Behavior of prices in periods of low and high demand .

Abstract

Aggregate prices fail to fluctuate significantly at both seasonal and business cycle frequencies. In order to look at the issue of cyclical price rigidity at a disaggregated level, I collected daily price information for several consumer products during the months surrounding Christmas in 1987. Patterns in the survey data suggest that prices of individual goods are not nearly as inflexible as seasonal analysis of the overall CPI might indicate. It appears that prices of various goods respond differently in high demand states (some increase, some decrease) so that fluctuating price patterns wash out in the aggregate. Evidence from the survey data suggests that price adjustments do not always facilitate market clearing. The data contain examples of rationing. I also found evidence of retail stores lowering price during peak demand periods (specifically before Christmas and on weekends). I developed a very simple model of shopping in which stores reduce price in order to induce demand . The key feature of the model is that because consumers do more bulk shopping during peak demand , individual stores face more elastic demand . Evidence of price fluctuations responding to elasticity changes is only consistent with a world in which firms possess some monopoly power. Clearly prices are not the sole market clearing mechanism. Since altering price is not costless, Carlton (1987) and Okun (1981) contend that firms use long-term knowledge of the customer rather than price to more efficiently allocate goods during peak demand . How such a relationship is established in the retail market is unclear. At least in the retail goods market the explanation for price rigidity appears more likely to be some sort of market coordination failure. Kahneman, Knetsch, and Thaler (1986) suggest that consumers prefer stores that appear to treat the customer "fairly", and freely fluctuating prices responding to changing market conditions are perceived to be unfair. Such an explanation seems to be the most consistent with my data and the existence of stockouts. More formal modeling incorporating the theme of "fairness" in future work may prove fruitful in understanding cyclical price fluctuation.