The Marvel effect: cinematic universes and their impact on box office receipts

ABSTRACT

This paper investigates the effects of cinematic universes in films on box office receipts using data from 250 films released between 2009 and 2019. Regarding the non-Marvel Cinematic Universes, there is a statistically significant impact on the opening weekend box office receipts, but the same is not true with respect to a filmʻs initial run. Alternatively, the Marvel Cinematic Universe has a statistically significant effect on box office receipts over opening weekend and during the filmʻs first run. In addition, female star power remains statistically significant with the inclusion of more independent variables while male and director star power become statistically insignificant regarding the opening weekend box office receipts.

KEYWORDS:

• Motion picture industry
• box-office revenues
• consumer preferences
• cinematic universe
• Marvel

JEL CLASSIFICATION:

• L10
• L80
• L82

Acknowledgments

I would like to thank Prof. Heather L.R. Tierney for her generous guidance and comments and the participants of the 2021 Annual MBAA International Conference for their helpful comments and suggestions as well.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Oxfordʻs A Dictionary of Film Studies defines a cinematic universe as “The extension of a filmic story world across a range of films, via characters and stories co-existing and interacting from one film to the next.” (Kuhn and Westwell 2020).

2 The links to the data websites are: www.the-numbers.com and https://pro.imdb.com.

3 Films that do not have a prominent male or prominent female role are left blank for this variable, which is interpreted as zero in the regression equation.

4 The review aggregator Rotten Tomatoesʻ website is: rottentomatoes.com.

5 Tarantinoverse refers to movies written and directed by Quentin Tarantino, which are connected in one form or another (Miller 2019).

6 Pre Aud = Prev Aud 2 + Cinema Univ All = Prev Aud 2 + Cinema Universe 2 + Marvel.

7 The empirical results for this paper have been estimated using Stata. If the collinearity between two variable is too high, then Stata will drop one of the variables automatically. No variables have been dropped for any of the regressions for this paper.

8 Elberse (2007) uses this measure because it matches the StarBond calculation on the Hollywood Stock Exchange.