IV. ISSUES THE EFFECT OF PERFORMANCE ROYALTIES ON THE RADIO INDUSTRY 25 The Congress has recently considered legislation conferring of the consequences of performance royalties in sound recordings, The Werner analysis of the effects of performance royalties consists of two steps, a discussion of theoretical issues followed by empirical work which is interpreted in light of the theoretical analysis. While the theoretical section is described as a textbook presentation of relevant economic principles for noneconomists, the subsequent empirical results are explained with reference to this theoretical analysis and a more sophisticated statement of the argument is not 26 included for the benefit of professional readers. Therefore a critique of the theoretical section of the Werner study is not inappropriate. A major failing of the Werner theoretical analysis is that the issues discussed are oversimplified, ignoring important factors which are crucial to the analysis. For example, the theoretical discussion is applicable only to stations with annual revenues exceeding $200,000. However, according to the proposed revision in the copyright law, four classes of radio stations would be recognized for the purpose of collecting the recording license fees which would be disbursed as performance royalties to record companies and recording artists. Stations with annual revenues of less than $25,000 would pay no fee. For stations with revenues between $25,000 and $100,000 the annual payment would be $250 and stations with revenues between $100,000 and $200,000 would be obligated to contribute $750 a year. Stations with revenues in excess of $200,000 would have a choice of paying one percent of revenues or making payment according to a schedule which takes into account the proportion of a station's commercial time which is devoted to recorded material. While acknowledging that most stations would probably choose the prorated formula, the Werner analysis is based on the assumption that stations will be paying a fixed percentage of revenues. This assumption also makes the Werner theoretical analysis inapplicable to stations with annual revenues below $200,000 which are obligated to a fixed payment of either $250 or $750. The Werner study fails to make the distinction between a radio station's marginal and inframarginal programming units. 27 A station's marginal programming is programming which just barely covers its costs with the advertising revenue it generates. Additional programming offered by the station could only be offered at a loss. The effect of a recording license fee on marginal programs is discussed in detail in version 2 of the mathematical model in the appendix. It is shown that the imposition of a license fee would make some marginal programs extramarginal and they would be dropped by their stations. In addition, the reduction in total programming offered by stations would come partly at the expense of recorded material resulting in an overall reduction in the number of recordings broadcast. A probable negative side effect of the loss of marginal programs would be a general reduction in the diversity of programming provided by the broadcasting industry since it is usually the marginal programs which cater to small audiences with more esoteric tastes. Not all stations have programs which are marginal in the sense just defined. It is possible that for some stations all programming broadcast generates revenues more than sufficient to cover its marginal costs and additional programming is not offered only because all time available for broadcasting is already being used. In cases such as this it is quite possible that the imposition of a recording license fee would not affect the total amount of programming broadcast but would lead to alterations in the programming mix selected. Since the analysis for this case is complex the formal analysis is presented in version 1 of the mathematical model of the appendix while the basic results and reasoning of that analysis are given a more general exposition here. Recall that the recording license fee is calculated as a 28 percentage of station revenues and that the percentage, if a station so elects, is determined by the fraction of commercial time devoted to recorded programming. For this reason a station can influence the size of its fee by varying the proportions of recorded and nonrecorded material in its programs. Since the fee increases with the fraction of air time devoted to recorded material, stations will attempt to reduce the size of the license fee by substituting nonrecorded material such as D.J. commentary, talk shows and other live formats for recorded programming. In addition the fee itself produces a bias in favor of low-cost programming which may either reinforce or work against the shift out of recorded material occasioned by efforts to reduce the size of the fee. The total effect can only be determined empirically with knowledge of the marginal costs of different types of programming. In the absence of such cost data it must be presumed that the most likely effect of a recording license fee would be to reduce the usage of recorded material even when the total amount of programming broadcast is not affected. As with any regulation induced change in economic behavior there would be both winners and losers from a shift by broadcasters out of recorded material. Principal winners would be the suppliers of the inputs for the nonrecorded programming with which recorded material would be replaced. To the extent that airplay is important in promoting the sale of recordings, the income of some recording artists and the recording industry could suffer. In addition, current recipients of performance royalties, such as authors and composers could see their incomes reduced. In the empirical portion of its investigation of the effect of the proposed performance royalties on broadcasters, the Werner study is primarily concerned with the ability of stations to continue operation after a recording license fee has been imposed. The presence or absence of high profits and the elasticity of demand for radio advertising are correctly identified as two factors determining the ability of stations to absorb the added cost. If profits are high, stations can absorb the added costs themselves. If demand for radio advertising is inelastic, (inelastic demand means that advertiser expenditures on radio commercial time are relatively invariant with respect to the price charged for the time) then radio stations can pass on to advertisers most of the added cost of a recording license fee in the form of higher prices charged for advertising. Of course, the above argument is not at all appropriate in analyzing the impact on the stations paying fixed fees of $250 or $750. It is a well-known economic proposition that higher fixed costs can not be passed on to advertisers in the form of increased prices. This point was ignored in the Werner analysis. From a study of ratio station profit and loss statements covering the five year period 1971-1975 Werner et al. concluded that there is evidence of high profits in the industry, in the form of "hidden profits" which do not show up on the balance sheet. If an industry is competitive, high profits are quickly competed away and one would expect profits and losses to be distributed randomly among firms with the identity of winners and losers varying from year to year. It was observed that among radio broadcasters, a small subset of stations accounted for most |