Volume 17, No. 2, August 2007
E-Government Failure and Political Timescales
Richard Heeks
richard.heeks@manchester.ac.uk
The question should not be "why do e-government projects fail?" Instead, it should be "why do e-government projects still fail when we've known for so long why they fail?" Let's take a massive example - quite possibly the world's largest e-government failure - the £12.4bn (c.US$24.5bn) spent on the UK National Health Service's (NHS) National Programme for IT. This is not a total failure; it's a partial failure - a project in which major goals remain unattained. Second, success/failure categorizations are never static - three or five or ten years later, this programme may be deemed a success.
|
The
NHS was set up in 1948 and is responsible for hospital services,
family practitioner services (doctors, pharmacists, opticians and
dentists) and community-based services. It has been recognized as
one of the best health services in the world by the World Health
Organization. In order to cope with the demands of the 21st
century, the NHS has consistently been making changes in its
structure; the way in which its different organizations relate to
each other; the way services are delivered; and in the way
patients are treated. But it was only with the publication of the
10-year NHS Plan in July 2000 that a full-scale modernization
programme was launched to transform the NHS. Looking
at the enormous improvements that can be brought about through ICT,
the NHS has tried to bring about a revolution in the health
service industry by delivering the following services with the
help of the information technology: -
Lifelong Electronic Health Records for every person in the country
- Online
access to patient records and support on best clinical practice
for all NHS clinicians -
Integrated care for patients through GPs, hospitals and community
services sharing information across the - Fast
and convenient public access to information, advice and care
through online information services such as NHS Direct and NHS
Direct Online. In 2003,
NHS entered into a 10-year agreement with Accenture
to provide a centralized database of medical records. But the £10
billion (€14.9 billion; US$18.9 billion) venture suffered delays
and cost overruns and the British government was urged to rethink
the plan after it emerged that there had been 110 glitches in the
last four months of 2006. Consequently, the NHS awarded a
nine-year contract worth about £1.97 billion (US$3.73 billion,
€2.91 billion) to Computer
Sciences Corporation, which entailed taking over the project
from Accenture. Source:
Website of National Health Services |
Nonetheless, at the time of writing, "failure" does appear an appropriate label. Results to date include only 10% of hospital administration and 0% of clinical record systems installed against target; hospital booking systems that have increased costs and referral times; and loss of patient data including child vaccination records.
So, why has this e-government project so far failed? A comprehensive review provides some answers[1].
-
Contracts were rushed: "The crucial contracts were advertised, bid for and signed at breakneck speed."
-
True requirements were not established: The center-piece of the programme is a nation-wide electronic patient record system yet "the government had no evidence that centrally held health records accessible anywhere in the country were needed."
-
End users were not properly consulted: Consultation processes were described by one insider as "a sham", and the implementation process was dominated by "IT whizzkids".
-
Software design did not match user realities: Suppliers made use of software systems designed for US healthcare: "These were built around billing for care and were unable to produce the data on which NHS management depends."
-
An aggressive and secretive management approach was adopted: Led by an IT director said to "talk in management-speak with added menace", with transparency replaced by "news management, avoiding awkward questions in favour of pumping out their own misleading information".
-
A heavily-centralized management approach was adopted: forcing hospitals and other health units to accept strongly-standardized solutions from monopoly suppliers which did not fit their real needs.
Give every e-government manager the back of an envelope and a pen. If asked to produce a list of bad practice to avoid, they could surely have produced most or all of the above without much prompting. Proper requirements analysis, user participation, fit between e-government design and user realities, a hybrid approach steering between centralised and decentralised - all of these have been recognised good practice for years[2] - in fact, for decades.
So why does this good practice keep being ignored on e-government projects? One answer must be: political timescales. In particular, the determination of elected ministers to see results either before the next election and/or before they are shunted to their next department. Thus it was with the NHS project. Conceived in 2002, it was pushed - quite unrealistically - to deliver before the next election.
From this flow all the departures from best practice - that massive contracts were put in place in less than one year compared to an industry average of 27 months; that there was no time to properly investigate requirements or user views; that immediately available albeit inappropriate imported software was used instead of home-grown varieties; that bad news or problems that might delay or divert the project were ignored; that hospitals were offered "any colour you like so long as it's black" because choice and attention to local needs and realities all take time. And time is one resource that most big e-government projects simply do not have.
Can we disconnect e-government projects from political timescales in order to allow for best practice? We probably couldn't insist that all truly major e-government projects had an initial timescale spanning at least one election. But we could try to depoliticise large projects by getting cross-party commitment and ongoing engagement. We could adhere to KISS - keeping e-government small and simple so that, even with the inevitable delays, it falls within political timescales. We could break large projects down into smaller incremental elements that, again, would fit inside the timescales.
But these ideas, too, are nothing new. Long before "e-government" was a twinkle in the eye, we were warned that public sector IT projects tended to be too large, too politicized, and with no financial bottom line[3]. Nothing much has changed today. The imperatives that attract politicians to big, risky projects which they demand must deliver by yesterday are written deep into the body politic. The NHS e-government failure may thus be the largest to date, but it will not be the last.
[1] Brooks, R. (2007). System failure! Private Eye, 1179, 17-24.
[2]
Heeks, R. B. (2006). Implementing and Managing Government: An
International Text.
[3] Margetts, H. and Willcocks, L. (1993). Information technology in public service: disaster faster? Public Money and Management, 13(2), 49-56
This article has been abstracted from a short paper by the author. The
original paper was published in April 2007 and is available on the website
of the Institute for Development Policy and Management (IDPM), University
of Manchester at
http://www.sed.manchester.ac.uk/idpm/research/publications/wp/igovernment/