percentage decreasing relative to music use. Other than that change, the SAMPRA formula remained largely
unaffected. It is against that determination that this appeal in terms of section 36 of the Act is directed.
[57] It is necessary at the outset to say something concerning the rather tortuous statutory scheme in terms of
which the Tribunal derives its power. As pointed out by Dean in Handbook of South African Copyright Law,10
the factual matrix set out in Chapter 3 of the Act has to be read "mutatis mutandis to accommodate the
adjudication of disputes arising out of s 9A". This means that one has to strain to make those provisions
compatible with those of section 9A, more particularly those of section 9A(1)(b) and (2)(c). Put simply, the
licensing scheme provisions are applied to the determination of the royalty rate. That notwithstanding, the
learned author rightly points out that to adopt a different view would render the provisions of section 9A
nugatory a consequence that should be avoided. Section 33(5)(b) therefore, applies mutatis mutandis and
requires the Tribunal, when it is determining a royalty rate, to make such order as it may "determine to be
reasonable in the circumstances".
[58] One other preliminary issue that needs to be addressed, albeit briefly, is the failure of the Regulations under
the Act to prescribe the procedure for the adjudication of a royalty rate. It is distressing that, despite a lapse
of more than three years since the report by the Copyright Review Commission in which it described the
regulations as vague and unclear and lamented the failure to prescribe a procedure for the adjudication of a
royalty rate, no progress has been made. This is an aspect the Minister should address urgently.
Page 278 of [2014] 2 All SA 263 (SCA)
[59] I now turn to deal with the manner in which the Tribunal made its determination. In my view, Sapire AJ
wrongly discounted the evidence that both parties took care to adduce to assist him in reaching a decision.
There are areas of commonality and divergence which ought to have been taken into account in arriving at a
conclusion.
[60] First, almost all the witnesses were agreed that in considering revenue, the financial statements of the radio
stations ought to be considered rather than the SAMPRA calculation of the rate card less a 15% discount.
[61] It is not irrelevant that many countries calculate a royalty rate based on the percentage of revenue with some
choosing to increase rates relative to revenue thresholds. I will in due course deal with comparable rates.
[62] That the broadcasting industry is extensively regulated is a factor to be taken into account in NAB's favour.
[63] It does not appear that royalty rates for sound recordings internationally exceed composer royalty rates. It is
arguable, though not definitive, that composers are the key component in relation to the production of music.
In Chile, sound recording rates are determined at half the rate of the composer royalty rate.
[64] Whilst NAB's proposal that revenue should be linked to time channels, distinguishing between peak periods
and the midnight shift in which listenership is minimal, is superficially attractive, the following oddities arise: It
discounts the fact that advertisers are offered free spots in the midnight shift in order to induce them to pay a
premium for advertising spots during peak time, and it ignores the fact that the greater amount of music is
used during offpeak periods.
[65] A concern expressed by the Advisory Committee on the Act in relation to the imposition of too high a royalty
rate was the financial implications for South Africa, which translated, means excessive negative currency
outflow. As pointed out, much of the money collected by SAMPRA will find its way to the USA and other
countries.
[66] In addition, consideration should be given to perverse consequences for the music industry by too prohibitive
a rate driving broadcasters to the alternative of using session musicians and the like.
[67] The problem with the introduction of profitability in the form of the industry average is that it does not
properly or necessarily give vent to the payforplay principle and might be punitive on people whose profits
are below the industry average. It might also lead to manipulation on an accounting basis in order to lessen
the royalty burden. There appears to be no precedent internationally for including profitability in the
calculation of a royalty rate. Dr Bishop, who testified on behalf of NAB, did not champion its inclusion and left it
to others to justify.
[68] Although there appears to be some logic to factoring audiencereach into a formula, one has to be mindful of
the difficulties of valuing an audience as became evident from the evidence set out above.
[69] There is force to the submission on behalf of NAB in relation to editorial content, namely, that the exclusions
by SAMPRA are too liberal. Whereas advertisements ought rightly to be excluded, it appears to me that
Page 279 of [2014] 2 All SA 263 (SCA)
programme promotions and other content such as charity drives or competitions ought not to be. They are
part of the total cachet of a radio station and can all be said to be part of the revenue generating effort.
[70] As far as can be ascertained, only 6 countries, who probably all qualify to be described as developed
countries, pay a rate of more than 5%. Only two developed countries pay more than 6% of total revenue.
India, which is probably the more closely comparable country, charges between 1 and 2% of total revenue. In
considering the international experience and practice, I am not unmindful of South African circumstances, more
particularly that South African performers have been clamouring for years for their due.
[71] It is clear that the factor of 10 representing a maximum rate of 10% proposed by SAMPRA is purely arbitrary.
Sapire AJ considered that, because the determination by him involved a value judgment, he was free to arrive
at a conclusion by a sense of what was reasonable without any real consideration of all of the factors set out
above. He appears to have considered that the SAMRO rate multiplied by two would be reasonable. In this