Supplementary Memorandum from Mr Simon
Maxwell, Director, Overseas Development Institute
Extracts from "An Evaluation of Development
Co-operation between the European Union and Ethiopia, 1976-94",
Final Report, Institute of Development Studies, University of
Sussex, UK, and Institute of Development Research, Addis Ababa
University, Ethiopia, June 1996
(a) Delegation of Authority
1. By comparison with the World Bank and many
bilaterals (though not USAID), the EC/EU maintains relatively
large field offices. Delegations play a lead role in defining
country policy and in managing programmes. We have already cited
cases in which Delegation staff have taken the lead in policy
dialogue or in adapting programmes to changing conditions. It
is surprising, therefore, to find criticism of over-centralisation.
2. The fact is, however, that project approval
lies almost exclusively in Brussels, mainly with the EDF Committee
of member states; and that most other decisions need to be referred
to Brussels. In the case of project approval, our enquiries show
that the delegated authority of the EC/EU Delegate, acquired only
in 1986, is limited to 60,000 ECU (less than $US 75,000) and even
this is for studies only, is subject to approval by the NAO, and
requires validation by Brussels. Food aid and NGO decisions are
all taken centrally. By contrast, the World Bank Representative
has considerable leeway as a task manager, the UNDP Resident Representative
can approve projects up to $US 1 million within an approved country
programme, and the USAID Director has pretty well unlimited authority
to approve projects within an agreed strategy. Similar constraints
apply to project modifications. The Delegation is an influential
voice, of course; but there is also no doubt that the need to
refer decisions to Brussels delays the implementation of the programme.
As the comparative study of European donors reported, `the structure
is decentralised while decision-making procedures are centralised'
(Cox et al 1995: 17).
3. We assume that the lack of decentralisation
responds in part to a political imperative on the part of the
member states. Thus, all EDF projects have to be referred to the
EDF Committee of member states in Brussels. Similarly, all food
aid decisions are referred to the Food Aid Committee. These committees
have some quick approval procedures, but perhaps not enough.
4. Over-centralisation is one of the principal
causes of delay in the system. Others include delays on the Government
side, and the need to negotiate project design with Government.
We do not have data on the average time taken to prepare and approve
NIP projects, though the fact that 80 per cent of the Lomé
IV NIP remained to be committed at the end of 1994 speaks eloquently
of problems in the system. However, the sectoral reports cite
the length of time taken to approve a project for NGO cofinancing
as being up to one year, up to a similar length of time for negotiating
a Framework of Mutual Obligations for Stabex, at least ten months
to prepare and agree a tender for EDF procurement, and up to seven
months to agree and deliver emergency food aid.
5. It seems to us that the term `Delegation'
is a misnomer for the EC/EU field offices, unless it is used in
the French sense to mean `represent', rather than in the English
sense of `depute' or `give authority to'.
6. The EC/EU maintains a "Delegation"
of nine professional staff in Addis Ababa, and has a full-time
desk officer for Ethiopia in Brussels, and it seems perverse to
complain of under-staffing. We were impressed by the calibre and
hard-work of the staff, who were frequently to be found in the
office late into the evening and at weekends. The motivation and
technical skill of individuals were also recognised and praised
by the participants in the focus group discussions. However, the
individuals were seen by the focus groups to be over-burdened
with administration, to the extent that their effectiveness was
impaired.
7. Our casual observations confirm the pressure
on individuals and the work-load. They also confirm the difficulties
Delegation Advisers have in combining the administrative work-load
with the professional advisory work for which they are nominally
recruited. In the case of the Delegate, these problems are compounded
by the additional duties of political analysis, liaison and representation.
We carried out some elementary work-load analysis which gave us
similar results. In our view, the term "Adviser" is
another misnomer: for most of their day, the Delegation staff
act more as what would be known in other agencies as "Programme
Officers", carrying out a combination of administrative and
routine project management tasks. They struggle to find time for
policy analysis and policy advice. In this context, the quantitative
analysis carried out for the comparative European study, which
found that staffing levels in DG VIII (though excluding food aid)
were comparable to those of the UK, do not provide a complete
picture.
8. Food aid provides an extreme case of lack
of professional input at the field level. As the sectoral report
makes clear, until 1991 there was at most one member of the Delegation
staff with responsibility for food aid, despite the fact that
it accounted for over 30 per cent of the programme throughout
the 1980s. At the time of the evaluation, there was one expatriate
logistics specialist, with no formal mandate to cover food aid
policy. At certain periods, different Advisers have made an input
into the food aid programme. And we were advised that the Delegate
in post at the time of the evaluation had had to spend 30 per
cent of her time on food aid. This is perhaps an indication of
the magnitude of under-staffing. A new food security unit has
been established in the Delegation, adding two new poss. However,
the terms of reference of this group had yet to be defined at
the time the evaluation was carried out.
9. There is one obvious reason why the staffing
problem arises, which is the lack of support staff to work with
internationally recruited advisers. Figure 20 gives the staffing
levels for four main agencies in Addis Ababa. It shows that the
EC/EU is unique in having no local professional staff involved
in programming. The size of the programme and the number of international
professionals both suggest that the Delegation needs a substantial
number of local staff. The same argument can be made at headquarters,
where the lack of junior professional and clerical support is
particularly marked.
10. It is perhaps worth pointing out that an
international staff member will cost very much more than a local
professional, when all costs are included. We were provided with
data by the Delegation which suggest that the salary and benefit
costs of a relatively young Adviser could amount to at least $US
150,000 per annum, including air travel, rest and recuperation
leave and other benefits. By contrast, the EC/EU could hire Ethiopians
of University professor level qualifications and experience for
about $US 8,000 per annum. Thus, without costing office accommodation
and other overheads, the EC/EU could employ about 20 local staff
for the cost of each expatriate.
| Figure 20
|
| Comparative staffing levels of aid agencies in Addis Ababa
|
|
| EC/EU | USAID
| World Bank | UNDP |
|
| Annual disbursement ($ million) | +/-200
| +/-100 | +/-120 | +/-10
|
| International staff | 9 |
20 | 4 | 9 |
| Local professionals (excluding admin., pr) |
0 | 25 | 5 |
7 |
|
11. The internal staffing problems lead directly to other
characteristics of the EC/EU programme, particularly the over-reliance
on consultants and the lack of technical continuity. In our judgement,
these features have a negative impact on the quality of the programme.
Participants in the focus groups made the same point.
12. Both problems arise because the EC/EU does not in general
have the internal capacity to prepare and manage projects. For
example, the Agricultural Adviser in Addis Ababa is kept occupied
by the many administrative demands of a large and varied programme.
He is supported by one Adviser in Brussels, who has to cover several
countries. As a result, any technical work on the programme has
to be contracted out. Because of the need to be accountable for
funds disbursed, consultants are chosen by open tender. As a result,
there is no guarantee of continuity. We found that at least six
reviews had been conducted of the Coffee Improvement Project,
each by a different organisation; and that five different bodies,
in addition to the EC/EU itself, had conducted major reviews in
the relatively short lives of the North and South Shoa projects.
13. It is not our intention to denigrate the quality of individual
consultants, most of whom seem to do a competent job (though there
were complaints in the Government focus group about the effect
on quality of member state quotas, informal or otherwise, in the
selection of consultants from Europe). In some cases, consultants
were found to do an excellent job: for example, the consultants
recruited to help manage counterpart funds were singled out for
praise in the programme aid report, although it was also recommended
that their job should be handed over to local counterparts as
soon as possible.
14. There is a more general problem underlying the staffing
figures. One of the strongest comments made in our Government
focus group discussion was that, unlike the World Bank and other
agencies, the EC/EU had no single line manager for each project,
responsible for follow-up and speedy decision-taking. A Delegation
Adviser is usually responsible for coordination and decision-making
in the field, but this is not equivalent to a task manager, with
overall responsibility for a project within the EC/EU as a whole.
15. One other general problem to which we wish to draw attention
is the lack of any social development expertise in the Delegation.
This affects the quality of advice in a range of sectors, perhaps
more importantly NGOs.
16. EC/EU procedures featured large in the focus group discussions,
particularly with participants on the government side who had
first hand experience of them. The EC/EU was described as being
burdened with excessive bureaucracy. Its procurement procedures
and its tying of aid to EC/EU suppliers were strongly criticised.
Our own observations confirm many of these criticisms, though
we are not always in a position to provide chapter and verse on
procedures which take place largely in Brussels.
17. The sectoral reports provide some evidence of procedural
delays. Those in project design and approval have already been
mentioned. Others relate to contracting and tendering procedures.
The programme aid report, for example, contains a table estimating
the duration of EDF procurement procedures. It suggests that from
preparation of tender documents to delivery of goods may take
from 13 to 20 months, with tender preparation taking 1-3 months,
approval of tender documents by the Commission taking 1-4 months
and evaluation of tender taking up to 6 months. Some of the delays
are on the Government side, but many are in the domain of the
EC/EU. For example, procurement processes could be shortened by
greater delegation on tender approvals and reduced time for flotation
of tenders.
18. The programme aid report also provides the best source
on the detail of why programme aid disbursements are slow—and
so much slower than those of other donors. It discusses the rules
of origin for imports, the requirement to provide a customs clearance
certificate for donors, and the restrictive sectoral targeting
of imports. Again, the report makes recommendations as to how
to short-circuit some of these procedures, without sacrificing
probity or accountability.
19. The EC/EU is sensitive to some of these problems and
has made some attempt to simplify procedures. For example, the
food aid programme is criticised for only allowing a single annual
request for food aid to be made, at a time of the year when the
final results of the harvest are not known and when new emergency
needs may develop. The flexible commitment of food aid to the
Food Security Reserve is one response to this problem. The Commission
has also made it possible to request additional food aid at two
or three points in the year.
(d) Aid and Political Relationships
20. Critics of the EC/EU charge that it behaved quite immorally
in providing aid to the Mengistu regime from 1976 to 1991. This,
they argue, was a regime that came to and held power by force
of arms, consistently ignored human rights, allegedly put to death
over 1.5 million of its own citizens, imposed resettlement and
villagisation on an unwilling peasantry, and fought a bitter war
against Eritrea, quite apart from the confiscation of property
and supposed loss of freedom that accompanied the introduction
of a socialist mode of development.
21. The Mengistu regime was regularly condemned, including
by individual states of the EC/EU and by the European Parliament,
which voted to suspend aid to Ethiopia in November 1982. Yet the
EC/EU, almost alone among non-UN donors, continued to provide
development aid on a large scale, signing National Indicative
Programmes in 1976, 1981, 1986 and 1991. Only the last of these
mentioned human rights, and then in rather general terms.
22. Difficult issues were involved here for the EC/EU. First,
Lomés I-III were deliberately non-political. Indeed, as
treaties in international law, it might be argued that they imposed
obligations on the Commission, regardless of the political complexion
of individual recipients. The contractuality of Lomé was
presented as one of its great virtues. Only in Lomé IV
did the principle of contractuality begin to be eroded and discussion
of conditionality, including over human rights, begin to appear.
23. In the 1990s, the trend has continued and "Good
Government" has become a major theme of development aid.
At the same time, the competence of the Commission in the field
of foreign policy has gradually begun to increase. Thus, the climate
for human rights conditionality on aid is quite different now
to before 1991. A more robust policy is both more likely and,
given the changes to Lomé, more feasible.
24. A second issue was that the case of aid to Ethiopia could
be said to derive from the impact it had on poverty and on the
policy leverage it could exert. Thus, poor Ethiopians were not
responsible for their government and should not be penalised by
the withdrawal of aid; and, in any case, the provision of aid
helped to shift the regime away from the more extreme of its policies.
The EC/EU, it could then be argued, was doing its best to find
"room for manoeuvre" in a difficult situation. Member
states could support its continuation for that reason.
25. The argument that EC/EU aid helped to protect the poor
is certainly true for emergency aid, which accounted for around
50 per cent of the programme through the 1980s (most of it in
the form of food aid). But emergency aid was never at issue: most
large donors provided emergency aid, even though development aid
was proscribed. The debate then centres on whether development
aid reached the poorest. In Lomés I and II, most of the
resources in the National Indicative Programme went to roads in
higher potential areas; and in Lomé III, most went to agriculture
and rural development, again with a strong, though not exclusive
bias, to higher potential areas south and west of Addis Ababa.
Coffee farmers, for example, were an important focus: and although
they were certainly poor, they were far from the poorest farmers
in Ethiopia. Furthermore, the foreign exchange they earned, as
well as the taxes they paid, helped to feed the war in the North.
A greater poverty orientation might have been possible, though
the EC/EU had limited say over the choice of projects, especially
in the early years.
26. The last argument is that a policy of "constructive
engagement" helped to secure political and economic change
in Ethiopia. Economic liberalisation certainly took place after
1988 and the EC/EU played a part in bringing it about (see discussion
in Section 13). The case for political leverage is much weaker,
however. In the end, the regime was brought down by the withdrawal
of support from the USSR, its own economic weakness, its lack
of moral and political legitimacy, and by the military efforts
of the EPRDF.
(e) Market liberalisation in 1988 and 1990: A case of
successful policy dialogue?
27. From 1974 until 1988, Ethiopian agriculture was subject
to a high degree of Government intervention in production and
marketing. State farms and producer cooperatives received the
lion's share of investment; land and labour markets were banned;
there were severe restrictions on private trade in grain; and
levies were imposed on peasant producers, to supply food at fixed
prices for the urban ration system.
28. In 1988, producer prices were raised and private marketing
was made a little easier. In 1990, a much greater degree of liberalisation
was introduced: the bias to producer cooperatives was eliminated;
private tree tenure was allowed; the hiring in of labour was made
legal, the quota system was abolished; and the private grain trade
was further encouraged. The 1990 reforms effectively sounded a
death knell for socialist agriculture in Ethiopia. They set the
stage for the deeper and faster liberalisation that took place
when the Transitional Government came to power in 1991.
29. These liberalisation measures were objectives that the
EC/EU had pursued for several years. In particular, they were
central conditions to the approval, in 1986, of a National Indicative
Programme for Lomé III, covering the period 1986-90. The
indicative programme was signed in March 1986 by Mersie Ejigu,
Minister for Planning (Ethiopia), and Dieter Frisch, Director
General for Development (EC/EU). It committed the EC/EU to provide
230 million ECU of aid to Ethiopia. For its part, the Government
agreed to partial liberalisation of the grain market, a review
of official procurement prices, a review of land tenure, and other
measures, all designed to "provide adequate incentives to
peasants to increase their production".
30. How far can the EC/EU claim credit for the changes that
eventually resulted in 1988 and 1990? Many other donors and outside
observers were pushing in the same direction not least the World
Bank, which suspended its own agricultural lending until reforms
were introduced. The long-standing denial of development aid by
some large bilateral donors could have been equally influential.
Within the Government itself, a strong reformist tendency was
arguing for the same changes. It is also relevant that the Ethiopian
regime's strongest supporter, the Soviet Union was undergoing
perestroika during the same period.
31. It is probably impossible to be certain in retrospect
how large a part the EC/EU played. However, there were certainly
factors which pre-disposed the Government to listen carefully
to the EC/EU's views. First, the EC/EU maintained development
aid relationships with Ethiopia at a time when many of its own
member states, and other large donors, did not. Secondly, a good
aid relationship was fostered, with an emphasis on dialogue rather
than conditionality. Dieter Frisch himself took an unusual interest
in Ethiopia, and Minister Mersie was a frequent visitor to Brussels.
Thirdly, the sums available were substantial; the EC/EU provided
around 100 million ECU a year in total to Ethiopia in the second
half of the 1980's (including non-Lomé instruments), or
around 15 per cent of all the aid received by the country. And
fourthly, the Commission was not averse to flexing its muscles,
delaying approval of agricultural projects under Lomé III,
for example, until the price increases had been announced.
32. On the EC/EU side, the process demonstrated some of the
conditions necessary for successful EC/EU policy dialogue: a clear
policy orientation, in this case the focus on agriculture and
rural development, written into Lomé III; good liaison
and preparation with member states and other donors; high-level
management attention; the ability to assimilate, if not undertake,
policy analysis; and the fostering of a good aid relationship
albeit one with teeth.
Simon Maxwell
Overseas Development Institute
|
