I like this short article, which is about using game theory to find the cheapest gas. It feels right. About the only problem I can easily see is this: what happens if gas stations don't come in clusters, or, worse yet, if there's only one? My guess is that the price of entry into the gas market is high enough that people won't just casually see a single gas station as an opportunity to build a new one, because then they'd have to overcome the startup costs that the first station no longer has while simultaneously charging enough less so that they'll make up in volume what they lose per gallon (and yes, I know that's the punch line of an old joke). That would keep the number of stations down, possibly to just one. Also, you have to figure the value of satisfying the need for gas -- how much of a premium in price would you pay?
Still, it's an interesting muse.
3 comments:
Rt104 in New Mexico illustrates your point quite well, I think. One gas station in about 100 miles. Literally! It's about 20 or 25 miles from the start, and the next gas is at least 75 miles distant.
There's also a road in South Dakota (I can't remember it's name), where you can drive for over 240 miles without seeing a single gas station; although one or two are within a mile or two toward Rt 80. And then there's The Rockies. You can easily drive 80 or more miles without seeing a gas station.
What the article writer neglects is the cost of gas near a highway. I presume they were driving on Rt15; it's a major highway. The cost of gas is always higher (in many cases much higher) than the cost of gas a mile or so away from the highway. Different customers, different prices. When you cater to the locals, you have to be competitive with the other gas stations around you. When your customers are coming off the highway, and are probably in need of gas, don't know the area and simply want to gas-and-go - you can charge more.
If there are a few gas stations that are visible to one another off the exit-ramp, the price will be lower than otherwise.
The perfect example of this is in Richmond, Virginia. Use one exit, and the cost of gas is about 40 to 50 cents higher than the surrounding gas stations. There are a couple, but you can't see the cheaper station from the one that is closer to the highway. All you have to do is drive about another 50 yards! (Around a bend, into what is clearly the start of a town.) But not many people will do that; they see the gas station, lament the price and fill up. More than a few will decide not to take a chance getting caught in some strange one-way system; others just want to get back on the highway (95).
It's basically: competition keeps prices down.
That limited access is the reason New Jersey regulates the cost of gas on the Turnpike. To use the Turnpike costs money; so people won't want to exit it. (The cost of entering and exiting is a little higher than staying on it, until you get to the NYC area.) This means that the gas stations could charge what they think the market would bear, but it's an artificially constrained market. So what it could bear would be much higher than it might otherwise be (who wants to run out of gas?). As such, NJ developed a system where the cost of the gas is regulated, and is changed once a week. (When gas prices go haywire, as they did in '08, I think a different system is used.) If the average cost of gas goes down, the gas station gets to keep the difference. If it goes up, they have to eat the increase until the next Thursday. It's an interesting, and artificial, mix of government regulation of a market.)
Another interesting tidbit is that he doesn't tell us the name of the gas supplier. Was it one of those cheap-gas places? Or a supplier of known quality? (BP, Exxon, Shell, etc). For instance, I will not use cheap gas in the Ducati, but will (with reluctance) in the Royal Enfield. Even if it means I'm paying 20c more (I've seen that level of difference in premium), I want a gasoline I can have some measure of trust in. It might not seem rational, but there is something to such thinking.
So the writer didn't (really) solve a math or statistical problem. He did in the sense that he applied a statistical model to the problem of finding gas at a reasonable-ish cost. What he really did was come across a commercial truism, and example of how capitalism works. :-)
Carolyn Ann
Oops. I meant to say "he solved an economics problem."
I recall driving the highway in SoDak. Didn't EVER want to run out of gas -- even filling up at Wall was better than that.
Post a Comment