Towards Gasoline Market Efficiency

For the past year or so, I’ve been using the same website to save money on gasoline. The parent site of the one I use is Gas Buddy. Commuting to work I spend about $130 per month on gas, or roughly $1,550 per year. There are several reasons for this. Gasoline is one of my biggest work-related expenses, California gas prices are consistently among the highest in the nation, and I’m also an economist. I feel impelled to fill up at the station offering gasoline at the cheapest price, without going significantly out of my way to get there, of course.

Economic theory would typically classify a local gasoline market as a competitive market, yet, I often see differences of 20-25¢ per gallon for the same gasoline grade among nearby stations. Why does the standard model of competition not seem to apply here? Because most consumers probably accept the notion that gasoline of the same grade is nearly identical regardless of the station, competition should drive prices to the same competitive market clearing price. However, gasoline retailers often try to differentiate their product through methods such as affiliated credit cards, which give the holders a discount when they purchase gas with the card from a retailer that is part of the corporate chain. Another strategy they use is to offer a discount on a car wash to consumers who have purchased gas at their station. Nevertheless, it doesn’t seem like such differentiation would be important enough to keep the market from a perfectly competitive equilibrium.

What else might explain these facts? One possibility is that some gas stations employ a strategy of luring customers into their stores with gasoline sold below cost, to sell them high margin convenience goods. Another possibility is that some stations enjoy location advantages that allow them to command higher prices, such as the first station located off of a high traffic freeway exit. Nevertheless, the explanation that I prefer is that gasoline consumers do not have all of the information regarding prices of gasoline in surrounding areas. Websites like Gas Buddy help alleviate this informational deficiency in a nearly costless way thanks to its gas price maps and price lists. As more people use the site, the local gasoline markets covered should theoretically approach a perfectly competitive equilibrium.

Where does the website get its price information? People who are interested in either winning gas cards or making the gas market more efficient have accounts on the site and post gas prices there. Although there are obvious benefits to the information provided by Gas Buddy, there may also be drawbacks to the site. Besides the obvious damage to the profits of gas station companies, there are likely to be people who misuse the information. For example, imagine the user who drives several miles out of his way to fill-up on gas that is only 5 cents cheaper per gallon than the nearest station. This person may save $.75 or so, but environmental costs of the extra driving distance, the cost of the additional gasoline used and vehicle wear, and the value of the person’s extra driving time are likely to sum to significantly more than $.75. So, while getting the cheapest gas is great, remember that there are more to costs than just retail prices.

Author’s note 10/19/09: During her review of this post, Kasie Jean mentioned the possibility that consumers may have gasoline brand loyalties. The author found this unlikely but later received advice from a trusted mechanic regarding the benefits of Chevron with Techron gasoline. The author owns no securities issued by the Chevron corporation.


Discussion Questions

1. Now that you are aware of a gasoline price website, would you use one to locate the cheapest nearby gas prices? Why or why not?

2. Think about the characteristics of perfectly competitive markets. Do you believe that gasoline markets are perfectly competitive? If not, what are some aspects, besides those described above, that keep them from perfect competition?

3. In 2007, a study concluded that the optimal tax on gasoline was $2.10 per gallon. What is your opinion of this conclusion? Do you think that gas price websites would be viewed more if gasoline taxes were significantly higher?

4. In what other ways has the internet made markets more efficient or perhaps less efficient?

Slicker than Oil: The Debate Over California’s Proposition 87

Proposition 87, a hotly contested measure on the ballot in California, nicely illustrates the key lessons about tax incidence--who actually pays a tax. The political debate about this Proposition shows that these lessons, which are typically covered in an introductory microeconomics course, are not well understood by the voting public. Is education powerful or what? Armed with knowledge of just a few economic principles, you will be able to analyze important policy issues better than most.

To raise funds for research on and development of alternative fuels, Proposition 87 taxes oil produced in California. (Read a summary of the provisions of this act on the Secretary of State's web site.) Opponents of the tax claim that this tax will increase the price of gasoline at the pump. The following figure shows why this claim is wrong. The market for oil is global, and the price of oil is set by global supply and global demand. California produces less than 1% of all the oil produced in the world, so changes of a few percent in its output would be far too small to have a noticeable effect on global supply and demand. Taken together, all the producers in California are in the same position as a single firm in a competitive industry. They face a residual demand curve for their oil that is horizontal, or perfectly elastic, at a price equal to the world price of oil plus the cost of transporting it to California. Call this price P*.
To understand why the demand curve faced by local producers is horizontal, consider the decisions made by the typical manager of a gasoline refinery. She already has barrels of imported oil delivered to her at the price P* and could order many more at that same price. If the price of oil delivered from a local producer was more than P*, she wouldn't buy any. If it was less, she'd buy as much as she could and cut back on purchases of oil brought into the state from Alaska and the rest of the world. Local producers, understanding this, would be foolish to charge a price less than P*. They'd be leaving money on the table. Even if they did, this wouldn't reduce the price that the refinery manager would charge for gasoline. She'd happily buy all the cheap local oil she could, but still charge the same price for gasoline that other refiners, who continue to import foreign oil, are charging--the price determined by P* and the cost of refining.

What happens if the state of California imposes a tax of T% on the oil produced by the local producers? The figure shows that the price of oil purchased by the refineries is still equal to P*. Local producers receive (1 - T) times P* per barrel of oil. Because they face a horizontal demand curve, they bear the full cost of the tax. They can't pass the tax increase on to consumers.

As the figure shows, local production of oil will fall by a small amount after the tax is imposed, which means that imports of oil will go up. The amount of this reduction depends on the elasticity of the supply curve. As usual, the supply response grows larger over time. Initially, there should be very little supply response, but over time, wells will be retired from service sooner and fewer new wells will be drilled.

What might the size of this impact be? One estimate cited by opponents of Proposition 87 is that local production will fall by an average of about 3.4% or about 22,000 barrels per day over the first 10 years of the life of the tax. Oil produced in California meets only 37% of consumption there. The balance is supplied by oil imported to California from Alaska and the rest of the world. This means that imports of oil into the state would increase by about 1.8%. A gradual increase in imports of this magnitude would be too small to have any perceptible effect on the world price of oil or the transportation costs of oil to California.

The local producers do understand the economics of tax incidence. Chris Hall, an oil producer based in Torrance, CA, was quoted in the San Diego Union-Tribune as saying of the tax that “I can't pass it on…I'm a price taker and not a price maker. I don't determine the market.” Understandably, these producers have raised a lot of funds to support a campaign to oppose the imposition of this new tax. The tax would raise up to $4 billion before it expires, so for oil producers as a group, it makes sense to spend millions of dollars to defeat the tax. What is interesting about the campaign ads that they support is their repeated claim that the tax will raise the price of gasoline. No doubt, their campaign consultants have found that raising this irrelevant issue is the most effective way to get people to vote against this proposition.

Reasonable people can differ about whether Proposition 87 is a good idea. See the web page, by one well-known economist who studies energy, Severin Borenstein, for an even-handed discussion of its plusses and minuses. Or read this critique by Greg Mankiw. Ironically, he is opposed to the proposed tax because it will not increase gasoline prices and therefore will not encourage conservation. This kind of discussion clarifies the issues and helps people understand the underlying economic principles.

The other kind of political discussion, the kind that goes on between battling pundits or in the back and forth of campaign ads, sways some voters because they don't have the benefit of an introductory course in economics. They don't understand something that you do: a tax on a small subset of firms in a competitive industry will not affect the price paid by consumers.

1. Greg Mankiw correctly points out that the tax in Prop 87 is not a Pigovian tax--that is, a tax on oil for the purpose of reducing oil consumption to socially optimal levels. However, the revenues from Prop 87 are intended to subsidize the research and development of alternative energy. Because the marginal private benefit of R&D is less than the marginal social benefit, the market does not allocate enough resources to R&D on alternative fuels. Appropriate government subsidies could encourage a socially optimal level of R&D. Does this mean Prop 87 is in fact a Pigovian subsidy? Should Mankiw and the other members of his Pigou Club support such a policy? (Thomas Friedman and Al Gore, both of whom Mankiw counts among the membership of the Pigou club, already do. Other members of the Pigou club are encouraged to comment…)

2. Would a Pigovian tax on gasoline consumption be a better way to fund research on and development of alternative fuels? Why or why not?

3. How useful is it to enact public policy through ballot measures? On the one hand, polls show that the public isn’t well informed as to the economic consequences of Prop 87 and other ballot measures. On the other hand, the policies developed by legislators are also imperfect. Some people argue that elected officials face few political incentives to tackle long-term problems like global warming. Is Prop 87 an example of “direct democracy” achieving something the legislative process could not? Or is it an example of why setting public policy is best left up to the legislature?

Cheap Gas Hurts

Economists rarely advocate higher taxes on a good or service because higher taxes often increase the price that consumers pay and lower the price that producers receive--a "lose-lose" situation for both consumers and producers. However, Pigovian taxes, which are used to correct situations in which the free market produces an inefficient result, might actually increase social welfare. Greg Mankiw, an economist at Harvard and founder of the Pigou Club, argues that such taxes are currently needed on gasoline, due to the negative externalities that accompany gasoline consumption.

A negative externality is a cost imposed on a third-party by the consumers and the producers of a good or service. Take for example, gasoline. Oil companies produce and distribute large amounts of gasoline to satisfy America's desire to drive. How does a person who uses gasoline hurt other people? First, burning gasoline emits toxic chemicals such as carbon monoxide and carcinogens that damage public health. Second, cheap gas contributes to excessive driving which wears down our country's highways and causes traffic congestion. Third, as Al Gore argues, burning gasoline produces carbon dioxide, which contributes to global warming. Fourth, as Thomas L. Friedman has argued, high oil revenues actually support regimes like Iran and Venezuela, decreasing freedom in those countries as well as our own national security.

If the consumption of gasoline imposes a negative externality, then economists say that the marginal social cost (MSC) of gasoline exceeds the marginal private cost (MPC). The invisible hand fails to bring the market to an optimal outcome because the free market equates demand and private supply, and does not take external costs into account. Ideally, the market would equate demand and social supply, but rational consumers would not take into account external costs because they feel someone else should reduce their consumption of gasoline (free-rider problem). The free market leads to an almost shocking result: the price of gasoline (P1) is below the socially-optimal price (P2), and the quantity of gasoline consumed (Q1) exceeds the socially-optimal quantity (Q2).

In other words, in a free market, Americans consume too much gas! The government may remedy the situation by increasing the per-unit tax on gasoline. Higher gas taxes would increase marginal private cost and reduce the gap between social supply and private supply.

1. In a free market, the price of gas is P1 and the quantity of gas consumed is Q1. In this case, what is consumer surplus plus producer surplus minus total external costs?

2. Suppose the government imposes a per-unit tax on gasoline that forces the market to price and produce the socially-optimal quantity (Q2). What is consumer surplus plus producer surplus plus government revenue minus total external costs?

3. An action should be taken if and only if the benefits outweigh the costs. What are the costs of the gas tax in this example? What are the benefits? Which one outweighs the other?

4. The above example assumes the government has perfect information about the size of the externality caused by gasoline. But in reality, measuring the costs and benefits (especially when it comes to things like climate change or the effects on national security) can be difficult. Does this problem of imperfect information mean we should not impose Pigovian taxes? If you think we still should impose Pigovian taxes, what does the problem of imperfect information imply about the optimal level of taxation?