(Part 2 of 5) The Fine Line Between an Alternative Basis and Simple Dicta
The question we’re asking is whether it’s appropriate to rely on legislative materials generated by Congresses who did not enact but did oversee the drafting of the legislation in question. Prof. Peter Menell thinks it is, at least where the Copyright Act of 1976 (“1976 Act”) is concerned. I’m pretty sure it’s not. The question matters because Prof. Menell’s re-write of a crucial passage of Nimmer on Copyright relies on such materials. I have criticized this re-write on several grounds, including this unexamined reliance on pre-enactment legislative materials. See the last post for further details.
Ordinarily, you can’t rely on pre-enactment legislative materials. Ordinarily, when this issue arises, Congress has drafted but failed to pass a bill, but the next Congress manages to pass substantially the same bill. This is not unusual. In these situations, the courts have been clear that it is inappropriate to rely on the previous Congress’ legislative materials when construing the statute. Normally, this doesn’t pose a problem because the later (enacting) Congress will simply adopt the previous Congress’ work—why reinvent the wheel?
But the 1976 Act was not typical. First, it might’ve been “just a bill” at one point, but it was an enormous, sweeping bill, too big and too complex to have been drafted all at once. Important insights generated early in the process might have been lost before the bill was completed. Second, it was substantially complete by 1966, but it wasn’t passed for ten years in 1976. Normally, the process of drafting and enacting a bill are close enough in time that we don’t need to distinguish between drafter’s and enactor’s intent, but here, five Congresses (with commensurate turnover) intervened between final draft and passage. Third, it was neither drafted by Congress nor wholly by lobbyists. Instead, it was drafted under the auspices of the U.S. Copyright Office (which is nominally part of Congress itself, being set beneath the Library of Congress), with the input of many industry groups (which didn’t have to fight for access, so there were little or no K-street advantages). Congress had, in a sense, outsourced the drafting of this monumental legislation to people who not only knew copyright law but understood the industries that relied on copyright law.
My view is that none of this makes a difference. A bill doesn’t become a law simply because experts worked hard at it. It becomes law because enough members of Congress were persuaded to pass it. Thus, what matters is not what the drafters thought but what those voting in the bill’s favor thought.
The Role of the Supremes
Still, as Prof. Menell points out, the U.S. Supreme Court has, at a minimum, referenced the in the 1976 Act’s older, pre-enactment legislative history, in discussing the 1976 Act, on several occasions (he names six). Since the U.S. Supreme Court is the ultimate arbiter of what a federal law means, it’s also the ultimate arbitrator of how to interpret federal statutes, including when and how to use legislative history.
The problem is that the Court has never squarely addressed the issue of whether it’s appropriate to rely on 1960’s-era legislative history of the 1976 Act. It either assumed it was appropriate, or has assumed that it wasn’t. The probable reason for this is that no one has explicitly argued against using the 1960’s-era legislative history. My theory for why that is is that legislative history is so malleable that both sides think it will help them, so neither is motivated to argue against its inclusion (and waste precious pages).
But that doesn’t stop us lawyers. It’s not unusual for the Court to touch on but not directly address a legal issue. The Court can only squarely address matters put before it, and there are only so many opportunities for that. In the absence of a definitive holdings, lawyers will engage in a kind of exegesis or tea-reading of Court opinions. It is not unusual for stray pronouncements in footnotes to attain the status of established law. Sometimes these stray pronouncements presage a real change in the law (as in the famous “footnote four” of Carolene Products), and sometimes they are just forgotten. (Fun article on Supreme Court footnotes.)
In this case, we’re looking for instances where the Court relied on—or refused to rely on—the pre-enactment legislative history in construing the 1976 Act. If the Court merely notes its existence, that doesn’t tell us anything because it’s not being used to interpret anything. Of the six Supreme Court decisions mentioned by Prof. Menell, only three actually can be said to meet this requirement, and only one supports the propriety of relying on such legislative history: Community for Creative Non-Violence v. Reid(“CCNV”).*
* In addition, I found another Supreme Court decision that Prof. Menell might have cited that arguably supports his position, Mills Music, Inc. v. Snyder (“Snyder”), on which CCNV partially relied. I’ll discuss Snyder in a separate post along with its doppelgänger, Stewart v. Abend.
Background: Works Made for Hire
CCNV is one of the best-known copyright decisions issued by the Supreme Court because it directly addresses a common and crucial question in copyright law: when someone commissions a creative work, who initially owns the copyright in the work, the commissioning party or the commissioned party? In copyright parlance: is the work a “work made for hire” or not? If it’s a “work made for hire,” then the copyright would vest automatically in the commissioning party; if not, it would vest in the artist.
The 1976 Act defined a “work made for hire” as including two distinct types of works:
- “a work prepared by an employee within the scope of his or her employment.”
- “a work specially ordered or commissioned,” but only if (a) there is a written agreement to that effect, and (b) the work falls within one of nine categories of works (such as motion pictures, maps, test questions—but not including sculptures).
In CCNV, the plaintiff non-profit organization had commissioned a sculptor to make a statue depicting homelessness in America for a Christmas pageant (i.e., there’s no room at the inn). The organization was heavily involved in conceptualizing what the sculpture would depict, would check on the progress of the work regularly, offer suggestions and state preferences, and reject proposals from the sculptor. In short, the organization controlled much of the work product at a very high level. But everything else—the execution of the plan—was controlled by the sculptor. He used his own tools, procured his own materials, worked in his own workshop, etc.
There was no written agreement. Further, “sculpture” doesn’t fall into any of the nine categories of works set forth in the second type of “work made for hire.” Thus, if the organization was going to prove that the statue was a “work made for hire,” it would have to show that the sculptor was, in effect, an employee of the organization acting in the scope of his employment. While this might sound counter-intuitive, the organization had several lower-court decisions in its favor.
More Background: The Control and Agency Theories of Employment
The organization argued that the sculptor was its employee because it had the ultimate right to control or had exercised actual control over the final product. Under this view, the fact that it imposed its own creative vision on the sculptor was sufficient to render the statute a work made for hire. For his part, the sculptor argued that an “employee” meant a formal, salaried employee—i.e., what you and I probably think of as an “employee.” Since the sculptor was an independent contractor, by this reasoning, the statue was not a work made for hire.
The Court rejected both these views. First, the court noted that, when Congress doesn’t define a term, it assumes the traditional, “common-law” definition of the term (and Congress is aware that this is how courts will interpret the term). Before our modern concepts of employment (as found in anti-discrimination laws), an employee was just a kind of agent, i.e., one who acts on behalf of another. For example, when an employee at a store sells you something, that employee acts the store’s agent for purposes of a sales relationship with you. Similarly, a company might commission an independent contractor to go overseas to purchase some raw material, giving him or her broad authority to strike the best deal he or she can. Both are “employees” for purposes of work made for hire, even though the overseas agent is clearly an independent contractor.
The Court rejected the “control” theory because that just re-stated the whole problem with commissioned works. What was a commissioned work (usually) but a work created by one person but to the approval (in terms of taste, specifications, etc.) of someone else, and anything subject to another’s approval is, in a practical sense, controlled by that person. The Court reasoned that commissioned works were already covered by the second type of “work made for hire.” If you defined “employment” to include commissioned works, then you’d be rendering that second type of “work made for hire” superfluous. What’s more, you’d be defeating the purpose of limiting that second type of “work made for hire” to nine categories of works, since the first type has no such limitation.*
* It’s useful to turn this analysis around. What if a “formal, salaried” employee was so trusted that she was given complete carte blanche to create certain works however she wanted with no real supervision, and the employee would pay her regardless? Under a control theory, the employee would retain the copyright in her works, which is highly counterintuitive.
The Court concluded:
We therefore conclude that the language and structure of § 101 of the Act do not support either the right to control the product or the actual control approaches. The structure of § 101 indicates that a work for hire can arise through one of two mutually exclusive means, one for employees and one for independent contractors, and ordinary cannons of statutory interpretation indicate that the classification of a particular hired party should be made with reference to agency law.
We’re done, right? If you think that, you haven’t read enough Supreme Court opinions…
Enter Legislative History
Having reached a definitive holding on the issue, the Court nevertheless went into an extensive analysis of the legislative history of the definition of “work made for hire.” In doing so, it relied on pre-enactment legislative materials, emphasizing that the distinction between employee and commissioned independent contractor existed as early as 1961, then was the subject of a tug-of-war between publishers (who favored a broad definition of “work for hire”) and authors (who favored a narrow definition), resulting in 1965 in a compromised brokered by the Copyright Office—and in language almost identical to that eventually passed in 1976 (the main difference being the number of categories in the second type of work made for hire). This language was specifically endorsed by one of the key Congressional committees, and much of that committee’s report was borrowed by the 1976 Congressional report.
After all of this, the Court concludes:
Thus, the legislative history of the Act is significant for several reasons. First, the enactment of the 1965 compromise with only minor modifications demonstrates that Congress intended to provide two mutually exclusive ways for works to acquire work for hire status: one for employees and the other for independent contractors. Second, the legislative history underscores the clear import of the statutory language: only enumerated categories of commissioned works may be accorded work for hire status. The hiring party’s right to control the product simply is not determinative. Indeed, importing a test based on a hiring party’s right to control, or actual control of, a product would unravel the “ `carefully worked out compromise aimed at balancing legitimate interests on both sides.’ ” [Citing the 1966 Congressional report, which in turn quoted the Register of Copyright’s Supplemental Report].
At the end of this conclusion, the Court added the following footnote 14: “Strict adherence to the language and structure of the Act is particularly appropriate where, as here, a statute is the result of a series of carefully crafted compromises.”
For all of the Court’s extensive discussion of the pre-enactment legislative history, all the Court is really able to glean from it is what it already knew: that there are two different types of “works made for hire,” and one type shouldn’t occupy the other’s territory (or else, why have two types?); and only certain types of commissioned works are eligible under the second type of “work made for hire,” something one can gather from simply reading the statutory language.
“Your Argument Is Invalid”: It’s Not Done at the Supreme Court
Why, then, even bother with the legislative history? Probably because the organization and its allies made a big deal of it. They argued that, by 1966, there were a number of appellate-level decisions embracing the “control theory,” and Congress must have known about them and, by its silence, incorporated them into the definition of “work made for hire.” The Court responded by making two points. First, this interpretation doesn’t square with the language (or “structure”) of the definition. Second, the definition was already in its final form by 1965, before the first of these “control theory” decisions were handed down: “Congress certainly could not have ‘jettisoned’ a line of cases that had not yet been decided.”
Except for this last point, nothing the Court finds in the legislative history has any independent weight. It just confirms what the Court understood in the statutory language. That makes sense since language does a pretty good (if imperfect) job of conveying intent (but that’s actually an argument against over-reliance on legislative history—if the best evidence of legislative intent is the actual statutory language, then why look beyond it?). The Court recognized this by the way it structured its argument. By the time it discussed the legislative history, it had already reached a definitive holding about how to interpret “employee” and “employment.” The Court could have stopped there, but it must have felt obligated to address the organization’s arguments in a manner more satisfying (and, candidly, more respectful) than just saying that its argument is invalid.
That last point—that it matters when the statutory language had achieved stability—has independent weight because it doesn’t just parallel what you can read in the statutory text. You can’t tell from the statute when the definition of “work made for hire” had achieved (more or less) final form. But even then, the argument is secondary to arguments based on the statutory text.
That footnote 14 is certainly odd if you think about it. It basically says the more carefully balanced the compromise that resulted in into a statutory provision, the more carefully we should read the statutory language (well, the stricter we should adhere to the language and structure of the provision, which I think means the same time). While that makes some intuitive sense, it implies there are times when we need not adhere strictly to statutory text. Such as, when? Must we look behind every statutory provision to see whether it was the result of compromise, and if so, how carefully balanced that compromise is, and then re-read the statutory text in light of that balance? That defeats the purpose of having statutory text in the first place: to explain to the citizenry what the law is.
Conclusion: CCNV Supports the Use of 1960’s-Era Legislative Materials, but Not that Much
If the statutory text is sufficiently clear, then, under Supreme Court precedent*, there is no need to “look behind” it and review its legislative history. If so, strictly speaking, the Court never should have reached the legislative history—not even to see if the statutory text was the result of compromise. I strongly suspect (the briefs are not readily available) that both sides argued in terms of legislative history: that the organization brought up legislative history because it knew its textual arguments were weak; and that the artist saw that he could turn the legislative history raised by the organization to his own advantage. Neither side, I am guessing, argued against the use of statutory history. And if neither side argued against it, the Court (like any court) was going to analyze the arguments in the terms presented by the parties.
* Koons Buick Pontiac GMC, Inc. v. Nigh, 543 U.S. 50, 60 (2004); Robinson v. Shell Oil Co., 519 U.S. 337, 340–42 (1997); Conn. Nat’l Bank v. Germain, 503 U.S. 249, 253–54 (1992); United Sav. Ass’n of Tex. v. Timbers of Inwood Forest Assocs., Ltd., 484 U.S. 365, 371 (1988); see also Comm’r v. Ewing, 439 F.3d 1009, 1013 (9th Cir. 2006) (court may not use the legislative materials to create an ambiguity in the statutory language); United States v. Wildes, 120 F.3d 468, 470 (4th Cir. 1997) (same).
On balance, CCNV’s treatment of the 1960’s-era legislative history of the 1976 Act is a citable point in favor of the propriety of using such materials to construe the 1976 Act. But it’s not as supportive as Prof. Menell assumes. The Court’s reasoning based on the 1960’s-era legislative history is virtually dicta, since it does nothing more than mirror a conclusion the Court reached without the legislative history.
Next time, we’ll look at the Supreme Court opinion most dismissive of the propriety of using pre-enactment legislative history of the 1976 Act, the recent decision, Kirstaeng v. John Wiley & Sons, Inc.
Thanks for reading!