitting in this place before, seemed to express a lot of doubts about whether such transparent mechanisms exist. Well, they do in the EIB, at least. I was able to introduce a results framework, so it should be pretty clear. We are not able in every case, for example, to get data in relation to gender impact, and I know we can improve in this area, but we have made a start.

Dr West: Just to make a general comment and come back to your specific one, Fiona, if I were to look over the last decade, the transformational change that I think is happening, certainly from our perspective, is that there is a growth in social entrepreneurship-businesses who have, as a mission, a blended social and financial return core to their purpose. That is one transformation of the last decade. I think equally there has been a blending of how the public sector and private sector can co-operate to support these ventures growing, and that has led to impact investment; it has led to an increase in private capital from foundations and others.

Possibly misusing Owen’s analogy about the football field at the beginning, where I see the problem at the moment is that you have, if you like, one layer of people who have grant funding to help these enterprises set up and start, and then there is a valley of death before they get into other types of organisation that have a finance-first perspective on life. The real gap we have in this market is how we bridge those two worlds. This is where I find your question about a development bank or other body hugely important. I do not know KfW well enough, but if I were to typify most finance-first organisations as being typically commercial banks, in my experience most development finance institutions are, too. We therefore find a rather unfortunate position: you can catalyse these new, really scalable social entrepreneurs, who are wedded to generating this blend of social and financial value, but they hit this wall where there is no one with the risk appetite, patience and skill support to get them to the level where they can go on to the balance sheet of, for example, the EIB. My anxiety about the use of concessional finance is that, whether it is cheap money or expensive money, if you do not have a viable partner in which to invest, that money will be wasted. I do not think the issue is the concessionality.

The issue is: can you blend together a grouping of people who can offer the skill support, the appropriate finance, in many different forms, and the market linkages to help this growth sector really emerge, until it gets to the balance sheet position of commercial banks or others? That, for me, is what is missing at the moment. Using Owen’s analogy, I do not think there is a midfield in this football team at the moment. There is a grant-first, development-first sector, and then there is a finance-first sector. What we need is to straddle that. Foundations to some extent are playing that role, but I cannot pass all our successful partners over to someone who will do balance-sheet lending with a finance-first perspective.

Dorothee Fiedler: I would very much like to add to that, because what we have not spoken about a lot is the huge difference between a development bank and a bank. KfW is a knowledge bank as well, so they know a lot about what is going on in developing countries. They have very longstanding experience, and they do not just give money; they do not just give grants or loans. It is the other way around: they know a lot about developing countries and developing policy; they have a lot of experience. For example, right now the World Bank is saying, "We want to be seen as a knowledge institution." The countries we are co-operating with, even the bigger ones, say, "Okay, what we want from KfW is not just their money. We get money at many different banks. We want that special expertise that you can offer." They do it either by themselves or through people they hire, but that is what we really like very much about the idea of having a development bank, and not just a bank that gives out grants or loans.

Marc Engelhardt: That is really a big advantage of a development bank, because you have the holistic assessment of a project; you have the banking perspective, the development perspective, the real financial risk, but also the reputational risk in one institution, at least. We also have a credit risk department, which tends to set pretty hard standards. We have an environmental department, which is as hard, probably, as the one at EIB, but we have the whole assessment and the whole risk burden, in the end, in one institution. That helps a lot.

Coming back to the knowledge point of view, that may be another argument for a bilateral bank. You will not be surprised that I am in favour of a bilateral bank, because you have a specific knowledge in each country. That is, you have real value added. Some multilaterals surely have good projects, and of course there are a lot of good arguments for having multilateral banks and multilateral agencies, but especially higher developed countries ask for specific knowledge and value added, and you can better combine that with the knowledge you have from your own country, because we have it at first hand.

Tamsyn Barton: Could I also pick up this missing midfield point? It does seem to be pretty critical. It is the job of DFIs to fill that space, and we do at the EIB, in cases where we take a micro-finance institution that has begun with NGOs and grants and develop it into a more commercial one. The more we do that very labourintensive work, the more you come back to the question raised a lot in the last session about administrative costs, because it is very labourintensive. Also, with public money, there is the issue of the appetite for risk, which was raised by Peter Young. There has to be a balance. We have, in the EIB, a revolving fund, and it is the gift that keeps giving, in the way that you would like the private sector to. We had to agree with the member states that capitalised it through the European Development Fund that we would operate a certain standard of risk management, so that their money would come back to them. We have now been given an additional window through which we can basically take more risk for higher social returns, so the balance has shifted a bit between finance first and social first. It is really a question of the mandate of the institutions, and the appetite of those putting in public money to provide the flexibility for it to play that crucial midfield role.

Q95 Jeremy Lefroy: Following up on what Tamsyn and Chris were saying, the biggest problem with regard to the middle, which is still a massive problem, seems to be the transaction cost, which basically means the cost of the people. It is often highly paid, developed country people doing the job. How do you see us being able to overcome that? To me, it is a major problem, and often I have seen people saying, "We would love to do this, but the transaction cost is too high." In effect, you have a perfectly good thing that is sunk because it is costing too much in terms of highly paid people. How do we get round that? There must be ways to get round it. Obviously, volunteers can overcome that, or training people locally, where the costs are perhaps lower, to do all that kind of professional work.

The second thing is the tools available. They seem to be either grant or loan, and there is always a problem-Chris will know a lot about this-with equitytype tools in small businesses, particularly where the governance structure is not set up so that equity makes much sense. How do we get around that? Again, it seems to me that a lot of very good SMEtype projects are simply not financed in developing countries because of the question of the lack of an equitybased tool, or the cost. How would a development bank fit into that, or how could something that DFID could do fit into that?

Dr West: For this market segment, if I look at social enterprises growing, to get through that midfield level one has to accept that whatever products and services they need, they need skilled support, and they need finance in the right form, which therefore means the transaction cost of servicing that market is high and the risk is high. That is why the end result is a lower yielding return out of investing in that. It is a combination of both cost and risk. Now, there might be smarter ways of doing it, but I think fundamentally that is why you need a bridging instrument beyond the finance first, which will try to reduce cost and therefore reduce risk, and therefore invest in things with a slightly more commercial return expectation. If we do not have a higher-transactional-cost, risktolerant vehicle in the middle, I do not think you will get the graduation of these initiatives to the scale we all hope for.

On your second point about the range of instruments, again, there is a lot of liquidity in a lot of emerging economies, as was alluded to in the last conversation. A lot of that is locked up in banks that have hugely conservative lending rates, and of course it is often provided in shortterm debt. If you are a startup growing business in any country in the world, you really need some form of patient, flexible finance that adjusts to your cash flow income. It is not necessarily a short-term debt instrument. Equity is usually not very attractive to the entrepreneur, and it is also not necessarily attractive to the investor, because there is not a very clear exit route from investing in equity in small ventures like this.

You really need different finance forms. You need mezzanine finance forms related to the cash flow performance of the business that are much more patient and much more flexible in tenor. Again, they are not difficult to construct, so I think there is some financial flexibility here, on the type of instrument that is most suited to this. On the grant side, again, grant historically has been provided as an input base support. It will buy down the costs. I think impact bonds were alluded to in the last conversation. There is an opportunity to use grant as an outcomerelated tool: can you buy down and offset the ability to generate social outcomes that will leverage in the private sector?

In all these instruments, you can be far more flexible than "It is a grant for buying down cost, and it is shortterm debt or private equity." That is the bit we need to engineer in the middle. I think there are various people doing it at the moment, but I cannot think of a single onestop shop that one would go to with a portfolio of deals like this to say, "Does this meet your investment criteria?" It is certainly not a development finance institution.

Q96 Jeremy Lefroy: Following on from that, it appears from what we have seen of DFID’s work that they are doing quite a number of those things in different parts of the organisation, whether it is in their involvement through the Africa Enterprise Challenge Fund or through support for micro-finance. However, I have not yet got-maybe you can help us on this-a sense that there is a place in DFID to which somebody, as Chris says, could come and say, "This is the problem; do we have a solution?" Either they are all over the place, the particular instruments they have, or, in the case of concessional finance, they do not really exist at the moment. This is what we are really trying to get to, and ask whether you think that this would be a good thing for DFID to do, given that DFID seems to have a lot of the expertise anyway, but rather scattered throughout its organisation, and perhaps some of it is in CDC, which we will come on to.

Tamsyn Barton: I wonder if I could link that to your former question about admin costs, and how we can get around them, which is the perennial hope. I am afraid the bad news is that I have worked in DFID, I have worked in NGOs and now I am within the EIB, and there is no way around admin costs. You can transfer them to others, but at the end of the day they do have to be picked up. I rather agree with the former witnesses who pointed out that that tends to be done more efficiently in multilaterals.

Having said all that, if your ambition is not for the UK to fill the entire field, but rather for DFID to broaden its understanding of this area, particularly using DFIs to lever private investment, then DFID would need to get its hands dirty a bit more, and perhaps be in one place. There is the CDC already, which was formally spun off from DFID and focuses on equity. It now has the ability to lend, and there are obviously options there as to how DFID relates to it. I am not up to date, but I believe it is not so closely involved at the moment. I would say that there is no substitute for facing the tradeoffs that we have been talking about, in terms of risk and administrative costs in relation to objectives, and actually doing it. DFID’s strength up until now has been that it is alone, I think, in having covered the field-multilaterals, bilaterals, humanitarian relief, research-all in one place, so that it is not just an aid player but a development player. I have a certain sympathy with it getting involved, but probably only in an incremental way, because it is costly. I do not see how you can have a virtual development bank, except with an arm’slength relationship that already exists in the way that it sits on the governance of our investment facility. There would need to be some more direct involvement, I would think. That involves having risk people, specialists, and so on.

Marc Engelhardt: I agree 100%. It is costly because you need qualified staff. Nevertheless, you should not be afraid that; when you establish a development bank, you have to pay salaries for investment bankers. Clearly for KfW, this is not the case. We are lucky to attract the bestqualified people from universities, but they are not coming to us because of the salary; they come to us because they look for a meaningful job.

Q97 Jeremy Lefroy: This is the point we made to CDC when we did the report: people would be very willing to come and work for an organisation like CDC without looking for City of London salaries.

Marc Engelhardt: There is exactly that point. I have one other point: in the session before, you were referring to the local structure. KfW increased its local structure tremendously over the past 10 years. We meanwhile have 80 offices abroad, but with only 80 to 100 permanent staff from Frankfurt, plus another 300 local staff. Per office, it costs us roughly only €300,000 to €400,000 per year. The reason for that is not so much that we use local furniture, as you were referring to before; it is because we have a very lean structure. We have normally one German or international representative-sometimes two in the bigger countries-and then four to five, sometimes 10, highly qualified local staff, who are also not cheap but are, of course, far less expensive than international experts. We can thereby afford a very powerful and important structure for our quality assurance that is not that costly.

Dorothee Fiedler: Yes, I would like to underline that. I spoke about the German system with KfW and GIZ. We have people going to both sides, so people who have worked with GIZ now work with KfW and the other way around. As Marc said, in both institutions we do increasingly employ highly qualified staff in those countries. In more or less all of Latin America, you sometimes get better qualified people from those countries than from our own country. I would still say that the people-be they from KfW or GIZ in Germany, or from those countries-are cheaper than the experts from the big development banks, the World Bank and other banks. We think this is a very important point.

The other thing is that we do a lot in human resource development in the field of technical cooperation. The people that we do the training for in technical cooperation could then be used in the field of financial cooperation, which is KfW. This makes a lot of sense, and with your DFID experience, knowledge, structure, and whatever, you will already have a lot of things that you have also done in human resource development, so you could very easily find well qualified people in the countries you are cooperating with.

Q98 Hugh Bayley: My first question is to Marc and Tamsyn. What proportion of your loan book is accounted for by loans to least developed countries, and what proportion to EU accession states, past or present-that is to say, countries that have recently joined the EU, or states, like Turkey, that are in the course of negotiations.

Tamsyn Barton: I cannot give you the figures now; I would have to do so later. I would think for least developed countries they would definitely be small. I would say that the investment facility as a whole you would see more concentrated on lowincome countries, but mostly not the least developed. It is focused on ACP countries. I did check out the figure for the Commonwealth in the ACP, and it was 52% of our investment, but I will have to get back to you on that one. Accession is huge. That is one of our key mandates as the EU bank, and making it past and present is definitely going to complicate getting it, but I will get it for you. I can tell you about last year: last year, our lending side outside the EU was €7.4 billion, and €4 billion of that was nonaccession, so it is a substantial proportion.

Q99 Hugh Bayley: Outside the EU was €7.4 billion, of which €4 billion was nonaccession. So getting on for half is accession.

Tamsyn Barton: That is indeed how it is at the moment. Croatia, of course, acceded yesterday, and that was one of the more significant countries, but Turkey is indeed the largest, in terms of annual lending. That is what we do as our mandate, but the cost is far less in relation to lending in Turkey-both the internal cost and the use of public funds is much less, so outside the ACP, where we have the European Development Fund, we just operate on the basis of a guarantee from the EU budget. We do not use it in relation to the private sector, except for political risk, and at least half of our lending in Turkey does not involve any public money at all.

Q100 Hugh Bayley: And for KfW?

Marc Engelhardt: I do not have the absolute exact figures, but when it comes to the accession and preaccession countries, it is less than 10% of the new commitments and also the overall portfolio. We still have a portfolio in Turkey, and a little bit in Romania, Bulgaria and Croatia, but all in all, this is far less than 10% of our portfolio. As for the LDC portfolio, all in all, new commitments to LDCs-concessional loans and grants-is approx 20%. What is interesting is that this already used to be the overall figure 10 years ago, when we operated almost exclusively with Government funds. 20% went to LDCs. Meanwhile, nearly 50% of the, Government funds go to LDCs, because we can use those funds in MICs much more efficiently. We use much less Government funds for MICs and higher-developed countries.

Q101 Hugh Bayley: If you think back to the time, 10 or 15 years ago, when there was widescale multilateral and bilateral debt relief, what proportion of your respective portfolios did you write down or write off? Perhaps I should add, I think for both banks but particularly for the EIB, what impact has the eurozone crisis had on the value of investment? For instance, have you had to write down or write off historical investments in Greece? The first is a more general question about the debt relief exercise 10 or 15 years ago.

Tamsyn Barton: Again, I will have to go back and check out the figures for you, but we are fully implicated in all the debt relief for HIPC countries. It is not always straightforward to pay it off. We have just had to provide €50 million to Ghana in relation to debt relief, which went straight into a Millennium Development Goal contract there. There was a little wrinkle, which was that we had not dealt with the reflows from the Lomé Convention, but while I was in DFID I managed to sort that one out so that they also went back to the countries from which they came, because it was not initially involved in the first cut of debt relief. I am not aware of an instance where we did not fully conform to what other multilaterals were doing.

Q102 Hugh Bayley: It is not so much about whether you conformed. I am just trying to think about what would happen if we were to set up some UK instrument for a loan window, whether within DFID or within a separate institution. Thinking over a 25 or 30year period, if that is the course of the loans, we ought to have an idea of how many loans inadvertently turn into grants, because inevitably you make bad investment decisions. That is the nature of it.

Tamsyn Barton: That is a very different thing. The organised debt relief is not at all the same as losses taken in the course of normal business.

Q103 Hugh Bayley: Is it not conceptually the same thing, just dealt with in a different way?

Tamsyn Barton: That would be a very, very problematic line to cross. What is crucial to all of us MDBs is that we have preferred creditor status, so we will be paid back. That is the only way we can operate. To answer your question, now I understand what you are getting at, if you take the use of the guarantee from the EU budget for our external lending, since the 1980s there have only been three cases where we have had to use it: Argentina, the former Yugoslavia, and, most recently, Syria. Syria is obviously ongoing, but with Argentina and Yugoslavia we got all the money back, and at a very decent rate of interest, by today’s standards.

Hugh Bayley: Today’s standards-oh, yes.

Tamsyn Barton: It is a benefit, ultimately, to the grants available for ODA, so member states have not lost any money in that way.

In relation to the investment facility, as I mentioned, it is ringfenced, and it does take a bit more risk within the limits of the risk parameters. In more risky countries, such as Zimbabwe and Côte d’Ivoire, we did have some losses, but we generally recover them over time. We also have complications when there is political conditionality or sanctions imposed, as in Syria, but in the end, we get all the money back with a good rate of interest.

Marc Engelhardt: At KfW, we were not affected that much by the huge debt rescheduling in the HIPC countries in the late 1990s, because back then we operated almost exclusively with grants or very soft loans that were restructured. In the last 10 years, when we really started concessional lending with market loans, I recall only one case where we had a write-off: in Central Asia in 2009, we had to write off, for that fiscal year, approx, €50 million, but we regained a good part of it the next year due to a restructuring exercise, so there were net losses of maybe €20 million or €30 million but over a period of 10 years, with market funds amounting to €20 billion, so this is a very low rate of losses.

Q104 Hugh Bayley: There are two conceptual questions that it would be very interesting to hear about, if either of you were able to think about them and send us a further note or evidence. If you create a loan instrument, as you have done over the last decade, does that skew your partnerships more towards lowincome and middleincome countries, rather than least developed countries, and, if so, to what proportion? That would give us a little bit of a guide as to what the implications of unlocking this Pandora’s box would be.

Secondly, there is a question over what proportion of your overall portfolio turns out to be a bad investment and is, in effect, written off. It sounds like, from both of your cases, you can manage things so that it is a very small proportion, but I would be interested to know. If DFID were to create a loan window, would it make it easier, Marc, for KfW and DFID to join together in a joint venture on a scheme?

Marc Engelhardt: It would open new opportunities. As a matter of fact, we are already cooperating on a number of projects, together with DFID. I have heard that this is functioning pretty well, but, of course, it would open new perspectives. One model could be a model established by the EIB, AFD and KfW, and formally launched this year, which is the Mutual Reliance Initiative, where these three development banks recognise each other’s procedures when it comes to project appraisal and monitoring. That is really an effective mechanism of efficient cofinancing, because then one takes the lead and the others do not have to double the effort, so you really have efficiency gains. The procedures are pretty elaborate. We have a 50page document with all these procedures laid out, so it is not just a politicallipservice document; it is a really matteroffact coordinating of nittygritties and nuts and bolts. That is why it is functioning, in our opinion. If DFID were willing and able to follow those procedures, one could open up new forms of cooperation, or venture into new projects together with DFID.

The other window, somehow linked also to the Mutual Reliance Initiative, is EU blending. As you probably know, the European Commission is planning to enhance the part of EU grants blended to loans from development banks, with the same interest I reckon the German Government had 10 years ago, to increase the leverage and still be able to use the budget funds in a more efficient way. This is happening on a large scale, projected for the next sevenyear cycle of the European Commission. There, of course, DFID would be a highly welcome partner, as an additional European development bank in that circle.

Tamsyn Barton: Could I pick up your point on skewing and your worry about skewing? I would say it is the opposite. There is a new opportunity to make sure that scarce grant funds are not used for the wrong projects or the wrong countries. I certainly can see that, within the context of the European Commission’s funds, they have been able to focus a bit more on lowincome countries. They know that, in effect, they can ensure that there is not a gap left, because there can be an increasing move into more lending, or blending where loans are the larger part. I would not say it is skewing. It can achieve different goals with different instruments. One of the key things that we do for the EU is in relation to the BRIC countries, for example. There is clearly a mutual interest in climate mitigation, and those are the countries with the biggest emissions. If we can bring the expertise from Europe with the latest technologies to bear, that is probably the most useful use of those funds that you could find for that goal. You need to determine what your priority goals are.

Q105 Hugh Bayley: That is the point. If you want to do middleincome countries, set up a more appropriate mechanism. If you want to remain a DFID niche for the poorest countries, then probably you do not need this additional instrument.

Tamsyn Barton: It is partly about countries. It is also about projects, because even in poorer countries, you can really only get the scale of investment in terms of a loan, and it is a more appropriate use of the scarce grants just to blend, for example, for the project feasibility work, without which it would not happen. So it should end up being more effective, and, again, if DFID were involved in both grants and loans, it might be better able to determine, in the context of what it does multilaterally, where the best uses are.

Q106 Hugh Bayley: Chris, you said some time back that DFID is lacking business DNA. Is that still the case?

Dr West: Let me first define what I mean by "business DNA": how to assess and take risk; to look at a market opportunity and provide the best form of product or service to satisfy that; and to look at it in a way that you escape subsidy dependence and create something that can stand on its own two feet. This goes back to Jeremy’s point. Clearly, DFID has evolved in the last decade or 15 years in many ways. The whole world has evolved in terms of how it wants cooperation between the public sector and the private sector. Going back to the point about the bank, I am very loth to recommend ever setting up a new institution when you do not look to see whether the existing institutions can do what is needed, but I would say that there is a gap here. As I see it, and given the market I am supporting, the present institutional landscape or framework is not sufficient to allow these social enterprises to go to scale. Despite what everyone might say, there is a gap.

Then I look at what the instruments are, in a UK context, that could help there. If I look at CDC, it has had a private equity focus in the past, although it is admittedly getting into loans now. The risk appetite and the return appetite from that is simply not adequate to address this gap. We need impact funds. DFID has very generously put an impact fund through CDC. Personally, I doubt whether that will influence the balance sheet of CDC, because it has such a different lending mandate to it. It might well be well used, but then what? Is it going to influence fundamentally how the CDC gets into this space? I doubt it.

Going back to your point, DFID has very credibly got a lot of instruments to support these types of activities. There are challenge funds. You mentioned the Africa Enterprise Challenge Fund. From the private sector angle, challenge funds are incredibly clunky, inflexible, processdriven and timebound. We have had partners where we have just thought, "Let’s not bother going down that route. I will absorb the cost, because it is just quicker." They are good instruments in name, and they have had some successes, M-PESA in Kenya being an obvious one. Could they do more? Absolutely, with a different structure and mandate to taking risk, and operating in a more flexible way.

DFID equally has the Business Innovation Facility, of which I chair the selection committee. Again, it is extremely good at taking risk and outsourcing it to a consultant to manage. It is really bringing in the private sector to do things that it would not normally have done. Excellent. But it is outsourced to a consultant. It is not really part of DFID.

Then I start thinking, "Is there anything in the UK system at the moment that I can take our partners to that operates that bridging support of skill support, appropriately structured finance, and ways of leveraging in others, like KfW?" and that does not exist. It is not just the UK. From my angle, there is not one, whether I find it in Nairobi or whether I find it in Brazil. These entities are few and far between. Maybe I should get to know KfW more, because the idea of someone playing a proponent role, developing deals to the point where they are financeable, is what is needed.

If we go back to the pre-accession days in the European Union, what was the real problem in preaccession? It was not the fact that there was not money available, but that we needed to develop deals that could be bankable to secure that money under the various programmes at the time. We have the exact same problem in the development space at the moment. We need a project preparation facility, coupled with finance that will take a loweryielding, marginal return on those deals, and leverage off that, and then bridge into other types of finance, which are amply developed. Business DNA, moulded with development DNA, is what this entity needs to have. It is those two sets of DNA. At the moment, we have got one that is development DNA-grantbased-and financial DNA. That is how we have to mix that up.

I am just worried about a model that is outsourced to a consultant, not because consultants are not great, but because you want a manager with an incentive framework that is judged on how he or she delivers outcomes from such a vehicle. Consultants are often judged by whether they have met various input targets. This is why I feel you need a mechanism where somebody is incentivised to make things work. Part of that incentive framework then captures the things that Fiona was talking about: can you generate sufficient measurable, auditable social value, as well as financial value? Can you leverage in capital from others? To me, that is a much more interesting structural route than the routes we have at the moment in the UK.

Q107 Hugh Bayley: I would like to go back, if I may, to the German experience, Dorothee. It is clear to me that this move towards using KfW increasingly to provide loan finance is something that has happened over the last decade or so. What have been the considerations within your Ministry, in terms of German Government policy, that underpin this? In other words, has the policy been formulated by the Government and by Ministers, and a brief passed to KfW, or has it happened without a direct policy initiative being taken by the Government?

Dorothee Fiedler: This is the way we share our work between the Government and KfW. That means there was a general agreement between the Government and KfW, which has been in existence more or less ever since. Then we have guidelines for financial and technical cooperation, which have been very clearly set by the Ministry; of course, that is in discussion with KfW, but they have been set by the Government. Then there is a contract letter for each and every single project we have. What we are doing, and what we do increasingly, is really the policy governance of our financial and technical cooperation very much jointly with the organisations, but it is most definitely the Ministry’s task to set those guidelines for, for example, financial cooperation.

Q108 Hugh Bayley: Why did the Government decide that Germany should increase the proportion of its ODA that took the form of loans?

Dorothee Fiedler: Because we had the impression that for many of the partner countries we have, given their increasing wealth-think of a country like Brazil-there is no need simply to go on giving them grants. It is the same for other countries. We think that we could be much more efficient and effective by increasing the loan part, because there are many projects for which the loan can and will very easily be repaid by those countries, and it is not only the likes of Brazil and South Africa; it is many African countries and many different Asian countries. We have the impression that, with our partner countries, this can very easily be increased.

The Ministry is about 50 years old now, and you may remember, more or less, what the development world was at that time. Of course, it was mainly grants, and we called it development aid. Now we call it cooperation, so I think it is a very clear decision definitely to give more loans.

Another very important underlying idea is that you get more ownership when you give out loans. You definitely discuss it with a partner; it is not just that you come with a big heap of money. You have to ask the Government of this or that country if they are willing to pay it back, even at a modest interest rate. We have the impression that in many cases the ownership is maybe a little higher than in other cases.

Q109 Chair: Just perhaps drawing the threads together, you see from the previous witnesses, and indeed the overall structure of DFID, that the UK has concentrated on trying to reach the poorest people in the poorest countries. In that context, grants have been the vehicle for doing it. As the world changes, people’s incomes are rising and the number of very poor countries is thankfully falling, although we have a debate about where you make the cutoff point. Is a country that has $2,500 as a percapita income a rich country? I do not think that many people would think it was, so we could have a debate about that. What I am interested in is the extent to which you feel your operation-the loan operation, the banking and financial operation-helps to lift people out of poverty, whether they are in the poorest countries or in emerging economies. If I give you the example of India, which we have talked about before, the UK has decided that the grantaid approach to India is no longer appropriate, but we are potentially walking away from 400 million people living absolutely at the poverty level. A lot of what you do goes to middleincome countries and into projects that are much closer to the market. We are interested in the extent to which, if DFID was to go down this route, it could add value to DFID’s very sharp focus on poverty reduction, whether it is the poorest people in emerging economies or the poorest people in the poorest countries. Do you feel that you can achieve that in the way you operate, and do you think that there is a model that would enable DFID to achieve that?

Dorothee Fiedler: There is still a very strong focus on the poorest in German development policy, as well. However, we feel that in those more emerging countries, the countries themselves very often are even better able to do something directly for their very, very poor than we are. I happen to know a bit more about Latin America than about other parts of the world. If you look into those huge social programmes that Brazil and Mexico, for example, do, they do it very well. We could not really assist them, but we cooperate with them in other fields. If you want, we give some of those countries just leeway. We leave their budgetary money really to cooperate directly with the poor in those countries. We have the impression in many of those more emerging countries that, in a way, it becomes increasingly difficult for us to assist with the fighting of poverty, and we think that we should concentrate on other fields where we think that we are better able to do it. Of course, this is different in other countries. As I said, a lot are subSaharan African countries, where we still concentrate very much on direct poverty reduction. However, in those more advanced countries, we increasingly get out of those programmes and get into other programmes. If you look at the portfolio of KfW, I think a lot of KfW’s money goes into climate programmes, water programmes, and infrastructure programmes, as well as social infrastructure. This a bit different from what you do to immediately work for the poor.

Q110 Chair: What about the EIB?

Tamsyn Barton: You have heard a lot from other witnesses about how the landscape is changing, and I think a lot of people have been given pause for thought by looking at what happened in terms of the MDGs, where the major numbers lifted out of poverty were in China, and that was obviously an exportled route and an indirect route. More and more in the development landscape we are thinking, "How can we be part of a much largerscale, transformational effort with smaller amounts of money?" It is often harder to show what our grants have achieved beyond a very shortterm, direct impact, and I think that, given DFID’s strength up to now in being an allrounder, it needs to be, in some sense, present at both ends to understand where the most difference can be made, and to use that knowledge in policy in relation to the multilaterals and what it does more generally.

In EIB we do some direct poverty lending with much higher levels of grants-water projects, for example, or micro-finance investments-but the bulk is on the indirect side, and that is all very clearly set out in the way that, for each project, we anticipate what the results will be, how we are going to get there, and what the specific EIB addition to the project is.

Q111 Chair: Do you both lend to the public and private sectors?

Tamsyn Barton: We certainly do, and obviously are involved a lot, for example, in support to PPPs, which were raised earlier, where we have specialist expertise.

Marc Engelhardt: We lend mainly to the public sector, but sometimes we can also lend, via financial intermediaries, to the private sector. By the way, we also support social entrepreneurs in India, so that is also possible if you have an intermediary. There may be one footnote that may help: you have to differentiate between lending and onlending conditions. The lending conditions to MICs, to a country like India or Brazil, should be more than a grant, but then of course, if, for example, you have a water sanitation project, you can formulate onlending requirements for remote areas or small municipalities that are very favourable, or even grantbased, and for bigger or richer municipalities, you have perhaps onetoone onlending conditions, so there you can play to get tailormade financing, not only for the country but also for the end purpose of the project, and differentiate it according to financial strength as well as performance. We had projects where we linked the on-lending conditions also to the performance of the municipalities, which proved to be a good instrument.

Q112 Jeremy Lefroy: First, I am thinking particularly of a KfW project that I saw over many years, involving rural water supply in northern Tanzania, which seemed to me to be a particularly effective project. One of the advantages was that it was a longterm project over many, many years-perhaps 10 or more. Do you think that there is a restriction in the kind of time scale over which DFID is able to interact, because it does not have a loan window and is constrained effectively by fouryear spending periods? We may have reached 0.7%, but no Government is bound by its predecessor, and therefore the next Government could cut aid spending drastically, and therefore aid programmes only really have a time window, at the moment, up to 2015. You could not undertake directly a project such as the one that KfW did, although you could do it through a multilateral. That is my first question: do you think that there is an advantage in the ability to deal with longterm projects in countries, perhaps, where the multilaterals are not particularly engaged? We have to admit that they are not engaged in every single country, much as we would like them to be.

The second question is really focused on what Chris was saying. From my own experience-obviously I declare the usual interest, having been involved in this area for many years-I would absolutely endorse what Chris was saying about the inability to get all the kind of finance and all the kind of legal stuff available in one place, to work with small social enterprises or small and medium companies. It is hugely timeconsuming to do that, and I wondered, Chris, if you saw that as an area that a development bank or a virtual development bank, or some kind of financial institution within DFID, could make a core competence, since it does not seem to exist anywhere else.

Dorothee Fiedler: I really think it is a big advantage that we have longer time frames for our projects. Even though politicians may only be elected for four years, we make commitments that go very far beyond that. For example, with KfW, we might plan a project that extends much further than those four years. We also have the advantage that technical cooperation via GIZ may come into that same project, which even makes it a little longer, so that after KfW has left, after maybe eight years, you may still have a part of technical cooperation that goes on and helps make sure that the results are really very good. I think our instruments are good.

Every two years, we have intergovernmental negotiations, and then we commit ourselves to projects, but those projects do not necessarily have to last for four years, or even less; they could go for much longer. Sometimes, we have a preparation period, which goes for maybe two years, and only then will it start. So this is one of the advantages, maybe, of our system: whoever is the Government, those commitments are for a much longer term, in both financial and technical cooperation.

Marc Engelhardt: What helps, also, is the attitude of the development bank. It does not take too long. Coming back to what Ms Bruce asked earlier, about whether we need more time, I think that, yes, sometime we need more time, and definitely more than four years. On the other hand, we should not wait for ages, and the attitude of the development bank that has to see results, after a certain time, and also the attitude of the recipient country that has to repay the loan-and has to pay, in most cases, or at least in our case, a commitment fee-can also speed up some processes. This has to do with the ownership issue mentioned earlier, so there a developmentbanking perspective may help to make things happen, at least after 10 years, and not after 20 or 30 years.

Dr West: From my side, I have the fortune of probably having some of the highestrisk capital under management around, through a foundation. However, having said that, investing in these and catalysing and creating some of these ventures has taken us a huge length of time. We are still partnering with some after 10 years. Patience is the name of the game here. It is right that those with the highrisk, earlystage capital play that patient, longerterm game. The question then is about who then takes them on. To answer your question, if you look in Africa, as an example of one region at the moment, most job creation is going to come through the SME sector. Of that SME sector, it will particularly be the "S"-the small enterprise. It is completely unbanked and unserviced.

Over a decade ago, we cofounded a partner, GroFin, that specialised in providing the skills and finance for that. It is doing well, and has now raised $170 million from various organisations around the world, including the EIB. It is doing it in a way that is financially viable, with measurable social outcomes. We are now at a point where the marginal returns from that business mean the credit committees at some of the DFIs are writing it off. There we have a dichotomy, where we need an intermediary of some sort that can provide some degree of subsidy to this, but equally accept the marginal returns of doing this highcost, highrisk business, to create the jobs in Africa that we all know are needed. It does not lend itself, as a vehicle, to balancesheet lending from a DFI. It does not have privateequitytype returns. It does not lend itself to a commercial bank that does debt finance, so we struggle where something has been nurtured and is doing really, really well, and many of the people that we had to go to, to raise the capital to support it, do not accept the levels of returns from it. That is the tragedy of this space: that we cannot then bridge it. That is not in any way to say that those institutions not accepting that level of risk are not right. It is just that they are fit for a different purpose than taking these social entrepreneurs through that bridging phase. That is what we really need: someone who can do that now. Whether DFID can do it, or whether it can help cocreate others like this or support the few that do exist, I think it is absolutely the right thing for a subsidy provider to be asking itself.

Tamsyn Barton: I would really like to come in on this point about the long term, because it is a very important part of the decisionmaking. The EIB was set up for the purpose of financing regional infrastructure to join up the European community. We take that beyond the EU, in what we do outside and to support to regional infrastructure, generally and particularly. We can lend, in exceptional cases, for up to 50 years. Obviously, there is a huge mismatch between the speed at which development really happens and the normal donor habit of doing threeyear projects, and that is a real problem. It is striking, whether it is in NGOs or in any grantmaking sector, that the interest in the evaluations dies off somewhat quickly after the end of the project. All the interest is in the design stage, so certainly that is a mismatch.

However, the difficult issue for DFID, if it does go into this, is that it is a commitment for 20 years, or more if it takes longer to get something off the ground. You were asking about the writeoffs that we have. Most of the writeoffs that we have are ancient heritage-preLomé portfolio from the 1980s and 1990s-and come back to haunt us many years later because of political changes or new risks. There is an awful lot of monitoring work to do for 20 years, whereas with a grant, normally, the file can be pretty much closed apart from for the evaluators, so the administrative costs of this are high, but it is a huge gap in the market. Everywhere we go, we are able, for example, to increase the terms at which the banking sector can lend, because they can only lend shortterm.

Chair: I think that is really what we are teasing out: does DFID stick to what it knows, or does it help to fill a gap between where the multilaterals are operating comfortably and simply giving grants? I completely accept the point. I think we have found both these sessions extremely interesting. They are confusing in a way, because they are contradictory, but that is fair enough. We are really trying to tease these things out, and I think we are quite a long way from settling on what we would choose to recommend, but then we have quite a little while to go on this. We have had interesting informal sessions, just to share this with you. For example, we had JICA last week, which was a very interesting session; we also had a session with the president of the World Bank. Other donors and multilaterals have said that they believe that if DFID had a little bit more flexibility, then the scope for productive cooperation between DFID and other agencies might expand. That is one area that perhaps we need to tease out: the extent to which that would or would not happen, and what mechanisms we need for it.

Can I thank you very much? It has been really interesting for us to hear how these things operate in other spheres, and also your thoughts about what DFID should be focusing on. Thank you very much indeed.

Prepared 11th February 2014