DOJ ANALYSIS OF RADIO MERGERS
JOEL I. KLEIN
February 19, 1997
I'd like to start by thanking you for giving me the opportunity to come here and discuss radio mergers with you. I know this is an important issue for you and I want you to know that it's important for the Antitrust Division as well. Since the 1996 Telecom Act, there has been an explosion of radio mergers -- more than a thousand, I am told -- and, under the requirements of the Hart-Scott-Rodino Act, about 140 of these have automatically come before our agency for review. The process of merger review can be complicated, even frustrating, and I know it's never fun for those of you who are doing deals. I mean, you've struggled to get the thing done, negotiated your hearts out, and then, all of a sudden, some federal agency comes along and asks a lot of questions, wants documents, takes testimony, and sometimes tells you that the deal can't go through quite as planned. Worst of all, perhaps, this whole process takes time and, in business, time is money (and so are legal fees). So going through the Hart-Scott process can fairly be said to be like getting a medical check-up -- somebody's looking for bad things about you, and you're being asked to bear the costs of the inquiry.
But even if not pleasant, the Hart-Scott process is a critically important one. Indeed, in my view, it is essential for protecting American consumers from mergers that can create market power and result in non-competitive price increases or other anticompetitive effects. To be sure, most mergers don't lead to those results, but sometimes, no matter how experienced you are in this business, it still takes very careful study and analysis to find out if a given merger is likely to have anticompetitive effects. And our job is to make sure that the analysis is done properly and, when necessary, thoroughly. To go back to my medical-examination analogy, some tests may be unpleasant (as well as costly) and though you may not like going through them at the time, you can still appreciate that they need to be done. The analogy breaks down, of course, in that the results of additional medical testing inure to your benefit, whereas additional Hart-Scott review can only benefit someone other than you. So our medicine is even harder to swallow than the doctor's.
And, given all that's going on in radio mergers, you folks have had to take a lot of our medicine lately. That's largely due to the fact that, prior to the 1996 Telecom Act, the amount of radio consolidation that was allowed by statute was so small that the antitrust laws never really came into play and, as a result, whatever other problems you had, you didn't have to worry about us. Indeed, the primary reason for the huge radio merger wave we're now experiencing is the enormous pent-up demand that resulted from the previous statutory limitations on radio ownership.
Now, from where I sit, I fully applaud the changes put into place by the 1996 Telecom Act and believe that it has led to a lot of healthy consolidation in your industry. But, with the benefits of substantial consolidation, you must also take on the traditional burdens of antitrust review and exposure. Most other industries have been going through this for years and, while I'm sure they're no more thrilled about the experience when they undergo it than you are, I think they've come to understand what it's all about, are more able to take account of our concerns at the front-end of a deal, and, quite frankly, often have had the experience of opposing certain mergers as well as having been subject to merger review, and, so, I think they tend to have a somewhat more balanced view of what we do.
That's why I wanted to come here and speak with you today. I'll admit we haven't done a perfect job but I think we've done a good job in handling the huge number of radio cases that have come before us. And I believe that if I talk candidly, you will come to understand and perhaps even respect what we're doing, which is not to say that you'll always agree with it. But if our disagreements are based on a full understanding of each other's views, I can't help but believe that we will all be better off.
I have spent a lot of time focused on radio mergers since I became Acting AAG in October and I have specifically reached out to the industry in an attempt to engage in a constructive dialogue. Early on, I made some informal remarks at a law firm program where several radio executives had the opportunity to ask questions. I followed up with a formal meeting with NAB President Eddie Fritts and his legal team, where I agreed to do a lengthy interview for the NAB, answering a wide-range of questions, which is now done and available for all to read. I also agreed to Eddie's requests that I come here today and that I speak at the NAB Convention in April. And, I have also met with other association representatives and individuals to obtain their views about radio mergers. In fact, I'm a bit worried that it might look as if I'm trying to bore you to the point where you'll leave me alone and never want to see me again. But the real truth is that I've found the process helpful in terms of improving what we're doing and I believe that it has sharpened my thinking in this area. Let me also say publicly that I appreciate the spirit in which Eddie and his staff have approached this. While I know there are areas of disagreement and concern that remain, Eddie Fritts and the NAB team have been constructive and professional throughout the process and I want to thank them for that.
Before turning to the substance of our analytic approach, there's one other point I'd like to place in perspective at the outset. And that is that, overall, I think the numbers show that our impact on radio consolidation really have been quite modest. There have been well over a thousand radio mergers since the passage of the Telecom Act a year ago. We have conducted investigations in only 50 of those -- 41 under Hart- Scott, and 9 on our own initiation. Of these, we have thus far brought three cases -- each resulting in a consent decree: (1) the Jacor/Citicasters merger in Cincinnati, where we required that one station be divested, resulting in a post-merger market share of advertising dollars of 46% (as compared to 53% without the divestiture); (2) the ARS/Lincoln Group merger in Rochester, in which we insisted that three stations be divested and that a Joint Sales Agreement (JSA) be terminated, thereby reducing the merged party's market share from 63% to 38%; and (3) the Westinghouse/Infinity merger, which involved 9 separate geographic markets and a total of 83 stations, and in which we required divestiture of one station in Boston and one in Philadelphia, lowering market shares in both cities from something above 40% to just below that number. Finally, as we sit here today, the Division has 25 radio mergers under investigation. So, while we are not yet finished, as of now we've required divestiture of 6 (or 7 if you include the JSA) of literally thousands of stations that have been the subject of mergers.
All right, after that enormous wind-up, I suspect you're wondering whether I'm ever going to throw the ball. Well, here goes.
I start from the premise that radio mergers, like all mergers, can be a desirable thing. From the point of view of the merging parties, at least two good things can happen. First, the merger can result in efficiencies, which in turn will make the merging parties' goods or services more attractive and thus lead to greater profitability. For example, a handful of radio stations that consolidate staff through a merger may be able to operate more efficiently than a stand-alone station. Second, mergers can lead to the creation of market power resulting from the restriction of consumer choices -- for example, due to the merger of the only two gasoline stations within 100 miles (in an area where zoning restrictions prevent the creation of any new gas stations), the merged entity could raise its prices merely because of the fact that customers who need gas have no option but to buy from the merged station. That kind of market power, while good for the merging parties, is nevertheless anathema to the antitrust laws and directly forbidden by Section 7 of the Clayton Act when it comes about through merger.
Our job, in a nutshell, is to separate the good mergers from the bad ones -- those that are efficient from those that result in naked market power. Of course, a little of both may be going on so sometimes we have to balance these competing effects. And frankly, if nothing bad's going on, we take a pass on the merger even if we conclude that nothing good's really going on either. So, in a sense, in our business, like in baseball, a tie goes to the runner.
Now, when we get down to applying these general principles to a radio merger, while there are often case-specific disagreements, the overarching point that seems to divide us from radio owners can be summed up in one question -- is radio a market? That is, in terms of a potential advertiser's options, can it fairly be said, as I've heard industry people frequently say, that an advertiser can always buy around radio, which I take to mean that if radio prices go up the advertiser can use newspaper, broadcast, cable, or some other effective substitute. And, before I explain why I believe that the answer to that question is no, let me make clear that if you disagree with my view on this fundamental point, then you believe that, even if a single person owned every radio station in the country, he or she would have no market power as a result and that, if he or she raised prices, say by 5%, enough advertisers would go to other media to make such a price increase unprofitable. I just don't believe that would happen, though I do admit that some advertisers might move to another medium as a result of price increases that resulted from an anticompetitive merger.
In other words, unlike some of our critics, we don't go all the way and say that for all advertisers radio is really no different from these other media and that for a 5% market power price increase they would all just pick up and go to another medium where they could then get more advertising bang for their buck. Given the differences between the various media and the way they are looked at, or listened to, as well as the differences in prospective customers -- young, old, have cable, don't -- a claim of perfect (or even close-to-perfect) substitutability across different media simply isn't convincing. And this isn't just me speaking -- although I can tell you from personal experience that if you think advertisers can always sell rock CDs as well through newspapers as through radio, you haven't had teenage kids. But even aside from personal experience, my view about the difference in these various media are based on the work we've done, reading the files and testimony of people in the industry, of advertisers, and of investment bankers, as well as from considering studies and analyses, including some that we've conducted ourselves.
In fact, some of the strongest arguments in support of our position have come directly from the documents of major radio station owners. For example, in a recent filing with the FCC, Citicasters asserted that "radio and television compete in distinctly different markets. The peak audience for radio is during the morning drive time while the peak viewing audience for television is during evening prime time. The demographics of the audience is also different, with radio stations tending to be much more focused in their demographic appeal." Viacom echoed these views in the same FCC proceeding, stating that "experience in the marketplace . . . confirms that advertisers generally do not treat broadcast television and radio advertising as direct substitutes."
And moving away from the generalities of public filings, there is one document we uncovered stressing, as part of the pitch during a roadshow for investors, that the right combination of radio stations makes it "difficult [for advertisers] to buy around." Similarly, another document touted that a principal advantage of consolidation was that "back-side profits [would] result from aligning multiple properties in a such a way as to eliminate today's competitors, while deterring tomorrow's." Now, I want to make clear that I wasn't born yesterday and I know that people can be boasting when they make such arguments so we look behind documents like this to examine the specifics of the marketplace. I hope you take some comfort from that. Still, it's nice to know that our views appear to be shared, at least some of the time, by important players in your industry as they consider making significant investments.
Let me next reiterate, and elaborate on, a point I mentioned earlier -- that is that our view of radio as a distinct market doesn't mean that there are no advertisers who can divert their advertising to other media to avoid a price hike, but only that such behavior will not ultimately defeat an anticompetitive price increase. A key reason that leads us to this conclusion is that radio owners can, and routinely do, charge different rates to different customers depending on the customer's demand for radio. That is, radio stations raise prices for those customers who don't have other realistic options available, while they maintain prices for customers who do have such options. For example, if an advertiser is interested in reaching a particular demographic group -- let's say females aged 18 to 34 -- an owner who has all the stations that cater to that group will have more market power with respect to that advertiser than with respect to an advertiser who isn't aiming at that particular demographic group. When price negotiations take place, both sides are aware of these considerations. And since radio advertising rates are negotiated with each advertiser individually, the radio station owner is able to charge a higher price to the advertiser with fewer options, while keeping prices low to the advertiser with more alternatives. Again, I want to stress, we're not making this stuff up; our investigations have found business strategy memos indicating that this is precisely the kind of activity that takes place when it comes to pricing decisions.
Finally, if I can make one last point on this issue of whether radio is a separate market, I'd like to explain why it doesn't matter to us that only 7% of all advertising dollars go to radio. Unless I am wrong about the substitutability points that I just made, the question remains one of market definition, not of overall advertising revenue. It's no different from a situation where all soft drink manufacturers would seek to merge and control 100% of that market. We wouldn't walk away from such a merger -- and if you like soft drinks I should think you wouldn't want us to walk away -- merely because there are lots of other beverages out there, such as milk, juice, beer, wine, etc. And I say that even though I know a price hike in soft drinks would lead some people to drink more juice and others to drink more beer.
Before moving on to discuss how we assess market power within the market as we define it, I should briefly mention that our overall market definition requires us to consider geography as well as the relevant product. In that way, we set the boundaries in which we will look to determine what alternatives are available to radio advertisers. Someone who sells cars in Topeka can't profitably advertise in Miami if the cost of advertising goes up in Topeka. I mean, there aren't a lot of people in Miami who are going to buy cars in Topeka no matter how good the advertising is.
In radio cases, the question of geography generally has not been a highly disputed issue. We usually rely on a metropolitan area as the appropriate market and most people agree that's reasonable. Occasionally, however, we run into a problem when people try to argue that the mere fact that a station can be heard in a given area is sufficient to include that station in the geographic market. In our view, for example, San Francisco and Sacramento can't properly be considered part of a single geographic market simply because there are some San Francisco stations that reach Sacramento. For a supermarket in Sacramento, it wouldn't normally make sense to advertise on a San Francisco station since you're paying to reach a lot of listeners that you don't have any interest in. On its face, that makes such an alternative appear to be too inefficient for serious consideration; as a result, when we go out and look at the actual practices of advertisers and find that local businesses in Sacramento don't advertise on San Francisco stations, we limit our geographic market to Sacramento.
Once we've properly defined the relevant market, we then move on to consider whether the merger will create or enhance market power within that market. There are two well-established ways in which that can happen -- first, through unilateral action by the merged party, and second, through coordinated effects in a concentrated, post-merger market. Let me say a word about each of these methods of analysis.
To begin with, let me describe the theory of unilateral effects, which is the approach that has dominated our analysis in radio merger cases. This theory recognizes that the products or services in question are not simple commodities -- such as wheat or iron ore -- but rather are differentiated so that consumers often have somewhat different preferences for each product, even though they may ultimately be in the same market. We believe that this kind of product differentiation affects the radio market. And what that means is that an advertiser that prefers one station to another (but still finds the second station to be a reasonable option) can often play the stations off against each other in an effort to get a better deal. But if this advertiser's first and second choices merge, her ability to negotiate prices may be diminished because her third choice may be only a so-so alternative.
So that I don't entirely lose you in economic jargon, let me offer a concrete example of how this would work in the radio industry. Suppose, for example, that Owner A currently owns two of the four rock stations in a given market and that he seeks to purchase the only other two rock stations out there. I know that many of you are thinking that other formats might also be able to reach the same audience that listens to those stations, but, for simplicity's sake, let's just assume that no other formats reach this same audience. (I could make the hypothetical more complicated but it wouldn't change the point.) In the circumstances I've posited, then, a merger of all four rock stations, for at least certain advertisers, would seriously diminish their options and, those advertisers who need to reach those people who listen to rock stations would have little choice but to pay the increased advertising rates -- and certainly less choice than when the two other rock stations were in the hands of a different owner.
We've seen this theory implemented in practice. Let me once again refer to documents from merging parties. First, I'll quote from a document describing what actually happened immediately after a merger:
"I have already put in a 20% rate increase for 6a[m]-7p[m] time periods, which agencies typically purchase. . . . Many buyers work [station 1] against [station 2] [the two stations recently merged] to get the lowest rate possible, since both stations are `must' buys in many cases. . . . I will use our combined stations' cooperation to get some of our long-term, low dollar contracts raised to a higher rate."
And, second, I'll read from a document describing post-merger plans, which talks about:
"[w]orking in conjunction with [the station to be acquired] to raise rates. . . . One of the biggest reasons our rates are so low is the direct format competitor [the station to be acquired]. . .. Simply raising our rates by 50% which I think is possible, will accomplish our goal."
Now, let me again reassure you that such documents don't determine what we will ultimately do in any given case. Only our conclusions as to the actual market impact of the merger decides that. Still, the candid, contemporaneous views of parties to the transaction are certainly worthy of consideration when we're making that assessment.
Before moving off the topic of unilateral effects, let me take a minute to outline a variant on the "next best substitutes" theory that I've been discussing. This variant, which we call the "buy around" theory, starts from the fact that many advertisers have told us that they prefer to use multiple stations to reach their target audience. If, as a result of a merger, a significant number of advertisers cannot effectively avoid dealing with the merged firm because no set of alternative group of stations can offer as cost-effective a package, that firm is likely to have acquired the ability to raise price. Or, stated from the advertisers' perspective, if, as a result of a merger, a significant number of advertisers can no longer effectively "buy around" a single owner in assembling their advertising package, that merger would likely be anticompetitive.
The second way a merger can lead to undue market power is through anticompetitive coordinated effects. This kind of harm arises when firms in a market come to an understanding among themselves that results in limiting competition on, say, certain formats or which customers they'll deal with. Mergers can increase the likelihood of such coordinated anticompetitive effects because more concentrated markets facilitate anticompetitive cooperation between competitors. In our experience, high concentration often results in a live-and-let-live attitude where duopolists, for example, readily decide to stay out of each other's backyard. At present, we have not challenged any radio mergers because of their ability to facilitate collusion, but that does not necessarily mean that we would not bring such a case in the future.
To conclude this discussion of market effects, let me reiterate that, in every investigation, we try to put factual meat on the theoretical bones I've just outlined. We recognize that formats and demographics are really just short-hand ways to identify advertisers who may be harmed by an anticompetitive radio merger, so we look behind the short-hand to determine whether an anticompetitive effect is real. We talk to advertisers, we talk to competitors, we talk to the parties through interviews, depositions and documents, and we look at the computer programs and other tools advertisers use to judge the efficacy of their advertising alternatives, all in an effort to determine whether the choices available post-merger would be sufficiently reduced to give rise to an anticompetitive price increase to a significant number of advertisers. Only after doing this work that can we say, in a given market, that there are indeed advertisers who can't buy around radio without paying more or getting less.
Now, if our analysis finds no likelihood of anticompetitive effects, we're finished and we go home. But if we do find such effects we then go on to see whether any supply-side responses -- that is, changed behavior by current vendors or the possibility of new entrants -- are likely to ameliorate the competitive harm. In short, we recognize that the competitive harm our analysis predicts in any given case may be short lived. For that reason, we continue the analysis to take into account the market dynamics that might satisfy our concerns.
In radio, the first candidate for ameliorating any harm is the potential for a station to improve its ratings. A low-rated station, which would not be an effective alternative to an advertiser even though it is formatted to reach the same demographic group that the advertiser wants to reach, could become an effective substitute if its rating were to improve. But there are often significant reasons why a rating increase won't occur, the most obvious being signal strength. If a station cannot reach a broad enough audience, there is usually little prospect for it to improve its ratings position significantly. Also, to mention another factor, in many markets the ratings of existing stations have not fluctuated all that much over the years. The brand awareness of the top station can be so significant that it is not reasonable to assume that it will be knocked off post-merger by a station that has never been a threat before.
Second, we consider the potential effect of reformatting. Going back to Owner A for a moment, one might suggest that she will not actually be able to raise prices because of the threat that other stations would take advantage of an opportunity to make a profit by undercutting A's unduly high advertising rates through reformatting itself to reach the same demographic group that she reaches. I admit this certainly seems like a possible result, especially because format changes do occur with some frequency in radio. But, while we recognize this possibility, we also recognize that there are some real-world obstacles to a successful format change in these particular circumstances. In the first place, we would expect successful reformatting to only occur where there is a demand for an additional station and, in any given market, that may not be the case. That is, if the market cannot profitably support five rock stations (rather than four) and the new entrant cannot be reasonably assured that it will be one of the surviving firms, no firm will have the incentive to spend the money necessary to change its format to enter rock. And, as we have learned through our investigations, the cost of these promotional expenditures and the loss of advertising revenue during the course of the format change while the station looks for new advertisers can be high.
Picking up on this last point, the theory that says radio stations will jump in with new formats to defeat price increases makes the questionable assumption that it's as easy to change formats as it is changing clothes. But that grossly overstates the situation. As a practical matter, almost any existing station has invested time, money, and effort to develop its format, audience, and advertising base. If it decides to change its format, it must abandon at least some of these ongoing relationships. A station owner may do so if he's losing money but he's not necessarily going to do it just because someone else is making a little extra.
Not surprisingly, radio-station purchasers pay attention to these real-world considerations and thus value "heritage stations" at a significant premium. On the other hand, if reformatting were all that easy, one would expect to see station valuation more closely tracked to signal strength and not to the station's past performance; and as we all know, this is not the case. Instead, radio stations are bought and sold on multiples of cash flow, which reflect the underlying strength of their built up audience share and advertising revenue.
The final supply-side factor that we traditionally consider in merger cases is the possibility of new entrants. When that contingency is a realistic one, we find that incumbents may be less willing to raise prices for fear of enticing additional players to the market, which is something they normally want to avoid. But this factor has had little impact in radio cases because, given the scarcity of spectrum availability in most areas, there is no realistic chance of new entrants.
Well, by now you know more about antitrust than you ever thought you'd need or want to know. And while what I've described may sound a bit cook-booky in description, it's not in action. My staff works hard at digging out the facts and assessing the competitive concerns on a case-by-case basis. And so, while I'd like to be able to announce a hard-and-fast rule -- say, no more than a 40% in a market -- that would be unfair to you and to us. We've already approved mergers resulting in more than 40% market share, such as in Cincinnati. And I can imagine a case, depending on format and signal strength, where even less than a 40% market share might allow a company to exercise market power unilaterally when demand and supply side factors are fully considered. But I still think it's fair to say that when you're doing a deal that's in the 35% and above range -- or that consolidates a large part of a particular format (even when that involves less than 35% of the overall market) -- you should bring in antitrust counsel early on, so that you can be fully advised with respect to any problems the deal might encounter and decide whether you want to take on these problems or restructure the deal to avoid them. In fact, I suspect this is already going on and, for that reason, I'm hopeful that your relations with the Antitrust Division will be even smoother in the future. In any event, I appreciate your patience and thank you for your attention.
DOJ Analysis Of Radio Mergers
Wednesday, February 19, 1997
Updated June 25, 2015