19 July 2012

Going down by the front end #nlpoli

In the late 1980s, the Progressive Conservative administration in Newfoundland and Labrador committed more than $11 million  - ultimately $22 million -  to a plan to grow cucumbers and tomatoes using hydroponic technology developed by Philip Sprung.

Sprung predicted his technology would grow almost seven million pounds of cucumbers and tomatoes in its first year of operation and upwards of nine million by the end of five years.  Sprung had little evidence to back the claim from his test facility in Alberta.

An assessment by provincial officials concluded that the Sprung’s projections were impossible to attain.  Aside from any technological miracles, the Sprung predictions would need the average daily sunlight of Cairo to stand a chance of coming true.  Mount Pearl  - the site chosen for the greenhouses – didn’t even come close to those light levels in the very best years.

Still, the government persisted.

“For once,”  Premier Brian Peckford told reporters, “Newfoundland will be first in new technology and not just in unemployment rates.”  The greenhouses would produce cheap, high-quality vegetables for the local market and supply much-needed work in a province where work was sometimes hard to come by.  What didn’t go to meet local needs would be available for export to the mainland.  Thus, the project was supposed to provide an ongoing source of revenue  and lay the basis for further economic development.

The greenhouses never produced as expected.  A crop failure prompted Sprung to accuse a phantom saboteur.  Costs continued to increase.  By the time the provincial government closed the whole project, the disaster had consumed $22 million of public money and dumped thousands of cucumbers onto the local market at prices far below what it had cost to produce them.

It’s not as though the government didn’t have plenty of experts inside and outside the walls of the Confederation Building raising questions about the project.  Peckford and his cabinet dismissed all of their concerns.  The politicians bought into the idea and persisted in it as much out of optimism as anything else. 

And optimism is a recognized decision-making bias that is especially prevalent among politicians and large projects.

Optimism isn’t the problem with the Muskrat Falls project, per se, although there is an excessive level of optimism about the project and the people behind it.  There are a bunch of other ideas at play there, as we’ve discussed before.

One of the biggest problems with Muskrat Falls is a variation on the sunk cost fallacy. That’s basically the idea that people will tend to keep pouring money into something after it no longer makes sense to do so.  In the Sprung example, the government put in its initial investment.  As problems came up, they dropped in more cash.  After all, went the argument, they were in already to such an extent that it was just as well to keep going with first this little amount and then this next little amount in order to make the project work.  

The Decision Gate Process
The argument is deceptive because all projects run into some cost over-runs or unexpected problems.  People who specialize in managing projects have even developed a way of avoiding those surprises and problems.  It’s called a “stage-gate” process.

In the diagram, the diamonds are decision “gates”:  this is the place when people can make the choice to stop the project,  do more study or cancel it.  The bits in between are stages. 

decision gates

The idea of this approach is to push as much of the planning work to the front end of the project as you can in order to solve problems when they are the least costly.  That’s a process called front-end loading and you will hear an awful lot about front-end loading from the gang at Nalcor.

The diagram below shows the decision gate process and front-end loading  in light of the time in the project life-cycle and how easy it is to make changes in the project.  The whole thing is wonderfully rational and sensible and it can work for plenty of projects.


Nalcor is proud of the fact they are using this approach for Muskrat Falls. Nalcor describes decision gating as a “phased quality assurance” approach:
The Gateway Process allows Nalcor’s leadership team to make sure the Project is ready before moving onto the next gate; in other words, to demonstrate that due diligence, checks and balances are being applied during the Project’s execution. In November 2010, the Project passed through decision gate two. At this point in the process, the decision to move forward with Muskrat Falls first was confirmed.
They even have their own diagram of the process.


If you go back and follow that link   for the “stage-gate” process you will wind up with a paper prepared by Westney Consulting Group.  Nalcor hired Westney in 2007 to consult on a portion of the Lower Churchill Project.

Westney’s 2008 piece on decision-gating specifically addresses the problem that decision-gating often doesn’t work:
The past decade has shown that even the best owners and contractors have been largely unable to scope, define, plan, estimate and execute these mega-projects with any sort of predictability. Cost overruns of 100% or more have been widely reported, along with years of schedule delays. In fact, such experiences are so common there is even a word for it: a “megawreck”.
Nalcor asked Westney to review some aspects of the project in November 2010 when, according to Nalcor, the project passed through Decision Gate 2.  The public utilities board cited a portion of this in their final report.  According to the PUB, Westney recommended that Nalcor set aside a strategic reserve for its capital costs on the project.
This recommendation was not accepted by Nalcor as in its view there had been a reduction in the key risks identified by Westney since its recommendation as a result of factors such as the commitment by the Federal Government for a loan guarantee and the selection of a conventional technology for the HVdc transmission line. Nalcor stated that it would reconsider the need for a strategic reserve amount at Decision Gate 3.
Now on the face of it that suggests Nalcor went through Decision Gate 2 missing another significant budget item for Muskrat Falls, but that’s really not what is so striking about all this. Nor is it really striking that they have not adopted Westney’s advice in their 2008 pamphlet on how to avoid tunnel vision often induced by the Decision Gate process.

If you read through Nalcor’s material they have scrupulously followed the best practice of Decision Gating to make sure the project is successful.  They will even mention, every once in a while, that they can stop if the next set of numbers don’t work out.

A Done Deal

What’s worth noting is that all that just ignores a crucial point:  this project has been front-end loaded to the ultimate degree. As Danny Williams put it on November 18, 2010:  “This project is a go.”  In November 2010, both the provincial government and Nalcor decided the project would happen regardless of the consequences or objections.  News media got the point.  When Williams retired, reports talked about the deal as done and credited him with doing what no one else had been able to accomplish. In the same way, in May 2011, Lisa Moore pronounced it as “an agreement for hydroelectric development of the Lower Churchill River, a drawn-out negotiation that ended with a profitable arrangement for Newfoundland.”

For the past 18 months, Nalcor has mounted a relentless marketing campaign to convince the public the decision to build Muskrat Falls was the right one.  That’s not what you do if project sanction – the decision to build the project – remains, as it turned out, at some undefined point in the future.  It is what you do when you have committed to the project, regardless of anything.

You see the same thing when you look at Kathy Dunderdale’s comment about the project after her recent meeting with Nova Scotia Premier Darrell Dexter:
“What’s the alternative? To either ration energy or sit up in the dark. We have to pay for energy.  Where’s the least-cost alternative?”
The first response to such a statement is to laugh at its hyperbole. it is silly.  The second response, though, is to realise that Kathy Dunderdale views her choice as made.  What is the alternative? she asks, not rhetorically but in all seriousness. Dunderdale is not joking when she offers no practical alternative.

There are practical alternatives, of course. Natural gas is one, proposed by knowledgeable local experts.  Nalcor claims it has studied natural gas and rejected it.  They have offered no copies of their review. They cannot offer what they do not have.

As it became painfully obvious that Nalcor had not done the work it was supposed to do, natural resources minister Jerome Kennedy leaped forward to hire consultants to review the alternatives.  Their task was not to explore the alternatives.  The consultants are to prove that there are no alternatives.

None of these actions are what you expect from project proponents who are following the stage-gate process as a prudent, rational way to manage a project.  None of these are actions you expect from proponents who can stop what they are doing when and if their analysis reveals insurmountable problems, problems like discovering that your project cannot produce the electricity you need.

Both the provincial Conservatives and Nalcor are ploughing ahead in what is really just a variation on the sunk cost fallacy. They believe – for whatever reason – that they cannot afford to turn back from the complete commitment they made in 2010.

Sure, they have spent tens of millions of dollars and will likely spend hundreds of millions before they start construction. None of that really matters. They are sitting on billions of dollars in assets. Money is not the issue.

The sunk costs for both Nalcor and its allies are entirely political, but they are as real a reason to make bad decisions as any others.