Examination of Witnesses (Questions 40-59)
DEPARTMENT FOR
INTERNATIONAL DEVELOPMENT
3 MARCH 2008
Q40 Mr Bacon: You were trying to
prevent there being a failed state by paying for the state.
Dr Shafik: For a brief period,
that is correct.
Q41 Mr Bacon: Basically what you
are saying is that you look at the benefits in the round compared
with the costs and see whether the risk is worth taking.
Dr Shafik: That is absolutely
right.
Q42 Mr Bacon: Can we go back to the
traffic lights that were mentioned earlier. You are quite right
to say that there are quite a lot of yellow and green as well
as the one or two reds, but if you look at page 56 you will see
under the heading: "Strengthening national systems"
that there are 30 traffic lights there. Two of them are bright
green which means that budget support has made a definite and
very significant difference; nine of them are pale green which
means that budget support has made a moderate difference; the
other 19 are pale yellow or bright yellow—most of them are
pale yellow—meaning that budget support has only made a
small difference or that it has made no difference at all. On
your own criteria, that is to say benefits, that chart would suggest
that budget support is failing to deliver these outstanding benefits
that would make it worthwhile going down that route. This is under
the heading: "Strengthening national systems" which,
so far as I understand this thing, is what it is supposed to be
all about.
Dr Shafik: Yes. We do have evidence
that public financial management and systems have improved in
many countries. There have been independent evaluations done in
Ghana, Rwanda, Mozambique, Tanzania, Uganda and Vietnam which
show that systems have improved. The NAO Report itself gives several
examples but it does take time. I believe this Committee has existed
for 147 years and it has taken a long time to develop the levels
of public accountability that we are accustomed to. The question
is: are the trajectories positive? I think the next appendix,
which is appendix seven, is the one which shows what the trajectories
are. We are starting with countries which start from a very, very
low base. Those trajectories show that most systems are either
stable or improving by and large. I think that is an important
factor to take into account.
Q43 Mr Bacon: In order to understand
really what is happening on the ground obviously measuring it
is pretty important. Paragraph 6.2 on page 34 says, "DFID
requires country teams to make an explicit assessment that `potential
development benefits justify the fiduciary risks' of using budget
support but only two of the appraisal documents we reviewed explicitly
pulled together analysis of risks and benefits to evidence this
assessment. In these circumstances we found it difficult to see
how the risks had been factored in to an assessment of potential
benefits to form an overall assessment." If only two of the
DFID teams out of ten that the NAO looked at there are providing
an evidence assessment that the potential development benefits
justify the fiduciary risks, how can you say that your methods
of assessment are up to the mark?
Dr Shafik: I think part of the
problem here was that some of this work was done but not necessarily
in the documents that the NAO is referring to; it might have been
done in other bits and pieces.
Q44 Mr Bacon: You mean documents
that were not shown to the NAO.
Dr Shafik: It is not that they
were not shown but they were not in a consolidated place and I
think we completely accept that it all needs to be in the same
place.
Q45 Mr Bacon: We are talking about
DFID systems now; we are now talking about strengthening systems,
that is right, is it not? I just want to be clear which systems
we are talking about when you say: "we need to strengthen".
The "we" in that is DFID.
Dr Shafik: Yes. We need clearer
and more systematic documentation. A lot of it is there and a
lot of these judgments and analyses and trade offs have been made
but they may appear in several documents and that actually makes
it quite difficult to assess the audit trail. Having said that,
we have also just issued new guidance to all of our teams specifically
on benefits assessment because I think we need much more quantitative
and consistent approaches to benefits assessment than we have
had in the past. We have tried to respond to this NAO recommendation
already.
Q46 Mr Bacon: May I draw your attention
to paragraph 5.4 where it says, "Country offices must also
produce an Annual Statement of Progress to demonstrate that there
have been no major changes in risk". It goes on, "Of
the ten country offices we reviewed we found that three quarters
of the expected update assessments were provided annually as required",
so a quarter were not. Could the reason for that be, as it says
in the previous phrase, that "only a sample of these statements
are reviewed by headquarters"? In other words, people in
the country offices do not bother because they know it will not
get looked at. Is that not right?
Dr Shafik: If I could I will ask
Rachel Turner, who is head of our office in Mozambique, to give
you a sense of the kind of accountability and scrutiny that is
provided by headquarters.
Ms Turner: I hope I am one of
the countries that did provide an assessment, but I think the
key point is that this is what we are doing all the time. As a
manager of an office that is very much what I am doing, I am taking
a view on risks and benefits. In the case of Mozambique I submit
to the Secretary of State twice a year, presenting information
about what is happening in Mozambique, about performance in Mozambique
and allow the secretary of state to take one decision early in
the year about how much money he or she wants to commit for the
next year and then before I actually sign the cheque—before
the money moves, if you like, from the accounts—I write
again to the Secretary of State with an update describing the
latest events in Mozambique, what is happening, what performance
has been achieved—
Q47 Mr Bacon: So this Annual Statement
of Progress is aside from all that and not relevant. Is that what
you are saying?
Ms Turner: The Annual Statement
of Progress is the formal part that goes with the submission.
Q48 Mr Bacon: So it is relevant.
Ms Turner: Yes, it is relevant.
Q49 Mr Bacon: But not everybody bothers
to complete it. That is what this says. It says basically that
only three quarters of the country offices bother to complete
the Annual Statement of Progress. That is right, is it not? That
is what it is saying.
Mr Sharpe: Some of them have been
a bit late but all countries have done fiduciary risk assessments
and subsequent reviews. Some of them were not done completely
within a year but most of them have been done within 18 months.
There were one or two specific circumstances in individual countries
why they were not done, but they are looked at carefully by my
team centrally.
Q50 Mr Bacon: Why do you not insist
on a minimum level of information on the use of government funds
before you start to provide budget support? If you look at paragraph
5.8 on page 29 it says that: "such assessments rarely include
quantified estimates of the impact of weak systems in terms of
possible wastage of resources". Why do you not insist on
a minimum level of information?
Dr Shafik: Again this goes back
to the issue of weighing risks versus benefits. The question is,
do we have enough information and are the benefits worth the kinds
of risks that we are talking about? Those are the judgments that
our teams have to make. The other element of the judgment is,
is this a government that is serious about improving its systems
or not and do we actually expect a positive trajectory of change?
Even if they are starting from a very low base like some of the
countries we work in are, if they are serious about improving
their systems we are willing to support them if they get a track
record.
Q51 Mr Bacon: It says later on in
this paragraph, "Six of eight DFID country offices providing
budget support used Public Expenditure Tracking Surveys"
(you mentioned public expenditure tracking earlier), why would
they not all use it?
Dr Shafik: Public Expenditure
Tracking Surveys are one very useful instrument for tracking very
narrow bits of public expenditure as to whether it is going from
the centre to the very local level, to where it is supposed to
go. It is not really a panacea in the end because in the end the
longer term solutions are around fixing national audit offices,
fixing PACs and having the whole system work properly. They are
an important check when we think that there is problem in a particular
country which, for example, we did in Sierra Leone where we thought
the health sector was a problem, drugs were not provided at clinics
and we did one of these analyses which resulted in going from
5% of drugs arriving at hospitals to 88%. It is more an additional
tool rather than a longer term solution.
Q52 Mr Bacon: Could you repeat the
sentence you said earlier about your accounts being qualified?
What is the current position again?
Dr Shafik: They have never been
qualified.
Q53 Mr Bacon: Have you read the document
The Responsibilities of an Accounting Officer?
Dr Shafik: Yes, I have.
Q54 Mr Bacon: Was it given to you
today or was it given to you before you became an Accounting Officer?
Dr Shafik: I read it before I
applied for the job.
Q55 Chairman: You just said the absolute
key was ensuring that they had adequate State Audit Institutions
and PACs, but I refer you to my original question: why only 20%
of countries where you give budget support do you give support
to PACs and State Audit Institutions?
Dr Shafik: In every country where
we provide budget support we do something on building local accountability
either through supporting national audit offices, PACs and parliaments
or through supporting civil society. What we actually do in each
country, though, depends on the local circumstances. For example,
in Malawi we provide general budget support to Malawi, we also
do some very specific things in sectors like agriculture and education;
we provide support to civil society in Malawi and we also provide
support to Parliament. However, we do not provide support to the
National Audit Office in Malawi because the Norwegians are doing
that. We often have a division of labour among donors as to who
is doing what, but in every country where we provide budget support
we are doing something to build and strengthen local accountability.
Q56 Chairman: Why does this independent
Report at paragraph 3.17 find that: "DFID had increased support
for policy and advocacy work by civil society, but it questioned
whether budget support really had increased recipient governments'
accountability to their citizens"? If what you say is right,
why does this independent Report say that?
Dr Shafik: That may be a point
of view by some UK non-governmental organisations but I think
quite a lot of them have become quite strong supporters of budget
support.
Q57 Mr Mitchell: Budget support in
this kind of way is going up; it has gone from £268 million
to £461 million over a five year period, so you must like
it. Why do you like it? Is it because it saves you staff and work?
Dr Shafik: No, it does not. Over
the long run it might, but that is not why we like it. We like
it because we think that the development benefits are greater,
that we can have a systemic impact on the government's entire
policy and public expenditure programme, because we think it gives
us a lever to improve accountability institutions, public financial
management, and we think it is a more cost effective way of delivering
public services.
Q58 Mr Mitchell: That is in theory;
you have not really done that because the accountability has not
improved all that much and you have not made strenuous efforts
to improve it.
Dr Shafik: I think there is evidence
in the NAO Report about where we are putting pressure on governments
to enhance accountability and there are many examples cited. In
Zambia people actually watch the PAC on television and there is
a high level of public interest in what the PAC does in Zambia.
In Ethiopia we have been able to build civil society so that it
is holding a very powerful and sometimes not very democratic government
more to account. So we do have very clear examples.
Q59 Mr Mitchell: To go back to the
original point, it is easier than sending stuff out, doing particular
projects, channelling aid into a particular purpose or a particular
social group or whatever. This is much easier, is it not?
Dr Shafik: Not really. In the
beginning it actually costs us more. The initial set up costs
in terms of the kind of analysis we have to do of government spending—the
coordination that is required with donors, the agreement and the
elaborate negotiations about what results we expect from budget
support—are actually greater. Over time, once you have set
up those systems and agreements, performance frameworks and monitoring
systems with the government it does reduce our administrative
costs, but that is not the reason why we do it.
| 
