By Alexander Budzier, March 15, 2015
The biggest project risks: Not what you might expect
Since 2009, my colleague, Bent Flyvbjerg, and I have examined more than 6,000 large scale projects (92% in the public sector and 71% focused on public sector IT change initiatives), comparing budgets and benefits estimates with actual results. The results have been surprising – and alarming.
When the figures were fully broken down and not simply averaged out, a staggering one in six projects ended up in the “fat tail” of budget overruns of more than 50% and schedule overruns of more than 70%. These statistics question previous studies that focus on average performance and talk of overruns as excusable error. Our data point to a picture where projects are generating a disproportionate – and disastrous - number of very large overruns that must be addressed.
Risk drivers: Are we getting it right?
Often, the size and duration of a project is considered the primary driver of risk. In the UK, it’s the largest 400 projects that sit on the Major Projects Authority register and are closely scrutinised. But our findings prompt other key risk considerations. Is the organisation strong enough to absorb the hit if budget overruns turn out to be extraordinary? Can the organisation still cope and deliver if 15% of its medium-sized projects exceed cost estimates by 200%? These numbers may seem unlikely, but 5 years of research has shown that they are frequent and realistic. Any publicly accountable organisation needs to know what its specific risk in the “fat tail” is.
Then we must consider the robustness and availability of data. If public sector executives don’t have full confidence in their data, that seriously impacts on their ability to make decisions. Particularly during the early project life cycle little is definitively known about the project because, by definition, the project design hasn’t reached maturity. But decisions that will lock the project into a course of action must still be made. We must be mindful of such shortcomings in the data and how to navigate them when examining risk and make decisions.
Can public transparency drive better risk management?
Of course, in the US, transparent publication of project details is de rigeur. Should we be striving for the same in the UK? And how would that new level of accountability drive behaviours and decision making in the public sector?
We’ve already made a move to de-risking projects through optimism bias uplifts, mandatory for big public projects in the UK. The MPA is moving towards more openness. Taking the principle of project comparison to the point of total transparency, though, carries inherent risks. The biggest of these is that you inevitably change the accountability landscape. That will influence behaviours – sometimes with negative or unintended consequences. We have studied in detail the beneficial and adverse consequences the move towards transparency had in the US.
Transparent public reporting also relies on robust, high quality data (which, as we have seen, can be problematic) and can involve a significant administrative burden. That creates a high level of resource consumption which could arguably be better deployed to maximise the benefits of the project. What’s more, an admission of failure in the public space can destroy the credibility of a project – and of other associated projects, potentially affecting investment, delivery and outcomes.
On the other hand, if data is made available across projects in a more consistent manner then that data can be used pro-actively. As well as reporting accurately, we can also start to map early warning indicators to identify high risk areas. That in turn will accelerate risk issues into the governance structure and alert people higher up in the hierarchy to possible areas of concern. Total transparency potentially equips us to widen the view from individual projects to the entire portfolio, prompting better decisions, innovation of project delivery and better outcomes.
We tend to assume that transparency is inherently a “good” thing. Certainly, one cannot address the risks on a project if they are not visible. What is undoubtedly important, given the results of our research, is that we engage in a new conversation: one that brings risk-related transparency to the table and honestly recognises the benefits and potential drawbacks of a total transparency approach.