DR Congo is a fragile post-conflict state which ranks at the very bottom of the UNDP Human Development Index (187th, UNDP 2011). Its gross national income is USD 167 per capita (OECD, 2011), 71% of the population live below the poverty line of $1.25/day. The population is 67.7million with an average life expectancy of 48 years.
DRC stands to achieve none of the Millennium Development Goals by 2015 (OECD 2011)
The Government’s fiscal year is January-December.
After the end of the “Second Congo War” in 2003, and before DRC’s first democratic elections in 2006, many bilateral and multilateral donors returned to DRC and are increasingly implementing development-oriented programmes in place of humanitarian assistance. ODA has steadily increased since the 2006 elections (which cost donors an estimated $500million).
According to nationally elaborated statistics, ODA reached USD 2.086 billion in 2009 with over half coming from the WB, China, the EC and the UK. The OECD accorded a figure of USD 2.354 billion for the same year (OECD 2011). An extra approx. $1 billion came in humanitarian aid in 2010
DP funding represents approximately 50% of the central government budget. (Ministry of Planning 2012)
DRC is highly aid-dependent, particularly for the continual functioning of basic social services. Geography, access, difficult working conditions, governance issues contribute to complex operating environments for donors
According to DAC statistics, in 2010 no donor was providing budget support (general or sector) to the DRC. The majority of aid to DRC is provided through ‘agences d’execution’, pooled fund mechanisms, or to third party service providers (NGOs and UN) (OECD 2011: 11-12)
Using country systems
DR Congo performs poorly against most Paris Declaration indicators. The 2011 Paris Declaration Survey reported 13% as the proportion of aid using country public financial management systems, against a survey average of 37%, and 9% using country procurement systems. Reliable PFM systems scored 2.5, no improvement on the 2005 and 2007 scores (OECD 2011). No donors provide regular budget support to DRC although the WB, IMF, EC, AfDB and Belgium provided exceptional general budget support in 2009 and 2010 in order to respond to the global food and financial crises, which had hit DRC particularly hard as the price of commodities dropped significantly, and to maintain macroeconomic stability.
Most bilateral donors only returned to DRC in the mid-2000s, having been absent during the end of the Mobutu regime and consequent years of conflict. The World Bank was most prominent in the early years of 2000 and was instrumental in setting up the numerous ‘agences d’execution’, parallel implementation agencies created primarily to ring-fence aid investments, and which continue to play an important role in DRC donor portfolios. Most donors straddle humanitarian contributions with ‘transition’ programmes and more recent development programmes, reflecting the reality within the country, where some provinces enjoy relative stability and high growth (Katanga), while others continue to see conflict, mass human displacement, an absent state and fledgling peace processes (the neighbouring Kivus). Government faces serious challenges and donors are still hesitant to work with counterparts beyond basic developmental courtesies. Senior GoDRC officials also do not consistently engage with development aid and donors, despite external resources supporting the majority of the country’s basic services and over half of its budget. Coordination among donors is weak. China signed an agreement with the GoRDC in 2007 to provide $9billion worth of infrastructure in return for copper, cobalt, and other minerals. The deal was renegotiated in 2009 to $6billion. Very little is known about the details of the project, and no financial imputations of the swap deal appear in national budget documents.
1. COMMON CLASSIFICATION
1.1 Would more comprehensive, more accurate data but in a format that still requires some work to align perfectly, be preferable to how the country receives information currently, in respect of formats, accuracy, timeliness etc.
The Common Classification depends on a few assumptions which are not necessarily in place in DRC, where aid information provision, aid coordination more generally, and national budgetary capacity more generally still, are weak. The CC would build on a certain level and quality of aid information which is not currently available in DRC. It also assumes some kind of existing and functional translation of information from donor agency to national budget, which is also not present. However at the same time this represents some level of opportunity, because a plausible IATI Standard with CC could provide numerous solutions to problems which currently seem intractable.
There was a low level of awareness of IATI, its potential and current status in DRC. Few donor respondents were aware of IATI and those that were, were still sceptical. Firstly, it is seen as a top-down initiative and little information had trickled from aid effectiveness sections in headquarters to colleagues in Kinshasa. Secondly, as yet emerging practices of regular publication to IATI result in scepticism at country level about its usefulness. Thirdly, engaging with government is not currently as high a priority for all the donor agencies interviewed as it might be in other countries. Providing aid information involves interactions with GoDRC which many donors are not fully able to take on board, and many mix up too easily this type of engagement with use-of-country-systems issues. Fourthly, many donors argued - not irrationally - that a competent transfer of aid information to government – and ensuing improvements to the national budget systems – should only come after they have invested in coordination of their own portfolios, coordination with other donors, and become more comfortable in a complex, fledgling and currently rather dysfunctional donor environment. Finally, and related, donors wonder what the point would be of providing higher quality aid information into a system which is itself still emerging and not able to use such data competently.
The Ministry of Planning has been keen to embrace much of the agenda for aid effectiveness, has signed up as an IATI pilot in addition to undertaking a number of local initiatives to prioritise aid effectiveness. However, currently the Ministry of Planning is far from being able to produce reliable data on aid, be it forward planning, current implementation, or past execution. It also does not yet provide to the Ministry of Budget, who is charged with preparing and monitoring the national budget, accurate or useable figures on donor activity. This is compounded by the fact that the Ministry of Planning’s aid information management system (Plateforme pour la Gestion de l’Aide et des Investissements: http://pga.ministereduplan.cd, PGAI) is the only repository for aid data in DRC. The mechanism is mistrusted, misunderstood or just ignored by most donors, which undermines its effectiveness. As such, donors are difficult to incentivise to make improvements to the quality of their inputs to the system.
Overall, therefore, implementation of the IATI standard (not only the common classification) would on balance provide an improvement over the current situation, in which aid information is poorly shared and badly managed. Timeliness, accuracy would surely improve. Detail would likely improve. Formats are currently messy in DRC and the IATI Standard would represent a clear improvement. There are a few caveats however:
All donors provide less information than necessary or useful to the GoRDC, with the exception of the World Bank. It has (largely through the initiative of individuals) taken to providing aid information to the GoRDC with high levels of granularity, within correct timescales, with local context and classifications integrated. IATI data (particularly the data currently supplied by the WB centrally to IATI for the DRC) would represent a drop in quality for the World Bank particularly if switched today (the biggest donor to DRC). The Ministry of Planning, having been given the choice between switching to IATI automatic data exchange and retaining current practice, have chosen the latter in the case of the WB. For all other donors, though, reliable IATI data would represent an improvement over the current provision of aid information
The amount of local work required would still be significant in order to identify the granularity and particular details important to the GoDRC. Geocoding was said to be particularly important to GoDRC stakeholders, and the data made available by donors in country on this was particularly poor. IATI geocoding would also need a strong push. This weakness likely comes from an overall weak M&E framework in DRC, where donors become separated from implementation on the ground. Given that many donors currently devote negligible resources to providing aid information in DRC, refining and verifying IATI data would actually represent an increase in workload for some donor staff.
The IATI mission of April 2012 proposed to introduce automatic data exchange between data for DFID and the World Bank, and the PGAI in DRC. The mission established that figures published by DFID to IATI differed from funding statistics on the DFID projects website (http://projects.dfid.gov.uk/) which differ again from the data provided to the PGAI by DFID in DRC. Subsequent investigation revealed that the IATI figures were accurate.
China is a significant actor in the DRC (worth some $6 billion), and as a non-signatory to IATI, a future IATI-influenced system would still suffer from gaps. However China currently provides no aid information to the GoDRC anyway so this is not necessarily an issue for IATI
Finally, it bears underlining that national systems are so weak in DRC that improved aid information from IATI would only be one first step in a huge challenge of making aid information useful to the DRC, and making PFM and democratic systems and the budget cycle more broadly incrementally more fit for purpose.
1.2 Would the proposed common classification system enable better / faster / more comprehensive / more efficient translation, alignment and absorption of aid information for budget preparation purposes, budget execution decisions and budget reporting purposes at both or either central and line ministry level? Does it provide a better fit with the country main vote structure than current classifications used by donors?
As above, the process of putting aid information into the budget process would require numerous local improvements in capacity and practice before IATI would be able to make a noticeable impact. However IATI might be able to act as a catalyst for such improvements. Donors do not currently report aid to the PGAI for budget preparation with a classification that maps to the budget – in fact it is quite rare for donors to report useful forward aid information at all. Any translation is done by PGAI staff working with the Department de Preparation et de Suivi Budgétaire (DPSB – Dept for Budget Planning and Monitoring) from the Ministry of Budget. The level of aid which gets onto the budget accurately, or is monitored by the Ministry of Planning, is low. This is not simply a classification issue. Donors either provide spreadsheets for the PGAI to key in their aid data, or they enter it into the PGAI web database themselves (some do neither), but in all cases the sector information requires manual mapping by PGAI/DPSB. Economic classifications for projects have been requested on the initiative of the Ministry of Budget of donors, but as yet no donor has provided this information.
IATI could potentially lighten the workload, reduce error, and increase comprehensiveness for stakeholders in DRC. IATI’s publication of future budgets, if reliable, would contribute significantly to budget preparation, where current practice is practically non-existent. More reliable and up to date disbursement figures would be an improvement over current practice (IATI proposes to publish every 3 months – such regularity is rare in DRC). There would still be much local work needed to make information relevant, useable, and, critically, understood by its users.
There is not a lot of donor interest in translating aid onto the budget, because as yet they don’t see the budget as a credible document or the process as credible. Ongoing PFM reforms in DRC aim among other goals at increasing the credibility of the budget.
1.3 Would the proposed common classification aid parliamentary ex ante and ex post oversight, i.e. by providing an independent flow of information that can be aligned with budget formats by parliament itself?
Aid information is presented in the Investment budget but it is of very poor quality and it would be hard for Parliament to exercise critical oversight of the aid environment based on what is provided in the budget documents. The same is true of budget execution reports.
However significant problems centre not only on the quality of the aid information but the credibility and comprehensiveness of the budget document and process. Furthermore, parliamentary interaction with aid is virtually non-existent in DRC.
It seems improbable then that IATI can have an impact on parliamentary activity in the short to medium term. Parliament exercises weak oversight over aid information in the budget cycle, primarily because the information it receives in the budget documents is inaccurate, also because the budget more broadly is relatively unrelated to GoRDC spending, thirdly because Parliament has no capacity or incentives to engage with aid information. If a future IATI CC could bring aid information into budget discussions with some reliability, it may incentivise parliamentary actors to take a closer look, but for the time being this seems remote. Parliamentary discussions are also still highly politicised and driven by the divides between parties, with less attention to a unified oversight role over the executive.
1.4 Does the current format and/or process in which aid information is received from donors -- either for the aid management central unit, central budget unit, or line ministry aid management and budget officers – hinder/slow down/render inaccurate its absorption into budget preparation, execution and reporting processes?
Yes it does, to the extent that aid information cannot be said to be absorbed into any budgetary processes with any degree of accuracy. Donors are very weakly incentivised to provide timely, comprehensive, reliable information. Typical problems with the current process include:
Late forward budget forecasts - N+1 forecasts are asked in June, and are rarely provided before October (forcing PGAI and DPSB to scramble information desperately from donors as budget deadlines approach, or in many instances, guestimate), if at all. Poor predictability of aid means that often only broad indicative figures are made available. For instance USAID is only able to provide sector envelopes for the next budget year during budget preparation (June/July), which are of limited use to the Ministry of Budget (USA fiscal year is Oct 1 – Sept 30).
Incomplete information (deriving from a mix of unmotivated donor inputs, and overambitious PGAI demands)
Incorrect information (e.g. in respect of geographical location, sector, disbursements)
Not all donors provide info (France and CHINA do not provide information)
Donors do not fully understand why they are providing their information so they pay scant attention to formats and details which could be more useful to the GoDRC. The aid coordination environment is also still confused, with clear communication required on what is expected from donors and needed by the GoDRC.
As a result, only a fraction of external aid is captured in the budget, and what is there is generally of poor quality. Reasons for the poor information – despite formal procedures for its capture -- include that donors lack the capacity, time and motivation to provide information; the complexity of the PGAI with ambitious data aspirations combined with still insufficient donors trained mean that the database can be difficult for donors to negotiate; in practice with data being keyed in up to 3 times (by the donor, the PGAI and the Ministry of Budget) before the budget document errors occur; and the server is occasionally down with the result that donors cannot access the online database and when it is up it is slow.
1.5 What is the process for converting donor information into country budget information?
1.6 Who converts aid information from the classifications used by the donor (by programme, project, activity) into information for inclusion in budget processes and in budget documentation (e.g. donors, aid management unit, line ministries, budget office)?
Projects are entered into the PGAI when they come online, with a bare minimum of information as described below. Donors are encouraged to capture as much information as possible about the projects, including financials (committed, disbursed, future years), narratives (including M&E objectives and results), alignment details (to national priorities, DAC sector codes), partners, geography, modality, risk. In reality, the bare minimum is the most which can be expected in many cases.
As per the Arreté Interministeriel of December 2011 (see below), the following information is the bare requirement (Article 39):
- Identification number
- nature of aid: (BS, project grant or loan);
- disbursement timetable
- amount in currency and equivalent in Congolese Franc (CFR)
- Beneficiary Service (e.g. Ministry dept, NGO etc)
- Activity Sector (from DAC)
- Geographic location
Donors are encouraged to input commitment data twice per year (indicative and confirmed, April and Oct), and update disbursement data every 3 months. This rarely happens.
Projects are analysed by the PGAI team and passed to the DPSB in Ministry of Budget, where they are entered onto the investment/development budget, although there are serious discrepancies and inaccuracies here, not least with regard to what constitutes aid “for the benefit of the structures of the state”, which is the requisite for putting aid on budget according to the Ministry of Budget (2012). The vagueness of this definition is not lost when it comes to putting aid on budget. DPSB are left to guestimate many pieces of information for their needs because they are not provided by PGAI data.
There is no automatic transfer of data from the PGAI to the DPSB, therefore information on forward aid budgets (collected in theory in June, in practice nearer to October), actual disbursements (in theory trimestrial, in practice on average once or twice a year, sometime not even), and other relevant narrative data (if ever it exists) is transferred to the DPSB manually. The mapping process for translating donor data onto the national budget is not transparent, has not been agreed by donors, is poorly understood by them, and even less trusted.
1.7 When does it occur (how many times in the life time of a project, in the budget cycle?)
Donors should input their basic project data when they come online. Donors are asked for N+1 aid forecasts twice per year, in April and October. This fits with the national budget calendar which aims to produce the budget document by October N-1. Donors are asked to provide disbursement reports every 3 months (in theory this would ease the problems of overlapping fiscal years). These recommendations are laid out in the Arreté.
1.8 By what process?
Donors each have a PGAI-nominated focal point who has a PGAI username and is responsible for inputting data directly onto the internet database. Some do this, but many provide spreadsheets, tables etc, often drawn down from their own data systems, if time is tight/ they can’t access or use the PGAI site/ they are not sufficiently committed to providing the PGAI with the data they need. In many cases the PGAI end up having to come to donor offices to procure whatever information they can related to expenditure.
1.9 What is the burden on donors? What are the problems they experience?
This depends on the donor. Firstly, of the PGAI’s focal point list, some individuals are project managers, some programme managers and some are finance officers. Those donors with Finance Officers inputting data tend to find the demands of the PGAI more manageable, but access to some data at required times within their own organisation can be problematic (i.e. DFID’s focal point, the Finance Officer, was good with the figures but had trouble getting programmatic detail from the programme managers for inputting into the PGAI). Programme manager focal points have the data, but are less likely to have the time or perhaps capacity for inputting it. Sometimes granularity is lost in the process of getting the information to the focal point. Because of a widespread absence of motivation for providing aid information to the GoDRC, higher placed programme personnel are less likely to provide information. Nevertheless, because of poor general levels of understanding among focal points about what information the PGAI needs and how to input it correctly and efficiently, (lack of training on how to use the PGAI is a problem) there is a burden on donors. This they argue is problematic, as they do not see the benefit of providing information given the opacity of how this information translates into budget information and the weakness of reports the PGAI produce given the low level of uptake among donors.
Among the problems donors can face:
lack of coherent approach by the PGAI, while the PGAI database remains a work in progress and while the PGAI staff work out what are the priority areas of information that they need from donors. For instance PGAI staff caused confusion in late 2011 by asking for disbursement information at monthly intervals before returning to quarterly demands. Depending on the context, PGAI staff might claim that the absolute priority of the PGAI is capturing execution one day, capturing forward commitments the next, and collecting data on impact of aid the next. The PGAI database allows for high levels of granularity, and broad areas of detail, and the Ministry of Planning has difficulty sometimes in prioritising certain data sets over others.
the server can often be down and it is terribly slow. For donors who allocate a specific time to sit and input aid information, this can be a source of frustration
1.10 How accurate is the translation? Would the common classification assist in terms of the accuracy of recording aid against the budget?
The translation of donor information onto the budget is fraught with error. Donors have for instance been accredited with providing budget support in the budget when they did not. Elsewhere, because of the nature of using Agences d’Execution and of confused reporting requirements, cases of double and triple recording of aid data by the PGAI have been noted. France was accredited in the PGAI with spending $30million more than it had done in a given year. There is something of a mutual blame-game to explain the deficiencies: PGAI staff blame poor quality, last minute, or careless data (or none at all which PGAI then have to guestimate, or draw from the CRS 18 months later which makes it quite useless) while donors decry lack of capacity in the PGAI (whose admin staff often end up keying the info into the PGAI system off poorly-understood or incomplete donor information formats) and low commitment to verifying and sourcing data (France argued the PGAI staff should be physically coming to donor offices to get aid data instead of email and telephone reminders).
To take an example, (difficult to do because data is difficult to identify in the PGAI and the budget documents) of a DFID project for support to the DRC Humanitarian Pooled Fund.
On the DFID projects website, this project was said to have budgeted and disbursed (approximately – exchange rate not known) $35million in 2011
In the PGAI, $15.6 million is forecast for this project in 2011 on Jan 1, and 35 million is entered as disbursed on 31/12
Some 44 billion Congolese Francs (FRC), about $48million, is indicated in the investment budget for this project in a total of 51 billion FRC under ‘Humanitarian Action and National Solidarity’ credited to the General Secretariat (chapter 70003). Synthesized versions of the budget are available classified by economic nature, but there is no indication of how such donor financed projects could have been broken down this way.
The budget execution report lists the voted amount for chapter 70003 as 70.2 billion FRC (about$75million) in the execution report of end March 2011 (the latest publicly available at the time of writing). The national contribution to chapter 70003 is approx $103,000.
Such humanitarian pooled funding does not belong in the budget documents according to the Arreté in any case, the government having no say in how the funding is spent (the humanitarian pooled fund is coordinated and allocated by the Special Representative of the Secretary General/Humanitarian Coordinator of the UN, mainly to UN agencies and INGOs). Arguably such information would nevertheless inform Parliament and their vote of funding allocations, and therefore deserves to be made available ‘on parliament’, but the point is that there is no clear understanding of what goes where, who can use it, and why. Capacity for PGAI to decide correctly what aid should go on the budget is weak and often usurped simply by availability i.e. if we can get the information, we’ll put it into the budget.
It is clear that with this level of disarray, IATI CC can help bring accuracy and traceability to the process of recording aid on the budget.