«Jonathan Haskel Imperial College Business School, CEPR, Ceriba, IZA and UK-IRC Keywords: copyright, IP, innovation, knowledge, investment, ...»
As we have outlined in our discussion of the industry structure, the innovative sector in film actually consists of two industries: production and distribution. In order to derive an estimate of the upstream, it is certainly not correct to simply sum them together, since the revenues of the production industry are largely made up of the production fees received from the distributor. Therefore we shall consider just the distribution industry, since that is where the majority of asset rights are held. Note that output in the distribution industry is partly made up of the same costs that exist in the production industry (paid for by the budget provided by the studio), plus additional costs for advertising, marketing and other overheads, plus additional revenues that are due to the innovators market power as the owner of a unique asset.
In calculating a very rough proxy for GFCF based on industry aggregates we first take ABI 53 data on the gross output of ‘Motion Picture Distribution’ (SIC 92.12). We then adjust this using data on international trade, sourced from the FTV survey. 54 That is, we add on imports and subtract exports (for each we use the sum of production fees and royalties). Note it is not correct to treat the data on exports and imports in this way since production fees and royalties ought to be treated differently. Within royalties there are data on asset purchases and genuine rentals, but we have little information on the proportional split. Since we are using the data for the Distribution industry, we also subtract intermediate consumption, which will also include other forms of knowledge investments already capitalised in our wider intangibles framework, as well as additional costs and overheads. Finally we subtract the change in inventories or stocks. Since the data are based on revenues, any estimate will implicitly include a mark-up, where μ1.
Figure A2.3: Derived GFCF in Film Originals, CP £m (PnN)
For years prior to the introduction of the ABI (pre-1995), data are from the Service Trades Sector Review Trade receipts and payments split into fees for production and royalties are only available from 1998.
Therefore we used the average share in 1998-2000 to backcast each series.
-500 Source: Derived from macro aggregates from the ABI, Service Trades Sector Review, and FTV Survey.
An interesting feature of the series is the strong growth from 2004 that is sustained until 2006.
Unfortunately the ONS are unable to provide us with any information on a breakdown of industry revenues to allow us to better understand the source of this, but we note that data from the UK Film Council, shown in Figure 16, also show a spike in UK-certified productions in 2004-5.
To re-iterate, there are a number of issues with the underlying data:
- the FTV trade data is a mix of outright fees, asset transactions, and rentals. A split is available for both 2005 and 2006, but the composition of payments differs greatly in each year and so does not provide suitable information to use in any assumption to split the data
- we have not been able to find suitable data on production subsidies or tax breaks to subtract from industry output. This could be significant in the case of Film. National Accounts series are for taxes less subsidies at the aggregate level and the underlying component series do not appear to be available.
- the output data implicitly includes capital compensation, which should be adjusted for consumption of fixed capital (depreciation)
- it is possible that there is some element of double-counting, for instance as one example, ‘special effects output’ may already be counted in software GFCF Due to these data issues, we consider this series to be an inaccurate approximation of GFCF in Film Originals, and for both conceptual and practical reasons do not consider it suitable to present as a final estimate in the main body of this report. We re-iterate that there are many practical and conceptual barriers to accurately estimating GFCF in this way. Therefore data are presented largely to just illustrate what industry data are available and we do not attach a great deal of weight to the result.
2. Film: Estimates of Aggregate Output according to the Input-Output tables and DCMS As discussed in the main report, one potential way of estimating the output of the upstream is to collate measures of aggregate sector output and estimate a factor which would allow the data to be split into components for the innovative sector and the final goods sector.
Data for aggregate film output, P(G)G, are available in an ONS analysis of the creative sector based on the I-O tables (Mahajan, 2006). As we have already stated, there are serious drawbacks in the use of such data to value ‘creative output’ and it is even less useful when estimating investment in artistic assets. Additionally note that the output measures presented here are gross value-added.
Figure A2.4: Output of the ‘Film Industry’, P(G)G, CP GVA (£mns) 4,000 3,500 3,000 2,500
1,500 1,000 Source: UK I-O Analyses, Creative Sector 1992-2004, and DCMS estimates of the creative sector Note to figure: For the ONS data, the film industry is defined as: i) Motion picture and video production ii) Motion picture and video distribution iii) Motion picture and video projection iv) Reproduction of video recording v) Reproduction of computer media.
The data show that, according to the ONS, in 2004 ‘Films’ represented £3.3 billion of output (GVA), but we re-iterate that this also includes parts of the final goods, or downstream, sector, such as projection and video re-production. The inclusion of both production and distribution also means that there is some double-counting of the upstream. It also, of course, includes the production costs and revenues for assets that are not UK-owned (exports). This is the reason why we, in concordance with Eurostat and OECD, advocate an approach based on the input costs of the innovators.
The DCMS recognise that not all output in a ‘creative industry’ is necessarily ‘creative output’, and so exclude a proportion of output considered non-creative. Therefore the estimates produced by DCMS are lower than those from ONS although the exclusion is not sufficient, and nor does it attempt, to remove all downstream activity. Since DCMS estimates for film also include photography, they are not presented here due to differences in classification. Provided the output of the photography industry was removed from the data, it would be possible to attempt to remove the remainder of downstream film output, using some factor. However, with no information on the size of such a factor, and bearing in mind that a reasonable estimation would pose many problems already discussed with regard to identification of products and asset ownership, we are unable to apply this approach to Film.
3. TV & Radio: A proxy of GFCF based on industry aggregates As noted in the main body of the report, there are number of limiting factors to the aggregate data. The following chart presents data for the television industry largely for illustrative purposes to show what data is available. Adjusting ABI data on the output and inputs of the Television and Radio Industry using trade data from the FTV, results in an estimate of GFCF in TV & Radio Originals of £1,991m in 2007, compared to the estimate of £2,632 currently recorded in the National Accounts.
Figure A2.5: Derived GFCF in TV & Radio Originals, CP £m (PnN) 20,000
4. TV & Radio: Estimates of Aggregate Output according to the Input-Output tables and DCMS Data published in the I-O analysis of the creative sector (Mahajan, 2006) shows that in 2004, GVA in the radio and television industries was approximately £8.9bn. Such measures will be familiar to those interested in measures of ‘creative output’ but for reasons already explained, such measures are not suitable for use as measures of investment due to inherent doublecounting.
Figure A2.6: Output of the ‘TV & Radio Industry’, CP GVA (£mns) 10,000 9,000 8,000 7,000
3,000 2,000 1,000 Source: UK Input-Output Analyses, Creative Sector 1992-2004 The DCMS correctly recognise that a measure of the output of the television and radio industries is not a valid measure of ‘creative output’ in those industries. Therefore they adjust the output data to remove what they consider to be non-creative output, resulting in estimates presented below.
Figure A2.7: α.
P(G)G: ‘TV & Radio Industry’, DCMS (£mns) With no information on the construction of α and the adjustments used by DCMS we are unable to judge the quality of the results. It should be noted that to fit into our framework, any estimate of α should be done separately for the TV and Radio industries. In the case of Radio, the revenues of most radio stations would be generated from the rental of music originals. That is, radio stations are also a significant downstream user of the upstream music sector. Additionally, the activity of presenters on these stations does not meet the capitalisation criteria, and the same is true of talk radio stations. Therefore the α for the radio industry would likely be extremely small, whilst that for the television industry would be larger. Also note that the conceptual reasoning behind the adjustment factor used by DCMS, and the factor in our framework, is different. DCMS is not attempting to derive an estimate of ‘upstream output’ in its analyses.
5. Books: Estimates of Aggregate Output according to the Input-Output tables and DCMS Data published in the I-O analysis of the creative sector (Mahajan, 2006) shows that the Publishing industry recorded output of approximately £10bn in 2004, although the industry definition used includes a range of activity that is not asset creation, including the publication of goods with short services lives such as news. Additionally the data refers to both the innovative and downstream sectors.
Figure 37: Output of the ‘Publishing Industry’, CP (£mns) 12,000 10,000 8,000
4,000 2,000 Source: UK Input-Output Analyses, Creative Sector 1992-2004 Note to figure: Data actually include all of Publishing, defined as: i) Publishing of books ii) Publishing of newspapers iii) Publishing of journals and periodicals iv) Other publishing v) News agency activities Adjusted estimates produced by DCMS are very similar to those in the I-O analysis, implying a very high value for α, although as mentioned the adjustment made by the DCMS is not to back out the value of upstream output.
6. Music: Derived GFCF in Music Originals, CP £m, P(N)N The following chart adjusts ABI output for the music industry using international trade in royalty payments in the UK Balance of Payments. We use data on exports and imports of ‘Other Audio-visual fees’ which we assume to be largely made up of payments for the rights to use UK music originals. 55 Since the sound publishing industry also includes a significant part of the using sector, we assume intermediate consumption refers to the downstream and subtract it from GFCF. For the same reasons discussed in the context of Film and Television, these estimates have been produced largely for illustrative purposes and we do not consider them adequate to measure investment in artistic originals.
Figure A2.9: Derived GFCF in Music Originals P(N)N, (CP £mns), ABI Audio-visual fees will also apply to other asset categories such as literary works but we have no information on the types of assets to which the fees apply.
7. Music: Estimates of Aggregate Output according to the Input-Output tables and DCMS Data published in the I-O analysis of the creative sector (Mahajan, 2006) do not present datasolely for Music. Instead the data are combined with Photography and other forms of artistic creation termed ‘The Arts’. Therefore the data includes elements of both the upstream and downstream of music creation, as well as data for the upstream and downstream of other asset categories, and production of goods outside the definition of artistic originals.
The DCMS also publish a series on the output of music industries, adjusted to remove what they consider to be non-creative output. However, as with the I-O analysis, the industry is defined more broadly than just music to include the ‘visual and performing arts’.
8. Miscellaneous Art: Estimates of Aggregate Output according to the Input-Output tables and DCMS Since our definition for this category is a diverse group of specific assets, an estimate of aggregate industry output from either the ONS I-O or the DCMS analyses is unavailable to us. In the case of the I-O analysis, photography and ‘the arts’ are aggregated into the output measure for ‘Music and The Arts’, whilst in DCMS publications, photography is aggregated into ‘Video, Film and Photography’ whilst the ‘Miscellaneous’ category is largely made up of data for antiques. Output data are available from the ABI however, and have already been presented in Figure 34.