• The relevance and importance of aid is as great as ever. Official development assistance (ODA) is unique in being able to target poverty directly. But more ODA needs to be mobilised, better targeted to the poorest people and countries, and better focused on the right mechanisms, channels, sectors and modalities that build human capital and strengthen institutions.
• Aid continuing to stagnate should not be accepted. Donors can, and are, making new commitments. An additional US$1.5 trillion can be mobilised to support sustainable development if OECD DAC donors meet their 0.7% GNI pledges by 2030.
• The proportion of ODA being transferred into countries has fallen significantly such that less than two thirds of gross ODA is actually allocated to a particular country.
• The proportion and volume of ODA that is never transferred from a donor country is on the rise and stood at a peak of 18% total ODA in 2016, driven mainly by a four-fold increase in spending on refugee hosting since 2013.
• ODA is not benefiting the poorest populations nor the countries with the lowest government revenues proportionately to their needs. Just 35% of countryspecific aid (or 22% of all aid) goes to countries that account for three quarters of people known to be in extreme poverty. ^[1] Countries with government revenues above $4,000 per capita receive more than three times the ODA per poor person than countries with less than $400 per capita.
• ODA to the group of countries identified as being left behind has fallen by 6% since 2010, while ODA to all other recipients has risen by 32%.
• ODA focusing on specific vulnerabilities is not effectively targeting countries with the greatest needs. Climate change adaptation finance, for example, is not always going to countries known to be the most vulnerable.
• The ways in which ODA is delivered has seen a shift towards loans and away from channels that can empower countries to lead their own development agenda. Over the last decade, ODA loans have grown more than three times faster than other forms of ODA. Almost a third of loans to low income countries went to nations deemed to be in debt distress or at high risk of being so. Investments in general budget support have fallen and aid not channelled via the public sector is mainly implemented by international rather than local actors.
• ODA is not consistently targeting the sectors most critical to the poorest people. On aggregate, health and education spending is stagnating, while small volumes of aid to social protection is growing slowly and remains at just 1.4% of total aid. Nor does the distribution of sector aid always reflect the needs of the poorest people: in countries with the highest poverty levels, spending on education was US$1.2 billion lower than that given to developing countries with the lowest levels of poverty.
• Development cooperation from government providers outside the OECD DAC has seen significant increases in recent years, though to maximise its potential in the wider concessional financing landscape, improved accountability and transparency remain key.

Since the 1960s, ODA has been the principal resource used to stimulate development and alleviate global poverty by members of the OECD DAC. ^[2] It plays a unique role in the growing mix of resources available to developing countries as it is the only flow that has poverty reduction as its core purpose and therefore has a critical role to play in reaching and helping the people furthest behind. ^[3]

In recent years the development finance landscape has changed rapidly. While national public and commercial resources remain the primary source for development investments for many countries (see Chapter 3), total volumes of finance are increasing and the range of international finance available to many countries has expanded. ODA is a comparatively small part of this growing mix and so many donors are starting to re-evaluate what its role can and should be.

First and foremost, ODA will continue to be needed and it is important to move away from pervading pessimism about ODA stagnating with little scope to increase: this is misplaced and undermines efforts to end poverty. As Chapter 3 argues, aid cannot be automatically substituted by other types of finance and the same outcomes achieved. Crucially therefore, donors need to be more strategic about how ODA is spent. The scale and variety of non-ODA resources at the global scale creates more space for ODA to do what it does best – reach the people at greatest risk of being left behind. ^[4]

Apart from a decline during the 1990s, ODA volumes have generally followed an upward trend since 1960. In fact, in 2017, net ODA from DAC members was four times higher in real terms than in 1960. While ODA fell slightly in 2017 – the first drop in ODA levels since 2011 and 2012 when ODA fell in the aftermath of the global financial crisis – such crises come and go and should not undermine the counter-cyclical advantages of aid.

However, while ODA quadrupled, the GNI of DAC members rose more than six-fold in real terms between 1960 and 2017. This means ODA has not increased as quickly as growth and the share of DAC donors’ national income spent on ODA has fallen from just over 0.5% in 1960 to just over 0.3% in 2017.

Major donors are taking steps to address this and setting new, specific and time-bound commitments to increase proportions of national income spent on aid. ^[5] France has committed to spending 0.55% of GNI on ODA by 2022 and reaffirmed its goal of reaching 0.7% by 2030. ^[6] These commitments are in line with the 1969 Pearson Commission ^[7] proposed target of 0.7% of donor GNI to be reached “by 1975 and in no case later than 1980” – a target taken up in a UN resolution on 24 October 1970. OECD DAC members have generally accepted the 0.7% target for ODA, at least as a long-term objective, except for Switzerland (which only became a member of the UN in 2002) and the US (although the US did state that it supported the general aims of the UN resolution).

Despite this near-unanimous support from donors 47 years after the adoption of the UN resolution, just five DAC members disbursed at least 0.7% of GNI as ODA in 2017. Eighteen DAC members had ODA levels that were 0.3% of GNI or less and ten of these gave less than 0.2%. Australia, Canada, the US and France have seen the most significant falls in ODA as a percentage of their GNI since 1970.

While on aggregate the GNI/ODA proportions of the original DAC members were the same in 2017 as in 1970, newer members of the DAC give much lower proportions and this accounts for the marginal fall overall (0.31% in 2017 compared with 0.33% in 1970). These newer donors joining the global development effort are typically newly industrialised countries without a long tradition of involvement in international development. Recently joined DAC members from Central and Eastern Europe typically have small bilateral aid programmes and a significant proportion of their ODA comes from their contributions to the EU. The average amount of ODA as a percentage of GNI for donors who joined the DAC after 1970 is just 0.19% in 2017.

Low ODA to GNI ratios have significant implications: if all DAC donors had dispersed 0.7% GNI then net ODA would have been US$318 billion in 2017 – twice the current figure of US$144 billion. ^[8] Crucially, if all DAC members increased their ODA/GNI percentage year on year (so by 2030 they reach 0.7% GNI), US$1.5 trillion additional ODA would be disbursed between 2018 and 2030 compared with ODA/GNI percentages remaining at current levels. The impact that could be had if all donors followed those already reaching or pledging their commitment to 0.7% GNI by 2030 should put real impetus on other donors to do the same.

ODA accounts for a wide range of activities, expenditures and investments and not all aid actually goes to developing countries – whether it never leaves the donor country (a non-transfer, such as, for example, debt relief, costs for hosting students from developing countries, building development awareness, some administrative costs and in-donor refugee costs), or is not allocated to specific countries (a transfer, but not to countries, for example investments in research, global initiatives, and regional aid).

The share of ODA allocated to countries has fallen rapidly in recent years and in 2016 US$54 billion, nearly a third (32%) of gross ODA was not reported as going to a specific country. A further US$6 billion was reported as being allocated to specific countries but was in a form that resulted in no transfer of resources to those countries (e.g. debt relief). Large proportions of ODA not directly allocated to countries is a relatively new phenomenon, and before 2007 this form of ODA rarely accounted for more than 20% of the total.

When coupled with a significant volume of ODA being spent in donor countries (non-transferred aid), a worrying trend can be seen. ODA going directly to countries is falling as a proportion of total ODA – with non-transfer and non-country allocated aid compounding one another. In-donor refugee costs are the largest factor causing this – up from just over 1% of total ODA in 2006 to almost 10% in 2016.

Of the US$166 billion in total gross ODA disbursed by DAC members and multilateral bodies in 2016, just US$107 billion was allocated to specific countries and in a form that resulted in an actual transfer of resources from the donor. This accounted for 69% of ODA in 2013 and had declined to 65% by 2016.

In 2016, US$30 billion of ODA was not transferred from donors and most of this (US$24bn) was not allocated to a specific recipient country. Indeed, the amount of ODA spent in donors’ countries and not reported as allocated for a specific country has risen sharply, from 9% of ODA in 2013 to 15% in 2016, driven by an almost four-fold increase in spending on refugees in donor countries. The proportion of ODA that never left the donor country rose from 15% of total gross ODA in 2013 to 18% in 2016. These trends need to be reversed if ODA is to address the needs of the poorest people first.

The ODA that does result in transfers to specific countries is not always targeted at the poorest places. While the largest proportion of ODA allocated to specific countries goes to sub-Saharan Africa (37% in 2016), few of the countries in this region receive the largest amounts of ODA overall and only Ethiopia is one of the largest eight recipients of aid (US$4.2bn). The others, apart from Turkey, are all in Asia and the Middle East: India (US$5.3bn), Afghanistan (US$4.1bn), Viet Nam (US$3.8bn), Pakistan (US$3.6bn), Bangladesh (US$3.2bn) and Syria (US$2.9bn).

Looking at the prevalence of extreme poverty ($1.90 per day) ^[15] across countries, the amount of ODA received by countries with larger populations of people in extreme poverty is nowhere near proportionate. The countries that contain 75% of the world’s poorest people ^[16] received 35% of the ODA disbursed in 2016, and countries with less than 1% of the world’s poorest people received 25% of ODA.

When looking at ODA per poor person, it is far from shared equally among people in extreme poverty. In countries with less than 5,000 people in extreme poverty, it is over US$230,000 – more than 3,500 times higher than the ODA per poor person in countries with more than 10 million people in extreme poverty. Considered in this way, ODA allocations appear regressive – with proportionately more resources per poor person going to the countries with the least poverty.

Similarly, countries with the lowest levels of government revenue are not proportionately benefiting from ODA

There is significant variation in the government resources available in developing countries. Some ODA recipients only have a few hundred dollars or less per person to spend on providing services, infrastructure and governance in their countries; others have thousands of dollars per person in government revenues. While more ODA does go to countries with less government resources, ODA disbursed to the low-revenue countries is not proportionate to the level of poverty in these countries. Countries with revenues of between $2,000 and $4,000 per capita get seven times as much ODA per poor person as countries with government revenues of less than $400 per capita. Countries with government revenues of above $4,000 per capita get three times as much ODA per poor person as countries with the lowest government revenues. This reflects a similar picture to ODA disbursed based on extreme poverty levels.

ODA to the least developed countries has not kept pace and has flatlined in recent years

Another indication that ODA is not being targeted at countries with the greatest need is donors not prioritising the least developed countries (LDCs). LDCs are described by the UN as: “lowincome countries confronting severe structural impediments to sustainable development... highly vulnerable to economic and environmental shocks and have low levels of human assets”. ^[17] While ODA to LDCs grew fairly sharply in the early years of the century, this growth then slowed and has now flatlined, falling from 35% of ODA in 2010 to less than 30% in 2016.

ODA is not being sufficiently targeted to countries identified as being left behind

ODA allocations should also consider where poverty is likely to persist in the future. The countries referenced in Chapter 1 as at risk of being left behind currently account for 23% of global poverty, and on current projections will be home to around 80% of the world’s poorest people by 2030. Yet recent growth in ODA has mostly bypassed these countries. In fact, gross ODA to this group of countries has fallen by 6% since 2010, while ODA to all other recipients has risen by 32%.

There is not much variation between countries classed as having medium ^[19] or low ^[20] levels of vulnerability. They received an average of US$49.8 million and US$42.2 million respectively. The countries with the greatest ^[21] levels of vulnerability receive relatively little – on average US$65.3 million per country. Some of the most vulnerable countries received the lowest amounts – including Eritrea (US$11.3m), Chad (US$24.6m), the Solomon Islands (US$15.9m) and Micronesia (US$0.2m). Furthermore, certain countries with high levels of vulnerability and relatively high levels of extreme poverty are not prioritised, for example Guinea-Bissau and Central African Republic.

Such gaps and inequality in the distribution of adaptation-related ODA, considering patterns of vulnerability and poverty, reveal significant opportunity for DAC donors and multilateral institutions to improve the targeting of their climate-specific resources.

Various means (often described as modalities or instruments) are used to deliver ODA. Modalities describe how aid is managed and disbursed and can influence what, where and who is targeted. ^[22] The impact of changes in modalities, particularly on those people and places furthest behind, can be hard to see immediately. Yet there is little doubt that how aid is used matters almost as much as what aid is used for, with who and where . The global figures may also obscure important distinctions and variations at the level of each donor. Here the use of ODA as loans demonstrates those variations and the potential long-term impact on the poorest people.

A significant dividing line in types of modalities is how concessional aid is – grants are 100% concessional while other forms such as ODA loans or other financial instruments vary in their concessionality, requiring some form or proportion of repayment and can thus create debt.

Just as gross ODA has fallen for a number of the poorest countries since 2010, so grants directly to projects, a key modality for such countries, have also declined. They have fallen as a proportion of total ODA – from 42% in 2010 to 36% in 2016, as loans and non-transfer aid have taken a larger share of ODA. And when only grants allocated to specific countries or regions are considered (thus excluding those that have no specific geographical focus), volumes have flatlined. For LDCs they have actually fallen in real terms by 12%, at a time when lending to such countries has risen (see below). ^[23]

Core grants to NGOs have risen quite rapidly in recent years, up by over 50% since 2012 to US$3.6 billion in 2016. The great majority of this spending (US$3bn) was targeted at international NGOs and NGOs based in donor countries. Core grants to specific-purpose funds and pooled funds have grown faster than total ODA – by 32% since 2010 to US$18.4 billion in 2016, 11% of total ODA. However, as discussed in depth in this chapter, loans have also risen significantly.

Some equity investments in companies in developing countries, usually made by development finance institutions (DFIs), are eligible to be counted as ODA. These investments peaked in 2010 and 2011 at US$1.6 billion in both years. From 2012 to 2014 gross equity investments reported as ODA stood at US$1.4 billion per year. Since 2014, tracking the use of these instruments has been complicated by some donors choosing to report capital sums into DFIs from donor governments as ODA, rather than the actual investments made by the DFIs. In 2014, the UK reported US$445 million of equity investments from CDC Group (the UK DFI) as ODA. In 2015, CDC ceased to report the detail of its investments to the OECD; instead the Department for International Development reported a US$688 million investment into CDC as an ODA grant. As a result, total equity investments reported as ODA fell to US$1 billion in 2015 and 2016.

Another important dividing line is whether the modality results in a transfer of real resources from the donor country and, as noted earlier, non-transfer ODA has been on the rise in recent years. Finally, some forms of ODA do not result in a transfer of ‘cash’ resources but a transfer of knowledge, experience and capacity, for example technical cooperation. Since 2007 this has fallen by 2.7% as a proportion of total ODA but grown in dollar value by 16.7%.

The most significant trend over the last decade has been the increasing use, by some large donors in particular, of concessional loans as a way of delivering aid. Loans now make up almost a quarter of total ODA, growing three times faster than other ODA modalities combined from 2007 to 2016 (Figure 2.15). This has meant a rise from 15% of gross ODA in 2007 to 24% in 2016 despite compelling evidence on the negative or counter-productive effects on the poorest countries, for which low absorptive capacity and unsustainable debt can seriously affect their current and future economic growth. ^[24] , ^[25]

Just four countries – Japan, Germany, France and Korea – supply 98% of ODA loans from DAC member countries. Japan has long been the lead provider of ODA loans among the DAC donors, however Germany, France and Korea have rapidly expanded their lending programmes between three and five-fold between 2007 and 2016. Institutions of the EU, the only multilateral donor that is a member of the DAC, increased levels of ODA loans by a factor greater than 40 over the same period.

A substantial amount of lending goes to countries deemed in debt distress or at risk of being so, indicating donors are not consistently showing due diligence

Crucially, the rise in loans in recent years has been evenly spread across countries of different income groups, with no significant changes in the proportion of loans going to any particular income group between 2007 and 2016. This means many countries with high poverty levels and low government revenues are receiving large amounts of ODA loans. Looking further at low income countries, over 30% of the loans going to these countries were assessed by the International Monetary Fund as being lent to countries in or at high risk of being in debt distress.

In Africa significant volumes of loans went to countries at high risk of debt distress in 2016, including Ethiopia (US$1.5bn) Ghana (US$656m) and Cameroon (US$391m). Mozambique, a country rated as actually being in debt distress, received loans totalling US$417 million in 2016.

As well as being delivered in a variety of modalities, ODA is channelled by and through a complex network of agencies and actors. Most donors have a dedicated aid agency or development specialists in foreign affairs ministries to disperse ODA. However, certain donors such as the UK are making a conscious effort to push spending increasingly via other departments. ^[26] ODA being administered by such ‘other’ government ministries accounted for 30% of ODA reported by DAC donors in 2016. Interestingly, small amounts of ODA from some donors also come from local, regional or municipal government bodies – just over US$2 billion in 2016.

With growing emphasis on the role of the private sector in promoting development, it is expected that bilateral DFIs may in future account for a larger proportion of ODA (see Chapter 3). At present, however, their contribution is fairly small – accounting for just under US$6 billion of ODA in 2016.

Most ODA is channelled via public sector institutions, but significant portions bypass local or national actors

Over half of all ODA is channelled via public sector institutions. Of this, a third (33%) is implemented by the donor government directly, with a larger 57% via the government of the recipient country. For aid not channelled via the public sector, the great majority is implemented by international rather than local actors. ^[27] International multilateral agencies account for 17%, the second-largest amount of ODA being channelled, two thirds of which is via the UN. NGOs channel a similar amount (12%); almost all of this is channelled through organisations headquartered in donor countries. Finally, the private sector only accounts for 4% with over half (54%) via firms in the donor country. These proportions of total ODA have remained almost static since comprehensive data became available as channel of delivery began to be reported by donors in 2009.

Donors and implementing agencies have committed to providing more humanitarian funding as directly as possible to local and national actors who are regularly the first to respond in humanitarian crises. To date, volumes of reported international humanitarian assistance are well below the 2020 target agreed in the Grand Bargain of 25% of all assistance. ^[28] In 2017, 3.6% of international humanitarian assistance reported to UN Office for the Coordination of Humanitarian Affairs (OCHA)’s Financial Tracking Service was delivered directly or through one intermediary to local and national responders. ^[29]

There are variations in aggregate totals when looking at different sets of countries, with less ODA to countries at risk of being left behind being channelled through government institutions than is the case for other countries. In fact, the proportion of ODA channelled via the public sector in countries being left behind has declined noticeably since 2011. Between 2008 and 2011 around 50% of ODA to countries being left behind went through the public sector, but this proportion declined in each year from 2011 to 2016. ODA to this group of countries is more than twice as likely to be channelled through a multilateral body or NGO than other ODA. To some extent this is unsurprising as the poorest and most vulnerable countries may have weaker government institutions, if present at all, via which ODA can be administered. However, this is not the case for all countries being left behind, some politically fragile countries may have functioning subnational institutions. Yet their low levels of government revenue make ODA channelled via their government institutions all-the-more vital if they are to be empowered to lead their own development agendas. ^[30] There is clearly substantial scope for donors to work better through local institutions – state and non-state. ^[31]

ODA supports a wide range of activities and health, infrastructure and governance-related sectors consistently receive the largest amounts. In 2016, these three sectors combined accounted for more than a third of gross ODA disbursements (health 13%, infrastructure 12% and governance 10%).

Humanitarian aid has risen notably faster than total gross ODA, with proportions of total aid almost doubling from 6% in 2006 to 11% in 2016 (see Chapter 5 for a discussion of implications). Another significant rise is in spending on refugees in donor countries, which typically accounted for around 2 to 3% of total ODA before 2014 but has increased almost four-fold to US$16 billion in 2016, 10% of total gross ODA. ^[32]

Focusing in on development ODA, a broad range of sectors receive significant amounts of aid, with notable trends related to a number of the sectors that should be the greatest priorities to ensure no one is left behind. Spending on health has stagnated since 2013 after a 250% increase between 2002 and 2013. ODA to education has also remained fairly static, only rising by 6% in real terms between 2010 and 2016. This means the share of total ODA going to education slipped from 8.4% to 7.3% over that period.

Other social services (social services excluding health and education) saw the slowest growth, with spending rising well below the rate of ODA as a whole. In fact, spending on this sector was over 20% lower in 2016 than a peak year in 2008. Although there is no data specifically on ODA to social protection, ODA to social and welfare services, a subdivision of other social services that largely comprises spending on social protection, is tracked. ODA spent here peaked in 2010 at US$2.6 billion before dropping to US$1.7 billion in 2012. Since then such aid recovered to US$2.4 billion by 2016 – although still only 1.4% of total ODA. Data on ODA support for social protection programmes may improve in coming years as the OECD has updated its data to allow donors to track spending on social security, pensions and other social protection schemes in the form of cash or in-kind benefits.

Finally, general budget support ^[33] has been cut substantially – falling from over 5% of ODA in 2009 to just 1.7% in 2016. This is particularly pertinent since the Busan Partnership for Effective Development Cooperation made country ownership one of its stated principles for effective development. ^[34] Budget support was widely viewed as a key mechanism through which aid would be delivered in such a way as to increase country ownership of development outcomes. ^[35]

When focusing in on different countries, there are clear variations from the aggregate trends on sector allocations. Many of these variations are to be expected. For example in countries with high levels of extreme poverty (more than 20% of the population living on less than US$1.90 per day) 30% of developmental ^[36] ODA is spent on health and 13% on infrastructure projects. In countries with lower levels of poverty (less than 5% of the population living on less than US$1.90 per day) the reverse is true, with 27% of developmental ODA spent on infrastructure and less than 5% on health.

However, looking at sectors that are particularly significant to ensuring the people furthest behind are lifted out of poverty, there are some surprising trends. ODA spending on education in countries with high poverty levels was US$1.2 billion lower than the amount given to countries with the lowest levels of poverty. Spending on water and sanitation projects is also higher in low poverty developing countries (US$2.5bn) than in high poverty countries (US$2.0bn).

With increasing focus on donors attracting private sector investment in developing countries, it is also important to acknowledge trends in ODA spending on business and industry in developing countries. At present countries with the lowest levels of poverty are receiving large amounts of ODA targeted at these sectors, meaning investments are primarily targeting where there is the least risk and greatest chance of receiving a profitable return. ^[37] This could call into question such ODA allocations on the basis that billions of dollars of donor funding may be going to the same areas that could instead benefit from private investment. If this is the case, it should incentivise donors to redeploy ODA to countries and sectors that are otherwise underfunded.

Countries being left behind receive the least investment in infrastructure and business and low amounts to education – with potential ramifications for human capital

Focusing in on the countries being left behind, allocations are largely as expected. Humanitarian aid and health dominate. They comprise almost half of gross ODA received and account for the largest increases (see Figure 2.25). Meanwhile the proportion of ODA allocated to infrastructure, business and industry is substantially lower than for other countries. Similarly, the trend in education follows the same surprising disparity seen when comparing countries based on income poverty levels. In fact, the amount being allocated to education in countries being left behind is falling.

Another worrying trend is the decline in general budget support. As reflected globally, the relative significance of this modality has continued to fall, and accounts for one of the largest proportional decreases, from 14% in 2002 to just 3% in 2016.

A number of countries being left behind are characterised by fragility and political insecurity. However, despite large and growing volumes of aid to humanitarian assistance, aid for conflict prevention and peacebuilding is low, accounting for US$1.1 billion, or 3% of aid to countries being left behind in 2016. This reflects trends of funding to fragile states more widely, where commitments in these areas have largely stagnated since 2010. Conflict prevention now accounts for just 2% of aid to fragile states, and peacebuilding just 10%.

Levels of development cooperation from government providers outside the OECD DAC, including South–South cooperation ^[38] and ODA from other government providers, have increased significantly since 2000 and stood at US$25 billion in 2016. It is therefore important to acknowledge these volumes alongside ODA as, while classified differently, their purpose is aligned. Figure 2.26 shows development cooperation from three types of provider: developing countries not reporting ODA to the OECD (South–South cooperation), developing countries reporting ODA to the OECD (South–South ODA) and non-developing countries which are not members of the OECD DAC, but which report ODA to the OECD (ODA from other government providers).

South–South cooperation and South–South ODA combined have grown in particular from just US$1.7 billion in 2001 to US$16.2 billion in 2016, of which Turkey (US$6.9bn), China (US$6.6bn) and India (US$1.8bn) were the largest donors. ^[39] ODA from other government providers – from developed countries reporting to the OECD but not DAC members such as the United Arab Emirates, Kuwait and European donors like Bulgaria – has also increased overall to US$8.4 billion largely due to spending by Gulf donors.

Data remains sparse, patchy and incomplete for much South–South cooperation and volumes may be underestimated. There are also significant challenges in assessing where it is going, what it is being spent on and what impact it is achieving. While volumes are smaller than ODA or other resources such as foreign direct investment, they are clearly rising and these actors are likely to play an increasing role in the future.

While Southern providers consider such flows as unique relationships based on solidarity and expressed through collaboration rather than reflecting a ‘donor–recipient’ transaction, this does not mean that accountability and transparency are any less important. Understanding better the contribution such actors are making, or could make, will be important to identify where they can most add value to countries’ needs in relation to other types of assistance. It could also ensure they are mobilising these flows effectively in support of Agenda 2030 in a transparent and accountable way.