I compare the outcome when firms semicollude on advertising to the outcome in the Grossman and Shapiro (1984) model of informative advertising. I show that advertising is lower but prices and profits are higher under semicollusion on advertising. I also show that semicollusion on advertising is detrimental to welfare. Although firms earn higher profits when colluding on advertising, fewer consumers are informed, and as a result, welfare is lower. Compared to semicollusion on price, semicollusion on advertising is not always less profitable. Hence I lend theoretical support to empirical studies that find evidence of collusion on advertising rather than price.
by 