The Real, Dark Secret About the HPE-Juniper Settlement Isn't About Lobbying
It is that the complaint is extremely dubious. Without a real competition problem to solve, Tunney Act review may put on a show but is bound to approve any settlement, including this one.
Welcome back to Competition on the Merits!
I had a few weeks off to attend to some personal business. Namely, I got married to an incredible woman and spent a few weeks in Italy with my family. But you did not come here for personal updates. You came for the antitrust. So let’s do that.
First, as always, please subscribe, upgrade to paid, and send to your friends.
So what did I miss? The antitrust story of the summer has been the various ups and downs of the DOJ’s challenge to the merger of Hewlett Packard Enterprises and Juniper Networks. It would be more accurate to say that the story has been the political economy of Trump antitrust – in particular, at the Antitrust Division. That larger story is what appears to be a battle between various forces within and close to the Trump administration that has played out in the context of the DOJ’s review of the HPE-Juniper merger, its lawsuit to challenge it, and its eventual consent decree.
The HPE- Juniper merger has a peculiar procedural history. The merger was proposed and investigated by Jonathan Kanter’s Antitrust Division during the Biden Administration but dropped into the lap of the new Trump DOJ to make the final decision. A merger that straddles an old and new administration is not that peculiar. But if we rewind back in time to January 2025, it was unclear what approach the Trump DOJ would take to mergers. The transaction was approved by major regulatory bodies around the world, it did not trigger any of the “Big Tech” indicators that would surely attract extra scrutiny from Trump antitrust authorities, the post-merger shares alleged (even in the DOJ Complaint, accepting its market definition) were relatively low. HPE-Juniper would argue the merger was necessary to intensify competition with the WLAN market leader (Cisco).
But none of that is particularly interesting from an antitrust perspective other than the fact that the deal sort of always had extra political and public salience because it straddled administrations and carried with it some extra symbolic value for those interested in reading the tea leaves about which way the administration would go on mergers. Q1 2025 was fully of mixed – or at least difficult to interpret – signals in so far as predicting the direction of Trump merger enforcement: keeping the Biden-Khan merger tax in the form of new and cumbersome HSR forms, keeping the Biden-Khan 2023 Merger Guidelines that increased antitrust risk for mundane M&A activity, but on the other hand, insisting upon the return of early termination.
By now, my very well informed COTM readers will know how the next chapters in the HPE-Juniper story went:
DOJ sues under the leadership of Acting Assistant Attorney General Omeed Assefi. Current Assistant Attorney General Gail Slater had not yet been confirmed by the Senate.
Post-Complaint, according to all sorts of public reporting, HPE appeals to the Antitrust Division, the White House, the Department of Defense, and the Department of Justice to make the case that the merger challenge should be resolved without litigation – and perhaps with some licensing-based remedy.
AAG Slater is confirmed. In a series of speeches from either Slater or Deputy Assistant Attorney General Bill Rinner, they offered up the standard view that merger remedies that do not solve competition problems will not be accepted and that structural remedies are preferred to behavioral ones (like licensing) because the latter often do not solve competition problems. Neither were earth-shattering propositions or positions. But in the backdrop of concerns the Trump FTC and DOJ would follow the Biden-Khan-Kanter hostility to M&A generally, they were impactful in the marketplace.
Slater and the Antitrust Division appeared not to like the licensing-based settlement offered by HPE-Juniper, but others in the Department of Justice (outside the Antitrust Division) believed a deal should get done to allow the merger to go forward. Public reporting has emphasized the role of various HPE consultants - including important and influential Trump allies – in making successful inroads with DOJ officials to get a settlement across the finish line.
There is ample evidence of conflict between Slater’s Antitrust Division and the larger DOJ in the form of the terminations of two of her key deputies, the aforementioned Bill Rinner and academic Roger Alford. Rinner is of particular interest because, while both were key figures in Slater’s front office, he was the deputy covering merger enforcement, is well-known and respected by the antitrust bar, had just offered up an important speech about merger enforcement, and would have been key to shaping Slater’s merger enforcement strategy and approach moving forward.
The settlement is now signed and delivered and has only Tunney Act review in front of it before it becomes final.
So that is the timeline. So why the big story and heavy breathing in the reporting? Dozens of stories have dropped about the terminations of Rinner and Alford and the success of HPE’s consultants in persuading the Department of Justice to override AAG Slater on the decision to accept HPE’s settlement. From the NY Times to Laura Loomer and Matt Stoller, antitrust news has read like Page 6 for the past week or so.
The NY Times reports a civil war:
Sohrab Ahmari frames it as a civil war inside MAGA between the populists and the lobbyists. Matt Stoller calls it an attempted coup:
You get the point. Critics and commentators have been absolutely obsessed with the palace intrigue surrounding the cast of characters around the Antitrust Division, the DOJ, the HPE consultants and so forth: Bondi! Mizelle! Slater! Davis! Rinner! Scandal! Oh my.
You can go read Ahmari or Loomer or Stoller for that palace intrigue. It is interesting stuff. Read it at your own risk. And with a few grains of salt. But I want to cover something else here. You know, antitrust.
And the next big antitrust question surrounding the HPE-Juniper deal is the Tunney Act (the Antitrust Penalties and Procedures Act if we are being formal), a well known statutory provision that requires the Antitrust Division to engage in a number of procedural steps ranging from publication of the proposed judgment to entering a competitive impact statement. Ultimately, a federal court must conclude that the Antitrust Division’s consent decree is in the public interest in order for it to take effect.
While the Tunney Act judicial review of DOJ Antitrust Division consents is well-known, it is getting some special treatment this week as the Hipster Antitrust left and some other observers call for its invocation to throw out the HPE-Juniper settlement.
Senator Warren too wrote a letter to Judge Pitts encouraging him to use the full scope of the Tunney Act’s various procedures to obtain evidence concerning the settlement and even reach a conclusion that it is not in the public interest.
Ahmari asks whether the Tunney Act will undermine the settlement:
Ditto Stoller, who says that the Judge Pitts should use the Tunney Act to undo the settlement and uncover the details associated with any lobbying effort associated with it:
That’s a lot of hopes and dreams being poured into the Tunney Act review process. I’m afraid those waiting for the Tunney Act to undo the merger settlement are barking up the wrong tree. And while I do not expect coverage of our antitrust institutions to stop sounding like Page 6 any time soon, COTM readers come up for something else. So let’s talk about what the Tunney Act review process actually is and what to expect.
The punchline: the secret weakness of the DOJ’s process here is not rumors of lobbying, it is not DOJ on DOJ crime, it is not petty fights within the Antitrust Division or between the Antitrust Division and folks in the Department of Justice or the White House. That is not to say those conflicts do not exist. They almost always show up on one case or another throughout an administration. And it is unfortunate to see the focus on our antitrust institutions turn from substance to politics. And it is not a good thing that the Division will not have the benefit of talented folks like Rinner and Alford. But if you have been watching for the past decade or so – please do not pretend to be “shocked, shocked to see that gambling is going on in here.”
The secret weakness of the DOJ’s position here has nothing to do with HPE’s and its consultants, political infighting at the Division or the DOJ, or anything else you are going to read about from Laura Loomer, Matt Stoller, CBS News, or Bloomberg.
The weakness is that the DOJ’s complaint alleging that the HPE-Juniper transaction violated Section 7 of the Clayton Act was not very good on the merits and was very likely to lose if litigated for failing to show that the merger was likely to substantially lessen competition in the WLAN market. That’s it. The complaint was not just weak sauce, it was historically weak for reasons I’ll explain below. And that matters for the Tunney Act. The Tunney Act process fundamentally inquires whether the settlement satisfies any competitive injury alleged in the Complaint. And when there is no competitive injury – even on the face of the Complaint – there is little work for the remedy embodied by the consent decree to do.
TL;DR: if there is no competitive problem to solve, the remedy does not have much work to do.
Such is the case here. And so while the Tunney Act process might play some role in revealing who talked to whom about the merger, critics’ demand for undoing the consent decree is going to go unsatisfied.
What the Tunney Act Process Is and Is Not
The Tunney Act requires judicial review of DOJ’s antitrust civil actions and a notice and comment period before any consent decree can be made final. It has a fun and interesting history, of course. The DOJ settled its litigation with International Telephone & Telegraph Corporation (“ITT”) in 1971 and required divestiture of some subsidiaries of the Hartford Fire insurance Company. In 1972, ITT contributed $400,000 to fund the Republican National Convention, leading to suspicions that the contributions that the settlement was the result of a quid pro quo unrelated to competition principles and contrary to the public interest. While there was no proof of any quid pro quo, the sequence of event and interest in them motivated calls for increased transparency into DOJ settlements. Thus, the Tunney Act was born in 1974.
But what does it actually do? The Tunney Act lays out a set of procedural requirements meant to bring “sunshine” to DOJ Antitrust settlements. Congress amended the Tunney Act in 2004 to be a bit more clear about the factors that courts “shall” consider during the review. But the review remains rather limited.
The Tunney Act requires the government to prepare a competitive impact statement to be filed simultaneously with the consent decree. The CIS is a familiar document for most antitrust lawyers. But it must spell out the nature of the proceeding, the consent order, the alternatives considered, and how the consent decree resolves competitive concerns laid out in the complaint.
The defendants must file (within 10 days) a description of all communications on its behalf concerning the consent decree “with any officer or employee of the United States,” other than communications by its counsel of record with the Attorney General or employees of the DOJ. It is certainly possible that this process identifies consultant communications that are not yet public. Fine. This is what folks like Stoller, Dayen, Ahmari, and Senator Warren are hoping for. They may even persuade Judge Pitts to be interested in this sort of stuff – who talked to who when about the merger?
The Tunney Act also requires the CIS and the proposed consent decree to be published in the Federal Register and open a 60 day comment period.
But while the public comments (now stoked with press coverage of rumored interference) may certainly focus on conversations adjacent to consideration of the merger — what can the court actually do?
The bottom line Tunney Act inquiry is for the court to determine “that the entry of such judgment is in the public interest.” No more. No less.
What does that mean? The 2004 amendments overturned an old line of cases that limited a judge’s grounds to reject a proposed consent to the situation when the consent “made a mockery of judicial power.” But courts have described the amendments as making “only minimal changes, and that this Court’s scope of review remains sharply proscribed by precedent and the nature of Tunney Act proceedings.”
The 2004 amendments make mandatory the court’s review of “the impact of the decree on competition in the relevant market.” The court may NOT reject a decree because other remedies might be preferable to those the government found satisfactory. The court may accept or reject, but not alter any settlement. Its options are inherently limited by the Tunney Act process. As a practical matter, if the court rejected a settlement, the DOJ and the parties can negotiate a new one, or HPE-Juniper could refuse and force the DOJ to litigate. Given the history here, DOJ litigation is highly unlikely. Judge Pitt’s rejection of the settlement would likely result in the merger closing without any settlement at all.
But the bottom line here I want you to focus on is that despite all of the various sunshine and disclosure elements of the Tunney Act – the substantive focus of the review is focused squarely “upon competition in the relevant market or markets” and on “the public generally and individuals alleging specific injury.”
That takes us out of Page 6 and palace intrigue and back to antitrust. As it should be. So the important insight here is that while Judge Pitts can order evidentiary hearings, or even permit a party to intervene, the substantive question at the core of Tunney Act review is about competitive harm in relevant markets. It is an antitrust question that should be understood in the context of the DOJ’s antitrust complaint against HPE-Juniper.
The Deep, Dark Secret of the DOJ’s Merger Challenge Isn’t About the Lobbying It is the Weak Complaint
Once one understands that the actual substantive focus of the Tunney Act proceeding sounds in antitrust and not politics, the substance of the DOJ complaint against HPE-Juniper takes center stage. The problem with HPE-Juniper as an antitrust merger challenge under Section 7 of the Clayton Act has always been that the Complaint was weak on its face.
Now I do not know if the politics and odd procedural history here play a part in that. The case sat around with the Kanter-Biden DOJ and was passed off to Acting AAG Omeed Assefi who had to make a “litigate or settle” decision within a very short time frame. But what was released as a complaint was simply not very good by professional antitrust standards. Not was it very convincing. This is not the place to do a wholesale analysis of the merger and its competitive effects but I want to point out some highlights here to help you understand the relative weakness of the complaint.
Here it is. All 21 pages of it. It reads, quite frankly, like it was rushed out the door. And it may have been.
There is an old bit of antitrust conventional wisdom that says there are three parts of an antitrust merger litigation: economic evidence, customer complaints, and hot documents. Cases are very likely to win if they have 3 for 3, but you can get away with going 2 for 3. Cases with only 1 of the 3 or none of the above are very likely to lose. I forgot where I first heard that – but I’m going to credit it to Tim Muris, because it sounds like something he would have said and I learned a lot of things by watching him and reading him.
Now of those three foundational pieces to antitrust litigation – economic evidence is king. But in the merger context we interpret “economic” evidence a bit loosely. Why? Because the Clayton Act has what antitrust lawyers describe as “the structural presumption,” or “the PNB presumption,” after the holding of US v. Philadelphia National Bank (1963) in which the court declared that a post-merger market share of greater than 30% was sufficient to shift the burden of production in a merger challenge. And so in the merger context, while the government can also bring “real” economic evidence to bear, and often does, it can also rely on the “PNB presumption” to do the economic work for it. If the FTC or DOJ can show post-merger shares >30%, the merger is legally presumed to substantially lessen competition.
I’ve written a LOT about the PNB presumption. I don’t think it is very good law. And I think it is horrible economics. The basic problem is, as Judge Ginsburg and I have explained:
The problem for today’s courts in applying this semicentenary standard is that the field of industrial organization has long since moved beyond the structural presumption upon which the standard is based. The point is not that 30 percent is an outdated threshold above which to presume adverse effects upon competition; rather, it is that market structure is an inappropriate starting point for the analysis of likely competitive effects. Market structure and competitive effects are not systematically correlated. Nor does the rebuttable nature of the 30 percent presumption reduce it to a mere annoyance, an exercise the clutters up litigation but is soon enough dispatched by the defendant’s showing; the practical effect of beginning the analysis of a merger with an essentially irrelevant presumption is to shift the burden of proof from the plaintiff, where it rightfully resides, to the defendant, as though the law prohibited all mergers except those that could be proved acceptable by their proponents.
But it is the law. The basic problem is that the legal presumption embodied by PNB requires untrue economic beliefs. That is, a merger that pushes post-merger shares >30% does not make a merger more likely than not to produce competitive harm. But that false statement is exactly what the presumption embeds into the Clayton Act. What is worse is that as a practical matter once the PNB presumption is invoked in court it is very hard to come back from for defendants, even if the merger is procompetitive. For better or worse, the PNB presumption plays an outsized role in modern merger litigation because so many mergers where it might be invoked are procompetitive or competitive neutral but the presumption is used to render them presumptively unlawful.
You might expect that agencies would use this power presumption to their advantage in modern merger litigation. You would be right. Most modern horizontal mergers that attract any scrutiny at all from the government involve post-merger shares above the PNB level. Even Lina Khan did not bring a single horizontal merger case in which the complaint alleged shares below the PNB threshold level of 30%. You have to go back to the 1960s to find such a horizontal merger brought by the federal government. Yes, the 1960s. The era in which Justice Potter Stewart noted in his dissent in US v. Von’s Grocery (1966) that the only consistency with Section 7 of the Clayton Act “is that the government always wins.”
What does this have to do with HPE-Juniper? You guessed it, the post-merger shares do not even trigger the PNB presumption. Let me show you the Complaint itself:
So Cisco is the market leader. And HPE and Juniper WITH Cisco add to up 70 percent of the market. And no doubt, that meets the FTC/ DOJ Merger Guidelines standard for what mergers courts “should” (in the agencies’ view) deem presumptively lawful. But the Merger Guidelines are not law. PNB is Supreme Court doctrine and the agencies never fail to cite it when the post-merger share is above 30 for precisely that reason. Here is the operative paragraph of the DOJ Complaint — notice anything missing?
That’s right. The Complaint relies upon IDC estimates of shares for wireless access points and asserts that shares of “U.S. enterprise-grade WLAN market are roughly in line with their shares of the U.S. market for access points alone.”
So what do IDC data say about those shares?
You can do the math. Most estimates of HPE and Juniper combined shares end up around 20-25%, short of the all important 30% PNB presumption. And that is even assuming the DOJ’s market definition prevails in court – a dubious assumption as it ignores a broader set of players competing for WLAN business as customers shift to AI and cloud-based strategies. But the real point is that the DOJ complaint, on its face, read in the light most favorable to it, does not trigger the PNB presumption.
This is a really big deal. I cannot think of an FTC or DOJ complaint with post-merger shares this low. Not even from Lina Khan or Jonathan Kanter, much less agencies any time from the 1970s to current. You have to go back to the 1960s to find such a case. No wonder the Kanter DOJ passed this one on to let someone else make the final call.
On the critical factor of economic evidence that the merger would be anticompetitive, the complaint fails miserably. And it fails in a historic manner. There are other arguments to be made here about merger efficiencies, market definition, and whether the merger would intensify competition between HPE and Cisco. But even ignoring those additional arguments the truth of the matter is that the complaint just does not cut it. It does not establish that the merger is presumptively anticompetitive or unlawful.
I can already hear some of you saying “but the complaint says it is presumptively anticompetitive!” And it does. But again, dear COTM readers, this is a classic error made by law students and antitrust policy folks who do not understand Section 7 of the Clayton Act. The agencies have LONG promulgated HHI thresholds to encourage courts to deem mergers above those thresholds as likely to be anticompetitive. Sometimes those thresholds persuade. Sometimes they do not. But they are not the law and never have been. Only the PNB presumption, not the merger guidelines, have the force of law. That is why the FTC and DOJ rely upon the PNB presumption so heavily. It is their greatest weapon in Section 7 enforcement. Without it they often lose.
But wait, Josh, didn’t you just tell us that the agencies haven’t brought a case under the PNB threshold since the 1960s. Yes and no. They have not brought any horizontal mergers under the PNB threshold. But you know where the FTC and DOJ litigate without the benefit of the PNB presumption? Vertical mergers. Somebody remind me how the agency record looks in those cases!
You get the point. The HPE-Juniper merger is an outlier. The post-merger shares are so low as to be the first litigated horizontal merger challenge to fall short of the PNB presumption in 50 years.
Now, as I said, there are other factors. Things like hot documents and customer complaints. The complaint itself goes through a handful of business documents talking about close competition between HPE and Juniper’s MIST product. But nearly every litigated merger case has a handful of those. Nothing special there. And there was significant reporting that the DOJ did not have any complaining witnesses ready to testify. Though one cannot be sure that is accurate. So while we cannot be sure the other factors would also fall short, there is nothing in the complaint to suggest they would carry the DOJ to victory given the dramatic shortcoming on economic evidence and post-merger shares.
Let’s bring this all back to the Tunney Act, shall we?
What is Judge Pitts Going to Do in the Tunney Act Review?
We might well see Judge Pitts, given the public attention paid from Senator Warren and others, the high likelihood of an active comment period, call for some sort of evidentiary hearing or inquiry to further examine the decision-making process at DOJ in approving the settlement. I would not be surprised if there was some evaluation of that process.
But at the end of the day, Judge Pitts’ task is to examine the competitive impact of the proposed settlement and the competitive effects in any relevant markets. In other words, Judge Pitts is charged by the Tunney Act to ask the question if the proposed settlement solves the competition problem in the relevant market as described in the Complaint. But the Complaint articulates one of the weakest theories of competitive harm in Section 7 litigation in the modern era, if not the weakest. The Complaint does not even trigger the PNB presumption, a relic of the pre-economic era of antitrust during which the government always won.
It is hard to imagine finding the proposed settlement lacking when it comes to solving a competition problem the government could not bother to identify with any precision or persuasive force in its Complaint. While the political process that led from litigation to settlement has gotten all of the attention, the process that led from investigation to a historically poor complaint has received little attention. The latter is beyond the scope of the Tunney Act.
So yes, we are going to see a lot more press about the palace intrigue at DOJ, the White House, and within the Antitrust Division. Fair enough. Some of that is interesting and the transparency is probably good for the institutions in the long run.
But when it comes to the Tunney Act process and the ultimate fate of the HPE-Juniper settlement – it is going to be approved. And it is going to be approved not only because the Tunney Act procedures are limited in scope, but because they ultimately ask a court to evaluate the proposed remedy relative to the harm identified in the Complaint. And here, COTM readers, the DOJ complaint stacks up poorly against others over the past 50 or so years. Judge Pitts will have little choice but to approve the settlement – but that will ultimately be because there likely was never a competitive problem in the first place.
See you next week and maybe sooner.