FREE SPEECH COALITION WEIGHS IN ON NEW 2257 REGULATIONS

In response to Attorney General John Ashcroft’s announcement in June that the U.S.C. Title 18, Section 2257 adult material record-keeping and labeling law is being expanded to include the Internet, Free Speech Coalition has submitted comments to express our concerns about the new regulations and their application to this new medium. Drafted primarily by Board of Directors’ member and First Amendment attorney Reed Lee, with assistance from Board Chairman Jeffrey J. Douglas, the 23-page document outlines in detail the constitutional infirmities and practical difficulties in the proposed regulations.

Beginning with the impropriety and unfairness of a retroactive effective date for the proposed rules, Lee outlines seven main points in making the argument that these new policies are unjustly burdensome on protected expression and on members of FSC who produce it. One of the key issues as these regulations apply to the Internet is the definition of a “producer” of content and the requirement that “secondary producers” also keep original records. As Lee points out, those who are not a primary producer should not be required to keep extensive records, particularly since this issue of secondary producers has already been litigated in Sundance Associates, Inc. Reno, and was decided contrary to these rules. On a constitutional note, Lee says, “the volume and complex­ity of those records will very likely chill the willingness of many to republish material, thus limiting the reach of constitutionally protected expression.”

As the trade association of the adult entertainment industry, representing all the diverse creators and distributors of adult products and services, Free Speech Coalition continues to serve as the voice of this industry in fighting against regulations that unfairly target our industry. Our timely and cogent comments to the Justice Department in this matter are another example of that important work. A copy of the Comments is published below.

Re: Docket No. CRM 103
AG Order No. 2723-2004
RIN 1105-AB05
Comments on Proposed Rule concerning
Inspection of Records Relating to
Sexually Explicit Performances

To whom it may concern:

The Free Speech Coalition is the adult industry’s trade association, founded in 1991. The organization represents all creators and distributors of adult entertainment, information, products and services, including the interests of the creative artists as well as their employers. The organization refrains from litigation except in extraordinary circumstances (see for instance, Ashcroft –vs- Free Speech Coalition, 535 U.S. 234 (2002).) The Free Speech Coalition shares the Justice Department’s resolve to detect and deter child pornography. Our organization is the only entity known to us which offers up to $10,000.00 annually for information leading to the arrest and conviction of traffickers in child pornography.

With hundreds of members, representing every aspect of the entities affected by 18 U.S.C section 2257 and the related regulations, we trust that the Department will take the expressed concerns with the seriousness they are offered.

Much of the material attached mirrors comments offered by Reed Lee, submitted individually and separately. Mr. Lee, a Director of the Free Speech Coalition, is the primary author of these comments.

Kat Sunlove, Executive Director
Nicholas T. Boyias, President
Jeffrey J. Douglas, Chair of the Board of Directors


Contents of Comments

I. The regulations should adopt a purely prospective effective date
A. Retrospective application will raise constitutional questions which are otherwise moot
B. Congress has wisely rejected retrospective application in amending the statute
C. Prospective application is the only fair way to implement a complex record-keeping
requirement
D. No obligations should be imposed respecting images first made prior to the effective date

II. The definition of “producer” is unwarranted by the statute and burdens too many people
A. There should be only one primary producer
B. No “secondary producer” should be required to maintain extensive records

III. The record-keeping requirements are unclear and are too burdensome
A. Records should be acceptable if they are in digital form instead of on paper
B. The requirements for a picture identification card are confusing and inappropriate
C. It is extraordinarily burdensome to require that the records contain a copy of each work
D. The proposed rule is too burdensome in creating multiple, duplicate records
E. The proposed rule confuses the indexing requirements for the records
F. The record segregation requirement is unduly harsh
G. The proposed rule requires that the records be maintained for too long

IV. The record-shifting requirements are impermissibly burdensome
The record-shifting requirements seriously burden and chill
the dissemination of constitutionally protected expression
Substantially less restrictive regulatory alternatives are available

V. The inspection provisions are in some respects improper and in other respects incomplete
A. The final rule should fully and properly define “investigators”
B. Advance notice of inspections should be provided
C. The proposed rule excessively defines “reasonable times”
D. The proposed rule allows unnecessarily frequent and insufficiently justified inspections
E. The proposed rule is unclear concerning photocopying
F. The proposed rule should expressly clarify the right to remain silent
G. The proposed rule should address the seizure of the required records

VI. The disclosure statement requirements are confusing and, in many cases, too burdensome
A. Requiring the open disclosure of the custodian’s name and address
is impermissible in many situations
B. The requirements concerning the placement of the disclosure statement are confusing
C. The regulations concerning the date of production are inappropriate or incomplete
D. The typographic requirements for the disclosure statement are absurd
E. Provisions governing an “exemption statement” are unnecessary and improper
F. The regulation should specify that “transfer” is limited to commercial transfer

VII. The proposed rule contains some minor grammatical errors and drafting problems
A. The final rule should correct some minor grammatical errors
B. The final rule should avoid a troublesome drafting error

I. The regulations should adopt a purely prospective effective date.

The proposed rule retains an initial effective date of November 1, 1990, §§ 75.2(a), 75.6(a), 69 Fed. Reg. 35551, 35553, with other requirements phased in as of May 26, 1992, §§ 75.2(a)(1), (2), 75.6(a), 69 Fed. Reg. 35552-53 and still others phased in as of 30 days after the publication of the final rule contemplated here, § 75.2(a)(1), (d), 69 Fed. Reg. 35552. The final rule should adopt a single, prospective effective date (no earlier than 30 days after publication of the final rule), and the Department should avoid retrospective investigations and prosecutions.

A. Retrospective application will raise constitutional questions which
are otherwise moot.
By its own terms, of course, section 2257 became effective on May 17, 1989. P.L. 100-690 sec. 7513(c), 102 Stat. 4488. Yet the initial version of the Section was declared unconstitutional by the United States District Court for the District of Columbia, American Library Association v. Thornburg, 713 F. Supp. 469 (D. D.C. 1989) and, after the appeal in that case was argued before a panel of the United States Court of Appeals for the District of Columbia, American Library Association v. Barr, 956 F.2d 1178, 1186 (D.C. Cir. 1992), Congress amended the section in order to – to use the term adopted by Congress – “restor[e]” the record-keeping provisions, P.L. 101-647 sec. 301(a), 104 Stat. 4816. Both Congress and this Department thus ultimately, and wisely, refrained from adopting May 17, 1989, or any earlier date, as the effective date. For the same reasons, this Department should now avoid articulating any effective date earlier than April 30, 2003. Any effort to inspect records or prosecute violations dating between February 6, 1978, and November 29, 1990, will reraise all questions concerning the initial constitutionality of Section 2257 – questions which Congress plainly tried to moot in 1990. Moreover, since the magistrate in American Library Association v. Thornburg (Civ. A. No. 89-0661) designated July 3, 1995, as the effective date, a ruling which the Government never challenged, any attempt to enforce the law prior to that date will raise significant due process questions. Similarly, any effort to inspect records or prosecute violations dating between November 29, 1990, and April 30, 2003, will raise questions concerning the constitutionality of the Section during that period. Since Congress recently amended section 2257(d)(2), P.L 108-21 sec. 511(a)(1), 117 Stat. 684, in order to avoid a constitutional problem (though it may have created another, cf. Message from President Transmitting Child Protection and Obscenity Enforcement Act of 1987, H.R. Doc. 100-129 at 66-68), this Department should avoid the necessity of litigating the Section’s validity prior to its most recent amendment by promulgating a purely prospective effective date.

A. Congress has wisely rejected retrospective application in amending the statute.

Although what became Section 2257 was originally proposed with a 25-year retroactive effective date, Message from President, H.R. Doc. 100-126 at 8, Congress was always concerned about the validity and workability of such a proposal. And although it initially enacted the section with a ten year retroactive effective date, P.L. 100-690 sec. 7513(a) (original § 2257(a)(1)), 102 Stat. 4487; but see Id. sec. 7513(c), 102 Stat. 4488 (180-day prospective “effective date”), Congress essentially altogether abandoned any effort at imposing any retroactive date when it amended the section in an effort to save its constitutionality. P.L 101-647 sec. 301(b), 104 Stat. 4816. Since valid and comprehensive implementing regulations were plainly both contemplated and required, see P.L. 101-647 sec. 312(1), 104 Stat. 4817, and since this Department now apparently recognizes the shortcomings of the regulations as initially promulgated, the spirit of purely prospective application mandates an effective date no earlier than 30 days after the publication of the final rule.

B. Prospective application is the only fair way to implement a complex
record-keeping requirement.

Section 2257 and, especially, its implementing regulations purport to impose upon those who produce and disseminate constitutionally protected expression a genuinely novel and quite elaborate record-keeping requirement. That requirement extends far beyond the verification of each performer’s age; it extends to the production, organization, maintenance, transfer, disclosure, and inspection of a potentially extensive set of carefully cross-referenced records. Even assuming that it is constitutionally permissible to impose such burdensome requirements upon those who are merely exercising their constitutional rights in order to catch others who may have exceeded theirs, simple fairness dictates that producers and disseminators of the protected expression have a genuine chance to prepare to meet those obligations. This Department has been vigorously and successfully prosecuting child pornography cases for many years now without resort to a single inspection under Section 2257. Since it cannot be said that this Department showed any serious need or interest in inspecting the required records prior to this time, there is no basis – short of sheer harassment – for sanctioning those who have operated within their constitutional rights for not having undertaken the substantial record-keeping burdens prior to this time. Indeed, the regulatory history shows that this Department itself has sought to impose ever more expansive record-keeping requirements which it has felt necessary to phase in as of different dates. It is fair to suggest that this Department should “have its act together” before it expects compliance with novel and burdensome record-keeping requirements. The final rule should thus adopt a single, prospective effective date no earlier than 30 days after publication of the final rule, and the Department should avoid retrospective investigations and prosecutions.

A. No obligations should be imposed respecting images first made prior
to the effective date.

There is, of course, a potential distinction between the date on which newly enacted legislation attains the force of law and the date of the earliest matters affected by the new legislation. Ordinarily, the second of these dates is no earlier than the first, for very strong reasons. While Congress departed from this very sound ordinary practice by some ten years in originally enacting Section 2257 and, again, by some 28 days in amending it (although this appears inadvertent). It is quite clear, for both due process and free expression reasons, that the statute and regulations cannot now require that records be generated and maintained for images which were initially created when there was no valid, applicable record-keeping requirement. Indeed, this is the very reason why this Department has felt it necessary to phase in new and additional regulatory record-keeping requirements. This observation prompts one final comment concerning the effective date of the final rule: Sections 75.1(c)(1) and (2) and 75.2(a) of the proposed rule would add language designed to reach images in other than printed or analog form. 69 Fed. Reg. 35551. This is not itself a problem. But the proposed section could be read to impose record-keeping requirements when, in 2006, an individual digitizes an image from a photographic postcard print original photographed in France in 1923. A “digital image” would thus be “made” after the effective date of the statute and regulations. But the required records all properly apply to the initial image-making event, when the performers were present and their ages at issue. In this case there would be no hope of constructing the required records, and the proposed rule would thus prevent the republication of presumptively protected material in the public domain. There is no sound policy reason for regulating changes in the media on which an old image is stored. The final rule should make clear that no record-keeping or related obligations arise at all with respect to any image first made (i.e. actually formed from contemporaneous live conduct by actual performers) prior to the effective date, regardless of whether such image is later converted to a different expressive or storage medium.

II. The definition of “producer” is unwarranted by the statute and burdens too many people.

The question of the proper scope of the regulatory definition of “producer” has, of course, already been litigated and decided contrary to the position perpetuated in the proposed rule. Sundance Associates, Inc. Reno, 139 F.3d 804 (10th Cir. 1998). There can be little doubt that, as a matter of statutory and regulatory construction, that case was properly decided. Even though the judgment in that case was entered by only one of several intermediate appellate courts, this Department should avoid the need to relitigate that issue before other federal courts. Specifically, the final rule should delete the words “other than those activities identified in paragraphs (c)(1) and (2) of this section” currently in section 75.1(c)(4)(iii) in order to make the regulatory definition of “producer” consistent with the amended statutory definition of “produces,” Compare 18 U.S.C. § 2257(h)(3), with 28 C.F.R. § 75.1(c)(4)(iii). Such a change would render subsections 75.1(c)(4)(iv) and (v) of the proposed rule superfluous and would dramatically reduce the actual scope of subsections 75.1(c)(2) and (3) of the current and proposed rules. Such a change is necessary, though, because the foregoing regulatory language purports to narrow the scope of a statutory proviso. This would be improper in any event, but it is especially egregious where, as here, Congress deliberately added the statutory proviso in an effort to “restor[e]” Section 2257 by saving it from unconstitutionality. P.L. 101-647 sec. 311, 104 Stat. 4817. To the extent that the Department perceives any conflict between the original language of the statutory definition of “produces” and the subsequently added proviso, the latter must control. The final rule should also delete the term “assembles” from section 75.1(c)(2) since neither it nor any form of that word appears in the statutory definition of “produces.”

A. There should be only one primary producer.

Beyond the foregoing problems raised in the Sundance case, the regulations pose serious problems and substantial burdens by defining “producer” so generally that there may be more than one producer of a given image; indeed even more than one “primary producer.” 28 C.F.R. § 75.1(c)(1). The proposed rule seriously exacerbates this problem in at least two ways.

First, by multiplying the types of activities which will qualify one as a primary producer, section 75.1(c) (1) of the proposed rule makes it all but inevitable that most images will have more than one individual primary producer. 69 Fed. Reg. 35551. While the current rule lists three actions which will render a person a primary producer, that list is composed of alternative actions, i.e. any given image is likely to have been filmed or videotaped or photographed but (assuming that the terms are semantically distinct) not two or three of those things. In contrast, it is not at all unusual that a computer image was first photographed and then digitized, each action accomplished by different individuals. The same problem arises with respect to the proposed term “creates a computer generated image” since that term must refer to actual, not virtual, persons, see Ashcroft v. Free Speech Coalition, 535 U.S. 234 (2002), and those actual persons were probably first filmed, videotaped, or photographed.

Second, section 75.1(c), and (c)(5) of the proposed rule, read straightforwardly, would provide that a corporate producer employing an individual primary producer is itself a primary producer along with any parent and subsidiary corporations. Thus where a corporation employs a photographer and a digitizer and has a) a parent corporation which is a stock holding company investing in many varied business having nothing to do with entertainment and b) two subsidiary corporations one of which rents stage lighting to mainstream Hollywood productions and the other of which provides security services for special events, there are six primary producers! Whatever the intent of proposed section 75.1(c)(5), 69 Fed. Reg. 35551, it should be narrowed to avoid this result.

This multiplicity of primary producers is especially problematic because each must collect and maintain original records. Section 75.2(b)’s permissive provision concerning copies of records applies only to secondary producers. The final rule should carefully provide that there is only one primary producer per image. In keeping with the statutory proviso, that might be either the actual, initial image-maker or the agent who arranges for the performers’ participation, and the final rule should clearly spell out which one will be the primary producer in any common situation. Where this individual is employed, or perhaps otherwise retained, by a corporation or other business entity, it may make perfect sense to declare that that entity, not the individual, is the primary producer. Again, the final rule should clearly specify when this is so, and it should expressly relieve the employed individual(s) of primary producer responsibilities. It is utterly senseless and extremely burdensome to have more than one primary producer for any given image.

B. No “secondary producer” should be required to maintain extensive records.

Even beyond the Sundance issues concerning the permissible scope of a regulation implementing a statute, it is important to understand why the definition of secondary producer is so problematic. Given the presence of the (improper) limiting language in the present section 75.1(c)(4)(iii), the secondary producer definition is already very broad, and it already includes persons (and presumably corporations and their parents and subsidiaries) who have no hope of actually collecting and assembling record information directly. 28 C.F.R. § 75.1(c)(2). The additions to that subsection in the proposed rule would make matters even worse, by again multiplying individual and corporate secondary producers even for a single republication event. 69 Fed. Reg. 35551. If a previously existing photographic print is turned into a digital image by one person, inserted on a computer site by another, which site is managed by a third, and each is employed by the corporation envisioned in the previous subsection of these comments, then there are now seven more secondary producers arising out of a single republication!

And if that original photographic print had been in the public domain, it could, absent the proposed section 75.1(c)(2), have been freely posted on the computer site without any contact with the primary producer(s) or the performers. But the proposed rule would prevent this constitutionally protected republication apart from the unlikely happenstance that the computer site personnel can track down and obtain information from the primary producer or from each and every performer. To be sure, the disclosure statement might be of considerable help in locating the primary producer(s), but he, she, it and/or they need not permit inspection of any required records by anyone other than an “investigator” under the statute and the proposed rule.
Even beyond actually limiting some otherwise lawful and constitutionally protected republication, the secondary producer provisions very seriously burden all of it. This burden arises from the “record-shifting” requirements. This concept and these requirements are discussed in considerable detail in section IV of these comments. There it is established how these record-shifting requirements very seriously burden and chill constitutionally protected expression. There it is also demonstrated that far less restrictive means are available to achieve the statute’s legitimate purposes. In advance of detailing them, it is asserted that all of those comments apply here because the concept and definition of “secondary producer” vastly expand the number of individuals and entities who bear burdensome record-keeping duties and thus necessitate the record-shifting requirements in the first place. Indeed, this seems to be the practical motivation behind objections of the type raised in Sundance. It might well be that if a secondary producer were required to do nothing more than record and maintain the address of the individual or entity from whom the secondary obtained the right to republish the images and continue to use the primary producer’s disclosure statement, there would be few, if any, who would object to the regulatory provisions even if the technical Sundance defect had not been fully and formally overcome.

For all of the foregoing reasons, the final rule should require no more than that secondary producers record the address of the primary producer, or other copyright holder, where permission to republish was necessary. The final rule should also, for these and other reasons detailed in other sections, provide that third-party custodians may maintain original records at their own places of business on behalf of primary and secondary producers simultaneously.

III. The record-keeping requirements are unclear and are too burdensome.

The proposed rule contemplates record-keeping procedures which seem to be designed with a large producer of magazines or videotapes/DVDs in mind. Even there, the proposed rule is often too vague or too burdensome. But since a substantial effort seems to have been made in drafting the proposed rule to cover digital images on computer sites, it is surprising and disappointing that the record-keeping provisions seem to ignore the very different circumstances under which these producers must generate and maintain records. The final rule should thus take account of the following difficulties arising from the language of the proposed rule.

A. Records should be acceptable if they are in digital form instead of on paper.

At several points, the proposed rule seems to assume that the required records will be prepared and maintained as hard copies, i.e. printed on paper. See, e.g., §§ 75.1(a)(3), (e), 75.5(e), 69 Fed. Reg. 35551, 35553. This model might well suffice for a small magazine producer with a large file cabinet and three sets of index cards. But when it comes to computer sites, there are hundreds if not thousands of sites in this country each of which have many thousands (some have hundreds of thousands) of images, separately produced or gathered in small galleries which have been separately produced. Unless a good deal of generalization (not contemplated by the proposed rule) were permitted, these computer sites face truly awesome record-keeping burdens. Those burdens might remain acceptable if the required records could be generated, maintained, and inspected in digital form, but it is altogether unrealistic to expect that such records be printed onto paper in their entirety and then physically sorted and cross-indexed. The operators of these large computer sites are obviously computer savvy, and there is no reason not to permit them to use computers to reduce their record-keeping burdens to more manageable proportions should they choose to do so. The final rule should thus expressly permit records to be maintained in digital form. Such maintenance should be straightforward with the possible exception of the images of the identity documents, cf. § 75.2(a)(1), 69 Fed. Reg. 35552. Here, the final rule should expressly permit such documents to be scanned into digital form at a resolution reasonably attainable by producers using commer cially available consumer grade equipment.

B. The requirements for a picture identification card are confusing and inappropriate.

The proposed rule would alter the requirements for a valid “picture identification card.” § 75.1(b), 69 Fed. Reg. 35551. The reasons for the proposed changes are not at all apparent, and this Department suggested no reason upon publication of the proposed rule. It is not altogether clear why some of the most common picture identification cards, such as employer cards – now regularly handled with considerable attention to employers’ current very substantial security concerns – should not suffice. Other than a driver’s license, an employer or school identification card is likely to be the only picture identification card a typical performer has. And the potential problem with driver’s licenses, of course, is that they generally list the person’s home address. Since performers in widely disseminated sexually oriented productions often have a well founded fear of stalkers, requiring that their driver’s licenses be photocopied and that copies be maintained for years and disseminated to unknown secondary producers might well chill a reasonable performer’s willingness to appear in a sexually explicit image, thus limiting constitutionally protected expression. And this chill seems altogether unnecessary. There is apparently no valid governmental reason for recording and maintaining performer address information, since the proposed regulation does not require it. Thus no governmental purpose justifies the chill which is imposed when address recording is effectively required by an unreasonable limitation on acceptable picture identification documents.

In addition, the proposed requirement that the picture identification document “provide[] sufficient specific information that it can be accessed from the issuing authority” is altogether unclear. In the first place, accessed by whom? Does this proposed change support a suggestion that a primary producer now has a duty to check up on all identification documents with “the issuing authority”? If not, what is the purpose of the requirement? If a picture identification card contains a name and discloses an issuing authority, this Department and its investigators no doubt have the wherewithal to check its authenticity, so the requirement seems entirely superfluous in this regard. Moreover, how is a primary producer to know whether a picture identification card meets this requirement when the regulatory formulation is left so unclear?

The final rule should thus clarify the “sufficient specific information” requirement or abandon it. The final rule should also permit the use of school and employer picture identification cards or, at the very least, expressly permit address information on a driver’s license or similar state identification card to be obscured in the filed image.

C. It is extraordinarily burdensome to require that the records contain
e) a copy of each work.

The proposed rule would add a requirement that the records include a copy of each image. § 75.2(a)(1) (i), 69 Fed. Reg. 35552. Again, this requirement may make sense for a magazine producer whose press run typically numbers in the thousands or tens of thousands of copies. Here, the marginal cost of throwing one of those copies in the file cabinet in the proper folder may seem perfectly reasonable. And although the physical storage requirements might be greater for the videotape producer, the task still seems imaginable, even over the required maintenance period. But again, the final rule should fully and fairly consider the large computer sites as well. There are thousands if not tens of thousands of computer sites in this country each of which posts thousands and sometimes hundreds of thousands of images. Moreover, there is substantial commercial pressure on these sites to post new images on a regular basis. So even assuming still images of small size, mediocre resolution, and substantial compression, each image will take several tens of kilobytes of computer storage. Under these limiting assumptions, the largest sites must devote many gigabytes of computer storage to their current images. These current storage undertakings are economically and technologically feasible even when larger files – still images of higher resolution and better color depth or motion picture images of any length and resolution – are involved. But matters quickly become much less tractable when cumulative storage requirements are considered for the required record maintenance periods.

One entity reports 45,000 new images created for website distribution monthly. If a copy of each of the 45,000 images were required to be printed and placed in physical file cabinets, the recordkeeping burden becomes virtually unmanageable.

The following situation is not at all uncommon (i.e. many dozens or hundreds of computer sites in this country are similar to what is described here): a computer site offers to its willing adult customers sexually explicit but nonobscene live streaming video feeds on twenty channels, 24 hours per day, seven days per week. Even assuming substantial compression and mediocre image size and resolution, each of these video streams is likely to generate something on the order of five or ten megabytes each minute. The twenty streaming video channels might then generate something like 150 megabytes per minute. To be sure, this requires fast processors and communications channels of wide bandwidth, but, so far, the digital storage requirements are not particularly daunting, because the computer site narrowcasts, but does not record, the live streaming video. Individual audience members may or may not record the performances for later replay, but if they do, they use their own digital storage resources for the purpose. But the proposed rule would require that all of these streaming images be captured and recorded on site so that they can be maintained as part of the required records. Now the storage requirements quickly become unfathomable. Since there are 1,440 minutes in each and every day, one day’s streaming images from this twenty channel site will require over 200 gigabytes of storage. And since there are 2,556 days in the shortest minimum required storage period, it will take over 500 terabytes (i.e. 5 . 1014 bytes) of storage to maintain these images even if old images are removed from the required records as soon as legally possible. This storage requirement is in addition to whatever would be required for the other images – still or motion picture – offered on the site over the years. This would be a truly staggering storage requirement even if my file size estimates were high by one or two orders of magnitude! Matters are certainly no better if the records are required to be kept in “hard copy” rather than in digital storage. Even assuming that the streaming video recorded at the lowest standard VHS video resolution would be acceptable to this Department, each streaming video channel would fill four standard video cassettes per day. At that rate, the twenty channels would generate over 200,000 separate video cassettes over the shortest record maintenance period. This is certainly no less staggering than the digital storage requirements. It is difficult to imagine what legitimate governmental purpose could require such burdensome storage, especially since it is quite likely that not one of the dozens or hundreds of customers who may have seen a particular live streaming video on a particular day would have saved that image for anything like seven years. Furthermore, it bears noting that, even if the cumulative storage requirements were more objectively feasible, any storage devoted to seven year old images is storage which cannot support current expression, so even more realistic requirements in this regard will substantially burden or limit much expression. The final rule should operate to impose only realistic storage requirements.

The final rule should also take account of the possible reoccurrence of a very rare but devastating event. An underage performer might present identification documents which genuinely deceive many reasonable primary and secondary producers, as well as fellow performers, into believing that the performer is of proper age. This performer could then use those identification documents to work on many “shoots,” so that over even a short period of time, a very large number of this performer’s images would be published and republished. Following the proposed rule, a copy of each of these images would properly find its way into the records of the many primary and secondary producers involved. Now, one way or another, this performer’s fraud is revealed. Those who were innocently duped nevertheless absorb the cost of destroying all of the circulating and warehoused images now know to be illegal and unprotected by the constitution. Retailers and other distributors conscientiously do the same thing. But what of the images which are required in the records? Can they be destroyed or not? If not, how are the custodians to be protected against state laws which typically, and constitutionally, Osborne v. State of Ohio, 495 U.S. 103 (1990), criminalize even the private possession of child pornography? Any final rule should expressly address and resolve these questions.

D. The proposed rule is too burdensome in creating multiple, duplicate records.

Assuming that the new regulations are intended to cover the streaming video performances and/or video-conferencing described above, the regulations appear to require that a new record be created each time a performer works a new shift. Since in most cases the performer is not a contractual employee, but works at will, there is no predictable means to determine the period of employment in advance. If a new set of records, declarations and inspection of identification is required for each work shift, the paperwork creation and storage problems are virtually insurmountable. As described in the earlier hypothetical, a producer of twenty channels operating 24 hours daily will typically have 80 performers daily (that is four 6 hour shifts, for each of the twenty channels. If each day eighty new records must be created, checked and indexed, the recordkeeping will likely require at least one fulltime employee to create, check and maintain the records. Furthermore, the physical space necessary for the recordkeeping would be overwhelming. Lastly, the generation of such voluminous records, the vast majority of which will be perfectly duplicative of records generated each and every previous day, provides the Department with no added benefit, other than the increased likelihood of misfilings resulting in felony convictions and five year sentences for persons who have created all the records as defined by statute and regulation necessary to establish that all performers are of age.

A different problem exists for record creation involving individuals employed to engage in videoconferencing from remote locations. That is, after initially inspecting the original identification documents in person, there is no reason for the videoconferencing to occur from one central location. Any person with the appropriate computer and access to the internet could perform from home. If new records must be created for each shift, a performer must work in a location where the producer can reinspect the identification and observe the completion of the 2257 declaration. This serves no legitimate government purpose, but is an anomaly of the apparent requirement of daily, duplicative record creation.

E. The proposed rule confuses the indexing requirements for the records.

In a number of places, the proposed rule threatens to confuse the cross-indexing requirements, perhaps as an unintended consequence of its inartful use of the term “URL.” That term is quite properly and understandably defined by the proposed rule, § 75.1(h), 69 Fed. Reg. 35551. It then appears to be misused as a species or example of expressive “matter.” § 75.2(a)(2), (3), 69 Fed. Reg. 35552. The proposed rule even speaks, at one point of “a copy of any URL.” § 75.2(a)(1)(ii), 69 Fed. Reg. 35552. A “uniform resource locator” is an alphanumeric symbol which is mapped to an IP address in order to serve as a mnemonic device. It is thus akin to a name for an internet node. Such a node might, of course, be occupied by a computer containing images and other files, and the URL might also contain directory information for such files. But the ULR itself is not a digital file of any sort, let alone an image or any other sort of expressive matter. The term’s use as such in sections § 75.2(a)(2) and (3) is confusing at best. One who understands precisely what a ULR is could read section 75.2(a)(3), for instance, to require that images or works posted to the internet be indexed by URL as well as the name or identifying number of the work. That is, a list of URL’s would serve as a fourth cross-indexing list. If this is what is intended, it does not appear to be required by the applicable sections. §§ 75.2(d), 75.3, 69 Fed. Reg. 35552. The final rule should clarify this point one way or the other and take care to confine technical terms to their respective technical senses.

F. The record segregation requirement is unduly harsh.

In the most abstract sense, perhaps, one can sympathize with this Department in worrying that a records custodian might delight in keeping perfectly arranged and cross-referenced records amidst so many other items that this Department’s investigators will need to spend an inordinate amount of time, thrice yearly, cf. § 75.5(d), finding the required records in order to examine them. Whether this possibility is a likely one is open to serious doubt, since it would do nothing more than simply prolong the inspections. Perhaps one can imagine a lonely records custodian. In any event, the proposed rule’s “cure” here may well be far out of proportion to any imaginable disease. § 75.2(e), 69 Fed. Reg. 35552. The basic performer information required by the statute and regulations will invariable track materials in a producer’s personnel files quite closely. Under the proposed rule, if a stray model release, 1099 form, or I-9 form winds up in the 2257 records instead of in the more general personnel file, the custodian or the producer faces five years’ incarceration. If the problem contemplated by section 75.2(e) is worth worrying about at all, a much kinder, gentler solution is called for in the case of inadvertent misfilings.

G. The proposed rule requires that the records be maintained for too long.

Assuming for present purposes that the recent amendment of Section 2257(d)(2), P.L. 108_21 sec. 511(a), 117 Stat. 684, did not create serious constitutional difficulties, cf. Message from President, H.R. Doc. 100-129 at 66-68, the reason for fixing the basic record maintenance period at seven years, as proposed § 75.4, 69 Fed. Reg. 35552, is fairly obvious, although the Meese Commission, which originally formulated the idea for Section 2257, suggested only five, 1 Attorney General’s Commission on Pornography, Final Report 621 (1985). Any fixed period is an improvement over the provisions of the current rule, 28 C.F.R. §75.4 (“as long as the producer remains in business”). There is no principled objection to defining the maintenance period with reference to a statute of limitations running from the initial production of an image. Nevertheless, there remain several problems with the maintenance period prescribed by the proposed rule. First, in connection with the proposal concerning the frequency of inspections, § 75.5(d), 69 Fed. Reg. 35553, the seven year period seems quite excessive (or, more likely, the routine inspections much too frequent). The proposed rule as a whole would require the primary producer to maintain the basic records concerning each image or work through twenty or twenty one inspection cycles. I can think of no reason why this Department’s investigators would need that many inspections to assess any particular record. If the records themselves disclose that a performer had been underage, that fact will appear as soon as the record is first seen. If there is reason to doubt the authenticity of a recorded identification document, far less than seven years would be required to check it out. Second, the proposed rule may or may not contemplate culling seven year old information out of an ongoing cross-referenced set of records. If it does not, it is really no change from the present rule, and it poses very substantial cumulative storage requirements. The final rule should make it clear that the records concerning an image or work can be disposed of seven years after the image’s creation and that a producer’s records concerning a performer can be disposed of seven years after the performer is last imaged by the producer.

This raises two additional problems with the proposed rule. One is that it is drafted in such a fashion that the basic seven-year maintenance period could turn into an almost seventeen year period, which would plainly be excessive by any stretch of the imagination. If a corporate producer goes out of the business of producing sexually explicit images – say, to provide security services for special events instead – on the day before the basic seven-year period runs, it is required to keep the records for an additional five years. If the corporation is then dissolved the day before that five-year period runs, the individual custodian is required to keep the records for another five-year period. The final rule should make it clear, using “whichever occurs first” language, that the operative period is the shortest of whichever of these three possibilities obtains in fact.

Of far more basic concern is the apparent assumption underlying the proposed rule that secondary producers will be required to maintain records for seven years after they reproduce an image. If a secondary producer reproduces an image one day short of seven years after it was first made, the records concerning that image or work will have to be maintained for just shy of fourteen years, and this process could be extended indefinitely. At first blush, this might seem entirely reasonable. But as I have already observed, see section I-D, the information in the required records relates to the initial production, not to the reproduction, so there really is no good reason to start the maintenance clock running anew upon each republication. Any rule requiring records to be maintained as long as an image is in circulation would be so cumulatively burdensome as to unconstitutionally throttle expression. And consider the foregoing situation changed by just two days: If a secondary producer decides to reproduce an image one day after seven years after it was first produced, how would the proposed rule operate? The primary producer has now lawfully disposed of the original records, so he, she, it and/or they cannot transfer copies thereof to the secondary producer under section 75.2(b). May the secondary reproduce the image at all without the records or is further publication to be restricted, at very substantial constitutional cost, to the primary (who is also now quite lawfully without the records)? The upshot of these observations is, I think, that no one should be required to keep records concerning a particular image more than seven years after the image was first created. Similarly, no producer – primary or secondary – should be required to keep records concerning a particular performer for more than seven years after that performer last performed to create any image for any work published or republished by that producer.

IV. The record-shifting requirements are impermissibly burdensome.
As suggested in section II-B of these comments, a great deal of the burden on expression imposed by the regulations at issue results from the “record-shifting” requirements. That is, the regulations expressly require that copies of the full set of required records must follow any image to any secondary producer, broadly defined, who assists in disseminating the constitutionally protected expression. § 75.2 (a), (b), 69 Fed. Reg. 35551-52. That is akin to saying, first, that no one may write a newspaper article critical of a public figure without first recording written proof that the article is not unprotected defamation. Such a requirement would be problematic enough, but it would be made dramatically worse if that very elaborate written proof had to be transferred to others before they could republish the article. Such a requirement would dramatically increase both the number of individuals and entities burdened by the novel and arcane record-keeping requirements as well as the overall costs, to publishers and to society, imposed by those requirements. And so it is here. These record-shifting requirements will chill the willingness of others to redistribute images, and they will thereby restrict the dissemination of constitutionally protected expression.

A. The record-shifting requirements seriously burden and chill
the dissemination of constitutionally protected expression.
This Department has quite properly recognized that secondary producers cannot be required to assemble their own records from “scratch.” It has thus provided in section 75.2(b) that secondary producers may satisfy their own record-keeping requirements by receiving copies of a primary producer’s records so long as the secondary producer also obtains, records, and maintains the primary’s business address. 28 C.F.R. § 75.2(b). That is, the primary producer generates and then “shifts” copies of records to each secondary producer for each republication event.

This solution to the secondary producer’s otherwise intractable record-making problem is certainly no panacea. For one thing, the volume and complexity of those records will very likely chill the willingness of many to republish material, thus limiting the reach of constitutionally protected expression. Where a primary producer licenses some but not all of his, her, its and/or their images to a secondary producer, for instance, it will likely be daunting work to untangle the cross-references so that the secondary producer receives only the required records, cf. § 75.2 (e), 69 Fed. Reg. 35552 (subjecting anyone with extraneous matter in the required records to five years’ incarceration). The records of the primary producer often comprise invaluable business secrets; i.e., the ability to contact talent by means of the information contained in the 2257 disclosures can be critical for a producer. Sharing this information with existing competitors, or would-be competitors, which meet the definition of “secondary producer” could undermine the free enterprise system we all so highly value. Far more importantly, many performers might agree to part with identity information to a primary producer whom they know and trust, but (since the required records will likely contain a copy of a driver’s license or other identity document containing a home address) those same performers may well refuse to allow such records to go to unknown secondaries who may openly publish their identities, steal their identities, attempt blackmail, or worse: performers in sexually explicit works have well-founded fears of stalkers. In this situation, a performer either will not perform or will insist on a binding agreement that the primary producer will never use a secondary to republish. There is at least one documented case of identity theft of an actress by a producer based exclusively of the required 2257 documents and disclosures. This problem would be severely aggravated by increasing by several magnitudes the number of persons maintaining these private records, especially persons whom the actors have no knowledge and no professional relationship. In all of these cases, constitutionally protected expression may well be chilled, if not frozen out of existence.
These serious burdens and chills must be evaluated in light of the governmental purpose in generating them. Presumably, this Department seeks to avoid a situation where investigators have to trace back through a long line of secondary producers to find the primary in order to examine the original records. It is difficult, however, to see how such a situation would arise or needs to arise. First of all, if there were no such thing as a secondary producer at all, each republisher would have to include the primary producer’s disclosure statement on every republished copy. 18 U.S.C. 2257(f)(4). So an investigator obtaining any copy would immediately and directly know where to find the primary producer. It is easier to find this from a glance at a disclosure statement than from an inspection of the secondary producer’s records. It is also possible that this Department reasons that the more individuals and entities required to keep the records, the more likely it will be that someone actually follows the law and keeps them. The same reasoning, of course, would support requiring that copies of the records follow all distribution of an image or work. That way, this Department’s investigators could expect to inspect records in the hands of anyone possessing the image or work, even an ultimate retail consumer. But this reasoning merely compounds a serious constitutional problem underlying the entire statutory scheme: it deliberately and seriously burdens the constitutionally protected expression of the law abiding on the assumption that others will violate applicable laws.

B. Substantially less restrictive regulatory alternatives are available.
It is important to recognize that the entire record-shifting difficulty (together with important problems concerning disclosure statements, raised in section VI-A of these comments) arises from the requirement that records be maintained at the producer’s own place of business. If other locations were permitted – if, for instance, third-party agents were permitted to act as records custodians at their own places of business, properly disclosed by the required statement – the record-shifting problem would be resolved: both primary and secondary producers could rely on the same records in the hands of the same third party custodian using the same disclosure statement. To be sure, this Department’s investigators would have fewer choices of places to visit, but, as noted at the end of the previous section of these comments, that is only because fewer lawful producers, acting within their constitutional rights, would be burdened by the regulatory scheme. The record-shifting burdens could be eliminated entirely and the record-keeping burdens could be minimized by concentrating those burdens on third-parties who are ready, willing, and technically able to receive the required basic information from primary producers and then organize it, maintain it, and make it available for proper inspection. Under such a system, this Department’s investigators might become more familiar with these dedicated, professional custodians, and they might develop relationships with them – perhaps including some training requirements and a competence showing – which might better insure proper cross-referencing and full observation of other peculiar record-keeping details.
This Department is entirely free to implement such a system, as the one suggested in the preceding paragraph, by regulation. The statute itself expressly provides that the required records be maintained at the producer’s “business premises, or at such other place as the Attorney General may by regulation prescribe . . ..” Cf. 18 U.S.C. § 2257(c) (emphasis added). The business office of a competent third-party custodian is certainly an available regulatory option. It may be that a very small producer – such as one of the many thousands of individuals operating very small, personal websites – would prefer to hire a competent third-party record-keeper rather than to undertake the task of learning how to cross-reference the records and then personally arranging the records maintaining them for a long time. This would also eliminate the problem of having a lawful home-based internet business be subjected to inspections in private residences, but also disclose that home address to the entire world. This reduction of privacy has a chilling effect on constitutionally protected expression. It would also be possible for producers to effectively share competent custodians. That is, dedicated records custodians, could provide record-keeping services for many small producers at once, thereby allowing the custodians to make a full-time business out of it at a known address with fixed hours and without the need to divide time between maintaining records and producing expression. Whether such a system would result in many individuals seeking to do business as records custodians or in a very few large and highly computerized operations dominating the “records custodian” market is, perhaps, worth careful thought. In any event, industry groups, such a those representing adult websites and the producers of widely distributed videotapes and DVDs, might well be interested in providing seed resources to establish training programs and to produce compliant record-keeping software packages for dedicated records custodians.
Under such a system as proposed here, this Department’s investigators would still know where to find the required records, but many of the chilling burdens imposed upon producers would be eliminated altogether, and the remaining burdens could be concentrated upon those competent individuals or entities willing to undertake them voluntarily. All that would be necessary to permit such a system is for section 75.4 of the final rule to provide that the required records may be maintained at a competent third-party custodian’s place of business. Such a change would provide for professional third-party custodians who could not only better serve this Department and its investigators, but could also serve the needs of the thousands or tens of thousands of very small primary producers in this country by performing the arcane record cross-referencing and maintenance functions, and by avoiding the need for these individual primary producers to risk using their personal names and their only address on the disclosure statements. The overall burdens imposed by the current and proposed regulatory scheme, especially upon small producers, by the substantial record-keeping and record-shifting requirements would be eased dramatically if they could lawfully be transferred to those who are willing to specialize in the required services and to provide those services to many producers at once. Under such a system, in short, primary producers would be left free to produce (provided they check identification and copy identification documents and send the required information along to their custodians), secondary producers would be left free to republish (provided they include the primary’s disclosure statement), and custodians would be free to specialize in the business of cross-referencing, maintaining, and permitting proper inspection of the required records.

V. The inspection provisions are in some respects improper and in other respects incomplete.

It is evident that the drafters of the proposed rule devoted considerable attention to the contemplated inspection process. In one sense, at least, this attention is welcome. It is certainly better that potential questions surrounding the inspection process be addressed in advance than that they be left to arise as conflicts during individual inspections. Nevertheless, section 75.5 of the proposed rule, 69 Fed. Reg. 35552-53, contains provisions which are inappropriate and, in some respects, it leaves out other provisions which would be appropriate. The final rule should address the following difficulties.

A. The final rule should fully and properly define “investigators.”

The first and most basic problem with section 75.5(a) of the proposed rule, 69 Fed. Reg. 35552, is that it assumes that the statute authorizes the Attorney General to designate inspectors other than him- or herself. It does not. Under the statute, a producer has an obligation to make the required records available only to the Attorney General. 18 U.S.C. § 2257(c). In the very same clause imposing this obligation, Congress expressly left it to the Attorney General to alter or expand other provisions of the statute. 18 U.S.C. § 2257(c) (“at his business premises, or at such other place as the Attorney General may by regulation prescribe”). As the common lawyers of old used to say: expressio unius est exclusio alterius. Black’s Law Dictionary at 692 (Rev. 4th Ed. 1968). Even assuming that this was an oversight, it should be remedied by statutory amendment rather than by a regulation exceeding statutory authority. Indeed, amendment of section 2257 would be welcome if the process at last included committee hearings which would finally permit an opportunity, not yet afforded to anyone, to the tens of thousands of affected persons to comment on the statutory scheme, to indicate the burdens it imposes, and to suggest less restrictive and less burdensome alternatives.

Even assuming that this Department intends to authorize investigators other than the Attorney General absent statutory amendment, the proposed rule leaves the matter entirely too unclear. The ideal way of designating such investigators would be “by regulation,” as the parallel provision of Section 2257(c) suggests. That would require that the final rule designate who the authorized “inspectors” are not of course, by name, but by designating a class of persons (e.g. attorneys employed by CEOS or FBI agents assigned to a particular unit) as authorized investigators. In the alternative, the final rule could provide for some mechanism a search warrant is what comes most readily to mind – by which the Attorney General’s designation is reduced to a verifiable writing. How else is a custodian to know whether to answer the door when a stranger comes knocking claiming to be a designated inspector? Cf. § 75.5(c)(2)(i), 69 Fed. Reg. 35552 (“credentials”). This is not an abstract question. Although Section 2257 does not currently prohibit records custodians from freely disseminating the required records to the world at large (it should surely be amended to do so), a conscientious records custodian will realize the important policy reasons behind limiting inspections to those absolutely required by statute and by valid implementing regulation.

B. Advance notice of inspections should be provided.

At first blush, it is perhaps not surprising that section 75.5(b) of the proposed rule, 69 Fed. Reg. 35552, expressly provides that no advance notice of records inspection will be provided. Law enforcement officers are understandably trained that surprise is an important part of most inspections because it deprives suspected wrongdoers of the opportunity to destroy or to conceal evidence of illegality. But even a moment’s reflection will indicate that this principle does not operate here. This Department’s inspectors are not, as in the case of a “drug bust,” looking for items that should not be present (and so could be hidden or destroyed). Here, they are looking to see that items which must be present and in order are, in fact, just that. Even in the very unlikely case of a record which actually indicated that a performer had been underage, the records custodian could not avoid a problem simply by destroying that record: that would still leave a required record missing, also an unlawful situation. And even if a producer keeps somewhat messy records after six and one-half years but uses advance notice of an inspection to put them in proper order, absolutely no legitimate purpose underlying the record-keeping requirement would be undermined or disserved in any way. Finally, advance notice would serve to limit the occasions on which an investigator is put to the choice between knocking down a locked door – because no one happens to be present on the premises at the time the investigator arrives – or chalk the attempted inspection up to wasted time and effort. The final rule should expressly specify precisely what will happen if no one is present when an inspector arrives. Section 75.5(b) of the final rule should also flatly reverse the proposed course and provide for reasonable advance notice of routine inspections.

C. The proposed rule excessively defines “reasonable times.”

Section 75.5(c)(1) of the proposed rule, 69 Fed. Reg. 35552, is altogether unrealistic in defining reasonable times. Few offices of any kind are regularly staffed for public access from 8:00 a.m. to 6:00 p.m., local time, even on every weekday, let alone on a Sunday. The Department of Justice itself publicizes no offices which are so staffed. To be sure, emergency operations such as police stations and firehouses may be so staffed at these hours and more, and some retail locations relying on customer traffic for business may be open for such hours – though, even here, seven days per week and all legal holidays would be unusual. These last examples, though, are hardly analogous to a producer’s business premises. The ten-hours-per-day, seven-days-per-week-even-on-holidays expectation is particularly unrealistic as it applies to the thousands or tens of thousands of small producers in this country, such as those who operate small or personal websites. Not only would it be difficult for these (typically) one-person operations to staff an office for all reasonable times contemplated by the proposed rule if this website were their only endeavor in life, but these individuals also very likely work full time at other jobs, making it very difficult to staff their relevant place of business consistently even during more reasonable business hours.

Moreover, proposed section 75.5(c)(1)’s additional provision concerning “reasonable times” will operate with consequences which, whether or not intended by this Department, are altogether unnecessary and constitutionally unacceptable. If a producer is actually engaging in a photo shoot or if live models are posing and acting for live streaming internet video at 2:30 a.m, local time, the proposed rule would permit inspections at that time. There are at least two problems with this proposal. First, even assuming that the production work in question is occurring on the business premises, this is the situation in which inspection is most likely to interrupt production, since spare personnel, let alone the designated records custodian, are unlikely to be present at those hours. More importantly, perhaps, the proposed rule would permit inspections at the place of business even if the late-night/early-morning shoot is “on location” at some distant place. These problems seem especially unnecessary since such late-night/early-morning inspections could have nothing to do with the on-going production because some time is surely required before all of the cross-indexing is done, and because it may not even be clear which performers on the set will actually appear in any published image until the editing is complete. On the other hand, if there is probable cause to believe that an underage performer is actually participating in an on-going off-hours’ production, the courts of this country are well equipped to issue warrants on child pornography or child abuse charges without the requirement for any late-night records inspection at all.

Section 75.5(c)(1) of the final rule should specify much more realistic “reasonable times” for regularly-staffed places of business, and it will also have to deal with the very difficult issue of inspection timing where a producer has an entirely separate full-time job elsewhere. It is difficult to envision how this latter set of problems can be dealt with absent adoption of some of the other proposals advanced here, e.g. permitting record maintenance at a third-party custodian’s place of business or providing for advanced notice of inspections.

D. The proposed rule allows unnecessarily frequent
and insufficiently justified inspections.

As previously noted in section III-F of these comments, section 75.5(d) of the proposed rule, 69 Fed. Reg. 35553, together with section 75.4 thereof, would permit twenty or twenty-one sequential inspections of a given record. This seems very plainly excessive and unreasonably burdensome. But in general, it is difficult to evaluate proposed section 75.5(d)’s frequency provision as a whole. There may be production operations which are so large and so prolific that a four-month schedule for routine inspections would not be particularly unreasonable. Some such production operations may actually exist. Similarly, if third-party professional custodians are permitted, some of them may well have a sufficient client base to warrant such frequent routine inspections. But there are hundreds or thousands of production operations which are so small and so static that the required records are unlikely to change at all over any particular four-month period. Routine reinspection every four months for the sole purpose of repeatedly reviewing the same records is unwarranted. Presumably, the Department’s inspectors will realize that some record sets warrant more frequent routine inspection than others. Since emergency inspection is always available in any case, non-routine inspections should always require probable cause and a search warrant. Any inspection will be at least somewhat burdensome, and for the vast majority of cases, the proposed inspection frequency amounts to an invitation for harassment.

E. The proposed rule is unclear concerning photocopying.

It is not at all clear how proposed section 75.5(e), 69 Fed. Reg. 35553, is supposed to work in practice, even if the required records were kept on paper and index cards. Many small producers may not have a photocopier on the premises and, even if they did, the rule fails to specify how an investigator would reimburse the producer for its use. Does the proposed rule contemplate that this Department’s investigators will bring portable photocopiers to the inspections with them? If so, this is another respect in which the entire proposed record-keeping scheme is ridiculously expensive, except that here the ridiculous expense falls upon the American taxpayers. Even assuming that CEOS, or some other agency of this Department, has already budgeted for a fleet of portable photocopiers, proposed rule 75.5 (e) fails to address the situation where the required records are kept in digital form – unless, of course, these portable photocopiers also operate as a universal printers and optical/magnetic media readers. The final rule should clarify what this Department has in mind here. As the proposed rule stands, it is truly difficult to figure out what that is.

F. The proposed rule should expressly clarify the right to remain silent.

Since section 75.5(c)(4) of the proposed rule goes to all of the trouble of specifying what the investigator may say at the conclusion of an inspection, and since it goes even further to permit the producer to speak to the investigator at that time, perhaps the final rule ought to take the trouble to also specify that no one need say anything at all to an investigator or answer any questions during an inspection, i.e. that the authority to search does not include the authority to require that questions be answered. It should also specify that anyone and everyone on the premises is free to leave before the commencement of or at any time during a records inspection. If this Department thinks that there would be anything wrong with the producer and the records custodian leaving the premises during an inspection, leaving behind only an uninformed security guard to watch the investigator, then the final rule should say so and include the constitutional safeguards appropriate for custodial interrogation situations.

G. The proposed rule should address the seizure of the required records.

It is apparent that many peace officers and magistrates have forgotten the strict specificity requirements and other procedures which are constitutionally required when search warrants and other seizures affect expressive materials. It has happened more than once recently that, upon executing a search warrant concerning obscenity or child pornography, peace officers have seized original records which the statute and regulations currently require be kept at a producers place of business. If left to their own devices, these peace officers would typically hold these records for many, many months – certainly more than the four-month routine inspection period, cf. § 75.5 (d), 69 Fed. Reg. 35553. The final rule should expressly address this situation in a number of particulars. First, this Department should consider legislation forbidding anyone other than a custodian or one of this Department’s investigators from moving, disturbing, or interfering with the required records in any way. Local peace officers rarely know what these records are, and their routine search and inventory procedures will very likely destroy the required cross-referencing and may very well otherwise disturb the integrity of the records. Second, and far more importantly, the final rule should make it clear that the seizure – and, for that matter, the theft – of the some or all of required records does not require the cessation of any on-going or previously planned expression. If seizure of the required records did have that effect, then the seizing authority would, of course, be required to provide copies of such records back to the producer within 24 hours so that expression could promptly resume. Cf. Heller v. State of New York, 413 U.S. 483, 492-93 n. 10 (1973). In the alternative, the final rule should provide that the seizing authority thereupon becomes the lawful custodian of the records and is required to provide the producer with an individual custodian’s name and street address so that proper disclosure statements can be affixed to any new copies produced. These alternatives may be amusing to contemplate, but the problem is serious and real. The final rule should address it.

VI. The disclosure statement requirements are confusing and, in many cases, too burdensome.

The proposed rule’s requirements concerning the contents and location of the disclosure statement largely retain the requirements articulated by the current rule, making some additional effort to adapt them to expression posted on computer sites. §§ 75.6, 75.8, 69 Fed. Reg. 35553. The proposed re-organization of section 75.8 into subsections is welcome, as it will be of some help to producers and custodians, who may not be legally sophisticated. Beyond this, the proposed rule both retains and adds provisions which are confusing at best, ridiculous at times, and unconstitutional in very many situations. The final rule should address the following difficulties.

A. Requiring the open disclosure of the custodian’s name and address
is impermissible in many situations.

As with many of the record-keeping requirements, the present and proposed requirements concerning the contents of the disclosure statement, § 75.6(b)(3), (c), 69 Fed. Reg. 35553, seem to have been drafted with a few large producers of magazines or videotapes in mind. For such producers – invariably incorporated and operating large warehouse and, perhaps, production facilities – most of the current and some of the proposed requirements may be perfectly acceptable. Even for these producers, the proposed rule is entirely unclear concerning which or how many producers must be named if, as recognized in section II of these comments, there is more than one primary or more than one secondary producer. Corporate parents and subsidiaries, § 75.1(c)(5), 69 Fed. Reg. 35551, for instance, will generally not share the same business address. Moreover, if there is no legitimate need to keep records with respect to a work beyond seven years after the images it contains were first created, see section III-F of these comments, there is certainly no need to continue printing any disclosure statement at all after that date. The final rule should clarify which producer should used for the disclosure statement, and it should expressly “sunset” the disclosure statement requirement when the relevant records may be discarded.

One might assume that the same general statements could be said about the operators of at least a few of the largest adult websites. But even there, looks may be deceiving. The internet, and particularly the world wide web, has made it possible for individuals of even modest means and business sophistication to publish expression which rivals that of the largest corporations in its design sophistication. Thus, even the largest adult mega-websites may be operated by a single individual or by a very small group of individuals organized in the loosest manner imaginable. But whatever the picture concerning the largest sites, it is certain that there are thousands or even tens of thousands of very small adult world wide web operations, many operated by a single individual out of his or her residence, in full conformity with local ordinances. In terms of sheer numbers, in fact, these producers overwhelmingly predominate over large, well-organized corporate producers operating from fixed places of business. So it is fair say that the vast majority of the producers covered by the statute and regulations will be very small concerns, some operating as nothing more than sole proprietorships or informal partnerships.

With respect to these very small producers, the requirement that the disclosure statement – which will be, in any event, available for all the world to see – include a street address and the name of an individual records custodian imposes a substantial burden and a very serious chilling effect on constitutionally protected expression. While a large corporation may well be able to search far and wide to find a highly-paid records custodian willing to see his or her name plastered on thousands of items of sexually explicit expression, such a possibility lies far beyond the reach of the typical small producer. Many thousands of internet operations run by one or a few individuals have no business address apart from the producer’s residential address. So in many thousands of cases, the rule effectively requires the disclosure of this address to the world. Similar observations apply to the name of an organization’s records custodian. Not only do these provisions interfere with constitutionally protected rights to anonymous speech, McIntyre v. Ohio Elections Commission, 514 U.S. 334 (1995); Brown v. Socialist Workers ’74 Campaign Committee, 459 U.S. 87 (1982); Talley v. State of California, 362 U.S. 60 (1960), they expose producers – who in small operations are very often also performers – to the very real dangers of identity theft, blackmail, and stalking. And these dangers undoubtedly chill constitutionally protected expression.

As with the record-keeping and especially the record-shifting burdens, the burdens and chills imposed by the disclosure requirements could be substantially reduced – or eliminated entirely – simply by permitting third-party records custodians to maintain records at their own business premises. While it is unrealistic to assume that each small producer, some living in small, rural hamlets, will be able to find a custodian willing to a) be named publicly and b) actually do business at the producer’s address, it is not at all unlikely that there are individuals willing to be so named somewhere in this country. If those individuals are permitted to market their record maintenance services to many small producers at once, they could easily afford to establish their operations at their own, secure business addresses, which would reduce the “stalking” and related concerns to the level experienced by the largest magazine or video producers. Again, an incredible amount of constitutionally significant burden arises from the baseless requirement that the required records not be separated from the producers’ business premises. The final rule should alter that requirement.

B. The requirements concerning the placement
of the disclosure statement are confusing.

The provisions of the current and proposed rules concerning the placement of a disclosure statement within motion pictures, § 75.8(b), (c), 69 Fed. Reg. 35553, on any medium (i.e. film, videotape, and presumably also DVDs, and internet files such as .avi, .mpg, or streaming video), are unnecessarily confusing, and will pose traps for the only moderately wary. While the contrast between subsections (b) and (c) is a nice one – in the very old-fashioned sense of “nice” – there is simply too much at stake to maintain it. There are fairly sophisticated producers who include end credits in their productions but who assumed for a long time that their disclosure statement was required where all of the other video producers put it: at the beginning. With five years’ imprisonment at stake, unnecessarily Byzantine requirements have no place in any final rule. The final rule should choose between subsections (b) and (c), and make clear that it applies to all moving images which have a beginning and an end.
While attention in the final rule to the placement of the disclosure statement on computer sites, § 75.8(d), 69 Fed. Reg. 35553, is welcome, the proposed rule is too simplistic. If the disclosure statement were likely to be a one or two line affair in these cases, the proposed rule would suffice. But almost every adult website is a composite work, consisting of thousands or sometimes hundreds of thousands of images organized into separately produced galleries. The disclosure statement in such a case could remain simple only if a great deal of generality were permitted. If the final rule adopts such a strategy, it should plainly spell out the permitted generalizations. If, on the other hand, a separate disclosure statement, or a separate line in a disclosure table, is to be required for each separately produced work, then it is absolutely unrealistic to require that all of this be placed on the front or main web page. Again, it seems that such a requirement is altogether unnecessary. One of the distinguishing features of the hyper-text mark-up language (HTML) is its ability to support intricately linked documents. It is thus perfectly possible for moderately sophisticated webmasters to link their complex disclosure statements, or disclosure tables, together in a coherent fashion – enabling this Department’s investigators to follow the links simply by “clicking” on them – without compromising the artistic design or literary integrity of their websites. But for this possibility to satisfy the regulations, the final rule would have to permit the front or main page to display an appropriately labeled link, which would then open to perhaps several pages of disclosure statements or to an elaborate table of disclosure statements. A webmaster who wished to do so could use a series of links to keep individual disclosure statements close to the galleries to which they belong, while at the same time saving this Department’s investigators the trouble of sifting through all of those galleries. The final rule should choose between the possibilities outlined here, and perhaps those suggested by others, but as it stands the proposed rule falls far short of the required mark.
Finally, it is worth noting that most digital image file formats offer the capability of including a disclosure statement in a way which avoids interfering with the aesthetic integrity of the image itself. Where, for instance, someone is posting many thousands of altogether unrelated images on a computer site, the proposed rule is silent about where the many thousands of disclosure statements should go. Under some internet and related protocols, the poster would face complicated choices in including the disclosure statements. The final rule should avoid a requirement that the statement appear, in such a case, on the face of each image, both because such a requirement would unnecessarily increase the size of the image files and because it would interfere with the integrity of the image itself. Again, a solution is readily at hand. Most if not all image file formats provide for comments – which do not affect the displayed image at all – to be placed in the digital file. The final rule should provide, as an option, that the disclosure statement can be placed in such a comment field, at a defined location, where the producer will know where to put it and where this Department’s investigators will know where to find it. Indeed, such a placement in a motion picture file will make for a smaller file, since it is much more efficient to store the disclosure statement as ASCII characters in a comment field than as thirty seconds or so of displayed video. There is no legitimate governmental interest which would forbid such “buried” disclosure statements.

C. The regulations concerning the date of production are inappropriate or incomplete.

Both the present and the proposed rule exceed the statutory mandate by requiring that a disclosure statement specify the date of production. Compare 18 U.S.C. § 2257(e)(1) with § 75.6(b)(2), 69 Fed. Reg. 35553. Perhaps for this reason, the regulations have never been particularly clear about what date is contemplated. Cf. 28 C.F.R. § 75.6(a)(2). Even assuming that the statute required a date of production in the disclosure statement, there is no reason why that date should be anything other than the date on which the images in the work were first recorded from live conduct by the performers. This is the only date which matters for child pornography purposes and, as previously noted, see Section I-D of these comments, all of the required records properly relate to that event alone. Similarly, that date, or something very close to it, will determine when the record-keeping obligations expire and so when the disclosure statement no longer serves any useful purpose at all. Again, it would be easy to preserve the initial production date by preserving the initial disclosure statement; which in turn would be easy to do if the required records, once generated, could stay in one place for the required seven years, such as in the hands of a third-party records custodian.

Also requiring clarification is the date of production for a production which transpires over several days. That is, some filming (approximated at over 30% of adult motion pictures) takes more than one day to complete. If the filming of a motion picture begins on December 30, 2003, and is completed on January 2, 2004, what is the “date of production?”

There are different theories which could result in different definitions of “date of production” under such circumstances. The regulations must address this unambiguously.

D. The typographic requirements for the disclosure statement are absurd.

The proposed rule makes an understandable effort to specify the appearance of the disclosure statement. § 75.6(e), 69 Fed. Reg. 35553. Unfortunately that effort, as it stands, falls flat, resulting in truly absurd requirements. The first requirement is that the disclosure statement appear in the same typeface and type size as the largest of the performers’ names. Assume that the video box cover for, say, “Annette Ginger’s Greatest Hits” contains images covered by the statute, 18 U.S.C. § 2257(h)(1). The box cover itself would then, of course, have to include a disclosure statement. It is not at all unreasonable to expect that the Annette Ginger’s name would appear in 60-point type or larger. Especially since a “greatest hits” tape is likely to be a compilation of separately produced scenes, the rather elaborate disclosure statement simply could not fit on the rest of the box. Other provisions of the proposed rule also pose problems for images displayed on a computer site. Since the size and appearance of any text posted on a computer site will depend on the settings of each remote computer monitor displaying it, the proposed rule’s 11-point type minimum size requirement is meaningless. While HTML provides for a standardized set of text headings for world wide web pages, distinguished by their relative size and weight, they too will appear differently on different monitors. Reference to such standardized headings for websites is preferable, however, since it would allow this Department’s investigators to configure their own monitors to avoid eyestrain. Finally, the requirement that the disclosure statement appear in black type on a white, untinted background is overkill at best. It would also have the effect of attracting the ultimate consumer’s attention to the disclosure statement, aggravating the stalking concerns of small producer/performers. While this Department’s concern here is understandable, there seems little reason to forecast that a requirement that the disclosure statement appear in the same typeface and size and against the same background as the copyright statement would not prevent the microscopic blue-on-dark-blue disclosure statements which this Department seems to fear.

E. Provisions governing an “exemption statement” are unnecessary and improper.

The regulatory provisions concerning a so-called “exemption statement,” § 75.7, 69 Fed. Reg. 35553, are unnecessary, unauthorized, and, in any event, seriously under-inclusive. If, as will be the case very many situations, no disclosure statement is required, then no statement of any related kind is required either. The absence of an exemption statement will not relieve the government of its burden of proof on any element of an alleged offense under Section 2257(f)(4). If it did, serious problems would arise concerning the validity of the regulation. Since section 75.7 can and should be ignored altogether, it is not particularly important that it is far from comprehensive in listing the circumstances under which no disclosure statement can be required. To be sure, images created prior to the effective date, as I have used that term here, are subject to no record-keeping requirements and need no disclosure statement. But the same is true for later-created images after the record-keeping period has expired for them. And the same is also true, unless the statute is read to impose serious and constitutionally problematic obstacles to the importation of images, of images created in foreign lands by foreign producers using foreign performers. Finally, neither the record-keeping nor the disclosure statement requirements can constitutionally apply to amateur videotaped images made, for instance, by a married couple of themselves for their own personal use. Since there will surely be this and similar “privacy-type” exemptions – whether or not the statute and regulations recognize them – this is another aspect in which section 75.7 under-enumerates the exemption circumstances. The final rule should abandon all provisions concerning an exemption statement.

F. The regulation should specify that “transfer” is limited to commercial transfer.

For some of the reasons articulated in the preceding subsection of these comments, the final rule should also specify whether the term “transfer,” as used in subsection (f)(4) of the statute is, like the term “distribute,” 28 C.F.R. § 75.1(d), limited to commercial transfers. Since the term appears in a statutory context which might suggest otherwise, 18 U.S.C. 2257(f)(4) (“sell or otherwise transfer”), and since extravagant claims have been advanced for the reach of the statute, Message from President, H.R. Doc. 100-129 at 60, the final rule should specify whether a disclosure statement is required for purely private expression. If, for instance, one of the spouses in the married couple envisioned in the preceding subsection of these comments sent their videotape to the other spouse who was traveling on a long business trip, would the statute and regulations require a disclosure statement? Section 75.1(d) of the regulations is on the right track in suggesting – to the extent that it actually does so – that the entire record-keeping scheme is restricted to commercial production operations. Any contrary suggestion will, of course, seriously exacerbate the constitutional problems surrounding both the statute and the regulations.

VII. The proposed rule contains some minor grammatical errors and drafting problems

While the notice-and-comment process in informal administrative rulemaking is undoubtedly most beneficial when it leads to the consideration of practical and policy concerns which the agency has overlooked, such as those raised here, there is no reason why the process should not also be used to correct grammatical and drafting errors in proposed rules. Indeed, Congress has always been so hasty and sloppy in promulgating and amending Section 2257 that typographical errors remain in the statute to this day, even after an effort at technical correction. P.L. 103-322 sec. 330004(14), 108 Stat. 2142. But the same need not be true of the regulations issued thereunder.

A. The final rule should correct some minor grammatical errors.

In section 75.1(a) of the proposed rule, 69 Fed. Reg. 35551, the word “technologies” is either out of place or in improper form. Definitions and defined terms simply do not include or exclude technologies. The drafters may have intended something like “technological,” but this would inexplicably contrast with the “commonlanguage guidance” previously mentioned. The problem is not a major one, but the final rule should clearly state whatever this Department means here.

Section 75.1(f) of the proposed rule, 69 Fed. Reg. 35551, contains a grammatical error, albeit a common one. “Myriad of” is simply not proper English.

B. The final rule should avoid a troublesome drafting error.

Section 75.1(d) of the proposed rule contains, 69 Fed. Reg. 35551, a common drafting error which can be the source of unnecessary confusion and, therefore, unnecessary litigation. After restricting certain defined terms to a “commercial” meaning, the section then expressly states that the terms do not embrace any non-commercial meaning whatsoever. While that second recitation is essentially redundant, there is nothing wrong with it; and the final rule should retain it. The problem arises because the proviso then proceeds to list examples of what the defined terms do not mean. But placing a restriction on a listed example does not re-expand the meaning of the defined terms themselves. It is thus pointless to add the term “bonafide” to the preexisting term “lending libraries,” since the defined terms will not reach even sham lending libraries so long as they are non-commercial. If this Department believes otherwise, it should clearly say so in the final rule and face the constitutional consequences.

Respectfully Submitted,

The Free Speech Coalition