Ebook Sticky Prices, Inventories, and Market Power in Wholesale Gasoline Markets
From January 1999 to March 2000, the price of crude oil tripled from about $10 per barrel to over $30. Significant crude price volatility has continued through 2000. In addition, changes to reformulated gasoline in some parts of the U.S. have been associated with further volatility in gasoline prices. Oil companies argue that gasoline price movements just reflect oil prices, costs of adjusting supply, and the normal interaction of supply and demand.
Many consumers and politicians have interpreted these gasoline price increases as the oil companies seizing an opportunity to exploit their market power. While there are idiosyncratic aspects to local price fluctuations, recent events have highlighted two of the underlying issues that affect all gasoline markets: the cost of adjusting the supply of gasoline and the relationship between price adjustment and market power.
In this paper, we examine one piece of this puzzle: the responses of gasoline prices to cost shocks and how those responses might be affected by market power. This research is motivated by a puzzling empirical pattern in the movements of gasoline prices: wholesale gasoline prices do not respond fully and immediately to changes in the price of crude oil. The lag in adjustment occurs even even though the price of crude oil is a major determinant of the cost of gasoline and crude price changes are public information. Borenstein, Cameron and Gilbert (1997), for example, documents that crude oil price shocks are eventually fully passed through to wholesale gasoline prices, but full adjustment takes many weeks. Because wholesale prices are formed in well organized markets where one would expect market clearing prices to incorporate quickly all available information, these lags are surprising. A change in crude prices changes the opportunity cost of the primary input, and under most standard models of firm behavior would lead to an immediate change in the equilibrium price.
Consider, for instance, a competitive refiner who realizes that the price of crude oil has increased by an amount sufficient to cause a 5/c increase in gasoline prices in the long run. If the firm had been producing where short run marginal cost was equal to price, then its marginal cost is now above price and it has an incentive to reduce production. Since all refiners have a similar incentive, each would withhold supply and the price would increase immediately by 5/c. Whether the price of crude oil rises or falls, the price of gasoline should adjust immediately. An anologous argument holds if refiners have some market power. The magnitude of the gasoline price adjustment to a given crude oil price change might be differnt, but the response should still be immediate. The purpose of this research is to explore possible explanations for why immediate adjustment does not happen.
We consider two classes of explanations. One comes from theories of supply adjustment costs. Because adjusting levels of production is costly, firms in these models spread the adjustment over time. A reduction in the price of crude oil, for example, will imply some long run increase in the supply of gasoline. But because there are adjustment costs that increase with the absolute size of adjustment per period, firms will spread the adjustment over time, gradually achieving the full quantity increase implied by the decline in cost. Models of this type have been proposed by Pindyck (1993, 1994) and Thurman (1988) among others. Importantly, while it takes some time to achieve the new long run equilibrium in these models, price adjusts to clear the market at every point in time. There is no non price allocation of supply or demand. This explanation is consistent with the arguments made by many refiners over the last few years.
The other class of explanations comes from a large literature that spans industrial organization and macroeconomics and focuses on differences between the market clearing price and the price at which spot transactions actually occur. In this literature, the problem is to explain why the transaction prices charged do not change (or do not change quickly) when the market clearing price changes. The models offer a variety of mechanisms that lead to a spot transaction price that does not clear the market and, thus, to non price allocation. Menu costs, for example, that make changing the transaction price costly, have been offered as an explanation for infrequent price changes even when market clearing prices change frequently. Other models point to information imperfections as the source of the divergence. When information is imperfect and demand is relatively inelastic, quantity adjustments might be preferable to price adjustment. Or buyers and sellers might minimize transaction costs and reduce price risk by entering long term agreements that limit price flexibility.
Download
PDF Ebook Sticky Prices, Inventories, and Market Power in Wholesale Gasoline Markets
Posted in :
