‘Shared Data,’ a short story from an alternate future

“We make enough stressful decisions at work,” Toba agreed. “No need for extra.”

“Exactly! But then when I looked at moving it was that very reassessment and all that made it pretty much impossible, no one wants to buy here any more, what with the extra cost.”

They were on one of the Assuring Respectful Telework Act-mandated coffee breaks, although Genevieve was drinking rooibos and Toba had shared a picture of the lovely swirly boba tea she’d gotten from the café below her apartment. “Almost enough to make me wish I had moved,” Genevieve joked.

“Oh no you don’t,” Toba answered. “We’ve got the same thing here, only it’s wildfire season. And earthquakes year-round.”

Genevieve clucked sympathetically. “Then you know what I’m talking about with the insurance stress.” Because of increasing volatility in weather patterns — or maybe just because they wanted to, Genevieve thought, but the former was what all the mailings said — insurance policies no longer covered more than a season. (Unless you were really, really rich. And then you had someone to manage it for you anyway.) Every year the terms changed; every year Genevieve had to wade through the piles of offers, each couched in different parameters and with different exceptions and technicalities and terminology.

“Oh yes,” Toba agreed. “I hate it. You know, we have a new brand here, they call it simplified insurance, and they just saw, oh, if a fire crosses the highway you get a payout. If an earthquake is stronger than x and closer than y you get a payout. They don’t check your house or anything, just immediate pay.”

“Huh.” Genevieve sipped her tea and tried to imagine how that would work for hurricanes.

“The payout isn’t as much, but the premiums are cheap—they don’t have to employ all the assessors. But,” Toba’s voice deepened a little, and Genevieve imagined her leaning in towards her microphone, “I’m going to do that this year, and not because it’s cheaper. I just can’t stand…”

“I know,” Genevieve said, when the pause lengthened. The details, the calculations that the insurance companies were always going to be better at than you were, the loopholes, the paperwork.

“Coupla years ago we had some pretty bad smoke damage, and some property stuff—fence got burned.” Toba sniffed. “Lucky not to lose the house, really. And Gen, let me tell you, watching the insurance company do its business was almost worse than the losses themselves. Poking at everything, asking whether we were sure that the damage was from that exact fire and not something we did accidentally with the barbecue, as if it even could have been! Sneering, prodding, looking for lies. It was like they were trying to make our lives harder, like we didn’t deserve their help, even though we’d paid for it.”

“That’s what I was going to tell you,” Genevieve said, glancing at the timer in the corner of the screen. Their employee handbook said explicitly that five minutes over the ART Act-required minimum every once in a while wasn’t penalized, but everyone suspected the office manager tracked those minutes and that at some point they were going to come back to haunt you. “We set up a data cooperative here last year – no, two years ago. At first it wasn’t much, but once enough people joined and we were able to get some of the companies to contribute the data they’d collected—”

“You got the companies to share data?” Toba sounded skeptical, and well she might.

“That—the lawyers got some national agreement on it, so that they had to share the data on the same principles that we share it. Which I guess made it more amenable to them because we get less of the data than they have, because of privacy and stuff. And I don’t even know if it’s a good deal, really, because it meant they wouldn’t be prosecuted for some of the ways they messed up on privacy in the past. But anyway, whether or not that turned out fair, we have a lot of data now, and it really does help in making these decisions, I can’t tell you.”

“I’ll look into it,” Toba said, her voice taking on that winding up tone; Genevieve knew she was watching the timer too. 

“Let me know if you have questions.”

Two weeks later, a massive blackout hit Arizona during the worst heatwave in recorded history, and Genevieve’s brother and his family had to evacuate. 

“Oh, my niece is there too!” Toba exclaimed, when she heard during a mandatory Self-Aware is Self-Care: Addressing Non-work Stressors meeting, and the two made an appointment to chat. Since the catastrophe had already been designated an official non-work stressor, they could use self-care minutes instead of a coffee break, although they both brought comforting drinks anyway, as one of the strategies recommended for self-care. Genevieve was also, on paper where it couldn’t be seen, doing some of the scratch mathematics for tax form N-294201, about deductions for non-monetary contributions to civic good; technically they weren’t supposed to be doing anything work-related or stressful during self-care minutes, but she also couldn’t properly de-stress with the filing deadline hanging over her, so she did it anyway and tried to make sure it didn’t show up on the camera.  

“How are you holding up?” Toba read flatly off of the Suggested Conversation Operners screen that popped up with they initiated the meeting.

They both chuckled, but Genevieve’s ended with a sigh, and she put down her pencil for a moment. “I’m fine,” she said. “My brother is having a hard time. They spent hours with the kids in the car stuck in traffic, and eventually had to abandon it when they ran out of charge and gas — it’s an old car, only partially electric — and they were walking in the heat, fortunately it was night time, but it was terrifying, and they finally got to a FEMA staging area but it was awful. He finally got them out by paying a stranger to come pick them up, you know through one of those ad hoc social media matching things people were setting up for all the people stranded? But it was hard because with the kids they needed someone with a larger car and a car seat, and his wife was so worried about getting into a car with a stranger. But at least now they’re in a hotel.”

“Oh my gosh!” Toba sounded shocked. “That sounds like an absolute nightmare!”

“Yeah.” Genevieve felt drained just talking about it. “And they don’t know if they’ll get their car back, or how long they’ll have to pay for the hotel. They’re going to try to get into a shelter but because they took non-FEMA transport out of the area and are paying for themselves now, it seems like that makes it hard…” She had to stop because of the tightness in her throat, anger or grief.

“At least they got out,” Toba said. “The images from Phoenix are so awful.”

“Is your niece still there?”

“Oh yes.” Toba sighed. “She’s a student, and we told her to leave when the heat got so bad but you know, she didn’t want to miss classes and I guess people weren’t taking it seriously, and then the blackout hit and…but she’s young and healthy, so far she’s all right, she’s with friends in the dorm and they say they have enough water and food, and they just lie around trying to stay cool during the day, but I’m so worried if something happens…”

“What can I do to help?” Genevieve knew the answer was probably nothing, but it had to be said.

“I was going to ask the same about your brother,” Toba said. “Money, I guess? My niece is fine, don’t worry about her.”

“Well…hang on.” Genevieve grabbed her personal device and pulled up the data cooperative’s depressingly crowded ongoing disasters dashboard. Because of her brother, her priority topics were all focused on the displacement crisis outside of Arizona: a running counter on FEMA regulations and updates; an anonymized aggregation of the tens of thousands of complaints to FEMA registered by members of the cooperative, along with any responses; tallies of the populations in various shelters, official and informal; and of course the conversation board and the new app area, where people posted any new tools they had developed for crunching the available data.

She swiped over to the section on Phoenix itself and the people who remained there. “Do you think she would join the data cooperative? They have a bunch of different engagement models. Or maybe you could, as a way to stay aware of what she might need? They can help with sharing information. It sounds like the big worry for you is if something happens.” Illness. Shortage of water. People with guns. “If you don’t know what’s going on. I know,” she added hurriedly, “there’s official information, but it’s never enough. And there are rumors, but you can’t trust them. This is almost like…in between.”

Toba hemmed. “I mean, she has us…and her friends’ families…and those online matching systems like you mentioned.”

“Yes, but those matching systems are so scary, at least with the cooperative we have a better idea who’s involved. And I think it’s more efficient. Plus even if she doesn’t need a ride, just…the info about where to get water or medical care.”

“I’ll take a look,” Toba said, and Genevieve sent her the link. “Ah…if I join, does that—that is, do I owe them…fees, or, data, or?”

“The ongoing disaster stuff is all free. If you want to join later, there’s different rates and sharing plans, especially good stuff for students. But it’s all pretty reasonable. Do you see the site for Phoenix?”

“Um.” A pause. “You know what, let me switch the interface. This has a lot of dataviz for me.”

“You prefer text?”

“Text or audio. Ah, now that’s better. Hm. Oh my!”

“What did you find?” Genevieve was still poking at the Phoenix data, looking at the greatest needs reported in the city and the average wait times on calls to FEMA, 911, the fire department, appliance repair. 

“It’s got her dorm.” Toba sounded a little dazed. “It shows how many people are there right now — phone data, I think — and what shops are open nearby, and the average temperature inside buildings in that neighborhood. And there must be someone there who’s more involved, because there’s stuff on the composition of the paint in the dorm — maybe someone’s chemistry homework? — and its high temperature specifications, and when the lightbulbs were last replaced, and…”

Genevieve chuckled. “It’s a lot, isn’t it?” She remembered when she had started using the cooperative, how surprisingly granular and detailed it was. “And the worst is…”

“…knowing that this data has always been out there,” Toba finished after a pause. 

“Except worse,” Genevieve said. “Because the cooperative has much stricter privacy rules than most of them, and fewer members. You know one time a company made an overture to our part of the cooperative, the municipal area I mean, asking for access to our data, and some members who are lawyers looked at their terms and explained it to everyone and we voted and turned it down.”

“What was wrong with it?”

“Oh, nothing all that serious. Just…they only wanted the data to sell more of their stuff. And none of us felt like that was a good reason. Especially considering how much time we all spend watching ads that supposedly have access to tons of data and still aren’t for anything we want.”

There was another silence, while they both thought about the work they did and the company they did it for.

Toba cleared her throat. “I’ll tell my niece. About the cooperative. Does your brother use it?”

Genevieve sighed. “He joined, but…right now he’s just mainly worried about the cost of this disaster. None of the insurance covers hotels if your home isn’t destroyed, but also they can’t go back. I told him to come here and we’ll sleep on the floor. So far he says no but it will probably come to that. And they keep thinking of things they forgot to bring…”

“That’s something I’ve been thinking about myself,” Toba said. “Emergency bags. I used to keep one, you know, a first aid kit, but it just feels like things are changing so fast and I don’t know what I might need.”

“Oh yeah, the data cooperative has a whole section on that,” Genevieve said. Her screen blinked, reminding her that their self-care minutes for the day were almost up, and she talked faster. “And they’re reaching out to everyone who has evacuated to try to add information about what people should have put in but didn’t. Also there’s a lot of new information about things like what cars held up best, and house features that helped, stuff like that.”

“Kind of wish we didn’t have to organize everything long-term purchase around possible disasters.”

“I kind of wish that too. But since we do, we better try to do it right.” Her screen blinked again, and then Genevieve was back in her standard dashboard view. She swallowed the last gulp of her tea and got back to work.

In response to ‘Shared Data’

By Don Marti, open-source software advocate; Caroline Sinders, machine-learning-design researcher; and Rebecca Weiss, Project Lead rally.mozilla.org

There are few things that are shared on a planetary scale, linking us as cultures, peoples, governments, and societies across the globe, but one of those things is climate change. 2020 saw the most hurricanes and tropical storms on record, along with record highs for heat. Forest fires raging across the West Coast in the United States and fires across the Amazon rainforest are happening at greater rates. The world is changing, and not exactly for the better.

Shared Data exists as an eco-futurist hypothesis of a world existing and persisting throughout climate disasters that have now become commonplace and woven into the fabric of the characters’ everyday lives. But this story isn’t just about climate change — at the heart of Shared Data is a hypothesis of a future project and cooperative that could have numerous benefits for consumers- a shared data hub and cooperative that provides real time data, insights, and social interaction spaces for communities to provide mutual aid in real time. 

Let’s look at Toba’s experience with her insurance company — partly the result of information asymmetry between consumers and companies. Markets that suffer from information asymmetry, where buyers have less information than sellers, often result in a “market for lemons” where high-quality products are eventually forced out of the market and only “lemon-quality” goods are left behind. In Genevieve and Toba’s world, it’s likely that all insurance pricing and premiums will be entirely algorithmically determined; in such a scenario, the side with all the data will always have the lion’s share of negotiating power.

Data asymmetry is only part of the problem, of course. The insurance company is a party to many disputes and can afford to lose some, while each customer has a substantial fraction of their assets depending on their one claim. And while an insurance company has a lower cost of capital and can afford to wait out a dispute, the customer will require a sustained data advantage over the insurance company to offset the company’s cost of capital, risk diversification, and time advantages — not just data parity. The open question is whether this data advantage needs to be so great, to offset the company’s other advantages, that it would be impossible to run an insurance company on so little data. 

The co-op that Genevieve describes is a direct intervention to reverse the information asymmetry. Data cooperatives, where people pool valuable information together and jointly determine how to use it, are an appealing solution for multiple reasons:  

• corporate accountability and government oversight – verification of government information, or validation of corporate transparency initiatives
• collective power – negotiated rates for better quality services, establish access rights conditional on purpose
• potential for innovation – faster response to critical needs should produce better quality information, such as new apps the coop are creating to create more utility and value from the data

The beauty of the co-op is that it creates an official entity of representation on behalf of the good of the collective members, an authorized agent. This authorized agent represents and advocates for the wellbeing of the collective and its members.

Today, the ‘average’ person’s data interests aren’t represented by a collective, a group, or union, so the ‘institution’ working on behalf of an individual is often a bureaucratic representative or elected official. This idea of another entity working on behalf of a person creates a unique third space.  

Like in the data cooperative and Shared Data scenario, we can leverage this new model and offer more support to individuals as a collective whole. The data collective bargaining, advocating and creating boundaries by using an authorized agent is key.  This new structure and communal decision making is core to creating collective power on a scale that most consumers have not had before, outside of electoral processes. 

The current landscape of data policies have been used to secure and concentrate valuable data among increasingly fewer corporations.  As a result, these corporations can increasingly leverage their network effects to their advantage, crushing potential new entrants into the market at critically early stages of product exploration. Insurance regulators can learn from other markets: patterns of anti-competitive behavior seen in the fast-moving web advertising business today will show up in slower-moving markets such as insurance in the near future.

Privacy policy can work against the interest of competition policy when intra-company data transfers are under-regulated compared to (easier to observe) inter-company data transfers. The best-known privacy law, Europe’s GDPR, recognizes this and requires that information only be used for the purpose collected–but intra-company sharing is harder to enforce than obvious sales or transfers between companies.

Although many companies have expressed that they will be confused by the many state privacy laws, this is likely to be addressed by the ability of corporate data professionals to learn and develop best practices that span sites. For example, there are more than 3,000 state and local health departments in the USA, and chain restaurants that operate in every state are able to train their workers and stay in compliance. Data handling best practices will be able to work like food safety best practices, reflecting a safe set of decisions that are clear and work across states. Some positive policy trends include:

• Google’s Adtech Antitrust Case in France: regulators are gaining expertise in analyzing complex data-intensive business models
• Resistance to interpretations of CFAA for entrenched corporate interests: more decisions such as Van Buren in favor of accountability and transparency.
• CCPA limited private right of action in the event of a data breach
• State laws that limit the ability of employers to require arbitration in harassment cases

Cybersecurity, competition policy, privacy, and tax policy are all related. Policies that work in isolation may be counterproductive when considered across all four areas.

• Governance – legal recognition of new corporate structures, and existing ones such as mutual insurance companies and co-ops, that protect and enforce community data ownership models
• Data portability – ownership of personal data on existing platforms and the right to share with whom you choose
• Interoperability – data that works in one system should work other similar systems (no vendor lock-in)
• Taxation – corporate databases that contain information about individuals have significant negative externalities, making them a textbook example of an activity that could use a Pigovian tax. 
• Safe harbor standards for organizations acting as authorized agents.

Along with the policy updates we need to build on the current momentum of citizen led data initiatives. PotholesOfNewOrleans, where citizens of New Orleans document potholes so that authorities can take action to fix them, or advocacy groups like Eye On Surveillance that have created citizen led projects to map surveillance cameras are examples of citizen led data collection projects that could spur a movement for citizen-control over data. These examples are just a start; we need more robust versions of these citizen led data initiatives that can grow, scaffold, and support all kinds of data and datasets, and that prime and train consumers to work together for a common data good.