Governments in Canada - and around the world - have a very difficult time with large-scale information technology projects.
This is not the cronyism and mendaciousness that led to the Toronto c omputer leasing scandal.
Rather it is the challenge of administering massive transformation of the information systems within government.
Examples are legion.
In the United States, the Veterans Affairs department is under fire for spending more than $120-million over eight years without producing usable software.
Federally, the RCMP spent more than $100-million to implement two computer systems that can't speak to each other , one in British Columbia and one in the rest of the country.
The provincial government in Ontario recently had to ensure months of criticism for improper procurement in its eHealth initiative, which is building hundreds of thousands of electronic health records.
Each of these projects endured major challenges, but what lessons can be learned from past cases to inform how IT should be managed in the future?
The two cases that will be examined in detail are eHealth and the Integrated Justice Project.
Integrated Justice (or IJP) was a shared responsibility project between the civil service and a private-sector contractor. An initiative of the Progressive Conservative government of Mike Harris, the goal was to introduce electronic records to the justice system, reducing the vast armies of court reporters and clerks required to manage the 19th century paper-based records of Ontario's courts.
The plan was in essence for the private-sector partner to be paid only when the benefits were realized. The challenge was that the project management was kept completely in the hands of the civil service. Politicians were wary of involving themselves in the process, for fear of being accused of a conflict of interest. The lack of supervision from elected officials produced an environment of diffused responsibility, where ownership of the file changed at several times and was split between two Ministries. In the end, decisions were at best delayed, and at worst avoided, and the very changes needed to produce savings (cutting full-time positions, etc.) were never acted upon.
According to the province's auditor, the result was more than $250-million spent by the government and the private-sector partner to realize fewer than $10-million in actual "benefits."
The private sector partner sued the government for damages, stating the civil service failed to make the decisions needed to achieve success.
The Achilles heel of IJP was the lack of ownership of the project and little political oversight of the civil service, an emphasis on process over results, and diffused responsibility.
In response to this and other IT project failures, the province commissioned a task force that recommended a more robust oversight process, including quarterly reporting to Management Board and a "gateway review process" to force projects into manageable, sequential chunks.
While the new methodology kept new projects on the rails, the Ministry of Health's Smart Systems for Health - started by the PC government under Ernie Eves - was already well underway at this point. The same process challenges hampered SSH so badly that the government finally pulled the plug and went to a new model.
In many ways, eHealth was an overreaction to the experience on IJP and other process failures.
Rather than hand control over to the civil service, the government created an independent agency that reported directly to the minister. The agency was staffed primarily by external people who fell outside the traditional hierarchy of the civil service. The intent was to avoid the stultifying process-orientation and diffused responsibility that bogged down past projects. eHealth would drive pell-mell for results.
Unfortunately, the intense drive for results compromised process so much that consulting firms were overpaid, open competition was avoided, and expense accounts were approved without thought.
By the time the eHealth machinery had run for a few months, it was already too late. Expensed choco-bites and thousand-dollar-an-hour fees were in the past.
Unlike IJP, eHealth continues to deliver tangible results for the money spent. Just last week, another 10,000 patients had their records digitized, meaning fewer inaccurate prescriptions, digital x-rays and faster test results.
But compared to the relatively minor hit the Conservatives took for IJP, the digestible process stories in the eHealth drama were far more damaging to the McGuinty government because they were on a scale people could understand.
The lesson of both IJP and eHealth is government is not all that great at large scale projects. Finding the balance between process and results is hard, and the risk of failure is always present.
It is time to change the approach to IT projects and examine the benefits of complete outsourcing at a fixed price.
In this model, the government and the service provider would agree on the policy outcomes desired, the basic methodology that would be used, and the measurements that would determine success. A fair price is negotiated and 100 per cent of the risk is transferred to the company. If they fail to deliver what was promised, they don't get paid. But with that risk also goes control. The private sector provider enjoys responsibility for design and the freedom to decide what route to take.
IJP shows us that too much process will kill a project. eHealth shows us that too little process will also hurt the government.
The best thing might be for government to hire people who know what they are doing, negotiate a fair price, and pay only when a successful installation is completed.
Update eHealth Ontario has been quietly moving in the suggested direction. eHealth is working hand in glove with Infrastructure Ontario on alternative methods of procuring IT, including a "no money until satisfied" clause on the recent RFP for the diabetes registry.
(Editorial cartoon by Brian Gable/The Globe and Mail)