Who will reap the benefits of a gas tax holiday?

Where have you been, Megan, my readers cry! Here it's been days since the gas tax became an issue, and you haven't weighed in. What are we to make of all these confusing arguments? How can you abandon us when our head aches so?

Okay, well, I did get several emails asking me to explain the debate.

Like about a zillion economists and most other people who were not whacked upside the head with a stupid stick, I think it's worse-than-useless pandering. There's not so much a debate as a bunch of economists saying "this is bad policy" and two campaigns sticking their fingers in their ears, saying "lalalalalalalalalal I can't HEAR you!"

However, there is apparently some interest in a) how we know who bears the cost of the gas tax and b) more posts with charts. And there's nothing I like doing more than drawing charts. It's rather Econ 101, so economists can skip it, as can anyone who likes to snot in a sophisticated manner about how simplistic Econ 101 is. For those who are interested, it's after the jump. The rest of you go and write indignant letters to your congressmen telling them not to reduce the gas tax.

The Federal gasoline tax is what's known as an excise tax: it is 18.4 cents applied to each gallon of gasoline you buy. The general rule of supply curves is that they slope upwards: if the price drops, suppliers want to offer less of it for sale, while if it rises, they want to sell more of it. Look at a simple example of a natural price of gas at $2.00 a gallon, and a $1.00 excise tax (I'm using big, round numbers instead of the actual numbers because they're easier to graph with my exceedingly crude tools):

Here's what's happening in the graph above. We start off on supply curve S0, where at a price of $2.00 a gallon, suppliers are willing to sell some arbitrary amount of gasoline every year--call it 150 billion gallons, which is about where American demand would probably be now if prices hadn't increased. When you slap an excise tax on gasoline, the fundamentals don't change for the suppliers, but at every point, they need the final price to be $1.00 higher than it is now. So where they were willing to supply 150 billion gallons when gasoline was at $2.00 a gallon, they're now only willing to supply 150 billion gallons if gasoline is selling for $3.00. If the price stays at $2.00 a gallon, only people who could make a profit when gas was at $1.00 a gallon will still want to sell into the market, so the supply will be lower. Essentially, the supply curve has just shifted upwards, as you can see above.

Of course, supply isn't the only factor in the marketplace, though too often economic commentary reads as if it was. At left, we add in demand. Demand curves are basically the opposite of supply curves; they slope downwards, because when something is more expensive, people want to buy less of it.

When we add in our pretty, purple, downward sloping demand curve, you can see that things change. The market will clear where supply and demand intersect. Consumers don't like paying $3.00 a gallon for gasoline; they reduce their consumption. So instead of moving from Point A to Point B, we move to Point C: the price rises a decent amount, and consumers reduce the amount that they use, until the market clears at 130 billion gallons sold at $2.66 per.

In this graph, the cost is split between consumers and suppliers. Suppliers get 44 cents less per gallon than they used to ($1.66 instead of $2) and sell fewer gallons, while consumers pay 66 cents more, and drive less. It's important to note that driving less is also a cost--no, please, I don't want to hear any lectures on the spiritual benefits of simpler living.

Of course, how the tax is split between suppliers and consumers matters, politically. Most politicians would like to live in a world where big, mean oil companies pay the whole tax, while wholesome American voters get off scott free. The only way that this would happen is if supply is what we call "perfectly inelastic". Price elasticity measures how much supply (or demand) changes in response to a change in price. If a small change in price produces a big change in output, this means supply is very elastic; contrarywise, if a big change in price produces a small change in output, this means that the supply is very inelastic. The ideal for politicians would be a situation in which oil companies are unable to change their output at all, so the tax has no effect on consumer's pocketbooks.

From a politician's point of view, the worst possible outcome is that supply is highly elastic. If supply is very responsive to price, then suppliers will bear very little of the tax, especially if demand is inelastic--consumers will start bidding up the price of gasoline to keep the supply high, and the tax will take a lot of money out of voter pockets without reducing driving much. Elastic supply curves are very flat, so you don't have to raise the price very much for supply to fall dramatically. As you can see at left, when you have flat supply curves combined with steep demand curves, consumption falls very little, and the price rises a lot, so that the suppliers are selling nearly as many gallons of gasoline as they did before, at very little reduction in price--most of the cost is directly borne by consumers.

The current debate is an empirical question: how elastic are the supply and demand for gasoline? Paul Krugman has argued that the refineries run flat out during the summer driving season, so there's no excess supply on the market. Meanwhile, consumer demand is rapacious, so they'll just bid up the price of however much gasoline refiners can supply, handing the oil companies a windfall. In our original example, if you reduced the excise tax by $1.00, consumption would move from Point B to Point A, with the amount drivers use rising somewhat, and the price falling somewhat. The suppliers reap most of the financial gains: they sell more gallons, and get 55 cents more for every gallon they sell.

We don't know exactly how elastic supply and demand are right now. For one thing, it changes over time. And for another, elasticity depends on what time horizon you use. If gas rises over a period of ten years, consumers buy more fuel efficient cars and move closer to work. If gas spikes over a month, you can either pay more or keep the car in the garage.

Nonetheless, we can make a decent guess. Over the short term, demand is generally very inelastic. Supply of gasoline is more elastic--you can always store the stuff for later. But it's capped on the upside by the capacity of the refineries. Regulatory requirements for gasoline formulations make it hard to import the stuff from other countries, or for that matter, other states. Chicago has its very own gasoline mix, which is why prices there are prone to spike whenever there's trouble at the refineries that serve it.

But the summer driving season is a little different. I actually think that Paul Krugman is a little too pessimistic--consumer demand has finally started to edge downward in response, which means that the refineries have a little more spare capacity than usual. If there were a gas tax holiday, I would expect to see some more fuel come on the market, though I'm not sure how much more. And summer driving is somewhat more discretionary than commuting.

That said, I'm betting on the refiners. Partly this is because gas prices are so high; 18.4 cents isn't as big a change in one's gas bill as it used to be. And partly this is because gas prices aren't high enough to make large numbers of people stay home from the beach. Consumer demand is very close to perfectly inelastic, and supply is capped.

But this is not the only reason, or even the best reason to be against lowering the gas tax. Remember Megan's Third Law: to spend is to tax. Unless the federal government is going to lower spending to match the lost revenues, we're just going to take the money from somewhere else, now or in the future.

Hillary Clinton's proposal is particularly stupid, in my humble opinion, because it tries to get the money back from the oil companies with a windfall profits tax. Tax incidence is tax incidence: if the oil companies can make consumers pay most of the excise tax, then probably consumers can stick them with your windfall profits tax too. Meanwhile, the instinct to mess with the oil companies every time prices rise is thoroughly counterproductive. We (at least, those of us who want cheaper oil) want the oil companies out foraging for more supply. If you lower the returns on finding new oil, you kill their incentive to do so--more importantly, you kill the incentive of investors to give them capital to do so. All her plan does is make us take the trouble to build new administrative capacity to collect the tax, while keeping all the old administrative capacity for collecting the excise tax (since, after all, it's not actually going away permanently), while scaring the bejeesus out of investors. It's lose-lose-lose.