Friday, May 03, 2019

Legal Applications of Marketing Theory, part 3

Lorin Hitt, Vildan Altuglu, Samid Hussain, & Matteo Li Bergolis, Wharton & Cornerstone Research, Cornerstone Research, Cornerstone Research, Valuation of Privacy: Assessing Potential Harm from Unauthorized Access and Misuse of Private Information in Consumer Class Actions: Disputes over use of data allegedly beyond consented use and over data breaches. Need a method for computing damages classwide if you’re going to have a class action.  Reliable, feasible, and common is required. 

Different methods have been proposed. Reliability is determined by ability to address variation across consumers in nature of private info and preferences about this info; ability to address unique aspects of privacy (privacy paradox).

What is misuse of private info?  Consumers mostly don’t even look at their contracts. Should it be dependent on what the contract says?  So should we do a cost-benefit tradeoff? Look at expected uses v. unexpected uses? Consumers have different preferences over things like retail loyalty cards in return for tracking, office surveillance (54% would accept a camera on them to prevent theft), free social media (33% think it’s acceptable in exchange for targeted ads even though 80% are on FB which is that).

Theory 1: intrinsic value of privacy: inherent value to society. Independent of consumer and type of info. As a nonmarket good, you’d estimate value from surveys (e.g., contingent valuation). Issues: it’s an old school concept, not so consistent w/class actions where harm is alleged to class. Inconsistent w/heterogeneity in preferences & info; ignores cost-benefit tradeoffs. Expressed preferences methods are especially unreliable for privacy (privacy paradox).

Q: why isn’t there a market for this information?

A: it’s been tried and failed.  A lot of people will just give you permission if asked, it turns out. You could try secondary markets as well (e.g., value to landlords, to info aggregation services). Another answer: Other markets for personal info do exist, but the big ones aren’t legal.  On the dark web, you can use secondary market data to value privacy.  Highly volatile, though in persistence, quality, price.

Data breaches have same problem with heterogeneity, but there’s no compensating benefit to consumer. Weirdly, consumers sometimes feel safer after learning of breach. But causality is difficult b/c there are so many breaches—even if you experience identity theft, which breach led to that theft?

Possibilities: difference-in-difference analysis of rate of identity theft incidences before and after breach; same for changes in credit scores, credit lines, bankruptcies. There doesn’t seem to be much effect & it’s shortlived; financial loss is small.  But most studies have looked at small data breaches, not 140 million people as w/Equifax.  Still remains a Q: life cycle of effects.  Alternatives: hedonic models.  Contingent/conjoint valuation.  Q: is privacy actually an attribute? (focalism bias; the privacy paradox)  What is the price of a zero price good (given that zero price is special)?  Privacy paradox makes expressed preference methods unreliable. Existing studies show widely, perhaps implausibly, varying results.

Increased consumer vigilance following breach is an issue.  Studies also show that PII can be inferred from other info: I can guess your SSN with reasonable accuracy if I have other significant info about you. Main point: the research here is just starting. In doing research must be mindful of heterogeneity/accounting for it.

Orly Lobel, Samuel Becher, & Yuval Feldman, San Diego, Victoria University of Wellington &  Bar Ilan University, Poor Consumer(s) Law: The Case of High Cost Credit and Payday Loans

Marshmallow test/delayed gratification as predictor of success. Revisited recently (w/more emphasis on what child’s world has already taught them about trusting promises about the future).  What is poverty?  Hard to compare different countries. A billion people by one metric count as wealthy (over $32/day, adjusted by cost of living by country), while 2 billion are $8-32/day, 3 billion $2-8, 1 billion less than $2 and in extreme poverty.  46 million people in the US.  Experience of poverty has effects on the brain: poor mothers predictably have less emotional regulation in response to baby’s cries.  There are 5,500 Walmarts in the US, 14,000 McDonald’s, and 18,000 payday loan places in the US. 12 million borrowers in the US w/ave. interest rate of 391% compared to credit card average of 15%. $9 billion in fees alone. 80% of the loans are rolled over.

Bertrand et al.: a photo of attractive woman has the same effects on demand for bank loan offers in an ad as reducing the interest rate by 5% in the ad. Overoptimism can be a significant driver of excessive borrowing, but poor people are less overoptimistic than non-poor people. Present bias: poor suffer more from that—give too much weight to the present at the expense of the future—they are very stressed.  Behavioral economics: info overload as a problem; confirmation bias; the ostrich effect/information avoidance.  [The Fyre Festival is a prominent example of all of this happening, with the difference that a bunch of people participating in the marketing or trying to attend weren’t poor and therefore didn’t suffer catastrophic consequences—though the workers who were unpaid suffered more and differently.] Financial worries, time pressure, negative stereotypes, and emotional distress all take up cognitive resources leaving less for evaluating offers. Poverty is punishment for a crime you didn’t commit. Poor children also hear 30 million fewer words than wealthy kids—affects brain development.  Poverty does not grant vacations, so you don’t get a break.  And poverty is expensive: fees and interest rates paid (or consider Desmond’s Evicted, which notes that rents in poor areas are often about equal in dollars to rents in rich areas nearby, but the properties rented are much different).  Self-control in and of itself can’t overcome economic & social disadvantages. Farmers in India: IQ test results were correlated with the harvest—after the harvest they “gained” 9-10 points, a big difference.

We don’t argue that payday loans should be outlawed, as in 12 states.  Balance: these loans may indeed offer some aid in extreme circumstances, especially with no other source of credit. But: 70% of people who take payday loans don’t use it for extreme/unforeseen circumstances, but for everyday bills.  Industry claims that these loans are expensive b/c of low repayment, though some states already cap extensively w/no real problem, as in Colorado.

Recommendation options: (1) Large scale policies like ex ante fixes (e.g., Universal Basic Income); raising minimum wages. (2) Lender regulation for responsible lending. (3) Improving borrowers’ financial decisionmaking.  We can try to manipulate system 1 with nudges, encourage use of system 2. Personalization/differentiated regulation a possibility. Nudges: default requiring payment of whole amount & not rollover; to rollover requires counseling; or rollover that defaults to paying 80% of the whole next time.  Bank apps offer reminders about saving—we could do the same with payday loans. Could present information in other ways, not just interest rate but “dollars owed” to evoke loss aversion.  Can compare the rates to alternatives, e.g., credit cards or postal banking. System 2: one main reason people say they choose payday loans is the ease: there’s little paperwork. [This reminds me so much of Tressie McMillan Cottom’s Lower Ed.] But if it’s too easy they don’t think about it enough. Create more roadblocks: video tutorials, repayment plans that people would have to fill out, vanishing option test (if you didn’t have this, what would you do?).

Q: how is this going to work for people who already are stressed and depleted?

A: the point is to get them to stop and not just use system 1. We’re not expanding their bandwidth, but directing them to use whatever they have in this situation. Can also test readability of disclosures.  Disclosure of rollover risks; ask people why they think they’re different or what they’ll do if they have to roll over the loan.

Marshmallow test: it wasn’t the ability to delay gratification that predicted success, but environment and background of tested children.  Those who didn’t fear that it was now or never for the marshmallow, who were confident that two marshmallows would appear later, were more willing to wait. Without changes, people will not be able to think clearly about taking payday loans.

Q: Why aren’t businesses competing with lower interest rates to take market share?

A: the market is not competitive; these businesses are centered around poor, minority consumers. 

Q: one of the options: offer people alternatives (online lenders)—make that available.

A: Yes; sometimes people prioritize ease of getting the money over searching.

Q: unfair to blame poor people for being stupid. People at our income level are giving away trillions of dollars. It’s irrational for rich people to vote, or to pick stocks instead of investing in Vanguard. Rich people go to church, which is irrational.  [I feel that the presenter was not blaming poor people for being stupid.]  We look at others’ bad choices. Why play on our guilt by ignoring rich people’s errors?  [Because rich people’s errors don’t immiserate them?]

A: I’m not saying that we’re smarter than them. There are enormous positive benefits from getting poor people out of poverty, which is hard to get out of without structures.  Other market failures grounded on irrational behavior exist, but in this case, even rich people should endorse helping poor people get not poor.

Q: but it has nothing to do with payday loans.  They have payday loans because no one will sell them stocks. [?]

A: Payday loans are reinforcing poverty by trapping people into loans they can’t pay. Let’s solve other problems too!

Q: Colorado example: what happened to the supply side in Colorado?

A: many branches closed, but people still have access to credit, with less interest and fees. 

Q: was there displacement into other states?

A: doesn’t know.  Online lending is also a Q.  But payday loans are still available in Colorado, capped at 35%.

Q: predatory practices generally are a problem: studies about microloans in India & interest rates being quite high given low default rates.

Tom Wesson, Mark Pelofsky, David Heller, & Erich Schaeffer, York University, Global Business Experts Group, Voluble Insights, Voluble Insights, Social Media Evidence in Commercial Litigation

Social media are useful for lots of reasons: (1) contemporaneous reactions, not harmed by decay of memory; (2) don’t have to distort the environment with a survey question.  You can also risk losing control of a narrative where the other side uses social media evidence and you don’t.  Complements consumer surveys. Already being used by attys and experts in all kinds of cases; courts are still figuring out how to make sense of it, but most courts have found it probative. USPS sued Lance Armstrong for violating his agreement by using performance enhancing drugs.  Harmed the Post Office instead of helping.  Challenge: show that goals of partnership, which ended in 2004, were harmed by 2014 confession.  We looked at social media (Twitter) for posts that talked about Armstrong & doping, and found spikes corresponding to timing of events. Can also get data for industry-specific bulletin boards. For Armstrong, peak activity was much more than for Alex Rodriguez & Maria Sharapova who had similar scandals.

Do people connect that to USPS?  Some do—and it’s timestamped.  Even two and three years after the interview, people are still mentioning USPS in connection w/his doping.

Q: [but does it matter how many people talk about Armstrong & doping without mentioning the USPS?  OK, it turns out I misinterpreted that question, which I now have.]  The question asked was really whether there was any connection to negative events and the mention of the USPS. For example, if he has a new girlfriend, does that cause a spike in attention?  And do people mention USPS when talking about that?

A: could look. But there are also memes featuring Armstrong in USPS gear, so the negative association continues to be enforced.

Social media evidence from both sides: pink slime, or lean finely textured beef, made by centrifuging beef trimmings.  Safe, leaner, cheaper and gets 20 lbs more meat per cow; or a cheap filler used only in dog food at first that didn’t belong in people food.  Spring 2012: ABC ran a bunch of stories about it.  BPI, the major maker of pink slime, took exception and sued for defamation, claiming $1.9 billion in damages subject to trebling because of South Dakota’s Agricultural Food Products Disparagement Act.  Blamed ABC for the term “pink slime.”

Social media for BPI: very clear increase of use of the term b/c of ABC’s reporting. Stories on ABC’s FB page received a lot of consumer engagement; they were very popular ABC stories in terms of number of comments received relative to other ABC stories.  Allegedly false claims: not nutritious; not meat or beef; BPI committed fraud/impropriety in acts w/USDA; not safe for consumption.  All these claims were repeated/discussed 1000s of times on social media. 

However, before ABC news mentioned pink slime, there were more than 18,000 posts in the first three months of the year mentioning the term, such as when McDonald’s and other restaurants announced they weren’t using it. The line starts to go up the day before ABC’s report, b/c a number of other outlets reported that USDA bought 7 million pounds of pink slime for use in school lunches.  Hard to say all the negative attitudes were driven by ABC’s reporting. Nor was an alternative for the phrase: LFTB wasn’t used to identify this product in social media before; 98% of mentions used “pink slime.”

Q: correlation w/sales?

A: BPI closed 3 of 4 plants.

We work to clean the data: false positives, spam, irrelevant posts, duplicates. Standards are evolving. Newness of field offers us opportunity to apply cutting edge techniques.

One issue: expert may equate social media w/survey evidence, but there are important differences. Not asking the same Qs, like what percentage of overall population is confused—represents people on social media.  Other challenges: distinguish sarcasm from truth; you have to read them.  Social media is heterogeneous in a lot of ways.

RT: I wonder about the relationship between the use of this data and courts’ already troubling tendencies to rely on anecdotes over data. Consider for example the Armstrong fans or antifans or Armstrong-indifferent people who don’t tweet even if they aren’t mad at the Post Office; how many people talk about Armstrong & doping without mentioning the USPS; and related baseline questions seem to me to make this stuff extra risky, which is not in any way to say it shouldn’t be done.  Teach people the differences from surveys!

A: look for correspondences w/other evidence.  And sometimes using search with words won’t work given how much is done online w/images.

Scott Hemphill & Jacob Gersen, NYU Law School & Harvard Law School, Evergreening and the Coca-Cola Bottle

Bottle was introduced in 1917 when CC was really struggling with copycats/fraud like Coke-Ola, Koke, etc.  Sold at the pharmacy soda fountain, the syrup was often replaced w/something cheaper/sweeter.  First, they go to court a lot—it created and distributed three volumes that look like court reporters of the Coca-Cola cases!  Inherent vulnerability was the initially descriptive name from the coca leaf and kola nut, which appeared on the label at first.  By 1903-05, the cocaine had mostly been removed and the kola nut was only used in trace amounts.  So the suit against Koke led to claims of unclean hands.

The bottle might help—Coke’s legal department came up with that, not the marketing department, so go legal!  There used to be just a diamond-shaped label, which was easy to emulate.  Contest chose a new bottle in 1915, then they used a bunch of design patents in 1915 (ungainly and prone to tip), 1923, 1937 (looks more like a modern Coke bottle).  According to lore, they decided to use cocoa pods to inspire the ridged/bulgy design.  If true, sort of misdescriptive. 
1915 design patent and production 1916 model

Eventually, 1960, PTO allows registration of container shape as a mark.  Two forms of evergreening: additional patents w/later expiration dates; trade dress protection as an end run around patent protection.  May be either bad or good. There are some additional design patents of doubtful validity.  1916 production bottle looks a lot more like the 1923 patent than the 1915 design, so is the 1923 patent valid? (see this great article on tracking the shape of the bottle from which the pictures above are taken).  It’s possible, though also a stretch, to think that the 1916 bottle would infringe the 1915 design if unauthorized.  But once that 1915 patent expired, shouldn’t folks have been free to sell their own bottles in that shape?  [I was expecting a discussion of the on sale bar, maybe anticipation; that 1916 bottle sure looks like it ought to trigger the on sale bar for the 1923 bottle.]

Evergreening is a common strategy in other areas, such as pharma. Is there a link?  It may depend on what we think we’re protecting. Aesthetic features of a bottle? Brand investment/source identification? Exclusivity on attractive design?

Should Kellogg treatment for functionality be applied to expired design patents as well as expired utility patents?  Courts have unhesitatingly said no, but perhaps they should hesitate.

Q: we have no idea what optimal incentives are in any IP field.  Better to say: isn’t it interesting to look at how law evolves when it has no idea what the welfare answer is?

Perpetual exclusivity was probably not what Congress wanted under utility patent system; what about design/TM systems?

RT: [Role of incontestability is really significant here.  If you convince the PTO once to register your design and wait five years, you can get perpetual protection even without actual secondary meaning persisting over time.  Congress wasn’t thinking about trade dress & incontestability’s interactions (in part b/c it’s pretty clear they didn’t expect registration of trade dress) and that creates a significant “perpetual patent” problem.] [And by the way, you can get a registration through a presumption of secondary meaning from showing five years of substantially exclusive use … which a design patent will have given you, plus more years beside.  The PTO need not accept exclusive use as sufficient evidence of secondary meaning, and it does not necessarily apply the presumption in all product design cases, but the statute gives it a bunch of freedom.]

[my love of TerraCycle as an example: TerraCycle puts its fertilizer in recycled bottles, including recycled Coca-Cola bottles.  It’s an example of defendant-side functionality; even if the bottle isn’t functional for people who are making their own new bottles, it is for the business model that incorporates recycling.]

Q: from this morning: look at internal and external evidence about their intent to create secondary meaning and their success in doing so.

Q: possible rule providing that you can’t register a TM for a design patent for five years after the expiration of patent protection?  [Also would have to know, as Sarah Burstein has emphasized: What is the patented design?]

A: yes, that would affect incentives. The hesitation is that maybe it’s ok to allow registrations.  Source identification is often a good thing. The anxiety is that one form of transient/temporary protection turns into something else.

Q: the problem is strategic behavior where a party seeks the first right (design patent) to secure the second (TM).

A: does it distort claims made about the product during the period you’re seeking protection? That would be a reason to try to avoid this behavior.

Q: related to whether there’s intent—are firms trying to earn a return on investment on brand?  One should perhaps not be allowed to fall into trade dress protection, but should have to show investments were made.  [This would be an easier rule for courts to tolerate if we recognized a separate role for unfair competition.]

No comments: