Rockonomics

by Alan B Krueger

  CHAPTER 4

  The Economics of Superstars

  The winner takes it all

  The loser’s standing small.

  —Benny Andersson and Björn Ulvaeus

  The hit song “The Winner Takes It All” was released by the Swedish pop group ABBA in 1980. That year also happened to be a turning point for economic inequality. Since 1980, more than 100 percent of the total growth in income in the United States has gone to the top 10 percent of families. A whopping two-thirds of all income gains have gone to the top 1 percent. The bottom 90 percent saw their combined income actually shrink.1 Europe and Asia have also experienced a rise in the share of income accruing to top income earners in recent years.2 Why has the economy become more of a winner-take-all affair?

  The income distribution has become more skewed in the direction of top earners for a variety of reasons. Studying the music industry helps shed light on one key factor: the role of superstar markets. Labor markets have been transformed into superstar markets in many professions, where a small number of top performers earn an outsized share of income. Music has long—though not always—been a textbook example of a superstar market, and holds many lessons about the economy as a whole.

  Serious study of the economics of superstars began with the great English economist Alfred Marshall in the late nineteenth century. Marshall was interested in understanding why a small number of successful businessmen were earning greater and greater salaries. As you’ll see later in this chapter, his answer related to improvements in technology that enabled businesspeople to command far-flung empires and reap the rewards of operating on such enormous scale. Ironically, he used music as a counterexample to this phenomenon, as a case study of a profession where superstar effects were limited, because only people within earshot of a singer’s unamplified voice could enjoy his or her performance.

  Times have changed. Marshall didn’t envision speakers, microphones, amplifiers, digital recordings, streaming media, or Jumbotrons. It is a powerful reminder of the central role that technology plays in determining economic winners and losers. But technology exists within a legal and social framework that also influences economic success or failure, the size of the economic pie, and the distribution of rewards in an economy. For example, the rather meager royalties paid for song publishing rights restrains incomes for songwriters. For musicians, and those in many other fields in our increasingly celebrity-driven culture, the process by which popularity grows or decays is a critical gatekeeper for entry into superstar status, and for all that superstardom entails. Finally, norms of fairness, which constrain salaries, concert prices, and other economic variables, also play a role in determining the rewards to achieving superstardom.

  Making a Superstar Market

  Alfred Marshall (1842–1924) was the most influential economist of his generation. He marveled at the shifting income distribution taking place during his lifetime. Commenting on the 1870s, Marshall wrote, “A business man of average ability and average good fortune gets now a lower rate of profits…than at any previous time, while the operations, in which a man exceptionally favoured by genius and good luck can take part, are so extensive as to enable him to amass a large fortune with a rapidity hitherto unknown.”3 Sound familiar? The same might be said of Jeff Bezos, Bill Gates, and Mark Zuckerberg today.

  Marshall’s explanation for the growing income gap between superstar businessmen and everyone else rested on new developments in communications technology—namely, the telegraph. The telegraph connected Great Britain with America, India, and even places as far away as Australia. As a result, Marshall recognized, top entrepreneurs were able “to apply their constructive or speculative genius to undertakings vaster, and extending over a wider area, than ever before.” In other words, technology increased the scale of the market, an essential ingredient for a superstar to be able to earn a supersized income.

  Marshall, as I mention in Chapter 1, pointed to the example of Elizabeth Billington (1765–1818) to make his argument. Mrs. Billington was widely regarded as the greatest soprano of her era. But, Marshall observed, “so long as the number of persons who can be reached by a human voice is strictly limited, it is not very likely that any singer will make an advance on the £10,000 said to have been earned in a season by Mrs. Billington at the beginning of the last century, nearly as great [an increase] as that which the business leaders of the present generation have made on those of the last.” Because Mrs. Billington and other outstanding singers could not reach a large audience, they lacked the scale required to become superstars.

  Thanks to the work of Sherwin Rosen (1938–2001), a hard-nosed University of Chicago economist who possessed a dry wit and sharp mind, we have subsequently learned that scale is necessary but not sufficient in and of itself to create superstars. Rosen developed a formal economic model of superstars that showed that a second essential ingredient is needed for a market to be dominated by a small number of superstars. The contenders for the top of the market have to be imperfect substitutes, meaning that each superstar has his or her own unique style and skills that raise or lower profitability.

  If all managers were identically talented—that is, if they were perfect substitutes—it would not matter whom a company appointed as its CEO or top executives. The top businesspeople would be unable to command a premium because they would face stiff competition from equally qualified rivals. Likewise, if every musician sounded the same, it wouldn’t matter which ones we listened to on our smartphones and radios, and all musicians would be paid the same. It is only when the top contenders can distinguish themselves from everyone else in some relevant dimension that there is a chance for a market to be dominated by a superstar. And it must be the case that combining the efforts of workers of lesser talents does not overtake the output of the most talented. As Rosen put it, “Hearing a succession of mediocre singers does not add up to a single outstanding performance.”4

  In other words, both scale and uniqueness have to be present to create a superstar market. Mrs. Billington’s voice was unique, but she lacked scale. Scale magnifies the effect of small, often imperceptible differences in talent. With the ability to scale, the rewards at the top can be much greater for someone who is slightly more talented than his or her next-best competitor, because the most talented person’s genius can reach a much greater audience or market, in turn generating much greater revenue and profit.

  Because small differences in talent can generate large differences in economic rewards, and talent is often hard to judge and predict, luck also plays a significant role in determining winners in superstar markets. Luck refers to the myriad random factors and chance occurrences that can lead one person to rise to the top and another, equally talented person to fall behind. This is particularly the case in the arts, where talent is difficult to assess and requires subjective judgment, and where tastes are fickle. People in the music business often call this “the X factor.” In fact, Alfred Marshall specifically noted the importance of good luck as well as talent for achieving superstar status. In my view, the power of luck is so important that I have devoted Chapter 5 to the role of good or bad luck in determining success or failure in the music business, and in life in general.

  From Billington to Beyoncé

  Nineteenth-century contemporary accounts described Elizabeth Billington as a singer with a singular voice of great sweetness, compass, and power.5 She performed in all the great opera houses in London, Dublin, Milan, Venice, Trieste, and Paris. Rumors of turbulence in her marriage were the subject of tabloid publications and widespread gossip. In other words, Billington was the Beyoncé of her day. Citing figures from the Encyclopaedia Britannica and Grove’s Dictionary of Music and Musicians, Sherwin Rosen reports that Elizabeth Billington earned between £10,000 and £15,000 in the 1801 season singing Italian opera at Covent Garden and Drury Lane. (With characteristic wit, Rosen adds, “No information is given on endorsements.”)


  In current dollars, Mrs. Billington earned between $1 million and $1.5 million a year—a significant sum, especially in her day, but less than 2 percent of the $105 million that Beyoncé earned in 2017, according to Forbes.6

  What changed from Billington to Beyoncé? Apart from general economic growth that has raised living standards across the board, the obvious development that propelled superstar musicians’ incomes was the invention of sound recordings and amplification. First physical records, and subsequently digital recordings, have vastly increased the scale that musicians can achieve. Beyoncé is able to reach a practically infinitely large audience through streaming platforms. And once a song is produced, the cost of distributing it to a mass audience is nearly zero.

  Technology has also increased the scale that musicians can achieve in live performances. On a crystal clear night in July 2016, I joined 56,368 other fans to hear Bruce Springsteen perform at Circus Maximus in Rome.7 No one left the four-hour show disappointed; most were singing “Thunder Road” on their way out. And while the venue was available back in Billington’s day, it is likely that fewer than five hundred fans would have been able to hear her had she performed at Circus Maximus in her day. Today, the number of people who can be reached by a human voice has become virtually unlimited. Just imagine how many people could have heard Elizabeth Billington sing if she had had access to a microphone and amplifier, let alone to MTV, CDs, Apple Music, Spotify, YouTube, and Tencent; her own station on Sirius XM Radio; and a private jet to transport her around the globe.

  The ability to record and replay musical performances has enabled today’s top artists, those most in demand, to dominate the music business. Before recordings, many restaurants and bars would hire live performers to play the hits from sheet music. Now live performances are less common, and restaurants and bars instead play music recorded by the stars. After all, why would you pay for a cover band to play “Born to Run” when you can hear Bruce Springsteen and the E Street Band perform it themselves on a recording? The technology that made it possible to achieve scale in the music industry has made it a challenge for musicians below the very top echelon to earn a middle-class living. While demand for superstars has risen, demand for lesser performers has declined.

  Album sales and digital streaming clearly reflect the superstar phenomenon. The top 0.1 percent of artists accounted for more than half of all album sales in 2017. Song streams and downloads are similarly lopsided.8

  But a funny thing happened to musicians’ incomes as recorded music became more accessible. Because digital recording technology and the rise of the Internet made it easy to copy and share recordings, artists’ and record labels’ revenue from selling recorded music has plummeted since the 1990s. Touring, always an important source of musicians’ incomes, has become even more important with the demise of record sales.

  Today, recorded music is primarily a means for musicians to become popular, so that they can earn superstar incomes from touring and related merchandise sales. As David Bowie predicted in 2002, “Music is as available as running water.”9

  Touring has become more of a superstar market over the last forty years, with a rising share of revenue tilted to the top performers. Figure 4.1 shows the share of worldwide concert box office revenue accruing to the top 1 percent and top 5 percent of performers, based on my tabulations from the Pollstar Boxoffice Database. The underlying data are reported to Pollstar magazine by venues, promoters, and managers. Information is reported for box office revenue; tickets resold in the secondary market are not counted, but that money (overwhelmingly) does not filter back to the artists anyway. The Appendix provides a statistical evaluation of the strengths and weaknesses of the Pollstar Boxoffice Database, and efforts I made to adjust the data for underreporting. Although there are some lapses in the Pollstar data, and coverage was particularly incomplete in early years, the Pollstar Boxoffice Database remains our best and most comprehensive source for historical data on concert revenue.*

  The Pollstar data indicate that the top 1 percent of artists increased their percentage of total concert revenue from 26 percent in 1982 to 60 percent in 2017. The top 1 percent now take in more revenue than the bottom 99 percent combined. And the top 5 percent of performers increased their percentage of total concert revenue from 62 percent to 85 percent over the same period. The top 5 percent of performers earn almost six times as much revenue as the bottom 95 percent combined—a superstar market if ever there was one.

  Figure 4.1: Percentage of Total Ticket Revenue Accruing to Top Performers, 1982–2017

  Source: Author’s calculations based on Pollstar Boxoffice Database.

  Some have argued that the music industry has become more egalitarian because of streaming, computer music production technology, and lower entry costs. But as far as artists’ incomes are concerned, it is becoming more unequal. The driver of this rise in inequality is the rapid increase in ticket prices for the superstar performers. The top performers’ share of tickets sold and of shows has remained more or less constant since the early 1980s. What has changed is that superstars are commanding higher prices for their live performances.

  And the Pollstar Boxoffice Database probably understates the degree of inequality in artists’ net incomes, because costs vary. Although data on revenue net of costs are not available, it is likely that costs are a lower share of revenue for superstars than for less popular artists, because the superstars have greater bargaining power and can strategically organize touring locations to minimize costs. As Q Prime’s Cliff Burnstein told me, “When you’re coming to town with a hot show, the venue pretty much has to do what you want.”10

  Merchandise sales, sponsorship income, endorsements, and other potential sources of income would likely skew the industry further in the direction of a superstar market, as the superstars receive a greater share of such income.

  Before turning to the question of how superstars become so popular, it is worth emphasizing how the rules of an economy constrain or amplify superstars’ incomes. In the United States, most top musicians are independent contractors, who, together with their managers, are free to negotiate their own compensation and touring schedule.11 In South Korea, however, K-pop groups sign long-term contracts with their management companies, which essentially prevent them from receiving a large share of the profits that they generate.12 Similarly, in Japan, musicians typically work on a work-for-hire basis, which restricts their upside earning potential. And in China it is common for record labels to take a large slice of performers’ touring revenue. Economic forces affect the types of contracts that musicians are able to negotiate, but history and local customs matter as well.

  The Power of Power Laws

  A total of 33.2 million different songs were streamed in 2017, according to BuzzAngle Music, which gathers data from the various online streaming services.13 Spotify itself offers 35 million tracks. It would take six lifetimes for any single individual to listen to all those songs just one time.14 Obviously, with such an enormous volume of music available, we cannot possibly sample every song to decide which ones we like or dislike. Instead, in forming our musical preferences, we rely heavily on the advice of our friends, family, and associates; on what we hear on the radio or other media; and on expert opinion (such as the Billboard Hot 100 and curated playlists). Musical tastes do not develop independently across individuals.

  This process creates bandwagon effects, or a tendency for what is popular to become even more popular. Bandwagon effects can come about because we gain information about the existence of particular songs or musicians from our friends, or because our friends encourage us to like certain types of music. Bandwagon effects are likely to be particularly strong in this instance because music is often a social activity. Listening to music is a common experience that people like to share with others. We want to be familiar with the music that our friends are familiar wi
th because it strengthens bonds of friendship and enhances the experience of listening to (or dancing to) music. Thus, there is a tendency for music that is popular in our networks to become more appealing to us. And bandwagon effects are further reinforced by the well-established psychological tendency for people to grow to like a particular song the more they hear it.15

  There are two key economic implications when the popularity of a good is determined, in large part, by transmission through social networks. First, the distribution of what is popular will be highly skewed, with the most popular items absorbing most of the oxygen. Second, the determination of what is most popular is highly susceptible to random perturbations in the ways in which new products are introduced to a market and ripple through networks of potential customers.

  In statistical jargon, the cascade of information and musical preferences through networks of fans generates a power law distribution of popularity—the popularity of the most popular item is a multiple of the next-most-popular item, and so on down the line. As a result, a small number of players—superstars—come to dominate a market. The so-called 80/20 rule (Pareto’s law), where 20 percent of a firm’s customers are responsible for 80 percent of sales, is an example of a power law that is common in business.

  To conceptualize the way in which social influence operates, suppose each person who is considering making a purchase of a song sometimes follows her own judgment and at other times follows the behavior of a friend. Specifically, in p percent of cases she buys the recording if, in her own, independent judgment, she believes the recording is worth purchasing. In the remaining cases, she simply follows the decision of her friend. So in (100 −p) percent of cases, she either buys the recording if her friend bought the record or she doesn’t buy the recording if her friend didn’t buy it. Her friend acts the same way, relying on her own judgment part of the time and on another friend’s judgment part of the time.16 This decision-making process would generate a power law distribution of song purchases. Even if all the people in this example like every song equally, the reliance on the behavior of others in forming one’s own decision creates a bandwagon effect and a highly skewed distribution of song purchases. Some unlucky songs that were not purchased initially will wither away with hardly anyone purchasing them, because no one recommends them, and others will grow in popularity because the fact that they were already purchased by many early consumers leads additional consumers to purchase them. Mathematically, this copycat procedure will result in a multiplicative relationship in popularity, with a small number of songs responsible for most of sales. Such is the way that popularity snowballs up- or downhill.