On the morning of May 3, 2000, in front of an unremarkable San Mateo, California, office building with a bank on the ground floor, the music industry collided with the technology business head-on. At 10:30 a.m., a chauffeured Chevy Blazer pulled up carrying Metallica’s drummer, Lars Ulrich, the band’s lawyer, Howard King, and two men to help them deliver thirteen document boxes to Napster’s fifth-floor headquarters. The boxes held a list of 335,000 people who had downloaded Metallica songs from the first file-sharing service that made such widespread piracy possible.
The delivery of the boxes was intended to be theatrical, and the band had tipped off reporters to hear Ulrich speak about how musicians had to stand up to technology companies trying to profit from their work. But Napster staged a rival performance. The fledgling company produced its own supporters, who shouted down Ulrich and called him a greedy rock star; one even smashed his Metallica CDs with a sledgehammer. Most wore matching T-shirts with the Napster logo, and some seemed to know company staffers.1 But journalists saw the chauffeured Blazer opposite the upset fans and concluded that Metallica was on the wrong side of a new generation gap.
To Napster’s users, the real rock star that day was Shawn Fanning, the nineteen-year-old hacker who had written a revolutionary file-sharing program in his Northeastern University dorm room. He had formed a company with his uncle, John Fanning, who had secured initial funding and moved the operation to California in September 1999. By then, the service, which let music fans copy files from one another’s computers, had more than 100,000 users. When the RIAA sued Napster three months later, the resulting publicity only made it more popular. By the time Ulrich delivered his boxes of documents, three-quarters of U.S. college students had used the program.2 The sheer breadth of music available, combined with the opposition of major corporations, made Napster seem less like a technology company than a youth movement, complete with slogans and T-shirts.
Metallica tried to tell the other side of the story. This wasn’t about the money—they had plenty already—and, unlike most bands, they had allowed fans to record and trade their performances for years. What upset them was the leak of an unfinished version of “I Disappear,” a song they had recorded for that summer’s Mission: Impossible II.3 The previous month Metallica had sued Napster for copyright infringement and racketeering. The company might look like the underdog, Ulrich said of Napster at the time, but “they’re looking at [the business] like one day there’ll be a major IPO or an AOL-type company is going to come and buy Napster out for a gazillion dollars.”4 (As it happened, about a month later Napster received a venture capital investment.)
Metallica was also counting on promised backup that never appeared—a fact that has never been reported. “We thought it made sense to put an artist’s face on the litigation—Metallica didn’t want to do it alone, and there were commitments from three or four other major artists to do this, including a major country act,” King says. But they dropped out when they saw fans threatening to boycott Metallica. King only managed to bring in his client Dr. Dre, who, like most hip-hop artists, isn’t exactly sheepish about wanting to make money. No other major artists came forward. “The band knew there would be a negative reaction,” King says, “but they were surprised at how big it was.”
Ask most people what humbled the mighty major labels and you’ll hear the same answers: they failed to negotiate with Napster, missed their chance to turn file sharing into a legitimate business, and got saved by Apple’s iTunes Store when Steve Jobs dragged them, kicking and screaming, into the digital age. This makes for a compelling narrative about how a college student revolutionized a calcified business that didn’t give consumers what they wanted. But while the labels certainly moved too slowly, the real story is far more complex.
The labels did negotiate with Napster, even as they fought in public, and each side walked away from talks at different times. Even if they had made a deal, a legal version of Napster would have faced competition from second-generation file-sharing services like Grokster and Kazaa, which would have offered for free what Napster charged for. And while Apple gave the labels a workable digital business model, its iTunes Store also replaced sales of $15 albums with ninety-nine-cent songs.
The real story is that for-profit technology companies deliberately set out to make money from piracy and never came up with a workable plan to pay artists. Although Shawn Fanning sincerely wanted to find a better way to distribute music and Napster negotiated with labels, Grokster and Kazaa—based in the offshore tax havens of Nevis and Vanuatu, respectively—didn’t really try. Even with the growth in online distribution, sales of recorded music in the United States were worth $6.3 billion in 2009—less than half their 1999 value of $14.6 billion.5 The music business became trapped in a downward spiral: starved for revenue, major and indie labels alike have laid off thousands of staffers, dropped hundreds of artists, and cut investment in recording, marketing, and tour support. Emerging artists face a choice between signing new contracts that give labels revenue from live performances and merchandise sales or trying to fund their own careers in a touring market dominated by superstars.
In 2000, Napster had no problem presenting itself as the underdog. Over the previous decade the recorded music business had grown and consolidated into five corporate behemoths—Universal Music Group, Warner Music Group, Sony Music Entertainment, Bertelsmann Music Group, and EMI Music—that among them controlled more than three-quarters of the U.S. market. All of them generated enough cash to run as independent fiefdoms within conglomerates that seemed to barely understand how they operated. One key to their success was the rise of the CD, which spurred consumers to replace their old records and cassettes. Labels also phased out singles, which meant that novelty hits like “Thong Song” and “Who Let the Dogs Out?” drove sales of $15 CDs. And as the cost of manufacturing CDs declined, labels didn’t lower prices—a decision that came back to haunt them as music buyers began to see whole packages of blank discs on sale for a few bucks.
By the time Fanning released Napster in June 1999, the major labels had been dealing with online piracy for years on a smaller scale. When ordinary consumers started accessing the Web, full audio files were too big to download easily. But hackers embraced the MP3 format, which compresses songs by leaving out sounds unnoticeable to most human ears. At first record labels viewed sites that featured pirated MP3s the same way they saw bootleg record-pressing plants: a problem that could be contained with the right combination of lawsuits and lobbying prowess. But technology advanced faster than they expected. “There was a perfect storm around 1999 when manufacturers started putting CD burners in computers, blank discs went down to under a buck, and MP3 came along,” remembers Ted Cohen, a technologically savvy executive who shaped the digital strategy at EMI Music.
Like other students in his dorm, Fanning liked downloading MP3s, but it took time to find the ones he wanted. So he decided to write a program that would index the MP3 files on every computer running it. As users logged in to the system, the songs on their computers would become available to everyone else signed in to the network. Fanning’s main innovation was that the central server which ran the program wouldn’t store the files themselves—just the list of where they were located. He called the program Napster, after a high school nickname inspired by his hair.
The RIAA had been dealing with pirated MP3s by sending Digital Millennium Copyright Act takedown notices to the sites that hosted them. But that wouldn’t work with Napster. Since the music could come from any computer on the network, labels had to either contact individual users or deal with Napster directly. But the company, then run by the former venture capitalist Eileen Richardson, was in no hurry to talk. On December 7, 1999, the RIAA sued, seeking $100,000 in damages per song infringed.6
Napster’s reluctance to negotiate may have been deliberate. A company strategy document written around that time said, “We will use the hook of our existing approach to grow our user base, and then use this user base coupled with advanced technology to leverage the record companies into a deal,” according to Joseph Menn’s All the Rave: The Rise and Fall of Shawn Fanning’s Napster, the definitive book on the subject.7 Essentially, Napster was using free music to assemble an audience big enough to give it a negotiating advantage. It was a strategy that seemed to inspire scores of other start-ups, including—Viacom would later argue in its lawsuit—YouTube.
Before starting at EMI, Cohen had consulted for Napster, and he remembers bringing Napster’s vice president Bill Bales to meet EMI’s executive vice president Jay Samit at the Capitol Records Building in Hollywood. “Bill laid out what Napster was all about, and Jay said, ‘Let’s figure out an experiment,’ ” Cohen remembers. As he walked Bales out of the office, Cohen asked when they would get back to Samit with a proposal for how the two companies could work together. “And Bales said, ‘We’re not—we just need to stall for a while.’ He said, ‘I don’t think we’re going to have to make deals with the labels—we think there’s going to be enough people using Napster that they’ll change copyright law.’ ”
In public, Richardson emphasized Napster’s legitimate uses, such as downloading tracks by emerging bands that wanted to offer their music for free. At the same time, the site advertised that “you can forget wading through page after page of unknown artists.”8 The company also sought out endorsements from established acts. Public Enemy rapper Chuck D defended Napster on television and in a New York Times opinion piece—and received a payment from the company for another project, according to All the Rave—while Limp Bizkit got the company to sponsor a tour of free concerts.9
Record companies have a well-earned reputation for not dealing fairly with performers, of course. But by the 1990s, the most exploitative contracts were behind them, thanks to corporate image concerns, savvier artist managers, and lawsuits over past behavior. But standard contracts still worked in favor of labels, which used byzantine accounting to pay out as little in royalties as possible. For years, to cite one example, companies deducted money from artist royalties for “breakage”—a holdover from the vinyl business that wasn’t much of an issue for more durable CDs.
When the RIAA sued Napster, tensions between artists and labels had just been exacerbated by the passage of a 1999 bill with an obscure amendment that classified music recordings as “works made for hire.”10 That wording change—which Billboard reported was quietly made by Mitch Glazier, majority chief counsel to the House Judiciary Subcommittee on Intellectual Property11—would have prevented artists from ever filing to regain their copyrights after a certain number of years.12 (“Nothing can actually be ‘snuck’ into an IP bill,” says Glazier, who points out that another office held the final copy of the bill.) Although that amendment was repealed, musicians and managers were outraged when the RIAA then hired Glazier as a lobbyist.13 “On the one hand, RIAA creates all this flap about Napster and copyright infringement,” Eagles frontman Don Henley said at the time, “while with the other hand, they’ve taken away artists’ copyrights.”14 This didn’t exactly inspire artists to speak out on behalf of labels.
At the end of May 2000, about a month after Metallica delivered its boxes, Hummer Winblad Venture Partners invested $13 million in return for 20 percent of Napster and installed one of its partners, Hank Barry, as interim chief executive, a position he ended up keeping for a year and a half. An intellectual property lawyer who had done work for A&M Records as well as numerous Silicon Valley start-ups, Barry understood both businesses and had an easy time talking with label executives, many of whom shared his professional background. Within a week of starting at Napster, he met with the Universal Music digital executives Albhy Galuten and Lawrence Kenswil at the Palo Alto office of Wilson Sonsini Goodrich & Rosati, the law firm where he worked before going to Hummer. “The thing I was really trying to communicate to them was, I want to try to have an industry-supported model,” Barry remembers. He believed Napster was legal, but it was far from certain that courts would agree; he knew a case like this had never been decided before. And since a hearing that could shut down the site until a trial started was scheduled for July 26, he knew he might have less than two months to strike a deal.
Napster and the labels put more effort into negotiating a deal to make the service legal than most people realized at the time. The week after Barry spoke with Galuten and Kenswil, he flew out to New York to meet Edgar Bronfman Jr., then the chief executive of the Seagram Company, which owned Universal Studios and its music operation, the industry’s largest. In public Bronfman took an aggressive stance toward online piracy, saying he would defend intellectual property with “a Roman legion or two of Wall Street lawyers.”15 In private he was willing to make a deal, though, and both he and Barry remember their meetings as cordial. During one meeting, Kenswil says, Barry showed him a draft Napster press release attacking the labels for suing the company, and Kenswil jokingly corrected the grammar. “There was no ideology,” Kenswil recalls. Like almost everyone on both sides of the debate, Barry mostly just wanted to make money for his company.
Bronfman and Barry met twice more, and their discussions went well enough that Bronfman arranged for a larger meeting the following week—at Herb Allen’s annual Sun Valley, Idaho, media business summit—with chief executives from the other media companies that owned labels. Ironically, just when the labels and Napster were closest to making a deal, they squared off in public at a July 11 Senate Judiciary Committee hearing about online music.16 Label executives pointed out that Napster was building a business based on their music, while Barry said studies showed consumers used the service to sample music they then bought on CD.
Both sides were under pressure at the Sun Valley summit—but they also saw opportunity. The labels had a fighting chance to get file sharing under control, and Barry was waiting for a court decision to establish whether he was running an illegal operation or the fastest-growing technology business in the country. “In order to assume they had a lot of value, they had to wait for the judge’s decision,” Bronfman says. “But I said, ‘I don’t think you want to risk that, because the odds are overwhelming that you’ll lose.’ ”
Two days after the Senate hearing, in a conference room at the Knob Hill Inn in Sun Valley, Barry and his boss, John Hummer, met with Bronfman; Bertelsmann’s chief executive, Thomas Middelhoff; Sony’s chief executive, Nobuyuki Idei; and Sony’s U.S. chief, Howard Stringer. They discussed a deal that would give the companies that owned labels a two-thirds stake in Napster, as well as greater control of the company, in exchange for settling their lawsuit. Idei and Middelhoff gave Bronfman permission to negotiate for them, with the idea of organizing a potential deal to present to Time Warner and EMI, then in merger talks. “We would have ended up somewhere where we probably could have gotten a deal done,” Bronfman says. Middelhoff remembers the meeting differently: he says he left the room thinking the two sides were too far apart to make a deal.
After this, accounts of the negotiations diverge. Bronfman says Hummer “went radio silent for a week,” then called six days before the court hearing to tell him that another company wanted to buy Napster. Hummer gave Bronfman the opportunity to match an offer of $2 billion—a startling sum for a company that had no revenue, let alone profit. “I said I don’t need to think about this, I’m not going to buy Napster for $2 billion,” Bronfman remembers. “If you can get this deal, you should lock the door and take hostage whomever is going to write you this check.” An executive who worked with Bronfman at the time confirmed that Hummer tried to sell Napster for a very high price.
A few hours later Bronfman says he got a call from Yahoo!’s co-founder Jerry Yang, who told him his company might buy Napster to forestall a bid from America Online, which Hummer had told him was interested. Bronfman told Yang he thought Hummer was bluffing: America Online wouldn’t risk its pending merger with Time Warner with another complex transaction. Yahoo! passed, and an America Online bid never materialized. “I think John [Hummer] just miscalculated,” Bronfman says. “If you remember 2000, those were the headiest days of the technology boom, and I think people just got stars in their eyes.”
Barry and Hummer say this didn’t happen. But they did have a very different idea of what Napster was worth. “Napster was the fastest-growing application in the history of the world, and there was every reason to believe that its value would have been in the billions of dollars,” Hummer says. The record companies believed that the secret to this growth was the fact that it was giving away something people usually paid for—and that obviously played a major role in its success.
Less than two weeks after the Sun Valley meeting, on July 26, the U.S. district court judge Marilyn Patel handed down a preliminary injunction that ordered Napster to stop users from trading copyrighted works, even if that meant the service had to shut down entirely.17 As for negotiation, Bronfman remembers, “it was kind of done.”
Middelhoff, who was fired from Bertelsmann in 2002, thinks both sides overplayed their hands. As venture capitalists during the dot-com boom, Barry and Hummer thought they could turn Napster into an online empire. But Middelhoff thinks the labels fundamentally failed to understand that they were entering a new world where they could no longer control distribution. “In the past, they had the production of CDs; they had all the value chain,” he says. “My position was, when the Internet came up, this is the end of this behavior—you cannot control the Internet.”
It’s tempting to believe that a deal with Napster could have saved the major labels, and many technology executives see the labels’ inability to reach one as their undoing. In the thoroughly reported Appetite for Self-Destruction: The Spectacular Crash of the Record Industry in the Digital Age, author Steve Knopper suggests that a legitimate version of Napster could have convinced half its 26.4 million monthly users to spend $10 per month—to bring in $1.58 billion a year.18 But it’s hard to see how Napster would have been able to bring in that much revenue. Few consumers buy $10 worth of music per month at the iTunes Store, which has sold an average of fewer than a hundred songs over the life of each iTunes account.19 Napster also discussed selling subscription access, but services that do so now have struggled to find a paying audience. Most businesses that entice consumers with free products set a goal of converting 5 percent of them into paying customers. That’s a little less than the free-to-paid conversion rate of the online music service Spotify, which at the end of 2010 had about 10 million users and 750,000 subscribers.
“The only way [a Napster deal] would have worked is if you had free-to-paid conversion rates that were unheard of,” remembers Bertelsmann Music Group’s chief executive, Strauss Zelnick, who understood such business models from overseeing the company’s old five-albums-for-a-penny music club. And Napster had several other substantial disadvantages: Not all of its users had credit cards, and at least some of those who did would have been reluctant to use them for online purchases in 1999. More important, Napster would have had to change the expectations of an audience it had conditioned to expect free music, and it would have had to do so while competing with illegal services that were still free—and free to offer copyrighted movies and unreleased music when it couldn’t.
That hasn’t kept technology pundits from insisting the music business missed its big chance. In The Perfect Thing: How the iPod Shuffles Commerce, Culture, and Coolness, the former Newsweek technology correspondent Steven Levy blames the labels for not making their content free online the way newspapers did, although that didn’t work out very well for them (or for Newsweek, for that matter).20 Levy writes that when he interviewed Barry, he saw a look in his eyes that said, “Why didn’t they work with us?”21 But Barry is hardly as naive as Levy makes him sound, and even he doesn’t think the labels were as clueless as some people say.
“To this day it bugs me when people say, ‘They’re so backward,’ ” Barry says. “They’re not backward at all. They’ve got a physical model that’s a high-margin, high-revenue model. So it’s a perfectly rational business decision for them to stay in the physical world as long as they possibly can.” The labels wasted time they should have spent setting up legal online services, and they made plenty of other mistakes. But why would any company rush to turn $15 transactions into ninety-nine-cent sales, let alone ones worth nothing at all? Hummer says a music business lawyer told him that he worried the record industry could shrink to $1 billion a year and that his job was to delay that for as long as possible, which sounds depressing but might well make sense from a financial perspective. If you really believed music would inevitably be free, the logical thing to do would be to sell what you could for as long as you could, and use the time and cash flow to sign new contracts with artists that let you make money on live performances or merchandise sales.
Labels that wanted to cooperate with Napster also faced a potentially serious legal problem: they didn’t have all the rights they needed to operate a file-sharing service. Older artist contracts didn’t include permission to sell music online, and labels worried that working with Napster while it still ran its original service would make them vulnerable to lawsuits from performers or other companies. Technology executives tended to dismiss this. As it turned out, however, when one media company did make a deal with Napster, it found out the hard way that this was more than an idle concern.
As Bronfman and Barry were negotiating politely, almost everyone else in the music and technology businesses was issuing angry pronouncements about the future of artistic creativity, technical innovation, or civilization itself. Even as the RIAA’s chief executive, Hilary Rosen, privately urged label chiefs to prepare for a digital future, she played the music industry’s bad cop in television interviews. John Hummer told Fortune, “I am the record companies’ worst nightmare.”22
“Napster upped the PR ante, so we upped the PR ante, and there was a very hostile environment created,” Rosen says. “I had bodyguards at one point, from death threats.”23 That was years before the major labels sued individuals, so this intense ill will stemmed only from the RIAA’s lawsuit against Napster.
By that point it had become clear that, whatever the company’s intentions, Napster was essentially building a business on piracy. The labels presented a study that looked at more than a thousand users and found they were all trading copyrighted files, although not all of the files they shared were copyrighted.24 The legal discovery process also found that Napster’s co-founder Sean Parker—the Facebook executive portrayed by Justin Timberlake in The Social Network—had sent an e-mail to Shawn Fanning about convincing the RIAA that Napster was “not just making pirated music available.”25
Barry hired the superlawyer David Boies, who had just represented the government in its antitrust case against Microsoft and would soon work for the Democrats in Bush v. Gore. Boies argued forcefully that Napster should be legal as long as it was “capable of substantial non-infringing uses,” the standard the Supreme Court established in the Betamax case. He said the labels were holding back innovation by refusing to license their content for sale online. And in July 2001, he convinced the U.S. Court of Appeals for the Ninth Circuit to stay Patel’s injunction.26
As Boies shaped Napster’s defense, anticopyright activists weighed in separately to support the service. The free culture activist Lawrence Lessig, then teaching law at Stanford University, submitted an “expert report” to Judge Patel that argued Napster would have legitimate uses, even if they hadn’t emerged yet.27 John Perry Barlow, who had written “A Declaration of the Independence of Cyberspace,” contributed a manifesto to Wired claiming that Judge Patel’s injunction had turned “millions of politically apathetic youngsters into electronic Hezbollah.”28 Barlow said he thought Napster should have been “Napster.org”—a nonprofit—even though everyone at the company but Shawn Fanning was motivated by monetary gain.29 In Barlow’s view, the future would involve voluntary payments to artists “without the barbaric inconvenience”—italics his—“currently imposed by the entertainment industry.”30
Other pundits simply decided the law didn’t matter. In her book Digital Copyright, the law professor Jessica Litman asserts that “if forty million people refuse to obey a law, then what the law says doesn’t matter.”31 But most of those people didn’t have a philosophical objection to the law; they just violated it because they knew they wouldn’t get caught. About forty million speeding tickets are issued every year, according to some estimates, and few drivers would argue that traffic regulations are irrelevant.32 Like copyright laws, they are often inconvenient, occasionally annoying, and plainly necessary for the greater good.
On Halloween 2000, Napster finally got what it wanted: a lifeline from an old-media executive who believed that file sharing was the future. At a press conference in Manhattan’s Essex House hotel, Thomas Middelhoff announced that Bertelsmann had made its own deal to lend Napster $60 million, convertible into equity, in order to create a legitimate version of the service. Bertelsmann had a reputation as a risk-averse company with roots in the book-club business and a headquarters in sleepy Gütersloh, Germany, but Middelhoff had made his reputation with an extremely profitable investment in America Online, and he wanted to make another bold move. But he didn’t share his plan with Zelnick, who ran Bertelsmann’s music business, until just before he announced the deal.
“I was recovering from oral surgery, and I got a phone call” from Bertelsmann’s digital chief, Andreas Schmidt, Zelnick remembers. “I said, ‘No, listen, we can’t do it this way.’ ” Zelnick, who has a background in technology as well as a law degree, worried that Bertelsmann could share liability for Napster’s copyright infringement. “Thomas felt that there was only one path forward, which was digital, and he was right about that,” Zelnick remembers. “He felt that it was taking a long time to develop anything and that we were losing ground in the meantime, and that was true. In many ways his vision was apt. The problem was that his answer to that vision was not a good answer.” Zelnick wrote Middelhoff a memo with ten things that could go wrong—which, he says, more or less predicted the course of events—but his concerns were dismissed. “He said, ‘I invested in America Online because I had a certain feeling about where it would go and I have the same feeling now,’ ” Zelnick says. “I’m not a believer in prayer as a business plan.” Zelnick resigned within a week, although he says he didn’t make the move because of Napster.
Middelhoff says legal liability shouldn’t have been a problem. “What we did was give money to Napster to develop a legitimate service,” he says. That was the only thing the company was allowed to do with the money. “It was for nothing else: not to run their business, not to buy new servers, it was just about that.” As Middelhoff took meetings with the other labels, though, he found them less receptive to the idea of a legitimate Napster service than he thought—perhaps because the parent company of one label group would end up with power over the others. Middelhoff says the chief executives of the media conglomerates that owned labels understood that the world was changing, but the record executives beneath them wanted to hold on to what they had, and missed an opportunity as a result. “Napster could have been the early iTunes,” he says.
On February 12, 2001, the U.S. Court of Appeals for the Ninth Circuit mostly upheld Patel’s injunction, on the grounds that, unlike the makers of the VCR, Napster had an ongoing role in copyright infringement.33 A week later, Barry and Middelhoff called a press conference to offer the other labels $1 billion over five years from future Napster subscription revenue in exchange for settling the suit. It was a grand gesture but a paltry sum, especially considering how much the service would have cut into CD sales. On July 2, Napster went off-line, unable to filter out copyrighted content as well as Patel wanted.
As Napster ran out of legal maneuvering room and operating capital, it was every venture capitalist for himself. Negotiations for an outright Bertelsmann acquisition stalled as John Fanning argued with Hummer Winblad over who would get the proceeds or receive protection from liability.34 And when Bertelsmann tried to buy the company out of bankruptcy, a judge ruled against the sale on the grounds that Konrad Hilbers, a former Bertelsmann executive and friend of Middelhoff’s who had replaced Barry as Napster’s chief executive, had a conflict of interest in making the deal. Napster was eventually sold to the software company Roxio and then to Best Buy, both of which operated it as a legal music site without much success.
As for Bertelsmann, because of its role in helping Napster stay afloat, it was sued for copyright infringement in February 2003 by songwriters and the publishing companies that control their copyrights, much as Zelnick warned.35 During the course of the bankruptcy proceeding, evidence emerged that Bertelsmann exercised some degree of control over Napster and its loan had been used for everyday operating expenses. The other major labels sued too, and settling all the cases may have eventually cost Bertelsmann more than €390 million.36 Middelhoff, who was fired in 2002 amid disagreements over the direction of Bertelsmann, points out the company was never found liable for any wrongdoing and says it would have prevailed in court. “There is no way I would have settled this,” he says.
The only way the major labels could have invested in Napster without taking a legal risk would have been to shut down the company’s illegal service before launching a legitimate one. In the meantime, though, illegal competitors would have taken many of Napster’s users. But problems like this would soon come to seem quaint. The next generation of file-sharing services decided that free online content represented the future of the music business—whether the music business liked it or not.
Silicon Valley executives have always seen the music business as a mess of inefficiency. As music became cheaper to record and distribute, labels spent most of their money generating demand—marketing and promoting performers they wanted to make into stars. Aside from Apple, the technology world has never placed a high value on consumer marketing, and most entrepreneurs saw it as inefficient. Why couldn’t artists record on a computer, give away their music online, and then find something else to sell?
From the outside, this makes sense, since—as technology executives kept pointing out—most performers don’t make much money on album sales anyway. After a major-label artist receives an advance for signing a contract that assigns copyright of his recordings to the record company, he doesn’t receive any further royalties until he’s “recouped”—earned back—the money spent on recording, promotion, and other label expenses. Due to high spending and what might charitably be called imaginative accounting, most acts never see any additional money, while those that do make $1.25 to $1.50 an album in royalties—a pittance compared with the retail price. Major-label recording contracts have been compared to mortgages that leave the bank owning the house after it’s paid off.
While record-label contracts leave some artists feeling exploited, file-sharing services don’t replace record companies; they just replace the trucks that deliver CDs to stores. For all their faults, labels advance artists money to live on, pair them with songwriters and producers, and help find audiences for their albums. At their best, they also serve as creative partners, such as when the Atlantic Records executive Jerry Wexler brought a modestly successful gospel and rhythm-and-blues singer named Aretha Franklin to FAME Studios in Muscle Shoals, Alabama, and helped reinvent her as a soul powerhouse. At their worst, labels ignore projects that aren’t priorities and throw money around as recklessly as the hip-hop stars in their videos. But although some artists regret signing with labels, the choice is theirs, and they can sue if they feel they’ve been wronged. File-sharing services give them no such options.
To understand why the music industry has had so much trouble operating online, it helps to understand that “the music industry” is only a general term for several interrelated businesses that coexist in the same world but run very differently. When a label releases an album, it controls only the right to the recording. Separate copyrights cover the songs as compositions, which are usually owned or administered by publishers; they collect money when songs are played on the radio, recorded on albums, or licensed to movies or television. And the concert and merchandise businesses have their own approaches to making money, and hence their own priorities.
The modern music business evolved during the 1960s, when more artists began to write their own songs and albums started to replace singles as the dominant format. Both shifts gave labels the money and power to organize every aspect of an artist’s career around new releases—concerts would promote albums, instead of the other way around. “Historically, record companies were the marketing engine for all of this—your live career, your songwriting career, and your merchandise career,” says Danny Goldberg, president of the talent agency Gold Village Entertainment and a former chief executive of Warner Records. “They had the publicity and the promotion and the people who would get you onto the soundtrack.”
So far, the acts that have most successfully used free music to promote major tours—Radiohead and Nine Inch Nails—have benefited from millions of dollars’ worth of marketing from their respective major labels. “Almost never does an artist establish a national or global career without a lot of money being spent—six or seven figures—with a team of experienced professionals,” says Ron Shapiro, a former president of Atlantic Records, who now manages Regina Spektor and other acts. “And I do not see that changing.”
Although young artists like Lily Allen and the Arctic Monkeys found fans on Myspace, they quickly signed with labels to advance their careers. So did Hollywood Undead, which had 400,000 Myspace friends but chose to release their album on A&M Octone. “It’s great to have Web visibility, but the things that have helped to break this band are the old-school things—radio, retail, and that stuff,” says Mike Renault, their manager. Like most bands, Hollywood Undead also needed “tour support,” money provided by labels to help with touring expenses. Before performing live becomes a source of revenue, it can be a considerable expense.
Consider an actual budget provided by the manager of a developing rock act that released its first album on a major label in the fall of 2009.37 For its first proper tour, the band played the East Coast and the Midwest—sixteen shows in nineteen days at small venues with a capacity of two hundred to five hundred. It would take in $1,000 to $2,000 a night, or around $24,000 for the whole tour—not bad for less than a month’s work.
That’s before costs, though. This band has six members; it’s a big group, but that means they don’t have to pay a roadie or a driver. The budget calls for them to share rooms with an average cost of $100 per night and spend $25 a day each on food; $3,000 is allocated for insurance and replacement equipment, while transportation should cost only $1,500, since the group has its own van. That’s more than $12,000 worth of expenses.
That leaves the band with $12,000 for three weeks of work. Assuming the manager makes a 15 percent commission, each member will take home about $1,700 before taxes. That’s not a great living—especially because it’s hard to play the same cities more than a few times a year. But it could be a wise investment. If the band uses these shows to build a following and gets popular enough to play venues that hold a thousand people, it could make a nice living.
But that could take some time; it’s not easy to build a following. And while the Internet creates opportunities for bands to get their music to a mass audience, that doesn’t mean people will listen. Even in online music stores, with their impressive selection, 5 percent of songs account for 80 percent of downloads.38 This may have less to do with major-label marketing than with the fact that people gravitate toward experiences they can share. This may not be a bad thing. The Beatles sold hundreds of times as many albums as Gerry and the Pacemakers—contemporaries who also worked with the manager Brian Epstein and the producer George Martin—but few music fans would call that an injustice.
Can musicians make money without reaching a big audience? Some technology pundits predict a world of niche culture, but understand how tough that could be for artists. In 2008, the former Wired editor Kevin Kelly wrote a blog post titled “1,000 True Fans.”39 Assuming each dedicated fan spent $100 per year on an artist’s work, he could bring in $100,000 a year, Kelly wrote, “which minus some modest expenses, is a living for most folks.” But the products Kelly believes acts should sell, such as vinyl records and T-shirts, cost money to produce, so an artist might keep only half of that $100 after expenses. After setting aside $10,000 for instruments, recording equipment, and a van in which to tour, the artist might end up with $40,000—less if he has a manager, lawyer, or booking agent.
Then there’s the issue of how to find those fans in the first place. Some bands get inexpensive exposure by encouraging fans to freely share their music or live performances, like Phish and the Grateful Dead did. But that takes a while, and most acts need income in the meantime. (Early in their career, before they signed to Warner Records, the Grateful Dead were funded by the LSD chemist Owsley Stanley.40) And artists who offer their music for free online still need to control their intellectual property rights—much as the Dead did—in order to sell vinyl records, T-shirts, and other merchandise.41
Many of the acts using innovative strategies to make money without selling recordings also rely on the old music business in one way or another. In his blog, Kelly cites the example of the singer-songwriter Jill Sobule, who raised $75,000 to make an album by selling various items, from $10 downloads to a $10,000 guest vocalist spot. But many of Sobule’s fans first heard her music over a decade ago, when Atlantic Records gave her money to make a funny video with Fabio and helped get one of her songs on the soundtrack to the movie Clueless. The most expensive work—getting an artist enough exposure so some listeners become fans—had already been done.
The Internet offers plenty of inexpensive ways to promote music, from Myspace to Twitter. The problem is that those platforms have become so full of musicians that it’s hard to stand out from the crowd—and harder to convert an online following into a career. The rapper Soulja Boy started as a YouTube phenomenon, then had a radio hit with “Crank That (Soulja Boy),” which sold 4.6 million copies to become the fourteenth most popular digital download ever.42 He gradually amassed 2.5 million Twitter followers. But when he released his third album in November 2010, it sold a disappointing 13,000 copies its first week in stores.43 All that online activity didn’t make up for the lack of a hit single.44
Even by the end of 2001, most record label executives didn’t realize how profoundly file sharing would affect their business. They had vanquished Napster, and CD sales had dropped less than 4 percent compared with 2000—and in a weakening economy at that. Perhaps consumers really were downloading music to decide what CDs to buy.
If that was the case, it was the second generation of file-sharing services that decisively shifted the listening habits of teenagers and twentysomethings. Gnutella launched in 2000, with a design that let users download files from one another’s computers without relying on a central directory, as Napster did. Other services further refined that model.45 That meant the precedent set in the Napster case might not apply, and the fact that some new services were based outside the United States further complicated matters. In October 2001, twenty-eight major labels and movie studios sued the companies that owned Grokster, Kazaa, and Morpheus.46 In the technology world, the move was viewed as a direct attack on the precedent set by the Supreme Court in the Betamax case.
As the case proceeded, the conflict moved to Washington, where film companies started to flex their muscle. Since DVD players were built with copy-protection technology, studios started to ask why computers couldn’t include it as well. In February 2002, Michael Eisner, the chief executive of Disney, testified in front of the Senate Commerce Committee that “there are people in the tech industry who believe that piracy is the killer app for their business.”47 Eisner was mocked for this comment, with some justification. But Napster certainly believed that piracy fueled its growth; it said so in its own strategy document.
The movie business is bigger than the music industry, and it received more sympathy in Washington, where the then-recent controversy over gangsta rap hadn’t won labels any fans. But the technology business commanded respect as well, and companies like Intel argued that mandating content protection would interfere with innovation.48 (Intel apparently changed its philosophy, since it later developed copy-protection systems of its own.) And Silicon Valley had an easier time appealing to consumers, who liked getting content for free online. “The entire theme of the copyright community is that downloading off the Web is both illegal and immoral,” said the Consumer Electronics Association’s chief executive, Gary Shapiro, at a technology conference in September. “It is neither.”49 With a nod toward the Home Recording Rights Coalition, the venture capitalists Joe Kraus and Graham Spencer started the organization DigitalConsumer.org to promote a “Consumer Technology Bill of Rights”—which would also help protect their investments. Public Knowledge, an advocacy group that receives funding from technology and electronics companies, started to criticize lawsuits against file sharing as well. “It became a consumer movement,” says the RIAA’s Hilary Rosen.
Both sides dug in for the Grokster lawsuit. This time, though, there was no sympathetic young inventor in the mold of Shawn Fanning. Kazaa installed “malware” that served up advertising on users’ computers.50 Kazaa also complained that Kazaa Lite, a modification of its program that removed the malware, violated Kazaa’s copyright.51 Wired said the Vanuatu-based company, which was eventually dropped from the case, had such a byzantine financial structure that it resembled a “corporate nesting doll.”52
Like Napster, the file-sharing companies argued that their services had legitimate uses, such as distributing free software or music that was no longer covered by copyright. If so, those uses never caught on. The entertainment companies presented evidence that 97 percent of all files downloaded from Grokster infringed copyright.53 So far, no reputable study has found a popular file-sharing network where more than 10 percent of downloads are legitimate, and most show that fewer than 5 percent are. (The percentage of available files that infringe copyright can be very different; it was 90 percent in a study presented in the Grokster case.) Many of these studies have been criticized for bias. But a 2010 study by a Princeton student also found that 85 to 99 percent of files distributed on the BitTorrent file-sharing network infringed copyright, and his work was supervised by Ed Felten, a computer science professor, who has criticized copyright laws.54
In April 2003, a U.S. district court ruled in favor of the file-sharing services, in part because they were set up differently from Napster. In a summary judgment, the U.S. district court judge Stephen Wilson held that “Grokster and StreamCast are not significantly different from companies that sell home video recorders or copy machines, both of which can be and are used to infringe copyrights.”55 Although the services were covered by the Betamax precedent, Wilson ruled, their users were still liable for copyright infringement. In August 2004, after the U.S. Court of Appeals for the Ninth Circuit mostly upheld the decision, the media companies appealed to the Supreme Court.56
By then, the RIAA had started suing individuals. In September 2003, a few months after it lost the Grokster case in district court, the RIAA filed 261 lawsuits against users who distributed copyrighted music on file-sharing services. (The RIAA’s lawsuits against individuals depended mostly on illegal sharing, not downloading.57) The organization had considered this for some time, but Rosen and Warner Music Group’s chief executive, Roger Ames, stopped the suits from moving ahead until Apple’s iTunes Store provided a convenient way to buy music online. “I think we had sued almost every service around that was a real threat,” says Rosen, who had left the RIAA by the time the suits against individuals were filed. But when the courts ruled in favor of the services, the labels felt they had no choice but to sue individuals.
From a public relations standpoint, the lawsuits were a disaster. U.S. copyright law sets statutory damages for infringement at $750 to $30,000 per work—or up to $150,000 if “infringement was committed willfully”58—but that penalty was devised with moneymaking pirate operations in mind. Although the RIAA offered to settle for pennies on the dollar—the lawsuits, however ill-advised they may have been, were never intended to generate profit—the potential damages still shocked most observers. And as the RIAA continued to file lawsuits based on Internet protocol addresses—a standard but not reliably accurate way of identifying Internet users—the public relations hits just kept on coming. Among others, it sued a grandmother who didn’t know how to download and someone who had died.59
Almost every individual accused of illegally sharing copyrighted songs settled with the RIAA, although some challenged the lawsuits on various grounds. The first case to reach a jury involved Jammie Thomas, a single mother from Brainerd, Minnesota, who said she hadn’t downloaded the songs in question.60 Journalists portrayed her as a victim of media conglomerates, but the jury didn’t take long to find her liable for willful copyright infringement and awarded statutory damages of $222,000 for sharing twenty-four songs—more than $9,000 a track.61 In a retrial, ordered because of an error in jury instruction, Thomas was found liable for damages of $1.92 million, presumably because the jury didn’t believe her account of events. (The law sets a range of damages that’s absurdly high, but both juries awarded well over the minimum; this indicates that either Thomas didn’t come across well in court or ordinary Americans don’t see file sharing as a victimless crime—or, probably, both.) The judge reduced this award to $54,000, and the RIAA offered to settle for $25,000 if Thomas would ask the judge to vacate that decision, so it wouldn’t set a precedent.62 Thomas declined this offer, and the labels rejected the judge’s remittitur. In November 2010, a third trial solely to determine damages found Thomas liable for $1.5 million—a verdict she has said she will appeal. In December 2010, Thomas made a motion to set aside that verdict as unconstitutionally high, and the RIAA plans to respond. The labels don’t want to jeopardize their ability to ask for high damages, and Thomas seems to be fighting on principle.
Did all of these legal fees and publicity missteps actually accomplish anything? The lawsuits alienated fans, tied up the legal system, and arguably made it harder for the RIAA to push antipiracy legislation that might have been more effective. Meanwhile, the number of tracks downloaded illegally continued to rise. But at least some of that increase is the result of faster connections and larger hard drives. Between 2006 and 2009, the percent of Internet users who downloaded music illegally declined slightly, according to the research company NPD Group, no small thing given how fast it had been rising.63 It’s almost impossible to stop committed pirates, but the RIAA’s lawsuits probably kept their numbers from growing—although at too high a price.
In a few years, the major labels managed to destroy the cultural cachet they had spent decades building. In the late 1960s, CBS Records used to advertise its counterculture acts with the tagline “The man can’t bust our music.” Now most teenagers believe the major labels were trying to do just that.
Some of the negative publicity came down to poor planning and bad luck. “By putting Metallica out there, it became a farce,” says Bob Mould, the former Hüsker Dü frontman and solo artist, who thinks the band members’ hearts were in the right place. It might have been different “had it been a newer artist who said, ‘I think it’s great that you’re sharing my music, but I have to get paid or I can’t do this full-time like you want me to.’ ”
Some younger fans tend to sneer at musicians who want to get paid, partly because one of rock’s defining myths is that no one is in it for the money. This is unique to rock: country performers value material success, and rappers go to absurd lengths to show it off. But it was never in anyone’s interest to discuss the details of the music business, which weren’t very glamorous anyway. “Artists don’t like to talk publicly about getting paid,” says Danny Goldberg of Gold Village Entertainment. “Rock and roll is a rebellious art form, dealing with a teenage sensibility, and through the lens of that attitude it’s easy to look at record company executives as uncool and shallow.”
The labels tried to project a simple message: file sharing copyrighted music was theft. But as technology business lawyers rushed to point out, they’re not the same thing at all: different laws cover copyright infringement and stealing, and copying a file doesn’t deprive its original owner of anything. The implication was that theft was wrong, copyright infringement less so.
It makes more sense to compare copyright infringement to a theft of service—sneaking into a movie or jumping a subway turnstile. (In the case of uploading files, it’s more like breaking a door or turnstile so others can do so.) In all of these cases, nothing is literally taken; as is the case with music, enough other people pay that it seems like a minor infraction. But both are still illegal, because if everyone stopped paying, we would no longer have movies or public transportation of the same quality. And while file sharing isn’t theft, it’s not exactly “sharing” either.
After CD sales fell 10.7 percent in 2002, labels heard a “Geek Chorus” of suggestions on how to reinvent their business. While it’s hard to know for certain if they could have prevented their decline, the labels’ biggest failing was not organizing a legitimate, convenient online retailer before Apple launched its iTunes Store in April 2003. Such a store would have struggled to compete with Napster, but it would have appealed to consumers troubled about taking songs without paying for them and put the labels in a better position to lobby for laws that encouraged legal commerce online. When the five major labels finally introduced two online stores among them—Universal and Sony launched Pressplay, while EMI, Warner Music Group, and Bertelsmann Music Group started MusicNet—they were plagued by confusing prices, confusing copy protection, and compatibility problems.
The music industry’s own online stores were such disasters that Napster postponed its demise by getting Judge Patel’s permission to gather evidence of whether the labels had colluded to suppress competition. The record companies certainly wanted to gain more control over how their product was sold, but they may have quarreled more than they cooperated. “They mistakenly thought this was an opportunity to change the equation,” Rosen says of the labels. “Maybe they wouldn’t just invest, record, promote, and market—maybe they could even sell.” But disagreements between labels, publishers, and artists made it hard to put together offers that interested consumers, Rosen says. “Everybody was fighting among themselves for a couple of years rather than coming together.”
The labels also missed an opportunity in the late 1990s to replace the CD with a secure music format they could control. Any copy-protection system will eventually be cracked, but they didn’t need to make piracy impossible—just inconvenient. The labels tried to do this in 1998, when they gathered scores of electronics, software, and technology companies to form the Secure Digital Music Initiative. The idea was to give music files protection like that on DVDs. But the organization’s work stalled as labels argued for more security and electronics companies pushed for convenience—rightly so, since a legal format has to be as convenient as piracy. When unprotected MP3s became the de facto standard, though, negotiations broke down as electronics companies realized they could make more money selling gadgets that would play pirated music; they had nothing to gain by cooperating. Technology and consumer electronics companies maintain that consumers hate copy protection, but that may depend on whether it gets in their way; it doesn’t seem to have held back Apple’s iTunes Store.
Once the labels realized they couldn’t establish a secure music format, they should have eliminated copy protection entirely. This isn’t a contradiction—just an acknowledgment of reality. As long as labels released CDs with unprotected music, it didn’t make sense to protect the same songs just because they happened to be sold online. In 2005, Sony BMG Music Entertainment released CDs that, unbeknownst to users, installed rootkit software on Windows PCs that limited copying but also made the computers vulnerable to viruses. The public outcry was immediate: copy protection had gone too far. Sony BMG was forced to recall the affected CDs and settle several class-action lawsuits. More important, bad publicity over the incident angered consumers who hadn’t known much about copy protection until then. The labels never made another serious attempt to protect CDs.
The labels could have also tried to postpone their decline in other ways, including cutting wholesale prices, although Universal Music tried this in 2003 without much success. In retrospect, though, it’s hard to know if any of these ideas would have worked. As other media businesses began to face the same problems, executives talked confidently about their plans to avoid making the same mistakes as the major labels. But they didn’t even agree on what those mistakes were. It’s not as easy as it looks to run a business based on giving away what you once sold, especially in a medium where ads can be stripped out.
“There were plenty of stupid record executives, there was plenty of short-term thinking, and there were some incredibly awkward PR moments,” Goldberg says. “But even if they had done everything right, the end result would have been basically the same. Because how about the newspaper business, how about all these other content businesses that didn’t sue their customers? They’re just as screwed up.”
For about two months in the summer of 2002, the University of Texas economics professor Stan Liebowitz became a hero to everyone who hated record labels—by then a large enough group to win him some media attention. That spring, he published a paper with the libertarian Cato Institute suggesting file sharing would hurt the recording industry.64 But he started questioning his conclusion once he saw sales weren’t actually falling much.
Liebowitz stepped back from his earlier beliefs and suggested the recession could have caused the slight decline and that music fans might be sampling songs online and then buying the ones they liked—just as Napster had argued in court. He pointed out that labels had also worried about piracy in the early days of the audiocassette, although the format eventually expanded the market: some consumers taped albums rather than buying them, but others bought new cassettes for their cars. “I try to let data tell me what’s actually happening in the world,” Liebowitz told Salon.65 “And when the theory says one thing and things don’t work that way, then I say something’s missing in the theory.”
Liebowitz, already respected among academics for his work on the economics of copyright and software, suddenly had a wider audience. Music Web sites linked to his paper. The Dallas Observer called him “the journalist’s go-to guy on the issue of online music distribution.”66 And Lawrence Lessig praised his work in an opinion piece for the Financial Times. “When the data say something different from the party line,” Lessig wrote, “the academic should do as Prof Liebowitz did: speak the truth even if that means changing your mind.”67
And then Liebowitz changed his mind again, because the data started saying something else entirely.
In July he saw that the midyear numbers for music sales had dropped 9.7 percent from the previous year. Examining three decades of music sales, Liebowitz found that previous recessions hadn’t caused such significant declines and concluded the falloff probably did come from illegal downloading, even though he didn’t think every pirated copy of a song represented a lost sale. In a second Salon interview, a little more than two months after the first, Liebowitz predicted that file sharing would lead to a 20 percent drop in music sales.68
At that point, “I stopped getting calls from reporters,” Liebowitz remembers. “People liked the story that [file sharing] wasn’t causing harm. People don’t like big companies, they don’t like the recording industry, and they liked the idea that they could get something for free.”
This wasn’t the first time Liebowitz had researched the economic effects of copyright—or caused controversy with his findings. Before 2002, most of his studies had concluded that unauthorized duplication actually helped copyright holders, and he believes he was among the first economists to suggest this.69 He came up with a theory called “indirect appropriability,” which said the ability to make copies of a work could raise its value to a consumer. The example he used was the way photocopy machines allowed libraries to get more value out of scholarly journals. “Needless to say, I was not very popular with organizations representing copyright holders,” Liebowitz wrote on his Web site.70 (He has since done some research for the International Federation of the Phonographic Industry, the global recording business trade organization.) “In those days there was not an army of copyright critics to embrace my work and make me a hero.”
Since the Napster lawsuit, any evidence that file sharing doesn’t hurt the music business has found an eager audience. In March 2004, when a paper by the economists Felix Oberholzer-Gee and Koleman Strumpf concluded that file sharing had little or no effect on music sales, journalists found the counterintuitive take as irresistible as they had found Liebowitz’s.71 The economists analyzed the relationship between album sales and file sharing and found that uploading increased during school vacations in Germany, which at the time supplied a substantial amount of the music downloaded by file-sharing service users worldwide. This surge prompted U.S. Internet users to download more songs, but album sales didn’t decline accordingly when they did.
Liebowitz questioned Oberholzer-Gee and Strumpf’s research, pointing out that German students would upload a lot of music that wouldn’t interest Americans and that, contrary to what Oberholzer-Gee and Strumpf said, genres that were more popular on file-sharing services saw larger sales declines and sales were falling around the world.72 (“We were studying file sharing in 2002,” Strumpf says. “We’re now in 2011, and it’s a very different world.”) In May 2007, Industry Canada released a study by the economists Birgitte Andersen and Marion Frenz that found that file sharing actually helps sales.73 Liebowitz challenged that paper as well, and Andersen and Frenz later concluded there wasn’t a strong relationship between piracy and CD sales after all. Asked about the study, Andersen says the Pirate Bay file-sharing service doesn’t offer much copyrighted music from major-label acts—although even a cursory glance at its Web site shows that isn’t true.
Many factors have hurt music sales, including the closing of so many record stores. But almost every other study has concluded that file sharing played a role,74 and anyone who believes otherwise is running out of alternate explanations. Several studies have shown that individuals who download music illegally also buy it, but that proves only correlation, not causation. Some suggested CD sales fell because music fans are no longer replacing their old records, but “catalog” sales of older releases declined less than overall sales from 2004 to 2009.75 Others speculated that DVD sales cut into the CD market, but now they’re declining as well.
Music sales have also declined disproportionately in countries where file sharing is more common. In Spain, where fully 45 percent of Internet users get media from pirate services—about twice the average rate in Europe76—CD sales declined 77 percent since 2001, compared with a continent-wide average decline of 54 percent.77 And Japan is now the No. 1 market for CDs, despite having one-third the population of the United States, partly because many people there access the Internet with mobile phones that don’t run file-sharing programs.78
In June 2010, with U.S. music sales less than half what they were a decade earlier, Oberholzer-Gee and Strumpf published another paper which conceded that file sharing had affected sales, but not that much and in a way that helped society.79 They said that 60 percent of online traffic consisted of file trading—an astonishing statistic that implies illegal downloading accounts for more than half of all Internet use—but that all this piracy accounted for 20 percent or less of the decline in music sales. And they pointed out that despite weaker copyright protection, artists were creating more work than ever: the number of new albums issued each year had doubled since 2000, and book and movie releases rose substantially as well. “Artists often enjoy what they do, suggesting they might continue being creative even when the monetary incentives to do so become weaker,” they wrote. “In addition, artists receive a significant portion of their remuneration not in monetary form—many of them enjoy fame, admiration, social status, and free beer in bars.”
Essentially, two respected economists argued that people will work for beer. (This may be literally true, of course, but that doesn’t mean they’ll work as hard as they otherwise might have.) Oberholzer-Gee and Strumpf also pointed out that the music business has grown over the last decade if you count iPod sales. Since in the United States none of this revenue goes to anyone who makes music, this is like saying the clothing business is booming if you count the installation of custom closets—both technically accurate and entirely beside the point.
What Oberholzer-Gee and Strumpf really proved was that more music than ever is available for sale in a way that can be measured. They got their numbers on new album releases from Nielsen SoundScan, which tracks new titles, whether or not they contain new music, which means they counted reissues, new compilations of old material, and foreign releases available only online. (Probably because there’s only so much more old music to sell online, 2010 saw a 22 percent decline in album releases.) They also counted thousands of digital-only releases on iTunes from bar bands, who a decade ago might have sold their albums only at concerts, where SoundScan couldn’t count them. And much of the increase over the past decade seems to have come from hobbyists: the number of U.S. residents employed as musicians has declined 19 percent since 2001.80
That’s important because the purpose of copyright is to “promote the Progress” of the arts, not simply encourage participation. It’s hard to see any relationship between the quantity and the quality of work produced. It’s even harder to argue, as Oberholzer-Gee and Strumpf do, that society benefits much from all this new music, given that so few people hear it. Of the seventy-five thousand albums released in the United States in 2010, sixty thousand sold fewer than a hundred copies, and many of those sold fewer than ten.81 If a band makes an album and no one is there to hear it, does it still make a sound?
In the grand narrative of the music business collapse, the labels were rescued from their own incompetence by Steve Jobs’s iTunes music store, which made piracy a marginal problem. Apple certainly pushed the labels into doing something they were unable to do themselves, and its iTunes Store has become the biggest music retailer in the United States. But legitimate online music stores like Apple’s have hardly stopped piracy: more music is downloaded illegally than legally, according to the NPD Group.82 (Not all of those songs represent lost sales, of course, but surely some must.) And, like the industry’s attempts to turn file sharing into a legitimate business, the real story of Apple’s effect on the music business is more complicated than most people realize.
In October 2001, Apple was a second-tier technology company that controlled about 3 percent of the personal computer market and was slowly regaining its relevance with high-design candy-colored Macs.83 When it introduced the $400 iPod just as the economy slowed, consumers didn’t exactly line up in front of Apple stores (of which there were only a few at the time anyway). Record labels were even less enthusiastic. In October 1998 the RIAA had sued Diamond Multimedia to stop the sale of the Rio, one of the first MP3 players. The organization argued that the device didn’t comply with the 1992 Audio Home Recording Act, which mandated that digital recording devices incorporate copy-protection technology and pay a royalty that would go to labels, performers, publishers, and songwriters. After a court ruled the law didn’t apply since the Rio didn’t record songs, other companies followed Diamond into the MP3 player market, including Apple.
Later that year, following the introduction of the iPod, Jobs called America Online’s chief executive, Barry Schuler, to ask about selling music from the company’s Warner Music Group. “He had the device, but devices without content don’t sell well,” remembers Paul Vidich, a Warner Music executive vice president at the time. So Vidich went to Cupertino, California, where he initially spoke with Jobs about DMX, a proposed secure music format. “I had someone giving a PowerPoint presentation and Steve was rocking back and forth in his chair, obviously agitated, and he basically said, ‘I don’t want to talk about this—I want to talk about this other thing,’ ” Vidich remembers. “And he launched into a tirade against the music industry. The room went dead silent. And I looked at him and said, ‘Steve, you’re absolutely right, but that’s why we’re here. We’re here because we think you—Apple—can help us solve this problem.’ And that sort of broke the ice.”
Over the next six months, Vidich met with Apple executives as they designed the iTunes Store. By then he knew consumers wanted an easy experience, and Warner agreed to a copy-protection system that allowed buyers to burn songs on CD. Jobs insisted on selling individual songs for one set price, and Vidich suggested ninety-nine cents. Jobs next approached the Universal Music Group, where the chief executive, Doug Morris, and Interscope Records’ chairman, Jimmy Iovine, loved the store’s elegant simplicity. Approval from the other majors and most big indies soon followed. Since the deals to sell music ran for a year and applied only to the Macintosh, executives figured they could always renegotiate later. “If this was really successful,” Vidich remembers thinking, “we’d find a way to differentiate the prices and charge more.”
Ironically, the one decision the labels usually get credit for making correctly may have been among their worst. Their eagerness to find a decent digital rights management system led them to accept one they didn’t control, which allowed Apple to dominate the market. They successfully pressed Apple to get seventy cents out of every ninety-nine-cent sale, which doesn’t leave Apple with much of a profit after the costs of processing credit card transactions and running the iTunes Store. The company makes so much on the iPod and related products that it essentially runs its online retailer to market them. Like Walmart, it sells music to promote more profitable goods, which means it has the leverage to sell music however it wants.
“We pushed hard to make sure all the constituencies around the table grabbed a lot of margin, which meant there wasn’t enough left for the retailer,” says Strauss Zelnick, who believes this was one of the industry’s biggest mistakes. (By then, he had left Bertelsmann.) “We would have been vastly better served to allow someone to grab ten points of net margin to have a bigger business than what we have today.” Allowing retailers to make more money could have created more competition.
In September 2005, Bronfman, by then chief executive of Warner Music Group, gave a speech saying that the price of a song should vary according to its newness and popularity, much like that of a DVD. “We are the arms supplier in the device wars,” he said.84 To some extent, this was true. The labels had given Apple a powerful weapon—its digital rights management system, FairPlay, which it refused to license to other companies. The iPod would play only MP3s and songs purchased from the iTunes Store, and songs from the iTunes Store would not play on other digital music players. That discouraged iPod owners from using other online music stores, and they couldn’t switch to another kind of digital music player without losing the songs they bought from iTunes. It also let Apple sell the iPod for a premium. Eventually, it helped Apple use the iTunes Store to develop the App Store that helped it lead the smartphone market, introduce the first successful tablet computer, and ultimately become one of the most valuable companies in the world.
By the end of 2004, the iTunes Store had sold 200 million songs. Until 2009, sales of digital downloads grew dramatically every year. But iTunes also encouraged fans to buy ninety-nine-cent songs instead of $10 or $15 albums.
Even if they continue to grow, those ninety-nine-cent-song sales won’t come close to making up for the corresponding decline in CD sales. “When they let Steve Jobs roll over us, that was the end,” says Peter Mensch, cofounder of Q Prime Management, who works with Metallica, the Red Hot Chili Peppers, and other acts. “They thought, ‘It’s another way to sell music.’ But now I’m selling singles when I should be selling albums.” Mensch dismisses the idea that labels had to make the deal to compete with file-sharing services. “The 40 or whatever percent of people who steal music—they’re gone and they’re never coming back.”
This goes against the conventional wisdom of the technology world, which has pushed labels to cut prices in order to turn pirates into consumers (and, in the process, help their own businesses). But trying to compete with free in this particular way might not have been a smart move. In January 2009, Apple announced that labels had agreed to eliminate copy protection from iTunes songs in exchange for more power to set song prices, which would range from sixty-nine cents to $1.29 in the United States. Technology executives predicted sales would decline, which they did. At the same time, though, revenue increased. In the following six weeks, the revenue generated by most of the $1.29 songs increased, according to a Billboard study, and the online store’s overall revenue went up 12 percent.85 Since label studies say between 10 and 15 percent of music fans are dedicated pirates, in most cases it makes sense to ignore them and concentrate on consumers who are willing to spend money.
More surprisingly, some of the most successful acts of the last few years aren’t on iTunes at all, and one reason for their success may be that they sell more entire albums. Most technology writers have a hard time believing this: in November 2010, when the Beatles finally made a deal to sell their music on iTunes, the Washington Post wrote that by holding out, they had “proved themselves to be lousy capitalists.”86 But in the first decade of the new century, the Beatles sold more albums in the United States than any act besides Eminem; their hits collection, 1, was the best-selling album of the decade.87 And although the Post suggested that the Beatles had ceded the market to pirates, illegal downloads of the band’s songs actually increased after their iTunes debut.88
The Beatles are an exception to almost every rule. But other acts may also have received a sales boost by staying off iTunes. In 2008, when iTunes was already the largest music retailer in the United States, the country’s No. 4 and No. 5 best-selling albums were by Kid Rock and AC/DC, neither of whom sells music on iTunes in the United States.89 (Kid Rock sells music on iTunes in Europe and in online stores that don’t break up his albums; AC/DC doesn’t currently sell any music online.) The success of that Kid Rock album, Rock n Roll Jesus, was driven by the anthemic radio hit “All Summer Long.” Since it wasn’t sold as a single, some people downloaded it illegally, but many others apparently decided to spring for the album.
To some extent, Kid Rock just found a new take on a time-tested strategy. In the 1970s, when the music business was still based around singles, some rock acts deliberately didn’t release them, in order to stimulate album sales. The most famous example is Led Zeppelin, which never put out “Stairway to Heaven” as a single in the United States; fans who wanted the song had to buy the band’s fourth album. Other hard rock bands, like Metallica and AC/DC, never even put out greatest-hits albums.
AC/DC’s attitude toward their music might be best summarized as “Have it our way.” The hard rock group doesn’t authorize samples and rarely licenses songs to advertisers. As CD sales have plummeted, AC/DC albums do as well as or better than ever, and the group sells more catalog albums than any other act but the Beatles.90 “They felt like they made albums and they wanted them dealt with in that way,” says Steve Barnett, chairman of Columbia Records, the band’s label. “My opinion is you just have to look at the numbers—that should tell you what they do is working.” Even though AC/DC’s music isn’t legally available online, it was actually downloaded illegally less than that of Led Zeppelin in 2007.91
“Popularity on file-sharing networks really only correlates to one thing, and that’s popularity,” says Eric Garland, founder of the media measurement company BigChampagne. “Is there more Kid Rock file sharing because he’s not on iTunes? Maybe. But we’ll probably never know, because a much larger factor is what’s going on with Kid Rock in general: Does he have a new album, is he on tour, did he perform at the VMAs [the MTV Video Music Awards]?”
All the performers who succeed without iTunes are “album acts,” with followings interested in more than their latest hit. Pop singers—“singles acts”—wouldn’t succeed with this strategy. And online sales are becoming more important: data provided by Nielsen SoundScan showed that Def Leppard, another iTunes holdout, would almost certainly be better off selling its music online. At some point—soon, if not already—it will be impractical for any band not to sell music on iTunes.
But several studies suggest that “unbundling” albums—the selling of individual songs—contributed significantly to the decline in music sales. A 2007 study by the consulting company Capgemini funded by the U.K. music industry found that 18 percent of the labels’ 2004–2007 revenue loss stemmed from piracy, while the rest was the result of selling music by the track.92 And a May 2010 study by the Harvard Business School professor Anita Elberse found “strong evidence of the negative consequences” of unbundling and estimates that it caused about a third of the revenue decline during the period of her study.93
Media companies have always relied on bundling to sell products for a price that can cover their fixed costs, and they’re not alone. Many businesses bundle goods in some way, like clothing stores that don’t break apart suits to sell jackets and pants separately. Although some consumers take their business elsewhere, this can be an effective strategy. Just ask Apple, which doesn’t sell separate replacement batteries for the iPod.
By the end of 2004, when the U.S. Supreme Court agreed to hear the Grokster case, it was obvious to everyone that the recorded music business was hurting, and the people who ran it knew things would get worse. That year $12.3 billion of music was shipped, up slightly from 2003 but down 16 percent from 1999.94 The companies that owned the labels were preparing for the worst as it became clear that legal action alone couldn’t stop online piracy from seriously cutting into sales. Sony and the Bertelsmann Music Group combined their operations in a joint venture. Time Warner sold its music business to a group of investors led by Bronfman. And all the labels laid off employees in what became a grim annual ritual.
The Supreme Court started hearing oral arguments in the Grokster case at the end of March 2005, and the stakes were high. If the Supreme Court turned a blind eye to businesses that essentially encouraged illegal behavior, it would render copyright meaningless in the digital age. But if it outlawed certain kinds of technologies, it would give the entertainment industry veto power over inventions like the iPod.
The justices didn’t want to do that, but they also seemed uncomfortable with the idea that Grokster depended on illegal activity. “I know perfectly well that I can buy a CD and put it on my iPod,” Justice David Souter said at one point. “But I also know if I can get music without buying it, I’m going to do so.”95
On June 27, 2005, the Supreme Court handed down a unanimous decision that Grokster and Morpheus were liable for “inducing” copyright infringement. After two successive defeats, media companies had added to their case a new argument based on “active inducement,” a concept from patent law. The Court allowed the Betamax precedent to stand, but its decision outlined a doctrine that would hold companies liable for promoting infringement. “One who distributes a device with the object of promoting its use to infringe copyright, as shown by clear expression or other affirmative steps taken to foster infringement, is liable for the resulting acts of infringement by third parties,” Souter wrote in the majority opinion.96 The decision sketched out how companies could be found liable for inducing infringement if they promoted illegal uses of a product, made no effort to reduce these uses, and depended on them for profit.
Many legal scholars praised the decision for its fairness, since the justices seemed to find a way to protect copyright without jeopardizing technological innovation. But anticopyright activists resented any limits at all. The Electronic Frontier Foundation issued a white paper, “What Peer-to-Peer Developers Need to Know About Copyright Law,” which basically outlined how file-sharing services could stay on the right side of the law.97 And while technology groups predicted the Grokster ruling would “chill technology innovation,” one could just as easily argue that the continued operation of file-sharing services would prevent the development of legitimate new businesses.98 It’s hard to imagine Apple would have spent the money to build its iTunes Store if the Napster case had made it legal to give away music online. The same goes for Spotify and almost every other online music service.
For all the anger on the other side, the Grokster case didn’t exactly save the entertainment business. As the decade went on, the value of U.S. music sales went from bad to worse, dropping 9 percent in 2007, 18 percent in 2008, and 12 percent in 2009. (These numbers include digital sales.) As record stores closed and big-box retailers cut shelf space for music, sales of CDs continued to fall at least partly because there was nowhere to buy them.
“I call it the death spiral,” says Russ Crupnick, an analyst at NPD Group. “While everyone was so focused on price, what people wanted was a shopping experience—flipping through the racks, looking for stuff.” Crupnick sees the same spiral in the way labels operate: less revenue leads to less spending, less promotion, and, ultimately, even less revenue. As much as the music business has changed, you still have to spend money to make money—look at Lady Gaga or Taylor Swift.
“I don’t know what will happen to the record labels, and I don’t think they’ll look anything like they do now,” says Charlie Walk, who was president of Epic Records from late 2005 to 2008. “But a funny thing happens when you don’t promote a project: nothing.”
The major labels are no longer the only companies that can promote projects effectively. For the last couple of decades, they had three key advantages indies couldn’t match: marketing budgets, efficient distribution, and access to radio. None of those are exclusive to them anymore, although this has less to do with the Internet than with other changes: start-ups have money, the majors now distribute music from other companies, and layoffs have created a pool of industry veterans who can be hired on a freelance basis. Established acts can now strike deals directly with retailers—as the Eagles and Journey did with Walmart—that can spend more on promotion than most labels.
“Record labels, at a time when they should be expanding their personnel, they’ve gone in the other direction,” Peter Mensch says. “So there’s less and less reason to be with them.” Mensch’s Q Prime started its own label, as have several other management companies with the staff and cash flow to provide the services only labels once could. As management companies start releasing albums, major labels are acting more like managers. Labels now sign most artists to “multi-rights deals” that let them share in revenue from songwriting copyrights, merchandise sales, and even concert performances. The British pop star Robbie Williams signed a contract like this with EMI in 2002, and Paramore and Lady Gaga have similar arrangements with Atlantic Records and Interscope Records, respectively. These deals give labels an incentive to emphasize career development over quick hits, but some acts worry that these contracts will leave them with less money in the long run.
In the half decade since the Grokster decision, all four major labels have spent significant money and resources diversifying their businesses—with multi-rights deals, merchandise businesses, and strategic investments in technology companies. But they’ve also seen their share of chaos. Warner Music Group, which went public under Bronfman in 2005, has seen its stock price lag. In 2007, EMI was sold to a private equity firm, which has since lost it to Citigroup. In August 2008, Bertelsmann sold its half of Sony BMG Music Entertainment to Sony and got out of the recorded music business entirely. Only the Universal Music Group has had a relatively smooth run, thanks to dominant market share, but it has come under pressure to cut spending.
The major labels also made some difficult decisions about how to sell music online, with varying degrees of success. Start-ups say they asked for advance fees that were too high, and perhaps they did. But those payments kept start-ups from giving away music in order to gain market share, and destabilizing the rest of the market as they did. Once the labels identified iTunes as a steady source of cash, they had an interest in not undercutting it. And so far at least, no label of any size has found a way to make much money giving away music. Online streaming services like imeem never sold enough advertising to meet their expenses, although Vevo, a joint venture in which three of the four major labels share ad revenue with Google, shows promise.
A decade after Napster, labels of all sizes are still struggling to re-invent their businesses, with consequences for the culture business as a whole. “Years ago, if you were really talented, you didn’t have to work a normal job—you were a full-time musician,” Mensch says. “We live in fear that the next Kurt Cobain is sitting around saying, ‘Fuck this, the music business is over—I’m going to join my dad at the printing plant.’ ”
Jack White is adjusting to the new music business about as well as any artist out there. The former White Stripes frontman owns his own label, Third Man Records, which releases his albums through a distribution deal with Warner Music Group. He runs a subscription service called the Vault, which sells rare songs and concert recordings for up to $20 per month. He also records and releases material from other acts, both online and on vinyl records that appeal to fans and collectors; recent products include limited-edition, multicolored vinyl, “Texas-size” thirteen-inch LPs, and Halloween singles that glow in the dark.
White has the clout to follow Radiohead and Nine Inch Nails and distribute his music online instead of relying on Warner. But he says he’s had better experiences with major labels than with indies, and he worries that the same technology which makes it easy for him to run his own label could hurt other artists. He also loves records as tactile objects.
As White sees it, music has become so easy to copy that listeners forget how hard it is to create in the first place. “I think a lot of people think that because of digital technology things happen in two seconds and you don’t go into the studio and spend six months working,” he says. “I did a lot of interviews with the Dead Weather, and people kept saying, ‘Shouldn’t music be free?’ I was giving it right back to them: ‘Food should be free, movies should be free, so should paintings.’ I really just don’t understand that mentality—it’s an insult to the artist.”
White isn’t personally concerned about losing money to illegal downloads, but he worries about how piracy has devastated labels. “It’s one of those things where people think it’s no big deal because they don’t see the drastic effects,” he says. These days a band like the Dead Weather might get a $5,000 budget from its label to make a video, he says, while a half decade ago the White Stripes might have been able to play with $300,000. It’s easy to scoff at such a budget now that artists get fans to make their videos for free. But the White Stripes got a big boost in popularity from the clip for “Fell in Love with a Girl”—one of the last truly iconic music videos—for which Michel Gondry (Eternal Sunshine of the Spotless Mind, The Science of Sleep) simulated animation with Lego bricks. Working with a director like Gondry costs money, but the investment paid off. “What I’m worried about,” White says, “is that ten years from now there will be no videos and you won’t be able to record in a studio if you want to because you can’t afford it.”
The idea that artists will give away their music assumes they’ll create it cheaply—at home, on a computer, by themselves. But making an album can take time and outside expertise, neither of which has fallen in cost the way the price of equipment has. Even if a computer can replicate the equipment at Abbey Road Studios, it cannot replace producer George Martin, whom the Beatles relied on for musical arrangements that brought their vision to life. And how would a music business based on performances and merchandise sales compensate songwriters like Jimmy Webb, or Jerry Leiber and Mike Stoller?
Artists also need time to focus on their work. Many musicians begin their careers performing for admiration and beer, much as Oberholzer-Gee and Strumpf suggest. Some record albums without much hope of getting paid, even indirectly. But artists who create great work usually do so after a few years, and musicians who can’t make money might not keep at it that long.
“How is the up-and-coming band going to quit their jobs and focus on music if they’re not signed?” White asks. “If the music is free and no one’s buying the record, who’s paying for it to be made? Who’s paying for it to be recorded? Who’s paying so that artist doesn’t have to have a day job anymore and can dedicate himself and go out on tour?” Right now, many artists signed to labels will never make a dime beyond their signing bonuses, but they’ll get the recording budget and tour support that could help them build an audience. And if their labels drop them, that audience won’t.
The idea that recorded music is little more than a way to promote concerts will also inevitably affect the quality of albums that get made. Since the mid-1960s, recording has been its own art form—hence the term “recording artist.” “I realize that now a record is nothing more than a postcard to get people to buy a ticket,” says Bob Mould. “Albums are still everything to me, but I’ve had to cut corners—I don’t make money on them anymore. I hear a lot of great music, but not a lot of great records, where you listen to it and say, ‘What a great production.’ A lot of that is because of budgets and home recording.”
Artists who don’t sign with labels have to decide what to charge for, what to give away, and how to keep Internet companies from turning the first into the second. “Today, I would have to say that I find the huge Internet companies to be more of a threat,” says Eagles frontman Don Henley. “I’m still not a fan of the large record companies, but on this issue of copyright and protection of creative works I find myself on the same side of the table with them.”
Proponents of file sharing often point out that the overall music business is fairly healthy if one counts marketing deals, publishing income, and concert ticket sales. But much of that business is generated, indirectly, by the major-label marketing of the past. Endorsement deals work best for stars, and publishing revenue depends partly on licensing hits from the past. And the concert business depends disproportionately on higher ticket prices from older stars.
Digital song sales by themselves won’t save record labels, and many executives believe they represent a transitional business model that will soon seem as outdated as CDs. “I believe that music will evolve from a product-only business—meaning you buy a CD, a video, or some combination of the above—to a product and service business,” Bronfman says. He means some consumers will continue to buy CDs or vinyl, while others will buy music subscriptions the way they now pay for cable television. “In order to have access to music—to be able to playlist, to share, to do things that the community wants to do—you’ll pay X amount per month or year.”
It’s a promising vision—the ability to hear any song from a store like iTunes for a single monthly fee. So far, though, the services that offer these subscriptions haven’t caught on with consumers. The largest in the United States, Rhapsody, has about 750,000 paying subscribers.
The one subscription service that has caught the imagination of consumers is Spotify, a well-designed online streaming service that’s available in the U.K., France, and several Scandinavian countries. Spotify announced that it would launch in the United States before the end of 2010—after planning and postponing a 2009 debut—but couldn’t arrange deals with the major labels in time. The company uses a “freemium” model: users in the U.K. can hear a limited amount of music with ads for nothing, pay £4.99 a month to eliminate the restrictions and commercials, or pay £9.99 to use it on a mobile device. The major labels like the company’s subscription business—it has been reported that they own shares in the company99—but they’re concerned that its free service is good enough that consumers won’t feel they need to pay for it. “Our perspective is that you really need to restrict free, so that it’s basically a customer acquisition vehicle and not a service alternative,” says a major-label executive who deals with digital services. “If the free service is really good, what’s the incentive to convert?”
It’s a more important question than most observers realize. As of the end of 2010, Spotify had sold subscriptions to 7.5 percent of its users, and label executives say the company wants to get to 10 percent by promoting its mobile subscriptions. But a service like Spotify could hurt labels if users who don’t subscribe choose to buy fewer CDs. As an example, let’s imagine a million music fans who spend $60 a year on CDs and iTunes songs—representing $60 million in retail revenue—but might cut that amount by a third once they start using Spotify. If the company can sell subscriptions to 10 percent of its users for $10 a month, it would generate $12 million in fees; those 100,000 customers would spend another $4 million a year buying music, for a total of $16 million. But the other 900,000 consumers using the service for free will spend only another $36 million. That adds up to $52 million—only $8 million less than before—except that the first users of Spotify will be the consumers who now spend the most on music.
The problem with Spotify is all too common in the online world. In order to compete with pirate sites, legitimate services have to aggressively cut prices or offer some media for free. But this can hamper the growth of companies with the potential to generate more revenue. It’s the new-media catch-22: you have to give away content to attract an audience that turns out to be worth less than you thought because they’re attracted to free content.
“I just think it’s very hard to compete with free—the reason people pay $50 or $100 for a concert ticket and won’t pay $10 for their CD is because you can’t get into the concert for free,” says Danny Goldberg. “What happened is this extraordinarily powerful financial juggernaut entered society and changed the rules about intellectual property. But that was not necessarily inevitable, and it wasn’t driven by all of these consumers—it was driven by people who made billions of dollars changing the rules.”