Financing The Sustainable Development Goals: Lessons From Government Spending On The MDGs

RESEARCH REPORT
FINANCING THE SUSTAINABLE

DEVELOPMENT GOALS
Lessons from government spending on the MDGs
2015 REPORT
Government spending is falling one-third short of MDG needs and the SDGs
will require at least US$1.5 trillion extra a year. Based on lessons from
tracking country budgets, this report recommends how the SDGs should be
financed: by doubling tax revenue, through radically overhauling global tax
rules; doubling concessional development cooperation, and improving its
allocation and effectiveness; and raising US$500 billion in innovative public
financing. In addition, all spending must be dramatically reoriented to fight
inequality, and needs to be much more transparent and accountable to the
worlds citizens.
Development Finance International (DFI) and Oxfam have collaborated on this Government Spending
Watch report to share research results and contribute to public debate on public financing for
development, financing the SDGs, and the Financing for Development processes. The report does
not necessarily reflect Oxfam or DFI policy positions. The views expressed are those of the authors.
ACKNOWLEDGEMENTS
This report was written by Matthew Martin and Jo Walker. The research and data team who produce the GSW
data consists of Maria Holloway, Lance Karani, Jeannette Laouadi, David Waddock and Jo Walker. Special
thanks go to David Waddock for acting as interim data team lead. Rebecca Simson and Paolo de Renzio played
a key role in the research underlying Section 5. Earlier data and research contributions, without which GSW
would not be possible, were made by Richard Watts, Alison Johnson, Katerina Kyrili and Hannah Bargawi.
Key contributions to the compilation and interpretation of the data were made by budget officials in the 66
countries for which GSW has compiled data too many to mention individually.
Oxfam provided valuable comments and input, including from Paul OBrien, Max Lawson, Greg Adams, Ceri
Averill and Claire Godfrey.
Finally, DFI and Oxfam are most grateful to the Global Campaign for Education, the International Budget
Partnership, Save the Children International, UNESCO, UN Women and WaterAid, who have commissioned
work using GSW data during 201315, which helps us to continue to develop and deepen our analysis and
understanding of trends in government spending.
2 Financing the Sustainable Development Goals: Lessons from Government Spending on the MDGs
EXECUTIVE SUMMARY
THE AIM OF THIS REPORT AND ITS DATA
Six months remain until the end of the Millennium Development Goals (MDGs) the framework
used to measure global development progress since 2000. Government spending is a key way for
countries to achieve the MDGs. Yet astonishingly, throughout the MDG period, the international
community has conducted no comprehensive monitoring or analysis of spending.
Since 2009, Development Finance International (DFI) has compiled the latest data on MDG spending
through investigative data gathering with government officials across seven key sectors: agriculture
and food, education, environment, health, social protection, water, sanitation and hygiene (WASH)
and womens rights. These data have been published in reports and on the Government Spending
Watch website (www.governmentspendingwatch.org), and have helped to increase spending levels
and transparency.
The 2015 Government Spending Watch (GSW) report aims to take stock of progress on MDG
spending (and less desirable spending on debt and defence) as the world moves towards the finish
line for the MDGs. However, 2015 is also the year when the international community will agree a
new set of more ambitious Sustainable Development Goals (SDGs) for 201630, and will decide how
financing for development (FfD) should fund them. This report aims to influence these decisions, by:
Analysing whether current spending trends will suffice to achieve the SDGs;
Examining how spending has been funded since 2008, and what needs to change in FfD;
Identifying what needs to be done to ensure that government spending combats inequality; and
Assessing how ready countries (and the international community) are to track SDG spending,
and to hold governments and funders accountable for its levels and results.
MDG AND SDG SPENDING

Section 2 of this report looks at recent spending on the MDGs and the implications for potential
spending on the SDGs. It demonstrates that government spending in developing countries rose
rapidly in 201214, but revenues did not. This has led to growing deficits, resulting in sharp increases
in debt service. As a result, debt service is crowding out MDG spending in 21 of 66 countries, and
MDG spending has not risen to the same degree as overall government spending because debt
service has absorbed 40% of the extra spending, and infrastructure 35% only 25% of additional
finance has been allocated to MDG sectors.
In terms of specific sectors, based on 2014 data, what is the country spending performance and
what implications does this have for spending on the SDGs?
Agriculture: Only 1416% of countries are meeting financial targets, and average spending is
only halfway to the target and has fallen since 2012. The SDGs for zero hunger and sustainable
agriculture will require doubled spending, and tracking of what is anti-hunger and sustainable.
Education: Only 1922% of countries are meeting Education For All (EFA) targets, though
average spending is 80% of the target and most countries are increasing spending levels.
Education is the sector closest to reaching its MDG targets, but the SDGs lifelong high-quality
learning agenda will require US$161 billion more.
Environment: This sector has no MDG financial targets and received less than 1% of spending,
and the average is falling. The environment and climate change SDGs require US$261 billion
more a year, and there is an urgent need to set targets and ensure that all spending is
sustainability-compliant.
Health: No African country is meeting its targets, though 40% of all countries are meeting World
Health Organization (WHO) per capita spending targets. Average spending is only half the
targeted level, and recent trends have been mixed. Universal free healthcare will require an
increase of US$5080 billion, and a major effort is needed to monitor spending split by disease
and beneficiary group.
Social protection: Only Timor Leste meets any of the international finance targets on social
protection: across all countries the average spending is less than 1% of GDP though most
countries have increased spending in recent years. Higher spending will be vital to target zero
extreme poverty, full employment and decent work, and reduced inequality: even a cash
transfer programme would cost US$6590 billion extra a year. It will also be vital to invest in
capacity to disaggregate social protection spending by target and beneficiary.
WASH: Only 10% of countries are meeting targets, spending averages less than 1% of GDP and a
majority of countries are reducing spending as a percentage of GDP. Reaching universal access
to WASH will require US$24 billion, plus more to ensure sustainable water management, and
much better monitoring of spending for both purposes.
Overall, countries should be spending close to 60% of their budgets on the MDGs, but current
allocations are only 38% and falling. Total additional public spending needs for the SDGs (including
the sectors above as well as access to modern energy and infrastructure) could be as high as US$1.5
trillion a year. New sectors will also pose extra challenges for tracking spending, as much of it will be
implemented by state-owned enterprises or will use complex finance mechanisms such as public-
private partnerships (PPPs).
FINANCING THE SDGs

Section 3 analyses how the MDGs have been financed. It finds that government revenue currently
funds 77% of spending, which has been more stable, aligned with government priorities, balanced
between investment and recurrent and easy to implement than donor-funded spending. The SDGs
therefore require a massive step up in domestic resource mobilisation, which requires:
Major changes in international tax rules and practices to give fair treatment to developing
countries, including in current G20/OECD tax initiatives, by allocating taxation rights primarily to
source countries of raw materials, redesigning tax treaties and sharply reducing tax exemptions;
Major reinforcement of developing country capacity to receive, analyse, audit and supply tax
information, prosecute evaders and renegotiate contracts and agreements with corporations;
Agreement on inclusive global governance of cooperation in tax matters, via the FfD process and
a reinforced UN Tax Committee, to give developing countries equal decision-making power.
Whatever the increase in government revenue, it will not suffice to fund all the SDGs. Concessional
international public finance will still be vital. We need to mobilise US$1 trillion extra a year from:
All DAC donors recommitting to reach 0.7% of gross national income (GNI) in official
development assistance (ODA), by 2025, which could mobilise an additional US$250 billion a
year, bringing ODA to around US$400 billion;
SouthSouth cooperation providers accelerating increases in concessional flows. These rose by
300% during 200015, and a similar increase for the SDGs would raise them to US$80 billion;
Innovative financing of US$450550 billion a year, including taxes on carbon, bunker fuels and
air travel (US$250300 billion), financial transactions and currency (US$100150 billion) and
issuance of IMF Special Drawing Rights (SDRs) (at least US$100 billion).
There is also a need to focus 90% of concessional flows on the countries with the greatest needs
low- and lower-middle income countries, including 50% on countries in special situations (fragile
and conflict-affected, least developed, landlocked and small island states) channelling international
public financing for development where it is most needed because lower-income and special needs
countries can least afford to fund the SDGs using their own revenues.
Public non-concessional finance and PPPs will have a role to play in financing the SDGs. However,
rising debt burdens diverting spending from the MDGs, and many current national debt
restructurings (notably in high- and middle-income countries), limit the scope for non-concessional
finance, and it will be vital that the FfD agreement includes a dramatic strengthening of current
measures to prevent and resolve debt crises. There are also major doubts about the scope for and
potential impact of public-private blended finance in many SDG sectors, and especially in reducing
poverty and inequality. To maximise their contribution to the SDGs, we must set effectiveness
indicators for non-concessional public finance and blended and private finance.
The post-2015 context needs more rational accountability for sectoral aid allocation. This should
involve:
An overall target for allocating concessional funding to SDG sectors of 60% or higher;
Targets for global sector allocations of concessional funding, similar to those agreed for spending
by developing country governments but updated to reflect SDG funding needs;
Using markers for SDG spending in each sector in global and national aid databases; and
Screening all aid projects for positive impacts on the three pillars of the SDG agenda (e.g.
increasing income and gender equality and fighting environmental crises).
This report also makes recommendations to improve financing effectiveness in each sector.
Finally, the report stresses the need for dramatic improvement in the effectiveness of financing,
going beyond the 2011 agreement on a Global Partnership on Effective Development Cooperation,
by:
Targeting 80% of aid to reach developing countries, and 85% of this to be in their budgets;
Tracking these trends automatically in global and national aid databases;
Monitoring the investment/recurrent spending within aid to ensure sustainable spending;
Putting a much higher share of aid through government systems (as agreed in Busan).
SPENDING TO FIGHT INEQUALITY

Section 4 examines how public spending can be used to fight inequality. Income inequality has
increased sharply since 2000, undermining growth and poverty reduction efforts. The SDGs
therefore focus strongly on reducing inequality, with many goals designed to leave no one behind
and help the most marginalised. Moreover, failing to tackle inequality will add hundreds of billions of
dollars to the spending needed to end poverty: failure to tackle inequality will mean failure to
deliver the SDGs.
In the search for policy measures to combat income inequality, many analysts have identified more
equitable government spending and more progressive and redistributive tax systems as crucial
interventions. Investment in public services especially by covering their costs on education, health
and social protection can help by lifting the poorest people out of poverty. Much more analysis is
needed to ensure that government spending tackles all forms of inequality simultaneously: it will
require a laser-like focus on ensuring that all spending in government budgets (as well as all aid)
addresses inequality issues, as well as rebalancing all spending to tackle specific forms of inequality
and the design of more specific and focused programmes to reach the marginalised.
Some of the key areas which GSW has identified through its research are:
Increasing sector financing and spending envelopes to eliminate user fees, insurance systems
and profit-making provisions which discriminate against the marginalised;
Ending major inequities within sectors by increasing spending on pre-primary and primary
education, maternal/child and reproductive health, sanitation and smallholder agriculture; and
Combating geographical/spatial inequality by allocating more to areas which are making less
SDG progress, using equitable spending formulas, to overcome huge current inequities.
Finally, reducing inequality among beneficiary groups (by gender, age, ethnicity, disability or
sexuality) will require a more comprehensive and integrated analysis of all intended beneficiaries of
spending (building on gender-responsive and child-oriented budgeting experiences) and its impact
on inequality, as a basis for targeting reallocation to more marginalised groups.
Overall, the world is woefully ill prepared to use government spending to fight inequality. We need
much higher spending levels; more detailed data and impact analysis; capacity-building and political
buy-in to anti-inequality allocation formulas; leadership by governments and local stakeholders
rather than donors in fighting inequality; and close national-level monitoring of implementation to
combat the diversion of spending to the needs of more powerful interest groups.
ACCOUNTABILITY FOR RESULTS

The final crucial element to ensure that public spending fights inequality and reaches the SDGs will
be citizens in developing and developed countries holding governments and donors accountable.
This process begins with greater transparency, especially of data on spending and revenue/aid in
budget-related documents, allowing them to track increases in means of implementation for the
SDGs, which are in turn likely to accelerate SDG results.
This report makes many recommendations to increase transparency but section 5 shows that:
Current budget transparency is moderate but increasing rapidly, although it is lower for actual
spending, types of spending and donor financing, and for spending on new SDGs;
Transparency varies considerably by country, with 20% of 124 countries performing very well,
23% well, 25% moderately and 32% relatively badly;
Budget transparency makes tracking MDG spending and financing more feasible, by
encouraging governments to improve the quality and details of their data;
When accompanied by government will/capacity and parliamentary/civil society demand, and
sustained across the whole planning and budgeting cycle, transparency brings accountability
and leads to higher MDG spending allocations;
With similar conditions, accompanied by efforts to improve spending effectiveness, higher
allocations lead to higher actual spending, and in turn to better MDG results.
This evidence suggests a number of low-cost quick wins tailored to country circumstances to
increase accountability and results, as part of the data revolution needed to support the SDGs:
Publishing budget documents and data which governments already produce;
Improving data and documents on spending by making reporting gradually more disaggregated
and moving towards programme budgets; publishing regionally disaggregated spending;
publishing preliminary actual data more rapidly; and publishing budgets by beneficiary;
Improving data and documents on revenue by publishing breakdowns by tax, sector and size of
enterprise; analysing revenue losses from exemptions and incentives, and publishing lists of
companies granted exemptions; publishing and analysing national tax codes to prevent harmful
tax competition; analysing tax incidence and progressivity to combat inequality; and publishing
tax revenues mobilised/exemptions received by donor projects;
Improving and publishing data on aid and other budget financing, by making the International
Aid Transparency Initiative (IATI) and national budget aid systems compatible; reporting
automatically via Development Assistance Committee (DAC)/IATI and national systems whether
aid is on the national budget; enhancing efforts to collect data from non-DAC funders including
SouthSouth partners and civil society organisations (CSOs)/foundations; and publishing and
tracking all loan agreements and off-budget contingent liabilities such as PPPs.
However, transparency on its own will not suffice to ensure accountability. This also requires:
Dramatically scaling up capacity-building support to governments to improve budget data,
through low-cost interventions designed to achieve the above quick wins; and to parliaments
and civil society to demand more transparency and accountability;
Setting SDG monitoring goals and mandating relevant UN agencies to monitor budgets and
means of implementation (spending, revenue and aid) for the SDGs within three months of the
end of each calendar year, through processes similar to those used by GSW;
Governments agreeing national SDG-compatible development plans and financing compacts
with their development partners, against which all can be held accountable by citizens for
results.
The recommendations in this report set out a detailed and ambitious Financing for Development
agenda bold enough to match the ambition contained in the SDGs. Without this level of ambition,
many of the SDGs will be dead at birth, as countries fight over a stagnating pool of tax revenue and
concessional resources and stumble forward in semi-darkness as to how much is being spent on the
SDGs. To reach the world we want, the recommendations in this report must be adopted.
1 INTRODUCTION
Six months remain until the end of the Millennium Development Goals (MDGs) the key objectives
used to measure global development progress since 2000.1 Government spending is a key way for
countries to achieve the MDGs. Yet astonishingly, throughout the MDG period, the international
community has conducted no comprehensive analysis of whether government spending is high
enough to achieve the goals.
Since 2009, Development Finance International (DFI) has been compiling the latest data on MDG
spending across seven critical sectors, seen as particularly important for the achievement of the
goals: agriculture and food, education, environment, health, social protection, water, sanitation and
hygiene (WASH) and womens rights.2
In 2013, these data were analysed to give an up-to-date and unique analysis of spending across
these seven key sectors in 52 countries towards achieving the MDGs, in the Government Spending
Watch 2013 report.3 The report gave a clear message: that spending was falling way short of MDG
needs, and had stagnated or fallen during the period 200913.
The 2015 Government Spending Watch (GSW) report aims to update this analysis and take stock of
progress on MDG spending, as the world moves towards the finish line for these goals. It extends the
2013 dataset and analysis by:
Updating data to cover 201415 budgets and, where possible, actual expenditure levels;
Expanding coverage to 66 low- and middle-income countries, for which budget spending data on
the MDGs are readily available (up from 52 in 2013);4
Comparing MDG spending with less desirable spending on defence and debt service.
All these additional data are available free on the GSW website for all to use.5
However, 2015 is also the year when the international community will agree on a new set of more
economically, socially and environmentally ambitious Sustainable Development Goals (SDGs) for the
period 201630, with a strongly focus on fighting inequality. In July, leaders will also meet in Addis
Ababa to decide how financing for development (FfD) should fund these goals.6 This report has been
launched to coincide with these processes, in hope that the analysis will help to inform the
discussions by giving solid evidence to:
Analyse whether current spending trends will suffice to achieve the SDGs;
Examine how spending has been funded since 2008, and what needs to change in FfD;
Identify what needs to be done to ensure that government spending combats inequality; and
Assess how ready countries (and the international community) are to track SDG spending and to
hold governments and funders accountable for its levels and results.
The rest of this report is organised into an additional five areas of analysis. Section 2 looks at
government spending on the MDGs, assessing for each sector whether countries are meeting
spending targets or needs and the additional spending that will be needed for the SDGs. Section 3
looks at how this spending is being financed and the influence of external financing on spending
patterns. Section 4 looks at how ready spending is to combat inequality (as a critical priority in the
SDGs). Section 5 looks at how greater transparency and accountability of government spending and
financing could be vital to achieve the SDGs. Section 6 concludes with practical suggestions to help
deliver the SDGs.
Figure 1. Countries covered by the GSW database
Note: This map shows the data available for all 67 countries currently on the GSW website. The online database includes South Africa, but analysis of the countrys data was not completed in time to be included in
this report. The number of countries in the database is currently being incrementally expanded, with a view to having all low-income countries (LICs) and middle-income countries (MICs) online in the future. Figure
5.1 in this report shows the current expansion plans.
For more information, please visit: http://www.governmentspendingwatch.org
Box 1: The GSW website and database
The Government Spending Watch (GSW) website provides comprehensive, up-to-date data on
government spending on the MDGs. It was the first global, comprehensive and up-to-date
database: several other organisations publish MDG spending data, but with two years delay and
for only some sectors or countries.
The GSW dataset covers 66 low- and middle-income countries (see Figure 1). It covers seven
MDG-related sectors: agriculture, education, environment, health, social protection, water,
sanitation and hygiene (WASH) and womens rights, as well as data on defence and debt service
spending. The data cover planned and actual expenditure for 2008 to 2014.
Compiling the dataset required a lengthy exercise of investigative data-gathering from public and
semi-public budget-related documents. The data available in many of these make identifying
MDG spending very difficult, so DFI has worked with a network of government officials to
interpret and classify data using a complex but consistent methodology and ruthlessly excluding
7
data where they are unclear.
Data are disaggregated by type of spending (recurrent vs investment) and source of funding
(government vs donor). They are presented in national currency, US$ equivalents, constant and
current prices, as a percentage of GDP and total spending, and per capita. They can be viewed
as graphs or tables, or downloaded in Excel format.
As discussed in Section 4, some data are not available, and one of GSWs aims is to encourage
technical support to countries and support advocacy and campaigning, so that data are more
readily available. Due to lack of resources, data on the GSW website are currently limited to the
period 201214: all data will be made publicly available as soon as possible.
The website also contains a series of analytical research reports and briefings, some of which
focus on specific sectors or regions. The aim of these is to help policy makers in governments
and international institutions make the right decisions in allocating spending, by giving them clear
analysis of spending trends, comparing them with other countries and the promises they have
made and analysing case studies in detail to show how spending has affected progress towards
the MDGs.
The GSW website also contains a brief guide to campaigning and advocacy, to help civil society
and parliaments hold governments and donor agencies to account for spending on the MDGs
and other goals. This includes a summary of targets set and promises made by governments,
inspiring stories and videos on key campaigns which have increased spending and made a
difference to the lives of citizens across the world, and links to global, regional and national
coalitions working on increasing spending and transparency. The GSW site will continue to grow
in 2015, adding additional countries and 2015 budgets, a toolkit for influencing government
spending and financing, and more analysis. From 2016 it will be adapted to cover spending on
the newly agreed Sustainable Development Goals (SDGs).
GSW is jointly sponsored by Development Finance International (www.development-finance.org)

and Oxfam www.oxfam.org). These two partners are gradually building a wider coalition of
sponsors for this vital initiative. All comments and feedback, and expressions of interest in joining
this coalition, are welcomed. Please use the websites feedback page
(http://www.governmentspendingwatch.org/feedback-and-suggestions) to contact us.
11. Financing the Sustainable Development Goals: Lessons from Government Spending on the MDGs
2 GOVERNMENT SPENDING:
TARGETS, TRENDS AND SDG IMPLICATIONS
This section evaluates 2014 spending against internationally agreed MDG-related targets, drawing on
GSW data on 66 low-income and lower-middle or middle-income countries.8 It begins by assessing broad
overall trends in government spending and financing across these countries. It then looks at the
proportion of government spending that is going to the MDGs and to other less desirable spending on
debt service and defence. Next it analyses six key sectors,9 comparing spending levels with targets set to
meet the MDGs, analysing recent trends and looking at additional spending and monitoring needs for the
SDGs. Finally, it summarises progress across all six MDG sectors, and looks at spending needs for
additional SDG sectors.
TRENDS IN OVERALL SPENDING AND FINANCING

Figures 2.12.4 show trends in government spending for the 66 GSW countries and offer Rising spending
explanations on trends in spending.10 On average, spending has risen by 3.2% of GDP and deficits
since 2008 (Figure 2.1), with a sharp increase in 2009 largely caused by efforts to
stimulate economies to combat the global economic crisis, followed by stagnation in have been
200911 and a renewed rapid rise in 201214. funded mainly
by borrowing
This spending increase has been only partly matched by a rise in budget revenues (Figure
2.2). Revenues fell sharply in 2009 as a result of the global crisis, rose consistently in
201012 but have stagnated since, and are only 0.9% of GDP higher than in 2008. As a result, budget
deficits (Figure 2.3) grew by 2% of GDP in 2009, fell back in 201013, but widened sharply to 8% of GDP in
2014. These deficits have been financed (Figure 2.4) by 5% of GDP in aid grants each year but also by a
16% cumulative increase in debt/GDP levels resulting in sharp increases in debt service.
OVERALL MDG SPENDING TRENDS
This overall rise in spending has not brought a similar increase in MDG spending.11 MDG Increases in
spending rose in real terms every year during the period 200813, but stagnated in 201214. overall
In addition, as discussed in more detail below, the international community has set targets for government
spending levels needed to reach the MDGs as a percentage of GDP or total government spending are
spending, and spending is falling well short of these targets. As Figures 2.5 and 2.6 show, since not being
2012 spending on the MDGs has stagnated at just over 11.5% of GDP, after rising sharply in matched by
2009 but falling during 201012. As a proportion of total government spending, MDG spending spending
also fell by 0.5% in 201214, continuing an earlier decline since 2008. MDG spending is not increases in
keeping pace with broader spending increases. What is crowding it out?
MDG spending
DEBT AND DEFENCE: CROWDING OUT MDG SPENDING?
While analysing government budgets, GSW researchers have identified a clear link between high levels of
debt service and defence spending and lower levels of MDG spending.12 Spending on these areas is
clearly crowding out spending on MDG sectors in a considerable number of countries. On average,
combined debt service and defence spending is reaching 21.3% of total government spending, compared
with 37.9% going to the MDGs. Almost three-quarters of this less desirable spending is going to debt
service, which averages 14.9% of total spending.13 Just over one-quarter is going to defence, which
averages 6.4%.14
Debt service crowds out MDG spending in one-third of countries

Spending on debt service is crowding out MDG spending in a considerable number of countries. Twenty-
one spend more than 15% of their budget on debt service, and 12 more than 20%. Six have to divert
more than one-third of spending to debt service, with the highest burdens occurring in Sri Lanka (69%),
Jordan (66%), Jamaica, the Occupied Palestinian Territory (48%), Colombia (45%) and Ghana (38%). All
these countries are spending more on debt than on the MDGs: Sri Lanka is spending 3.3 times as much on
debt service, Jamaica 2.5 times as much and Colombia and Jordan twice as much. In addition, average
debt service has risen since 2008 by 1.2% of GDP across the sample of countries, reflecting the high levels
of borrowing to fund government spending discussed above.
At the other end of the scale, some countries are spending very little on debt service, providing greater
fiscal space for spending on the MDGs. Fourteen countries are spending less than 5% of their budget on
debt service and, of these, 10 are able to allocate more than 30% of their budgets to MDG spending. In
particular, Kiribati is spending more than 70 times as much on the MDGs as on debt, Afghanistan 50 times
and the Solomon Islands 21 times. Ethiopia, Mali, Moldova and Rwanda also stand out as countries that
are close to meeting various MDG spending targets, due in part to low levels of debt service, which
accounts for less than 10% of their total spending.
Defence spending is worrying in at least four countries

Defence spending is crowding out MDG spending in far fewer countries. Of the 44 countries for which
data are available, only eight spend more than 10% of their budget on defence, and only two
(Afghanistan and Armenia) more than 15%. Unsurprisingly, Afghanistan is spending the most on defence
(22%): this represents 85% of the total amount spent on the MDGs. Armenia, Colombia and Sri Lanka also
spend around half as much on defence as on the combined MDGs.
On the other hand, six countries spend less than 3% of their budgets on defence (Cape Verde, Ghana,
Jamaica, Moldova, Papua New Guinea and Timor Leste), and Cape Verde and Moldova are spending more
than 20 times as much on the MDGs as on defence. In addition, defence spending has fallen slightly on
average (by 0.7% of GDP between 2008 and 2013), reducing its negative impact on MDG spending.
When the combined effects of debt and defence spending are taken into account, Sri Lanka emerges as
the country suffering the worst diversion from the MDGs, with the less desirable spending more than
four times as high. In contrast, Moldova spends more than 11 times as much on the MDGs as it does on
defence and debt service combined.
Overall, higher revenue, increased borrowing and a fall in defence spending have all provided additional
fiscal space for spending. However, only a quarter of the extra spending has gone towards the MDGs:
more than 40% has gone to debt service, and the remaining third to infrastructure reflecting developing
countries growing concern to accelerate growth as well as reaching the MDGs. The next section explains
trends for each MDG sector in terms of meeting internationally agreed targets, and demonstrates how
these overall trends are leading to stagnation in spending on these key sectors.
MDG SECTOR-SPECIFIC SPENDING
AGRICULTURE AND FOOD

MDG target 1.C aims to halve extreme hunger by the end of 2015. According to the latest figures, there
are 805 million chronically undernourished people in the world, about one in nine of the global
population.15 This is down by more than 200 million since the start of the MDGs.16 However, overall
hunger levels have fallen only from 23% to 13.5% in developing countries, narrowly missing the MDG
target of halving hunger,17 while one in four children around the world remains stunted due to chronic
malnourishment in the first two years of their lives.18
Are countries meeting their agriculture spending targets?

GSW tracks agriculture spending to map government efforts to meet MDG 1.C, as this tends Average
to be the most direct budgetary commitment and has proven capabilities in reducing spending on
hunger.19
agriculture in
In 2003, African governments meeting in Maputo committed themselves to spend at least Africa is 5%
10% of their budgets on agriculture within five years.20 Figure 2.9 shows how the 31 African half the amount
countries for which data are available performed in 2014. Five countries met the target needed to meet
(Burundi, Guinea-Bissau, Malawi, Mali and Niger), but this was down from six (including the 10% CAADP
Burkina Faso) in 2013 and nine in 2012. Malawi is the star performer, with an average of
spending target.
15% over the past five years. On the other hand, Angola, Ghana, Lesotho, Liberia and
Nigeria spend 2% or less of their budgets, and need to enhance efforts considerably.21
Though there is no formal target for other regions, GSW also assesses the performance of
non-African countries against a 10% target, given that the World Bank has found this level essential for
countries to transform their agricultural sectors.22 Figure 2.10 shows that, of 26 non-African countries,
only three (Bhutan, India and Nepal) are reaching 10%. Colombia, Ecuador and Jordan spend 1% or less.
This means that across all 57 countries for which data are available only eight, or 14%, are meeting the
target.
What are the recent trends in agriculture and food spending?
In 2013, the GSW report showed that after a small increase in the aftermath of the food price crisis of
200809, agriculture spending fell slightly in 201012.23 This has continued since 2012, with only 48% of
countries increasing spending as a percentage of GDP, and 41% as a proportion of total spending.
Average spending is only 5% of the budget (half the amount needed to meet the 10% target).
The SDGs require doubled spending and closer monitoring
The new hunger-related SDGs go way beyond the current targets: they commit not only to ending hunger
everywhere by 2030, but also to improving nutrition and promoting sustainable agriculture. Against a
backdrop of several interlocking crises including changing climate, increasing resource constraints and
population growth reaching zero hunger levels, sustainably, will require much higher spending on
nutrition and sustainable farming. Additional public spending needs are estimated at US$61 billion a year
(US$46 billion for zero hunger and US$15 billion for sustainable agriculture).24
They will also entail reinforced commitment to track spending more closely: by separating out
programmes which address hunger and nutrition and by analysing whether agriculture spending is
promoting sustainability. The Scaling Up Nutrition (SUN) movement is leading efforts to improve the
tracking of spending on nutrition,25 and GSW will work closely with them in future.
EDUCATION
MDG target 2.A commits that, by 2015, children everywhere will complete a full course of primary
schooling. In 2000, the Dakar Framework for Education For All (EFA) also set out six goals, and a broad
commitment to ensuring quality education for all.26 Progress has been patchy. There have been huge
strides in getting children into primary school; some improvements in youth literacy; a narrowing of
gender gaps; and more children than ever completing basic education. But 58 million children are still
missing out on primary school, expansion of access to primary education has been slowing,27 the global
early school primary leaving rate (25%) is the same as in 2000 and (due to insufficient focus on quality)
many children end basic education without basic literacy and numeracy.28
Are countries meeting their education spending targets?

There are two international benchmarks for spending on education: governments should spend 6% of
gross national product (GNP) and 20% of their overall budgets on education. Hence, GSW tracks total
education spending as a percentage of both overall spending and gross domestic product (GDP).29 The
2014 data show that only 11 of 59 countries with data available for 2014 met the percentage of GDP
target: Cape Verde, Kiribati, Lesotho, Malawi, Moldova, Mozambique, Niger, Samoa, Senegal, Solomon
Islands and Zimbabwe. A similar number, 13, met the percentage of spending target: Benin, Cte dIvoire,
Dominican Republic, Ethiopia, Guatemala, Honduras, Madagascar, Mali, Nicaragua, Senegal, Solomon
Islands, Vanuatu and Zimbabwe although the two groups contain different countries due to different
government spending/GDP levels. Across both targets, spending is nowhere near the levels needed for
the MDGs.
What are the recent trends in education spending?
Overall, governments have increased spending on education by 0.6% of GDP since 2008. However, after a
sharp rise in 200910, spending fell in 2011 and has broadly stagnated thereafter at 16.8% of government
spending and 5% of GDP.30
The country-level position is slightly better, as shown in Figures 2.15 and 2.16: between 2012 and 2014,
59% of countries increased spending as a percentage of GDP, and 54% as a percentage of total spending.
Implications for the post-2015 education agenda

The draft SDGs set out a far broader agenda for education than the MDGs: the new The total cost for
agenda commits to lifelong learning, going beyond pre-primary/primary/junior meeting the new
secondary education and including upper secondary, vocational, technical and tertiary
post-2015 education
education and full access for youth and adults. Alongside the drafting of the SDGs, there
has been a process to create an EFA 2 agenda (known as the Muscat agreement). goals will require a
Both of these agendas have a far greater focus on learning and quality than the tripling of spending
MDGs/EFA, and aim to ensure access to education for the most marginalised children. in low-income
This will be far more expensive than existing achievements and requires new, ambitious
countries
commitments.31
However, the Muscat agreement commits governments only to spending 46% of GNP,
and 1520% of total government spending lower levels than contained in the current targets discussed
above.32 As a result, the current targets contained in the Muscat agreement fall well short of what is
required to fund the SDG-related expansion.
In March 2015, UNESCO estimated that the annual total cost of achieving universal pre-primary, primary
and lower secondary education in low- and lower-middle-income countries is projected to increase from
US$100 billion in 2012 to US$239 billion between 2015 and 2030. The projected increase reflects a
combination of greater numbers of students and higher per student expenditure to improve quality and
address marginalisation,33 and is broadly consistent with the Muscat targets, though it is unclear how this
relates to reduced percentages of GNP or total spending.
In addition, it is clear that even these UNESCO costings fall well short of the SDG ambitions, notably in
aiming to finance only pre-primary, primary and lower secondary education and thereby omitting all the
other levels of education covered by the SDGs. As a result, no reliable costings exist for meeting the new
post-2015 education targets, and it is urgent that these be established as a top priority.
ENVIRONMENT
MDG target 7.A is to integrate the principles of sustainable development into country policies and
programmes and reverse the loss of environmental resources; target 7.B is to achieve, by 2010, a
significant reduction in the rate of biodiversity loss. Although MDG7 has galvanised some political
attention, bringing minor progress, the degradation of ecosystems, unsustainable use of natural
resources and loss of biodiversity have continued at an alarming rate since the 1990s. In the last decade
alone, 13 million hectares of forest have been lost worldwide each year.34
Over the lifetime of the MDGs it has become increasingly clear that these targets are a weak reflection of
the ambition necessary to address the environmental challenges facing the planet, provide too little
clarity for governments to implement them, fail to cover linkages with other development goals and
target a narrow range of symptoms rather than underlying causes.
Are countries meeting the environment spending targets?

Tracking spending on the environment targets is highly complex, partly because the targets cover diverse
sets of systems (forests, oceans, wetlands, urban biodiversity, etc.), which require changes and spending
across a broad range of government agencies if development policy is to be compliant with sustainable
development.
Robustly tracking spending is difficult, but GSW builds an initial picture by analysing spending by
environmental agencies and relevant budget lines in other ministries. Figure 2.17 shows that of the 39
countries with data, only three planned to spend above 1% of GDP in 2014 (Haiti, Papua New Guinea and
Samoa) and these are all highly vulnerable Small Island Developing States (SIDS) who are already feeling
the worst impacts of environmental devastation and climate change. On average, across all countries,
spending is only 0.3% of GDP, clearly nowhere near the amount needed to address the global
environmental crisis.
What are the recent trends in environment spending?

Average spending has actually fallen by two-thirds from a peak of 0.9% of GDP in 2012, which reflected
spending to combat natural disasters (mostly in SIDS). In terms of numbers of countries, rises in 51% have
been more than offset by falls in 41% (with no change in 8%), as shown in Figure 2.18.
Average spending
on the
environment is
only 0.3% of GDP
nowhere near
the amount
needed to address
the global
environmental
crisis.
Sustainability-compliant spending targets a priority for the SDGs

The complexity of allocating spending against the MDG environment targets may well explain why there
are currently no global targets or reliable estimates for costing the MDGs related to the environment: in
turn, this may explain why spending on the environment has fallen behind other types of MDG spending.
In addition, assessing whether government spending is ensuring greater environmental sustainability
would require an analysis of all spending similar to the gender-responsive spending concept (see Box 4
in Section 4). Improving this will take considerable efforts to track the SDGs.
The current business-as-usual approach to environment spending shows an appalling lack of will to deal
with the environmental crisis facing the planet. One of the largest increases in financing needs for the
SDGs will be for the environmental goals. The draft SDGs acknowledge increasing environmental
degradation and natural disasters, the urgency of reducing carbon emissions sharply by 2030 and the
need to ensure that policies and spending across a wide range of sectors are sustainability-compliant.
They include goals on protecting biodiversity, combating desertification and deforestation, conserving
oceans and climate-smart targets for agriculture, poverty and energy. Estimates indicate that this will
require annual public spending increases of US$80115 billion for climate change mitigation,35 US$60
100 billion for climate change adaptation36 and US$1541 billion for the other environmental goals.37 It
will also be urgent to set targets for and to track both core environmental spending and the degree to
which all government programmes are sustainability-compliant as a top priority for the post-2015
framework.
HEALTH
Health is the most prominent sector in the MDGs, with three goals dedicated to it, but on which progress
has varied. MDG target 4.A aimed to reduce by two-thirds the under-five mortality rate, but only a 50%
reduction has been achieved. MDG target 5.A (to reduce maternal mortality by three-quarters) and target
5.B (universal access to reproductive healthcare) will be missed by wide margins. Efforts to combat key
diseases (target 6) are mainly paying off, with a 25% fall in deaths from malaria and lower rates of new
HIV infections, but many people still do not have access to anti-retroviral treatment.38
Are countries meeting the health spending targets?

Ensuring that healthcare is free and universally provided takes significant Average
government spending. Over a decade ago, two clear targets were established. In
2002, at a Special Summit in Abuja, Nigeria, African Heads of State committed spending on
themselves to allocate 15% of government expenditure to health. Globally, the WHO health in low-
Commission on Macroeconomics and Health in 2001 concluded that a minimum of income countries
US$34 per capita was required to be spent on health to reach the health-related is below US$20
MDGs.39 More recently, the Taskforce on Innovative International Financing forecast
per capita.
that, by 2015, US$60 per capita was needed to strengthen health systems and
provide essential services in 49 low-income countries (LICs).40
In 2014, GSW data show that no sub-Saharan African country met the 15% Abuja target.41 Malawi met it
in 2011, and was the highest-spending country in 200814 (averaging 13%). The only other country to
meet the target in the GSW database was Togo in 2008. Ten countries exceeded 10% of allocations in
either 2013 or 2014 Burkina Faso, Burundi, DRC, Ghana, Lesotho, Liberia, Malawi, Swaziland, Mali and
Zambia. Senegal was the lowest-spending African country in 2014, spending below 5% well below the
Abuja target.42 Among non-African countries, Kiribati met 15% in 2010, and El Salvador and Nicaragua are
at 17%.
In relation to the target of US$60 per capita, 23 countries of 59 with data available planned to spend
above US$60 in 2014. However, these included only four LICs (for whom the target was set), and average
LIC per capita spending is below US$20, with the lowest spenders being Haiti (US$6), Madagascar (US$4)
and Zimbabwe (under US$1). Unsurprisingly, lower-middle-income countries (LMICs) and middle-income
countries (MICs) tend to spend much more per capita, with 13 spending over US$100. Some MICs are
spending very low amounts, such as India at a shocking US$18 per person. By far the highest-spending
country per capita in the GSW database is Samoa, at $315 per capita; since 2008, spending on health has
doubled as a percentage of total spending, although this appears to be temporary spike, possibly as a
result of post-cyclone reconstruction. Overall, most countries are spending much less than is needed.
What are the recent trends in health spending?
Overall, the latest years GSW data indicate virtually no rise in health spending as a percentage of GDP
since 2012, with spending standing at 2.7% of GDP, the same as the 2009 level cited in the GSW 2013
report. However, the average proportion of total government spending has fallen since 2012, from 9% to
8.6%. During 201214, only half the countries increased health spending compared with total spending,
as did 52% as a percentage of GDP. These trends are very worrying, given the major shortfalls in meeting
the health MDGs.
Health SDGs require a trebling of spending and careful tracking
The draft SDGs are much more ambitious than the MDGs on health: they target universal access to
healthcare (UHC) and focus on accelerating reductions in maternal and child mortality and increasing
access to sexual and reproductive health services. The consensus in the international health community is
that UHC-related spending needs to be predominately public.43 Ensuring that healthcare is free at the
point of delivery will also be critical to achieving equity (as will be shown in the next section).44 This will
require a significant scale-up of spending which makes the context of current stagnant spending very
worrying.
In particular, this will require a vast scaling up of public spending. The Lancet Commission on Global
Health 2035 has suggested that LICs and LMICs would need to spend US$51 billion more a year from
2015, rising to $80 billion in 2035, saving an additional 10 million lives.45
One further major issue will need to be resolved in monitoring spending on the SDGs: the degree and
nature of disaggregation. As research carried out by GSW has found, data in the health sector are very
difficult to disaggregate. Whether considering beneficiaries for maternal and child health, spending on
particular diseases or spending by objective such as sexual and reproductive health, it is fiendishly
difficult to disentangle spending by intended outcome or with a focus on equity. Though overall health
spending will need to be tracked to reach UHC, only more detailed data can assess whether governments
are adequately tackling the main barriers to UHC and achieving higher life expectancy, as well as
targeting those most in need.
SOCIAL PROTECTION
MDG target 1.A is to halve the proportion of people whose income is less than US$1.25 a day.46 This
target was met in 2010, lifting half a billion people out of poverty in the process. However, concerns
remain about the fact that many countries, especially in Africa, did not meet the target.
All government spending which boosts economic development for the poor and promotes inclusive and
employment-intensive growth can help meet this goal. However, GSW data focuses on the direct
government interventions that have been most effective in reducing poverty and providing employment,
known as social protection spending.47
Are countries meeting the social protection spending targets?

Social protection can be classified in two major groups, according to their financing mechanisms:
contributory and non-contributory. To track efforts on MDG target 1.A, GSW looks only at how much
governments are spending on non-contributory social protection. This is because GSW focuses on
mechanisms which target extreme poverty (and hunger)48 or can help to mitigate the poorest against
shocks which may push them into deeper poverty. Given that contributory funds are almost always for
those in formal (especially government) employment, they do not tend to target the poorest or most
vulnerable, and hence are not included by GSW as a means to track spending on the MDGs.
Two targets have been established for spending on social protection. In 2008 at the African Union (AU)
Windhoek Conference, African governments committed to a basic Social Protection Floor, the cost of
which was determined at 4.5% of GDP. In addition, the International Labour Organization (ILO) and
others have estimated the level of government spending needed to provide basic social protection at
between 2.9% and 5.2% of GDP.49 Given that GSW tracks only non-contributory spending, we have
chosen to measure countries progress by the lower end of the ILO target (i.e. 2.9% of GDP), although
clearly this is a somewhat crude measure.50
For the 45 countries covered by the GSW database which have information available on social protection,
spending is well below both these targets, as shown in Figure 2.23. No African country is anywhere near
meeting the Windhoek or the ILO target. Timor Leste is the only country which meets the lower end of
the ILO target range. Armenia is close, spending 2.3% of GDP, as is Jordan (2.2%). Only six countries
(Dominican Republic, Guatemala, Haiti, India, Papua New Guinea, Samoa) spend 1%, and overall there is
a woefully insufficient focus by LICs on social protection spending.
What are the recent trends in social protection spending?
Around 60% of countries have either increased their spending on social protection or kept it at the same
level as a percentage of GDP since 2012, while 53% have increased spending as a percentage of total
spending, which shows a slight increase comparative with other sectors. This builds on previous GSW
analysis which demonstrated slight rises from 2009 to 2012. Hence this appears to show a steady
incremental increase in spending over the past five years on social protection at the very least, the
sector is staying relatively buoyant in comparison with other sectors, which is in line with recent findings
in the ILOs World Social Protection Report.51 GSWs analysis of budget documents in collecting these
data has revealed a number of countries that have introduced new social protection budget lines, which
also appears to corroborate the reports findings that a number of MICs are expanding their social
protection programmes, while LICs are beginning to more explicitly focus on social protection.52
SDG implications: US$100 billion spending and more detailed data for
social protection
The draft SDGs reflect a consensus that the MDGs have targeted only extreme poverty, that inequality is
rising across the world and that the poor increasingly require life-cycle social protection to provide
resilience against natural disasters and other shocks. As a result, the SDGs include zero extreme poverty;
inclusive and sustainable economic growth, full and productive employment and decent work for all; and
reduced inequality. Investment in social protection will be essential to make progress on all these goals,
moving towards a lifelong social protection floor reflected in comprehensive sector-wide spending
programmes for social protection. Currently, the ILO estimates that 70% of the worlds population is at
risk of falling through the cracks because they are not adequately covered by social protection.53 Ensuring
improved coverage will take a huge scale-up in finance. Few estimates exist of the cost of such spending,
apart from cash transfer programmes to end extreme poverty, which have been costed at US$6695
billion a year.54 More comprehensive social protection floors would have costs several times higher.
The SDG agenda will also make it more vital to disaggregate social protection spending to understand
how it is impacting on different goals in terms of benefits (income, food, employment, etc.);
beneficiaries (children, families, elderly or disabled people) and the degree to which it reduces poverty
and inequality by distinguishing clearly between contributory and non-contributory systems.
WATER AND SANITATION
The final sector covered in this report is water, sanitation and hygiene (known as WASH). MDG target 7.C
is to halve the proportion of the population without sustainable access to safe drinking water and basic
sanitation. Between 1990 and 2012, 2.3 billion people gained access to an improved water source, and
the water target was met in 2010 globally (though not in sub-Saharan Africa).55 But sanitation is one of
the most off-track of all MDGs: 2.4 billion people will not have access to adequate sanitation at the end of
2015, virtually the same number as in 1990 and a total of one-third of the global population.56
Inadequate progress here has also undermined progress on the health MDGs.
Are countries meeting the WASH spending targets?

GSW tracks spending on water and sanitation, using a target of 1.5% of GDP.57 This is based on two
components: the agreement in 2008 at the eThekwini meeting of AU ministers to spend 0.5% of GDP on
sanitation and hygiene;58 and studies, including by UNDP, which have suggested that meeting the MDG
water goal requires 1% of GDP annually.59
In 2014, only three of 31 countries (Kiribati, Samoa and So Tom and Prncipe), or fewer than 10%, met
this target, as Figure 2.26 shows. In 2013, an additional four countries met the target: Angola, Niger,
Timor Leste and the Solomon Islands. While the Central African Republic came very close at 1.4% in 2014
(due to a huge increase in donor projects which may not be sustainable in the long term), 20 other
countries are spending much less than 0.5% of GDP the amount needed for sanitation alone on all
aspects of WASH. The average level of spending across all countries is just above 1%, nowhere near the
levels necessary for getting the sanitation MDG on track.
What are the recent trends in WASH spending?
Over the period of the GSW database (200814), average WASH spending has stagnated at around 0.9%
of GDP and 2.3% of total spending. As figure 2.27 shows, 50% of countries saw their spending go down as
a percentage of GDP and 47% saw this go up after 2012. Overall these trends demonstrate that spending
on WASH is largely stagnant and well short of the necessary target levels.
WASH spending needs to double and be tracked for impact on access

The draft SDGs are much more ambitious than the MDGs on WASH, intending to reach universal access to
water, sanitation and hygiene for all, which implies higher unit costs to reach marginalised groups,
especially in informal settlements and urban slums.60 In addition, the SDGs are broadening the focus to
include sustainable water resources management given
increasing water scarcity, and specify more detailed goals for
hygiene. UNCTAD also suggests that expanding equitable
coverage to all will require public investment in services to
meet the needs of the poorest.61
There is no doubt that vastly more investment is required for

the sector.62 According to the Sustainable Development
Solutions Network (SDSN), US$24 billion could be required
annually to ensure universal access to safe water and
sanitation.63 It will also become even more vital to distinguish
between spending which is increasing access and spending
which is only increasing water provision and infrastructure to
those with access already.
$24 billion could be

required annually to
ensure universal access
to safe water and
sanitation.
OVERALL SECTOR SPENDING PROGRESS

Comparing the different MDG sectors, Table 2.1 shows that progress in spending in Countries are over
virtually all MDG sectors is at best mixed. The percentage of countries meeting spending one-third short of
targets ranges from none on social protection to 39% for per capita health spending, with the spending
most sectors showing 1020%. Education appears to be doing best in that 20% of needed to deliver
countries are meeting at least one EFA target. on the MDGs
Education also receives the highest average proportion of total spending, at 16.8%, which
is twice the level of health, three times that of agriculture, 5.5 times social protection, 7.5 times WASH
and almost 20 times as much as spending on environment and climate change. This also leaves education
on average three-quarters of the way to meeting its targets, whereas health is only halfway, and
agriculture and social protection have only reached one-third of the necessary levels.
In terms of trends, agriculture performs worst, with spending falling on average and in the majority of
countries as is also reflected in the falling number of African countries meeting the Maputo AU target.
WASH also does poorly, with averages stagnant and spending by most countries falling. Environment and
health have seen stagnant or falling averages but marginally more countries increasing spending. The
best performers have been education and social protection: though averages have been stagnant, a
majority of countries have seen spending rises.
Overall, countries should be spending 57% of their budgets on these core MDGs, according to the
financial targets outlined above, but their annual budget allocations are closer to 35% and not rising to
any degree. They are well over a third short of the spending needed to deliver the MDGs: no surprise
then that many of the MDGs are not going to be met.
TABLE 2.1. PROGRESS ON MDG SECTOR SPENDING

Targets Average Recent trends
Sector spend
= all countries meeting target = upward trend
KEY
= some countries meeting target = mixed trend/stagnant
= no countries meeting target = downward trend
14% meet World Bank estimate 5.2% of total Averages stagnant as %
Agriculture 15% of African countries meet spending GDP (50%) or falling
Maputo target (10% of budget) (57%) as % total spend
Averages stagnant
16.8% of total
spending
19% meet EFA goal of 20% of 54% of countries
5% of GDP
Education government spending increasing spending as %
22% meet EFA goal of 6% of GDP of total: 59% are
increasing spending as %
of GDP
0.97% of total Averages fell sharply
Environment No target! Very low levels spending 51% of countries
0.3% of GDP increasing
No African country meets Abuja 8.6% of total Average stagnated (%
target of 15% of spending spending GDP) or fell (% spend)
2.7% of GDP 52% of countries

Health 39% of countries meet the WHO
increased spending as %
target of spending US$60 per
GDP: 50% increased/
capita
decreased spending as %
of total spending
No country meets Windhoek target 3.1% of total Averages mainly stagnant
Social (4.5% GDP) spending 53% of countries
protection One country meets ILO estimate 0.95% of GDP increasing either as %
(2.9% GDP) GDP or total
2.3% of total Averages stagnant
10% meet eThekwini target + spending 50% of countries
WASH
UNDP estimate (=1.5% of GDP) 0.9% of GDP increasing/decreasing
spending as % GDP
SECTOR SPENDING NEEDS FOR THE SDGs
Looking forward to the SDGs, Table 2.2 shows the overall scale of public spending Additional public
needed to achieve the new goals. In addition to the MDG sectors discussed in detail
above, for which the total needs are between US$227 billion and US$313 billion spending needs for
annually (excluding a large portion of education spending which remains uncosted), it the SDGs could be
shows the costings estimates for additional SDG sectors. These include access to as high as US$1.5
modern energy, climate change adaptation and mitigation, and major infrastructure trillion a year.
spending on power, transport, telecommunications and large infrastructure for WASH,
which total somewhere between US$562 and US$923 billion. These costings are based
on the assumption that climate change is limited to 2C: if not, financing needs could rise by several
hundred billion a year. It also shows potential costs of more comprehensive and timely monitoring of
SDG
Table 2.2. ADDITIONAL PUBLIC SPENDING FOR THE SDGs
Sector Extra public spending Private spend

(US$ billion annually) assumed (%)
Low High
MDG SECTORS
Agriculture 51 77%
Education 22+++64 2040%
Environment 15 41 16%
Health 51 80 0%
Social protection 66 95 0%
WASH access 22 24 1020%
MDG TOTAL 227 313
NEW SDG SECTORS65
Access to modern energy 23.5 30%
Climate change
Of which: adaptation 60 100 0%
mitigation 80 115 80%
Large infrastructure (power, 398 684 50%
transport, telecoms, WASH)
SDG TOTAL 561.5 922.5
Costs of monitoring SDGs 7.5 10
OVERALL TOTAL 796++ 1,245.5++
progress, including some of the budget transparency and monitoring issues discussed in Section 5. Once
full education costings are included, and especially if climate change is not limited to 2C , total extra
needs could be as high as US$1.52 trillion a year.
The additional sectors will also pose extra challenges for tracking spending, including distinguishing
energy and water spending designed to broaden access from that intended to increase provision; and
tracking spending in sectors where much of the public spending would be undertaken by state-owned
enterprises or using complex financing arrangements such as public-private partnerships (PPPs).
3 FINANCING THE MDGs
Government Spending Watch also tracks how budgets are financed and the impact this has on how
spending is implemented. This section examines how spending on the MDGs has been financed, and how
this has affected the types of spending (investment and recurrent) and the degree to which planned
spending has been implemented on schedule. Finally, it draws together lessons for financing the post-
2015 agenda and the current Financing for Development negotiations.
FINANCING THE MDGs

The MDGs contained a promise to forge a global partnership for development, reflected in the eighth
goal, and a set of related targets for high-income countries (HICs) to increase aid and cancel debt in order
to increase funding available to spend on the MDGs.
In 2002, the Monterrey Financing for Development conference established a series of concrete
commitments to finance the MDGs. These included:
A commitment to mobilising government revenue through effective, efficient, transparent and
accountable systems for mobilizing public resources and managing their use by Governments,
including equitable and efficient tax systems and administration (para 15);
A commitment to provide more aid to support the MDGs, by urging developed countries to allocate
0.7% of their GNI to aid and 0.10.2% to aid to least developed countries (LDCs) (para 42);
A commitment to increase the effectiveness and results of aid, notably by supporting development
frameworks that are owned and driven by developing countries (para 43);
Exploring the use of innovative finance to supplement ODA;
Balancing the provision of new sustainable debt financing and debt relief;
Commitments to support national development efforts with an enabling international economic
environment (para 6), including the use of ODA to leverage additional financing for development,
such as foreign investment, trade and domestic resources (para 43).
What has been the progress on these commitments since 2000?

In terms of government revenue, there
have been major steps forward. Overall,
developing countries increased their tax
revenue by 5.5% of GDP between 2000
and 2014,66 and low-income countries by
5%. For the GSW group of countries, as
shown in Figure 2.2 (section 2), the
revenue/GDP ratio has risen by almost
2% of GDP since 2008. This has occurred
in spite of trade liberalisation, which cut
revenue from trade, and the negative
impact of the global crisis, which cut
revenue by 1% of GDP. As a result of
these efforts, LICs have managed to
increase the proportion of their spending
funded by government revenue from only
46% in 2000 to 71% in 2014. As Figure 3.1 shows, GSW countries were funding 78% of their spending
from revenue by 2008. Due to the crisis, this fell to only 71% in 2009, but countries reached 77% again in
201213.
On the other hand, progress with aid has been mixed. By 2014, only six countries had met the 0.7% gross
national income (GNI) target for official development assistance (ODA) (Denmark, Luxembourg,
Netherlands, Norway, Sweden and the UK) and only five the target of 0.2% to LDCs (Denmark, Ireland,
Luxembourg, Norway and Sweden). Many other OECD member countries managed to increase aid flows
sharply during 200008 but since then aid flows have increased much more slowly, due initially to the
impact of the global economic crisis on their budgets, but more recently due to anti-aid changes of policy
in several major countries. Nevertheless, there has been a sharp and continuing increase in both South
South cooperation (aid from non-OECD countries) and private sector cooperation (aid via private
foundations and non-governmental organisations (NGOs)). As a result, overall aid more than trebled after
2000, exceeding US$200 billion by 2013.67
However, only a relatively small amount of aid is allocated to sectors related to the MDGs. As Figure 3.2
shows, within aid tracked by the OECD the total allocated to the sectors discussed in this report more
than quadrupled, from US$13 billion in 2002 to US$57 billion in 2013. In real terms (allowing for inflation)
aid to these sectors doubled, and their share of total aid also rose from 24% to 34%. However, the share
of MDG-related aid has stagnated since 2008, due in part to a 50% rise in aid for infrastructure, especially
transport, energy and the financial sector. As Figure 3.3 shows, all individual sectors except agriculture
saw an increase in their shares in 200208, with health gaining most. Yet since then only agriculture, the
environment and health have increased their shares.68
Government Spending Watch aims to enhance the accountability of governments to their citizens, and to
show how the quantity and quality of aid influence developing country governments efforts to meet the
MDG spending targets. One way to make this easier would be for the international community to have
clear evidence of how much of its aid goes through developing country budgets, and is therefore subject
to accountability to their parliaments and citizens. However, it remains impossible to calculate this
amount. All we know is that around half of Development Assistance Committee (DAC) member aid is
actually programmed to be spent in a planned way in developing countries (i.e. excluding emergency aid,
debt relief and spending in the donor countries): this is known by the OECD as country programmable
aid (CPA). Of this amount, only around two-thirds (i.e. a third of total DAC aid) is channelled via
developing country budgets.69 A slightly higher proportion of SouthSouth cooperation, and virtually no
private sector development cooperation is spent on-budget, implying that overall only around 30% of
global aid can be held accountable by developing countries. To make spending on the SDGs more
accountable, it is vital to increase the share of on-budget aid.
In addition, there are no published global statistics on any breakdown of this on-budget aid by sector or
type of spending (recurrent/investment). It would be extremely easy for OECD DAC donors (at least)
when reporting their aid to the OECD or the International Aid Transparency Initiative (IATI), to indicate
whether each project is on-budget thereby automatically allowing analysts to classify it by sector. It
would be somewhat more difficult, but not impossible, for them to break down the spending in each
project or programme depending on whether it is investment or recurrent.
Fortunately, GSW monitors the amounts of aid recorded in 66 developing country budgets i.e. on-
budget aid to track how much goes to each of the MDG-related sectors. This is the only global analysis
of trends in on-budget aid and, given the very low share of aid which goes through country budgets, the
trends in amounts and proportions going to MDG sectors via budgets can be very different from those
reported within global aid as above.
Figures 3.4 and 3.5 show the difference between the two sets of data. According to DAC data (Figure 3.5),
around 35% of ODA is spent on the MDG sectors, with health accounting for around 40% of this,
education 21%, agriculture 14%, WASH 11%, the environment 8% and social protection only 6%.
Comparing these allocations with governments own budget choices in Section 2, we can see that the
proportion of government spending allocated to the MDGs is somewhat higher, at almost 38%.
According to GSW data for on-budget aid identified as going to the MDGs (Figure 3.4), WASH gets the
largest amount, followed by education and health: this difference is because high proportions of health
aid are spent off-budget, compared with much lower proportions for WASH. Agriculture also shows
relatively lower shares on-budget, while environment and social protection have higher shares spent off-
budget. The post-2015 agenda puts particular emphasis on integrated government planning in the three
sectors with relatively low shares of on-budget aid to reinforce integrated healthcare systems for
universal health coverage, ensure fully environmentally sustainable development plans and build
comprehensive social protection floors. It is therefore a particular priority to make sure that more aid is
brought on-budget for these three sectors.
Figure 3.4: Percentage of donor funding by MDG sector Figure 3.5: Percentage of donor funding by MDG sector
(GSW data) (DAC data)
Environment
4%
Agriculture Education Education
18% 25% 21%
Environment
8%
Agriculture
14%
WASH
WASH 11%
27% Health
24% Health
Soc.Prot.
Social Protection 40%
6%
2%
GSW data also indicate that, since 2012, on-budget aid for the GSW countries has fallen substantially for
agriculture, education (especially primary education) and (to a lesser extent) health. It has stagnated for
the environment and social protection; only the WASH sector has seen increases in on-budget donor
spending. This reinforces other analysis showing that donor aid overall is volatile, which reduces its
effectiveness in producing results by 1520%.70 In addition, this more detailed volatility for MDG sectors
has worrying implications for donor commitment to key SDG sectors where much higher spending will be
needed post-2015 and, should revenue not continue to rise, for governments capacities to plan
sustainably and make long-term progress with the SDGs by replacing declining donor funding.
What does this mean for each sectors budget? Figure 3.6 shows the percentage of budget spending in
each sector that is funded by government and donors. There are wide variations across the different
sectors. In education and social protection, government revenues are more than 85% of the funding; for
health three-quarters, the environment two-thirds and agriculture over half. On the other hand, in WASH
donors fund more than three-quarters of spending (and the donor share has gone up by 3% since 2012).
This raises worrying concerns about government commitment to WASH spending and overdependence
on donor aid but also about low levels of donor on-budget spending on education. It also implies widely
varying degrees to which donors are following government priorities and supporting national ownership.
IMPLEMENTING THE SPENDING

GSW also tracks two other breakdowns of data which allow us to analyse how spending is being
implemented: the types of spending (split between capital and recurrent spending), and the degree to
which planned spending is actually implemented. In both cases, we have conducted extensive analysis of
what influences spending patterns and donor aid emerges as the most important factor.
Types of spending
The GSW database disaggregates spending according to its type either capital (investment in buildings,
equipment, etc.) or recurrent (for wages, maintenance and other goods and services). Both are needed to
ensure that new investment continues to improve the quality of equipment, extend coverage in
marginalised regions and deliver buildings and equipment in new spending areas under the SDGs; and
that recurrent spending is sufficient to ensure adequately motivated workforces, maintain investments
and provide other supplies such as schoolbooks and drugs. Too much investment can lead to it not being
maintained, and underinvestment can lead to reduced quality and insufficient coverage.
Donor funding distorts the types of spending in three ways:

Donor aid distorts
All donor project money is counted as capital spending in government budgets,
regardless of whether it funds capital or recurrent items. This is largely because relative shares of
most donors do not provide governments with a breakdown of their spending investment and
between capital and recurrent items. This makes it hard to get a true picture of recurrent spending.
the capital/recurrent split, especially in aid-dependent sectors, and shows that
interpreting spending trends would be much easier if donors did report an
accurate capital/recurrent split to governments.
Sectors with high donor funding have much higher percentages of investment spending (or appear
to have given inaccurate classification). As Figure 3.6 shows, more than 83% of funding in the WASH
sector is for capital investment. This may well be representative of actual spending in that sector:
experts have for years complained that there is a lack of recurrent spending to maintain water and
sanitation facilities once they are constructed; it also emphasises the need to ensure that adequate
recurrent funding is included in sector plans.
Sectors with lower donor funding appear to have much lower investment spending. This may also be
accurate because they have genuinely high recurrent bills (salaries, drugs, books, social transfer
grants) and because most of the necessary capital investment in new schools and clinics took place
earlier during the MDG period. However, this raises concerns about how recurrent spending will be
maintained when the SDGs demand a massive new wave of investment spending in new areas.
One vital disaggregation by type of spending is missing from this analysis. GSW (as well as many sector
specialists and labour unions) would love to be able to separate out wages to analyse trends in sector-
specific wage bills. However, our work for UNESCO found that only 12% of the GSW countries publish
separate and consistent data on wage levels for the education sector (one of the most transparent
sectors), and therefore this has not been possible for the 2015 GSW report.
Implementation rates
GSW tracks not only planned spending but also actual spending implemented (though as will be discussed
in Section 5, data on actual spending are less available). This is vital in order to assess and correct any
implementation shortfalls in a particular sector. Many authors have raised questions about the capacity
of developing country governments to absorb increased spending, or to execute budgets as planned, and
it is true that some countries (especially conflict-affected states) regularly spend 2030% less of their
budget than planned. However, other analysts and developing country representatives have indicated
that the fragmentation into small projects and volatility of donor aid, as well as the cumbersome and
lengthy nature of donor procurement and disbursement procedures, are mainly responsible for
implementation delays and that the absorption problem is less significant for government-funded
spending.
The evidence appears to support the latter position. Figure 3.7 shows two important
Absorption rates
facts:
Shortfalls are considerably less of a problem that often implied. All sectors are are less worrying,
keeping their underspends below 10%, and in 2013 agriculture, education and but donor aid
social protection were below 5%. So overall absorptive capacity is much less of an reduces them
issue indicating that countries could absorb increased spending on the SDGs. In
particular, around one-third of countries have seen overspends in social
protection, because needs have outstripped plans for spending.
There has been a strong correlation between donor aid and shortfalls in implementation, most
notably in the WASH sector. However, compared to the 2013 GSW report, implementation has
improved considerably in donor-funded sectors, implying that procedures have become somewhat
more flexible. This trend needs to continue, to improve SDG implementation rates.
Figure 3.7: Average underspend by sector as a percentage of GDP
0%
-5%
-10% Education
% planned spending
Health
WASH
-15%
Social Protection
Agriculture
-20% Environment
-25%
-30%
2011 2012 2013
HOW SHOULD THE SDGS BE FINANCED?

What lessons can we draw from the financing of the MDGs for how to fund the SDGs better?
Quantity of financing71
First, this section has shown that government revenue is a more reliable funding source than aid, because
it tends to be more aligned with government sectoral priorities; more likely to be stable; more able to
fund investment/recurrent spending in a balanced way; and produces higher implementation rates. So
the primary priority for financing public spending on the SDGs ought to be from increasing government
revenue.
Many developing countries have achieved large revenue increases in recent years, with the average
increase across LICs and LMICs during the MDG period reaching 5% of GDP. As a result, the countries
analysed in this report have already been funding 77% of MDG spending themselves the main
exceptions being countries which have had protracted internal conflict and where taxpayers have little
faith in government.
How has this been achieved? There have been three main types of country strategy:
Strong improvements in administration and the introduction of new taxes such as value-added
tax (VAT) (for example in Ghana, Rwanda and Senegal);
Changing policy to cut tax exemptions dramatically (for example Mauritius); and
Renegotiating key contracts with mining companies (for example Guinea, Liberia and Niger).
The first two of these methods have produced increases of 45% of GDP, and the third has gone as high
as 10% of GDP. But in all three types of country these strategies have been stagnating in recent years: the
LIC average has stagnated at about 2223% since 2011, and the IMF forecasts no significant increase
between 2015 and 2020. The only countries which still have scope to increase revenue through these
traditional strategies are post-conflict countries with revenue/GDP ratios well below 20%.
Yet the scale of extra spending needed for the SDGs requires a proportionate increase in revenue
mobilisation a massive step change increasing revenues by up to 10% of GDP during the SDG period,
and thereby (together with the effect of growth on the revenue base) doubling budget revenues on
average over the same period. Efforts to improve tax administration and squeeze more revenue out of
existing (largely domestic) taxpayers will not suffice. How can this increase be delivered?
As many have analysed, it is essential that the international community delivers on its Monterrey
commitment to support revenue mobilisation with an enabling international environment. One
component of this involves increasing aid which helps to catalyse budget revenue from its current
pathetically low level of only 0.2% of total aid. However, without more fundamental changes to the
global tax system and rules, developing countries will not be able to fund the SDGs.
These changes include ensuring that the G20/OECD initiatives on base erosion and profit shifting (BEPS)
and exchange of information among tax authorities are implemented in ways which maximise the
increase of tax revenues for developing countries. This means focusing on tracking practices and
exchanging information that will increase taxpaying in the source countries of the raw materials which
are the basis for company profits, rather than in the headquarters country of the multinational investor.
However, even if these are oriented to the maximum to help developing countries, and they receive all
the technical assistance needed to implement these complex initiatives, this is likely to increase tax/GDP
ratios by only around 2%.
Much more far-reaching measures are needed, including:

Revising tax and investment treaties in the same way, to ensure that they give fair emphasis to
paying tax in source countries, using the UN model of tax treaties. This could mobilise up to 3% of
GDP;
Dramatically reducing tax exemptions, which, as a forthcoming IMF/World Bank report to the G20
will say, have mostly been proven not to influence investment flows. This could increase revenue by
a further 3% of GDP but requires commitments by OECD governments (through their aid agencies
and development financing institutions), as well as international organisations and major enterprises,
that they will not request such exemptions for private sector enterprises;
Clamping down on tax evasion and avoidance and illicit financial flows (IFFs) by individuals, by
providing information urgently to developing countries (without obliging them to reciprocate, given
that virtually no tax evading or IFFs go to developing countries which are not tax havens),
dramatically increasing developing country capacity to exchange, analyse and audit information and
to prosecute evaders, and guaranteeing non-intervention by OECD governments or international
organisations on behalf of companies accused of evasion. This could raise a further 2.5% of GDP;
Increasing tax collections from large enterprises (especially multinationals in the natural resources
sector but also those in other high-profit sectors such as finance, telecoms, fishing, tourism and real
estate) by providing impartial technical and legal assistance to countries to renegotiate contracts and
tax agreements with these companies.
The current G20/OECD discussions on tax issues cover only a portion of these issues, and largely from a
G20/OECD member government perspective. It is therefore essential that the role of the UN which
gives a louder voice to developing country governments and other stakeholders, and is already dealing
with the wider issues raised above be expanded by upgrading the ECOSOC UN Committee of Experts
into a full intergovernmental ECOSOC commission which in cooperation with the OECD, IMF and other
organisations would make regular annual decisions on global tax rules.72 All of these tax issues
therefore need to be strongly debated in the FfD process and reflected in the Addis Ababa conference
through a special roundtable discussion and in the final outcome communiqu.
However, it is also clear from the estimates of additional financing needs presented in Section 2 that
government revenue itself would not be remotely adequate to finance the SDGs. The Overseas
Development Institute (ODI) has recently come to this conclusion especially for LICs, based on the
theoretical maximum capacity of countries to raise revenue given their income levels and economic
structures (as assessed by the IMF and World Bank).73 However, it is also true for most LMICs, given that
attaining their maximum tax capacity will be very difficult and that ODI looked at spending needs for only
three SDG sectors education, health and social protection. The SDSN has also found, especially in the
social sectors (education, health, social protection and WASH) but also for broadening access to energy,
and smallholder agriculture and nutrition, that the degree to which private financing can replace public
financing is very low; and that even in environmental, climate change and infrastructure interventions,
the public sector will need to spend a lot more (see also Table 2.2 in Section 2). Equally, as shown in
Section 2 via growing debt burdens, non-concessional public finance and extremely expensive public-
private blended financing arrangements, cannot fill the financing gap.
Concessional international public finance will remain vital to funding the SDGs with as much as US$1
trillion being needed. To mobilise and channel this money, the FfD agreement needs four clear targets:
All DAC donors recommitting to reach the goal of 0.7% of ODA/GNI, by 2025, and halving the current
gap between the levels of GNI they provide and 0.7% by 2020. These steps could mobilise an
additional US$250 billion a year, bringing ODA to around US$400 billion;
SouthSouth cooperation providers committing to rapid increases in concessional flows. These rose
by 300% during 200015, and a similar increase for the SDGs would bring them to US$80 billion;74
Innovative financing of US$450550 billion a year, including taxes on carbon emissions, bunker fuels
and air travel (which could easily raise US$250300 billion a year), on financial and currency
transactions (US$100150 billion) and regular annual issuance of IMF Special Drawing Rights targeted
to supplementing developing country reserves and fiscal space (at least US$100 billion);75
Focusing 90% of concessional flows on lower-middle-income and low-income countries, and 50% on
countries in special situations (fragile and conflict-affected, least developed, landlocked and small
island states) countries which can least afford to fund the SDGs from their own revenues.
Taken together, these measures could double the availability of international concessional public
financing to support the SDGs and, when matched with a doubling of budget revenues, would therefore
represent a genuine commitment to an equal partnership between developing countries and developed
countries in financing SDG progress across LICs and LMICs.
Even with higher public financing, private financing will have a key role to play in supporting the SDGs, as
will catalytic use of concessional funding to mobilise private financing. However, to maximise its positive
contribution to the SDGs, it is vital to set and monitor standards for the quality and effectiveness of
catalytic and public-private finance, and of private financing intended for the SDGs.76
Sectoral issues and allocations

This section has also shown that in terms of sectoral allocation of concessional flows to support the
MDGs and the future SDGs donors are falling short. Figure 3.8 shows that for the three sectors where
African governments are setting themselves targets to allocate a percentage of government spending,
lower shares of ODA (and only half as much in education) are being allocated to these sectors by donors
than overall in government budgets, and ODA is falling even further behind sectoral targets.
As we approach the period when new money will be needed to fund the SDGs, many different groups are
arguing for more funds for their sector without thinking about the negative impact this could have on
funding for other sectors (especially as all of the SDGs are inter-related and spending on one impacts on
many others). The post-2015 period needs a more rational accountability and decision-making process on
sectoral allocation. This should involve:
Setting an overall target for allocation of aid to SDG sectors, at least to match the 60% targeted for
government spending by developing country governments in Section 2;
Setting broad targets for global sector allocations of concessional funding, comparable to those
agreed by developing country governments (currently 20% education, 15% health, 10% agriculture
but revised to reflect the relative funding needs of the SDGs);
Inserting markers for SDG spending in each sector and overall into the OECD Creditor Reporting
System (CRS) and IATI databases, as well as all national aid databases, to make providers accountable
for progress;
Screening all aid projects for their compliance with the three pillars of the SDG agenda, as reflected
for example in positive impacts on reducing income inequality, increasing gender equality and
fighting environmental degradation and climate change (by expanding and tightening the definitions
of the current DAC screening marker systems for gender and Rio).
There are also crucial implications for

potential SDG funding for each sector:
Agriculture has seen a marginal decrease
in the proportion of donor/total spending
in the last few years, from 47% to 43%,
with a 20% reduction in on-budget aid. A
temporary rise in on-budget aid to
agriculture after the 2008 food crisis has
been followed by a decrease since 2011.
Much more aid funding will be needed for
agriculture, with higher recurrent
spending on agricultural extension,
purchase of food stocks and recurrent
supplies for nutrition, and investment in
sustainable agriculture if SDG progress is
to be achieved.
Education has also seen a fall in the share funded by donors, from 18% to 14%. This matches overall
DAC data showing falls in aid to education.77 Even more alarmingly, aid to primary education has
fallen more sharply, leaving governments to pick up 90% of the bill (up from 80% in 2012). These
trends need to be dramatically reversed, with much higher aid commitments to education. Very high
levels of recurrent spending raise concerns about long-term sustainability for teachers salaries and
other maintenance/materials costs and low levels of capital spend currently, against the need to
dramatically increase both spending for implementing lifelong learning under the SDGs.
Environmental aid (and its share in total environmental spending) has been stagnant. In the context
of the needs to address environmental challenges in the SDGs, this demonstrates that donors have
not really begun to scale up on-budget interventions to make development environmentally
sustainable. Huge commitments will be needed in this sector.
Health on-budget aid has fallen by around 20% from an all-time high in 2012, requiring governments
to spend much more from their own revenues to maintain the relatively slow MDG progress in this
sector. This is a worrying trend given the massive need to scale up budget funding for the SDG of
universal health coverage discussed in Section 2.
Social protection has the lowest amount of on-budget donor support of any sector, and among the
lowest donor-funded proportions of total spending. Both have been stagnating since 2008. This is
concerning given the massive increase in social protection spending that will be required post-2015
for comprehensive social protection floors to eliminate poverty and reduce inequality.
WASH has been the only sector to register large increases in donor funding since 2012, but this has
left the sector highly dependent (75%) on donor funds, spending too much on investment and with
larger spending shortfalls. A series of collaborative projects between GSW and WaterAid have
pointed to high donor funding as reducing recurrent spending and hampering absorption. 78
Governments will need to commit more of their own resources to this sector especially to recurrent
maintenance spending if progress is to be sustained and expanded under the life of the SDGs.
Aid quality and effectiveness
Finally, the GSW analysis of donor performance also brings lessons for quality and effectiveness:
A much higher share of aid (at least 80%) needs to be reaching developing countries, and at least 85%
of this amount needs to be recorded and implemented on-budget (the target set in the 2011 Busan
agreement on effective development cooperation), to enhance accountability between governments
and their citizens and improve budget planning and implementation.
DAC and IATI recording systems need automatically to track these trends by introducing markers for
each project.
Donors and governments need to monitor carefully the balance between investment and recurrent
spending, and how their aid may be distorting such priorities, especially in the likely context of a new
capital spending boom which could accompany the SDGs. To facilitate this, they need to report their
aid spending to developing country governments in ways which separate investment and recurrent
components.
Donors need to put a much higher share of aid through developing country government procurement
and disbursement systems, again in line with their Busan commitments, and to simplify and
accelerate their procurement and disbursement procedures for remaining aid.
The post-2015 SDGs require much greater amounts of financing than the MDGs. This means much
stronger efforts at domestic resource mobilisation, which can occur only if international tax rules change;
much greater efforts to mobilise international public finance (through SouthSouth ODA cooperation and
innovative finance) and improve its allocation; clear standards for private and blended finance;
monitoring and implementation of better sectoral allocation and spending practices; and dramatic
improvements in the quality and effectiveness of development cooperation. A business as usual
approach or an attempt to rely excessively on private or non-concessional financing will not suffice. If
these measures are not taken, then many of the SDGs will be dead at birth.
4 SPENDING TO FIGHT INEQUALITY
Fighting inequality is vital to post-2015
Though the MDGs made significant progress in reducing extreme absolute poverty in most countries,
they did far too little to address the needs of the very poorest and most marginalised people, or to
address spiralling inequality of income and wealth.
Income inequality has increased by 11% in poor countries since the beginning of the MDGs,79 while a
significant majority of households in developing countries more than 75% of the population are living
in societies where income is more unequally distributed than it was in the 1990s.80
Inequality within individual developing countries, or across regions, is often extremely high and getting
worse. In 2010, six of the ten most unequal countries in the world were in Africa.81 In China, the gap
between rich and poor surpassed the US last year.82 In India, income inequality has doubled over the last
two decades,83 while South Africa is now the most unequal society in the world, and the two richest
South African citizens now own more than the bottom 50%.84 In Zambia, in spite of average growth of 6%
every year between 2004 and 2013, the number of people living below the $1.25 poverty line grew from
65% in 2003 to 74% in 201085 income inequality is now the worst on record in the country.86
Figure 4.1: Growth in inequality by Gini index of household income inequality87
Source: UNDP calculations using data from Solt (2009)
Such rising inequality is not only bad for those left behind there is a growing body of evidence which
shows that it can also have a corrosive impact on democracy, security, poverty reduction and overall
growth. Evidence shows that, beyond a certain threshold, inequality harms poverty reduction and can act
as a brake on economic growth. In recent years, the Asian Development Bank, the IMF, the OECD and the
United Nations Development Programme (UNDP) have all cautioned that inequality can act as a drag on
growth and negatively impact on poverty reduction.88
Without action, extreme income inequality could hold back progress on poverty eradication, and on the
SDGs. The Brookings Institution has recently estimated that 463 million more people worldwide could be
lifted out of poverty in a scenario where inequality was reduced, compared with a scenario where
inequality increases.89 Over the lifetime of the SDGs, the greatest challenge to eradicating extreme
poverty will be in Africa, with forecasts projecting the share of the worlds extreme poor rising to 80% or
above by 2030 in the continent. But if African countries continue on their current growth trajectory, with
no change in levels of income inequality, then the continents poverty rate will not Tackling
fall below 3% the World Banks definition of ending poverty until 2075.90 This inequality is
means that without action on inequality, and fostering of more inclusive central to
development pathways in Africa, there will be no way to meet the goal to end delivering the
extreme poverty by 2030. Moreover, failing to tackle inequality is estimated to add SDGs: without it,
hundreds of billions of dollars to the spending needed to end poverty.91 the SDGs will
fail.
Extreme inequality will also impact on the ability to meet other development goals;
for instance, countries with higher levels of income inequality experience higher rates of a range of
health and social problems compared with more equal countries.92 In a world grappling with how to fund
the SDGs, this message must be heeded: tackling inequality is central to delivering the SDGs. A failure to
tackle inequality will lead to costlier interventions and, ultimately, a failure to deliver the SDG promise.
The SDGs, therefore, place a much stronger focus on reducing inequality. They contain a goal explicitly
related to reducing inequality within and between countries (Goal 10). It also targets a number of
inequalities within nations, including:
10.1 By 2030, progressively achieve and sustain income growth of the bottom 40 per cent of the
population at a rate higher than the national average
10.2 By 2030, empower and promote the social, economic and political inclusion of all,
irrespective of age, sex, disability, race, ethnicity, origin, religion or economic or other status
10.3 Ensure equal opportunity and reduce inequalities of outcome, including by eliminating
discriminatory laws, policies and practices and promoting appropriate legislation, policies and
action in this regard
10.4 Adopt policies, especially fiscal, wage and social protection policies, and progressively
achieve greater equality.
Goal 5 also separately confronts gender inequality by aiming to Achieve gender equality and empower all
women and girls. In addition, there are also explicit references to reducing inequality in other sub-targets
of the framework (e.g. to eliminating gender disparity in education; equitable access to WASH; equal
pay for work of equal value; and equitable access to infrastructure).
Many of the other goals imply reducing inequality, by aiming for zero hunger and near-zero extreme
income poverty; or 100% access to education, healthcare, WASH, energy, financial services, housing and
transport. The SDG preamble also commits to leave no one behind, as stressed in target 10.3 above: this
means not just eliminating unequal access, but also reducing inequity in outcomes.
These universal goals are going to be far more difficult to achieve than the partial coverage targeted by
the MDGs, which allowed governments to focus on improving lives for relatively easy-to-reach
populations. They will require a focus on the hardest to reach. For instance, in education, those now left
out of primary education are the hardest-to-reach groups, such as children living with disabilities, who
need specific targeted (and more expensive) interventions if they are to be included.93 Similarly, in water
and sanitation, bringing infrastructure services to the poorest slum dwellers, living in chaotic informal
settlements, or those living in remote hard-to-reach rural areas is going to be harder (and costlier).94 In
order to achieve universal health coverage (UHC), it is important that equity is built in, rather than
coverage reaching the easier-to-reach populations through insurance schemes aimed at formal
employees, to the exclusion of the majority of poorer people initially with a view to later expansion to
these (harder-to-reach) populations, who will get left behind.95
Public services are a key tool in fighting inequality
In the search for policy measures to combat income inequality, many analysts have identified more
equitable government spending and more progressive tax systems as crucial interventions.
In particular, better and more targeted spending on essential public services can be a vital step to address
inequality. For instance, a 2012 OECD study found that public services (education, health and social
services) are worth 75% of the income of the poorest 20% of the population, compared with only 14% of
the income of the richest. They therefore have a huge immediate anti-inequality effect as well as a
longer-term effect by improving the human capital of the poor and increasing their opportunities and
capacities to earn higher incomes including providing a virtual income and redistribution (via taxation).
The study found that OECD countries which increased spending on services throughout the 2000s had
more success in reducing income inequality.
Oxfam used this analysis in 2014 to calculate that, between 2000 and 2007, the virtual income provided
by public services reduced income inequality by an average of 20% across OECD countries. In five Latin
American countries (Argentina, Bolivia, Brazil, Mexico and Uruguay), this virtual income from healthcare
and education alone has reduced inequality by 1020%.96 Moreover, the combined impact of taxation
(when progressive), coupled with spending on public services, has a double inequality-busting impact:
once when taxes are collected and, secondly, when is the funds are invested in public services (see Figure
4.2).
Figure 4.2. Impact on inequality of taxes, benefits and public services, in five Latin American countries97
So, investment in public services can help to tackle inequality by lifting the poorest out of poverty and
acting as a redistributive tool within the economy.98 This is being increasingly acknowledged in
mainstream development discourse. For instance, the IMF, through recent research and advice to
countries, is gradually recognising that spending on education, health and social protection has a key role
to play in tacking inequality.99 Ensuring that investment in the SDGs is both targeted at tackling inequality
and addresses pre-existing inequities will be vital to their success through GSW data and the knowledge
gained in gathering this, the research team has begun to identify lessons in spending to redress
inequality.
Spending to redress inequality: implications for the SDGs
Government spending on the SDGs which is targeted to redress income inequality can also reduce
inequality of access to, and outcomes, of government services. A recent Save the Children report on
equity in education concluded that the primary influence on disparity in educational outcomes is
inequality of income (rather than region, religion, urbanisation or gender) and that spending must
therefore, above all, be targeted to overcome and offset this inequality.100
Each sector has its own complex relationships between socio-economic status, access and outcome
inequality, which will need to be further analysed and addressed when designing equitable interventions
to deliver the SDGs. For instance, in health, gaps in inequality in access and outcomes by socio-economic
groups are widening rather than narrowing threatening the achievement of universal coverage.101
There is also a direct correlation between rising income inequality and health access: lowering inequality
of the richest 20% by just 1 percentage point could save the lives of 90,000 infants each year.102
Needless to say, significantly more analysis is needed to ensure that government spending tackles all
forms of inequality simultaneously. This will involve a laser-like focus on ensuring that all spending in
government budgets (as well as all aid) addresses (in)equity issues, as well as the rebalancing of spending
to tackle specific forms of inequality and the design of more specific and focused programmes to reach
the marginalised.
The remainder of this section provides preliminary evidence on three types of spending which can help to
fight inequality: inequality within sectors; geographical/spatial inequality; and inequality among different
beneficiary groups (notably along gender lines).
REDUCING INEQUALITY WITHIN SECTORS

Amounts and types of spending within different sectors can also have a profound impact on overall levels
of inequality and poverty. The size of the overall allocation to the sector (and how this compares with
total costs of reaching the relevant MDGs/SDGs) is extremely important, because any shortfall tends to
mean that spending that is most difficult to implement or least politically attractive (typically on the
marginalised) is foregone in favour of easier or high-profile political spending.
One key way in which spending envelopes exacerbate inequality has been the trend in many countries,
where public resources are short, to opt for contributory or privately run insurance systems or for private
(often profit-making) provision of services, or to insist that users should pay fees, in order to help fill
financing gaps. This has applied especially to education, health, social protection (e.g. pensions) and
WASH, while in many countries agricultural support (in terms of provision of inputs or training, marketing
and selling of products, or ensuring adequate food stocks) has simply been left to the market. All of
these mechanisms have been shown to exacerbate inequality.
Box 2. El Salvador: inequality in health spending

In El Salvador GSWs analysis of the budget shows that close to half of the total health budget is spent on
health insurance schemes supporting civil servants, including health workers, teachers and ex-servicemen.
Together these represent only 18% of households, meaning that nearly half of the budget is allocated to less
than one-fifth of the population. The remaining budget funds health services for the rest of the population
through the publicly provided health services although the richest almost all use private providers due to the
poor quality of public services, leading to an even greater inequality gap in the kinds of health service that El
Salvadorians can access.
It is not surprising then that, according to the WHO, 41% of the population are excluded from the health
system. The European Commissions Country Strategy Paper 200713 meanwhile states, the health sector is a
cause for concern there is inappropriate allocation of resources among public health institutions; health
spending is low and fragmented; the quality of care provided to citizens is poor.
Source: GSW data and Action for Global Health (2011) Health Spending in El Salvador: The impact of current
aid structures and aid effectiveness, available at:
www.actionforglobalhealth.eu/uploads/media/PolicyBriefing3_ElSalvador_LowRes-2.pdf
A number of Latin American and other countries have health insurance schemes which serve only small
groups (such as civil sector workers), leading to highly unequal health outcomes.103 GSWs analysis of El
Salvadors budget shows how this can occur (see Box 2). As also discussed in Section 2, many countries
also spend more on contributory social protection mechanisms which benefit wealthier citizens, which
again fails to expand coverage to the very poorest.
To tackle inequality comprehensively, such services need to be free at the point of To tackle
delivery and financed fairly through general taxation. For instance, the way a inequality,
healthcare system is designed and financed has consequences for inequality and services need to
poverty. For instance, PPPs can divert funds, such as a recent case in Lesotho (see Box be free at the
3). This is starving the budgets of health services in rural areas that are used by the point of delivery
poorest people, further widening the gap between rich and poor.104 In the absence of and financed
a free quality public healthcare system, patients often delay seeking treatment fairly through
which in turn deepens poverty. Health user fees have been estimated to push 150 general taxation.
million people into poverty each year.105 A study of 89 countries suggested that the
more a country relies on out-of-pocket financing the more of its poorest households face financial
catastrophe.106
Box 3. Health PPP trebles costs and cuts other MDG spending
The Queen Mamohato Memorial Hospital, in Lesothos capital Maseru, was financed by a public-private
partnership (PPP) that includes delivery of all clinical services, with advice from the International Finance
Corporation (IFC), the private sector investment arm of the World Bank Group. The promise was that the
PPP would provide vastly improved healthcare services for the same annual cost as before. Three years
on, the PPP hospital and its three filter clinics:
Cost US$67 million per year at least three times what the old public hospital would have cost today
and consume 51% of the total government health budget;
Are diverting urgently needed resources from health services in rural areas where three-quarters of
the population live and mortality rates are rising;
Are generating for the shareholders a 25% rate of return on equity and a projected cash income 7.6
times higher than their investment.
The cost escalation has necessitated a 64% increase in government health spending over the next three
years, with 83% of this increase accounted for by the PPP. This is a dangerous diversion of scarce public
funds from nurses, rural health clinics and other proven ways to get healthcare to the poorest and
reduce inequality. Moreover, GSW analysis shows that Lesotho has decreased its spending on other MDG
107
sectors, notably education and social protection.
Based on a study by GSW and Oxfam: A. Marriott (2014) A Dangerous Diversion: Will the IFCs flagship
health PPP bankrupt Lesothos Ministry of Health?, Oxfam, http://oxf.am/5QA
However, even leaving aside inequitable distribution among regions and beneficiaries (see the following
sections on these), there are often major structural inequities within sector spending. One good example
of this is in education. GSW work carried out for Save the Children and UNESCO indicates that
government spending on different levels of education (primary, secondary and tertiary) is often highly
unequal.108
GSW data show that in some countries there is low spending on pre-primary/primary education, as a ratio
of overall spending, while secondary or tertiary education may get far more money (especially per pupil)
than primary (see Figure 4.3 for a breakdown, which exemplifies differences in investment across levels of
education in 10 countries). This is likely to disadvantage the poorest, as the 25% of students who drop out
of primary school in developing countries are usually from poorer backgrounds, while those staying on in
education until later tend to be wealthier. This is part of the reason why 43% of public education spending
in sub-Saharan Africa109 and 73% in Malawi110 benefits the most educated 10%. Taking into account
retention and completion rates according to the level of family wealth, in most countries the most
educated 10% are almost unilaterally drawn from the richest 20% of the population.111
Source: Published using GSW data by DFI: Trends in government expenditure for public education, 201113. Background
paper prepared for the Education For All Global Monitoring Report 2015, Education for All 20002015: Achievements and
Challenges, available at: http://unesdoc.unesco.org/images/0023/002324/232476e.pdf
Other examples of inequality of spending within sectors relate to what is being funded. For example,
many countries have found themselves funding large programmes against specific diseases (especially
HIV and AIDS) while neglecting maternal and child healthcare, broader sexual and reproductive
healthcare or integrated health systems; or funding curative hospital care rather than preventive or
primary care (see Box 3 on Lesotho). Similarly, in the water sector often more is spent on more profitable
and visible water projects than on sanitation. In agriculture, spending may often be about financing
agribusiness or processing rather than extension or irrigation services for smallholders, or nutrition.
REDUCING GEOGRAPHICAL/SPATIAL INEQUALITY

Another major source of spending inequity is across geographical areas. Often allocations in each country
go disproportionately to areas with the largest populations, urban or wealthy areas, or areas which are
politically favoured by governing parties. This is in spite of the fact that it is widely accepted by global
experts that costs are higher in delivering services to reach rural and poorer areas, because of the lack of
infrastructure and the need to pay premiums to workers to attract them to poorer and more remote
areas so per capita allocations to these areas ought to be higher.
For instance, work by various organisations, including GSW, has found that in water and sanitation there
is huge inequity in spending, particularly for informal settlements in urban areas and poor and remote
rural areas.112 As a result, the number of people in rural areas compared with urban areas without clean
water is five times greater,113 and in South Asia poor people are over 13 times less likely to have access to
sanitation than the rich.114
A study carried out by GSW for Save the Children115 examined geographical inequality in spending for
children in a number of countries, and found that spending allocations in each country go
disproportionately to areas with the largest populations, and to urban and wealthy areas. This can lead to
highly unequal spending patterns (as seen in Figure 4.4, for an example of spending on education in
Uganda). The negative effects of this misallocation in terms of equity across the countries analysed were
confirmed by much lower school enrolment and completion rates and poorer health indicators
(vaccination, mortality) in rural, poorer areas. Even in South Africa a much heralded success story
where government has developed complicated and advanced formulas to allocate spending more
equitably there are no systematic data on geographical outcomes to compare with spending inputs. In
low-income countries, such as Mozambique and Uganda, one of the major causes of unequal spending
seems to be donors investing large amounts in projects in particular regions where they have a tradition
of working or where strong links have been established.
These results are confirmed by GSWs broader analysis of geographical spending patterns in 10 countries:
for example, in Guatemala and Peru, there is massive inequality in per capita spending across districts,
with some districts spending 1030 times as much as others, leading to per capita US$ spending
differences in Peru as high as US$400.
Figure 4.4. Inequality in education spending in Uganda
REDUCING INEQUALITY AMONG BENEFICIARIES

Another crucial type of unequal spending is among different groups of beneficiaries. Often, government
spending patterns reinforce disadvantages for particular groups for example by gender, age, ethnic
minority or disability. This is perhaps the area in which the least information on spending trends is
available, largely because most governments do not analyse spending by beneficiary.
The most widely analysed and discussed of these inequalities is based on gender. GSWs experience in
assessing gender-responsive spending is instructive here. GSW (and other expert bodies such as UN
Women) have struggled to obtain meaningful and robust data in a way which allows tracking of spending
oriented to women, as a historically disadvantaged group, or even in terms of addressing the MDG
commitments. Evidence of the gender equity of spending within different sectors and areas of budget
allocation (i.e. education or health) is woefully lacking. Where data is available, it is largely only
identifiable by specific ministries or lines for gender agencies (which represents only a small proportion of
government spending), as discussed in Box 4.
Box 4. Spending to combat gender inequality: evidence from GSW to shape more
gender-sensitive spending/tracking and outcomes in the SDGs
In addition to the six sectors already included in this report, the GSW database also tracks spending to combat
gender inequality. Our latest data show that less than 0.1% of spending on average across countries is
specifically targeted at addressing womens rights or tackling gender equality, and that this has not risen to
any degree since the GSW 2013 report. In contrast with 2013, we have not carried out a detailed analysis of
such spending for this report, as the data track only programmes executed specifically for women by gender
agencies/ministries, and are very hard to link to broad MDG goals such as improvements in womens rights or
greater empowerment of women, or even more specific targets such as gender parity in education or the
number of female parliamentarians.
Many donors and governments buy in to the view that a specific gender agency is essential to drive progress
more rapidly and to increase visibility of gender equality issues across government. However, such spending is
usually subsumed into broader social welfare ministries, with insufficient disaggregation to identify spending
on women. Furthermore, analysis by GSW for UN Women of 15 countries with detailed budgets for gender
ministries or agencies shows that only 11 are given a lead responsibility on gender mainstreaming across
government or womens rights; nine implement programmes on womens economic empowerment; eight
implement programmes on political and social participation; and eight support programmes in justice,
including combating violence against women.
As discussed in the GSW 2013 report and in work carried out for UN Women, there have been major efforts
across the world to promote gender-responsive budgeting and to analyse the degree to which spending is
directly or indirectly targeted to women. However, these efforts have not borne fruit to the degree that it is
possible to compare gender-responsive spending levels across countries, nor (with a few exceptions) to ensure
that countries are screening and programming all spending to maximise its impact on combating gender
inequality.
Given huge gender disparities in access to services and development outcomes, more and better spending
must be a touchstone for judging the success of the SDGs. It will be essential to target more spending directly
to women, such as maternal healthcare, reproductive rights and the fight against female genital mutilation,
sexual abuse and violence against women. It is also necessary to introduce gender-responsive budgeting and
spending/impact analysis, to ensure that spending is having the desired impact on equity and access. But this
must not be done as a stand-alone initiative it should be integrated with other efforts to ensure that such
spending is targeting inequality in all its forms.
However, women are not the only losers among beneficiaries. For example, on It is clear that
education, GSW found that only nine countries had specific lines available to be analysed the amounts
on spending for special needs students, who do not get anywhere near enough funding and kind of
to fulfil their extra needs. Similar conclusions have been reached about low levels of spending
spending on elderly people in developing countries, and GSW has also found that there necessary to
has been relatively little progress in budgets for children. Many other studies have tackle inequality
pointed to discrimination due to ethnicity or sexuality. All of these types of inequality through public
need to be assessed and targeted as part of the SDGs. services are
woefully lacking.
ARE WE READY TO TACKLE INEQUALITY?
It is clear that the amounts and kind of spending necessary to tackle inequality through public services
are woefully lacking. GSWs analyses for Save the Children, UNESCO, UN Women and WaterAid has
shown in various sectors a lamentable failure to target the poorest people or to address intra-sectoral,
geographical or beneficiary inequality using government spending. Specifically:
In terms of levels of spending, as shown in Section 2, current amounts are far too low even to meet
the relatively easier-to-achieve MDGs.
Most countries have not conducted the types of analysis needed to ensure that government
spending tackles all inequality simultaneously. This is partly because in many countries timely data on
inequality, spending breakdowns and impacts on household income/consumption are not available.
It also reflects a fragmentation of different types of analysis of impact on beneficiaries (gender-
responsive budgeting, child-oriented budgeting); on geographical areas (geographical/spatial
distribution of spending); and on intra-sectoral imbalances (analysis within sectors) and a lack of
tools which can simultaneously assess impacts on all types of inequality.
As for spending allocations, the degree to which countries apply their analysis by designing new
allocation formulas (see Box 5), and the degree to which these are objective rather than determined
by political preferences, varies dramatically.
The degree to which donors and other funders follow a governments/countrys preferences in
tackling inequality, rather than tackling it from their own (often equally fragmented) perspective (or
ignoring it completely) varies dramatically. In some country experiences donors are more oriented to
tackling specific inequalities and can help governments; but in many others, reliance on donor funds
has led to distortions.
Finally, implementing anti-inequality spending remains extremely challenging the beneficiary
groups, regions and sub-sectors/programmes which are most marginalised have the least ability to
resist any diversion of funding allocated to inequality by more powerful actors, especially when such
spending is decentralised.
All of these stages of the process in government spending need to be tackled, so that equitable spending
becomes a cornerstone of implementing the SDGs at national level. As a first step, it is vital to ensure that
allocation formulae deliberately channel spending away from the wealthiest and highest-spending areas
(generally capitals and other large cities) to poorer and rural areas which have the highest need, and
where the greatest impact on outcomes can be made. Box 5 draws on lessons from different
governments attempts to develop more equitable spending patterns.
Box 5. How to make spending more equitable

A recent review of country experiences in trying to promote more equitable sharing of education spending
comes to the following conclusions:
Governments need to adopt a redistributive approach to allocations, aimed at equalising opportunity and
outcomes.
Public spending formulae should reflect a needs-based approach to equitable sharing, striking a balance
between equal per capita transfers and weighting for disadvantage.
The poverty gap, as distinct from the poverty headcount and incidence, should be a primary indicator of
disadvantage.
More weight should be attached to the number of out-of-school children of primary school age and to
wider indicators of disadvantage in determining basic education budget allocations.
More equitable financing formulae in education should be linked to more effective policies for expanding
access and improving learning achievement.
Governments and their donors must invest in building national statistical capacity on inequality.
Source: K. Watkins and W. Alemayehu (2012) Financing for a Fairer, More Prosperous Kenya: A review of the
public spending challenges and options for selected Arid and Semi-Arid counties, Brookings Institution
5 ACCOUNTABILITY= RESULTS
One final critical part of ensuring that public services meet the needs of the poorest, and act as an anti-
inequality weapon, is ensuring that governments and aid donors are held accountable for their spending
and its results by citizens, for which budget transparency is essential. The draft SDGs acknowledge the
case for this by setting targets so that governments:
16.6 Develop effective, accountable and transparent institutions at all levels
16.7 Ensure responsive, inclusive, participatory and representative decision-making at all levels
16.10 Ensure public access to information and protect fundamental freedoms, in accordance
with national legislation and international agreements.
The need for greater transparency in the SDGs is also widely acknowledged through the post-2015
debate around a data revolution. However, much of the discussion focuses on outcome and results
indicators, and fails to specify the need for transparent budget information, so that the inputs, or means
of implementation, in terms of government and donor spending, can be tracked by citizens at the
earliest possible stages.
Sections 24 of this report have repeatedly stressed that tracking the spending on Much discussion
existing MDGs is difficult due to low transparency, that the new targets of the around the post-
SDGs require even greater transparency, that the behaviour of donors and other 2015 data
funding sources (including government revenue) also needs to be much more revolution focuses
transparent and that the need to tackle inequality of all types requires even on outcome
greater transparency and accountability on allocation decisions and actual indicators and fails
spending implementation. to emphasise the
need for transparent
This section, therefore, makes the case for why greater budget accountability is budget information
one of the most powerful tools in producing MDG (and potentially SDG) results. It to track the inputs.
begins by assessing the current state of budget transparency in terms of availability
of data on spending which can be linked to the MDGs. It then discusses the
qualitative and quantitative evidence supporting the argument that greater accountability increases
results (shifting the debate beyond transparency as an end in itself). Finally, it suggests how incremental
low-cost quick wins could rapidly advance budget accountability, increasing the amounts and
effectiveness of spending and making this one of the most powerful data revolution steps to achieve the
SDGs.
DATA AVAILABILITY IS IMPROVING RAPIDLY

Figure 5.1 shows the countries for which GSW has data, and the breakdown of levels of data available
in each country for the areas GSW analyses (i.e. for each sector, planned and actual, and split into
sources of financing and types of financing).116 GSW currently tracks 66 countries, and has given
priority to low-income and (although to a lesser extent) lower-middle income countries, whose
development plans are more closely related to the MDGs. This represents an increase of 14 countries
since the 2013 GSW report. Among these 66 countries:117
28 of the total 34 low-income countries are included. GSW continues to lack data for Chad, Eritrea,
Guinea, Myanmar, North Korea and Somalia. In addition, we have been unable to update data for the
Gambia due to a reduction in the level of budget transparency since 2013.
30 of the 50 lower-middle-income countries are included, with four new countries since 2013 El
Salvador, Guatemala, India and Swaziland. We have been unable to update data for Nigeria due to a
reduction in the level of transparency on decentralised state spending.
Eight are upper-middle-income countries, with six new countries since 2013 (Colombia, Dominican
Republic, Ecuador, Jamaica and Peru). South Africa will be added to these shortly.
Figure 5.1 GSW data by each country (% of data that can be extracted for each country by
GSW)118
Overall data availability by country

As shown in Figure 5.1, there is major variation across countries in the proportion of data available to
track and analyse spending on the MDGs.
The vast majority of countries (68%) have improved or sustained their data availability since 2013.
Four countries (Jamaica, Nepal, Peru and Swaziland) have 100% of data available.
Ten countries (Afghanistan, Armenia, Cte dIvoire, DRC, Guatemala, Honduras, Mozambique,
Nicaragua, Togo and Uganda) have 80% or more of their data available.
Fifteen countries have 6080% of data available (Bangladesh, Burkina Faso, Cape Verde, Colombia, El
Salvador, Guyana, Kenya, Liberia, Moldova, Rwanda, Solomon Islands, Sri Lanka, Tonga, Vanuatu and
Yemen).
Twenty-eight countries have 4060% of data available (Benin, Bhutan, Burkina Faso, Cambodia,
Cameroon, CAR, Congo, Dominican Republic, Ecuador, Ethiopia, Ghana, Guinea-Bissau, Haiti, India,
Jordan, Kiribati, Madagascar, Malawi, Mali, Papua New Guinea, Samoa, So Tom and Prncipe,
Senegal, Sierra Leone, Solomon Islands, Timor Leste, Zambia and Zimbabwe).
Six countries have 2040% of data available (Burundi, Lesotho, Niger, Tajikistan, Tanzania and
Yemen).
Five countries have 200% of data available (Comoros, Djibouti, Nigeria, Occupied Palestinian
Authorities and the Gambia).
In addition to the countries in the database, GSW has conducted scoping work to assess how countries
present their data and the prospects for expanding the GSW database for a further 52 countries,
including 15 high-income countries. This means that GSW has analysed data availability for a total of 124
low-, middle- and high-income countries. This gives a reasonably comprehensive picture across all
income levels. GSW aims to scale up to around 80 countries in its database over the next two years. In
terms of the way data are presented and the percentage of planned spending data available, the 124
countries fall into seven groups (as shown in Figure 5.2 below):
On the left-hand side of the chart are the 22 countries (16 low- and middle-income and six oil-
producing high-income countries) for which we have been able to source no data (apart in some cases
from overall total spending), and which have therefore been omitted from the GSW database. In many
of these, detailed spending data are secret and not subject to any freedom of information laws.
Next come 31 countries which have breakdowns available by sector or ministry, allowing them to
have 42% of planned spending data on average. They do not disaggregate ministerial or sectoral
spending, thereby making it very difficult to identify spending on gender, primary education, social
protection or WASH, which are split across several ministries or hidden as part of a ministry.
The third bubble represents 27 countries which split data by sub-sector or departments within
ministries/agencies. This allows them to show much more clearly splits for most of the MDG sectors,
bringing data availability to 63%. However, for this group, it is sometimes still not possible to analyse
gender, primary education, social protection or WASH, if institutions are not split this way.
The fourth bubble represents 20 countries which disaggregate spending by programme. This enables
access to an average 85% of data, but there are still some ways in which programmes are classified
(notably failing to identify beneficiaries clearly enough) which make the remaining data hard to find.
The fifth bubble represents a very small group of countries (only five) which are virtually MDG
ready. They present data by programme and by beneficiaries and potential results, allowing, for
example, identification of gender spending or different levels of education spending. They allow us to
find 95% of data on average, and are in many ways ready for the challenge of the SDGs and tackling
inequality.
The sixth and seventh bubbles represent many high-income countries, especially those which have
implemented budget transparency. They have too much data available and therefore we cannot
include them in the GSW database because it is impossible to piece the data together in a way which
shows totals in MDG sectors. Large parts of their spending is decentralised across multiple
government agencies. While a few aggregate this into a national picture, and therefore fall into
categories 35, some only aggregate one or two levels of government, and many produce no
aggregated numbers and often allow decentralised agencies to adopt different budget formats,
making it necessary to add up confusing numbers from 100+ spending units. Many of these are high-
or middle-income countries which do not acknowledge that their spending should have anything to
do with the MDGs, though in principle the SDGs will apply to all countries, regardless of income level,
implying that they need to do more to be accountable for progress. Therefore, theoretically, this may
imply the need to be able to better aggregate their spending in a way which demonstrates spending
according to the SDGs.
This classification also allows us to see what each group would need to do to advance further in making
their spending accountable, as well as how much progress in data availability there has been. From here,
identifying the relatively straightforward, rapid and low-cost steps that would allow countries in each
category to improve their data, the following must be priorities:
The first group should be encouraged to publish spending data preferably by programme (which
would increase their data availability by 85%, but at least by sub-sector (63%).
The second and third groups should be encouraged to disaggregate spending further (into sub-
sectors or programmes within agencies), which could increase their data availability by 2143%.
The fourth group should be encouraged to conduct full programme budgeting specifying the
beneficiaries and results, increasing data by 10%.
The fifth group should focus on moving on to becoming fully SDG ready, with more detailed
programmatic work (for example, on nutrition and smallholder agriculture), as well as assessing the
impact on inequality.
The sixth and seventh groups need to classify their data in line with the SDGs and to aggregate the
different levels of government (while of course also continuing to publish the more disaggregated
levels so that sub-national progress can be tracked). This would increase data availability by 80100%
per country.
WHICH DATA ARE AVAILABLE?

Overall, GSWs analysis of data availability for 2014 confirms the patterns described in The overall
the 2013 report, while showing major improvements in all categories. The overall availability of
availability of information across all categories of the GSW database rose from an information in the
average of 45% in 2013 to 60% in this report, representing a 15% or around one-third GSW database
improvement since 2013, and for a 25% larger group of countries. GSW estimates that has risen by
around 30 countries have improved their group (based on the above groupings), around one-third
moving closer to the centre of the chart and therefore are able to be better analysed, since 2013.
in the last three years.
Data availability by source of finance and type of spending

Data on total spending are the easiest to find across countries, which in understandable given that this is
the most basic of information. Of the 66 countries in the GSW database, all have information on planned
overall expenditure (though data for the Gambia and Nigeria are not available for the most recent years).
Less available at only 76% is information on total government actual expenditure, which is an
important barometer of commitment to implement spending plans.
Data on sectors and sub-sectors, which are vital to track MDG-related spending (and will also be for the
SDGs), are less readily available in many countries. This is generally because countries have not
structured budget classification systems to link to MDGs or national development objectives, but are
classified according to the ministry or agency implementing the spending.
Nevertheless, some sectors have relatively complete planned sector data, with education at 97% and
agriculture at 88%, reflecting the fact that these sectors are generally covered by clear separate
ministries or agencies. They have not improved much by 35% because they were already very high in
2013. Other sectors social protection (72%), environment (65%) and WASH (55%) are harder to
calculate because they are often split across multiple ministries and agencies, but there have been
increases of 710% in these sectors.119
Actual sector spending data are much lower on average (33%), and for all sectors (i.e. 76% education,
45% health, 40% agriculture see figure 5.4). The main reason for this gap is that some countries do not
publish actual spending data broken down by sector. Even more are subject to long delays, of 23 years,
because final actual data have to be approved by national audit courts before they can be published.
Data on types of expenditure should also allow citizens to judge whether their government is spending
more on recurrent costs or investment and, if disaggregated even further, would allow analysis of such
aspects as wage bills or equipment costs. Analysis of the split between recurrent and investment
spending is vital to assess whether investment spending is adequately increasing service provision or
productivity, and whether recurrent spending is supporting recurrent costs or maintenance.
However, sector-specific data disaggregated in this way are relatively unavailable. In most countries, the
lack of these data reflects: 1) a separation of responsibilities between recurrent and investment budgets,
with the former details being held in finance ministries and the latter in planning ministries or
commissions, which can hamper the compilation of joint data (though other countries with similar splits
manage to put data together successfully); and/or 2) lack of reporting from donors on the progress of
projects. There need to be greater efforts to compile overall disaggregated budgets and to enhance
donor reporting.
Data on sources of funds allow citizens to judge whether their government is allocating its own funds to
particular sectors, or relying on donor funding, allowing them to see whether their government is
reducing its aid dependence, the degree to which donors are supporting national priorities in different
sectors and (given aid volatility) the likely sustainability of spending.
Can spending be disaggregated to track equity?

Even more of a challenge is further disaggregation to track more detailed MDG (and potential SDG)
targets. Very few countries identify who they are targeting with spending sectorally, spatially or by
beneficiary the best performers being the five countries in group 5 in Figure 5.2. For example, GSW has
found that only 46% of countries split education by level in a way which means that primary education
can be identified as separate from total spending, even though primary education has been the main
MDG focus for the last 15 years. Similarly in the agriculture sector (as also discussed in Section 2), it is
very hard to track MDG-oriented spending, separating out nutrition or support to smallholders
(particularly vital as country studies have shown that government support is often mainly aimed at large
commercial export-oriented farming). In the health sector, it is virtually impossible to disaggregate
spending by type of disease or beneficiary related to the MDGs (e.g. maternal and child health); and in
the water sector, there is virtually no separate tracking of sanitation spending.
As also raised in Section 2, the extensions of this disaggregation needed to track the post-2015 SDGs are
even less prevalent. For instance, work carried out by GSW in 2014 to try to disaggregate education
spending in 45 low- and middle-income countries found that very few countries are ready for the
education SDGs in terms of tracking pre-primary education or early childhood development (only 13
countries); special education (only nine); vocational and technical education (15); or adult education and
literacy (seven).120
Some of these problems can be resolved by programme budgets, which match spending plans to
specific programme objectives, beneficiaries and outcomes. This would help to organise plans, budgets,
budget implementation reports and actual spending data, audits and impact reports around reporting
systems on intended (and actual) beneficiaries by age, gender, income, region and other classifications
designed to ensure equity. Only around 25 developing countries publish such budgets with varying
degrees of focus on beneficiaries and outcomes.
ACCOUNTABILITY DELIVERS RESULTS

GSW and its partners, the International Budget Partnership (IBP) and Oxfam, have recently conducted
research to examine whether increasing fiscal transparency and accountability have increased spending
and MDG results, in order to draw lessons to help inform the debate about the data revolution in the
post-2015 discussions.121 This research points to a growing body of evidence that suggests that
transparency, expenditure monitoring and accountability have contributed to increases in spending on,
and results related to, the MDGs.
Over the last decade, there has been a strong move by many governments to more open Greater budget
government, including a presumption that most governments will make all documents accountability
and data including those on plans and budgets transparently available to their is one of the
citizens; and a proliferation of right to information or freedom of information laws in most powerful
100 countries (up from just 12 in 1990). But transparency is not an end in itself for it to tools in
deliver results, it requires complex processes of accountability to work effectively. producing
MDG (and
Whether or not this occurs depends crucially not only on data availability, but also on potentially
space for civil society engagement, political will and government capacity. Case studies SDG) results.
across a number of countries show that increased transparency has often been a major
factor in increased and improved MDG budget allocations. This includes studies on
agriculture in Ghana and Nigeria; education in Argentina, Burkina Faso, the Dominican Republic, India,
Korea, Malawi and Tanzania; health in Armenia, Korea, Sierra Leone, South Africa and Zambia; maternal
health in Mexico, social protection in South Africa; water and sanitation in Sierra Leone; and marginalised
tribal groups in Gujarat and Dalits across India.122
Transparency has also often contributed to more effective and efficient spending, leading to improved
outcomes, through tracking surveys in more than 15 countries, including Malawi and Uganda, and quality
of service delivery surveys in more than 10 countries.123 Other social accountability tools have also
contributed to more effective and efficient spending: social audits in India and Kenya; citizen report cards
in India and Tanzania; procurement tracking in the Philippines; and auditing of actual spending by civil
society organisations (CSOs) in Mexico, the Philippines, Tanzania and South Africa.124 These cases show
dramatic increases in the share of funds reaching schools, clinics and water points, as well as delivery of
results.
These findings are supported by broader, multi-country studies on the impact of transparency and
accountability.125 However, they also indicate that higher spending and better outcomes depend on a
complex web of factors. Transparency is insufficient without accountability, which in turn depends on
both supply and demand factors:126
Demand factors include civil society space, access to information on plans and budgets, media
freedom and laws mandating participatory planning and budgeting processes. It also includes how
vocal and strong citizen voice and action to hold government accountable is, which depends on the
maturity and capacity of civil society actors, including their technical and advocacy skills.
Supply factors include state responsiveness, such as degree of democratisation, political will to
deliver spending and results, decentralisation and broader political accountability mechanisms.
Government capacity to deliver transparency, through parliaments, anti-corruption and public sector
procurement and auditing agencies, and procedures and skills for planning, budgeting and delivering
on the basis of performance goals are also vital supply side issues.
In general, demand and supply need to go together. However, cases such as Rwanda, where government
leadership and high capacity to deliver reforms have increased spending and results without major civil
society action, show that this is not necessarily the rule. In addition, although there have been few
successful civil society activities in countries with very low civil society space (as measured by the CIVICUS
Enabling Environment Index),127 accountability often produces results in countries with average-to-low
space. This indicates that budget accountability may be a vital lever to broaden civil society space.
Many case studies point to the importance of critical trigger events, such as corruption scandals,
economic crises, elections or changes of political regime.128 External influences, such as the MDGs,
participatory Poverty Reduction Strategies or donor inputs into social accountability, have also been
highly influential which should be a lesson when looking to the implementation frameworks for the
SDGs.
For sustained impact, it is also important that participation and accountability are fostered across the
whole planning, budgeting and delivery cycle to ensure that there are overall and sector development
plans. These plans require costed spending to reach the goals; medium-term outcome/performance-
based budgets with high allocations; in-year spending reports to track actual spending; participatory
mechanisms to track spending and fight corruption; transparent procurement and value-for-money
checking processes; and performance auditing and incidence analysis reports.
Box 6: Transparency can contribute to improved allocations and results
Budget transparency = allocation to disadvantaged groups

In India, the law requires that a percentage of spending should be targeted to support Dalits
(untouchables). In practice, however, this has often been disregarded. The National Campaign
for Dalit Human Rights (NCDHR) pushed the government to introduce a specific budget code to
track spending on programmes targeted to Dalits. Using this code, NCDHR helped to uncover
US$140 million of funds being diverted to cover the costs of the 2010 Commonwealth Games.
Following a public outcry, the government returned the funds to Dalit programmes.129
Revenue transparency = allocation to MDG spending
After discovering oil, Ghana passed a law that requires oil revenues to be managed
transparently. However, the US$2 billion of oil revenues subsequently raised did not translate
into increased public investment. In 2013, CSOs launched the Oil4Food campaign, which called
for oil revenues to be invested in supporting smallholder farmers. The campaign convinced the
government to commit 15% of oil revenues to smallholder agriculture.130
Transparency and accountability = allocation to health
In Zambia, the Vote Health campaign used a pre-election period to demand a dramatic increase
in health spending. This was based on government data which showed that the government was
falling way short of the agreed target (15% of spending). This generated significant media
coverage, widespread public engagement and political commitments by most candidates.
Following the elections, the new President raised spending by 45%, removed user fees and
employed 2,500 more health workers.131
Transparency and accountability = better allocations and results
In Malawi, the Civil Society Coalition for Quality Basic Education has a long history of tracking
education spending, including by conducting questionnaires with teachers and officials at the
community level. This has helped to increase funds to special education, reduce rural-urban
spending disparities, accelerate disbursement of teachers salaries and bring Malawi close to
achieving MDG2 on primary education.132
Quantitative evidence: accountability raises spending levels and

results
Despite severe data limitations, past quantitative analyses have provided some preliminary evidence
that transparency is associated with better health and water outcomes, and that higher spending can
improve MDG outcomes.133
More systematic analysis is now possible, due to new or recently updated datasets on spending and
transparency (see Box 7). Using these data (alongside the existing World Development Indicators for
MDG outcomes), GSW has examined three relationships: 1) whether greater budget transparency is
associated with better production of MDG data; 2) whether this is associated with higher spending on
MDG sectors; and 3) whether higher MDG spending is associated with better MDG outcomes.
Box 7. New data sources allow improved analysis
Since 2010, new or recently updated datasets have dramatically improved prospects for quantitative
analysis of the relationship between budget transparency, MDG spending and MDG outcomes.
On budget transparency, the International Budget Partnership updated and expanded the
coverage of its Open Budget Index (OBI) in 2012, to rank 100 countries based on the availability
of eight key budget documents.
On MDG spending data availability, Development Finance International has developed rankings
for 95 countries, based on the availability and level of detail of budget data in 2013. On MDG-
related spending, the Government Spending Watch database, developed by DFI and Oxfam,
brings together budget data for 70 countries for 200813, disaggregated by MDG sector. In
addition, the IFPRI Statistics of Public Expenditure for Economic Development (SPEED) database
has recently been updated to track expenditures through 2010 for 80 countries.
The main findings were that:
MDG expenditure tracking is more feasible in more transparent countries. Countries with more
transparent budgets on average also produce better-quality data for MDG tracking, as shown by a
strong correlation between the OBI and the DFI budget data availability index.
The link between transparency and MDG spending allocations is complex. Across the whole sample,
more transparent countries do not on average allocate a higher share of the budget to education,
health or water. However, countries which have recently significantly improved transparency have
also shown sharp improvements in MDG spending allocations.
Countries with higher expenditure on the MDGs have better MDG outcomes. There is a strong
positive correlation between per capita spending on education and health and MDG outcomes for
these sectors. This remains significant even after controlling for income levels. However, looking only
at low- and middle-income countries, income overrides this association.
Countries that have seen a strong improvement in budget transparency in the past decade have also
increased MDG spending faster and have seen faster MDG progress.
Nevertheless, the research also found that much stronger results might well be produced if spending data
were even more closely linked to each MDG result; if datasets were more comprehensive to provide a
bigger data sample; and if there were reliable ways to measure the enabling factors, including civil society
space, public participation in the budget process and government capacity and commitment.
IMPLICATIONS FOR MONITORING POST-2015

Discussions are currently under way on how progress should be monitored on the post-2015 framework
and the SDGs, and on success in mobilising the means of implementation (including financing). The need
for a data revolution has been reiterated throughout the post-2015 processes, but if the international
community and parliaments and citizens worldwide are to have a clear view of what is happening, this
revolution must ensure some low-cost, rapidly implementable steps to improve the monitoring of all
aspects of budgets to enable better scrutiny. This needs to also include information on government
revenues, aid and other external and domestic budget financing. As shown above, this can be expected to
have a major impact on both spending levels and ultimately (and importantly) on outcomes.
There are four sets of quick wins which could dramatically accelerate progress in this area:
1. Publishing documents and data that governments already produce: As shown by IBPs Open Budget
Survey and Tracker, timely, accessible and comprehensive publication of budget documents can
occur with no or little extra time and cost. These documents would need to include (if available) the
Pre-Budget Statement, Executives Budget Proposal, Enacted Budget, Citizens Budget, In-Year Report,
Mid-Year Review, Year-End Report and Audit Report.
2. Improving and publishing data and documents on spending:

Improved matching of spending with each of the post-2015 SDGs by each country improving its
disaggregation or aggregation in budgets, e.g. publishing sub-sectoral disaggregation with common
codes; programme/results-based budgets, etc. Plans to improve data must be tailored to what is
feasible in each country, along the lines of the improvements discussed in relation to Figure 5.2
above;
Publishing in-country regional disaggregations of spending so as to facilitate tracking of whether
allocations are combating in-country inequalities (e.g. urban-rural, poorest regions);
More timely publication of actual spending reports by accelerating validation and auditing
processes, and by publishing preliminary unaudited data where necessary;
Publishing budgets by beneficiary, combining gender-responsive and child-responsive data;
analysing the degree of gender/age equality in spending as well as inclusion of other groups such as
the disabled, elderly, etc.
3. Improving and publishing data and documents on revenue:

More detailed annual publication of revenue receipts by type of tax, sector, size of enterprise, etc.;
Systematic annual publication of revenue losses (otherwise known as tax expenditures) occurring
due to exemptions and incentives, as well as of lists of companies granted exemptions;
Publication of national tax codes and the compilation of a global database on tax rates and
thresholds to monitor global harmful competition;
Regular analysis of the incidence of tax (and spending) policies to assess whether they are
combating inequality;
Publication by all development financing institutions of the tax revenues mobilised by the projects
they are funding, and of the exemptions they have requested for projects (and the reasons for
these).
4. Improving and publishing data on aid and other budget financing:

Accelerating efforts at compatibility/similar codings between IATI and national aid and budget
reporting systems, to ensure that IATI and aid monitoring systems are useful for budget planning;
Automatic reporting via the DAC and IATI of whether specific aid and other official finance flows are
on-budget in recipient countries, to make global statistics more relevant to/compatible with
national accountability;
Accelerating efforts to improve country-level collection of data from providers of development
finance, including SouthSouth cooperation and CSOs/foundations;
Publishing and tracking all loan agreements and their implications for debt service and crowding out
of post-2015 spending;
Publishing off-budget contingent liabilities such as public-private partnership agreements which
could have major implications for potential debt service.
Finally, it is worth noting that a lack of data is mostly due not to governments lack of willingness to make
information available, but to a lack of technical capacity or institutional ability to change traditional
budget practices. There will need to be a dramatic scaling up of capacity-building support to governments
in this area in which GSW can play a part so that they are able to produce their own data for national
and global monitoring. As part of that, CSOs should focus more on comparative analysis and research.
There will also need to be a concerted process among UN agencies to ensure much more real-time
monitoring of budgets and means of implementation, rather than the current lengthy surveys which take
23 years to deliver results.
Lack of data also reflects insufficient demand from parliaments and citizens. Again there is no lack of will
to receive such data. Instead, it is often the case that such groups are not sufficiently informed of best
practice in other countries or of what is technically possible, or sufficiently well organised to demand
data powerfully and successfully. GSW will be working more intensively during 201516 on building
citizen pressure and voice, working with country coalitions to build their knowledge and capacity to more
effectively hold their governments accountable, using the data we collect to put tools in the hands of
citizens which can turn numbers into nurses and become a key weapon in the fight against poverty and
inequality.
NOTES
1
The eight MDGs range from halving extreme poverty rates to halting the spread of HIV/AIDS and providing universal
primary education, all by the target date of 2015. They form a blueprint agreed to by all the worlds countries and all its
leading development institutions. More information on the MDGs is available here: http://www.un.org/millenniumgoals/
2
For more information on the seven sectors that GSW tracks and their links to the MDGs, visit:
http://www.governmentspendingwatch.org/campaigns-and-advocacy/8-govt-spending-site/28-about-spending-goals
3
M. Martin and R. Watts (2013) Putting Progress at Risk? MDG spending in developing countries, available here:
http://www.governmentspendingwatch.org/research-analysis/latest-analysis/51-mdg-spending-in-developing-countries.
GSW data have also been utilised in numerous other reports and analyses by civil society organisations (i.e. Global
Campaign for Education, Oxfam, Save the Children International and WaterAid) and UN agencies (UN Women, UNESCO
GMR).
4
The GSW database examines all low-income countries (LICs) and lower-middle-income countries (LMICs) that have data
that are transparently available and analysable. It also analyses some upper-middle-income countries (UMICs). The long-
term plan of GSW is to incrementally expand data into all developing countries (where data are available). Of the 66
countries in the GSW database where budget information is available, 28 are LICs, 29 are LMICs and nine are middle-
income countries (MICS).
5
Available at: http://www.governmentspendingwatch.org/spending-data
6
The Third International Conference on Financing for Development will be held in Addis Ababa, Ethiopia, from 13 to 16 July
2015. More information is available here: http://www.un.org/esa/ffd/overview/third-conference-ffd.html
7
For a general overview of expenditure data definitions and sources, please visit the sources and data section of the GSW
website here: http://www.governmentspendingwatch.org/spending-data/8-govt-spending-site/7-definitions-and-sources
8
Low-income countries in the GSW database: Afghanistan, Bangladesh, Benin, Burkina Faso, Burundi, Cambodia, Central
African Republic, DRC, Ethiopia, Guinea-Bissau, The Gambia, Haiti, Kenya, Liberia, Lesotho, Madagascar, Malawi, Mali,
Mozambique, Nepal, Niger, Rwanda, Sierra Leone, Tajikistan, Tanzania, Togo, Uganda and Zimbabwe. Lower-middle
income countries: Armenia, Bhutan, Cameroon, Cape Verde, Congo, Cte dIvoire, Djibouti, El Salvador, Ghana, Guatemala,
Guyana, Honduras, India, Kiribati, Moldova, Nicaragua, Nigeria, Papua New Guinea, Samoa, So Tom and Prncipe,
Senegal, Solomon Islands, Sri Lanka, Swaziland, Timor Leste, Vanuatu, West Bank and Gaza, Yemen, Zambia. Middle-
income countries: Angola, Colombia, Comoros, Ecuador, Dominican Republic, Jamaica, Jordan, Peru and Tonga.
9
This section of the report carries out an in-depth analysis of only six of the seven MDG sectors that GSW tracks, and
leaves out an analysis of MDG-related gender spending (the seventh GSW MDG sector). After analysing spending on
gender, the GSW team concluded that the data available were too weak to draw robust conclusions or new trend analysis
(not already covered by the GSW 2013 report). However, this does raise serious concerns by the GSW team around data on
gender budgets and allocations by governments: therefore we cover gender more extensively in section 4, in terms of the
lack of gender-sensitive budgeting and the implication for inequality, and in section 5 in a discussion around the lack of
data transparency.
10
Data in this section are sourced from the latest IMF World Economic Outlook, Regional Economic Outlook and country
documents available on the IMF website as of 28 February 2015.
11
Spending on the MDGs is defined as spending on the seven sectors analysed in this report agriculture and food,
education, environment, health, social protection, gender and WASH. Due to a lack of data in some sectors, the total level
of MDG spending presented in this section may be a slight underestimate.
12
Spending on MDGs is used regularly throughout this report as shorthand to refer to spending across the seven sectors
that GSW tracks.
13
Data on debt servicing are available for 64 countries, and are drawn from IMF LIC-Debt Sustainability Framework and
MIC-Debt Sustainability Framework annexes, as well as other budget tables, in IMF staff country reports for 2014 and 2015.
14
Data on defence spending are available for 44 GSW countries. Information from the Stockholm International Peace
Research Institute (SIPRI) at: http://www.sipri.org/research/armaments/milex/milex_database
15
FAO (2014) The State of Food Insecurity in the World 2014, Rome: UN Food and Agriculture Organization. Available
here: http://www.fao.org/publications/sofi/en/
16
This is from 199092 levels, as measured by FAO.
17
FAO (2014) The State of Food Insecurity in the World 2014, op. cit.
18
In 2012, a quarter of all children under the age of five years were estimated to be stunted i.e. having inadequate height
for their age. United Nations (2014) The Millennium Development Goals Report. Available here:
http://www.un.org/millenniumgoals/2014 MDG report/MDG 2014 English web.pdf
19
Spending on agriculture is not necessarily the only measure of spending that can address hunger; however, in the
absence of other clear commitments, budgetary lines and sectoral spending which specifically aims to target hunger, this
currently serves as the best way to track spending across multiple countries. Moreover, analysts agree that, especially in
low-income countries, government support to agriculture can provide crucial support to addressing hunger. Considerable
work needs to be done on improving the ability to track spending targeted at addressing hunger and, in the SDGs, spending
aimed at reducing hunger, improving nutrition and improving sustainable agriculture. This will be touched on later in
Section 5 on data/transparency issues.
20
This was initially a five-year commitment and was later reconfirmed to be met by 2015. For more information, see:
http://www.nepad.org/nepad/knowledge/doc/1787/maputo-declaration
21
Data for what countries spend on agriculture vary from figures collected by other agencies, due to different
categorisations of spending. For instance, Ghana comes out as having higher spending in reporting to the Comprehensive
Africa Agriculture Development Programme (CAADP), as it includes spending on rural roads and development.
22
World Bank (2007) World Development Report 2008: Agriculture for Development. Washington DC: The International
Bank for Reconstruction and Development/The World Bank, available at:
http://siteresources.worldbank.org/INTWDR2008/Resources/WDR_00_book.pdf
23
See M. Martin and R. Watts (2013) Putting Progress at Risk? MDG spending in developing countries, op. cit.
24
J.D. Sachs and G. Schmidt-Traub (2014) Financing for Sustainable Development: Implementing the SDGs through
Effective Investment Strategies and Partnerships. New York: Sustainable Development Solutions Network (SDSN)
25
See the SUN movement document at: http://scalingupnutrition.org/wp-
content/uploads/2014/08/140703_EN_SUMMARY-3-STEP-APPROACH_Tracking_Domestic_Investments_for_Nutrition.pdf,
as well as Chapter 7 of the IFPRI-led Global Nutrition Report 2014, available at:
http://www.ifpri.org/sites/default/files/publications/gnr14.pdf
26
The Education For All (EFA) goals are six internationally agreed education goals that aim to meet the learning needs of all
children, youth and adults by 2015. An overview of the goals is available here:
http://www.unesco.org/new/en/education/themes/leading-the-international-agenda/education-for-all/efa-goals/. The
MDGs meanwhile aim to ensure that all children are in primary school by 2015; more details here:
http://www.un.org/millenniumgoals/education.shtml
27
UNESCO Institute of Statistics online tables. http://stats.uis.unesco.org/unesco/TableViewer/tableView.aspx
28
UNESCO (2011) Global Monitoring Report: The hidden crisis: Armed conflict and education. Highlights available here:
http://www.unesco.org/new/fileadmin/MULTIMEDIA/HQ/ED/pdf/gmr2011-highlights.pdf
29
Although the target is for GNP, GSW measures the commitment against GDP and how close or how far countries are
from meeting the 6% target, to keep consistency across the analysis.
30
See GSW 2013 Report: M. Martin and R. Watts (2013) Putting Progress at Risk? MDG spending in developing countries,
op. cit.
31
UNESCO (2010) EFA Global Monitoring Report 2010: Reaching the Marginalized.
32
Information on the Muscat agreement is available here: https://efareport.wordpress.com/2014/06/04/the-muscat-
agreement-new-proposed-post-2015-global-education-goal-and-targets-announced-today/
33
Education for All Global Monitoring Report Policy Paper, 18 March 2015, Pricing the right to education: The cost of
reaching new targets by 2030. Available here: http://unesdoc.unesco.org/images/0023/002321/232197E.pdf
34
These figures are from 200010, and are the latest available. United Nations (2014) The Millennium Development Goals
Report, op. cit.
35
This is based on a range of estimates used by Sachs and Schmidt-Traub (2014). They proposed using numbers cited by
UNCTAD (2014), which estimated total financing needs for mitigation of $550880 billion and a financing gap of some
$380680 billion, as broadly consistent with a number of other estimates on climate mitigation needs within a mid-point of
available estimates from other studies. They then used ratios projected by the Green Growth Alliance for private financing
and public finance to estimate that $80115 billion was necessary in public finance. They also noted: Clearly, though, this
situation remains unsatisfactory, and more work is required to understand the differences in estimates and to identify a
consensus range for the post-2015 agenda.
36
The Green Growth Alliance (2013) and UNCTAD (2014) project that at least $80120 billion must be invested annually in
adaptation to climate change, based on these estimates and current investments. Sachs and Schmidt-Traub (2014), op. cit.
37
Based on Global Environmental Forum figures from its 2014 replenishment conference (GEF, 2014) quoted in Sachs and
Schmidt-Traub (2014), op. cit.
38
United Nations (2014) The Millennium Development Goals Report, op. cit.
39
Commission on Macroeconomics and Health. Macroeconomics and Health: Investing in Health for Economic
Development. Chaired by Jeffrey D. Sachs. Report of the Commission on Macroeconomics and Health. 2001. Geneva,
World Health Organization.
40
This was estimated to be US$44 per capita needed to strengthen health systems as well as provide essential services in
49 LICs in 2009; with inflation and increased diseases this would need to rise to US$60 per capita by 2015. Taskforce on
Innovative International Financing for Health Systems (2010) Constraints to scaling up the health Millennium Development
Goals: costing and financial gap analysis, Geneva: Working Group 1 Report.
41
According to WHO, between 2010 and 2012 just six of the 43 sub-Saharan African countries for which there are data had
met or exceeded the Abuja target on average: Liberia, Malawi, Rwanda, Swaziland, Togo and Zambia.
42
African Union (AU) data suggest that five countries (Burkina Faso, Malawi, Niger, Rwanda and Zambia) are meeting the
target. However, these use a methodology which adds a proportion of off-budget aid into government spending. GSW does
not include this, because it tracks the amount of funding allocated to health in the budget. In its most recent assessment of
progress on the Abuja target, WHO states that Rwanda is the only low-income African country meeting the budget target.
43
R. Moreno-Serra and P. Smith (2012) Does progress towards universal health coverage improve population health?, The
Lancet 380(9845): 917-923; W.D. Savedoff (2012) Transitions in Health Financing and Policies for Universal Health
Coverage, Results for Development Institute; Jamison et al. (2013) Global health 2035: a world converging within a
generation, The Lancet, 9 382:9908, 1989-1955. Available here:
http://globalhealth2035.org/sites/default/files/report/global-health-2035.pdf; I. Agyepong, G. Liu, S. Reddy et al. (2014)
Health in the Framework of Sustainable Development; Commission on Macroeconomics and Health (2001)
Macroeconomics and Health: Investing in Health for Economic Development. Chaired by Jeffrey D. Sachs. Report of the
Commission on Macroeconomics and Health, Geneva: World Health Organization.
44
For instance, a study of 89 countries suggested that the more a country relies on out-of-pocket investments, the more of
its households face financial catastrophe.
45
Jamison et al. (2013) Global health 2035: a world converging within a generation, op. cit.
46
Initially this was US$1 per day, but was revised in 2010 to US$1.25. See:
http://econ.worldbank.org/external/default/main?theSitePK=469382&contentMDK=22510787&menuPK=574960&pagePK
=64165401&piPK=64165026
47
Social protection can include a range of policies that provide social safety nets, social funds, social welfare
assistance/services, labour market interventions and social insurance programmes (including pensions), which act as an
insurance policy against poverty, or a tool for helping the poorest or most vulnerable to manage risks, access basic services
or tackle very extreme poverty and hunger (essentially tackle the worst forms of deprivation). GSWs social protection
includes any programme which is a social transfer, either in cash or in kind, i.e. a transfer of income or services, from one
group in society to another, such as from the working population to the old, the healthy to the sick or the affluent to the
poor.
48
GSW deals with the hunger target of MDG1 through tracking food and agriculture spending. However, social protection
spending often tackles extreme poverty and hunger indirectly by allowing the poor to spend more income on food, or
directly by supporting improved nutrition as a sub-goal, especially for marginalised or disaster-hit groups. A lot of extreme
nutritional needs are addressed through social protection ministries; this is a very important issue going forward for the
SDGs, which will need extensive work.
49
In this report, social protection excludes all social services provided by government that could be classified as education
or health, nutrition or WASH. This is consistent with the social transfer element of the UN social protection floor concept;
however, this clearly excludes other areas included in the UN social protection floor, which guarantees a minimum that,
over the life cycle, all in need have access to essential healthcare and basic income security, including access to essential
healthcare, including maternity care; basic income security for children, providing access to nutrition, education, care and
any other necessary goods and services; and basic income security for persons in active age who are unable to earn
sufficient income, in particular in cases of sickness, unemployment, maternity and disability. This includes contributory
systems which are not included in GSW, as mentioned in the report.
50
Given that social protection programmes are a vital tool in ensuring greater equality and protecting against risk or
poverty (as well as fostering a contract between citizen and state), they will be vital to achieving the SDGs, and hence GSW
will move towards ensuring a more robust measurement for social protection as one means to track spending on the SDGs.
51
International Labour Organization (2014) World Social Protection Report 201415: Building economic recovery, inclusive
development and social justice. http://www.ilo.org/global/research/global-reports/world-social-security-
report/2014/lang--en/index.htm
52
This finding is in line with the ILOs World Social Protection Report 2014 15, which shows that a number of MICs are
expanding their social protection systems and many LICs are organising policy dialogue around how to build social
protection floors.
53
Ibid.
54
Lower-end estimate from L. Chandy and G. Gertz (2011) Poverty in Numbers: The Changing State of Global Poverty from
2005 to 2015, Washington DC: Brookings Institution, available at:
http://www.brookings.edu/~/media/research/files/papers/2011/1/global%20poverty%20chandy/01_global_poverty_chan
dy.pdf. Higher-end estimate from R. Greenhill, P. Carter, H. Manuel and C. Hoy (2015) Financing the Future, Overseas
Development Institute, London, available at: www.odi.org.uk
55
United Nations (2014) The Millennium Development Goals Report, op. cit.
56
WHO, UNICEF (2013) Progress on Sanitation and Drinking-Water 2013 Update, Joint Monitoring Programme for Water
Supply and Sanitation. Available at: http://apps.who.int/iris/bitstream/10665/81245/1/9789241505390_eng.pdf
57
It should be noted that this figure may be well below what is necessary, in 2011 WaterAid estimated that LICs in Africa
far from the target would need to spend 3.5% of GDP on WASH to meet the MDGs, and off-track South Asian countries
would need to allocate a minimum of 1% of GDP to sanitation. WaterAid (2011) Off-track, off-target: Why investment in
water, sanitation and hygiene is not reaching those who need it most.
http://www.wateraid.org/~/media/Publications/off-track-off-target-report-wateraid-america.ashx
58
This became known as the eThekwini Declaration commitment; available at:
http://www.unicef.org/wash/files/WA_eThekwini_ENGLISH_FINAL.pdf
59
Recommendation from the UN Human Development Report 2006.
60
WHO, UNICEF (2013) Progress on Sanitation and Drinking-Water 2013 Update, op. cit.
61
UNCTAD (2014) World Investment Report 2014. Investing in the SDGs: An Action Plan. UNCTAD.
62
WaterAid (2013), WHO and UN Water (2012), World Bank (2011) (see above for references).
63
WHO (2011). This is on the basis of a total need of US$535 billion to be spread over 20 years. UNCTAD (2014) projects a
much higher investment gap for access to water and sanitation of some US$260 billion.
64
As discussed in more detail in the section on education above, no reliable estimates exist for the funding needs for the
education SDGs. The US$22 billion figure quoted here is a very substantial underestimate.
65
The non-MDG sector estimates are based on two sources: 1) UNCTADs estimates UNCTAD (2014) World Investment
Report 2014. Investing in the SDGs: An Action Plan; 2) estimates/calculations within the SDSN report on financing needs
for the SDGs: J.D. Sachs and G. Schmidt-Traub (2014) Financing for Sustainable Development: Implementing the SDGs
through Effective Investment Strategies and Partnerships, op. cit.
66
Developing country data from IMF WEO database, October 2014. GSW country data from GSW database.
67
Estimate compiled from OECD and UN Development Cooperation Forum sources. Of this, US$151 billion represents DAC
countries gross ODA, US$24 billion non-DAC countries aid and US$33 billion private flows.
68
Data downloaded from the OECD CRS database on 10 March 2015, at http://stats.oecd.org/Index.aspx?ThemeTreeId=3#
69
The figure on country programmable aid is from the latest (2014) DAC survey on donor country aid plans, available at
http://www.oecd.org/development/effectiveness/aidpredictability.htm, and on the figure for aid channelled via budgets is
from the progress report on the implementation of the Busan Declaration, Making Development Cooperation More
Effective, available at: http://www.keepeek.com/Digital-Asset-Management/oecd/development/making-development-co-
operation-more-effective_9789264209305-en#page43
70
See A. Bulir and A.J. Hamann (2006) Volatility of Development Aid: From the Frying Pan Into the Fire?, IMF Working
Paper WP/06/65, at: http://www.imf.org/external/pubs/ft/wp/2006/wp0665.pdf; and H. Kharas (2008) Measuring the
Cost of Aid Volatility, Brookings Institution, at: http://www.brookings.edu/research/papers/2008/07/aid-volatility-kharas
71
A more detailed document showing all the calculations in this section and listing all the sources and definitions for the
data is available on request from DFI.
72
For more details on this, see also pp.66-67 of the Global Financial Governance & Impact Report 2014 produced by New
Rules for Global Finance, available at: http://www.new-
rules.org/storage/global_financial_governance_and_impact_report_2014.pdf
73
R. Greenhill, P. Carter, H. Manuel H and C. Hoy (2015) Financing the Future, Overseas Development Institute, op. cit.
74
Some authors have suggested that SouthSouth providers should agree to targets linked to GNI, but providers have long
rejected this as being a set of targets applicable to DAC/OECD donors, and this does not look likely to change for the FfD
2015 agreement.
75
These are by no means the only potential sources of global tax revenues others have suggested taxes on mobile
telephony, information technology and tobacco. It is important to note that they are also increasingly being implemented
by developing countries themselves, supplementing their own budget revenues.
76
For more on this, see ActionAid/BOND/CAFOD/Eurodad/Oxfam/WWF (2015) Delivering Sustainable Development,
discussion document; and M. Martin, Effectiveness Standards for Private and Blended Finance, policy brief for the
Development Cooperation Forum.
77
OECD DAC 2013 aid figures, latest available: http://www.oecd.org/dac/stats/
78
DFI/WaterAid (forthcoming); and WaterAid (2011) Off-track, off-target, op. cit., based on case studies by DFI.
79
On average and taking into account population size income inequality increased by 11% in developing countries
between 1990 and 2010. UNDP (2014) Humanity Divided: Confronting Inequality in Developing Countries.
http://www.undp.org/content/undp/en/home/librarypage/poverty-reduction/humanity-divided--confronting-inequality-
in-developing-countries.html
80
Ibid.
81
United Nations Economic Commission for Africa (2014) Assessing Progress in Africa toward the Millennium
Development Goals.
82 ,
Y. Xie and X. Zhou (2014) Income Inequality in Todays China, University of Michigan.
http://www.bloomberg.com/news/2014-04-28/gap-between-rich-poor-worse-in-china-than-in-u-s-study-shows.html
83
OECD (2011) Special Focus: Inequality in Emerging Economies (EEs).
84
E. Seery and A. Arendar (2014) Even it Up: Time to end extreme inequality, Oxfam. http://policy-
practice.oxfam.org.uk/publications/even-it-up-time-to-end-extreme-inequality-333012
85
Ibid. Data based on World Bank, World Development Indicators, http://data.worldbank.org/data-catalog/world-
development-indicators
86
Tax Justice Network Africa/Christian Aid (2014) Africa rising? Inequalities and the essential role of fair taxation.
http://www.christianaid.org.uk/images/Africa-tax-and-inequality-report-Feb2014.pdf
87
UNDP (2014) Humanity Divided: Confronting Inequality in Developing Countries, op. cit.
88
For instance, OECD (2011) Divided We Stand: Why Inequality Keeps Rising,
http://www.oecd.org/social/soc/dividedwestandwhyinequalitykeepsrising.htm; Asian Development Bank (ADB) (2014)
ADBs Support for Inclusive Growth, Thematic Evaluation Study, http://www.adb.org/documents/adbs-support-inclusive-
growth; and J. Ostry, A. Berg and C. Tsangardies (2014) Redistribution, Inequality and Growth, IMF staff discussion note,
http://www.imf.org/external/pubs/ft/sdn/2014/sdn1402.pdf; UNDP (2014) Humanity Divided: Confronting Inequality in
Developing Countries, op. cit.
89
L. Chandy (2013) The Final Countdown: Prospects for Ending Extreme Poverty By 2030, Washington, DC: The Brookings
Institution.
http://www.brookings.edu/~/media/Research/Files/Reports/2013/04/ending%20extreme%20poverty%20chandy/The_Fin
al_Countdown.pdf
90
Unpublished calculations based on the methodology and model developed in L. Chandy (2013) op. cit.
91
P. Edward and A. Sumner (2013) The Future of Global Poverty in a Multi-Speed World, Kings College London
International Development Institute, p.83,
http://www.kcl.ac.uk/aboutkings/worldwide/initiatives/global/intdev/people/Sumner/Edward-Sumner-
Version04March2013.pdf (NB: These estimates refer to US$2 poverty, not US$1.25 poverty.)
92
Wilkinson and Picketts research focused on OECD countries (a grouping of rich countries), yet the same negative
correlation between inequality and social well-being holds true in poorer countries.
93
UNESCO (2010) Reaching the Marginalized. EFA Global Monitoring Report.
94
WaterAid (2011) Off-track, off-target: Why investment in water, sanitation and hygiene is not reaching those who need
it most; WHO (2012) Global Costs and Benefits of Drinking-Water Supply and Sanitation Interventions to Reach the MDG
Target and Universal Coverage. Geneva: WHO.
95
.C. Averil (2013) Universal Health Coverage: Why health insurance schemes are leaving the poor behind. Oxfam.
96
Oxfam (2014) Working for the Many: Public services fight inequality.
http://www.oxfam.org/sites/www.oxfam.org/files/bp182-public-services-fight-inequality-030414-en.pdf
97
This figure is taken from Oxfam Internationals report Working for the Many (2014). it is based on work by N. Lustig
(2012) on inequality in Latin America, which found that investing in public services has a significant impact on tackling
inequality, even in countries where taxation is regressive and not fulfilling its redistributive potential, as shown in this
graph.
98
It is evident that similar analysis of spending on smallholder agriculture, nutrition and WASH would reach the same
conclusions, but such analysis has not yet been conducted, in part due to data problems.
99
J. Ostry et al. (2014) Redistribution, Inequality and Growth, op. cit.
100
Save the Children International (2014) More Is Not Enough: Achieving Equity in Domestic Education Financing.
101
The relationship between health and inequality is complex, but it is widely acknowledged that health outcomes are
worse for the poorest, and in more unequal societies. See M. Marmot (2008) Report of the Commission on Social
Determinants of Health. Meanwhile inequality in health is threatening progress on UHC and is worsening in many
developing countries. See Committee for Development Policy, Implementing the Millennium Development Goals: Health
Inequality and the Role of Global Health Partnerships.
102
T. Tacke and R. Waldmann (2013) Infant mortality, relative income and public policy, Applied Economics, Vol 45, Issue
22. Available at: http://www.tandfonline.com/doi/abs/10.1080/00036846.2012.705429
103
D. McIntyre, M. Thiede, G. Dahlgren, M. Whitehead (2006) What are the economic consequences for households of
illness and of paying for health care in low- and middle-income country contexts?, Social Science and Medicine
2006;62(4):858-65.
104
A. Marriott (2014) A Dangerous Diversion: Will the IFCs flagship health PPP bankrupt Lesothos Ministry of Health?,
Oxfam, http://oxf.am/5QA
105
Xu et al. (2007) Protecting households from catastrophic health expenditures, Health Affairs 26(4): 97283,
http://content.healthaffairs.org/content/26/4/972.full
106
Ibid.
107
Lesotho has been badly hit by the slowdown in the South African economy and by a large reductions in receipts from
the Southern African Customs Union (SACU).
108
M. Martin and R. Watts (2014) Investment in Children: A Global Policy Background Report, mimeo. UNESCO report.
Development Finance International: Background paper prepared for the Education For All Global Monitoring Report 2015.
Trends in government expenditure for public education, 201113, available at: http://www.unesco.org/ulis/cgi-
bin/ulis.pl?catno=232476&set=0054198FA8_0_36&gp=1&lin=1&ll=1
109
EFA 2013/14 Global Monitoring Report, available at: http://www.unesco.org/new/en/education/themes/leading-the-
international-agenda/efareport/reports/2013/2013-report-epub-en
110
J. Walker (2013) A Taxing Business: Financing Education For All Through Domestic Resources, Global Campaign for
Education. http://www.campaignforeducation.org/docs/reports/GCE_A TAXING BUSINESS.pdf
111
See the UNESCO World Inequality Database on Education (WIDE) for the completion rates of different economic
quintiles. http://www.education-inequalities.org/
112
WHO, UNICEF (2013) Progress on Sanitation and Drinking-Water 2013 update, op. cit.
113
UN (2012) The Millennium Development Goals 2012.
114
See, for example, UNICEF (2010) Narrowing the Gaps to Meet the Goals.
115
M. Martin and R. Watts (2014) Investment in Children: A Global Policy Background Report, mimeo.
116
GSW is rigorous in answering the question of how much data is available, excluding data which are not deemed to be of
sufficient quality or comprehensiveness to include in the GSW database especially where data include only a small sub-
component of spending on a particular MDG or where there is not enough disaggregation to make it possible for GSW to
identify data linked to an MDG.
117
Low-income countries in GSW database: Afghanistan, Bangladesh, Benin, Burkina Faso, Burundi, Cambodia, Central
African Republic, Comoros, DRC, Ethiopia, The Gambia, Guinea-Bissau, Haiti, Kenya, Liberia, Madagascar, Malawi, Mali,
Mozambique, Nepal, Niger, Rwanda, Sierra Leone, Tajikistan, Tanzania, Togo, Uganda and Zimbabwe. Lower-middle-
income countries: Armenia, Bhutan, Cameroon, Cape Verde, Congo, Cte dIvoire, Djibouti, El Salvador, Ghana, Guatemala,
Guyana, Honduras, India, Kiribati, Lesotho, Moldova, Nicaragua, Nigeria, Papua New Guinea, Samoa, So Tom and
Prncipe, Senegal, Solomon Islands, Sri Lanka, Swaziland, Timor Leste, Vanuatu, West Bank and Gaza, Yemen, Zambia.
Upper-middle-income countries: Angola, Colombia, Ecuador, Dominican Republic, Jamaica, Jordan, Peru and Tonga.
118
This is based on availability of data according to the seven different sectors which GSW tracks online, including
agriculture, education, environment, gender, health, social protection and WASH. We have measured each countrys
performance against each of these, in spite of not covering gender in great detail in the report (because the data
availability is so weak), and we have judged all countries across all these areas hence the percentage across these seven
areas is calculated in terms of data available for the different breakdowns in these sectors (i.e. capital/recurrent,
donor/government) and for budgeted and actual data.
119
Some apparently MDG-oriented spending may also not be relevant. For example, water spending may also be targeted
at a number of non-MDG-related activities including water conservation or large dams whose primary purpose is
generating energy so we exclude this spending where possible.
120
Development Finance International: Background paper for the EFA Global Monitoring Report 2015, Trends in
government expenditure for public education, 201113, op. cit.
121
This study is based on a joint paper produced by DFI, the International Budget Partnership and Oxfam in October 2014,
From Numbers to Nurses: Why Budget Transparency, Expenditure Monitoring, and Accountability are Vital to the Post-
2015 Framework, and a more detailed background study prepared for IBP and DFI by Rebecca Simson. More details of this
analysis will be available shortly in R. Simson (forthcoming) Transparency for Development: the relationship between
budget transparency, MDG spending and results, available at www.internationalbudget.org and
www.governmentspendingwatch.org
122
R. Carlitz and R. McGee (2013) Raising The Stakes: The Impact of HakiElimus Advocacy Work on Education Policy and
Budget in Tanzania, IBP Partnership Initiative Case Study Series No. 12, available at: http://internationalbudget.org/wp-
content/uploads/HakiElimu-Case-Study.pdf. On Argentina: http://internationalbudget.org/publications; Armenia:
http://governmentspendingwatch.org/campaigns-and-advocacy/health; Burkina Faso:
http://www.campaignforeducation.org/docs/CSEF; Dominican Republic:
www.governmentspendingwatch.org/campaigns-and- advcacy/education; Gujarat, Mexico and South Africa social
protectio:, M. Robinson (2006) Budget Analysis and Policy Advocacy: The Role of Non-Governmental Public Action, IDS
Working Paper 279, Brighton: IDS; India education: http://www.governmentspendingwatch.org/campaigns- and-
advocacy/education; Korea: Lee and You (2013) Country Report: South Korea, report commissioned by the Global
Initiative for Fiscal Transparency; Sierra Leone: http://www.governmentspendingwatch.org/campaigns-and-
advocacy/water-and-sanitation; and South Africa health: http://internationalbudget.org/publications. For Ghana, India,
Malawi and Zambia, see additional sources in notes 1619.
123
G. Baiocchi, P. Heller, S. Chaudhuri and M. Kunrath Silva (2006) Evaluating Empowerment: Participatory Budgeting in
Brazilian Municipalities, in R. Alsop, M. Frost Bertelsen and J. Holland (eds), Empowerment in Practice: From Analysis to
Implementation, Washington: World Bank; and A. Schneider and M. Baquero (2006) Get What You Want, Give What You
Can: Embedded Public Finance in Porto Alegre, Brighton: Institute of Development Studies.
124
African Development Bank (2009) Debt Relief Initiatives, Development Assistance and Service Delivery in Africa.
Oxford: Oxford University Press; on PETS and QSDS, see J. Dehn, R. Reinikka and J. Svensson (2003) Survey Tools for
Assessing Performance in Service Delivery, in F. Bourguignon and L.A. Pereira da Silva (eds) The Impact of Economic
Policies on Poverty and Income Distribution, World Bank/OUP; and for a more nuanced view from IBP, see
http://internationalbudget.org/wp-content/uploads/2011/01/newsletter44.pdf. See also
http://internationalbudget.org/publications, and especially http://internationalbudget.org/wp-
content/uploads/2011/01/IBP-Expenditure-Monitoring-Guide.pdf, for more discussion of social audits in India and Kenya,
citizen scorecards in India, Kenya and Tanzania, procurement monitoring in the Philippines, auditing of expenditures in
Mexico, the Philippines, Tanzania and South Africa, and other case studies on nutrition in Ghana and education and
health in Uganda. On participation in Kerala, see T. Isaac and R. Franke (2002) Local Democracy and Development: The
Kerala Peoples Campaign for Decentralized Planning, Lanham, MD: Rowman & Littlefield Publishers, Inc; and on Kenya,
J. Ouma (2012) Where Does the Money Go: Citizen participation in Turkana County, Kenya, Oxfam Insights Series.
125
J. Gaventa and G. Barrett (2010) So What Difference Does it Make: Mapping the Outcomes of Citizen Engagement,
CPADRC Working Paper 347, Brighton: IDS; A. Joshi (2011) Review of Impact and Effectiveness of Transparency and
Accountability Initiatives: Annex 1 Service Delivery, available at: www.transparency-initiative.org; S. Khagram, A. Fung
and P. de Renzio (2013) Open Budgets: The Political Economy of Transparency, Participation and Accountability,
Brookings Institution Press; R. McGee and J. Gaventa (2011) Synthesis report: Review of impact and effectiveness of
transparency and accountability initiatives, London: Transparency and Accountability Initiative; M. McNeil and C. Malena
(2010) Demanding Good Governance: Lessons from Social Accountability Initiatives in Africa, Washington DC: World
Bank; and World Bank Institute (2014) Evidence of Development Impact from Institutional Change: A Review of the
Evidence on Open Budgeting.
126
G. Bergh et al. (2012) Building governance into a post-2015 framework: Exploring transparency and
accountability as an entry point, London: ODI. http://www.odi.org/publications/6892-transparency-post-
2015-governance
127
See http://www.civicus.org/eei/ for more details of the index and its results.
128
L. Wild and M. Foresti (2013) Working with the Politics: How to improve public services for the poor, ODI,
http://www.odi.org/publications/7864- politics-service-delivery-institutions-aid-agencies; Khagram, Fung and de
Renzio, op. cit.
129
From IBP: http://internationalbudget.org/wp-content/uploads/LP-case-study-NCDHR-summary.pdf
130
From Oxfam: http://www.aceplive.com/wp-content/uploads/2014/08/ACEP-Report-PRMA-Final.pdf
131
From Oxfam: http://policy-practice.oxfam.org.uk/publications/a-healthy-influence-how-oxfam-convened-partners-to-
influence-zambias- elections-302290
132
From IBP: http://internationalbudget.org/wp-content/uploads/Civil-Society-Coalition-for-Quality-Basic-Education-
Carries-Out-Public-Expenditure-Tracking-Surveys-in-Malawi.pdf; and Global Campaign for Education:
http://www.campaignforeducation.org/docs/csef/CSEF%20CASE%20STUDY%20LEAFLET_AFRICA_JUNE2014_FINAL.pdf
133
S. Fukuda-Parr, P. Guyer and T. Lawson-Remer (2011) Does Budget Transparency Lead to Stronger Human
Development Outcomes and Commitments to Economic and Social Rights?, IBP Working Paper, No. 4; D. Kauffmann and
A. Bellver (2005) Transparenting Transparency, World Bank Policy Research Paper; DFI/Oxfam (2013) The 2013
Government Spending Watch Report: Putting Progress at Risk MDG Spending in Developing Countries; and ONE (2013)
The 2013 DATA Report: Financing the Fight for Africas Transformation.
For more information, or to comment on this report, email Jo Walker at [email protected]
Development Finance International and Oxfam International May 2015
This publication is copyright but the text may be used free of charge for the purposes of advocacy,
campaigning, education, and research, provided that the source is acknowledged in full. The copyright holder
requests that all such use be registered with them for impact assessment purposes. For copying in any other
circumstances, or for re-use in other publications, or for translation or adaptation, permission must be secured
and a fee may be charged. E-mail [email protected].
The information in this publication is correct at the time of going to press.
DEVELOPMENT FINANCE INTERNATIONAL
Development Finance International is a non-profit capacity-building, advocacy, advisory and research group
which works with more than 50 governments and international organizations worldwide to increase the
results of development financing. For more information, please visit
www.development-finance.org
GOVERNMENT SPENDING WATCH
Government Spending Watch is an initiative to provide the latest available information and analysis on
government spending which fights against poverty and targets the Millennium Development Goals, and to
increase advocacy and campaigning worldwide on these issues. The GSW site is jointly sponsored by DFI and
Oxfam, aiming to build a wider coalition of sponsors over time. For more information, please visit
www.governmentspendingwatch.org
OXFAM
Oxfam is an international confederation of 17 organizations networked together in more than 90 countries, as part
of a global movement for change, to build a future free from the injustice of poverty.
www.oxfam.org

Financing The Sustainable Development Goals: Lessons From Government Spending On The MDGs

Uploaded by

Copyright:

Available Formats

Financing The Sustainable Development Goals: Lessons From Government Spending On The MDGs

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Financing The Sustainable Development Goals: Lessons From Government Spending On The MDGs

Uploaded by

Copyright:

Available Formats

RESEARCH REPORT

FINANCING THE SUSTAINABLE

MDG AND SDG SPENDING

FINANCING THE SDGs

SPENDING TO FIGHT INEQUALITY

ACCOUNTABILITY FOR RESULTS

For more information, please visit: http://www.governmentspendingwatch.org

GSW is jointly sponsored by Development Finance International (www.development-finance.org)

TRENDS IN OVERALL SPENDING AND FINANCING

Debt service crowds out MDG spending in one-third of countries

Defence spending is worrying in at least four countries

AGRICULTURE AND FOOD

Are countries meeting their agriculture spending targets?

Are countries meeting their education spending targets?

Implications for the post-2015 education agenda

Are countries meeting the environment spending targets?

What are the recent trends in environment spending?

Sustainability-compliant spending targets a priority for the SDGs

Are countries meeting the health spending targets?

Are countries meeting the social protection spending targets?

Are countries meeting the WASH spending targets?

WASH spending needs to double and be tracked for impact on access

There is no doubt that vastly more investment is required for

$24 billion could be

OVERALL SECTOR SPENDING PROGRESS

TABLE 2.1. PROGRESS ON MDG SECTOR SPENDING

2.7% of GDP 52% of countries

Sector Extra public spending Private spend

Costs of monitoring SDGs 7.5 10

OVERALL TOTAL 796++ 1,245.5++

FINANCING THE MDGs

What has been the progress on these commitments since 2000?

IMPLEMENTING THE SPENDING

Donor funding distorts the types of spending in three ways:

HOW SHOULD THE SDGS BE FINANCED?

Much more far-reaching measures are needed, including:

Sectoral issues and allocations

There are also crucial implications for

Figure 4.1: Growth in inequality by Gini index of household income inequality87

Source: UNDP calculations using data from Solt (2009)

REDUCING INEQUALITY WITHIN SECTORS

Box 2. El Salvador: inequality in health spending

REDUCING GEOGRAPHICAL/SPATIAL INEQUALITY

REDUCING INEQUALITY AMONG BENEFICIARIES

Box 5. How to make spending more equitable

DATA AVAILABILITY IS IMPROVING RAPIDLY

Overall data availability by country

WHICH DATA ARE AVAILABLE?

Data availability by source of finance and type of spending

Can spending be disaggregated to track equity?

ACCOUNTABILITY DELIVERS RESULTS

Box 6: Transparency can contribute to improved allocations and results

Budget transparency = allocation to disadvantaged groups

Quantitative evidence: accountability raises spending levels and

Box 7. New data sources allow improved analysis

IMPLICATIONS FOR MONITORING POST-2015

2. Improving and publishing data and documents on spending:

3. Improving and publishing data and documents on revenue:

4. Improving and publishing data on aid and other budget financing: