The geography of music is a fascinating but understudied field of research in the creative economy. The presence of musicians is a key indicator of a location’s thriving creative environment; a signal of openness, diversity and creativity. These attributes manifest themselves unevenly across geography: some cities are specialists and home to one particular type of music, such as country in Nashville, while others host an increasingly eclectic collection of bands and genres such as New York. Music is a reflection of the people within a city and their creative expression. Music adds a large amount to a city’s creative attributes. Is there a relationship between geography and music? Or, is it a strategically good decision for musicians to move to a city with a strong scene in their genre or a vibrant, diverse music scene? And are more creative cities likely to specialize and succeed within one type of music or more likely to embody many types of genres? These are questions seldom answered. The goal of this paper is to test current assumptions and theories about established music clusters using detailed Myspace data from 2007, while also highlighting unexpected or emerging clusters of musical activity in the United States.

These data were initially gathered for the report, Chicago: Music City, using a custom algorithm created by Justin Savage. Details about the script and data gathering methods can be found in the report. Building on this initial work, Dan Silver, Assistant Professor of Sociology at the University of Toronto Scarborough Campus, who also contributed to the Chicago report, brought the data from Chicago to Toronto, and has been analyzing it from a number of angles, some of which were reported on the MPI Music Blog.

These Martin Prosperity Institute White Papers report preliminary descriptive findings using the Myspace data, examining metropolitan areas in the United States across several key variables, including the number of fans, plays, or views a particular genre is home to. The first white paper examines the Myspace data by metropolitan area, and the second by genre. Future research will examine several metropolitan areas in-depth, probing the musical composition of a particular music scene.

About the metrics
Below is a description of the key variables used for these Myspace data which were downloaded from Myspace.com on January 24 and 25, 2007. Myspace is an internationally known website popular for it’s social networking capabilities, of which a key use is for musicians seeking to promote their craft. There are over 3 million artists in this dataset, and 325 MSAs (based on 2000 definitions). It is worth noting, however, that a given band may or may not carry enough social credibility to be a band. Other recognized challenges with the data include: fan-generated duplicate band pages for popular bands, fake band pages, or bands sarcastically attributing a fake genre. it is also important to note that data cleaning was limited to addresses — specifically city name. Cities were identified using two digit state abbreviation and matched to a specific county by city name using the US Census Bureau’s place name file. Cities that were not matched directly, were next matched using SAS’s ‘soundex’ routine to match based on pronunciation, finally, all city names that had an identified state and were used 25 or more times, were manually investigated and matched to a US county when possible. Counties are then used to identify metropolitan areas.

  • Bands: The name of the band is as it was displayed on Myspace. It is important to note that not all bands are open to viewing, which can reduce the amount of data received. Also, most of the bandnames have odd characters in them, which is something Myspace allows them to do.
  • Plays: Myspace allows the band to post their music on the website and allow users to listen. The number of times users have played a band’s music is recorded under “Plays.”
  • Views: This is how many times a band’s Myspace page has been viewed.
  • Fans: On Myspace, each band acts an individual user. While a user has “friends”, bands have “Fans.” This is only the number of fans on the website.
  • Genre: Myspace allows a band to select up to three genres. On the page, they are displayed as Genre 1/Genre 2/ Genre 3. If the user chose to only enter one or two genres, the empty ones say “none”. This variable is the type of music the band self-identifies as. In this paper, genre is measured in two key ways. First, for a given location, the top twenty-five genres in that metro can be examined. Second,using the Ennis (1992) typology, we will present simplified genres.
  • City and State: This is the specified city and/or state. In a fair number of cases, users did not enter a city and/or state, or entered something that cannot be interpreted into a city and/or state except on a case-by-case basis. Of 2,485,564 band pages that were identified as being in the US, 2,331,357 were successfully matched to a US county. Of those, 154,207 were matched by ‘soundex’. Over 93.8% of bands were matched by city and state to a county.

Download this White Paper (PDF)