Copyright Board Canada

Commission du droit d’auteur Canada

 

Date	2017-12-01
Citation	File CB-CDA 2017-148
Regime	Copying for Private Use Copyright Act, subsection 83(8)
Members	The Honourable Robert A. Blair Mr. Claude Majeau Mr. J. Nelson Landry

Tariff of levies to be collected by cpcc in 2018 and 2019 on the sale of blank audio recording media, in Canada, in respect of the reproduction for private use of musical works embodied in sound recordings, of performers’ performances of such works or of sound recordings in which such works and performances are embodied

Reasons for decision

I. INTRODUCTION

[1] On March 29, 2017, the Canadian Private Copying Collective (CPCC) filed with the Board a Statement of proposed levies to be collected in 2018 and 2019 on the sale of blank audio recording media, in Canada, in respect of the reproduction for private use of musical works embodied in sound recordings, of performers’ performances of such works or of sound recordings in which such works and performances are embodied (“private copying”), pursuant to section 83 of the Copyright Act (the “Act”). [1] The proposed tariff targets recordable compact discs (CD-R, CD-RW, CD-R Audio, CD-RW Audio: together “blank CDs”). The proposed tariff sets the levy at 29¢ for each CD.

[2] On May 20, 2017, the Board published the proposed tariff in the Canada Gazette. On June 17, 2017, Mr. Sean Maguire filed an objection to the proposed tariff. No other objections were received prior to the end of the objection period, July 12, 2017.

[3] On August 8, 2017, the Board issued a schedule of proceedings, consisting of the filing of CPCC’s case, the filing of Mr. Maguire’s case, and the filing of CPCC’s reply case. [2] The Board did not contemplate an oral hearing at that time. When CPCC filed its case in chief, it confirmed that both Mr. Maguire and it agreed to a “paper-only” proceeding. [3] On August 28, the Board sent technical questions to CPCC. CPCC answered the questions, Mr. Maguire responded to those answers and CPCC replied to that response. Finally, CPCC replied to Mr. Maguire’s statement of case. The filing of CPCC’s reply, on September 29, 2017, perfected the record.

[4] CPCC is an umbrella organization that represents songwriters, recording artists, music publishers and record companies through its four-member collectives: the Canadian Musical Reproduction Rights Agency (CMRRA), Re:Sound Music Licensing Company (Re:Sound), the Society of Composers, Authors and Music Publishers of Canada (SOCAN) and the Society for Reproduction Rights of Authors, Composers and Publishers in Canada (SODRAC). CPCC has been mandated by its constituent member collectives to collect and distribute private-copying levies on their behalf and has been designated as the collecting body by the Copyright Board pursuant to paragraph 83(8)(d) of the Act. [4]

[5] Mr. Maguire is an individual who has purchased recordable CDs in the past. [5]

II. POSITION OF THE PARTIES

[6] CPCC’s position consists of two points. First, based on a conservative application of principles used by the Board, CDs continue to qualify as a medium ordinarily used by individual consumers to copy music. Second, the value of the private copying levy should remain at 29¢ per blank CD; this value is fair and equitable.

[7] Mr. Maguire also makes two points. First, he argues that the Board should revisit its definition of “ordinarily used;” by implication, a new definition of “ordinarily used” might not cover blank CDs, given the way CDs are used today. Second, he argues that if the overall cost of administering the private copying regime (including costs to CPCC, to payers of the levy, and to the Board) exceeds the revenues generated thereby, “maintaining the tariff would make no practical sense.” [6]

III. EVIDENCE

A. CPCC

[8] Ms. Laurie Gelbloom, General Counsel of CPCC, filed a witness statement. [7] Her statement discussed CPCC’s management and staff, its financial position, the distribution of royalties, and enforcement of the tariff.

[9] Ms. Gelbloom also presented data from a CPCC survey of retail prices of blank CDs. The survey was conducted annually from 2009 through 2017; it tracked prices for CDs sold singly, in spindles of 50 CDs, and in spindles of 100 CDs. From 2009 through 2012, the survey data reflect prices in six Canadian cities, collected two to three times per year. Both the frequency of data collection and the number of cities decreased after 2012. By 2017, data were collected only once in the year, and only from retail establishments in Toronto. [8]

[10] Mr. Benoît Gauthier filed an expert report, [9] updating the report by Audley, Freeman, and Gauthier filed for Private Copying 2017. [10] The purpose of this report is to marshal data and projections from those data to answer the question of whether or not CDs continue to qualify as a medium that is ordinarily used by individual consumers to copy music.

[11] Mr. Gauthier noted that the source of data for his report was the Music Monitor Survey he conducted from April 2015 through March 2016. Projections from those data follow technique 2, as described in Private Copying 2015-2016. [11] Technique 1 does no formal interpolation, but rather calculates implied growth rates from periods where the starting and ending points are known, extending the data beyond the ending point using that growth rate. Technique 2 interpolates data for missing years and then treats the interpolated data as if they were actual data for the purpose of forecasting. [12] The forecasts are based on a constant growth rate.

[12] Mr. Gauthier forecasts that 200.1 million tracks will be copied onto blank CDs in 2018 and that 183.9 million tracks will be copied onto blank CDs in 2019. [13] He forecasts that, of all tracks copied onto all media and devices, 7 per cent will be copied onto CDs in 2018 and 6.2 per cent will be copied onto CDs in 2019. [14] Mr. Gauthier forecasts that 7.15 million blank CDs will be purchased by individuals in 2018 and 5.7 million in 2019. [15] Finally, he forecasts that 33 per cent of blank CDs will be used to copy music in both 2018 and 2019. [16]

[13] Mr. Gauthier then examines these forecasts in the context of prior Board decisions. The figure of 200 million tracks is to be compared to 26 million tracks copied onto CDs in 1999. The figure of 7 per cent of copies being onto CDs is to be compared to 5 per cent of copies being onto CDs in 1999. Mr. Gauthier then multiplies the last two figures he forecasts, to obtain 2.25 million blank CDs to be purchased by individuals for the purpose of copying music in 2018; the corresponding figure in 2019 is 1.79 million. [17] This is to be compared to 1.05 million CDs purchased by individuals for the purpose of copying music in 1998. Finally, he notes that the Board has stated in 2016 that 33 per cent of CDs used for copying music is a significant measure. Accordingly, he notes that his forecasts are consistent with a conclusion that blank CDs will continue to be ordinarily used by consumers to copy music in 2018 and 2019. [18]

[14] Dr. Marcel Boyer filed an expert report on the value of the levy. [19] After summarizing the Board’s decisions since Private Copying 2011, [20] all of which left the levy unchanged from its 2010 rate of 29¢, he turned to the question of the economics of product life-cycles.

[15] As Dr. Boyer noted, the Board has been predicting the demise of the market for the blank CDs since Private Copying 2011; this has not occurred. Rather, the data appear to have stabilized in 2015-2016. [21] Furthermore, the Board has assumed that the price of a product drops at the end of its life-cycle; Dr. Boyer notes that this is not necessarily the case. Viewed from the lens of targeting a market subgroup (a type of price discrimination), it is possible for a seller to raise prices so as to target the diehard group of customers whose demand is relatively less elastic. [22]

[16] Dr. Boyer cited two papers in support of his assertions. First, Melser and Syed (2006) analyzed computer prices. While prices for desktops fell at the end of their life-cycle, the reverse was true for laptops. Second, Baker et al. (2010) developed the concept of the diehard market segment, noting that these buyers may face seemingly prohibitive switching costs.

[17] Dr. Boyer concluded that there is neither life-cycle evidence nor theory that suggests the Board should deviate from the 29¢ levy it has set since 2010.

[18] The responses to the questions posed by the Board were filed as Exhibits CPCC-5 and CPCC-5A; each question was answered by different people, depending on to whose report the Board’s question pertained.

[19] Question 1, relating to CPCC’s annual survey of retail prices, was answered by Ms. Gelbloom and Mr. Gauthier. The response contained 4 tables with additional details about the price survey. In particular, it contained data only from the Greater Toronto Area in June of the relevant year so as to allow intertemporal comparisons. The response also contained several measures of data accuracy and precision.

[20] Question 2, relating to the Music Monitor Survey, was answered by Mr. Gauthier and CPCC itself. The response showed a strong seasonal pattern for purchasing blank CDs and copying music onto them: copying/purchasing is strongest around the end of the year and weakest in the summer. CPCC added that there are no plans to conduct the Music Monitor Survey again because of the costs involved.

[21] Question 3 relating to life-cycle theory, was answered by Dr. Boyer. He draws an analogy between brand loyalty displayed by purchasers of Apple computers and format loyalty displayed by purchasers of blank CDs. Based on the analysis of Apple’s pricing strategy as described in Copeland and Shapiro (2016), it is consistent to observe little change in pricing of blank CDs from year to year.

[22] Dr. Boyer also addressed the concept of creative destruction. [23] In general, creative destruction has the effect of shortening the life-cycle of technological products. It has likely done so for blank CDs. However, the data show that the life-cycle is not over. Furthermore, most of the damage has been done: remaining buyers are stubborn in their loyalty to the CD technology.

[23] Exhibit CPCC-6 constitutes CPCC’s reply to Exhibit Maguire-2. CPCC makes three points. First, while Mr. Maguire claims the survey data are too old to be reliable, the survey data meets the standard set by the Supreme Court in Mattel, [24] as noted by the Board in Private Copying 2005-2007. [25] Second, Mr. Maguire misunderstands the purpose of the retail price survey data: he seems to contend that they are used for showing that CDs are ordinarily used for private copying when they are in fact used for showing that the retail price of CDs is stable. Finally, while Mr. Maguire pointed to CPCC financial data and mused as to whether or not it continued to be worthwhile to collect the levy, CPCC noted that it would not have filed proposed tariffs for 2018 and 2019 if it had not thought that its revenues would exceed its costs.

[24] Finally, Exhibit CPCC-7 constitutes CPCC’s reply to Mr. Maguire’s statement of case. CPCC refuted Mr. Maguire’s arguments (discussed below) relating to “ordinarily used,” based on Board decisions. CPCC addressed Mr. Maguire’s contention that the Board should take wider costs to society into account. It argued that the Board should not do so for three reasons. First, the purpose of the private-copying regime is a public purpose. Second, CPCC is not a profit-maximizing corporation but rather a non-profit entity. Third, taking wider costs into account exceeds the Board’s jurisdiction.

B. Mr. Maguire

[25] Mr. Maguire conceded that if the Board maintains its standard for ordinarily used, CDs qualify. However, he argues that the Board’s standard is a “low Threshold.” [26]

[26] Mr. Maguire argues that CPCC must state which data are new to these proceedings. By implication, if there are no new data, this is a fact that the Board should take into account. In his reply to Exhibits CPCC-5 and CPCC-5A, Mr. Maguire continues to emphasize this point. He notes that CPCC has reduced the number of retail surveys to a single one, and did not conduct the Music Monitor Survey in 2016-2017. He argues that the Board may not even have enough current data to show that blank CDs are “ordinarily used.” While Mr. Maguire does not mention burden of proof, his argument is intrinsically linked to burden-of-proof considerations.

[27] Mr. Maguire files data from CPCC’s website relating to its revenues and expenses. He concludes that CPCC’s expenses are approximately $1 million per year and relatively fixed, while its revenues have continued to drop over the past several years, implying that its expense ratio is rising. From these data, Mr. Maguire muses that it may no longer make “practical sense” to operate the private copying regime, especially when taking into account the expenses of other parties, such as the Board, and manufacturers and importers of blank CDs.

IV. LEGAL ISSUE: “ORDINARILY USED”

[28] Mr. Maguire raises the question of “ordinarily used.” Is the Board’s threshold too low? Does the definition of “ordinarily used” adopted by the Board in previous decisions remain relevant?

[29] In Private Copying 1999-2000, [27] the Board first addressed the term “ordinarily used.” It explained that it should include all non-negligible uses, so long as it is used by more than a few eccentrics. While the Canadian Storage Media Alliance (CSMA) preferred a different definition of “ordinarily used,” [28] the judicial review of the Board’s decision found that the decision was not patently unreasonable. [29]

[30] Later in that decision, it referred to the “threshold of ordinariness.” [30] The term threshold was next mentioned in Private Copying 2003-2004, [31] when the Board explained that it did not think its threshold was too low. In Private Copying 2010, the Board referred to the “threshold of ordinary use.” [32]

[31] In essence, the Board’s approach has been data-driven. Data filed by CPCC for each proceeding are used to determine whether each medium is “ordinarily used.” These data are compared to comparable data for other media for which the Board has certified or declined to certify a private copying tariff in the past. In respect of any medium, the comparison is done for a set of relevant variables, such as the number of musical tracks copied onto CDs, or the number of blank CDs purchased for copying music. Of course, which variables are relevant and which data are compared may differ for each proceeding. The key point is that it is up to CPCC to file data to show that each medium for which it wants a tariff is “ordinarily used.”

[32] Mr. Maguire suggests that the Board revisit the definition of “ordinarily used.” In AVS, the Federal Court of Appeal held that the issue was one of statutory interpretation. As such, it confirmed that the Board’s approach to interpreting “ordinarily used” in Private Copying 1999-2000 was not patently unreasonable. Although the Board is not bound by its previous decisions, Mr. Maguire has not put forward an alternative definition the Board can analyze.

[33] We find once again that the Board’s approach to “ordinarily used” as explained in Private Copying 1999-2000 is appropriate to use in these circumstances.

[34] This approach can be summarized as follows. First, the definition emphasizes consistency rather than frequency. [33] Second, the definition focuses on uses by individual consumers rather than all uses. Finally, the application of the definition can draw on quantitative considerations but need not be exclusively quantitative.

[35] The second implicit proposal of Mr. Maguire is harder to discard. It is possible to conceive of other data that could be more relevant than the ones we use here. The Board has focused in the past on absolute and relative measures of quantity of private copying; the objector could have supplied other such measures. While Mr. Maguire’s observation that the Music Monitor Survey data are imperfect may have some merit, the absence of data for which to substitute makes that observation difficult to put into practice. Without any alternative measure being proposed, we refrain from reinterpreting the term “ordinarily used.”

V. ANALYSIS OF THE QUALIFICATION QUESTION

A. The Relevant Variables

[36] As mentioned above, there are two types of data that have been collected in the past – absolute measures of the amount of private copying and relative measures of the amount of private copying.

[37] The number of tracks copied privately onto CDs is an absolute measure. The number of blank CDs bought by individuals for the purpose of copying music is an absolute measure. By contrast, the fraction of tracks copied privately (using as denominator the number of tracks copied onto all media or devices) is a relative measure. So too is the fraction of CDs used for copying music (using as denominator the total number of CDs bought by individuals).

[38] The advantage of absolute measures is that they do not depend on data from non-leviable media. Looking at the use of other media and devices to determine whether or not CDs qualify introduces an inconsistency. Depending on which media and devices are included in the universe of study (that is, mentioned in the questions of the Music Monitor Survey, for example), the relative measures give different results.

[39] Viewed from a dynamic perspective, this introduced a potential unfairness. Suppose there is no change in the use of CDs for private copying of music. However, if the use of non-leviable media increases substantially over a period of a few years, a relative measure might suggest that CDs no longer qualify, even though there has been no decrease in their usage for private copying. This strikes us as unfair.

[40] This principle can be further clarified by use of an example. At present, about 2.9 billion tracks are copied onto all media and devices each year, whereas CDs make up about 200 million tracks. The remainder, 2.7 billion tracks, are tracks copied onto the following media and devices: DVDs, memory cards of all types, digital audio recorders, cellular telephones, and others.

[41] CPCC has proposed tariffs in the past for DVDs, memory cards (later limited to microSD cards), and digital audio recorders. DVDs were never accepted by the Board as “ordinarily used.” [34] MicroSD cards were expressly removed from the regime by regulation. [35] The Federal Court of Appeal determined that digital audio recorders are devices, not media, and therefore cannot qualify as a leviable medium. [36] CPCC has never filed a tariff for cellular telephones.

[42] The ratio of 200 million to 2.9 billion is 6.9 per cent. But if the denominator did not include tracks copied onto the internal memory of cellular telephones, for example, the ratio would be considerably higher.

[43] The same is true for the second relative measure. CDs purchased by individuals include those used to store photographs, movies, music, and other content. About 33 per cent of CDs purchased by individuals are used to copy music. Since everything except music is outside the regime, a great portion of this relative measure depends on data from non-leviable file types.

[44] In the past, the Board has not made this distinction explicit. In Private Copying 2008-2009, the Board referred briefly to non-leviable media.

For the same reasons as in the past, copies made from paid downloads and promotional tracks obtained from the Internet should be removed from the total. In Private Copying IV, the Board set at 6 per cent the amount of copies from paid downloads and 3 per cent the amount of promotional tracks. This time, CPCC puts those numbers at 3 and 2 per cent respectively. These percentages become 3 and 3 per cent when only leviable media are considered. [37] [our emphasis]

[45] The last sentence of the quote is based on information filed by CPCC in response to the following question from the Board:

[TRANSLATION] The Circum report generally provides numbers on private copying activities not only on medium already subjected to a levy (cassettes, CDs and MiniDisc) but also on all other media (personal digital audio recorders, removable memory cards and DVDs).

Is it possible to recalculate for 2006-2007, only as it relates to copies made on medium subjected to a levy (cassettes, CDs and MiniDisc):

- the Tables 4.1, 4.3, 4.4, 4.5, 4.6, 4.10, 4.11, 4.12, 4.13; […] [38]

[46] In that matter, the Board did not use the revised version of Table 4.10 for the purpose of determining whether any media for which CPCC proposed a tariff qualify for one. However, by asking the above question, the Board signalled its willingness to differentiate between data relating to all media (and devices) and data relating only to leviable media.

[47] The argument for focusing on leviable media is stronger today than it was in 2008-2009. In the early 2000s, leviable media constituted almost everything onto which private copies were made; today, the reverse is true. In the present context, relative measures disturb the Board’s understanding of the private-copying landscape, due to the problems with inconsistency and unfairness explained above.

[48] Fundamentally, the issue with using relative measures comes from the wording of the Act. The definition of audio recording medium, based on which the Board is required to determine if a medium qualifies is:

audio recording medium means a recording medium, regardless of its material form, onto which a sound recording may be reproduced and that is of a kind ordinarily used by individual consumers for that purpose […] [39]

[49] The plain reading of this section is that the Board should focus on the medium itself. If the medium is used by individual consumers ordinarily to copy music, the medium qualifies, regardless of how much individual consumers copy music onto non-leviable media and devices.

B. The Forecasts by CPCC

[50] As mentioned in the previous section, we focus only on absolute measures of private copying. There are two such measures. First, there is the number of tracks copied onto blank CDs. Second, there is the number of blank CDs purchased by individuals for the purpose of copying music. [40]

[51] CPCC forecasts that 200.1 million tracks will be copied onto blank CDs in 2018 and 183.9 million tracks will be copied onto blank CDs in 2019. CPCC notes that the threshold for this variable is about 26 million tracks. It obtains this figure in two ways. First, it notes that the Board certified a tariff for audiocassettes when only 26.1 million tracks were copied. [41] Second, it notes that the Board certified a tariff for CDs when two million copies were made. In 1999, copying onto CDs constituted copying from one CD to another (digital downloads were not yet available for sale, and copying from other media to CDs was relatively difficult). On average, prerecorded CDs contained 13 tracks each, [42] implying that 2 million copies were copies of 26 million tracks. [43]

[52] In our view, the number of tracks forecast to be copied in 2018 and 2019 indicates that CDs qualify.

[53] CPCC forecasts the number of blank CDs purchased by individuals for the purpose of copying music to be 2.25 million in 2018 and 1.79 million in 2019. [44] By way of comparison, in Private Copying 1999-2000, the Board’s calculations implied that individual consumers purchased at least 1.05 million blank CDs in 1998 for the purpose of copying music. [45]

[54] Once again, the number of blank CDs forecast to be bought by individuals for the purpose of copying music indicates that CDs qualify.

[55] Based on these forecasts, CDs qualify as a medium ordinarily used for the purpose of copying music.

C. Forecast Methodology

[56] In Private Copying 2015-2016, the Board was presented with two forecast methodologies (as described above). The Board noted that “[in its] view, technique 2 is the better technique.” [46] It is noteworthy that the Board did not state that technique 2 was the best technique.

[57] It is time to revisit the question of forecast methodology, for several reasons. First, several years have passed since the Board chose technique 2. Second, the data for 2015-2016 have replaced the data for 2013-2014 as the most recent actual data available.

[58] In its implementation of technique 2 in the present proceeding, CPCC uses data from 2006-2007 onwards. This may be problematic. Beginning from that year, the number of tracks copied onto blank CDs is non-monotone. [47] It is far more difficult to perform time-series forecasts on non-monotone series than on monotone ones.

[59] There is a single year’s data which disturbs the monotonicity, namely 2009-2010. To be specific, while the number of tracks so copied decline continuously from 2006-2007, the number of tracks copied in 2008-2009 are 294.5 million and the number of tracks copied in 2009-2010 are 464.1 million. CPCC has not explained this anomalous year.

[60] In our view, beginning the forecast from the 2009-2010 year would solve this problem. This problem is also present for the number of blank CDs purchased by individuals in those years; furthermore, we feel it best to start all forecasts in 2009-2010 for the purpose of consistency.

[61] Both technique 1 and technique 2 [48] are based on percentage-change calculations and filling in missing data. But these techniques assume the growth rate stays the same each period. A regression model uses the data differently; it can allow the growth rate to speed up or slow down. (This is especially true if the regression line is curvy)

[62] These regression models not only produce forecasts, but a set of diagnostic data that tell the forecaster how good the forecast is. [49] They can also be used to test hypotheses about the trend the forecaster is trying to model (is the life of the CD ending, or is the use of CDs for private copying stabilizing).

[63] In respect of the number of tracks copied, we would use the curvy line (a better fit than the straight line); in respect of the number of blank CDs bought, we would use the straight line (the curvy line does not fit at all).

[64] Given the forecast specifications above, we would forecast 229.4 million tracks to be copied onto blank CDs in 2018 and 194.3 million tracks to be so copied in 2019. We would also forecast the number of blank CDs purchased for the purpose of copying music to be 6.1 million in 2018 and 4.13 million in 2019.

[65] Given these forecasts, as modified, we come to the same conclusion we would have come had we used CPCC’s forecasts as filed. CDs would qualify as a medium ordinarily used for the purpose of copying music.

D. The Issue of Data Quality

[66] Mr. Maguire first raised the issue of data quality in his statement of case. [50] He echoed this point in his response to CPCC’s answers to the Board’s questions. [51]

[67] CPCC’s reply [52] to the latter contained a quote from Private Copying 2005-2007, in which the Board wrote:

CPCC presented two surveys. The first is the Music Monitor Survey. CPCC has filed its equivalent since the very first private copying hearings. It has none of the characteristics that would lead a court, applying Mattel, to conclude that it should be excluded. The survey is prepared under the direction of a seasoned professional and has withstood the test of three cross-examinations. It is well designed and impartially administered. The questions it asks are directly relevant to understanding the private copying market in Canada; so are its findings. Its results are overall reliable and valid. [53]

[68] Mattel is a Supreme Court of Canada decision from 2006. It dealt with the likelihood of confusion between a famous trademark and another mark. Before the motion judge, the appellant had sought to introduce new evidence, in the form of a public opinion survey. Binnie J. stated that the survey was not responsive to the point at issue and thus not relevant. In addition, he stated that provided that the question asked was relevant, the results should also be reliable and valid. Mattel therefore addresses the question of when the use of survey data is appropriate.

[69] CPCC claims that “[t]his most recent iteration of the Music Monitor Survey was conducted using the same methods and the same care that have been recognized by the Copyright Board as reliable and valid.” [54]

[70] Just because the Board found, in Private Copying 2005-2007, that the Music Monitor Survey lead to reliable and valid results does not mean it would necessarily do so in all future private-copying proceedings. As an example, assessing the reliability and validity of a survey requires the examination of certain facts, which may not be the same every time the evidence is introduced in a new proceeding.

[71] Also, and of even greater concern, for the most recent data points, the Music Monitor Survey asked about usage of various private copying technologies during 2015-2016. In respect of the tariff covering 2018 and 2019, we are prepared to accept that those results are acceptable, albeit undesirable. In effect, the data is not as up to date as it could be. While there may be good reasons for this, there is nevertheless a possibility that in respect of a tariff covering 2020 and beyond, it may no longer be the case. Questions about usage of various private copying technologies in 2015-2016 may no longer be the right questions to ask.

VI. THE PRACTICAL-SENSE TEST

[72] Mr. Maguire noted that if the costs of the private copying regime exceed the revenues therefrom, maintaining the levy would make no practical sense. The Board has applied a version of the practical-sense test once before. In ERCC 2013, the Board noted that:

“[...] Costs have continued to exceed revenues, and debts have always largely exceeded any amount available to the collective. As a result, nothing has ever been distributed to rights holders [...] ERCC’s board of directors [...] has [...] voted [...] to dissolve ERCC and to write off [...] loans as well as any accumulated interest.” [55]

[73] It should be noted that the two situations are not at all comparable. ERCC, faced with no objectors, applied to the Board to vary its tariff by eliminating the final few years, so it could wind up its operations. In the present case, CPCC exists and continues to file proposed tariffs; Mr. Maguire as objector is asking the Board (implicitly) not to certify a tariff, on the grounds that costs may exceed revenues.

[74] It is also noteworthy that Mr. Maguire applies two traditional principles of economics. First, he notes that if CPCC’s costs exceed its revenues, this would justify not having a tariff. Second, he considers the costs of operating the entire regime (including those of CPCC, the tariff payers, and the Board). If this extended measure of costs exceeds revenues, this would justify not having a tariff.

[75] In effect, the first principle is private, microeconomic efficiency. The second is public, macroeconomic efficiency.

[76] In response to the first principle, CPCC argues that it is not a profit-maximizing entity. It is rather a non-profit umbrella organization that represents songwriters, recording artists, music publishers and record companies through its member collectives, which have mandated CPCC to collect and distribute private copying levies on their behalf. [56]

[77] In our view, whether or not CPCC maximizes profits is not the relevant consideration. Facing constant costs and declining revenues, CPCC may well come to the decision that filing a tariff with the Board is no longer worthwhile. That decision, however, is CPCC’s alone, and not the Board’s.

[78] In response to the second principle, CPCC claims that the private copying levy has a public purpose. [57] Presumably, the implication is that the public purpose overrides concerns of macroeconomic efficiency.

[79] In our view, costs to the Board should not be part of the consideration whether or not to certify a tariff since this would lead to an aberrant result. Otherwise, unopposed tariffs should always be certified since their costs are low. But the same logic would lead to the conclusion that the costliest tariffs (i.e., large tariffs with many objectors, that are sometimes sent back to the Board for redetermination from judicial reviews or appeals) should not be certified.

VII. ANALYSIS OF THE LEVY

[80] In Private Copying 2010, the Board set a levy of 29¢ using the Stohn-Audley model. [58] Since that time, the Board has issued four private-copying decisions, each without the benefit of a model (or a proxy), each setting the rate at 29¢ (2011, 2012-2014, 2015-2016, and 2017).

[81] The justification in each of these four decisions is approximately the same. As the Board noted last year, “the current levy is a reality in the marketplace.” [59]

[82] One aspect of being a reality in the marketplace is that the price of the levy is expected and is passed on, in whole or in part, to consumers. The measurement of pass-through is difficult in the present context, since we do not have data on wholesale prices. However, so long as the retail price exceeds 29¢, we cannot preclude a non-zero pass-through of the levy.

[83] Ms. Gelbloom filed evidence of retail prices, consisting of a survey of retail establishments conducted in Toronto in June of each year. The survey looked at the prices for three items: a single CD sold with a jewel case, a spindle of 50 CDs, and a spindle of 100 CDs.

[84] We ignore the pricing information for the single CD sold with the jewel case. The jewel case is a piece of molded plastic that is convenient for storage. The single CD sold with the jewel case is a bundle of two products – the blank CD and the jewel case. We do not know the value of the jewel case that is sold with the blank CD, even though it is possible to purchase jewel cases separately. Jewel cases come in several levels of quality, each with their own associated price.

[85] Over the years, there is sometimes a difference between the per-unit price of CDs when sold in a spindle of 50 and when sold in a spindle of 100. However, these differences are smaller than one standard error of the sample. [60] In short, these differences can be attributed to sampling variation of the survey.

[86] The data (averaged as between the per-unit price for spindles of 50 and spindles of 100) are filed for the years 2009 through 2017. The price ranges from a low of 29.3¢ (2014) to a high of 52.8¢ (2017). The time series of the data is decreasing from 2009 to 2014 and is essentially flat thereafter (about 51¢).

[87] Four points are noteworthy about these data. First, all of these average prices exceed, if weakly, the levy at 29¢. Second, for 2017, we cannot reject the hypothesis that the levy is being fully passed through to retail prices, using a standard, 95-per-cent confidence level. Third, the fact that the price is higher in the later period is consistent with the theories advanced by Dr. Boyer about life-cycle pricing. [61]

[88] Finally, the fact that the price of blank CDs is higher in the later period is consistent with the Board’s finding that a differently calculated levy would almost certainly lead to a levy higher than 29¢. [62]

[89] We rely on three reasons to set the levy at 29¢. First, as explained above, the price of 29¢ is a “reality in the marketplace.” Second, as shown by the analysis of the survey data, the price of blank CDs is consistent with full pass-through of the levy. Third, since a levy calculated by other means would almost certainly be higher, 29¢ is a lower bound. [63]

VIII. THE TARIFF

[90] In accordance with the above reasons, we certify a tariff of 29¢ per blank CD.

[91] We retain the existing apportionment among collective societies: 58.2 per cent to authors, 23.8 per cent to performers and 18 per cent to makers.

Gilles McDougall

Secretary General