Antitrust Law Daily Wrap Up, ANTITRUST—N.D. Cal.: CRT antitrust litigation converging on $1 billion damages for two remaining Irico defendants, (Mar 12, 2026)

By Justin Marcus Smith, J.D.

The court explained why and how it would adopt a single damages model where three competing expert models produced different results.

In the long-running multidistrict antitrust litigation about cathode ray tubes (CRTs), the federal district court in Oakland, California, selected one of the competing expert damages models, applied modifications, and ordered the parties to meet and confer to calculate damages. A modification of the dynamic model of the direct purchasers’ expert, which included separate regressions for CPTs and CDTs, presented the most reasonable estimate of the overcharges in the case. The court held the estimate should apply to both the DPP and IPP classes. The court said it was not persuaded it should add annual conduct variables in lieu of dividing the conspiracy period into only two periods. The court also ruled firm-specific data would provide a better estimate of missing sales data. The court ordered the parties to meet and confer for final calculations and to resolve the deferred issues of whether damages should be apportioned between the IPPs and DPPs and whether treble damages are warranted (In Re: Cathode Ray Tube (Crt) Antitrust Litigation, No. 4:07-cv-05944-JST (N.D. Cal. Mar. 11, 2026)).

Background. This multidistrict litigation (MDL) arose from an alleged conspiracy to fix the prices of CRTs used in the manufacture of televisions and computer monitors. Over the years, the court approved various settlements that encompassed indirect purchaser plaintiffs (IPPs), including with certain exceptions for an Omitted Repealer State Subclass (ORS Subclass) and a Non-Repealer State Subclass (NRS Subclass) to remove potential conflicts of interest (see Antitrust, Aug. 31, 2020). Some excluded purchasers were unsuccessful in a bid for certiorari (see Antitrust, Jun. 13, 2022). The court granted a motion for class certification by direct purchaser plaintiffs (DPPs) in 2022.

The only remaining defendants in early 2026 were Irico Group Corp. and Irico Display Devices Co., Ltd (collectively, Irico). However, after adopting a Special Master’s recommendation to grant terminating sanctions for Irico discovery misconduct, the court ordered that Irico’s answers and affirmative defenses be stricken and that the clerk enter default against Irico.

The court then held a three-day evidentiary hearing to ascertain damages on judgment by default. The court reviewed how it could consider probable or inferential proof in determining damages even though the resulting award might be an approximation. The IPPs, DPPs, and Irico each presented an expert witness.

The parties did not challenge counterparty expert qualifications, and the court concluded all three were qualified experts. The parties also agreed that the IPP and DPP class periods ran through November 25, 2007. The court previously ruled, over Irico objection, that the plaintiffs could recover damages from Irico dating back to March 1, 1995. The parties also stipulated there were two types of CRTs: color display tubes (CDTs) and color picture tubes (CPTs). CDTs were typically used as computers monitors. CPTs were typically used in color televisions.

IPP expert. The IPP expert developed a static regression model to estimate overcharges for CRTs. She concluded CDT prices would have been 22 percent lower and CPT prices would have been 9 percent lower for most of the relevant period were it not for the conspiracy. Irico did not oppose her conclusion that all of these overcharges were passed on to IPPs. The IPP expert arrived at nearly $2.7 billion in damages for the IPP class. Display glass was the most important cost input for CRTs.

The Irico expert proposed to modify the IPP expert’s model by adding several supply and demand variables, like shipping costs, and by using separate overcharges for 1995 and 1996 to account for the increased demand for desktop computers due to the introduction of Windows 95. The IPP expert did not accept these modifications.

DPP expert. The DPP expert, unlike the IPP expert, created a single regression model for both CPTs and CDTs. This model also differed in that it was dynamic, rather than static, and included a control for the effect of past prices. The DPP expert calculated cartel effects in two time periods, with Q3 2006 as the start of the second period based on an apparent change in conduct, a change in meeting frequency. He estimated CPT prices were elevated on average by 6.1 percent during the class period and CDT prices were elevated on average by 10.7 percent. He calculated damages of nearly $1.1 billion.

The Irico expert criticized the DPP model for using a single regression instead of separate CPT and CDT regressions and for breaking the class period into two periods instead of by calendar year. The DPP expert rejected both criticisms. The proposed modifications would have yielded an average CPT overcharge estimate of 1.5 percent and an average CDT overcharge estimate of 7.0 percent.

The Irico expert also criticized the DPP expert method for determining sales volume. The DPP expert accepted it would have been more reasonable to assume the DPP class share was the same, whether the data identified a customer or not, and that it would have been reasonable to use a separate dataset that indicated a cartel member’s sales to the U.S. instead of inferring sales from customer names. These two changes reduced the DPP expert’s damages estimate to about $1.0 billion under the DPP expert model and only $423 million under the Irico expert’s model. Another modification to identify cartel member sales data would reduce the damages calculations to $960 million and $391 million, respectively.

Discussion. The court decided it would adopt a single model because the three models produced different results, the case law supported adoption of a single model, and no party argued that doing so would be improper. The court considered the testimony of all three experts together to determine a reasonable estimate of overcharges caused by the conspiracy.

The court ultimately found that a modification of the DPP expert’s dynamic model to include separate regressions for CPTs and CDTs presented the most reasonable estimate of the overcharges in the case, and that this estimate should apply to both the DPP and IPP classes. The court rejected the Irico expert’s suggestion of adding annual conduct variables, but it adopted three proposed modifications, two of which the DPP expert accepted, about the volume of commerce calculations as to the DPP class.

The court considered what the experts had to say, or declined to say, about each other’s models. Whether a model was easier or more straightforward to modify had no bearing on reliability. The court looked to the IPP and DPP models themselves because there was no meaningful expert testimony comparing them. The court considered the two main differences: (1) whether the model should be static or dynamic; and (2) whether there should be single or separate regressions for CDTs and CPTs.

The court concluded that the effect of target prices went beyond the immediate quarter. The court therefore found it appropriate to adopt a dynamic model to estimate the overcharge effects.

As for single or separate regressions for CDTs and CPTs, the DPP expert’s own work, among other circumstances, revealed differences between CDTs and CPTs. The court was ultimately not persuaded that the DPP expert’s use of a single regression with different effects for CDTs and CPTs from LCDs was sufficient to model the behavior of the two different product types. The court concluded instead that the overcharge model should include separate regressions for CDTs and CPTs.

Having determined that the overcharge model should be dynamic, and that CDTs and CPTs called for separate regressions, the court next considered which model best satisfied both criteria. No one proposed modifying the IPP expert model to include a lagged price variable or why it would otherwise account for the changes supporting adoption of a dynamic model. The court therefore chose the DPP expert model as its starting point, as modified to run separate regressions for CDTs and CPTs.

As for the Irico expert’s remaining proposed modifications to the DPP expert model, the court was not persuaded it should add annual conduct variables in lieu of dividing the conspiracy period into only two periods. The IPP and DPP experts both agreed there was no economic justification, for example, for estimating conduct effect separately for 12 calendar years to capture variability. Disaggregating the conspiracy effect by calendar year, as the Irico expert proposed, would be arbitrary. Among several other circumstances, the court also found the DPP expert’s testimony persuasive that adding annual variables made estimation of the conduct “imprecise and unreliable.” The court therefore rejected the Irico expert’s proposal to add annual conduct variables to the DPP expert’s model.

The last dispute was over how to estimate missing sales data for three manufacturers. Given the choice between relying on firm-specific data from different years, or industry-wide data from the same years, the court chose firm-specific data as a better estimate of missing sales data.

The court ordered the parties to meet and confer to attempt to reach agreement on a final damages calculation for each class, as well as the two issues the IPPs and Irico agreed to defer: whether damages should be apportioned between the IPPs and DPPs and whether treble damages are warranted. If the calculations adopted by the order result in damages to the DPP class greater than the $1.011 billion the DPPS requested, and the DPPs seek the higher amount, then the court ordered the DPPs and Irico to meet and confer about whether the DPPs are limited to a damages award of $1.011 billion.

The court also ordered the parties to file a joint case management statement setting forth their agreed or respective proposals about any further necessary proceedings, including potential referral to a magistrate judge for settlement.

The Case is No. 4:07-cv-05944-JST.

Judge: Tigar, J.

Attorneys: Guido Saveri (Saveri & Saveri, Inc.) for Crago, Inc. Eliot A. Adelson (Morrison & Foerster, LLP) for Hitachi America, Ltd.

Companies: Crago, Inc.; Hitachi America, Ltd.

Cases: Antitrust CaliforniaNews