Sales: Tests of theories on causality and timing

Modern theories of sales make conflicting predictions about the temporal pattern of sales, which we test using grocery scanner data. We examine both frozen orange juice, which consumers can store, and refrigerated orange juice, which is more perishable, to determine what role-if any-durability plays in the pattern of sales. We examine (1) whether sales are more frequent for national brands than for private label brands; (2) whether manufacturers, retailers, or others determine sales; (3) the distribution of prices; (4) whether sales and their effects differ for durable (frozen) or nondurable (refrigerated) products; and (5) the temporal ordering of sales. We use correlations, runs tests, probit regressions, and vector-autoregressive analyses to test our hypotheses. Some of our more striking findings are that retailers rather than manufacturers determine sales, private labels have sales as often as national brands, and that a sale of a national brand is more likely to "cause" sales of other products than is a sale of a private label product.