The Boycott Was Never the Threat. The Playlist Was.

Based on the research ofDaniel Winkler, Nils Wlömert, and Jūra Liaukonytė, "Separating the Artist from the Art: Social Media Boycotts, Platform Sanctions, and Music Consumption," Journal of Marketing Research, 2026

Daniel Winkler, Nils Wlömert, and Jūra Liaukonytė tracked what actually happened to streaming demand around several of the last decade's biggest music-industry scandals, including R. Kelly, Morgan Wallen, Rammstein, and Sean "Diddy" Combs. Their finding cuts against how both sides of the culture war describe cancellation. They found no evidence that public backlash alone produced a sustained drop in an artist's streams when the platform kept the artist's music visible in playlists and recommendations. In several cases, the scandal itself coincided with short-term increases in listening. The declines that did stick appeared only where platforms actively pulled an artist out of curated and algorithmic discovery. The moral judgment of millions of individual listeners, it turns out, is not what determines whether a scandal-hit artist keeps getting played. A ranking decision made inside a streaming company is.

The Policy That Outran the Outcry

R. Kelly is the paper's clearest case because the mechanism is so exposed. In April 2018, a coalition of women of color within the Time's Up movement publicly called on labels and streaming services to cut ties with Kelly over long-standing allegations of sexual abuse. Spotify's response, announced May 10, 2018, was not to ban his music outright but to remove it from the platform's own curated and algorithmic playlists under a new "Hate Content and Hateful Conduct" policy, his songs stayed available if a listener searched for them directly, they simply stopped surfacing anywhere else. Winkler, Wlömert, and Liaukonytė's data shows streams of Kelly's most popular songs fell by around 25 percent afterward. But the paper's more pointed result is about attribution: for Kelly songs that were not pulled from Spotify's curated playlists, the authors found no comparable pullback in listening. As Liaukonyte put it, the drop "was driven primarily by reduced platform visibility... the platform's discovery tools made it harder for listeners to encounter R. Kelly's music," not by fans deciding, en masse, to stop choosing him.

The Same Lever, Aimed at a Bigger Fanbase

Three years later, Spotify and its rivals reached for the identical tool against a very different artist, and it did not work at all. When a video surfaced in February 2021 of country star Morgan Wallen using a racial slur, the backlash was immediate and structurally similar to Kelly's case: radio stations dropped him, the Academy of Country Music disqualified his album from award consideration, and streaming services pulled him from their curated playlists, 14 on Spotify and 21 on Apple Music, more than 30 total across Spotify, Apple Music, Amazon Music, and Pandora. By the theory that visibility loss drives listening loss, Wallen's numbers should have sunk the way Kelly's did. Instead, in the week ending February 4, 2021, sales of his album "Dangerous: The Double Album" rose 102 percent to roughly 25,000 copies, song downloads climbed 67 percent, and on-demand streaming ticked up 3 percent to around 160 million streams, according to MRC Data figures reported by Billboard and Rolling Stone. Country radio and playlist curators had cut off the passive discovery channel completely, and Wallen's audience simply routed around it.

Why One Scandal Sinks and the Other Floats

The paper's answer is that platform delisting only works on the kind of listening it actually touches: passive, discovery-driven plays generated by playlists and recommendation algorithms. It has no purchase on intentional listening, a fan opening the app and searching for an artist by name, adding a song to a personal library, or buying an album outright. Kelly's post-scandal audience leaned heavily on the passive channel the platform controlled; once Spotify's algorithm stopped surfacing him, a large share of his listening had nowhere to regenerate. Wallen's fanbase did the opposite: mobilized, loyal, and in some documented cases explicitly buying and streaming the album as a show of support against what they read as unfair punishment, they supplied intentional demand large enough to swamp the lost discovery traffic. The delisting mechanism is real and measurable, but it is not a universal switch. It is a tool that removes one channel of demand, and its total effect on an artist's numbers depends entirely on how much of that artist's audience was ever using that channel to begin with.

Streaming platforms hold real power over which scandals fade and which explode, but only over listeners who were never searching for the artist in the first place.

The Lever Is Political, Not Just Technical

Spotify's own history with the Kelly decision shows how unstable this governance tool actually is. The Hate Content and Hateful Conduct policy lasted barely three weeks: after musicians and industry groups objected that the language was vague and its enforcement looked selective and racially uneven, Spotify rolled back the "conduct" half of the policy on June 1, 2018, stating plainly that "our role is not to regulate artists." Yet the reversal was itself selective. Rapper XXXTentacion, removed under the same policy, was quietly reinstated to playlists; Kelly was not. A tool marketed as a neutral, algorithmic response to misconduct turned out to be adjudicated case by case, under public pressure, with outcomes that did not track consistently with the severity or nature of the underlying allegations. Any platform that leans on delisting as a governance response should expect the same dynamic: the decision of who gets cut off from discovery, and who gets quietly restored, becomes a political artifact of the platform's own choosing, not a mechanical output of its content-moderation policy.

What Any Attention Platform Should Take From This

The mechanism here is not specific to music. Any platform whose product mixes algorithmic discovery with direct search, an app store, a social feed, a marketplace, faces the identical split between passive and intentional demand, and the identical asymmetry in how much control it has over each. Shadow-banning a creator, delisting a seller from search results, or excluding a developer from a featured-apps carousel will visibly suppress activity among users who were never seeking that party out by name; it will do essentially nothing to a base of users who already know exactly what they want and can type it into a search bar. Before treating a visibility penalty as an effective sanction, an operator needs a real answer to how much of the target's audience is discovery-dependent versus intentional, because a scandal that looks identical in the headlines can produce a 25 percent collapse in one case and a sales spike in the other, and the difference is not measured in outrage. It is measured in whether the audience was searching before the platform ever touched a single recommendation.