Explorando a dinâmica de popularidade de músicas sampleadas, regravadas e remixadas

Abstract

Digitalization of music has transformed its production, distribution, and consumption, introducing new forms of composition, among which "musical borrowings" stand out. In this practice, composers and producers use elements of pre-existing songs to create new works. This process often includes collaboration between artists, a strategy increasingly used to expand audiences and maintain success. However, these practices raise significant legal questions, particularly regarding copyright and fair use, with debates centered around the impact such borrowings may have on the market for the original work. Some studies suggest that musical borrowings can even boost sales of the original songs, indicating that, in certain cases, such practices could be classified as fair use. In this work, we investigate how samples, covers, and remixes affect the popularity of the involved songs. Using data from WhoSampled, which maps around 700,000 borrowings, we explore two main hypotheses: (i) derived songs capitalize on the popularity of the original work, and (ii) the release of a new version impacts the popularity of the original version. To test these hypotheses, we compare YouTube views and a Spotify popularity metric, applying Regression Discontinuity Design (RDD) to measure the Average Treatment Effect (ATE) in the short term. Granger Causality is used to examine the long-term impacts. The results indicate limited evidence that musical borrowings increase the popularity of the original works. The RDD analysis suggests that the release of new versions can boost interest in older songs. However, the long-term effects, as measured by Granger Causality, are harder to establish due to the influence of multiple factors on interest over time. Our findings provide new insights into the complex economics of musical borrowings and may contribute to fairer compensation in the music market.