Accountability and Leadership in IT
An interview with Denis Desautels

by Paul Crookall

DENIS DESAUTELS WAS THE AUDITOR GENERAL OF CANADA FROM 1991 TO 2001. HE RECENTLY CHAIRED A SPECIAL TASK FORCE ON THE MANAGEMENT OF LARGE-SCALE INFORMATION AND INFORMATION TECHNOLOGY PROJECTS FOR THE ONTARIO GOVERNMENT. HE SPOKE WITH US ABOUT ACCOUNTABILITY AND THE MANAGEMENT OF IT.

Accountability seems to be an unreachable star in the public service. We talk about it a lot. Elsewhere there is less talk but it is more a reality. I did not have the ultimate answer in 1991 when I began as Auditor General, and I still don't. But I had a vision.

I knew the principle goal was to make accountability more of a reality, to emphasize accountability for results over process. Ole Ingstrup challenged me on that. He asked, "What do you mean by that? What are the implications?" One of the things Ole did, as head of CCMD (now the Canada School of Public Service) was organize a dinner with several senior bureaucrats to chew the fat, to have me on the hot seat. It was a very useful exercise; it forced me to think through the concept. It turns out accountability for results has gained acceptance.

To achieve accountability, we need to be more realistic. Accountability is not the same as blame. It is an obligation to render an accounting of our stewardship given our resources and mandate. Assuming we all agree on the need to take results into consideration along with compliance with processes, then you have to ask whose accountability we are talking about – we tend to confuse personal and organizational accountability. In government, we should talk more about the latter. More programs are now being delivered in partnerships among organizations, resulting in joint accountability. This makes some people nervous.

People keep asking, "Who's to blame?" But people don't hang around long enough and turnover is so high it is virtually impossible to hold one person responsible. There is limited continuity from start to finish. Deputy ministers are responding to recommendations from reports and inquiries into actions
that were taken before their watch. Things move slowly. It is difficult to ‘pin’ the responsibility on individuals, so I emphasize organizational responsibility.

Personal accountability, as we are finding in the Gomery Inquiry, is still needed. But we also need a higher level of organizational accountability to our mandates, our mission. Individual actions tend to be more short term. Consider a private sector analogy: If performance results are not there, the stock price shows it, the market makes a judgment, the organization is penalized; in the public service, if we fail to deliver on objectives, maybe the organization needs to pay a certain price, which could be a reduction in resources it might otherwise receive. There is always competition among departments for resources. Performance results should be one criterion in making these decisions. Accountability is an elusive concept, yet fundamental to our business; we need to keep pursuing it.

MANY OF OUR READERS ARE NOT IT EXPERTS – SOME ARE EVEN INTIMIDATED BY THE PROCESS. WHAT ADVICE DO YOU HAVE FOR MANAGERS WHO MUST DEAL WITH IT PROJECTS?

First, senior executives and the minister need to recognize that large IT projects are often a business transformation as well, and calling it an IT project is the first mistake. By raising it to the level of a business transformation you give the project the same direction and attention and accountability for delivery as other major business transformations, and impose accountability on the IT part as well. Recognize the project as transformation and treat it as such.

Second, take into account the capacity of the organization to undertake a project of this size at this time. Banks do this well – they have a quota on how much change will be attempted at once, and spread it out over time. They will not have too many projects underway simultaneously.

Third, ensure the project can be done in a reasonable, measurable, limited amount of time. This is more than just putting in ‘gates’ to assess and give a go/no go decision. Most organizations should undertake large projects that can be done in one or two years at the most. If longer, it loses impetus. This is another capacity issue. If needed, define the larger project into separable projects each with a one- to two-year time frame. Projects over two years suffer partly because of the churn in leadership. It is difficult to keep and track ownership, which makes it more difficult to track accountability.

Fourth, project management is important as a professional discipline within government. Recognize that, build project management skills, and apply project management to IT projects, as well as other projects. Keep the team intact over the development and implementation period – one benefit is that people will strive to complete the project successfully, with attention to results/deliverables, without worrying about where they are going next or if they will be ‘traded’ off the team. Give them a commitment to the management of the project as a professional activity and with a promise that another good assignment will follow completion.

It is important to be committed to the project, to have the professionalism of commitment. In Quebec, they seem to do this well – to be able to look back and have the pride of accomplishment. Invest more money and time in project planning before approval and funding. For example, instead of spending 10% of the budget on planning before approval of the full project, we should be spending 20%. IT projects are different from construction, for example, in that more planning is needed. Invest more effort in assessing projects – some will be rejected on further inquiry for not meeting the four principles.

The result will be approval of fewer projects that may not achieve their objectives. Take more of a portfolio approach. The DM or CEO should not be making decisions one at a time, but as part of an overall portfolio consistent with the organization's objectives. The portfolio approach is superior to backing and bankrolling only one big project or investing only in the project that has an early lead.

Take an organizational approach; the whole management team, not just the chief information officer, should be involved. Procurement is a tough nut to crack; there is no perfect solution. But we can do each of the pieces better. Plan, manage the project, have shorter do-able timeframes and a committed team.

Build in off-ramps. During the life of a major IT project, needs may change; you may find the capacity is not there. Be ready to modify or put an end to the project. Separate design from build. BC seems to be having initial success with this approach.

Have a mechanism for better dialogue between the IT industry and government executives. The IT industry now seems interested in a more consistent ongoing dialogue with government. On the government side, we need people who can maintain that dialogue, who have the interest and the time. Deputy ministers are pulled in many different directions; they need to define who should represent government in the dialogue with the IT industry. It could be the CIO, or an ADM, or central agencies like Public Works and Treasury Board. The whole idea is not to have a social club but to have an agenda at the core of which is effective leadership and implementation of technology.

We need to benchmark others. Although our report was for Ontario, we benefited from a study done by KPMG on leading practices across the country and internationally, and research on recent OPS IT projects by Jean Pierre Soubliere and Solly Patrontash.

WHAT CAN A NON-IT MANAGER DO TO BETTER MANAGE IT PROJECTS?

Think about it, reframe your approach, consider and apply these four principles. Realize that managing big IT projects that are business transformation issues can no longer be delegated and ignored. Keep in mind that the chances of success are better if you think modular – smaller projects building on each other are more likely to get funding and more likely to succeed.

Learn more about the IT business, not to become an IT expert, but to learn how to manage these projects better. Ensure you get the right resources and support – the budget and the up-the-line commitment that this project is good for the organization. Then get the right project management team working for you. The process is like driving a barge. As you go along, people keep dumping extra expectations on you, so be careful to manage the scope or it will become unmanageable. Stay involved without micromanaging. Full ownership rests with you.



BIO: Denis Desautels is executive-in-residence at the University of Ottawa, and a member of the board of CARE Canada, Bombardier, Alcan, Jean Coutu and the Laurentian Bank. He is a member of the Parliamentary Precinct Oversight Advisory Committee. Honours include the Order of Canada and two honourary doctorates.