Findings & Recommendations

The report concludes that aid is not transparent enough; despite the fact that information is often produced, much of it is not published systematically or not easily accessible. There is potential for higher levels of aid transparency to be achieved across the board, with a number of organisations proving aid transparency is possible and can be done rapidly when there is sufficient political will. Those donors that are leading on aid transparency need to set an example and encourage others to follow their lead.

Conclusion 1: Most aid information is not published

The index indicates that the vast majority of aid information is not currently published, with only a handful of organisations publishing more than 50% of the information types surveyed. The average overall score across all organisations is only 34%.

It is striking that some donors who are traditionally perceived as leading on issues of aid effectiveness or transparency score particularly disappointingly. These include Australia (26%), Canada (31%), Finland (38%), Ireland (29%), New Zealand (30%), Norway (32%) and U.S. PEPFAR (34%). This is because they performed poorly at either the country or activity level, or both.

The average score of the larger organisations included in the ranking is only 37%[1]. The performance of some of these organisations is particularly problematic given the amount of aid they give and therefore the relative impact of their lack of transparency. This notably includes the U.S. (with U.S. Department of the Treasury (scoring 10%) and U.S. Department of Defense (14%)), Germany GIZ (25%), France (31%) and Japan (36%). An astonishing 23 organisations do not systematically publish any country information – country strategies, forward budgets, evaluations and results – which is over a third of the 58 organisations surveyed and includes major donors such as Canada, Germany, Norway, USAID and U.S. Treasury.

Perhaps unsurprisingly, agencies performed best on information about their own organisations, although none of the 58 organisations included in the ranking publish all the organisation level information items covered in the survey. The AfDB, DG Enlargement and ECHO came closest, although even they do not publish annual audits.

The lack of comprehensive, comparable and timely information makes demonstrating the representativeness of the data sampled difficult in a number of cases. Perhaps the most fundamental illustration of the lack of current levels of aid transparency was for a handful of organisations for whom it was a struggle to determine which country was the current largest recipient country for each agency. This was the case for the UNDP, Canada and Germany, for which two-year-old DAC CRS reporting had to be used to select the largest recipient country but which the organisation later informed us was no longer the largest recipient[2]. Similarly, for France’s AFD, Côte d’Ivoire was originally included in the survey for the activity level indicators because last year it was the largest recipient country of aid. There was almost no information available on the AFD website about Côte d’Ivoire, except for a project commemorating 20 years of French research into chimpanzees. After consultation, the AFD informed Publish What You Fund that AFD’s (rather than France as a whole) largest recipient country last year was actually South Africa. Consequently the activity level indicators in the survey were re-answered based on a project in South Africa, which as it happens resulted in a significantly improved score.

There are a number of organisations where lack of disclosure made it challenging to be sure whether the sampled information is representative. Some organisations may have over-performed in this index because the transparency of information on their largest activities in their largest recipient countries may not be representative of their activities as a whole. This is particularly true of U.S. agencies where Afghanistan was selected as the largest recipient country, because there are dedicated websites on Afghanistan – USAID has an Afghanistan project website and the Department of Defense has information on the SIGAR (Special Inspector General for Afghanistan Reconstruction) website. Another example is Spain, which has a country level project database for Peru but it is not clear if similar databases exist for other countries that receive Spanish aid.

Generally, the level of information availability is disappointingly low. Six years after the Paris Declaration, there is a distinct lack of political and technical leadership, particularly given the series of “beginning now” commitments made on aid transparency in 2008 in the Accra Agenda for Action. With the exception of overall activity or project costs, very few donors are systematically disclosing detailed activity level information.

A good gauge or proxy for whether donors are delivering on the AAA commitment on publishing “actual disbursements”[3] would be the transactions indictor. However, only six organisations are currently publishing transaction-level information: AfDB, Global Fund, Hewlett Foundation, Netherlands, UK DFID and World Bank IDA. Similarly, only seven organisations delivered[4] on the specific AAA commitment on country forward budgets[5], and Conditions and Memoranda of Understanding (MoU) have only been disclosed by six and five organisations respectively[6], with only World Bank IDA and U.S. MCC doing both[7].

Conclusion 2: Information is produced but not always published and is far too hard to access and use

The survey results point to the fact that a far greater volume of information is being produced than is being published by organisations (see chart 4 in Annex 2). For 11 information items, less than 10 organisations systematically publish them. These 11 information items include a mixture of basic information (actual dates, contracts) as well as areas where there are extensive commitments (forward budgets, conditions, transactions) and current donor priorities around results and value for money (organisation level audits, activity impact appraisals and design documents). Most of these information items relate to monitoring results and impact and it is likely that organisations collect this information internally. There is no defensible reason for why there is not a presumption towards the publication of this information – particularly given donor commitments relating to monitoring for results, mutual accountability and conditionality.

In terms of accessibility, what is produced is not made available systematically and is hard to find. Even if you speak the organisation’s language, are computer literate with a good Internet connection and are extremely familiar with the organisation’s policies and operations, it can still take considerable time and effort to locate basic information on, for example, organisation procurement procedures, budgets, contracts and evaluations. Websites are often difficult to navigate and information is sometimes provided on more than one website but in varying levels of detail depending on the language and format. The different systems and formats used for publishing the information means that comparing information even within the same donor organisation is a challenge.

For example, DG Enlargement has several websites with varying information across them. It was difficult to know whether the information provided was complete and it was not clear why the information was on one website rather than another; there were even different project IDs and titles across the several websites. To gain a complete picture of all of DG Enlargement’s activities in Turkey, it was necessary to consult four different websites. Another example worth highlighting is Switzerland, which has an online country level project database for Nepal, but (like Switzerland’s online central project database) it does not appear to be comprehensive. It is not clear why SDC has two partial databases instead of taking a more comprehensive and systematic approach.

The pervasive use of PDF formats and some unsearchable databases (e.g. France and Austria) is a major obstacle to accessibility, locking away information which then can only be made accessible by time consuming manual copying and pasting, or complicated and unreliable website ‘scraping’ techniques.

Other issues that undermined the accessibility of information included basic technical failures that should be picked up during standard testing phases. These included links breaking after websites are updated (e.g. GAVI and Norad) unstable URLs which means it is impossible to link to information (e.g. Ireland) or masked generic URLs (e.g. Korea KOICA) and USAID Afghanistan’s extremely slow website which would be impossible to use without a fast computer and a broadband connection.

The value of the investment of some of the best performers is reduced by the lack of information provision by large organisations that are performing below average, for example France, USAID and Germany’s GIZ. This lack of coherence is also evident within donor governments and between their agencies: within the U.S. MCC ranked 7th and DOD and Treasury ranked 46th and 49th respectively; for Germany, KfW ranked 21st but GIZ came 39th; and in Korea, KOICA came 22nd yet EDCF came 43rd. If one agency within a donor country can publish information the failure of other agencies to do so undermines the value of that transparency.

Conclusion 3: Achieving aid transparency is possible

The level of transparency of a number of organisations demonstrates that aid transparency is possible across all three levels assessed; these leaders include the World Bank, the Global Fund, the AfDB, the Netherlands, DFID, Sweden and the MCC.

There is good and bad performance across the spectrum, and a range of different organisations and agencies perform well. While some patterns emerge, it is clear that an organisation’s size, how established they are, or whether they are a multi- or bilateral organisation does not predict or determine the level of their transparency. The top 20 organisations are almost evenly split between multilaterals and bilaterals, and while most are more established donors, this does not mean that established donors are all doing well. As a group their average score is 39%; but there are some poor performers in this group, in particular Spain, Portugal, two U.S. agencies (DOD and Treasury) and Italy. This demonstrates that it is possible for a different range of organisations to achieve greater aid transparency.

The size or amount of aid an organisation provides does not necessarily mean it will be more or less transparent than another organisation that provides more or less aid. Although many of the newest members of the DAC and emerging donors generally perform poorly, achieving an average overall score of just 18%, there are some notable exceptions. Estonia performs particularly well compared with other much larger organisations that have been providing development assistance for considerably longer periods. Other notable exceptions are the Czech Republic, which ranks higher than its neighbour Austria (23rd and 25th respectively); as does Korea’s KOICA compared to Japan (22nd and 23rd).

Some organisations have shown that it is possible to improve their levels of aid transparency extremely quickly. For example, since the beginning of 2011, the Netherlands, Sweden, the UK and the World Bank have published considerably more information about their aid activities. In particular, the Netherlands demonstrated how rapidly progress is possible (perhaps providing an example for donors using the extended CRS format integrated within their systems) as it was lifted up the ranking from joint 30th to 4th place through a major release of new data to the IATI Registry and the publication of their biennial “results in development cooperation”[8]. Most recently, United Nations Office for Project Services (UNOPS) has demonstrated that, with the required political will and good underlying data management systems, it is possible to publish information to the IATI Registry very quickly[9].

Recommendations for organisations on improving aid transparency

In addition to the cross-cutting recommendations below, on the country pages, specific suggestions are made for each of the 58 organisations included in this pilot index.

Recommendation 1: Increase political will and action – using the Aid Effectiveness Agenda as a springboard

There is potential for higher levels of aid transparency to be achieved across the board. A handful of leading organisations have shown it can be done, and they include large, small, new, established, multilateral and bilateral donors. However, the common characteristic they share is political will and technical leadership.

Those committed agencies, politicians and civil servants need to continue to set an example by driving forward, addressing existing information gaps, but also encouraging others to follow their lead. The November 2011 HLF-4 meeting in Busan is clearly a springboard for making concrete time-bound commitments. Post-Busan monitoring of aid effectiveness needs to include a specific aid transparency indicator. The future aid effectiveness monitoring framework should also use the major increase in up-to-date information becoming available through IATI to streamline monitoring and reduce the burden of recurrent surveying of information already in donors systems.

A scale-up of political and technical engagement by organisations that are dragging their feet has the potential to deliver significant improvements in a short space of time. Donor governments should seek out fora in which to drive forward commitment, learning and collective action on aid transparency and the common standard, whether in the Open Government Partnership, through the Commonwealth, within the EU, UN Development Co-operation Forum or through the growing focus on open data. Partner countries and CSOs need to continue working together to intensify the pressure on donors to ensure IATI implementation delivers for the recipients of aid.

Recommendation 2: Organisations should publish what they have, build systems to collect what they don’t and make sure it is all accessible

Since much more information is produced than is currently made available, and even more is published only sometimes, an obvious step is to publish those collected information items and to do so systematically. This particularly applies to basic information (actual dates, contracts) as well as areas where there are extensive commitments (forward budgets, conditions, transactions) and current donor priorities around results and value for money (organisation level audits, activity impact appraisals and design documents).

All organisations periodically make system upgrades and these opportunities should be used to fill in the current gaps, adding data fields into systems or establishing the processes required to produce these documents consistently. Those organisations that do not have a Freedom of Information or equivalent disclosure policy should address this urgently.

Many of the accessibility issues are basic, obvious and cost-free and should be picked up during standard web testing phases, including ensuring links continue to work following website updates or redesigns. Donors need to move rapidly away from publishing information in restrictive formats such as PDF – even just publishing the same document in the original Word version would be a step forward. Another clear accessibility gain would be improving basic website structure and navigation so that information can be found without extensive searching. That means common sense structuring and sign posting to types and levels of information, particularly for policy documents and project databases.

Recommendation 3: Aid actors must rally round the common IATI standard and increase its coverage

All donors should sign up to and implement the International Aid Transparency Initiative and partner country governments should endorse it, ensuring IATI works for their needs. There is also an important role for other aid actors and agencies to support and engage with the standard and ensure flows such as climate finance, humanitarian aid and private aid are included. Large contractors and grantees who spend foreign assistance also need to be engaged.

Comparability of information is what turns more information into better information. IATI is the common standard that is making this possible. The standard was agreed in February 2011, building on existing formats and specifically designed for the comprehensive publication of current aid information in ways that are comparable, timely and accessible. Crucially, IATI enables users to map, search and re-use information, allowing aid agencies to publish once but use many times. This reduces the cost and inefficiency of serial re-issuing of information in slightly different formats for the needs of different audiences and users.

Since the publication of the 2010 Aid Transparency Assessment, donors have proven the feasibility of IATI – with over 50% of aid flows represented by the signatories and over 30% to be published by the end of 2011. IATI is the tool required to turn the rhetoric of the Accra and Paris commitments into practice. The U.S., Japan, France and Canada are countries notable for their failure so far to sign up to IATI and should do so without delay. Engaging with the standard at this stage in its development is essential to ensure it works for them, particularly for the complexity of systems like those of the U.S. The result of such major donors remaining outside the fold is that the common standard is undermined and is not able to deliver its full potential.

[1] Defined as organisations providing more than USD 10bn per annum in ODA as reported to the DAC CRS for 2009. This average score includes Germany (KfW and GIZ), the six U.S. agencies (MCC, PEPFAR, USAID, States, DOD and Treasury), the UK (DFID and CDC), the EC, France and Japan.
[2] In these cases, DAC aggregate statistics from 2009 were used to determine largest recipient rather than CRS activity data from 2009 as the aggregate data provides more complete statistics on amounts disbursed.
[3] Donors will provide full and timely information on annual commitments and actual disbursements so that developing countries are in a position to accurately record all aid flows in their budget estimates and their accounting systems. (§26b, Accra Agenda for Action.)
[4] Organisations found in the survey to be publishing country level three year forward budget included: AsDB, Belgium, Denmark, DG Enlargement, Ireland, MCC and DFID.
[5] Donors will provide developing countries with regular and timely information on their rolling three to five year forward expenditure and/or implementation plans. (§26c, Accra Agenda for Action.)
[6] Donors and developing countries will regularly make public all conditions linked to disbursements. (§25b, Accra Agenda for Action.)
[7] The following organisations were found to publish conditions – AfDB, EuropeAid, DG Enlargement, Global Fund, MCC and WB IDA, while for MOUs the list is Hewlett Foundation, Ireland, Norway, MCC and World Bank IDA.
[8] The initial data for the Netherlands was collected in March–April 2011. During the final data verification process, the survey was updated to reflect the September IATI data release.
[9] UNOPS is not included in this index; however, it became a signatory to IATI in October 2011 and published data within a few weeks, becoming the sixth multi- or bi-lateral organisation to publish their information to the IATI Registry.