News today that the UK Government has “put the brakes on” Network Rail’s modernisation plan resonates with my experience of major projects. As a project planner myself, I was frequently amazed by disingenuous declarations of surprise whenever a major delay or cost overrun was announced. How stupid do the politicians think the Public are? Of course, such problems will have been known for a long time.
As the great Fred Brooks (author of the seminal The Mythical Man Month) asked:
How does a project get to be a year late?. . . . One day at a time.
Brooks calls this “hatching a catastrophe” (Chapter 14) and he cites Sophocles:
None love the bearer of bad news.
No doubt there were sensible people at Network Rail, warning of the difficulties, who were told to “just shut up, you damn fool”. So problems festered, until it became impossible to conceal them any longer.
A. Canova, “Ercole e lica,” 1802. Hercules hurls to his death the messenger Lycas, who innocently brought the death-garment.
Thus it is now reported that the Great Western project is running a year late and its cost estimate has tripled from £640m to £1.7bn. Is anyone still gullible enough to believe that new estimates for time and cost have a greater likelihood of being achieved?
As the Financial Times points out:
Network Rail has come under increased scrutiny since its £38bn debt was transferred to the government’s balance sheet last year as a result of European Union pressure.
Michael Dugher, shadow transport secretary, accused ministers of sitting on a report last September which had suggested that the projects needed to be delayed. “They have pretended to the public that everything was fine until after the election,” the Labour MP said.
Government fiddling the books and suppressing the facts. Unsafe to talk. Nothing new.
I picked the subject of cost and time overruns of major projects for my MIT Master’s thesis Owning Up or Covering Up: Motivation for Meaningful Project Audit. My depressing finding was that there are few incentives for incisive reviews of the lessons learned from major projects gone wrong. On the contrary, in almost every case, there is huge political pressure for any kind of audit to be a whitewash: that is, to cover up not own up.
Right now, in the UK, we need look no further than the repeated delays to the Chilcot Inquiry into the Iraq War for an example of exactly the same phenomenon happening today. In 2009, David Cameron was quite right to call it an “an establishment stitch-up“… it still is.
The reasons we tend to underestimate the time and cost of our projects are the meat and drink of popular TV shows like Grand Designs and Homes Under the Hammer. The difference is that while small scale building and renovation projects like these tend to overrun by 20 or 30 percent, major projects like Network Rail’s tend to overrun by 200 or 300 percent, or more. You would think that when there is so much at stake project managers would be more thorough and therefore more accurate in their estimates.
My conclusion was that there are three kinds of bias that creep into such estimates, and that technical or “methodological” causes of bias are the least significant.
Sure, there are methodological reasons we screw up time estimates. Fred Brooks alludes to a key one of these in the title of his book: the” terrible simplification” that time and resources are interchangeable. This might be true of straightforward tasks, but some jobs are indivisible. Two women can’t “do pregnancy” in four and a half months, he points out. Adding more resources to complex tasks actually makes them take longer as people fall over each other, spending more time on necessary interactions to coordinate their activities than in doing the work.
Pieter Bruegel the Elder – The Tower of Babel (Vienna)
Another frequent source of “methodological bias” is misuse of Project Network Analysis. This “model” of the time schedule suggests that the overall duration of a project will be determined by its longest sequence of activities: its “critical path”. While this may be the case for certain kinds of projects, like pipelines and high-rise buildings, the duration of most projects is determined by too many (individually non-critical) activities left too late: a “bow wave effect”.
But a good planner understands such pitfalls and compensates for them by building contingency into their schedule. The next two sources of inaccuracy in project time and cost estimates – managerial and political – are far more insidious and difficult to compensate for.
Managerial bias occurs wherever project management start to trade “realism” for “reach”. That is, when they start to say things like:
Yes, we know you estimate that the job will take three years but in our experience these types of jobs usually run, say, six month’s late. So we are going to set ourselves a “stretch target” of completion in two-and-a-half years and that way we might actually get it done in three years not three-and-a-half years.
This would be fine if project management always remembered its manipulation of the time estimate. But it is easy to delude ourselves, and all too often two-and-a-half years becomes the new planned completion date and three years (which was always how long the job would take) becomes a six-month delay.
I am willing to bet Network Rail’s planners initial schedules were far more realistic than their managerially-adjusted versions.
Finally, the schedules of major projects are warped into pure fantasy by the baleful influence of politicians, “positioning” estimates to “manage stakeholders”. This was brought home to me when I worked for Westinghouse Nuclear Europe in Brussels.
Back then, Nuclear Power Stations took ten to twelve years to build. This was easy to demonstrate because Westinghouse’s Nuclear Steam Supply System projects were fairly standard and key project milestones for more than a hundred such plant were tabulated and updated monthly in the trade press (Nuclear Engineering International).
Nevertheless, “political bias” meant that Westinghouse always quoted their duration as six years. When I asked about this, I was told that it was as long as any client government was prepared to commit for. This might be OK if everyone tacitly conceded the deception. But Westinghouse manufactured its expensive reactor vessels as if this six-year timescale was a reality, delivered them to site and proceeded to charge their clients an arm and a leg in demurrage for the vessel and its skip. A nice little earner for Westinghouse, a rip-off for the client government and its taxpayers.
In the current Network Rail case, you can be sure that unconscionable corporate interests will have found ways to profit hugely from the cowardice and muddle-headedness of the UK’s politicians.
Yet again, we will not deal with such issues until we make it safe to talk.