European Court of Auditors response to
the House of Lords European Union Committee report on Financial
Management and Fraud in the European Union: Perceptions, Facts
and Proposals
Your Committee's recent report on the EU budget
has been read with considerable interest here in the Court of
Auditors. My colleagues and I thought that it might be of interest
to set out the Court's reactions to a number of its observations
and recommendations, which in some cases go beyond the subject
matter discussed when three of my colleagues and I met the Sub-Committee
in Luxembourg on 16 May. Specifically, we should like to comment
on:
— the scope and nature of the "DAS"
audit;
— "national declarations"
and the role of national audit institutions in the audit of EU
expenditure.
THE DAS AUDIT
The Committee's report raises a number of issues
about the interpretation and practical implementation of the "DAS"
provisions in Article 248 of the Treaty establishing the European
Communities.
First, the Committee recommends that "the
annual examination and audit of all revenue and expenditure of
the Commission should be separated from the broader objectives
of the Statement of Assurance" (paragraph 17).
This recommendation appears to flow from the
analysis of Article 248 in paragraph 17 of the Committee's report.
Paragraph 17 describes the requirement to provide a statement
of assurance, set out in the second paragraph of Article 248,
as additional to (and by implication separate from) the requirement
in the first paragraph of Article 248 to examine the accounts
of all revenue and expenditure.
The Court was not the author of the "DAS"
paragraph inserted into the Treaty at Maastricht: but it has assumed
that the two paragraphs are to be read together and interpreted,
as it were, in apposition. The Court is to audit revenue and expenditure
and, on the basis of that audit, provide a statement of assurance.
It is difficult to see how else, in practice, the two paragraphs
concerned could be interpreted.
Second, there is the question whether there
should be one or many statements of assurance. The Committee concludes
that "the Court's decision to give a single Statement of
Assurance on the accounts as a whole means that a positive Statement
is difficult to achieve" and recommends instead "statements
on each of the spending areas; in much the same way that the National
Audit Office in the UK issues separate audits for each Government
department" (paragraph 23). In paragraph 35 of its report,
on the other hand, the Committee supports "the recent decision
of the Court of Auditors to produce a Statement of Assurance giving
details of each of the areas analysed" and recommends "that
this should be developed into a Statement which concentrates on
an analysis of the audits conducted in each expenditure category
and Member State rather than on the single Statement of Assurance
on all the accounts."
The Court would wish to make four points in
reply.
— The form of the statement of assurance
cannot be settled independently of the structure of the Commission's
accounts. The Court is enjoined by the Treaty to make a statement
of assurance on the reliability of the accounts (and on the legality
and regularity of the underlying transactions). There is a single
set of accounts for the Community Budget: logically there can
only be a single statement of assurance for the Community Budget.
(There is of course a separate statement of assurance for the
European Development Fund, which has separate accounts).
— The Court has, for every year since
1998, provided "specific assessments" giving more detailed
information about the main areas of the budget and, in the Committee's
words, "concentrating on an analysis of the audits conducted
in each expenditure category" (and revenue)—though
the Court has not, pace the Committee's report, produced
separate statements of assurance for each category.
— Within this framework the Court has
commented on the legality and regularity of different categories
of transaction, giving a positive assessment on commitments, receipts,
administrative expenditure and pre-accession aid, but not on payments
for internal policies, the structural funds etc; and, in the 2004
and 2005 annual reports, commending the Integrated Administration
and Control System (IACS) for agricultural spending where it is
properly applied. It is not the form in which the Court publishes
its findings that leads to the absence of a "positive DAS":
it is the facts as we find them.
— Where appropriate, the annual report
contains references to findings in specific Member States, either
to illustrate the Court's findings or if the Court's audit has
identified a particular national problem. In addition the chapter
on agriculture contains detailed information at Member State level
provided to and by the Commission. But the Treaty does not ask
the Court to provide separate assessments for all programmes in
all Member States; and doing so would be impractical without a
massive expansion of the scale of the Court's activities and resources.
Third, the Committee's report encourages the
Court "to put in place measures clearly to distinguish between
irregularity and fraud and to publish separate figures for the
level of fraudulent transactions and administrative mistakes."
(paragraph 31).
The Court of course shares the Committee's view
that it is wrong to regard all errors in payment as a reflection
of fraud, ie as a result of deliberate and criminal attempts to
misapply EU funds. At the same time "administrative mistakes"
perhaps gives the impression that clerical failures by those administering
EU funds are the principal problem uncovered by the Court's audit.
In fact, as the Court's annual reports have repeatedly made clear,
most errors identified by the Court appear at final beneficiary
level, in the amounts claimed from the Community budget. Such
errors may be caused by the very complex nature of the relevant
EU legislation.
Whether in any particular case an error occurs
as a result of fraud or from mere confusion or carelessness is
a question which it is rarely possible to answer from an audit
alone. I am afraid that our existing practice of sending to OLAF
details of suspected frauds uncovered by the Court's audits is
far more practical than attempting ourselves to estimate fraud
and irregularity separately. OLAF are best placed to estimate
the impact of fraud.
AUDIT METHODS
The Committee's report appears to contain two
strains of criticism of the Court's audit methodology, both relating
to the Court's use of sampling, and both concentrating on the
second aspect of the DAS—the legality and regularity of
transactions.
On the one hand the Committee appears to ask
for more and better sampling of transactions. Paragraph 28 of
the report says "Due to the small number of transactions
actually looked at each year we do not consider that this methodology
can lead to an accurate picture of financial management. The Court
should aim to improve the methodology behind the Statement's production
so as to provide more accurate data. We consider that these weaknesses
must be remedied as a matter of priority so an accurate picture
of the error rate can be obtained." Paragraph 38 also calls
for "the development of a sound basis for payment transaction
sampling" in the context of "naming and shaming"
of individual Member States (see above).
Paragraph 29 however puts the emphasis on other
sources of evidence. "We are pleased to see that the Court
now conducts an assessment of supervisory systems and controls;
reviews the Annual Activity Reports and Declarations from each
of the Directors General in the Commission; and evaluates the
results of other auditors. We consider that these need to be further
developed to give a more rounded picture of performance over the
year. In particular, greater use by the Court of the Annual Activity
Reports could add positive pressure for their development into
proper accounting tools."
It is perhaps useful to go back to first principles.
— The Court uses a number of sources
of evidence in arriving at its audit judgment.
— The extent to which the Court can
place reliance on ("take assurance from") analysis of
systems, the reports of Directors General etc depends on the quality
of the systems concerned and the extent to which they assist in
answering the Treaty DAS questions. In a good many areas of the
EU budget the Court has found that financial control systems are
weak; it follows that the Court can place little reliance on them
as a source of assurance for the DAS. In some areas of the budget
there are financial reporting arrangements which work reasonably
well (for example the annual "financial clearance" of
CAP paying agencies' accounts); but they do not necessarily provide
a satisfactory answer to the DAS questions (for example the "certifying
bodies" which audit the accounts of CAP paying agencies do
not necessarily investigate transactions all the way to the final
beneficiary).
— However good the management systems
in place, there must be some direct testing of transactions. For
the reasons set out above the Court has to rely on transactions
testing to a greater extent than might be ideal.
— Two arguments might be advanced for
increasing the sample size. The first is that a larger sample
size might lead the Court to change its view as to the level of
irregularities present within the payments for a particular year
(ie lead us to conclude that a level of error that appeared at
first material would appear immaterial if a further three or five
hundred transactions were tested). In practice both the nature
of the results the Court finds and the corroboration available
from sources such as the IACS inspections suggest that our results
can be considered as robust.
— The second argument is that a larger
sample would assist Member States, Parliament, and the Commission
in identifying where things go wrong (by type of transaction,
or by Member State). There is some force to this, but it is necessary
to consider several factors here: the sample is not the only source
of information; for most transaction streams risks change little
from year to year; the Commission has access to our cumulative
results; and the Commission itself is responsible for monitoring
the implementation of control systems, and for proposing penalties
when Member States fail to implement systems.
— Mr Gray pointed out in his evidence
to the Committee that the Commission has more than 100 auditors
to examine Structural Funds expenditure and more than 100 to examine
agricultural expenditure. My understanding of the situation is
that the Commission has the information it needs, but that there
could be an increase in accountability if the Commission found
some way of summarising the results of its examinations and proposed
corrections, and of publishing this in the Annual Accounts. This
would be in line with the Single Audit principles, and would enable
the Court to comment when it had reason to believe that the information
was out of line with its own findings.
It is perhaps worth underlining the point that
audit methodology is not something that the auditor can freely
choose, and that the best way to more economical audit by the
Court is through improved management by the Commission. Ms Andreasen
(page 111 of the Committee's minutes of evidence) makes the point
very well:
"The question here is do we ask the European
Court of Auditors to hire 10,000 people to do the audit or do
we ask the European Commission to strengthen the controls. My
response is you should ask the European Commission to strengthen
the controls."
NATIONAL DECLARATIONS
AND NATIONAL
AUDIT INSTITUTIONS
The Committee is "strongly in favour of
a national Statement of Assurance on the monies disbursed in each
Member State. As in the Dutch pilot project, such a Statement
should be sent to national parliaments as well as to the Commission
as we consider that this will encourage the Member States to take
responsibility for the systems and controls they operate"
(paragraph 105); though it does not consider that a political
signature is necessary (paragraph 106) or that a single national
account is desirable (paragraphs 111 to 112).
The Committee also expresses the hope that joint
or single audit work including Supreme Audit Institutions (SAIs)
and the Court can be developed (paragraph 121) and envisages arrangements
in which the Court would have a supervisory role over the audits
of European expenditure conducted by the Supreme Audit Institutions
in the Member States" (paragraph 123).
How Member State governments and SAIs can best
contribute to improving the management and audit of EU funds is
a large question. Articles 274 and 280 of the Treaty place clear
responsibilities on Member States for the management of the budget
and the combat of fraud. Individual SAIs already produce reports
of considerable interest and value on EU spending (for example
the UK NAO's recent report on the Rural Payments Agency, and its
report of a few years ago on the case of an individual farmer
whose claimed holding included part of the North Sea). And there
have been valuable joint audits involving the Court and SAIs,
for example a parallel audit on the structural funds carried out
in 2005-06 (whose findings were in line with the Court's previous
observations on the funds).
As for the potential contribution which "national
declarations", perhaps audited by SAIs, can make to the management
of EU expenditure and to the DAS, the Court would wish to underline
the following points.
— Declarations that expenditure has
been properly and legally undertaken are a normal feature of systems
of financial management. Such arrangements already exist in a
number of contexts within the EU budget; the Court has commented
from time to time on their appropriate form and substance.
— It is not clear at the moment how
some of the ideas for global national declarations would fit together
with and add to the effectiveness of these sector-specific declarations.
For example there is current discussion of amending the Financial
Regulation in order to require Member States to produce a national
summary of sector-specific declarations relevant to EU expenditure
within their territory. It is not clear how in itself this change
would directly improve financial control.
— The Court has to carry out the DAS
on an annual timetable (specified in Community law) and in order
to answer the DAS questions in the Treaty. The Court can, does
and will take account of professionally conducted audits of EU
spending carried out on the appropriate timetable and which address
those questions—but the qualifications are important.
CONCLUSION
The Committee's report at a number of points
compares and contrasts EU and UK public spending and the Court
and the UK National Audit Office (NAO). I understand that the
NAO has repeatedly qualified the accounts of the Department of
Works and Pensions (and its predecessor bodies) which has a budget
of £124 billion, and where most of the expenditure concerns
transfer payments to benefit claimants. The NAO quotes an error
rate of 2.2 per cent, which it judges to be substantial and of
serious concern. In the case of agriculture expenditure (outside
IACS), structural measures, internal policy and external aid the
audits of the Court have identified levels of irregular or illegal
expenditure which are significantly higher than that. These results
are derived on the basis of similar statistical techniques to
those applied by the NAO and other European SAIs. They are not
one off results but, rather, have been consistent over a number
of years. There is sufficient and compelling evidence that public
funds are not being spent according to the regulations and the
intentions expressed by the legislator.
My colleagues and I trust that these comments
will be of assistance to the Committee in any further discussion
of the subject matter of its report.
15 December 2006
| 
