«Jonathan Haskel Imperial College Business School, CEPR, Ceriba, IZA and UK-IRC Keywords: copyright, IP, innovation, knowledge, investment, ...»
revenues earned by UK artists. Since the UK is considered a net exporter in music, it is likely that data adjusted for international trade would result in higher estimates.
Figure 21: Breakdown of capital income from music originals, P(R)R, 2009
Source: PRS for music Note to figure: *adjusted for potential double-counting, overseas payments and collecting society commissions; **may conceptually be more appropriate to treat as revenue from branded capital, rather than from artistic originals.
Therefore we can build on ONS estimates by including the additional capital compensation earned by UK music originals, including those received by the upstream partners of musicians, namely record labels and publishers. Let us apply the relationship between P(N)N and ΣP(R)R, as set out above. In 2009, aggregating across revenues distributed by PRS and PPL for mechanical rights, performance rights as well as licensing and synchronisation revenues, with each adjusted to allow for payments sent to overseas artists and commissions taken by the relevant Societies, the total cross-section of royalty payments was approximately £650m. Our estimate of the growth rate is based on our ASHE series (1997-2008), at 3%.
We estimate a geometric rate of depreciation rate based on Soloveichik (2010) and the rate used in the ONS VICS at 13.33% (Long, Turvey and Wallis, 2010). For the real rate of return we use an estimate of 10%. Application of this method yields an estimate of investment of £931m.
We consider this estimate to represent a lower bound because the data do not fully account for all the compensation earned by artistic originals. First they do not fully account for revenues earned from sales of recordings, since PRS payments for mechanical rights only go to Songwriters. With the assistance of PRS we have derived an estimate of total compensation for mechanical rights and adjusted our final estimate accordingly. This income is split between the record label, artists, songwriters and publisher who each own a share of some of the rights. Further income is earned from live performance, with a typical industry arrangement being that the performers earn some percentage or all of ticket revenues, with the booking fee used to pay for remaining inputs. All of this income accrues to the artists and songwriters.
Including estimates for recorded sales and live performance in our estimate of total capital compensation yields an estimate of ΣP(R)R of £1,835m, and P(N)N of £2,701m. We consider these data our central estimate using this method. It is worth noting that conceptually the income earned from live performance is probably closer to ‘Mixed Income’ rather than pure capital compensation. That is, it includes a return for both capital and labour.
Splitting the figure using an approximate ratio of ‘Compensation of Employees’ to ‘Gross Operating Surplus’ would result in an estimate of approximately £1,487m for ΣP(R)R and £2,189 for P(N)N. However, in the National Accounts, the convention is to include Mixed Income within Gross Operating Surplus.
Finally we consider a higher bound of investment using this method. As we have already discussed, the upstream in music includes a number of agents with some share of ownership to (some) asset rights. These include the record label, artists, songwriters and publishers. The investment made by the record label is in the form of funding asset creation (through an advance) and building the reputation or brand of the artists and songwriter. The created brand is also used to generate upstream revenues which we consider here in our higher bound. 43 Including such revenues in the valuation of investment yields an estimate of £2,833m for P(N)N. Although we would argue that such revenues are a return to an asset created in the upstream music sector, brands are not considered to be assets in the SNA framework. For this reason we count them in our higher bound but exclude them from the central estimate, which is designed to be consistent with the SNA.
Note, we are referring to payments for the use of the brand or image of the artists, and not to payments for the use of music in advertisements which are already counted.
Figure 22: Approximation based on cross-sectional aggregate of Music Royalties 3,500 3,000 2,500 2,000 Source: own calculations.
Of course this calculation has relied heavily on the strong assumptions we have made by invoking steady-state conditions. It is also worth noting that data on sales royalties are based on UK sales or recordings; ideally they would be based on worldwide sales of UK recordings but such data are not available. Since the UK is one of the few countries that is a net exporter of music, we do not feel that the use of UK sales introduces any upward bias to the estimate.
Therefore although figures derived from this method should not be seen as robust estimates of investment in music originals, they are useful for indicating the scale of investment that is potentially missing in the official data, and for supporting our view that the ASHE does not provide a representative estimate of the value of asset creation by musicians. A more informed estimate requires the use of royalties paid to individual assets over their lifetime in a longitudinal study. We hope that such data will be available in the near future
11.3. P(N)N, Upstream Output. Music The industry classification for music industry does not easily fit the upstream/downstream
distinction in our framework. The music industry is essentially made up of four activities:
2) Songwriting and publishers
3) Live performance
4) Artist management As usual the industry structure is complicated. Whilst initially it may seem that the innovators, located in the upstream, are the artists and songwriters, the important issue is who owns the final asset. In practice, ownership is split between the artist(s), record label, songwriter(s) and publisher, but the split of IPRs varies in individual agreements and by specific right (e.g. mechanical, performance, synchronisation etc.). In general, it seems that for less well-known acts, it is the record label that owns the majority of the asset rights, whilst larger acts retain a higher proportion due to their greater market power.
Figure 29 presents data for ‘Publishing of Sound Recordings’ (SIC 22.14), ‘Re-production of sound recording’ (SIC 22.31) and a combined aggregate for the two. We use the latter as our definition of the music industry although note that numerous musicians and artists will be recorded in Artistic and Literary Creation (SIC 92.31) alongside writers, authors and other creators of artistic assets.
Figure 23: Output of the ‘Music industries’, (CP £mns), ABI 1,400 1,200
Source: ABI published aggregates
11.4. P(Y)Y, Downstream Output. Music Estimating a series for the downstream revenues for users of music originals would require information on the revenues of all users of UK music IP, that is those that rent from the capital stock of UK final originals, such as the television and radio stations/networks, merchandisers and organisers of live performances, including all those users that are located outside the UK. Additionally, this should include implied rentals of UK recording companies that are renting from their own in-house stock. Such detailed data are not available, however we do have industry data on sales from the British Phonographic Industry (BPI), a trade association for the British record industry, giving us an incomplete measure of P(Y)Y, as shown below in Figure 30.
Figure 24: Record Industry Revenues, BPI, CP (£mns) Source: BPI Statistical Handbook, 2010, collected in BPI surveys. Values are wholesale and so do not include VAT, and are net of returns. In this figure, other is defined to include Mobile income which includes ringtones, ringback tunes and music videos, Subscriptions, Ad-supported, and Other Digital Musical Content. Mobile Income prior to 2008 is included in the data for Online income.
As we have discussed, the official method for estimating asset creation in Music is to apply a factor (λ) to final downstream output and thereby estimate the component that represent royalty payments, leaving just that income that accrues to the owners of artistic assets. This is the basis of the official method used by the ONS, where a λ is applied to recording sales.
Whilst conceptually annual capital income is not the same as investment, we have also shown that by invoking golden rule steady-state conditions and some strong assumptions on lifelength and the mix of vintages in the capital stock, a measure of λ.P(Y)Y can be seen as a proxy for investment.
The data presented in Figure 31 below compare the two alternative sales series from the British Phonographic Industry (BPI) and the ABI respectively.
Figure 25: Alternative series for Gross Output (Sales), Recordings, CP (£m) Source: BPI data taken from their Statistical Handbook, with gross output defined as total income from sales of albums, singles, tracks, DVDs, subscriptions and other content, sold digitally or as physical formats. ABI data are for ‘Publishing of sound recordings’ and ‘Re-productions of sound recordings’, as defined by the SIC, combined.
There are two ways of interpreting the method of using a constant percentage of sales. The first is that the ONS is seeking to split output into respective components for the upstream (P(N)N) and downstream (P(Y)Y). The second is that the percentage is based on the royalty rate earned by the upstream, and that rate is applied to sales to derive some estimate of the annual capital income generated by music originals.
The method raises a number of issues. First, whilst application of the factor may generate an estimate of the annual cross-sectional sum of royalties, that is not the same as a discounted sum of lifetime revenues which accurately takes account of the vintage of assets in the UK stock. Second, the sales figure to which it is applied includes the sales of copies of non-UK assets. A more appropriate series would include sales of copies of all UK-owned originals sold worldwide, and exclude sales of copies of originals owned in the ‘Rest of the World’.
This is important, especially if the UK has a positive trade balance in this industry. Third, the sale of copies is just one the downstream activities that use music originals. Others include live performance, merchandise, TV or Radio broadcast, sheet music and advertising. Fourth, the factor should take account of all royalties earned from music originals, by all (co-)owners in the upstream, that is, artists, songwriters, publishers and record labels, not just the earnings of the songwriter for mechanical rights.
A potential way of improving the estimates in the National Accounts but maintaining a
similar method would involve:
- replacing imputed estimates for sales with BPI data
- adjusting the royalty rate to account for the capital income earned by record labels, artists and publishers, since there is effectively joint ownership of the music original.
- estimating the implicit capital income earned from live performance, using a series on ticket revenues and a factor, θ
- an adjustment for net exports Through our work with ‘PRS for music’ we have access to detailed data on downstream sales of recordings and live performance, and various margins and components that apply to those revenues. To summarise, we estimate an improved measure of λ=0.57, which implicitly includes the capital compensation earned by artists, songwriters, publishers and record labels.
The high value of λ is a reflection of low share for intermediates, partly due to technical progress. We also estimate a provisional value of θ=0.4 which we apply to live performance revenues 44.
The following chart applies this method and compares the ONS series with some alternatives that use improved estimates for component series’. The middle line is the ONS estimate of investment.. The lower red line uses the same royalty rate, but an estimate of the trade value of sales as published by the BPI. The dotted red line removes various margins and costs from the BPI trade value to derive and estimate of λ=0.57. It also adds in an estimate of the revenue earned from live performance, where θ=0.4.
Figure 26: Alternative measures for GFCF when applying ONS method, CP (£m)
We are grateful to Will Page and Chris Carey of PRS for providing us with these data. θ=0.4 is calculated by estimating the proportion of revenues that flow to artists from live performance. The aggregate figure for live revenues includes revenues from ‘at the event spend’ and all margins, which are removed. The remaining figure our estimate of capital compensation, which we use to derive θ.
For recorded sales, the process is similar, with revenues adjusted to remove components such as retailer margins and costs for manufacture, distribution and marketing.
Source: Sales (BPI), Live revenues (PRS), ONS sales (ONS) ONS λ (ONS), Alt λ (our estimate based on BPI and PRS data), θ (our estimate based on PRS data)
12. Miscellaneous Artwork From our discussion of the Eurostat Taskforce report and their criteria for inclusion, as well as the development work by Soloveichik of the BEA, we know that any original that is covered by copyright, can be commercialised, has a service life of greater than one year and is not already recorded as an asset elsewhere in the National Accounts, can be counted as GFCF in artistic originals. Outside of the four assets already discussed, no other investments in artistic assets are used in the calculation of GFCF in the UK Accounts. This leaves a wide range of potential candidates for inclusion, such as photography/images, artwork, choreography and cartography, although in the case of the latter, it may be that these are already at least partly included in estimates for literary originals. The difficulty when considering such asset types, where less data are typically available, is to ensure that what is being counted is the production of assets rather than intermediate goods.