Why do gas prices rise quickly, but fall slowly?
The first lesson in economics is that prices move based on supply and demand. But the second lesson is that they don’t always do it well. Sometimes prices are “sticky,” and don’t quickly respond to changing market conditions.
A common example is gas prices. From the The Economist’s online econ glossary:
Petrol-pump prices do not change every time the oil price changes… Sticky prices are slow to change in response to changes in supply or demand. As a result there is, at least temporarily, disequilibrium in the market.
Stickiness is easy to see in the data. Below is a chart of California gas prices from 1996-2000. The pink line is wholesale prices, and the blue is retail. Note how quickly retail prices rise when wholesale prices jump, and how slowly they come down when they fall—a classic case of asymmetric price adjustment.
There are many reasons for stickiness—“menu costs”, contracts that lock-in prices, firms’ unwillingness to irritate customers with constant price changes, etc. But a common one that’s false is that price stickiness like in the above chart is evidence of monopoly pricing power by firms.
It’s an easy mistake to make. From the chart, it’s clear that when wholesale prices rise, gas stations jack up retail prices immediately. But when they fall, retail prices don’t drop as fast. Firms keep prices high, and profit margins fatten. Looks a lot like a case of monopoly pricing to gouge consumers. Right?
Not so fast. A new job-market paper from Matt Lewis, a Ph.D. candidate at U.C. Berkeley, shows that this asymmetrical price adjustment has nothing to do with market power. Instead, it’s caused by consumers, and the way they act differently when prices are rising vs. falling. [Download the paper here (PDF)]
The paper builds a “reference price search model”, in which consumers don’t search much for better prices when they’re falling, but search a lot when they’re rising. When consumers search, markets are more efficient. And when they don’t, markets are less efficient. So when wholesale gas prices fall, gas stations drop prices just enough to keep people from searching much. That makes markets sloppy, and retail prices get pushed down more slowly than up.
It’s a great paper, and it makes intuitive sense—as one professor quipped in a recent listserv discussion, “This phenomenon is no more surprising than the fact that no one ever gets a second opinion when a doctor delivers good news.”
This pricing phenomenon is very widespread, and the paper is one of the first attempts to model it at a micro level. Lewis finds pretty good empirical support for it, too. Email him here.
So, gas prices are on the rise now… if you’re in the Washington, D.C. area, try this tool to make your own search for cheap gas more efficient.
Posted by Andrew on Thursday May 27, 2004 | Feedback?
