Aaron Fernandez / BuzzFeed News

On Monday, Uber released a report on driver earnings, conducted by the company's head of economic research Jonathan Hall and Princeton economist Alan Krueger. Hall and Krueger used data from a diverse group of drivers in 20 markets. They found that the majority of Uber drivers drive 15 hours or fewer per week, and that the average Uber driver earns an average wage of $18.75 an hour, before expenses.

(Krueger was previously a paid consultant for Uber, but the company says he was not paid for his work on this study.)

Jonathan V. Hall and Alan B. Krueger / National Bureau of Economic Research

Uber released a survey of driver earnings in 2015; this new report is an update of that survey, revised to include estimations of driver expenses — how much drivers spend on gas, car maintenance, insurance, and taxes. It also looks at whether the amount of time drivers spend working per week influences their hourly earnings, and how driver earnings change over time, especially following price cuts.

According to the report, which used data from AAA, a full-time Uber driver driving a small sedan would have about $3.76 in expenses per hour, while the same person driving an SUV with four-wheel drive would have expenses of around $5.94 per hour. These conclusions about driver expenses are similar to those made in a June BuzzFeed News investigation, which found that in late 2015 in three cities, working for Uber cost drivers $3.50 to $5 an hour, and which didn't take car size or full-time versus part-time into account. However, this report doesn't take geographical location into account, which is important because the amount of rain or snow a vehicle experiences can impact the cost of maintaining it.

Jonathan V. Hall and Alan B. Krueger / National Bureau of Economic Research

Despite the fact that Uber was able to estimate driver expenses for the first time in this report, its authors chose not to publish estimates of driver net earnings, instead publishing estimates of drivers' earnings before expenses.

However, it's possible to draw an illustrative hypothetical using Uber's estimations. For example, a part-time driver of a minivan in Chicago earning an average $15.48 an hour would, based on Uber's model, incur $4.02 an hour in expenses, for a net hourly earning of roughly $11.46 an hour. A full-time driver in Washington, DC, earning an average $18.21 an hour driving a four-wheel-drive SUV would have expenses of around $5.94 per hour, for net hourly earnings of $12.27.

In June, BuzzFeed News found that Uber drivers in late 2015 earned approximately $13.17 per hour after expenses in Denver, $10.75 per hour after expenses in Houston, and $8.77 in Detroit.

It's interesting to note that while the report does look at how driver earnings change over time, it doesn't look at how driver expenses change over time. Specifically, the report addresses the issue of fare cuts and driver utilization, and ultimately supports Uber's argument that even when customers are paying more, lower prices increase demand, meaning drivers' earnings remain stable.

While that may be true — and Uber's data suggests it is — it doesn't take into account the fact that a driver's expenses will increase as he or she is driving more. So, although the report does measure the rise and fall of driver earnings as prices fall and demand increases, it doesn't measure the impacts of that increase in demand on how much a driver is spending.

Robert Gauthier / Getty Images

Facing criticism from customers and threats of a strike from its on-demand workers, Instacart is rolling out a software update intended to better explain the recent changes it has made to the pay and tips earned by the people that pack and deliver the groceries on which it has built its business.

The move follows unrest among some of Instacart’s on-demand workers, who are frustrated by a recent decline in take-home earnings, as well as by new contracts that include arbitration agreements that would preclude them from jointly suing the company. This growing coalition of Instacart workers is orchestrating a movement against the company via social media that includes calling in complaints to the Department of Labor and a possible strike intended to disrupt the company's delivery operations during two busy shopping days leading up to the Thanksgiving holiday.

“Part of me feels bad because we may or may not be ruining people's holiday, but this is a strike on Instacart,” said Melanie Sutton, a shopper in Chicago who plans to participate in next week’s boycott. She personally knows of seven other workers who plan to participate. “They'll hopefully re-evaluate everything after that. There's only 20 (if that) shoppers in my zone. So if only a couple people are working, it'll be a huge blow to Instacart out here.”

In September, Instacart changed how tipping works for the independent contractors — “full service shoppers” in Instacart parlance — who both shop for and deliver groceries to Instacart customers. Instead of defaulting to a tip that goes directly to the worker filling an individual order, the system now defaults to a service fee that is pooled among all Instacart workers, with the option to add an additional tip later on. Instacart had originally planned to kill tipping altogether, but outcry from workers — and customers, according to the company — caused them to alter course just before the changes rolled out.

The goal of this update, according to Instacart CEO Apoorva Mehta, was to stabilize pay and make the company more sustainable — a change he conceded would inevitably cause some workers to lose money. “We're going to pay a fair and competitive wage in every single market,” Mehta told BuzzFeed News at the time. “If you're getting more than that, it’s going to go down. And if you’re getting less than that, it's going to go up.”

Workers who experienced a decline in wages were unsurprisingly upset by the loss in earnings, which they’ve estimated was around 30 to 40% of their Instacart income — a figure corroborated by a BuzzFeed News analysis of paystubs. Some reached out to Instacart customers to note that the change was hurting their livelihoods. “This service amount is not a tip, and won’t go directly to the shopper delivering your order,” reads a flyer that workers have been handing out to customers, which includes step by step directions to opt out of the service fee in favor of a tip.

Some workers posting screenshots of their paystubs to two Instacart-themed Facebook groups say the tactic is working. Workers anecdotally report that they’ve received more tips recently than they did in the weeks immediately following the change.

But not everyone is satisfied. In addition to mourning a decline in earnings, some Instacart workers are taking issue with the fact that Instacart is asking some of them to sign updated contracts, which reiterate Instacart's existing arbitration agreement.

Instacart was the subject of a class-action lawsuit from workers last year, but the case was sent to individual arbitration. Arns Law Firm, which represented the workers, tried to get the case reconsidered, but on Thursday, the judge denied their motion. Uber, Lyft, Airbnb, and WeWork have all previously used arbitration clauses to evade class-action lawsuits.

Arns has distributed information to workers encouraging them to take advantage of their opportunity to opt out of the arbitration agreement in the new contracts, but Instacart says the new arbitration agreements aren't a change to its existing policy. “We update and revise shopper agreements from time to time; they've always included an arbitration agreement,” Instacart communications director Rebecca u told BuzzFeed News in an email. “All new shoppers have been on the latest version of that agreement since March, right now we're getting existing shoppers on the same agreement.”

But Jonathan Davis, an attorney with Arns, said the new arbitration clause is a "different agreement with new language that impacts rights." It's unclear whether workers who opt out of the arbitration agreement will pursue a new class action lawsuit going in the future.

Boycott Instacart

The Instacart workers dismayed by the company's decisions regarding their compensation and legal rights are now working to orchestrate a strike on Nov. 22 and 23. Their stated goal is to jam up Instacart’s system in the days leading up to Thanksgiving. Organizers, some with experience working in movements including the Fight for $15, OUR Walmart, and Occupy Wall Street, are pushing workers to walk out on shifts to which they've committed or refuse to fill orders they've accepted in hopes of sending a pointed message to the grocery delivery service.

"I am boycotting because my income is 25% of what it used to be."

Danielle Gage is a shopper in Boulder, Colorado, who plans to participate in the Thanksgiving strike. "I am boycotting because my income is 25% of what it used to be. Now $8 an hour is a "good” day,” she said via email. "Every time I have told a customer how the new service fee works they were outraged and Instacart simply doesn't care."

Though anti-Instacart Facebook groups have been growing, it’s unclear how many workers intend to boycott the platform next week. Some workers are planning a picket instead with a date tentatively set for early December. Still others have been calling the Department of Labor to complain, or reaching out to Instacart’s commercial partners to put the pressure on them. This momentum has led to the creation of a “National App Workers Guild” Facebook page in support of "Instacart shoppers who were voicing dissent over the removal of tips by Instacart.”

Instacart twice declined comment on the prospect of a pre-Thanksgiving boycott.

Meanwhile, some Instacart customers are also voicing frustration, saying they were unaware of how the company’s shift to a pooled service fee would impact the people that pack and deliver their groceries. They argue Instacart should have done a better job explaining what was going on with tips.

“I read the sentence that described the service fee three times,” Instacart customer Danielle DeRuiter-Williams told BuzzFeed News, “and I still couldn't figure out exactly what I was paying for, and how it was different from the tip.” DeRuiter-Williams said if Instacart doesn’t take steps to address the confusion, she would seriously considering leaving the platform.

“I read what it said in the app, and I was confused."

DeRuiter-Williams' complaint was echoed by other Instacart customers across Facebook and Twitter. Indeed, BuzzFeed News conducted an informal (and, yes, unscientific) survey of Instacart customers, to which around 5,000 people responded. The overwhelming majority reported that they were unaware of the changes the company had made to its tipping structure and how it had impacted Instacart's on-demand workers.

Instacart told BuzzFeed News it has so far refrained from emailing customers about the change in tipping procedure because it claims very few customers actually open and read its emails. Instead, this week, the company has been gradually rolling out language in its checkout flow that says “NEW: We now use a service amount to pay the shoppers that work on your order.” The system still defaults to a 10% service fee, however, and there’s no specific mention of tips until you click on “Learn More.” Customers also have the opportunity to tip after the delivery is completed.

Instacart

Elise Tribbey lives in Denver, where she uses Instacart to order groceries around once a month. The last time she used the platform, she noticed the introduction of the service fee, but “was unclear on where I was supposed to put the tip.”

Tribbey said she understands the importance of tips to workers because her husband works in the service industry. She said she’ll continue to use Instacart as long as tipping instead of paying the pooled service fee is an option. But if the fee were to become mandatory she “might reconsider.”

“I read what it said in the app, and I was confused,” she told BuzzFeed News. “So I googled and I think I learned more from the articles and Facebook posts than I did from Instacart.”

A worker pushes shopping carts in front of a Wal-Mart store in La Habra, Calif.

Jae C. Hong / AP

OUR Walmart, a digital labor network of more than 100,000 Walmart employees, is using an artificially intelligent chatbot called WorkIt to spread information among workers ahead of the holiday shopping season.

According to Walmart workers asked to speak on an OUR Walmart press call, employees are only able to access information on company policies at work if they use a manager’s computer using a system called “the wire.” As a result, they said it was sometimes difficult to research company policy on issues like sexual harassment and sick time. The WorkIt app, which includes discussion forums as well as a Watson-powered AI chatbot, is intended to address this problem by allowing employees to seek out answers to HR questions via their smartphones. Currently, the app is available only on Android devices, but OUR Walmart plans to bring it to iOS later this year.

“Being unable to access policy online when you are at work makes difficult to know when you are breaking policy — or understand how managers are using policies against you,” said Mississippi Walmart worker Joanna Chambers. “Through the WorkIt app, we’re amplifying the strength of OUR Walmart networks.”

Initially funded by the United Food and Commercial Workers union and managed by labor organizers Andrea Dehlendorf and Dan Schlademan, OUR Walmart subsequently broke off on its own. It now manages a 44,000 plus member Facebook page that isn’t controlled by a traditional labor organization. But if Walmart employees embrace WorkIt, the app could supplant Facebook as a nexus of worker communication, putting control more squarely in the hands of those running the OUR Walmart organization.

“What we’ve seen with discussion happening online — on Facebook and on Reddit — are conversations that are jumbled and disorganized,” Dehlendorf said on the press call today. “They don't allow people to get the answer quickly and efficiently, and from someone they trust.”

In contrast, WorkIt’s algorithm is designed to draw from hundreds of pages of company policies and employee guidelines to deliver accurate and reliable answers to worker questions. The tool is being trained by “experts,” employees who volunteered to teach WorkIt how to provide the most accurate and useful questions to employee queries.

Walmart disputed the notion that its workers do not have unfettered access to the employment documentation they need. “Our associates already have anytime-access online to the company’s most current and accurate Paid Time Off policies and there is no way to know if the details this group is pushing are correct,” Walmart spokesperson Kory Lundberg wrote in a statement provided to BuzzFeed News. “Our people are smart and see this for what it is, an attempt by an outside group to collect as much personal and private information as possible.” OUR Walmart told Bloomberg that it doesn’t collect location-based data and won’t sell user information to third parties.

Catherine Huang, OUR Walmart’s chief technology officer, said the group is currently using Watson's natural language processing abilities to match employee queries with material in an existing training database. Going forward, she said she'd like to expand those capabilities and is considering a Spanish language version for early 2017.

OUR Walmart will be taking WorkIt on the road over the next two weeks, visiting stores in over a dozen cities in an effort to get workers to sign up. Traditional unions require things like membership and fees, but OUR Walmart is defined by a looser structure. That makes WorkIt — an app that’s a central hub for communication and independent silo of user data, including contact info and most frequent questions and complaints — especially valuable if it works.

Though OUR Walmart created the app, Walmart workers are only a test case for the technology (albeit a major one — Walmart is the second largest employer in the United States). For a price, WorkIt can be repurposed by any organization looking to connect workers to each other and information. Troy Burton, a labor organizer in Australia who said unions there are looking forward to using WorkIt, suggested that some data gathered in the app could even be used at the bargaining table when it comes time for workers to negotiate new contracts.

For Burton, the project is a blend of the mechanics of collective bargaining that undergird the “philosophy of unions” and “the long held promise of the internet to democratize information and overcome the barriers of geography and cultural separation that people face.”

ニューバランス「オバマ政権は私たちの話を聞いてくれませんでした。トランプ次期大統領のもとで、いい方向に物事が動くと期待しています」

ニューバランス社の広報部長、マット・ルブルトンは、ジャーナリストのサラ・ジェルマーノがツイートした問題引用について、引用自体は間違っていないが、本来の文脈から外れて引用されたものだ、とBuzzFeed Newsに話した。

「（サラ・ジェルマーノに引用された）発言は、貿易という文脈においては正しいのですが、大規模な地政学的なことについてではありません。環太平洋戦略的経済連携協定（TPP）について話していたのです」

「引用が不正確です。私の発言は、すべて貿易との関連において語ったことですから」

ジェルマーノのツイートは、人気のスニーカーブログ、 Sole Collectorで取り上げられた。

その誤った情報はスニーカーファンたちの激しい怒りを買い、Twitterで広まった。

ニューバランス社には、今年4月以来、TPPに反対してきた歴史がある。ルブルトンによると、ニューバランスは「大統領選挙の政治に巻き込まれる」ことには関心がない、という。

「私の発言は政治的なものではなく、TPPに関する発言なのです。私は ボストン・グローブ紙や他のメディアの取材を受け、オバマ大統領の履物に対する税やTPPにニューバランス反対していることについて話しました」

ニューバランス社は、トランプの勝利を祝っているわけではない。それでも、大勢の人々がこの報道に反発するのを防げなかった。

ニューバランスの靴をゴミ箱に投げ入れる動画をツイートした人々もいれば、ニューバランスの靴に火を点けようとする動画を共有する人々もいた。

New Balance vice president of public affairs Matt LeBretton, quoted above by journalist Sara Garmano, told BuzzFeed News that while the quote was correct, he was taken out of context.

"The statement [quoted by Sarah Germano] is correct in the context of trade, not talking about large geo-political anything, but in the context of the Trans Pacific Partnership agreement."

"It's inaccurate. Everything I've said is in the context of trade."

Germano's tweet was picked up by Sole Collector, a popular sneaker blog:

The error led to outrage among sneaker fans across Twitter.

New Balance has a history of opposition to the TPP dating back to April, but LeBretton said the company is not interested in "getting involved in the politics of presidential elections."

"My statements aren't political, this is policy related solely to TPP. I've spoken on the record to The Boston Globe and other sources about our opposition to Obama's footwear tax and the TPP."

While New Balance isn't actually celebrating Trump's win, that didn't stop plenty of people from reacting to the news.

Some tweeted out videos of themselves throwing New Balance shoes in the trash, others shared pictures of themselves trying to light the shoes on fire.

BuzzFeed News spoke to Mary Katherine Ashmore, originally from Georgia, outside of the popular San Francisco gay bar El Rio about how she felt about a Trump presidency.

Jed Holtzman, speaking to BuzzFeed News inside El Rio, said that he had spent the past four days in Nevada canvassing in support of Hillary Clinton and that he felt that his work had been defeated. He was crying as he spoke. He told BuzzFeed News he was "completely shocked" by Trump's success in the election and called the candidate an "authoritarian racist, rapist, and fascist."

"The mood at El Rio tonight is a little celebratory," he told BuzzFeed News, "And I think you're seeing a bit of the shallowness of San Francisco on display. We live in a highly regionalized country, and the people here think that whatever happens in Washington or in the rest of the country won't affect them. A lot of queer people elsewhere are being grotesquely affected by the decision America made tonight."

Rick Hauptman turned 70 two weeks ago; he served as a delegate for Clinton at this summer's Democratic National Convention, and for Barack Obama before that. "I lived to see the first African-American president, and I thought I was living to see the first female president, " he told BuzzFeed News from El Rio. "I will never see that in my lifetime," he continued, choking back tears.

Jane Martin, also inside El Rio, said she was scared and sad at the prospect of a Trump Presidency. She works for the San Francisco janitors' union and had also been canvassing in Nevada for Clinton. She was worried about the safety of her colleagues who are Latino immigrants.

Martin told BuzzFeed News, "People here are in shock. Hopefully we'll be in the streets soon — maybe tomorrow — showing that we believe in a country where everyone is safe."

"People on the street were just weeping and holding each other," Tarin Towers told BuzzFeed News.

People online weren't any happier than the people at El Rio.

They're afraid for their futures under a possible Trump presidency.

Intersectional hate is particularly strong. Many LGBT people of color on Twitter were not hopeful about their prospects.

Trump's running mate Mike Pence has supported "institutions which provide assistance to those seeking to change their sexual behavior" and LGBT people haven't forgotten.

How the LGBTQ community will fare under a Trump presidency remains to be seen. Much has been made of been made of the Gays for Trump movement in the media, but Trump has indicated that he will effectively reverse the LGBT-friendly attitude of President Obama by taking a more hands off approach to the community's issues.

A Latina supporter speaks on stage with Republican presidential candidate Donald Trump during a campaign rally at the Venetian Hotel on Oct. 30, 2016, in Las Vegas.

John Gurzinski / AFP / Getty Images

Hillary Clinton's campaign fought hard to mobilize Latino voters against Donald Trump, as he promised to build a wall along the US–Mexico border, compared undocumented Mexican immigrants to criminals and rapists, and said he would stop immigrants from sending money back there.

But now, Latino voters throughout Arizona and in San Francisco told BuzzFeed News that they're scared after Trump's win.

In Phoenix, Latino and immigrant rights organizers reveling in their defeat of Sheriff Joe Arpaio were beginning to come to terms with a Trump presidency.

“We always knew it was going to be a fight,” said Neidi Dominguez, director of worker centers and deputy director of community engagement for the AFL-CIO. “Now we have to think about what the clash is going to look like with an empowered immigrant and Latino community against a group of empowered racists.”

Dominguez said she was worried about how communities of color will "survive in a nation led by the most racist and bigoted leader the country has seen in decades."

"The fight is now going to be a fight for years to come and it's going to manifest itself in 100 different ways under a Trump presidency," Alfredo Gutierrez, the former Arizona senate majority and minority leader, told BuzzFeed News.

David Campos, a member of San Francisco's Board of Supervisors, told BuzzFeed News that "for every positive step the country has taken in its history, there's always been a backlash."

"That's what we're seeing here: a backlash following the election of the first African-American president, following same-sex marriage. There's a group of people who want to turn back the clock," Campos said. "The fight continues and we're not going to give up."

Carmen Elias, who lives in the Mission neighborhood in San Francisco and has owned a Mexican bakery there for 24 years, told BuzzFeed News her husband keeps waiting for her to cry about the outcome of the election, but she's staying strong.

Carmen Elias

Caroline O'Donovan

"I feel really depressed," she said. If Trump wins, she said, "life continues. You cannot stop that. But you really have to start thinking about things that are going to happen," she said, explaining that a Trump win shows a greater number of people share his views on immigrants.

"He said, like, he's gonna report Mexican people," Rogelio Tocohua, a 25-year-old in the Mission, told BuzzFeed News. "Some people are fixing papers right now, getting visas. I'm scared and disappointed about what's going to happen."

Instacart rolled out some changes to its platform recently regarding how its drivers get paid. Initially, the company said the move was to help stabilize or increase earnings. But after some shoppers complained about losing earnings, and an analysis of their pay stubs backed them up, CEO Apoorva Mehta told BuzzFeed News the wage decrease for top shoppers was necessary for the company's continued growth. You can read more about that here.

One of things these full service shoppers were most upset about was losing tips, which are no longer set up as a default. Many of them say customers have no idea their tips aren't going directly to the person who dropped off their groceries. We want to find out if that's true. If you're an Instacart customer and have been for longer than two weeks, you can take the survey below.

Please only take the survey if you're an Instacart customer. If you've never used Instacart, or are a shopper, or just like taking online surveys, we kindly ask that you refrain from participating in order to protect the integrity of our findings. Thank you!

View Entire List ›

Bloomberg / Getty Images

Instacart workers are earning a lot less money after changes to the company’s pay structure — changes CEO Apoorva Mehta told BuzzFeed News are necessary for the company’s continued growth, but that hundreds of vocal Instacart shoppers say are threatening their livelihoods. According to a Buzzfeed News analysis of 15 workers’ pay stubs, shifts that once earned shoppers $100 or more in 4–8 hours have dropped closer to $60 to $80 for similar shifts. These shoppers estimate their earnings have fallen by around 30% so far.

When Instacart announced in September that it would be updating the way shoppers get paid, replacing its tips-based pay system with a pooled “service amount,” a company blog post described in detail how the changes would both increase driver pay and stabilize their earnings. But now, one week after the update rolled out in most markets, many shoppers who spoke with BuzzFeed News said the change has made working for Instacart non-viable.

Remington Donovan, an LA filmmaker, told BuzzFeed News he relied on tips to make an average hourly earning of around $25 an hour, which is twice LA’s minimum wage. But Instacart replaced those tips, which went directly to the individual worker delivering your groceries, with a fee that is collected by the company and distributed among all shoppers. Customers can still tip on the platform after an order is delivered, but that’s in addition to the service amount, and shoppers say the new tipping feature is hard for customers to find.

“Out of seven orders on one day, I only got two tips,” said Donovan. “My income decreased enough in just a couple days that by my assessment, it just won’t be worth it for me.”

The shoppers’ main complaint is that, although it is still possible to leave tips via Instacart’s platform, the tip feature is now hard to find, and customers, who were never explicitly told about the change, don’t realize they’re no longer leaving tips. Donovan called the move “knavish subterfuge.” Mehta says, from his perspective, “tipping is extremely, clearly, visible” in the app.

But some customers disagree. “I'm a longtime user of Instacart and I love the service, but I'm very frustrated. I wanted to leave a tip, but I couldn't figure out how, and I work in tech!” said Nicole Sullivan, a San Francisco based project manager. Some shoppers are taking time to explain how the new tip system works when they make deliveries, but even that might not solve their problem. “I finally found a way to add it,” Sullivan said, “but I can't add a 15% tip on top of a 10% service fee on top of groceries that are more expensive than they would be in the store. It’s too expensive!”

While some low-earning shoppers will benefit from the pooled service amount, Mehta told BuzzFeed News that for shoppers who were earning significantly above market rate wage, “the wages are reduced.” Indeed, pay stubs shared with BuzzFeed News by more than a dozen shoppers do suggest tipping has decreased significantly since the changes rolled out.

A flyer distributed to customers by Instacart shoppers explaining how to leave a tip.

(The vast majority of shoppers who spoke with BuzzFeed News for this article asked to remain anonymous out of concern that their accounts would be deactivated for speaking with the press; Instacart said it has never deactivated workers for speaking publicly about their experience with the company.)

For example, a female driver in Denver earned around $90 on Instacart during a four-and-a-half-hour shift on a Monday in October prior to these changes; over $50 of that came from tips. The following Monday, the first day the changes took place, she earned only $60 in six and a half hours, and received only $19 in tips. She said if the trend continues, she’ll have to find another job.

Like many other shoppers, she believes that Instacart is intentionally allowing customers to believe that the automatic 10% “service amount” is a tip that goes directly into her pocket. “I will not continue working for Instacart at this low of pay, and while customers are being tricked into thinking they left a tip,” she said in an email. Other shoppers had similar experiences. Six of the shoppers who shared their pay stubs got no tips at all, and three more made less than $5 in tips following the change. In the week before, most of them earned around $50 in tips, with three of them earning more than $100 in tips.

Only a portion of all Instacart shoppers — those who do both shopping and delivery and are independent contractors, not the ones who work in-store and are employees — were impacted by the pay structure changes, but the group has nonetheless been notably vocal on social media and successful in building momentum around their cause. There are two active Facebook groups, each with a couple hundred members, plus an Instagram page, a popular hashtag —#WheresTheTipInstacart — and an anonymous open letter on Medium. The shoppers use these methods to share information about what they’re earning, as well as distribute materials like flyers to explain what’s going on to customers. There’s even a website, Instawtf.com, where almost 1,000 people have pressed a big red button to tweet “wtf” at Instacart’s Twitter account.

These shoppers are hopeful that their combined efforts will convince Instacart to revert to a tips based system. But Mehta says that’s unlikely to happen. The purpose of changing the system, he told BuzzFeed News, was to “bring consistency to our model as we go from tens of thousands of shoppers to doubling that by the end of next year.”

The goal, Mehta said, is to have shoppers earning an average “market clearing wage” — so, in San Francisco, between $16 and $22.50 an hour. “If you are making more than that as a shopper, then your wage will go down,” said Mehta. “If you were making less, your wages will go up.” Instacart is confident that plenty of shoppers will be willing to work for the new rate, which Mehta said changes daily based on supply of workers and demand for orders in a particular market. In fact, since the new dynamic pay model started on Instacart, that base rate has already started to fall. In two Bay Area zones, for example, the new base pay rate rolled out at around $10 per trip, but was already below $9 per trip within a week, according to screenshots shared with BuzzFeed News.

As independent contractors, Instacart isn’t required to pay its full-service shoppers a particular amount — not even minimum wage. Like any free labor marketplace, Instacart only has to pay whatever it takes to keep its supply of workers steady. Even if some of the shoppers who are now earning less money quit, it doesn’t seem like that will be a problem for Instacart, at least not right away.

“We want to continue to hire, and double the fleet that we have,” said Mehta. “The way we’re going to be able to do that is to make sure it’s a great place for shoppers to earn a fair and competitive wage.”

But while it’s undoubtedly true that there are people willing to work for what one Instacart shopper on Facebook called “McDonald’s wages,” those who have been in the on-demand game a little while are concerned this move is just the beginning of a continuous crunch for Instacart shoppers, and other independent contractors who aren’t protected by employment law.

“Instacart is a great idea and a great company. It can be frustrating, but overall it's been a fun job for me, part time,” Donovan said. “But I don’t think this is the right way to do business. We need to send a message to other on-demand businesses: This isn’t right.”

During a conference call today, Airbnb announced plans to directly police illegal activity on its platform.

Starting Nov. 1, Airbnb hosts in San Francisco and New York City will be unable to list more than one home on the platform. Listing multiple homes has been banned by legislation in both of these cities, but those prohibitions have proved hard to enforce. This is the first time Airbnb has changed the way its platform is engineered in order to assist with regulation.

Airbnb critics in high-rent cities like New York and San Francisco have long argued that some hosts are renting out to tourists multiple units that could otherwise be rented out affordably to residents of those cities, thereby driving up housing costs.

“If you build this into the front end,” said Airbnb Head of Global Policy Chris Lehane Wednesday morning, “the problem takes care of itself.”

Starting Nov. 1, hosts will go through “a rigorous process where we figure out who you are,” said Lehane on the conference call. Airbnb did not go into further specifics about how this process works.

In an email, a spokesperson from anti-Airbnb interest group ShareBetter said the change is "a tacit acknowledgement by Airbnb that they're dominated by commercial operators with multiple listings who are stealing our supply of affordable housing and they're finally admitting it's a problem.”

In a memo released Wednesday morning, Airbnb also made other policy proposals specific to New York hosts, including a three-strikes policy for hosts who break the rules, a revenue sharing scheme in which landlords of multi-unit buildings could use some of the profit from tenants hosting Airbnb guests to cover maintenance costs, and a “simple, streamlined registration system” built by Airbnb, rather than by the city.

These suggestions from Airbnb come just as New York Gov. Andrew Cuomo is weighing whether or not to veto a bill that would heavily regulate short-term renting, including fines of up to $7,500 on individual hosts. The offers made by the company on Wednesday follow an aggressive ad campaign opposing that bill, as well as threats by Airbnb to sue the state of New York. The company has already sued San Francisco, Anaheim, and Santa Monica in California over similar regulatory issues.

It’s not yet clear whether Airbnb plans to make similar changes to its product outside of New York and San Francisco. It’s also not clear whether and how hosts will find new ways to circumvent the rules.

These won’t be the only changes to the platform going into effect on Nov. 1. According to a report on discrimination on Airbnb published in September, all Airbnb users “will be asked to affirm and uphold the Airbnb Community Commitment before they book or share their space on the Airbnb platform.” According to the report, users who don’t agree with the discrimination policy will be booted from the site starting Nov. 1. Other features — including an email survey sent to guests who are denied rooms and online anti-bias training — are set to launch the same day.

Governor Cuomo has until Oct. 29 to veto the short-term rental bill.

Via Flickr: 14936293@N03

Glassdoor is putting its trove of user-reported salary data to use with a new tool that estimates an employee’s “market value.” Called Know Your Worth, the tool relies on a combination of personal information — including location, industry, education level, and years of experience — as well as the local demand for labor. It then spits out an estimated dollar amount workers can reasonably expect to earn at that moment in a given market.

The tool, which is available in beta as of Tuesday night, is powered by several million salary reports and over 4 million job postings, and is supposed to give Glassdoor users a sense of where they stand in comparison to peers with similar levels of experience.

It also compares this figure to average pay for people in your position, and provides a chart showing how market value for workers of your experience in your location has changed over time.

BuzzFeed News calculated market value for a small group of people contacted via Twitter and Facebook prior to the tool’s beta release.

“I know I’m underpaid,” said a 26-year-old project manager working for a tech company in New York City. She earns a base salary of $95,000 a year, but Know Your Worth suggested her market value was more like $108,000. Though the tool only confirmed suspicions she already had, it could serve another purpose, she said. “I would definitely bring this into a negotiation.”

In Chicago, a project manager at a startup also making $95,000 a year echoed her sentiment, saying he plans to bring this Glassdoor data into a salary negotiation later this week. “It's always a bit unsettling to see it presented so badly: You are making less money than you should be. That being said, I knew I was a bit below market,” he said. “Seeing it just output like that with only a comparatively small amount of information needed from me is pretty neat!”

Others were more surprised by the results they received. According to Glassdoor, a 26-year-old male software engineer at Uber in San Francisco with two years of experience and a bachelor’s degree from Stanford is worth more than $140,000 a year. Given these results, an engineer who is actually making $128,000 a year said he “knew the base salary was less than market, but that is a more substantial difference than I expected.” He, too, said the data from Glassdoor could play a role in future salary negotiations.

Know Your Worth was less helpful for those in less straightforwardly corporate jobs. For example, a microbiologist with a doctorate working for the USDA in the Bay Area and earning $70,000 a year was surprised to find his market value was actually lower than what the government was paying him. “I had figured I was more underpaid in order to work on what I value. But it turns out, I'm not worth shit,” he said. “I know if I left my research job to do strictly data stuff, I'd be worth much more.”

Other highly educated workers felt Glassdoor’s tool underestimated their potential earnings. For example, according to Glassdoor, a 30-year-old public defender with a law degree from Harvard earning $60,000 a year in the Bronx has a market value of around $108,000. But she says the tool wasn’t as useful as word of mouth. Plenty of the lawyers she graduated with are already making $160,000, she said, which she knows “because in law school, all everyone talks about is the starting salary at law firms.”

Glassdoor’s new tool was also ineffective for people who aren’t traditionally employed. For example, a 35-year-old marriage and family therapist who charges clients $150 an hour in a private practice was unable to get any results from Know Your Worth. Glassdoor said that the tool should work for someone who is self-employed, but did say, for now, it won’t work for a freelancer who works multiple jobs for multiple employers. “Our goal in future iterations of the tool is to take into account even more information to help all people and give them even more personal information,” said Corporate Communications Director Scott Doboroski via email.

Of course, Know Your Worth is only as strong as the data Glassdoor receives, which means people in certain jobs might not be able to get an accurate quote. The company says its data set is constantly being updated, and that those who don’t find it useful can still make sense of the more basic Salary Explorer feature.

Glassdoor says Know Your Worth will continue to learn over time as new data enters the system, and employees can sign up for regular updates as conditions develop.

Postmates

Postmates, the anything on-demand startup, has run afoul of labor regulators in Chicago.

On Thursday, the National Labor Relations Board's regional office in Chicago announced it had found "merit to the October 2015 charge filed against the on-demand food delivery service."

The board issued a complaint alleging Postmates violated its drivers' employment rights "by prohibiting them from discussing terms and conditions of employment, including safety, with other drivers."

The complaint also alleges Postmates violted the rights of its employees by requiring them to enter into arbitration agreements.

The San Francisco-based Postmates has raised more than $238 million in venture capital since it was founded in 2011. NLRB documents list the "number of workers at dispute location" as 10,000.

WeWork, an office space rental startup that helped popularize co-working, recently faced similar allegations from the NLRB over arbitration clauses. That company agreed to amend its employee handbook as part of a settlement filed in September.

If Postmates fails to reach a settlement with the employee who filed the initial charges with the NLRB, there will be a hearing in Chicago on January 26, 2017.

Postmates declined comment on the NLRB complaint citing company policy on "pending litigation or similar matters."

Evan Vucci / AP

RENO, Nevada — Donald Trump supporters cheered Mike Pence Wednesday at a campaign rally in Nevada following the first vice presidential debate. But despite Pence's oratorial success where Trump was deemed a loser, supporters said they'd stick with Trump over the Indiana governor.

Pence was widely perceived to have been the victor in Tuesday's sparring match, and Trump went so far as to claim credit for the win:

Attendees of the rally were accordingly unanimous in their support for Pence's performance in the debate. "On a scale of one to 100, Pence got I'd say about an 85, as high as a 90," said Louis Entrekin of the debate. "The other guy, he just kept regurgitating himself. He got a five."

Louis Entrekin

Caroline O'Donovan / BuzzFeed News

Compared to Democratic rival Tim Kaine — who Trump supporters said repeatedly interrupted their candidate — Pence came off as "dignified, honest and appealing," according to Gordon, who drove from Sacramento to Reno for the rally. "You could put him in the first chair," he said.

But, while Trump has lagged in the polls following his first debate with Hillary Clinton, his supporters said they wouldn't want to see the two switch places.

"I don't ever see Trump being number two," said Beverly Shimizu Krisch, a Reno resident and 71-year-old daughter of Chinese and Japanese immigrants.

Another Reno resident, a 70-year-old contractor named Bob, said while Pence "did a great job" and "kept his cool," he doesn't inspire the same feeling that Trump does. "Pence has experience in Congress. He's done a good job for his state," Bob said. "But Trump is as if I was going to run for president."

Like Trump, Bob said he'd both made and lost a lot of money in the contracting business, and felt he had "a lot in common" with the candidate.

Bob

Caroline O'Donovan / BuzzFeed News

In fact, supporters said that what's most compelling about Trump is that he doesn't sound like a polished, professional politician. "I take Trump as what he is — a hardcore, hard-driving business man," said Entrekin. "He don't know any other way."

Chris Vaughn, a Reno resident who initially supported Ben Carson, echoed that sentiment. "I don't care if the language is flowery as everyone would like. People are pissed off at career politicians and 'looking presidential'," she said. "I think it's more about acting presidential than looking and sounding like one."

Chris Vaughn

Caroline O'Donovan / BuzzFeed News

Many supporters felt that the Trump-Pence ticket works best with Trump as the charismatic persona leading the way, and Pence as the more experienced legislator.

David, who attended the rally along with his wife Julie, said Pence's "experience is foundational for providing the other half. He completes Trump. It's like marriage."

Caroline O'Donovan / BuzzFeed News

Scott from Sacramento put it more simply: "Trump's the driver. He can initiate, and Pence can do his legwork."

But just because Trump's supporters prefer him in the driver's seat doesn't mean they don't think he could learn a thing or two from Pence, especially with the next debate coming up on Sunday. "I think [Trump] should have started a little earlier in being a bit more presidential," Gordon from Sacramento said.

Julia and Cherie, two sisters from Reno who attened Trump's rally together, both like Trump precisely because he's not a seasoned politician. But Cherie said would like him to lay off the "midnight, early morning tweets." And Julia agreed.

"On the Sunday debate, I want him to look at her and say, I'm here to talk about the problems of the country," she said. "But that's not his personality I guess."

Ultimately, there's something of a conflict between what Trump's supporters like about him — namely, that he's not a politician — and what they think he needs to do to win the election. Michael and Jan, enthusiastic and appropriately dressed Trump supporters from Sacramento, wrestled with this inherent contradiction and how it will come into play during the debate.

Michael and Jan

Caroline O'Donovan / BuzzFeed News

"I want him to be Pence-y," said Michael. "He shoots from the hip too much. He's not a politician."

"That's what people want," Jan countered. "They're tired of the bullshit and lies and backstabbing."

Michael agreed, but said if they want to win, Trump fans "want to get voters from the undecideds and independents and ex-Democrats."

Jan, nodding, assented. "Yes," she said. "He needs to keep his cool."

LINK: Trump Repeatedly Mispronounces “Nevada” While Insisting On Pronouncing It Correctly

John Macdougall / AFP / Getty Images

日本でも急成長する民泊仲介サイト「Airbnb（エアビーアンドビー）」。同社は9月8日、差別防止策を公開した。なぜ対応を迫られたのか？海外で浮上するシェアリング・エコノミーの深い闇とは。

「俺はニガー（黒人の蔑称）は嫌いだ。だからお前の申込みは無効にする。ここは南部だよ、きみ。別の場所を見つけな。ニガー野郎」

シャニ・テイラーさんは5月31日、ケロッグ経営大学院MBAの同級生がこんな人種差別に遭ったとTwitterで明かした。

憎悪に満ちたホストの言葉。SNS上で瞬く間に非難が高まった。Airbnbのブライアン・チェスキーCEOがツイートで釈明する事態に発展した。

このホストを永遠に禁止した、という。

Airbnb報道担当のニック・パパス氏はBuzzFeed Newsの取材に「このような言動は受け入れられない。我々の方針にも我々が信じるものにも反する。安心して過ごせる部屋の用意といったサポートをするため、ゲストに連絡した」と話した。

訴訟も

Airbnbのホストによる黒人差別問題は根深い。

グレゴリー・セルデンさんも5月4日、Airbnbで遭った差別をTwitterで暴露した。

フィラデルフィアのある部屋を予約しようとしたセルデンさん。だが、ホストに「空きがない」と断られた。そこでこんな実験をしてみることにした。

「トッド」「ジェシー」という名前の白人男性のフェイクアカウントをつくって、同じホストに空き状況を聞いた。すると「予約できる」と返事。

公民権法と公正住宅法に違反するとして、セルデンさんはAirbnbを訴えた。

SNSで問題提起

セルデンさんが使った#AirbnbWhileBlack（黒人がAirbnbを使うとき）というハッシュタグは昨年7月、カーティナ・クリテンデンさんが始めた。自身も差別を経験した黒人女性だ。

#AirbnbWhileBlackを使って、差別が次々と報告されている。

「プロフィール写真で陸軍の制服を着ている。退役軍人なら差別されることが少ないから」

「空き表示が出ているのに、予約の問い合わせをしても返事をもらえないことばかり。『差別的』シェアリング・エコノミー」

代替サイトを起業

起業家ローハン・ギルクスさんも5月13日、人種差別エピソードをImgurに投稿した。

ギルクスさんはアイダホ州の友人を訪ねるため、7月1〜5日に友人宅近くの部屋を申し込んだ。だが、満室だと断られてしまう。友人は6月でもいいというので、別日程で予約しようとしたところ、これも断られた。

不審に思ったギルクスさん。白人の友人に同じ予約を頼んだ。すると「予約可能」という返事をもらえた。

これで終わらないのが起業家。「Innclusive」という民泊仲介サイトを立ち上げた。「Inn＝宿」と「inclusive＝多様性の受け入れ」を掛け合わせている。

ギルクスさんはBuzzFeed Newsの取材に「投稿が注目されて初めてAirbnbから電話があった。Airbnbの対応が早ければ、ここまで至らなかった」などと説明する。

相次ぐ起業

ラッパーで起業家のステファン・グラントさんも昨年10月、音楽フェスの出演のため、Airbnbを使ってジョージア州の部屋を借りた。こぎれいな離れの建物だった。

庭でフェスの準備をしていると、パトカーがやってきた。「近所の人が俺たちを強盗だと思ったようだぜ」とグラントさんは動画ツイート。

最後には「確認」という名目で勝手に部屋に入ってくる白人まで現れた。

ホストが差別したわけではない。だが、この経験を契機に、黒人がもっと安心して旅行できるようにグラントさんは民泊仲介サイト「Noirbnb」を始めることにした。Noirは仏語で「黒」を意味する。

差別を暴く論文

「Airbnbで、黒人特有の名前のゲストは、白人に比べて、16％ほど予約申込みを受け入れられる確率が低い」

こんな計量経済学の論文を1月、ハーバード経営大学院の准教授らが発表した。

論文の実験では、名前以外は条件が同じ架空アカウントを20人分作った。黒人と白人に特有の名前が10人ずつで、プロフィール写真は載せなかった。

全米5都市の約6400部屋で、この架空ゲストたちが予約申し込みをして、受け入れられるかを調べた。結果、黒人が受け入れられたのは42％、白人は50％。差は8パーセンテージポイント、つまり16％あった。

筆者らは「差別は、より規制されているオフライン市場では減ったが、オンライン市場では逆に現れ、続いている」「Airbnbの設計は広範囲な差別を可能にしてしまう」と指摘する。

オンラインプラットフォームはデータ分析で差別を見つけやすいはずで、特性を生かした監査をすべきだと提案する。

差別を許すビジネスモデル

Airbnbで差別防止が難しいのは、そのビジネスモデルに要因がある。ホストはAirbnbの社員ではない。もちろん米公正住宅法はこうした差別から消費者を守るためにある。だが、ビジネスではなく、個人の行動にどう適用するか。課題が残る。

LGBT差別

差別は人種に限られない。

BuzzFeed News / Getty Images

LinkedIn co-founder Reid Hoffman would like to see Donald Trump's tax returns, and he's willing to pay up to $5 million for the opportunity.

On Monday, Hoffman pledged his support for a Crowdpac.com crowdfunding campaign aimed at pressuring Trump into releasing his tax returns. The campaign was started by a US military veteran named Pete Kiernan, who says he’ll donate the cash — almost $5,000 so far — to 10 veterans affairs groups if Trump releases his tax returns.

“Trump claims to love veterans,” reads Kiernan’s Crowdpac.com page, “and so we’re asking him to put his money where his mouth is.”

If Kiernan succeeds and Trump releases his tax returns, Hoffman — whose net worth increased by $800 million in a single day this year when Microsoft acquired LinkedIn — says he will quintuple the sum raised. So, if the Crowdpac campaign meets its goal of $25,000, and if Trump makes public documents that he’s so far vehemently insisted on keeping private, Hoffman would donate $125,000; the more money the campaign raises, the more Hoffman will donate, with a cap at $5 million.

In a post on Medium published Monday, Hoffman noted that $5 million is the same amount that Trump himself pledged to donate to charity during the 2012 election if President Obama agreed to his request to release college records and passport documents.

“Given Trump's vocal support of veterans, I imagine he will recognize the great good that can come from Kiernan's proposal,” Hoffman writes. “But taking Trump's own 2012 offer to President Obama into account, I'd like to assist Kiernan in his campaign.”

It’s worth noting that Hoffman was an early investor in Crowdpac, which bills itself as a crowdfunding platform designed for political campaigns. Hoffman is a partner at Greylock Partners, but participated in Crowdpac’s $6 million Series A in early 2016 as an independent investor.

It’s also worth noting that this is not the first Crowdpac campaign Hoffman has publicly involved himself in — or even the first this month. Last week, Hoffman announced that he would donate $25,000 to a campaign to recall Judge Aaron Persky, who presided over the Stanford sexual assault case. As of this writing, that campaign has amassed $40,000 toward its $250,000 goal.

Though Hoffman, as an investor, stands to profit from Crowdpac’s success, some Silicon Valley luminaries see his investment in progressive political causes as a worthy use of his wealth:

The deadline for the Crowdpac campaign is Oct. 19, the date of the final presidential debate.

Via Flickr: 65487073@N03

Nextdoor, a location-based social network for neighbors that has more than 10 million registered users, is rolling out a new tool today that the company says has reduced incidents of racial profiling on its network by 75% during tests. In recent years, so many people have used Nextdoor to report things like black men driving cars or hispanic women knocking on doors as suspicious or even criminal that the site has become known as a hub for racial profiling.

The new tool, an algorithmic form for reporting crime and safety issues, has been in beta for an ever-increasing portion of Nextdoor’s 108,000 neighborhood groups since May. This feature, which automatically identifies racially coded terms and prevents users from posting without supplemental descriptors, goes live for all users today.

“The impact of being racially profiled in general is terrible,” CEO Nirav Tolia told BuzzFeed News on Tuesday. “It runs counter to the mission of Nextdoor. It’s something we feel morally obligated to take seriously.”

Racial profiling became an issue for Nextdoor in 2015, when a number of news outlets reported on the frequency of posts about crime or suspicious behavior that mentioned an individual’s race, but little or nothing related to actual criminal activity. In many cases, these posts would refer to people of color doing things such as talking on the phone or walking a dog.

Tolia said it wasn’t the bad press, but the work of civic groups in Oakland that brought the issue to his attention. Nextdoor touts its collaborations with police departments, city governments, and other public agencies. Last fall, Oakland Vice Mayor Annie Campbell Washington encouraged Oakland city departments to stop using the app to communicate with citizens until Nextdoor addressed the issue of racial profiling. By October, Tolia’s team was holding working groups with advocacy groups and city officials, and together they came up with a solution.

The idea, which Tolia credits to members of a group called Neighbors for Racial Justice, was to change the way crime and safety issues are reported on Nextdoor. Instead of a blank text box and subject line, it was suggested that Nextdoor design a form that more closely resembles a police report or 911 dispatcher questionnaire. By explicitly requesting details about height, clothing, and age, they would discourage people filing reports from focusing exclusively on the race or ethnicity of the subject.

Nextdoor features a wide variety of post categories — Classifieds, Events, etc. — but it’s the Crime and Safety section where people tend to focus on race to the exclusion of other salient details. As of today, Nextdoor neighbors posting a “crime” or “suspicious behavior” to the site will be warned against allowing an individual’s race to color their interpretation of events. And if their post focuses too much on the race or ethnicity of the subject, they’ll be prevented from publishing it. “When race is invoked, we create a higher bar,” Tolia explained.

For example, try to post about your car windows being smashed, and you’ll be prompted with this message:

Try to describe someone with just a racial characteristic, and you’ll see this prompt, asking you to be more descriptive:

Nextdoor claims this new multi-step system has, so far, reduced instances of racial profiling by 75%. It’s also decreased considerably the number of notes about crime and safety. During testing, the number of crime and safety issue reports abandoned before being published rose by 50%. "It's a fairly significant dropoff," said Tolia, "but we believe that, for Nextdoor, quality is more important than quantity."

Vice Mayor Washington said she’s “thrilled” with the results Nextdoor has achieved. “I don’t think a lot of technology companies would have taken the steps they did, and made significant changes to their platform,” she said.

When tech companies come under fire for failing to take race issues seriously — Snapchat, Twitter, and Airbnb are examples — critics often assert that, were minorities better represented on the staffs of those companies, the same mistakes might not have been made. But Tolia, who himself identifies as a person of color, said that while he's working actively to diversify Nextdoor's staff, when it comes to racial profiling, "we believe we get the best information from our members, and in this case, our advisers.”

But not all of Nextdoor’s advisers on the racial profiling project are satisfied with how the process turned out. Two founding members of Neighbors for Racial Justice, Audrey Williams and Shikira Porter, said Nextdoor left them out of the development process after a flurry of early interest. Porter told BuzzFeed News that she continues to see instances of racial profiling in her Nextdoor neighborhood despite the rollout of the form. Nextdoor confirmed that the company has not met with Neighbors for Racial Justice since the test pilot began in April, but said it was made aware of only two instances of racial profiling that had slipped through its algorithms in the last few months.

“We've been doing the work of consultants for them, and they’ve been taking it as free, pro-bono, volunteer advice from the community,”said Williams, who works in digital marketing. “And we’ve been happy to give it, because it makes our lives better. But over time, it began to feel a bit like exploitation.”

Nextdoor will hold a conference call for local stakeholders on Wednesday, but neither Williams nor Porter will be able to attend. Porter said it “didn’t feel right” to have Neighbors for Racial Justice attached to a project they felt the organization hadn’t been given a chance to sign off on.

"We appreciate working with [Neighbors for Racial Justice] to create these improvements,” said Tolia in a follow-up email. “We are encouraged by the progress, but know there is still more work to do."

However, it’s clear that the contributions of Neighbors for Racial Justice and other local organizations to Nextdoor’s efforts were integral to the design and execution of the final product. Some of the copy Nextdoor ended up using in the form — such as, “Ask yourself, 'Is what I saw actually suspicious, if I take race or ethnicity out of the question?'” — came at the suggestion of Neighbors for Racial Justice.

The new racial profiling form isn’t the only change Nextdoor has made in service of tempering racial profiling on its platform. Last November, the company introduced a checkbox that allows users to flag posts for racial profiling. More recently, it's trained the group of in-house customer service representatives that reviews such posts in conflict resolution and “cultural humility” with an eye towards helping users understand why their posts were flagged, and how racial profiling negatively impacts whole communities.

It’s unusual for a tech company to take such an active role in policing its users, or to make an investment in educating them about social and cultural issues. In addition to relying on community members, Nextdoor also hired consultants to help, among other things, define what racial profiling outside of a police setting even is. Debo Adegbile, a civil rights attorney with the NAACP who was nominated for US assistant attorney general by President Obama, and Grande Lum, a race relations expert with the Department of Justice, both worked as advisers to Nextdoor.

Because the new form makes it less likely that users will post to Nextdoor, Tolia said there's a "business cost to doing this." But given the way Airbnb, Twitter, and other tech companies have struggled with issues of race in recent months, it’s easy to see Tolia’s decision to meet the racial profiling problem head-on not just as morality, but also as good business sense.

“Let’s not be fooled,” said Audrey Williams of Neighbors for Racial Justice. “It’s a win for us, and it’s a win for them.”

An Uber driver in San Francisco, California, on May 7, 2015.

Robert Galbraith / Reuters

A federal judge has rejected an up-to-$100 million settlement Uber had agreed to pay to two class action lawsuits from drivers who argued they should be classified as employees rather than independent contractors, calling it “not fair, adequate, and reasonable.”

The drivers had argued they should be entitled to benefits like reimbursement for expenses such as gas while driving for Uber. The settlement, which would have allowed Uber to continue classifying drivers as independent contractors, would have left some of the about 385,000 Uber drivers in California and Massachusetts with as little as $10 each in settlements.

A sticking point for US District Judge Edward Chen was how part of the settlement would resolve an estimated potential $1 billion in liabilities drivers could sue for under California’s Labor Code Private Attorneys General Act (also known as PAGA) for only $1 million.

“The settlement, mutually agreed by both sides, was fair and reasonable,” Uber said in a statement. “We’re disappointed in this decision and are taking a look at our options.”

Shannon Liss-Riordan, the drivers’ attorney, told BuzzFeed News, "It is possible the parties could reach a revised agreement that satisfies the court’s concerns. ... But if not, as I’ve said before, I will take the case to trial and fight my hardest for the Uber drivers."

The lead plaintiff in the case called the settlement “disastrous” in early May and dismissed Liss-Riordan as his lawyer. Following outcry from other class members over their disappointment in the terms of the settlement, Liss-Riordan offered to reduce her own fee by $10 million, leaving her firm with between $11 million and $15 million. (The settlement offered $84 million to drivers, plus an extra $16 million if Uber were to go public.)

Earlier today, Uber said it was preparing to launch a pilot program in Pittsburgh for passengers to hail autonomous vehicles – a move that will eventually remove drivers from the equation entirely. For now, the 100 autonomous Volvos will still have humans in the driver’s seat.

The rejection of the settlement opens the class up to “substantial risk” regarding arbitration clauses signed by some drivers. “This risk would have the effect of substantially altering – if not effectively terminating – the class action in this Court,” Judge Chen wrote. Because of these arbitration clauses, Liss-Riordan told BuzzFeed News the scope of the case and number of drivers involved could be limited to about 8,000 drivers.

“The drivers are in a holding pattern for now, while we see whether Uber will return to the settlement table or not,” Seattle University labor law professor Charlotte Garden told BuzzFeed News. “If not, and if the arbitration issue goes badly, then most drivers will be left to individually arbitrate their claims.” If that happens and the class is disbanded, Garden said it’s “unrealistic” that many drivers would pursue individual arbitration. That would be a win for Uber.

As part of the settlement, Uber had also agreed to clarify its tipping policy by allowing drivers to place signs in their cars noting that tips are not included. The judge was “not convinced” making the tipping policy clearer would actually substantially increase drivers’ income, according to a court filing, particularly because Uber has “actively discouraged tipping, arguing that it is inconsistent with its business model, drivers’ interests, and a positive rider experience.”

Last month, Uber argued in federal court that two subpoenas from the National Labor Relations Board (NLRB) — which is investigating the company’s labor practices — should be stayed, because this class action settlement would have resulted in the withdrawal of the underlying charges of the NLRB's case. Now, however, that settlement has been rejected, so Uber’s argument against the subpoenas is moot. In March, Uber was accused of impeding the NLRB’s ongoing probe into drivers’ employment status.

Uber has in recent months added several measures to placate drivers, including expanding a fee for passengers who make drivers wait longer than two minutes and giving drivers the ability to cash out instantly whenever they wish.

Caroline O'Donovan / BuzzFeed News

What do you call a typewriter with no keys? Though it sounds like a riddle, Stanford historian Tom Mullaney told BuzzFeed News, it’s not. A Chinese typewriter has 2,500 characters but zero keys, and Mullaney owns more of them than anyone else in the world.

The Chinese language has over 75,000 characters — way too many to fit on a keyboard with a single key per character. The inventor of the original Chinese typewriter slimmed that number down to just the 2,500 most used characters, but even still, that’s too many to type out as you would on an English-alphabet typewriter.

On a Chinese typewriter, those 2,500 characters correspond to 2,500 tiny metal squares, which lie side by side in a tray bed. Instead of pressing individual keys with individual fingers, the typist uses a knob in each hand to move a scroll of paper up and down, left and right over the keys. When the machine is on top of the character you want, you press down on a lever. Then, Mullaney explained, the metal piece gets pushed (or sucked up) into the type chamber, inked, and struck the surface of the paper, before being ejected and spit back out into the spot in the grid where it was.

Here's what that process looks like:

youtube.com

That’s a lot of steps for just one character. Each stroke of the lever takes just a second, but the thing that slows Chinese typists down, and for years made them much slower than Western typists, is the distance between characters — especially because the typewriter's characters were organized according to what Mullaney called “dictionary order.”

“The problem is, just like a dictionary, words don’t necessarily go together,” he said. “'Aardvark' and 'apple' don’t show up in sentences together, despite being in the same part of the dictionary.”

So in the '50s, typists started dumping out all of the characters and building their own custom tray beds from scratch, using tweezers to array characters based on how commonly they were used together.

A tray of metal characters that came with a Chinese keyboard in so-called "dictionary order."

Caroline O'Donovan / BuzzFeed News

The thinking was, “If I have to write the name Mao Zedong, which is three characters, over and over, why would I move across the tray bed?” Mullaney said. “Let’s put them right next to each other.” The result was “completely personalized, completely individualized, completely idiosyncratic” tray beds.

This innovation made typing much faster, according to Mullaney. “If the top speeds in the 1930s and '40s was 20 characters per minute, and that was for a very fast typist, after the 1950s ... you had top speeds of 55, 60 characters per minute,” he said. “That’s a three-time increase in the speed of the machine just by rearranging the characters in the tray bed.”

But he also sees it as an early, handcrafted, analog version of the kind of predictive text we now find in Google autocompletes and advanced smartphone keyboards. "Predictive text of that sort was already baked into Chinese typewriters in the mechanical realm, and then gets brought into the realm of Chinese computing in the '60s and '70s.” Nowadays, he said, in China, pretty much every interaction a person has with a computer interface — from word processors to search bars — has predictive text built in.

Caroline O'Donovan / BuzzFeed news

These days, the QWERTY keyboard is standard in China — but instead of each key corresponding to a single letter in the alphabet, each key corresponds to a sound, and predictive text suggests a character that goes with those sounds. As a result, he said, after decades of relying on slower technologies, "the fastest Chinese computer inputter using a QWERTY keyboard input ... is faster than the fastest alphabetic typist.”

Stanford historian Tom Mullaney explains how *not* to use a Chinese typewriter.

Caroline O'Donovan / BuzzFeed News

Both alphabetic and Chinese typewriters had an indelible impact on the way we communicate today, but only the Chinese typewriter has been almost entirely forgotten. Only a few institutions in the United States, including the Huntington Library and the Museum of Chinese in America, have Chinese typewriters, and three Chinese speakers I spoke with about this article weren’t aware they had ever existed.

So Mullaney, who amassed one of the largest collections in the world more or less by accident, is trying to “Save the Chinese Typewriter," starting with a (successful) Kickstarter campaign. His first Chinese typewriter, which is seafoam green and was the leading model in China throughout the 1970s, was given to him by a man who worked at a church in San Francisco. The typewriter had been used to print Chinese-language bulletins, but its services were no longer needed, and the man didn’t know what to do with it.

“If these were Western typewriters, you’d have the pick of the litter in terms of where to donate a machine like this or sell it on the antique market,” Mullaney told BuzzFeed News on a hot afternoon outside on Stanford’s campus, where he is a history professor. “There’s no such thing for East Asian information technology and Chinese typewriters.”

In the end, Mullaney's Kickstarter campaign raised more than $13,491, and his traveling exhibition will — “outside of maybe the National Library of China and the National Diet Library in Japan” — feature more Chinese typewriters than any other collection in the world. The tour kicks off with an exhibition in the San Francisco Airport in 2017.

View Video ›

Facebook: video.php

Flickr / HJL

On Wednesday, Kate Vershov Downing, a corporate lawyer for an enterprise cloud company called ServiceNow, resigned from the Palo Alto Planning and Transport Commission. The reason, she said, was that she and her husband — a software engineer — could no longer afford to live there.

“We rent our current home with another couple for $6200 a month,” Downing wrote in a post on Medium. “If we wanted to buy the same home and share it with children and not roommates, it would cost $2.7M and our monthly payment would be $12,177 a month in mortgage, taxes, and insurance. That’s $146,127 per year — an entire professional’s income before taxes. This is unaffordable.”

Like dozens of her friends have already done, Downing will be relocating outside of the Bay Area — a financial decision a new report says is becoming a trend.

“Twenty six percent of software engineers in San Francisco are searching for jobs out of state,” said Indeed’s Paul D'Arcy, who presented this research to San Francisco’s Chamber of Commerce on Wednesday afternoon. D’Arcy said that, specifically, they’re looking in Seattle, Portland, Austin and Denver — cities that have emerged as tech hubs, but where salaries go a little further.

According to Indeed, tech workers in San Francisco are making an average $113,497 a year. In Seattle, that figure is lower, at $98,215, and in Austin, lower still, at $94,025. But when those salaries are adjusted for cost of living, which is much higher in California than it is in Texas, that order is reversed. While San Franciscan techies are spending around 37% of their income on rent, Austinites are only shelling out 23% of their earnings on housing.

“We've seen, globally, tech salaries equalize,” said D’Arcy during his presentation on Wednesday. “It used to be people in San Francisco made much more, but that gap is decreasing. With the dramatic increase in cost of living, the economics are becoming less favorable to workers.”

As a result, according to Indeed’s data, San Francisco has has fallen behind Washington DC and Austin when it comes to desired destinations for people searching for jobs in tech. (San Jose is still in the lead, though.)

If the city is interested in reversing this trend, D’Arcy said, it will need to invest heavily in transportation systems that would help commuters access cheaper housing; if companies want to help, he said, they’ll need to offer workers based in San Francisco increased flexibility — aka, permission to work from home.

San Francisco is the third most unequal city in America after Bridgeport, CT and New York City, according to Indeed’s report, and that gap is widening. As the top earners get richer, the lowest earners — those in the bottom 20th percentile — are actually seeing their income decrease. It’s a problem Kate Downing is thinking about, even as she prepares to bail on Silicon Valley proper.

“It’s clear that if professionals like me cannot raise a family here,” she wrote, “then all of our teachers, first responders, and service workers are in dire straits.”

Reich testifying on income inequality before the Joint Economic Committee in 2014.

Win Mcnamee / Getty Images

Former United States Labor Secretary Robert Reich always suspected Uber driver pay was low — but he didn't know how low until recently, when he read a BuzzFeed News report that found that Uber drivers in three cities made less than $14 an hour on average.

“It was about where I thought they'd be,” he told BuzzFeed News on a recent sunny afternoon in Berkeley, California. “But I was surprised the payment was so low after costs.” He also called Uber’s price cuts, which tend to infuriate drivers despite Uber’s claim that they increase demand for rides, a “blunder.”

“Tech is affecting the entire workforce very dramatically.”

Reich has spent about half of the 19 years since he left the Clinton administration and the Department of Labor working at the University of California, Berkeley, where he’s a professor of public policy. Between writing best-sellers, stumping for Bernie, and maintaining an active Facebook presence, he has also been observing the impact of the tech industry on the American workforce.

“Tech is affecting the entire workforce very dramatically,” Reich said. “A lot of people don't know how much they're being affected. It's not just Uber and Airbnb — it's really the whole employment relationship.”

Here’s what he told BuzzFeed News about why the gig economy — a group he said makes up 30% of the American workforce — is bad for workers, how it’s making labor worse, and what we can do to fix it.

Gig workers are the new blue-collar workers.

Not only are wages for gig workers low, but the work is also unpredictable, according to Reich — "not just in how much you earn, but when you will be needed." For example, it used to be that it was an employer’s job to predict demand; if customers didn’t come into a store, the people staffing the cash registers still had to get paid. Nowadays, it’s the Uber drivers of the world who take a hit if demand is slow. Said Reich: “If you have a lot of fixed costs that you have to bear as a person with a family and a home and utilities and auto costs and so on, that unpredictability can really cause havoc.” The result of this instability and frustration, he continued, is that gig workers are becoming part of the “angry working class.”

“The new blue-collar workers are gig-economy workers. If you define gig economy broadly to include part-timers, the self-employed, independent contractors, freelancers and temps, they’re on a downward escalator.”

The gig economy is both a product and a cause of income inequality.

"We are actually reducing the quality of life for millions of people.”

Airbnb and Uber say they’re helping middle-class Americans get by, and Uber justifies paying drivers at a rate, that is in Reich’s words, “not all that far removed from minimum wage” by saying it’s only supposed to be additional income on top of another job. But 35 years ago, Reich said, employed Americans didn’t need supplemental income just to pay the bills. If people still had good wages and labor protections, they wouldn’t need to turn to these platforms for extra cash. And ironically, Reich said, it’s companies like Uber, which popularized the idea of a digital on-demand labor marketplace, that are driving down the value of work, causing people to need extra jobs.

“I think the digital platforms are making it easier for companies to move in this direction,” he said. “And it's happening in all sorts of ways. I'm not blaming tech companies. They are doing what the economy wants them to do. There's all sorts of money to be made in making it easier for everyone else to move from fixed costs to variable costs. The problem is, if you believe, as I do, that the heart of the economy is a good job, with predictable wages and benefits, we are actually reducing the quality of life for millions of people.”

The solution is more and better benefits.

Thanks to Obamacare, many gig workers now have health insurance. But Reich thinks we need income insurance, too. Here’s how it would work: “If you, next month, are earning half as much as you earned this month, the wage insurance or income insurance would give you, say, half the difference between the two,” he said. “It smooths out your income stream over time.”

Reich also think we need portable benefits — in which tax advantages normally given to employers are instead attached to individuals. (On Thursday, the Department of Labor announced $100,000 to fund portable benefits research. Many tech companies already support the idea.)

But the easiest way to make sure workers get the benefits they need is to help them retain employee status. To that end, Reich wants to make it harder for companies to classify workers as contractors — even Uber. Said Reich: “I think Uber made a terrible strategic blunder. A few years ago, a lot of drivers really liked the company. They felt they were getting a fair deal. These price changes coming at them with little or no warning has made them far less generous with regards to the company. They feel that they are being treated badly, and the fiction that they are not employees becomes even more obvious.”