APM responds to NAO report on Crossrail

7 comments

Infrastructure projects are notorious for escalating costs and timescales, and I know many committees have angsted about why this happens. Too many people have personal gain from the vast sums of money being spent, and optimism bias (a polite way of saying "economical with the truth") obfuscates the real problems and real drivers. What if we change the focus? What if, instead of a focus on WHAT is going to be done, the programme focuses on WHY, ie the benefits and value to be realized, and make decisions to optimise benefits? Would it work? All project management involves hitting obstacles and unexpected changes, and making decisions to keep the project on track. At the moment, a project staggers to the next milestone within tolerance (or, more usually, to change the milestone so that the milestone is where the project is going to end up - "shoot first, then name your target"). If the next milestone were a benefit milestone (we will make progress towards something that logically means we're on target to deliver benefits - ie a target within the timescale of project delivery - a proxy or lead measure), then that frees up the project leadership team to change the specification on the fly, within their competence, and the sponsor board can hold them to account to deliver benefits. Crossrail - it's going to benefit some people a lot. In particular, it's going to benefit some landowners with rapidly escalating value of land and properties which are near to Crossrail stations. Commuters will be spared a bit of misery. Of course some questions have reared their ugly heads about how much people want to commute once the Covid-19 lock-down is over, but there are obvious and quantifiable benefits. So the whole project could be managed from a benefits perspective. Since "what gets measured, gets done" the results would probably be very different.

One significant failing in Crossrail was the Assurance. Why were the problems the project faced not highlighted by the Assurance regime? Perhaps they were, but Assurance was ignored? Quite a bit has been said about this by various reviews, and I'm sure there was no single Assurance failing; rather, I suspect it was to do with behaviours, which is a common downfall of Assurance. Was Assurance not extensive enough? Was it not looking at the right things? Were Assurance outcomes not communicated effectively? Was Assurance simply ignored as being 'inconvenient'? Is there a tendency for Assurance to be only seen as important when the decision-makers are worried, and when they're no longer worried Assurance is dropped? The way to stop that is to ensure the voice of Assurance is at least as loud as the project manager's.

Key findings of the January 2018 report include: 1. The Elizabeth line will help reinforce London’s status as a global city, it has created the capacity for major HQs including Facebook, Deutsche Bank, Capita and Societe General, allowing for the accommodation of 300,000 new jobs in key employment hubs Outside central London, town centres including Ealing, Woolwich, Ilford and Romford are being rejuvenated as the construction of new homes and offices attract more businesses and residents 2. 90,599 new homes along the route are predicted by 2021 and 180,000 by 2026 – far greater than the 57,000 new homes (by 2021) predicted in previous report; 3. Over 4.4 million square feet of commercial office and retail space expected along the route by 2021 – an increase from 3.35 million square foot predicted in 2012; 4. £10.6 billion increase in property values within 1km of a station expected by 2021 As we are now halfway through 2021, it would be interesting to see an update showing benefits realisation progress.

Thanks to Hugo, Roy and Adrian for insightful posts (above) that I completely agree with. I have always thought that Crossrail should have been initiated and defined as an Outcomes, Benefits & Vision driven programme in the first place (the programme having some whopping great construction projects within it) - and ALL Benefits defined (not just the financial ones). Maybe it was, but I have never seen it described in this way.

Three things: 1. I disagree with the inference in Hugo's comment that "optimism bias" is "(a polite way of saying "economical with the truth")". Cognitive biases happen at a unconscious level: they are therefore not deliberate attempts to be be untruthful. What you need to do is take conscious steps to counter-act those unconscious biases of which there are at least a 100! 2. All projects should be delivered with the intent to maximise value e.g. some combination of benefits versus whole life cost and the sooner the benefit is delivered, then, unless it costs a stack load more, more value will be delivered on NPV/discounted basis. However, the delivery of value should be done in a planned way, not an opportunistic way which buggers up the delivery plan and hence costs a lot more: in construction, a change introduced post-contract typically costs 3 times as much than if in the original contract. So I challenge Hugo's comment that "the whole project could be managed from a benefits perspective." It should be managed from a Value perspective. 3. While I was aware that the earlier phases of Crossrail were by no means perfect, what seems to consistently bugger up tunnelling projects now is not the civil aspects which can be delivered on a traditional basis i.e. waterfall. It is the M&E integration where different tangible assets have to be installed on top of each other and then made to work together: there is simply too much interdependency to deliver it under a traditional infrastructure delivery model. A more agile approach, which would have to be reflected in the contracting strategies, would probably be more appropriate.

Following Jon Broome's excellent comment. M&E these days isn't just Mechanical and Electrical, but layers of software and systems. Perhaps the appropriate acronym is MESS! I'm not sure Agile is the right genre of approach, but the systems integration aspect of Crossrail does appear to have been underestimated. The initial concept of the line is part of this: as a system of systems, integrating at least two legacy railways and one entirely new one, along with the signals and controls interoperability headaches that crop up all over UK rail (i.e. this wasn't an unpredictable type issue).

The views below makes some good points, but the APM comments, unfortunately add to the confusion that seems endemic in this report from the PAC. Confusing projects with programmes is a fundamental error as they involve completely different approaches, different governance and different assurance. The PAC report continues to refer to Crossrail as both as if the terms were interchangeable, despite £100m being spent on “strategic programme management”. The value from this is not clear, but what is clear is that the commercial strategy increased risks at the same time as the relentless focus on time to complete escalated costs. Instead of prioritising value to stakeholders, the programme created damaging tensions and 22,800 notices of compensation events. Flaws in the commercial and delivery strategy indicate a systemic failure of governance. The risks were not managed effectively and the complexity exceeded the capability of the delivery teams, through its 36 separate contracts to manage. (Section 2.13). It was plainly too big and took too long to be delivered in this way. That the contingency was massivly insufficient, and that changes in scope and the state of assets caused delays, conflicts and cost-overruns looks like inadequate preparation, fact-finding and not learning the lessons of Bent Flyvbjerg and his team on Megaprojects. The particular failure of governance is laid bare as an inadequate commercial stategy, lack of managing dependencies and interfaces, and not heeding the evidence that the programme was failing. Governance particularly is culpable for its lack of focus on work to be done nstead of falling for the confirmation bias of what had been delivered (Section 2.22). This refects a “good news” culture and lack of critical thinking, more comfortable for all concerned than hard decisions and managing conflict. The skills, time and knowlege of the people in charge must be open to question. , Finally, the report admits that this is based on the progress reports submitted to governance, so it is by no means certain that this is the end of the revelations. There are lessons to be learned but only if we face the facts.