Conclusions and recommendations
1. The
Department and the Rural Payments Agency failed to implement the
single payment scheme effectively. By the end of March 2006 it
had paid farmers only 15% of the £1,515 million due, compared
with its target of 96%, causing significant hardship. Taxpayers
will have to pay additional implementation costs. In addition
to a provision of £131 million included in the Department's
accounts for 2005-06, the Department has had to secure a supplementary
estimate of £305 million to meet the potential cost of disallowance
of expenditure by the European Commission arising on the 2005
and 2006 single payment schemes and the previous schemes administered
by the Agency.
2. At the end of October
2006, some 3,000 cases for the 2005 scheme remained to be settled.
The Agency subsequently managed to progress some of these outstanding
payments, but 911 claimants had yet to receive anything and 2,184
claimants were awaiting a final 'top-up' payment by the time payments
started to be made under the 2006 scheme on 1 December 2006. By
May 2007 there were 24 claims, mostly probate cases, which remained
unpaid, but the Agency was still reviewing the accuracy of a substantial
number of claims already processed and making adjustments both
for over and under payment.
3. There are a number
of lessons to be learned by the Agency, the Department and government
bodies more widely.
| |
|
|
| |
i. |
The scheme is small, covering only some
116,000 claimants, but the Department made it unnecessarily complex
by choosing to adopt the most demanding implementation options.
It selected the 'dynamic hybrid' option for calculating entitlement,
a one year implementation timescale, and no de minimis threshold
for claims. Scheme parameters should not be chosen in isolation,
but with due regard to the overall complexity and risk they will
jointly present. |
| |
ii. |
Because the government sought to implement
the single payment scheme at the same time as a wider business
change initiative, the Agency shed too many of its experienced
staff and their knowledge at a time when it needed them most.
It then spent some £14.3 million on agency staff in 2005-06
to process 2005 single payment scheme claims. Before combining
projects, their interdependency and the potential for compounding
risk should be assessed as well as the risks of the individual
projects. |
| |
iii |
Implementation of the project started
before the specification of the single payment scheme was finalised.
The aim was to meet the March 2006 payment
deadline, but the result was that the Agency had to make assumptions
on what the final regulations from the European Commission would
contain. It subsequently had to make 23 substantial changes to
its computer systems to reflect policy and regulatory revisions.
The risk of having to make changes later in the development of
the scheme could have been given more weight in determining the
implementation timetable at the outset. |
| |
iv |
Processing capacity had not taken sufficient
account of the number of maps and mapping changes that would need
to be processed. The scheme was based
on land area managed, and incentivised farmers and new claimants
to register additional land. A proper estimate of the scale of
the work should be made by appropriate modelling and testing. |
| |
v |
The Agency tested each key element
of the IT scheme before introduction but testing in isolation
did not fully simulate the real world environment and problems
emerged later. Failure to test computer
systems completely and adequately is a problem we have often seen
with government IT projects. Time should be built in to test the
IT systems as a whole as well as the individual components within
it to obtain adequate assurance that components are fully compatible
and deliver the required business process. |
| |
vi |
Without an individual or small team processing
a whole claim end to end, claimants found it difficult to obtain
advice and information on the status of their claim and Agency
staff were hampered in their attempts to resolve claimants' queries.
The Agency had instead decided to adopt a task based design for
claims processing to enable staff in different offices to work
on any tasks relating to any claim, but it did not adequately
consider the customer interest in following their claims through
the process and the consequent impact of the new way of working
on customer service. The development of new business processes
should take the customers' requirements into account in the design
of the proposed system and any potential contingency arrangements. |
| |
vii |
A lack of information was the principal
cause of frustration and complaint within the farming community.
Automated telephone lines provided unhelpful responses such as
"there is nothing that the call centre staff can tell you
about your payment". Farmers were discouraged from pursuing
queries by being told that "If you contact us, this will
divert resources away from the urgent tasks of completing validations
and making full payments". A communications strategy should
be developed which keeps all concerned but particularly customers
in touch. |
| |
viii |
The Agency could not easily determine
how much work remained outstanding on claims each week and how
long it would take to complete them. The
Agency had deferred development of software to draw out key information
on the progress of each claim to focus resources on other parts
of the system it considered to be critical. Those with oversight
of the project thus found it difficult to distinguish between
real progress and inherent optimism within the project team. Specific
measures should be developed from the outset to enable implementation
progress to be assessed objectively, and make sure management
information systems enable appropriate data to be tracked. |
| |
ix |
The Agency mothballed one contingency system
on the basis it would have experienced the same data accuracy
problems as the main system, although it would have allowed processing
on a claim by claim rather than a task by task basis.
The Agency also decided not to invoke partial payments available
from the end of January 2006 because it expected to make full
payments in March 2006, but in the end was unable to do so. Genuine
and workable contingency arrangements commensurate with the project
profile and risk need to be factored into the business case and
developed from the outset. |
| |
x |
The Department did not recommend specifying
a minimum claim size, unlike Germany which specified a minimum
claim of 100 Euros (around £68).
Adopting a similar approach would have reduced the number of claims
by almost 14,000 (12%) and saved administration costs which may
well have exceeded the sums claimed. In designing processes and
in supporting documentation such as application forms, cost effective
opportunities to simplify should be identified and implemented. |
| |
xi |
The Department and Agency established separate
boards to provide technical programme management and critical
challenge but there was a lack of clarity as to which Board or
individual was ultimately responsible for decisions.
The challenge board took a greater role in decision making as
the project proceeded, blurring its scrutiny role. The departmental
Permanent Secretary at the time, Sir Brian Bender, bears responsibility
for administrative failure leading to additional costs that together
risk exceeding £400 million. There needs to be a clear distinction
within project governance structures between those responsible
for oversight and challenge and those managing the decision making
process, even when a project reaches a crisis point. |
| |
xii |
The Agency's management team recognised
the risks to delivery of the project, but the Chief Executive,
Johnston McNeill, felt unable to show that it could not be delivered.
At issue, however, was not just the feasibility
of the project, but the acceptability of the risks, which were
acknowledged to be high. If Accounting Officers believe that their
assessment of risk is being discounted, the proper course of action
is to seek a direction from the departmental Accounting Officer
or Minister concerned as to whether they should proceed. |
| |
xiii |
The structures originally set up to oversee
the project included two Senior Responsible Owners, one for policy
(in the Department) and one for implementation (in the Agency).
Splitting the role of Senior Responsible Owner was bad practice, and undermined the Department's ability to challenge the Agency's progress reports. Every project should
have one Senior Responsible Owner so that lines of accountability
and responsibility are clear. The Department should agree progress
milestones with the Senior Responsible Owner, whose pay and performance
bonuses should be directly linked to performance objectives and
programme delivery. Every project should have objective targets
and progress data so that any corrective action can be triggered
quickly. |
| |
xiv |
The implementation of the single payment
scheme was subject to four Office of Government Commerce Gateway
Reviews between May 2004 and February 2006, and three of these
Reviews assessed the programme as "red". Development
work on the computer system nevertheless continued and no contingency
plan was invoked, despite limited confidence that the system would
be ready on time. If 'red' reviews are to be taken seriously,
departments need to be explicit about the circumstances in which
they would lead to fundamental review or termination of a project. |
| |
xv |
We are disappointed that it took the present
Accounting Officer, Mrs Helen Ghosh, nine months to provide the
Committee with a full account (in the form requested) of the total
cost to public funds of removing Mr Johnston McNeill from office
as Chief Executive of the Rural Payments Agency and from employment
as a civil servant. We expect Accounting
Officers to show better co-operation with such requests by the
Committee. |
|
