Monday, March 3, 2008

The Utility of 'Ceteris Paribus'

While throwing around Latin for the sake of throwing around Latin irritates me, I had an introductory micro prof who insisted firmly on the use of 'ceteris paribus' instead of 'all else equal'.

Anyway, there's an op-ed in today's Globe that's basically saying consumers won't change their behaviour because of a carbon tax because it's only a few pennies.

As evidence, the following is submitted:
History tells us they won't. The carbon tax rises to 7 per cent of current prices over four years. In the past five years, the average inflation-adjusted retail price of gasoline in Canada went up 40 per cent, and gasoline consumption rose in every province.

Aside from the fact that gasoline consumption is a long-run decision, since most aren't willing to buy a new car to save on gas (yet), as time passes, we get richer. The question to ask is: "if the real price of gasoline had stayed constant over the last five years, by how much more would gasoline consumption have risen?" I'm betting a lot.

If he wants to show that a carbon tax won't cut down emissions, it needs to be shown that as price increases, gasoline consumption increases or remains unchanged - holding income, the amount of urban sprawl, the differential in urban and rural real estate prices, and so forth - constant. Which, of course, is impossible. Econometrics will lead us towards there, but only a controlled experiment can bring us all the way.

I can now answer his head-scratcher:
As a believer in the economic concept of inverse price to demand elasticity, the resiliency of energy demand in face of big price increases has me scratching my head.

Well, if you look at the aforementioned econometrics literature that uses our flawed tools to help estimate the price elasticity of gasoline - holding those other bothersome factors constant - it, surprise! - comes out to be a negative number. Price goes up, we buy less gas, ceteris paribus. See, for example, here, here, or here. (Possible paywalls, depending on your degree of journal access.) He partially cops to this point, but only somewhat.

I won't pretend that there's consensus on the exact number, but the overwhelming majority of the literature suggests that gasoline is inelastic, yes, but certainly does obey the law of demand. As such, there's every reason to believe that a carbon tax will reduce gasoline consumption relative to the situation without a carbon tax.

Yes, I too would like total carbon emissions to fall. But if he advocates a different plan, it can't be because of a reason like 'a carbon tax won't work because it's only pennies and we can't make gas cost even more'. Unfortunately, when he starts to talk about reasonable alternatives to a flat carbon tax - nonlinear pricing and whatnot - all the blather beforehand ruins the message.

UPDATE: More evidence that people respond to gas prices at Environmental Economics.

UPDATE: Rejected at Brown. And here I was thinking it was a relative safety. Word on the street is that they're only looking to admit 10 for September.

1 comment:

Tony Trepanier said...

I'm with you on the Latin, "all else equal" "ceteris paribus" you're not saving a lot of time. As for the carbon tax, I agree you will see a reduction in carbon emission, but it will be too small to have any effect on climate change. It will probably have more benefits improving traffic in Vancouver than anything else