They affect consumers the same either way, and upstream is simpler and more transparent
In his post on conservatives and carbon taxes, David said:
First, we have to remember all the places the price signal created by an upstream tax can be diluted or stymied on the way to consumers — i.e., those who can change their behavior in response to prices. Not every industry or business will pass an increase in operating costs directly on to the next link in the chain. Information failures and split incentives abound. Price signals that begin strong, catholic, and clear become fragmented and faint downstream. For all the hype, an upstream carbon price will deliver fairly little incentive to where the carbon is used.
There are two problems with this: It is overstated, and it places blame in the wrong place, i.e., the fact that the tax is levied upstream.
Both points can be substantiated by looking at a 2004 study of gas taxes [PDF]. This study shows that federal gas taxes, the closest thing the U.S. has to a carbon tax, are borne about 50/50 by consumers and producers/distributors of gasoline. This means that 50 percent of a carbon tax does in fact reach the consumer. However there is an even more important point. You can’t levy a tax much further downstream than a gas tax. It is paid at the gas pump by the consumer. Yet it turns out that consumer pushback still ends up placing almost half the cost on the producers and distributors. That is true even though the oil industry is an oligopoly, controlled by few players.
In fact, how much of the tax is born by people other than consumers turns out to be primarily determined by long-term demand elasticity in response to price. (Elasticity is a number representing how much less consumers of a good or service consumers use as prices rise.) Elasticity determines how much of an emissions tax is borne by consumers vs. producers and distributors — not whether the tax is levied upstream or down, but to what extent consumers will tend to lower demand in response to price increases.
That makes a strong case for levying carbon taxes upstream rather than down. You have the simplicity and transparency of levying the tax at far fewer points against far fewer actors. And levying the tax downstream does not change the tax incidence. Consumers, producers, and distributors will split the cost about the same way regardless of to whom it is charged.
I will note that this is also a good argument for using public investment and standards-based regulation as the primary means to lower emissions, where practical. Where elasticity is high, so will be consumer pushback, tempering consumer response as producers and distributors bear much of the cost. Where elasticity is low, consumer pushback won’t be strong because consumers won’t be able to lower demand easily in response to higher prices. So consumers will bear the highest tax incidence when it will be least effective. Note that this applies equally to a carbon tax, an auctioned permit, or a cap-and-trade system.