The tormenting of Guantanamo detainees by subjecting them to round-the-clock barrages of blaring rock music has raised a thorny, if thus far hypothetical, legal question: Is torture a "fair use" under the Copyright Act?

Several commentators have recently pondered whether songwriters who object on moral grounds to use of their music in this fashion might be in a position to enjoin such activity. (See the Patry Copyright Blog here, and an article in the U.K.'s The Register here.) I decided to take the legal analysis to a more prosaic level by simply inquiring whether royalties might not already be due and owing songwriters and music publishers for infringement of the performance rights in their works.

After all, songwriters need the money. Music-as-torture seems to offer rare growth potential in an industry otherwise in distress, ravaged by revenue losses brought on by digitization and its consequences: downloading, file-sharing, and the demise of the album. In previous years, we would, only once in a great while, see our government use copyrighted music -- mainly hard rock and heavy metal classics -- to break down a foe's will to resist. We saw it used in Panama, for instance, to drive Noriega from his palace and then, years later, in Waco, Texas, against David Koresh and his Branch Davidians (with unanticipated results). But with the arrival of the War on Terror and our liberation from previously crabbed interpretations of international human rights commitments, high-decibel music may now be becoming a fairly routine interrogation tool used in Guantanamo, Iraq, Afghanistan, and, perhaps also, an archipelago of secret C.I.A. prisons across Eastern Europe. I'm speculating, but for Van Halen, Metallica, the estate of Jimi Hendrix, and certain rap artists we might already be talking real money here. Likewise, Barney the Dinosaur's excruciatingly monotonous "I Love You" theme (see Patry post and links therein) has apparently been found by military intelligence officials to possess powerful, yet so far entirely unrecompensed, coercive properties. (I should acknowledge up front that the Bush Administration denies any use of "torture" anywhere in the world; whenever I use that term in this article I am using it only in its nontechnical, lay sense, the way the rest of the world seems to understand it.)

For my research, I asked three eminent copyright scholars the pertinent, if ordinarily un-askable, questions. Though sometimes ill at ease, they obliged me in the spirit of grappling with a preposterous law-school exam question.

At the outset, of course, my experts all noted that the U.S. Copyright Act only applies to infringements in the U.S., and they were uncertain of the nuances of the applicable laws abroad (particularly in the secret prisons, if they exist). William Patry, the author of a magisterial new treatise on copyright law (see here and my earlier post here), reminded me, however, that thanks to former ambassador Paul Bremer's puzzling priorities, Iraq does now have a copyright law much like our own. Both Patry and professor Jane Ginsburg, of Columbia Law School, also thought that Guantanamo would be treated as a part of the United States for these purposes, bringing the largest potential source of new royalty streams within the purview of the Act. Ginsburg noted, too, that one might even be able to argue that, to the extent U.S. officials were using music on U.S. bases in this fashion, the U.S. copyright law should "follow the flag," though she acknowledged that this would be an aggressive interpretation. (While the U.S. would probably not be protected from royalty demands by sovereign immunity, songwriters would have to sue in the U.S. Court of Claims, where there are no juries and damages are circumscribed, both Patry and Ginsburg observed.)

The key hurdle to royalty recovery that all the experts homed in on first was the songwriters' obligation to establish that bombarding prisoners with loud music in locked cells inside walled prisons could be considered a "public" performance. Nevertheless, the Copyright Act provides that the copyright holder's performance right will attach even to a performance held in a private place "where a substantial number of persons outside of a normal circle of a family and its social acquaintances is gathered." Certainly the Branch Davidian compound near Waco would seem to meet that definition. My panel of experts felt that even an interrogation held inside a sweltering container car in Guantanamo or Iraq might conceivably qualify as "public" under that provision, depending on how many interrogators and detainees were present at the time. "One soldier, one detainee?" says Patry, "I would think not." But if you had a small group of interrogators in there it would present a more difficult question, he says. "It's not a group of friends or social acquaintances."

If the playing of music as an adjunct to interrogation did count as a public performance, the obvious next question would be whether the government could defend its nonpayment of royalties by arguing that it was making a "fair use" of the works in question. Here my experts parted company with one another. Patry thinks fair use would be a "tough" argument for the government to make. It's certainly not one of the usual fair use exceptions, he says, like excerpting a work for purposes of scholarship, teaching, news reporting, or creating a new work of art. "They're not using the work for criticism or comment," he continues. "Quite the opposite."

But Fred Koenigsberg of White & Case -- who also serves as ASCAP's general counsel -- thought the fair use argument to be a weightier one. Referring to some of the key factors to be weighed in any such an analysis, he notes that interrogators are clearly not using the music for a commercial purpose and that the government could argue that, indeed, they're using it for purposes "relating to public safety and welfare, and to aid law enforcement" -- all factors that would weigh in favor of granting a fair use defense. Likewise, he notes, the effect of military use of the music on its existing commercial markets seems to be negligible. "Fair use would not be a laughable claim" in this context, he concludes.

In hypothetical rebuttal, Ginsburg suggests that the songwriters might contend that "if word gets out that their music is being used for these purposes, it might [well] have a deleterious effect on the commercial market for the music." (This might be a stronger argument for the publishers of Barney the Dinosaur's songs than for the publishers of the Metallica catalog.)

Clearly, no songwriter should imagine that collecting royalties for use of their works in Gitmo interrogations will be a slam dunk, let alone a windfall or future annuity. Yet as more traditional revenue streams for artists dry up, songwriters can't afford to be leaving money on the table -- or the gurney, as the case may be. Music-as-an-instrument-of-torture does at least seem to offer the promise of a growing market segment: a rare bright light amid the general gloom.

Candidly conceding that its ruling "may not coincide ... with notions of justice and fair play," the U.S. Court of Appeals for the Fifth Circuit yesterday decertified the shareholder class action against the banks that allegedly helped perpetrate Enron's frauds -- a case that has already produced more than $7 billion in settlements. The ruling, if it is not overturned by the full appeals court or the U.S. Supreme Court, will benefit Merrill Lynch (MER), Credit Suisse First Boston, and Barclays (BCS), which had not settled, saving each potentially billions in liability.

Though the issues are complicated, and have lots of twists and turns, the heart of the problem has actually been apparent to most lawyers watching this case from the moment it was filed in 2002. Back in 1995, in a 5-4 ruling of the U.S. Supreme Court that shocked lawyers at the time -- it ran counter to what every appeals court that had faced the question had previously assumed -- the Court found that the securities laws did not create liability for those who "aid and abet" fraud (i.e., knowingly help others to commit fraud), as opposed to those who act as "principals" in such schemes. Even as they rendered that ruling, several of the justices in the majority acknowledged that the outcome of the ruling was unjust, and they urged Congress to fix the problem by amending the law to include aiding and abetting liability. In almost every other legal arena -- including the criminal arena -- aiders and abettors are treated as every bit as responsible as principals. (Indeed, the distinction between the two is often hard to draw.)

Congress never fully fixed the problem, however. It did allow the Securities and Exchange Commission to go after aiders and abettors, but not private plaintiffs attorneys. The reason is simple: it did not trust the latter to use good judgment in doing so; rather, it anticipated -- no doubt correctly -- that allowing aiding and abetting liability would result in banks, accountants, and law firms being routinely named in nearly every shareholder class action suit filed, no matter how frivolous. (Fittingly, the Enron case is brought by Bill "Partner B" Lerach, who is king of both the wheat and the chaff when it comes to class actions. He is lead counsel in the Enron case, and yet earlier in his career, for example, his firm also brought a series of civil RICO class actions against baseball trading card manufacturers for allegedly promoting gambling among children by giving away bonus trading cards in some, but not all, packs.)

But the omission leaves fraud victims uncompensated in obviously legitimate cases, like the Enron case, and may even help encourage such frauds to the extent that aiders and abettors are emboldened by their apparent immunity. The Enron case is the ultimate example. Here, banks allegedly -- and, by now, the evidence against certain banks seems overwhelming -- knowingly helped Enron manipulate its financial statements by engaging in numerous shady deals and sham transactions. It may well be that the banks felt free to do so at least in part because they thought they were immune from aiding and abetting liability.

When these suits were filed, the banks immediately brought motions to dismiss, claiming that the complaint only accused them of being, at worst, aiders and abettors in these frauds on Enron shareholders. (The bank officials had no direct fiduciary duty to Enron shareholders, the way the Enron officials did.) U.S. District Judge Melinda Harmon found a way around the obstacle at that time, ruling that the banks were so deeply involved here that they could be considered principals. And it was on the basis of that early ruling that shareholders have recovered, to date, more than $7 billion. (Which they can keep.) But yesterday that ruling finally received, in effect, appellate scrutiny (though in a slightly different procedural context), and was overturned.

The Fifth Circuit panel majority (one judge would have decertified the class on other grounds) was quite candid about the dilemma it felt the courts face, and that it believed the Supreme Court had already resolved in a direction that is harsh toward plaintiffs: "the rule of liability must be either overinclusive or underinclusive so as to avoid what [has been] called “in terrorem settlements” resulting from the expense and difficulty of, even meritoriously, defending this kind of litigation." It's a difficult call, but I actually come down on Lerach's side on this one.

How about you?

On Friday, in a terse, one-paragraph statement that was even more opaque than the very similar one Apple (AAPL) issued in December, the audit committee at Walt Disney Co. (DIS) cleared Steve Jobs of any "intentional or deliberate acts of misconduct" in connection with the options backdating that concededly occurred at Pixar (PIXR). Disney, which acquired Pixar in 2006, shed no light on how the backdating came to pass, but cleared anyone "currently associated with Disney" of wrongdoing.

The backdating at Pixar was in some respects even more extreme than at Apple: top officers were repeatedly granted options on dates when Pixar stock was at its lowest point for the year. On the other hand, unlike the situation at Apple, none of the Pixar options went to Jobs himself. At Apple, the special committee also acknowledged that Jobs may have actually been involved in "recommending" dates for options grants -- a circumstance that, on its face, suggests involvement in either backdating or springloading. (The latter term refers to the suspect practice of executives granting options to themselves or others just before new market events are about to give the stock price a big boost, as the executives know due to inside information.)

The story line at both Apple and Pixar appears to be that Jobs, the notorious micromanager who headed both companies at the time of the backdating, did not understand the legal or accounting ramifications of backdating. Interestingly enough, virtually Jobs's only compensation at either company during this period was coming from the Apple options whose workings he so poorly understood. (During the relevant years, his salary at Apple was famously just $1 per year, while at Pixar it was about $55 per year.)

The thing that puzzles me most is this: If you don't understand the accounting ramifications of backdating, why do it? Why not just issue the options dated as of the actual date you're issuing them, and simply choose whatever strike price you think is appropriate -- even though it may not correspond to the current stock price?

Take, for example, the situation at Apple, for instance, where Jobs received a grant of 7.5 million options options that wasn't finalized until December 19, 2001, when the stock price was $21.01. The special committee found that these options were backdated to October 19, 2001, when the price had been $18.30. (Phony documents were created to reflect a board meeting on October 19 which never really occurred.) Well, if you don't understand the accounting implications of backdating, why not just issue the options as of December 19, but announce a strike price of $18.30?

The answer, of course, is that someone at the company certainly did understand that if you did that, you'd be granting "in-the-money" options, which have onerous accounting and income tax repercussions for the company. The whole point of backdating is to avoid those consequences. To do that, you pretend that the options aren't in-the-money, even though they really are. So why go through that deception if you don't understand the accounting implications of granting in-the-money options?

And if Jobs himself didn't understand the accounting implications that were driving the deception that he himself (at Apple, at least) was benefiting from and, to the extent he was "recommending" grant dates, participating in, why didn't the underlings who did understand ever try to explain to Jobs the perilous legal situation he was getting himself into? Were they too petrified to tell him something they assumed he didn't want to hear, or already knew, or both? What a guy to work for.

Can readers explain to me why a company's executives would engage in backdating when they didn't understand its accounting implications?

Thanks be to a state superior court judge, former California Attorney General Bill Lockyer's ghoulish felony prosecution of former Hewlett Packard (HPQ) Chairman Patricia Dunn and three others for their involvement in HP's telephone pretexting scandal ended yesterday -- not with a bang, but a whimper.

Judge Ray Cunningham dismissed all charges against Dunn outright, and will dismiss all charges against her three co-defendants in about six months, after they do some community service. Though each had been charged with four state-law felonies, none will ultimately stand convicted of anything. The three whose charges were not dismissed outright -- a senior HP lawyer and two outside investigators -- were required only to plead "no contest" to a single misdemeanor, though the misdemeanor will ultimately be dismissed too. (A "no contest" plea is an admission of nothing that nevertheless permits the state to treat you as if you were guilty.) To be clear, none of the four will be left with any criminal conviction stemming from this prosecution. On the other hand, each still faces potential exposure in an on-going federal criminal inquiry arising from the same events.

These were the charges that arose from HP's outlandishly overblown, but legally murky attempt to ascertain which director on its board had been leaking inside corporate information to the press, in violation of HP policy and, possibly, the director's fiduciary duties to the corporation. In the course of the inquiry, HP's outside investigators obtained private phone records of directors and journalists by pretending to be the people in question.

There had never been any doubt that Dunn had acted in good faith throughout the inquiry, having received advice from two senior HP attorneys -- then HP general counsel Ann Baskins, who was never criminally charged, and senior counsel Kevin Hunsaker, who was -- that the investigatory techniques were lawful. Nevertheless, Lockyer, who was then running for state treasurer, evidently theorized that ignorance of the law was no defense, which is, indeed, the rule for some white-collar offenses. Still, the prosecution would have been wildly inappropriate on these facts even had Dunn not been suffering from late-stage ovarian cancer, as she was.

California had no law specifically aimed at telephone "pretexting," as the practice has come to be called, until September 29, 2006, long after the events at issue had occurred. When that law finally was enacted, it was a misdemeanor. Nevertheless,Lockyer charged each defendant with four felonies under preexisting laws that he theorized could be applied to this situation, though none ever previously had been. (For an earlier post questioning the appropriateness of these charges, see here.) In light of yesterday's consensual resolution, we'll never know if any of his theories would have panned out.

The recently enacted California misdemeanor that does prohibit "pretexting" was signed into law, of course, just two days after Dunn herself testified before a Congressional subcommittee exploring HP's runaway leak probe. Bravely, Dunn had chosen to testify rather than invoke her Fifth Amendment right against self-incrimination, the way Hunsaker, Baskins, and all the others in Lockyer's gun sights had.(Dunn's attorneys at Morrison & Foerster were roundly second-guessed in the press for "letting" her do so -- mainly by attorneys who were seeking free advertising for their own criminal practices -- but the decision was, of course, ultimately Dunn's.)

Federal law on pretexting during the period in question was also murky, so the defendants await with trepidation the outcome of the ongoing federal inquiry, which appears very much alive. In a recent New Yorker article, journalist James Stewart said that pretexting could "obviously" constitute federal wire or mail fraud. When a journalist of Stewart's caliber says something like that, it may well mean that he has spoken on background to the federal prosecutors, and that they are, indeed, proceeding on that theory. Still, I disagree with Stewart and/or his sources; a mail or wire fraud on these facts would be unique, pioneering, abusive, and highly suspect as a matter of law. Under federal precedents, deceit alone does not constitute "fraud," which usually requires that the defendant intend to realize some personal pecuniary gain by depriving someone of money or property. (Though there are exceptions -- see this article on "honest services" fraud for a general discussion -- none seem to apply here.) In this case the victims were being deprived only of their privacy rights, and the intent of the defendants (though ill-considered and betraying terrible judgment) was not to enrich themselves but to catch an alleged wrongdoer. Even if the conduct should be made illegal, people should not be prosecuted retroactively for conduct that wasn't clearly designated as criminal at the time they engaged in it.

COMMENT FROM STATE TREASURER BILL LOCKYER (received March 19, 2007):

The Hewlett-Packard pretexting case certainly has culprits. Neither I nor lawyers at the Department of Justice are among them. The defendants in this case broke the laws of our state and, in the process, violated constitutionally-protected privacy rights. Combined, the prosecutors in this case possessed more than seven decades of experience. They evaluated the evidence and determined they could win a unanimous jury verdict on the charges filed.

HP officials' legal investigation of the conduct essentially consisted of consulting a lawyer in Massachusetts connected to a defendant and surfing the Web. That lack of due diligence is matched only by some pundits' lack of credibility. The Stanford professor frequently quoted as a critic of the charges didn?t even know phone companies are utilities in California, a key facet of the case.

This case was not cobbled together. It rested on a solid foundation of evidence and state laws. I had a duty to pursue justice, regardless of the status or special circumstances of any particular defendant. If I had wanted to take the politically correct action, I would have ignored the evidence and dropped the case. I'm proud I didn't place election-year comfort above duty.

Bill Lockyer
California State Treasurer

The most important broadcast decency case in a generation is due to be decided any day now by a federal appeals court in New York. I wrote an article about the case, called "Bleep Deprivation," which is available on page 53 of the current ("March 19, 2007") issue of Fortune magazine (the one with "America's Most Admired Companies" on the cover). Alternatively, you can also see it by going to page 53 of Fortune's free (and pretty neat) digital edition, which is available here.

The last time the U.S. Supreme Court reviewed the power of the FCC to punish "indecency" on broadcast TV was a generation ago, in the landmark 1978 case known as FCC v. Pacifica Foundation. The Court was then ruling on whether the FCC was empowered (consistent with First Amendment free-speech guarantees) to punish a New York radio station for having broadcast an expletive-studded George Carlin skit at two in the afternoon. (It was his 1973 "Filthy Words" skit, which was an expanded version of his 1972 monologue, "Seven Words You Can Never Say on Television." You can read the skit in question if you click on the Court's Pacifica ruling here, and scroll to the Appendix at the end of Justice Stevens's opinion.) The Court ultimately decided, by a bare 5-4 vote, that the FCC could punish the radio station, but the precise meaning of the ruling was always confusing, since no opinion won support from a majority of the justices.

Perhaps because of the narrowness and opacity of the ruling, the FCC had invoked its power to punish dirty words sparingly thereafter -- until March 2004. Then, after a series of incidents in which celebrities like Cher, Bono, and Nicole Richie used expletives on live award shows, topped off by the February 2004 Wardrobe Malfunction incident during the half-time of Super Bowl XXXVIII, the FCC adopted a tougher stance. Its crackdown prompted the case that was argued before the U.S. Court of Appeals for the Second Circuit last December -- involving the Cher and Richie incidents -- which has raised the question of whether Pacifica is even still good law. (A lot has changed since then, including social mores, the composition of the Court, and technology.)

If you do look at the story, let me know how you think the Court should rule.

You'll recall that in late December, when a special committee of independent Apple (AAPL) directors exonerated CEO Steve Jobs of any wrongdoing in connection with the options backdating that it admitted had occurred at that company, it stressed, among other things, that Jobs had not "financially benefited" from any of the backdating.

Others have already questioned the special committee's interpretation of financial benefit (see, e.g., this Washington Post article, focusing on the fact that the options were later exchanged for 5 million shares of restricted stock), and I have previously argued (here) that whether someone ultimately succeeded in realizing a financial benefit is irrelevant to the question of whether he attempted or conspired to violate various securities, tax, or accounting laws or rules.

But there's a simpler reason, still, that, I think, Jobs did "financially benefit" from some of the backdating at Apple. It stems from a very unusual fact about Jobs's two enormous options grants, which I only stumbled on recently when re-reading some of his SEC filings: large portions of each grant vested immediately.

Usually, of course, options don't start vesting for at least a year. (Under a typical vesting schedule, about a quarter of the optinons vest after a year, and the remainder vest on a pro rata monthly or annual basis over the next three or four years.) The delayed vesting is what makes them so hard to value: no one knows whether they'll be worth squat to the recipient by the time they start vesting. That's why people come up with complicated statistical estimation formulas for valuing options grants, like the Black-Scholes test.

Now I'm not criticizing Apple for making an exception from the usual vesting customs for Jobs, because Jobs had famously been receiving only $1 per year in compensation at Apple since returning to the company in 1997, so he was clearly entitled to some immediate compensation for past performance.

But the immediate vesting does seem to me to change the analysis of whether he "financially benefitted" from options backdating. (So far as I know, the committee was not using "financial benefit" as any term of art, with a specialized definition; they seem to have been using it the way it's understood at a gut-level in ordinary English.) The special committee acknowledged, remember, that the 7,500,000 options that Jobs was ostensibly granted on October 19, 2001, when Apple's price was $18.30 a share, had not really been "finalized" until December 18, 2001, when the price was $20.01 per share. (In addition, the committee found, the Apple board had not really met at all on October 19, 2001, as board minutes falsely indicated.) Accordingly, the committee determined that the accounting for these options had been improper, and it included a correction for this mistake in Apple's restatement. (The commmittee also found, in fairness, that the Apple board had first approved this grant to Jobs in August 2001, at a time when Apple's price was even lower than it was on October 19, 2001.)

Well, one-quarter of these admittedly backdated options -- representing 1,875,000 underlying shares -- vested immediately, according to a Form 4 Jobs filed with the SEC on March 20, 2003. (Similarly, fully one half of the 10 million shares ostensibly granted to him on January 12, 2000 vested immediately. But I'm ignoring those, because the the special committee concluded that that earlier grant -- notwithstanding some eyebrow-raising circumstances -- was properly accounted for.)

The value of Jobs's 1,875,000 vested options as of the day on which we now know that they were really granted are easy to calculate. On December 18, 2001, when the grant was finalized, the value of those vested options was the fair market value of the underlying shares ($39,393,750) minus the cost of exercising them at the strike price ($34,312,500). That's $5,081,250. All of that money was available to Jobs for the asking, then and there, and therefore I don't see why all of it wouldn't be considered "financial benefit."

How much of that financial benefit came from backdating? All of it. Had those options' strike price been set as of December 18, 2001, as the special committee says they should have been -- instead of as of October 19, 2001, when the phantom board meeting took place -- their intrinsic value to Jobs on that date would have been zero.

At this point in my argument, I'm sure, members of the Apple special committee would protest: there was no financial benefit, because Jobs never exercised those options. (He cancelled all his options in March 2003, when they were underwater, and received restricted stock in their place.) And I would reply: No, he did financially benefit, but he then took a calculated gamble with his gains -- the way we are all constantly taking calculated gambles with all of our assets -- which did not pay off in the short term. He chose not to exercise the options immediately, hoping for even bigger gains down the road. His gamble didn't happen to pay off, or at least hadn't as of March 2003, when he took another calculated gamble, and traded in the options for restricted stock. (My colleague Geoff Colvin analyzes here why agreeing to the trade-in was actually Jobs's biggest mistake of all.)

Say someone gives me a gift of $5,081,250. Say I sink it all into an uninsured beach house which later gets leveled by a flood surge stemming from a hurricane. Or say I put it all into the next Enron, and it soon plummets to worthlessness. Would anyone say that I had experienced no "financial benefit" from the $5,081,250 gift? I don't think so.

Apple spokesman Steve Dowling says: "The most important point is that both options grants, which were underwater, were cancelled. No options were ever exercised. . . . Following exhaustive independent investigation, the special committee found that Steve [Jobs] was not aware of any irregularities associated with [these grants]."

Of course, from previous posts on this subject, I understand by now that any reader of this blog who owns Apple shares does not care one whit whether Jobs backdated options or benefited from the practice; they just want him to stay as CEO and make them some more money. But what do people think about my reasoning? Did he experience "financial benefit" from backdating, or didn't he?

Last month, on the eve of a scheduled trial, the accounting firm of Deloitte & Touche quietly settled a malpractice suit that accused it of having approved a stock option pricing program that amounted to backdating.

The plaintiff in the case, San Jose semiconductor maker Micrel Inc., disclosed the fact of the settlement -- though none of its terms -- last week in its 10-K filing with the SEC. In an interview, Micrel general counsel Vince Tortolano declined to give any additional information about the deal, but suggested that some clues about its size might be discernable from the company's 10-Q for the first quarter of 2007, which will be filed in April or May.

"Deloitte & Touche has denied, and continues to deny, any responsibility for the claims in the lawsuit brought by Micrel," a spokesperson for the accounting firm said in an e-mail Monday. "The lawsuit has been settled to avoid the substantial expense and inconvenience of continued litigation."

I had been hoping it would go to trial, because it would have finally shed some light on how Silicon Valley accountants were analyzing options backdating issues in the late 1990s. (Many of the court filings in the case were lodged under seal, and are consequently unavailable.)

The case was also intriguing because it was one of only two I know of so far -- the other being Microsoft -- where a company is believed to have actually secured the blessing of some outside "gatekeeper" ( i.e., an accounting firm or law firm) for a "backdating" scheme. At the outset of the options backdating scandal, there was much speculation that such gatekeepers had to be at fault, given how widespread backdating was prior to passage of the Sarbanes-Oxley law in 2002, but that theory has not yet borne much fruit.

What allegedly happened at Microsoft (according to Microsoft's SEC filings and a June 16, 2006 Wall Street Journal article by Charles Forelle and James Bandler [behind a paying firewall]) and at Micrel (according to the complaint Micrel filed against Deloitte in April 2003) sound startlingly similar. Here is the chronology of the intertwined events at the two companies.

In 1992, Microsoft instituted an options pricing policy that seems to have aimed at solving a problem many tech companies were then facing -- one that only grew more intense during the dot-com bubble years. Due to extremely volatile prices, new employees often complained about receiving options with dramatically different strike prices from those of similarly situated colleagues who had joined the company just a few days earlier or later. Under Microsoft's program, new employees were granted options that, at first, conformed to the norm: they had an exercise price equal to the price of a share of Microsoft at the market close the day before. But for the next 30 days, if the stock price fell, the option price would be "reset" to the lower price. In the end, new employees would get their options at the lowest trading price for the 30-day period following their date of hire. Using a similar mechanism, existing employees regularly received additional options priced at the lowest price for the month of July. An unidentified Microsoft official told the Journal that Microsoft's accountants at Deloitte had approved the programs, evidently opining that such options grants could be considered "at-the-money" grants -- options priced at fair market value as of the day of the grant. Such options would not need to be expensed at all under the accounting rules of that time. (A Deloitte spokesperson says in an email that the firm "can't comment on advice the firm has given to a client.")

In 1996, Micrel, in San Jose, began using a nearly identical 30-day pricing program, also allegedly blessed by its lead outside accountant at Deloitte, according to Micrel's complaint.

Then, in mid-1999 Microsoft, for reasons that are not clear, decided that its program did not comport with generally accepted accounting principles after all. Since the pricing process did not really conclude until 30-days after it started, the company apparently concluded, the strike prices were effectively being backdated to the lowest price during that 30-day period. (Duh.) Accordingly, they were in-the-money options, which needed to be expensed. Microsoft then discontinued the program and took a $217 million charge against earnings.

Meanwhile, Micrel kept using its own 30-day pricing policy for two more years, until November 2001. By that time, however, according to Micrel's complaint, the lead Deloitte partner on the Micrel account had changed. The second Deloitte partner then allegedly disavowed his predecessor's opinion, and advised Micrel to discontinue the policy and make a restatement, which the company did. (The revisions added more than $50 million in expenses to its financial statements.) In 2003 Micrel sued Deloitte for malpractice, seeking reimbursement for all the costs that resulted from the first Deloitte partner's allegedly bad advice.

It's a tantalizing set of facts, isn't it? But, alas, due to the settlement, that's all we know.

We're quite used to hearing that question posed (and answered "yes") by champions of Linux and other forms of free and open source software (FOSS). But those people have an obvious axe to grind: users and developers of FOSS fear being sued for patent infringement by big corporations, like Microsoft (MSFT), that have amassed huge software patent portfolios.

Remarkably, however, those very folks with the huge software patent portfolios -- Microsoft, the Business Software Alliance, the Software and Information Industry Association, Intel (INTC), Amazon (AMZN), Yahoo (YHOO), Autodesk (ADSK) and others -- now seem to be posing that question themselves, although one has to read between the lines a little bit. I'm talking about the message that emerges from the array of briefs those companies and groups have filed in connection with the recently argued U.S. Supreme Court case Microsoft v. AT&T.

The focus of that case is a narrow one. It has to do with a provision of patent law that was enacted two decades ago in order to discourage a pretty obviously unfair practice. (See earlier posting here.) U.S. manufacturers who wanted to make here, but sell abroad, a hardware device that would infringe U.S. patents had been manufacturing all the components here and then shipping them abroad for final assembly--doing an end-run around our patent laws. A law was enacted that forbids that practice. The question posed by the Microsoft v. AT&T case, is whether that law should apply to situations where the only exported "component" in question is software written in the U.S. but copied and reinstalled in computers abroad. Interestingly enough, the near unanimous view in the software industry--judging from the briefs submitted by the parties I've listed above--is that it shouldn't. (The specific facts of the case are these: Microsoft concedes that its Windows operating system infringes a U.S. patent belonging to AT&T (T) relating to coding and decoding human speech. It is willing to pay royalties on copies of Windows sold in the U.S., but contends that it shouldn't have to pay for copies installed on computers abroad and sold there. AT&T, and the court below, say it should.)

In other words, all these parties with mighty software patent portfolios would rather, on balance, not be allowed to enforce those valuable assets abroad, so long as they could be assured that, in exchange, they also wouldn't have to worry about being sued for infringing anyone else's U.S. software patents abroad. That doesn't sound like a ringing endorsement of the U.S. patent system, at least as it relates to software patents; it sounds like the opposite. (Patent law is supposed to benefit industry by spurring innovation; yet it sounds like the software industry regards it as a net drag on in its industry.)

Emery Simon, counsel to the Business Software Alliance, assures me that none of the software companies in his group agree with my reading of their position. "All of our members believe that intellectual property rights generally, and patents specifically, are essential to promoting their continued development and distribution of innovative products and services," he writes in an email. "The problem is how the lower courts have interpreted [the statutory provision concerning exportation of components], it is not the fact that the patent law protects software." He continues: "The way the law has been interpreted is inconsistent with the objective of the law: the purpose of the law is to promote innovation, the effect of the lower court's decisions is to create disincentives to US-based innovation. So, if the current interpretation of the law stands, US tech companies will have a disincentive to do their development in the US."

What he's saying is that the law, as it's currently being applied, may drive U.S. software makers to move their operations abroad to ensure that they can sell their products abroad without running afoul of the U.S. patent system. Still, it sounds to me like the "disincentive to US-based innovation" he's complaining about is the burden of having to operate within the U.S. patent system itself. And a patent system is supposed to provide an incentive to innovate, not a disincentive.

I wonder if Simon's position and mine might be compatible in this sense: Maybe the companies with big software patent portfolios believe that software should be protected by a patent system; they just don't believe it should be covered by the patent system we happen to have in place in the United States at the moment.

Can others make sense of this situation?