This is an utterly content-free rant in which I express my anger at recent Internet Explorer preview releases requiring installation of what is effectively an entirely new operating system. I would like to know how new IE behaves on various testcases. But I don’t want to potentially hose my primary functioning Windows system to do it, especially if I then lose access to a working IE9 installation. And I am really not interested in wasting a bunch of time to spin up a virtual machine just so I can waste a bunch more time to “upgrade” it to test a new version of IE.
Here’s a novel concept: what about shipping a browser that doesn’t have to insert itself deep into operating system guts? Maybe you could even install and uninstall it distinct from the OS. But that’s crazytalk, nobody would ever do that, right?
So, yeah, whatever the latest IE10 does, meh. Someone who cares can be the sacrificial lamb and find that out, if it actually matters.
(“Content-free rant”, indeed. Future posts will return to substantive form.)
After much line-standing, it was now time to hear actual argument.
For a number of reasons, I’m not going to go into much detail. First, I was pretty sleep-deprived. While I did a reasonable job of concentrating on the arguments, my mind sometimes just wasn’t capable of keeping up with the discussion out of sheer exhaustion. This also means my memory of the arguments is spotty and may have forgotten particular interesting exchanges. Second, while I read petitioner/respondent briefs for both cases, quickly skimmed the United States’s briefs in both, and read a couple amicus briefs of particular interest in Microsoft v. i4i, I’m far from expert in either area of law, so anything I say isn’t going to be the best-informed commentary. Third, if you want full detail, you can always read transcripts and listen to audio (simultaneously, even, thanks to The Oyez Project) from the oral arguments for Tapia v. United States and for Microsoft v. i4i. (Count your blessings: until Chief Justice Roberts joined the Court in 2005, transcripts and audio weren’t released until the end of the term, around the end of June. The Court then started releasing transcripts shortly after argument. And until this year, with very rare exceptions, they didn’t release oral argument audio until the end of the term, whereas now it’s released at the end of the week the argument occurs.)
For the most part, then, I’ll limit discussion to the impressions that stuck with me. Due to lack of time I haven’t gone over the argument transcripts, so this is all raw recollection diluted by a week’s delay in scribing.
The justices sit in high-backed chairs that appear very solemn, rigid, and somber. They are that — when nobody’s sitting in them and testing them. It turns out the chairs recline quite considerably, which diminishes their impressiveness a bit. Justice Thomas and Justice Scalia sit next to each other in this iteration of the Court (the chief justice sits in the center, and the remaining justices alternate sides in order of seniority, so they sat on opposite sides before Justice Kagan joined the Court), and they had a tendency to lean back so far in the chairs that the sense of dignity the chairs conveyed was rather disrupted. Not that it really makes any difference, of course: they are who they are. Still, it was kind of funny to see that the chairs’ veneer of gravitas wasn’t as deep as the angle they reclined.
Justice Thomas is frequently noted as remaining unusually silent on the bench. At the moment he hasn’t spoken in oral argument in just over five years — to counsel making argument, that is. From time to time he and Scalia would turn their chairs and confer with each other, as a lawyer for one side or the other continued his argument, presumably discussing the case being argued. His involvement’s certainly not passive even if that involvement includes none with counsel making an argument. This is about what I’d read various places before going to argument.
Justice Ginsburg is definitely the most frail-looking justice. Even taking into account her physical position, she seemed to look downward at counsel more than appeared to be necessary, and she seemed to hunch over the presumed papers in front of her when asking questions. The other justices had a physical presence of sorts which Justice Ginsburg seemed to lack. (Again, not that it makes any real difference, she being who she is.)
Some people question the importance of oral argument, suggesting that the written briefs settle the issue, the justices have already made up their minds, and so on. Among the usual arguments against this is that the justices use argument as a sounding board to figure out what their fellow justices think and to in a sense passively-aggressively make their cases to each other. If that happened here, I didn’t have the Court-reading experience to see it, or perhaps these cases were simply not clear-cut enough for it to happen in any obvious way.
Tapia v. United States
Justice Sotomayor asked a number of questions, at various points in the argument, which essentially took the point of view of a district court judge tasked with sentencing. Her questions seemed to suggest that she thought judges should have sentencing discretion to consider rehabilitation programs for the defendant’s own good. She was a trial judge at one time, as I recall, and I found it interesting that, for the first time I can recall reading, (dare I say it?) empathy due to her past experience as a trial judge seemed to strongly affect her approach to a case. I also found it interesting because from what I can remember, historically her position has been generally pro-defendant, and (facially) this sympathy for the judge would run counter to that.
Justice Scalia, well known as an advocate of textualism, repeatedly questioned about how a judge could “recognizing that imprisonment is not an appropriate means of promoting correction and rehabilitation” while lengthening a sentence to rehabilitate. As I recall he drew a laugh once for suggesting a judge could “recognize”…and then say he was going to give a longer sentence anyway. He seemed to find the statutory text pretty clear.
I have no strong memories of the questions and answers of the other justices (save Justice Thomas, who asked no questions), although I’m sure on reading the oral argument transcript more memories would spring to mind.
Microsoft v. i4i
The very first thing I noticed about this case was that, in the very brief recess between this case and Tapia, Chief Justice Roberts had disappeared. This was expected: he’d recused himself because, as I recall, his family owns Microsoft stock. At the same time his disappearance was so abrupt that I didn’t even see him leave, despite making half an effort to watch and see it happen (because I was wondering if the other justices would rearrange their seating around Justice Scalia, who acted as chief justice for the argument).
I found this argument more technical and peppered with citations than the previous one, which made it harder to follow. The justices spent some time talking about RCA v. Radio Engineering Laboratories, Inc., the opinion in which Justice Cardozo made perhaps the strongest statements supporting the clear and convincing standard of proof. It was more time than I’d have expected, given the other opinions that could have been discussed as well.
At one point one counsel cited an opinion made by then-Judge Alito when he had been judge in a lower (circuit, I think?) court. I’m not certain, but I suspect it was the same side that had also cited an opinion of his in their brief. I had to read briefs containing bald assertions that previous cases cut in exactly opposite directions, and skeptic that I was at the time, I found this good for a laugh when reading that brief. Justice Alito seemed to think similarly, as he immediately made a self-deprecating comment about how he wasn’t nearly as smart at the time, or something to that effect. However, I don’t remember thinking he was seriously hinting which way he was leaning in doing so.
When oral argument in Microsoft finished (Microsoft’s counsel still had rebuttal time, as I noticed because the white light on the podium had turned on [indicating five minutes remaining] but the red light [indicating no time] had not), my mind felt overwhelmed with what I’d seen and heard. I knew I’d followed less than half of what had been said, yet at the same time I knew I’d basically achieved my goal: to know enough to be dangerous while hearing the arguments.
Alejandra Tapia and a friend were passing through customs while driving from Mexico to the United States. An immigration official noticed they were acting nervously and took a closer look at their car, discovering two illegal immigrants that Tapia was attempting to smuggle into the United States. She was arrested for this but was soon released, subject to court restrictions, while proceedings continued. When she failed to appear in court for a hearing, a warrant was issued for her arrest. After later apprehension Tapia was convicted on counts related to her smuggling of illegal immigrants and for jumping bail.
The judge responsible for sentencing her thought she would benefit from a particular drug treatment program, offered only in certain prisons. In order to get her into that program, he recommended Tapia be sent to a particular prison, and he increased the length of her sentence explicitly to give her the opportunity to take part in that program. In doing so he potentially fell afoul of this statutory language:
The court, in determining whether to impose a term of imprisonment, and, if a term of imprisonment is to be imposed, in determining the length of the term, shall consider the factors set forth in section 3553(a) to the extent that they are applicable, recognizing that imprisonment is not an appropriate means of promoting correction and rehabilitation.
Before enactment of the law adding this language, the “rehabilitative ideal” as implemented by U.S. law held that isolation from society and prison routine would rehabilitate criminals. Parole boards determined when prisoners were ultimately released based on subjective assessments of the progress of inmates. The Sentencing Reform Act changed the system to be more objective, while still attempting to leave judges discretion to determine individualized sentences.
The sticky wicket is in the phrase, “recognizing that imprisonment is not an appropriate means of promoting correction and rehabilitation”. Does “recognizing” this mean that when a district court sets a length and term of imprisonment, he can’t consider rehabilitation programs in determining the sentence length? If he can’t consider them, then perhaps Tapia’s sentence was wrongly lengthened. On the other hand, section 3553(a) explicitly says, “The court…shall consider…the need for the sentence imposed…to provide the defendant with needed educational or vocational training, medical care, or other correctional treatment in the most effective manner”, which arguably includes drug treatment programs like the one recommended for Tapia here. How should the “recognizing” clause be understood?
Further wrinkles in this particular instance include that Tapia wasn’t sent to the recommended prison (because doing so would place her in too close proximity to another inmate deemed likely to be a bad influence) and that she wasn’t entered into the recommended drug treatment program (as I understand it, because she refused to be enrolled). So the judge in question (more generally, any judge, as I understand it) couldn’t actually have required her to enroll in a treatment program as part of her sentence.
Laws and decisions to consider
This is mostly a case of statutory interpretation of existing laws, so there’s not much in the way of prior Supreme Court decisions to consider. It mostly comes down to how you read the text of the relevant statutes. One statute beyond those above that I find particularly interesting is 28 U.S.C. § 994(k), which states:
The Commission shall insure that the guidelines reflect the inappropriateness of imposing a sentence to a term of imprisonment for the purpose of rehabilitating the defendant or providing the defendant with needed educational or vocational training, medical care, or other correctional treatment.
The guidelines in question specify sentencing ranges from which judges generally can’t depart. Interestingly, this section seems pretty clear that sentencing to a term of imprisonment for the purpose of providing the defendant with access to “correctional treatment” (like a drug treatment program) is not acceptable. If §3582’s text is ambiguous, perhaps this text might clarify it.
Microsoft v. i4i Limited Partnership
In the 1990s i4i received a patent on a method of storing a document whose contents are structured, say, by formatting, nesting, and so on. Broadly speaking, the patent proposes storing document contents separate from document structure to facilitate easier manipulation. (The exact trick is pretty much this: store all the textual contents of the document somewhere, then store a list of pointers into that text with associated semantic meanings, like begin-element or end-element, separately.) The idea apparently works well when used in editors, and Microsoft once (no more, due to this litigation) used it in the custom XML editor built into Word. i4i sued them for doing so, and while Microsoft claimed the patent was invalid, the jury didn’t buy it. Microsoft was required to pay i4i $200 million for infringing their patent, then $40 million for doing so willfully. (Apparently Word developers had seen a presentation from i4i about software practicing the patent, asked questions about how it worked, received marketing material from i4i that mentioned the patent by number, and discussed i4i’s marketing material in email.) With interest the total’s now around $300 million.
Microsoft pled their case up through the courts, losing at each level. Finally they appealed to the Supreme Court — but because Microsoft’s particular travails aren’t necessarily interesting to the Supreme Court as it can only selectively right wrongs (it reviews around 1% of ten thousand or so appeals every year), Microsoft tried an interesting tack: they challenged the standard of proof required to demonstrate invalidity of the patent. This got the Supreme Court to bite and hear their appeal.
The district court told the jury that Microsoft had to prove invalidity by clear and convincing evidence, a phrase not found in statutory text but rather inferred from 35 U.S.C. § 282. Microsoft argues that when the standard of proof is unspecified, the default (with rare exceptions not relevant here) is a preponderance of the evidence. Clear and convincing evidence basically means the claim is “highly probable”; a preponderance of the evidence basically means the claim is “more likely than not”, or “50% plus one”. It could be much easier for Microsoft to prove invalidity if they only had to get a jury to say that the patent in question was more likely than not (by even the smallest fraction) to be invalid.
(What was Microsoft’s evidence of invalidity, you ask? Here it turns into a mess. i4i sold software called S4 over a year before applying for the patent, and S4 might have implemented the patent in question. If S4 implemented the patent, then an on-sale bar in 35 U.S.C. § 102(b) would invalidate the patent. Microsoft and i4i disagreed about whether S4 practiced the patent. [S4's source code, and presumably S4 itself, was no longer available to be inspected, as S4 had been obsolete before litigation commenced.] Microsoft said S4 implemented the patent, citing the software’s manual, testimony from a former i4i employee, and a letter i4i’s founder wrote to investors indicating that S4 implemented the idea he later patented. i4i’s founders said it didn’t, and one stated in court that the musings in the letter to investors were “an exaggeration, and as I said, it could be said to be a lie” and that the idea was conceived after S4 went on sale. Without S4’s source code Microsoft couldn’t prove invalidity by “clear and convincing evidence”. But it’s quite possible Microsoft could prove it was “more likely than not” that the on-sale bar applied, and that the patent was invalid.
Microsoft also attempted to claim invalidity of the patent by dint of its being obvious through combination of a couple other patents, one of which the examiner who granted i4i’s patent hadn’t seen. Should a patent be “presumed valid” in the language of the statute, and in the lower court’s understanding thus requiring clear and convincing evidence to prove invalid, if the evidence for invalidity wasn’t considered when the patent was granted? Here too could be another avenue of attack under a favorable ruling.)
Laws and decisions to consider
This case, as like Tapia, is mostly a matter of statutory interpretation. But beyond the clauses of patent law at issue there are also a few Supreme Court decisions of relevance.
First, in 1934 in RCA v. Radio Engineering Laboratories, Inc. Justice Cardozo variously described the burden to overcome a patent’s presumption of validity as “clear and cogent evidence”, that the challenger “bears a heavy burden of persuasion, and fails unless his evidence has more than a dubious preponderance”, and that “the countervailing evidence is clear and satisfactory”. (Microsoft distinguishes this language as referring to particulars of that case, specifically that the evidence being tried there had been tried once by one party and found wanting, then in that case was being tried again by an entirely separate party — so “clear and convincing” makes sense only in that rare situation.) Justice Cardozo’s opinion, and its language, precede the statutory text used to justify the clear-and-convincing standard: does the text merely codify that precedent? (Of course, Microsoft and i4i furiously disagree about what the precedent says, and about how courts interpreted that standard.) This case cuts for i4i.
Second we have Grogan v. Garner and Herman & MacLean v. Huddleston. In Grogan the former Justice Stevens, analyzing a case where the statutory text was silent about the standard of proof, determined the standard to be a preponderance of the evidence. In Justice Stevens’s words, “This silence is inconsistent with the view that Congress intended to require a special, heightened standard of proof.” In Huddleston Justice Marshall determined the standard of proof in a similar situation to be a preponderance of the evidence. These cases cut for Microsoft.
Third we have KSR v. Teleflex, a 2007 Supreme Court patent case. In passing in that case, Justice Kennedy wrote in dicta for a unanimous Court that “the rationale underlying the presumption—that the PTO, in its expertise, has approved the claim—seems much diminished” in a case where the PTO hadn’t considered evidence being used to challenge a patent. So perhaps the presumption of validity that Microsoft was challenging shouldn’t apply (or should apply with lesser force) because the PTO didn’t know about the software Microsoft claims implemented the patent. This would cut for Microsoft.
Past this you have the usual arguments for and against patents: Microsoft saying that granting patents too easily is dangerous (an interesting position for them, perhaps), and i4i saying that the patent system stimulates innovation and that invalidating patents hurts that. (Each side’s argument pretty much completely ignoring the patent system’s inherent balancing act between incenting progress and retarding it.) How much weight you want to give this depends to an extent on what sort of partisan you are.
(I speak on behalf of Adam Smith and free market economic ideas alone in this post.)
In a recent blog post, Gen Kanai, in discussing the EU browser ballot, at one point opines that South Korea’s monopoly browser market is a consequence of a free market. (I derive this from the first sentence of his second paragraph, in which I interpret, “To those critics of the browser ballot [supporting a free market through the invisible hand], I would present to you the example of South Korea”, as saying South Korea’s situation is a negative consequence of the invisible hand. It seems to me that either this interpretation is correct, or presenting the example would be a non sequitur.) Specifically, by his words he deems South Korea’s situation a consequence of “Adam Smith’s invisible hand” (link in original). This doesn’t seem to me to be the case, and I would like to correct a few mistaken assumptions implicit behind his assertion.
Did Adam Smith’s invisible hand approve of monopolies?
Through specific invocation of Adam Smith it is implied that, in his support of a free market, his acceptance of laissez-faire economics also means he would have condoned monopolies. This is not the case. Adam Smith was certainly no fan of monopolies, and his magnum opus, The Wealth of Nations, repeatedly noted the bad outcomes engendered by monopoly.
Europe, however, has hitherto derived much less advantage from its commerce with the East Indies than from that with America. The Portuguese monopolized the East India trade to themselves for about a century, and it was only indirectly and through them that the other nations of Europe could either send out or receive any goods from that country.
Smith goes on in subsequent sentences to explain that the exclusion of other countries, and the widespread use of nationally-chartered companies such as the East India Company to only permit exclusive access to the East Indies, made trade with America of greater overall importance than trade with East India — this notwithstanding the greater sophistication of potential trading partners in the East Indies, capable of supplying more value than “savages and barbarians” of the New World. This is hardly acceptance of monopoly. One might also consider this later statement impugning monopolies:
Country gentlemen and farmers are, to their great honour, of all people, the least subject to the wretched spirit of monopoly.
True, Smith permitted the possibility of monopoly in the case of two narrow exceptions: with respect to industries where a national defense interest might come into play (see Book IV, Chapter 2), and with respect to an industry resulting from “a dangerous and expensive experiment” such as autonomous colonization and fortification (see Book V, Chapter 1). These exceptions, however, certainly do not swallow the rule. It is simply inaccurate to say that Adam Smith’s support of a free market implies support of monopolies. (In any case, while Adam Smith’s The Wealth of Nations may have been a seminal work of free-market theory, it is not its highest point, its ne plus ultra, but merely its starting point.)
(With respect to the particular situation of South Korea and encryption, it’s plausible a national security interest was an initial factor in the development and use of a proprietary algorithm. Such an interest seems highly ephemeral given the development of widely used standards a short time later, however, so Smith’s acceptance of this particular monopoly on that basis would have been fleeting at best.)
Is the browser market in South Korea the consequence of a free market?
I freely admit that I am mostly uninformed on this topic; my knowledge comes primarily from Gen Kanai’s first post on the cost of monoculture in South Korea’s browser market. If I have misconstrued the narrative given in that post, I welcome correction.
Quoting from that post:
The history goes back to 1998, when the 128 bit SSL protocol was still not finalized (it was finalized by the IETF as RFC 2246 in Jan. ‘99.) South Korean legislation did not allow 40 bit encryption for online transactions (and Bill Clinton did not allow for the export of 128 bit encryption until December 1999) and the demand for 128 bit encryption was so great that the South Korean government funded (via the Korean Information Security Agency) a block cipher called SEED. SEED is, of course, used nowhere else except South Korea, because every other nation waited for the 128 bit SSL protocol to be finalized (and exported from the US) and have standardized on that.
In the early years of SEED, users downloaded the SEED plugin to their IE or Netscape browsers, either an Active X control or a NSplugin, which was then tied to a certificate issued by a Korean government certificate authority. Can you see where this is going?) When Netscape lost the browser war, the NSplugin fell out of use and for years, S. Korean users have only had an Active X control with the SEED cipher to do their online banking or commerce or government.
SEED, then, was the direct result of government funding of encryption and protocol research. The two (later one) SEED plugins used to implement client-side support for the protocol and encryption were part of a package provided exclusively by the government, as Gen Kanai says in a comment on that post:
Each Korean citizen is issued a nation ID number. This is embedded into the certificate issued by the Korean CA. Thus non-Koreans in Korea (such as US military in Korea) cannot make secure transactions like online banking or online commerce. The ‘package’ (including SEED, the national ID, and the Active-X cert.) that the CA’s distribute is Active-X based, and thus only works in Windows and IE.
Assuming the above is true, I’m not sure how the free market was responsible for the establishment of Microsoft’s monopoly on being able to provide browsers supporting South Korean encryption. The South Korean government brought about the single-protocol monopoly there through its funding of SEED and its ubiquitous use in governmental settings. It alone provided the packages implementing SEED in browsers. The result may be a Microsoft monopoly (if one partially due to a global decline in Netscape market share), but it seems one sanctioned by the Korean government. As Gen Kanai later notes,
Korea will only get beyond this problem by 1) applying Korean laws on open standards to the certificate authorities, 2) reassigning new certificates which work with open web standards to all Koreans, 3) reprogramming all Korean websites to support 128 bit SSL which will allow for a heterogeneous marketplace of operating systems and web browsers.
The third component of the solution requires actions by private entities, but the first and second components require governmental action. Indeed, they are direct steps to overcome previous governmental missteps, and the third component arguably doesn’t make sense except in anticipation of the first two occurring imminently. In what way did the invisible hand, or free markets, bring about South Korea’s encryption monoculture and monopoly? True, Netscape’s declining market share turned a duopoly into a monopoly — but the Korean government chose not to offer plugins for other browsers after that point (and indeed had not offered any for Internet Explorer on Windows Vista, either). Did the problem truly lie with the free market, or did it lie with the Korean government? Whose action brought about the monopoly? Whose inaction sustains it? Whose action is necessary to move toward proper competition?
It also seems worth noting that, later in Gen Kanai’s recent post, he mentions that until recently South Korea required Korean-specific hardware in smartphones. Further, South Korean government guidelines and regulations still require SEED encryption and signing for financial transactions, perpetuating SEED’s monopoly in phone browsers. These governmental restrictions on the cell phone market simply further show how governmental interference in South Korea has impacted Koreans for the worse.
A return to Adam Smith
That the Korean monopoly is the result of government action is particularly important when considering what views Adam Smith might have had of it. While Smith disdained monopolies, he particularly disdained government-created monopolies such as the East India monopolies noted previously. He writes:
The legislature…ought…to be particularly careful neither to establish any new monopolies…nor to extend further those which are already established.
It seems clear that Adam Smith would have been doubly suspicious of government-created monopolies of the sort present in South Korea, precisely because they are government-created.
A brief note on choice and browsers
None of this is to say that choice, or browser choice in particular, is bad — it’s good, precisely for reasons demonstrated by South Korea! However, South Korea’s experience doesn’t really seem to demonstrate a free market-induced shortcoming which might add support for a non-free-market action like a browser ballot. Neither the legitimacy nor the illegitimacy of a browser ballot is increased by reference to the history of encryption protocols and browsers in South Korea, even if South Korea emphatically demonstrates the ills of monopolies.