AT&T Lawyer Grills Star U.S. Witness on Data Tinkering in Study

(Bloomberg) -- The U.S. government’s star witness in the trial over AT&T Inc.’s planned takeover of Time Warner Inc. admitted he erred in earlier testimony about data used to predict that the merger will raise pay-TV prices for consumers by hundreds of millions of dollars a year.

Professor Carl Shapiro of the University of California at Berkeley, whose report forms the backbone of the Justice Department’s lawsuit to block the $85 billion merger, has repeatedly clashed with defense lawyers over the accuracy of the data used to predict the impact of the deal on competitors and customers.

Shapiro, who returned to court Tuesday as the last witness in the trial, was grilled about a data point he said was crucial to his findings and was dramatically contradicted for the second time.

He said a hypothetical blackout of Time Warner’s Turner Broadcasting content at a competing pay-TV service would help AT&T’s DirecTV unit retain at least 10 percent of the customers who otherwise would have considered switching. The prediction is one of many the U.S. has used to claim the merger will give the combined AT&T-Time Warner an unfair advantage.

AT&T and Time Warner claimed the 10 percent figure was the result of last-minute tinkering designed to bolster the government’s case. Shapiro testified earlier in the antitrust trial that the figure -- taken from a consulting firm’s study on behalf of Charter Communications Inc. -- hadn’t been changed.

But on cross-examination by Daniel Petrocelli, the lead attorney for AT&T and Time Warner, Shapiro was shown evidence that the 10 percent prediction was originally 6 percent, and that the “final” number had been changed just around the time Charter was working with the Justice Department to block the merger.

Data Change

“Yeah, I made a mistake,” Shapiro said before clashing with Petrocelli on a number of alleged inconsistencies in the data he used. At one point in the exchange, U.S. District Judge Richard Leon interrupted Shapiro’s defense of the data and asked him not to editorialize -- a bad sign for the Justice Department which has already struggled in the case.

Shapiro’s use of the larger 10 percent number led to the conclusion that the AT&T-Time Warner deal would lead to higher prices for consumers. The lower figure would have resulted in no change, which he conceded on the stand. The academic said he stands by his report, noting that the dispute was over the low end of a range of numbers rather one set figure.

Petrocelli also showed Shapiro evidence that he’d used 2016 data in his report, when information from 2017 would have resulted in projections of smaller price increases. Again, Shapiro stood by his decision and said he was being criticized for being “conservative” in his calculations. Shapiro complimented Petrocelli for his “flare” in presenting evidence but said he stood by his calculations.

Lawyers for AT&T and Time Warner have portrayed the study used by Shapiro as unreliable because of an earlier last-minute change in which the predicted number of subscribers who would drop Charter during a blackout of Turner channels such as CNN and TBS was arbitrarily increased to 9 percent from 5 percent after the study was completed. In that case as well, the price increase in Shapiro’s model would have been zero if he’d used the original number.

To contact the reporter on this story: Erik Larson in New York at elarson4@bloomberg.net.

To contact the editors responsible for this story: David Glovin at dglovin@bloomberg.net, Rob Golum, John Harney

©2018 Bloomberg L.P.