Journal of Cultural Economics 22: 249–270, 1998.
© 1998 Kluwer Academic Publishers. Printed in the Netherlands. 249
Movie Stars and the Distribution of Financially
Successful Films in the Motion Picture Industry
STEVEN ALBERT
The Open University Business School, Walton Hall, Milton Keynes, MK76AA, U.K.
Abstract. Audiences choose to see films using information from previous films. A stochastic con-
sumer choice process based on this assumption generates a particular distribution of financially
successful films among film types. Movie stars can be used to mark these successful film types.
Thus, their star power originates not only from “box-office appeal” but also from “marking power”.
Evidence of 960 top 20 films released in the United States and Canada between 1940–1955 and
1960–1995 is consistent with this model.
Key words: Hollywood, film, consumer choice, movie stars, star-power
1. Introduction
It has been argued that in the 1930s the popularity of films starring Mae West saved
Paramount Studios from financial ruin, and that a succession of Deanna Durbin hits
similarly kept Universal Studios afloat (Walker, 1974; Morin, 1960). In the 1940s
and early 50s, a majority of MGM’s financial success was due to a string of films
starring Clark Gable (Shipman, 1979). Stars of the 1990s who, like Harrison Ford,
are attached to a series of financial hits are still highly sought after inputs (Brown,
1995). Indeed, today the power wielded by movie stars underscores much of the
motion picture industry. Producers, directors, writers and the stars themselves un-
derstand this. Saul Zaentz, independent producer of One Flew Over the Cuckoo’s
Nest and The English Patient relates, “The studios? . . . the first question is, ‘who’s
in it?’ ” (The South Bank Show, 1997). Says John Landis, director of hits The
Blues Brothers and Coming to America, “. . . You ask them, ‘What do you think of
the script?’ And they answer, ‘If you can get Harrison Ford then it is a good script.
If you can’t then it’s a bad one’ ” (Thomas, 1996). Gary Fleder, director of Things
to Do in Denver When You’re Dead, comments on how he was able to finance this
hip, off-beat U.S. independent thriller: ‘. . . actors embrace it, and Miramax em-
braces that . . . ” (Cherity, 1996). Peter Howitt, director and writer of Sliding Doors,
also understands the actor’s power to get movies made. He had no luck selling
his screenplay until “an agent showed the script to Gwyneth Paltrow. He thought
she would like it. She did. Suddenly every major studio wanted to distribute the
film” (Sunday Times, 1998). Superstar Barbara Streisand comments, “the audience
buys my work because I have complete control” (Vincent, 1996), while Robert

250 STEVEN ALBERT
DeNiro describes his power to get films made for his long time colleague and
friend, acclaimed director Martin Scorsese: “I told Marty, if he had trouble getting
it off the ground and if he really wanted me to do it [The Last Temptation of Christ],
I would’ve done it. It was something I was telling him as a friend; if he needed me,
I was there” (Paris, 1989).
However, no star guarantees a successful film. For example, the films Havana
and The Two Jakes starred Hollywood giants Robert Redford and Jack Nicholson,
respectively, and yet, both were financial failures. With this seeming paradox in
mind, we must ask: Is there any evidence to support a system that relies so heavily
on Hollywood movie stars?
Past economic and statistical analyses seem to answer, “no”. Recently, DeVany
and Walls (1996) have suggested that there is no known combination of release pat-
tern, star, story, etc. which will ensure a hit. They maintain that “the crucial factor
is just this: nobody knows what makes a hit or when it will happen” (p. 1493).
As far as stars are concerned, Kindem (1982) suggests only a weak impact of
marquee value (a combination of star and other factors) to box-office success, and
this only during the studio era. Simonet (1980) and Garrison (1971), both using
multiple regression analysis, conclude that movie stars, after the studio era, have
very little impact on box-office success. Miller and Shamsie (1996) suggest that
stars during the studio era, while under long-term contracts, are weakly associated
with revenues, while in the post-studio period they are not. Austin (1989) shows,
using an audience survey, that the presence of stars is unimportant when deciding
to attend a movie (See Austin, 1989; cited in Jowett and Linton, 1989, p. 39).
However, Prag and Casavant (1994) report a positive relation between the pres-
ence of stardom and revenues. They also report a positive relation between rev-
enues and “quality”, revenues and the film being a sequel, and revenues and the
film having won an Academy award. When advertising is also included in their
regression, the presence of stardom is no longer significant; only “advertising”,
“quality” and “sequel” have a positive relation with a film’s revenues. However,
every independent variable shown by Prag and Casavant to have a significant posi-
tive relation to revenues is also shown to be related to stars (including advertising).1
From this, one can argue that the relation between stars and revenues is the most
important conclusion of Prag and Casavant’s study.
Prag and Casavant, although contributing a very good start, use methods that,
perhaps because of the exploratory nature of their study, do not reveal the full
picture of why stars are so important to the film industry. Many questions remain,
like: What is the relation between stars and consumer behaviour? Is their any
particular way in which stars are related to successful films? Do particular stars
have more positive influence than others do? Why do stars appear to be the most
powerful influence over the film industry, in terms of which movie is made, who
will contribute to its production and how it will be advertised?
This paper explores movie stars and successful films from a perspective that
is significantly different from those cited above. Whereas those who employ re-