[This post originally appeared in the MusicTechPolicy Monthly Newsletter before the announcement of the MLC vendors.]
Title I of the Music Modernization Act is clear about a few things, one of which is the creation of the “musical works database.” That database is for a specific purpose-it is to be used by the mechanical licensing collective in fulfilling its statutory mandate to “collect and distribute royalties from digital music providers for covered activities” under the new blanket license. Among other things, the MLC is also mandated to “[a]dminister a process by which copyright owners can claim ownership of musical works (and shares of musical works)”.
But the MMA also establishes a cohort of “special” people who can get <em>copies of the entire database</em>, a “bulk copy.” And there’s the conflict of interest rub. Not to mention the data integrity rub because as you’ll see, the last thing the Congress wants is to have inconsistent “official” databases floating around the ether, each claiming to be authoritative and more authoritative than the next. (Without, of course, undermining competition for accuracy or efficiency with the “official” database.)
Authority of the Copyright Office to Create Database Conflict of Interest Rules
As the legislative history teaches us, the Congress is taking a critical look at how the musical works database is in fact utilized. Congress is also onto the old database arbitrage rope a dope: “Music metadata has more often been seen as a competitive advantage for the party that controls the database, rather than as a resource for building an industry on.” (H.R. Rep. No. 115-651 and S. Rep. No. 115-339, at 8.)
Unsurprisingly, Congress also empowered the Copyright Office with the authority to establish regulations regarding “usage restrictions of the musical works database.” The Copyright Office could take up that issue in the current rulemaking. As the Copyright Office noted in the notification of inquiry for MMA regulations, “[t]he Office seeks public input on any issues that should be considered relating to the oversight of the MLC, including but not limited to conflicts of interest….” 84 Fed. Reg. 49966 (No. 185 Sept. 24, 2019).
The musical works database is to be used for a specific purpose to further the permitted work of the MLC and to be a national asset that is the definitive resource for copyright ownership of songs. Congress did not intend for the database to be a leveraged byproduct of Title I that would allow special people special privileges. The integrity of both the data and the database is of great importance to Congress and to the Copyright Office.
It is also clear that nothing in the MMA grants to anyone any ownership right to the musical works database or any data acquired in its establishment, not MLC, DLC or anyone receiving a copy. Neither does MMA grant any right to sell or resell that data. The Copyright Office clearly has an oversight role for any sales, a role the Congress takes seriously based on the legislative history as well as a host of public statements by Members and Senators.
So it will be entirely within the brief of the Copyright Office to establish rules to protect the public against both threats to database integrity and efforts at database arbitrage. (In the eminently readable MMA, you will find this at 17 USC Sec. 115(d)(ii)(D)(bb)(E)(v) “Accessibility of Database” for those reading along.)
Who Can Get a Copy of the Entire Musical Works Database?
Title I requires that when established, the musical works database “shall be made available to members of the public in a searchable, online format, free of charge.” Users of the blanket license, among others, are to get a free copy of the database in a “bulk, machine-readable format”. This language appears to draw a distinction between members of the public looking things up onesey and twosey compared to getting a copy of the entire musical works database.
Congress clearly was drawing a distinction between casual users in the public and those whose needs relate to the operation of the blanket licensing system-users who are already paying for the creation of the database through the administrative assessment and don’t want to pay twice.
Fine so far, although given the siloed nature of reporting and accounting systems at the music services it remains to be seen just how much any of them would actually use a copy of the musical works database unless that “machine-readable format” was highly customized. The cost of that customization would seem a cost more appropriately borne by the entity requesting the copy rather than socialized across all licensees.
The Vendor Conflict
But-Title I goes further. And here is where the need for the Copyright Office to develop conflict of interest rules becomes acute.
“Authorized vendors” of users of the blanket license are also entitled to a copy of the database free of charge, but anyone not a user or a vendor of a user can pay for a copy of the database. That’s consistent as far as it goes because vendors of users are working for someone who pays for the MLC under the administrative assessment (and indirectly pays for the establishment and maintenance of the database) and others do not. Again, given that those vendors run existing systems that already talk to the systems of the licensees, the jury is out on just how many will want or be able to use this “free” copy as a practical matter. Even so, what should not happen is that the data standard sinks to the lowest common denominator to accommodate a vendor’s legacy systems (#nofoxpro #nodos).
However, Congress drew a sensible distinction between (1) those who pay the administrative assessment or their vendors and (2) those who do not. Congress is clearly making that distinction to fairly match benefits with burdens.
What Congress did not intend was for vendors of users to acquire a free copy of the musical works database that they then could use for their own purposes unrelated to the work these vendors performed for users under the blanket license that justified their free copy in the first place.
It also seems clear that if vendors of the MLC are to obtain their own copy of the database they would be treated like anyone else. They would be included in the group that must pay a fee for a copy of the musical works database “not to exceed the marginal cost to the mechanical licensing collective of providing the database to such person or entity”. Congress could easily have given those vendors a free copy as they did with the vendors of users but didn’t. (It seems that the situation may actually be going the other way–vendors are selling their data to the MLC.)
Extracting Data from Indie Labels
It is important to note that Title I also empowers users of the blanket license to obtain from indie labels all their song information, or use “[g]ood-faith, commercially reasonable efforts” to do so. Do such efforts include suing to get the data? Must the label bear the cost of providing the data? Does this include updates? Does the label warrant the accuracy of their data when used for a purpose for which it was not collected? If the licensee must use “commercial reasonable efforts” to do so, does that imply that the label need not provide the data if it is not commercially reasonable to do so? Is charging a market price reflective of the cost of the data to the label as well as updates commercially reasonable? Is this another “taking,” this time from labels?
Congress clearly states that anyone purchasing a bulk copy of the musical works database must pay the marginal cost to the MLC. But who pays the labels for the cost of the label data that is to be included in the musical works database and then sold by the MLC? Do the labels get compensated? At what price?
Clearly the Congress did not intend to set up a free rider issue here, so this is another area that is ripe for regulation.
Reselling the Musical Works Database
There is nothing in the statute that allows those obtaining a copy of the musical works database to resell that database or services derived from that database, including when those services do not relate to the purpose for which the database copy was obtained. Neither does MMA permit the subsequent distribution, alteration, sale or reproduction of a permitted copy, i.e., all such copies must originate with the MLC as required by the statute. In fact, Congress addresses the issue by allowing the Copyright Office to restrict such arbitrage.
Neither does the MMA restrict how a copy of the musical works database once purchased is to be valued on the books of the buyer, or whether it can be transferred to a new owner in the sale of that business or pledged as collateral. Again, important issues for the Copyright Office.
Issue Spotting the Conflicts
The transfer back and forth of all of this data seems to create exactly the kind of moral hazard that cries out for conflict of interest rules of the road to accomplish the proper goals of Title I. This would, of course, have to prohibit co-mingling of existing data of a vendor with data acquired during the normal course of that vendor’s services. This is easy to accomplish through the use of a clean room and is a best practice and common procedure in the tech industry (e.g., former employees with knowledge of competitor IP or business practices).
Clear conflict of interest rules would assure Congress that their mandate is not creating unintended enrichment of any particular vendor capturing data while working for either the MLC or users of the blanket license.
One might say that song data acquired by the MLC is just information that was essentially public anyway (we’ve heard that kind of argument before) so why restrict it. I would suggest that argument is a bit of a dodge because even if true, the vendors didn’t in fact get the data through public means as available to anyone else as it was to them. They got it because the statute required the information be given up in order to enjoy rights of a copyright owner-not to create an arbitrage opportunity. All the more reason why conflict of interest rules are essential.
And of course such regulations would give the public comfort that the entire supply chain had proper oversight and transparency as well as a corrective remedy should anything go astray.
Therefore, it is entirely within the mandate of the Copyright Office to establish conflict of interest and data integrity rules of the road for all concerned.
Conflict of interest rules would address and require:
-Notification of the Copyright Office of each copy of the musical works database distributed.
-The Copyright Office could easily publish that notification in the Federal Register for transparency. Any transfer documents should be recorded with the Copyright Office and made available to the public for search
-“Flow down” language from the Copyright Office that must be included preemptively in any click through-agreements to insulate users
-Disclosure of pre-existing data held by vendors rendering services
-Disclosure of data submitted by rights owners (including labels) to the MLC that is made available by the MLC to its vendors
-Establishing a fee schedule of sums to be paid to rights owners (including labels) providing data
-Prohibitions on pledging the musical works database as collateral
-A mechanism for copyright owners (including labels) to opt out of being included in any copies of the musical works database transferred to vendors or sold to others
-Establish restrictions on ownership of the musical works database (which might come in handy if it is necessary for the Copyright Office to designate a new MLC)
-Clarifying the Copyright Office FOIA policy as it relates to the database
-Prohibiting co-mingling pre-existing data with copies received
-Prohibiting modification of permitted copies of the database (for the obvious reason of maintaining the integrity of the MLC as the authoritative source)
-Requiring all copies of the database be watermarked with origination and destination
All these guideposts are compassed by the Copyright Office mandate and should not require any amendments to the MMA. If these issues remain unaddressed, I fear a real mess consistent with that old legal principle snafu fugazi fubar.