Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
Using data consisting of domestic and foreign automobiles over the period 1985‐1990, the hypothesis that advertising serves as a signal of higher quality is empirically tested. This is accomplished by examining how advertising levels vary with a quality measure that is not observable at the time of purchase but becomes available subsequent to a model's release. Above average quality results in expenditures that are 15% higher than average quality outlays. Lending further support to the signalling hypothesis, the positive advertising–quality relationship holds strongest when a model significantly improves in quality relative to the previous year's model.