Room 204 Christopher Buccafusco (w/ Joseph Blocher), Firearms, Innovation, and Regulation
How do law and markets affect the pace and direction of
innovation for firearm related safety in the US? Costly inefficiencies in
supply and demand.
Virtually no one thinks the US has the right amount of gun
violence, and the firearms industry has been and continues to be enormously
innovative—AR-15 customization, ghost guns. But innovations to make the guns
themselves safer have failed to appear. E.g., smart guns: thousands of people
are killed/injured each year by guns fired by someone other than the owner. 1
of 6 police officers shot are shot by an officer’s own gun. 250,000 guns stolen
annually and disproportionately used in crime. Shootings by children, suicides,
etc. 2016 survey says 59% of Americans would be willing to buy a smart gun,
including 56% of political conservatives and 4 of 14 gun owners.
Since the 1990s, manufacturers have been producing
functional prototypes for user authentication of handguns, shotguns, and other
long guns—facial ID/biometrics, codes, etc. But none reached the market despite
millions in R&D from National Institute of Justice. Colt and Smith &
Wesson had functional prototypes by 2000, followed by dozens of startups. But
only last week did Biofire offer the first public sale of a smart gun.
VCs: Liberal, not interested in funding guns; more
interested in software than hardware.
Demand: no buy-in from institutional investors like police
forces; some purchasers are deeply hostile to smart guns b/c they fear gov’t
will come for all other guns. Partly in response to NJ’s 2002 law—once there’s
a smart gun, manufacturers have to switch to it w/in 30 months, though NJ
backed off and just required retailers to stock it, but still infuriated gun
rights advocates who boycotted Colt and Smith & Wesson who then got out of
the market entirely. Established firms pulled out of the market, so startups
can’t expect to be acquired and must go all the way to direct sales; there’s
also fear of tort liability. Biofire isn’t submitting for registration in NJ to
avoid triggering law.
Microstamping: tech that imprints a gun’s serial number on
discharged rounds. 2007: Cal. required firms to included microstamping once DOJ
certified that there was no patent on it; challenged under Bruen and dct
overturned the law b/c it prevented people from buying state of the art
handguns/not consistent w/historical tradition which was light on
microstamping. Cal. isn’t even appealing the loss on microstamping.
Limits on internal innovation created opportunities for
external innovation on safety (which is probably bad)—e.g., installation of
Shot Spotter all over. $21 million in Louisiana to “harden” schools, half a
billion in Texas; schools designed with curving walls to decrease damage done
by active shooters; bulletproof backpacks.
Sometimes external environment-level innovation is more
efficient: ramps v. stair climing wheelchairs. But many external innovations
come with huge social costs.
We don’t have solutions; everything sucks. Maybe: subsidies
for smart gun purchases? Institutional commitments from institutions like
police forces?
Q: Europe?
A: some innovation does come from Europe. Is there demand in
these other countries?
Betsy Rosenblatt: one story seems to be that
innovation-forcing laws can inhibit that. Is that unique to this field?
A: No—seen something similar in disability as well. Tort
tries to get people to make safer products to encourage innovation, but also
leads to anxieties about creating new stuff b/c it risks litigation v. doing
what everyone else is already doing. Boycotts are inhibiting Biofire—they have
had to go fully direct to consumer b/c dealers won’t stock them for fear of
boycotts.
Q: federalism story?
A: this is part of the challenge—innovation folks usually
don’t have to think about public law and state v. federal.
Q: what about the military, which seems to be missing from
this story? Probably a big buyer w/an interest in having still-lethal weapons
that limit friendly fire and suicide. Where are they?
A: we are looking for an answer. They were very interested
very early with Colt and Smith & Wesson. These are professionals who know
how to handle guns, so the safety needs may seem less pressing, though that’s
probably wrong. It’s become so politicized, and military is disproportionately “gun
rights” folks. But senior leadership could decide to prioritize safety (maybe
with negotiating not to trigger the NJ law).
Q: VCs are less liberal than people think, and politics take
a back seat to economic opportunities, so why not?
A: the big problem is exit—Colt and Smith & Wesson don’t
want you; you have to be Tesla and go to production. But the story that many of
these people are telling is “we can’t get funding.” Maybe VCs don’t believe in
the tech; the story: there’s money for mental health, victims, and school
hardening, but people don’t want to be in the gun industry.
RT: [So one factor that might be silent here is
outsourcing/contracting: b/c the military no longer makes its own stuff and
seems institutionally incapable of imagining that it might, it is dependent on
outside contractors, and if they won’t do it, too bad. That seems bad for
reasons beyond guns.]
Mark Schultz, Video on Demand Services: New Frontiers in
Regulation of Cultural Policy, Industrial Policy & Copyright
Streaming took off, and regulation was close behind, motivated by concerns
about culture. New cultural policy in Australia, 2023: fear of voices being
drowned out. Unlike free over the air TV, no requirements to make Australian
content available.
Rising wave of interventionist cultural policies: European
AV Services Directive, Australia, Canada have passed regulation and others are
on the way. Argument: likely to fail both as cultural and industrial policy.
Building on other work on cultural and economic policy by Pager, Park &
Messerlin.
A better way: decentralized policies that promote local
capabilities—the strong preference for local content can be competitive. Korean
success story.
Two models of cultural policy/industrial policies. First,
interventionist: on the content side, subsidies (French film industry), content
requirements, language requirements. Industrial policy: local content quotas;
local investment of profits (streamers must invest percentage of revenue, not
profits, locally); local production requirements for location, personnel, financing;
terms of trade (retention of copyright, exclusivity limits—streamer can only
have license for limited amount of time, investment limits on how much
streamers can invest in local companies); import quotas; screen quotas;
prominence requirements (local content must rise in search results).
Second, decentralized—there is no country that is purely
noninterventionist; everybody does something. But broadly, market based, focused
on private investment, content neutral; largely hands-off except for granting ©.
Some countries like Korea invest in building creative skills and technical
skills, building studios and other infrastructure; tax breaks; promotion,
marketing, and other related capability policies.
EU AV Services Directive requires streamers to include at
least 30% “European” content. Permits member states to require re-investment of
streaming revenue locally. France has required 20-25% reinvestment, Italy
considering similar marks. Regulating terms of trade also permitted, including ©
ownership/exclusivity/restricting investment in local productions.
Canada, Bill C-11 passed. Regulations in progress: local
content, if similar to broadcast will be 35-50%. Local production requirements:
not enough to film it here. Must have Canadian producer making decisions; a
point requirement where you get points for, say, screenwriter, which leads to certain
market distortions.
Hasn’t worked well as cultural policy and thus fails as
industrial policy. Sean Pager’s work: as France increased subsidies, its share
of its own box office relative to American share went down. The argument has
been that the French had incentives to create content based on guaranteed
subsidies so there was no incentive to create material that was appealing to
audiences, especially in comparison to American films. The subsidy trap: the
bureaucrat is your audience, which leads to a certain type of filmmaking (not Scorcese
or Spike Lee; willing to take bureaucrat’s suggestions). Censorship isn’t the
main problem—even when the cultural bureaucracy is insulated from politics, the
office culture has its own office politics and may not be interested in what’s
appealing to the local public. There are also cronyism and quota problems: when
you have a quota, people may take advantage of that to make quickies on the
cheap w/low production values.
The Emily in Paris problem: The dodge where you make the
content in the country but not for the country. The Falcon & the Winter Soldier—set
in Eastern Europe but about who is the right person to have Captain America’s
shield.
Distorting local investment: local filmmaker complains it
makes it harder for locals to compete. Streamers are paying local actors more
than local producers—maybe that’s good but it does divert from local-inspired
content. Can pigeonhole locals/block them from opportunities. Margaret Atwood:
book by a Canadian, filmed in Canada, but it didn’t count b/c scriptwriters
weren’t Canadian.
Korea as success story: Language unique to Korean peninsula;
relatively unique culture. First swept through Southeast Asia, Japan, China;
then US, Latin America, Europe. $12 billion/year in exports, plus soft
power/tourism.
Decided in 1993 to focus on culture as strategic sector. Indirect
support: tax credits and incentives for private investment, including
micro-investment; pushed chaebols to be involved, which they were until the
financial crisis when most spun off those parts. Direct support focused on infrastructure
and human capital: production facilities, training, export promotion. This is
the model with the fewest unintended consequences. It’s lowbrow/mass culture,
sure. Most countries do mild subsidies to preserve certain forms of culture;
but media sector shouldn’t be dependent on those subsidies, and ultimately time
tells what it is highbrow or lowbrow.
To avoid the mistakes of interventionist policies of the
past, national governments should promote cultural industry capabilities, but
avoid picking winners in ways that make creative industries complacent.
RT: I find this convincing but I’m interested in what a
French bureaucrat would say in response and your answers.
A: France would say: We make great stuff. We make real art. True:
The Francophone Africa film industry has produced beautiful movies, but no
Africans ever saw them: Nollywood is more popular and tells “African” stories. French
might also point to the fact that, in countries where they dropped controls,
American movies flooded in—as in Mexico, where film industry struggled. Mexican
gov’t didn’t do Korean-style policies, though it did promote telenovelas.
Revealed preferences: if people don’t go see it, does it
matter?
Q: why not delegate to experts about what would be good, not
necessarily popular?
A: experts tend to have their own strong preferences. German
cultural bureaucracy would finance either old German operas or really
avant-garde productions.
Rosenblatt: What’s the role of unions?
A: good question—maybe some interaction.
Rachel Landy, The Innovation Void in Downstream Content
Markets
Music: Same product, at same price, from three main
companies in our lives plus Spotify. Live online TV has more variation in
price, channel options, etc. What about a $4.99 monthly for nothing but
catalog, no playlists? What about a jazz service or a metal service? Record
industry’s role in suppressing innovation. Labels’ conditions prevent
innovation. High concentration—3 dominant labels with must-have catalogues.
Each can veto an entire business. They are complementary oligopolists: Cournot
complements—you get even more market power and leverage. Coupled w/desperation
to get back to pre-digital levels of control. You see ratesetting and other key
license provisions.
Each label enters into an agreement “independently” with
each service, but there are standard terms. A large up-front minimum guarantee
payment, often in the hundreds of millions. There’s a revenue share for recoupment
against minimum guaranteed; the revenue shares have most favored nation
principles, which allows them to know how the other labels are pricing. Labels
keep any overage, known as “breakage,” and it’s unclear how much if any is
shared.
Super-narrow © license. Services have to get permission from
each label for any new feature or functionality that invokes the catalog, and
again this facilitates information sharing.
Trust: repeat players, reputational sanctions, reciprocity—deter
innovative options. Result: higher end-user prices. Barriers to entry, and less
innovation by incumbents b/c so much is being extracted: 55% of revenue. Also harms
indie artists; incentive to promote major label content to recoup the
guarantee. The labels have seen their own costs go down—no pressing records;
the services pay for the infrastructure.
Solutions? Consent decree frameworks; statutory licenses;
antitrust law reform against tacit collusion; MFN clause ban. Incentives to
defect? We could tax the breakage that can’t be tied back to any content on the
service. Could tax the surplus made through the MFN; might encourage labels to
drop guarantees to a level where they could actually be recouped. Transparency of
parallel contractual provisions might also help.
Tang: There’s more innovation than you say in music—you’re
only talking about premium streaming services, but not iHeartRadio, free
Pandora, Amazon Prime bundling.
A: for webcasting, there’s a statutory license with some
protections built in though they also limit innovation. Those are also
controlled by the labels and subject to more restrictions than the premium
services are b/c the labels want to funnel people to the paid subscription.
There are other parts of the industry where innovation is flourishing—where the
labels can’t do this—TikTok or YouTube UGC.
Kristelia Garcia: the tax thing is intriguing—are there
similar examples?
A: tax as a tool we often look to for encouraging/discouraging
behavior; cigarette and other sin taxes. Congress loves to amend tax code and
not so much ©.
Blake Reid, Copyright’s Periphery
Copyright on a dying planet. Looking at 1201 triennial
rulemaking: 15 years of trauma before the Copyright Office discussing far
ranging areas of law and policy: environmental regulation, disability rights,
medical devices—far from concerns of ©. We’re trying to do serious policy and
somehow we’re funneling it through the distorted lens of ©. © routinely infects
policy areas outside its core of incentivizing creative works. The 1201 review
illuminates the problem.
1201 creates paracopyright liability for circumventing tech
protection measures that control access to © works. There’s no protection for
circumvention aimed at noninfringing or fair uses in most circuits.
2021, there were 21 distinct exemptions, including
traditional categories (motion pictures, video games), but a lot of literary
works as computer software. Worth emphasizing that a lot of these exemptions
are not new, but have persisted across multiple rounds of rulemaking. They show
the © periphery.
Intended beneficiaries are often small/individual—film critics,
documentarians, disability services providers, people w/disabilities, farmers,
repair techs; often public-facing, they often care about complying with the law
and require degree of legal certainty; they often produce public goods like
privacy, education, agriculture, data security.
Many of these uses are functional and uncontroversially
noninfringing: functional uses and modifications. Unlocking, jailbreaking
involve only glancing uses of protected works and are incidental to the use:
the fact that you need to use the software on a tractor engine to repair the engine
is just incidental. Facilitates uncontroversial uses like reading that might be
required by other laws like ADA. Or exposes/tests vulnerabilities of TPMs and
software—used to evaluate and diagnose software. Many uses are noninfringing
but not subject of direct case law.
1201 requires Copyright Office, which doesn’t look at fair
use particularly charitably, to determine that uses are likely noninfringing.
B/c these uses and users are public facing and often chilled ex ante, there is
often no case law on point.
Why do rightsholders object to exemptions? The review is
really contentious despite the noninfringing nature. Objection 1: speculative
abuse of exemptions—encourage infringing behavior adjacent to but beyond the
bounds of the exemption. Across decades, no instance of this has been
identified where a bad actor purports to rely on an exemption. Objection 2:
non-© policy motivations for regulating: 1201 is a proxy for other policy
issues like DOJ treating 1201 as belt and suspenders for CFAA defense against
hackers, as if Russian gov’t cares. Concerns about vehicle modification violating
pollution regulations; FDA worrying about medical devices. Explicitly beyond scope
of © and institutional context where Copyright Office isn’t capable of evaluating;
it tries to do so under 1201’s catchall provision. Why is the Librarian of
Congress in charge of deciding which cellphone you can use? It’s supposed to
consult w/NTIA, but routinely rejects NTIA’s recommendation.
Objection 2.5: non-copyright policy interests of TPM
deployers/rightsholders: disclosing security flaws might be embarrassing; right
to repair might allow independents to compete with authorized repair.
Objection 3: non-copyright micromanagement of user/circumventor
activity. That’s a result of a sense of entitlement from exemption opponents to
control how users behave. CO builds a miniature regulatory scheme into the
exemption, e.g. for text and data mining—security practices and accreditation
requirements for researchers.
What could we do to wall off the periphery?
Lower the bar for securing/renewing/expanding exemptions; encode
more in statute.
Cover development of tools.
Eliminate 1201 or require an infringement nexus.
What would bear on copyright more broadly?
We should consider specific exemptions/limitations for categories
of users likely to engage in the production of public goods, knowing their uses
are likely to be especially sensitive to liability risks; specific exemptions
for functional uses; new institutional contexts for assessing fair use ex ante—declaratory
judgment attempt in CASE Act was an unsuccessful but interesting trie.
More muscular policymaking in non-copyright congressional
committees—don’t defer. Even where there are complex fair use and doctrinal
issues, AI is an example where the policy equities are far beyond © and fair
use’s capacity to address, like labor and privacy.
Zahr Said: Why are you conceding that this is on the
periphery? Post-colonial theory makes it feel like a concession.
A: experience of triennial review, which represents a long
dedication from a lot of communities who band together. All that engagement
with the core gets indifference from actors who are at ©’s core. They’ve tried
but there is so much skepticism and distrust for exemption proponents.
Charles Duan: it’s cheap to use DRM and get the value of
excluding people from an entire device. “Cheap Exclusion”—relevant paper. Value
disconnect.
A: I’m skeptical that they’re really getting much value out
of this [Duan and Rosenblatt: They think they do!] Some of their claims are
just ridiculous—using 1201 to stop password sharing on Netflix is never going
to happen. There are no 1201 lawsuits about that, or about anything really.
Used for B2B disputes. But it’s cheap to send someone to the CO.
RT: (1) It’s true that the participants sincerely want to
comply with law/are often risk averse, but the perverse thing is that
participation has to be combined with cynicism about tools/distribution: everybody
agrees to ignore the distribution. (2) As for the characterization of “periphery”:
The terminology makes sense to me because these issues are beyond c’s
boundaries: none of c’s business. Maybe there’s another discourse about
boundaries—this is an invasion. (3) One thing that’s valuable to the industry
is not to lose, ever, and that may explain some of the dynamics.
Buccafusco: dividing the world into software and nonsoftware
might make more sense—allowing © in software is a key problem. Ripping DVDs is
at least plausibly in c’s wheelhouse.
A: yes, probably an original sin, but probably a bad idea to
design 1201 to protect distribution of video as well.
Posts
No comments:
Post a Comment