LIPEZ, Circuit Judge.
"The Game of Life" is a classic family board game, introduced in 1960 by the Milton Bradley Company to great success. This case involves a long-running dispute between Rueben Klamer, a toy developer who came up with the initial concept of the game, and Bill Markham, a game designer whom Klamer approached to design and create the actual game prototype. Eventually, their dispute (which now involves various assignees, heirs, and successors-in-interest) reduced to one primary issue: whether the game qualified as a "work for hire" under the Copyright Act of 1909. If it did, Markham's successors-in-interest would not possess the termination rights that would allow them to reassert control over the copyright in the game. After considering the evidence produced at a bench trial, the district court concluded that the game was, indeed, such a work. Plaintiff-appellants, who all trace their interest in the game to Markham, challenge that determination. We affirm.
We begin with a summary of the facts, as found by the district court. In 1959, Bill Markham, an experienced game designer and the head of a California-based product development company, was approached by Rueben Klamer, a toy developer with extensive industry contacts. Klamer had just visited Milton Bradley's Massachusetts headquarters, where he had been asked to develop an idea for a product that would commemorate the company's 1960 centennial. While searching for inspiration in the company's archive, he discovered a copy of the company's first board game: "The Checkered Game of Life," created by Milton Bradley himself in 1860. The original game was intended to instill its youthful players with lessons about vice and virtue. Klamer saw potential in an updated version, modified to reflect contemporary American society and values. On the trip back to California, Klamer developed the concept, even scribbling some thoughts on the flight home. Klamer was more of an ideas person, though, and he needed help developing the concept and creating a working prototype that could be pitched to Milton Bradley. Klamer chose Markham's firm partly because of two talented artists who worked there: Grace Chambers and Leonard Israel.
Markham and his team started work on the project in the summer of 1959. To ensure that a product launch coincided with Milton Bradley's 1960 centennial, they rushed to produce a prototype in just a few weeks. Markham and Klamer together contributed key features of the game: play would advance along a track winding through a three-dimensional game board, with a spinner determining how far players would move on each turn (thereby progressing through various "life milestones"). Klamer visited Markham's firm once or twice per week to offer feedback on the development of the physical game board and the box cover. Chambers built most of the prototype board. She constructed houses, mountains, and the elevated track out of balsa wood, cardboard, and paper. Israel focused on the art for the prototype's box cover. He produced various sketches, Markham and Klamer chose the one they liked best, and Chambers integrated it into a box cover. As the game took shape, Markham, Klamer, Chambers, and Israel would all play the prototype together, suggesting (and vetoing) various rules and refinements. Sue Markham, Bill's wife and a copywriter by trade, memorialized the agreed-upon changes in what became the prototype's rulebook.
After approximately six weeks, the prototype was ready. At a meeting at Chasen's (a famous Hollywood restaurant), Klamer and Markham pitched it to a group of Milton Bradley executives. Also present was an associate of Klamer's, Art Linkletter, a well-known radio and television personality. Klamer and Linkletter were co-founders of a company called Link Research Corporation, which developed products and used Linkletter's celebrity to promote them. Part of the pitch was that Linkletter could help market the game. The pitch worked. The Milton Bradley executives liked the game and thought that it had commercial potential.
The parties subsequently entered into two agreements regarding rights to the game. The first was a license agreement between Link Research and Milton Bradley. It gave Milton Bradley the exclusive right to make and sell the game and noted that Link Research "ha[d] had ... [the game] designed and constructed." The license agreement also gave Milton Bradley the right to use Linkletter's name and image in promoting the game. In exchange, Link Research would receive a six percent royalty on sales, including a $5,000 non-refundable advance. The second was an assignment agreement between Link Research and Markham. Stating that Markham had "invented, designed[,] and developed [the] game," it assigned "all of [Markham's] right, title[,] and interest in and to the Game to LINK." In exchange, it gave Markham thirty percent of Link Research's six percent royalty, including a $773.05 non-refundable advance. It also noted that Markham would be paid $2,423.16 to cover the costs of producing the prototype. In fact, Klamer had agreed at the beginning of the project to cover Markham's costs, and Markham had already billed Link Research for his expenses (including the salaries of Chambers and Israel and the cost of the materials used to create the prototype). Klamer ultimately paid Markham's bill from the $5,000 Milton Bradley advance.
Milton Bradley, meanwhile, began refining the prototype and made some design changes, often with input from Markham and Klamer. It ultimately published the game in early 1960. Milton Bradley applied to register copyrights in the game board and rules later that year, identifying itself as the author of both. Separately, Link Research applied for copyright registration of the game's box, and likewise identified Milton Bradley as the author. The game was a hit, and even today remains a money-maker for Hasbro, which acquired Milton Bradley (and rights to the game) in the 1980s.
In the decades following publication, however, Markham and Klamer clashed (in and out of court) over who deserved credit for creating the game. Generally speaking, Markham felt that he was not given proper public recognition for his role, and that his share of the royalties under the assignment agreement was unfairly low. Markham passed away in 1993.
This litigation is the latest chapter in the dispute over the origins of the game. Markham's successors-in-interest sued Klamer, the heirs of Art Linkletter, and Hasbro, seeking (among other things) a judicial declaration that they possess "termination rights" under the 1976 Copyright Act. Such rights give the authors of works the power to terminate the grant of a copyright after a certain period of time, see 17 U.S.C. §§ 203, 304(c), and 304(d),1 thereby permitting them to extricate themselves from "ill-advised" grants made before the "true value" of their work was apparent. Mills Music, Inc. v. Snyder, 469 U.S. 153, 172-73, 105 S.Ct. 638, 83 L.Ed.2d 556 (1985). With termination rights, Markham's successors-in-interest would be able to cancel the original assignment agreement and presumably negotiate a more lucrative royalty deal. There is, however, a crucial qualifier. As all parties agree, termination rights do not extend to "work[s] made for hire." 17 U.S.C. § 304(c); see also 3 Melville B. Nimmer & David Nimmer, Nimmer on Copyright § 11.02[A] (2020) (noting that the law "disallow[s] all works for hire from termination"). Accordingly, whether the game qualified as a work for hire became the focal point of the case.
After a bench trial (which included testimony from Klamer, Chambers, and Israel), the district court concluded that the game was a work for hire under the so-called "instance and expense" test. Specifically, the court found that Klamer "provided the instance for and b[ore] the expense of the prototype's invention." As a result, according to the court, Markham's successors-in-interest lacked termination rights under the 1976 Copyright Act. They now challenge that conclusion on appeal, arguing that the district court erred in using the instance and expense test, and, even under that test, reached the wrong conclusion. They also challenge the court's failure to strike one of the defendants' discovery responses.
A. Standard of review
When reviewing a district court's judgment following a bench trial, we defer to the court's findings of fact (unless clearly erroneous), but not to its legal conclusions (which we consider de novo). See Rojas-Buscaglia v. Taburno-Vasarhelyi, 897 F.3d 15, 23 (1st Cir. 2018). A more flexible standard governs so-called mixed questions of fact and law. See In re IDC Clambakes, Inc., 727 F.3d 58, 64 (1st Cir. 2013) ("The more fact intensive the question, the more deferential the level of review (though never more deferential than the `clear error' standard); the more law intensive the question, the less deferential the level of review.").
B. What work-for-hire test applies?
1. Doctrinal background
American copyright law has long recognized that a work created by an employee belongs to the employer, who is then viewed as the author and copyright holder. See Bleistein v. Donaldson Lithographing Co., 188 U.S. 239, 248, 23 S.Ct. 298, 47 L.Ed. 460 (1903). This judge-made doctrine was "later codified in the Copyright Act of 1909." Forward v. Thorogood, 985 F.2d 604, 606 (1st Cir. 1993). However, the 1909 Act did not provide much detail. It indicated that "[t]he word `author' shall include an employer in the case of works made for hire," 17 U.S.C. § 26 (1976 ed.) (1909 Act), but did not define "employer" or "works made for hire." As a result, "the task of shaping these terms fell to the courts." Cmty. for Creative Non-Violence v. Reid, 490 U.S. 730, 744, 109 S.Ct. 2166, 104 L.Ed.2d 811 (1989).
Initially, courts limited the doctrine to "the traditional employer—employee relationship," that is, to "a work created by an employee acting within the scope of employment." Forward, 985 F.2d at 606. Later, however, courts extended the doctrine "to include commissioned works created by independent contractors." Id. In these situations, courts would "treat the contractor as an employee and creat[e] a presumption of copyright ownership in the commissioning party at whose `instance and expense' the work was done." Id.; see also 1 Nimmer on Copyright § 5.03[B][a][i] (noting that, under the 1909 Act, "the courts expanded the definition of `employer' to include a hiring party who had the right to control or supervise the artist's work"). In practice, this test often favors the hiring party. See Roger E. Schechter & John R. Thomas, Principles of Copyright Law § 5.2.1 (1st ed. 2010) (noting that, "[e]ven in situations very far removed from the typical employer-employee case," the test "was often satisfied because the hiring party was the one who was the `motivating factor' for the project and who had at least a theoretical `right to supervise' the work").
In the Copyright Act of 1976, Congress introduced a more explicit, two-part framework that applied to works created on or after January 1, 1978 (the effective date of the Act). 17 U.S.C. § 101; Forward, 985 F.2d at 605. The 1976 Act defined a "work made for hire" as either:
(1) a work prepared by an employee within the scope of his or her employment; or
(2) a work specially ordered or commissioned for use as a contribution to a collective work, as a part of a motion picture or other audiovisual work, as a translation, as a supplementary work, as a compilation, as an instructional text, as a test, as answer material for a test, or as an atlas, if the parties expressly agree in a written instrument signed by them that the work shall be considered a work made for hire.
17 U.S.C. § 101. By adopting this two-part definition, Congress seemingly "meant to address the situation of the full-time or conventional employee in the first provision, and the situation of the independent contractor in the second." Principles of Copyright Law § 5.2.2. Significantly, Congress's new approach was friendlier to commissioned parties than under the 1909 Act, at least in certain ways. In the absence of an employee-employer relationship, only specific kinds of works could be treated as works for hire, and then only if there was a written agreement to do so. See id.
The latest relevant development, for our purposes, came in Community for Creative Non-Violence v. Reid, 490 U.S. 730, 109 S.Ct. 2166, 104 L.Ed.2d 811 (1989). Reid dealt with the proper interpretation of "a work prepared by an employee within the scope of his or her employment" — that is, the first way in which a work can qualify as a work for hire under the 1976 Act. 490 U.S. at 738, 109 S.Ct. 2166 (quoting 17 U.S.C. § 101(1)). Noting that the Act did not define "employee," Reid explained that the term should "be understood in light of the general common law of agency." Id. at 739-41, 109 S.Ct. 2166. In so holding, the Court rejected an approach to § 101(1), adopted by some circuits, that had deemed a hired party an "employee" if the hiring party had "a right to control" or "actual control of the work. Id. at 742, 109 S.Ct. 2166.
Because The Game of Life was created long before the 1976 Act took effect, there is no question that the standard for a work for hire under the 1909 Act governs. See Forward, 985 F.2d at 606 n.2 (noting that the 1976 Act "altered the works for hire doctrine," but only "prospectively"). However, appellants claim that the instance and expense test — the prevailing approach under the 1909 Act for determining whether a commissioned work is a work for hire — is no longer applicable, even as to pre-1978 works. This is so, they argue, because of Reid. Appellants acknowledge that Reid addressed the 1976 Act, but they maintain that its underlying logic applies equally to the 1909 Act. They argue that Reid requires courts to read the 1909 Act's reference to "employer"2 in light of standard agency principles, and thus forecloses the instance and expense test. In other words, according to appellants, the work-for-hire doctrine under the 1909 Act is limited to works produced under a traditional employer-employee relationship defined by principles of agency law, and does not extend to commissioned works, for which the lower courts developed the instance and expense test. In that circumstance, Markham would retain his status as the original author, a status precluded by the work for hire doctrine, and enjoy the termination rights that go with that original author status. Appellants thus urge us (or the district court on remand) to apply the agency law factors set forth in Reid in order to determine whether Klamer qualifies as an employer. Upon doing so, they say, it would be clear that he does not, and the game would therefore not qualify as a work for hire.
Even if we were disposed to appellants' view, however, it does not account for our own precedent. In Forward, which was decided four years after Reid, we applied the instance and expense test to a work governed by the 1909 Act, noting that the test controlled whether a commissioned work qualified as a work for hire. See id. at 606. Under our law of the circuit doctrine, we are bound to apply a prior panel decision that is closely on point. Tomasella v. Nestlé USA, Inc., 962 F.3d 60, 83 (1st Cir. 2020) (citing San Juan Cable LLC v. P.R. Tel. Co., 612 F.3d 25, 33 (1st Cir. 2010)). Facing this obstacle, appellants argue that Forward is not binding precedent on the validity of the instance and expense test because the applicability of the test was not contested there (and, indeed, both the test and Reid were "barely mentioned" in the opinion). We disagree. As we have often observed, "`when a statement in a judicial decision is essential to the result reached in the case, it becomes part of the court's holding.' The result, along with those portions of the opinion necessary to the result, are binding." Arcam Pharm. Corp. v. Faría, 513 F.3d 1, 3 (1st Cir. 2007) (quoting Rossiter v. Potter, 357 F.3d 26, 31 (1st Cir. 2004)).
The facts of Forward plainly demonstrate that the instance and expense test was essential to the result there. John Forward was a music aficionado and record collector who became a fan of a band — George Thorogood and the Destroyers — after seeing them play at a Boston nightclub in 1975. Forward, 985 F.2d at 604. Drawing on his industry contacts, Forward arranged and paid for two recording sessions for the band at Rounder Records, with the aim of producing a demo tape that would get the attention of the label. Id. at 604-05. Besides suggesting specific songs to be recorded, Forward's input was limited to arranging and paying for the sessions. Id. at 605. Rounder Records liked what it heard and signed the band to a contract; the band agreed that Forward could keep the 1976 demo tapes for his own use and enjoyment. Id. More than a decade later, after the band had achieved wider success, Forward informed the band that he was planning to sell the tapes as part of a commercial release. Id. The band objected, and Forward sought a declaratory judgment that he held copyright ownership in the tapes. Id. In part, he argued that the tapes were commissioned works for hire under the 1909 Act because they were created at his instance and expense — and, thus, he was the presumptive copyright owner. Id. at 606.
Applying the instance and expense test, the panel rejected Forward's argument. Id. The panel found that the evidence supported the district court's conclusion that "although Forward booked and paid for the studio time, he neither employed nor commissioned the band members nor did he compensate or agree to compensate them." Id. In short, "Forward was a fan and friend who fostered [the band's] effort [to secure a record contract], not the Archbishop of Saltzburg [sic] commissioning works by Mozart." Id. Put simply, Forward applied the instance and expense test to reach the outcome it did. Accordingly, the panel necessarily held that, post-Reid, the instance and expense test remained applicable to commissioned works under the 1909 Act. That holding is binding on us here.
Anticipating that we might conclude that Forward is binding, Markham's successors-in-interest also argue that we are somehow free to "correct" it because the instance and expense test is inconsistent with the Court's analysis in Reid. That argument misses the mark. Although a "controlling intervening event" — such as "a Supreme Court opinion on the point" — can allow a panel to depart from our court's precedent, United States v. Walker-Couvertier, 860 F.3d 1, 8 (1st Cir. 2017) (quoting United States v. Chhien, 266 F.3d 1, 11 (1st Cir. 2001)), that is not the situation here. Reid was decided before Forward, and, indeed, the Forward panel cited Reid three times. See 985 F.2d at 605, 606, & 606 n.2. Hence, as a panel, we are not free to abandon Forward. See United States v. García-Cartagena, 953 F.3d 14, 27-28 (1st Cir. 2020) (rejecting a party's attempt to cast doubt on an applicable panel decision based on a case decided before that panel decision); United States v. Troy, 618 F.3d 27, 36 (1st Cir. 2010) (noting that Supreme Court cases that precede prior panel decisions are "impuissant against the law of the circuit rule").
Moreover, even if we had authority to abrogate a prior panel opinion on the ground that it misconstrued then-existing Supreme Court precedent, we would be disinclined to do so in this case. While appellants' view of Reid has at least one influential adherent,3 we are skeptical that the Supreme Court, in construing the 1976 Act, casually and implicitly did away with a well-established test under a different Act. We also note that the Second and Ninth Circuits have determined that Reid does not require abandonment of the 1909 Act's instance and expense test. See Twentieth Century Fox Film Corp. v. Entm't Distrib., 429 F.3d 869, 878 (9th Cir. 2005), abrogated on other grounds by Rimini St., Inc. v. Oracle USA, Inc., ___ U.S. ___, 139 S.Ct. 873, 203 L.Ed.2d 180 (2019); Estate of Burne Hogarth v. Edgar Rice Burroughs, Inc., 342 F.3d 149, 162-63 (2d Cir. 2003).4 And finally, although the Eleventh Circuit noted that some of Reid's observations were in tension with the instance and expense test, see M.G.B. Homes, Inc. v. Ameron Homes, Inc., 903 F.2d 1486, 1490 (11th Cir. 1990), abrogated in part by Reed Elsevier, Inc. v. Muchnick, 559 U.S. 154, 160 & n.2, 130 S.Ct. 1237, 176 L.Ed.2d 18 (2010), those comments were made in a case, like Reid, that was governed by the 1976 Act and thus are dicta. See id. at 1490 n.10.
In sum, we stand by the approach in Forward and reiterate that the instance and expense test applies to works governed by the 1909 Act.
C. Application of the instance and expense test
Even under the instance and expense test, Markham's successors-in-interest insist that they prevail. They offer two arguments, both of which were considered and rejected by the district court. First, they maintain that the game fails to satisfy the second prong of the test because it was not made at Klamer's expense. Second, arguing that the test creates only a presumption that the work qualifies as a work for hire, they contend that language in the assignment agreement between Link Research and Markham is enough to rebut the presumption. We construe these arguments as raising fact-intensive mixed questions, which we review with some deference to the district court. See In re IDC Clambakes, Inc., 727 F.3d at 64.
As to the first argument, the evidence amply supports the district court's finding that the game was created at Klamer's expense. In general, the expense requirement looks to the parties' relative investment of resources in the work and the related financial risk. See Marvel Characters, Inc. v. Kirby, 726 F.3d 119, 140 (2d Cir. 2013) (noting that the overall purpose of the expense requirement is to "reward with ownership the party that bears the risk with respect to the work's success"). Here, Klamer promised at the outset to pay Markham any costs incurred — regardless of whether Milton Bradley ultimately liked the game and paid for the rights. Hence, if the dinner at Chasen's had gone poorly, Klamer still would have been obligated to pay Markham's costs.5 As a result, Markham's downside was limited.
Appellants argue that the game was in fact made at the expense of Milton Bradley, not Klamer, with the result that Klamer cannot satisfy the instance and expense test. They seize on the district court's passing remark that "[a]n argument could have been made (but was not)" that the game was made at the expense of Milton Bradley, as "it was Milton Bradley that, once it accepted the Game, paid Klamer $5,000 and bore the risk of its failure to sell to the public." Markham Concepts, Inc. v. Hasbro, Inc., 355 F.Supp.3d 119, 129 n.5 (D.R.I. 2019). But the district court's remark focuses on a later stage in the chronology, after the creation of the work. No doubt, after Milton Bradley paid for the rights to the game, it ran the risk of not recouping its investment. But at the more relevant time period — when the prototype was being developed — it was Klamer who bore the primary risk, as he was on the hook for the costs if Milton Bradley passed on the game.
As for Markham himself, it is true that he was paid in the form of a royalty, rather than a sum certain, which "generally weighs against finding a `work for hire' relationship." Urbont v. Sony Music Entm't, 831 F.3d 80, 90 (2d Cir. 2016). However, the form of payment is "not conclusive," Warren v. Fox Family Worldwide, Inc., 328 F.3d 1136, 1142 (9th Cir. 2003), and distinguishing between a royalty and fixed sum payment can be "a rather inexact method" of determining which party bears the main financial risk. Marvel Characters, 726 F.3d at 140. In this case, we think it significant that Markham's initial royalty payment ($773.05) was a non-refundable advance, meaning that he could keep the money even if the game did not sell a single copy. In that respect, the arrangement resembled payment of a sum certain plus a running royalty, rather than a pure royalty deal. See Warren, 328 F.3d at 1142-43 (finding a work-for-hire relationship when the hired party was paid a fixed sum and a royalty); cf. Picture Music, Inc. v. Bourne, Inc., 314 F.Supp. 640, 651 (S.D.N.Y. 1970), aff'd, 457 F.2d 1213 (2d Cir. 1972) ("[T]he fact that the author was obliged to repay advances on royalties which were never accrued is an indicant that the relationship was not an employment for hire[.]" (emphasis added)). Overall, we find no error, clear or otherwise, in the district court's determination that the game was made at Klamer's expense.6
The second argument — that the assignment agreement rebuts the presumption created by the instance and expense test — presents a closer question. Some cases suggest that a contemporaneous agreement can clarify that a work, even if made at the instance and expense of another, is not a work for hire (and therefore that the hired party remains the "author," entitled to termination rights). Assuming that a contemporaneous agreement could indeed alter the game's work-for-hire status,7 the independent contractor bears the burden of showing that such a contrary agreement was made, see Playboy Enters., Inc. v. Dumas, 53 F.3d 549, 554-55 (2d Cir. 1995), and courts generally demand clear and specific evidence of such an agreement, see Lin-Brook Builders Hardware v. Gertler, 352 F.2d 298, 300 (9th Cir. 1965) (requiring "an express contractual reservation of the copyright in the artist" to rebut the presumption); see also 1 Nimmer on Copyright § 5.03[B][c] (requiring "persuasive evidence" of a contrary agreement).
Markham's successors-in-interest point to two parts of the assignment agreement which, they say, overcome the presumption. First, the agreement recited that, "[a]t the request of LINK, MARKHAM has invented, designed[,] and developed a game tentatively known as `THE GAME OF LIFE.'" But that language falls well short of an express reservation of copyright. In fact, insofar as it makes clear that the work was done "[a]t the request" of Link, it supports, rather than undermines, the idea that the game was a work for hire.
Second, the agreement provided that
[u]pon the request of LINK, MARKHAM will pursue any copyright, trade-mark and patent applications ... to which he may be entitled as the inventor, designer and developer of the Game.... MARKHAM will assign any such copyright, trade-mark, patent or application therefor to LINK, provided that said assignments will revert to MARKHAM upon the termination of this agreement.
We agree with the district court that this language is best read not as a reservation in Markham, but as a kind of failsafe for Link. That is, it makes clear that if, contrary to expectations, Markham were entitled to the copyright in the game, he would, at Link's request, assign it over. See Marvel Characters, 726 F.3d at 143 (suggesting that a freelancer's assignments could be "redundancies insisted upon by [the hiring party] to protect its rights" rather than an indication that the hiring party "did not already own the rights"). This reading is supported by the tentative, open-ended language ("to which he may be entitled," "any such copyright") (emphasis added), which appears to be an attempt to cover all conceivable bases without acknowledging that any rights actually belong to Markham. Regardless, this language is not the required "express contractual reservation of the copyright in the artist." Lin-Brook, 352 F.2d at 300. The district court thus supportably found that the assignment agreement did not overcome the presumption that the game was a work for hire made for Klamer. As a result, Markham "never owned the copyrights to assign," and "there are no rights the assignment of which his ... heirs may now terminate." Marvel Characters, 726 F.3d at 137.8
Because the evidence amply supports the district court's conclusion that the game was created at the instance and expense of Klamer and that there is insufficient evidence to rebut the resulting work for hire presumption, we need not address the defendants' alternative theory for affirmance: that the game was a work for hire created by Chambers and Israel — with Markham as the "employer." This alternative argument — essentially, another way of establishing that the game was a work for hire — would also mean that no termination rights exist and would similarly spell defeat for Markham's successors-in-interest.
The outline of this alternative theory seems to have emerged in supplementary interrogatory responses made after the close of discovery and shortly before trial. Plaintiffs unsuccessfully moved to "preclude" this new theory and strike the underlying responses. They challenge the district court's rejection of their motion on appeal. But discovery rulings are reviewed only for abuse of discretion, and reversal requires a showing that the ruling was both "plainly wrong" and resulted in "substantial prejudice." In re Subpoena to Witzel, 531 F.3d 113, 117 (1st Cir. 2008) (quoting Saldana-Sanchez v. Lopez-Gerena, 256 F.3d 1, 8 (1st Cir. 2001)).
Even assuming the district court erred, and we are not suggesting that it did, we fail to understand how its ruling caused substantial prejudice. As appellants essentially concede, the district court did not adopt the alternative theory — and neither do we. Appellants' real concern, as we understand it, is that the interrogatories introduced novel testimony from Israel and Chambers indicating that they had a much more prominent role in the creation of the game than previously disclosed. But, as noted above, to the extent the updated interrogatory responses were inconsistent with earlier depositions of Israel and Chambers (or their ultimate testimony at trial), appellants had the opportunity to cross-examine them at trial and impeach them with any inconsistencies. See supra note 6.
Accordingly, the judgment of the district court is affirmed. So ordered.