Relative Prices, Concentration, And Money Growth: Comment
评论Chappell和Addison关于行业集中度与价格行为关系的论文，指出其方法缺陷，并修正后支持管理价格假说。
In a recent article in this Review (1983), Henry Chappell and John Addison provide new evidence on the thesis that links an industry's price behavior over the business cycle to its structural features, most notably, some measure(s) of the monopoly power possessed by firms in that industry. The approach taken by Chappell and Addison (C-A) is both sensible and novel. Essentially, they regress aggregate inflation measures for different groups of manufacturing industries on a distributed lag in monetary growth. The industry groups are distinguished by their degree of concentration-low, medium, or high-and the validity of the administered-pricing thesis is then tested by comparing the time pattern of response in each group to monetary impulses. Chappell-Addison conclude that this test reveals no clear response differences between the sectors of varying concentration, and hence reject the hypothesis of concentrationrelated administered prices. The purpose of this comment is to show that the C-A conclusions are overly strong. Their focus on the mean lag as the sole basis of interindustry comparisons obscures important differences in the pricing performance of the lowand high-concentrated sectors. More fundamentally, the C-A test is risspecified in that it fails to differentiate between the expected and unexpected components of money growth as suggested by Robert Barro (1977). When the analysis allows for these modifications, the C-A data reveal considerable support for the administered-pricing hypothesis. 1. The Chappell-Addison Model