Falling off the edge of a flat world? - Internet gossip on P2P filesharing research – a response from Professor Birgitte Andersen

This blog challenges an attempt by entertainment industry stakeholders to discredit the Andersen-Frenz report (2007): The Impact of Music Downloads and P2P File-Sharing on the Purchase of Music: A Study for Industry Canada.

Findings from the report have been published in a peer-reviewed journal, and used as expert evidence in two British landmark court-cases dealing with P2P filesharing (Oink’s Pink Palace and one other). It also plays a central role in the debate on copyright reforms internationally.

Campaign to discredit

You have got to ask yourself, - who will spend time going through years (across 2005,06,07) of internal and informal emails (obtained under the under the Access to Information Act) between myself and one of my consultancy contractors (Industry Canada), a past reviewer’s report of an early draft, selectively reading my publications since 2001, and producing about 20 printed pages (in two parts) of menacing rant appearing to try to undermine my (Andersen’s) research credibility and the Andersen-Frenz (2007) study for Industry Canada on “The Impact of Music Downloads and P2P File-Sharing on the Purchase of Music”.

Amazingly, this two-part diatribe (published in May and June 2010) attacking me and my research is published as an apparently ‘independent’ analysis representing facts and not taking sides in the ongoing international debate on how policy should respond to unauthorized filesharing on the Internet and the digital economy of the future.

The blogger is in fact Chris Castle from ‘Christian L Castle Attorneys’ based in Los Angeles and San Francisco USA, and the firm’s web site lists that they represent the record labels, film studios, among others. Thus, one could consider they have a clear financial interest in the debate on intellectual property (IP) policy in the digital economy.  Another blogger spreading this unsubstantiated gossip on the Internet is Bob Tarantino from Heenan Blaikie’s Entertainment Law Group in Toronto.

Is this type of work commissioned? Peer- reviewed? (not likely!), and what interests surround the making and publication of it?

Is it hypocritical, or even dishonest, that firms so closely linked to the interests on the one side of the copyright debate on digital economy policy present themselves as neutral carriers of ‘truth’ while accusing a neutral academic study of taking sides?  Do these firms have a conflict of interest?

Interestingly, I was not informed of the publication of this work (but realized its existence as I received a number of unusual emails over the summer), which explains my late response.

The accusations

Information (citations and quotes from emails and past publications) listed in the blogs is presented out of context and outrageous interpretations, twisted arguments and a catalogue of wholly untrue statements are made regarding myself (lack of capabilities and honesty in research process), my work (lack of research history in areas I consult on), and my relations with my contractors and academic colleagues/co-authors (dubious hiring process and the suggestion that we are somewhat ‘activists’). I will give some examples to illustrate my point:

(1)

Bizarrely, the Chris Castle blogs flag standard practice or normal processes regarding consultancy project development, which are adopted in the project leading to the Andersen-Frenz 2007 report, as “old-fashioned scam”. This is untrue.

It is useful to know that the same processes which were used in the consultancy leading to the Andersen-Frenz 2007 report (based upon independent research, academic refereeing, interactive discussions with contractors, copy-editing for publication) are commonly used in academic consultancy for national and international governments. This was/is identical to my experiences when consulting for other Government bodies such as, for example, the UK Intellectual Property Office, UK Department for Trade and Industry – DTI /UK Department for Business, Innovation and Skills - BIS, the OECD (Organisation for Economic Co-operation and Development), and the Danish Ministry of Science Technology and Innovation, among others. 

In my opinion the 20 pages of criticism demonstrates no clue of how robust interactive review-processes are common in academic studies (and how this contributes to ground-breaking research) or how knowledge flows from research base into use by government or industry.

For example, the blogs seem to suggest that the use of academic refereeing on an early or first draft, and my willingness to collaborate around the comments in order to subsequently strengthen the robustness of the results “rather than going into long discussions” , is an indication of a weakness of me as a consultant and that it makes me “underqualified”. Furthermore, it suggests the organization of copy editing (which is simply language proof-reading and suggestions of where to shorten the length of the text) by the contractor (Industry Canada) on a final-draft as an example of the research results and report not being my work: “So who actually wrote this paper? Isn’t it at least a collaboration”, and later he refers to it as “...rewriting done by Industry Canada”.

This campaign to discredit also appears to disregard my own intellectual property rights over the Andersen-Frenz (2007) report, by deliberately not recognizing me as the intellectual architect behind the research. For this reason I will simply now repeat what is also listed in the Andersen-Frenz (2007) report ( - how could Chris Castle miss it?), and in another publication emerging from it (Andersen, B. and Frenz, M (2010). Don't blame the P2P file-sharers: The Impact of Free Music Downloads on the Purchase of Music CDs in Canada. Journal of Evolutionary Economics. Vol 20, No 5, Pp 715-740 –available online since 23 March 2010):

(i) Andersen developed the methodology underpinning the design for the subsequent data analysis (I here explain this includes literature review, hypothesis development, variable decisions for subsequent data collection, research design including theoretical underpinning for subsequent data analysis). (ii) Andersen initially provided Industry Canada with the questionnaire developed for the survey. The final version was shaped in accordance with the recommendations by Industry Canada and Decima Research, and in accordance with the results of the pilot survey conducted by Decima Research. (iii) Decima Research conducted 2,100 telephone interviews with Canadian households, and provided the raw data. (iv) Industry Canada prepared the survey database. Plus, (v) Analysis and writing process was in collaboration (co-authored) with Frenz. Andersen was responsible to Industry Canada for the content and the delivery of the report.

(2)

I was accused of having “never published a study for which she had designed or developed surveys, or analyzed survey data”.  This is untrue. 

My coordination of the design and data collection for a survey on IPR strategies in IPR marketplaces (including patents, copyright, open source and non-protected inventions) as part of an EU Specific Targeted Research Project (STRP), called U-KNOW (Understanding the relationship between knowledge and competitiveness in the enlarging European Union), under the EUROPEAN UNION 6TH FRAMEWORK PROGRAMME, illustrates an example. The research design for this competitive tender started in 2004 and the research was conducted 01/03/2006 - 28/02/2009. Several peer reviewed journal publications, based upon quantitative analysis of this survey data, have emerged from this, and the data are also feeding into a consultancy report for the OECD Working Party on SMEs and Entrepreneurship (WPSMEE) and a consultancy report for the UK Intellectual Property Office, Strategic advisory board for intellectual property policy. (Please check my website for further information)

Furthermore (as evidence of experience in analysing huge data sets), in 2001 I published a book on patent statistics entitled “Technological Change and The Evolution of Corporate Innovation: The Structure of Patenting 1890-1990” (Edward Elgar, 285 pages). Favourable reviews of this book (with direct comments to the statistical analysis) include those made by: Brian S. Silverman (University of Toronto, The Academy of Management Review, Vol. 27, No. 2 ., Apr., 2002) and Richard R. Nelson (Columbia University, Journal of Technology Transfer (27) 2002).

(3)

By lifting a single line from one of my publications “It is well-known that artist are not driven by reward incentives for their creative expressions, but instead driven by their inborn spirit”, the blog suggests (by sarcastic repeating a historical comment “Let them eat cake”) that I do not recognize that artists need to be financially rewarded from their creative skills. This is untrue.

If my published papers had actually been read, it would be clear that whilst the papers acknowledge that individual artists (and individual researchers and inventors) would always be creative (e.g. creative talent and creative expressions exist in all global societies, including those with weak or no IPR systems), the challenge to industry and the current intellectual property rights system is to capture this diverse creativity and enable artists and researchers to push their creativity and inventiveness to a higher level, and to become commercially successful, and to build an industry in which all artists are fairly rewarded for their talent. It is useful to know that the current situation is that empirical studies of cultural markets consistently show a winner-take-all distribution of earnings. For example the Monopoly and Mergers Commission's study of the UK collecting society, the Performing Right Society - PRS (1996), revealed that the top 10% of composers/songwriters earn over 80% of total earnings, and there is no evidence that this skewed income distribution has changed.

That the IPR system supposedly functions as a stimulus to the venture capitalist to invest in creativity, invention and innovation, rather than direct stimulus to the creativity among individual artists and researchers, who will always be creative, is a well-known argument by those in the field of intellectual property rights and was presented as long ago as in the nineteenth century when the rationales for the intellectual property right system was debated among the economic philosophers. For an overview of this debate see Machlup and Penrose (1950) The patent controversy in the nineteenth century, Journal of Economic History, X(1), May.

I find it a concern if entertainment industry representatives are not interested in knowing about the sources of creativity, and even try to ridicule this debate. How do they pick artists to promote, - or do they simply manufacture them?

(4)

Several completely out of context citations in the blogs appear to suggest that I am campaigning against the industry. This is untrue.

As an independent academic, I have no interest in campaigning against an industry (which can potentially enrich our cultures and lives and improve welfare), but also, I have no interest in protecting the specific interests of individual firms.  As an economist, my interest is in helping to shape policies that allow businesses, markets and economies to grow, through which socio-economic equality is improved. This is what normally would be considered as a progressive economy.

An aim for music industry representatives should be to stimulate a true global representation of the music industry, and the digital economy presents an opportunity for businesses, cultures and society to develop. For years it has been a concern that less developed countries (who are not poor in musical talent) are not able to break into the world music market where the major four labels (Universal, EMI, Warner, SONY-BMG) dominate three quarters, and that even in many developed countries artists are not able to make a living from their talent due to the star-dominated industry creating a skewed income distribution (see point (3) above). These issues have been discussed in depth within the United Nations and the International Trade Forum, and my published papers which the blogs refer to (by – again – picking citations out of their contexts) are adding to this debate and feeding into international development work. Frankly, to ignore such concerns is protectionist and naive. 

(5)

It was accused that the Andersen-Frenz (2007) report for Industry Canada was based upon “pre-ordained conclusions”. This is untrue.

This accusation was in particular based upon the following statement “Andersen is a strong critic of “mainstream economics” and “copyright law” in particular”. It strikes me that Chris Castle does not know what this means, so I will explain.

Departure from the mainstream:

This study, which measures the extent to which P2P file-sharing activities act as substitutes or complements to music purchases in markets for CDs and electronically delivered music, breaks with the mainstream economics approach which dominates the music file-sharing discussion. Whereas such models ‘assume relationships’ at the micro level (e.g. between file-sharing and purchases) based on observations made at the macro level, the approach in the Andersen-Frenz (2007) paper ‘measures the direct effects’ using micro data representative of the Canadian population. In particular, the analysis aims to assess the behavioral patterns associated with music consumption, as well as the motivations behind such behavior.

Previous large scale studies examining the impact of P2P file-sharing on music purchasing/sales have employed macroeconomic (i.e. aggregated) data and assumed certain relationships (See, for example, Liebowitz (2008) in Management Science, Vol 54, No 4), which is based on city-level data, and where a relationship is assumed between record sales measured for an entire city and Internet penetration in a city (used as a proxy for file-sharing)).  Our micro study has a strong advantage over such existing studies, as the data are collected by asking individuals directly about how much music they purchase and how much they download for free, and what their incentives are when they engage in free downloading. The behavioural incentives underpinning free music downloading, novel to this research, are the multiple effects of: ‘unwillingness to pay’ (market substitution), ‘hear before buying’ (market creation), ‘not wanting to buy a whole album’ (market segmentation), and ‘not available in the CD format or on electronic pay-sites’ (market creation). 

Thus, this paper, which methodological underpinning and survey-data is based upon measuring 'direct effects' at the micro level in order to explain the macro situation (rather than 'assuming relationships'), can only be regarded as an improvement of research in this area. To my knowledge, this is also the first large scale empirical study to employ representative microeconomic data derived from direct survey responses. 

Not about being a “strong critique of copyright law, in particular”:

As an independent academic at a university, and a Professor of the Economics and Management Innovation, one of my research themes is to investigate if policy instruments (such as the intellectual property rights) perform in accordance with their objectives. If they don’t, is there problem in the IPR marketplace? (e.g. problems with enforcement), or a problem in the operational aspects of the management of IPRs in firms and sectors? (e.g. our business models), or do our IPR laws need amending? (and if so, in which direction?).  Radical changes to our technology (e.g. from analogue to digital) challenge all these areas.

To openly and broadly evaluate and assess, via evidence based research, if our policy instruments and management strategies perform in accordance with the objectives of government and industry and their overall rationales, is not about being opinionated or a ‘strong critic’. This is what academic economists do (including those specialized in the area of IPRs), and this is also what national and international governments are concerned with, and this is also what the entertainment industry should be open-mindedly interested in.

Digital entrepreneurship:

Chris Castel also picked up on another point in his blog, which in his opinion seems to contribute to making my conclusions (regarding the effect of P2P filesharing on music markets) pre-ordained or disqualified. Before entering the P2P filesharing project for Industry Canada I had expressed a concern in one of my papers (discussed at a Music Industry Workshop at the UN Least Developed Countries (LDC)-III Conference in 2001) that the music industry should try harder to learn from digital entrepreneurship within the industry (e.g. from the NAPSTER case and like) and incorporate such ideas into forward looking business models, instead of fighting in court the new opportunities which arise from digitalization.

I cannot see how this view automatically means that I had some pre-ordained conclusion when entering this research project on what effect downloading of free music may have on music pay-markets, and which I needed to prove. They are unrelated subjects.

My research for Industry Canada was in response to the unknown effect of free music downloads and P2P filesharing on music markets and the music industry. Different industry representatives claim opposite effects, and there was a need for a large-scale survey to provide more profound evidence of what the true effect of free music downloading and P2P filesharing really is. Presumably, I was found suitable to conduct the study on the market effect of free music downloads, including P2P filesharing, because of my previous work with economists in the United Nations (UNCTAD) on the global music industry, and because of the fact that my career since the beginning 1990s has mainly been within the area of IPR research (in relation to corporate innovation, the rationales for IPRs, the performance of IPR systems, and patent statistics). Also, despite my work on music copyrights, I had never claimed an opinion on what the effect of free music downloading and P2P filesharing is likely to be, and I am independent, in the sense that I have no interest in what the analysis would conclude.

It is pointless to go further into the campaign trying to discredit me and my research.

I believe that Christ Castle’s blogs throw an unflattering shadow on the behaviour of some entertainment industry representatives in the planning our future digital global society.

Jumping on the bandwagon

Unfortunately, the manner or style of the attack on the Andersen-Frenz 2007 study by Chris Castle (spread by Bob Tarantino) is not unique. Interests groups and their representatives have pushed for Industry Canada to remove the study from their website.

One can only assume that experienced entertainment lawyers and other interest groups know that ‘fabricated fairytales’ of this style cannot successfully challenge the Andersen-Frenz 2007 study, - neither in an academic or legal context.  But also, that is not the point.  It is about creating perception. They appear to have no interest in evidence based research or in a serious academic debate. However, is their aim to influence the political process dealing with copyright-reforms in Canada and elsewhere, by trying to create doubt and prejudice against the credibility of the Andersen-Frenz 2007 report and me (Andersen) as a researcher?

Vicious and ruthless scare-campaigns are not about finding the truth, as long as they work. Scare-campaigns of pressure groups have been launched against research contributing to the progress of policy-reforms dealing with climate change, health sector reforms in the USA, and anti-smoking movements.

The problem with policy reforms on copyright (just as those relating to climate change, the USA public health sector, and anti-smoking movements) is that even if the overall socio-economy can experience progress, some established firms may find their market position weakened, and they now act as forceful lobbyists to enforce their market position. Stakeholder interaction behind the Digital Economy Act in the UK (came into force June 2010) represents an example. A power struggle has emerged across technological regimes between (i) the old economy dominant players whose business models are developed around content production for analogue distribution channels (e.g. Warner, Universal, EMI, and Sony) and (ii) the Internet service providers who have grown up with digitalization and the Internet and who are not developing content or products but selling services using the technological and business opportunities of digitalization (e.g. Google, Facebook, Yahoo, eBay, TalkTalk and BT). You can read more about this power struggle in my blog which debates the policy measures in the UK designed to curb P2P file-sharing on the Internet

The INTUITION:  Falling off the edge of a flat world?

Worryingly, it is also argued by some that their ‘intuition’ (thinking that P2P filesharing of free music substitute music markets) should be treated as more valid than the ‘evidence based research’ by Andersen-Frenz 2007 (which found opposite conclusions to their intuition), and that this is considered sufficient to dismiss the research findings in the Andersen-Frenz 2007 report. This view is put forward in the blogs by Chris Castle. This view has also been echoed in many responses to our study directly from the music labels and other bloggers. 'Intuition' also plays a role in the academic research. Immediately after the release of the Andersen-Frenz (2007) report Prof Liebowitz of the University of Texas gained media headlines pronouncing that “without going into details of the study we can ask whether this result is even remotely plausible” and that “the result is so counterintuitive that I think it fails the laugh test.”(Our reply to Liebowitz)

Of course intuition has its place, e.g. for short-cuts or if we do not have concrete evidence to rely on.  However, if we are unable or unwilling to free our minds, ‘intuition’ can also imprison our thought and lead to prejudice and ignorance.

For example, although the world seems to be flat (by pure intuition), then falling off the edge of a flat world is not among my fears!  Similarly, although it seems that the sun rotates around the Earth (by pure intuition), then research has proven that it is the other way around. However, we shall not forget that after Galileo announced these research results, which were counter-intuitive for the general public and the belief of the Catholic Church, he was forced to retire as a scientist and live in house arrest.

What do you think?  Should policy-making and court-trials be based upon ‘intuition’ (e.g. of stakeholders and their representatives) or evidence based research? Why has much intellectual property right policy lacked a foundation of evidence based research?

The EVIDENCE BASED RESEARCH results:  This is how it works!

Given some cannot grasp by ‘intuition’ how P2P filesharing of free music can increase music purchase, I will now summarize how it works. The (Andersen-Frenz 2007) analysis was carried out for 2100 Canadian households and the data are representative (age, gender, region, and downloading status) of the Canadian population age 15 and above.

P2P filesharing creates more markets than it substitutes:

(a) Comparing P2P filesharers of free music with those who are not engaging in such activity:

• Free music downloading, including P2P file sharing, is neutral to CD purchase. That means that there is no difference in CD purchase between those who fileshare and those who do not.
• However, those engaging in free music downloading are more likely to purchase electronically delivered music.

(b) Focusing only on the behaviour within the P2P file-sharing group:

We found that the more people engage with P2P file-sharing, the more music they purchase. And this is even after we adjust for the effect of ‘music interest’ so the result do not simply reflect the fact that music lovers engaging in more P2P file-sharing also buy more music.  So, how do we explain this increased purchasing behaviour of active file-sharers? For this we need to assess, for each person, the various effects of P2P filesharing, as there are many reasons for sharing music freely.

• Market substitution effect: We found with statistically significance that people download some music freely as they are unwilling to pay. 
• Market creation effect:  In particular, we also found with statistically significance that people engage in P2P file-sharing because they explore, and this leads to subsequent purchases due to a ‘hear before buying’ effect.
• Market creation effect:  Another statistically significant market creation effect is that people look for music which is ‘not available elsewhere’ (e.g. in the mainstream outlets) boosting new markets.
• Market segmentation effect: Another reason for people to engage in P2P file-sharing is that people are ‘wishing to not to buy ‘whole album'’, but prefer the single digital file.

Overall, we found that the increase in music purchases of more active P2P file-sharers is explained by the fact that the significant ‘market creation effect’ of P2P file-sharing outweighs the significant ‘market substitution effect’, where people download music freely as they are unwilling to pay. 

There seems to be a misguided obsession by some on the substitution effect of P2P file-sharing, ignoring the market creation effect. The situation is complex and this is not acknowledged in many industry claims (as those discussed above). 

(c) A change in technological paradigm, and the demise of the CD format, explain the fall in CD sales (and the increase in sales of electronically delivered music)

• People who own MP3 players are less likely to purchase CDs.
• People who own MP3 players are more likely to purchase electronically delivered music.  

For more detail (INFORMATION, RESULTS, and DIGITAL ECONOMY POLICY IMPLICATIONS):

• Link to study Andersen-Frenz  (2007) The Impact of Music Downloads and P2P File-Sharing on the Purchase of Music: A Study for Industry Canada)

Blogs:

• This blog debates the policy measures in the UK designed to curb P2P file-sharing on the Internet. Shackling the Digital Economy means Less for Everyone: Why the DIGITAL ECONOMY ACT is bad for the music industry and the digital economy.
• This blog contains the Andersen-Frenz response to the critique by Liebowitz: The Impact of Music Downloads and P2P File-Sharing on the Purchase of Music.

Impact:

• Part of the Andersen-Frenz (2007) report (focusing on the effect of P2P filesharing on CD sales taken over the entire population) is published in a peer-reviewed economic journal: “Don't blame the P2P file-sharers: The Impact of Free Music Downloads on the Purchase of Music CDs in Canada”. Journal of Evolutionary Economics. 2010. Vol 20, No 5, Pp 715-740. Link: http://www.springerlink.com/content/d3h27828tx887556/
• I (as defence witness) and the research findings in the Andersen-Frenz (2007) report were cross-examined in a landmark British court case (Oink’s Pink Palace) where the findings were accepted and used as key expert evidence. It has also been used in another British court case. Both walked free.