Jake Linford & Aaron Perzanowski
Volume 75, Issue 2, 293-372
When Donald Trump descended the escalator of Trump Tower to announce his
2016 presidential bid, Neil Young’s “Rockin’ in the Free World” blared from the
loudspeakers. Almost immediately, Young’s management made clear that the
campaign’s use of the song was unauthorized. Neil Young was not alone. Trump
drew similar objections from dozens of artists during his first two presidential bids.
But as a matter of copyright law, it is unclear whether artists can prevent their
songs from being played at campaign rallies.
Putting the intricacies of copyright licensing aside, what motivates artists to object
to the use of their songs by political campaigns? This Article identifies and
measures three types of harm artists may reasonably fear. First, an artist may
worry that campaign use of their song will harm its market value and popularity.
To test that theory, we examine a novel set of industry streaming data to identify
any meaningful shifts in streaming consumption after well-publicized campaign
uses. Second, campaign use may falsely lead the public to believe that an artist
supports or endorses a candidate. And third, an artist may fear a tarnishment effect.
That is, consumers may negatively associate the artist or their music with an
unpopular candidate even in the absence of any perceived endorsement. We test
the endorsement and tarnishment theories through an experimental design that
measures consumer reactions to a set of hypothetical campaign uses.
Our data paint a complicated picture. We find some evidence that songs used by
the Trump campaign suffered a drop in streaming consumption, but we cannot
conclude that campaign use drove that reduced popularity. We also find strong
evidence that an artist’s perceived support or endorsement of a candidate is
material to consumers. But consumers do not appear to infer that an artist endorses
a candidate when their campaign uses that artist’s song. Finally, we found that less
well-established artists are most likely to suffer from tarnishing associations when
their songs are used by divisive politicians. Our results do not fully resolve the
thorny doctrinal and normative questions at the heart of these controversies, but
they do offer a crucial empirical grounding for a recurring policy debate.