CE Research Report: Aid Quality Advocacy

By weeatquince @ 2022-03-16T12:46 (+39)

This report is the result of Charity Entrepreneurship’s research to identify the most promising charities to start in the area of global health and development. Aid quality advocacy is one of our top ideas for 2022. A new organization operating in this space would focus heavily on policy and advocacy work in the countries that currently provide the most resources. 

We are currently accepting applications to our 2022 Incubation Program, through which we hope to launch this charity along with others that focus on road traffic safety, postpartum family planning, tobacco taxation, and EA Meta. The 2022 Incubation Program will take place from late June to late August. The deadline to apply is March 31, 2022. We provide a 2-month, full-time, cost-covered training program (online), cash stipends to cover living expenses, considerable funding (of up to $175,000), and a wide range of ongoing support. Visit our website to learn more.

The CE team is happy to answer any questions you may have about the following research report, and we welcome critical feedback and conversation.

Health and Development Policy –
Aid Quality Advocacy

Review: Vicky Cox and Melanie Basnak
Initial research: Ruby Dickson
Date of publication: January 2022
Research period: 2021

Thanks to Ruby Dickson for her assistance with the early research for this report, and to Antonia Personette and Urszula Zarosa for their contributions. We are also grateful to the experts who took the time to offer their thoughts on this research.

For questions about the content of this research or about the research process, please contact Sam Hilton at sam@charityentrepreneurship.com.

Executive summary

Government aid (or development assistance) is the flow of resources from the government of a wealthier country to a developing country with an objective of improving human welfare or promoting economic development. Each year, about US$170 billion is given in aid. However, there is broad consensus from experts and think tanks in this space that aid effectiveness could be improved. The intervention idea is policy advocacy work to ensure this flow of resources is directed as effectively as possible.

Exactly what changes should be advocated for depends on the country, but we identify a few key focus areas for a new organization, including: advocating for more aid to be given, improvements to development agency decision-making, and better targeting of giving to poorer countries, top multilateral agencies, or effective best-buy programs. The potential scale of such changes is huge. Successful campaigns may be able to influence government funds from £10 million–£bns and have the potential to improve how resources are spent by up to an order of magnitude.

Based on an analysis of six case studies, expert views, one review of this topic by a charity evaluator, and one academic paper, we think that a well-run advocacy campaign in this space has a reasonable chance of being successful, maybe 20-30%.

Our highly tentative cost-effectiveness analysis suggests that a campaign spending $1.5m could expect to have an influence equivalent to directing $300m of additional funds to the world's poorest. 

However, this space is reasonably crowded with think tanks, such as the Center for Global Development, and campaign groups, such as the ONE campaign, working on this topic across the US and Europe. A new charity would need to think carefully about its approach and strategy to determine how it could differentiate itself from the existing actors in the space in order to find an unoccupied niche. Locations to focus on might be the US, India, South Korea, Japan, the UAE, the EU, the UK, Germany, France, and other smaller EU countries.

Our main concern with this intervention derives from our analysis of high-level trends. The data indicates that, even though individual campaigns can be successful, global aid effectiveness does not appear to show much improvement over time. The reasons for this are unclear, but it suggests there may be some strong countervailing effects. For example, one donor giving effectively might lead to other donors giving less effectively or a political backlash that lowers aid or aid effectiveness.

Overall, our view is that work to improve aid effectiveness is an idea worth recommending to future charity founders.

1     Introduction

This report has been produced by Charity Entrepreneurship (CE). CE’s mission is to cause more effective charities to exist in the world by connecting talented individuals with high-impact intervention opportunities. We achieve this goal through an extensive research process and our Incubation Program. In 2021, our research process focused on the top policy and advocacy interventions within global health and development. 

Advocacy to improve aid quality and quantity was chosen by CE research staff as a potentially promising intervention within this category. This decision was the result of a 9-month process designed to identify interventions that were most likely to be high-impact avenues for future charity entrepreneurs. This process began by listing nearly 250 ideas and gradually narrowing down, examining them in more and more depth.

In order to assess how promising interventions would be for future charity entrepreneurs, we use a variety of different decision tools such as group consensus decision-making, weighted factor models, cost-effectiveness analyses, quality of evidence assessments, case study analysis, and expert interviews.

This process was exploratory and rigorous, but not comprehensive – we did not research all 250 ideas in depth. As such, our decision not to take forward a charity idea to the point of writing a full report should not be seen as a view that the idea is not good.

2     Background

2.1   What is development assistance/aid?

Government aid or development assistance is the flow of resources from the government of a wealthier country to a developing country with an objective of promoting economic development or improving human welfare. 

Aid can be bilateral, i.e., directly given to developing countries, or multilateral, i.e., given to intergovernmental institutions for them to use to support or distribute to developing countries (Wikipedia contributors, 2021a).

ODA (Official Development Assistance) is a category used by the OECD (Organization for Economic Co-operation and Development) to clarify what they consider development assistance and to measure the amount of development assistance given globally. It typically takes the form of grants but does include some low-interest loans.

In 2019, the OECD estimated that total ODA was $168bn. The majority of this sum, $152bn, was given by the 30 members of the OECD’s Development Assistance Committee (DAC). The biggest donor countries are listed in the table below (Wikipedia contributors, 2021d).

Non-DAC countries that give large amounts include China, India, and the UAE.

The map below shows the countries that have been the biggest recipients of development assistance in 2019.

Net official development assistance and official aid received, 2019 (Net official development assistance and official aid received, n.d.)

Various evidence suggests that there is room to improve aid effectiveness. Measures of aid effectiveness (such as QuODA ^[1](QuODA 2021, 2021)) show that some countries’ programs seem to be run noticeably better than others, which suggests there is scope for improvements. 

Interviews with experts suggest there is significant room for improvement on the quality of state-run aid programs, although there are limits to what can be changed, as aid agencies tend to have various political objectives. 

There is also some evidence of sector-wide failings, such as poor coordination, or some countries receiving significantly less support than would be expected (Mitchell and Hughes, 2020).

2.2   Scope of this report

The focus of this report is on the flow of resources by governments as opposed to aid given by foundations or individuals.

The focus is also on financial development assistance policy. Numerous policies that countries can take impact the world’s poorest and are not development assistance, such as changes to trade, finance, climate change, military commitments, and so on. We do not consider these in this report.

The report takes it as a given that effective development assistance is a net positive force in the world and does not look into this higher-level question.  

2.3   Terminology

For this report, we use “development assistance,” “foreign aid,” and similar terms interchangeably. Exactly how funds are categorized into these different buckets will vary from country to country and, as such, pinning this down is not crucial for this report.

For this report, we use “advocating” to refer to a set of actions where the goal is to influence government policy. We use “campaign” to refer to a specific set of actions with a defined goal, for example, an advocacy group could launch a campaign to increase the overall development budget in a country. These terms as used include both behind-the-scenes lobbying of decision-makers and generating influence through public mass mobilization. 

3     Theory of change

The theory of change for a charity advocating for the government of a donor country to improve the effectiveness or scale of its development assistance programs is illustrated here:

The key assumptions, corresponding to each step (i.e. “→”) in the theory of change, are ^[2]:

There are a number of different policy asks a charity could make, set out in Evidence that development assistance could improve.

4     Geographic assessment

4.1   Where existing organizations work

This space is reasonably crowded with think tanks and campaign groups working on this topic.

A number of think tanks provide advice on development effectiveness. For example, in the US/UK context, the Center for Global Development (CGD) is particularly influential. CGD carries out high-quality analysis into how development assistance can be more effective (Center for Global Development, n.d.). It mostly provides a research and advisory function and does not actively campaign for policy change beyond writing papers and engaging with policy makers. It has advised us that it sees value in more campaigning-focused work. Other think tanks and research groups in this space include Overseas Development Institute (ODI), IDInsight, J-PAL, Copenhagen Consensus Center (CCC), and Brookings (ODI: Think change, n.d., Improving lives with data and evidence, n.d., Copenhagen Consensus Center, n.d., Brookings - Quality. Independence. Impact, n.d.; The Abdul Latif Jameel Poverty Action Lab, n.d.).

There are also a number of campaign groups, the biggest of which is the ONE campaign. It is based in the US but also works across Canada, Europe, and Africa. Publicly, it focuses on big campaigns on topical issues, but also does behind-the-scenes work promoting effective development. Smaller groups around the world include RESULTS (UK and US), Bond (UK), Save the Children (Global), Concord (EU), and others. 

At the level of trans-national institutions, the OECD’s DAC (and the associated Development Co-operation Directorate) also encourages and supports aid effectiveness among its member states.

4.2   Geographic assessment

To assess locations where this intervention would be effective, we created a weighted factor model taking into account the total amount each country spends on development, how easy advocacy would be in each country (based on corruption scores and similar metrics and the size of the country), and the room for improvement in development effectiveness (based on metrics such as QuODA scores and Commitment to Development Index score). We ran this model a few times, taking into account different parameters in order to suggest countries that would be good focus areas for different campaigns. The results are set out in Evidence that development assistance could improve.

5     Expert views

In producing this report, we spoke to 12 relevant experts. Many of these actors (6 of 12) were people who had worked in government development agencies. Overall:

• Almost all experts believed that government development assistance could be improved.
• Views on exactly what improvements are the most crucial differed considerably. We think a large part of this variation can be explained as a result of different policy issues being more or less politically tractable in different countries.
• Experts generally believed that an external group could have some effect on development assistance, however, views varied considerably on how easy or difficult it would be and how long it might take to have an impact.
• Experts, especially in the UK, felt that this space was crowded and that it could be hard for a new organization to usefully contribute.

The table below summarizes each expert's key views.

Note: Organizations listed above are the most relevant current or former organization the interviewed person has worked at.

6     Evidence that development assistance could improve

We are confident that there are ways to improve both the quality and quantity of development assistance.

The list below sets out 10 ways that development agencies can improve development spending (at least insofar as the goal is the alleviation of poverty and suffering of the world’s poor). We outline the evidence that such changes are impactful and the degree to which there is room for improvement. The list is in a rough priority order in terms of the expected impact of a new charity focusing on these topics.

Criteria are categorized as: Very low, Low, Mixed / Medium, High, Very high

6.1.   Prioritizing poorer countries

Issue

Donations can be significantly more effective if given to the world's poorest. Poorer global citizens have fewer resources so can often be helped more at a lower cost.

Quality of evidence: High

That it is more cost-effective to provide support in less-developed countries appears to be a widely held view across the development sector with strong theoretical and empirical justifications. The theoretical case is based on the ideas that:

• Money goes further in less-developed countries. Nominal GDP per capita for rich countries is roughly 100x nominal GDP per capita for poorer countries (GDP per capita (current US$), n.d.), meaning you could roughly increase household income in a poor country by 100% for the amount it would cost to increase household income by 1% in a rich country.
• Need is greater in less-developed countries. Given the diminishing (potentially logarithmic) marginal returns of income for improving wellbeing (Pettinger, 2018), more impact can be had by supporting the poorest.

In-depth research into philanthropic effectiveness shows that this theory holds up and that there is a 100x multiplier between donations in poor and wealthy countries (GiveWell’s top charities are (increasingly) hard to beat, 2019; Your dollar goes further overseas, n.d.; Adleberg et al., 2021). Experts we spoke to and the international development literature also expressed this view. 

We did not do an additional search of the academic literature to seek further justification for this conclusion. We do, however, note that there are challenges in giving to many of the world's poorest countries, such as corruption and conflict. Our understanding is that this effect does not appear to negate all possible advantages of giving in less-developed countries, but can limit either the magnitude of the effect or the countries that resources can most usefully be shifted to. 

Room for improvement: High

There is room for improvement. The amount of ODA going to upper-middle-income countries ranges from 8% (UK) to 28% (Spain) according to Donor Tracker (Donor Tracker, n.d.). 

Looking at the QuODA for 2021, it seems that almost no countries’ aid scores are in the 75th+ percentile for directing aid to "under-aided" countries (QuODA 2021, 2021). 

There also appears to be some degree of consensus that there are "under-aided" countries/countries that miss out on foreign aid (Aid allocation within countries: does it go to areas left behind?, n.d.; Mitchell and Hughes, 2020), and that donors are bad at targeting the world's poorest (Briggs, 2017; Harcourt, 2019). 

The explanation for donor countries not giving where it is most needed is that they choose where to give for political reasons rather than for poverty prevention reasons.

Potential scale (quantified): High

Given the above, it seems plausible that ongoing effective advocacy on this topic could shift up to 10% (~$0.5-$1bn) of a donor country’s development assistance from upper-middle-income countries to lower-middle-income and lower-income countries. This estimate is based on closing the aforementioned gap of 8% to 28% of ODA going to upper-middle-income countries by half. 

Such a shift could increase the effectiveness of those resources by about 3x – this is based on GNI (Gross National Income) per capita data comparing upper-middle-income to lower-middle-income countries (Wikipedia contributors, 2021b), ignoring lower-income countries, which are likely to be more unstable.

Tractability: Medium

Changing the beneficiary might be more difficult than changing the kinds of interventions that are being funded, as country decisions seem likely to be more political.

Promising countries

Places where this could be a key focus (based on QuODA (QuODA 2021, 2021)) might be: Germany, United States, Japan, Slovenia, France.

6.2   Increasing the quantity of development assistance

Issue

Activists could advocate that governments spend more on development assistance or not make cuts to existing development assistance funding.

Quality of evidence: Medium–High

That there is value to development spending and that development spending can be done effectively is an underlying assumption of Charity Entrepreneurship’s work. An investigation of this belief is outside the scope of this report. That said, it should be noted that there is debate around these topics. Some key points to be aware of are:

• All of the evidence in the section above applies to making the case that there is value in shifting funds from a rich country to a less-developed country.
• The evidence that development assistance has led to economic growth is very weak.
• There is, however, some consensus and evidence that global health outcomes have significantly improved (Bharadwaj et al., 2021) and that progress is being made towards targets such as the UN’s Sustainable Development Goals (Edouard and Bernstein, 2016).
• There are concerns that if development assistance is done ineffectively it can cause more problems than it solves, so it would be beneficial to focus any such interventions on countries that spend aid effectively.

Room for improvement: High

The OECD has built an international agreement on the idea of donor countries spending 0.7% of GNI on development assistance. The OECD says “In 1970, The 0.7% ODA/GNI target was first agreed and has been repeatedly re-endorsed at the highest level at international aid and development conferences” (The 0.7% ODA/GNI target - a history - OECD, n.d.). Yet, despite this agreement, OECD DAC countries spent an amount close to 0.3% of combined GNI on development assistance in 2019 (OECD, 2020), and there is significant variation in how much different countries give (Net ODA provided, total (% of GNI), n.d.).

Potential for scale (quantified): Very high

Given the above, it seems plausible that ongoing effective advocacy on this topic could shift $1-5bn from a high-income country to lower-income countries – this is based on a high-income country giving an additional 0.2% of GNI per capita (i.e., closing half of the gap between current average DAC spending of 0.3% and the target of 0.7%). 

Such a shift could increase the effectiveness of those resources by about 10x. This is based on GNI per capita data comparing high-income to lower-middle-income countries (Wikipedia contributors, 2021b) and assuming the impact is reduced by 50% due to inefficiencies as a result of spending resources abroad rather than at home. 

Tractability: Low

Spending more on development assistance has a high cost. There is often competition for government funds, and spending resources abroad may not be seen as a vote-winning policy. It may be more tractable to work to prevent cuts to development assistance, although knowing when and in which countries cuts are on the table may be difficult.

Promising countries

Places where this could be a key focus might be: Finland, Denmark, Sweden, Canada, Ireland, Japan

6. 3   Improve development agencies' decisions

Issue

There are a broad set of ways to ensure that development agencies’ decisions are more in line with the goals of maximally improving global human wellbeing. These include improving:

1. oversight & transparency,
2. monitoring & evaluation of overseas programs,
3. the use of evidence in designing programs,
4. understanding of recipient's needs (e.g., mechanisms for country feedback),
5. staff skills.

These are all necessary, but not sufficient, steps towards ensuring governments make good decisions about development spending.

Quality of evidence: Medium

Our views on this topic have been largely informed by talking to experts. There seems to be a broad expert consensus that development agencies could make better decisions, although experts’ views on the value of trying to advocate for changes were mixed, ranging from highly skeptical to highly supportive. 

In addition to evidence from experts, there are numerous examples of ineffective aid (Failure in international aid, n.d.), guides to good practice in decision-making, and think tank support for better decision-making (such as the work of CGD and ODI) (UNDP, n.d., Aid effectiveness, n.d., ODI: Think change, n.d.). However, a brief literature review failed to find relevant academic literature on this topic. 

Overall, our informed view is that there is reasonable evidence to suggest that changes of these types are tractable, useful, and impactful, but that there is high variation; for example, from country to country.

Room for improvement: Medium-high

Experts from a number of countries pointed out cases of governments managing development budgets poorly.

The QuODA scale suggests room for improvement. For example, there is a clear split in monitoring and evaluation, with some countries performing this well and others poorly (QuODA 2021, 2021). There are also clear cases of sub-optimal development programs, such as the monetization of food aid from USAID (Wikipedia contributors, 2021c).

Potential scale (quantified): Medium-high

Given the above, it seems plausible that an effective campaign on this topic has the potential to affect up to 70% (~$4-$5bn) of a donor country’s development assistance – based on an estimate of how much of a country’s development assistance is controlled by a single organization. 

It is unclear how to quantify how changes to the decision-making process would affect development spending effectiveness. Quantifying it might need to be done on a case-by-case basis.

Tractability: Medium-high

Improving the effectiveness of development spending should be in line with both donor-country and recipient-country interest. However, any changes to development spending and processes will likely lead to changes in budget and upset some vested interests. 

Much of the insight into how to do this well comes from conversations with experts, and anyone working in this space would need to have a good network of advisers, experts, and insiders within government agencies to make progress. Some experts pointed to cases of progress and some pointed to where attempts to create improvements had not succeeded. 

Promising countries

Places where this could be a key focus are: Germany (who could better use evidence, based on expert views), France (who could have an oversight body, based on the views of CGD), India (who could do more monitoring and evaluation, based on expert views), and the United States (ensuring local country ownership of projects where appropriate).

6.4   Increase funding to top multilaterals

Issue

Aid can be bilateral, i.e., directly given to developing countries, or multilateral, i.e., given to intergovernmental institutions for them to use to support or distribute to developing countries. 

The best multilateral institutions are generally considered to be more effective than the best bilateral institutions. Some donor countries are choosing to give increasingly through multilateral institutions. Pushing for more funding to go through effective multilateral channels could generally increase aid effectiveness.

Quality of evidence: Mixed-high

There seems to be a substantial amount of evidence on this topic, although the results are mixed. A literature review of 45 papers on the relative effectiveness of multilateral vs. bilateral aid found that “9 studies find that bilateral aid is more effective … 13 find that multilateral aid is more effective … 13 find no significant difference in effectiveness … 10 studies report mixed findings” and concluded that there is “no consistent evidence that either bilateral or multilateral aid is more effective overall" (Biscaye, Reynolds and Anderson, 2017). 

Think tank evidence suggests that top multilaterals are more effective than bilateral donations, within certain limits. For example, the Overseas Development Institute (ODI) states that "multilateral channels appear less politicized, more demand-driven, more selective in terms of poverty criteria and a better conduit for global public goods” (Gulrajani, 2016). 

The results from QuODA suggest that all the most effective donor institutions are multilateral, but many multilateral institutions are worse than bilateral giving (QuODA 2021, 2021). Therefore, a push for more spending via multilaterals should only focus on the top few multilaterals. 

Expert views were mixed, with experts who felt that their own country's aid agency was ineffective being more optimistic than experts who felt their own country's spending was effective. One expert also suggested that governments should not scale up 100% of their spending of some type (e.g., health) to be given through multilaterals, as it is useful for governments to maintain some level of in-house expertise in order to hold multilaterals to account.

Room for improvement: Medium-high

The amount of ODA provided through multilateral institutions varies considerably, roughly between 20% to 70% within DAC countries, with an average of about 40% (QuODA 2021, 2021).

Potential scale (quantified): Low-medium

Given the above, it seems plausible that ongoing effective advocacy on this topic could shift up to 20% (~$1-$2bn) of a donor country’s development assistance to top multilaterals – this is based on a country shifting from providing 20% to providing 40% through multilaterals. 

The effect of this is unclear, and the intervention is likely only of high impact in countries that spend development assistance poorly and where the specific multilateral institutions that the donations are going to are highly effective.

Tractability: High

There have been a number of successful campaigns (see Evidence that advocacy can lead to improved aid) by advocacy groups for governments to increase spending through multilateral institutions, and this topic is currently on governments’ radars.

Promising countries

Places where this could be a key focus are unclear, but could be France, India, and the US.

6.5   Transfer funds between differently impactful programs

Issue

Advocates could encourage countries to shift resources from ineffective programs to more effective programs.

Quality of evidence: High

There is considerable academic evidence that some programs are more effective than others, and decent evidence in favor of specific and neglected programs. The Disease Control Priorities (DCP) Project collated academic data on this topic, involving hundreds of experts and referencing thousands of papers. The results suggested that, at least within the field of global health, some interventions may be hundreds of times more effective than the median intervention (not counting regression to the mean) (Ramanan Laxminarayan, Jeffrey Chow, and Sonbol A. Shahid-Salles, 2006; Cotton-Barratt, 2016). 

Evidence from philanthropic actors has shown an ability to put this into practice to identify effective programs and shift resources towards them (GiveWell’s top charities are (increasingly) hard to beat, 2019).

Room for improvement: Medium-high

Many of the most well-known effective large-scale interventions, such as childhood vaccines and TB treatments, are largely funded. However, as Charity Entrepreneurship’s research has shown, there remain many effective interventions that are not currently being delivered by development agencies.

Potential scale (quantified): Low-medium

Trying to improve development assistance effectiveness one program at a time may not be the most effective way of leveraging resources to impactful programs, yet some programs in development may be many times more effective than others.

Given the above, it seems plausible that an effective campaign on this topic could shift up to $10-100m – which is the amount a development agency could spend on a large program. 

Based on the evidence set out above about how programs vary in cost-effectiveness, it seems plausible that such a shift could increase the effectiveness of those resources by up to 10x or even 100x.

Tractability: Medium

The most effective large-scale cases of this happening to date tend to be advocacy towards funding top multilaterals (as above) or funding for very specific health projects. 

However, it can be very difficult to identify highly effective interventions, so this may be difficult for advocates to work on.

Promising countries

Places where this could be a key focus are the United States, Germany, Japan, the United Kingdom, and France.

6.6   Coordination between donors and recipients

A key aspect of ensuring that development assistance is effective is ensuring that there is input from the recipients of the assistance. This means the recipient countries and the individuals in need who may be helped by an intervention. This ensures projects are well designed to meet recipients’ needs and that there are good feedback loops to confirm projects are helping. 

This is currently standard practice in development, yet it may be lacking or done poorly. One expert we talked to suggested that working to guarantee this happens, and is done well, could be a highly impactful approach to take.

We have not looked at this topic in depth in order to limit the scope of the report.

6.7   Investing in the most useful global public goods

Countries may invest official development assistance into creating global public goods. In particular, it may be used for the research and development of public goods such as vaccines, improved crops, low emission technology, and so forth. 

Advocates could assess which public goods are in need of investment and work with governments to ensure that funds set aside for R&D are spent well. 

We have not looked at this topic in depth in order to limit the scope of the report.

6. 8   Improve coordination between aid agencies

Many different actors provide development assistance, and actors do not always coordinate well with one another. We believe that this consideration is important. However, it would likely be difficult for a new advocacy group to work on improving coordination as an initial campaign. 

We have not looked at this topic in depth in order to limit the scope of the report.

6.9   Reduce tied aid

Some development assistance is tied. This means it is offered “on the condition that it be used to procure goods or services from the provider of the aid” and this “can increase the costs of a development project by as much as 15 to 30 percent” (Untied aid - OECD, n.d.) 

We believe it is important for aid to be untied, but have not rated this intervention highly. Firstly, this is due to the significant progress that has already been made on this issue in recent years (QuODA 2021: Aid Effectiveness Isn’t Dead Yet, n.d.) Secondly, this is due to the potential benefits of untying aid being relatively low in comparison to the potential benefits of spending aid on more effective rather than less effective programs.

We have not looked at this topic in depth in order to limit the scope of the report.

6.10   Address other inefficiencies

There are a large number of other inefficiencies that can be identified in the development sector that do not fall under the categories listed above. Examples might be the use of food aid in the US (Wikipedia contributors, 2021c) or the culture amongst staff at development agencies (an issue raised by experts). 

We have not looked at this topic in depth in order to limit the scope of the report.

7     Evidence that policy change can be effectively implemented

Given the range of types of interventions in this space, it is difficult to give a definitive answer to the question of whether these are the types of interventions governments can effectively implement. Spending more or giving via multilaterals are the kinds of actions a government can take easily, but improving effectiveness is much harder. Some details on how easily governments can carry out such changes are in the “Tractability” subsections above.

There is some macro-level evidence suggesting that over the long run, there has been very little improvement in global aid effectiveness. This could suggest that governments struggle to make improvements to development spending. We discuss this further in the Macro-level considerations section below.

More could be done to investigate this area.

8     Evidence that advocacy can lead to improved aid

Advocacy is a high-risk strategy. Sometimes millions of dollars can be spent with no impact or, conversely, a small team can leverage vast amounts of government resources.

To assess whether advocacy for improved or greater aid and development is tractable, we have talked to experts, reviewed an independent charity evaluation, looked into the academic evidence, and examined case studies of particular campaigns and organizations.

Overall, our analysis of this evidence suggests that the chance of success of a well-run policy advocacy campaign is not high, but not negligible, perhaps about 20-30%. 

Moreover, a successful campaign could leverage significantly more government resources to be spent effectively than would be spent on the campaign. The evidence suggests that an advocacy organization (including the costs of unsuccessful campaigns) could see a ratio of charity cost to government money moved of around 1:50. 

There are likely some diminishing marginal returns to how much advocates spend on influencing development budgets in a single country, with cost-effectiveness dropping off when advocacy organizations spend over a few million dollars per year.

8.1   Expert views

Experts had a range of views on how possible this work would be, ranging from the very positive (e.g., 50% chance of success) to the very negative (e.g., <1% chance of success). It is unclear if this demonstrates a genuine lack of consensus or simply reflects different backgrounds, countries, and ideas about what to campaign for. The general view seems to be that change is possible but could be quite difficult.

8.2   Independent charity evaluations

There is one in-depth independent charity evaluation of an organization working on development policy advocacy: a 2012 review by Giving What We Can (GWWC) of RESULTS US.

GWWC’s report concludes, “we think there is a reasonable chance that RESULTS’ past cost effectiveness was high mostly because of their campaigns to increase funding for the treatment of TB. We estimate that their past cost effectiveness is under $20 per DALY averted, potentially significantly under. This would be better than our current top recommended charities, which achieve around $20-50 per DALY” (RESULTS U.S. Cost Effectiveness Assessment, 2012).

However, it did not recommend donating to RESULTS given uncertainties regarding RESULTS’ future cost-effectiveness, the counterfactual impact of a shift of government resources to TB, and the extent to which RESULTS played a role in these policy successes.  

Success rate: ~50% (based on success rate of campaigns run at this time).

Money moved: ~1:66 ratio of cost to government money moved, based on GWWC’s data.

• Note another source suggests that the ratio is closer to ~1:230 (Back of the Envelope Guide to Philanthropy, n.d.).

Bias: This data is likely affected by a selection bias effect, as GWWC would have chosen to evaluate the most impactful-looking organization in this space at that time. 

8.3   Academic research

There is not much academic research on advocacy’s effects on changes to development policy. 

Montes-Rojas (2013) found that “foreign lobbying in the US has a statistically significant effect for attracting US foreign bilateral aid.” By tracking developing countries' own lobbying for aid, they show that increasing lobbying expenditures by 1% increases aid receipts in the long run by 0.07% (Montes-Rojas, 2013). This means that, on average, this lobbying has roughly a 1:10 cost to money moved ratio. As such, we might expect that a campaign advocating for a shift in development resources toward under-aided or less developed countries might similarly have a 1:10 ratio of costs to government money moved.

Along similar lines, Anwar and Michaelowa (2006) and Lahiri Raimondos (2000) found evidence that interest groups have had an effect on US aid spending decisions (Lahiri and Raimondos-Møller, 2000; Anwar and Michaelowa, 2004).

Money moved: ~1:10 ratio of costs to government money moved. 

8.4   Case studies

Specific case studies

We analyzed three specific campaigns. This was a convenience sample based on campaigns where we had good contacts and data available.

1. USA REACH Act – failed (Reach Every Mother and Child Act - Save the Children Action Network, 2017)
   • A big campaign in the US to make aid more effective
   • Campaigners likely spent $1m+ (unclear)
2. Swiss ballot initiative – passed (EAF’s ballot initiative doubled Zurich’s development aid - EA Forum, 2020)
   • The initiative cost around $25,000 in financial costs and around $190,000 in opportunity costs. Depending on the assumptions, it raised a present value of $20–160 million in development funding
3. 2021 UK 0.7% campaign – failed
   • Expect it cost around £50k (unclear)
   • Aimed to prevent £4.5bn of aid cuts
   • Seemed to be going well, but then had no impact

Success rate: ~33%
Money moved: ~1:475 ratio of costs to government money moved for ballot initiative. 

• Perhaps closer to 1:50 across all three case studies

Organizational case studies

The ONE Campaign has run numerous lobbying and mass public engagement campaigns, including a few possible successes:

• Electrify Africa, 2016 – Act passed on a significant bipartisan basis, part of Obama’s Power Africa initiative (Electrify Africa, 2020). Nearly all references to the Act reference the ONE campaign, so it likely played a significant role
• Note that this is not direct aid, rather, it is providing loan guarantees
• Global Fund, 2019 – A global campaign for the global fund. Claims responsibility for Canada increasing contribution by CAD$930.4 million (The Global Fund to Fight AIDS, Tuberculosis and Malaria, 2020)
• BUILD Act, 2018 – Reorganized existing finance administration into DFC within USAID (BUILD Act, 2020)
  • Note that this is not direct aid
  • ONE’s role is unclear

Other campaigns listed on the website do not have clear policy victories; nothing more concrete than meetings with government officials. There are 7 such other campaigns, looking only at those focused on the developed world. These were: Poverty is Sexist Campaign; EU One Vote 2019; Trillion Dollar Scandal; Agit8; GAVI campaign; Girls count; and IMF Gold Campaign. ONE may be trying to achieve broader goals than policy change.

Success rate: ~30% (only considering policy change)

Money moved: ~1:3 ratio of cost to government money moved (based on a ~US$30m per year budget). This is not counting the effects of the 2 Acts passed.

Other organizations have had traction in this space, such as:

• The Development Policy and Finance team at the Gates Foundation secured US$4 billion in long-term funding commitments from donor governments for the International Finance Facility for Immunization (IFFIm) (Development policy and finance, n.d.)
• Concord Europe has, in 2019-20, increased the European Commission’s focus on sustainable development and civil society actors and had some impact on the European Commission’s Gender Action Plans (GAP) II & III. It spends about  US$1.9 million per year (Annual Report 2020: Co-creating the future with clear messages, 2021)
• RESULTS US and RESULTS UK: though it is not clear exactly what impact they have had, they seem to have helped direct $bns to the Global Fund and potentially stop the Trump administration from cutting the aid budget (Annual Successes 2018, 2018)

9     Cost-effectiveness analysis

We carried out a cost-effectiveness analysis modeling an advocacy campaign focused on persuading the German government to adopt cash benchmarking as a metric for effectiveness. This metric would then improve decision-making within the Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ). We chose this as an intervention to model based on a conversation with an expert.

Overall, the intervention appears to be highly cost effective, with a small advocacy group spending $1.5m over 8 years having an impact similar to $300m given in direct cash transfers to the world’s poorest.

10     Implementation

10.1   Talent

Any co-founder team working on this intervention would need to be able to do developed-world policy advocacy work. There are a variety of approaches to such work, including public campaigning, political networking, and producing reports. The ideal co-founder team would have the skills needed to take any of these approaches, although there is some flexibility within advocacy for the co-founders to adopt an approach that works best for their skill set. In all cases, the ability to understand complex political systems and potential levers for changing those systems will be key.

Depending on the country and method of advocacy work, it may be significantly advantageous for co-founders to be from the country they are doing advocacy in, and in some cases already be well-connected there. This applies in particular to India and the UAE.

Working out exactly how development assistance could be improved is complicated and requires policy research skills. This may prove a challenge for some co-founders. Those who feel less confident in addressing this challenge may want to focus on campaigning for increasing the quantity of development assistance rather than for improving the quality.

10.2   Access

Information

Developed countries are more transparent about their data and methods, so information should be reasonably accessible. 

Government

Charity founders may need good networks to have easy access to the government.

Funding

EA funding

We expect there to be EA funding available for work of this type from Open Philanthropy. There may be other sources of EA funding. 

Non-EA funding

Finding non-EA funding may be challenging.

10.3   Scale of the problem

Based on our geographic assessment and evidence review, there seems to be 10-15 top countries where this is likely to be worthwhile. There may also be an additional 10-15 countries where this could potentially be worthwhile.

10.4   Neglectedness 

We think there is space for new actors campaigning on development effectiveness, although an area for further research would be to engage more actively with existing campaign groups and map out the neglected topics.

This remains one of our key uncertainties.

10.5   Externalities

There are a number of risks with many of the possible actions that could be taken in this space. In particular:

• Public criticisms of the effectiveness of aid budgets could lead to criticism of aid and development. This could result in budgets being cut rather than spent more effectively. Campaigners should consider any negative messaging very carefully.
• Shifting a limited pool of resources from ineffective spending to effective spending still requires moving resources away from some group beneficiaries. This could cause harm to that group.

10.6   Macro-level considerations

Although there is good evidence that individual development agencies could improve the quality of their spending, there is also a chance that the effect of these improvements is significantly reduced when taking into account the entire development sector. 

One of the experts we spoke to flagged that even though individual campaigns can be successful, global aid effectiveness does not appear to show much improvement over time. The expert pointed to the lack of success in hitting targets set at OECD meetings (e.g., in Rome, Paris, Accra, and Busan). Furthermore, results from CGD’s QuODA suggest that, although some aspects have improved, on many metrics of effectiveness, governments are not getting better at spending development assistance effectively over time (QuODA 2021: Aid Effectiveness Isn’t Dead Yet, n.d.).

Possible explanations for this lack of improvement are that:

• Expectations have risen. QuODA is based in part on surveys in recipient countries. It is possible that respondents' expectations of what effective aid looks like have risen.
• There is a regression to the mean effect across the whole sector. For example, one development agency focusing on effective interventions could lead to other actors giving less support to such interventions. This in turn could lead to a regression to the mean effect where the overall effectiveness of such changes is small in the grand scheme of things.
• Improvements are reversed. Some policy changes involve stakeholders who benefit from the policy and therefore have a vested interest and who also have some voting power or other political influence. These policies are very hard for politicians to reverse. Development policy affects the world's poorest, who have minimal political influence and no voting power in donor countries. As such, development policies may be easy to reverse. For example, the UK recently (albeit temporarily) reversed its policy to give 0.7% of GNI to ODA.

Overall, we think that some amount, but not all, of the effect of an organization working in this space will be offset due to wider macro-level trends and a regression to the mean effect.

It may also be the case that development assistance, by its nature, does not actually improve the lives of those whom it aims to help. For example, it may breed dependency, facilitate corruption and rent-seeking, and compress growth. If true, this consideration would negatively impact the expected value of both this intervention and a significant amount of the development interventions that Charity Entrepreneurship investigates. Given the role of this report is to be comparable to other reports by Charity Entrepreneurship, we have chosen not to investigate this broader topic in this report.

10.7 Remaining key uncertainties

A number of key uncertainties still remain that could be useful areas for further research. In particular:

• What are the counterfactual impacts of shifting very large amounts of funds?
• What are the marginal benefits of additional funds to large pots of resources and is there a risk that giving to a cause prompts other actors to donate less (is there a regression to the mean type effect)?
• How well can governments make these types of changes?
• How neglected is this issue in various countries?

11     Conclusion

Overall, we recommend this intervention to future charity entrepreneurs. We think there is room for improvement within existing aid policy and that advocacy in this domain can succeed. Exactly what policy to advocate for depends on the country of focus. Potential founders would benefit from a good understanding of politics and policy making.

Thank you for taking the time to read this report. 

Visit https://www.charityentrepreneurship.com/ to learn how you can launch this intervention through our Incubation Program.

References

Adleberg, T. et al. (2021) ‘Charity begins at home; shouldn’t we solve our own problems before helping others?’ Giving What We Can, 2 May. Available at: https://www.givingwhatwecan.org/post/2021/05/charity-begins-at-home-shouldnt-we-solve-our-own-problems-before-helping/ (Accessed: 22 December 2021).

Aid allocation within countries: does it go to areas left behind? (no date) ODI: Think change. Available at: https://odi.org/en/publications/aid-allocation-within-countries-does-it-go-to-areas-left-behind/ (Accessed: 22 December 2021).

Aid effectiveness (no date) Center For Global Development. Available at: https://www.cgdev.org/topics/aid_effectiveness (Accessed: 22 December 2021).

Annual Report 2020: Co-creating the future with clear messages (2021). Available at: https://concordeurope.org/resource/annual-report-2020-co-creating-the-future-with-clear-messages/ (Accessed: 22 December 2021).

Annual Successes 2018 (2018). RESULTS. Available at: https://results.org/wp-content/uploads/Annual-Successes-2018-1.pdf.

Anwar, M. and Michaelowa, K. (2004) ‘The political economy of US aid to Pakistan’, Review of Development Economics, 10(2), pp. 195–209. Available at: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=625944

Back of the Envelope Guide to Philanthropy (no date). Available at: https://www.beguide.org/index.html (Accessed: 22 December 2021).

Bharadwaj, S. et al. (2021) ‘Can foreign aid and international charity make a difference?’ Giving What We Can, 12 April. Available at: https://www.givingwhatwecan.org/post/2021/04/can-foreign-aid-and-international-charity-make-a-difference/ (Accessed: 22 December 2021).

Biscaye, P. E., Reynolds, T. W. and Anderson, C. L. (2017) ‘Relative effectiveness of bilateral and multilateral aid on development outcomes’, Review of Development Economics, 21(4), pp. 1425–1447. Available at: https://onlinelibrary.wiley.com/doi/full/10.1111/rode.12303

Briggs, R. C. (2017) ‘Does Foreign Aid Target the Poorest?’. Available at: https://www.cambridge.org/core/journals/international-organization/article/does-foreign-aid-target-the-poorest/24010BE037DC8F4C6AF04DED5D94E4C9

Brookings - Quality. Independence. Impact (no date). Available at: https://www.brookings.edu/ (Accessed: 22 December 2021).

BUILD Act (2020) ONE. Available at: https://www.one.org/international/take-action/victories/build-act/ (Accessed: 22 December 2021).

Center for Global Development (no date) Center For Global Development. Available at: https://www.cgdev.org/ (Accessed: 22 December 2021).

Copenhagen consensus center (no date). Available at: https://www.copenhagenconsensus.com/ (Accessed: 22 December 2021).

Cotton-Barratt O (2016) Prospecting for Gold. Available at: https://www.effectivealtruism.org/articles/prospecting-for-gold-owen-cotton-barratt/.

Development policy and finance (no date) Bill & Melinda Gates Foundation. Available at: https://www.gatesfoundation.org/our-work/programs/global-policy-and-advocacy/development-policy-and-finance (Accessed: 22 December 2021).

Donor Tracker (no date) Donor Tracker. Available at: https://donortracker.org/ (Accessed: 22 December 2021).

EAF’s ballot initiative doubled Zurich’s development aid - EA Forum (2020). Available at: https://forum.effectivealtruism.org/posts/dTdSnbBB2g65b2Fb9/eaf-s-ballot-initiative-doubled-zurich-s-development-aid (Accessed: 22 December 2021).

Edouard, L. and Bernstein, S. (2016) ‘Challenges for Measuring Progress towards the Sustainable Development Goals’. Available at: http://www.bioline.org.br/pdf?rh16037

Electrify Africa (2020) ONE. Available at: https://www.one.org/international/take-action/victories/electrify-africa/ (Accessed: 22 December 2021).

Failure in international aid (no date) GiveWell. Available at: https://www.givewell.org/international/technical/criteria/impact/failure-stories (Accessed: 22 December 2021).

GDP per capita (current US$) (no date). Available at: https://data.worldbank.org/indicator/NY.GDP.PCAP.CD (Accessed: 22 December 2021).

GiveWell’s top charities are (increasingly) hard to beat (2019) Open Philanthropy. Available at: https://www.openphilanthropy.org/blog/givewells-top-charities-are-increasingly-hard-beat (Accessed: 22 December 2021).

Gulrajani, N. (2016) Bilateral versus multilateral aid channels: Strategic choices for donors. Overseas Development Institute. Available at: https://cdn.odi.org/media/documents/10492.pdf.

Harcourt, S. (2019) Opinion: Aid isn’t going to the world's poorest people. Why?, Devex. Available at: https://www.devex.com/news/opinion-aid-isn-t-going-to-the-world-s-poorest-people-why-94747 (Accessed: 22 December 2021).

Improving lives with data and evidence (no date). Available at: https://www.idinsight.org/ (Accessed: 22 December 2021).

Lahiri, S. and Raimondos-Møller, P. (2000) ‘Lobbying by ethnic groups and aid allocation’, Economic journal , 110(462), pp. C62–C79. Available at: https://www.jstor.org/stable/2565723

Mitchell, I. and Hughes, S. (2020) Which Countries Miss Out in Global Aid Allocation? Center for Global Development. Available at: https://www.cgdev.org/sites/default/files/Mitchell-Hughes-Aid-Allocations.pdf.

Montes-Rojas, G. V. (2013) ‘Can Poor Countries Lobby for More US Bilateral Aid?’, World development, 44, pp. 77–87. Available at: https://ideas.repec.org/a/eee/wdevel/v44y2013icp77-87.html

Net ODA provided, total (% of GNI) (no date). Available at: https://data.worldbank.org/indicator/DC.ODA.TOTL.GN.ZS?end=2016&view=map (Accessed: 22 December 2021).

Net official development assistance and official aid received (no date) Our World in Data. Available at: https://ourworldindata.org/grapher/net-official-development-assistance-and-aid-received (Accessed: 22 December 2021).

ODI: Think change (no date) ODI: Think change. Available at: https://odi.org/ (Accessed: 22 December 2021).

OECD (2020) Aid by DAC members increases in 2019 with more aid to the poorest countries. OECD. Available at: https://www.oecd.org/dac/financing-sustainable-development/development-finance-data/ODA-2019-detailed-summary.pdf.

Pettinger, T. (2018) Diminishing marginal utility of income and wealth, Economics Help. Available at: https://www.economicshelp.org/blog/12309/concepts/diminishing-marginal-utility-of-income-and-wealth/ (Accessed: 22 December 2021).

QuODA 2021 (2021) Center For Global Development. Available at: https://cgdev.org/quoda-2021 (Accessed: 22 December 2021).

QuODA 2021: Aid Effectiveness Isn’t Dead Yet (no date). Available at: https://www.cgdev.org/blog/quoda-2021-aid-effectiveness-isnt-dead-yet (Accessed: 22 December 2021).

QWIDS - query wizard for international development statistics (no date). Available at: https://stats.oecd.org/qwids/ (Accessed: 22 December 2021).

Ramanan Laxminarayan, Jeffrey Chow, and Sonbol A. Shahid-Salles (2006). Disease Control Priorities in Developing Countries. Available at: http://dcp-3.org/sites/default/files/dcp2/DCP02.pdf

Reach Every Mother and Child Act - Save the Children Action Network (2017). Available at: https://www.savethechildrenactionnetwork.org/articles/reach-factsheet/ (Accessed: 22 December 2021).

Research on programs (no date) GiveWell. Available at: https://www.givewell.org/research/research-on-programs (Accessed: 22 December 2021).

RESULTS U.S. Cost Effectiveness Assessment (2012). Giving What We Can. Available at: https://assets.ctfassets.net/dhpcfh1bs3p6/5AsulAmVXiGo26CwskmK8Y/a824339cf6496b7a545ce5d9d3572016/Results_Report.pdf.

UNDP (no date) Available at: https://www.undp.org/search?q=good+practice&form_build_id=form-tYsExLQhQm3GsYFC48Ijf3proidNxMsU9qGO2PBgyk0&form_id=undp_solr_search&op=Submit (Accessed: 22 December 2021).

The 0.7% ODA/GNI target - a history - OECD (no date). Available at: https://www.oecd.org/development/financing-sustainable-development/development-finance-standards/the07odagnitarget-ahistory.htm (Accessed: 22 December 2021).

The Abdul Latif Jameel Poverty Action Lab (no date). Available at: https://www.povertyactionlab.org/ 

The Decision Lab (2021) The COM-B Model for Behavior Change. Available at: https://thedecisionlab.com/reference-guide/organizational-behavior/the-com-b-model-for-behavior-change/ (accessed 6 January 2022)

The Global Fund to Fight AIDS, Tuberculosis and Malaria (2020). Available at: https://www.one.org/international/take-action/victories/the-global-fund-2019/ (Accessed: 22 December 2021).

Untied aid - OECD (no date). Available at: https://www.oecd.org/dac/financing-sustainable-development/development-finance-standards/untied-aid.htm (Accessed: 22 December 2021).

Wikipedia contributors (2021a) Development aid, Wikipedia, The Free Encyclopedia. Available at: https://en.wikipedia.org/w/index.php?title=Development_aid&oldid=1060900844.

Wikipedia contributors (2021b) List of countries by GNI (nominal) per capita, Wikipedia, The Free Encyclopedia. Available at: https://en.wikipedia.org/w/index.php?title=List_of_countries_by_GNI_(nominal)_per_capita&oldid=1061532059.

Wikipedia contributors (2021c) Monetization of U.S. in-kind food aid, Wikipedia, The Free Encyclopedia. Available at: https://en.wikipedia.org/w/index.php?title=Monetization_of_U.S._in-kind_food_aid&oldid=1040661748.

Wikipedia contributors (2021d) Official development assistance, Wikipedia, The Free Encyclopedia. Available at: https://en.wikipedia.org/w/index.php?title=Official_development_assistance&oldid=1036608578.

Your dollar goes further overseas (no date) GiveWell. Available at: https://www.givewell.org/giving101/Your-dollar-goes-further-overseas (Accessed: 22 December 2021).

1. ^{^}

   QuODA is the Quality of Official Development Assistance tool produced by the Center for Global Development. It compares the quality of aid given by the largest aid agencies across a range of relevant factors such as prioritization, transparency, and evaluation.

2. ^{^}

   COM refers to the capability, motivation, and opportunity for change from the COM-B model for behavior change (The Decision Lab, 2021).