Aug 04, 2025
On July 24, The Capitol Forum held a conference call with Jeremy Sandford of Econic Partners for a conversation on his recent paper on the FTC’s challenge to the 2015 Steris/Synergy merger. The full transcript, which has been modified slightly, can be found below.
NATE SODERSTROM: Hi, good morning and welcome to our conference call on “Steris/Synergy a Decade Out.” I’m Nate Soderstrom. I’m a senior editor at The Capitol Forum.
With us today is Jeremy Sandford, who is a partner at Econic Partners. Prior to private practice, Jeremy spent seven years at the Bureau of Economics at the Federal Trade Commission, where he was the lead economist on a number of matters, including the 2015 investigation into Steris’s acquisition of Synergy Health, which gets us to our topic today.
Jeremy earlier this month published a paper entitled, “Steris Synergy a Decade Out: Retrospectively Assessing this Century’s First Litigated Potential Competition Matter,” which I would certainly encourage folks to read it, regardless of where you fall on the enforcement spectrum. I think everyone agrees we need more merger retrospectives. So, just from that perspective alone, tremendously important work.
So, with that intro, Jeremy, thank you for the paper and perhaps more importantly, thank you for speaking with us today. Very excited to have you.
JEREMY SANDFORD: Yeah, happy to be here.
NATE SODERSTROM: Before we get started, I wanted to mention that if you are on the call and would like to submit questions, type them into the questions pane of the control panel. We will collect them throughout the call and address them at the end of today’s conversation, if they are good questions.
So, to set the stage, Steris, in October, 2014, agreed to acquire a firm called Synergy Health. The deal was valued at $1.9 billion. The deal would combine two of the world’s three largest providers of sterilization services. If you’re a large manufacturer of medical devices or labware or something along those lines, you want to sterilize those products before use. You do not want to implant a medical device teeming with bacteria into somebody’s body, which is where these sterilization providers come in, certainly provide an important service.
The deal attracted an in-depth investigation. The Commission, in late May of 2015, authorized a lawsuit challenging the transaction. The three-day evidentiary hearing came in the Northern District of Ohio before Judge Aaron Polster in late August of 2015. On September 24th, 2015, he issued a decision denying the agency’s motion for a preliminary injunction. The parties closed a month or so afterward. And here we are.
Jeremy, could you walk us at a high level through why the FTC brought this case? What were the facts that gave the Commission reason to believe the merger might substantially lessen competition?
JEREMY SANDFORD: Sure, and if I can give a disclaimer at the beginning. Obviously, my views are my own. They’re not those of the Econic Partners, my employer. And they’re certainly not those of the FTC, my former employer, or any present or past commissioner. While I did work on the case as a government employee, I was not compensated for writing the paper.
So, at issue in this case, there are various ways to sterilize medical products. At issue in this case was radiation sterilization. You can also sterilize products with poison gas, but that was separate. That’s a separate group of products that are suited to that method. Steris was one of two large providers in the U.S. of gamma sterilization. And gamma means you put the products in a room with cobalt-60, which is a manufactured in a nuclear reactor. And through the natural process of radioactive decay, it emits gamma rays that bombard proteins and any life forms in the boxes and kills them or renders them sterile.
Synergy was in the process of evaluating a U.S. entry using X-ray sterilization, which is very similar to gamma. It uses an electrically powered accelerator to produce X-rays that are similar to gamma rays. They’re slightly higher energy, but they have the same effect. They bombard proteins of any life forms that may exist in the boxes containing the products and kill them or render them sterile.
And X-ray is a close substitute to gamma. That was true at the time based on forward-looking evidence and it’s even more true now based on looking backwards. It has some advantages over gamma that Synergy documents indicated that Synergy might be able to leverage to sell X-ray to current gamma customers and take business away from Steris. And those advantages are it’s lower cost. It’s electrically powered. There’s no radioactive waste. Machines can be turned on and off.
The supply of Cobalt-60 was in question at the time. One of the two main suppliers was a Russian firm, potentially subject to U.S. sanctions. Another of the two large suppliers was a Canadian firm that had just been purchased by the other large U.S. gamma supplier and there were concerns about foreclosure. So, there were concerns about availability of gamma.
Because X-ray has even higher energy than gamma, you can use it to sterilize entire pallets of packages without having to depalletize the boxes and run them individually through a gamma scanner. And so, there’s less handling, less risk of damage from handling. And so, it did look like there was some potential for a provider of X-ray to take on the incumbent gamma providers and compete successfully. And it looked like Synergy was pretty far along in doing that.
Synergy had purchased an X-ray plant in Europe in 2012. So, about three years before they were sued by the FTC. And beginning in about 2013, they began to look at bringing X-ray into the U.S. They hired a U.S. CEO and tasked him with leading the X-ray expansion effort into the U.S. They did all sorts of market work. They were talking to customers and this got as far as a presentation to Synergy’s management board where they presented the X-ray strategy.
The board approved the strategy. The Synergy board of directors met the next day and approved down payments for the X-ray machines and sort of sent Synergy management out to go implement the plan. So, finding sites, seeking tax incentives, working on getting costs down and driving up customer interest.
And so, all of that, going by the FTC’s complaint, indicated to the FTC that it looked like there was a good chance that Synergy was likely to enter the U.S. with X-ray and take on Steris and that that competition may be lost with the transaction. Because the post-merger Steris may have different incentives to pursue Synergy’s X-ray plan than the pre-merger Synergy did.
NATE SODERSTROM: Right. It all makes sense. As you mentioned, the court denied the agency’s motion for preliminary injunction. And the FTC had proposed, and the court accepted, a four-part test for determining whether the acquisition of an actual potential competitor would violate Section 7. The second prong of which is that the potential competitor probably would have independently entered the market but for the merger.
As you note in the paper, the court directed counsel to focus entirely on this second prong, whether Synergy probably would have entered the U.S. contract sterilization market. The court concluded that they probably would not have or wouldn’t have probably entered in part because Synergy claimed that it had already kiboshed plans to pursue the expansion. And I want to get into the reasons behind Polster’s conclusion in a moment. But before we do that, what was the FTC’s theory about why Synergy discontinued its expansion plans, and in particular, how that decision may have intersected with the agency’s investigation?
JEREMY SANDFORD: So, again, this is reading the FTC’s complaint and public trial record and not relying on any private information. Synergy did kill the X-ray expansion. The timing though, I think was a bit suspicious to the FTC.
The merger was announced in October of 2014. Synergy continued making steps towards entering with X-ray following the merger announcement, including announcing to its investors that it was planning to enter the U.S. with X-ray, and then continuing to develop sites, seek tax incentives, draw up customer interest, get customers to text X-ray sterilization at European site. That continued, in the FTC’s view, until February of 2015.
What happened in February? On February 19th, Synergy came into the FTC to meet with staff. So, it’s a standard meeting where parties under investigation get a chance to tell the FTC their views. And in the course of that meeting, the FTC shared their views that it was concerned about the potential for the deal to affect Synergy’s incentives to enter with X-ray.
Six days after that meeting, Synergy officially killed the X-ray project in the form of a declaration provided to FTC staff saying they were killing the project. And again, going by the FTC’s complaint and other public documents, I think circumstances surrounding that killing raised some suspicions.
So, for instance, five days after the meeting between FTC staff and Synergy, there was an internal Synergy email saying, “gents, this whole FTC inquiry is going down a rat hole and I’m going to have to communicate to IBA (our X-ray supplier) that we can’t proceed to the Americas.” A reply on that same email thread said, “Well, that’s definitely a switch, but I’m not surprised based on our approach with the FTC.” So, that, I think, raised some concerns within the FTC that the declaration was perhaps designed to get the FTC to go away more than reflecting business realities.
I should also say Synergy provided five customer emails with its declaration saying the customers were not interested in X-ray. And the declaration said, we’re not continuing because we haven’t been able to generate customer interest. Now, the topic of “were customers interested in this” was discussed extensively at trial. It’s probably best evaluated in 2025 by looking backwards and seeing what has happened since 2015. But we’ll say more about that later.
NATE SODERSTROM: Absolutely. So, I certainly understand why the FTC had some questions around the timeline there. The defendants had an alternate explanation as to why Synergy discontinued the project. It was not related to the FTC investigation. It was related to commercial realities – an alternate explanation about why Synergy was unlikely to effectively ever enter the U.S. contract serialization market, at least in the near term. What were the components of their argument, their theory? And why did the court find that compelling?
JEREMY SANDFORD: So in its opening arguments at trial, Synergy said that, yeah, we acknowledge we had efforts to look into X-ray. But, it never really got past the “wouldn’t it be great phase?” They said businesspeople are always trying to develop new ideas and to bring new products to market. Some of them work, some of them don’t. And this just never got past the investigatory phase where hopefully this will work out. Maybe it will, maybe it won’t.
And Synergy characterized “woeful financial returns associated with the project” and said that, yeah, people were looking into this. But when you really dug down into the details, there just wasn’t much there. There wasn’t a compelling financial case and customers weren’t interested in using the technology. And why would we bring new facilities online if no one wants to use them? That doesn’t make any business sense.
And so, Synergy said, well, look, facilities are expensive. They cost something like $20 million each to bring an X-ray facility online. That’s a big ask for us. We’re a relatively small company and we’re not going to make such a big investment unless we have key customers lined up that we can kind of count on revenues. And unless it looks like there’s a compelling financial return associated with the project as well, in their view, the return wasn’t there.
Synergy also raised process points at trial, principally that the Synergy Board hadn’t approved the strategy. They had announced it to investors, but they hadn’t really committed to anything. And the Synergy Board of Directors—they’re the ones that have to sign off on a large capital expenditure product—and they haven’t done it, which is true.
And so, until they have that signed off from the Board of Directors, actually writing the check on the machines that are going to power the plants, it’s still in the investigatory phase. And in Synergy’s view, the evidence was lacking that it was likely to be signed off by the Board of Directors because of the poor financial returns and because of the lack of customer interest.
And the court, in its decision, cited executive testimony talking about Synergy’s financial model and the executives’ view that the returns were “woeful” and that the modeling was incomplete and speculative and the judge put quite a bit of stock in that. The judge put quite a bit of stock in testimony that there was insufficient customer interest, including a lack of “take or pay contracts,” meaning customers that are obligated to make purchases using X-rays.
And as for timing, I should note the judge cited the timing of Synergy’s killing X-ray as evidence that it was done for legitimate reasons. Because he said, if it was really done for competition reasons, they would have killed X-ray as soon as the merger was announced, right? And so, doing it six days after meeting with FTC staff is evidence it was done for legitimate reasons because—
NATE SODERSTROM: I had not read the decision since 2015. I briefly re-read it in the lead up to this call. Yes, that is what the court concluded on the timing.
So, as you alluded to, the court denied the FTC’s motion for preliminary injunction. The deal closed in 2015. We now have a decade’s worth of post-consummation evidence. You describe in your paper three different sets of evidence, post-consummation evidence, each of which you contend support the FTC’s theory of harm. The first bucket is Steris’ X-ray expansion since 2015, that it’s been substantial, that it’s been successful, and that Steris promotes it.
Could you talk about the evidence there, what you’re seeing, and the idea that this evidence supports the FTC’s theory of harm?
JEREMY SANDFORD: Sure. So, a central question at trial, I think a central question on the judge’s mind, is just, is anyone going to use this? Do customers actually want X-ray?
Because there are costs associated with switching what you’re doing. You have to make sure everything’s kosher with the FDA and make sure—you have to do testing on a new site, much less a new modality. And so, there was a question at trial, is anyone going to go through those efforts to certify a new modality or a new site when they’ve been using gamma to sterilize stuff for a couple decades and it’s working fine?
So that was a central question at trial. I think the post-trial evidence says something about that question. And in brief, post-merger Steris has relied almost entirely on X-ray to build out its irradiation sterilization network.
So, at the time of the merger closed, Steris had 12 gamma facilities, all in North America… Since the merger… and Synergy had, I think 17 in—or, 16 in Europe and Asia. Since the merger has closed, Steris has announced and/or opened 11 X-ray facilities. So, a fair number. In that time, they’ve opened one gamma facility in Thailand and I’m not aware of any expansions of gamma facilities, like adding a second line or something, at an existing facility.
So, what I take from that is almost all of Steris’ post-merger expansion has been in the form of X-ray. I’m thinking of, again, irradiation sterilization, leaving the poison gas aside because that’s a different set of products that is sterilized with that. And so, that seems to me to say something about customer acceptance, right? If Steris wants to make money, they wouldn’t be opening these facilities if customers weren’t using them.
We also have some evidence on use. So, the first X-ray facility that post-merger Steris opened up was in Venlo, Netherlands. On an earnings call, they said that Venlo X-ray had a significant positive margin in its first quarter of operation, which is not necessarily the norm for the industry. Basically, there was enough demand that they made money right off the bat. And in February of the following year—so, eleven months after the facility opened—Steris announced that it would expand Venlo by adding a second X-ray line. So, suggesting they’re already looking at a capacity constraint for the first X-ray line that was being built. So, that suggests to me that the Venlo X-ray plant was doing pretty well. There was significant customer demand for it.
And same thing with the U.S. X-ray facilities. Steris announced three U.S. X-ray facilities and it would not have been incentivized to use X-ray as opposed to gamma or a different modality, unless they forecast sufficient customer demand to support the facilities. So, I think the post-merger evidence all in all suggests that there is—we can be more confident that there is meaningful customer acceptance of X-ray technology now than there was in 2015.
NATE SODERSTROM: That certainly seems clear. The second bucket of evidence you reviewed were Steris’ margins post-merger and in particular finding that they increased – and in fact increased pretty significantly. Could you talk about what you saw there?
JEREMY SANDFORD: Yeah. Look, here the evidence is limited. I still think it should cause us to shift our beliefs a bit. But all I know is what’s in Steris’ public financial reports. They’re broken down by business units. Steris presents results for their sterilization unit as a business unit. Okay, so that’s good: we’re not mixing in non-sterilization stuff. But sterilization includes the poison gas and it includes—well, that’s the main other thing it includes. It includes the poison gas as well as gamma and X-ray business. And also, stuff changes over the years, right? Materials could get cheaper and there could be other things going on that could cause the change.
But, in short, what I find is the margins increased. They increased meaningfully. They increased by 28 percent. So, their publicly reported margins were 32 percent in the first full year following the merger and they were 42 percent in fiscal year 2024.
So, that’s a pretty big increase in margins. Over the post-merger period, Steris’ sterilization revenue grew at a 10 percent annual growth rate while their costs grew at a 7.8 percent annual growth rate. So, that suggests that Steris’ sterilization business is healthy and their revenue is growing more than their costs. Which suggests that they have some room on price, more room now than they did in 2015. I looked for evidence, kind of direct evidence on price. But again, I don’t have any non-public data here from any party and that’s a pretty important limitation. I should be clear about that.
The one thing I could find, which is not—it is what it is. It’s something. But one of Steris’ smaller competitors is a company called eBeam Services. And so, eBeam is kind of a more distant substitute for gamma and X-ray that can, it doesn’t, it shoots electrons instead of photons and electrons can’t penetrate matter. So, you can only sterilize individual boxes of products. So, everything has to be depalletized and sent one box at a time. So, it’s kind of a relatively small share of radiation sterilization.
Anyways, eBeam Services is the provider of eBeam. And they have a blog post where they talk about—they try to sell customers on the advantages of eBeam over gamma. And they periodically update the blog posts with information about the price of gamma. And so, in 2015, their blog posts that gamma costs $1.60 per cubic foot, so a three cubic foot box would costs $4.80 to sterilize. And that price got updated to $4.20 per cubic foot in 2021. Which is about an 18 percent annual growth rate.
So, that’s the data point I have. It’s not perfect. We would need to be measured in the weight we assign to it, of course. But that one data point does suggest that the price of gamma has been going up and that’s consistent with sort of relaxed competition following the merger.
And one might wonder, well, this was a merger of potential competitors, right? So, it was just the prospect of future competition that was of concern to the FTC. How would removing a potential competitor affect margins?
Well, I think economics teaches that firms price with an eye not only towards competition today, but what competition might look like tomorrow too. And firms might take a view that, well, if I get too carried away on price, that’s going to encourage entry. And it makes it more likely I’m going to face more competition in the future as well. And so, the removal of the potential competition, the competitor, does potentially remove a competitive constraint on a firm and could potentially incentivize them to raise price.
NATE SODERSTROM: On the margin question in particular and on the margin increase, is there any argument that these have been, to some extent, driven by merger related efficiencies or synergies by virtue of the 2015 acquisition?
JEREMY SANDFORD: Yeah, that’s a good question. I think the short answer is no. Efficiencies were discussed briefly at trial. Steris’ CEO was one of the parties’ witnesses and he was asked about efficiencies. He identified $2.5 million in savings for sterilization. That’s all sterilization, including the poison gas. And that was the parties’ number. It wasn’t like—it didn’t go through the FTC’s vetting process where they figure out what is merger specific and what is marginal cost versus fixed cost and all that. And $2.5 million isn’t – well, it’s not a very big number. It’s certainly not enough to drive the increases in margins that we see.
I’d also say just thinking about the nature of this business. I mean, you kind of have a plant over here. It takes boxes in, sterilizes them, and sends them off to hospitals and providers. A plant over here that does the same thing. It’s not really clear why having both of those plants in the same network results in cost savings. You can maybe get some freight rationalization, but freight’s not a huge component of their costs. And so, just thinking about the nature of the business, this isn’t one where I think we would expect there to be significant efficiencies from combining.
There’s also geographic differentiation between the parties. So, Synergy was planning to enter in the U.S., but their actual facilities at the time of the merger were in Europe and Steris’ facilities were in the U.S. So, it’s even less clear why there’d be efficiencies from combining European plants and U.S. plants. So, I think even the parties would agree, if they were asked, that significant efficiencies were not likely to result from the merger and probably did not result from the merger.
NATE SODERSTROM:. All makes sense. Going back to the sets or buckets of evidence you reviewed, the third was that you went back and looked at the claims Synergy made at the hearing, including replicating Synergy’s financial model for the X-ray expansion that they presented and that the court found compelling. And the model showed that this expansion would not have met required IRR targets, which, as you mentioned in the paper, were lightly documented, but by virtue of this woeful financial prospects here, this is not something that would have ever been implemented. The financial model simply didn’t support it. You went back and took a look at it. What did you find?
JEREMY SANDFORD: Right. So, Synergy had prepared a financial model to help their management and their executives assess the financial case for X-ray. The financial model was a spreadsheet. It was saying “here are the revenues we project and here are the costs we project.” And then you can kind of add those up and calculate the return on the investment. And so, that’s what they were looking at.
And so, IRR is an accounting term. It stands for internal rate or return. You can think of it as like a return on investment metric where a higher number is better. It means a higher return on investment. And IRR was the output of the model. So, the spreadsheet is here’s the revenues. Here’s the costs. And then there’s the cell over here that calculates the IRR.
And the model became, I think, of central importance at trial. Synergy’s, in my view, most effective witness was their CFO who was up there to discuss the model and why it showed a “woeful” return and why it was speculative. And even the return it did show, in his view, was speculative and unsupported and couldn’t put much stock in it.
And I think the judge credited this testimony and testimony of other people saying that, I mean, Synergy’s story at bottom was we looked at the model and the return just wasn’t there. So, why would we – we have other ways of making money; why would we allocate capital towards this? And so, it was very effective testimony based on the financial model.
The model was presented at trial through the testimony of Steris executives. There was no expert testimony on this topic. So, keep that in mind. So, I thought it would be useful to evaluate the claims made by the executives in relation to the model. To do that, there’s enough information in the trial record to recreate the model. The model is just inputs, projected revenues, projected costs and outputs like IRR or a couple of other outputs, like when the plants reach full capacity as an input. So, for instance, what the price of X-ray is that they projected? Well, that’s in the trial record.
So, I get as many inputs as I can and I get the outputs from the trial record and I try to match everything up as well as I can. And where there’s stuff I don’t know, I tweak it to match the outputs that are discussed in the trial record. And that’s my replication of the Synergy financial model. And this allows me to analyze claims made by Synergy.
So, one of Synergy’s claims about the financial model was that it showed a low return in the form of IRR. They said it showed a return of six percent. They said Synergy requires a return of 15 percent to pursue large CapEx projects. Well, the six percent return comes from looking at the first ten years of life of the X-ray facilities.
So, literally assuming they will burn down on day one of year eleven and the facilities and just be worthless. And that’s what they call the ten year IRR. And Synergy’s testimony was that, well, we only look at the ten year IRR. We require the ten year IRR to be over 15 percent. This was not documented at all. There were oblique references to a 15 percent IRR threshold in some documents, but it didn’t say whether it was a ten year—whether you would take into account year eleven and beyond of the firm. And, in fact, if you account for the full lifespan of the facilities, which is 20 to 30 years, well, you get an IRR—using Synergy’s financial model—of about 15 percent.
So, that result allows me to contextualize the testimony that, well, we only look at the first ten years of the facility. Well, that doesn’t make much sense. If you look at the full lifespan of the facility where, of course, the facility is generating returns beyond year ten. So, it makes sense to account for those returns. You get a 15 percent IRR, which is Synergy’s supposed threshold.
Another of Synergy’s concerns about the financial model was that, well, we asked management to reduce CapEx of the facilities, just try to get the spend down, and they weren’t able to do so. And so, I’m able to look at that through my replication of the financial model.
One of the points I make in the paper is that the main CapEx associated with an X-ray facility is buying the X-ray accelerator. Synergy’s supplier was located in Europe. Synergy is a U.S. company. So, they would pay for the X‑ray accelerator with dollars. The dollar to euro—and the accelerator was priced in euros—the dollar to euro exchange rate became dramatically more favorable between the time that management were told to reduce CapEx and when they killed the X-ray project. And the effects of the dollar becoming stronger against the euro was to reduce CapEx by $1 million per facility.
So, Synergy was asked to reduce CapEx by $1.5 million per facility by the executives who were evaluating the project. Exchange rate fluctuations took care of two-thirds of that on their own. And you can go into the model and say, okay, what happens if we reduce CapEx by $1 million, which happens through exchange rate fluctuations alone? Well, that raises the IRR to 16 or 17 percent.
The model is also conservative in some ways. So, the model projected the price of X-ray to be at a 20 to 40 percent discount relative to gamma. Okay. Well, you might imagine they priced it low initially to kind of encourage customers to take a look at it. But the point of X-ray was this is a better technology than gamma. It involved less handling, less damage to the products, faster turnaround times, and no risk of supply disruption from not being able to get cobalt.
So, at least over time, you would expect the price of X-ray to rise and maybe at least get to parity or something closer to parity with gamma. But Synergy projected the price of X-ray to stay flat over the lifespan of the facilities and not rise relative to gamma. So, if that’s a conservative assumption. You could see why a business might make a conservative assumption on a valuable investment, sure, but it certainly is true it was conservative and that would cause the IRR to be understated relative to, say, allowing the price of X-ray to catch up to gamma over time. And I do this in the paper. If you assume two percent annual growth rate in the price of X-ray, that raises the IRR by another couple of percent.
I don’t want to bog us down in details, but there was testimony about a double counting error. Well, I’m able to evaluate the significance of that in the context of the model, and I didn’t think it was really a double counting error either.
So, anyway, the takeaway from all of this is the Synergy financial model is not rocket science. It was just a list of projected revenues, projected costs, that result in an IRR. I think the testimony that the financial returns were “woeful” is just, in my view, not supported by the financial model, by my replication of it. And I think to some extent the testimony relied on a kind of cherry picking. Maybe that’s too strong of a term. But characterizing the model in a way that made the returns seem more unfavorable than they probably were.
NATE SODERSTROM: Okay. So, those are the three buckets of evidence as far as thinking about a retrospective and what we’ve seen ten years since the deal closed. I want to talk a bit about implications for other potential competition cases.
You mentioned in your paper, and it’s in the decision, Polster applied what was ultimately a binary test. If Synergy is probably going to enter, I’m going to enter the injunction. If they’re not, then I’m not. He concluded they were not and the deal closed.
You mentioned that that test may be consistent with the law, or at least the law as the FTC described it. But that test is not necessarily consistent with the economics. I’m going to offer a couple of observations. I’m going to integrate a listener question, and I’ll see if I can do so effectively.
But the first is on the 2023 guidelines, which you mentioned in the paper propose a slightly different test for evaluating the question of whether the acquisition of an actual potential competitor might violate Section 7. As you observe in the paper, the 2023 guidelines discuss a reasonable probability of entry, and cites there are to Penn-Olin and Marine Bancorp. So, two 1960s Supreme Court cases. Steris is unsurprisingly not cited in the 2023 guidelines, and the guidelines don’t reassert that four-part test the FTC suggested in 2015.
So, the 2023 guidelines seem to leave open the possibility that you could bring this case, and you don’t need to show concrete plans from the B-side to enter or probably enter. You simply need to show feasible means, capabilities, incentives.
So, again, there’s the question there. If you had the same fact pattern in 2015, you’d probably see the agency propose a different legal test for actual potential competition. Would that push the court in a different direction? I don’t know.
This gets to the listener question, which may be challenging for you to answer. But I will ask it anyway. Do you have a view on the question of whether this case would be decided differently under the current merger guidelines? And I think as part of that, you will assume that the court would accept the 2023 guidelines test for whether the acquisition of a potential competitor might violate Section 7.
JEREMY SANDFORD: That’s very legal, and I don’t want to get too far outside of my lane. Would it be decided differently? I don’t know. I think the—this gets into my view of the guidelines. My view of the guidelines is their job is to distill economic learnings and law in a way that is helpful to practitioners and courts. I don’t know that the economics underlying potential competition have changed or that our understanding has advanced relative to, say, the 2010 guidelines. And so, I don’t know.
I think, to the extent the guidelines have changed, they reflect a different characterization of the same law and economics than the 2010 guidelines did. And that’s fine. People have different views on this. But I don’t know that changing the guidelines without an underlying change in law or economics is going to be enough to produce a different outcome.
That’s kind of my view in general of guidelines. They should distill law and economics in a way that’s helpful to practitioners. And so, if those don’t change, I wouldn’t expect different wording and guidelines to produce different outcomes. But this is very legal, though. I know that the case law here has always confused me, right? When I first heard about Steris, I’m like, well, we have a close competitor who is seriously contemplating entry. That seems bad.
And I was told, oh, but it depends on whether the court accepts the actual potential competition doctrine. It’s like, what does that mean? Like, there’s no doctrine that needs to be accepted when we have two current competitors that are competing against each other. Why does that change when one of them is nearly on the verge of entering? So, the law here has always kind of confused me. You may have probably more informed views on that than I do, Nate.
NATE SODERSTROM: Yeah. I don’t know that I do. Although, as you recall the defendants suggested that an actual potential competition case could not actually implicate Section 7 because Section 7 only deals with existing competition, which eventually the court declined to take up. Left to the Supreme Court. We did not get there, of course.
One other point, and we’ll get to perhaps some Monday morning quarterbacking on what the FTC could have done differently in 2015, but you made some points about the decision. One other thing I noticed just in reading—which, again, I hadn’t done for a decade or so—just the court’s treatment of 13(b) of the FTC Act, which allows the agency to go get a PI in the first place. And in each case, when the court refers to 13(b), they refer to it as 13(b) of the Clayton Act, not the FTC Act. So, I’ll leave that to the appellate lawyers. But there’s getting the law wrong and there’s, as is the case here, quite literally getting the law wrong.
Look, I know in 2015, 13(b) was not a point of emphasis. I think more recently the agency has attempted to emphasize the 13(b) standard and the idea that the threshold to get a PI to get this to an ALJ should be pretty low. Again, had there been more emphasis on 13(b) in 2015, would we have reached a different outcome? I don’t know, but it’s, of course, another question one might ask.
But as you alluded to, we’re not here to talk about the law. We’re here to talk about the economics with the benefit of a decade’s worth of hindsight. And I imagine you’ve—I won’t say you’ve had many sleepless nights over this, but I’m sure you’ve thought about it. Is there anything the FTC could have done back in 2015, as far as its presentation, as far as its emphasis in front of the court, that you think potentially could have pushed the court in a different direction?
JEREMY SANDFORD: Right. So, I think the FTC was massively hamstrung at trial by its inability to put on expert testimony. Why did it not put on expert testimony? Because the judge said in an order that any economic testimony would be heard through expert reports. And the judge said explicitly he didn’t want to hear from economists on the day of the trial. And I think that was unfortunate, given how central the financial model turned out to be. And again, the financial model is not rocket science. It’s revenues and costs and some assumptions. And you can talk about the reasonableness of the assumptions and why they were made and what they showed.
But I think it was difficult. The Synergy executives in their testimony put forward a version of the financial model that supported their view that the returns were dismal and “woeful” and speculative. And a little bit of unpacking of the financial model would have gone a long way in contextualizing those claims. But it’s difficult to do that unpacking in cross-examination. It’s just not the right forum in which to say, “okay, you said this. But isn’t it true that this is…”—I mean, witnesses are trained to deflect those questions and a questioner—even questioners, of course, we economists were working with the attorneys that were doing cross-examination, but it’s not realistic to expect an attorney to be so well-equipped with all of the facts and details of the financial model that he or she could effectively use cross‑examination to essentially rebut factual claims made by a witness.
I think the way to do that is through expert witness testimony. If the judge didn’t want to hear from economists, maybe, and I don’t know exactly what the judge said, could we have put forward an accounting expert or a financial expert, someone who’s not called an “economist”? And the FTC has those people or sometimes they hire them from outside firms as well to kind of walk through, here’s what the financial model is trying to say. Here are the assumptions. Here’s what it does say. And then here’s some context to the claims made by Synergy executives.
And so, I think it would have helped the FTC’s case had an expert economist been able to testify. Your audience is probably thinking, this is not shocking. The economist wants more economics in a merger trial. Oh, shocking. But I think that would have helped them. And it was unfortunate the judge said he didn’t want to hear from economists. And maybe there was a way around that in putting forward a non-economist expert witness, whether an industry expert, an accounting expert, a finance expert, to help kind of rebut some of the claims made by Synergy executives. And I think the paper is in part an effort to sort of fill that gap as well. Here’s what expert testimony might have looked like and here’s how it might have been effective at contextualizing claims made about the financial model.
NATE SODERSTROM: Yeah, like you said, Polster did not want to hear live testimony from an IO expert, but accepted expert reports. Certainly, in his decision, he cited those expert reports, I imagine, quite a few times?
JEREMY SANDFORD: I don’t think he did. I don’t think he mentioned either of them.
NATE SODERSTROM: Understood. Okay, so again, implications for future potential competition cases. You mentioned that Polster ultimately applied a binary test – if Synergy is going to enter, injunction. If they’re not, no injunction. You mentioned that may be consistent with the law. But you also make the point that the four-part test the FTC suggested is not necessarily well-supported by the economics and in particular having an oversized emphasis on prong two, this question of probability of entry. And an overemphasis on that prong could lead to some suboptimal outcomes. Why do you hold that view?
JEREMY SANDFORD: Yeah, and the judge said explicitly that he would decide the case in the FTC’s favor if he thought it was likely that Synergy would have entered the U.S. market by building one or more X-ray sterilization facilities within a reasonable period, but for the merger. So, the court set the standard of, “I will give the FTC the injunction if I’m convinced Synergy was more likely than not to have entered.” I take that as a—if the judge thinks there’s a 50 percent chance of entry, I’ll grant the injunction. I think that has a couple of risks if we take that standard and apply it to future cases.
One is, I think we should also be concerned about the effects of entry. The judge kind of spotted the FTC that the merger would have effects. And to be honest, my view at the time was that the FTC’s case was relatively stronger on probability of entry than on effects of entry. There was some concern that—especially at the time without the benefit of hindsight—is anyone going to use this? So, it could take a while for customers to accept the new technology. It could take a while for the plants to ramp up. And so, it could be a while before X-ray plants are a competitive constraint on gamma. The judge assumed all that away and just said, I’m going to grant you effects. I’m going to decide this for the FTC if there’s a 50 percent chance of entry.
Well, that was a favorable stance to take, favorable to the FTC. I’m not sure that stance is helpful to apply more generally. I think we should be concerned about effects. And I think, more importantly, we should be concerned jointly about probability of entry and effects. It seems to me that, whether a merger has a 49 percent chance of killing entry or 51 percent chance of killing entry shouldn’t really affect anything. It seems like a minor difference.
And it seems like you can imagine a hypothetical where a merger lowers the probability of entry by 49 percent. And so, it falls short of Judge Polster’s standard. But if entry occurred, it would have been really beneficial. Let’s say there’s a 49 percent chance a competitor to NVIDIA is going to come in with lots of advancement in AI chips and price competition versus a 51 percent chance that a laundromat is going to enter, something smaller scale. I think we should be more concerned about the former case than the latter case.
And the judge’s standard seems to miss that entirely, which, again, was favorable to the FTC. When he said, I’m going to rule for the FTC if I think entry is more likely than not, I thought the FTC was going to win. That didn’t turn out to be the case but it was it was a favorable stance. But it seems like the wrong stance to take in general. And it can lead to both under-enforcement and over-enforcement. You’re going to over‑enforce the laundromat mergers and under-enforce the mergers involving, say, a potential competitor to a large company like NVIDIA.
NATE SODERSTROM: Yeah, understood. And those, I think, hypotheticals make the point well. But more broadly, the idea is over-emphasizing this question or solely emphasizing this question, of probability while ignoring potential effects or de-emphasizing potential effects, wrong way to think about it. You propose an alternate approach, which you term the expected value approach. Could you talk about that? What would that look like in practice?
JEREMY SANDFORD: Sure. So, an expected value approach looks at the expected value of competition that would be lost if the merger is consummated. So, an expected value is, well, it’s a mathematical term. It basically means you multiply the probability by the outcome. So, if I flip a coin and I get $10 if it comes up tails. The expected value is $5. This is a half chance of tails and multiply one half by 10.
And so, expected value would look at the expected outcome, probability weighted outcome, of the merger. And so, it would take into account both the probability that the merger would kill entry and the effects of entry conditional on it happening. And the expected value approach would be more likely to enforce mergers that have a higher expected harm than those that have lower expected harm.
So, an expected value approach would definitely be more concerned with my hypothetical involving 49 percent competition with NVIDIA than with my hypothetical involving a 51 percent chance of laundromat competition. Because the expected benefit of that competition from a serious competitor to a large company would be much larger than the expected competition between laundromats. And so, even though the probability is a little bit lower, the expected value approach would be more concerned with the larger merger.
And it doesn’t have to be a 49 percent versus 51 percent. I mean, one of the things I found, as I looked into potential competition mergers, nobody really knows what the standard is. It’s 50 percent. It’s 40 percent. It’s 30 percent. It’s 60 percent. Different people have different views on this. But the point applies regardless of what it is.
I mean, if a court has a view that the correct standard is 30 percent—there needs to be a 30 percent standard of chance of entry for a merger to be illegal, that still kind of misses—to the extent that the court evaluates the merger on the basis of whether the probability of entry is over 30 percent, it still kind of misses the point that you need to look at both entry and the effects of entry, initial entry occurring.
NATE SODERSTROM: Yeah, certainly makes sense. I have one final question from the audience that the questioner phrased as – well, it seems to be a prompt to ChatGPT. So, I’m going to attempt to divine an English language question out of this.
The question is something along the lines of, in your view, what are the best quantitative strategies for assessing agency intervention retrospectively? And maybe to phrase that slightly differently, you’ve done this retrospective on Steris/Synergy, unique in the sense that it was potential competition, not existing competition. But when you’re thinking about the opportunity for other retrospectives, especially in scenarios where we had existing competition, where the agency did intervene or ultimately didn’t intervene, maybe we’ll phrase it a bit more broadly.
What’s the best quantitative way to perform these merger retrospectives? Again, this was a unique situation with unique questions. To make it a bit more general, though, how do you think about merger retrospectives? What’s the best kind of evidence? What’s the best way to approach it?
JEREMY SANDFORD: I think merger retrospectives are incredibly important. I think they’re incredibly informative to the agency’s work and to the IO researchers more generally. And there’s a rich literature of merger retrospectives. A lot of it is done by FTC or former FTC economists. BE has a website of merger retrospectives that’s done by people who are currently or were once affiliated with BE.
I have another paper on that list. It’s a retrospective of a casino merger in Missouri. And there are people that have written much more than I have. Two of my Econic colleagues come to mind. Dan Hoskin and Nate Miller have both written important merger retrospectives where you look at what happened with a given merger. And both of them have found instances where price goes up.
I mean, look, data availability is a driving factor in any research paper. And so, it’s certainly nice if you have information on price. That’s more possible when you have mergers of supermarkets or CPGs (consumer packaged goods) where there’s companies that track prices. You can measure price before and after. There’s a set of empirical tools that I think are appropriate and very powerful in looking at merger effects, difference in difference regressions.
Basically, the idea is you look at how post-merger trends compared to pre-merger trends for an appropriate control group. And if you see, following the merger, prices started to go up and the control group prices stayed the same, that could be evidence that should certainly cause us to think it’s more likely that the merger caused those prices to go up. There are other ways of looking for causality and trying to see did anything else change that might have caused prices to go up right after the merger happened? For instance, if a merger occurred contemporaneously with COVID, well, you’d want to control for COVID because COVID caused lots of prices to go up. And so, there’s a robust set of empirical tools that economists use routinely to assess those things.
And yeah, there are hundreds of recently written merger retrospectives. And they’re some of my favorite papers. And I’ll say from experience, folks at the agencies take retrospectives very seriously. Researchers should continue to pursue them because I think there’s a high value added from continuing to write those papers.
NATE SODERSTROM: I certainly appreciate that and appreciate the coherent answer to what was not a particularly coherent question. Jeremy, we’re almost at the half hour. Before we go, any last minute thoughts on the paper, on the merger more broadly? Anything else we didn’t hit that we should have?
JEREMY SANDFORD: I think we hit the main points. I will say certainly if anyone has any questions or any thoughts or how I can improve the paper or what else I can add to it, please do reach out. I’m available through email or phone. I’m happy to talk to anyone that has questions or thoughts about the paper. And to certainly appreciate anyone who came to the talk and listened; thanks for coming. As I said, I’m happy to talk further about this topic the next time we have a chance to do so.
NATE SODERSTROM: Wonderful. I will echo that again. Thank you to everybody who listened. And most importantly, Jeremy, thank you for joining us. This was tremendous. Really enjoyed it. And with that, we will sign off. Thank you again.
JEREMY SANDFORD: Great, Nate. Thanks for having me. Thanks for the questions. Really appreciate the conversation.
NATE SODERSTROM: All right. Thanks, everybody.
JEREMY SANDFORD: All right. Bye.
