Best remake since Body Snatchers.

I wrote a paper on "acting" see below for a summary. I'm eager to hear your input on which subjects I didn't understand.


THE POWER OF STARS:
DO STAR ACTORS DRIVE THE SUCCESS OF MOVIES?
ABSTRACT (215 WORDS)
Is the involvement of star actors critical to the success of motion pictures? Film studios,
which they regularly pay multimillion-dollar fees to star actors, seem driven by that belief. I
shed light on the returns on this investment using an event study that considers the impact of
over 1,200 casting announcements on trading behavior in a simulated and real stock market
setting. I find evidence that the involvement of stars impacts movies’ expected theatrical
revenues, and I provide insight into the magnitude of that effect. For instance, my estimates
suggest that, on average, stars are worth about $3 million in theatrical revenues. In a crosssectional
analysis grounded in the literature on group dynamics, I also examine the
determinants of the magnitude of stars’ impact on expected revenues: among other things, I
show that the stronger a cast already is, the greater is the impact of a newly recruited star with
a track record of box office successes or with a strong artistic reputation. Finally, in an
extension to the study, I do not find that the involvement of stars in movies increases the
valuation of film companies that release those movies, thus providing insufficient grounds to
conclude that stars add more value than they capture. I discuss implications for managers in
the motion picture industry.
3
THE POWER OF STARS:
DO STAR ACTORS DRIVE THE SUCCESS OF MOVIES?
“'A guy stranded on an island' without Tom Hanks is not a
movie. With another actor, [the movie Cast Away] would gross
$40 million. With Tom Hanks it grossed $200 million.
There's no way to replace that kind of star power.”
Bill Mechanic, Former Chairman of
Twentieth Century Fox (Variety 2002)
“Had Troy opened impressively, one can be sure [Brad] Pitt
would have gotten the credit. In this way, movie stars are like
chief executives fattened on stock options. If the stock price goes
up, the boss reaps the rewards (…). If the stock price goes
down, well, then there must be some other reason.”
Dan Ackman, Reporter (Forbes 2004)
The importance of stars permeates the motion picture industry. While there are thousands of
aspiring actors and actresses, the small group that has risen to the top of their profession can
command fees of millions of dollars per movie in salaries, perks, and profit participation deals. A
handful of high-profile stars—including Jim Carrey, Tom Cruise, Tom Hanks, Brad Pitt, and Julia
Roberts—have been paid salaries as high as $25 million per picture. Profit participation
arrangements, so-called back-end deals, can sometimes amount to yet higher fees. Tom Cruise, for
example, reportedly earned more than $70 million—a 22% share of the total receipts—for the
movie Mission Impossible, and another $92 million for the sequel (Epstein 2005). Moreover, highprofile
stars can have a powerful influence on movie development: some even trigger the
“greenlight” by generating commitments from investors, producers, distributors, and exhibitors.
Is the involvement of star actors critical to the success of motion pictures? Are they worth
the star treatment? The returns on the investments in stars are heavily debated in the trade and
popular press. While some practitioners argue that stars are the “locomotives behind some of
Hollywood’s biggest blockbusters” (Variety 2002), many others doubt whether the high level of
compensation for stars is justified. A Forbes (2004) article entitled “The Myth of Brad Pitt”, which
compared more than 200 recent films, revealed that fewer than half of the highest-grossing hits
featured an actor who had top billing in at least one hit movie previously. The top three movies—
4
Star Wars, E.T. the Extra-Terrestrial and Titanic—had no stars. This shows, some insiders claim, that
“it is the movie itself—not the star—that makes the hit” (Forbes 2002). There are signs that the
doubts regarding the returns on investments are causing existing contractual arrangements with
talent to come under pressure, making the relationship between the involvement of stars and the
success of movies a critical research issue (Eliashberg, Elberse and Leenders forthcoming, Wei
forthcoming). The stakes are high—not just because of the high fees paid to stars, but also because
movies themselves are enormous financial bets.
An extensive academic literature exists on the question whether star creative talent impacts
the financial performance of movies, but the findings to date are mixed. Even when extant research
finds star power to be significantly related to movie performance, it is difficult to draw conclusions
about the direction of causality, because that research typically does not account for the possibility
that studios may employ bigger stars for those movies that are expected to generate higher revenues,
or that the most powerful stars may be able to choose the most promising movie projects. Moreover,
motion pictures are the result of the work of many actors and other workers, and research to date
has largely ignored the effect of individual stars and the interdependencies between stars.
I re-examine the relationship between star participation and movie revenues in a setting that
addresses these limitations in extant research, and build on those findings by considering the
determinants of that relationship. I use an event study that revolves around more than 1,200 casting
announcements that cover a total of about 600 stars and 500 movies. I assess the impact of those
announcements on the behavior of participants of the Hollywood Stock Exchange (HSX,
www.hsx.com), a popular online market simulation in which players (or “traders”) predict box office
revenues. HSX’s reasonably high accuracy makes it a suitable setting for an examination of this kind.
The event study reveals that HSX prices respond significantly to casting announcements,
and thus provides support for the hypothesis that the involvement of stars positively affects revenues.
5
For instance, my estimates suggest that the average star in the sample is “worth” about $3 million in
theatrical revenues. In an extension of the analysis involving the “real” stock market performance of
film studios listed on the NYSE, I examine whether stars also impact the valuation of film companies
that recruit them. I fail to find support for this idea. I use a cross-sectional analysis, grounded in
relevant literature on group dynamics, to assess whether certain characteristics of the star and the
other cast members are determinants of stars’ impact on revenues. I find that a star’s economic and
artistic reputations are positively related to his impact, and show that the roles of a newly recruited
star and the other cast members are linked—the more “A-list” a cast already is, the greater is the
impact of a star with a track record of box office successes or with wide recognition among critics
and peers. This result contributes to the group dynamics literature by providing an example of
increasing returns to recruiting stars in the context of the motion picture industry. The study draws
attention to the multiplicative nature of the production process and the existence of a classic team
problem (Alchian and Demsetz 1972) in recruiting and compensating stars (Caves 2000; 2003).
HYPOTHESES
The Impact of Stars on Revenues
Several researchers have studied the effect of star power on revenues. The findings are mixed: some
studies have not detected a relationship between revenues and talent involvement (Austin 1989;
Litman 1983; Litman and Ahn 1998; De Vany and Walls 1999; Ravid 1999), while others have found
evidence that a movie’s likely cumulative, weekly, or opening-week revenues increase with the rank
of the star talent associated with it (Faulkner and Anderson 1987; Litman and Kohl 1989; Wallace,
Seigerman and Holbrook 1993; Prag and Casavant 1994; Sochay 1994; Sawhney and Eliashberg
1996; Albert 1998; Neelamegham and Chintagunta 1999; Basuroy, Chatterjee and Ravid 2003;
Elberse and Eliashberg 2003; Ainslie, Drèze and Zufryden 2005).
6
The role of stars in the performance of their team or organization is also a general theme in
the academic literature on group dynamics. The prevalent view is that, all else equal, groups with
more talented individual members should outperform groups with less talented members. For
instance, Tziner and Eden (1985), who studied military tank crews, show that group productivity
relates positively to the summed abilities of the group members. Groysberg, Polzer and Anger
Elfenbein (2006), who examined Wall Street equities research analysts, find that groups benefit from
having members who achieved high individual performance. They note that stars’ contributions
could directly increase the team’s performance, but may also indirectly drive success, for example by
enhancing the group’s perceived standing in the eyes of external constituents.
The latter observation fits the context of motion pictures. As Albert (1998) indicated, actors
can be characterized as “stars” for a number of reasons: they may have critically acclaimed acting
skills, possess personality traits that appeal to the movie-going audience, attract a lot of free
publicity, have the ability to secure investment, or simply have been lucky. Accordingly, academic
researchers have measured “star power” in different ways. For example, Sawhney and Eliashberg
(1996) use a dummy that is based on a list of stars possessing “marquee value” published by trade
magazine Variety. Elberse and Eliashberg (2003) and Ainslie et al. (2005) measure star power using
The Hollywood Reporter’s Star Power Survey, in which executives and other insiders were asked to rank
talent. Ravid (1999) classifies stars based on, among other things, whether they have been nominated
for or won an Academy Award, or have participated in a top-grossing film in the previous year.
Ravid’s (1999) measures directly relate to the two kinds of reputations that can generally be
the source of stars’ power: an economic reputation, derived from their box office success, and an
artistic reputation, derived from the recognition of critics or peers (Delmestri, Montanari and Usai
2005). A star’s historical box office record has been found to be an indicator of his future potential
in some studies (e.g., Litman and Kohl 1989, Sochay 1994, Ravid 1999, Lampel and Shamsie 2003)1
,
7
and is a valued source of information for studio executives (Chisholm 2004). A star’s artistic
reputation, which in the motion picture industry is primarily revealed by means of awards or
nominations, is a sign of quality for audiences, executives, the media, and other constituencies (e.g.,
Wallace et al 1993), and thus also a likely predictor of that star’s future box office record.
Both reputations can be viewed as dimensions of “status” as commonly defined in the
literature on group dynamics: “the amount of respect, influence, and prominence stars enjoy in the
eyes of others” (Anderson, John, Keltner, and Kring 2001). Status is an integral element of team
composition (West and Allen 1997) but has received surprisingly little attention as a possible
predictor of team performance (Groysberg et al 2006). I test both dimensions, and hypothesize:
H1: The impact of a star on a film’s box office revenues positively depends on (a) the star’s
economic reputation, reflected by his or her historical box office performance, and (b) the
star’s artistic reputation, reflected by his or her awards or award nominations.
The Role of Other Cast Members
Movies are “complex creative goods” that are the result of teams of creatives working together
(Caves 2000), which makes it worthwhile to consider the role of one star in relation to other cast
members. Managers face important questions in this area, for instance whether a film producer can
successfully economize with a lower-ranked “B-list” lead actress after paying a high fee for a higherranked
“A-list” lead actor, or whether it makes economic sense to further invest in “A-list” stars.
Again, the body of work on group dynamics and group composition offers useful theoretical
considerations. In their study of tank crews, Tziner and Eden (1985) find significant interaction
effects: each member’s ability influenced crew performance differently depending on the ability
levels of the other members. Specifically, they find that a high-ability member achieves more in
8
combination with other uniformly high-ability members than in combination with low-ability
members. Furthermore, uniformly high-ability crews impressively surpassed performance
effectiveness anticipated on the basis of members’ ability. Egerbladh (1976) reported similar findings
in a laboratory setting. Groysberg et al (2006) also note that the benefits to group performance of
assembling talented individuals could extend beyond the simple aggregation of their separate
contributions. However, in their recent study of Wall Street analysts, they find that the marginal
benefit of stars decreased as the proportion of individual stars in a research group increased, and
that the slope of this curvilinear pattern becomes negative when teams reached a high proportion of
star members, leading them to conclude that the outcome of a group of highly-ranked
interdependent stars may be less than the sum of its parts.
What could explain the divergent findings? On the one hand, Groysberg et al (2006) offer a
number of reasons for decreasing returns to recruiting stars. For instance, once a team has an expert
who, to put it simply, “knows the right answers,” more experts may add little value. Also, for a starstudded
group that is already highly visible to stakeholders, adding an additional star may add only a
negligible increment of visibility to the group. Expectations or egos may even impede the willingness
to share information or engage in other behaviors that are necessary for the team as a whole to work
together and succeed (Hambrick 1994). Groysberg et al (2006) find specific evidence of
dysfunctional team processes in environments with too many stars.
On the other hand, there are valid reasons to expect increasing returns: individuals may benefit
from working with talented colleagues (Cummings and Oldham 1997), for example because they are
motivated to maintain informal esteem among highly productive team members (Zuckerman 1967)
or may be stimulated by ideas and suggestions that arise from close contact with talented peers
(Allison and Long 1990). Highly-ranked individuals make their teams more visible: if an individual is
prominent in the eyes of others, and others associate the individual with a particular group, then that
9
group should be more prominent by virtue of their association. The individual star effectively
acquires free advertising for his or her team. This, in turn, may increase the group’s ability to secure
resources and therefore help it succeed, particularly in contexts in which groups compete for
resources (e.g., Podolny 1993, Groysberg et al 2006).
In the context of the motion picture industry, increasing returns may be more likely than
decreasing returns. For instance, Tziner and Eden (1985) largely attribute their finding to the nature
of the group task at hand. Like in their setting of military tank crews, the tasks of individual movie
cast members are complex, require a close synchronization, and are difficult to evaluate in isolation
(also see Porter, Lawler and Hackman 1975, Wageman 2001). The hedonic nature of the product is
important in this respect. Caves (2000) claims that differences in various films' gross rentals can be
explained by the number of well-known actors in each film and a combined measure of the
performance of those actors (also see Baker and Faulkner 1991, Faulkner and Anderson 1987,
Zuckerman 1967). The core idea here is that a project may be no better than the least capable
participant involved. In other words, the effects of individual talents' qualities may be multiplicative:
if a B-list participant is replaced by one from the A-list in an otherwise A-list project, its value rises
by more than if the replacement had occurred in a B-list project (Caves 2000).
One underlying premise may be that two superstars are cast in the hopes that their joint
cachet will pay off, which one studio executive referred to as the “one plus one equals three” model
(Variety 2002). It may also be that each actor exerts effort proportional to the efforts exerted by
others—actress Renée Zellweger has to raise her game if she is playing opposite, say, Tom Cruise
instead of a less skilled actor. The intuition here is in line with Taggar (2002) who, in his work on the
relationship between individual and group creativity, argued that involving others may improve
social facilitation and increase the production pressure coming from other group members
(Hackman and Morris 1975). This is also supported by Fisher and Boynton (2005), who studied
10
“virtuoso teams” comprised of the elite experts in their fields which are specifically convened for
certain projects, including the Broadway musical West Side Story and the 1950s era television hit Your
Show of Shows. They ascribe the latter team’s success partly to its tendency to engage in high-energy
contests that raised their performance levels: “[E]very day, each [writer] tried to top the others for
the “best of the best” title” (p. 116). Consequently, I hypothesize:
H2: The impact of a star on a movie’s box office revenues positively depends on (a) the number
of other stars cast for the movie, (b) the economic reputations of those other stars, and (c)
the artistic reputations of those other stars.
DATA
The Hollywood Stock Exchange
The primary data source is the Hollywood Stock Exchange (HSX, www.hsx.com), an online market
simulation focused on the movie industry. As of January 2005, it had over 500,000 registered users
(“traders”), a frequent trader group of about 80,000 accounts, and approximately 19,500 daily unique
logins. New users receive 2 million 'Hollywood dollars' (denoted as "H$2 million") and can increase
the value of their portfolio by, among other things, strategically trading “MovieStocks” and
“StarBonds”. The trading population is fairly heterogeneous, but the most active traders tend to be
heavy consumers and early adopters of entertainment products, especially films.
HSX is not a real stock market: it revolves around fake money, all traders start with the same
capital when they first join, the pricing of “stocks” is governed by a “market maker” based on
computer algorithms (as opposed to pure supply and demand dynamics), and “stocks” have a finite
horizon in that trading comes to an end when the movie has played in theaters for some time.
However, the simulated market appears reasonably efficient and accurate in its predictions, which
11
makes it a suitable setting for this study. For instance, Pennock, Lawrence, Giles and Nielsen (2001a;
2001b), who comprehensively analyzed HSX's efficiency and forecast accuracy immediately before a
movie’s release, find that arbitrage closure on HSX is quantitatively weaker, but qualitatively similar,
relative to a real-money market. They also show that HSX forecasts perform competitively in direct
comparisons with expert judges. Elberse and Eliashberg (2003), Spann and Skiera (2003), and, for a
larger pre-release period, Elberse and Anand (2005) provide further evidence that HSX traders
collectively produce relatively good forecasts of actual box office returns (also see Hanson 1999,
Gruca 2000, and Wolfers and Zitzewitz 2004).2
(I return to HSX’

hahahah yah the 'acting' is great! how did the oscars overlook these actors?! no I know how delusional you are BeOneOfUs