Friday, 18 June 2010

Junky cost benefit studies: Canadian edition

William Watson fisks a dodgy cost-benefit study purporting to show the benefits of large-scale solar panel power in Canada:
do you think the federal government and Ontario should build a $2.4-billion state-of-the-art photovoltaic solar cell manufacturing plant and hand over the keys, free of charge, to a national champion producer in that industry? Does that sound maybe a little implausible? Not if you believe the cost-benefit analysis presented recently in the journal Energy Policy by two researchers in Queen’s University’s Department of Mechanical and Materials Engineering.

In fairness, the cost-benefit analysis they used is typical of lots of cost-benefit analysis in this area, i.e., a little junky. What are the benefits? They don’t actually calculate the all-in social benefits and costs, which are often hard to estimate but are what should really determine the decision. Instead, they calculate the “government return,” the cash return the government makes on its $2.4-billion investment. What form does the return take? Income taxes, corporate taxes, sales taxes and health and environmental costs saved because the solar cells the plant produces replace polluting coal-fired electrical plants. When the Queen’s researchers estimate all these future flows of income to the government it ends up their present value at plausible interest rates exceeds the $2.4-billion investment.
Watson then goes through the usual laundry list of dodgy practices:
  • Lowballing project costs
  • Not counting deadweight costs of taxation used to support the project
  • Conflating the benefits of switching from coal to solar with the benefits of building solar in Ontario when panels could be imported at lower cost
  • Counting taxes on employee wages against a counterfactual that those employees would never have otherwise ever been employed
Dodgy nonsense all over.

A good cost benefit analysis is pretty powerful. I'm starting to wonder though whether the whole cost benefit analysis industry would pass cost benefit analysis, given that the majority of studies are dodgy nonsense designed to come up with a politically appropriate number.

No comments:

Post a comment