Newsbeat

Popular Science

On March 18, 2018, Elaine Herzberg, 49, was crossing a road in Tempe, Arizona, when a Volvo SUV traveling at 39 miles per hour hit and killed her. ­Although she was one of thousands of U.S. pedestrians killed by vehicles every year, one distinctive—and highly modern—aspect set her death apart: Nobody was driving that Volvo. A computer was.

A fatality caused by a self-driving car might not be more tragic than another, but it does encourage the wariness many of us feel about technology making life-and-death decisions. Twelve months later, a survey by AAA revealed that 71 percent of Americans were too scared to zip around in a totally autonomous ride—an eight percent increase from a ­similar poll taken before Herzberg’s death.

Self-driving cars are already cruising our streets, their spinning lasers and other sensors scanning the world around them. Some are from big companies such as Waymo—part of Google’s parent conglomerate Alphabet—or General Motors, while others are the work of outfits you might not have heard of, including Drive.ai or Aptiv. (Uber operated the Volvo involved in Arizona’s fatal crash and took its self-​­driving cars off the roads for about nine months afterward.) But what makes some of us so wary of these robotic chauffeurs, and how can they earn our trust?

To understand these questions, it first helps to consider what psychologists call the theory of mind. Put simply, it’s the recognition that other people have brains in their heads that are busy thinking, just like ours (usually) are. The theory comes in handy on the road. Before we venture into a crosswalk, we might first make eye contact with a driver and then think, He sees me, so I’m safe, or He doesn’t, so I’m not. It’s a technique we likely use more than we realize, both behind the wheel and on our feet. “We know how other people are going to act because we know how we would act,” explains Azim Shariff, an associate professor of psychology at the University of British Columbia, who has written about this issue in the journal Nature Human Behaviour.

But you can’t make eye contact with an algorithm. Autonomous cars generally have backup humans ready to take control if necessary, but when the car is in self-driving mode, the computer’s in charge. “We’re going to have to learn a theory of the machine mind,” Shariff says. What that means in practice is that self-driving cars will need to provide clear signals—and not just turn signals—to let the public know what that machine mind is planning.

One solution comes from Drive.ai, a company ­running self-driving vans in Texas. The bright-orange-and-blue vehicles have LED signs on all four sides that respond to the environment with messages. They can tell a pedestrian who wants to cross in front of the car, “Waiting for You.” Or they can warn them: ­“Going Now/Please Wait.” A related strategy is intended for passengers, not pedestrians: Screens in Waymo vehicles show car occupants a simple, animated version of what the autonomous vehicle is seeing. Those displays can also show what the car is doing, like if it’s pausing to allow a human to cross. “Trust is the willingness to make yourself vulnerable to somebody else,” Shariff says. “We engage in it because we can pretty easily predict what the other person will do.” All of which means that if the cars are predictable and do what they say they will do, people will be more likely to trust them. Sound familiar?

Communicating with the machine mind is important, but that doesn’t mean we want it to mimic exactly how humans think and act while driving. In fact, the promise of traveling by autonomous car is that silicon brains won’t do dumb things such as text and drive, or drink and drive, or rocket down the highway while upset after a breakup. (Cars don’t date.) “I believe that they have the potential to be safer” than regular cars, says Marjory S. Blumenthal, a senior policy ana­lyst at the RAND Corporation think tank who has researched the vehicles. But she says there’s not enough good data yet to know for sure.

One practical way to create a reputation for safety is to start slow. The University of Michigan’s pair of self-driving shuttles go just 12 miles per hour. Huei Peng, a professor of mechanical engineering who oversees the little buses, says the research team behind the project is building trust by not asking too much: The predetermined route is just about a mile long, so they’re not exactly speeding down a highway in the snow. “We’re trying to push the envelope but in a very cautious way,” Peng says. Like other experts, Peng compares self-​­driving cars to elevators: an initially frightening technology that people eventually got used to.

Ultimately, not everyone will have to trust driverless cars enough to go for a ride, and especially not at first. Indeed, the public isn’t homogeneous, says Raj Rajkumar, who directs the Metro21: Smart Cities Institute at Carnegie Mellon University. He notices three categories of potential users: tech skeptics, who know that their computer crashes and worry about getting into a vehicle controlled by one; early adopters, who are delighted by the promise of new tech; and people who are stressed by driving and would rather not do it if they don’t have to. The early adopters will buy in first, followed by the folks who just dislike driving, and then finally the skeptics, he argues. “So it’s a long process.” Trust grows like a self-driving shuttle drives: slowly.

The Atlantic

America has long had a fickle relationship with homework. A century or so ago, progressive reformers argued that it made kids unduly stressed, which later led in some cases to district-level bans on it for all grades under seventh. This anti-homework sentiment faded, though, amid mid-century fears that the U.S. was falling behind the Soviet Union (which led to more homework), only to resurface in the 1960s and ’70s, when a more open culture came to see homework as stifling play and creativity (which led to less). But this didn’t last either: In the ’80s, government researchers blamed America’s schools for its economic troubles and recommended ramping homework up once more.

The 21st century has so far been a homework-heavy era, with American teenagers now averaging about twice as much time spent on homework each day as their predecessors did in the 1990s. Even little kids are asked to bring school home with them. A 2015 study, for instance, found that kindergarteners, who researchers tend to agree shouldn’t have any take-home work, were spending about 25 minutes a night on it.

But not without pushback. As many children, not to mention their parents and teachers, are drained by their daily workload, some schools and districts are rethinking how homework should work—and some teachers are doing away with it entirely. They’re reviewing the research on homework (which, it should be noted, is contested) and concluding that it’s time to revisit the subject.

Hillsborough, California, an affluent suburb of San Francisco, is one district that has changed its ways. The district, which includes three elementary schools and a middle school, worked with teachers and convened panels of parents in order to come up with a homework policy that would allow students more unscheduled time to spend with their families or to play. In August 2017, it rolled out an updated policy, which emphasized that homework should be “meaningful” and banned due dates that fell on the day after a weekend or a break.

“The first year was a bit bumpy,” says Louann Carlomagno, the district’s superintendent. She says the adjustment was at times hard for the teachers, some of whom had been doing their job in a similar fashion for a quarter of a century. Parents’ expectations were also an issue. Carlomagno says they took some time to “realize that it was okay not to have an hour of homework for a second grader—that was new.”

Most of the way through year two, though, the policy appears to be working more smoothly. “The students do seem to be less stressed based on conversations I’ve had with parents,” Carlomagno says. It also helps that the students performed just as well on the state standardized test last year as they have in the past.

Earlier this year, the district of Somerville, Massachusetts, also rewrote its homework policy, reducing the amount of homework its elementary and middle schoolers may receive. In grades six through eight, for example, homework is capped at an hour a night and can only be assigned two to three nights a week.

Jack Schneider, an education professor at the University of Massachusetts at Lowell whose daughter attends school in Somerville, is generally pleased with the new policy. But, he says, it’s part of a bigger, worrisome pattern. “The origin for this was general parental dissatisfaction, which not surprisingly was coming from a particular demographic,” Schneider says. “Middle-class white parents tend to be more vocal about concerns about homework … They feel entitled enough to voice their opinions.”

Schneider is all for revisiting taken-for-granted practices like homework, but thinks districts need to take care to be inclusive in that process. “I hear approximately zero middle-class white parents talking about how homework done best in grades K through two actually strengthens the connection between home and school for young people and their families,” he says. Because many of these parents already feel connected to their school community, this benefit of homework can seem redundant. “They don’t need it,” Schneider says, “so they’re not advocating for it.”

That doesn’t mean, necessarily, that homework is more vital in low-income districts. In fact, there are different, but just as compelling, reasons it can be burdensome in these communities as well. Allison Wienhold, who teaches high-school Spanish in the small town of Dunkerton, Iowa, has phased out homework assignments over the past three years. Her thinking: Some of her students, she says, have little time for homework because they’re working 30 hours a week or responsible for looking after younger siblings.

As educators reduce or eliminate the homework they assign, it’s worth asking what amount and what kind of homework is best for students. It turns out that there’s some disagreement about this among researchers, who tend to fall in one of two camps.

In the first camp is Harris Cooper, a professor of psychology and neuroscience at Duke University. Cooper conducted a review of the existing research on homework in the mid-2000s, and found that, up to a point, the amount of homework students reported doing correlates with their performance on in-class tests. This correlation, the review found, was stronger for older students than for younger ones.

This conclusion is generally accepted among educators, in part because it’s compatible with “the 10-minute rule,” a rule of thumb popular among teachers suggesting that the proper amount of homework is approximately 10 minutes per night, per grade level—that is, 10 minutes a night for first graders, 20 minutes a night for second graders, and so on, up to two hours a night for high schoolers.

In Cooper’s eyes, homework isn’t overly burdensome for the typical American kid. He points to a 2014 Brookings Institution report that found “little evidence that the homework load has increased for the average student”; onerous amounts of homework, it determined, are indeed out there, but relatively rare. Moreover, the report noted that most parents think their children get the right amount of homework, and that parents who are worried about under-assigning outnumber those who are worried about over-assigning. Cooper says that those latter worries tend to come from a small number of communities with “concerns about being competitive for the most selective colleges and universities.”

According to Alfie Kohn, squarely in camp two, most of the conclusions listed in the previous three paragraphs are questionable. Kohn, the author of The Homework Myth: Why Our Kids Get Too Much of a Bad Thing, considers homework to be a “reliable extinguisher of curiosity,” and has several complaints with the evidence that Cooper and others cite in favor of it. Kohn notes, among other things, that Cooper’s 2006 meta-analysis doesn’t establish causation, and that its central correlation is based on children’s (potentially unreliable) self-reporting of how much time they spend doing homework. (Kohn’s prolific writing on the subject alleges numerous other methodological faults.)

In fact, other correlations make a compelling case that homework doesn’t help. Some countries whose students regularly outperform American kids on standardized tests, such as Japan and Denmark, send their kids home with less schoolwork, while students from some countries with higher homework loads than the U.S., such as Thailand and Greece, fare worse on tests. (Of course, international comparisons can be fraught because so many factors, in education systems and in societies at large, might shape students’ success.)

Kohn also takes issue with the way achievement is commonly assessed. “If all you want is to cram kids’ heads with facts for tomorrow’s tests that they’re going to forget by next week, yeah, if you give them more time and make them do the cramming at night, that could raise the scores,” he says. “But if you’re interested in kids who know how to think or enjoy learning, then homework isn’t merely ineffective, but counterproductive.”

His concern is, in a way, a philosophical one. “The practice of homework assumes that only academic growth matters, to the point that having kids work on that most of the school day isn’t enough,” Kohn says. What about homework’s effect on quality time spent with family? On long-term information retention? On critical-thinking skills? On social development? On success later in life? On happiness? The research is quiet on these questions.

Another problem is that research tends to focus on homework’s quantity rather than its quality, because the former is much easier to measure than the latter. While experts generally agree that the substance of an assignment matters greatly (and that a lot of homework is uninspiring busywork), there isn’t a catchall rule for what’s best—the answer is often specific to a certain curriculum or even an individual student.

Given that homework’s benefits are so narrowly defined (and even then, contested), it’s a bit surprising that assigning so much of it is often a classroom default, and that more isn’t done to make the homework that is assigned more enriching. A number of things are preserving this state of affairs—things that have little to do with whether homework helps students learn.

Jack Schneider, the Massachusetts parent and professor, thinks it’s important to consider the generational inertia of the practice. “The vast majority of parents of public-school students themselves are graduates of the public education system,” he says. “Therefore, their views of what is legitimate have been shaped already by the system that they would ostensibly be critiquing.” In other words, many parents’ own history with homework might lead them to expect the same for their children, and anything less is often taken as an indicator that a school or a teacher isn’t rigorous enough. (This dovetails with—and complicates—the finding that most parents think their children have the right amount of homework.)

Barbara Stengel, an education professor at Vanderbilt University’s Peabody College, brought up two developments in the educational system that might be keeping homework rote and unexciting. The first is the importance placed in the past few decades on standardized testing, which looms over many public-school classroom decisions and frequently discourages teachers from trying out more creative homework assignments. “They could do it, but they’re afraid to do it, because they’re getting pressure every day about test scores,” Stengel says.

Second, she notes that the profession of teaching, with its relatively low wages and lack of autonomy, struggles to attract and support some of the people who might reimagine homework, as well as other aspects of education. “Part of why we get less interesting homework is because some of the people who would really have pushed the limits of that are no longer in teaching,” she says.

“In general, we have no imagination when it comes to homework,” Stengel says. She wishes teachers had the time and resources to remake homework into something that actually engages students. “If we had kids reading—anything, the sports page, anything that they’re able to read—that’s the best single thing. If we had kids going to the zoo, if we had kids going to parks after school, if we had them doing all of those things, their test scores would improve. But they’re not. They’re going home and doing homework that is not expanding what they think about.”

“Exploratory” is one word Mike Simpson used when describing the types of homework he’d like his students to undertake. Simpson is the head of the Stone Independent School, a tiny private high school in Lancaster, Pennsylvania, that opened in 2017. “We were lucky to start a school a year and a half ago,” Simpson says, “so it’s been easy to say we aren’t going to assign worksheets, we aren’t going assign regurgitative problem sets.” For instance, a half-dozen students recently built a 25-foot trebuchet on campus.

Simpson says he thinks it’s a shame that the things students have to do at home are often the least fulfilling parts of schooling: “When our students can’t make the connection between the work they’re doing at 11 o’clock at night on a Tuesday to the way they want their lives to be, I think we begin to lose the plot.”

When I talked with other teachers who did homework makeovers in their classrooms, I heard few regrets. Brandy Young, a second-grade teacher in Joshua, Texas, stopped assigning take-home packets of worksheets three years ago, and instead started asking her students to do 20 minutes of pleasure reading a night. She says she’s pleased with the results, but she’s noticed something funny. “Some kids,” she says, “really do like homework.” She’s started putting out a bucket of it for students to draw from voluntarily—whether because they want an additional challenge or something to pass the time at home.

Chris Bronke, a high-school English teacher in the Chicago suburb of Downers Grove, told me something similar. This school year, he eliminated homework for his class of freshmen, and now mostly lets students study on their own or in small groups during class time. It’s usually up to them what they work on each day, and Bronke has been impressed by how they’ve managed their time.

In fact, some of them willingly spend time on assignments at home, whether because they’re particularly engaged, because they prefer to do some deeper thinking outside school, or because they needed to spend time in class that day preparing for, say, a biology test the following period. “They’re making meaningful decisions about their time that I don’t think education really ever gives students the experience, nor the practice, of doing,” Bronke said.

The typical prescription offered by those overwhelmed with homework is to assign less of it—to subtract. But perhaps a more useful approach, for many classrooms, would be to create homework only when teachers and students believe it’s actually needed to further the learning that takes place in class—to start with nothing, and add as necessary.

The Verge

“Prison labor” is usually associated with physical work, but inmates at two prisons in Finland are doing a new type of labor: classifying data to train artificial intelligence algorithms for a startup. Though the startup in question, Vainu, sees the partnership as a kind of prison reform that teaches valuable skills, other experts say it plays into the exploitative economics of prisoners being required to work for very low wages.

Vainu is building a comprehensive database of companies around the world that helps businesses find contractors to work with, says co-founder Tuomas Rasila. For this to work, people need to read through hundreds of thousands of business articles scraped from the internet and label whether, for example, an article is about Apple the tech company or a fruit company that has “apple” in the name. (This labeled data is then used to train an algorithm that manages the database.)

That’s no problem for articles in English: Vainu simply set up an Amazon Mechanical Turk account to have people do these small tasks. But Mechanical Turk is “not really that useful when you want to do something [with the] Finnish language,” Rasila says, and the company had only one trainee tagging lots of data in the Finnish language. “We saw that and said, ‘okay, this is not going to be enough,” he adds. The Vainu offices happen to be in the same building as the headquarters of the Criminal Sanctions Agency (CSA), the government agency that oversees Finnish prisons, and so, says Rasila, the founders had an idea: “Hey, we could actually use prison labor.”

The partnership started about three months ago and Vainu is now working with two different prisons, one in Helsinki and one in Turku. Vainu shipped 10 computers to these prisons and pays the CSA for each task the prisoners complete. The amount is comparable to how much the startup would have paid for a task done on Mechanical Turk, though the CSA is responsible for figuring out how much of that goes to the prisoners, as well as selecting the inmates who do the data classification.

Officials at the agency were excited to partner, according to Rasila, especially because the new jobs don’t require anything other than a laptop. “There’s no risk for violence,” he says, adding that when it comes to other forms of prison labor, like metalsmithing, access to tools that can be turned into makeshift weapons can make a prison workspace “a dangerous place.” Rasila estimates that, currently, a little less than 100 prisoners are working on Vainu’s project for a few hours a day. (A representative from the CSA did not respond to a request for comment.)

Right now, Vainu and the CSA have an annual contract based on the number of tasks. The Vainu team hopes to expand elsewhere in Finland, and other countries where it can be hard to find people willing to do this type of work in local languages. To them, it’s a win-win situation. One motivation for the inmates is to make money, of course, but “a selling point of this was that the demand for training AI is actually increasing significantly, globally,” Rasila says. Similarly, the CSA wrote in a release that the program is part of its efforts to develop work activities that match “the requirements of modern working life,” and a PR representative pitched the partnership to The Verge as “an opportunity for inmates to have a job that can empower them.”

It’s not surprising, either, that there would be an especially high demand for this type of work in other countries, according to Lilly Irani, a professor of communication at the University of California at San Diego. AI algorithms need to be trained in culturally specific ways, she says, and most Mechanical Turk workers are in the US.

Though Rasila says that this is an example of developing skills that can be useful in the future, he also says that the tasks have “zero learning curve” and only require (presumably pre-existing) literacy, which calls into question how useful this skill is. This type of job tends to be “rote, menial, and repetitive,” says Sarah T. Roberts, a professor of information science at the University of California at Los Angeles who studies information workers. It does not require building high level of skill, and if a university researcher tried to partner with prison laborers in the same way, “that would not pass an ethics review board for a study.” While it’s good that the prisoners are being paid a similar wage as on Mechanical Turk, Roberts points out that wages on Mechanical Turk are extremely low anyway. One recent research paper found that workers made a median wage of $2 an hour.

For Irani, there’s nothing special about AI in this story. In the US at least, prison labor has long been controversial, with some saying that it economically exploits workers while others argue that it can help rehabilitate them. To her, the public relations push around the collaboration is more surprising than the fact that digital work has become part of prison labor. “The hook is that we have this kind of hype circulating around AI so that we can masquerade really old forms of labor exploitation as ‘reforming prisons,’” Irani says. “They’re connecting social movements, reducing it to hype, and using that to sell AI.”

The Guardian

Nobody was supposed to see Yovana Mendoza eating the fish. The 28-year-old influencer, also known as Rawvana, has amassed more than 3 million followers across YouTube and Instagram by extolling the life-changing properties of a raw vegan diet. She has built a lucrative brand around veganism. But a couple of weeks ago, Mendoza was recorded eating seafood in a video posted by another vlogger. Realising she was being filmed, she tried to hide the fish, but the jig was up. “It was one of the worst days of my life,” Mendoza told the Daily Beast, recounting the ensuing vicious backlash from “vegan YouTube”. “I felt like someone had died.” Well, mate, someone did die. The fish. That’s why the vegans were angry.

Happily, the fish died for a noble cause. As the controversy escalated, Mendoza posted a 33-minute video titled This Is What Is Happening, where she admitted she had stopped being a vegan for health reasons. Her periods had become irregular, and she had been having digestive problems, so she had started eating animal products to see if that helped. In a perfect universe, Mendoza’s digestion would have been of no interest to anybody but herself. But we live in hell; her video has been watched more than 850,000 times, and the postlapsarian drama (dubbed “Fishgate”) has made international news, covered by everyone from the Washington Post to the South China Morning Post.

It’s not difficult to unpack this story’s virality: people love laughing at vegans. But there is a more serious side. Most obviously, it highlights the artifice of the internet: everybody filters their life to some extent, selling a more aspirational version of themselves.

Fishgate is also a worrying reminder that people are increasingly getting health advice from unqualified online influencers. The registered nutritionist Rhiannon Lambert told the Telegraph that she had seen a rise in clients coming to her clinic with symptoms resulting from poor nutrition and, in severe cases, with eating disorders, having taken the advice of social media stars.

Not only was Mendoza promoting a restrictive diet that was making her sick, she was extolling dangerous practices, such as 25-day water fasts, to her millions of followers. And she is far from the only influencer promoting extreme eating. Jordan Peterson, a prominent psychologist, has been outspoken about his all-beef diet, claiming it cured his depression and his gum disease. (Unfortunately, it hasn’t cured his pseudo-intellectual prattling.)

The internet seems to have ushered in an unhealthy age of food fundamentalism. There are, for example, “bitcoin carnivores”: a small but intense community of cryptocurrency enthusiasts who believe that you should only eat meat and only use bitcoin. The internet’s ad-centric infrastructure has accelerated this march towards food fundamentalism.

Last year, the sociologist Zeynep Tufekci wrote a piece for the New York Times headlined YouTube, the Great Radicalizer, which argued that the video platform was pushing people towards more extreme content because it helped generate clicks. “Videos about vegetarianism led to videos about veganism. Videos about jogging led to videos about running ultramarathons,” she wrote. “It seems as if you are never ‘hardcore’ enough for YouTube’s recommendation algorithm.”

“You are what you eat,” has never been so true as it is in the digital era. Whether we are vegan or carnivores, it seems that a lot of us are hungry for community, and will go to extreme lengths to find it. Fishgate may seem funny on the surface, but it’s a tiny part of a much larger problem.

The New York Times

“I was sure she was dead,” Lottie Mackinnon said quietly.

Ms. Mackinnon was sitting huddled in the corner of a cafe with her two children, sipping hot chocolate as she described the day three years ago when she was walking with her Border collie, Bonnie, over the Overtoun Bridge in Dumbarton, Scotland.

“Something overcame Bonnie as soon as we approached the bridge,” Ms. Mackinnon said. “At first she froze, but then she became possessed by a strange energy and ran and jumped right off the parapet.”

A bewitched dog lured to leap off a bridge by a malevolent force? It sounds like a preposterous scene straight from an old “Twilight Zone” episode.

But Ms. Mackinnon’s dog is one of hundreds that Scots insist have suddenly been compelled to throw themselves off the gothic stone structure since the 1950s. Many have ended up dead on the jagged rocks in the deep valley bed below.

Residents of Dumbarton, which is northwest of Glasgow, began calling Overtoun, a century-old bridge that stretches across a 50-foot gorge, the “dog suicide bridge.”

Ms. Mackinnon, who grew up in the neighboring village of Milton, winced at the memory of scurrying down the gorge through the trees and the bushes in a desperate hunt for Bonnie. But when she approached the dog’s body, Bonnie started to whimper and eventually tried to stand up.

“It was a miracle that she survived,” she said.

In a land rife with superstitions, myths and monsters the bridge has been at the center of an enduring mystery. Why do so many dogs jump?

Local researchers estimate more than 300 have sailed off the bridge; tabloid reports say it’s 600. At least 50 dogs are said to have died.

Some say there are rational explanations involving the terrain and the scents of mammals in the gorge that may drive the dogs into a frenzy.

Other explanations take on a more paranormal tone.

The bridge’s location, hushed, lush and sometimes still, fits the description of what the pagan Celts called a “thin place,” a mesmerizing spot where heaven and earth overlap.

“People in Dumbarton are very superstitious,” said Alastair Dutton, a local taxi driver. “We grew up playing in the Overtoun grounds, and we believe in ghosts here because we’ve all seen or felt spirits up here.”

The leaps inspired an episode of the American TV series “The Unexplained Files.” An entire book is dedicated to exploring the phenomenon.

But despite all this attention, the mystery lives on, unsolved.

From a distance, it seems as if the ornate Victorian bridge, built in 1895, is a mere extension of the driveway of an adjoining 19th-century manor built in Dumbarton by a wealthy industrialist, James White.

Closer still, one can make out the bridge’s three archways spanning a small river, the Overtoun Burn. Standing in the middle on the bridge’s blackened granite parapets, it is easy to forget the space beneath falls away into the deep gorge.

In the manor nearby, the current tenant, Bob Hill, said he and his wife had seen several dogs suddenly dive off the bridge since they moved into the property, now called Overtoun House, more than 17 years ago.

But Mr. Hill, a pastor from Texas who runs a local center for women in crisis, had an earthbound explanation: The smell of small animals scurrying around in the gorge below the bridge drives the dogs into a frenzy; then they break free of leashes — if they’re on any — and jump.

“The dogs catch the scent of mink, pine martens or some other mammal and then they will jump up on the wall of the bridge,” Mr. Hill said. “And because it’s tapered, they will just topple over.”

Still, he allowed, the Overtoun grounds are “more spiritual than other parts.”

“Scotland is kind of a place where there is a lot of the supernatural, and it is very common in people’s lives,” he added.

Paul Owens, a teacher of religion and philosophy in Glasgow, grew up in a town close to the bridge and recently published a book about the mystery. When it comes to an explanation for the leaping dogs, he is firmly in the supernatural camp.

“After 11 years of research, I’m convinced it’s a ghost that is behind all of this,” he declared, while sitting outside a pub on a drizzly day in Glasgow.

Mr. Owens’s theory is popular among some local residents, who grew up hearing stories about the “White Lady of Overtoun,” the grieving widow of John White, James’s son.

“The lady lived alone in grief for more than 30 years after her husband died in 1908,” said Marion Murray, a Dumbarton resident. “Her ghost has been lurking around here ever since. She’s been sighted in windows and walking around the grounds.”

In 2010, the animal behaviorist David Sands investigated the phenomenon and ruled out the possibility that the animals were deliberately killing themselves.

His experiments at the bridge found that dogs — especially long-nosed breeds — were drawn to the scent of mammals below. Dr. Sands theorized that the dogs’ limited perspective, their ignorance that the path changes from level ground to a bridge spanning a deep gorge and the smells wafting through the air probably enticed the dogs to jump.

But even he acknowledged that the bridge had a “strange feeling.”

Some residents found his theory plausible, but many here still take the position that the leaps are inexplicable. They question why the phenomenon does not occur at the same rate at other bridges in Britain where mammals roam below.

“Other bridges don’t have troubled spirits lurking about,” Ms. Mackinnon insisted grimly.

Despite the macabre reputation, the Overtoun grounds remain a popular dog-walking area and many of the animals are off leash.

“Many people don’t believe in the story until they see it for themselves, and even then they don’t think it will happen to them,” said Mr. Hill, the pastor.

One day, Emma Dunlop, who said she had heard “the horror stories,” took her Labrador retriever, Ginger, for a walk to Overtoun anyway.

She did not let him out of her station wagon until he was on a leash.

“He’s never tried to jump,” she said, “but sometimes he freezes or hesitates when he gets on the bridge, so I’m always careful.”

Ginger jumped from the car, raced around his owner and headed straight toward Overtoun Bridge, crossing it without any hesitation.

But then Ginger froze, looking back intently at something on the bridge, which appeared empty to human eyes.

“Aye, there she is — there’s the White Lady,” Ms. Dunlop said with a laugh, suggesting Ginger had seen the bridge’s ghost.

Then the pair continued their walk.

The central idea behind the Brexit referendum was for Britain to “take back control”—over its laws, its money, its immigration system. For those who campaigned in favor of the United Kingdom leaving the European Union, Brexit would mark the beginning of a new, more global Britain. By leaving the EU, they argued, they would be returning power from Brussels back to lawmakers in Westminster and, by extension, to the British people themselves.

But if the past two years have demonstrated anything, it’s that Britain hasn’t taken back control. It has lost control entirely.

More than 1,000 days after the U.K. voted to leave the EU, the country’s future is still shrouded in uncertainty. Its exit date from the bloc is still unknown. Its parliament is hopelessly divided. Who will be leading the country in the weeks and months ahead is no longer clear. No one, least of all British lawmakers, seem to know how Brexit will happen, or even whether it will happen at all. Nearly every political figure or institution of note in London—including Prime Minister Theresa May, who has lost control of Brexit to both her party’s right wing and to Parliament, and the House of Commons itself, which has not been able to agree on any single option for leaving the EU—has, when seeking to exert control, proved to have none.

Much like Brexit itself, Britain’s loss of control wasn’t a moment, but a process. It began two years ago, when May embarked on a series of negotiations with the bloc during which her government, without a plan or, soon, a governing majority to support one, appeared at best unprepared and at worst unsure of what it was even asking for.

The British wielded little control over the negotiations, from the timetable that dictated how they would take place (the EU ruled out beginning talks until London formally triggered Article 50, the bloc’s time-limited exit procedure) to the sequencing of the talks themselves (which was also set by the bloc). What little control Britain did have, it often squandered. May’s decision to preemptively invoke Article 50 is now widely regarded as a strategic error. Her failure to pursue a Brexit strategy that could pass muster with both her colleagues in Westminster and her allies in Northern Ireland’s Democratic Unionist Party, who prop up her Conservative Party’s wafer-thin majority, is another.

This loss of control didn’t end when British and EU negotiators reached a provisional agreement on the terms of the U.K.’s withdrawal late last year. Instead of fighting with the EU, May’s government began fighting with itself. The prime minister’s deal—the only one the EU said it would consider—was twice rejected in the House of Commons by historic margins. Facing the prospect of a third defeat on her deal, and with no parliamentary support for the country’s crashing out of the EU, May was forced to go back to Brussels and ask to delay Britain’s March 29 exit date.

Here, too, the decision-making power belonged to the EU: A short extension will be granted until May 22 if the House of Commons agrees to support May’s deal by March 29; if it does not, the extension will last until April 12. A longer extension is conditional on Britain taking part in the upcoming European Parliament elections, which May has so far ruled out.

Throughout the negotiations, various interests have tried to exercise influence—from the hard-line Brexit proponents within May’s own party and her Northern Irish partners advocating against her deal, to the opposition members of Parliaments calling for a second referendum. Though their efforts have worked to stymie parliamentary support of May’s deal, they have done little to achieve their respective aims.

The chaos was compounded this week when the British Parliament, in a bid to end the Brexit stalemate, made the historic decision to strip the government of its powers to set the parliamentary agenda so that it could hold a series of nonbinding votes on Brexit alternatives. The plan ultimately backfired Wednesday night after eight different options—including maintaining a close economic relationship with the EU, a second referendum, and revoking Article 50—failed to achieve the support of a majority.

Having lost control of her party and her cabinet, as well as her authority in Parliament, May did have one last card to play: her own leadership. On Wednesday night, the prime minister pledged to her Conservative Party colleagues that she would end her premiership early in exchange for her opponents’ agreeing to support her deal. “I know there is a desire for a new approach—and new leadership—in the second phase of the Brexit negotiations,” May told Conservative lawmakers at a meeting in Parliament, “and I won’t stand in the way of that.”

With few lawmakers signaling their willingness to shift in support of the prime minister’s deal, even her own resignation seems to be out of her control. Now it’s up to Parliament to reach a consensus on May’s deal or find an alternative.

When Chris Rackliffe, a motivational speaker in New York, met a potential friend at a bar last weekend, it never occurred to him to exchange phone numbers. Instead, the two swapped Instagram handles, and have been liking each others’ posts. Rackliffe said they'll probably meet up in person again soon.

"It's so much more casual to give someone your Instagram handle and keep in touch through stories and DMs,” Rackliffe said. “Swapping numbers feels so serious and stiff nowadays."

As Instagram has grown to more than a billion monthly users, it has also morphed into people's default public internet profile and communication method. "I cannot imagine preferring phone number to Instagram handle," said Ziad Ahmed, founder of Juv, a Gen Z consulting agency.

While nearly many people message frequently throughout the day with their closest friends, Instagram is particularly great for connecting with people you don't know super-well, or have just met.

"Usually someone hands you their phone and you follow yourself on their Instagram," said Rachel Schultz, who works in the advertising industry. "It’s nice because you can then share things with them like events or funny things or whatever they’re into, which makes nourishing new friendships easy."

Crucially, Instagram provides much more context—and conversation fodder—than a random string of ten digits. Because iMessage or SMS don’t come with public profiles, when someone texts you from a number outside your contacts, you have little to no information on who that person is or where you met. Sometimes a person will provide this information themself via text, but often you're left guessing.

But adding someone on Instagram is like scanning a digital business card into your address book. You get the person's full name, their bio, and a direct line of contact through Instagram DM. Plus you have the added benefit of scrolling back on their profile for additional context on who they are and what they're into.

A person's Instagram posts provide a stream of conversation prompts: Maybe you’ll just reply to their Story, asking where they got that pasta they're eating, or notice they're nearby and asking to meet up for a drink. One of the hardest parts of fostering a new connection is figuring out how to reach out and start a conversation out of thin air. Instagram makes it easy.

John Colucci, a social strategist in Seattle, said that he's even seen the insta swap become more popular in business settings. In fact, he has all but done away with his physical business cards. Instagram is "a less intense, light touch connection," he said. It’s notably less uptight than LinkedIn, said Allison Winer, a public relations strategist in Baltimore. "LinkedIn is so stale when it comes to being interactive, and then exchanging numbers is a little too formal and close for comfort," she told me.

And of course, because Instagram is such a good way to get a sense of someone's personality and interests, it's also a great way to suss out dating opportunities. Winer said that men nearly always ask for her Instagram handle as opposed to her phone number. "To be honest, I always try to look them up that way, too," she said. "It gives you a little insight into their life, and of course, selfies." Many dating apps, such as Tinder and Bumble, even allow users to sync their Instagram profiles so suitors can browse their recent pics. In 2017, New York Magazine coined the term Tindstagramming, for the growing phenomenon of men who slide directly into women's DMs on Instagram looking for a date.

Revealing your phone number to too many people can also be a security concern. Once someone adds your number to their contacts it's easy for them to find your profiles on other apps. Given the prevalence of two-factor authentication, keeping your phone number private is an important part of thwarting SIM card hackers. "I avoid giving my [phone number] out unless I know a person very well," said Nima Gardideh, a tech worker and entrepreneur, citing security concerns.

When things go south, the fluid nature of Instagram makes it easy to disconnect. Whereas on Facebook or LinkedIn unfriending someone is a default two-way interaction, on Instagram, it's easy to simply click the unfollow button.

Instagram swapping doesn't work in all contexts. Many users, particularly older ones, still consider the platform to be more personal and are uncomfortable being followed by, for example, more junior members at work. But there are signs the practice is spreading. After I recently posed the question in a Facebook group for Bay Area parents, many moms revealed that they preferred connecting via Instagram before giving away more detailed contact information.

"I asked to exchange phone numbers for texting with another mom and I felt a weird vibe, like I was being intrusive," commented Tina Grove. "I think I should have asked for her Instagram handle instead."

The New York Times

When Facebook and its family of apps experienced a daylong malfunction last week, millions of people got a taste of what life would be like if the social network were out of their lives for good.

I can tell you more about that: I permanently deleted my Facebook account five months ago.

So what has happened in the aftermath?

The social network’s long-stated mission has been to connect people so that we can live in a more open world. But after being off Facebook since October, I found that I did not feel less connected and that my social life didn’t suffer, even though I was no longer seeing status updates and pictures on my News Feed.

My friends and I continued making plans over email and messaging apps. So did my family. Same old, same old.

There were some differences, though — including some strange experiences with online ads. Facebook has long used information that it collects on its users to target people with the most relevant ads. So after a few months of deleting the social network, I began seeing random ads pop up on sites like Instagram (which Facebook owns). Among them: promotions for women’s shaving products, purses and bathing suits.

Instagram might have started thinking I was female, but my wallet thanked me. I realized I was spending considerably less money on my usual guilty pleasure of buying clothing and cooking gadgets online because I was no longer seeing the relevant Facebook ads that egged me on to splurge. Over the past five months, my online shopping purchases dropped about 43 percent.

So what about FOMO, otherwise known as a fear of missing out, typically induced by social media? That is often one of the biggest reasons people are afraid to quit Facebook. What if they didn’t see that post about an outing with a distant friend? Or a party invitation shared on the social network?

For me, it turned out that without Facebook, there wasn’t much I really missed out on — except targeted ads. Here’s more of what I learned.

I don’t miss my ‘friends’

Over the 14 years that I used Facebook, I accrued about 500 friends. Most were former classmates whom I had lost touch with.

In my real life, I have about 20 friends I talk to on a regular basis. So when I finally deleted Facebook, the fallout was underwhelming.

Those same friends kept in touch over iMessage, Signal or email. We still get dinner or go to the movies together. I can think of one friend who exclusively used Facebook Messenger to communicate — we email now and talk less than we used to, but when we meet in person we are as close as we always were. And I can’t remember the last time I attended an event that I was invited to via Facebook, so I never had a case of FOMO.

I can also tell you what I absolutely don’t miss about Facebook: the people who frequently posted online quizzes, political news stories or their thoughts about current events on the site. Occasionally, there were funny or interesting posts, but ultimately most were time wasters.

Recently, I also started reading more books. Could it be because I’m no longer expending my energy reading Facebook?

Brands have long been able to target us with ads through Facebook’s tools. You might see an ad on Facebook for a watch, for instance, because a watch company used the tools to upload your email address and leverage other data that the social network has on you — like your age or the fact that you follow Timex’s Facebook page.

When you browse sites outside Facebook, the company can still track your browsing activity to help brands serve you targeted ads. After visiting a website for a pair of shoes, for example, you might see an ad for those shoes — or similar ones — when you go to another site.

The social network uses a variety of approaches to collect information about web users. One involves Facebook pixel, an invisible tracker that brands can embed into their websites. When you load a website for a brand, Facebook pixel sends information about the device and its browsing activities back to the company. The social network can then use that information to help brands target you.

When I deleted Facebook, I wanted all of that ad targeting to go away. So not only did I erase my Facebook account, I also installed tracker blockers on my computer browser and mobile devices to prevent advertisers from using web cookies and invisible tracking pixels like Facebook’s.

The extra steps worked. The onslaught of targeted online ads stopped.

“If you have the tracker blocker and deleted your Facebook account, you’ve exited,” said Gabriel Weinberg, the chief executive of DuckDuckGo, which offers internet privacy tools including a web browser that blocks trackers.

Facebook says it does not build profiles on people who are not on the social network, nor does it serve targeted ads to them. “Sites and apps send us information about the people who visit them, regardless of whether that person has a Facebook profile,” the company said in a statement. “If you aren’t a Facebook user, we don’t know who you are and don’t build any kind of profile on you.”

Advertisers still have methods other than Facebook to chase me around, but there are economic reasons for them to give up. With Facebook’s tools, it was relatively affordable and effective for them to track and target me. Without those, it gets a lot more costly.

“You might be too expensive for them to chase,” said Michael Priem, the chief executive of Modern Impact, an advertising firm in Minneapolis.

My spending dropped. A lot.

Facebook has often defended targeted ads by saying that internet users are annoyed when they see irrelevant ads. I disagree. Yes, the ads I now see have nothing to do with me — but the benefit was watching my spending drop immensely.

About a year ago, I recall shopping on the site for Taylor Stitch, a men’s clothing retailer. I looked at a coat and closed the window after deciding not to buy it. Then over the next weeks, an ad for that coat loaded on Facebook, inside Instagram and on other websites. Guess what happened? I bought it.

After deleting Facebook, I have more often canceled online purchases after asking myself: Do I need another plaid shirt, frying pan or cocktail shaker? The answer was always no. And as there is now nothing to change my mind, my credit card remains sheathed.

I opened Mint, the budget tracking tool, to get a clear picture of how much I was saving. From October to mid-March, I spent about $341 on clothing and $1,100 on Amazon purchases.

That was a significant drop from my internet shopping sprees before I deleted Facebook. From October 2017 to mid-March of 2018, I spent $1,008 on clothing and $1,542 on Amazon. Gulp.

For years, I saw ads on my personal Instagram account for men’s boots, briefcases and video games. Now, I’m getting ads for products like women’s razors and brassieres.

I have a theory for the change, which Facebook didn’t dispute: Having no Facebook account means Instagram is missing big pieces of accompanying information about who I am and what I like. So it may be lacking signals to serve the correct ads.

And because my fiancée and I both run our dog’s Instagram account and interact with it from our personal accounts and our own devices, some wires may have been crossed and Instagram now thinks I’m female. (Instagram does not require people to share their gender when signing up for accounts.)

That’s a bit odd, but I’m not annoyed. Instead, I find the irrelevant ads amusing. At least they give me ideas for future gifts.

It’s been just four days since the president learned that Robert Mueller found no evidence of collusion, but Rudy Giuliani, Donald Trump’s lawyer, is no longer in the mood to celebrate. He’s thrilled about the outcome, of course, as is his client. Trump told the former New York mayor that “he’s happier than he thought he would be.” But Giuliani, sipping a Diet Coke on Wednesday morning at the Trump Hotel, said it’s time to focus on the next mission: Find out who started all this—and why.

“We’re now trying to prove who did it,” Giuliani said. “The premise is, somebody had to have started the ‘He colluded with the Russians’” narrative. (Never mind that Trump’s posture toward the Kremlin has long been strange.) Asked whether the president himself wants an investigation to examine this question, Giuliani said, “Goddamn right he [does]. This is not ‘Oh, gee, it’s over, let’s forget about it.’”

A reliable barometer of Trump’s moods, Giuliani offered a glimpse into the future. Mueller might be done with his investigation, but Trump and company are loath to let it drop. They want to capitalize on the president escaping criminal charges and make Mueller’s findings a core piece of 2020 campaign messaging. In their view, Attorney General William Barr’s four-page summary of the report is a gift that vindicates Trump, undercuts Democratic investigations, and repudiates critical news coverage. There’s time enough to talk policy on the campaign trail. Team Trump first wants to showcase the special counsel’s conclusions: According to Barr, Mueller reported no evidence that the Trump campaign colluded with the Russians, and he couldn’t make a judgment on obstruction of justice. Even though the probe has led to 215 criminal charges and five prison sentences, Trump and his allies have framed Mueller’s findings as total exoneration.

Yet many Republican lawmakers and strategists fear that Trump would be fixating on the wrong message at the wrong time. They worry that Trump risks repeating the same strategic blunder he made in the midterm elections, which culminated in Republicans losing control of the House. Rather than spotlight economic gains rung up on his watch, the president might wind up dwelling on collateral issues of scant interest to voters. In the midterms, Trump locked onto migrant caravans making their way north from Mexico, warning of a national-security threat that never materialized, and ultimately made little mention of the bread-and-butter issues that some strategists believe would have bolstered his party’s odds for winning.

Tensions over Trump’s campaign message underscore a split within the Republican Party that’s existed since the day he announced his candidacy for president in 2015. Then and now, he’s trampled long-held convictions about how a presidential candidate should behave, relying on his own instincts and feel for the voters’ mood, to the dismay of mainstream GOP figures. Though the establishment and Trump wings of the GOP have joined forces since the 2016 campaign, the next election will test their alliance anew, with the Mueller report emerging as an early flash point.

Trump allies see Barr’s letter as a kind of Swiss Army knife—a tool useful in all kinds of situations. Not only is it exculpatory, they say, but it also implicitly rebukes the press for its coverage of the Russia investigation, inoculating Trump from any future scandal that reporters might unearth. According to a source familiar with internal discussions at the Republican National Committee and the pro-Trump super PAC America First, both organizations are “geared up for any nonsense to come.”

They’re already prepared to attack reporters. “Any reporter who tries that will be hit with 30-second spots of all their ridiculous claims about collusion,” said the source, who, like others interviewed for this story, requested anonymity to describe private conversations. “Their tweets have all been screencapped. It’s all ready to go.”

Earlier this week, Trump’s campaign previewed the in-your-face tactics they have in mind. A campaign official sent a letter to TV producers cautioning them against booking certain guests who had alleged that Trump colluded with Russia, including Democratic Senator Richard Blumenthal of Connecticut, Representative Adam Schiff of California, and Democratic National Committee Chairman Tom Perez.

“It’s not hard to figure where we’re going to go with this,” a current campaign official told The Atlantic. “We’re still in victory-lap mode, but it will turn into a message that [Democrats] will say or do anything to stop us from making America great again, including making up lies about the president and ruining a lot of people’s lives.”

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

White House officials suggested that the president has no plans to “move on” from the report because Democrats aren’t moving on. Instead, “they are doubling and tripling down,” says the White House spokesman Hogan Gidley.

At some level, letting go would be out of character. From the first, Trump has personalized the presidency. He still obsesses over the crowd size at his inauguration, along with perceived betrayals from Senator John McCain of Arizona, who died last summer. The Mueller investigation shadowed Trump for nearly two years. Now that it’s over, he is indulging in a bit of triumphalism.

But the president’s allies on Capitol Hill take a more clinical view. Having lost their House majority in 2018, they’re not persuaded that the Mueller report is the path back to power. “We have to be able to pivot to something,” one House Republican leadership aide told The Atlantic. “You can certainly use the findings to shine a light on any kind of frivolous investigations, but at the same time, we have to start thinking about moving on.”

Representative Mark Meadows of North Carolina, a close Trump ally, suggested that focusing on the press in particular would be unproductive. “Calling on the media to apologize is not something that most people would advocate for, as much as hitting the remote and changing to a different channel,” Meadows said.

No less than the leadership of a pro-Trump political-action committee is ready for a new script. “What we’re gonna try to do is get things back on the economic front—I don’t think we want to stay in the mud slopping for another year,” says Ed Rollins, the lead strategist of the Great America PAC. “I hope he can stay focused on his agenda, keep things positive.”

Fresh developments could keep the Mueller probe alive in any event. Barr’s letter isn’t the last word on the subject. Mueller’s report clocked in at more than 300 pages, The New York Times reported Thursday, raising the possibility that Trump’s handpicked attorney general elided more troubling conduct on the president’s part. Congressional Democrats are demanding to see the full report, raising the prospect of a standoff with the Justice Department. But should more damaging material come out, Trump’s legal team says they have a lengthy rebuttal at the ready.

Trump’s lawyers had drafted a counter-report in preparation for Mueller’s findings. Coming in at nearly 90 pages, it was kept locked in a safe in the Trump attorney Jay Sekulow’s office, ready to be released if the Mueller report found that Trump colluded with Russia or obstructed justice.

Throughout Mueller’s investigation, Trump’s lawyers were far more concerned about the obstruction part of the inquiry than the collusion question. Giuliani summarized parts of their defense in the counter-report to The Atlantic. If Mueller wrote that Trump obstructed justice after allegedly telling then–FBI Director James Comey to “let go” of his inquiry into former National Security Adviser Michael Flynn, Trump’s lawyers would respond by laying out a timeline that suggests Comey inexplicably sat on the information. Comey said the conversation with Trump took place in February, but he didn’t raise alarm until May, Trump’s lawyers note. “If I’m the director of the FBI, and you call me in, and you are obstructing my investigation, I know it right away,” Giuliani said. “How can he decide … three months later, that this is obstruction?”

Trump’s lawyers could still put out some of the counter-report if, in the coming weeks, the Justice Department turns over more material from Mueller’s findings that suggests malfeasance on Trump’s part. Indeed, Mueller’s indecision on obstruction leaves many questions unanswered, which Trump’s team finds frustrating. “It’s kind of absurd that he couldn’t decide,” Giuliani said. “‘Difficult issues of law and fact’—that’s what you’re there for, pal! It’s like saying, ‘Well, I’m a brain surgeon, but I’m not gonna operate because this is a difficult brain situation.’”

In the coming weeks and months, the president will not only press for answers on how the investigation began, Giuliani said, but he’ll also potentially consider pardoning its casualties: Though “he hasn’t decided” if he wants to take that step, “that doesn’t mean down the road he’s given up the power to do it.

“I mean, are there cases that are worthy of it? Probably. People have been pardoned for far worse,” he continued. “Flynn is a very sympathetic case, and in some ways [the former Trump campaign chairman Paul] Manafort is, because he’s already spent a year in jail.”

It’s those emotions that will likely motivate Trump moving forward: a conviction that he and his allies were victims of a devastating miscarriage of justice. Appearing on Sean Hannity’s show on Wednesday evening, the president made clear that the nation should remain riveted to all things Russia, saying that “this was an attempted takeover of our government, of our country.” And his biggest fans affirmed that they have no intention of letting him let go. As Hannity put it, “The deep state’s day of reckoning has now come.”

The New York Times

The dream of a “Blue Texas” has captured the imagination of Democrats for nearly a decade, and Beto O’Rourke has come closer than anyone to making a statewide victory a reality.

His strengths as a candidate in his narrow loss in a 2018 Senate race against Ted Cruz — by 2.6 percentage points — led his supporters to push him to run for president, and he obliged them Thursday morning.

But his performance may have demonstrated something else: Texas is on the doorstep of emerging as a battleground state, and any number of Democrats might stand a chance to compete there in 2020 for the presidency or the Senate.

His relatively close loss is promising for the party because he did not take full advantage of the longer-term trends that might put it over the top sooner than later. His strength came almost exclusively from white voters, not from the growing Hispanic population in the state.

None of this is to take away from his accomplishment. He did better than every Democrat running statewide in Texas in the 2018 midterms. It seems unlikely that many Democrats would have fared as well as he did, and you could argue it bodes well for his chances in a presidential race. But on balance his success was a reflection of deeper trends.

Mr. O’Rourke’s close result wasn’t because of an exceptional turnout that will be hard for other Democrats to repeat in 2020. Republican voters, defined as those who have participated in a recent Republican primary, turned out at a higher rate than Democratic ones. Neither the Hispanic nor youth voter share of the electorate was higher than it was in 2016, when President Trump won the state by nine points.

On the contrary, Democrats in 2020 can be expected to enjoy a more favorable turnout because presidential races tend to draw in more young and Hispanic voters. Mr. O’Rourke might have won Texas last November if turnout had been at the level of a contested presidential race, based on an Upshot analysis of Times/Siena poll responses, actual results and voter file data from L2, a nonpartisan voter file vendor.

The data yields an estimate of how every registered voter in Texas would have voted, based on a long list of geographic and demographic factors that predicted vote choice in the Times/Siena polling. Importantly, turnout in 2018 is among those factors, which allows us to fully untangle how much of Mr. O’Rourke’s strength was because of strong turnout among his supporters.

The data indicates that two opposing turnout trends influenced the results. The electorate was older, whiter and more Republican than the state as a whole — or than the 2016 electorate. But an O’Rourke supporter was generally likelier to vote than a demographically and politically similar supporter of Mr. Cruz. This was the pattern nationwide, so it is not obvious that this can be attributed to Mr. O’Rourke specifically; it could have been the favorable Democratic environment more generally.

Either way, the extra turnout boost probably cut Mr. Cruz’s margin of victory by two points.

Mr. O’Rourke might have won with a turnout of around 10 million voters. (The actual turnout was around 8.4 million.) Without the extra edge of a Democratic wave year, it might have taken 11 million votes, a number that is not out of the question in 2020 if Texas is contested as a battleground state.

So how did Mr. O’Rourke fare so well? He did it through old-fashioned persuasion, by winning voters who had voted for Republicans and for minor-party candidates.

The results themselves make it clear that he won a lot of voters who supported Republicans in other races. He ran three points ahead of the overall Democratic state vote for the U.S. House in 2018 (adjusted for uncontested races), and ahead of every down-ballot Democrat running statewide.

Mr. O’Rourke’s personal appeal was probably a factor, and there’s no guarantee a different Democrat can replicate it. His strong favorability rating (plus-10 in the exit polls, 52 percent to 42 percent) is consistent with that possibility, though he might have also had the benefit of a relatively unpopular incumbent in Mr. Cruz. Only 50 percent of voters had a favorable impression of Mr. Cruz in the exit polls (48 percent had an unfavorable one).

But Mr. O’Rourke’s personal appeal is not the whole story. After all, many Democrats running for the U.S. House or other statewide offices posted noteworthy performances.

The tide lifting all Democratic boats in Texas was an anti-Trump rebellion.

Over all, President Trump’s approval rating was at 49 percent in the exit poll and 50 percent in the large AP/Fox Votecast poll. This is consistent with a variety of other survey data, including a recent Quinnipiac poll that put the president’s approval rating at minus-3 among registered voters in the state, 47-50. Gallup, measuring the much more diverse pool of Texan adults, put the president’s approval rating at just 41 percent in 2018.

It is notable that Mr. O’Rourke, with 48.3 percent of the vote, fell a bit short of the disapproval number of the president. That could reflect that he was facing a well-known incumbent, which offers on average a meaningful benefit to the officeholder (despite Mr. Cruz’s relative unpopularity). It could also reflect the state’s G.O.P. tradition; some number of Republicans may disapprove of Trump but not be ready to vote for a Democrat.

No matter how you explain it, the president’s disapproval rating in Texas would seem to imply that there’s at least some additional upside for Democrats there, beyond what Mr. O’Rourke pulled off. And the president’s far lower approval rating among all adults (as opposed to among registered voters) hints at another opportunity for Democrats: mobilizing unregistered voters. In both cases, Hispanic voters could represent the upside for Democrats.

Mr. O’Rourke’s strong showing had essentially nothing to do with the initial vision of a Blue Texas powered by mobilizing the state’s growing Hispanic population. The Texas electorate was only two points more Hispanic in 2018 than it was in 2012, but President Obama lost the state by 16 points in 2012, compared with Mr. O’Rourke’s 2.6-point loss.

At the same time, Mr. O’Rourke fared worse than Mr. Obama or Hillary Clinton in many of the state’s heavily Hispanic areas, particularly in more conservative South Texas. This could reflect Mr. Cruz’s relative strength among Hispanic voters compared with a typical Republican.

Instead, Mr. O’Rourke’s improvement came almost exclusively from white voters, and particularly college-educated white voters. Whites probably gave him around 33 percent of their votes, up from a mere 22 percent for Mr. Obama in 2012.

There’s clearly additional upside for Democrats if they could pair their recent gains among white voters with improvement among Hispanic voters (through some combination of persuasion, higher turnout among registrants and newly registered voters).

Can Democrats really turn out Hispanic voters in huge numbers in Texas? There are real barriers, and you wouldn’t expect Hispanic Democrats to turn out in particularly high numbers given their low turnout elsewhere in the country. It is far-fetched to suggest Hispanic mobilization will fundamentally transform the electorate and turn Texas into a blueish state anytime soon.

But a more modest increase in turnout or support among Hispanic voters is fairly likely if Texas sees the campaign dollars and attention that come with a contested state in a presidential year. When Mr. Obama contested Virginia, North Carolina and Colorado in 2008, the turnout there increased by an average of 13 percent to 23 percent over 2004 levels, which included a mix of higher turnout among registrants and thousands of newly registered voters. Given the extraordinary turnout in the 2018 midterm election nationwide, it is plausible that turnout will be up substantially in Texas regardless of whether the state is a battleground.

Is Texas a true battleground state? Mr. O’Rourke and House Democrats fared much worse in Texas than in clear battleground states like Wisconsin, Pennsylvania and Michigan. They also fared worse than in the next tier of Sun Belt battlegrounds, like Florida, Georgia, Arizona and North Carolina. And Texas is substantially more expensive to compete in than any of those states, save Florida.

But if you squint at the midterm results, it is not obvious that Texas will be so far to the right of those Sun Belt battlegrounds. The president’s approval rating in the two election surveys was basically the same in Texas as in Florida, Georgia and Arizona (there were no surveys in North Carolina, where there was no statewide contest). Democrats in Texas, including Mr. O’Rourke, fared worse in the major statewide contests, but it was also the only state where the Democrats all faced incumbents in such contests.

It is also the state where the Democrats have the most obvious upside in terms of the mobilization of additional voters; the pace of demographic shifts; and the trend toward Democrats in the Trump era.

Put it together, and Texas is on the cusp of being a true (if Republican-tilting) battleground state. It might not be immediately and vigorously contested, as Arizona or North Carolina will most likely be, given the greater expense of campaigning in Texas and the fact that it starts out to the right of those states. But if Democrats chose to contest it seriously in 2020, there wouldn’t be anything crazy about that.