On October 1, 2026, a federal district court for the Northern District of Illinois (Eastern Division) certified a class of rural telecommunications carriers (the “Class”) in a case against T-Mobile, relying in part on the impact and damages models developed by Dr. Hal Singer. The Class asserts that T-Mobile failed to maintain its network quality, allowing call competition rates to decline, and inserted fake ringtones to mask the alleged network infirmities (the “Challenged Conduct”), in violation of sections 201 and 202 of the Communications Act.

This decision marks the 18th occasion on which a federal or state court certified a class in reliance of Dr. Singer’s models.

Dr. Singer developed three models to show impact and estimate damages to the Class owing to the Challenged Conduct. The first model was a “Contractual Obligation Model,” which held T-Mobile to the performance standards in its contracts with Class members. The second model was a “Call Completion Model,” which used a during-after regression to isolate the impact of the Challenged Conduct on call completion rates. The third model was a “Fake Ringtone Model,” which identifies certain unanswered calls that would have been answered but for the fake ringtones.

Regarding Dr. Singer’s during-after regression model, the court explained why it was admissible:

“Ultimately, this criticism does not belong in a Daubert motion, particularly when the relevance of these variables is disputed. [Dkt. 598 at 10.] See also Manpower, 732 F.3d at 808 (“selection of the variables … goes to the probative weight of the analysis rather than to its admissibility”). Indeed, T-Mobile’s rebuttal expert never even tested hypothesized control variables, id., but instead applied Singer’s model to a “placebo” group that, in T-Mobile’s opinion, should have showed no during-after effect, but did. [Dkt. 616 at 9–10.] T-Mobile says that this “proves the point,” id., but the court thinks it quite clearly proves a different one: that T-Mobile’s criticism is with the persuasiveness of his conclusions, not the reliability of his method. Once again, this is not an argument for exclusion. See Daubert, 509 U.S. at 595 (“[F]ocus, of course, must be solely on principles and methodology, not on the conclusions that they generate.”).”

The court concludes that “Using the available [call detail records] data, Singer established a mechanism to measure the amount of revenue lost based on false ringtones, and for that matter as it relates to other counts, calls not completing, and substandard [network effectiveness ratio].”

The class certification order can be read in its entirety here.

Dr. Singer was supported by Ted Tatos and Logan Summerlin.