The collective memory of some departments is short. Few keep records of IT failures that pre-date the general election of 1997.
So officials at the Department for Work and Pensions say they cannot recollect the aborted "Camelot" system in the 1970s to computerise welfare benefits, or its successor, the £2.6bn "Operational Strategy" in the 1980s, which was set up to give people information on all their entitlements "at the touch of a button". Opstrat, as it was called, cost three times the original estimate and the DWP has yet to integrate its welfare benefit systems.
Indeed the Department is working on its "third attempt" at integrating welfare IT systems, said the National Audit Office in a report earlier this year.
Below are excerpts from a report of the Committee of Public Accounts, 21 June 1984.
The fact that the PAC published the report 25 years ago, and has been saying much the same thing in numerous reports on IT-based change projects ever since, suggests to me that the assumption of a happy ending and a concomitant underestimation of complexity, potential problems, costs and risks is congenital to large IT projects within central government.
The 1984 PAC report was deftly entitled: "Management and Control of the development of Administrative Computing in Government Departments".
It said:
"In February 1984 the C&AG [Comptroller and Auditor General - head of the National Audit Office] reported the results of a National Audit Office examination of a selection of computer projects in four departments which had encountered serious difficulties ... the C&AG identified a number of common weaknesses and concluded that across government departments as a whole much more needed to be done to improve:
strategic planning
project management
system design and development"
There was a need for "early warning of delays and difficulties" so that timely action could be taken.
These were the things that brought down the £513m C-Nomis project for prisons, and the IT £350m IT project which supported the Rural Payments Agency's Single Payments Scheme.
On those projects there were indeed warnings - but not as a basis for timely action.
The 1984 PAC report had an innovative idea which, perhaps, was too innovative to be taken seriously and hasn't been adopted even today: moving talented IT staff between departments, to where they were most needed, and rewarding them accordingly .
The 1984 report also lamented a lack of departmental action on its previous findings (the very point made by the committee in its report last month on the C-Nomis project). Other excerpts, which are relevant today, from the 1984 PAC report:
- "We were surprised that over six years only eight of the 22 major computer-using departments had completed acceptable baseline plans for their computer development programmes..."
- "[Departments should] concentrate their existing resources on those projects which offer the highest benefits and, if need be, cancel or defer lower priority schemes."
- "We noted the dangers of general over-optimism about the benefits and timescales of computer projects." [my italics]
Links:
Something must be done to break the cycle of IT failure - The Times (opinion piece by me)
Are briefings to the minister on the NPfIT too optimistic? - IT Projects Blog
Anatomy of an IT disaster - IT Projects Blog
10 [US] corporate IT failures - Computerworld
