The recipe for failure | The Press

The object did not go unnoticed by the National Assembly. We saw it in the hands of Geneviève Guilbault and Jonatan Julien, among others. I’m talking about the book How Big Things Get Done by academic Bent Flyvbjerg and journalist Dan Gardner.

Several ministers received it as a gift from François Legault’s chief of staff. At the end of a sophisticated journalistic investigation, I obtained a copy of this book available in bookstores.

Here is the message, because it deserves to be relayed.

The short version: the vast majority of projects fail, and it’s not because they get derailed along the way. This is because they are going in the wrong direction from the start.

Summarized like this, one might believe it to be another fashionable management book. But lead author Bent Flyvbjerg doesn’t play guru. He is an economic geographer and researcher.

Over the past 20 years, he has amassed a database of more than 16,000 projects. His conclusions are frightening.

Barely 8% of them are on budget and on schedule. And only 0.5% do so while complying with the original mandate – for example a metro line with the planned number of stations. In other words, out of 200 projects, only one will be successful.

The key, according to the authors: be slow in planning and fast in execution. However, the opposite is more common.

An optimism bias encourages people to underestimate risks. And often, politics makes this trend worse.

If Quebec examples don’t yet come to mind, it shouldn’t be long before they do.

For a politician, bridges, stadiums and other big projects can become means, not ends. They are used to win votes and elections. Their finish line is actually the starting point: the precious photo with a construction helmet and a shovel, with the dignitaries and the cameras. But will these projects reduce congestion or will they attract a professional team at a reasonable cost? When the answer comes, they are no longer there to be held accountable.

According to Flyvbjerg and Gardner, the most important question should be “why”. It should be repeated during each step. What exactly is the objective of this project? To do this, you must first document the needs and look for the best way to satisfy them.

With the third link to Quebec, the opposite was done. The Lévis Chamber of Commerce promoted a new road axis before verifying whether it was technically feasible, whether it would reduce congestion in the long term and whether it was at a reasonable price.

The politicization of major projects also prevents them from being adjusted along the way. The authors give the example of Ottawa’s rapid light rail. To get elected in 2010, Mayor Jim Watson promised to limit the bill to 2.1 billion. However, the cost estimate was not finalized. This ceiling imposed questionable choices, such as doors that could be blocked by users.

The example of Ottawa also shows the danger of segmenting risks too much. In principle, this seems prudent. The City, the developer and its subcontractors each assume their role.

In practice, it doesn’t work well. Instead of thinking about the purpose of the project, everyone protects their backs by transferring the risk to others.

Except that in the long run, this tension between partners becomes harmful, demonstrates the Flyvbjerg database. Conflicts are becoming judicialized and lawyers are the only winners, as we have seen in particular with the SRB Pie-IX^⁠1.

It is better to manage projects in an integrated manner. And this, from the beginning, maintain the authors.

How to predict and reduce risks?

By drawing lessons from similar cases already carried out, the authors respond.

In particular, they distinguish reversible projects, such as the design of software, from those which are irreversible, like a bridge that cannot be erased and rebuilt in the middle of the work.

For the first category, audacity is better justified, they demonstrate, without however repeating the cliché of Facebook which boasts of “acting quickly and breaking things (Move fast and break things)”.

For the other, caution is required. Simulate, test and retest, they say. This would contribute to the success of the architect Frank Gehry, to whom we owe, among other things, the Guggenheim Museum in Bilbao. He reproduces hyper-precise computer models of his constructions, even if it takes a long time, in order to minimize errors.

The Pixar studio is also cited as a model, with its 20 blockbusters consecutive. The designers’ recipe: write a few pages, show them to their colleagues, discuss them as a team, correct them, then start the process again. The opposite of the cliché of the tortured genius who works alone in his corner.

The counterexample is the Sydney Opera House. Although celebrated today, this iconic work was a disaster. It cost 14 times more than expected. We even had to demolish part of it along the way in order to start again.

The causes of failure are numerous, but Flyvbjerg and Gardner put forward one that was predictable: tailoring. The more atypical a project, the more risky it is. This was also the case for the Olympic Stadium, for which we are preparing to pay once again for the renovation.

The projects with the best success rate are scalable. They can be divided into segments that are preassembled or, at least, have already been manufactured elsewhere. A lesson for the energy transition: solar and wind energy come together in this way. This explains the rapid decline in their costs over the last decade (despite a more recent increase due to inflation).

Among the 25 project categories studied by Flyvbjerg, wind and solar are those that are the least financially risky. The worst: nuclear power, where each plant presents particular challenges. This pitfall could be circumvented by the new technology of modular nuclear reactors, they say. The second worst: the Olympic Games.

The authors are obviously not saying that the projects should be abandoned. They simply plead for a more careful analysis of their risks.

This, in short, is what we learn in How Big Things Get Done. That CAQ ministers have read it is good news. What will be even more encouraging is to see concrete proof that they have learned from it.