<html><body>American Enterprise Institute Wednesday, May 16, 2007 [Edited transcript from audio tapes] <TABLE cellSpacing=1 cellPadding=1 width="100%" border=0> <TBODY> <TR> <TD> <DIV class=BodyText>3:00 p.m. </DIV></TD> <TD>Registration</TD> <TD> <DIV class=BodyText> </DIV></TD></TR> <TR> <TD> <DIV class=BodyText> </DIV></TD> <TD> <DIV class=BodyText> </DIV></TD> <TD> <DIV class=BodyText> </DIV></TD></TR> <TR> <TD> <DIV class=BodyText>3:15</DIV></TD> <TD> <DIV class=BodyText>Presenter:  </DIV></TD> <TD> <DIV class=BodyText>Richard A. Epstein, University of Chicago Law School</DIV></TD></TR> <TR> <TD> <DIV class=BodyText> </DIV></TD> <TD> <DIV class=BodyText> </DIV></TD> <TD> <DIV class=BodyText> </DIV></TD></TR> <TR> <TD> <DIV class=BodyText> </DIV></TD> <TD> <DIV class=BodyText>Discussant: </DIV></TD> <TD> <DIV class=BodyText>A. Douglas Melamed, WilmerHale</DIV></TD></TR> <TR> <TD> <DIV class=BodyText> </DIV></TD> <TD> <DIV class=BodyText> </DIV></TD> <TD> <DIV class=BodyText> </DIV></TD></TR> <TR> <TD> <DIV class=BodyText> </DIV></TD> <TD> <DIV class=BodyText>Moderator: </DIV></TD> <TD>Michael Greve, AEI</TD></TR> <TR> <TD> <DIV class=BodyText> </DIV></TD> <TD> <DIV class=BodyText> </DIV></TD> <TD> <DIV class=BodyText> </DIV></TD></TR> <TR> <TD> <DIV class=BodyText>5:00  </DIV></TD> <TD> <DIV class=BodyText> <DIV class=BodyText>Adjournment</DIV></DIV></TD> <TD> <DIV class=BodyText> </DIV></TD></TR></TBODY></TABLE> Proceedings: Michael Greve:  Good afternoon and welcome to this event with Richard Epstein and about his book on antitrust consent decrees, which looks like this.  This is a new AEI publication by Richard Epstein, but it is not the newest AEI publication by Richard Epstein.  There is also this one on federal preemption.  Professor Epstein will be back on June 12 in this forum for another book event on Federal Preemption, States Powers, and National Interests.  Today we will talk about antitrust consent decrees.  There are two pieces of good news.  A)  We will end this debate at quarter-of-five and not five.  There was a slight mistake in the announcement.  B)  We have copies of Professor Epstein s book outside.  We have no bookseller.  Therefore, if you are so inclined, you get to take one copy for free.  We ll proceed as follows:  Richard will talk about his book and related issues for about 20 minutes.  Then we re very pleased and grateful to have Doug Melamed with us.  He s one of the of the nation s premier antitrust authorities.  He s the head of WilmerHale s antitrust practice and of special importance for today s event.  He has served as Assistant Attorney General for antitrust matters.  He s seen these issues, especially consent decrees, from two sides of the fence.  Again, we re very grateful that he s agreed to comment on Richard s book.  Richard, please. Richard Epstein:  I ll try to be brief because I always like to hear questions and criticisms.  Doug and I have actually already been at this once before.  We had an interesting meeting, which was sponsored by Sapience in December of last year.  We sort of talked about the whole general Section 2 question on antitrust.  It s always a pleasure to be with him.  As someone who only works on this issue in a spasmodic nature, which is the way I would describe my involvement, always learns from people who have devoted their entire life to mastering a particular field.  The issue of consent decrees and of structural remedies generally seems to be a matter which has attracted somewhat greater attention in recent times.  This past March, the DOJ and the FTC organized a joint hearing, which had some fairly prominent academics and practitioners speaking about the issue.  There was I think, at that particular occasion, a tone which I thought was really quite welcome.  It was a cautionary tone, which essentially said that to the extent that the government is involved in various litigations of this sort, the presumptions should move away from structural decrees, which are designed to reshape the entire organization of particular industry and incline itself to the alternative to try to get much narrower fixes with respect to the subject, often stressing the way in which certain kinds of contractual provisions might be removed from the kinds of agreements which otherwise take place.  That is basically the conclusion that I reached when I did this study.  I should mention to you that the study was commissioned by Microsoft, which has an obvious and direct interest in this question.  Its own consent decree is up for reexamination in November of this year.  The question is whether or not it terminates or whether it continues.  It was with that theme that they asked me to write it.  I m happy to say that they asked me to write it and then left me alone to write it in whatever way I saw fit, which I of course did.  Let me see if I can go through the basic situation with these consent decrees and then kind of relate this topic to larger antitrust issues.  When I start talking and thinking about antitrust, I think that essentially it is as easy as one and two.  Section 1 and Section 2 to me basically embrace very different portions of a legal and moral universe.  Section 1 deals with contracts and combinations and restraints of trade, and I think it is reasonably possible to develop a coherent theory as to why there is the use of government force or to limit the kinds of contacts that are involved.  No reason in this crowd to give a long and detailed explanation of how it is that monopolies can raise price, reduce outputs, and decrease the sum of consumer and producer surplus.  I think those propositions are true, and the sole question you have to ask with respect to the antitrust law is whether or not the cost of enforcement, which also includes the errors from enforcement, is sufficiently high, such that the gains that you get from an improvement in market structure are, in fact, more than offset by the source of intervention.  I think in general you can make, at least in some cases, a credible argument that that is not the case.  This, in fact, if you have a clear theory of liability that then shapes the remedy, if in fact its collusive arrangements that are the thing that is the danger.  What you do is you ban the collusion -- perhaps impose private damage that remedies past collusions and the consequences it creates, and it is a kind of in-and-out remedy that the government is going to have without the pitfalls and the dangers of long-term systematic supervision with respect to how an industry works.  Section 2, it seems to me, is really a very different kettle of fish because its primary focus is on those unilateral strategies taken by a single firm, albeit one with a dominant position, which is said in fact to decrease the overall level of consumer welfare.  Now one possible way to take this particular thing on is to say, Frankly, my dear, I don t give a damn.   What you really ought to do is to just repeal Section 2 on the grounds that if you stop companies from engaging in certain sorts of practices but leave them with their dominant position because there is no Section 1 violation, what they will do is they will simply change the price term of an agreement to offset whatever restrictions they could put into place, so that in the end you are perhaps better off by giving them free carte blanche and hoping that new entry by other firms into various kinds of industries will be the solution.  In dealing with consent decrees, it seems to be to be inappropriate to make that kind of argument for two reasons:  one is that you are trying to figure out how it is that you implement substantive commands, not how to nullify those substantive commands, and I don t think that in the portion of the discourse which goes into the proper remedies what you should want to do is make systematic arguments with respect to appeal; and secondly, it seems to me that in fact the unilateral practice set is not empty with respect to serious concerns.  One way in which to put this is to think of antitrust as simply a subset of the entire remedial approaches that governments might have with respect to firms that occupy dominant positions.  In that, for example, comes other forms of direct regulation stemming from the common carrier tradition, which says that to the extent that you have a single firm that has a dominant position over a certain kind of industry, what you have to do is to regulate the rates in the, perhaps futile but nonetheless principled, effort to see if you can return it to competitive pricing standards at the same time you do not impose sort of very, very high costs.  When you are dealing, therefore, with the kinds of situations that come up in the consent decree area, many times you are dealing, in fact, with antitrust efforts that deal with the same kinds of network industry and monopoly problem, and it therefore follows that you can t take the sort of no remedy ever solution is going to be appropriate.  In some cases, for example, foreclosure may well be possible, but often customers blanket licenses so that they have to take it for everything they use, whether or not they use it on a particular piece of equipment, and there are, at least since the ASCAP case, some fairly strong arguments that foreclosure that takes place from these sorts of practices is something that ought to be discouraged and that the antitrust consent decrees are appropriate.  So essentially when you want to work within the area of antitrust, I think the appropriate way to look at the problem is to say that Section 1 is always going to be easier than Section 2 because the remedy flows quite naturally from the breach, at least if you do things correctly.  In Section 2, the remedial side of this problem is going to be absolutely critical and getting it correct is going to be very, very important.  Now in order to sort of make this case out, what I did was to go back and not write an abstract abstruse book dealing with theory -- that is not my comparative strength -- but I looked at a number of consent decrees to see the extent to which I thought that it met the kinds of standards that I wanted.  Those were just, to basically repeat:  keep it simple, stupid, if you possibly can.  Try and have remedies that are relatively tailored to the particular infraction.  Try to keep them short with respect to the period that are going; try to keep them simple with respect to administration; try to make sure that you have only one set of moving parts, not multiple moving parts; and if you can put all of that stuff together, you are likely to come out better in the day than if you go very ambitious with respect to the ways in which these things operate.  So let me just mention a few things about some of the cases and where I think they went wrong.  One case I like, which is the one that Doug actually disagrees with:  the Microsoft, I thought, consent decree was much better than many of the alternative ones that have happened before.  It is probably worth spending a minute or two explaining what institutional features led to that result in terms of both design and in terms of basic litigation.  And so the first of the cases that I shall mention in effect is one that was quite important in the 1920s and had to do with the break up, not of a particular company, but of the cartel that had existed in the meat packing business - Swift and Armour and so forth.  The entire litigation was pre-staged.  The settlement was announced.  If I expire, it has nothing to do with Swift meats.  Well, I ve got to make something about that.  The entire settlement decree was announced the day that the litigation was brought and the whole thing was obviously pre-negotiated.  The most striking feature about this decree was for these purposes that it just didn t stop at the proposition that we are not going to let you continue with your various kinds of price-fixing abuses, but rather sought to impose all sorts of line-of-entry restrictions on what it was each of these countries can do unilaterally instead of just simply concentrating on what happened collectively.  The upshot of this was that virtually every major company inside the industry was essentially blocked from participating in other industries in which it didn t have any kind of market dominance, so these firms could not diversify.  They couldn t take advantage of adjacent technologies.  And the reason that we know that the decree became highly anti-competitive in this regard is when Swift sought to get relaxations subsequently of some of the restrictions on line-of-entry, the opposition that they received was not only from the Justice Department.  It was also from the National Retailers Association, of course, who were competitive merchants, who feared the fact that Swift and Armour, who had developed real expertise with refrigeration and so forth, would be able to take that technology and move it across the barriers and damage the way in which they behaved.  The legal point, I think, relatively unimportant for these purposes, and that is what the standard is whereby consent decrees ought to be revised:  to me the key feature is that if you have one that goes too broad by way of its original design, it turns out that in an effort to promote competition, you can easily allow the forces that are so familiar to public choice theorists to restrict the kinds of competitions that would take place.  The last bit of litigation that took place over the Swift decree, ironically enough, was the question of what to do with respect to Greyhound bus lines, another failing company, and this was 40-odd years later.  This is not the way in which we one wants to go.  Of more structural interest are the three cases I am going to mention next, which are the United Shoe case, the Bell Telephone case and the big merger and the break up there, and then finally the Microsoft situation.  The reason I mention that is that in each of these cases, you do have fairly strong elements of network industries vertical integration and so forth where they are real strong economies that get [indiscernible] place from cooperative behavior.  It is not like the Swift case where keeping them apart has no efficiency cost.  There are really very great differences that take place.  The first of these cases, I think, is actually a really fascinating saga.  It involves the United Shoe Machinery Company, which was formed by a merger in 1899, and the entire saga finally reached its more-or-less conclusion in 1968, a mere 69 years later, when the Supreme Court essentially ratified the breaking up of a company just at a time when real competitive forces were coming in from overseas.  One of the interesting features about the early United Machinery case was the sophistication that was brought to the question of why this as opposed to other mergers.  Right now, I think in modern economic theory, there is a very powerful awareness of what we sometimes call the double marginalization problem, which in fact is not double -- it is triple; it is quadruple.  This is basically the phenomenon that existed on the Rhine River before the Treaty of Westphalia.  If you ve got an integrated asset in which people can put various tollbooths at various points on it, the non-coordinated behavior of separate owners will in fact very radically reduce the surplus that could be acquired from an integrated business.  There is some of these sort of mathematical proofs of inefficiency which have very small empirical payoff.  Everything about Ramsey pricing -- you can t give illustrations that show more than a 2 or 3 percent difference in output, so forget about it.  But with respect to this one, huge declines in social welfare from non-coordinated activity.  Essentially the way to understand what was going on in the United Shoe Machinery case was that this was a former patent integration of vertical components, so that if they were put under single management, you could make a shoe much more efficiently than you could if you had separate lines.  In addition to this, there may well have been some other chicanery going on, because it is one thing to have a vertical integration.  It is another thing to have horizontal integration at any step, which starts to look as though it was a kind of standard restraint in trade.  But this merger was not attacked so much on those grounds, because those people were sort of spun off, but the whole question went to the kinds of contractual practices that you could put into place with respect to your customers once the sort of integrated product was put out there.  Here there is both a legal and a business tale, which is well worth recounting, which is what happened is the government kept winning lawsuits and this contractual restriction and this exclusive -- if you buy this part, you have got to buy this part; if you take on good then you ve got to take on bad -- all the tie-in stuff they managed to knock out very early on in the litigation.  It had absolutely no effect with respect for the market performance that took place with respect to United Shoe.  It was quite remarkable they were maintaining 90 percent margins 25 years after they lost the 1922 consent decree, for example, on this subject.  Why is that the case?  It is the case I think for the most part that the contractual restrictions are not in addition, but rather simply reinforced what would be a natural predilection on the part of buyers anyhow.  Economists often tend to think about potential restrictions in pricing theories.  Customers aren t thinking about that at all.  They are thinking that if the machine breaks down, they want one guy to call and make sure that he will fix it.  The moment you start allowing alien elements to intrude on this particular line, you might be able to get a better product, but if the darn thing breaks down, you are going to have to fight yourself from one end of the day to the next to find out who is going to be responsible, and everybody is going to try to ditch the responsibility on everybody else, so that in the end, people stick with the single customer or the single supplier.  They may pay something of a premium, but they find that it is worth it.  What is the advice that you give?  Well, the first thing that you want to do is to tell regulators to go easy with respect to these contractual invalidations.  They may not be important, but more importantly the advice that I give to every firm and your man Thorn basically accepted it in public is: unilaterally surrender on every term that is attacked with respect to anti-vertical arrangement.  Somebody says this thing is a restraint of trade:  get it out; repeal it tomorrow.  Remove it.  Don t fight them.  What happens is, the business change is very small, and if you keep the term in there, what you will do is you will always lose on the causation front.  Somebody will look around and say, Well, the only reason why United Shoe was able to maintain its dominant position was they had these clauses.   So they postulate all sorts of economic losses based upon these vertical restraints.  If you remove them and the things don t change, you realize that even though the restriction may be illegal -- let's just assume it is -- that the causal element under antitrust law is going to be highly restrictive.  In the end, what happened is the government kept trying these remedies.  They kept on failing.  Wyzanski, who was a pretty smart guy, didn t want to break the companies up until eventually he was goaded by the Justice Department, and the moment they broke up United Shoe, it meant they had to divide a single factory -- the one they had in Massachusetts -- into multiple components and the company basically cratered thereafter:  a lesson about how it is that you can do very badly in this business.  When it comes to the Bell Systems, I don t know how much to say about it, but here there are two things going on at the same time.  You have direct regulation, and you have antitrust regulation.  Bill Baxter was a wonderful guy in many ways, but he was a man with a big bee in the bonnet, and his idea was that the vertical integration allowed the companies that controlled the local exchange carriers to exert dominant power over everybody else, and he was determined to break that out through a structural solution.  It turns out I think that this was a colossal mistake in terms of the way in which you wanted to deal with the Bell Systems.  The remedy that I would have suggested then had I known enough to even talk about it was one that essentially was much more tailored to the risk.  My view about certain kinds of industries like, in effect, Telephony back at that time is that you are probably going to have single hubs of huge strength, and then the question is which way to counter.  Breaking them up doesn t do any good because what happens is people have to coordinate through two hubs instead of one and they are going to have to form some contractual arrangements to replace the single ownership position that they had.  The way in which you should handle it is the way in which the FCC in its own halting, bumbling, and incompetent way -- it is that way from time to time -- particularly in the 1970s was eventually moving, which was to get rid of all of the prohibitions on interconnection for which the Bell Systems had fought for and to essentially say that when anyone wants to hook up their lines to Bell lines, they should be able to do so, the only technical requirements being that they do not bring the entire system down because of various kinds of incompatibilities based on the operation.  What happens is, if everybody can get into the same network by forcing interconnection agreements, this would produce in my judgment the kind of cohesion that allowed entry on the one hand and preserved integration on the other.  Sure enough, sooner rather than later, the dominant monopoly position of the local exchange carrier would disappear or wither away.  In fact, as late as 1995 when I working on some cases for Verizon, and SBC as it then was, everybody s assumption was that the local exchange monopoly was going to remain dominant for another 10 or 15 years.  By 2002, it was completely clear that that was just wrong so that the erosion takes place much more rapidly, which means that the dramatic intervention should take place on a much more cautious and circumspect kind of basis.  What does one therefore want to do with respect to this business?  Essentially, you want to, even after the statute is introduced, force the interconnections and get rid of any other coercive remedy so that, for example, when you re talking about the question of system design, I would make this observation: if you wanted to find $1 trillion bundle of the 1996 act, it was creating a parity between the purchase of unbundled network elements with the questions of forced interconnections under arbitration.  The interconnection remedy tends to work pretty well.  The forced exchange stuff on the sales things, it turned out they could never figure out what units you could compel and what prices you could compel them.  We could go through the details later on, but essentially this thing cratered time after time after time.  It is because they did not tie the remedy to the questions of system design that they made such costly blunders.  I can t think of anybody, either the new entrants or the Bell incumbents, who is better off in consequence of what has happened.  The lawyers, we did just fine, thank you.  We made a fortune on this one.  Microsoft I have exactly the same reaction to.  The truth about the matter is that it is a less important case than the Bell case because the industry was less large and the sort of total assets that were involved were much smaller.  My basic attitude about it is that in equilibrium, you are going to have a dominant person with respect to service, because all of the people who make applications and all of the people who use them want essentially the same single focal point, much as they did in the United Shoe case.  The only question was whether that was going to be Microsoft in perpetuity or somebody else.  When we talked about this at the DOJ hearings on FTC, I think it was Frank Fisher  who was a government expert on that case, wasn t he, on liability?  His attitude was, well, break them up, have another competition, and then you will end up with another company that is going to be a single dominant supplier.  My attitude about it is, I just don t understand why it is you want to run that experiment when you have no confidence that it is going to come out in a way in which the guy at the end is going to be better than the guy at the beginning.  I think exactly the same kinds of arguments should have been made.  Essentially you get rid of all the foreclosure strategies that you could see taking place by Microsoft.  That was done fairly early on in the process.  I think the great strength of the second tier was not to facilitate entry by creating substitutes for Microsoft, but by trying to be really good with respect to the interconnection obligation.  The irony -- I will end on this note -- is if one wants to figure out what is going on in the European Union with respect to the same questions about the monopoly position that Microsoft has, it is the same battle that went on with slightly terminology that you had in the telephony business.  That is, the Microsoft remedy that they are working for in Europe is essentially give trade secrets out to other people to allow them to interconnect but not to be used for any other purpose.  What essentially the governments there or that this European commission is urging is that you require them to surrender the intellectual property so that people can develop substitute systems using somebody else s trade secrets.  That battle, it seems to me, is going to tell a lot of what is happening.  In the end, when you deal with all of this stuff, it seems to me that the Section 2 cases, that is the structural ones that come up through the antitrust laws, you have to think of them very much as close-kissing cousins to the common carrier cases and to structure the remedies accordingly.  With that, I will be silent at least until Doug is finished at which point I will probably talk again -- maybe. Michael Greve:  Thank you, Richard.  Douglas Melamed:  I am guessing that he will talk after I talk.  Richard referred to the dinner that we had weeks ago.  I said there that I had known him for more than 30 years and have always found him delightful and provocative and all that, but that we had disagreed about almost everything. Well, that s true.  But after reading his book, I think that we agree on a lot actually.  There is a lot in this book that I think is really quite terrific.  I do not profess to have the degree of expertise about many of the cases that Richard studied, but certainly what he said for the most part is very consistent with my understanding and it is a compelling rendition.  It is a very good account of the somewhat, sometimes dismal history of equitable remedies in antitrust cases. And the case studies from the Standard Oil case, Microsoft case, Bell Telephone are, not surprisingly, the same story: the comprehensive or complicated antitrust remedies are often ineffective and are counterproductive -- not in the sense I take it of exacerbating the problems they have set out to cure but in the sense of having the cure be worse than the disease or at least sufficiently awful that you wish you never had the disease.  Richard s principal conclusions, I think, less is more, and new entry beats comprehensive regulation, are surely correct as a general matter.  The difficulty of antitrust decrees, Richard says in the book, principally are those of duration, ambition, and complexity.  I have long believed that he is surely right, although I didn t know that he thought that.  I long believe that duration is the real problem with consent decrees.  Some of them are nutty; the litigation 15 years ago over the Kodak decree which was then 75 years old made no sense.  If decrees after a few years are still needed -- after a few years either decrees have already succeeded or they are never going to and presumptively at least should not last long.  Ambition and complexity is a more difficult question, it seems to me, because you always have to: ask compared to what?  It is not enough to say, Gee, this was a mess -- Telecom decree or the Microsoft decree or whatever it may be.  If there is maybe a quarrel I had with the book, I think maybe it didn t do enough asking of compared to what.  I want to pick up on something that Richard said just now about that.  He used, what I believe is, a false dichotomy to prove that structural remedy and a narrow fix, and those were his terms.  I think it is a false dichotomy because the real dichotomy, in my view, is between remedies that attempt to restore markets to the competition that would have existed but for the wrong or otherwise to repair the damage on the one hand versus remedies that are sometimes called sin-no-more remedies that simply say don t do what you did that was found to be wrong.  The reason I think that is a better statement of the dichotomy is that within the category of restorative remedies, it is not clear to me that a structural remedy is always more intrusive or more burdensome, more ambitious or more complex than a conduct remedy.  Certainly in the merger cases we don t think that.  We think structural remedies are preferable because they can restore or preserve market structures that will enable competition to exist without ongoing regulatory supervision.  To me, the real question is not between structural and narrow remedies; it is over a continuum that asks how regulatory the remedy is.  Put that way, I would certainly agree with what Richard said that there has been a lot of bad stuff -- a lot of bad regulation done in the name of antitrust remedies.  A note from my personal experience both in the government and in private practice about the source of the problem.  I am going to get in a couple of minutes to the issue of the sort of underlying substantive antitrust laws, which I don t think are the source of the problem.  I know Richard s secret -- I'm going to say in the minute -- and that s the secret is that is the source of the problem.  I think an important source of the problem lies elsewhere.  When you negotiate a consent decree with the government, leaving aside broad questions, are we going to bust up Microsoft to do something else?  You are dealing for the most part with lawyers and economists who in my experience are bright and very dedicated and honorable people who are trying to do the right thing, and who, if they are antitrusters, particularly if the antitrust division -- there is an institutional difference I will get to in a second with the FTC -- believe that regulation is a bad thing.  They believe in the antitrust laws, and one of the core reasons for that is that they think it is a better way to deal with the problems of economic welfare, lost efficiency that is regulation.  They don t set out to be regulators but they often wind up being regulators.  I believe part of that comes from two understandable fears that they have:  one, they don t know as much about the problem as the defendant does, so they have to tread very cautiously for fear they will be duped, to put it bluntly; secondly, a fear that the world is going to change.  They all recognize markets are dynamic and we don t want to be the heavy thumb on the scale of technology and all, but if markets change, maybe the remedy is going to prove to be ineffective and there work was done naught or they will look foolish or something like that.  Those two forces, I think, interlocked [sounds like] inch by inch cause consent decrees sometimes to be too complex, too loaded up with loophole closing or anticipating provisions, and too cumbersome.  The FTC s problem, I believe, is an additional one.  There are five heads instead of just one, so you wind up getting sometimes a little bit of the old perennial sort of Christmas tree: put this in for Commissioner So-and-so and put this in for Commissioner So-and-so.  In other words, I think the problems are rooted in bureaucracy, not in antitrust.  Compared to what?  I want a digression on that in terms of are decrees too complex or too ambitious?  The Baxter Telco decree -- Richard said, well, look, instead of the incredibly burdensome divestiture and all of the rules trying to enforce that and prevent the boundaries and the millions of times people went down to see Judge Greene, why not just require interconnection?  That was the problem in the case -- that there was no effective to enforce interconnection that wouldn t demonstrably -- I think the Justice Department thought this advantaged rivals.  That is the problem, the alleged problem -- I have no view of the facts in the Microsoft disputes in Europe.  I think saying let s just force interconnection is like saying let s just bust up Microsoft.  It ignores all the hard work and all the heavy lifting.  I think Richard has, in a powerful way, demonstrated the harm that can come from a heavy-handed, overly-complex and ambitious regulatory decree.  I am not as enthusiastic, however, about some of Richard s somewhat impassioned comments about substantive antitrust rules.  I won t get into all of that here.  A couple just leapt out at me.  It is not true that the broad objectives of antitrust law are not attainable in network industries.   That s true only if define antitrust pre-enlightenment as all about deadweight loss.  If you define antitrust as all about a competitive process likely to maximize welfare in a dynamic sense over time, network industries are industries about which antitrust has a lot of constructive things to say.  Nor is it true that dynamic competition precludes a network monopoly like Microsoft from having and exercising destructive market power.  It is no doubt true that Microsoft was concerned about the prospect of future competition.  All monopolies face downward sloping demand curves.  They face some constraints.  In fact, that was the premise of the government s case.  It was so concerned about competition that it set out to nip rivals in the bud.  This takes me to my more basic concern.  I was going to talk about Microsoft, but that s [inaudible].  This is a wonderful book about remedies.  It is an antidote to the triumph of hope over experience when agencies set out to do these complicated remedies.  But the book seems implicitly to be in the middle of a conversation in which it accepts the premise that antitrust is only as good as, or is fundamentally about, equitable remedies that result from antitrust cases.  I want to read four sentences from page two that strike me as ones he probably thought more about than the ones on page 150 because he was fresh, all from the same paragraph.  I m going to read them in a different sequence from the sequence in which they re in the book.  I don t think I m doing an injustice to Richard, but he can tell me if I am.  I m doing it because I think it will make the point that I m trying to make clearer just by juxtaposing them a little differently.  Here are the four sentences:  Antitrust exuberance has frequently led to an expansive view of the remedial role of court cases, which all too often has proved counterproductive.  Government lawyers proposed, and judges eagerly embraced, concerted efforts to restructure entire industries to advance the needs of consumers and the public at large.  All too often however, antitrust decrees have suffered from an excess of ambition by which too much was too attempted, and too little was achieved.   My objection to that way of doing business is not to its implicit standard of social welfare, but rather to the na�ve assumption that aggressive antitrust enforcement is the invariant to that laudable end.  Note the equation of aggressive antitrust enforcement with heavy-handed equitable remedies.  In my view, this is a common but fundamental error.  Antitrust is fundamentally, and most importantly, not about the remedies and cases and individual defendants found to have violated the antitrust laws or take the consent decree to avoid litigation of that question.  Rather, it is about the behavioral rules that antitrust articulates and the signals it sends that it sends to the business community; signals that inform the millions of decisions that millions of businesses make everyday that never get to court.  It is about the rules of lawful versus unlawful conduct, not about the sanctions or consequences for being found to have engaged in unlawful conduct.  Remedy is the second-order consideration.  Richard said early on in the book, and I think he s clearly right when he said this, The strongest defense for an open economy lies in free entry and exit of firms.   I believe that antitrust enforcement, backed perhaps by financial penalties or damage remedies where equitable remedies might be regarded as a cure worse than the disease, can be a vital part of ensuring that entry barriers are not artificially raised by incumbent monopolies.  Interestingly, the Microsoft case was a case about raising entry barriers.  In Richard s rendering of the remedies, he rightly says that the U.S. remedy hewed pretty closely to the wrongful conduct and prevented its recurrence.  Of course, a remedy that focuses just on wrongful conduct is inherently a backward looking remedy.  It is a little bit like saying the remedy for 9/11 should be to ban al Qaeda from flying the plane, or maybe to ban them from flying planes into the World Trade Center.  It might not be a complete answer.  One might ask, if a company like Microsoft engages in conduct to raise entry barriers, are there remedies that will reduce entry barriers to compensate for that harm; and if not, is there a deterrent we can build into antitrust law that ex ante will prevent that kind of conduct in the future without the specter of counterproductive equitable remedies?  Richard says in the book that monetary penalties founder and cannot effectively be used where the magnitude of the effects cannot accurately be assessed.  Surely, a situation after a case like Microsoft that involved the kind crib death theory -- that is to say they were nipping in the bud potential entrance -- how do you measure those damages?  The implicit premise seems to be that fines should be equal to what Richard calls the social loss from the anticompetitive conduct.  I haven t studied this question a great length, but it s not self evident to me that that s correct.  Why, using jargon loosely, shouldn t antitrust laws be enforced by property rules where the liability is?  Here s my point:  Antitrust is fundamentally about law enforcement, not about regulation.  Richard s book powerfully demonstrates that antitrust can get into trouble when it becomes regulation rather than law enforcement in the form of complex ambitious, sometimes excessively long-lasting, equitable decrees.  As a practical matter however, a complete laissez faire attitude is a nonstarter.  Politically, it wouldn t be a desirable one either.  It would be a shame therefore, if Richard s compelling explanation of the problems of antitrust as regulation through equitable remedies spawned a debate about what kind of regulation is best, rather than a debate how to do the law enforcement part of antitrust without regulation. Richard Epstein:  Thank you, Doug.  I thought those were terrific remarks.  I actually agreed with some of them.  Part of it really is the question about the focus and emphasis with respect to this volume.  On the first point, there is no doubt that the primary objective of any effective legal system, whether you re talking about the criminal law or torte law or antitrust law, is in the unobserved violations that do not take place in virtue of the fact that the system has been sufficiently clear that people can avoid them.  I think anybody who wants to sort of assume that you measure the success of a legal system by definition of its failures is somebody who is, every time they see contracts, will say, all contracts are breached and therefore we are against markets.  It s not the way in which one wants to play.  Starting from that theme, I think it only reinforces the difference that we have between Section 1 and Section 2 of the antitrust law.  In Section 1, you know what it is that you are not supposed to do.  I do not have to spend an enormous amount of time to say trips to remote chalets in order to have ski trips at which you fix prices are the sorts of things you can t do; ski you may, but fixed prices you may not.  People will understand all that, but when you start talking about Section 2 and you start talking about network industries, it is a really different game.  You cannot tell a company that the only way you can avoid a charge of monopoly power is not to sell your goods at all.  It turns out, I think it s fair to say that virtually every substantive provision that you put in which has to do with service or restrictions is, under current law, potentially a risk for evaluation.  The difficulty that you have is that there is now clear safe harbor in any Section 2 case, in my judgment, which allows firms to say with confidence this we can put in and that we cannot.  That s a real sort of deterrent because now it s the opposite effect.  People will essentially avoid efficient relationships for fear of massive retaliation.  The second point relates to the first one.  I will stand by the proposition that there is no way in network industry that you can get a competitive solution.  That means that the whole effort of this business, whether through antitrust or direct regulation, is try to figure out which of the imperfections that are going to take place in the market has the smallest reduction in the overall social surplus that is going to be created.  That s really a question of institutional design.  One of the reasons why I think that the antitrust decree and AT&T case had such a disaster associated with it is that it was as though the antitrust displaced a regulatory arrangement.  In fact, it necessarily required the regulators to do things which they hadn t done before.  By the same token, it meant that you had two people at the wheel.  You never did get what I think is the effective thing that you want there, which is a coherent effort to figure out how it is that you facilitate interchange agreements.  I do think that Doug is wrong on this.  If you go into the post-1996 stuff, correct me if I m wrong, I don t believe there s been any serious interconnection dispute which comes within a thousand-fold of the difficulties that you had with respect to the pricing of unbundled network elements.  Is that correct or am I wrong about that? [Inaudible male voice] Richard Epstein:  So I m wrong.  It s a hundred-fold or a ten-fold difference.  It s at least an order of magnitude.  Let me put it to you this way.  In all the time that I worked on these areas, the only time we had to work on network interconnection was on the bill-and-keep situation, when there were radical price differentiations between the cell phones on the one hand and the land lines on the other.  Essentially what you did was have a takings type of arrangement.  The truth about the matter was the fixes were relatively easy once you understood the problem.  The issue disappeared within five years anyhow, when people started to leave their cell phones on.  The second thing I want to say is about antitrust regulators.  They re worried about being duped.  I agree that that s a problem.  They re worried about changed circumstances.  That I don t think is as much of a problem.  I m very hard pressed to think of any series of technical innovations in any area which have actually strengthened the hand of a prominent player.  Essentially what happens is that all technical changes either expand the scope of geographical markets, you get global instead of local competition with more entrants, or that products are now reduced to a common technological base so that you could substitute electricity for wire, for example, in the way in which you distribute.  I think that it would be most unwise for any antitrust regulator to come forward with the general proposition that they ve got to be extra careful for fear that new technological innovation will undermine the effectiveness of their decree.  I think exactly the opposite situation is going to work.  The general theoretical point was made, without reference to these industries, when Harold Demsetz wrote wide and regulated public utilities in 1968.  I think for the most part that thing has worked.  The third point I sort of want to mention is the question about strategies to nip rivals in the bud.  There is no question that Microsoft certainly tried to do that in some way.  The early decrees got rid of the per price and license example, which was a big effort to try to run these exclusions.  That s a clause that probably did matter.  It mattered there and it probably mattered in ASCAP.  Then the question is diminishing marginal returns from additional forms of regulation.  You can say all sorts of things about how it is that you don t allow Netscape to get on the top of the desktop.  Maybe it s right and maybe it s wrong, but when 250 million copies of the could be downloaded at zero cost it seems to be that the ability to nip in the bud by stopping one mode of distribution is in fact undone by technology, which allows you to essentially download this stuff online.  I m not going to argue that there s no disadvantage to Netscape, which has only one avenue relative to Microsoft with two.  But I will put out there that you have a little bit here of the same patent problem.  One of the things that you want to do is not only look at the static model and figure out after the whole technology is developed what shares are going to go to Microsoft and to Netscape or anybody else, but the more that Microsoft can internalize gains by putting this stuff early on, the earlier the innovations will take place.  There is the same battle that you have in the patent law, in which, if you really give strong rights to exclude, you have ex post in efficiencies and ex ante in advantages.  Once that trade off was put in there, it becomes much more difficult to think that this is going to be the overall and the end situation.  I agree with much of what Doug says and certainly I do think that the remedial side is the second-order question in antitrust.  I thought I had specified that.  If the first-order question is worth trillions, then the second-order question is worth hundreds of billions of dollars.  Even in this lofty crowd, it s worth a little bit of attention to be paid for it in order to get it about right.  I think we should open it up unless Doug has anything. Michael Greve:  No?  We ll open it up.  There is a microphone.  Please, when you ask a question, let us know who you are and then proceed.  Thank you. Greg Sidak:  Working now?  I m Greg Sidak.  I m at Georgetown University Law Center.  The other area where antitrust consent decrees are used is in merger approval, of course.  In particular, I would like to know your view, Richard, on the use of conditions in merger approvals as reflected in consent decrees.  In horizontal cases where the merging properties would be expected to achieve market power but offer to forebear from exercising market power by consenting to price regulation, is that good or bad?  How does it play into your analysis, number one?  The second part is: Is the imposition of price regulation through a consent decree constitutional? Richard Epstein:  On the first question, I didn t talk about mergers and I think it s probably worth mentioning a word about them.  My view about it is, the simplest remedy in a merger is, first of all you can undo it or block it, because remember this is going ex ante.  It s not going through litigation after the fact.  The unraveling problem if you have an ex ante procedure is going to be much less than it would be if you had the merger and then tried to disagree with it.  There is a cost if you slow these things down when in fact they are appropriate, but the actual remedies that come forward in the bank cases, for example, is Bank A wants to join with Bank B; we look at one local market and we realize they are going to have to have an 80 percent concentration there.  All those banks in Tennessee, Kentucky, and Alabama have to be spun off into an independent company.  That is a remarkably clean remedy in which it meets all of the standards that I was talking about.  It s once and for all, and it s a little bit like, and it seems to me to be much closer to the sort of injunctive relief that you talk about in the antitrust price fixing cases, than it is about the complicated efforts to force all sorts of interactions through a combination of direct price regulations and antitrust stuff that you get in the AT&T cases.  I think that s the way one goes.  The second thing is, I think, very different.  I am not at all happy with the prospect of allowing a merger on condition that there be generalized rate regulation through the antitrust laws.  My sense about that is it would be generating the same kinds of real confusions that you have in cases like ASCAP, where for 50 years they are trying to figure out how to regulate two competitive combines and they never subject them to the same sets of regulations.  One danger you will have is if you get two or three of these mergers in the same industry and you get different feels with respect to each of them going through, you re going to really upset some kind of competitive parity that you would otherwise want, so my sense about this is I would for the most part not like to see it done, and if it is, duration would be a very strong limitation that I would put on it.  On the other question, the Congress, on the one hand you are asking whether this stuff is absolutely indispensable for the regulation to go through, and on the other hand is it unconstitutional if you do it, right?  I mean, it s a kind of opposition.  I think in effect that the standards that they tried to develop in the late 19th Century to -- as Bob Bork put it when he wrote in Jersey Central -- avoid the twin dangers of monopolies expropriation on the one hand and confiscation on the other are the appropriate measures by which you do this.  Therefore, some form of price regulation can certainly be appropriate under some cases, but it is going to have to meet certain kinds of rates of return sorts of requirements.  Some of them.  Most of them don t.  I, for example, when I thought about this after the 1994 decree, 1996 statute, it occurred to me that we may have been better off leaving AT&T in place, having the interconnection remedy, and then using the price caps whereby every year you lower the effective rates 3 percent and then raise them 3 percent for inflation, which meant that they couldn t move them one way or another.  My sense is that competition would have eventually driven the prices below the caps.  If you change the form of regulation something not like price caps, which is a cost base situation, I think you get very different kinds of results, so I don t think there is a uniform answer to that question on the constitutional stuff, except for the following.  I do not treat the world of rate regulation and the world of unconstitutional takings as though they have no overlap, but under the current environment the presumption is going to be against it, not for it.  Doug, do you have something to say? Douglas Melamed:  No, I agree. Greg Sidak:  My question, actually, on the constitutionality was heading in a different direction.  I was thinking of a separation of powers problem where you have the executive branch and judiciary engaged in rate regulations, which the Supreme Court has said is a legislative function. Richard Epstein:  That was the argument that was raised, of course, by Justice Rehnquist in his descent when they denied cert in the Bell cases, and it didn t gain much traction then.  I actually have some sympathy for that particular line of argumentation.  I actually do believe in separation of powers.  In the current Supreme Court you have no traction with respect to that argument.  Their basic view is that the administrative state is the fourth branch of government and to the extent that it is allowed to engage in hybrid function, it has taken out all the steam from people who have talked about the purity with respect to the line.  In fact I think one of the real difficulties with modern separation of powers doctrine, speaking more generally, is that the Supreme Court, except perhaps for the special prosecutor cases, maybe, now doesn t believe in it internally as a serious kind of constitutional constraint.  And generally speaking, whenever you don t believe that a doctrine has a function it becomes a rational basis test, at which point you can just watch it fly away with the evening sun.  I don t think it is going to have any real legs today because the moment you introduce that seriously, the NORA, the SEC, the EPA, you name it, they are all going to be subject to some degree of hybrid communication because remember the NORA model is first we pass the rules and then we prosecute the cases.  That is clearly unconstitutional in my view, only nobody listened to me when I said that seven years before I was born in 1935. Doug Melamed:  I had a brief thought about the current Supreme Court, about which in general I m not an expert, but to a lot of us in the antitrust world who do begin with the notion that regulation is a terrible thing and that we should try to avoid it and better to do it by antitrust-type regulations and so on, Justice Scalia s opinion for an unanimous court in the Trinko case was quite striking because of its apparent subtext that regulation is really a good thing and antitrust ought to maybe forebear where regulation seems to be a viable alternative.  It was surprising to me and to others as an antitrust to becoming from Scalia, but I think a lot of people read it that way.  Alden doesn t seem to read it that way. Richard Epstein:  I was not intimately involved in the Trinko case, but I certainly had flirtations with Verizon when this thing came up.  I thought it basically was trying to reinforce the lesson that we had from the original Bell decree.  I regard antitrust regulation as regulation.  I regard rate regulation as regulation.  The only thing worse than one system of regulation is two systems of regulation, which tend to work at war with one another, and the reason why the Trinko case was so attractive is you had all of these elaborate arrangements with respect to the rates that could in fact be charged.  Now somebody is trying to upend them on the strength of a weak sort of substantive theory on the merits.  The case was actually interesting for two reasons.  They said we are only taking on this structural issue; we re not taking on the privity issue and then three justices wrote on the privity issue anyhow, notwithstanding the fact that they refused to hear it.  To me the more important case in the long run, I thought, is going to be what happens when they decide this recent thing that, oh my God, the name is slipping my mind.  What?  Twombly?  The Twombly case -- the problem is, this is why I don t like the antitrust stuff.  You get two things going on.  One is, you have a duty to conspire.  What I mean then is, within a network industry you must talk to your competitors in order to be able to integrate so that the usual tests -- don t meet at the ski club -- can t work to separate these guys out.  You just have to have them speak.  Secondly, the usual argument is you have got to have the private antitrust discovery stuff because otherwise you will never find out whether these guys were doing something, but in my view in a regulated industry, where you have already had a Congressional investigation, there has been a lot of discovery that is available, and they don t have zip in order to bring that particular case.  Essentially, allowing discovery before summary judgment can be ordered is basically $10 million a case for God-knows how many cases somebody wants to bring.  I think that the integration in that particular case should be, so long as you believe that the FDA is anything other than comatose, then in effect what you want to do is prevent a second bite of that apple on principles that have to do as much with comity and race judicae, particular since the underlying substantive theory in Twombly is woefully weak, precisely because there are so many reasons why people would stay out independent of the fact that they are colluding to divide market territories. Doug Melamed:  No, not really.  Were I writing those opinions, I would write opinions for the defendants in both cases.  Certainly Trinko in my view reached the right result, but I just wouldn t do it with reference to we already have a regular on the beat.  I would attempt to give him the back of the hand as much as I could do. Richard Epstein:  I don t see how you can do that.  The FCC, I don t like it very much, but it s here to stay and it s a constant presence.  My own view about it is, Michael and I have written -- do you have the other book up there?  The basic theme of this book is remarkably consistent with the basic theme of that book, which is that we don t like concurrent powers essentially.  We are clean guys.  Do it one way or do it the other way, and the reason you don t need to get regulators into the Microsoft stuff is not that it isn t the common carrier case, but there is no crisis involved, so courts can actually handle that.  They can t handle price administration.  The ASCAP case is a quite good evidence of that.  I just want to have one poison.  Let the government pick it.  I think with respect to the telephony business, it has got to be a system of direct regulations on interconnections.  It would be the stuff.  I disagree with Doug.  I think in effect if you cared enough to get the interconnection stuff right, that would be soluble.  They could have done it.  What happened is, Green didn t believe in any of that stuff because his leading guide on the antitrust was Karl Marx.  He cites him.  No, no, no. If you look at what Green has said -- not Baxter -- his view was that the indirect, adverse consequences of large concentrations of power have genuinely bad, dysfunctional effects upon the overall political system and those things that we have to cap with the antitrust law, and he cites the people who cited the Karl Marx, whom he also cites.  It was very much the 1960s antitrust law Ralph Nader view.  This is a cure-all for every ailment in political society because big is bad.  My view about it is when these guys get big, they become punching bags for every small, two-bit group that wants to hold them up whenever they want to get a new license.  If it s a fight between Americans for concerned ecological care with respect to the condition of the cactus plant in Texas as against Verizon, I m going to vote with the cactus guy when it comes to the question of who got the power.  It seems to me that what Fred McChesney wrote years ago about how it is years that large guys could be held captive  by small people has been proved in all of the telecommunications stuff, whether it is with broadcast licenses or any other kind of licenses, conditional renewals.  These guys are masters at playing the extraction games, and you guys are masters at capitulation. Michael Greve:  Yes? Frank Fletcher:  Frank Fletcher [phonetic].  A couple of questions to the panel.  Would any of you like to talk about what you think about the Office Depot/Staples merger that apparently was blocked?  Then I was thinking about Loews Cinemas.  Didn t they used to be in Hollywood and make movies and they were driven out?  The final point would be, back 25 or 30 years ago, people who made computers -- I m thinking of Wang, Prime Computer -- they made both the hardware and the software proprietary.  Computers couldn t talk to each other.  Now as technology has evolved, you don t see the computer makers also manufacturing the software, so maybe that goes to Professor Epstein s point that somehow economically it doesn t make any sense, even the chip makers, too. Douglas Melamed:  A little bit on the first couple.  Sound antitrust is all in the details.  It s not about Marx, and it s not about greed is good.  It s all in the details.  I wasn t involved in that case.  My understanding is that the econometric details were compelling for the proposition that even though paper clips can be bought at a gazillion outlets and so forth, in communities in which there were -- I forget it was -- two or fewer or three or fewer or whatever it was of the office superstores, prices were higher.  Douglas Melamed:  Okay, so the move from three to two was material to prices.  Given those kinds of data, the intuition that I can buy a paper clip anywhere is trumped by empiricism and that has become a competitive merger so the right result was reached I believe.  Just one thing briefly on the Paramount decree, the Loews, and all of that stuff:  it reminds me that I actually wrote an article many years ago that is in some obscure journal that probably got as wide a readership as it deserved that began with something about in ancient Greece they used to shoot the messenger who brought the bad news on the theory that the messenger was too powerful.  Here we had a different approach to powerful messengers.  We just strangled them by bad antitrust.  It basically went through starting with radio and then motion pictures and then television and so on, a series of how we have been particularly imprudent in unsound, vertical theories directed at the media.  I think for a while we toyed with that in cable and then backed off.  Certainly the television consent decrees and the [indiscernible] rules are at peace with that.  I think the Paramount decrees fall into that category as well. Richard Epstein:  I think there should be a strong presumption against attacks on vertical kinds in of arrangements because the double marginalization theories are really powerful and so they are strong efficiencies.  You have to be able to show some really powerful effects that move in the opposite direction.  I don t think there is any particular great advantage in ending the studio system, which is what they have done by virtue of those decrees, so I would think they are a mistake.  On the other point I think that basically what Doug said is, again, we re going to state the same thing.  Empirically, the basic intuition is four to three we can live with it; three to two we cannot.  If you have something like Staples and Office Depot coming together, the problem that you have is when you get multiple different markets -- in some cases it would be nine to eight and in other cases it will be three to two and in some cases it will be two to one, so that my sense is to sort of do the pockmark approach, which is to figure out where you re going to divest these things, is not going to work in that case because divesting individual units from one of those companies is not going to give them any of the advantages of advertising amd mass purchasing, which made them distinctive.  Essentially, blocking the merger is probably the most efficient situation, which I think is very different from what it is in the banking cases where sometimes you can take a fairly large and viable situation and spin off 10 or 12 units or market sections into another company, which itself is going to have a billion dollars capitalization and real market presence within its own particular area.  I think that from what I have seen I would do exactly the same thing, but the olden days, remember, when people didn t know how to do the Herfindahls, and they would start getting worried about eight to seven and stuff like that.  If I m not mistaken, I don t think there are many people in the world who differ from the proposition that three into two is bad.  There would be a few who would argue about four into three.  I don t think there s anybody now who is much concerned about five into four, is there?  [Indiscernible cross-talking] Female Voice:  The FTC just brought a case against two petroleum companies in Albuquerque, New Mexico, and depending on how you slice that market, it is probably a six to five. Richard Epstein:  That doesn t strike me as being a winner.  Not only could that -- that s a market where I assume relatively easy entry could take place.  It s not like a national chain.  It s much harder to gin up an Office Depot than it is to find a way to get BP into a market in which it hasn t moved at all.  I m not wildly sympathetic with that.  It s only one market.  Maybe the FTC will grow up on this issue; maybe not. Michael Greve:  Can I ask you a political science-y question?  It s about the consent in consent decree.  If you get away from antitrust, there are a lot of consent orders, consensual arrangements, where businesses effectively invite lawsuits from governments or turn current pending lawsuits around for the purpose of cartelization and for their own benefit.  It s just lack of demand for regulation.  Just because Guido Calabresi cranked that up God-knows how many years ago doesn t mean it s necessarily false. Examples:  Tobacco, the research settlement agreement, and on and on.  You can theoretically see how that might happen in the antitrust content, too.  Start with one of these structural remedies.   The one thing that the FTC or whoever administers these decrees over time probably won t do is actually put you out of business or weaken your competitive position to the point where that becomes the net effect because it s their sandbox now and they want to keep playing in that sandbox.  My question is simply, have either of you seen evidence to that effect or is antitrust regulation even under the auspices of the structural long-lasting decrees wholly unlike direct regulation under the auspices of the FTC or, of course, somebody else? Richard Epstein:  I think it harder to cartelize through the antitrust laws than it is through direct forms of regulation, but it is not impossible.  If you can impose, as they did with the Swift case, a line of restrictions on entries in other lines that are unrelated to the area of monopolization,  you, in effect, have done exactly that.  My own sense about it is that is usually sufficiently obvious and I don t think, given the current level of sophistication in the business, that that s likely to succeed.  I think the whole antitrust culture is sufficiently sophisticated that I put that as a relatively small situation.  Other decrees are not.  One of my favorites I will just mention is years ago the government, I think it was the FTC, put a very tough decree on the cigarette industry saying they could no longer use celebrity advertisement.  The industry was thrilled.  Essentially what happened is no one needed to use them anymore because nobody else could use them, so essentially what it was, it was a way of essentially reducing their costs under the guise of having an anti-smoking campaign.  That s where you re going to start to see it happen. Douglas Melamed:  I don t know about Swift.  I m going to limit myself to decrees since the electric typewriter was invented or something within my memory.  I ve never seen anything like that and I would be very skeptical for the broader reasons that Richard raised, but also for this reason:  I mentioned that one of the problems I thought was the caution of the government lawyer who is trying to cross all the Ts and close all the loopholes.  One of the things that government lawyers increasingly do now to make sure they re not missing something is that they market test a proposed remedy.  They go to other people in the industry, customers, suppliers, and competitors, and they ask if this makes sense.  That strikes me as a pretty good check on that kind of outcome. Cecile Lindell:  I kind of wanted to follow up on that.  I m Cecile Lindell [phonetic].  I m a reporter with The Daily Deal.  I cover antitrust regulation.  In 1999, the FTC came out with a divestiture study that looked at the efficacy of consent decrees in merger cases.  One thing that came out of that study was that a lot of the companies that are sophisticated have realized that they could come up with a proposal that would get blessed by the regulators at the FTC and DOJ, and they could divest some property that would then wither, and then decrease the amount of competition going forward.  You mentioned this market testing as being one way to kind of avoid that happening.  Are there other things going on?  And it doesn t it sort of say that companies have really gotten sophisticated about trying to manipulate the antitrust system and consent decrees to their own advantage? Richard Epstein:  The answer to the last question is yes, they ve become extremely sophisticated.  I thought I tried to give an answer to the fact of why it is that in the Staples/Office Depot situation you would not allow partial divestitures because the stand-alone units could not survive, given the fact that integrated purchasing, advertisements, and marketing were essential the operations of it.  I don t think there s any answer to that particular question except equal and opposite sophistication on the other side of the arrangement.  In fact, there is some real question now, although I m not an expert on it.  They ve revised the Tunney Act.  It s clear that the level of scrutiny that s supposed to take place under it is somewhat than it s been before.  What?  [Inaudible voice] Richard Epstein:  It s not?  The Bell South merger was tied up inside the Justice Department for a very long time on the question of figuring what things went on.  The sense that I got from a fair of number of people who actually knew something about this was that they thought the additional scrutiny had many anticompetitive features about it, by virtue of the fact that it allowed competitors to take advantage of the system and to slow things up.  It also created an immense amount of uncertainties.  I ve not seen this recent opinion, but if they in fact decide to ratchet back the significance of the [indiscernible], that s the shift I think we re talking about.  That would probably be a good thing.  I think it s okay to consult competitors, but to start to giving them formal rights of access to hearings and a potential veto position is a great step backwards.  I would rather rely on government expertise to gather this information than to rely on any format which essentially gave other people strong standing rights and blockades on the theory that most of these things will probably have to go through, and that the antitrust divisions are sufficiently sophisticated now that regardless of political administrations it s going to work well.  I m curious.  Do you agree with that or not? [Indiscernible] MichaelGreve:  Yes. Alden Abbott:  Doug, you raised briefly the issue about divestitures in merger cases.  You asked, Why do you have structural divestitures there and why not in appropriate monopolization cases?   But doesn t that get back to the issue of the difference between Section 1 and Section 2, which Professor Epstein raised?  We find the famous language in Copperwell  that there is good reason to be much more risk-averse in pursuing a Section 2 case than a Section 1 case.  In the Section 2 case, you have integrated entities and have developed a variety of markets integrated vertically, and put together a system is furthering efficiencies.  That set of considerations is absent in the typical merger, which is examined ex ante; you have two independent entities.  Even if ex post you get a merger before there s been too much scrambling of the egg.  My underlining question is, given the differences between Section 1 and Section 2, and I know people diverge on that, what does that have to say about consideration of appropriate remedies and decrees? Douglas Melamed:  Certainly as a general matter, Alden, what you said is correct.  If you have a merger where the divestiture is, basically, saying don t merge, it s less likely to be disruptive than on that takes a company that has been operating as a single corporation for some period of time and says to undo what you ve been doing.  One could combine that generally correct statement with a lack of confidence in the ability of government enforcement agencies to get more fine-tuning.  You could have a general rule that says never to do a structural remedy with a company that has already been together.  I think that would be going too far.  It would be taking a generalization and turning it into an absolute that s not warranted.  For one thing, in merger cases, it is not uncommon for companies to do kind of mix-and-match mergers where the acquiring company will say, okay, I will divest Plant A from the acquired company and Plant B that I already own.  You do that sometimes when in fact the surgery isn t that onerous because maybe it s Jeep and Chrysler operating -- I don t know -- sufficiently separately that one could be spun off to private equity or some kind of LBO and if so, why not consider a candidate for divestiture if that fits in appropriate antitrust theory.  I just think it s overly broad to say structural relief of preexisting companies is so disruptive that the cure is going to be worse than the disease and worse than the alternative.  I think it s generally going to be true, but I wouldn t make it an absolute rule. Richard Epstein:  I think it should be an absolute rule, and the question is the strength or the presumption.  I think that the two of us essentially think that the line between 1 and 2 is dark grey and light grey or near white, and Doug is sort of pushing them a little bit closer together.  I m hard-pressed to think of a situation in a kind of integrated case in which the breaking up has made sense.  I can think of many cases in which conditional mergers might make sense, and in fact once you introduce private equity, that s an extremely important consideration in my judgment, because those guys are not going to pay top dollar in order to get lump fragments that are going to disintegrate and fall apart.  In many ways, I think the answer to your question as a reporter, one of the ways in which you stop the spin-off racket from taking place is you don t allow the spin off to become a standalone company.  You require that they sell it to somebody else who can then treat it as part of their integrated operations or to some sophisticated private equity firm and then have a market check on what s going on.  I don t think that would be a perfect answer, but I certainly think it s one that would transform the nature of the inquiry, and I think in that sense the arrival of big, private equity companies with billion dollar capacities actually makes all of this market much more efficient than it was in the pre-Sarbanes-Oxley days.  Michael Greve:  Yes? Paul Larkin:  Paul Larkin.  [Inaudible few sentences]  Also now you have a large number of foreign who are going to be superintendent [inaudible sentence] a reasonable way of trying to deal with this [inaudible] without some rough number of years [inaudible] and what would that be and why?  I m just curious.  I m guessing that I know what your answer is.  Why do you think five years [inaudible]? Richard Epstein:  Because my view is five years is at least one generation of technological innovation and in some cases as much as two and that what happens is I think after that point the predictions are going to be completely unreliable and so let the whole thing rip.  In the European context, my thing is probably five days is the optimal period for a consent decree, but that s on a different proposition.  One of the things that troubles me about the European administration of the antitrust, and I think the Microsoft case is an illustration of that, is that if you fashion decrees on an ad hoc basis it becomes essentially an adjunct protectionist trade policies, which they pursue by other means.  That is, there is nobody else in the Microsoft position or on the Honeywell position going over there, and what I see in them is an effort to use ostensible neutral principles for very partisan ends, and I think frankly on that issue, although God knows America is grateful on many trade policy issues, particularly direct restrictions, I don t think that we have quite the anti-foreign bias with respect to the antitrust decrees.  Maybe Doug can inform us to whether I m wrong, but I do think that s a differential risk going one side of the pond as against the other. Douglas Melamed:  A lot of differential risks are going against the pond, and xenophobia wouldn t be top on my list, although it s probably in the mix. Richard Epstein:  What s number one?  Ignorance? Douglas Melamed:  No.  I think they are reading the same economics text.  I think it is a statist tradition.  I think they actually believe government knows how to do it and I think we have a healthy skepticism about that.  On duration, I am tempted to say, Paul, for some decrees five years make sense but for some decrees, and I can think of a hypothetical, it might not.  What we really ought to do is have a longer kind of default duration with a kind of vigorous sense that the agency will entertain a motion to terminate if circumstances are such that it comes no longer to be needed.  My problem is that based on a handful of experiences I ve had, I don t have any confidence in the ability of agencies to terminate decrees.  There is just too much inertia.  They find some non-compliance and then think, for understandable reasons or institutional reasons, you can t terminate a decree until the party to it is clean as a whistle and it just never gets done. Richard Epstein:  First the presumption.  I thought one of the nice features about the Microsoft decree is that it ends after five years unless somebody can show good cause as to why it ought to be continued.  One of the other tactics that Judge Kollar-Kotelly used, which I think was also admirable, is that you require interim reports.  If there are signs of flyspecking noncompliance, you could sort of address them there.  I haven t read the last of the Microsoft ones, but I did read the earlier ones before I wrote the book, and essentially they were very dull reading because there is precious little by way of kind of antagonism that was taking place.  My sense is that the relatively sclerotic performance by Microsoft in the last five years against Google is in part attributable to fact that they have to change their corporate culture to be responsive to the outside risks.  There was some guy from KK&R who wrote this nice piece in The Wall Street Journal about the difference between business boards and compliance boards.  I don t know if you saw this.  A business board is where people have technical expertise and they re trying to tell folks in the business what the hell they ought to do in order to get their product out the door.  A compliance board is essentially somebody who slaps you ever time it turns out you re doing something which might provoke a regulatory response.  If you re Microsoft and every one of your activities, it s not just that you have a legal situation; you have a good reason to switch your mix of people to compliance people. Folks like me. I cannot think of anything more deadly for a corporation than to have lawyers with oversight with respect to innovation.  That s another kind of very substantial cost that you have to take into account.  That one you could quote me on. Male Voice:  The idea that Microsoft is somehow intimidated by the government is just simply belied by the current facts.  To have a company come out with an operating system that is incompatible with everything except the Microsoft software is not exactly a pro-competitive stance.  I m talking about Vista.  That s obviously a calculated step on the part of Microsoft, and the Antitrust Division ought to be doing something about it. Richard Epstein:  The one thing I will plead on that is ignorance with respect to the Vista situation because it s post the book, which I m happy to provide.  The answer is that it certainly doesn t do anything with respect to the consent decree.  The point that you make is surely correct.  If there are anticompetitive features with respect to system in which the incompatibility to design to frustrate interchange, that goes to the core of the theory I m about.  You don t need to have the consent decree only; you could start an initial kind of investigation.  On the other hand, before one makes that judgment, you want to hear the other side and figure out what s going on.  I think there s actually a huge risk to the Microsoft strategy.  Folks like myself who are low-tech types, have no incentive whatsoever to shift onto a system like that if it s more bulky and more complicated.  I want good word processing.  I can assure you that Vista can t improve that very much over what is that I have running right now, which I don t even know. Michael Greve:  On that cheerful note, please join me in thanking Richard and Doug for a terrific discussion. [Applause] [End of seminar] [End of transcript]  </body></html>