This week, Press Progress and the Vancouver Sun’s Peter O’Neil reported that Kinder Morgan had quantified the economic benefits of oil spills in its application to expand the Trans-Mountain Pipeline from Edmonton, Alta. to Burnaby, B.C. Yes, the benefits of oil spills—they create jobs in the communities where the spills occur. Beyond being a public relations disaster for the company, this should lead us to re-examine what we accept as proof of the economic benefits of projects in general.
Companies facing regulatory hearings like Kinder Morgan, groups promoting events like the Vancouver Olympics, or politicians justifying public expenditures like the new hockey arena in Edmonton generally talk about the benefits of their projects using a tool called economic impact analysis. This technique takes spending, runs it through a statistical model, and predicts impacts of that spending on both GDP and employment.
There are, basically, two issues with this method, and both are wonderfully illustrated by the oil spill example. First, the method is agnostic as to what the money is spent on, so money spent cleaning up an oil spill looks remarkably similar to money spent building a new hospital or school in terms of economic impact. Second, the technique generally does not account for where the money comes from, or how it would otherwise have been spent. Money spent on a hockey arena in Edmonton might otherwise have been spent on light rail or tax reductions, both of which would also have economic impacts. In the case of an oil spill cleanup, the costs are likely to be directly incurred by an insurance company, but the premiums paid for that insurance come at the expense of the value of the oil transportation service—the higher the expected clean-up costs from oil spills, the higher insurance premiums will be, and this will mean higher pipeline tolls, which in turn implies lower profits, taxes, and royalties on the products shipped. Simply-put, there’s no free lunch.
Economic impact analysis leads to what economists have known for decades as the broken windows fallacy—the idea that rebuilding existing things creates economic growth by requiring people to spend money. The oil spill is a perfect example of this fallacy, but of course it’s one in which we all know intuitively that oil spills are bad, so it seems ridiculous to talk about the benefits of cleaning them up. The fact that the technique we use to evaluate the benefits of spending in general only makes sense when we know the spending is beneficial should raise more than a few flags.Economic impact analysis is a wonderful tool for confirming your pre-existing biases and for justifying anything. For this reason, it’s a terrible tool for analyzing economic benefits because it never yields negative numbers—all spending is good spending as far as an economic impact analysis is concerned. All spending creates jobs. The trick, it seems, is to only listen to the results when it’s a project you like.
Kinder Morgan was quick to say that they know that oil spills aren’t actually beneficial even though their analysis suggests there would be benefits. But, the absurdity of economic impact analysis isn’t limited to oil spills. Some of the benefits they cite for the pipeline such as jobs and direct and indirect GDP impact would be higher if they built the pipeline, ripped it all up, and built it again- twice the jobs—twice the GDP impact. Of course, the costs of doing so would be paid for by the shippers, so it would never happen, but why rely on a technique which ascribes benefits to something so obviously ridiculous?
As Stephen Gordon has written before, I don’t believe there’s a technique both more universally loathed by academic experts in the field and at the same time widely used in public policy than economic impact analysis. Let’s not just look at Kinder Morgan’s citing benefits of oil spills as a PR blunder, but rather let’s look at what it really is: the best evidence we’ve had in a long time that we’re doing project evaluation wrong.
Update: Kinder Morgan responds
I am writing in response to Andrew Leach’s post. In it, he questioned Trans Mountain’s choice of words in our Application related to spills and the economy and the underlying methodology for calculating the value of our project. Regrettably, some have taken a statement about oil spills in our 15,000-page Application to the National Energy Board out of context. Spills are not part of our economic benefits analysis, nor do we in any way say that money spent on spill response would be justification for our project. No spill is acceptable to us anywhere, any time, for any reason.
We do, however, as our interpretation of the NEB’s filing manual, list, in depth, the various socio-economic effects of a worst-case-scenario spill. We all know that at the end of the day the total effect of a spill is negative and every effort must be expended to prevent such a thing from happening.
There are many positive economic benefits that will result from our project – jobs, local tax dollars and national benefits for all Canadians – and we have provided extensive studies of those benefits in sections of our Application that are unrelated to those in which the effects of a spill are analyzed.
Our description of the socio-economic impacts of spills should have been more sensitive to people who might read it without context. However we are confident that our Facilities Application demonstrates our commitment to transparency, safety and that moving oil in an expanded Trans Mountain pipeline benefits oil producers, government, and Canadian taxpayers.
Scott Stoness, Vice President, Regulatory and Finance, Kinder Morgan Canada