American Enterprise Institute

May 28, 2008

[Edited transcript from audio tapes]

8:45 a.m.	Registration and Breakfast
9:00	Presenter:	Michael Perino, St. John’s University School of Law
	Discussant:	Peter J. Wallison, AEI
	Moderator:	Theodore H. Frank, AEI
10:30	Adjournment

Proceedings:

Theodore H. Frank:  We will go ahead and get started.  I’m Ted Frank, director of the AEI Legal Center for the Public Interest.  Thank you for fighting the crowds to be here today.  Milberg Weiss in its various incarnations at one point had a majority of the market for bringing securities class actions; they literally dominated the field and struck fear in the heart of every corporation whose stock happened to have an unfortunate decline.  When the indictment came out in 2006, I was asked to testify before Congress on the subject and Congress holds congressional hearings and brings corporate executives in front of it all the time, baseball players -- investigates various scandals at all times and this time they want to look at Milberg Weiss in 2006. 

The first thing that Democrats on the Financial Services Committee did was move to hold the hearing in executive session and closed it to the public and when that did not happen, they essentially berated the witnesses, including me, for even discussing an indictment when somebody was innocent until proven guilty, and they probably did not do anything and all they were doing were trying to help investors. 

That is an empirical question.  Was what Milberg Weiss was doing just a matter of indifference for investors?  Were the kickbacks a means to a greater end to help investors?  Were they returning more to investors at a small expense and the benefits outweigh the costs in pursuit of the so-called greater good?  You can look at that question.  You had an indictment that followed the financial transactions and could identify where the kickbacks were made and what cases; and you can compare those to the other cases that are out there, to the other law firms that are out there. 

We have been very fortunate that Professor Michael Perino has done that and has come out with a very interesting paper that we hope to have in hard copy for you today, but we have instead in soft photocopy formats in your conference packets and he will be discussing it.  He is currently the Dean George W. Matheson Professor of Law at St. John’s University School of Law in New York, where his interests are securities regulation and litigation, corporations, and judicial decision-making.  He has also been a visiting professor at Cornell Law School and at Colombia Law School, as well as in Stanford Law School.  He has authored numerous articles on securities regulation and class action litigation, and his article on securities litigation was of critical importance in an acting Securities Litigation Uniform Standards Act of 1998.  Professor Perino.

Michael Perino:  I’m told I have to use this microphone for the overwhelming crowd here, so I will do so. 

Good morning, thanks for coming today.  And thank you for your efforts in getting everything out as quickly as you did; I appreciate it. 

Mel Weiss, no doubt the Dean of the Securities Class Action Bar, is scheduled to be sentenced, I believe, on Monday.  And Mr. Weiss, of course, pleaded guilty to conspiracy charges stemming from a 25-year scheme and that scheme was to kickback a portion of his firm’s fees to what have been dubbed “professional plaintiffs”, individuals who bought tiny stakes in likely litigation targets in order to serve as the required injury investor in any subsequent lawsuits.  Now, he is going to join three of his partners - or, I should say, former partners - who have also pleaded guilty in connection with this scheme. 

The prosecution against the firm is still pending.  They have dropped the Weiss name, it is now Milberg LLP.  Apparently the joke around the firm was that if Larry Milberg had not died in 1989, they have to call the firm LLP. 

In any event, as Ted points out, this paper examines an empirical question that is at the heart of what at times had been a relatively vigorous debate about whether this firm should have been prosecuted or not, whether the government should have pursued this investigation.  And that question simply is this:  Were the firm’s clients, the members of the class of investors that the firm represented harmed by this kickback scheme? 

Now the government, of course, says, “Of course they were.  They had to have been harmed by this scheme.”  The reason they say that - and these just are a couple of quotes from the indictment -- is they said that the kickback “created a conflict of interest between the plaintiffs and the absent class members,” who are out there in the lawsuit.  Now, the lead plaintiff in these cases, the way these kickbacks worked was that Milberg Weiss would kick a percentage of their fees back to the lead plaintiff who had been willing to go out by these securities and serve as the lead plaintiff in the case.  What that meant was that the lead plaintiff could boost his recovery in the case by allowing the lawyer to charge a higher fee. 

So the bottom line for the government is that the kickbacks led to excessive fees, every dollar that was paid in kickbacks was another dollar that could have gone into the class suit’s pockets, according to the government.  And of course the firm, well up until recently anyway, said, “No, no that is not true at all.”  Now after Mr. Lerach pleads guilty, he was quoted in the Wall Street Journal as saying this, “The clients that we represented, cared about, and fought for were not harmed by this wrongful conduct.”  And for a little while, actually for quite awhile, the firm had a website up called, milbergweissjustice.com and now in their more contrite phase, they have taken the website down.  But while it was up - they had this website - it was basically devoted to attacking the government’s theory of the case and their basic argument summarized in that quote is this, “Kickbacks were slice of the attorney’s fees.  So this was money that was coming from the firm’s pockets and not from the pockets of the class.”  That was what they said. 

They also added, “Look, kickbacks are percentage of fees and because the dollar amount of fees generally rises with the dollar amount of settlements, now the representative plaintiffs, theoretically, have an incentive to maximize settlements, right?”  “Class members,” according to Milberg Weiss, “thus, may have benefited from the kickbacks if they led the representative plaintiffs to hold out for larger settlements.”  So kickbacks were a good thing. 

In any event, they say, “Look, it is the judges who award the fees at the end of the day.  The judges are going to look at any request; if it is an excessive request, they are going to slash that request.  And so, the judges are there; we couldn’t have gotten excessive fees even if we wanted to because the judges would have knocked down any fees that were too high.”

Well, those two positions basically set out the research questions that I wanted to explore in this paper and it is really four questions here.  First, addressing Milberg Weiss’ take on the effect of kickbacks.  In the indictment cases -- the settlement in the indictment cases, all that is being equal, higher than settlements in the other cases in the database.  Second, is there any correlation between being named in the indictment and either fee requests or fee awards?  We also might want to look not just at the indictment cases versus the non-indictment cases, but also at the Milberg Weiss cases versus other cases.  Is there any correlation between Milberg Weiss and any fee requests or fee awards?  And then finally, what about judicial oversight, was judicial oversight sufficient and effective at reducing excessive fee request?

What I did was I looked at a sample of 731 settlements in securities class action lawsuits.  Now, those settlements are in a time period.  The underlying cases were filed from 1984 to 2005 and settled from 1991 to 2007; that time period includes the time period in which the case is alleged and the indictment occurred.  The individual units of analysis I focus on are individual settlements.  So some cases will have multiple settlements, right?  You will settle [indiscernible] with a couple of different defendants and that is going to lead to more settlements than there are cases.  So in this sample, I have 695 non-indictment settlements and 36 indictment settlements.

Now, the key to figuring out whether it is the allegations of the kickback versus something else is controlling for the other variables that are going to affect either settlement size or fee awards.  When you can think about it very simply, a case like Enron, is going to have a lot higher settlement than a case involving a small biotechnology start up or something like that.  And so, what I’m going to control for -- I’m not get into the details, they are all in the paper but basically what you want to control for are various characteristics about the issue or the underlying issue or the other defendants in the case; the case characteristics, what are the allegations, that sort of thing and then also, characteristics about the settlement.  Was it in cash or was it not in cash, that sort of thing.  So I’m going to control for all of those factors in trying to figure out what, if any, effect these kickbacks had.

So let’s start with my first question.  The first question had to do with settlement size.  Now, what this figure is showing are mean average and median settlements for the non-indictment cases on the left and the indictment cases on the right.  Now, the average, in blue there -- the average non-indictment settlement is larger than the indictment settlement, but the difference is not statistically significant.  There are a bunch of outliers there, which inflates those averages the way they are.  If you look at medians, you get exactly the opposite result.  The median settlement in the indictment cases is higher than the median settlement in the non-indictment cases, but again, those simple comparisons do not tell us very much about the impact of cases because there are a lots of variables out there like the case quality or the size of the damages in the two different sub-samples that might explain these results.

So what we have to do is we have to control for these other influences and I do that using what is called the multiple regression analysis.  I will not bore you with the details but basically, the key here is we are going to include a variable in that analysis called indictment.  An indictment is going to take a value of one for cases alleged in the indictment and zero otherwise. 

So start with this first regression here.  What is it telling us?  If you think about what Milberg Weiss is saying, they are basically saying what we should see is this: if plaintiffs receiving kickbacks had incentives to obtain higher settlements and they were actually able to do so, then there should be a significant positive correlation between settlement size and that indictment variable.  But that is not what the data shows.  What the data shows is that indictment has a negative correlation.  For some are actually lower, but, of course, the differences there is not statistically significant.  It could be due to the kickbacks or something along those lines, or could be just due to random variation in the sample.  Essentially what that shows is that, for all intents and purposes, the settlements in the indictment versus non-indictment cases are statistically indistinguishable - no difference really between the two of them. 

Now, one potential objection to relying on the cases alleged in the indictment is we do not know the criteria that government used to choose which cases they were actually going to name in the indictment.  For all we know, the government sat down with a long list of cases, in which it thinks kickbacks were paid, and picked out all the ones with low settlements.   If that were the situation, then all we will be measuring with the Indictment variable is whether the government was good or bad in picking out cases with low settlements or not.  So to deal with that problem, what I did was I went through the database and I found all of the instances in which a case contained one of these repeat professional plaintiffs - that is the Repeat variable; and I used that instead of Indictment.  On the theory that basically the same deal should have been in the works in each of these cases involving the same plaintiffs. 

Now, here we have the sign flips, right?  We have a positive correlation here, which suggests that settlements are little larger, but again, statistically insignificant.  There does not appear to be any link between settlement amounts and professional plaintiffs and that result is inconsistent with Milberg’s claims about the effects of paying kickbacks.  They do not seem to have led to any higher settlement amounts.

 Take a look at regressions four and five for a second; I’ll come back to three and six in a minute.  The first set of regressions measured -- our dependent variable there -- was the total size of the Settlement.  There is another way to measure recoveries in these cases.  The other way to measure recoveries is to look at the ratio of the settlement to a proxy for the damages in this case, which is MDL, Maximum Dollar Lost, which is used in a number of papers.  And the idea here is: did you get more of the available dollars in the case? 

 And theoretically, again, if the plaintiffs getting kickbacks were holding out for higher settlement values then that number should be higher as well.  But again, if you look at four and five, whether you use the Indictment variable or the Repeat variable, you get a number that is negative but, again, statistically insignificant.  Again, it does not seem to be any correlation here and this is again inconsistent with Milberg’s claims about the effect of paying kickbacks. 

The only time we get a significant correlation is in regression six.  Now what regression six does is to just look at this subset of Milberg Weiss cases.  Let’s compare Milberg Weiss cases alleged in the indictment versus other Milberg Weiss cases in the sample.  Are there differences between the Milberg Weiss cases?  And what you see here is a negative and significant correlation.  It seems that Milberg Weiss actually recovered less of the available dollars at stake in the indictment cases versus other cases they handled, and that is significant. 

 Now, the problem with that result is that we do not know what is driving it.  I did not include it in this table, but if you replace Indictment with Repeat and run the same regression, you still get a negative result, but now it is insignificant.  It could be driven by the kickbacks, but I think the most conservative view to take of this data and the conclusion I reached is, there is simply no correlation whatsoever between settlements size and recoveries for the plaintiff and paying kickbacks.  There is just seems to be no correlation at all, completely inconsistent with the claims that Milberg Weiss had made in their defense.

 Now, there is something that is worth pointing out here.  Include a variable in here with Milberg, which is why it disappears in three and six because that is only the Milberg Weiss cases.  And if you look there, that variable is positive and significant throughout, which suggests, all else being equal, if Milberg is in the case, the settlement values were higher.  Now what does that mean?  Does it mean that Milberg produces better results than other law firms?  I suppose it is possible. 

It could also be though, that the cases that Milberg Weiss participated in were, on average, of higher quality than the cases the other law firms were participating in.  And it was that higher quality that is explaining the result, not anything that Milberg Weiss happened to do.  I cannot distinguish those two things in this analysis; although, it is worth pointing out that the whole point of paying kickbacks was to do what?  To get a leg up in getting the best quality cases.  All of that are settlement values.

 The crux of the government’s argument about harm and the conclusion there seems to be there does not seem to be much correlation between the kickback cases and settlement values; although, there seems to be this correlation with Milberg.  The crux of the government’s theory was fees, not settlement values.  So let’s turn to fees.  We are going to start with fee requests.  Now, we are going to look at fee requests in the non-indictment cases versus the indictment cases.  First thing we see here, average and median fee requests in the indictment cases were respectively little over 29 percent and 30 percent, compared to nearly 32 percent and one-third in the indictment cases.  Those differences are statistically significant.  So we have evidence of higher fee request in the indictment cases than in other cases in the database. 

 We can also look at what Milberg Weiss asked for versus what other law firms ask for.  And if we compare the Milberg cases to the non-Milberg cases, we see that the Milberg and the medians are exactly the same but the Milberg cases have a significantly higher fee request - 30 percent - than the non-Milberg cases, just under 29 percent.  So Milberg seems to be asking for higher fees than other law firms. 

And then finally, we can look at Milberg Weiss’ fee request behavior in the cases alleged in the indictment versus the other cases that are in the database.  And what we can see here is that Milberg Weiss asked for significantly lower fees in the non-indictment cases than they did in the indictment cases.  Now of course, that, again, could be because the government only chose to name in the indictment cases with high fee requests.  So if we look at that Repeat variable instead of the Indictment variable - and I do not have a slide for that here - but we see precisely the same relationship; fee requests in the cases with repeat plaintiffs are significantly higher than those without - 31.7 percent to 29.8 percent.  

 Now, to get a better sense of how kickbacks may have affected fee requests, take a look at this slide, which plots settlements on the bottom versus fee request on the top - this is logged to deal with outliers and things like that.  I think a couple of things jump out in this picture.  First of all, there is obviously a strong relationship between settlement size and the size of the fee request; it is almost a straight-line linear relationship between the two of them.  Second, the red dots in there - maroon, red, whatever - represent the indictment cases and what you see for the most part is - look at where those red dots are - they are clustered in a narrow band at the high end of fee request for any given settlement value and they do not vary very much.  Unlike the other cases which have a lot of variations.  So they are all clustered at the top of the high range.

 Finally, we have two lines in here; this is just a linear prediction of what the relationship between settlement and fee request is.  This bottom green line is for the non-indictment cases and that orange line is for the indictment cases.  At the bottom line, they are close together, so there does not seem to be much of a difference between fee requests in the two sets of cases, but the lines obviously start diverging as settlement size increases.  What that means is that as settlements grew larger, the size of Milberg Weiss’ fee requests grew in a faster rate in the indictment cases than in the non-indictment cases.  So this picture is consistent with the government’s theory of harm, that when there were kickbacks here, the fee requests were higher than when there were no kickbacks. 

 Obviously, you want to test that using a regression analysis and there are a lot of numbers on this page and I’ll break them down a little bit for you.  Let’s start with two and three.  The way you test whether there is a differential relationship of this kind in the data, is by using what is known as an interaction term.  That is just multiplying the Indictment variable by the Settlement variable and that is Indict-Settle there on the third line of the chart.  You are going to include both of the Indictment variable and that interaction term in there, to test whether there is a differential relationship here.

 What we see here is that relationship - at least in regression number two there, which is for the full sample - is significant but only at the 10 percent level.  If regression three drops us some cases that are severe outliers and there you have a relationship that is significant at one percent, but regression two, it is only significant at 10 percent.  Social science convention says, “It is not really significant unless it is five percent.”  It turns out we just missed the five percent threshold here.  The reason why is probably this: we have a very small sample of indictment cases and we have a very small standard deviation in fee requests.  You put those two things together and basically what they mean is you need a much larger sample of cases in order to find statistical significance. 

Now, there are various statistical techniques that you can use to deal with that problem - I will not get into the details but I did them, they are all in the paper - and if you use those techniques what you find is that Indictment variable and the interaction term for the full sample are significant at about the one percent level.  So it really does seem that there is a difference in the fee requests in the indictment cases versus the non-indictment cases.  So what does it mean?  What it basically means is this, what is the size of this relationship?

For every one percent increase in settlement size, Milberg Weiss asked for about .08 percent larger fees in the indictment cases than in the non-indictment cases.  So what is that going to mean?  It is going to mean, as we saw in that picture before, in the small cases, the fees are virtually identical but as the cases get larger and larger, the fee requests start to diverge from one another and the fee requests in the indictment cases are significant and larger.  So, consistent with the government’s theory of harm to the -- although, to a limited extent the government said, “this led to a higher fee request and fee awards across the board,” that is not quite sure -- it seems to have lead to a higher fee requests but only in the largest cases.

Now, the Milberg Weiss variable here is positive and significant at 10 percent; if you use those same statistical techniques I was talking about before, it comes positive and significant at less than five percent.  So what this suggests is that, all else being equal, Milberg Weiss asked for about five percent larger fees than other law firms.  I want you to notice for a second how well that finding fits in with the findings on settlement size.  Remember there, what I said was that the Milberg Weiss variable was positive and significant, and Milberg Weiss says, “Hey, that is because we got better results than other law firms.”  Let’s say that is true for a second.  Assume it to be true for a second.  If that were the case, would it be terribly surprising that Milberg says, “Hey, we should be awarded for our better efforts and get higher fees than other law firms”?  No.  So that would explain this positive coefficient we see with the fee request.

But those first regressions on Settlement size also found no statistically significant correlation with the Indictment variable, but the firm still asked for increasingly higher fees as Settlement size increased, even though there seems to be no basis for claiming better results in those cases.  Now again, we want to run it with the Repeat variable rather than the Indictment variable and if we do that, we get results that are virtually identical.  So this does not seem to be a by-product of the cases the government alleged, it seems to be consistent with the government’s theory of harm in these cases. 

Okay, so we can quickly go to fee awards because fee awards look remarkably like fee requests, because it turns out judges do very little when it comes to monitoring the request that attorneys make.  If we compare fee awards in the indictment versus the non-indictment cases, we see these significantly higher fees in the indictment cases versus the non-indictment cases. 

If we look at Milberg Weiss cases, we see fee awards are higher in the Milberg Weiss cases; although, the difference and means is only significant at the 10 percent level within the Milberg Weiss cases, fee awards in the indictment cases significantly higher than fee awards in the non-indictment cases.  We can plot again settlement values versus fee awards and the picture looks remarkably the same.  As settlements grew larger, the fee awards were an increasingly larger portion of the total settlement in the indictment cases versus the non-indictment cases.  The regression results look pretty much the same.  Although, here, we get a significant result just using standard linear regression; we do not have to do anything fancy here.  It looks like that for each one percent increase in settlements, Milberg Weiss was awarded about 0.1 percent more in fees in the indictment cases than in the non-indictment cases.

Milberg is, again, significant here although most of the time only at the 10 percent level.

This is one difference that we get from the results on fee requests.  Again, this regression is just for the subset of Milberg Weiss cases.  So we are comparing Milberg’s behavior in one set of cases versus Milberg’s behavior in another set of cases.  Compared to the other cases it handled, on average, as settlement size increased, Milberg Weiss received increasingly larger fees in the indictment cases versus the other cases it handled.  So it seems like the kind of plaintiff may have made a difference in these cases.

We can do it for all of the repeat plaintiffs rather than in the indictment cases - same results.  So again, they are not a byproduct of the cases that the government chose to allege in the indictment.

All of this suggests that courts did not do much to monitor fee request since the fee awards look almost identical to the fee request.

There is another way to look at that.  We can look at the ratio of the award to the request, so if we get a ratio of one, the judges gave the lawyers exactly what they asked for.

So that is a histogram of the ratio of award to request, and you can see that in more than half of the cases, the judges gave the lawyers exactly what they asked for.  On average, judges awarded lawyers over 90 percent of the fees that they requested.  So Milberg’s claim about judges’ judicial monitoring of the excessive fees does not seem to pan out very well.

We can take a quick look and compare the non-indictment to the indictment cases in terms of the ratio.  Although the ratio is a bit higher in the indictment cases, there is really no statistical difference between the two of these.  So there is nothing to suggest that judicial monitoring did anything to alleviate these effects.

Bottom line - basically, if you look at these data, they appear to be consistent with the government’s theory of harm.  There is no statistically significant correlation between kickbacks and class recoveries in the indictment cases.  Milberg Weiss asked for and got a greater share of the settlements than it did in the non-indictment cases, and the judges did not appear to do much to reduce excessive fee requests.

Theodore H. Frank:  Thank you, professor.  Commenting will be AEI’s very own Peter J. Wallison, the Arthur F. Burns Chair in Financial Policy Studies here, where he co-directs the Institute’s program on financial market deregulation.  He previously practiced banking law at Gibson, Dunn & Crutcher in Washington and New York and was general counsel of the Treasury Department, where he had a significant role in the development of the Reagan administration’s proposals for deregulation.  During 1986 to ‘87, Mr. Wallison was White House counsel to President Ronald Reagan.  Mr. Wallison?

Peter J. Wallison:  Thanks very much.  I guess I need a mic.  Thanks, Ted.

I guess I will start with all kinds of reasons why I’m not really qualified to review this paper.  I did not follow this case very closely in the press, so I really do not know much of the details.  I never was a criminal lawyer in practice.  I tried to avoid indictment when I was in the government.  That is basically my experience with the criminal law.  I’m not a quant, and so, a lot of these numbers that Michael has thrown around are going to go by me, but I did think this was a terrific piece of scholarship and work.  As we go through it, I’ll just raise a few questions that occurred to me as I was reviewing the paper.  I’m sure all of you have your own views and questions you would like to raise.

The paper does to me, at least, seem to show that in certain circumstances, the kickbacks could have created incentives for larger fee requests, and if there were larger fee requests, that probably meant that if those additional amounts had gone to the plaintiff class, they were hurt by the fact that Milberg Weiss had to reward the people who were receiving kickbacks for their assistance in creating the class or becoming the first class plaintiffs.

It did seem to me there were some unanswered questions here though that made the whole thing just very difficult to demonstrate in a way that, at least to my mind, made it conclusive.

The increased size of the fee requests, were they commensurate in any way with the size of the kickbacks?  We do not really know.  At least, I could not find in the paper any indication of what the size of the kickbacks was.  So I felt a little uncomfortable knowing whether the fee request that Milberg Weiss was getting were in some way related to - in a significant way - what they were actually paying to their kickbacked, bribed - whatever you want to call them - plaintiffs, and it would have helped if there was some way to know that.  Maybe it is in the paper and I could not find it, but that was one area that it seemed to me would be helpful in establishing Michael’s premise if we could have seen a little bit more indication of a relationship between those two.

Michael talked about this, but I’m still not quite satisfied with the idea of these higher fee requests being approved because they were intended to provide a fund out of which they could pay off their class plaintiffs.  For one thing, it is very hard to demonstrate these things, of course, or even to know, but there could have been a superior lawyer in here.

Milberg Weiss was a very fine firm, and they were exceedingly successful over a long period of time in getting and extracting very large settlements from companies.  I know a little bit from my limited corporate practice that when a company faced Milberg Weiss, it was of a different level of conflict from any other old class action plaintiff firms.  Milberg Weiss was the tiger of the industry, and it may well be that they were simply just better, more aggressive lawyers.  That could explain in part why they were able to ask for and receive higher fee requests apart from whether they had a class plaintiff they had to kickback to.

Also, there is no way of knowing whether the cases were tougher.  I’m not saying it could be known.  I’m saying it probably cannot be known without a lot of investigation and a lot of judgmental activity to know whether the cases were tougher that they were handling.  They might well have been, and that would also have justified a higher fee request for whatever the result was that they received for the class.

The paper notes that where large kickbacks are combined with small holdings, the incentive of the plaintiff is to increase the fee request, and that was a pretty persuasive thing, I thought, but there is no data on the size of the lead plaintiff’s holdings.  So we do not really know in each case whether the lead plaintiff had a significant holding or a small holding.  I think we are all assuming that these lead plaintiffs had bought very small amounts of shares in each of these defendant companies so that they were prepared when the moment came to be the class plaintiff, but we do not really know that and it would be helpful, given the argument that Michael is making, to know what their holdings were.

Finally, this question of the degree to which the court disciplines the fee requests, this is a particularly puzzling one for me.  Because if the courts are granting fee requests, 90 percent, almost much of the time, just what the lawyers asked for, I do not understand why more of these law firms that were competitive with Milberg Weiss were not asking for higher fees.  If the conduct of the courts is well known or obvious that they mostly grant a higher fee request, why were these other firms also not asking for higher fee requests even if they did not have a class plaintiff they had to kickback to?

So that part of this argument did not make a lot of sense to me, and I’m sure Michael has some good explanation for it, but I was bothered by it.

There are a couple of other things that I thought were really interesting about this.  The article did not explain the actual legal basis for the prosecution.  I do not know anything significant about the criminal law, but I assume there is a statute of some kind that was the underlying basis for the act - for the indictment - which was an anti-kickback law or something of that kind.  I do not know what it was.

But assuming that there is such a statute that made this a prosecutable case, the question of whether these kickbacks were helpful to the plaintiff class or not helpful to the plaintiff class does not seem to me to be particularly relevant.  There are such things as harmless kickbacks.  They are frequently known as discounts, and those things are perfectly okay.  On the other hand, there are harmful kickbacks, and they are known as bribes.

The difference, I think, and I have not really had time to think about this in too much depth, but the difference I think is that an unlawful kickback or a bribe will distort the economic relationships which the law presumes will be free of influence. 

So let’s consider, for example, the question of the government official who awards a contract and gets a kickback.  We would automatically say that is obviously prosecutable, but would it be a defense that the contractor had - that is, the person who won the contract - the best product?  That, in fact, no one was harmed here because, even though I got this kickback, I awarded the contract to the best party to receive it.  I think most of us would say that is ridiculous.  That cannot be a defense.

Why do we say that?  We say that, I think, because we understand that an anti-kickback statute is intended to have a deterrent effect of some kind, and we do not want to leave the case in a state where it is necessary for the government to prove not only that the bribe was given but that the bribe was not actually harmful.  We just want to prosecute bribes.  So we do not really have to show harm to anyone in most of these cases.

So if we needed justification, we could argue that the kickbacks, let’s say for this particular case, the Milberg Weiss prosecution, we could argue that the kickbacks increased the likelihood of litigation making it easier to find plaintiffs.  The common law litigation system, I think, assumes that lawsuits will only be brought where there is a real interest on the part of the plaintiff.  Congress and the courts that have let this problem process go on for a long time could well have assumed that class plaintiffs would have substantial interest in themselves and would not simply be recruited and paid for the purpose of being class plaintiffs. 

So there is an underlying fault here in allowing people to be paid to be plaintiffs - barratry or champerty or those things that, in the old common law, people used to be indicted for.  If no one did that, then the case probably should have not been brought.

In other words, the harm was done to the economy, to the companies who were charged and attacked, to the litigation system, to the courts, but it seemed to me somewhat irrelevant whether the harm was done to the class plaintiffs.  The reason, I think, I’m troubled by this is that I do not understand yet - and I’m sure Michael will make it clear - what the underlying statutory provision was that caused the Justice Department to bring this case.

So if under those circumstances, ultimately, I come out here a little bit dissatisfied after what is really an excellent paper with the thought, “Why was all this brilliance devoted to an issue that, even if it demonstrated that the kickbacks caused the class plaintiffs not to receive as much money as they otherwise would have, even if that had been demonstrated or the opposite had been demonstrated that it did not have any effect, it would not have affected the prosecution because the prosecution was brought under an entirely different theory?”  This seemed to me to be something that was just added on by the government.  Why was so much time and effort spent to prove that the government was correct in its theory?

That is where I come out on this, Michael.

Theodore H. Frank:  Mike, just as some background, the kickbacks are what, about 10 percent of the attorneys’ fees?

Michael Perino:  Actually, for purposes of time, I left this out, but let me put this up because I think it will help.

Basically, the way this works -- do I have to come back to the microphone?

Theodore H. Frank:  Yeah.  You probably will have to.

Michael Perino:  Basically, the way this scheme worked is Milberg Weiss develops a relationship with somebody like Seymour Lazar.  Seymour Lazar was a retired lawyer in real estate development from Palm Springs, California - yeah, the kind of poor impoverished plaintiff that Milberg Weiss always says they represent.

Lazar would go out and buy, presumably, small stakes in likely litigation targets.  We do not know from the indictment, and I cannot tell from the documents that I was able to access, exactly what the size of those was.  I agree.  It is something I would like to know.

Presumably, it is going to be irrational for someone like Lazar to buy big stakes in these companies.  He wants to spread out his money as widely as possible.  He does not know which companies are going to have the stock price suits.  At the time that these practices were most prevalent, the size of your stake in the case was an irrelevant consideration in terms of being named the lead plaintiff in the case.  So it really does not make much sense to buy more than 100 shares or so or maybe even less than that.

Theodore H. Frank:  One question.

Michael Perino:  Sure.

Theodore H. Frank:  What are the requirements to be a lead plaintiff?

Michael Perino:  At the time, there really were not any -- the court would consider whether you were adequate to serve as the plaintiff, but that was typically a question of whether your claim was typical of the other class members’ claims, which, of course, it always was.

Most of the time, during this time period, judges appointed lead counsel in the case as the law firm that files first.  Not true now.  We have the whole lead plaintiff provision under the Private Securities Litigation Reform Act.  We have a whole different system now, but then, first to file made a difference.  Being lead counsel was quite advantageous.  Lead counsel ran the case, divvied up the work and collected the lion share of fees at the end of the day.

So it made economic sense for Milberg Weiss or any other law firm for that matter to do anything they could to get to the courthouse first, which is why having somebody like Lazar ready, willing and able to serve as the plaintiff was so useful to them.

So we file a lawsuit, the lawsuit gets filed, assuming we get a settlement at the end of the day, Milberg Weiss is awarded 20 to 30 percent - 20 to a third - of the settlement, and then Milberg Weiss kicks back between 5 and 10 percent to somebody like Lazar or one of the other professional plaintiffs.

Theodore H. Frank:  Five to 10 percent of its fees?

Michael Perino:  Five to 10 percent of its fees.

Theodore H. Frank:  You talked about the ratio of the settlement to the maximum damages law.  What kind of ratio are we talking about there?  Are the plaintiffs getting back 80 percent of what they lost or is it much lower than that?

Michael Perino:  Much lower than that.  We are talking -- I do not have the number right in front of me, but 10 percent or less is more likely the number.

Peter J. Wallison:  NERA had a study that was two to three percent.

Michael Perino:  They could get less than 10 percent.  I just do not remember what the number is off the top of my head, but I want to come back to your point about what the effect of this was.

Assume a hypothetical.  Assume that we have Lazar.  Lazar owns 100 shares that have been theoretically damaged in this lawsuit.  There are a total of 200,000 shares.  These are just numbers that I picked to show the hypothetical.  All the affected shares - just to make the math easier - equally affected, and Lazar is going to receive a 10 percent kickback.

This, Peter, goes to your question about, “Was the increase in fees matching what the lead plaintiffs were getting?”  So if you look, first of all, at scenario number one, let’s say we have a US$10 million settlement.  Attorney’s fees are going to come off the top.  Let’s assume an attorney’s fee of 25 percent which is a fairly common attorney’s fee.  So the net recovery to the class is going to be US$7.5 million.

Let’s look at the representative plaintiff’s recovery.  Lazar only owns 100 shares.  So his recovery as a class member is just going to be US$3,750.  His kickback is going to be US$250,000, so more than 99 percent of his recovery comes from the kickback and not from his recovery as a class member.  What Milberg Weiss says is, because fees rise with settlement size, this is going to even cause the plaintiff to holdout for higher fees.  That is scenario number two.

Let’s say we holdout for another year.  We get another US$2 million in settlement, same 25 percent fee, is the class better off?  Yeah, the class is better off.  If this was what happened, Milberg Weiss was right.  Instead of giving US$7.5 million, the class gets US$9 million.  Sure, the lawyers are better off, and Lazar is better off too, but so is the class.

So in this scenario, kickbacks would actually -- or we could say if we want to put a less pejorative term on it, revenue sharing would be a good thing for the class members.

Notice though the recovery is here.  Lazar’s recovery as a class member is just going to be US$4,500.  He has only gone up a few hundred dollars because he has such a small stake in the underlying action.  His kickback though goes from US$250,000 to US$300,000 for a total recovery of US$304,500.

Now, what if there is an easier way to get the same recovery?  Instead of rejecting the US$10 million settlement, I take the US$10 million settlement, and now, I say, “Okay, charge 30 percent instead of 25 percent.”  Does it matter whether the recovery to the lead plaintiff, the kickback, matches up with the attorney’s fees?  It does not really, because now, the lawyer and the representative plaintiff have exactly the same incentives.  Because their recovery is largely determined by the size of the fee and nothing else, they both want to charge the maximum possible fee.

So here, the class is much worse off.  They get US$7 million as a total net recovery.  Lazar’s recovery as a class member is virtually unaffected.  He gets US$3,500 but his kickback is US$300,000.  He gets virtually the same recovery under scenario three that he got under scenario two.

The lawyer and the lead plaintiff were probably going to prefer this scenario to scenario two for a variety of reasons.  First of all, they get their money sooner.  They do not have to continue to litigate the case.  They avoid all sorts of opportunity costs and the additional cost of litigating the case.

In terms of the representative plaintiff, if he wants to do this, and they seem to want to do this again -- Steven Cooperman apparently appeared in 70 different lawsuits.  At one point, the courts called him the unluckiest investor in the world.

Well, if you show a little flexibility on fees, which is, by the way, your natural incentive anyway because that is how you are going to get most of your recovery, the law firm is going to want to use you again in the future, right?

So scenario three seems to present the most rational way for the plaintiff receiving a kickback to increase their recovery from the case, and I do not think it matters at the end of the day whether the amount of the kickbacks actually matches the rise in fees because the lawyers and the representative plaintiffs both had an incentive to make sure that the fees were as high as possible.

Theodore H. Frank:  Can you talk about some of the variables you used to measure the quality of the cases?

Michael Perino:  You raised, “Are these tougher cases?”  What I try to do is to control for everything that some studies somewhere on settlement values showed was correlated with settlements or with fee awards.  So I controlled for the size of the damages, the length of the class period, which obviously the longer the class period - theoretically, all odds being equal - the higher the damages should be in the case.  I controlled for the kinds of allegations in the case.  Was this an accounting fraud case?  Was it a case involving a restatement of earnings?  Cases that could be correlated with case quality.  I controlled for the kinds of defendants that were in the case.

Obviously, if you are facing the company and the accounting firm and the investment bankers, that is going to be a tougher case and a more complicated case to litigate than a case involving just the single issuer, for example.  So I controlled for that as well.

Basically, I’m explaining what this regression about a little more than 70 percent in the variation in settlement size which is as good as any model that is out there for settlement values.  Have I controlled for everything -- probably not.

Again, this kind of regression analysis shows correlation.  It does not show causation.  Is it possible at the end of the day that the cases that Milberg Weiss litigated were tougher than the cases -- whatever difference between the two of them?  Sure, it is possible.

Again, I come back to how nicely the results fit together.  If it was something about the underlying cases or something that Milberg Weiss did, then it makes sense that Milberg Weiss asked for and got higher fees, so we should see that correlation with the Milberg Weiss variable.  But we do not see any difference between the indictment cases and the non-indictment cases in terms of outcomes, and yet, we see the higher fees and the higher requests in those cases.  That, to me, suggest at least it is much more consistent with the government’s theory of harm than it is with Milberg Weiss’s theory.

Why does any of this matter, as you say?  What is the statutory basis for this?  The charge that everybody pled guilty to is, of course, conspiracy to do what?  Well, conspiracy to do a couple of things - wire and mail fraud is one of them but also obstruction of justice.  This scheme required various misrepresentations to courts over time that the plaintiffs were not getting any additional recovery.  That, you can say, is harmful to a litigation system and they do not need to worry about any additional harm to the absent class members.

But if you look at the motions to dismiss that were filed in this case and the government’s theory in the indictment, there is really a stark difference between the two of them.  The government is arguing, “No, this is not just us going after Milberg Weiss because we do not like class action lawsuits.  There really was some harm here.”  Milberg Weiss contending, “No, no, no, there is no harm,” and it is that dispute rather than the underlying legal theory that I wanted to address in this paper.

Peter J. Wallison:  I see.  So in a sense, it was political because Milberg Weiss had a significant amount of political support in Washington for what it was doing.  The Justice Department had to show that some ordinary people were being harmed in some way, not something intangible like the judicial system because of a mere lie to a court.

Michael Perino:  I think there is something to be said for that.  Many tried to portray the Milberg Weiss prosecution as simply Republicans going after class action lawyers, and they did not like class actions.  Now, put aside the issue that the investigation actually started during the Clinton administration, and they have collected a few dollars over the years from plaintiff’s lawyers, but put that aside for a second.

I think you are right.  I think there is an underlying political issue here.  The government had to come up with some reason why they decided to both pursue the prosecution in general and also indict the firm.  The other big dispute here is whether the firm should have been indicted in this case or not.

The government’s theory, I think, and the theory that separates this from a case like Arthur Andersen, you can see Arthur Andersen was a case where you had, theoretically, a limited activity confined to the Houston office and really one episode of shredding documents and destroying evidence in connection with Enron.  Does it make sense to prosecute the firm in the entirety for that kind of situation - probably not.  This appears to be the way Milberg Weiss did business though, and did it for a long period of time.

The other difference between the two of them, quite frankly, is if you look at the market - we are now down to the final four in terms of accounting firms - if you look at plaintiff’s class action lawyers, there are many people willing to step into the shoes of Milberg Weiss.

I think there are real differences between the prosecutions in those two different cases, but I think there is an underlying political reason here why the government took the position they took and why Milberg took the position they took.

Theodore H. Frank:  Let’s take some questions from the crowd.  We are recording this.  Is there a microphone there or are we just going to shout out?  We will just shout out.  Are there any questions?

Michael Perino:  I’ll try to repeat the question.

Theodore H. Frank:  Yes, sir?

Male Voice:  Hi.  I’m perplexed at the amount that Milberg Weiss paid the professional plaintiffs.  I worked on the Securities Litigation Reform Act, so I was aware of these kickbacks, but I did not realize how much they were.  Is there a legal exposure for the professional plaintiff in accepting a kickback?  Is there exposure for them?

Theodore H. Frank:  They are in jail.

Michael Perino:  Well, some of them are in jail.  Yeah.  Cooperman was already in jail.  Cooperman had faked the theft of a Monet and Picasso from his art collection and had been convicted of insurance fraud, and was trying to wrangle a somewhat lighter sentence and said, “By the way, I have evidence on Milberg Weiss.”  The John Grisham aspect of this case is quite high.

But Lazar pled guilty.  I do not recall exactly what he pled guilty to but conspiracy, obstruction of justice or some falsified document, something along those lines.

There were various lawyers who served as intermediaries to try to disguise the fact that the kickbacks were actually going to the plaintiffs.  So the lawyers would get money as consulting on this case, and the money would then get funneled to the plaintiffs.  Some of those have pled guilty as well.

So there is definitely a legal exposure, and I guess the real irony in this case is that somebody has brought a class action against Milberg Weiss on a RICO theory alleging that, as members of the class, they were harmed by this activity.

Male Voice:  They knew they were engaged in illegal conduct.

Michael Perino:  Presumably, yes.

Male Voice:  In these kickbacks, okay.

Theodore H. Frank:  Yes?

Male Voice:  First of all, one question we must ask of this event is incomplete.  Is more research needed?

Michael Perino:  I think I will be drummed out of academia if I say no more research is needed.  I think it would violate -- I can get my tenure revoked if I said there is no more research needed on the subject.

Peter J. Wallison:  It is an agreement in restraint of trade.

Male Voice:  I wanted to see it just one time where someone says, “Not only is no more research needed but we should not have done the research that we did.”

I came up with a theory of a brand premium for Milberg Weiss that I think Peter alluded to, that they are the Fannie and Freddie of the business.

Michael Perino:  It is possible.  As I said I cannot -- one thing I suppose that we could do would be to try to construct some matched sample analysis where we had cases that were as identical as we could possibly make them, and have one set of Milberg cases and one set of non-Milberg cases, and see whether the outcomes are really any different between the two cases.  That is not something I did with this analysis, but I suppose it is something you could do, and you could test whether there is actually - Milberg did obtain, I should say.  Well, Milberg is still around - did obtain better results for their clients than other law firms.

Again, I come back to the results of this analysis though.  Yeah, that might explain the higher fees that Milberg got at the end of the day, but it does not explain why the fee should be higher in the larger indictment cases than in the cases that were not alleged to the indictment.

Male Voice:  They do not seem statistically significant to the naked eye.  I know that you found, according to your criteria are significant and I do not mean to dispute that, but it is just difficult to figure out why a criminal case brought, and you start looking -- I did not realize that was brought during the Clinton era, if you are a cynic as I am, you then say, “Well, maybe someone in the plaintiff’s bar wanted to take out this competition had enough political juice to do it.”

I have just one observation about Andersen because I have followed that closely.  I believe that had Andersen taken proactives early on - a chief one being removing the leadership of the firm - the firm would have been saved.  That is a conjecture but I just --

Michael Perino:  But a reasonable conjecture, I think.

Male Voice:  Put their head down when -- and just reorganized enough because that was not the first option.  That raises the question, “Given the prosecutorial discretion that is always involved, why not -- given that there were also violations of law and practice -- use one of the vast array of sanctions, one or more, I’m sure of a criminal prosecution?”  It is unsatisfying.

Michael Perino:  Well, you will have to ask the prosecutors that.

Male Voice:  Pardon?

Michael Perino:  I said you will have to ask the prosecutors that.

Theodore H. Frank:  We would also have to ask the state bar associations that -- these things just do not get disciplined.

Male Voice:  [Off microphone] Given that so much, it does not even get disciplined that you go to the other extreme to have a criminal prosecution or something that is -- in other words, neither [inaudible].  It looked like you had a violation of some statute.  For whatever reason, someone decided that they needed to be able to make a good showing and then looking at what was produced; it does not look like they really did it.  I still raise the question, “Why?”

Michael Perino:  Well, there is one other thing I omitted to mention, which is, that the investigation started, and for a little while, Milberg continued to pay these kickbacks.  Also, the Private Securities Litigation Reform Act barred these payments in 1995, but there are a few, albeit it is not the majority, but a few of the cases where these kickbacks were paid even after they were banned by the statute and even after the firm knew that the government was investigating for these very practices.

Male Voice:  [Off microphone] That is likely Andersen [indiscernible] when you know that [indiscernible] at least Ronald Mack [phonetic] [indiscernible].

Theodore H. Frank:  Of course, Andersen’s conviction was reversed but that is --

Male Voice:  [Off microphone] That does not mean this will not be [inaudible].

Theodore H. Frank:  Well, these are all guilty pleas so there is nothing to appeal.  Are there other questions?  Yeah?

Male Voice:  Actually, I have two questions.  One is one of the points that Peter brought up.  Why is it that other firms, if the judges are awarding high fees on a regular basis, not -- maybe because they are not the tiger and they cannot get that fee from the judge, but whatever the case, why are more of these firms not asking for higher fees similar to what Milberg Weiss was getting?

Michael Perino:  I just do not know the answer to that question.

Male Voice:  The other one is you mentioned that the guilty pleas were for such crimes as obstruction of justice, wire and mail fraud, et cetera.

Michael Perino:  Conspiracy to commit those, yes.

Male Voice:  Okay.  So if as Milberg Weiss alleged during the case, and again, I did not follow it, either that these kickback schemes really did not harm anyone, perhaps that plaintiffs got money that they otherwise never would have received but for the filing of the suit, why was all this conspiracy to commit all these acts committed to hide this activity?  Is it because the 1995 Act forbids the kickbacks?  Was that it?  If you understand my question --

Michael Perino:  Sure.  I think there are two questions buried in there.  One question is, “Why did they disguise the underlying activity?”  The other question is, “Why is a conspiracy charge at the end of the day?”

It is a conspiracy charge at the end of the day because that is what they negotiated at the end of the day to plead to.  That is the answer to that question.

Why did they disguise this at the end of the day?  It was sort of an open secret that this activity was going on.  People had suspected for years that individuals were getting paid to serve as the plaintiffs in these cases.  If they are getting paid to serve as the plaintiffs in these cases, that raises the whole host of questions that Peter raised.  Why do we have these cases in the first place if we have to pay somebody to be willing to bring this case?  So that is, I think, why they tried to keep it under wraps.

Actually, if you go back to the legislative history leading up to the Reform Act, Mel Weiss and Bill Lerach were asked these questions at their hearings, and they did their best not to answer the questions.

I think it was Senator Domenici who asked Mel Weiss directly, “Provide us any information about any payments that are made to lead plaintiffs.”  And Mel Weiss said, “Well, you mean court-awarded bonuses, right?”  Domenici for whatever reason said yes.  So they answered about court-awarded bonuses.  Mel Weiss was very good.  Somebody else, and I do not remember who it was, made an allegation that the law firms has a computer disk with a list of all these plaintiffs on them and they get paid for serving in these cases, and Lerach wrote an angry letter back saying that this was libel and, “We have no such disk,” and left out the part about the payments.

Male Voice:  It was a negative [cross-talking] he made.

Michael Perino:  There were rumors and allegations about these kinds of practices all the time, and I think the lawyers for quite natural reasons tried to do their best to keep them under the table.

Theodore H. Frank:  Just an attempt to answer Peter’s question about the Milberg premium.  My recollection was that the race to the courthouse was really the key element to being the lead plaintiff.  And because Milberg had this stable of professional plaintiffs ready to go in most cases, and they did good research in terms of stock price drops, that they actually got the easy cases; the ones that were the highest potential settlement which was what really made a good case or a bad case.

So rather than getting a higher fee for better work, there really was not that much work involved because they took the cream cases and left the other cases for everyone else.

Michael Perino:  Again, that is also a possibility I raised during the presentation and from the data we have right now, we just cannot sort that out.

Theodore H. Frank:  Of course, when these cases are settling for 2 to 10 cents on the dollar that reflects not so much a meritorious case but a way to get out of the harassments of being deposed by Mel Weiss at relatively low expense.

One more question.

Male Voice:  This closes the circle.  What was alleged, I think, in some of these underlying cases, was the IPO bunch of cases.  There were all these tech dot bomb IPOs were the IPO would come out and then you would have the equivalent of these outhouse plaintiffs – it would be the purchasers of these stocks who would undertake and return for getting access to the IPO that they would then buy more and help goose up the performance of the IPO.  So you have the issuers allegedly doing the same thing that Milberg Weiss was doing.  Culture manipulation --

Theodore H. Frank:  That was a completely different litigation on these litigations.  In fact, those IPO litigations were dismissed.

Male Voice:  Would they have been the cases that Milberg Weiss was trying to --

Theodore H. Frank:  No, that is --

Michael Perino:  That is a different group of litigations.  That litigation happened largely after these practices, at least, according to the indictment, had petered out.

Male Voice:  Further research --

Michael Perino:  Well, by that point -- so those litigations came down, of course, after the bubble burst around 2000 or so.  So these litigations came down in about 2001.  At that point, we were under the Private Securities Litigation Reform Act, and at that stage, we have this lead plaintiff provision.

The way the lead plaintiff provision works is that the plaintiff who volunteers and who has the largest financial interest among the volunteers to serve as lead plaintiff is the presumptive lead plaintiff.  It is the provision that was intended to encourage large institutional investors to serve as the lead plaintiffs in these cases.

If you look at litigation today, it tends to be dominated by institutional investors - two groups, largely, public pension funds and union-affiliated pension funds.  I have done some additional research on this, and I find that, at least, with public pension funds, the outcomes appeared to be better; the fees appeared to be lower than in cases involving other kinds of plaintiffs.

The interesting question here is, “How has the game shifted?”  If the game is now not first to get to the courthouse, if the game now is develop the relationship with the institutional investor, what if anything is going on to develop those relationships with the institutional investors?  Suddenly, plaintiff’s law firms are making campaign contributions to the comptroller of various states.  Now, they may be interested in those comptroller races, but they also may be trying to do a pay to play kind of thing.

There have been stories in various publications about some of the relationships between law firms and union-affiliated funds.  So I think what we need to look at now is not this question of paying kickbacks to professional plaintiffs, but now look at the question of whether under our new system, we have created other potential problems that we need to address.

Theodore H. Frank:  I have promised -- [audio abruptly ends]

[End of file]

[End of transcript]