Thanks for the kind introduction Vic. I figured I would start off with a more substantive post about payola before I try to sell folks on our Poker and the Law project.
I spend a lot of thinking about payments for product distribution like slotting arrangements in grocery stores or payola in the music industry. NY AG Eliot Spitzer’s investigaion of Sony BMG Music Entertainment resulted in an agreement prohibiting Sony BMG (others may follow) from making payments in exchange for radio airplay. Others in the blogosphere have responded to the investigation (see Ribstein and Picker commenting here and here emphasizing the potential for Satellite Radio to alleviate competitive problems in the industry).
Competition for product distribution is crucial to a variety of industries: slotting allowances for grocery store shelf space, payments for inclusion in mutual fund “supermarkets,” and for listing preference in search engine results. Despite the widespread use of payments for distribution in markets, payola can lay claim to the most colorful history of regulation and controversy. The arguments against payola appear to come in two flavors:
(1) payola is bad because it disfavors small music publishers (the “competition objection”); and
(2) payola is bad because it deceives the listening audience,who believes that music is chosen based on merit rather than payola (the “deception objection”).
Each objection has taken center stage at various stages of the payola debate. Spitzer’s investigation explicitly adopts the deception theory, stating that “Sony BMG and other labels present the public with a skewed picture of the country’s ‘best’ and ‘most popular’ music.”
Ronald Coase responded to both objections in his seminal paper “Payola in Radio and Television Broadcasting.” With respect to the “competition objection,” Coase argued that small publishers had thrived under payola and protested attempts to ban it by the FTC. Further, it is not likely that the competition objection would amount to much in the presence of healthy capital markets where investors are willing to get behind artists likely to produce popular music.
As to the second objection, Coase employed economic analysis to show that radio stations’ would respond to incentives, which for the stations meant playing the mix of records that would maximize the popularity of the station and therefore maximize advertising revenues (as well as future payola payments). Attempts to select music that the audience did not ultimately approve of would result in losing radio station traffic, which would in turn, reduce advertising revenues and future payola. The point of Coase’s analysis is that payola introduced a price system that would efficiently allocate resources in the music industry, and that a ban on payola would therefore reduce efficiency and community wealth. An ancillary point of Coase’s analysis was that a payola ban “may result in worse record programs” because station song selection will depend solely on maximizing advertising revenues. I am not confident that I know what the mix of songs that attracts the best demographic audience for radio commercials looks like. Nor am I confident that I, or any regulator for that matter, could figure out whether such a mix of songs would be better or worse than what we have now.
But Coase’s pathbreaking analysis of payola is not complete. While pointing out the benefits of a price mechanism for radio spins, it does not address a fundamental economic question: why do music publishers have to pay radio stations for playing music at all? The answer “because radio station spins are scarce” is not enough. Radio stations also benefit from playing a mix of songs that attract a larger audience. If radio stations are interested in selecting the songs that will maximize advertising revenues (and future payola), why does the music publisher need to pay at all? Why doesn’t competition result in the efficient level of spins (that which maximizes joint profits) without payments? This is an important and unanswered question that sheds light on the pro-competitive role of payola in the music industry.
The answer is that music publishers must pay radio stations to promote singles (give more airtime) than they would otherwise because radio stations do not take into account the marginal profits earned by the publisher as a result of record sales produced by the additional airtime. The publisher earns substantial profits from record sales with each incremental spin granted by the radio station. The radio station does not take into account these profits when determining the number of spins to give a particular record and will therefore undersupply spins of a particular song without further compensation. Therefore, the music publisher must pay for the additional spins in order to achieve the efficient (jointly profit-maximizing) solution.
The radio station cannot just take payments to play horrible music and ultimately survive. This is implicit in the popular claim that payola is the cause of the rise of “bad” music, or is only necessary for such music. This was the claim with respect to rock n roll in the 1950s, and more recently, rap music (where payola is quite popular) and the musical stylings of J-Lo. But payola has always been a part of the music industry, even when the songs performed did not raise the same type of criticism. This suggests that economic forces other than bribery must be at work. The radio station is constrained by selecting those singles that will not cause a substantial reduction in its audience, while it does retain some discretion over its overall song mix because it faces a downward sloping demand curve (i.e. the station is able to allocate spins such that playing song X rather than song Y will typically not result in a substantial loss of radio traffic).
The key economic point is that supply of these incremental spins (promoting sale of one record over another) are not likely to have a large impact on inter-radio station competition, but they are significant forces in competition between music publishers. Of course, radio stations do not have unlimited discretion to play any record they want for an unlimited number of spins --- at some point, the radio station’s disregard for its listening audience will induce closely monitored station switching. In general, however, radio stations have an insufficient incentive to promote a particular single without compensation from the publisher. One can think of payola as the record company sharing the profits created by the incremental record spins with the radio station.
Benjamin Klein and I lay out the details of this theory in our forthcoming analysis of slotting allowances (payments for shelf space) in the grocery industry (an older draft is available here).
The analogy to grocery store shelf space is instructive. Like radio stations, grocery stores face downward sloping demand curves (and therefore can change the allocation of products without losing all customers) and seek to maximize store traffic. The manufacturers of brand name products sold in grocery stores where we observe slotting allowances earn large incremental profits on additional sales created as a result of the supermarket allocating the product preferential shelf space (because the wholesale price is significantly greater than the manufacturer’s marginal cost) as the music publishers earn additional incremental profit from the record sales created by incremental spins. In each case, it is the distributor that does not take into account these additional profits in choosing the promotional resources to dedicate to a particular album or product. Therefore, the manufacturer must compensate the distributor for this additional promotion in order to achieve the efficient solution.
Once one understands why record company and radio station incentives do not coincide with respect to airtime, the logic of payola payments and their important role in the competitive process for music distribution is easier to digest. But other questions remain. Would a ban or mandatory disclosure of payola improve the outcome?
Most economists would be suspicious of an argument that more disclosure is a bad thing, although the Spitzer settlement goes beyond mandating disclosure and bans payola. What is interesting is that many feel more “deceived” by payments in the music industry then by slotting allowances or other payments for product placement. Why is this? Perhaps there is something different about radio. It could just be that consumers feel differently about music than about their groceries. From an economic standpoint, one could argue that consumers perceive the radio station has having something closer to a fiduciary duty to select the best products than does the supermarket. But at first glance it appears that radio stations ought to be more concerned with offering poor products since a consumer listening to that station can switch stations with the push of a finger rather than driving to a new supermarket. Perhaps paradoxically, both consumers and regulators seem to believe that competition is more likely to deceive consumers in the music industry. Let me be clear that I am not against a regulation demanding disclosure of these payments, but am merely suggesting that there are some very interesting questions surrounding this issue that have not been resolved.
The fact that payola is an important part of the competitive process did not escape Coase. Coase’s account includes a detailed history of payola tracing back to 1867 when public performers were paid to perform songs from a publisher’s catalog. The most striking feature of the history of payola is the series of unsuccessful attempts, each initiated by the music industry, to stop payola on their own (at least one attempt in 1890, 1916-17, 1933, the more well-known attempts to amend the Communications Act in 1960, and the shortlived suspension of independent promoters in 1986 following a 1984 Senate investigation).
Because radio airtime is a substitute for advertising, it is completely unsurprising that music publishers desired to collude to stop advertising --- an important dimension of competition for record sales. Collusion is notoriously difficult to accomplish in the first instance, and even harder to sustain because members of the cartel increase profits by deviating from the collusive agreement. Successful collusion often takes a third party to regulate the agreement and punish defectors. Occasionally, would-be cartel members are able to persuade the government to take the job. It appears that Spitzer may succeed where the recording industry has failed for over a century by stepping up to police the industry restriction on competitive payments for spins.
TrackBack URL for this entry:
Links to weblogs that reference Some Economics of Payola:
I noticed this article on the payola settlement the other day, which dusted o ..." [more] (Tracked on August 4, 2005 @ 16:09)
1. Posted by Vic Fleischer on August 2, 2005 @ 15:23 | Permalink
Did Coase address agency costs at the radio station level? Seems like the payola often goes to DJs, not the stations. And so the stations then have to spend more time and money monitoring what the DJs are doing. That hardly seems efficient.
2. Posted by Joshua Wright on August 2, 2005 @ 16:00 | Permalink
Interestingly, payola went primarily to the DJ's during the early years (until the 1950s) and subsequently to the Program Directors at the station. Payments to either would raise the type of agency cost problem you are concerned with.
Coase did consider these costs in his analysis, arguing that "the lack of any serious effort on the part of the broadcast industry in the 1950s to prevent payola suggests that there was a broad congruence of interest between the operators of radio stations and the disc jockeys." He also argues that DJ's accepting payola would reduce the salary for DJ's and increase station profits.
I certainly agree with you that the station/DJ (or Program Director) would involve monitoring costs. But I suspect that these costs would be relatively small (it is easy to tell what songs the agent has selected) compared to the benefits to the station (reduced salary and the portion of the payments which accrue to the station itself). The fact that stations willingly allow PD's to select these songs is suggestive that the benefit to the station outweighs the costs and total wealth is increased as a result of the arrangement.
3. Posted by jallgor on August 4, 2005 @ 16:24 | Permalink
I don't think the market for determing what "mix of records that would maximize the popularity of the station" is an efficient one. I think one phenomenon that is not taken into account in this analysis is the ability of pop culture to literally make something popular. I don't know if this is documented but I think that most music can be made popular simply by playing it enough. It's the "everyone else likes it so I have to like it too phenomenon." Remember that much of the music industry is catering to adolescents and teens where this phenomenon is greatest. Buying shelf space at grocery stores is not an apt analogy because no matter how much manure you put on the shelves people still won't eat it. In addition, since so many stations are owned by just a couple companies they all have the same playlists. The result is that the stations don't have to worry about what they play because a) if they play it enough people will think they like it and b) there is no alternative. Satellite radio's effect ont his should be interesting. Most people I know switching to satellite radio aren't switching because they dislike advertising they are switching because of the better programming.
4. Posted by Joshua Wright on August 4, 2005 @ 21:52 | Permalink
"The result is that the stations don't have to worry about what they play because a) if they play it enough people will think they like it."
I am not quite sure what this means. My analysis assumes that consumers make decisions with some rhyme and reason, specifically, they maximize utility based on a set of preferences (songs, records, etc.). Stations clearly are worried about playlists and compete vigorously for competitors' listeners. It is quite possible that social interactions (the fact that somebody else like's a particular song) may increase my demand for it, but this is not irrational.
5. Posted by markm on August 7, 2005 @ 10:19 | Permalink
Payola is simply payment for advertising. The record companies hope that those who hear their songs on the air will then buy the recordings in the store. Whether this has increases or reduces competition depends upon the way the payola market is structured. Can a small company desiring to get a single song on the air get an appointment with the program manager to make their offer, or do the major labels with huge playlists monopolize the program manager's time?
The first reason people get emotional about this in a way they don't about grocery stocking fees is that music involves emotion much more than food does, at least among the vast majority of Americans that have never been really hungry. Both songs and groceries are promoted by a system that bears a passing resemblance to bribery, but people care more about what's on the radio than whether Post or Kellog gets more space in the breakfast cereal section. However, I do not think this is the whole story.
Many radio stations will relentlessly self-promote their playlists as being more popular than other songs. However, if I understand correctly, that "top 40" list is based simply on the number of plays on the air. Whether listeners like the song or not is unknown, and I think unknowable with any precision until CD sales results come back many weeks after the CD is introduced. At best, this means the top 40 (or whatever) are popular with station program directors, who had better have a fair idea of what the public likes if their stations are going to keep their listeners. At worst, it might mean the top 40 were selected entirely according to the size of the payola - which makes any pretense to playing the most popular songs close to fraudulent.
Note however that there are strong economic motives for the record companies to pick their best songs for promotion through payola, and for stations to turn down songs that would repel their audience, no matter the amount of payola. With or without payola, songs that get played are ones that all the experts involved think will go over well with the audience. Sometimes the experts are all mistaken...
Finally, someone who hates most of what is on the radio might prefer to think that, "the record companies bribed the station to play trash", rather than "my taste must differ from most of the listeners", or "my tastes are so unpredictable the experts keep getting it wrong".
6. Posted by Alex Wright on May 1, 2007 @ 4:50 | Permalink
But you must take into account the brains natural ability to gain a liking to/ or gain an affinity for a piece of music that it hears more often than not.
The more the person hears a particular song, the more that person enjoys that song, therefore a definite link can be made between a song put on the air under a certain payola to a company that pays for, or "sponsors" this song.
Therefore whether the music is "bad" or not the more you hear it the less "bad" it becomes, so a direct link can definately be made between sponsors and songs; leading to the radio stations not repelling their listeners at all, but keeping their market base (or even increasing it), whilst recieving more cash benefits.