2003 Copyright Exam Sample Answer by Eric Goldman

2003 Sample Answer
Eric Goldman

About Grading

The exams were consistently good and, as usual at Marquette, tightly bunched together.  There were 6 As, 5 ABs, 24 Bs and 1 BC.

I changed my grading system a bit.  Before, I used a 100 point scale that I used in 5 point increments with a 60 baseline.  This semester, in keeping with the minimal grading choices at Marquette, I tried a 5 point scale where I only used whole integers between 1 and 5, with 3 as the baseline.  I think this system worked OK (better in Cyberlaw than Copyrights, frankly).  I share your frustration about there being so little grade dispersion, leading to the ubiquitous B.  Sorry I couldn’t avoid that outcome.

About Using the Computer

21 of you took the exam by computer.  14 of you handwrote.  About 10 of you downloaded the exam electronically, and 11 of you returned the exam by email.

There was almost no difference in terms of grades.  The average grade for handwritten exams was 3.16 (out of 5) and the average for computers was 3.21.  I don’t believe the difference is statistically significant.  Having said that, I MUCH preferred typewritten exams as they took only 1/3 to 1/4 the amount of time to grade, and some handwritten exams were painful to read.  Further, I heard a lot of enthusiasm—and no complaints—from those who chose to use the computer.  Some students thought it was less physically stressful than handwriting, and a lot of students liked the ability to organize and reorganize their thoughts.

Because I was able to capture word counts from the typewritten exams, I got some interesting data about how many words students actually used.  In general, there was little correlation between number of words and score (either way).  Certainly I got cranky with longer answers that were padded with cutting ‘n’ pasting from class notes, so in some cases length worked to the student’s disadvantage.

Q1: average 2,021 words.  High: 3,417 words.  Low: 1,032.  One student was able to get a 4 on this question, and an A on the exam, using as few as 1,272 words.

Q2: average 274 words.  High: 593 words.  Low: 105 words.  One student got a 5 on this question with as few as 186 words.

Q3: average: 423 words.  High: 600 words.  Low: 183 words.  One student got a 5 on this question with as few as 400 words, and one student got a 4 on this question with as few as 183 words.

My point is: it’s about picking the right words, not saying more of them!  I definitely cannot discourage you enough from slavish cutting ‘n’ pasting of notes, which often raised more questions in my mind than it answered.

In the future, I expect to adjust the word limits down, perhaps as low as 20 words per minute instead of 30.

Question 1

This is your typical issue spotting question.  Insight, organization and accuracy were keys to success.

What is Copyrightable, and Who is the “Author”?

Because we are dealing with a heavily factual work, I would start by looking at the “minimum unit of copyrightability” and identifying the author of these units.  Once we’ve identified the various copyrightable units, we can more precisely discuss ownership and infringement.

  • Information in an individual category.  Most of the information is too factual to contain the requisite originality.  Perhaps the information in the “purpose” or “eligibility” categories has enough “stuff” to be copyrightable but this is marginal at best.  If there is any copyrightable material here, and Julie just cut ‘n’ paste the contents of those categories, the author would be the program sponsor.  Julie’s edits to the information would be unlikely to transform these facts into copyrightable works.
  • An individual program description.  There is enough material here that the whole description could be copyrightable.  OK, but who is the author?  If Julie cut ‘n’ paste the description verbatim, then the description’s copyright would be owned by the program sponsor.  Otherwise, it depends on whether the description is a literary work or a compilation.  If the description is a compilation, then to the extent Julie rearranged the material by taking a program description from the program sponsor and placing the relevant content into the applicable categories, she has applied her own selection/arrangement/coordination to the facts (the information in each individual category) and would be the author of that (but see taxonomy discussion below).  This conclusion is only reinforced by the edits Julie makes.
  • “Taxonomy.”  Almost everyone discussed the copyrightability of taxonomies, but few of you evidenced any thoughtfulness on this point.

In this case, two different things might qualify as a “taxonomy.”  First, the relationships of each program description to other program descriptions, which were organized by religion and then alphabetically.  As a two-layer taxonomy, this may be too thin to be copyrightable.  Contrast the complex hierarchies in the biological taxonomy or even the dental procedures taxonomy in the ADA case.  Instead, this structure is better thought of as a methodology for selecting, arranging and coordinating items, and thus applies to a compilation copyright (as discussed above), not as a separate copyrightable work.

Second, the categories used within each program description (“scholarship name,” “sponsor name,” etc.).  Again, I think of this more as a methodology for selecting, arranging and coordinating items and thus should be analyzed as a compilation copyright.  However, if this structure for relating facts to each other is copyrightable, then the author is CollegeFunding.

In that case, CollegeFunding could claim that program descriptions were “joint works” because any copyrightable contribution of Julie was inseparably merged with CollegeFunding’s copyrightable taxonomy.

  • Selection/arrangement/coordination of program descriptions.  Arranging program descriptions by religion and then alphabetically is somewhat obvious.  However, there are multiple ways the records could be organized, including by geography, by deadline date, by amount of award, by type of financial aid (loans v. scholarships v. internships), or just alphabetically.  Therefore, unlike Feist, where there are so few ways to organize white pages information, this organization probably has enough creative spark to be copyrightable.  This conclusion is reinforced by the judgment of what programs to include and exclude (such as what religions to cover and what programs should be associated with each religion), unless Julie actually catalogued every single program for every single religion.  Therefore, I believe the overall organization of program descriptions in the book is protected by a compilation copyright.  Because Julie made these judgments, she is the author of the compilation.

This section was a big mover in the scores.  Those of you who thoughtfully and specifically described each type of copyrightable work were usually rewarded.  A cursory discussion in this area pushed you out of an honors grade.

As an aside, fixation is also a prerequisite for copyright, which is amply met here.  A number of you said that fixation was established by the book’s publication.  In fact, any copyrights should have been fixed no later than when Julie reduced them to a tangible form.

What Copyrights Does CollegeFunding Own?

This question was a little imprecise, and I really apologize for that.  A literal reading of the question suggests that you were to analyze the question from Jane’s perspective instead of CollegeFunding’s, an extra layer of complication that I did not intend.  Fortunately those of you who spotted this issue were able to assume away the problem pretty deftly, but I apologize if this was unnecessarily confusing.

Looking at the question from CollegeFunding’s perspective, CollegeFunding needs to establish ownership of the applicable copyrights.  There are two problems here.

First, it may be infringing on copyrights owned by the program sponsors.  On this front, CollegeFunding is caught between a rock and a hard place.  If it argues that the program sponsor’s data is factual or subject to a defense like the merger doctrine or scenes a faire, then this argument should also apply with equal force in any litigation against the students.  On the other hand, if it argues against the students that each individual category entry is copyrightable, then it will likely have to acknowledge that it infringed the copyright in these entries owned by the program sponsors.  This is an inherent problem among any owner of a compilation copyright and is magnified by cases like CDN v. Kapes that lower the size of a copyrightable unit.

Second, CollegeFunding needs to establish it has ownership over the work that Julie did.  There are a few ways that CollegeFunding could do so:

  • It could argue that Julie was a scrivener like the Lindsey case.  CollegeFunding would make this argument by saying it provided a template, thus making all of the judgments, and Julie merely robotically followed its instructions.  This argument doesn’t seem very persuasive because Julie had to make lots of judgments herself about what went where.
  • It could argue that the program descriptions were a joint work (see above).  This argument isn’t great but the statutory definition of “joint work” could support this—the resulting product was interdependent.  This argument, if successful, would give CollegeFunding standing to sue but with the duty to account.  It is unclear if the 20% royalty would satisfy or negate that duty, in which case this approach could end up being more expensive than initially bargained-for.
  • It could argue there was a transfer.

In order to obtain standing to sue, CollegeFunding needs to obtain either ownership or an exclusive license.  However, it cannot obtain either of those without a written agreement (Sec. 204) unless the work was done by an employee.  The facts naturally begged you to apply the CCNV v. Reid factors to determine if Julie was an employee, which many of you did but some of you skipped or only mentioned in passing, which invariably hurt your score.

  • skill required.  You could go either way on this.  The work could be performed by a college kid with no obvious specialized training, so it’s not super-skilled and thus more like an employee. On the other hand, some careful judgments were required, which would push this towards an independent contractor.
  • source of instrumentalities and tools.  The facts don’t say specifically, although a natural inference is that Julie provided her own tools because it was in her dorm room.  Some of you noted that CollegeFunding provided an “instrumentality” by providing the categories for her to use.  I’m not exactly sure what an “instrumentality” is (dictionary.com isn’t helpful), but I think this factor generally turns on tools, supplies, etc., and I think of the categories as instructions about the work, not supplies to do the work.
  • location of the work.  Takes place in Julie’s dorm room.  Favors independent contractor.
  • duration of relationship between the parties.  One summer.  Favors independent contractor.
  • whether the hiring party can assign additional projects to the hired party.  Facts are unclear but the ad hoc and custom arranged nature of this summer relationship suggests this favors independent contractor.
  • hired party’s discretion over when and how long to work.  Facts are unclear but it’s natural to assume that Julie is free to do the work whenever she wants during the summer, which favors independent contractor.
  • hired party’s role in hiring and paying assistants.  The facts are unclear about this.  You could have assumed that Julie was free to do whatever she wants as long as she delivers a book to CollegeFunding, in which case this factor would favor independent contractor.
  • is work part of regular business of hiring party.  It is, so this factor favors employment.
  • if hiring party is in business.  It’s natural to assume yes, so this factor favors employment.
  • employee benefits.  Facts don’t say.
  • tax treatment of hired party.  Facts don’t say.

Several factors clearly favor independent contractors, and a few clearly favor employment.  Going back to the touchstone agency principles underlying these factors, the overall goal is to assess whether Jane had the “right to control the manner and means by which the product is accomplished.”  With that perspective, I think the more natural reading of the facts is that Julie is an independent contractor. But certainly CollegeFunding would try to argue otherwise.

CollegeFunding can also get a copyright assignment in writing now.  Because Julie is family, she would probably be willing to assign.  CollegeFunding could also try to reduce to writing an agreement characterizing this as a specially commissioned work (if the Second Circuit position applies) but the only benefit of doing so (avoiding the 35 year termination of transfer right) seems unnecessary given the time-sensitivity of the work (i.e., all info will be moot by then).

Did the Students Infringe?

First, we need to establish copying-in-fact.  We do not have any actual evidence of copying, so we need to see the likelihood of establishing copying circumstantially.  This can be done by showing access and substantial similarity.

The access showing is weak.  The book was published but in a geographically-distant location and only 1 day prior to the students’ publication.  However, the book was published, and in the age of the Internet, can one argue that data can move so fast nowadays that any publication anywhere in the world should be enough to show sufficient access?

CollegeFunding could also try desperation arguments that perhaps Julie’s dorm room was broken into, a la the Arnstein plaintiffs.

With a weak showing of access, CollegeFunding may still be able to establish copying-in-fact with a strong enough showing of similarity, such as the “striking similarity” in the Ty v. GMA case.  Here, we may very well have the same degree of striking similarity found in the Ty case.  The copies were identical, but arguably this is because the source material was the same.  Whether a judge would find striking similarity in such a fact-intensive work likely depends on the judge’s predispositions, not the law per se.

Assuming a judge accepts the inference that copying occurred, we turn next to whether the copying was wrongful.  I was disappointed how many of you cut corners on this analysis.  Some of you combined the wrongful copying analysis with the copying-in-fact analysis, but most of you took the same approach of the CCC court–I know something is copyrightable, and I know something was taken, but I won’t be specific about what exactly was taken and how that overlaps with what copyright the plaintiff actually owns.  Here, given the copyrightability and authorship/ownership discussion above, I expected you to be a little more careful.  Exactly what right was infringed and did CollegeFunding own it?

Let’s start with the copyright in the book.  Assume CollegeFunding has a compilation copyright in the book as a whole.  Arguably the students did not infringe this because they took such a small portion of the book.  Certainly you cannot rely on CCC here because, in that case, so much of the second database overlapped with the original database.

CollegeFunding would argue that the compilation copyright protects the association of programs with a particular religion, and by taking all programs associated with Scientology, this association was infringed.  However, if the protectable copyrightable interest was taking a pool of programs and associating each one with a specific religion, does this get infringed when all of the programs are taken from only a single religion?

On this point, you might use the various forms of infringement tests to determine the answer.  I think the dissection and abstraction (and AFC) tests would almost invariably lead to no infringement.  The dissection test would carve away non-copyrightable elements which should raise the infringement level above taking programs from a single religion.  The abstraction test probably does the same (although using the abstraction test makes one wonder if there is anything copyrightable in the first place, as it is tough to find a copyrightable abstraction level).  Under the ordinary observer test, I think it’s a completely open question of how a jury would evaluate this question.  You probably have the best chance of infringement under the total concept and feel and intrinsic-extrinsic tests, as these tests may permit the fact-finder to consider the similarity between the non-copyrightable elements.

As discussed earlier, it’s not clear there is any independent copyright in the taxonomy.  If there is, and the student guidebook used the exact same categories, then the students infringed that right’s reproduction and distribution right (not a derivative work because it’s an identical copy, and not a public display unless the taxonomy ends up being styled as a literary work).

By making identical copies, the students arguably infringed (reproduced, distributed and maybe publicly displayed) any copyright interest in the program descriptions (or any copyrightable component thereof).  Here, defenses like the merger doctrine and scenes a faire would be strong, because both the book and the guidebook may draw from the same sources, as well as an argument that CollegeFunding’s purported copyright is really just infringing the program sponsors itself.

Fair Use

As with the infringement analysis, the fair use analysis properly should be customized for each different infringed work.

  • Nature of Use.  On the spectrum of commercial v. educational, the students can make a good argument that their use is educational (being in a University setting and with a goal to help with the educative function of the University).  On that basis, this factor could weigh in favor of the students.  Some of you assumed that the guidebook was made available for sale and thus concluded that the use was commercial, and that was OK too.

Whether the use was transformative is a tougher question.  If the copyrighted work is the book compilation, then arguably there was something new added (6 programs).  However, CollegeFunding could argue that the purpose was the same (resource to find financial aid information).  The students would counter-argue that the financial aid information was put into a new context (a transition to college book) and thus had a different purpose and character.  Each argument is supportable by case law, so a judge can decide on his or her own biases.

If the copyrighted work was the taxonomy or the individual programs, then there’s a better argument that nothing new was added.  Still, under Kelly v. Arriba, a court could find that there still was a different character/purpose and thus transformation without anything new being added.

CollegeFunding might also argue that the infringement was exploitative like the Napster case.  It’s not clear how the courts would consider that argument.

  • Nature of the Work.  All of the copyrightable works are pretty factual in nature and have been published.  Thus, however this factor is interpreted, it should weigh in favor of the students.  A few of you believed that published works weighed against fair use; this evidenced a solid misunderstanding of Harper & Row.
  • Amount Taken.  One argument is that the students only took 18 of 1000 program descriptions, or less than 2%.  It’s pretty hard to argue that these are the “heart of the work” because unlike Harper & Row, there’s no basis to assume that any program description is more important than any other program description.  Of course CollegeFunding can try the circular argument that it was the most important part of the work because they chose to take that.

Another argument is that the program description was the copyrightable unit, so the students took 18 of 18 copyrightable units.  Or if the taxonomy is the copyrightable unit, it was taken 100% 18 times.

  • Effect on the Market.  As usual, we encounter lots of variances about defining the market.

If the market is for the books in total, arguably the guidebook will interfere with that market because some group of Scientology students will treat the guidebook as a substitute.  The students could counter-argue that they are just a small group at a single university and thus the market effect is de minimis, but CollegeFunding can counter-counter-argue that this factor aggregates up all like conduct, and if every university student group did the same, there would be no market for the books.

CollegeFunding could also counter-counter-argue that the students have usurped a traditional, reasonable, or likely to be developed derivative market that it might want to exploit (Texaco) or a derivative market that it has the right to choose to not exploit (Castle Rock).  It does seem logical that a book publisher might break apart its books into sub-constituencies; on the other hand, this splintering may not be cost-effective.  Nevertheless, some judges will buy either argument.

The students could alternatively argue that, if there are programs in the book that are available across multiple religions, then the guidebook may increase the demand for the book because students who are interested in religious-directed financial aid may learn that there are more programs to pursue and seek the book out to fill in the gap.  This doesn’t seem very persuasive, certainly not without attribution to the book or something that would connect the two works.

Many reference books are not designed to be sold retail but rather make their bread and butter in the library and institutional market.  If that’s the case, the students could argue that the guidebook would not affect library purchases.  CollegeFunding could counter-argue that if student groups for every religion did the same conduct, then libraries wouldn’t buy the books, a logically-accurate but silly argument.

If the market is each individual program (or, worse, a component of the program), then CollegeFunding may have a harder time convincing a judge that there is an exploitable market.

In all cases, CollegeFunding can try the circular argument that the guidebook interfered with CollegeFunding’s licensing rights.  Actual evidence of having a licensing program can make this argument a little more solid, but there is really no defense to this argument by the students if a judge is willing to accept it.

On balance, I think that if the book is characterized as for an educational purpose, a court will likely drive the other factors to find fair use.  Otherwise, this would not be fair use.

While I have gone into a lot more detail on fair use than I expected from you, I did want you to do some of the back ‘n’ forth arguments.  Some of you only had a single sentence on each factor, which at minimum left a bad taste in my mouth.  I also did still find a few circular arguments that did not acknowledge their circularity, which definitely made me cranky.

What Should CollegeFunding Do?

Kudos to all of you for actually answering this portion of the question!  Suing the students may not be a good idea for several reasons:

  • CollegeFunding may not even have standing to sue
  • Probability of winning is unpredictable (questions about ownership, infringement and fair use)
  • there’s a risk that a lawsuit will expose CollegeFunding’s potentially sloppy database-building practices
  • there’s a risk of establishing adverse precedent that CollegeFunding owns nothing (before the suit, there may still be enough mystery to inhibit competitors from stealing more aggressively)
  • the quantum of infringement is small, so there’s a good chance that the litigation cost will exceed any benefits
  • this is especially true because students are quintessential deadbeats (no offense)

On the other hand, at some point CollegeFunding has to stop second-comers or its business will ultimately be hen-pecked to death.  I’m not sure this is the right case, though.

A Little History

This question is based on my mom’s company Reference Service Press.  In the particular situation I dealt with, we decided that I would write a stern lawyerly letter to the second comer.  The letter recipient took quite a while to get back to me.  Meanwhile, because the infringement occurred online, I also went after her IAP by sending 512(c)(3) notices (which proved to be worthless because her IAP was also non-responsive, giving me an added deadbeat defendant if I wanted it).  However, when I finally got a call from the second comer, she was sufficiently scared by the lawyer letter to agree to take everything down, which she did in a matter of days..

Word Count Discussion

My answer above, which definitely took me more than 2 hours and includes commentary, has about 3,500 words.

Question 2

This should have been a very easy question!  Some of you missed this completely and I penalized you pretty hard for that.  Some of you thought this question was entirely duplicative of the material covered in question 1, which it was not supposed to be (as you’ll see in a moment) but I can see how that happened.

Because there is no actual written agreement, Pretentious has not acquired ownership rights (either an assignment or as a specially commissioned work for hire) or an exclusive license.  (204).  The article could not have qualified as a specially commissioned work for hire in any case because it was written on spec.  Sorry the facts were not clearer, but nothing suggested that Jane was an employee.

Despite the lack of a written agreement, could Pretentious find some way to obtain ownership or an exclusive license based on some form of promissory estoppel?  I believe the Statute of Frauds trumps even a promissory estoppel doctrine.

So what does Pretentious have?  Under Effects Associates v. Cohen, Pretentious should have obtained some form of nonexclusive license.  The scope of that license is unclear but, at minimum, includes the magazine publication.  However, this license is still terminable after 35 years (203).

Pretentious may also have rights under Section 201(c) as a contribution to a collective work (the magazine).  Indeed, as we discussed in class, 201(c) seemed to be designed exactly for this purpose—as a backstop statutory license when the contracting process fails.  Under 201(c) Pretentious has the right to use the article as part of the magazine, any revision of the magazine (which does not include databases where the article in displayed “out of context”—Tasini) and any later magazine in the same series.

(I could see a technical argument being made that 201(c) preempts the implied nonexclusive license, but let’s reserve that arcane issue for another day).

Only a few of you noted the 201(c) issue, which almost always improved your score on the question.  Failing to address 201(c) did not necessarily lead to a reduction in score.

This question is a typical one for me, as I have often delayed signing onerous publication agreements knowing that a periodical will publish the article even if I don’t.  (Although I’ve never lied about signing it when I know I won’t).  My word count, including commentary, is about 250.

Question 3

In the future, I don’t think I will double up topic coverage like this.  However, I decided to make a last minute swap of questions, so it did create some redundancy.  Because of the doubling up, I was really hoping you would mix it up with the facts, but a number of you instead gave me one sentence responses on each factor, which downgraded your score.

Nature of Use.  Joe’s use is not educational, but it’s not fairly characterized as commercial because he’s not trying to sell anything or otherwise derive a commercial benefit.  However, using the Napster court’s circular reasoning, his use could be characterized as commercial because he is saving the cost of copying.  Thus, under the traditional statutory language, it’s hard to predict how Joe’s use will be interpreted, but I think it should slightly weigh in favor of Joe.

Using the transformative approach, the comic owner could argue that nothing new was added—Joe just took a copy of the comic verbatim.  On the other hand, Joe could argue that he added his personal sentiments to each letter.   That argument seems a little silly to me.  Alternatively, using Kelly v. Arriba’s standard, Joe could argue that even if he didn’t add anything new, the comic was being used for a different purpose (personalized social expression instead of media entertainment) and with a different character (a single comic in letterhead, not in a series of comics like the newspaper).

As for Napster’s exploitative analysis, I personally think this varies with the number of copies Joe makes.  If he makes a single copy, I doubt anyone would call that exploitative.  But I could see a court feeling like Joe is getting away with something if he makes 1,000 copies.

Nature of the Work.  The comic was published (favoring fair use) but is more like fiction than fact (against fair use).  This could make the factor a wash or a court could pick whichever prong it thinks it more important.  I was underwhelmed by those of you who only referenced one of the prongs in your discussion and ignored the other prong entirely.

Amount and Substantiality of the Portion Taken.  Joe could argue that the copyrighted work is the series of comics and he took only one.  But that’s a pretty weak argument.  Each individual comic should be copyrightable, so this factor weighs in favor of the comic owner.  This makes the infringement presumptively unfair.

Effect on the Market.  If the relevant market is the newspapers, the original media for the comic’s publication, we would expect Joe’s use to have no effect on the market.

If the relevant market is purchasing stationery containing the comic, then Joe’s use has some effect on the market.  Joe might argue that he would not have purchased stationery unless it had the exact comic he wanted in the exact stationery format he wanted, which might very well be true.  However, this is where a court can circularly conclude that other like conduct will affect the market, which is also true.

Here, there is an explicit licensing market, so there’s no need to theorize about the pursuit of derivative markets as in Texaco.  However, Joe could argue that just because the comic owner has licensed the rights to make stationery to a third party doesn’t mean there was a licensing program he could take advantage of.  In fact, the opposite appears true.  This hearkens back to the dissent in the Princeton case, saying that there was no actual licensing market for the specific excerpts professors wanted to use.  The comic owner can counter-argue that it’s the copyright owner’s province to decide to leave a derivative market unexploited (Castle Rock).

Once again, I think the number of copies Joe makes can affect the outcome.  It shouldn’t, because the “aggregate up” theory should apply equally in both cases.  But I think a court would feel more comfortable saying that a single copy doesn’t affect the market, whereas 1,000 copies would.

On balance, I think Joe’s activity will be fair use if he makes one copy and will not be fair use if he makes 1,000 copies.

I asked this question because I used to make personal holiday cards using clipped comics.  I’ve since stopped (I’m pretty sure it’s been longer than the statute of limitations J) but this is a typical situation where copyright law permeates deeply into rather prosaic decisions.  My word count, including commentary: a little over 600 words.