ECONOMIC CONSIDERATIONS) FOR ALTERNATIVE DIGITAL) TELEVISION STANDARDS ) Volume:1 Pages:1 through 105 Place:Washington, D.C. Date:November 1, 1996 Before the FEDERAL COMMUNICATIONS COMMISSION Washington, D.C. 20554 ECONOMIC CONSIDERATIONS) FOR ALTERNATIVE DIGITAL) TELEVISION STANDARDS ) Room 856 FCC Building 2000 L Street, N.W. Washington, D.C. Friday, November 1, 1996 The parties met, pursuant to the notice, at 1:05 p.m. PANEL MEMBERS: JOSEPH FARRELL, Forum Chairman FCC 1919 M Street, Northwest Washington, D.C. BRUCE M. OWEN Economists, Incorporated Suite 400 1200 Nw Hampshire Avenue, Northwest Washington, D.C. 20036-6802 JEFFREY H. ROHLFS Strategy Policy Research Suite 810 7500 Old Georgetown Road Bethesda, Maryland 20814 LEE L. SELWYN President Economics and Technology, Inc. One Washington Mall Boston, Massachusetts 02108 I N D E X WITNESSES: None. Meeting Began: 1:05 p.m.Meeting Ended: 4:00 p.m. P R O C E E D I N G S MR. FARRELL: Good afternoon. I am Joseph Farrell, from the FCC, and welcome to our economists' panel on the economics of mandated standards for digital television. We have here three guests: Jeff Rohlfs, from Strategic Policy Research, who also filed a study on some of the issues we will be discussing today, on behalf of a group of interested parties, largely broadcasters; Lee Selwyn, at the end, from Economics & Technology, Incorporated, filed a study in this proceeding on behalf of a computer industry coalition; Bruce Owen, from Economists, Incorporated, filed a study on behalf of the National Cable Television Association. What I would like to do today is to try to bring out and discuss the economic issues involved in mandating a standard in general, and with particular reference to digital television. I am under no illusion that we will come to the final answer this afternoon. That will have to wait until tomorrow. But I hope we can at least clarify the issues somewhat. The way I would like to organize things is to ask each of our three guests in turn to give a presentation for on the order of 20 minutes, and after finishing the presentation the speaker is welcome to take a few minutes for clarifying questions from the other panelists and/or the people in the room. Or if he prefers, just stop. I am reminded of a famous academic story of the professor who said, "There will be no questions during class. Bad questions only waste time. Good questions are already covered in the lecture." So if anybody feels that way, they can -- then I would like to try my best on the fly a bit, since I don't know exactly what each of the speakers is going to say, to try to synthesize what I have heard and what I haven't heard, what seem to me to be the main economic questions that we need to address, and what has not been mentioned but perhaps should be. Obviously, that will be my own personal views or personal questions, and I certainly wouldn't want anybody to assume that it reflects more than my own thinking. Then I would like to invite our guests to give responses, rebuttals, replies, or whatever else they care to say for around 10 minutes in the reverse order. So in the initial presentations we will start with Dr. Rohlfs, then Dr. Selwyn, and then Dr. Owen. For the responses, we will go the other way around. And then after that we will, I hope, have free ranging discussion among the panelists, and then at some point start to take questions from the floor if we can. So that's the planned organization, and without any more from me I would like to introduce Dr. Jeff Rohlfs, who is going to describe for us his views on the economics of mandated standards for digital television. DR. ROHLFS: Thank you, Joe. I appreciate the opportunity to discuss the standards and the ATV process from an economic perspective. I think before starting the discussion of the economics, it's worthwhile though to review the interests of the parties to this proceeding. And I think it's very important to note that the supporters of the grand alliance standard all benefit if ATV succeeds. The said manufacturers will obviously sell more sets. Broadcasters have the opportunity to get a competitive advantage or to avoid a competitive disadvantage if ATV is deployed. The competitive disadvantage, of course, would be if cable and satellite use ATV, and they don't. On the other hand, the cable industry benefits if ATV fails. They benefit, if they were planning to deploy ATV themselves anyway, they would lose a competitive advantage if broadcasters also deploy ATV. If they were not planning to deploy ATV anyway, then broadcasters deploying ATV either would cause them to expend costs that they don't want to expend, or to suffer a competitive disadvantage. Both of those problems are avoided if broadcasters don't deploy ATV. The computer industry also derives no benefits from the success of ATV, and derives some benefits from its failure. There are Internet alternatives to broadcasting, and also computers are an alternative use for consumer dollars. In particular, if a consumer decides to get an advance television set this year, she may put off upgrading her computer system until next year. And I think that when the Commission considers this issue they should certainly consider the merits of the arguments that are brought up by all the parties, but they should also consider the incentives of the parties involved. What I am going to be using in my economic comments is an approach that I -- to standards, which is network externalities. That's an approach that's been used a lot in the literature on standards, and I will be referring to my own 1974 paper on network externalities which has been referred to a fair amount in the literature. The current situation in ATV, I think, and this is an important point to realize, is an equilibrium. No viewer has the incentive to buy an ATV set because if they got one there is nothing they could receive over it. Since there is no demand for the sets, no manufacturer has the incentive to produce any. And given that no sets are being produced, no station has any incentive to broadcast in ATV. And it's important to realize, no individual has an incentive to disturb this equilibrium. Everybody is maximizing their self-interest in this equilibrium, given what the other players are doing. Nevertheless, there is a possibility that everyone would be better off if ATV were broadcast and sets were produced and bought. And the problem of how to get from here to there is what I called in my paper the start-up problem, and that's what a lot of this -- a lot of the discussion about ATV is about, is how to solve the start-up problem. And a generality, in terms of solving the star-up problem in a situation like this, is it requires group action of some sort. And I think in terms of addressing what sort of action is required, it's very helpful to consider some instances where products have actually successfully surmounted the start-up problem and succeeded, notwithstanding the start-up problem and the network externalities. And let me just start off with a list here and see what implications they might have for ATV. One option, one example that obviously comes to mind is color TV, and there are two things to bear in mind here. One is that there was an FCC standard. Those of you who know that history know that it's not a pretty picture. Nevertheless, there was a standard at the time color TV was introduced. There is also another important fact. At that time RCA was producing television sets. It was integrated with NBC. And they were able to internalize a lot of the network externalities. That is, RCA, by producing sets, could benefit its broadcasting network partner. NBC, by broadcasting, could benefit its manufacturing partner. So I think that those are important factors that contributed to the success of color television. Another great success story in terms of recent products is the Internet. In the Internet the government played a very large role in priming the pump and getting the process started in the first place. In the case of Internet there definitely are standards. There are package standards. They were set by Bob Kahn and Vince Serve actually acting or essentially acting as agents of the U.S. Government. A third example where products have succeeded on their own, or two examples, are e-mail, commercial e-mail and VCRs. And in both of these cases the products have substantial value apart from the network value. For example, a single organization could get an e-mail system, and derive substantial value, even if no other organization in the economy had e-mail. Similarly, you could get a VCR and even if you were the only one who had one, and consequently there were no video cassettes, you could still derive substantial advantage from time shifting. And it's because of these non-network benefits that e-mail and VCRs were able to succeed and solve the start-up problem. And I think the point that I am trying to make is none of these offer a solution, none of these examples offer a solution that's applicable to ATV. We need to think farther in terms of how ATV can solve the start-up problem. Well, let's consider another example. What about fax? And I think the key point on fax is that they are small, and I will use the term I used in my 1974 paper, small equilibrium user sets. And I can think back to 1983, when I first went into the consulting business and joined Shooshan and Jackson, we had a fax machine, one of the original ones where you wrap the paper around the spool and then it spun around, and that's the way the faxed document got transmitted, and they were very expensive, even for that poor quality product. We were able to justify getting it largely because one of our major clients, Rogers Cable Systems at that time, also had a fax. And just the fact that we were able to communicate with one key client was largely enough to justify the cost of the fax. And I think that was true for a lot of users in the early days, and that's how the start- up problem got solved there. Actually, it solved itself because it has such large value that you can justify its use for relatively small user sets. Obviously, not the case for ATV, which is in an advertiser supported medium that requires mass audiences in order to be a commercial success. Well, this brings us now to one more example, which is CDs. How did CDs succeed? They also have a serious start-up problem. And the answer in the case of CDs is there was an agreement by all the major players in a consortium that was led by Phillips, and that's the way the standard got set there. That's a market process for setting the standard. More recently, much the same thing has happened in DVD, digital video disc, and I think it's useful reading some material I just pulled off the net with regard to digital video discs because this illustrates how a real world market process works. Let me just read some of this material. "Who is controlling DVD? The consortium. The DVD specification is controlled by a group of 10 companies referred to as "the consortium." These companies are some of the biggest consumer electronics, TV, computer and entertainment companies in the world. Typically in the $50 billion a year business volume range. Their accumulated annual budget is comparable to the U.S. Government. So their ability to bring DVD to market is formidable. All the companies are Japanese except for two European companies and one American company. They contribute patents and other technologies to the design and implementation of DVD, and agree to conform to the standard and receive a share of the licensing revenue that will be collected by an administrative entity to be designated by the consortium. Who is who? The leaders of the group have been Phillips and Sony, who developed and licensed the red book audio CD format; Matsushita, who joined Phillips and Sony in the development of the white book video CD format; and Toshiba and Time Warner, who focused on the needs of the movie and home video market. Pioneer is also important as the main manufacturer of existing analog laser disc technology. Hitachi, JVC and Mitsubishi are also included. Tompson, RCA, GE is also a member now, and has already made over two million MPEG-2 decoders for its DSF receivers. A year ago it looked like there would be two formats and a battle in the marketplace that would leave nothing but losers. At this point there is a rocky agreement to agree on a unified standard that would best serve the needs of all the companies in all the markets. Industry is involved. It is worth noting that this group represents most of the world's audio CD, TV and VCR manufacturing, with the notable exception of some Korean companies. While the standard setters of the computer industry -- Microsoft, Intel, IBM and Apple are all absent, but do have a technical interchange committee that has a strong influence. Representation of communication companies is also lacking, such as Telcos, Cablecos, DSS, Netscape, et cetera, and these companies are probably viewed as competitive content delivery providers, rather than partners in some vision of a network friendly DVD player set top box using the TV for a full range of anticipated multimedia applications. And Nintendo and Sega from the platform video game world are also absent." So this is who the players are in the DVD process. This is actually from several months ago. What has actually happened since then is they have gotten greater involvement from the entertainment industry, and have just recently signed an agreement in which the entertainment people are more satisfied that there is adequate security to prevent copying of copyrighted and widespread distribution of copyrighted material. So this is a market process in the flesh. I mean, this is what it looks like in the real world if you have a private market-setting process. And let's think about what would happen, could you do something like this for ATV. What's involved here would be that the major players get together and they reach a binding agreement. And in reality, this would not be dissimilar to the grand alliance process, except for some key differences. One is that the standard would reflect the commercial interests of the major players, not the public interest. How different that is in practice is something, I guess, that can be debated. But a second thing, I think, is a sharper difference is that there is openness in this process, in the market process, only to the extent that it serves the commercial interests of the players. And if this were to be applied to ATV, obviously the major players would include the broadcasters and the manufacturers. Probably the computer industry would play a role similar to what they played in the DVD standard setting process. Cable would have a role to the extent that it was willing to commit to deploy ATV technology, but would be given short shrift to the extent that their role was regarded as an attempt to delay or sabotage the process. So this is -- you can imagine what a market process looks like in this area. My view personally is that such a process is probably better and certainly quicker than the current process, if it were feasible, and there's the rub. And the problems with this approach, I think, are first of all, the FCC must set standards in this area. The FCC has spectrum- management responsibility. It must set standards with respect to modulation interference. That cannot be left to the market. So the FCC definitely has to have a role, unlike the DVD situation where there was no commensurate government need for involvement. A second problem is that if the market process that I discussed above actually did take place, there would be challenges under antitrust. Now, I am not a lawyer. I don't know whether those challenges would be successful. But it does occur to me that it may be no accident that the DVD process was dominated by foreign firms, and that American firms could not play a major role in that process. And a third issue, and this is perhaps the most important of all, is that once a market agreement is reached it would only be a matter of days before someone would be pounding on the Commission door and complaining that these standards that were reached by this market process do not serve the public interest, and "we want it changed". And I don't know. I mean, you can all have your, you can all, I guess, have your guesses about who would be the first parties to be knocking on the Commission's door in that event. But I would say that the real key here is that, because of the pervasive role of the government in broadcasting in particular, and also the strength of the U.S. antitrust laws, is that a private solution, if it's possible, it's very difficult, and this is -- and given that the government actions themselves prevent, or greatly inhibit if not prevent, a market solution, the government itself must offer a solution for -- have its actions cause an inevitable failure of ATV. Now, I think one of the -- one thing I think to emphasize here is that, although failure to set standards would surely doom ATV, at least in my opinion, it's by no means certain that if there were standards that ATV will be successful. One of the -- obviously, it's well understood that the sets are fairly expensive. Another thing is that the start-up problem is fairly difficult, even if standards are in place, and an important contributor in that regard is the fact that there is no -- there is not currently vertical integration in -- between manufacture and broadcasting. And it's interesting that in both CDs and DVDs, which were successful start ups, there was vertical integration between the content providers and the manufacturers. One thing that could be debated is should the government take further action to promote ATV. I mean, for example, policies have been talked about mandating that TV broadcasters use ATV, and that new TV sets be capable of receiving it. In principle, externalities can justify government intervention of this type, but one needs to carefully weigh the costs and the benefits. I mean, it's a very difficult question, whether such action is desirable, but I think it's a much easier question to set ATV standards and give the industry a chance to solve the start-up problem. With regard to the grand alliance standard itself, I think I just want to make just a few comments about that. One is on interlacing. Interlacing -- the grand alliance standard does not require, or at least what is recommended by the grand alliance is that receivers -- there be no constraints on receiver production. The receivers would not be required to receive any particular format. Broadcasters would have a wide variety of formats that they could broadcast in. One of those formats would be interlacing, which is actually the format that's used in NTSC today, but is not used in -- or which has been increasingly not used in computer terminals. As I understand the position of the computer industry, and maybe Lee can say something on this later, is their view is that the government should mandate that the broadcasters not be allowed to broadcast in the interlace format even though they may think it's in their best interest to do so, and even though it may be the result of some agreement between the manufacturers and the broadcasters. And the reason for this is that the computer industry down the line may have some application of the material and it will be difficult to -- it may be inconvenient for the computer industry to meet the needs to access this on computer terminals. This seems to me to be a clear case of the tail wagging the dog; that is, the computer applications down the line are certainly secondary. They may or may not occur at all. If they do occur, they may not have substantial value, and those considerations should not result in government mandating that the broadcasters can't broadcast in the format that they think best serves their interest. A second issue is the cost of flexibility. This is another point raised by the computer industry, and the claim is that flexibility is too costly. And I think that is always a possible concern in a public process like this which tends to go too far in trying to accommodate all the interested parties. But I think it is especially ironic in this case since much of the flexibility in the system that's currently in the standard was put in precisely to accommodate the computer industry. The choice in this regard, I think, depends on a key empirical issue, which is how great are the costs of the flexibility. I don't think economists have a lot of insight to offer on these kinds of engineering issues. But I think economists can offer insight on the following, and just consider this as a general proposition. If all the parties that stand to benefit from the success of a service favor the standard, and the major opponents all benefit from failure of the service, then I think an economist might be inclined to conclude that the odds are that the standard is pretty good. MR. FARRELL: Thank you, Jeff. Well, would you like to take clarifying questions or not? DR. ROHLFS: Well, okay, I'll take clarifying questions. MR. FARRELL: Probably not. It's pretty clear, I think. AUDIENCE PARTICIPANT: I have a question. Can you explain how the computer industry would benefit from the failure of HDTV? DR. ROHLFS: Well, the -- MR. FARRELL: Could you repeat the question back. DR. ROHLFS: The question is how would the computer industry benefit from the failure of HDTV, and I mentioned two things. I think that this is sort of a -- I think it's a combination of two things. One is they derive no benefit, and there are certain ways that they benefit from the failure. There are Internet alternatives to broadcasting as a use of people's time, the eight hours a day that they spend viewing television could be soaked up into Internet applications, some of it. And the second is the point I made about the alternative use of consumer dollars. Those are the two things I would point to. MR. FARRELL: Okay. With that let me turn to Dr. Lee Selwyn. DR. SELWYN: I do want to respond to some of the stakeholder issues that Dr. Rohlfs has raised, but I will reserve that to the rebuttal round. I thought it might be useful at the outset just to talk a little bit about the standard setting process, and what standards are for, and what they accomplish, what they perhaps don't accomplish. And in thinking about this I was reminded of a comment that I had heard a number of years ago, and I don't even remember precisely from whom, but I know it wasn't from me so I won't take credit for it. And in case it's wrong, I won't be blamed for it either. Apparently when -- in the old west when a cowboy would be out on the range and have a team of horses and want to camp for the night and not have a tree to tie the horses up to, that the solution that was adopted in that situation was to tie the horses to each other. And at least according to the way the anecdote goes, the horses, not being able to agree on running off in any one particular direction together, would basically stand in the same place. The suggestion from this, I think, is that there are many situations in which a standard serves a useful purpose in establishing a point of departure, a point at which things can come together, where the various potentially competing, conflicting, or disparate interests can nevertheless agree on something, or at least use some common point of departure in pursuing their own individual interests, innovations and ideas. The interesting thing about the particular analogy to tying up horses for the night is that basically the conclusion here is that it really didn't much matter where exactly the horses were tied up as long as they stayed still. And there are a lot of standards that have been set in a wide variety of industries and fields in which that really is the case. In other words, where the, particularly in a start-up situation, where no one has a prior investment that they are attempting to protect. Frequently the choice of any one particular specification of a standard really doesn't matter all that much as long as everybody agrees to it. For example, defining the gage for railroads, the uniformity is far more important than whether it's precisely one particular distance apart in terms of the rails. And where, for example, in Europe, in Australia, where the track gauges are not uniform, there are in fact costs that get introduced, and difficulties in transshipping goods and passengers across boundaries where trains can't travel from one railroad to another. On the other hand, if a standard is too encompassing in constraining individual decisionmaking and individual action, it can basically lock out ideas and innovations and experimentation. So what we need to do, I think, is to focus on establishing some minimum level of standard as a general matter, just enough so that the commonality will be beneficial, but not so much as to prevent innovation. And this sort of raises the idea of viewing the standard in terms of a platform. If we look to industries where this platform approach has taken place, and the personal computer industry is certainly a very good example of that, we find virtually no government-mandated standards. We find standards that are transitional; that is, a standard may exist for awhile, maybe a very strong and powerful standard, and then suddenly evaporate as something new comes along, and in each case the standards are really more market driven than they are collaborations. Hayes was the first one to produce a software controlled modem, and Hayes therefore set the standard for PC modems. As the modem technology evolved into higher speeds and data compression and other capabilities, new entrants into the market introduced new concepts -- new standards were introduced. Some of these were done collaboratively through industry associations. Some of them were done unilaterally and either succeeded or failed. If they succeeded, others followed. And if they failed, then they just died of their own volition. Modems are a perfect example of a product that has substantial network externalities because interoperability and the ability to communicate is extremely important. A modem that doesn't communicate with other modems is not particularly useful. So as we focus on the ATV issue the question basically is, are we looking at a market, are we looking at a technology paradigm that is closer to railroads, or is it closer to computers. And I would submit it's a lot closer to computers. And in that context I think it's important to try to establish a framework in which a platform approach that establishes some minimal baseline upon which various interests can build, upon which markets can experiment and either prove products to be useful or useless can proceed. Now, the basic idea -- well, before I get to that, that then raises a second order question. If the standard itself should not be all-encompassing, should the standard be government mandated or should it be industry set and enforced? And once again, the role of government is often useful in a start-up situation, particularly where there are stakeholders with disparate perspectives and where agreement on a single standard is really essential to proceed, and might not be expected to occur absent some external intervention by a disinterested party, such as the government. But the government set and government enforced standards have serious disadvantages. Not the least of which is their inflexibility. It takes a long time to get them established. It takes a long time to get them changed. Industry is really better able to develop and to modify standards either by consensus or simply in response to marketplace forces. So once again, it would appear as if what makes sense is to keep the role of government to that minimally necessary to satisfy what might be characterized as a police function; for example, to issues of interference and frequency assignment clearly are government functions. But going beyond that, the standardization should be left primarily to industry, which brings us to the subject of digital television. I totally agree with Jeff that this is an issue that is dictated by network externalities. It is -- they are pervasive in addressing this issue, and it's critical, I think, to understand how they will operate in order to evaluate the various alternatives. Put simply, a digital TV set that can't receive all digital broadcasts is not going to be terribly useful. And to the extent that the incentive to broadcast digital programming, and particularly high definition programming, is directly related to the number of receivers that exist in the marketplace. The inability of certain receivers to receive all formats will probably have the effect of slowing the development, or perhaps even preventing the development of, in particular, HDTV. Now, so we have two solutions that have been proposed, and they, I think, have some very significant differences that need to be explored. The grand alliance proposal would contemplate, depending on how they are defined, as many as 18 different formats that could be used by broadcasters. And these formats would have to be received, that is, receivers would have to be capable of decoding and displaying each and all of these formats since a priori the consumer does not necessarily know in advance which particular format or formats are going to be selected by individual broadcast stations within the consumer's market area. In effect, the grand alliance proposal, while professing not to require receiver standards, is requiring receiver standards. In fact, the grand alliance proposal just doesn't work. It doesn't make any sense without mandatory receiver standards. Now, let's sort of think about what happens here in terms of the way the market works. First, I don't think we can presuppose a priori consumer demand for high definition television. In fact, if one had to make a judgment based on revealed behavior of consumers, I think it would be an odds on bet that the demand for HDTV is very limited. Now, I base that conclusion on a number of things. First, if we look at the existing NTSC format, which has been around for a long time, what we find is that the vast majority of consumers are purchasing low end equipment that is capable of displaying only a portion of the potential resolution that is available in the NTSC signal. Your typical 150 - 200 dollar table top, 19-inch receiver does not display more than about 50 percent of -- 50 to 60 percent of the signal detail that is actually being broadcast. The typical $200 VCR certainly displays even less of the signal detail relative to what exists within the NTSC signal itself. Consumers buy VCRs by a factor of about 50 to 1 over laser disc players despite the fact that laser disc players provide a quantum improvement in image quality. And we don't have to look to the television field to see this same phenomena. Consumers buy cheap cameras rather than -- in overwhelming numbers compared to cameras with high quality lenses. If you impose a requirement that HDTV capability be -- at least from the perspective of decoding the HDTV signal rather than, let alone displaying it -- have to be incorporated into the receiver, then the receiver will be costly. Or if someone can come up with an inexpensive way of doing it, it will produce a low quality picture. If the receiver is costly, and doesn't display, and doesn't provide any particular added functionality to the consumer, the consumer won't buy it. If the consumer doesn't buy these things, then we don't get to HDTV. We may not even get to digital TV, and we certainly don't give back any of the existing NTSC channels. The computer industry proposal comes at this a somewhat different way. It says 'let's agree on a minimum baseline that every digital broadcast signal must conform with, and that every digital receiver or decoder must be capable of getting. So whether the broadcaster chooses to continue to broadcast in SDTV format or in an HDTV format, irrespective of what particular aspect ratio a broadcast is offered in, what particular aspect ratio the source material has been produced in, there would be a standard baseline format that anybody who purchases a digital television set or a digital decoder would be capable of receiving.' What this means is that the broadcaster would be assured of an audience whether or not an HDTV signal was actually being offered. Now, let's contrast this to the grand alliance proposal. The grand alliance proposal presupposes that manufacturers will voluntarily produce receivers capable of receiving all formats. Now, notwithstanding any assurances that may have been given in filings before the Commission, this is, in my view, an extremely unrealistic assumption. If consumer demand for high definition television is limited, and yet consumers confronted with the possibility of losing NTSC channels are being forced to purchase digital receivers or digital decoders, there will be demand for low end units, and low end units that are capable of receiving only the SDTV signals will go on the market. Broadcasters at that point will be confronted with the following dilemma. If they broadcast a high definition signal, it won't be received by some substantial chunk of their audience. And if the audience doesn't -- isn't capable of receiving the signal, advertisers won't be willing to pay for the transmission. The transmission simply won't occur. I would submit that the grand alliance standard actually has a lower probability of actually getting to a high definition world than the computer industry proposal. The computer industry proposal will get to HDTV if there is demand for HDTV. And if there is no demand for HDTV, it won't happen. If there is demand for HDTV, consumers will buy TV sets that will be capable of receiving them. Broadcasters will transmit programs that will be capable of being displayed in high definition format. And if this demand doesn't exist, then that simply won't happen. But we still can move away from NTSC into a digital format. We still can exploit other capabilities of a digital format. For example, an expanded aspect ratio in a standard definition mode which might be far more attractive to consumers, for example, than a high definition image itself. High definition images require, in order to be worth anything, a fairly large TV screen; much larger than about 90 or 95 percent of existing TV sets in the installed base currently have. It's not really clear that consumers are going to be willing to pay or will even in many cases have the space in their apartment for a large unit, a large piece of furniture of this sort. So what this does, it seems to me, is offer a result that brings us a lot closer to the objective and at potentially much lower cost to the consumer. Consumers will only be required, if the objective is to migrate them out of NTSC and into digital, to purchase baseline units. If a consumer purchases a baseline unit, which will cost roughly half or so of the typical high definition converter, the consumer will be assured the ability to receive every single minute of broadcast on a digital channel. If there is demand for high definition TV, some consumers will purchase higher end units. And if consumers buy them, TV broadcasters will offer this programming. If it's necessary to somehow stimulate demand, there may be other ways of doing that, such as by providing subsidies to advertisers to encourage the development of programming or whatever it takes to sort of overcome the start-up problem. But the main point is that where broadcast stations broadcast HTV signals, they will be received by all owners of digital receivers, whether or not those receivers are HDTV capable. Now, what's in it for the computer industry? I think it's sort of ironic that Jeff suggests that the computer industry views themselves as competing for the consumer's dollar. I think it's quite the contrary. There is clearly a convergence of computer technology and entertainment, and that convergence is centering on the display device. Computer screens are getting larger as graphical user interfaces Windows, Mac-type interfaces become more standardized and more established. The 14-inch computer monitor that has existed for the last 10 years or so, or even more than that, is giving way to 15 and 17-inch monitors. You are seeing more of these being sold. And the potential exists for the PC and the TV to be integrated into a single component, a single appliance. Internet-capable surfing devices will be offered on the market that can use the TV receiver as a display, and PCs will be capable of receiving and processing television pictures. In fact, there is a likelihood that by supporting convergence of demand for computer capability as well as broadcast-type entertainment, that demand for digital- capable receivers would actually be stimulated. These very tertiary income effects that Jeff describes, in terms of competing for the same dollars or competing for the same time, I think are really minuscule compared to the potential complementarities between these two industries. Computer text displays, particularly in high definition graphical interfaces, simply don't work on interlace screens. It's as simple as that. You can't look at windows desktop on an interlace screen unless you want to significantly restrict the amount of information that gets displayed. It would be very difficult to imagine doing any serious Internet access and Internet use with the kind of resolution that would be capable for text on an interlaced screen. So it's entirely reasonable to look to the future and not to the past. Not look to a standard that was set long ago, but to look -- or that was deigned long ago before people talked about computer networking and computer access, and to look to the future. Even the Internet standard that Jeff referred to, I think, is a perfect example. Even if we consider the basic packet transmission and protocols that have been government-mandated, I'm not totally sure that's true, but even if we accept that, it's for sure that the Worldwide Web is not the result of a government standard, government-imposed standard. And yet it's the Worldwide Web and Web browsers, such as Netscapes and Microsofts and others that are coming down and being made available to consumer, that are producing the demand, and these are being built on that basic Internet platform. It's precisely the kind of condition that we would like to see occur in the case of digital television. There is simply no reason to lock in today a format or a set of formats that won't even receive significant market acceptance for the better part of a decade under the best of and most optimistic of assumptions when we have on idea sitting here today what the consumers' needs will be at that point in time. A basic platform approach with flexibility to adapt that platform to respond to demand is clearly the only thing that makes real economic sense. MR. FARRELL: Thank you, Lee. Any clarifying questions? Chuck? AUDIENCE PARTICIPANT: Yes. I would like to ask one question. You described this sort of hierarchical different standard and I guess I can recall when I was in a special panel in '93, early '93, that suggested a banded analog approach to HDTV and going to digital. And I guess my question is this new standard you are proposing, would it take a comparable four or five years from today until it will be ready for delivery? And secondly, would it be the -- earlier this month the cable industry chose a digital standard, which is fundamentally a subset of the NTSC standard, would the standard you are talking about be compatible with those digital cable boxes and cable-ready TV sets that would be distributed between now and when the new standard was developed? MR. FARRELL: Okay. Chuck, you spoke in an admirably loud voice, but I was getting signals, I think, from the back that it still wasn't being picked up on the monitors, so let me try to repeat your question. The computer industry proposal, would it lead to substantial further delay because -- well, for whatever reason? And the second question was would it be compatible with what the cable industry recently decided on? DR. SELWYN: I can't answer the second question. Maybe Bruce can. And I will try to get an answer, hopefully before the end of the afternoon. And with respect to the first question, it is my understanding that the computer industry proposal has been demonstrated and has been shown to be workable. I am not an engineer, and I can't give you a precise figure as to what its status is in terms of development relative to the grand alliance proposal. But I would submit that I don't think that the difference is significant. The computer industry has coalesced around this proposal. I think a lot of the delays associated with the grand alliance standard were the result of simply reaching the compromise within the alliance itself. And, in fact, the grand alliance is not a standard at all. It is simply a collection of 18 or so standards, and doesn't even really reflect a true coalescing around a single thing. A slight delay here, if in fact there is one, and I'm not sure that's even the case, inasmuch as we're not going to see serious market penetration for many years, it would be easily worth it, rather than making a fundamental error as significant as this one, we have estimated the potential consumer cost of pursuing the grand alliance standard on a mandatory basis of something in the range of $100 billion if we assume a 10-year mandatory transition and shut down of NTSC channels at the end of a decade. Certainly before the economy imposes $100 billion worth of costs on consumers, the Commission ought to be absolutely sure that it's doing the right thing. MR. FARRELL: Thank you, Lee. Oh, further question? AUDIENCE PARTICIPANT: In terms of the cost analysis and cost to consumers, Dr. Selwyn, wouldn't you have to figure in not only, over the course of say, a 10- year transition, but ad infinitum those costs that would be borne by consumers as they need to upgrade their TVs over time to keep track of, keeping up to date on the new operating services and equipment and hardware and software. Doesn't that have to be figured into the costs that would be borne by consumers? DR. SELWYN: You're using sort of the example of the fact that we seem to go through computer generations every year or two. AUDIENCE PARTICIPANT: Exactly. DR. SELWYN: Well, you know, there is -- no one is shoving new computers down peoples' throats, or passing laws or promulgating regulations that simply state that 486 or 386 computers have to be thrown away. There are those of us, myself included, who still use DOS, and these are consumer choices. If in fact there are new capabilities, functionalities that come along down the road and that consumers want, then consumers will allocate their disposable income accordingly. But what -- and in the context of their computer purchases today, many consumers do that. But what we are talking about here is an imposed, a government externally imposed conversion that in many respects doesn't do anything for the consumer in terms of increased functionality or capabilities. It simply replaces an existing -- in fact, we based our calculations on the cost of converters. The assumption being the consumer would simply use the NTSC receiver, but put a digital converter box on top of it. And in that situation the consumer is not buying a whole hell of a lot for the $100 billion. The consumer is simply getting the capability to get signals because the existing receiver is incompatible. What you are speaking of are cases of voluntary choices where consumers are trading disposable income for increased functionality, and consumer surplus. If that happens, that's fine. I don't think that's a problem. It doesn't create a deadweight loss on the economy. A $100 billion that produces no particular consumer benefit, on the other hand, creates an enormous deadweight loss. AUDIENCE PARTICIPANT: But, in essence, if you are not going to be making the purchases for those upgrades, what you are left with is a minimal baseline standard that really gives you a product not any better than what you get on NTSC now. MR. FARRELL: Maybe you -- DR. SELWYN: But your solution simply says let's stop progress altogether. In other words, you know, God forbid a consumer should decide that they want -- that he or she wants something better and is willing to pay for it, we ought to simply prevent that form hitting the market, and not give the consumer the opportunity to make such a choice. That makes absolutely no sense. MR. FARRELL: Okay. Thank you, Lee. The third speaker is Dr. Bruce Owen. DR. OWEN: Thank you, Joe. I guess I can summarize my view within the context of the recent dialogue by saying that if the Commission were responsible for setting mandatory standards for computer operating systems we would all still be using CPM. For those of you who aren't as old as I am, that was the predecessor to DOS. (Laughter.) Let me make three or four affirmative points, and then at somewhat greater length comment on the points that were made in the reply comment filing by Jeff and by Chuck and their colleagues because that will capture a lot of my reaction to what Jeff said here a few minutes ago. In summary, my most basic point is that when a broad-based industry group comes to the government and asks the government to in effect pass a law making it illegal for either them or anybody else to do things in a way other than the way prescribed by the standard that they propose, we have all got to watch for our wallets. Something is fishy. If the standard is a great standard, it will surely be adhered to voluntarily. And if it isn't, we shouldn't pass a law that forces people to adhere to it involuntarily. It's perfectly appropriate, don't get me wrong, it's perfectly appropriate for the FCC to have facilitated the agreement that has been reached on the ATSC standard. There is nothing wrong with that. There are good reasons why private industry has difficulty, private markets have difficulty in reaching standards, agreements, in certain contexts. I don't agree that the antitrust is one of them, but things like externalities, network externalities and compatibility problems do exist, and I think there is a very useful role for the government to play in facilitating agreement on voluntary standards. Where I have a problem is when the government mandates a standard and makes it unlawful for anyone to deviate from that standard. Let me turn first from my first affirmative point to my own characterization of what the economics literature has to say about this. This isn't everything the economics literature has to say, but it strikes me as a fair summary of the relevant things it has to say. First, a particular standard can either support a monopolistic market structure, and inefficiency, or it can break markets open to competition, and promote economic efficiency. In other words, on a legitimate level a standard may either contribute or not to economic welfare. A competitive market may or may not develop useful standards. It may do so either too early or too late. There is no invisible hand that drives private markets to produce the efficient outcome here the way there is with respect to goods and services generally. When private markets fail to develop timely and useful standards, there is no practical basis for the government to conceive a remedial course of action, no practical basis. It isn't that one can't perceive that there is a need for a standard, and that would be better than the absence of a standard. It's that the government has no basis for knowing what standard or when to impose it or to suggest it. And, finally, government intervention in the market in imposing a particular standard, imposing a particular standard, is as likely to be harmful as helpful because, again, the government has no basis for knowing which standard is the right one, and therefore can easily make the wrong choice. It's one thing for the private market to make the wrong choice because there are forces that can remedy that mistake. It's quite another for the government to make the wrong choice in, in effect, passing a law, because it is much more difficult to change that. Now, I am not -- I want to make this clear -- I am not making the standard sort of council of despair, second- best argument that is often made with respect to government policymaking. As those of you who are economists know, our models tell us what an efficient outcome is and what the conditions are, but when something, or even a lot of things don't meet the assumptions we have made, as a practical matter it's often difficult to know what the right policy should be. Standards is different from say an environmental externalities, or monopolies, or other market failures because at least in those cases you know in what direction to go. With standards, there is no guidance as to what direction to go in. We know there is too much smog in L.A. It is not right to say we don't know whether there is too much or too little. We know there is too much. We know that a monopoly charges prices that are too high. That certainty, to the extent it's helpful, is not present with standards. It is necessary to distinguish compulsory standards from just plain old standards. Virtually everything, not everything, but virtually everything that Jeff said today about standards, all the examples have to do with standards are a good thing. I don't disagree that standards are a good thing. But with one exception, color television, none of them was a compulsory government standard. Not one. I think you have to look long and hard at the economics literature on standards before you can find anyone, any model, any commentator, who is even seriously considering, much less endorsing, the general benefits of the government imposing compulsory standards in cases where markets fail or have failed. And where you can find such examples, they are almost always in the context, and my reading, always in the context of models where the whole model is focused on the difficulty of coming to agreement on an efficient standard, and no consideration is given to what's going to happen next year. No consideration is given, in other words, to the possibility that the year after an efficient, an efficient, an appropriate standard is imposed by the government, a new technology arises or circumstances otherwise change in such a way as to make the old standard extremely inefficient. Just to drive the point home, a substantial change, a substantial technological change, or an innovation takes the form of products and services that break with old standards. That's what an innovation is. A compulsory standard in that world, in that context, is a formula for retarding innovation. Standards in general aren't, although they can also have that effect, but compulsory standards is a guarantee of retarding innovation. Why is that? Well, because they are difficult to change, much more difficult than voluntary standards. Voluntary standards are difficult to change. It's true that industries focus on standards privately agreed upon and adopted, whether through agreement or through decentralized decisionmaking, and adhere to those standards after the standards have become obsolete, even though some superior new technology has arisen. That can happen. It does happen. But with compulsory standards it's guaranteed to happen, because in addition to the various economic forces that keep industries from moving to new and better standards, or to adopting new and better technologies, are added the difficulties attendant upon changing a government policy. Interest group politics, I mean, that's part of -- we are one small atom in the interest group politics of this issue. And part of the reason it takes a long time to come up with a standard with respect to this particular technology is the government's involved in it. If the FCC didn't exist, if spectrum markets were private, we would be watching digital television today, assuming of course that it is a great innovation. I was tickled when I went to the Web page of Strategic Policy Research and discovered that the slogan and blazing on the top or the bottom of the page is somewhat unpoetic, but compelling, "Competition is preferable to regulation." I am going to turn now to the various criticisms that were made in Jeff and Chuck's reply comments of the comments that I submitted on behalf of NCTA in the first round of this phase of this proceeding. First, they say that I ignore the overriding goal of free over-the-air universal television service that makes broadcast television unique. According to Jeff, the market will take longer to settle on an ATV standard than the government. Well, I don't think that's obvious at all. There are lots of different players. There are compatibility issues. There are manufacturers of sets. There are broadcasters. There are programmers. And there are viewers. And none of them, as he said, seems to have an interest in doing anything, assuming the others are doing nothing with respect to ATV. Well, watch those assumptions. I mean, if that were an accurate description of the way these equilibria come about, we would be walking around with clubs, wearing tiger skins. Innovation does happen even when there are compatibility problems. Innovation does happen when the CD player manufacturers are not the same as the consumers or the same as the people who produce the music. I mean, it's a terrible overstatement to assume that government-mandated standards are required in order to avoid nothing, this equilibrium in which nothing happens. I don't really know anything about the AM stereo issue. Apparently there was a standard that was developed by the private market, or wasn't, and the whole thing cratered. I read an FCC staff report that said that the reason for that was lack of demand in the market. I don't know whether that's true or not. There is certainly out there in the world thousands, tens of thousands, maybe more, standards set without the benefit of the FCC's intervention, voluntary standards. Some of them set with help from government. Some of them voluntary government standards. Most of them private. And the world is working. Universal free over-the-air television service is, of course, a goal that economists don't automatically sign on to, but it may be a goal that this institution wants to pursue. I don't see any reason to suppose that it's threatened with a delay or threatened by a delay, assuming there were one, in the diffusion of digital broadcasting. I mean, think about it. There are -- they are the same four groups of players here. There are broadcasters who want to sell advertising. There are viewers who, we will assume, want to watch digital programming, HDTV programming. There are manufacturers who would like to sell television sets and broadcasting equipment, and program suppliers that want to supply program. It's in their interest to trade. There are gains from trade here. It's a strange equilibrium. Not impossible. But a very strange equilibrium which leaves them all sitting on their thumbs, particularly when the world around them is producing competitive challenges. I was struck when Jeff said that the cable industry stood to gain from the failure of ATV. I don't really understand that. It seems to me the cable industry here is helping the broadcasters save themselves. A majority of viewing of cable television is local over-the- air television signals. If those signals disappear, or become unpopular, cable has no way to differentiate itself from what's up there in orbit broadcasting directly to the home. And so cable's competitive advantage vis-a-vis DBS disappears. It's not in the cable industry's interest for anything bad to happen to local broadcasters. Far from it. I am accused of ignoring the vast flexibility of the ATSC/DTV standard. Apparently it includes 18 substandards. I am not an engineer. I have no idea what that means. That's very flexible. Apparently not flexible enough for the people in Hollywood or the people in Seattle, but still very flexible. They even say it's as flexible as the Internet protocols that were developed 20 years ago. Well, the Internet protocols were not imposed by the FCC or by any other government agency. It's true, as Jeff says, that development of the Internet was subsidized by the federal government, by the National Science Foundation, and by ARPA, but they didn't impose standards, and the world is free to adopt new standards with respect to the Internet. If it's really as flexible and useful as its proponents say, I don't see why there can be any serious fear that the new standard won't be voluntarily adopted. It's only if you fear that somebody won't adopt it that you require a law to require them to adopt it. Third, they say that I ignore the need to protect existing TV service from interference. Well, what do Senator McCain, Justice Breyer and President Clinton's Council of Economic Advisors have in common? What they have in common, among other things, I hope, is a preference for performance standards over design standards. That is, to the extent standards are useful, it is most useful to specify what you want them to accomplish rather than how to accomplish it. With respect to interference, I simply don't understand why the Commission can't permit people to do whatever they want to do with respect to digital broadcasting, so long as they don't impose unacceptable levels of interference on others. That's a performance standard. Allows a lot of flexibility, even more than 18 ways to do it. This standard, the ATSC standard is a design standard approach. The whole argument here in favor of the Commission adopting this standard as a mandatory standard requires you to believe several things that don't have very much basis. It could be true. They don't seem to have very much basis. For example, they require you to believe that the proposed standard is at or near the optimum, given present technological possibilities. Second, they require you to believe that a new technology or future circumstance won't arise shortly after the present standard is adopted, because, of course, that would be -- that would make the proposed standard inferior, but they would nevertheless be mandated. Or you have to assume that the services that could be provided by ATV under this standard are best provided by terrestrial broadcasting rather than some other means. They require you to believe that ATV ought to succeed. Actually, that assumption explains an awful lot of statements made in support of the standard. If you believe that ATV, by which I mean the standard, ought to succeed, then a lot of what's proposed makes perfect sense. If you don't accept that assumption, or think that it's capable of being questioned, then you may not. You have to believe that the present massive industry support in favor of this standard would wither or evaporate if the Commission did not make it a mandatory standard as opposed to a voluntary standard. And, finally, you are required to believe that the social costs of diffusion delays, delays in adopting the new technology, if any, that would attend the adoption of a voluntary standard exceed the social costs of creating a barrier to the adoption of new technology in a field where, as we all know, technological progress has been very rapid. So to sum up, I think standards can be good. The government may in some circumstances facilitate agreement on a useful new technology standard that might otherwise be delayed or worse. But I contend that some standards are inefficient and harmful, and the government may in some circumstances by imposing them coercively reduce economic welfare, and I do not understand how, except with hindsight, we can tell those two things apart. I don't understand if the ATSC standard is as good as its proponents contend, the government must impose it on the economy by fiat. I don't understand why a nonexclusive endorsement of the standard would not be sufficient to ensure its adoption. And, finally, I don't understand why it would be harmful to the public to permit other standards to be employed, as well as this one, without any further regulatory approval or consent, so long as they impose no greater burden of interference than that imposed by the present standard. Thank you. MR. FARRELL: Okay. Do you want to take clarifying questions? Okay, any clarifying questions? Yes? AUDIENCE PARTICIPANT: I wonder if Mr. Owen believes that the non-adoption, independent of some form of ATV, any form of ATV, by cable and by satellite providers should lead us to any conclusions about the level of demand for ATV services? MR. FARRELL: Let me just repeat the question. I'm sorry. AUDIENCE PARTICIPANT: Or, if I can. MR. FARRELL: Sorry. AUDIENCE PARTICIPANT: Or if this whole proceeding explains, adequately explains why those independent sources of programming haven't adopted it? MR. FARRELL: Okay. Why haven't nonbroadcast programming distributors adopted some form of ATV? What can we infer from that? What do you know about why? DR. OWEN: Well, it seems pretty clear that if there is a consumer demand for ATV products, then nonbroadcast media are more likely to adopt it if there is a greater supply of programming in that format, and there would have been a greater supply of formatting in that format if a standard had been adopted for broadcasting earlier, either by the market or by the government. So I don't think you can separate the incentives and the demand issue facing nonbroadcast media from the interference that's been going on from the government. I hope that's clear. The answer is I don't know. (Laughter.) AUDIENCE PARTICIPANT: Of course you don't know. MR. FARRELL: Yes? AUDIENCE PARTICIPANT: Two points you have raised, one about the imposing a compulsory standard at the end. You talked about a nonexclusive endorsement. Could you further distinguish between where you are in one camp and where you are in the other camp, or what the difference is? DR. OWEN: The question is would I distinguish between a mandatory standard and a nonexclusive endorsement of a standard? I am just distinguishing between an FCC rule that says that you've got to do it this way, according to the ATSC standard or any other proposed standard, and one that says you may do it this way, but you may also do it in any other way that doesn't cause interference. I realize that you can't just say it doesn't cause interference. It's technically more complicated than that. But one which is nonexclusive in the sense that it permits other entities to adopt standards other than the one embodied in this particular proposal. At the same time it gives the proponents of this proposal the benefit of an FCC blessing for the particular standard that they are -- that they have developed, which can't hurt at least in enhancing its chances of being adopted. Chuck? AUDIENCE PARTICIPANT: Can I follow up with a question on that point? One tradeoff you can think in the design of a digital modulation standard would be a tradeoff of short-range capacity versus service of longer distances say with equal interference capability. And do you think when the Commission is managing the spectrum under the model it does in the '34 Act, that they should necessarily leave the tradeoff between increased urban service and decrease rural service to the market, or should they impose that on the market? DR. OWEN: I will repeat the question, but I don't have much hope that by the end I will have an answer. The question was given the extent it would be possible to -- there is a tradeoff, a technical tradeoff between the number of channels you can produce for a local audience and the distance that you can broadcast, so that, for example, you could have 50 channels that would only go for an area two miles in the diameter, or you could have three channels for an area 200 miles in circumference. Is that -- AUDIENCE PARTICIPANT: That's kind of the idea. DR. OWEN: That's the idea. AUDIENCE PARTICIPANT: Not quite, but that's fine. (Laughter.) DR. OWEN: Please, I'm struggling as it is. Given that possibility, should the Commission under the '34 Act take account of that tradeoff in setting its spectrum policy? AUDIENCE PARTICIPANT: Well, in particular, I mean, do you think that is a tradeoff that should be left to the radio transmitter operator, or do you think that the Commission might wish to constrain the operator's freedom say to ensure rural service, or to ensure more urban service? Who should make that tradeoff, the broadcaster or the FCC? DR. OWEN: Well, as I hope I have made clear, I'm in favor of market decisions here. This is an area where there has been an awful lot of government intervention. And so market decisions are distorted by that intervention. But having thought about your -- the circumstances of your question for all of 20 seconds, I guess if I had to choose, I would choose the broadcasters. But I don't think that's an adequate period of consideration. MR. FARRELL: Okay. Thank you very much. Well, now I am going to perform the death defying act of trying to summarize on the fly at least some of what I heard this afternoon's speakers say, and perhaps even more death defying, try to identify how I would try to organize some of these things and possibly even questions that they haven't addressed. So it seems to me the right place to start for economists thinking about the possibility of mandating a standard is to ask what are the market failures that would make it not a good idea potentially to leave the choice up to the market? What are the government failures that might make it not a good idea for the government to choose a standard? And then based as far as possible on thinking about those things, in what circumstances does a government mandate potentially make sense? And do those circumstances seem to apply here? On the market failures, I think Jeff Rohlfs told some stories that, as Bruce Owen pointed out, were largely, perhaps not entirely, and you can discuss this in your reply period, largely about problems with initial coordination; that is, market actors, all of whom hypothetically would like to get to a certain agreement, but have trouble reaching such an agreement. And the horse's story, for which I think I have to accept responsibility maybe as the most colorful version of that, there is a slightly different, although closely related, possible market failure which has to do with subsequent splintering away from a generally acceptable consensus. The reason I say it's closely related is, of course, the splintering away might not really be subsequent. That is, as soon as you get into a world in which the incentives are not all perfectly aligned and it's more than a pure coordination failure, then you have to look at the incentives for splintering. And we do know, I think, from the academic economics literature that the incentives for splintering, the incentives for departing from what might otherwise be a general consensus standard are not necessarily optimal. This is part of what Bruce Owen was referring to in mentioning that there is no invisible hand in the -- at least so far as we have seen so far -- in the academic literature of standard's choice. And it's probably right, I think, to say that the splintering issue, the incentives to depart from what might otherwise be a consensus standard, is likely to be closer to the things that we need to think about in contemplating mandating a standard that is claimed to be -- to attract a lot of consensus. So perhaps we need to focus a little more on that and less on the coordination problems. Turning to government failures, I think the speakers identified two or three. One, of course, is the imperfect information. Imperfect information both about technology and about the degree of consensus required, and the direction, let alone magnitude, of the market failures to correct. And there is also the point that if the government chooses a standard and things later change, we may be unduly slow to change in response. Trying to bring those things together, well, obviously those are somewhat imponderable issues, and I think that's where we need to try to ponder the imponderable is what can we say about the relative importance of the market failure, the government failures and the directions perhaps of both in this case. Some of the speakers tried to do that. I would certainly like to hear more on that. A related question is what I might call the cream skimming question. Cream skimming has a bad name around here, but let me try to give it a good name. Can we do something by way of cream skimming in the sense of a partial mandate or a mandate with some set or something else that might potentially derive quite a bit of the benefits from one approach without too much of the cost? Lee Selwyn mentioned, I think, if I'm not misquoting you, that it could be the government intervention may help at first in getting an industry going, but retard later changes. If we thought that was the right model, that might argue for a sunset approach. Or if we thought that there was some aspect to the technology that are fundamentally an arbitrary choice on which there ought to be a coordination, and other aspects where there are real quality and innovation issues, could it make sense to mandate some and leave others aside? By the way, on cream skimming, I realize we usually think of taking the cream as getting a good part. I will put in a plug for the American Heart Association, and suggest maybe we should turn that metaphor around. But either way, you skim off the cream and take one part of the other. Okay, then let me try to organize some thoughts into thinking about incentives within an industry and incentives across different industries. Within an industry, let's say within the broadcast industry, you have an issue of coordination of initial commitments, and that gets, I think, to some of the vertical relationship points that Jeff Rohlfs was stressing. It also gets to the points about development of a consensus, what, if anything, is necessary or desirable to help the consensus develop? And does the government need to do something once there is a consensus, or is consensus a signal that the government can step out, perhaps never even needed to step in? Bruce Owen also raised the antitrust specter here of saying beware of industries asking for mandated restrictions because they may be up to no good. He didn't give, I think, any specifics, although thinking about subsequent innovation, obviously might lead one to the idea that asking for a mandated standard might be a way of asking to be spared the perhaps costly subsequent R&D competition to develop something different. I don't know whether that's what you had in mind or whether you have other fears from this mandated standard. Turning to incentives across industries, and here I am not talking about video versus sugar, but I'm talking more about broadcasting versus cable versus DBS versus perhaps computers, Jeff Rohlfs made an interesting claim about the incentives of the various parties; that those who would like to see broadcast ATV succeed, he told us, tend to be on the side of the grand alliance standard. And those that he argued would like to see broadcast ATV fail tend to be on the other side. That's certainly an interesting idea. A perhaps related question is different segments of the video industry presumably have different preferences about what a video marketwide ATV standard might look like. Just to take the most obvious example, nonbroadcast media probably have less binding, surely less rigid bandwidth restrictions than terrestrial broadcast. Does that lead them legitimately to want different things? And how should that difference in preferences -- I say preferences but, of course, to economists preferences are something that consumers have, and it's only some kind of derived preferences that industry segments would have -- how should those differences in preferences be taken into account in a private or a government standard's policy? And then might there be anticompetitive concerns along the lines that Jeff Rohlfs perhaps was hinting at in saying, well, if you do something different, maybe part of the motive or part of the effect, and who knows about the motive, might be a perception that it would harm broadcasters' attempts to develop this spiffy new broadcasting technology, and that might conceivably be good for cable, or as Bruce Owen pointed out, perhaps more plausibly be good for the DBS industry. Is that something that we should worry about? And if so, how? Changing the subject dramatically, compared to a lot of -- compared to a lot of standards choices it's been suggested, and it seems to me to make sense, that there might be an additional social benefit of speed in the current context because we may hope that once the DTV market develops fully in whatever sense that's turned out to be defined as we get some spectrum back for other uses. Well, what does that additional social benefit of speed tell us about how the process should be managed? Lee Selwyn, I think, argued that this is a consideration that argues for not burdening the transition with the requirement of high definition. Then he also argued, I think, that the grand alliance proposal doesn't make very much sense if you don't require high definition. But what might be said on the other side about the importance of speed? Finally, turning to some slightly more nitty- gritty, not very nitty-gritty, but slightly more nitty- gritty issues of possible standards, two things that have been raised in this docket, one of which was talked about more today than the other. One is standard definition digital television versus high definition. Lee Selwyn argued that most consumers don't even spend the money to maximize the definition they get in NTSC. Why should we think they want high definition? And should that choice be left up to market forces? And a second question that I know has been controversial, I don't think a lot of time was spent on it this afternoon, is progressive scan versus interlaced. And I think we need to ask a couple of questions here. One is if there is -- if there were to be a mandated standard audit to resolve those questions, and audit to ensure full compatibility in the sense that any receiver that conforms to the standard can receive any broadcast that conforms to the standard, kind of all option receiver, although that sounds more like a football term than I think I am used to saying. Well, Lee Selwyn, again, argued that most people perhaps don't really want high definition, so it might be a mistake to mandate a full all option receiver in the sense of including high definition. On the other hand, if you have a standard that consists of a lot of incompatible options, in the sense that broadcasts conforming to one option cannot be received, not just cannot be received optimally but cannot be received on receivers conforming to some other option, is that really a standard, and what exactly are we doing there? I think Jeff Rohlfs was saying that the all option receiver idea might be going too far. Lee Selwyn, I think, argued that not going that far would undermine the very network externality justification for a standard at all. All right, well, that was my, as I say, death defying attempts to summarize and provoke a little bit. The plan now, as I said at the beginning, is to go back to our panelists, go back to them in reverse order and ask each of them to take up to about 10 minutes to respond to various things, whatever they want to say, but perhaps particularly replying not only to the other panelists, but to my amateur summary of what they themselves said. So, Bruce Owen. DR. OWEN: Thank you, Joe. I've just got, I think, five brief comments. First, the fact that parties take positions consistent with their economic interest doesn't mean their arguments are wrong. Second, Jeff said that the FCC must set standards given the framework of the legislation that it operates under. With respect to spectrum allocation, it must set standards. And, of course, that's true. But I don't know of anything in the legislation that requires the FCC to set performance standards as opposed to -- I'm sorry -- design standards as opposed to performance standard. Third, the issue of antitrust has been raised in a number of different contexts. And as a non-lawyer, I can speak with perfect abandon on the subject, although it's true that I was once chief economist of the antitrust division. The first sense in which antitrust has arisen is in the context of whether or not a coalition of industry members seeking government mandates to enforce their agreement is a competitive problem or not. Well, of course, as you know, when competitors reach an agreement in restraint of trade each one of them has an incentive to cheat, and enforcing the agreement is a problem, and one object of the antitrust laws is to make it difficult to enforce those agreements, the agreements themselves being unlawful. One way to deal with that problem is to come to the government and have the government enforce the agreement. And, of course, there are lots of industries over the years who have done that -- optometrists, lawyers, others. But that's nowadays frowned on by enlightened policymakers. In other words, there is a presence dilemma problem, to be specific here. There is presence dilemma problem facing the members of the grand alliance, especially to the extent they are manufacturers. Each one of them may think they have a technology which is superior to the technology described in the grand alliance standard. Each one of them may think that they could do better. If the standard is adopted voluntarily, anyone of them might cheat, come into the market with their own Beta or VHS or the equivalent. And it's in the interest of the group as a whole to prevent that from happening, and so it is sensible for them to seek government regulations preventing anyone of them from cheating on the agreement. Now, as I said several times, it may very well be that this standard is the economically optimal standard. Okay, in which case one ought to discover this cheating for the moment. The problem is one doesn't know for how long. The second sense in which antitrust has come up has been in the context of whether or not it would be -- whether antitrust concerns prevent industry groups for agreeing voluntarily on standards; that is, just people worrying about antitrust prosecution stop them from agreeing on standards. And I really thinks that's not a serious concern. The Justice Department has brought very few cases and they have been very carefully targeted in this area. They have brought two kinds of cases. The first kind is where the standard in question is price, right. And the second is where the standard in question is explicitly adopted for the purpose of excluding a competitor, a particular competitor. And most people would agree that that's a sensible thing to prosecute. Beyond that, there has been very little government frowning upon industry groups that come together to set standards. For example. SAE has been doing it for years now, a real industry with no problem, Society of Automobile Engineers. At one point towards the end of his remarks Jeff said something which I wrote down as carefully as I could. If I have got it wrong, I apologize, but I am going to read it. And if it's what he said, I want to say I agree with it. He said, "We should not mandate that broadcasters can't broadcast in a manner that serves their interest." This is in the context of interlacing, but I think it's perfectly general. Yes, I agree. Finally, there is this issue of equilibrium, and getting stuck in an equilibrium where it's in nobody's interest to change, even though there is a superior alternative standard or technology. It's easy to think about how difficult it is to overcome that kind of problem and forget how often we solve it; how often the market has solved it, or we can think of the clever ways in which it is solved. One of the ways that happened recently that struck me is the way Netscape solved the problem of selling servers to people who use network service given that they were no clients out there with software that could use the servers. What did they do? They gave away the client software, and they are selling apparently a lot of servers. They actually had a profit. There is a way to overcome this network externality compatibility problem, but a very clever one, and I am sure there are many others. Thanks. MR. FARRELL: Thank you. Lee? DR. SELWYN: Thanks. I think what I need to do is take a cue from the tactics of the presidential candidates and try to use my time to first answer a previous question. Chuck Jackson raised the question about -- two questions with respect to the timing of adoption and development of the computer industry proposal versus the grand alliance proposal. And with respect to that, I am informed that basically there should be no difference in timing the computer industry proposal is -- to the extent that it is other than a subset of the grand alliance proposal, it has been demonstrated, it has been shown to be operative and should not produce any delay. And with respect to the question of compatibility with the cable industry, digital TV standard, I am informed again that, to the extent that the cable industry standard is compatible with the grand alliance -- with a segment of the grand alliance proposal, so too is it compatible with the computer industry proposal. So these are nonissue. I wanted to try to respond to some of your points, Joe, but also I wanted to address some of the suggestions with respect to stakeholder interests. I have absolutely no difficulty understanding the interests of the TV set manufacturers in their desire to sell $100 billion worth of television sets to the American public. That's pretty obvious. I have a great deal of difficulty understanding the genuineness or genuiousness (sic) of the broadcasters' interests in promoting high definition television. What do the broadcasters get out of high definition television? Well, what they get out of HDTV are additional frequencies, additional channels. If HDTV succeeds, in fact, if any digital TV format succeeds to the point where the NTSA -- NTSC channels can be returned, then what the broadcasters get out of that success is losing the very additional channels that they had previously been awarded. On the other hand, if HDTV fails, but enough HDTV receivers are, or at least digital receivers are out there to justify some continuation, perhaps for a further extension of the originally proposed transition period, or maybe even a definite transition period, what the broadcasters get out of the failure is they get to keep that additional channel for a longer period of time. So it's hard to imagine why the broadcaster would be supporting HDTV other than for the purpose of obtaining additional frequencies, and that their interest in doing that is also -- is as obvious as the interests of the TV set manufacturers in selling $100 billion worth of equipment. Your suggestion, and the suggestion of something short of a permanent full-blown standard such as a standard that sunsets, it seems to me that that is essentially what the computer industry is proposing. Not so much something in terms of sunsetting, and I will get to that in a second, but something that is less than pervasive. The computer industry is proposing a baseline, a platform. Satisfy that, and every digital television receiver and converter that is sold will be capable of receiving every digital television program that is transmitted, pure and simple. The grand alliance can't say that, pure and simple. Now, with respect to sunsetting, I'm not sure that sunsetting is per se important because the reason for getting the standard started is to overcome the start-up problem, and once it's overcome, whether or not the official standard is sunset may not make all that much difference if the standard had been accepted. It's kind of like, you know, starting a fire. Once you get it going it takes on a life of its own, and deciding that the fire should never have been started in the first place doesn't make it stop. So it seems to me that the computer industry is proposing precisely this notion of a minimal standard that is being put in place to just get things going to a point where there is some likelihood of success of digital television. I just can't see how the grand alliance proposal gets to any give back of channels unless one takes on faith that consumers are clamoring for high definition television. And we are so far from proving that or knowing anything about the complexion of consumer demand for high definition television that it seems to me that accepting something like that on faith is simply not possible at this point in time. With respect to the inter and intra industry incentives, the grand alliance proposal clearly benefits its two principal industry groups; for different reasons, but they clearly benefit. But as I think we have outlined here today, there are other industries that have a stake in this outcome, in the result. Let me give you an example. You had mentioned that nonbroadcast sources of software that are not restricted to the same bandwidth, bandwidth constraints as broadcast television might, might in fact be able to introduce services that couldn't be carried or introduced capabilities that couldn't be carried in an over-the-air format. For example, I an envision a situation in the future, and now I am going for the sake of this discussion taking on faith that at least some subset, some subset of consumers, enough to matter, are interested in not just high definition TV, but let's say ultra-high definition TV, where they want large wall-size screens with perhaps substantially more detail than even the HDTV formats are being proposed. Now, what you would like to be able to do in a situation like that is let the market, if the market wants to produce a video disc of some sort or some other media that can support that for movies or the like, and be a superset of an existing broadcast format, then that same equipment could be utilized very effectively to support a broadcast and a nonbroadcast application. And it's that multi-industry interoperability and compatibility that makes, it seems to me makes a lot of sense, and that's why we need flexibility and not locking in a specific set of capabilities and locking out all others. Similarly, the issue of aspect ratio. We have seen the coalescing of the grand alliance around 16 by 9, but in a recent letter from Leonard Nimoy to Commissioner Chong, he suggested that what the film industry really wanted was an aspect ratio of 2 to 1 or even greater than 2 to 1, and he signed that letter -- his salutation in that letter to the Commissioner was "Live long and prosper," which I thought was, as a confirmed Trekkie (sic), I was pleased to see that. In any event, it seems to me that we just don't know enough about demand, we don't know enough about technology to lock in anything more than the most minimal level of assurance of some success. The computer industry proposal is the only one that assures that all receivers get all broadcasts, and that doesn't impose inordinate costs on the consumer for functionalities that they may not particularly want or need. And I think it's consistent with precisely the kind of things that you are suggesting. MR. FARRELL: All right, Jeff? DR. ROHLFS: The examples I gave at the beginning were to illustrate how difficult the start-up problem is for ATV in particular. One of the issues that's always raised, well, there are a lot of other products that have succeeded without standards, what's different here. And what's different here is that in order to have success here it's necessary for two different industries to coordinate activities; namely, the manufacturers and the broadcasters, in addition to consumers. That's not the case in modems. That's not the case in Worldwide Web, but it is the case in two other examples that I cited, which were CDs and DVDs. And it's interesting to see that in a real world situation how is it that the market solves the standards problem. And the answer is they reach an agreement, all the parties sign an agreement that they will not use any standard other than the agreed upon standard. And there are time frames which are set in terms of the commercial interest of the players, but that's essentially what's happened in those industries, and that's the way standard setting actually gets set in a market when you have a difficult start-up problem like this one. Now, so far as I can see, there is very little difference between the Commission mandating a standard and the Commission simply blessing the process that the product parties have gone through. For example, if the parties were -- you could view the process as this, as we the industry, the grand alliance, has come up with a standard. We want to agree to that standard, sign on the line that we are going to abide by this standard and no other. We want you, the Commission, to determine that our doing that is in the public interest. Now, if the Commission were to do that, I don't think it would be very much -- I don't think the consequences would be very much different than if the Commission were to mandate a standard. What's essentially the problem is that if they are unable to make such an agreement or if they do make such an agreement that could be challenged before the Commission as not being in the public interest, then the standard -- then you don't have a standard. Then the industry can't -- then nobody can build a set because they don't know how to build it, and the industry never gets out of the null equilibrium. So the issue here, I think, is not -- is not government compulsion. The industry wants to agree to a standard. It's just that because of the pervasive government role in broadcasting the government needs to be a party to the agreement because otherwise it can be challenged as not being in the public interest, and the FCC can't avoid that role. That in fact if they were to now decide they are going to leave it to the market, the inevitable consequence is some years from now they would be having a proceeding on whether the market standard is in the public interest, being essentially where they are now. The second point I want to raise is that the grand alliance standard involves a lot of flexibility. No one is required to build a receiver that receives all 18 signals. Nobody is required to broadcast in any particular format. And what Lee said is that -- Lee essentially is saying that that's -- he is not fighting that particular standard, which is the one that's actually proposed, what he is criticizing is a much more pervasive standard where receivers are required to pick up all 18 signals. DR. SELWYN: That's not what I said. That's not what I said, but go ahead. DR. ROHLFS: The standard where essentially one can imagine things evolving in many different ways. For example, some broadcasters have already said that they intend to broadcast in progressive format. If it's true that it costs a lot of extra money to accommodate interlace, then it's quite possible that the sets will not be built to accommodate an interlace format, at least in the high definition range, although they probably would in the low definition range. That could be -- that could come out of discussions among the industry participants after the -- after an agreement is reached about setting the general parameters of the standard. There is a lot of flexibility built in that is actually agreed upon by industry participants. What is really being proposed here is that the computer industry, which is really a nonplayer in this process, is going to set the standards, and the government would then, by mandating these minimal requirements, would essentially limit the flexibility, and this is not an agreement of the players that are actually in the industry, broadcasting or producing the sets. It's an agreement that's being imposed on the outside by a nonplayer. It's a result that could never happen in a market-driven process. And I guess the final thing I want to sum up with is the -- it really is true that the people who support the standards are the people who will benefit if ATV succeeds, and I will say what I said again about the broadcasters. The broadcasters, the broadcasters benefit from broadcasting in ATV because they avoid a competitive disadvantage because they can't hold ATV back forever. ATV will be getting to consumers via cable. It will be getting to consumers via satellite ultimately. And they can be left out of the process with an inferior product unless they can rally behind some sort of an ATV standard. So I think the broadcasters have an interest in success. The manufacturers have an obvious interest in success. And the cable industry, perhaps the most likely scenario -- perhaps the most likely scenario for them is they don't want to see it because if ATV comes in in over- the-air broadcasting, then they need to expend a lot of money. Their cash flow is a little constrained lately. They would just as soon not face that problem. I accept what Bruce said, just because you have perverse interest doesn't mean that your arguments are wrong, but nonetheless, I think the Commission needs to bear in mind incentives in evaluating the arguments. MR. FARRELL: Thank you. Do you want to -- Lee? DR. SELWYN: Yes, I just want to clarify one point. My point with respect to the grand alliance receiver standard is this. If the receivers -- if the receivers are not capable of receiving all 18 formats as a matter of a mandatory requirement, then the market will drive the prices down, and the capabilities down to whatever consumers are demanding. And that will mean that in the marketplace less than all of these formats will become available. What is likely to happen then is that since the proliferation of the lower priced units that are not capable of receiving TDTV broadcasts at all, that will just display blank screens in that case, will simply then make TDTV unavailable even in standard definition format unless the -- under the grand alliance proposal. So what I am saying here is that if the Commission goes with the grand alliance proposal, it must enforce a mandatory receiver standard or else it will never get to TDTV, and let me give you a very good example of something that falls precisely in this category. SVHS is a much higher quality image than standard VHS, but SVHS takes cannot be played on standard VHS VCRs. They just don't -- they are not playable at all. As a result, there are very few, if any, I have never seen a rental or for sale pre-recorded tape that is in super VHS format, and consequently that entire technology is basically dies on the vine. There aren't very many of those machines that are sold because you can't get software for it, and there isn't any software because it's incompatible, it would be incompatible with the 90 plus percent of machine that it cant see. And I see exactly this outcome with respect to HDTV if receivers will go blank when an HDTV broadcast comes on. So if the standard goes with the grand alliance standard, it will have no choice, if it wants to get to high definition television, to impose mandatory receiver standards for all format. It just doesn't make any sense otherwise, and that's what imposes the largest cost on consumers. MR. FARRELL: Very good. Go to the floor. Jeff. AUDIENCE PARTICIPANT: Yes, I want to ask a follow-up question to that. I see Jeff Kraus in the audience, and I understand, if memory serves me right, did you ever purchase a pre-recorded super VHS tape -- AUDIENCE PARTICIPANT: Twice. As a matter of fact, "Raiders of the Lost Ark" is available in super VHS. I believe others are also. DR. SELWYN: Well, I haven't seen them. They certainly are not very -- they certainly don't proliferate. AUDIENCE PARTICIPANT: If I can just clarify something. Is it your position that you believe the higher bulk of services associated with VHTV standard, which equate at various levels -- is it your position that that would be entirely on manufacturers -- and will not receive certain signals, that they will base their product lines on those types of sets and be incapable of -- DR. SELWYN: No, that's based on the assumption that the -- that every product does receive all signals. What I am saying is that because consumers will not want to pay for a product that will receive a signal that doesn't do anything for them on a standard NTSC receiver, that they use this converter box to -- in connection with, that the market will ultimately force lower end less than fully capable ATSC units to be offered. AUDIENCE PARTICIPANT: Consumers would actually purchase a set that would go blank if -- DR. SELWYN: If consumers -- if consumers -- well, if -- AUDIENCE PARTICIPANT: Is that your belief? DR. SELWYN: Yes, that is precisely my belief. Consumer -- the only reason -- if consumer demand -- I am operating on the basis here that consumer demand for digital television is not demonstrated, and that the only reason that consumers would buy this at all is because they will have heard from their government that if they don't do something their NTSC receivers will go black permanently, and for all time. So they will have to do something. And when that happens consumers that don't see any specific benefit from digital television will obviously postpone the purchase as long as possible, and at such point in time as they are forced to make that purchase, if they don't see any benefit, if there isn't must HDTV around, they will look for the cheapest way to accomplish what they are trying to do, which is to simply preserve their ability to get any TV at all. AUDIENCE PARTICIPANT: It's not -- DR. SELWYN: Well, who says that -- broadcasters won't broadcast in HDTV if there aren't very many HDTV capable receivers. AUDIENCE PARTICIPANT: That just seems very unrealistic with that type of set [inaudible]. DR. SELWYN: Well, I think it's even more unrealistic to think that the American public will spend, whether it's 50 billion or 100 billion dollars, on something that they don't actually want. And I think long before anything of that sort is imposed on the American public somebody ought to really determine whether or not that demand exists. You know, just because you set a standard doesn't mean that people will flock to buy it. AUDIENCE PARTICIPANT: On that rate, do you figure capability or incapability? Which is it? DR. SELWYN: Well, we have estimated a transition cost over a 10-year period with replicating to the best that we could the transition pattern, the migration pattern for colored TV, but collapsing it into a 10-year time frame instead of the 20 years that it roughly occupied. For the fully capable ATSC device at approximately $100 billion, assuming that Moore's law halves the cost of the VLSI components every 24 months. When we modified that assumption to a 36-month process, the cost increased to about $150 billion. We also attempted to test the hypothesis that there would be both high end and low end units that would be offered, that is, high end, fully capable, and low end, SDTV only, and when we -- when tested that we made the assumption that 80 percent of consumers would buy the SDTV only units, and my recollection is -- I don't have it on front of me -- that I think our number there was in the 60 some odd billion dollar range, and we estimated the transition cost for units capable of receiving the baseline signal under the computer industry proposal at a little over 50 billion. MR. FARRELL: Okay. Let me ask for the panelists, when you take a question if you could repeat the question so that it goes on the tape and so on. And also ask questioners to phase their questions in a way that can readily be repeated. Yes? AUDIENCE PARTICIPANT: I'm a little concerns. Mr. Chairman, you raised the issue in your summary the notion that there are social benefits associated with the speed of the return the spectrum. I am a little concerned that none of these panelists have really -- have really looked at that issue. We have recently gone through an experience where the FCC has auctioned off 90 megahertz of PCS spectrum, and that brought in about $15 million. And we're talking here roughly about returning roughly that same amount of spectrum for auction but at a frequency that's 10 times lower in the VHS spectrum rather than two gigahertz. So the media is 10 times as valuable since it's 10 times lower frequencies. But maybe it's not quite 10 times as valuable. Okay, it's certainly in the tens of billions and maybe hundreds of billions of dollars of value to the country. That's with out regard to how that spectrum can be reused for some other purpose and benefit the country. I am little concerned that none of these panelists have addressed that. MR. FARRELL: Okay, let me try to repeat the question. I think the question is what about the potentially very large social value and perhaps monetary value of getting spectrum back? And how does that affect your analysis and your views? Let me ask any and all of the panelists to comment on that. DR. OWEN: Well, I will take a crack at it just to start. It seems to me that this is an issue that arises largely because of the certain rigidity in thinking about the nature of the problem that you face here. We're talking about whether or not the Commission should mandate a standard for a particular kind of broadcasting. Let's assume that the Commission doesn't mandate a standard, the Commission waits for the market to do it. There is a completely, and I mean completely separate issue of what, if any, broadcast spectrum to allocate to that service in addition to what's now allocated to broadcasting, and what the transition measures should attend, if any, should attend that. I don't think it has anything to do with whether or not it's a good idea to mandate a standard, and I think that the linkage between those two issues is one that need not constrain the Commission. DR. ROHLFS: The -- it seems to me if I were to, if I were to address the issue, what I would say is that the real problem here is that the Commission is going to take two years now and think over the computer industry filing and the cable industry filing, and everything will be delayed two years, including the return of the NTSC spectrum. And while that may serve the interests of certain parties, it doesn't serve the -- it doesn't serve the public interest. I think the important thing is that the -- the Commission has taken long enough to develop a standard. They have a standard now, and it's time to -- and the way to get the rapid return of the spectrum is to move forward instead of dithering. DR. SELWYN: I think the real problem here is that the broadcasters get their spectrum for free, and the alternate use that's being proposed would have to pay for it. So we don't really have the opportunity to have the marketplace make the kind of economic decision on the best use of this resource. If we really were serious about addressing this problem, we would make the broadcasters bid for those licenses against other uses. And if we did that, then we would really determine the correct way to allocate the resource and the correct social costs of the two alternatives. But the very fact that the broadcasters don't pay and other users would pay distorts the relationship and creates precisely the very perverse incentive that I have describe that the broadcasters have, which gives them a tremendous incentive to hold onto those free frequencies for which there is no out-of-pocket cost to them for as long as they possibly can, and the value of that far exceeds the dubious benefit of those small -- the small segment of the consumer market that actually wants HDTV. MR. FARRELL: Yes? AUDIENCE PARTICIPANT: Given the fact that -- bring about the establishment of HDTV, and in determining the players [inaudible]. MR. FARRELL: The question is wasn't the whole point of this to make sure there is high definition, and if there were to be no requirement for high definition, wouldn't that undo the whole purpose? DR. SELWYN: I think that gets to precisely the point that I have been getting at. The broadcasters have been very resistent to any mandatory minimum broadcast requirements for high definition signals. If that is the purpose of the additional channels, and the transition structure that is being described and proposed by the grand alliance is to overcome the start-up problem due to the presence of network externalities, then the solution is to give these licenses to the broadcasters expressly on condition that they use those frequencies for HDTV, because then at least the consumer would have some reason to want to go out and buy HDTV-capable receivers. But if the broadcasters have no obligation to broadcast HDTV, there is no particular reason to expect that they will if there are no receivers out there capable of receiving those signals. AUDIENCE PARTICIPANT: So why should we give them the spectrum? DR. SELWYN: I agree, I don't see why we should, not unless they are willing to commit to some specific minimum requirement. Now, you know, one of the possibilities under a more flexible standard, for example, is if the -- you know, there are really two separate issues here. We could migrate to SDTV to get HDTV entirely, and basically have the same amount of -- same number of broadcast channels occupying substantially less spectrum than it does today. So there are really two separate decisions that have to be made, and they really should be separated as economic matters. The first issue is the return of the existing analog frequencies and replacement of those frequencies with SDTV digital channels, which occupy perhaps a fourth of the bandwidth. Now, that could be treated in much the same way as the process by which, for example, users of private microwave have been compensated when their frequencies have been usurped or reassigned, and I suppose that one could overcome the social cost issue if the value of the released frequencies exceeded the cost of equipping consumers receiving units with SDTV-capable converters. Then some economic process could be established to flow a portion of the proceeds of the spectrum auctions back to consumers to, in effect, pay them for the equipment modifications that are required, and you could still end up perhaps with a substantial profit to the government. The second issue is do we want to allocate four time perhaps as much spectrum for the purpose of getting to HDTV, and that's a separate issue. And if the answer to that is yes, and the broadcasters are going to be given these licenses, then it seems to me that there has to be a condition that requires that they actually go forward and offer HDTV. And we haven't seen that. We haven't seen that kind of commitment forthcoming from the broadcasters. MR. FARRELL: Yes? AUDIENCE PARTICIPANT: Dr. Selwyn, if I understood you correctly, you have estimated that the additional cost of making digital televisions capable of receiving HDTV as opposed to SDTV is roughly twice. It's roughly $60 billion if only 20 percent of the sets sold will do that, and it's $100 billion if 100 percent can do that. Have I interpreted your analysis correctly, and do the other panelists agree that that's the right breakdown? DR. SELWYN: Well -- MR. FARRELL: Can you just repeat the question? DR. SELWYN: Yes, let me repeat the question. The question is, is my -- is it correct to interpret my analysis that the cost of converters, assuming 20 percent have HDTV capability, and 80 percent SDTV only would be about $60 billion over a 10-year transition versus $100 billion for 100 percent SDT capability? That is correct with respect to converters. That does not include the costs of SDTV display receivers themselves. In other words, we are speaking here only of devices that are capable of taking the SDTV or the HDTV, as the case may be, signal and transforming it so it can be displayed on an analog NTSC television set. It's consumer who actually want to see HDTV pictures in HDTV format would then have to spend additional money to acquire the large screen HDTV displays. That's not included in that estimate. AUDIENCE PARTICIPANT: And do the other panelists agree that the HDTV-capable converters for the NTSC set would be twice as expensive as the NDSC -- excuse me -- the SDTV-capable -- DR. OWEN: I have no idea. DR. ROHLFS: Yes, I have no idea either. AUDIENCE PARTICIPANT: Thank you. DR. SELWYN: And, you know, I am an economist, not an engineer. And the source of -- the source of my information, which is cited in the papers that were submitted with the computer industry comments, was studies that were conducted by industry, knowledgeable industry members who looked specifically at the components of this equipment, and attempted to project the change in costs over time based on the operation of Moore's law and the assumptions with respect to increased manufacturing efficiencies based on higher volumes of product. AUDIENCE PARTICIPANT: Can I make a comment -- MR. FARRELL: Behind you. Yes. AUDIENCE PARTICIPANT: Yes. Regrettably, I or anyone else who knows the answers to these questions were not invited to be on the panel. I won't even bother to phrase it as a question to Dr. Selwyn, but just note that numerous parties filed for advice to this study. People who genuinely do know something about putting these products that are showing that those costs are based on completely fallacious assumptions and have no basis in value. It pains me to hear them repeated as if they are the value and should not be the base of a discussion on them. DR. SELWYN: Okay. May I respond to that, please? MR. FARRELL: Yes, let me just repeat it. DR. SELWYN: Okay. MR. FARRELL: The comment was that the cost estimates are wildly, irresponsibly wrong. DR. SELWYN: We have looked at the reply comments and the most obvious observation about these criticisms of what we have done is that they sort of ignore the entire distribution channel, and go solely to manufacturing. The consumer electronics industry as a general matter experiences manufacturing costs running at about one- third of the cost of the final product when brought to market, to the retail market and sold to the consumer. And if you in fact take the estimates that have been provided for manufacturing costs in those reply comments to which you just referred, and you multiply them by three, you come up with -- which is what you must do in order to provide the conversion for manufacturing cost to delivered retail product -- then you come up with essentially the same price level that we have estimated. In addition, it also appears that in developing those replies assumptions were made with respect to the quality of the unit and the capabilities of the unit that perhaps reflected a lower quality than we had assumed. But our assumptions were consistent and based largely upon statements by members of the TV manufacturing industry with respect to cost relationships that are consistent with those, and the information that you have just referred to is simply incomplete, and therefore is irrelevant. MR. FARRELL: Yes? AUDIENCE PARTICIPANT: How did the marketing for all this hardware get arrived at? Do you have any [inaudible]. I go into buy a new TV set once every ten years or so. The last time I was in there maybe two years ago, and it was difficult to get any information at all about the quality of the [inaudible]. There is no information to make a choice. MR. FARRELL: The question is how does the retail marketing of television receivers factor into all of this? DR. OWEN: If you watch television, you will see a lot of ads, and they will explain to you which one is best. (Laughter.) MR. FARRELL: You have to have a television first. Yes? AUDIENCE PARTICIPANT: The discussion has focused largely on cost of receivers and receiver volumes, and we have a wildly -- wildly successful HDTV service. And I don't hear anyone thinking about 10, or 20, or 30 years down the road should that scenario prove out with the expected drop in the cost of receivers. But the other element, the spectrum element and the broadcasting service, how do you factor in the value of that public interest justification, so to speak, for those decades of service into this scenario as to whether or not a standard is worth the bang for the buck? MR. FARRELL: I think the question is what's the public interest component of your thinking, and perhaps also how do you think about what's happening 30 years rather than five to 10 years down. DR. ROHLFS: I think the focus in terms of standards discussion and a lot of the policies is really the start-up problem, and the issue is if we can -- if we believe that there may be an equilibrium in which a lot of people are -- purchase HDTV sets voluntarily, and they think they got a good buy, manufacturers are making a profit, broadcasters are benefitting as well, everybody is better off. But the question is how do we get from there to here. And getting from there to here, getting from here to there involves solving a very difficult start-up problem, which is what we have -- we have been focusing on here, I think, today. MR. FARRELL: Yes? AUDIENCE PARTICIPANT: Yes. Mr. Rohlfs, I have a question for you. It seems that one of the arguments in the document holds ATSC standards is it allows broadcasters the flexibility to pursue broadcasting in whatever particular format that they feel possible for the broadcast. However, you also stress that mandating receivers capable of receiving all of the HDTV format is not preferred. To me, that seems to preclude the argument that adopting a full standard would provide more flexibility to broadcasters. I was wondering if you could outline the benefits of adopting the full grand alliance standards and -- I'm sorry -- the NTSC standards, and not a receiver standard. DR. ROHLFS: The benefit is it would allow flexibility in outcomes, depending on how the facts evolve over time. For example, one way the facts might evolve over time is that Dr. Selwyn's estimates are completely wrong and the cost of providing flexibility are a tenth or a fifteenth of what he said. In that case, the natural equilibrium would be that the sets would probably accommodate a wide range of formats; maybe not the whole 18, but a wide range of them. Another is that the set manufacturers, when they are actually getting ready to produce, discover that it costs a lot of extra money to accommodate particular formats, and they talk to -- and then there could be discussions among the industry participants that it's going to cost a lot of money for us to produce these sets. How about if you don't broadcast in these formats that are very expensive for us to accommodate. And as a result of those discussions a preferable equilibrium could be developed. This allows flexibility in terms of the market processes beyond the initial standard setting process. Now, I am not sure I would really recommend -- I'm not sure I would recommend this, and I think I need further study, but it certainly has the advantage of providing flexibility and not locking -- not locking us into, prematurely into deciding which of the 18 or whether we want to all 18 now, that that can be postponed to the future. But it does -- it does require further coordination among the parties, and there are some costs associated with that as well. MR. FARRELL: Yes? AUDIENCE PARTICIPANT: If you will tolerate me with one additional comment in response to this discussion, the industry has already examined the issue very closely and found that an almost negligible additional cost, even with the first unit, all 18 formats to be supported. And the consumer electronics industry is saying there is no need to mandate the 18 formats because through an entity they all stand up and say they wouldn't dream of providing anything that did not include all of the 18 formats. So it's not an issue at all. MR. FARRELL: Okay, the comment is it will be cheap to provide all 18 formats, and so it really wouldn't be much of an issue. DR. SELWYN: Can I ask a question to the questioner? MR. FARRELL: Sure. He gave an answer. Yes. (Laughter.) DR. SELWYN: Now, you know, roughly speaking, in order to support HDTV you need more memory, you need more processing capability than for SDTV. And if I interpreted what you just said is that that doesn't cost anything. AUDIENCE PARTICIPANT: That is not true. I highly recommend that you read the reply comments in your own study. You could learn a great deal. It turns out that techniques have been demonstrated by Hitachi and others where you can do partial NPEG decoding, use the same amount of memory that is possible for an SDTV format, and a less capable processor to decode the full HDTV quality format and SDTV quality at about the same cost. There is no necessary additional cost. That should be great news. DR. SELWYN: My understanding of the -- of the response was that the Hitachi format offers a solution to this. My understanding of the Hitachi format is that it operates on an incremental image differentiation process that creates jerky images, particularly for fast action type of material, and that it probably is not an acceptable alternative. It's not -- it's not simply cutting down or stripping out the excess detail from the high resolution image, but it's sort of losing portions of the -- of the movement and then correcting itself periodically when things get too far out of whack. People we have spoken with who have seen it demonstrated report that it is not something that consumers would find acceptable. AUDIENCE PARTICIPANT: I have seen it, and I don't share that opinion. DR. SELWYN: Well. AUDIENCE PARTICIPANT: It's better now than their first try at it. They have improved it and they are continuing to improve it. MR. FARRELL: Okay, the comment was Hitachi is improving and it's continuing to improve. Yes, in the back. AUDIENCE PARTICIPANT: Just to respond to the statement of Hitachi. You might be interested to know that Glen Reitmeyer said the Hitachi quality is so low that it's not good enough to sell. And he's at Sarnoff Labs so that should probably -- MR. FARRELL: Okay, so disagreement on Hitachi. AUDIENCE PARTICIPANT: I might add that we saw the demonstration of Gary Demos' version of the baseline standard, and he thought it absolutely unacceptable, and didn't even show half [inaudible]. MR. FARRELL: Okay, and the Demos standard is also being maligned. (Laughter.) MR. FARRELL: Yes? Yes? AUDIENCE PARTICIPANT: Mr. Owen, you discussed the performance standard versus design standard, I believe. Would you distinguish more clearly the difference? DR. OWEN: Yes. A -- MR. FARRELL: Do you want to repeat the question? DR. OWEN: Yes. I am asked to distinguish more clearly between design standards and performance standards. A design standard says that the brick wall has to be made up of bricks that weight 3.7 pounds, and are two inches by six inches by 2.7 inches, and have the following clay content and titanium dioxide content, and the mortar has to be to the following exact ingredients, and the wall should be eight feet high. And a performance standard says the wall should be eight feet high. AUDIENCE PARTICIPANT: And in the RF application, what would it be? DR. OWEN: The equivalent -- the equivalent in an RF application is that the interference levels at particular points, geographic points, can't exceed certain levels at given frequency ranges. MR. FARRELL: Yes? AUDIENCE PARTICIPANT: I had just two simpleminded general questions. How does -- it doesn't seem to be be direct involvement in the decisionmaking process of the public about what costs will be imposed [inaudible]. The second part is has the economists estimated the cost [inaudible]. MR. FARRELL: Okay, just to repeat the questions, is it worrying that consumers don't seem to have an active role at the present point in time? And secondly, what is the costs, broadly interpreted, of delay in this decision? DR. ROHLFS: Well, the role of the consumers is very simple. Nobody makes any money if they don't buy the HDTV sets and watch the programming. So the consumers are, at least in that sense, are being -- are part of the picture all along. I mean, I can't think of -- it's unusual, for all I know, for consumers to participate directly in standard setting process, but if they don't like what comes out they don't have to buy it. And I think that's the general answer to the involvement of consumers. I mean, with regard to the cost of delaying a year, the question is could I estimate it? Yes. Have I done it yet? No. MR. FARRELL: Lee. DR. SELWYN: Well, this is not exactly a totally voluntary act by consumers. If the Commission and the administration decide that at some point in the future, some identified point in the future NTSC analog broadcasts will cease, consumers will be forced to either give up TV entirely, or make a purchase of something. Now, what we have going for us with respect to issues of delay, the costs of the electronic components that are needed to support converters and digital receivers are dropping, and they are dropping rapidly. So, in fact, from the consumer perspective, a delay of even a few years actually potentially reduces the transition cost, and it's very sensitive to the year in which purchases are made. If you delay it a few years, those cost do drop. Now, offsetting that is the delay then in the date at which the administration would be able to, whoever watch it is at that point in time, would be able to realize the revenues associated with the reclaimed analog frequencies. But the consumer decision is not a free decision. It's being distorted in large part because of the fact that these valuable frequencies are basically being used without payment by their present licensees, and at some point a payment will be received. But it really isn't obvious to me that there is very much social cost of delay if you get rid of the distortions or at least discount the distortions that exist in the present relationship between what broadcasters pay for the use of frequencies, and what other users will be required to pay once these frequencies are auctioned off. MR. FARRELL: Yes? AUDIENCE PARTICIPANT: I've got a question for Dr. Selwyn. You talked about the broadcasters having an incentive to keep the set frequency. Once you have a transition to digital made, why do they have that incentive? It seems like they would have costs. What benefit are they deriving from that of keeping the original analog channel if they have full-blown digital service? DR. SELWYN: Oh, if they have a full-blown service, then that's fine. But if that transition is never completed. You know, most people have multiple, two, three, four television sets in their homes. The average is something, I think, around 2.5 or thereabouts. And the consumer, you know, may be willing to spend some significant amount of money perhaps to convert the primary set, the one in the living room, but the 9-inch screen in the kitchen or in the bedroom, you know, might be delayed. And if the penetration rate of digital receivers is not sufficient to justify the elimination of NTSC channels, then the broadcasters get access to both the existing as well as the new channel. They can use the additional channel for multiple, potentially for multi- channel or digital frequency allocation to support multi- channel SDTV broadcasts, or HDTV, if they have that flexibility. If there is no market for HDTV, it's quite possible that they would benefit by actually being able to simultaneously offer free over-the-air service as well as perhaps several pay TV digital channels on that -- on the new digital frequency allocation, while at the same time maintaining their analog channel and the market that it serves. MR. FARRELL: Okay. For the benefit of our television audience, the question was what is the gain to the broadcasters from keeping the second channel if indeed there is an effect DTV transition. Of course, regular television viewers are good at guessing the question from the answer in any case. Yes, Chuck. AUDIENCE PARTICIPANT: Yes. A question for Dr. Selwyn. At the beginning of your response time you responded to my earlier question about the recent cable lab's standard for digital cable. And if I heard you correctly, that standard, which is general instruments symptom information, MPEG-2, main level, quality AC-3, and QAM. Did I hear you correctly? Did I hear you correctly when I thought I heard you say that that could be displayed by the Cicats display box? MR. FARRELL: The question is the cable lab's announced standard, can it indeed be displayed by the Cicats display box? DR. SELWYN: Kevin, do you want to help me out here? AUDIENCE PARTICIPANT: What he said was that Cicats transmission standard would be carried over cable system. It's the same -- it's the same [inaudible] levels. It's all the same. The only thing that's different is that you don't specify the profiles that a video format [inaudible]. MR. FARRELL: I'm sorry. I'm not going to try to repeat that. (Laughter.) MR. FARRELL: Yes? AUDIENCE PARTICIPANT: What was the income that we got from the direct broadcast satellite use of frequencies than from cable uses? MR. FARRELL: I'm sorry. What was -- AUDIENCE PARTICIPANT: Income to the government is gained by direct broadcast use of frequencies than cable use? MR. FARRELL: Well, cable doesn't use spectrum. The question is what income to the government is derived from satellite broadcast -- AUDIENCE PARTICIPANT: Direct broadcast. MR. FARRELL: -- direct broadcast satellite versus cable. Anybody know? AUDIENCE PARTICIPANT: I guess the thing is that broadcasters should be paying for their frequencies. MR. FARRELL: Oh, I see. Well, I think -- let's see. I think Dr. Selwyn's point was that -- DR. SELWYN: I think everybody should pay for frequencies. I mean, I don't know, you know, the history of when these frequencies were assigned. I assume by cable use you're speaking of wireless cable, MDS type? AUDIENCE PARTICIPANT: Cars and C-band phones. DR. SELWYN: Okay. Oh, all right, the satellite. AUDIENCE PARTICIPANT: In cars. DR. SELWYN: You know, if the decision is made that radio spectrum is a commodity, that it has economic value and that its use should be earmarked to the highest bidder, then the only way that you can truly reach an economic, an economically optimum result is to require that all users pay for the use of frequencies. Now, the -- you know, the only exception to that is that if there is some overriding public interest, such as reserving certain frequencies for public safety purposes where in an auction process the public safety agency would not be in a position to pay for frequencies in the same way simply because of the tax structure or the funding structure. But as a general matter we should be letting the marketplace make these decision. And I would certainly have no problem in imposing, if the obligation is imposed upon broadcasters to pay for frequencies, I certainly would agree that cable and satellite broadcasters should pay for theirs as well. MR. FARRELL: Okay, one last question in the back. AUDIENCE PARTICIPANT: How can you filter in economically the cost -- I don't know what [inaudible] the people who don't have faith to watch television, don't subscribe to any services they would pay for. Is it economically correct that if broadcasters pay for the frequencies and then the public don't have any preference for that service by paying, or is there some type of cost [inaudible] that portion of the [inaudible]? MR. FARRELL: Okay, I think the question gets at broadcasters paying for spectrum, and the question is if broadcasters paid for spectrum, would that imply that viewers had to pay to watch television? And if so, what are the welfare consequences of that? DR. SELWYN: Broadcasters do pay for spectrum. They pay for it right now, but they pay for it in the form of private auctions. They pay for it in terms of when -- when properties are sold the buyer is paying for spectrum to the seller. So there is -- there is already a market-driven process to pay for spectrum. It just happens that those payments don't go to the government. They go to the seller of the property, and then they create opportunity costs for those broadcasters who retain their licenses, retain their properties because they have a sense of what the signal is worth in the marketplace. It isn't clear that substituting a payment to the government for a payment in this private market would necessarily alter the costs and impose new costs on broadcasters. It would certainly change some of the values of these properties in terms of the private market because the government would not be taking a larger chunk of them. But I don't -- the price levels are market driven. They are based upon what revenues the broadcaster can collect from advertisers and others in its market. And I don't see that as necessarily being any different if we go to a government frequency auction versus a private frequency auction. AUDIENCE PARTICIPANT: How would you compensate the broadcasters that paid for their utilities now? How would you [inaudible] so that the next time they sold it the money wouldn't go to them, it would go to the government? DR. SELWYN: Well, that's a good question, and obviously you have to have some transition. However, these issues, you know, have come up in other fields with respect to land leases and land grants and other things. I am not -- by the way, I don't want to be recorded as recommending precipitously that, you know, a week from Tuesday that all broadcasters in the United States start paying for their frequencies. That's not my recommendation. I am simply pointing out that the disparity between the current obligation of broadcasters with respect to payment for frequencies and with respect to payment for the additional channels that would be granted to them to support ATV and HDTV, and the fact that the government is salivating over the possibility of $100 billion plus in revenues for the sale of the reclaimed analog frequencies is creating a very serious economic distortion in this whole process, and I believe that the consumer indeed is being left out of it in part because of that reason. MR. FARRELL: Okay. Well, let's see. I think we had better close. I would like to thank all of our distinguished panelists for generously giving us their time and their views, and thank everyone who came, and we appreciate it very much. (Whereupon, at 4:00 p.m., the forum was concluded.) // // // // // // // // // // // // // REPORTER'S CERTIFICATE FCC DOCKET NO.:N/A CASE TITLE:TELEVISION STANDARDS HEARING DATE:November 1, 1996 LOCATION:Washington, D.C. I hereby certify that the proceedings and evidence are contained fully and accurately on the tapes and notes reported by me at the hearing in the above case before the Federal Communications Commission. Date: _11/01/96_ _____________________________ Official Reporter Heritage Reporting Corporation 1220 "L" Street, N.W. Washington, D.C. 20005 Bonnie J. Niemann TRANSCRIBER'S CERTIFICATE I hereby certify that the proceedings and evidence were fully and accurately transcribed from the tapes and notes provided by the above named reporter in the above case before the Federal Communications Commission. Date: _11/12/96_ ______________________________ Official Transcriber Heritage Reporting Corporation Joyce F. Boe PROOFREADER'S CERTIFICATE I hereby certify that the transcript of the proceedings and evidence in the above referenced case that was held before the Federal Communications Commission was proofread on the date specified below. Date: _11/12/96_ ______________________________ Official Proofreader Heritage Reporting Corporation Don R. Jennings