Social Econ Blog

The great gas pumping debate continues.

To recap:

Bryce round 1

Tony V. round 1

Bryce round 2

Tony V. round 2

In his round 2, Tony V. asks, "what is the relevant market failure here?" He doesn't understand what produces the market failure, and thus can't believe that it exists. To this I respond, "how is the failure of a profitable market to exist in most of the country not sufficient evidence for market failure?" In many places, people cannot easily pay someone to pump gas for them even though collectively they are willing to pay more than enough to compensate those necessary to provide the service. They cannot do this because the owners of the gas stations do not allow this market to develop. Ultimately, I argue in this post that this occurs because the affected parties are unable to coordinate and "bribe" the owner (although at least part of this failure may be an entrepreneurial failure and not a failure of the market (unless we blame the market for failing to produce such an entrepreneur).

There are two important components to my argument. First, I argue that (in the absence of other considerations like convenience store bundling) this market should exist much more widely than it does currently. That is, the total value to consumers of this service exceeds the total cost to service providers (i.e., S>0 in the model below). My guess is that at any moderately busy gas station (particularly one in a location prone to inclement weather or in a moderately upscale area) an individual could easily earn more than minimum wage charging 25 cents per car to pump gas. In actuality, my belief is that many people would gladly pay a dollar (particularly if we called it a tip). At many hotels, airports, and train stations, people tip bellhops or skycaps a dollar per bag (and frequently more) to carry/watch bags. We tip waiters, baristas, cab drivers, … above the amount of their compensation that was already incorporated into the price. Now, I certainly understand that some of these are just random conventions and maybe the forces of convention have decided that paying someone to pump gas is unnecessary or unseemly (enough so S=0 in nearly all cases), but I find that implausible. Pumping gas and washing windows is dirty work (in that your hands and clothes can actually get dirty), and many people gladly pay others to do their dirty work.

Second, I argue that the economy as a whole is worse off when this market is not allowed to develop. That is, I argue that the marginal gain to owners from preventing people from providing this service (which comes entirely from the gain in profits from people who would not have purchased other (high profit-margin) goods and services if they weren’t forced to get out of their car) is not larger than what is lost to the would be consumers and providers of mini-serve and the would be recipients of the money spent by those whose behavior changed when they were forced to get out of their cars.

Now, this is a situation with clearly defined property rights (the owners of the service stations have the right to allow or not allow others to pump gas in exchange for money on their property). Thus the Coase Theorem indicates that we should be able to reach the efficient outcome. The three groups who benefit from allowing mini-serve (consumers of mini-serve, providers of mini-serve, and whoever loses profits from people whose choices change when they are forced to get out of their cars) should band together and bribe the owner fully compensating him/her for the lost convenience store profit and leaving everyone better off.

The problem in this case (and in most cases where the Coase Theorem fails) is coordinating that transaction. The obvious hurdle is identifying and involving the affected alternative suppliers. These people probably have no idea that they would have benefited by X if person A had not gotten out of their car back at the gas station. If they don’t realize they’ve been affected, involving them in the bribe transaction is clearly difficult. Further, given the high profit margins on convenience store items (while gas accounts for 66.5 percent of revenue it accounts for only 27.5 percent of gross profits), this group may be necessary to achieve the efficient outcome. (Technically, I guess I also need to account for the counter bribe from the people who ultimately get the dollars that would get spent buying gas pumping services in the model as well.)

Setting that group aside, if the surplus from just the market for pumping gas is large enough, it might be possible to improve efficiency simply by doing a better job of extracting the surplus from consumers (and to a lesser extent providers). In the previous model, I assumed that the sellers of gas pumping services could perfectly price discriminate in order to capture all of consumer surplus. The failure to allow service came if owners couldn’t extract enough of that surplus from the servers (i.e. b was small). In reality, extracting surplus from consumers is much harder than extracting surplus from the servers. It may be that the failure of this market to develop stems from difficulty price discriminating.

Typically, at mini-serve in Oregon and New Jersey and full-serve everywhere, the price of service is added to the per gallon price of gasoline. This single price allows consumers to keep a lot of their consumer surplus. It is certainly plausible that sufficiently compensating both service providers and owners requires extracting more of consumer surplus than is possible at a single price. As such, the failure may be entrepreneurial (i.e., the failure is figuring out how to do this).

I think the timing is ripe for someone to enter the market and supply gas pumping and window washing services and solve this problem. After 30 years many people (outside of Oregon and New Jersey) are sufficiently conditioned to combine gas purchases with sugar, fat, and nicotine purchases that I expect losses from people choosing to stay in their car are falling. Further, as pay at the pump technology spreads the number of self-serve customers buying convenience items may be falling. These two forces lower the expected amount an entrepreneur needs to pay an owner in order to supply service. In order to maximize the likelihood of success, I think the market should be organized similar to a strip club. Service providers pump gas, wash windows, and do whatever else (ok maybe putting “whatever else” in close proximity to a strip club reference wasn’t a good idea, but you get the picture) in exchange for tips. Maybe one would charge a small flat fee for the service, but most of the extraction of consumer surplus would come from tips. The owner of the station then charges a “uniform fee” in order to extract enough rents to compensate for lost revenues. If this uniform fee not so large that people are deterred from taking the job, everyone is better off. Consumers keep some surplus, service providers probably make more than minimum wage, and owners are happy.