Post-Project Research Protocol April 19, 2010 February 2, 2011 October 2014

From 2010 to 2014, the US-based nonprofit Future of Music Coalition conceived of and executed the broadest collection of primary-source data from US-based musicians and composers about their income conducted to date. The Artist Revenue Streams project has published over two-dozen data driven memos that examine the factors that affect musicians’ earning capacity, available at http://money.futureofmusic.org.

This document is a post-project version of the team’s research protocol. This includes commentary on some of the alterations we made to our methodology along the way. We hope that colleagues find this information useful in planning and executing similar research.

This draft is for Vienna Music Business Research workshop attendees. We will incorporate any additional feedback received at the workshop into this document and make a final copy available. Table of Contents

Purpose: The Value of Measuring Musicians’ Sources of Revenue...... 3 Research questions ...... 4 Existing work in the field ...... 4 Hypotheses ...... 5 Research methods ...... 7 Population of Study...... 10 Partnerships and participation ...... 12 Publication and dissemination strategies ...... 13 Objectives and Outcomes ...... 13 Future work ...... 15 Musicians as First Movers ...... 16

Appendix A: interview models and pre-qualifying survey ...... 17 Appendix B: Musician types ...... 20 Appendix C: Qualitative interview questions ...... 21 Appendix D: FAQ for financial case study candidates ...... 24 Appendix E: The 42 streams ...... 26 Appendix F: Population definitions and sample size for survey ...... 30 Appendix G: Money from Music survey construction ...... 34 Appendix H: Money from Music survey protocol/instructions ...... 36 Appendix I: Partnerships, Connectors and Advisors ...... 39 Appendix J: Research Advisory Committee ...... 41 Appendix K: Timeline ...... 43 Appendix L: Project privacy and permissions policy ...... 45

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 2 Purpose: The Value of Measuring Musicians’ Sources of Revenue

FMC’s primary purpose for conducting this research reflected our interest in documenting how the changes in the music landscape were affecting musicians’ earning capacity. New technologies like digital recording studios, digital aggregators, online music stores, streaming services and webcasting stations have greatly reduced the cost barriers to the creation, production, distribution and sale of music, and a vast array of new platforms and technologies – from Tumblr to Facebook to Twitter feeds – now help musicians connect with fans.

Many observers are quick to categorize these structural changes as positive improvements for musicians, especially when compared with the music industry of the past, in which access to commercial radio, mass media, and traditional retail was tightly controlled and difficult to navigate without major label support. A prevailing thought is that the barriers to entry are now so low (and so inexpensive), and the number of platforms through which musicians can create, promote, distribute and monetize their work are so vast, that these technological changes have had a net positive effect on the creative community. And it is true; musicians’ access to the marketplace has greatly improved. But how have they impacted musicians’ ability to make a living? Almost all analyses of the effects of these changes rest purely on assumptions that they have improved musicians’ bottom lines.

A handful of social researchers, academics and organizations have published studies that analyze the earning power or social capital of certain segments of the US music community, or how much musicians have earned from some specific activities. However, there has been no comprehensive effort to directly ask a wide range of musicians – working in many different genres and locations, and representing an array of career arcs – how their revenue streams are changing, if at all. This is why Future of Music Coalition developed a multi-stage research project to undertake this crucial work.

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 3 Research questions

The Artist Revenue Streams project collected information from a diverse set of US-based musicians1 about the ways that they are currently generating income from their music or performances, and whether this has changed over the preceding five years.

The fundamental research questions were:  What percentage of musicians’ income comes from each possible revenue source?  What is the ratio among different sources, whether it is royalties, money from gigs, t-shirt sales, or any of the 42 other meaningful revenue streams that FMC identified?  What are the demographic characteristics or external factors that influence music- related earnings?  Are the revenue stream ratios different for artists working in different genres and at different stages of their careers?

In addition to asking musicians about the percentage of income derived from “traditional” streams such as retail sales, public performance royalties and live performance fees, we collected benchmark data about participation in new streams like digital downloads, digital performance royalties, and a myriad of new licensing options. Over time, as the research is replicated, this will help us understand whether aggregate income from these new revenue streams offsets losses of revenue resulting from the erosion of traditional streams, or which, if any, of the new models are having a net positive impact on musicians’ bottom lines.

Existing work in the field

FMC’s research was informed by the excellent work done by a number of US-based social researchers who have been examining US musicians, income, hybrid business models and community networks over the past ten years. In 2003, Joan Jeffri from the Research Center for Arts and Culture published Changing the Beat, a detailed and comprehensive ethnographic survey of jazz musicians in three distinct communities: New York, San Francisco and New Orleans. Her work, which was supported by the National Endowment for the Arts, serves not only as a vivid snapshot of jazz players, but also serves as an excellent example of the Respondent Driven Sampling (RDS) model. Sociologist Daniel Cornfield from Vanderbilt University has done similar ethnographic

1 In this document, we use “musician” as an umbrella term that includes performers, recording artists, songwriters and composers.

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 4 work, interviewing 72 music professionals residing in Nashville, TN, with a chapter in Steve Tepper and Bill Ivey’s 2007 book Engaging Art. University of Massachusetts’ professor Pacey Foster has been examining the hip-hop community in Boston. Maria Rosario Jackson at the Urban Institute has been studying arts, culture, community and hybrid business models from an urban planning perspective, and Leveraging Investments in Creativity (LINC), a US-focused national service organization for individual artists has issued a number of reports about how artists build careers across commercial, nonprofit and community work. FMC’s work built upon these studies, but added a significant quantitative aspect to the research; something that has been lacking in the social research field.

In addition to academic sources, FMC examined the research published by industry sources such as Billboard and MiDIA Research, membership surveys of various organizations that serve musicians, and top-line data available from digital aggregators and rights societies that facilitate the flow of money to artists, including CD Baby, TuneCore, ReverbNation, Rightsflow, SoundExchange, CISAC, ASCAP, BMI, SESAC, Nielsen BDS, or Nielsen SoundScan. We also looked at NEA surveys, US Bureau of Labor Statistics information and US Census data in order to put the information from our interviewees and survey respondents in context.

Hypotheses

Given FMC’s work at the intersection of music, law, technology and policy – and our individual experiences as creators, musicians and independent label owners – we have some first-hand knowledge of the changes that today’s musicians are experiencing. Thus, we proposed the following hypotheses:

1. Musicians are relying on revenue from a variety of streams. Until recently, the vast majority of musicians’ revenue could be allocated among four possible buckets: live performances, record sales, royalties from public performances and/or licensing, and merchandise. For some musicians that specialized in a certain role – composers who don’t perform, for example – there were even fewer. While those revenue categories are still relevant, FMC suspects that today’s musicians are participating in more streams, both because there are more streams available – some of which didn’t exist until ten years ago – and because the traditional streams like retail record sales and mechanical royalties have been in serious decline. The big questions were: just how many revenue streams are today’s artists relying on? Has the number of revenue streams changed? And, do artists in different genres or at different stages of their careers participate in the

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 5 same number of streams?

2. The overall revenue pie is smaller for each artist, but more artists have access to revenue. Until about ten years ago, there was generally a more distinct split in the music industry: there were artists signed to major label deals that had access to recording advances, tour support and big budget promotion; and then there were independent and unsigned artists who were largely shut out of commercial radio/broadcast media (and the subsequent performance royalties generated by airplay), and had meager resources (in relative terms) to promote record sales. But because of the internet, many of the promotion and distribution avenues are much more accessible. Today’s independent artists can sell their music on the same digital platforms as any major label artist, 24/7, for a marginal cost. And, there has been an explosion of multimedia outlets and social networking services with negligible barriers to entry for the artists: satellite radio, specialized webcast stations, music discovery sites like Pandora and last.fm, and streaming services like Spotify, Rhapsody and Rdio. While it seems that technology has democratized the process and facilitated the creation of a “musicians’ middle class”, has the tide lifted all boats equally, or at all? Are there a greater number of musicians making money today than ten years ago, or fewer? FMC seeks to find out if there are, indeed, more working musicians, and whether the changes in distribution of revenue have been modest or marked.

3. For performers, money from touring or playing live shows is the biggest portion of their revenue pie. Even with the high costs of gas, and percentages paid to booking agents and managers, existing data and musician self-reporting indicates that live shows are where artists are able to gross the most money. However, is this true for all genres, or just for rock, pop and country artists where touring is expected? Is it as true for emerging artists as with established artists? And how do the costs of touring impact its reliability as an income source?

4. Songwriters and composers are seeing diminished mechanical royalties. Music industry statistics point to a continued and sustained reduction in traditional retail sales. For songwriters and composers, this corresponds to a reduction in mechanical royalties. But is this true for all genres? Are jazz and classical artists – who are somewhat insulated from the commercial retail environment – experiencing something different? Plus, the current music landscape presents many new possibilities for composition driven income, including ringtones, increased synch license opportunities, digital sheet music and publishing royalties from digital sales. Are these new revenue streams based on licensing compositions making up for the significant reduction in traditional

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 6 mechanical royalties?

5. Location doesn’t mean as much as it used to. There are traditional “music towns” in the United States: Los Angeles, New York, Nashville, Austin, Chicago, where musicians and the music industry have congregated. But now that the internet has largely removed regional barriers, especially for the promotion and distribution of music, what effect does an artist’s location have on his or her ability to make a living? We suspect that we’ll find that a number of artists are making a living outside of the traditional regional music industry structures, relying more on internet connections and virtual networks to sustain their careers. But has location been completely removed from the factors influencing an artists’ career?

Research methods

Because the musician population in the US is not well-defined, nor easy to reach, FMC chose to employ three data collection methods simultaneously:

 in-person interviews with a small but diverse number of musicians in 2010-11  a review of financial records of some musicians in 2011-12  a widely distributed online survey in fall 2011

We felt that this direct, multi-method strategy was the best way to get a snapshot of different musicians’ income streams. This mix of methods also allowed us to approach musicians who would otherwise be difficult to reach, or who would be unlikely to take an online survey.

Artist interviews: In step 1, FMC interviewed 80 different musicians (or in some cases, their managers or accountants) to seek their direct input about their revenue streams, and how these have changed over the past ten years. FMC identified about 20 musician types that we expected to have different musical revenue pies. FMC used networks and referrals to identify individuals that fit these musician types, then used a snowball sample method to interview peers.

In addition to the snowball sample, FMC is also sought interview candidates through a randomized sample of FMC newsletter subscribers. In November 2010, 1,000 random newsletter subscribers received an invitation to participate in this research. They were asked to fill out a 14-question pre-qualifying survey. Of the 1,000 possible candidates, 9

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 7 individuals completed the online survey and, of those, 6 met our criteria. We took this step to reduce researcher bias, and expand the pool of possible interview candidates.

FMC started conducting outreach and scheduling interviews with “seed” musicians, managers and accountants in Q2 2010, and interviews continued throughout 2011. Based on results from prior RDS researchers, we estimated that 55 percent of the referrals would be successful, giving us a grand total of 77 interviews. We ended up with 80 completed interviews.

The information gathered during this stage informed the survey-building process, and provided us with over 1,000 pages of rich, qualitative responses to the revenue questions. Portions of these interviews have been included in almost all our data memos.

For details about interview methodology and selection criteria, see Appendix A. For a list of musician types, see Appendix B. For interview questions, see Appendix C.

Financial data: FMC also reviewed individual financial data from six of the musicians that we interviewed, with an understanding that any financial information would not be identified with any particular artist, band or label. Actual financial data – Quicken/Quickbooks reports, accounting and royalty statements – helped us to quantify both the income and expenses generated in different categories, and changes from year to year.

All financial data was handled with the utmost discretion and sensitivity. Contributors were anonymous, and FMC did not report actual dollar figures, only the ratios among various categories of revenue.

The artists’ financial data was presented as six case studies, each with an individualized narrative/assessment that explain the reasons for fluctuations or changes in their income and expenses from year to year. The data is displayed primarily in two forms: 1. Revenue pies that illustrate the percentage of annual revenue from specific income categories for each musician. 2. Stacked column graphs that show a musician’s income and expenses from year to year.

The financial case studies were both the most difficult – and possibly the most beneficial – way to collect and present musician’s income data. For this collection method to work, participants needed to turn over years and years of detailed personal financial information to us. Even with a stated privacy policy and assurances that their data would remain

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 8 anonymous, it was very difficult to get musicians to trust us with their financial information.

Even if musicians trusted us, we encountered wildly varying methods of financial recordkeeping. Some participants were quite organized and were able to send us Excel outputs or Quickbooks reports of their financial data. Others only had paper records, check stubs and receipts. Data collection for some participants meant going to their homes or manager’s offices, sorting through paper files, and manually entering data.

One more barrier was participants’ own confidence in their financial picture. Most musicians declined to participate in this component; some stated it was because they felt their finances were nobody else’s business, some were uncomfortable with what the analysis might say, and some were wary of being judged for how much (or how little) money they made.

While the process was frustratingly iterative and extraordinarily complex at times, the stories we are able to tell with the financial portraits are incredibly valuable. First, they clearly show how different the income picture is for musicians working in different roles and genres. Second, they show more objectively than any survey or interview could do how various income streams have changed over time. Some of the shifts in income streams are due to technological changes but, in other cases, they reflect album cycles or personnel changes in management or label relationships. Third, they show a detailed accounting of musician-related expenses and, subsequently, make it possible for us to calculate a net income for each participant. Expenses matter immensely if we are to portray an accurate picture of musicians’ earning capacity, and there was no other way to gather expense information except through this direct method.

For details about the types of financial statements FMC examined, see Appendix C. For an artist-based FAQ about the process, see Appendix D or FMC website. For the project’s privacy policy, see Appendix L.

Online survey: For step 3, FMC created a 200-question online survey called Money from Music that presented relevant questions about artists’ revenue streams for US-based musicians. The survey used skip logic to create a survey experience that matched the subject’s role(s) (songwriter, composer, performer) and recognized the differences in how various communities operate and how creators are compensated.

FMC spent four months building and testing the survey instrument. We paid careful attention to language, and provided survey takers with three paths to choose from. Simultaneously, we did outreach to nearly 100 key organizations with a musician or

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 9 composer constituency, describing the survey and assessing their interest in helping us promote it amongst their membership. FMC put the survey in the field from September 6 to October 28, 2011, and mounted an aggressive marketing campaign that not only included outreach to partner organizations, but also social media updates, YouTube and Vimeo promotions, paid advertising, flyers at shows, posters and lanyards at festivals, and a lot more. The combination of partner support and aggressive marketing led over 7,000 musicians to begin the survey, and 5,371 to complete it, giving us a huge and robust sample to analyze.

Survey question set is posted online on ARS website. For population definitions and sample size – including risks and limitations – see Appendix F. For survey construction notes, see Appendix G. For survey protocol, see Appendix H. For partnerships, see Appendix I.

Population of study

One of the greatest challenges in doing this research was defining our population of study. There is no clear job description for “musician”, nor certifications or qualifying tests. In addition, there is no one organization in the US that represents the majority of musicians. Read a more detailed post about the difficulties in quantifying the musician population.

In the absence of standards, FMC established some parameters to define our population of study. First and foremost, this study was constrained to US-based musicians. This was a reflection of our capacity, and a recognition that many revenue steams are highly dependent on sovereign copyright law. An international or country-to-country comparison would be incredibly valuable, but was outside the scope of this project.

Second, this was constrained to participants 18 years of age or older.

The biggest debate amongst team members was whether to limit the population to “professional” musicians (a problematic and vague term with no consistent definition), or those who met some sort of minimum criteria. In the end, we decided not to impose any pre-set musician-related criteria on our participants. We welcomed any participants who self-identified as musicians.

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 10 As a result, there were only two qualifying questions prior to the start of an interview, and at the beginning of the survey. They determined that respondents were:  US citizens or permanent residents  18 +

However, we also asked interviewees and survey respondents dozens of demographic and descriptive questions about their role and career trajectory in the music industry. We asked:  What musical role(s) they played (performer, composer, recording artist, session musician, etc.)  How many workweek hours they spent being a musician  What percent of their annual personal income was derived from being a musician  Whether they were members of any of the PROs, unions, or one of nearly two dozen service organizations or associations  Whether they had any “teammates”, from attorneys to business managers to booking agents  Whether they had a relationship with a record label  Number of recording and/or composition credits in past year and during lifetime  Career length  Whether they had attended conservatory or music industry schools

This minimum threshold/maximum intake strategy was the appropriate choice for our work for two reasons. First, we were interested in gathering information from as many different types of US-based musicians as possible, so we did not want to create any unnecessary barriers to participation. Second, we worried that any criteria that we could come up with to restrict the population of study to “professional” musicians might erroneously exclude qualified participants. For instance, we thought about using a minimum number of recording and/or composing credits as a threshold for participation. But this ran the risk of excluding orchestra members, legitimate professional musicians who focus almost exclusively on high caliber performance, not recording or composing. After much discussion, the team realized there were no musician-related criteria we could devise in advance that would properly limit our population of study to “professionals” but not potentially exclude legitimate participants.

The battery of demographic information that we gathered from interviewees and survey respondents made it possible for us to create our own criteria and groups after the fact. For instance, any respondents who said they spent 36 hours a week or more, and also earned 90% or more of their income from music were categorized by us as “full time” musicians. Adding on a third variable, full time musicians who had fewer than 5 years of career experience were classified as “emerging”, while those with 20 years or more of

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 11 experienced were called “old school”. In addition, having a wide swath of musicians in the pool made it possible to compare “full time” musicians against those who are not full- time, to examine the factors and characteristics that may make a career in music possible. If we had limited our population of study to professionals at the outset, we might not have identified some of these factors.

For more about the population of study and sample size, see Appendix E. For survey protocol, see Appendix F.

Partnerships and participation

The ultimate value of this work depended greatly on the level of participation we could encourage among musicians. FMC first built partnerships with a small set of academic and research institutions, then sought the collaborative support from dozens of member associations and service organizations, who in turn urged their members to participate in the survey.

For a list of advisory committee members & research partners, see Appendix J.

FMC recognized that relationships and trust building among collaborators and musicians was a significant and crucial part of this research.

Given the sensitive nature of some of the conversations we’re having with musicians, FMC also recognized the need to respect the preferences of our interviewees, especially the anonymity of those granting us access to their financial data. FMC instituted a research protocol that provided a number of safeguards to protect our subjects, including:  an informed consent statement that was reviewed and signed by interviewees,  a pledge to share drafts of reports and case studies with subjects before publication, and  an internal project privacy document to which the FMC research staff committed.

To review the project privacy statement, see Appendix L.

The success of this project also hinged on trust building and collaboration with dozens of musician-focused organizations and associations. We thought that most service organizations would find the online survey to be very worthwhile, and would be keen to see the results. Other organizations were concerned that outcomes – which were unpredictable – could lead to questions about organizational objectives or policy positions, particularly how they can best serve their own communities.

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 12

We took all reasonable steps to ensure that the survey instrument was as accurate and relevant as possible to all types of musicians and composers, without appeasing narrow group requests. Time was spent with some partners imagining different outcomes and considering what musicians, the media and policymakers will glean from various results. But we were also clear that we would not filter the results or manage the outcomes to meet predetermined expectations.

Publication and dissemination strategies

Equally important to this project was a publication and dissemination strategy that would ensure that the results were widely available to the public. By December 2011, the ARS team had over 1,000 pages of interview transcripts and survey data from 5,371 respondents. Instead of publishing one enormous report, the ARS team have written and published about 25 data memos that combine qualitative and quantitative data. These look at the data through different lenses, and examine factors such as the impact of organizational membership, teammates, conservatory training and radio airplay on musicians’ earning capacity. Other data memos look specifically at certain genres, or at income streams themselves. These reports and the six financial case studies are all available online at a website dedicated to the project.

The ARS team also presented findings at dozens of music-related events and conferences, wrote formal law and sociology journal articles, and spoke frequently with the press. The project also spawned additional projects, including the Music and How the Money Flows infographic, and the Money from Music Quizzes that test musicians’ understanding of US copyright law and business practice.

Currently, we are building a browser-based data portal – a dashboard – that will give users the ability to examine the survey’s revenue data in detail, adjust various filters and see how the charts and graphs change for musicians working in different roles, genres, or at different stages of their careers. The dashboard should be complete by January 2015.

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 13 Objectives and Outcomes

Though we could not predict the actual results of the work at the outset, FMC proposed at least five uses of the research data:

Objective 1: First, and most obviously, the results could provide musicians, the media and the music community at large with a comprehensive analysis of how musicians from many different genres are being compensated in the digital age. This data could serve as a vital benchmark for understanding the shifting revenue streams for musicians.

Outcome: We feel that this has come to fruition through the various data memos and financial case studies that have been published to date.

Objective 2: Second, the data could help service organizations and advocacy groups understand how they can best serve their constituencies. We assume that most unions and service organizations survey their own memberships, but having data that captures the experiences of other musicians could help them identify trends, map policy objectives, and recruit more members.

Outcome: The ARS team produced organization reports, which provided a member- specific snapshot of the survey data to a handful of key organizations. We also wrote articles for various membership-based publications (like AFM’s Allegro Magazine) and presented findings at dozens of membership-oriented events.

Objective 3: Third, it could also help musicians and advocates (and the media) make a more informed case to the general public about the complex realities of being a musician in the current landscape. For instance, many music fans assume that “all artists make money from touring” so they don’t feel guilty when downloading songs for free. Perhaps if the public better understood the complex nature of musicians’ revenue (and the relatively small numbers we’re talking about), we can enrich the public dialogue.

Outcome: This objective is continuously achieved through our ongoing conversations with the media – including NPR, New York Times, Salon – where we urge journalists and writers recognize the complexity and diversity embodied in the US musician community.

Objective 4: Fourth, this research could serve as an external assessment of the value of new technologies and services available to musicians and fans, for musicians. Many new business models have launched in the past ten years that use music as the main attraction – iTunes, Rhapsody, CD Baby, MySpace, Pandora, webcasting, podcasting, various video games, even Napster. While many of these include a revenue component for

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 14 rightsholders, there has not been any systematic effort put into examining if, or how, musicians as a whole have benefited from participating in these new models.

Outcome: We met this objective by collecting benchmark data on participation rates in many of these platforms, and gauging survey respondents’ capacity to use emerging technologies to promote and distribute their music. However, it is difficult to proclaim that we were able to do an external assessment of these services, partly because musicians’ participation rates remain low, partly because musicians are just now figuring out this new landscape (see our memo about why “I Don’t Know” is an acceptable answer), and partly because the tech players in the space keep changing. There is nothing settled about this ecosystem yet, so it is premature to do an external assessment, but future assessments will be greatly aided by our 2011 data.

Objective 5: Fifth, the results of this research could have policy implications. We may find that certain types of compensation are not as robust as they should be because of policy decisions. For sample, there is currently no public performance right for sound recordings for airplay on terrestrial radio stations in the US, but there is digital performance right for airplay on webcast stations and satellite radio. Our research may highlight how policy decisions affect artists’ revenue, and serve as a way to leverage any necessary change. We may also realize that, despite the technological progress that these new business models represent, the vast majority of musicians live from gig-to-gig and struggle with middle-class issues like mortgages, gas prices and finding affordable health insurance. No matter what the outcome, FMC recognizes the immense value in undertaking this work as a fundamental part of understanding musicians’ earning capacity, now and in the future.

Outcome: FMC uses this data routinely to explain and translate complicated policy issues. The ARS team has presented its findings directly to policymakers at the US Copyright Office, the US Patent and Trademark Office, and the White House’s Intellectual Property Enforcement Committee (IPEC), and FMC frequently references the research in our public filings and in policy conversations. Indeed, the entire FMC team feels like the Artist Revenue Streams work was not only important and interesting, but also absolutely necessary if we are to be honest and objective advocates for musicians in the policymaking sphere.

Future work

Given the rapidly changing nature of the music and technology industries, FMC sees the value of repeating this work on a regular basis. The benefits of repeating this research

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 15 over time would be enormous; musicians’ service organizations and policymakers would have access to data, as well as the ability to measure change over time.

FMC is not equipped to manage such a project on its own. As such we are exploring partnerships with academic or government institutions that might house such an effort.

For a list of possible sources of institutional support, see Appendix J.

Musicians as First Movers

Today’s musicians are experiencing changes as profound as any since the invention of recorded sound in 1880. We are living in an era of unparalleled opportunity enabled by technology, where an individual musician can easily record a song, upload it to a server, distribute it across many platforms, and collect the revenue. It’s also a time of heady excitement among technologists, who have used music as the “flame” to attract millions of users to emerging services like YouTube and MySpace, and to the hardware developed by companies like Apple. Yet these very conditions breed a sense of foreboding and concern among many stakeholders in the music industry who have witnessed a dramatic drop in physical music sales, and who have thus far been unable to curb the unauthorized sharing of music files.

Technology has greatly improved musicians’ access to the marketplace but, to this point, measurements about the effect of these seismic changes on musicians’ ability to make a living have either been anecdotal or speculative. Even the most esteemed books about copyright in the digital age are largely based on theory, and lack qualitative data. FMC believes that this multi-method, cross-genre, musician-focused research is essential to understanding how changes in technology and the music landscape have actually affected US musicians. We hope it will not only educate the public about musicians as creators, but also inform the development of successful models of marketplace creativity in the future.

Project contacts: Kristin Thomson Future of Music Coalition Jean Cook Future of Music Coalition email: [email protected] email: [email protected] cell: 267.971.3088 cell: 646.489.8791

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 16 Appendix A: interview models and pre-qualifying survey

For the ethnographic portion of the work, FMC used snowball sampling to identify interview candidates. This is a technique where existing study subjects recruit future subjects from among their acquaintances.

In the scope of our interview work, we:

1. Identified about 20 different musician types that we’d like to interview. We knew that the revenue streams of an orchestra player, a touring musician, and a person who composes for film are each very different, and we wanted to capture these differences. See Appendix B for details on musician types.

2. Identified specific musicians who fit those types to interview. These were our “seed” interviews. FMC has a strong network of musicians, including connections in the jazz and classical world. Our seed subjects were identified through a combination of personal awareness, and recommendations from experts at key organizations. As noted below, we also sought interviewees through a randomized process.

3. Conducted interviews. Interviews happened in-person, via Skype, and over the phone. Interviews followed a pre-set list of questions (see Appendix C), and most lasted between 30 minutes and an hour. Interviewees who completed an interview were offered a $50 gift certificate to compensate them for their time.

4. Asked for referrals In each case, we ask the interviewee for 2 or 3 peer referrals.

5. Conducted three waves of recruitment. We repeated the request for peer referrals for three waves, as illustrated below.

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 17

FMC started conducting outreach and scheduling interviews with our “seed” musicians in Q2 2010, and interviews continued throughout 2011.

Based on results from prior researchers, we projected that 55 percent of the referrals might be successful, giving us a grand total of 77 completed interviews. We ended up with 80 completed interviews.

There were pros and cons of using snowball sampling as a method. In our favor, the method was appropriate for our work; we wanted to talk to a variety of different musicians and composers, yet there is no “directory” of such musicians that we can reference, and no way for us to widely advertise the research request that would allow candidates to come to us. Under any model, we would have needed to ask for referrals, and snowball sampling was a method that was both effective and affordable.

The largest drawback of using a snowball sample is the impact of researcher or subject bias. A common hazard of snowball sampling is that people who have many friends – a strong network – are more likely to be recruited into the sample. More specific to our work was the hazard of “picking winners”; in other words, over-sampling musicians who we were aware of because they have been successful or have a higher public profile.

Though we cannot quantify an absence of bias, the snowball sampling method did introduce us to dozens of musicians with which we had no prior contact and, in many cases, did not know at all. We were pleased that snowball sampling quickly led us away

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 18 from our existing network of friends and associates, and allowed us to interview musicians that were new to us.

The “picking winners” hazard was harder to avoid. Though, again, we cannot quantify this effect, the vast majority of interviewees referred us to peers who were also full-time, working musicians. Some were earning six figures, and some were driving themselves in camper vans from gig to gig, but nearly all interviewees were cobbling together a career in music in some fashion. It would be fascinating and quite useful to interview musicians who had exited the music industry in the past 10 years, but peer referrals slanted towards those who interviewees thought represented working musicians today.

To further reduce bias, FMC also sought interview candidates through a randomized sample of FMC newsletter subscribers. In November 2010, 1,000 random newsletter subscribers received an invitation to participate in this research. They were asked to fill out a 14-question pre-qualifying survey. Of the 1,000 possible candidates, 9 individuals filled out the survey, and of those, 6 met our criteria.

Due to constrained resources, we only completed one of these survey-solicited interviews, but the method itself was a useful way to identify more interview candidates.

The pre-qualifying survey is posted here.

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 19 Appendix B: Musician types

Before seeking interview candidates, FMC cataloged the different types of musicians that we expected to have different musical revenue pies:

 Band/artist on a major label  Band/artist on an indie label  Band/artist that’s self-releasing their work  Band/artist that has a lot of merchandise  Band/artist that gets played on commercial radio  Songwriter who does not record, perform or tour (writes songs for others)  Singer/songwriter who records own songs, but does not tour  Composer who primarily writes TV and film soundtracks  Composer who writes on commission  Chamber music group player  Player in a major orchestra  Player in a member-directed orchestra  Freelance classical musician (hired gun for tours or studio work)  Freelance non-classical touring musician  Freelance non-classical session musician  Jazz player who tends to play existing compositions  Jazz player who also composes  Hip hop musician who composes own work, including samples  Musician who has been heavily sampled  DJ

We also expected there would be variations within revenue pies based on:  How much time they’ve spent in the music industry  Whether they have a publishing deal or are self-published  The genre they work in  Their location

FMC made every effort to seek out and acknowledge these differences during the interview process.

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 20 Appendix C: Qualitative interview questions

July 2010 location question added January 2011 Value question added April 2011

Role and relationship  What’s your role? (manager, musician, composer, label person, etc)  What’s your level of responsibility with regards to your artists’ revenue (or your own revenue)?  What genre(s) of music do you work in?  How much time spent as songwriter/composer/performer  Percent of total annual revenue derived from being a songwriter/composer/performer  How many albums have you put out or receive royalties on?  Which organizations are you a member of? (PROs, unions, Grammy, etc)

What kind of musician are you? [Note: some questions below will be modified for musicians who work in pit orchestras or are essentially hired guns, either for studio work or on tour]  Do you write your own songs/compositions?  Do you record your own songs/compositions?  Do you perform live? How much? Is that more or less than 10 years ago? o When you perform live, what percentage of your set is your own compositions?  Do other artists perform or record your songs/compositions?  Are you on a label now? What kind? Were you on that label 10 years ago?  Do you have a publishing deal? What’s the split?  [for pit musicians/hired guns] how are you typically compensated for your work? How frequently do you get copyright credit for the pieces you play?

What differences do you perceive?  It’s been 10 years since Napster appeared:  Have you/any of your artists been active over the past ten years? What does that activity look like (# of records released, amount of touring, etc)  What do you perceive as the sources of revenue that have developed or grown the most for you over the past 10 years?  Which revenue sources have shrunk?  Do you think this is typical of other artists in your genre, or is yours a unique circumstance? What makes it unique?

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 21 What are your feelings about music-related technologies?  There are many new technologies that have developed over the past 10 years: Napster, iTunes Music Store, eMusic, Rhapsody, webcasts, satellite radio, ringtones, podcasts, social networking sites, Twitter. o What do you perceive are the most helpful technological developments for generating revenue, or as a driver to revenue? This can include things that are purely promotional, but that you think have been effective at driving sales or increasing royalties. o What do you perceive as the technologies that are the most disruptive: reducing your revenue, or acting as unpaid substitution for former revenue streams like CD sales.

 Internet-based technologies are one thing, but there’s also been growth in the number of licensing and branding opportunities for musicians in some genres. o Have you had any experience with licensing your music to film, TV, ads, games? Has this grown over the past 10 years?

 Are there any other sources of revenue that have developed over the past ten years for you? Examples: o Sample licensing o Web click thrus o Corporate sponsorship o Grants/commissions

 Do you get played on radio? Commercial radio? Non commercial radio? Webcast stations? What do you think its impact is on revenue and sales?

 If revenue is increasing, are you actually making more money, or are the expenses of operating also increasing (website hosting, van rental, staff, hired gun musicians, staging, gas, whatever)?

 How does your location affect your ability to make a living? In other words, how important is it that you live where you live? Has this factor changed over the past 10 years?

 Do you think your music is properly valued in the marketplace? In other words, do you think you are fairly compensated for your performances, your recordings, your compositions? Has its value changed over time? Does it have a non-monetary value as well?

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 22  Any final thoughts? What do you see at the future for musicians like yourself?

 Can you recommend two artists like yourself that you think we should interview for this project? Would you be willing to introduce us via email?

Documents that FMC would be interested in seeing in any years from 2000-2010. We encourage people to redact anything that is private – account numbers, name of label, individual’s names, band name. We really only care about the ratios, and the time frame.  Profit/loss statements based on your musician-related income  Record label royalty statements: # of CDs sold, mechanical royalties  Record label advances (if you’re on a label that gives advances)  (for unsigned artists) CD Baby, IODA or Tunecore statements  SoundExchange statements: digital performance royalties in US and worldwide  Summaries of lump payments and/or residuals for licensing arrangements (synch licenses, games, ads, movies, ringtones, etc)  Merch sales statements  Tour grosses (or tour nets)  Union statements (AFM/AFTRA revenue)  Anything else that shows up as a significant revenue stream over the past 10 years

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 23 Appendix D: FAQ for financial case study candidates

Thank you for considering participating in the financial case study phase of our research study. We are currently looking for a handful of musicians and songwriters to take part in in-person interviews and to share financial documents with us. Below are some frequently asked questions.

What kind of information are you looking for? In addition to the interview, we’re interested in reviewing any financial documents you might have – profit/loss statements from the past 10 years, ASCAP/BMI/SESAC statements, record label royalty reports, information you use to put together your tax returns, SoundExchange statements, Quicken or Quickbooks files, CD Baby reports, etc. Basically any information that will help us understand your income streams.

We know this is very sensitive information, and we will treat it with the utmost confidence. We are NOT interested in reporting the numbers – we’re simply interested in the ratios between different revenue streams, and whether those ratios have changed over time for you. For example, are you making more on licensing, or less on merch, than you were 10 years ago?

What will you do with my financial information? We’re going to take the information gathered during the interview, and build what we’re calling “revenue pies” that show how each artists’ revenue is allocated among different revenue streams each year. For those who have multiple years of data, we will create a time series chart as well. Anything we generate, we’ll share with you. If we're able to collect enough data, we also hope to be able to display comparative data from other musicians.

Who else are you approaching? We are hoping to paint a broad picture of the revenue make up for composers and performers, so we’re asking composers and performers from different genres, and at different stages in their careers to participate. In our pilot phase, we reached out to sidemen, symphony composers, platinum selling rock bands, hip hop collage artists, string quartets, and jazz soloists. Many are coming to us as referrals from our networks. If you have someone to recommend, we’d love to meet them.

I don’t have an accountant or business manager, and that is a lot for me to pull together We are happy to do the organizing for you. You can give us a pile or a box of papers for example – and it’s easy for us to pull what we need from it and send all the information

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 24 back to you neatly organized. We can come to your home to pick it up and ship it back to you. Or, if you can’t do it for the past 4-10 years, do you think you can give us information from the last year? Even seeing your taxes and information from 2009 can be helpful – because we can build a current pie from that information.

I’m not sure I’m a good example The deeper we look into this area, the clearer it becomes that there is no typical musician or composer. Unfortunately, there is a lot of misleading information out there, and journalists and pundits frequently try to generalize or suggest there is an average experience with respect to artist income – almost always without any data to back them up. We are hoping to counter this information and provide policymakers – and the industry – with a new perspective and some honest and thought-provoking data.

What are you expecting to prove with this information? FMC isn’t out to “prove” anything. Like all researchers, we have hypotheses that we’re testing. We don’t know the outcome of the research, so it’s hard to say what the conclusions will be.

We are clear about why we think this work is important. There have been meteoric transformations in how musicians create and distribute their music over the past ten years, and significant disruptions to the traditional music industry business model. Many observers are quick to categorize these structural changes as positive improvements for musicians across the board, but we think this is painting with too broad a brush. To this point, measurements about the effect of these new technologies on musicians’ ability to make a living have either been anecdotal or speculative. Even the most esteemed books about copyright in the digital age are largely based on theory, and lack qualitative data.

FMC’s research goal is to examine the issue from many different musicians’ perspectives, so we can get a more robust and realistic sense of what it’s like to be a working musician in the 21st century. FMC believes that this proposed research is essential to understanding how changes in the music landscape have really affected musicians, and how they could affect the arts and culture field in general, that will inform the development of successful models in the future.

Questions? Contact project co-managers, FMC’s Education Director Kristin Thomson and FMC’s Director of Programs, Jean Cook.

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 25 Appendix E: The 42 streams

In 2009, FMC began to catalog all of the ways that musicians and composers can earn money from their compositions, sound recordings, performances or brand. This first list of 29 Streams was posted on the FMC website in October 2009.

After gathering feedback from experts, this initial list was corrected and expanded, resulting in a longer and more organized 42 Revenue Streams. The list below is simultaneously the genesis of the Artist Revenue Streams Project, a framework through which we categorized income streams and presented questions to participants, and an outcome of the research. This list, plus a companion Revenue Streams: Existing, Expanded and New are available on the Artist Revenue Streams website, and remain the most visited elements of this project.

SONGWRITER AND COMPOSER REVENUE

Publisher Advance Bulk payment to songwriter/composer as part of a publishing deal. Paid to songwriter/composer by publishing company.

Mechanical Royalties Royalties generated through the reproduction of recordings of your songs – physical or digital. Paid to songwriter/composer by publisher, label, Harry Fox, or digital aggregator like CD Baby/TuneCore.

Commissions Typically a request from an ensemble, presenter, orchestra or other entity for a composer to create an original work for them.

Public Performance (PRO) Royalties Revenue generated when your songs are played on radio, TV, in clubs and restaurants. Paid to songwriter/composer/publisher by ASCAP/BMI/SESAC.

Streaming Mechanical Royalties Revenue generated when your songs are streamed on on-demand services (Rhapsody, Spotify, Rdio). Paid to publisher by Harry Fox or other mechanical licensing agent. Publisher then pays songwriter/composer.

Composing Original Works for Broadcast Typically a commercial request to compose an *original* jingle, soundtrack, score, or other musical work for a film, TV or cable show, or an ad agency. Paid to songwriter/composer by agency requesting the work.

Synch Licenses Typically involves licensing an *existing* work for use in a movie, documentary, TV, video games, internet, or a commercial. Paid to songwriters/composers either via publisher or record label, or via a direct licensing deal with the licensee (movie studio, ad agency, etc) if you are self- published.

Sheet Music Sales Revenue generated by the sale or licensed reproduction of songs/compositions as sheet music. Paid to songwriter/composer by publisher, or directly from purchasers if you are selling it on your website or at performances.

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 26 Lyric Display Revenue generated by the licensed display of song lyrics. Online lyric sites pay publishers, which then pay songwriter/composer.

Ringtones Revenue generated from licensing your songs/compositions for use as ringtones. Paid to songwriter/composer via your publisher, your label or Harry Fox.

Songwriter Awards Programs Awards given by ASCAP and BMI Foundation to writer members of any genre whose performances are primarily in venues outside of broadcast media.

Publisher Settlement Payment from publishers to writers for litigation settlements.

RECORDING ARTIST REVENUE

Record Label Advance Paid to artist as part of signing a deal.

Record Label Support Money from label for recording or tour support.

Retail Sales Revenue generated from selling physical music in retail stores or via mailorder. Paid to artist/performer by your label, or services like CD Baby or Bandcamp that help musicians sell physical product.

Digital Sales Revenue generated from selling music digitally/online. Paid to artist/performer by your label, or digital aggregator like CD Baby or TuneCore, or directly from fans via services like Bandcamp.

Sales at Shows Revenue generated from selling recordings of music at shows/live performances. Paid to artist/performer directly by fans.

Interactive Service Payments Revenue generated when your music is streamed on on-demand services (Rhapsody, Spotify, Rdio). Paid to performer by your label, or digital aggregator like CD Baby or TuneCore.

Digital Performance Royalties Revenue generated when your sound recordings are played on internet radio, Sirius XM, Pandora. Paid to performers by SoundExchange.

Master Use Synch Licenses Typically involves licensing an *existing* sound recording for use in a movie, documentary, TV, video games, internet, or a commercial. License fee paid to record label and/or recording artist via a direct licensing deal with the licensee (movie studio, ad agency, etc).

AARC Royalties Collected for digital recording of your songs, foreign private copying levies, and foreign record rental royalties, distributed to US artists by AARC.

Neighboring Rights Royalties Collected for the foreign performance of your recordings.

Film Musicians Secondary Markets Fund Paid to performers on recordings used in film, TV and other secondary uses by the Film Musicians’ Secondary Markets Fund.

Sound Recording Special Payments Fund Paid to performers for the sales of music recorded under AFM collective bargaining agreement by the Sound Recordings Special Payments Fund.

SAG-AFTRA Contingent Scale Payments Paid to non-royalty artists when a recording hits certain

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 27 sales plateaus

Label Settlements Payments from labels to recording artists for litigation settlements (MP3.com, Limewire).

PERFORMER AND SESSION MUSICIAN REVENUE

Salary as Member of Orchestra or Ensemble Income earned as a salaried member of an orchestra or ensemble

Shows/Performance Fees Revenue generated from playing in a live setting (for non-salaried players).

Session Musician/Sideman Fees for Studio Work Payments to studio musicians/freelancers/ sideman for work in recording studio. Paid by label, producer or artist, depending on situation.

Session Musician/Sideman Fees for Live Work Payments to studio musicians/freelancers/sideman for work in a live setting/on tour. Paid by label, producer or artist, depending on situation.

Non-Featured Artist Payments Payments from the AFM & SAG-AFTRA Intellectual Property Rights Distribution Fund, which distributes recording and performance royalties to non-featured artists.

KNOWLEDGE OF CRAFT: TEACHING AND PRODUCING

Music Teacher Revenue generated from teaching your musical craft.

Producer Payment for producing another artists' work in the studio or in a live setting. Paid by labels, featured artists, studios, presenters, or foundations, depending on situation.

Honoraria or Speakers Fees Payment for conducting a lecture, workshop or master class. Usually paid by school, conservatory, or presenting organization.

BRAND-RELATED REVENUE

Merchandise Sales Revenue generated from selling branded merchandise (t-shirts, hoodies, posters, etc.). Paid to artist/performer by fans.

Fan Club Money directly from fans who are subscribing to your fan club.

YouTube Partner Program Shared advertising revenue, paid to partners by YouTube.

Ad Revenue Miscellaneous income generated by your website properties (click-thrus, commissions on Amazon sales, etc.)

Persona Licensing Payments from a brand that is licensing your name or likeness (video games, comic books).

Product Endorsements Payments or free goods from a brand for you endorsing or using their

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 28 product.

Acting Payments for appearances in TV, commercials, movies.

FAN, CORPORATE AND FOUNDATION FUNDING and OTHER SOURCES OF REVENUE

Fan Funding Money directly from fans to support or pre-sell an upcoming recording project or tour (Kickstarter, Pledge Music, IndieGogo).

Sponsorship Corporate support for a tour, or for your band/ensemble.

Grants Foundation or public arts grants to support your work/project from foundations, state or federal agencies.

Arts Administrator Money paid to you specifically for managing the administrative aspects of a group that you are a member of.

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 29 Appendix F: Population definitions and sample size for survey

Population of Study

FMC seeks to study US-based musicians. In order to ensure that we’re focusing on musicians that have some credentials in the music community, all research participants must meet the following criteria:

1. US citizen or permanent resident 2. 18 + 3. Creative or technical credit on at least 6 commercially released tracks – physical or digital. Qualifying tracks can be on one album or on a combination of albums.

“Credit on 6 commercially released tracks" is the standard that the Recording Academy uses for membership. Practically, this means the musician has released something commercially, but the bar isn't set so high that only career musicians will qualify for research participation.

In addition, we will ask survey participants:

4. Whether they are member of one or more of the following organizations: ASCAP, BMI, SESAC, SoundExchange, AFM, AFTRA, AGMA, Recording Academy, Songwriters Guild, American Composers Forum, American Music Center, Meet the Composer, Chamber Music America, Americana Music Association, Folk Alliance, Just Plain Folks, Gospel Music Association, Blues Music Association, International Bluegrass Music Association, Fractured Atlas 5. How much of their time they spend being a musician 6. How much of their annual personal income is derived from being a musician.

While specific answers to questions 4, 5 or 6 would not eliminate any particular participant, we would isolate any responses for individuals who are:  not a member of any professional organizations; or  spend 0% of their time and make 0% of their money being a musician.

These definitions not only help us to determine the types of musicians taking part in our research, but they also help us to estimate population size and determine a valid sample size.

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 30 Risks and limitations to definition

FMC recognizes that adopting these criteria to define our field of study does mean that some musicians may be excluded.

FMC is comfortable limiting this by citizenship status and age. We’re also comfortable with the requirement that participants have at least 6 credits on commercial releases. As noted above, this aligns with the definitions established by the Recording Academy, and the threshold is low enough that we’re able to focus on musicians that have even a small body of commercially-released work.

We are aware that some musicians that would qualify otherwise are not members of any professional organizations. While we will offer a fairly extensive list of professional organizations, there are certain genres of music that do not have a trade group or service organization through which practitioners gather. Jazz musicians are a particular worry, since there is no one organization that represents their interests, and many of them make their living playing other composers’ music (thus making membership of ASCAP, BMI and SESAC less relevant). In addition, there are dozens of sub-genres of music – electronic, blues, metal, hip-hop – in which musicians aren’t participating in formal organizational structures.

To reduce the effect of this limitation, FMC would simply make the questions about membership non-disqualifying. In other words, we will ask participants about their memberships, but it will be a factor that we can use to sort and filter our results, as opposed to a strict requirement for participation.

Reducing Self-Reporting Problems

FMC recognizes that we are relying on our survey respondents to give us accurate information about their role as musicians. While we are able to control and assess who we speak to in the interview and financial documents stage, we are aware of the problems of self-reporting during the online survey stage.

FMC expects to minimize the problems of self-reporting by creating a number of different criteria that survey respondents must meet in order for their survey responses to be counted. In addition to the required criteria listed above (US Citizen, 18+, 6 recording credits), we will carefully look at the answers to questions 4, 5, 6 noted above. Any respondent who is deficient in more than one area will be tagged, and their credentials for recording credits will be checked manually.

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 31 Estimate on population size

As stated earlier, one of the challenges in doing this work is estimating our population size. There is no definition for “musician”, nor certifications or qualifying tests. In addition, there is no one organization that represents the majority of musicians.

According to the Bureau of Labor Statistics, there were 14,330 “music directors and composers” and 47,260 “musicians and singers” in May 2009. However, the BLS statistics do not include the self-employed, which we know many musicians and composers are.

FMC suggests additional metrics to estimate the size of our population. The membership organizations have the following totals:

ASCAP: 400,000 BMI: 350,000 SESAC: 34,000 (verify with SESAC)

Songwriters and composers tend to pick one of these three exclusively, so the possible total is 800,000, however (1) many folks are both the songwriter and the publisher, and therefore have two accounts at each PRO and (2) the PRO membership also includes many estates of now deceased songwriters and composers who still earn royalties. Though it’s impossible to know precisely, we’ll estimate that the number of songwriters/composers (who are alive) is 400,000.

SoundExchange: 45,000 AFM: 90,000 AFTRA: 10,000 recording artist members Recording Academy: 12,000 voting members Folk Alliance: 2,000

If each organization had exclusive membership numbers, the total would be 159,000. However, it’s safe to assume there’s some overlap between all of these performer-driven organizations, with some musicians belonging to two or more. If we assume a 25% overlap and some representations of estates, our field of study would be about 120,000.

There’s also significant overlap between the songwriter community and the performer community. So let’s say that 75% of performers also belong to one of the songwriter PROs:

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 32 400,000 + 30,000 (the 25% of performers who don’t belong to a PRO) = 430,000

So, for the time being, let’s determine our sample size based on the rough estimate that there are 430,000 US-based musicians that belong to one or more of the membership organizations listed in criteria 4.

Given the complexity of the survey and the diversity of our population, FMC will set a high target number of at least 1,000 respondents to an online survey, in order for us to discard or isolate any responses that don’t meet our criteria, and to give us a wider array of data for cross-tabulation.

Note that we greatly exceeded this survey target, with 5,371 completes.

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 33 Appendix G: Money from Music survey construction

FMC created a 200-question online survey called Money from Music that presented relevant questions about artists’ revenue streams for US-based musicians. FMC put the survey in the field from September 6 to October 28, 2011 to capture information about musicians’ revenue streams for 2011 and prior. We analyzed the results not only on the macro level, but also by genre, by time spent/income generated, and other cross-tab measures.

Survey platform

The ARS Team used the Platinum version of Survey Monkey to execute this survey. A leading online survey software platform, Survey Monkey offers various question types, display formats, skip logic and question piping. The Platinum level ensures unlimited questions, unlimited responses, enhanced security (SSL/HTTPS) and complete control over survey branding. Survey Monkey also offers reliable data collection and robust reporting, with the option of downloading collected data to Excel, .cvs, STATA or SPSS.

Survey construction

Because revenue streams are so complex, we needed to construct a lot of questions to capture all the possibilities. However, we took a number of steps to make it as easy as possible for musicians and composers to participate.

Skip logic

The survey used a significant amount of skip logic, which allowed respondents to skip past questions that were not relevant to them. All if/then directions and skip logic are noted in the PDF using Acrobat commenting tools.

Multiple paths

Survey respondents were also offered three paths through the survey. All respondents saw questions 1 – 17. After Q17, they were given one of three options:

1. Short: Respondents answered critical revenue questions, then were skipped to Q192 for some “big picture” questions.

2. Medium: After Q17, respondents were presented with questions related to their primary role as a musician, whether that was being a (1)

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 34 composer/songwriter, (2) recording artist, (3) performer (including salaried player), (4) session player or (5) teacher, then to Q192 for some “big picture” questions.

3. Long: After Q17, respondents were presented with questions related to each role they played as a musician, then were reunited with other survey takers at Q192 for the “big picture” questions at the end.

Results

Over 7,000 respondents started the survey, and 5,371 completed it.

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 35 Appendix H: Money from Music survey protocol/instructions

September 6, 2011

This information is also posted on the ARS website.

Goal: To gather crucial information about the ways that US-based musicians and composers are currently generating income from compositions, recordings, performances or brand, and how this has changed over the past five years.

Instrument: FMC used the premium version of Survey Monkey to field this survey.

Question set: The question set was developed between Oct 2010 and Aug 2011 by FMC and the Artist Revenue Streams research advisory committee, with additional advice from outside experts and printed sources.

Field testing: The question set was tested by about 45 different musicians and composers during two different testing stages (June 2011 and late July 2011). Feedback from the testers, which was gathered through feedback questionnaires, follow up calls, in-person interviews, and meetings, was highly influential on the final shape of the survey.

Collection period: The online survey was open from Sept 6 – Oct 28, 2011.

Capacity: There was no limit on the number of participants who could complete it. There was also no restrictions on the number of participants who could take the survey simultaneously.

Completions: High completion rates were critical to the success of this project. Thanks to widespread marketing, thousands of musicians and composers — working in many genres and many roles — completed it, so we have robust data to examine, filter and cross-tabulate.

Marketing: FMC implemented a robust and multi-dimensional marketing plan, which included paid advertising (print and web), op-eds and guest blogs, earned media, podcasts, events, interviews, and social media. We also cultivated partnerships with over 200 organizations and companies that serve musicians and composers, encouraging them to engage with their networks.

Eligible participants: This survey was for US-based musicians and composers who are 18 years of age or older.

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 36 Preparation: Many questions related to how survey respondents earn money as a musician and/or composer. Before they started, we urged survey takers to have financial documents or solid estimates of how they earned their musician-based money in the last year, as well as access to a calculator for one of the questions.

Accommodations: For participants who were unable to take the survey online, either because they don’t have internet access or they have difficulty with screen environments, FMC project staff was willing to schedule a time to administer the survey via phone to the participant.

Instructions: Survey participants were shown instructions on an introductory page. We also created a 3 minute video introduction that’s available on YouTube or Vimeo.

Anonymity: Survey respondents’ participation was both voluntary and anonymous. There were no questions where their response would identify them as an individual, and their data was aggregated with that of thousands of other musicians and composers. We did not collect names or email addresses in the survey itself.

Navigation: The survey did not “time out” if respondents stopped answering questions. As long as survey respondents kept their browser windows open, the survey accepted their answers.

Time estimate and options: Because the music industry is so complex, this survey was thorough. However, we took a number of steps to make it as easy as possible to participate:  Early on, survey respondents had a chance to pick a short (10 mins), medium (20 mins) or long path (30 mins) through the remainder of the questions.  If they didn’t participate in particular revenue streams or roles, they were skipped past those questions.  Throughout the survey, we provided links to a Google doc with definitions of music-industry specific words and terms that may be difficult to understand.

Incentives: In order to encourage participation and increase completion rates, we offered some incentives to survey takers, based on the path they chose to take.  Those who completed the long path were eligible for a random drawing for one of four Apple iPad 2′s (value $500).  Those who completed the medium path were eligible for a random drawing for one of 100 $10 gift certificates to Amazon or Guitar Center.  The first 100 to completed the short path were eligible to receive a discounted entrance to Future of Music Policy Summit in Washington, DC Oct 3-4, 2011.

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 37 Random drawing process: Participants who completed the long and medium paths were presented with a 4-digit PIN on a specific page inside their path. They were encouraged to write down this PIN number for later. After they completed the survey, all respondents were automatically re-directed to a separate follow up questionnaire, which asked them to enter their PIN if they would like to be added to the appropriate raffle. If yes, they had to leave a name and email address in this questionnaire. Because this is a separate data collection instrument, their email and name were not associated with their survey answers.

The random drawings took place on November 10, 2011. FMC used a random digit generator to choose the 4 recipients of Apple iPad 2, and the 100 recipients of $10 gift cards. Winners were contacted by email. We also made a video of the drawing for the iPads:

Results: This survey — along with interview and financial case study data — provides a rich snapshot of the complex nature of being a musician in the 21st century. We shared the data with organizations, advocates and musicians nationwide in 2012, and we hope it will help us to ensure that policymakers and the public understand the financial realities of musicians today.

If you have technical questions or feedback about survey design, please email [email protected].

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 38 Appendix I: Partnerships, Connectors and Advisors

Partnerships for participation

The success of this work depended greatly on the level of participation we could achieve among musicians. But we also knew that musicians are a difficult population to count and measure. There is no clear definition of what a “musician” is, and there is no one association or organization that represents their collective interests. Because of the dispersion, reaching these various groups took outreach to dozens of different liaisons: music managers, booking agents, record labels, performance rights organizations, unions, technology companies, social networking sites, service organizations and music schools – connections that FMC has been steadily developing for the past ten years through our event programming, collaboration on policy campaigns, educational efforts, and other shared values. FMC sought the collaborative support from dozens of member associations and service organizations, which in turn urged their members to participate in the survey.

Here are the organizations and associations that we reached out to before, during and after our research:

Performance Rights Organizations  ASCAP  BMI  SESAC  SoundExchange  AARC  AFTRA and AFM Payment Fund

Unions  AFM  AFTRA  AGMA  Actors Equity  Film Musicians Secondary Market Fund  Musicians Special Payments Fund  Recording Musicians Association

Labels  RIAA

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 39  A2IM  Direct connections with specific labels

Membership and Service Organizations  American Composers Forum  American Music Center  Americana Music Association  Chamber Music America  Folk Alliance  Fractured Atlas  League of American Orchestras  Meet the Composer  Nashville Songwriters Association  National Assoc of Performing Arts Managers and Agents  Opera America  Songwriters Guild of America  The Recording Academy  Country Music Association

Other connectors  Artist alumni from FMC’s Artist Activism Camps  Managers – we have working relationships with about 50 artist managers, plus some business managers  Booking agents – we have working relationships with about 20 booking agents  Aggregator services, such as TuneCore, CD Baby, RightsFlow, IODA

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 40 Appendix J: Research Advisory Committee

January 2011

FMC convened a 13-person advisory committee of social researchers, academics, and key music industry folks who supported and advised the project in the areas of research protocols, survey questions, and project scope. Our objective in assembling this advisory committee was to ensure this project research was as robust and meaningful as possible.

Bryan Calhoun, Vice President, New Media and External Affairs, SoundExchange Ann Chaitovitz Dan Cornfield, Sociology, Vanderbilt Peter DiCola, Assistant Professor, Northwestern University Law School Joe Karaganis, Vice President, The American Assembly Mary Madden, Senior Research Specialist, Pew Internet and American Life Project Barry Massarsky and Nari Matsuura Massarsky Consulting Ian Moss, Research Director, Fractured Atlas, and composer and performer Maria Rosario Jackson, Director, Culture, Creativity and Communities Program, Urban Institute Paul Sharpe Director of Freelance Services, American Federation of Musicians of the US and Canada John P. Strohm, Attorney, Johnston Barton and Proctor and Rose, LLP, and musician David Touve, Assistant Professor of Strategy and Entrepreneurship, Washington and Lee University Danilo Yanich, University of Delaware, expert on content analysis and survey design

Additional academic colleagues who helped us throughout the project

Steve Tepper, Curb Center at Vanderbilt Elizabeth Long Lingo, Curb Center at Vanderbilt Jen Lena, Vanderbilt Tina Piper, McGill University Jane Ginsburg, Columbia University Joan Jeffri, Columbia University John Kellogg, Berklee College of Music Jessica Litman, University of Michigan Kembrew McLeod, University of Iowa Walter McDonough, Suffolk University

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 41 Gabriel Rossman, UCLA Chris Bavitz, Harvard Law School/Berkman Center Devon Powers, Cultural Policy, Drexel University

Academic Partners

FMC sees considerable long-term value of this project, and hopes to identify an institutional partner that might be interested in housing a replicable study. Some US- based universities that do similar work are:

Belmont University Music Industry Program Berklee College of Music Curb Center at Vanderbilt CSU Chico Drexel University’s Music Industry Program Loyola University (New Orleans) Georgetown University Harvard University: Berkman Center Clive Davis School at New York University Northeastern University Northwestern University Suffolk University University of Chicago Center for Cultural Policy Joel A. Katz Music and Entertainment Business Program at Kennesaw State University University of Maryland William Paterson University Martin Bandier School at Syracuse University

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 42 Appendix K: Timeline

September 2009 – April 2010  Develop research proposal and examine protocols  Meet with social researcher peers to discuss work  Seek financial support from foundations and institutions

May – August 2010  Develop interview protocol  Pilot seed interviews and financial data collection  Continue conversations with peers  Continue to seek funding  Develop research advisory committee

September – December 2010  Continue with interviews based on respondent referrals  Continue collecting financial data  Begin coding interviews  Begin meeting with key US member associations to get their support for work  Begin meeting with possible sources of data  Continue invites for research advisory committee

January 2011 – April 2011  Continue with interviews based on respondent referrals  More formal development of survey instrument  Continue to work with partners to gain support for survey  SXSW and Re-Think get partner feedback about survey Q set  Continue coding interviews  Launch demo interactive charts

May 2011 – August 2011  Continue with interviews and financial data review  Finalize survey instrument  Implement survey marketing plan

September – October 2011  Complete interview phase  Launch survey in field with support from as many member organizations as possible

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 43

November – December 2011  Begin writing reports that combine survey data and information gathered from interviews

January 2012 – December 2014  Continue to write reports and present findings  Pursue additional funding to replicate work  Reach out to academic partners  Build data portal to expose survey data

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 44 Appendix L: Project privacy and permissions policy

January 2011

The following outlines our privacy policy protocol for the Artist Revenue Streams Project. All artists we work with default to this protocol unless we have their express permission to do otherwise.

Permissions

At each stage in the process, there will be an opportunity to check in with the artist before we develop a quote or chart or other info about an artist that we potentially circulate outside of FMC. At that time we will reiterate our privacy and permissions policy as it pertains to that piece of information and check with them how they want themselves attributed.

Permission to waive an artist’s privacy on one item does not necessarily imply permission waived on all items. At the same time, we will coordinate and use our best judgment so we don’t bombard the artist with permissions requests.

Artist Interviews

Artists who are interviewed are asked to review and sign a consent form, which outlines FMC’s responsibilities and their participation in the project.

All interviews are recorded and then transcribed by Chad Molter (our project transcriber), then uploaded to Basecamp. The only reason we record interviews is for accurate transcription. Kristin, Rebecca and Jean are the only ones that read the transcriptions, and we code them and pull quotes from them for our narrative report.

The names of artists interviewed are confidential at this point. Any artist who is quoted or referenced in the final report narrative will be sent a draft copy for their review. At that point, the artist will have a choice of how they are listed: by their real name, or by a pseudonym.

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 45 Artist Financial Data

A handful of artists have agreed to give us access to their musician-based financial data. This data is handled with the utmost care and discretion. The only people who see financials for artists are Kristin, Rebecca and Jean.

From the data table we will create pie charts and stacked column graphs. The pie charts will show the ratio of revenue streams in a given year. The data labels will show the category name and percentage of annual income, not the actual dollar figures. Similarly, on the stacked column graph, data will illustrate year to year changes. However, these will have no dollar amounts associated with them in any way, and the Y axis will not have any values. Data and charts will be sent to the artists to review before we release them.

Internal Coding

All interviewees and artists that we develop case studies for are given a code (currently letters – A, B, C, D etc), which we use to refer to the artist in case studies and in data tables. The only people who know the real names of individuals we are interviewing, and for whom we have financials, are FMC project staff Kristin, Jean and Rebecca.

General Communication

When we reach a major project milestone – such as a report release, interactive chart launch, blog post – that incorporates information from an artist that we have interviewed or been in touch with directly, we will give them a heads up before it goes public.

Survey Respondents

Online survey participants will be anonymous, and their responses confidential. Anonymous answers will not be traced back to the respondent. No personally identifiable information will be captured unless the respondent voluntarily offers personal or contact information in any of the comment fields. Additionally, responses will be combined with those of many others and summarized in a report to further protect anonymity. There may be a point in the survey where we ask participants if they would like to contribute financial records to the project, and if yes we will ask them for their contact information, but this would be completely optional.

Artist Revenue Streams: post-project research protocol. Draft for VMBR attendees. 46