Technology and the Recorded Music Industry sets the stage for this article. It records the development of a truly global music industry from its 19th century origins and how the industry during its first one-and-a-quarter century was based entirely on the technological development of physical recordings and the appropriate equipment to play them.
Two major innovations which both became commercial in 1995 were the World Wide Web and the MP3 encoding format for digital audio which quickly became the common format for consumer audio storage. This led to the advent of large-scale commercial digital file-sharing at the end of the 20th century. The 2013 annual report on digital recording by the International Federation of the Phonographic Industry (IFPI) is an important source on the position of the global industry at the end of 2012.
The advent of digital file-sharing radically changed the industry, providing it with opportunities for new product development at the expense of physical recordings with which the digital products competed effectively, especially on price, ease of access and diversity; but it also created a monumental problem as the opportunities for unlicensed operations escalated. The markets for physical recordings — albums, singles, and audiovisual products (music videos/DVDs) — suffered rapid decline especially between 2005 and 2010 as they lost the competition with digital products (the major producers of physical recordings themselves introduced digital products to stay in the race). Though there are signs that sales of physical products may be stabilising, the music industry and its organisations remain understandably concerned about piracy and copyright issues remain major. Australia is no exception.
In summary, it is no exaggeration that the industry in the first decade of the 21st century underwent what the Austrian-American economist Joseph Schumpeter termed “creative destruction” in the sense that new product technology to a large extent replaced the old range, leading to major structural change in the music industry.
This article covers the Australian recording industry, which is at the centre of what is commonly called the music industry. The main sales statistics are published by the Australian Recording Industry Association (ARIA), which is one of the two national recording industry associations. The other is the Australian Independent Record Labels Association (AIR). Both associations have a mixed membership ranging from small labels to medium-sized and large companies (as well as others who don't actually produce recordings, like the Australian Music Centre, Opera Australia, and distributors). ARIA was established by the then (1983) six major record companies in Australia. In contrast, none of the current "Big Three" (Sony Music Entertainment, Universal Music Group and Warner Music Group) are or have been members of AIR, which describes its membership as "artists, labels and distributors across the full spectrum of music genres, ranging from small sole traders to some of the biggest independent operations in the country."
Digital recording is well represented throughout both organisations, as the membership lists show. This goes for the worldwide majors too. For example, the US-based enterprise known as Universal Music Group Distribution covers "sales, marketing and distribution of Universal Music Group assets in all formats including digital, physical, streaming, subscription and mobile". Vevo, co-founded by Universal and touted to be the leading premium music video and entertainment platform, is available in many countries including Australia "through VEVO.com, the mobile web, Mobile and Tablet apps, connected television services, and user embeddable video players." At Warner Music Digital, to take another example, digital sales "grew to 38% of global recorded music revenue in fiscal 2012, up from virtually nothing in 2004."
Australia is clearly part of the global industry to which these examples apply. This article is designed to describe the industry, with an emphasis on available statistics, in terms of wholesale sales (the most comprehensive data). Other published statistics are very limited in scope as already established in the statistical roundup in Overview of Music Statistics: ABS and other articles in that series. The Australian Bureau of Statistics has published data on Internet activity since 1999 which do not specify music but sets the stage for the rapid progress made by the digital side of the music recording industry. These statistics have their own section in this article.
Table 1 presents a comprehensive summary of the ARIA statistics. The sales values are in the top panel (not in constant prices), sales in units in the middle panel, and average unit values in the bottom panel (dollars per unit sold). The available statistics in the left half of Table 1 show total physical sales in three main groups (singles, albums, and audiovisuals dominated by DVDs). The right half shows digital sales split into tracks, albums and other, the last classification including mobile master ringtones, digital mobile video, mobile ringback tunes, streams, advertising-supported income, unearned advances and one-off payments.
On the physical side, albums increasingly dominated falling total sales through the period shown in Table 1. In 2003, when digital sales were negligible, albums accounted for $545 million of total physical sales of $646 million (84.3%); in 2012 when physical sales had fallen to $214 million (-67%), albums accounted for 91.7%, increasing their share of a rapidly falling market. The decline in album sales over the nine years was 64%, compared with 97% for singles and 76% for DVDs.
Digital sales grew from virtually nothing in 2003 (no sales were recorded though some small quantities must have been sold and use of the Internet had started its growth towards prominence) to $184 million in 2012, 46.3% of the grand total. In 2005, the first year for which these statistics were recorded, single tracks accounted for 31% of total digital sales, albums for 12%, and DVDs/music videos for 57%. In 2012, the share of tracks had increased to 53%, albums to 34%, at the expense of DVDs which fell to 12% of the total despite a hefty increase in their total sales value.
Sales of "other" digital products are not known in unit terms except for mobile master ringtones, but both tracks and albums increased fast with an edge to albums until 2012. In annual percentage growth terms albums outpaced tracks from 2007 to 2011 but the volume of the latter shot ahead in 2012, though at a lower unit value as shown below.
Unit values would normally provide some clues to the real value of sales, assuming that the composition of the individual product groups remain reasonably unaltered over the years. The main physical item, albums, showed a fairly consistent decline in unit value from more than $10 in 2003 and 2004 to $7.11 in 2012. Compared with a fall in the total nominal value of sales from $545 million in 2003 to $195 million in 2012 (-64%), sales in unit terms fell from 51.2 million to 27.5 million (-46%), and unit values by 33%. The deterioration in sales, in other words, was partly due to declining prices, so it might be argued that the "real" sales value for physical albums declined not by 64% but by "only" 46%.
This is not particularly helpful for an industry undergoing the degree of structural change that the recording industry is experiencing, with every physical single, album or music video "replaced" by digital units at less cost — though this statement greatly oversimplifies a much more complex situation as should be evident from just looking at the physical and digital "units" counted in Table 1, and bearing in mind that "other" digital services include subscription and other services where downloads are not paid for individually. A CNet article dated 25 March 2013 by Lexy Sawides ("Australian music-streaming services compared") compares 15 music-streaming services available in Australia at the time. These services are typically priced at $6.99 to $7.99 per month for a basic subscription and $11.99 to $12.99 for "premium" services including mobile use and other additional features.
The demand pattern is changing so profoundly that it might be more helpful to relate sales of physical and digital products, respectively, to a general indicator of price change to get closer to an idea of the price-related competitive position of each. There is work to be done to clarify this.
Charts 1 to 3 illustrate the changes. Chart 1 traces the wholesale value of recordings from 2003 to 2012. The total black line declined from $528 million in 2005 to just under $400 million, stabilising from 2010 and increasing in 2012 (there was one prior rise in 2009 but then a 14% decline in 2010 associated with a 24% fall in physical sales). The total market, however, had already deteriorated from $646 million in 2003 (assuming negligible digital sales that year) to the $528 million in 2005. Chart 1 illustrates that physical sales have continued to decline, but the strong rise in digital sales more than compensated for that in 2012.
Chart 2 shows that in unit terms, the number of physical recordings declined fairly steadily from about 66 million in 2003 to 30 million in 2012, while sales of digital tracks increased from 2.4 million in 2005 to 110 million in 2012, and digital albums from less than 100,000 in 2005 to 6.8 million in 2012.
Unit values for physical albums declined by 33% between 2005 and 2012 as already noted (illustrated by Chart 3). On the digital side, unit values also declined but to a lesser extent than for physical sales. The fall in the unit value of digital singles in 2012 is noted, from $1.16 to 89 cents (-23%), coinciding with a strong rise of 61% in the number of units sold. No published explanation has been found for this; it is perhaps associated with changing marketing strategies well-known inside the industry.
In conclusion, there is little doubt that the price reductions for physical products were brought about by competition from digital products. The unit values of the latter fell to a lesser extent, as further technological advances in the digital area reduced production costs, and the structure kept changing as the range of services widened.
This overview is intended to give readers a general impression of these issues. For the full authoritative version, please go to the ARIA website.
The Copyright Amendment Act 2006 contains a number of important changes to copyright law relating to, inter alia:
The proposed levy is in effect an additional fee added to the price of various forms of blank recording media such as recordable CDs which would be paid to copyright owners as compensation for copying copyright material for private and domestic purposes. The supporters of the proposal maintain that there is little that can be done to prevent such copying so a compulsory levy seems the only effective solution.
The international recording industry opposes the measure. Apart from considering it generally impractical, the industry’s view is that copyright policy makers should focus their efforts on finding ways to improve copyright laws to deal with the challenge presented by the new technologies.
The original article of which the above is a summary was created by ARIA with cooperation from IFPI.
ARIA quotes a research study by Quantum Market Research, dated 2003, which showed that the number of people engaged in these practices was already substantial, especially among the younger age groups. For example:
In 2003, total sales of physical recordings were still high (refer Table 1), compared with today. It is difficult to establish whether illegal file-sharing has become significantly more prevalent and widespread (absolutely and relative to licensed services), or whether the greater sophistication of legal digital products coupled with the music industry’s own sustained efforts have had a significant effect on curbing the spread of illegal file-sharing. The growing demand for licensed digital services would itself have led to considerable change in the total income of the industry.
There is insufficient information on how the Internet is used, and the interactions between different uses. The most comprehensive collection of survey data is compiled by the Pew Internet and American Life Project, undertaken by the highly respected Pew Research Center, which described itself as "nonprofit, nonpartisan "fact tank" that provides information on the issues, attitudes and trends shaping America and the world. The Project studies the social impacts of the internet."
The right-hand box lists a range of surveys undertaken by the project, and shows the multitudinous uses to which the Internet is put. Two items are directly relevant to the issues facing the recording industry: downloading music files (37% of adult users in 2007) and downloading or sharing files using peer-to-peer file-sharing networks (15% in 2006). The proportions of the adult population (18+) are not as high as for many other uses — they are ranked 27th and 44th, respectively, in the list of 50 activities. However, these particular surveys are relatively dated (2007 and 2006), compared with those in the 15 top positions, and the US is likely to have seen a similar explosive growth in the amount of downloading as Australia (see Chart 7 below). Furthermore, even 15% P2P file-sharers is a formidable total number.
It is not surprising, of course, that the largest numbers of people use the Internet for searching, emails, exploring their hobbies, and getting driving directions and the latest weather forecasts. The proportion of the total population exploring the net for purposes of downloading or sharing files is bound to be smaller.
The proportion of Internet users in the US in 2012 was 81% (80% for males, 82% for females). The percentage fell with increasing age: 94% of 18-29 year olds, 89% for 30-49 year olds, 77% for the 50-64 group and 54% of people aged 65 or more.
The Australian Bureau of Statistics conducted a survey as part of its 2010-11 Multipurpose Household Survey (Table 2). It is not specific to music but gives a good view of the behaviour across the population of who downloads videos, movies or music (or some or all of these), and who listens to music or watches videos or movies online.
Downloading was practised by 37% of all survey respondents, 40% of males and 34% of females. The activity fell strongly with age from 65% of the 15-17 year group and 60% of 18-24 year olds to less than one-quarter of people aged 45-54 years and less than one-tenth of the oldest group.
The rate of downloading was highest for people with a university education and lower for people with lesser qualifications, with the seeming exception that a greater proportion of persons educated to Year 12 or less were more likely to download than people who attained a diploma or certificate. The group, however, would include 15-17 year olds who have not yet graduated from secondary school but have the highest download rates of all. Apart from that, level of educational attainment is at least to some extent a proxy for household and personal income — the survey found that download rates were lower for lower income groups. In the highest quintile (fifth) of household income, 42% downloaded, compared with only 32% in the lowest quintile. It was 42-43% for people with a personal income of $80,000 or more and 38% for those earning less.
Table 2 does not include geography for reasons of space. There were some minor differences among the six states and the Northern Territory with between 35% and 38% downloading, but the Australian Capital Territory stood out with 44%. Some of the differences may be due to different regional patterns: 39% of 9.6 million inhabitants of major cities were downloaders, compared with 32% in "inner" and "outer" regional areas (3.7 million), and 27% in remote Australia (163,000 inhabitants).
Similar patterns were found in the indicator shown in the right-hand column of Table 2 of persons listening to music or watching movies and videos online, except the participation rate was higher (overall average 53%).
The Australian Bureau of Statistics began to chart the growth in Internet usage in September Quarter 2000, when it ran its first Internet Activity Survey. Although there have been revised definitions along the way, which is less than surprising considering the unpredictable animal the ABS was trying to capture, and the timing of the regular surveys was changed as well, it has been possible to build up a number of time series for the period since 2000 (with judicious use of extrapolations and interpolations to arrive at a mid-year (June 30 or June Quarter) estimate for each year.
Charts 4 to 6 show the estimated number of subscribers to Internet service providers (ISPs) based on the obligatory reports that Australian ISPs have to submit to the Telecommunications Industry Ombudsman.
The charts cover estimates for all ISPs. This was the basis for collection up to 2005 and for some subsequent years when the ABS conducts a census of all ISPs (the next census is due in June 2013). However, by 2005 the small ISPs, with 1,000 subscribers or less and excluded from the regular survey on those grounds, had already become a tiny part of the total.
The estimated total number of Internet subscribers in Australia increased from an estimated 3.8 million in June 2000 to 12.1 million in 2012 — more than tripling in 12 years (Chart 4). In 2003, when these numbers were first collected, almost 90% of connections were dial-up, but broadband first exceeded dial-up in 2006 and by 2012 the overwhelming proportion of the more than 12 million Internet connections were broadband (96%).
The original type of broadband was DSL (digital subscriber line, which can be asymmetrical (ADSL) with the data throughput towards the ISP slower than the flow from the ISP). DSL in 2006, when the data can first be estimated in sufficient detail, accounted for 76% of all broadband connections. Wireless connections were in their infancy, accounting for 3% of connections. Other types including satellite and cable accounted for the remaining 21% (Chart 5). During the next six years, the DSL trend flattened out to reach around 4.6 million connections in 2012 (40% of the total). Wireless became the dominant type, reaching 5.9 million subscribers (51%) in 2012 compared to only 100,000 in 2006. Other connections reached 1.1 million subscribers in 2012 (9%), with cable accounting for 86% of that total.
The vast majority of wireless connections are mobile rather than fixed. They include USB modems and tablet SIM cards as well as other devices listed as datacards and dongles (generally described as copy-protection devices).
For many years, the ABS distinguished regularly between household use and use by business and government (Chart 6). It is possible to construct a reasonable time series indicating that the business and government component in 2000 accounted for 11% of total connections, and households 89%. Business and government increased their share slowly to 2006 when they reached about 13%. They then expanded more rapidly and by 2010 had increased to 20%, a level estimated to have been retained in 2011. They appear to have lost a bit of ground in 2012, if the estimate is correct that household connections had increased by 28% since 2010, and business and government by 21%. In total numbers, households remain dominant.
It is uncertain how this affects total downloads. The last time downloads were measured for the two components separately (September Quarter 2006), 826,000 business and government subscribers downloaded 6,733 terabytes (8.15 gigabytes per subscriber), and 6.14 million households downloaded 29,415 terabytes (5.05 GB per subscriber). These statistics were then discontinued, perhaps because the distinction between the two categories had become blurred.
The growth in the total amount of downloads, however, is truly staggering. It increased from 1,050 terabytes in September Quarter 2000 to 414,537 TB (or 414.5 petabytes, each equal to one billion megabytes) in June Quarter 2012. Chart 7 uses a logarithmic scale to convey how downloads have almost, if not quite, maintained a compound annual growth of 64.3% - the rate compatible with an increase from 1,050 to 414,537 over 12 years.
Average quarterly downloads per subscriber (household, business and government) were 273 MB when first measured in 2000. The one gigabyte average was reached in early 2004, 10 GB in late 2008, 30 GB in late 2011, and 34.4 GB in June Quarter 2012 when just over 12 million subscribers downloaded almost 415 TB — 414,537 GB to be exact.
An similar way of gaining a perspective of the trend is to look how long it took for the total downloads figure to increase by an order of magnitude (tenfold). Chart 7 suggests that it took from 2000 to just after mid-2004 to go from one to 10 terabytes (a little over four years), and from that time until 2009 to reach 100 terabytes per quarter (just below five years). On present trends, 1,000 TB should be reached in 2015, or in six years. This indicates that the actual rate of growth is reducing, but the actual compound annual growth rate from 2006 to 2012 was still a spectacular 52.9% pa. Few if any trends have ever shown such resilience.
To conclude, the Australian Internet statistics are not specifically about music, and the American data demonstrate the diversity of Internet usage. But the Australian statistics demonstrate how the opportunities for the music industry (and piracy), and for that matter for any other industry going down the digitizing path though the details would vary, have expanded over little more than a decade.
There are two other data sources that relate to the industry, both compiled by the Australian Bureau of Statistics, and both shown elsewhere in the knowledge base.
Household Expenditure shows that weekly expenditure per household of prerecorded CDs and audio records, at constant 2010-11 prices, declined from $2.02 in 2003-04 to $0.95 in 2009-10 (-53%). Total expenditure on these items for the years in question halved, from $814 million to $406 million.
Perhaps coincidentally, the weekly expenditure on music concert fees and charges increased from $0.88 to $1.95 (+122%), and totally for the year from $356 million to $833 million. The weekly total for the three items together was $2.90 in both years, while the yearly total for Australia rose from $ 1.17 billion to $1.29 billion as the number of households kept increasing (+10%). If this apparent substitution in the household budget of concert attendance for buying CDs and other physical records is real, it would suggest that concerts became more affordable because the closest substitute to physical recordings (their digital counterparts) became so much cheaper for the average household. This can only be posed as a hypothesis and may be a coincidence though quite a striking one.
According to Music Composition, Distribution and Publishing, there were 1,143 actively trading retailers of recorded music, and 678 actively trading businesses in recorded media manufacturing and publishing in June 2007. Many were small businesses that didn't employ staff: 40% of the retailers and 47% of the manufacturers and publishers.
The statistics have now been reclassified and updated (Table 3). The relevant new categories are "reproduction of recorded media" and "entertainment media retailing", as defined by the Australia and New Zealand Standard Industrial Classification (ANZSIC) 2006. The former group is a subclass of printing and includes all pre-recorded media including CDs, DVDs, CD-ROM and cassette tapes. The second group comprises retail sales of new goods, including computer games as well as CDs, DVDs etc. Table 3 shows that there are now less than 300 businesses classified as reproduction of recorded media, and the number has been generally falling. The apparent decline has been more consistent for the retailing group, from about 1,350 in 2007 to little more than one thousand in 2011 (a 25% fall in four years).
The music industry appears to be leading the counter-charge in the digital revolution, keenly watched by other industries from video stores to books and newspapers. All, of course, have their own lines of defence and strengths and weaknesses, but the music industry took it on the chin first and is probably ahead in the battle to rebuild.
The industry will be different when the dust settles though the major recording companies (of which at least one was amalgamated into another corporation in the process) will probably remain influential. Time will show the extent to which new business models will emerge to protect artists and the broader industry alike, and whether the intellectual property legislation which probably some share of responsibility for the decline, will be changed. There are some indications that artists yet again may be at the bottom of the pile when it comes to sharing the benefits — this needs to be explored in the light of the still emerging technology.
Hans Hoegh-Guldberg. Full version entered 5 April 2013. Statistics showing counts of businesses 2007-11 added 9 April 2013, together with minor editorial amendments. Note added to concluding section 3 August 2013.