may not be that large, especially if one does not internalize the reputational spillover to ones peers. This is ournalism profession ed an ethical sanction against advertising bias, because the returns to favoring advertisers might otherwise be high II. Data Our tests for advertising bias require data on advertising expenditures, media mentions, and mutual fund characteristics. Data on monthly mutual fund advertising expenditures by publication and fund family were purchased from Competitive Media Research(CMR). According to these data, the mutual fund in- dustry's annual advertising ditures averaged approximately $307 million during our 1996-2002 sample period, with $80 million(26%) going to national newspapers and $119 million(39%) going to consumer The publications we study include five of the six top recipients of mutual fund advertising between 1998 and 2002: the Wall Street Journal($48.5 million per year), Money($22.1 million), New York Times(S14.0 million), Kiplinger's Personal Finance($12.2 million), and SmartMoney($8.7 million).(We attempted to gather media mentions for Mutual Funds( S14.0 million)but were unable to access its content electronically. In total, the publications in our sample account for approximately 45 percent of the mutual fund industrys advertising expenditures. Naturally, mutual fund advertising is a more important source of revenue for the personal finance publications than the national newspapers. Whereas mutual fund advertising accounts for 3.8 percent of advertising revenues at the Wall Street Journal and 1.1 percent at the New York Times, it accounts for 15 percent at Money, 16 percent at SmartMoney, and 28 percent at Kiplinger's. We also gather edia mentions from Consumer Reports, which does not accept advertisin The media mention data vary across publications and are summarized in Table I. Since these data had to be hand collected, for several publications we restrict attention to particular articles or columns. For the New York Times, we track funds mentioned in a column from the Sunday Business section titled"Investing With. The column spotlights a particular fund, interviewing fund managers and providing details such For print publications, CMR tracks the size of each advertisment and estimates a dollar cost for the advertisement based on the publication's quoted advertising rates and any likely discount. Comparing CMR's estimates of total print advertising revenue for the New York Times and Wall Street Journal to figures reported in the parent companies'10K filings, the CMr estimates often differ from the actual revenues by less than 10 percentmay not be that large, especially if one does not internalize the reputational spillover to ones peers. This is precisely why the journalism profession developed an ethical sanction against advertising bias, because the returns to favoring advertisers might otherwise be high. II. Data Our tests for advertising bias require data on advertising expenditures, media mentions, and mutual fund characteristics. Data on monthly mutual fund advertising expenditures by publication and fund family were purchased from Competitive Media Research (CMR).5 According to these data, the mutual fund in￾dustry’s annual advertising expenditures averaged approximately $307 million during our 1996-2002 sample period, with $80 million (26%) going to national newspapers and $119 million (39%) going to consumer magazines. The publications we study include five of the six top recipients of mutual fund advertising between 1998 and 2002: the Wall Street Journal ($48.5 million per year), Money ($22.1 million), New York Times ($14.0 million), Kiplinger’s Personal Finance ($12.2 million), and SmartMoney ($8.7 million). (We attempted to gather media mentions for Mutual Funds ($14.0 million) but were unable to access its content electronically.) In total, the publications in our sample account for approximately 45 percent of the mutual fund industry’s advertising expenditures. Naturally, mutual fund advertising is a more important source of revenue for the personal finance publications than the national newspapers. Whereas mutual fund advertising accounts for 3.8 percent of advertising revenues at the Wall Street Journal and 1.1 percent at the New York Times, it accounts for 15 percent at Money, 16 percent at SmartMoney, and 28 percent at Kiplinger’s. We also gather media mentions from Consumer Reports, which does not accept advertising. The media mention data vary across publications and are summarized in Table I. Since these data had to be hand collected, for several publications we restrict attention to particular articles or columns. For the New York Times, we track funds mentioned in a column from the Sunday Business section titled “Investing With.” The column spotlights a particular fund, interviewing fund managers and providing details such 5For print publications, CMR tracks the size of each advertisment and estimates a dollar cost for the advertisement based on the publication’s quoted advertising rates and any likely discount. Comparing CMR’s estimates of total print advertising revenue for the New York Times and Wall Street Journal to figures reported in the parent companies’ 10K filings, the CMR estimates often differ from the actual revenues by less than 10 percent. 3